{"attribution":"Likelier — https://likelier.app","license":"https://creativecommons.org/licenses/by-sa/4.0/","support":"https://buymeacoffee.com/kgluszczyk?via=likelier&utm_content=api-fears","generated_at":"2026-08-17T00:55:03.205Z","count":543,"fears":[{"slug":"phone-at-gas-station","question":"What are the odds of a phone igniting fuel at a gas station?","quick_answer":"The odds of a phone igniting fuel at a gas station are effectively zero — not one confirmed case exists worldwide despite billions of fuelings a year over 25+ years of cell phone use. The warning stickers persist for liability reasons, not because any engineering body considers the risk real.\n","category":"tech","no_reliable_estimate":false,"perceived":{"description":"Warning stickers on gas pumps across the United States and much of the world instruct customers not to use cell phones while fueling. The fear traces to a 1999 email hoax attributed to Shell Oil that claimed cell phone signals could ignite gasoline vapors, and was amplified by sporadic media reports of gas station fires that coincided with phone use but were never causally linked to the phone. The Petroleum Equipment Institute, the American Petroleum Institute, and multiple fire investigation agencies have investigated the claim repeatedly. The result is uniform: zero confirmed cases worldwide. The warning stickers persist because the liability cost of removing them exceeds the cost of keeping them, not because any engineering body considers the risk real.\n","rough_estimate":"Many people believe there is a small but real chance a phone could ignite gas vapors","kind":"intuition"},"native":{"display":"0 in ~17 billion fueling events (no confirmed case worldwide)","numerator":1,"denominator":1000000000,"unit":"per fueling event","population":"Global gas station customers, all years of cell phone use"},"normalized":{"lifetime_us_adult":1e-9,"display":"Effectively zero (no confirmed case exists)","log_value":-9,"assumptions":"Zero confirmed cases of cell-phone-ignited gas station fires exist worldwide despite billions of fueling events per year over 25+ years of widespread cell phone use. Americans make an estimated 290 million gas station visits per week (~15 billion per year). The Petroleum Equipment Institute has been unable to document a single confirmed case. The native rate is set at 1 in 1 billion as a structural floor — the true rate may be literally zero. The normalized lifetime figure uses this structural floor applied over ~4,700 lifetime fueling events (one fill-up per 4.5 days over 59 adult years), yielding an effectively zero probability. The 1e-9 value is a placeholder to satisfy schema requirements; the honest answer is that no evidence of this risk exists.\n","uncertainty":{"low":1e-10,"high":1e-8},"scope":"us_adult_lifetime"},"sources":[{"url":"https://spectrum.ieee.org/cellphones-pose-no-gas-station-hazard","title":"Cellphones Pose No Gas Station Hazard","publisher":"IEEE Spectrum","source_type":"reputable_reference","statistic":"No confirmed case of a cell phone igniting gasoline vapors has ever been documented; laboratory testing has failed to produce ignition","excerpt":"\"No scientific evidence has shown that danger exists, and no confirmed incident has ever occurred anywhere in the world. A literature search found no evidence of fires or explosions at gas stations caused by a cellphone.\"\n","source_date":"2006-09-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260426205620/https://spectrum.ieee.org/cellphones-pose-no-gas-station-hazard","calculation_notes":"IEEE Spectrum is the flagship publication of the Institute of Electrical and Electronics Engineers. The article reviews the physics: a cell phone's maximum radiated power (~0.6 watts for GSM, lower for modern LTE/5G) is insufficient to produce a spark. The minimum ignition energy for gasoline vapor in air is ~0.2 millijoules — achievable by static discharge but not by RF radiation from a phone at any plausible distance. Even a phone battery failure (thermal runaway) would require the battery to rupture and produce an open flame in the presence of gasoline vapor at the right concentration — a scenario that has never been documented at a gas station.\n"},{"url":"https://pei.org/resources/stop-static-campaign/","title":"Stop Static Campaign","publisher":"Petroleum Equipment Institute (PEI)","source_type":"reputable_reference","statistic":"PEI has documented over 150 static-discharge-ignited refueling fires but has been unable to document any incident caused by a cell phone","excerpt":"\"The Petroleum Equipment Institute has not been able to document any incident that was sparked by a cellular telephone. While we are not aware of any scientific evidence that cell phones pose a hazard at the gas pump, we do know that static electricity can cause a flash fire.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421192450/https://pei.org/resources/stop-static-campaign/","calculation_notes":"PEI is the trade association representing fueling equipment manufacturers and has maintained a database of refueling fires since 1992. They have documented over 150 fires attributed to static electricity discharge — the actual hazard at gas stations. Static fires occur when a driver re-enters the vehicle during fueling, accumulates a static charge, and then touches the metal nozzle near gasoline vapor. Approximately 100 static-sparked fires occur per year at US gas stations. The contrast is instructive: 150+ confirmed static fires, zero confirmed cell phone fires.\n"},{"url":"https://mythbusters.fandom.com/wiki/Cell_Phone_Destroys_Gas_Station_(Myth)","title":"Cell Phone Destroys Gas Station (Myth) — MythBusters","publisher":"MythBusters / Discovery Channel","source_type":"reputable_reference","statistic":"MythBusters tested cell phones in gasoline vapor and were unable to produce ignition under any conditions; myth rated 'Busted'","excerpt":"\"The MythBusters placed cell phones in a chamber filled with gasoline vapor at optimal fuel-air mixture ratios and triggered incoming calls. The phones failed to ignite the vapors under any tested conditions. The myth was rated 'Busted.'\"\n","source_date":"2003-09-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251231053559/https://mythbusters.fandom.com/wiki/Cell_Phone_Destroys_Gas_Station_(Myth)","calculation_notes":"MythBusters Season 1 tested the claim under conditions more favorable to ignition than any real gas station scenario — enclosed chamber, optimal fuel-air ratio, multiple phone models. No ignition occurred. While MythBusters is entertainment rather than peer-reviewed science, the result is consistent with the physics (RF energy from phones is orders of magnitude below the minimum ignition energy for gasoline vapor) and with the PEI and IEEE findings. Included as a widely-known cultural reference point that has shaped public awareness of this myth.\n","independence_note":"MythBusters conducted independent empirical testing separate from the IEEE literature review and PEI incident database. All three sources arrive at the same conclusion through different methods.\n"},{"url":"https://www.snopes.com/fact-check/static-electricity-pump-fires/","title":"Static Electricity and Gas Pump Fires","publisher":"Snopes","source_type":"reputable_reference","statistic":"The claim that cell phones cause gas station fires is false; static electricity is the documented cause of refueling fires","excerpt":"\"Although you'll find 'No cell phones' stickers on gas pumps across the land, no one has yet documented a real-world case of a cell phone igniting fumes at a gas station. The real risk is static electricity.\"\n","source_date":"2022-08-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421192508/https://www.snopes.com/fact-check/static-electricity-pump-fires/","calculation_notes":"Snopes traces the myth to a 1999 email chain falsely attributed to Shell Oil that described three supposed cell-phone-ignited fires. Shell denied originating the email. Snopes' fact-check corroborates the PEI and IEEE findings: the documented refueling fire risk comes from static electricity, not cell phones. Included for provenance of the myth origin.\n"}],"comparison_anchors":[{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Asteroid impact death (lifetime)","lifetime_us_adult":7.4e-7}],"short_label":"Phone at gas pump","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The native rate of 1 in 1 billion is a structural placeholder, not a measured probability. The true rate may be literally zero — it is included to satisfy the schema requirement for a non-zero numerator. The warning stickers on gas pumps persist for liability and precautionary reasons, not because any engineering or fire-investigation body has identified a mechanism by which a normally functioning cell phone could ignite gasoline vapor at atmospheric concentrations found during fueling. The actual refueling hazard — static electricity discharge — causes approximately 100 fires per year at US gas stations and is addressed by the simple precaution of touching metal before handling the nozzle.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A gas pump nozzle and a phone silhouette separated by empty space, rendered in muted amber and slate tones, flat vector illustration."},"canonical_url":"https://likelier.app/phone-at-gas-station","api_url":"https://likelier.app/api/fears/phone-at-gas-station.json"},{"slug":"phone-on-plane-interference","question":"What are the odds of a phone crashing an airplane?","quick_answer":"Effectively zero — no aviation accident has ever been attributed to a passenger's electronic device. A substantial minority of travelers still believe a phone could interfere with cockpit instruments enough to cause a crash, but the true probability may be literally zero.\n","category":"tech","no_reliable_estimate":false,"perceived":{"description":"For decades, flight attendants instructed passengers to turn off all electronic devices during takeoff and landing, and the cabin announcement became one of aviation's most familiar rituals. The underlying fear — that a phone's radio signal could interfere with cockpit instruments and cause a crash — was never supported by a documented incident, but the regulatory posture implied a real danger. The FAA's 2013 decision to allow devices in airplane mode throughout all phases of flight effectively conceded the point, yet surveys show that many passengers still believe phones pose a meaningful crash risk. The FCC's separate ban on in-flight cellular calls (which remains in effect) is about protecting ground-based cell networks from airborne interference, not about aircraft safety, but this distinction is lost on most travelers.\n","rough_estimate":"A substantial minority of travelers believe phones could interfere with avionics enough to cause a crash","kind":"intuition"},"native":{"display":"0 confirmed crashes attributable to passenger electronic devices","numerator":1,"denominator":1000000000,"unit":"per flight with passenger electronics present","population":"Global commercial aviation, all years"},"normalized":{"lifetime_us_adult":1e-9,"display":"Effectively zero (no confirmed crash from PEDs)","log_value":-9,"assumptions":"No aviation accident has ever been causally attributed to passenger electronic device interference with avionics. The FAA Advisory Circular AC 91.21-1D and the 2013 PED Aviation Rulemaking Committee both concluded that modern avionics are sufficiently shielded. Globally, ~40 million commercial flights per year carry passengers using electronic devices (many with phones inadvertently left on). Over 25+ years of widespread cell phone use, this represents hundreds of millions of flight-segments with PEDs present and zero confirmed interference-caused accidents. The 1e-9 value is a structural floor to satisfy the schema; the true probability may be literally zero.\n","uncertainty":{"low":1e-10,"high":1e-8},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_91.21-1D.pdf","title":"Advisory Circular 91.21-1D: Use of Portable Electronic Devices Aboard Aircraft","publisher":"Federal Aviation Administration (FAA)","source_type":"govt_report","statistic":"FAA determined that airlines may allow portable electronic devices during all phases of flight, provided testing demonstrates the aircraft can tolerate potential interference","excerpt":"\"This AC provides guidance for use of portable electronic devices (PEDs) aboard aircraft. Operators may determine that certain PEDs may be used during all phases of flight, provided testing and analysis demonstrates the aircraft can tolerate the radio frequency emissions.\"\n","source_date":"2013-10-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260221112322/https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_91.21-1D.pdf","calculation_notes":"The FAA's 2013 advisory circular is the regulatory instrument that relaxed PED restrictions. It followed the PED Aviation Rulemaking Committee's finding that modern commercial aircraft meet RTCA/DO-160 electromagnetic compatibility standards and can tolerate RF emissions from consumer electronics. The AC does not claim that interference is physically impossible — it states that the probability of harmful interference from PEDs in airplane mode is acceptably low given modern avionics shielding. Airlines were required to verify their specific fleet's tolerance through testing before expanding PED permissions.\n"},{"url":"https://en.wikipedia.org/wiki/Mobile_phones_on_aircraft","title":"Mobile phones on aircraft","publisher":"Wikipedia (citing FAA, FCC, EASA, and RTCA sources)","source_type":"encyclopedia","statistic":"No aircraft accident has been attributed to interference from passenger mobile phones; the FCC ban on in-flight calls relates to ground network interference, not aviation safety","excerpt":"\"No one has officially or definitively linked cell phone usage to an airline accident. Researchers in the mid-2000s concluded that cell phones did have the potential to interfere with critical electronics in aircraft, though they couldn't find any instances in which it had caused an accident.\"\n","source_date":"2024-06-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260302194157/https://en.wikipedia.org/wiki/Mobile_phones_on_aircraft","calculation_notes":"Wikipedia's article synthesizes the regulatory history: the original FCC ban (47 CFR 22.925, dating to 1991) prohibits airborne cellular transmission to protect ground networks, not aircraft. The FAA's separate PED policy evolved from a blanket ban to the 2013 relaxation. The article notes that RTCA Special Committee 202 tested PED emissions and concluded that the theoretical interference pathway exists but at power levels well below those that would affect DO-160-compliant avionics. Used as a contextual source; the FAA AC is the authoritative primary.\n","independence_note":"Wikipedia synthesizes FAA, FCC, and RTCA sources rather than presenting independent evidence. Included for the concise summary of the regulatory timeline and the explicit statement about zero attributed accidents.\n"},{"url":"https://www.cnn.com/2013/10/31/travel/faa-portable-electronic-devices/index.html","title":"FAA allowing most electronic device use throughout flights","publisher":"CNN","source_type":"news_article","statistic":"FAA announced in October 2013 that passengers may use most PEDs in airplane mode during all phases of flight, including takeoff and landing","excerpt":"\"The FAA said that airlines can allow passengers to use portable electronic devices such as tablets, laptop computers, e-readers and cell phones in airplane mode throughout the flight — with some circumstantial restrictions.\"\n","source_date":"2013-10-31","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426205717/https://www.cnn.com/2013/10/31/travel/faa-portable-electronic-devices/index.html","calculation_notes":"Contemporary news coverage of the FAA's 2013 policy change. The article notes that a panel the FAA established to study the issue concluded that most commercial airplanes can tolerate radio interference signals from PEDs. Included as a timestamped record of the policy change and public communication. The panel's conclusion — aircraft tolerance of PED emissions — is the key engineering finding that underpins the \"effectively zero\" risk assessment.\n"},{"url":"https://www.rtca.org/do-160/","title":"RTCA DO-160: Environmental Conditions and Test Procedures for Airborne Equipment","publisher":"Radio Technical Commission for Aeronautics (RTCA)","source_type":"reputable_reference","statistic":"DO-160 is the standard for electromagnetic compatibility testing of all commercial avionics, including tolerance to external RF sources","excerpt":"\"DO-160 specifies a series of minimum standard environmental test conditions and applicable test procedures for airborne equipment. Section 20 covers RF susceptibility, and Section 21 covers emission of RF energy.\"\n","source_date":"2010-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260426205913/https://www.rtca.org/do-160/","calculation_notes":"RTCA/DO-160 (current revision G) is the engineering standard that ensures avionics can withstand electromagnetic interference. Section 20 (RF susceptibility) tests avionics against RF field strengths well above what consumer electronics produce. The standard has been updated progressively since 1975 to reflect evolving RF environments, including cellular, Wi-Fi, and Bluetooth frequencies. Modern aircraft certified to DO-160G are designed to operate correctly in the presence of consumer-electronics emissions. This standard is the technical basis for the FAA's 2013 policy change.\n"},{"url":"https://spectrum.ieee.org/unsafe-at-any-airspeed","title":"Unsafe At Any Airspeed?","publisher":"IEEE Spectrum","source_type":"reputable_reference","statistic":"RF monitoring aboard 37 US commercial flights (Sept-Nov 2003) detected unauthorized in-flight cell phone calls at a rate of one to four per flight, plus signals consistent with at least one passenger leaving a phone on during most flights","excerpt":"\"We measured the RF environment on 37 passenger flights in the eastern United States from September 2003 through November 2003... Results from our analysis imply that calls from on board scheduled commercial aircraft in the eastern United States occur at a rate of one to four per flight... We saw other signals that suggest that at least one passenger neglects to turn off his or her cellphone on most flights.\"\n","source_date":"2006-03-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260630013226/https://spectrum.ieee.org/unsafe-at-any-airspeed","calculation_notes":"Bill Strauss (Carnegie Mellon University, engineering and public policy PhD thesis \"Portable Electronic Devices Onboard Commercial Aircraft: Assessing the Risks,\" 2005) led the first documented study of in-flight RF emissions from passenger electronics, hiding a spectrum analyzer in carry-on luggage across 37 flights. The study found unauthorized calls occurring at a rate of one to four per flight and RF signatures consistent with at least one passenger leaving a phone powered on during most flights — the source for this entry's \"phones routinely left on\" and \"mid-2000s RF detection research\" claims. Strauss described the finding as suggestive rather than alarming, and no airline accident has ever been attributed to the emissions his team detected.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Asteroid impact death (lifetime)","lifetime_us_adult":7.4e-7}],"short_label":"Phone on plane","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 1-in-1-billion native rate is a structural placeholder. The true rate of PED-caused aviation accidents is zero in the historical record. The entry addresses crash risk specifically — low-level electromagnetic interference (EMI) events, such as brief static on cockpit audio, have been anecdotally reported by pilots but have never resulted in loss of aircraft control or a safety incident. The FCC ban on in-flight cellular calls (47 CFR 22.925) remains in effect as of 2026, but this regulation protects ground-based cell networks from airborne interference, not aircraft from phones. The FAA relaxation applies to devices in airplane mode; transmitting cellular calls from altitude remains prohibited for network-management reasons.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A phone silhouette and an airplane silhouette side by side with empty space between them, rendered in muted blue and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/phone-on-plane-interference","api_url":"https://likelier.app/api/fears/phone-on-plane-interference.json"},{"slug":"piranha-attack","question":"What are the odds of being killed by piranhas?","quick_answer":"Effectively zero — roughly 1 in 135 million over a lifetime, with no confirmed case of a healthy living person killed by piranhas. The cultural image of a flesh-stripping killer far exceeds any documented risk.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"The piranha occupies a unique position in the taxonomy of feared animals: almost universally recognized as deadly, almost never encountered outside South American river systems, and responsible for a body count that rounds to zero. Hollywood, sensationalist documentaries, and a century of exaggerated travel writing have cemented the image of a fish that can skeletonize a cow in minutes. The actual peer-reviewed literature describes an animal that delivers isolated defensive bites to toes and heels, almost always in the context of brood defense or dry-season crowding.\n","rough_estimate":"most people would guess piranhas are a significant mortal threat — the cultural fear far exceeds any documented risk","kind":"intuition"},"native":{"display":"~0 confirmed deaths per year globally; effectively unmeasurable","numerator":1,"denominator":8000000000,"unit":"per year","population":"global population"},"normalized":{"lifetime_us_adult":7.4e-9,"display":"1 in ~135,000,000 lifetime (global adult)","log_value":-8.13,"assumptions":"No peer-reviewed source documents a confirmed, unambiguous case of a healthy living human being killed by piranhas. The handful of reported fatalities involve individuals who were already dead or incapacitated (drowning, heart failure) before scavenging occurred. Using the user-specified native numerator of 1 death per 8 billion people as an upper-bound placeholder: annual rate = 1 / 8,000,000,000 = 1.25 × 10⁻¹⁰. Compounded over 59 years: 1 − (1 − 1.25e-10)^59 ≈ 7.4 × 10⁻⁹, i.e. roughly 1 in 135 million. Even this is almost certainly an overestimate. Uncertainty band spans from essentially zero (low: 1.0e-10) to a generous upper bound allowing for unreported cases (high: 5.0e-8).\n","uncertainty":{"low":1e-10,"high":5e-8},"scope":"global_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/14719860/","title":"Piranha attacks on humans in southeast Brazil: epidemiology, natural history, and clinical treatment, with description of a bite outbreak","publisher":"Wilderness & Environmental Medicine","source_type":"peer_reviewed","statistic":"Documented piranha-related human fatalities involve individuals who were already dead from other causes (drowning, heart failure); no confirmed case of a healthy person killed by piranhas","excerpt":"\"There are many tales describing ferocious schools of piranha attacking humans, but there are few scientific data supporting such behavior. [...] The characteristic profile of most injuries is a single bite per victim, generally related to the fish defending its brood.\"\n","source_date":"2003-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503092808/https://pubmed.ncbi.nlm.nih.gov/14719860/","calculation_notes":"Haddad and Sazima's epidemiological study is the most cited peer-reviewed work on piranha attacks. It establishes that injuries are minor, single-bite defensive events and that documented fatalities involved prior drowning or cardiac arrest — not predatory attack. This supports treating the annual death rate as effectively zero.\n","independence_note":"Independent clinical epidemiology study from southeast Brazil, methodologically separate from the media analysis below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12534048/","title":"Media information compared to scientific studies regarding piranha attacks in Brazil","publisher":"PMC / Wilderness & Environmental Medicine","source_type":"peer_reviewed","statistic":"82.27% of 711 reported piranha incidents were mild single-bite injuries; media coverage negatively reinforces popular fear","excerpt":"\"Piranhas are carnivorous fish that inhabit rivers in Central and South America, and they are popularly recognized as relentless hunters of continental waters. Their reputation as killers is fueled by folklore and cinematographic works. [...] 82.27% were classified as mild, with single 'punch-out'-shaped injuries.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503092756/https://pmc.ncbi.nlm.nih.gov/articles/PMC12534048/","calculation_notes":"This 2025 media-vs-science comparison reinforces the Haddad & Sazima finding that piranha fatalities are a media artifact. The study documents the systematic gap between sensationalist reporting and clinical evidence, supporting the use of an effectively-zero mortality rate. The paper also traces the popular \"killer fish\" reputation to Theodore Roosevelt's 1913-1914 Amazon expedition and his book \"Through the Brazilian Wilderness,\" which recounted piranhas devouring a wounded animal and popularized the Brazilian expression \"boi-de-piranha\" (\"piranha's cow\") — the cultural origin point referenced in the body prose below.\n","independence_note":"Independent media analysis study comparing journalistic accounts with clinical data, published separately from the 2003 Haddad & Sazima epidemiological study.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.0000065},{"label":"Death by asteroid impact (lifetime, global adult)","lifetime_us_adult":7.4e-7}],"regional_breakdown":[{"region":"Amazon and Pantanal river systems (local residents and bathers)","probability":5e-9,"notes":"Bites occur during dry season and spawning; documented injuries are minor single bites to extremities. No confirmed predatory fatality of a living person."},{"region":"Rest of world","probability":0,"notes":"Piranhas are not native outside South America. Aquarium escapes and isolated introductions pose no documented mortality risk."}],"short_label":"Piranha attack","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The native numerator of 1 is a placeholder upper bound, not an observed annual death count — no confirmed piranha-caused fatality of a healthy living human has been documented in the peer-reviewed literature. Cases reported in news media (e.g., Paraguay 2022) typically involve drowning victims subsequently scavenged by piranhas, or remain unverified. Piranha bites do occur, particularly during dry seasons when fish are crowded into smaller water bodies, but injuries are overwhelmingly minor single-bite events to feet and hands. The cultural fear of piranhas as pack-hunting killers is almost entirely a product of sensationalist media rather than observed animal behavior.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized piranha silhouette in murky river water, flat vector illustration in muted blue-green tones."},"canonical_url":"https://likelier.app/piranha-attack","api_url":"https://likelier.app/api/fears/piranha-attack.json"},{"slug":"banana-spider-eggs","question":"What are the odds of finding dangerous spider eggs in a banana?","quick_answer":"The chance of a dangerous spider bite from a store-bought banana is effectively zero — about 0.00001% over a lifetime, with no recorded deaths worldwide. The tabloid fear puts it nearer 0.1 to 1 percent, and spider eggs inside the fruit itself are biologically impossible.\n","category":"animal","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The \"Brazilian wandering spider in the bananas\" story is a perennial tabloid fixture and social media staple. The fear comes in two variants: deadly spiders hiding in banana bunches at the supermarket, and spider eggs embedded in the tip of the fruit itself, waiting to hatch when you peel it. Many people habitually cut off the end of each banana as a precaution. The Phoneutria genus (Brazilian wandering spider) has a genuine reputation as one of the most venomous spiders on Earth, and a 2015 viral video appearing to show a spider bursting from an overripe banana amplified the fear -- though the video was later confirmed to be CGI.\n","rough_estimate":"~0.1-1% chance of encountering a dangerous spider in store-bought bananas","kind":"intuition"},"native":{"display":"~7 Phoneutria specimens found in 135 banana-cargo spider submissions over 88 years (1926-2014); 0 deaths worldwide from banana-associated bites","numerator":7,"denominator":135,"unit":"proportion of cargo-submitted spider specimens that were actually Phoneutria (the dangerous genus)","population":"spider specimens from bananas and international cargo submitted to North American arachnologists, 1926-2014 (Vetter et al., 2014)"},"normalized":{"lifetime_us_adult":1e-7,"display":"~0.00001% lifetime probability of a dangerous spider bite from a store-bought banana (effectively zero)","log_value":-7,"assumptions":"Americans consume roughly 30 billion individual bananas per year. Over the entire 88-year dataset (1926-2014), Vetter et al. documented 7 Phoneutria specimens in banana cargo reaching North America. One confirmed bite from a store banana has occurred outside the endemic range (UK, 2005; victim survived). Zero deaths have been recorded from banana-associated spider bites worldwide, ever. The per-banana probability of encountering a Phoneutria is well below 1 in a billion. Over a lifetime of ~5,400 bananas (90/year × 60 years), the probability of encountering a genuinely dangerous spider rounds to effectively zero. We assign 1e-7 as the lifetime figure, reflecting \"one confirmed non-fatal bite in recorded history outside endemic regions\" against billions of bananas consumed. The specific claim about spider eggs inside banana tips is biologically impossible: no spider species lays eggs inside fruit, banana flowers are sealed tubes, and consumer banana varieties are parthenocarpic (seedless, closed ovary). As the Burke Museum puts it, \"there is simply no chance for spider eggs to get into a banana.\"\n","uncertainty":{"low":1e-8,"high":0.000001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26309299/","title":"Spiders (Araneae) Found in Bananas and Other International Cargo Submitted to North American Arachnologists for Identification","publisher":"Journal of Medical Entomology (Vetter, Crawford, Buckle)","source_type":"peer_reviewed","statistic":"135 spider specimens from banana/cargo submissions over 88 years; only 7 (5.2%) were Phoneutria; most were harmless Cupiennius or Heteropoda","excerpt":"\"We identified 135 spiders found in international cargo. A key for the most common species is provided. The most frequently submitted spiders were the pantropical huntsman spider, Heteropoda venatoria (L.) (Sparassidae), and the redfaced banana spider [Cupiennius].\" (The specimen span of 1926-2014 is confirmed by PBS reporting on the study; the \"7 (5.2%) Phoneutria\" count is in the paywalled full text.)\n","source_date":"2014-11-01","source_accessed":"2026-04-25","archive_url":"https://web.archive.org/web/20260504053821/https://pubmed.ncbi.nlm.nih.gov/26309299/","calculation_notes":"Vetter et al. is the definitive study on spiders in banana cargo. 135 specimens over 88 years means roughly 1.5 submissions per year to arachnologists across all of North America. Only 5.2% were the feared Phoneutria. The most common species found was Cupiennius chiapanensis (redfaced banana spider), which is frequently misidentified as Phoneutria by non-specialists and even some arachnologists. Heteropoda venatoria (pantropical huntsman) was the second most common -- also harmless. The study concluded that the danger from banana spiders is \"greatly overstated.\"\n"},{"url":"https://www.burkemuseum.org/collections-and-research/biology/arachnology-and-entomology/spider-myths/myth-spider-eggs-bananas","title":"Myth: Spider eggs are commonly found in bananas","publisher":"Burke Museum of Natural History and Culture (Rod Crawford, Curator of Arachnids)","source_type":"reputable_reference","statistic":"No spider lays eggs in banana flowers; banana flowers are narrow tubes and consumer-variety fruit grows from the ovary without fertilization; there is no chance for spider eggs to get into a banana","excerpt":"\"None of these spiders lays its eggs in flowers. Flowers change so fast that they would be poor places for eggs. [...] As shown in the photo (above, center), banana flowers are narrow tubes. In consumer varieties, the fruit grows from the ovary deep inside without fertilization. There is simply no chance for spider eggs to get into a banana.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-25","archive_url":"http://web.archive.org/web/20260210062426/https://www.burkemuseum.org/collections-and-research/biology/arachnology-and-entomology/spider-myths/myth-spider-eggs-bananas","calculation_notes":"Rod Crawford (Burke Museum arachnid curator) maintains the most comprehensive spider myth debunking resource. His analysis covers both the biological impossibility of eggs inside fruit (no spider reproductive anatomy supports this) and the botanical impossibility (parthenocarpic bananas have sealed, unfertilized ovaries). Egg sacs found ON banana peels occasionally occur from hitchhiking spiders during transport, but these are external, visible, and almost always from harmless species.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/10742722/","title":"Phoneutria nigriventer (armed spider) envenomation: clinico-epidemiological study of 422 cases","publisher":"Revista do Instituto de Medicina Tropical de São Paulo","source_type":"peer_reviewed","statistic":"Of 422 Phoneutria bite cases in Brazil: 89.8% mild (local pain only), 8.5% moderate, 0.5% severe (children only), 1 death (3-year-old child)","excerpt":"\"Of 422 patients bitten by Phoneutria spiders between 1984 and 1996, 89.8 percent presented with mild envenomation (local pain only). Moderate cases accounted for 8.5 percent and severe for 0.5 percent. Both severe cases were in children aged under 4 years. One death occurred in a 3-year-old child.\"\n","source_date":"2000-01-01","source_accessed":"2026-04-25","archive_url":"http://web.archive.org/web/20260223001337/https://pubmed.ncbi.nlm.nih.gov/10742722/","calculation_notes":"This is the foundational clinical study of Phoneutria envenomation. The 89.8% mild-only rate demolishes the \"world's most deadly spider\" narrative. In this series of 422 bites, a single death occurred (a 3-year-old child), i.e. deaths are extremely rare even among treated envenomations. Severe envenomation occurs exclusively in young children. An adult bitten by Phoneutria overwhelmingly experiences intense local pain and nothing more. The species most likely to arrive in banana cargo (P. boliviensis) is less toxic than P. nigriventer studied here.\n"}],"comparison_anchors":[{"label":"Dirty can illness (lifetime)","lifetime_us_adult":0.000005},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013},{"label":"Shark attack death (lifetime, US)","lifetime_us_adult":0.000004}],"personal_factor_multipliers":[{"factor":"living in Brazil near banana plantations","multiplier":100000,"notes":"~4,000 Phoneutria bites/year in Brazil, almost all from spiders encountered in homes and gardens near banana-growing regions, not from store-bought bananas"},{"factor":"working in a banana packing facility in Central/South America","multiplier":10000,"notes":"Packers have direct contact with freshly harvested bunches before washing/inspection; this is where actual spider encounters occur"},{"factor":"buying bananas from regulated US/EU supply chain","multiplier":1,"notes":"Multi-step supply chain (washing, cold transport, inspection, ripening rooms) makes spider survival to consumer vanishingly unlikely"},{"factor":"cutting off banana tips before eating","multiplier":1,"notes":"No effect on risk because spider eggs inside banana tips is biologically impossible; the precaution addresses a nonexistent mechanism"}],"short_label":"Banana spider eggs","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The Vetter et al. dataset of 135 specimens represents submissions to arachnologists for identification, not a random sample of all bananas imported. It overrepresents unusual or alarming-looking spiders (people don't submit tiny harmless spiders for ID) and underrepresents total spider encounters. However, even with this selection bias inflating the Phoneutria count, only 7 specimens out of 135 were the dangerous genus. Spiders occasionally found on banana bunches in supermarkets are almost always harmless Cupiennius or huntsman species misidentified by tabloid journalists as \"Brazilian wandering spiders.\" The one confirmed Phoneutria bite from store bananas (UK, 2005) resulted in full recovery. The \"spider eggs in banana tips\" claim has no biological mechanism and has been debunked by arachnologists, entomologists, and botanists independently. Cutting off banana tips is harmless but addresses a nonexistent risk.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-25","reviewed":true,"generated_at":"2026-04-25","image":{"alt":"A peeled banana on a clean surface next to a small magnifying glass, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/banana-spider-eggs","api_url":"https://likelier.app/api/fears/banana-spider-eggs.json"},{"slug":"kid-front-seat-airbag","question":"What are the odds of a child being killed by a front-seat airbag?","quick_answer":"In a modern (post-2007) car, a correctly restrained child's odds of being killed by the front passenger airbag are about 1 in 10,000,000 per trip — near lightning-strike territory. Parents often picture something closer to 1 in 100,000; the real hazard was first-generation airbags, largely engineered away by 2007.\n","category":"transport","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"\"Kids belong in the back seat\" is one of the stickiest pieces of parenting advice in the modern car-safety canon, and the airbag is usually cited as the reason. Parents who internalized the rule in the 1990s or 2000s often picture the front passenger airbag as an active threat to a buckled child on any given trip, rather than as a historical hazard that was largely engineered away by 2007. The perceived per-trip risk of a child being killed by an airbag deployment in a modern car is probably several orders of magnitude above the real one, even though the back-seat rule itself is still correct for independent crash-physics reasons.\n","rough_estimate":"~1 in 100,000 per trip feels plausible to most parents of small kids","kind":"intuition"},"native":{"display":"~1 in 10,000,000 per child-trip in the front passenger seat (modern vehicle, correctly restrained)","numerator":1,"denominator":10000000,"unit":"per child-trip in the front passenger seat","population":"children under 13 in post-2007 US passenger vehicles with advanced airbags, correctly belted"},"normalized":{"lifetime_us_adult":1e-7,"display":"~1 in 10,000,000 per child-trip (modern vehicle, correctly restrained in front seat)","log_value":-7,"assumptions":"NHTSA's Special Crash Investigations attribute roughly 290 US deaths to frontal airbag inflation in low-speed crashes between 1990 and 2008, of which more than 90 percent (~260) were children and infants — the great majority unbelted or in rear-facing child seats in the front passenger position of 1st-generation airbag vehicles. Spread across ~18 years and the subset of child-trips actually taken in the front seat of airbag-equipped cars during that window, the historical per-trip rate for a correctly-restrained child was already below 1 in a million. For a child in a post-2007 advanced-airbag vehicle (dual-stage deployment, weight sensors, suppression for child-seat profiles) the yearly pediatric airbag-inflation death count has collapsed to near zero, implying a per-trip point estimate around 1e-7. The uncertainty band is wide because the numerator in recent years is a handful of cases, not a stable rate.\n","uncertainty":{"low":1e-8,"high":0.000001},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.iihs.org/topics/airbags","title":"Airbags","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"NHTSA estimates >290 US deaths from frontal airbag inflation in low-speed crashes, 1990-2008; >90% were children and infants","excerpt":"\"NHTSA estimates that during 1990-2008, more than 290 deaths were caused by frontal airbag inflation in low-speed crashes. More than 90% of those were children and infants, most of whom were unbelted or in rear-facing child safety seats that placed their heads close to the deploying airbag.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250607032014/https://www.iihs.org/topics/airbags","calculation_notes":"Take ~290 total inflation-attributable deaths × 0.9 ≈ 260 child deaths across 1990-2008 (~18 years). US child-trips in the front passenger seat of airbag-equipped vehicles during that window are on the order of tens of billions; even on a conservative denominator of ~2e9 exposed child-trips the historical per-trip rate for unbelted or improperly-restrained children is roughly 1e-7 and for correctly-restrained kids is an order of magnitude lower. Post-2007 advanced airbags (mandated by the certified-advanced rule) then cut the numerator by roughly another order of magnitude. The 1e-7 headline is the midpoint of that range for a correctly-restrained child in a modern vehicle.\n","independence_note":"IIHS is summarizing NHTSA Special Crash Investigation (SCI) data; it is not an independent measurement. Treat the 290/90% figures as a single NHTSA-sourced estimate with two presentations.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16595421/","title":"Passenger seating position and the risk of passenger death in traffic crashes: a matched cohort study","publisher":"Injury Prevention (BMJ) — Smith & Cummings, 2006","source_type":"peer_reviewed","statistic":"Adjusted risk ratio for restrained children 0-12 years, rear vs front seat, vehicles with front passenger airbag: 0.62 (95% CI 0.48-0.81)","excerpt":"\"For restrained passengers in cars with a front passenger airbag, the aRR was 0.62 (95% CI 0.48 to 0.81) for children 0-12 years.\"\n","source_date":"2006-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174212/https://pubmed.ncbi.nlm.nih.gov/16595421/","calculation_notes":"The 0.62 adjusted risk ratio is not used to compute the native per-trip number; instead it anchors the statement that the back seat is still the better position for children in airbag-equipped cars as a matter of overall crash mortality, not just airbag-inflation mortality. It is the strongest peer-reviewed evidence that the \"kids in back\" rule has independent justification even after 1st-generation airbags were phased out.\n","independence_note":"Smith & Cummings draw on FARS/GES crash data, which NHTSA also uses. They are not fully independent of the IIHS/NHTSA inflation-death figure but answer a different question (overall fatality risk by seat position, not inflation-attributable deaths).\n"},{"url":"https://www.healthychildren.org/English/safety-prevention/on-the-go/Pages/Air-Bag-Safety.aspx","title":"Air Bag Safety for Children","publisher":"American Academy of Pediatrics (HealthyChildren.org)","source_type":"reputable_reference","statistic":"AAP recommends all children under 13 ride in the back seat; rear-facing infants must never ride in front of an active airbag","excerpt":"\"The safest place for all infants and children younger than 13 years to ride is in the back seat.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174251/https://www.healthychildren.org/English/safety-prevention/on-the-go/Pages/Air-Bag-Safety.aspx","calculation_notes":"Used as the authoritative pediatric-policy citation explaining why the back-seat rule persists regardless of the dramatic reduction in airbag-inflation fatalities since 2007. Not used in the per-trip arithmetic.\n","independence_note":"Policy document, not a primary data source — independent of NHTSA/IIHS crash data but also not a quantitative estimate.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811834","title":"Evaluation of the Certified-Advanced Air Bags (DOT HS 811 834)","publisher":"NHTSA, National Center for Statistics and Analysis","source_type":"govt_report","statistic":"Certified-advanced airbags (phased in 2003-2006) showed a 5% reduction in child right-front-passenger frontal fatality rate (12 and younger, per billion vehicle registration years; not statistically significant); SCI logged only one child right-front airbag fatality in otherwise-survivable crashes across CY 2004-2011 for MY 2002-2009 vehicles","excerpt":"\"The fatality rate, in frontal crashes per billion vehicle registration years showed a 4 percent reduction overall, 5 percent reduction for drivers, and 5 percent reduction for child right-front passengers 12 and younger, after vehicles were equipped with certified-advanced air bags. None of these were statistically significant. ... From CY 2004 to 2011 for MY 2002 to 2009, SCI has only one record of a child up to 12 years old sitting in the right-front seat with fatal injuries from air bags in otherwise survivable, low-to-moderate speed crashes (delta-V < 40.2 km/h [25 mph]).\"\n","source_date":"2013-09-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20250208042246/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811834","calculation_notes":"Primary NHTSA government evaluation (Greenwell 2013) of certified-advanced airbag systems. Its Special Crash Investigations tally — only one child right-front airbag fatality in otherwise-survivable crashes across CY 2004-2011 for MY 2002-2009 vehicles — is the direct evidence that modern-vehicle pediatric front-seat airbag deaths have collapsed to near zero, which anchors the entry's 1e-7 per-trip point estimate. This report does not restate the 290-death / 90%-children historical tally; that figure comes from the IIHS source (#1) summarizing NHTSA SCI data, not from this evaluation.\n","independence_note":"Government publication that measures the certified-advanced vs sled-certified comparison and the modern SCI child-fatality count directly. It draws on the same FARS/SCI system NHTSA/IIHS use, so it is not statistically independent of the historical inflation-death figure, but it supplies the authoritative modern near-zero count first-hand.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (per flight)","lifetime_us_adult":7.3e-8},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Modern vehicle (post-2007), child in back seat","probability":0,"notes":"Not exposed to the front passenger airbag at all; this is the default configuration recommended by AAP and NHTSA."},{"region":"Modern vehicle (post-2007), correctly-restrained child in front seat","probability":1e-7,"notes":"Advanced airbags with occupant sensors and dual-stage deployment; pediatric inflation fatalities in recent years are near zero."},{"region":"1990s vehicle, 1st-generation airbag, rear-facing infant in front seat","probability":0.001,"notes":"The canonical 1st-gen airbag hazard — the scenario that drove the back-seat rule into parenting culture."},{"region":"1990s vehicle, 1st-generation airbag, small child unbelted or out of position","probability":0.01,"notes":"Most of the ~260 historical child inflation deaths fell in this or the previous category."}],"personal_factor_multipliers":[{"factor":"rear-facing infant in front seat of any airbag-equipped vehicle","multiplier":1000,"notes":"Historical 1st-generation airbag scenario; still strictly contraindicated by NHTSA, AAP, and every vehicle owner's manual ever printed."},{"factor":"advanced airbag (post-2007) + correctly belted child","multiplier":0.1,"notes":"Dual-stage inflation, weight sensors, and suppression for detected child-seat profiles."},{"factor":"front passenger airbag manually switched off","multiplier":0.1,"notes":"Only relevant in vehicles that still ship with a manual on/off switch, which is rare outside pickups and two-seat vehicles. Note the mirror hazard: switching the passenger airbag off for a rear-facing child seat and then forgetting to switch it back on leaves a later adult or older teen front passenger unprotected in a frontal crash. That failure mode shifts risk onto a different occupant via a different mechanism (crash force, not inflation), so it does not raise this entry's child airbag-death number — see caveats.\n"}],"short_label":"Kid + front airbag","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is primarily a historical risk entry. The \"kids in the back seat\" rule originated from 1st-generation airbag deaths between roughly 1990 and 2000, but persists in parenting culture long after the specific mechanism (aggressive, single-stage, always-on inflation) was regulated out of new vehicles. In a modern vehicle the airbag-specific per-trip fatality risk to a correctly-restrained child is closer to lightning-strike territory than to car-crash territory. The back-seat rule is still the correct recommendation, but mostly for independent reasons — crash forces are lower in the rear, restraint geometry is better for small bodies, and adults up front aren't distracted turning around. Treat the airbag framing as the historical trigger for the rule, not its main modern justification. One flip-side worth naming: in the minority of vehicles that still ship with a manual passenger-airbag on/off switch, disabling it for a rear-facing child seat and then forgetting to re-enable it leaves the next adult or older-teen front passenger with no airbag in a frontal crash. That is a real, under-discussed failure mode, but it shifts the risk onto a different occupant through a different mechanism (crash force, not airbag inflation), so it sits outside this entry's headline number rather than inside it.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty child car seat on a pale background, flat vector illustration in muted tones, no vehicle, no airbag."},"canonical_url":"https://likelier.app/kid-front-seat-airbag","api_url":"https://likelier.app/api/fears/kid-front-seat-airbag.json"},{"slug":"shark-attack","question":"What are the odds of being killed by a shark?","quick_answer":"Being killed by a shark is astronomically rare — about 1 US death a year, roughly 1 in 5.7 million over a lifetime. Even being bitten is rare: around 61 unprovoked bites worldwide a year, and about 87% of them are non-fatal (ISAF). The shark \"attack\" most people picture is almost always a survivable bite, not a death.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Shark attacks are a cultural archetype for catastrophic risk — a single vivid film, reinforced by occasional summer headlines, keeps \"eaten by a shark\" near the top of many people's fear lists despite being one of the rarest causes of death on Earth.\n","rough_estimate":"34.6% of US adults report being afraid or very afraid of sharks (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~1 unprovoked shark attack fatality per year, United States (of ~8 worldwide)","numerator":1,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":1.76e-7,"display":"1 in ~5,700,000 lifetime (US adult)","log_value":-6.75,"assumptions":"The International Shark Attack File reports a 5-year average (2020-2024) of 8 unprovoked fatal shark attacks per year worldwide, with the United States accounting for approximately one of those fatalities in a typical year. The native rate uses the US-specific subset (~1 death/year among ~335 million US residents) since the normalized figure is a US adult lifetime risk. Annual rate: 1/335,000,000 ≈ 2.99 × 10⁻⁹. Compounded over 59 years of remaining adult life: 1 − (1 − 2.99 × 10⁻⁹)⁵⁹ ≈ 1.76 × 10⁻⁷, i.e. roughly 1 in 5,700,000. The uncertainty band reflects years with zero US deaths (low) and occasional years with 2-3 (high).\n","uncertainty":{"low":9e-8,"high":5e-7},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.floridamuseum.ufl.edu/shark-attacks/yearly-worldwide-summary/","title":"International Shark Attack File — Yearly Worldwide Summary","publisher":"Florida Museum of Natural History, University of Florida","source_type":"reputable_reference","statistic":"5-year average (2020-2024) of 8 unprovoked fatal shark attacks per year worldwide; 2025 saw 9 fatalities from 65 unprovoked bites","excerpt":"\"This number is also in line with the most recent five-year annual global average of eight unprovoked fatalities per year. [...] The 2025 worldwide total of 65 confirmed unprovoked cases is in line with the most recent five-year (2020-2024) average of 61 incidents annually.\"\n","source_date":"2026-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260411093742/https://www.floridamuseum.ufl.edu/shark-attacks/yearly-worldwide-summary/","calculation_notes":"ISAF reports a 5-year average (2020-2024) of ~8 unprovoked fatal attacks per year worldwide, with the US accounting for ~1 in a typical year. The US-specific annual rate (~1/335M) is compounded over 59 years for normalized lifetime risk: 1 − (1 − 2.99e-9)^59 ≈ 1.76e-7.\n","independence_note":"ISAF maintains its own international shark-attack case register, built from voluntary reports, media scanning, and expert verification. Entirely independent of CDC's ICD-10 death-certificate pipeline — the two describe overlapping events through different collection mechanisms.\n"},{"url":"https://wisqars.cdc.gov/","title":"WISQARS — Web-based Injury Statistics Query and Reporting System","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"WISQARS query tool for fatal and nonfatal injury data; ICD-10 coding does not isolate shark attacks from broader marine-animal contact (W56)","excerpt":"\"WISQARS is an interactive, online collection of analysis tools for fatal, nonfatal, and cost of injury data.\" [Note: WISQARS is a query tool, not a report. ICD-10 code W56 covers all marine-animal contact and does not distinguish shark attacks from other marine encounters. Individual query results are not quotable as static text.]\n","source_date":"2023-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260410040543/https://wisqars.cdc.gov/","calculation_notes":"WISQARS data confirms that ICD-10 W56 (contact with marine animal) deaths in the US are very rare. The coding does not isolate sharks from other marine animals, but the low total is consistent with the ISAF figure of ~1 US shark fatality per year.\n","independence_note":"ISAF and CDC draw from different case collections (voluntary international reporting vs ICD-coded death certificates), so this counts as meaningfully independent corroboration.\n"}],"comparison_anchors":[{"label":"Death by falling coconut (lifetime, global)","lifetime_us_adult":2e-10},{"label":"Death by bee/wasp sting (lifetime, US)","lifetime_us_adult":0.00001267},{"label":"Death by lightning (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"frequent surfer (100+ sessions/year)","multiplier":50,"notes":"ISAF reports surfers and board-sports participants accounted for 32% of unprovoked incidents (2025); frequent ocean users have dramatically higher exposure"},{"factor":"Florida Atlantic coast resident, regular ocean swimmer","multiplier":5,"notes":"Florida leads US in unprovoked attacks; regular swimmers have higher encounter rates"},{"factor":"landlocked state resident","multiplier":0.01,"notes":"near-zero ocean exposure eliminates encounter risk"},{"factor":"spearfishing with bleeding catch in open water","multiplier":15,"notes":"ISAF data identify blood and fish scent from spearfishing as a strong olfactory attractant for sharks; spearfishers are over-represented in provoked and unprovoked attack records relative to their share of ocean users"},{"factor":"ocean entry at dawn or dusk in known shark habitat","multiplier":5,"notes":"ISAF guidance and peer-reviewed shark-behavior literature document elevated shark feeding activity at crepuscular hours (dawn and dusk), increasing encounter probability for surfers and swimmers who enter the water at those times"}],"short_label":"Shark attack","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a population-level figure over the entire US adult population. Your actual risk is essentially zero unless you spend meaningful time in waters known for large apex shark species; surfers in certain regions face higher per-hour exposure but still tiny absolute lifetime numbers.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":5,"d8":3,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"injury-reframe-8d-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized shark fin silhouette against a calm horizontal water line, flat vector illustration."},"canonical_url":"https://likelier.app/shark-attack","api_url":"https://likelier.app/api/fears/shark-attack.json"},{"slug":"bear-attack","question":"What are the odds of being killed by a bear?","quick_answer":"The lifetime risk of being killed by a bear in the US is about 1 in 3,790,000, from roughly 1 to 2 fatal attacks a year. Vivid news coverage keeps the felt risk far higher — many people guess something like 1 in a few thousand.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"The bear is one of the classic archetypes of wild-animal danger — large, fast, occasionally predatory, and culturally locked in as the thing that might eat you in the woods. Fatal bear attacks are vivid enough to make national news whenever they happen, which keeps the felt risk running far above the recorded rate. We haven’t yet found a rigorous recent survey that isolates “fear of being killed by a bear” from the broader category of fear of large predators or the outdoors, so the perceived side of this page is marked intuition rather than survey.\n","rough_estimate":"most people guess something like 1 in a few thousand lifetime","kind":"intuition"},"native":{"display":"~1.5 fatal bear attacks per year, United States","numerator":1.5,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":2.64e-7,"display":"1 in ~3,790,000 lifetime (US adult)","log_value":-6.58,"assumptions":"Uses ~1.5 fatal bear attacks per year in the United States as the long-run central estimate, drawn from Gunther’s peer-reviewed Yellowstone fatality review, Bombieri et al.’s worldwide brown-bear attack compilation, and NPS-maintained park incident records, all of which converge on a US count between roughly 1 and 2 fatal incidents per year averaged over recent decades. Divided by a US population of ~335 million and compounded over 59 years of remaining adult life: 1 &minus; (1 &minus; 1.5/335000000)^59 ≈ 2.64 × 10^-7, i.e. ~1 in 3,790,000 lifetime. Black-bear attacks outnumber grizzly attacks in absolute US terms because black bears are roughly two orders of magnitude more numerous, not because they are more dangerous per encounter.\n","uncertainty":{"low":1.8e-7,"high":4.5e-7},"scope":"us_adult_lifetime"},"sources":[{"url":"https://digitalcommons.usu.edu/hwi/vol16/iss3/8/","title":"Bear-Caused Human Fatalities in Yellowstone National Park: Characteristics and Trends","publisher":"Human-Wildlife Interactions (Utah State University) / Gunther KA","source_type":"peer_reviewed","statistic":"8 bear-caused human fatalities in Yellowstone over 1872-2018; per-capita risk ~1 in 26.2 million park visits; 7 of 8 by grizzlies","excerpt":"\"The per capita risk of being killed by a grizzly bear was 1 fatality for every 26.2 million park visits... Seven of the 8 fatalities were caused by grizzly bears... Most fatal bear attacks involved men (75%) and small party sizes (88%)... Although the frequency of fatal bear attacks appears to have increased in recent years, the per capita risk of fatal bear attacks has declined.\"\n","source_date":"2022-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163612/https://digitalcommons.usu.edu/hwi/vol16/iss3/8/","calculation_notes":"Gunther’s Yellowstone review gives the strongest US-specific per-visit figure we have for fatal bear encounters and anchors the upper tail of the distribution (backcountry grizzly country). The 1-in-26.2M figure applies to park visits, not US adult lifetimes; we use it as a sanity check on the national central estimate rather than as the primary number. Eight fatalities over 146 years in a single park is consistent with a nationwide count near 1-2 per year.\n","independence_note":"Gunther draws on NPS park incident records; methodologically independent of the Bombieri et al. worldwide compilation below, which is built from media surveillance and national wildlife-agency reports.\n"},{"url":"https://www.nature.com/articles/s41598-019-44341-w","title":"Brown bear attacks on humans: a worldwide perspective","publisher":"Scientific Reports (Nature) / Bombieri, Naves, Penteriani et al.","source_type":"peer_reviewed","statistic":"664 brown bear attacks worldwide over 2000-2015, of which 183 in North America; ~14% of attacks globally resulted in human fatality","excerpt":"\"We investigated 664 brown bear attacks on humans between 2000 and 2015 across most of the range of the species... North America 183 attacks, Europe 291, East Asia 190... Half of the people were engaged in leisure activities and the main scenario was an encounter with a female with cubs... attacks have increased significantly over time and were more frequent at high bear and low human population densities.\"\n","source_date":"2019-06-11","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260415000000/https://www.nature.com/articles/s41598-019-44341-w","calculation_notes":"183 brown bear attacks in North America over 16 years is ~11 per year (non-fatal and fatal combined). Applying the ~14% global fatality rate and then halving to split North America between US and Canada gives roughly 0.8 brown/grizzly bear fatalities per year in the US specifically. Adding ~0.5-1 black bear fatalities per year (from NPS and state wildlife records; black bears are much more numerous but attack far less often per encounter) lands the central estimate near 1.5 fatal bear attacks per year in the US, consistent with Gunther.\n","independence_note":"Independent of Gunther and NPS incident records: Bombieri et al. compiled attacks from media surveillance and wildlife-agency reporting across multiple countries, not from park incident logs.\n"},{"url":"https://www.sciencedaily.com/releases/2008/03/080325171221.htm","title":"Efficacy of Bear Deterrent Spray in Alaska","publisher":"ScienceDaily (summarizing Smith et al. 2008, Journal of Wildlife Management)","source_type":"news_article","statistic":"83 bear spray incidents in Alaska 1985-2006; red pepper spray stopped undesirable bear behavior in close-range encounters; 98% of people carrying spray were uninjured","excerpt":"\"Of all persons carrying sprays, 98% were uninjured by bears in close-range encounters. The study by Tom Smith and colleagues, published in the Journal of Wildlife Management, reviewed 83 bear spray incidents in Alaska from 1985 to 2006 and found that bear spray was effective at stopping undesirable bear behavior across species.\"\n","source_date":"2008-03-25","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260216085616/http://www.sciencedaily.com/releases/2008/03/080325171221.htm","calculation_notes":"The species-specific effectiveness figures (92% brown bear, 90% black bear, 100% polar bear) come from the full Smith et al. 2008 paper in the Journal of Wildlife Management, not from the ScienceDaily summary. Used as the basis for the \"carries bear spray\" personal-factor multiplier. Smith & Herrero document that when bear spray was deployed in aggressive encounters, it stopped the bear’s undesirable behavior more than 90% of the time across species, and carriers were uninjured 98% of the time. A subsequent Smith paper on firearms in bear encounters found firearms perform notably worse than spray on the same outcome metric. The 10× risk reduction we use in the personal factors is a conservative reading of both.\n","independence_note":"Independent dataset from the fatality reviews above: field-incident reports on deterrent use, not mortality records.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by venomous snake bite (lifetime, US adult)","lifetime_us_adult":0.00000113},{"label":"Death by dog bite or strike (lifetime, US adult)","lifetime_us_adult":0.00000704},{"label":"Death by bee/wasp/hornet sting (lifetime, US adult)","lifetime_us_adult":0.00001267},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Yellowstone / Glacier / Denali park visitors","probability":0.000001,"notes":"concentrated grizzly habitat raises per-visit encounter probability roughly 100x over US-adult baseline"},{"region":"Alaska resident (year-round bear range)","probability":0.000002,"notes":"higher year-round bear density and more outdoor activity — Alaska has the highest per-capita bear-attack rate in the US"},{"region":"US resident outside bear range (northeast urban, southwest)","probability":5e-8,"notes":"near-zero encounter probability; deaths almost always involve captive or escaped animals"}],"personal_factor_multipliers":[{"factor":"backcountry hiker in grizzly range","multiplier":1000,"notes":"Gunther’s Yellowstone data shows the per-capita risk on multi-day backcountry trips is orders of magnitude higher than for frontcountry visitors, and the overwhelming majority of fatal US bear attacks occur in grizzly-range backcountry.\n"},{"factor":"carries bear spray","multiplier":0.1,"notes":"Smith, Herrero et al. (2008) found bear spray stopped aggressive bear behavior in &gt;90% of encounters across species and that 98% of carriers were uninjured. The 0.1 multiplier is a conservative reading of that effect size.\n"},{"factor":"food stored improperly at campsite","multiplier":50,"notes":"Gunther attributes most Yellowstone fatalities to bears conditioned to human food or garbage; improper food storage in bear country is one of the strongest single predictors of predatory encounters.\n"},{"factor":"urban/suburban resident, never in bear country","multiplier":0.01,"notes":"Essentially zero absent travel to bear habitat; a schema-safe floor rather than literal zero because the multiplier field requires a positive number.\n"}],"short_label":"Bear attack","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a population-level average over all US adults. Actual risk is extraordinarily non-uniform: almost all fatal bear attacks occur in a narrow set of contexts — backcountry hiking and camping in grizzly range (Alaska, the Northern Rockies, the Greater Yellowstone Ecosystem), hunting in black bear habitat, and a small number of predatory black bear incidents in rural areas. An adult who spends no time in bear country faces effectively zero risk. A frequent Alaskan backcountry hiker faces a risk several orders of magnitude above the headline number. The number also does not travel outside North America: brown-bear-caused fatalities are more common in parts of Romania, Russia, and Turkey, per Bombieri et al., and Asiatic black bears and sloth bears account for a separate category of incidents outside our scope.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized bear paw print on a muted forest-floor background, flat vector illustration."},"canonical_url":"https://likelier.app/bear-attack","api_url":"https://likelier.app/api/fears/bear-attack.json"},{"slug":"wild-berry-poisoning","question":"What are the odds of dying from eating a misidentified toxic wild berry?","quick_answer":"Dying from a misidentified toxic wild berry runs about 1 in 2.2 million over a US adult lifetime, roughly ten times rarer than a fatal lightning strike. Many people treat a single unknown berry as potentially lethal, but recorded berry deaths are near-nonexistent.\n","category":"food","tags":["food","kids"],"no_reliable_estimate":false,"perceived":{"description":"The image of the toxic berry — vivid, beautiful, and lethal — is a fixture of parental warnings, wilderness survival guides, and folk memory. Parents instinctively grab small children away from any unfamiliar berry on a walk, and foragers recount tales of deadly nightshade and yew with casual fluency. The fear is not irrational on its face: some berries genuinely are toxic, and a few species can cause serious harm in quantity. But the perceived risk for an adult who accidentally eats a wild berry vastly exceeds what the data support.\n","rough_estimate":"Many adults treat even a single unknown berry as potentially fatal","kind":"intuition"},"native":{"display":"~2 deaths per year from all plant ingestions (US, all ages)","numerator":2,"denominator":260000000,"unit":"per year","population":"US adults aged 18+, all wild plant/berry ingestion deaths"},"normalized":{"lifetime_us_adult":4.5e-7,"display":"~1 in 2,200,000 lifetime (US adult)","log_value":-6.35,"assumptions":"Krenzelok and Mrvos (2011) reviewed AAPCC annual reports 1983-2009 and found only 45 plant ingestion fatalities across the entire 26-year period — approximately 1.7 per year from all plant species combined. NPDS data for 2010 and 2012 each record 2 plant deaths, consistent with this average. We use 2 deaths per year as a conservative upper bound for all wild plant ingestion deaths in the US, noting this figure covers all plant species: Datura and Cicuta (water hemlock) alone account for 35.5% of historical plant fatalities, and most fatal cases involved intentional ingestion or root ingestion rather than accidental berry consumption by adults. Krenzelok et al. (1998) analyzed 11,237 unidentified berry exposure cases over 10 years and found zero fatalities; 99.6% of outcomes were \"no effect\" or \"minor.\" Berry-specific fatal risk is almost certainly lower than the all-plant figure, but zero reported deaths in the berry dataset makes a separate berry estimate unreliable — the all-plant upper bound of 2/year is used as the headline rate. Against a US adult population of ~260 million, this gives an annual rate of 7.7e-9. Over 59 years of remaining adult life: 1 - (1 - 7.7e-9)^59 ≈ 4.5e-7, or about 1 in 2.2 million. Normalized to 0.00000045. The wide uncertainty range reflects three sources of variance: (1) AAPCC data captures only reported cases — actual plant deaths may be modestly higher (though underreporting of fatal cases is much smaller than for nonfatal cases); (2) the all-plant figure includes Datura/Cicuta fatalities not driven by berry misidentification; (3) accidental adult berry ingestion is a subset of all plant deaths, so the true berry-specific rate may be substantially below 1e-7.\n","uncertainty":{"low":1e-7,"high":0.000003},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/21495882/","title":"Friends and foes in the plant world: A profile of plant ingestions and fatalities","publisher":"Clinical Toxicology (Krenzelok EP, Mrvos R)","source_type":"peer_reviewed","statistic":"45 plant ingestion fatalities recorded 1983-2009 (~1.7/year); 668,111 plant ingestion exposures in 2000-2009; Datura and Cicuta responsible for 35.5% of fatal outcomes","excerpt":"\"Only 45 fatalities were recorded between 1983 and 2009. Datura and Cicuta species were responsible for 35.5% of the fatal outcomes. [...] During the decade of 2000-2009, 668 111 plant ingestion exposures were reported. [...] Children ≤5 years of age accounted for 81.2% of plant ingestion exposures. [...] Morbidity was related directly to the reason for the exposure with the most severe outcomes occurring in those who ingested plants intentionally for self-harm or substance abuse.\"\n","source_date":"2011-03-01","source_accessed":"2026-05-01","archive_url":"https://web.archive.org/web/20260525162904/https://pubmed.ncbi.nlm.nih.gov/21495882/","calculation_notes":"45 fatalities over 26 years (1983-2009) = 1.73 deaths/year average from all US plant ingestions. This is the denominator for our central estimate. Combined with 2010 and 2012 NPDS data each showing 2 plant deaths, we use 2/year as the central estimate. Against US adult population (~260 million): 2/260e6 = 7.69e-9 annual rate. Lifetime: 1 - (1 - 7.69e-9)^59 ≈ 4.54e-7. The 35.5% Datura/Cicuta share confirms berry-specific deaths are a small fraction of even this tiny total.\n","independence_note":"This paper draws on AAPCC TESS and NPDS annual report data, same upstream as the Krenzelok 1998 berry study below. Treat as the same institutional data pipeline, used here for fatality counts vs the berry-specific outcome profile below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/9554066/","title":"Those pesky berries... are they a source of concern?","publisher":"Journal of Toxicology and Clinical Toxicology (Krenzelok EP, Jacobsen TD, Aronis J)","source_type":"peer_reviewed","statistic":"11,237 unidentified berry exposures over 10 years; zero fatalities; 99.6% of outcomes were no effect or minor","excerpt":"\"Unidentified berry exposures included 11,237 incidents, making it the 11th most common plant-related exposure. Children < 6 y-of-age accounted for 88.5% of the exposures, and 88.5% occurred during June-October. There were no fatalities, and morbidity included 1 major outcome in an infant and 26 exposures with moderate outcomes. In exposures with a known outcome, no effects (86.0%) and minor effects (13.6%) accounted for 99.6% of exposures. [...] Exposures to unidentified berries represent common inquiries to poison information centers. They are associated with low morbidity and no mortality.\"\n","source_date":"1998-01-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20250202004301/https://pubmed.ncbi.nlm.nih.gov/9554066/","calculation_notes":"10-year AAPCC dataset: 11,237 berry exposure incidents, zero deaths. This directly constrains the berry-specific fatality rate: if there were even 1 death per 11,237 exposures, we would expect ~1 death/year given ~1,000 annual berry calls (11,237/10 years). The zero-fatality result over a decade of exposures supports treating berry misidentification as a subset with a lower rate than the all-plant 1.7/year figure. The one major outcome occurred in an infant, not an adult forager.\n","independence_note":"Uses the same AAPCC TESS database as the Krenzelok 2011 paper above, covering an overlapping but distinct query (berry-specific vs all-plant fatalities). The berry dataset predates the 2011 all-plant analysis by a decade; both share the same institutional data pipeline.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000478},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death from carbon monoxide poisoning (lifetime, US adult)","lifetime_us_adult":0.0000714}],"personal_factor_multipliers":[{"factor":"Child under 10","multiplier":3,"notes":"Children under 6 account for 81% of all plant ingestion calls and 88% of berry-specific calls to US poison control centers (AAPCC NPDS), making their per-capita poisoning call rate roughly 3× the adult rate. Most calls result in no or minor effects, but children's smaller body weight means a given alkaloid dose is proportionally larger. Source: Krenzelok, Jacobsen & Aronis, J Toxicol Clin Toxicol 1998; Krenzelok & Mrvos, Clin Toxicol 2011."},{"factor":"Inexperienced forager (no botanical training)","multiplier":2.5,"notes":"The majority of serious wild plant poisoning cases in the AAPCC record involve misidentification — e.g. elderberry (Sambucus) confused with pokeweed (Phytolacca), or edible nightshade relatives confused with Atropa belladonna. Experienced foragers with botanical literacy represent a small fraction of cases. No formal multiplier study exists for berry-specific risk; the figure is conservatively estimated from the NPDS case-profile breakdown showing severe outcomes concentrated in unintentional ingestions by non-botanists. Source: Krenzelok & Mrvos (2011); AAPCC NPDS Annual Reports."},{"factor":"Endemic toxic species in local geography (yew, Atropa belladonna, water hemlock range)","multiplier":2,"notes":"Risk differs materially by the local flora. European yew (Taxus baccata) is common in UK, central European, and Pacific Northwest gardens and hedgerows; Atropa belladonna is naturalized in parts of the eastern US and Europe; water hemlock (Cicuta) grows throughout North America. Areas where these species are abundant have higher absolute per-encounter risk than regions where only relatively low-toxicity species (Solanum nigrum, holly) predominate. Source: Krenzelok & Mrvos (2011) — Datura and Cicuta account for 35.5% of US plant fatalities despite limited foraging interest."},{"factor":"Large quantity ingested (systematic harvest vs single berry)","multiplier":5,"notes":"The zero-fatality outcome in 11,237 berry exposures tracked by Krenzelok et al. (1998) overwhelmingly involved single-berry curiosity ingestions. Most highly toxic berries require a significant dose relative to body weight for lethal toxicity. Systematic harvesting and eating (confusing nightshade for elderberry, or pokeweed for blueberry) represents the scenario where fatal outcomes become plausible. Source: Krenzelok, Jacobsen & Aronis (1998)."}],"short_label":"Wild berry poisoning","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The 2/year all-plant figure is an NPDS-reported upper bound; actual plant poisoning deaths may be modestly higher due to underreporting, though fatal cases are captured more reliably than nonfatal ones. The figure covers all accidental plant ingestion deaths in the US — not just berry misidentification by foragers. Intentional Datura ingestion (drug-seeking) and Cicuta root ingestion (misidentified as edible root vegetables) together account for roughly a third of historical plant fatalities. Accidental adult berry ingestion by a casual forager or hiker is a much smaller fraction of that already tiny count. Children under 6 bear most of the exposure burden (81% of plant calls, 88% of berry calls) but the one major berry outcome in the 10-year dataset involved an infant, not a child who independently foraged. Risk is not zero for any specific highly toxic berry species (yew, water hemlock, Atropa belladonna) eaten in sufficient quantity — but the population-level lifetime rate reflects how rarely this scenario materializes.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A single cluster of small red berries on a thin branch, flat vector illustration on a pale background."},"canonical_url":"https://likelier.app/wild-berry-poisoning","api_url":"https://likelier.app/api/fears/wild-berry-poisoning.json"},{"slug":"japanese-encephalitis-travel","question":"What are the odds of contracting Japanese encephalitis as a traveler to Asia?","quick_answer":"For a short-term urban traveler the odds are about 1 in 2 million per trip — under 1 case per million travelers overall. Travel-clinic intuition treats it as a 1-in-1,000 to 1-in-10,000 trip risk, roughly a thousandfold too high; the per-trip death risk is rarer still, on the order of 1 in 6-7 million.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Japanese encephalitis is a staple of pre-travel clinic anxiety: a mosquito-borne brain infection, endemic to huge stretches of Asia, with a case fatality rate in the 20–30% range and serious neurological sequelae in a similar share of survivors. The framing is technically accurate and quietly terrifying, and the standard traveler response is to pay several hundred dollars for a two-dose vaccine before a two-week trip to Bangkok. We haven’t found a rigorous recent survey that isolates \"fear of Japanese encephalitis\" from general travel-disease worry, so the perceived side of this page is marked as editorial intuition rather than polled data. The working prior we observe in travel clinics is \"this is a real risk I need to protect against\" — which, for the typical short-term urban traveler, is roughly three orders of magnitude too high.\n","rough_estimate":"most travelers leaving a travel clinic treat this as a 1-in-1,000-to-1-in-10,000 trip risk","kind":"intuition"},"native":{"display":"~1 case per 1,000,000 travelers to endemic Asia","numerator":1,"denominator":1000000,"unit":"per trip","population":"short-term travelers from non-endemic countries to JE-endemic Asia"},"normalized":{"lifetime_us_adult":5e-7,"display":"~1 in 2,000,000 per traveler-trip (short-term urban)","log_value":-6.3,"assumptions":"The CDC Yellow Book and ACIP MMWR both put overall JE incidence among travelers from non-endemic countries at &lt;1 case per million travelers to JE-endemic Asia. Hills, Griggs & Fischer (2010), reviewing every published traveler case 1973–2008, arrived at the same estimate from a completely separate case pool. The point estimate here (5 × 10&#8315;&#8311;) represents a short-term (&lt;1 month) traveler whose visit is restricted to major urban areas — the group the CDC describes as at \"minimal risk\" — which sits a factor of roughly 2 below the headline &lt;1-in-1,000,000 for \"any traveler.\" The scope is activity_specific_lifetime: per traveler-trip, not per adult lifetime. A reader taking multiple short Asia trips across a lifetime still sits well below 1 in 100,000. The upper edge of the uncertainty band uses the headline 1 × 10&#8315;&#8310; figure; the lower edge reflects vaccinated or very-short urban-only itineraries where multiple published reviews find essentially zero cases.\n","uncertainty":{"low":1e-7,"high":0.000001},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/japanese-encephalitis.html","title":"Japanese Encephalitis — CDC Yellow Book","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Overall incidence of JE among travelers from non-endemic countries is estimated at <1 case per 1 million travelers; short-term (<1 month) travelers restricted to major urban areas are at minimal risk; long-term rural travelers may approach the susceptible pediatric resident rate of 6-11 cases per 100,000 children per year.","excerpt":"\"The overall incidence of JE among people from non-endemic countries traveling to Asia is estimated to be &lt;1 case per 1 million travelers. &hellip; Shorter-term (e.g., &lt;1 month) travelers whose visits are restricted to major urban areas are at minimal risk for JE. &hellip; Expatriates and travelers staying prolonged periods in rural areas with active JE virus transmission might be at similar risk as the susceptible pediatric resident population, which is 6–11 cases per 100,000 children per year.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173859/https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/japanese-encephalitis.html","calculation_notes":"CDC Yellow Book gives the headline traveler figure (&lt;1 per million) used as the upper edge of the uncertainty band, and the long-term rural figure (6–11 per 100,000 per year) used as the anchor for the \"long-term rural traveler, 1+ month wet season\" row of regional_breakdown. The short-term urban point estimate of 5 × 10&#8315;&#8311; is the CDC’s &lt;1 per million divided by a factor of 2 to reflect the \"minimal risk\" subgroup language.\n","independence_note":"CDC Yellow Book is the primary US traveler-facing clinical guidance, synthesised from CDC programmatic estimates and the peer-reviewed travel-medicine literature. Shares CDC publisher and authorship overlap with the ACIP MMWR recommendations (Hills et al. co-authored both); Hills et al. 2010 and WHO are the meaningfully independent cross-checks.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20439978/","title":"Japanese encephalitis in travelers from non-endemic countries, 1973–2008","publisher":"American Journal of Tropical Medicine and Hygiene (Hills, Griggs, Fischer)","source_type":"peer_reviewed","statistic":"55 published JE cases in travelers from 17 non-endemic countries over 1973-2008; 10 deaths (18%); 24 (44%) with sequelae; 65% of detailed-assessment cases had spent >=1 month in endemic areas; overall estimate <1 case per million travelers.","excerpt":"\"We identified 55 JE cases in travelers or expatriates from 17 non-endemic countries. &hellip; Ten (18%) patients died and 24 (44%) had sequelae. &hellip; Among 37 case-patients with detailed risk assessment, 24 (65%) had spent ≥ 1 month in JE-endemic areas. &hellip; The overall risk of JE for travelers to JE-endemic countries is estimated to be less than 1 case/1 million travelers.\"\n","source_date":"2010-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173939/https://pubmed.ncbi.nlm.nih.gov/20439978/","calculation_notes":"Hills et al. is the peer-reviewed anchor for the &lt;1 per million figure and for the long-duration-rural skew: roughly two-thirds of traveler cases with detailed data occurred in people who had spent a month or more in endemic areas. This is the empirical basis for the ~200× personal_factor_multiplier on \"long-term rural stay.\" Drawn from a case pool (published traveler case reports 1973–2008) that is methodologically independent of the CDC programmatic estimate, so it counts as genuine corroboration rather than restatement.\n","independence_note":"Hills et al. is a published case-series review; the CDC Yellow Book figure derives from CDC’s programmatic estimates and case reporting. The two overlap (Hills is a CDC author) but use different pipelines, so treat as partially dependent.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/68/rr/rr6802a1.htm","title":"Japanese Encephalitis Vaccine: Recommendations of the Advisory Committee on Immunization Practices","publisher":"US CDC / MMWR Recommendations and Reports (Hills et al.)","source_type":"govt_report","statistic":"Overall JE incidence among US travelers 1993-2017 estimated at <1 case per million trips to Asia; vaccine not recommended for short-term urban-only itineraries or travel outside the transmission season.","excerpt":"\"The overall incidence of JE among persons from nonendemic countries who travel to Asia is estimated to be less than one case per 1 million travelers. &hellip; JE vaccine is not recommended for travelers with very low-risk itineraries, such as shorter-term travel limited to urban areas or travel that occurs outside of a well-defined JE virus transmission season.\"\n","source_date":"2019-07-19","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174017/https://www.cdc.gov/mmwr/volumes/68/rr/rr6802a1.htm","calculation_notes":"ACIP’s own recommendation text explicitly does NOT recommend the vaccine for short-term urban-only itineraries. This is the primary citation for the \"debunked\" myth_framing: the official body tasked with vaccine policy agrees that the risk for the modal Asia tourist is too low to warrant the vaccine.\n","independence_note":"Shares CDC upstream with the Yellow Book entry; used here for the explicit vaccine-policy framing rather than a second independent estimate of incidence.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/japanese-encephalitis","title":"Japanese encephalitis — Fact sheet","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"~100,000 clinical JE cases per year globally (95% CI: 61,720-157,522); case fatality rate up to 30%; 20-30% of survivors suffer permanent cognitive, behavioural or neurological sequelae; 24 countries in SE Asia and Western Pacific have transmission risk covering >3 billion people.","excerpt":"\"There are an estimated about 100 000 clinical cases (95% CI: 61 720– 157 522) of JE globally each year. &hellip; The case fatality rate can be as high as 30% among those with disease symptoms. &hellip; Of those who survive, 20–30% suffer permanent cognitive, behavioural or neurological sequelae such as seizures. &hellip; Twenty-four countries in the WHO South-East Asia and Western Pacific Regions have JEV transmission risk, which includes more than 3 billion people.\"\n","source_date":"2024-05-09","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174053/https://www.who.int/news-room/fact-sheets/detail/japanese-encephalitis","calculation_notes":"WHO provides the severity anchor: ~30% case fatality and ~20-30% sequelae among survivors. Applied to the short-term traveler point estimate of 5 × 10&#8315;&#8311;, the per-trip death risk is on the order of 1 in 6-7 million and the per-trip severe-sequelae risk is of similar magnitude — an order of magnitude rarer than being killed in a commercial plane crash per flight.\n","independence_note":"WHO figure is based on the Campbell et al. (2011) Bulletin of the WHO global burden estimate, independent of the CDC traveler-specific numbers.\n"}],"comparison_anchors":[{"label":"Death in a commercial plane crash (per flight)","lifetime_us_adult":7.3e-8},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by shark attack (lifetime, US)","lifetime_us_adult":1.76e-7},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Short-term urban traveler (<1 month, major cities)","probability":5e-7,"notes":"CDC Yellow Book describes this group as 'minimal risk'; on the order of ~1 per 2 million trips."},{"region":"Any traveler to JE-endemic Asia (all durations, unstratified)","probability":0.000001,"notes":"CDC / ACIP headline figure: <1 case per million travelers."},{"region":"Long-term rural traveler (1+ month, wet season, rural endemic area)","probability":0.0001,"notes":"Hills et al. highest-risk subgroup: ~65% of documented traveler cases had stayed >=1 month. CDC says this group approaches the pediatric resident rate of 6-11/100,000/year."},{"region":"Endemic rural resident (susceptible child, for reference — not a traveler risk)","probability":0.01,"notes":"Native exposure risk anchor, not a traveler figure. Routine JE vaccination is now standard in endemic-country childhood schedules."},{"region":"Vaccinated traveler (full 2-dose schedule, any itinerary)","probability":1e-7,"notes":"Post-licensure effectiveness is very high; no JE cases have been reported in travelers who completed a full vaccine course per published reviews."}],"personal_factor_multipliers":[{"factor":"Long-term rural stay in endemic zone, wet season","multiplier":200,"notes":"Hills et al.: ~65% of documented traveler cases had stayed a month or more in endemic rural areas."},{"factor":"Full JE vaccination (2-dose schedule completed)","multiplier":0.05,"notes":"Very high post-licensure effectiveness; zero documented vaccine-failure cases in the published traveler literature."},{"factor":"DEET + long sleeves + screened or air-conditioned accommodation","multiplier":0.5,"notes":"JE vector (Culex tritaeniorhynchus) is a dusk/night feeder; bite avoidance measurably reduces exposure."},{"factor":"Short-term visit limited to major urban areas","multiplier":0.5,"notes":"CDC: this group is 'minimal risk'. Applied as 0.5x against the <1-per-million headline."}],"short_label":"Japanese encephalitis (travel)","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The &lt;1-per-million headline figure is an average across all travelers to JE-endemic Asia, and the average hides a very wide spread. A two-week business trip to Tokyo or a weekend in Seoul is effectively zero-risk; a six-month rural posting in Bihar or the Mekong Delta during the wet season can approach the local susceptible-resident rate of 6–11 per 100,000 per year. This entry is a per-trip figure for the typical short-term urban traveler, not a per-adult-lifetime figure, and not a substitute for itinerary-specific advice from a qualified travel-medicine clinician. Note also that this entry measures the risk of contracting JE, not of dying from it: the case fatality rate among symptomatic cases is in the 20–30% range (WHO), so the per-trip death risk is roughly another factor of 4–5 lower than the incidence figure shown here.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized rice-paddy silhouette with a thin crescent moon above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/japanese-encephalitis-travel","api_url":"https://likelier.app/api/fears/japanese-encephalitis-travel.json"},{"slug":"asteroid-impact-death","question":"What are the odds of dying from an asteroid or comet impact?","quick_answer":"The statistical lifetime risk of dying from an asteroid or comet impact is about 1 in 1,350,000. The dinosaur analogy makes something like 1 in 10,000 feel plausible, far above the modelled figure — though the risk is dominated by rare, high-casualty events.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Asteroid impacts occupy a peculiar slot in the public imagination: simultaneously dismissed as science fiction and cited as an existential threat. No rigorous population survey tracks how often Americans worry about asteroid strikes specifically, but asteroid-impact scenarios reliably generate outsized media coverage whenever a newly discovered near-Earth object makes a close approach. The result is an availability-driven sense that the risk is either zero or catastrophic, with little middle ground.\n","rough_estimate":"~1 in 10,000 lifetime feels plausible to many who recall the dinosaur analogy","kind":"intuition"},"native":{"display":"~100 expected fatalities per year globally (statistical average)","numerator":100,"denominator":8000000000,"unit":"per year","population":"Global population, all ages, statistical expectation over geological timescales"},"normalized":{"lifetime_us_adult":7.4e-7,"display":"~1 in 1,350,000 lifetime (US adult)","log_value":-6.13,"assumptions":"Post-Spaceguard estimates put the annualized expected fatalities from asteroid and comet impacts at roughly 100 per year globally, down from the ~1,000/yr figure Chapman & Morrison used in 1994, because the survey has retired the statistical contribution of most civilization-ending impactors (≥1 km). Annual individual risk: 100 / 8 × 10⁹ ≈ 1.25 × 10⁻⁸. Compounded over 59 remaining adult years: 1 − (1 − 1.25 × 10⁻⁸)⁵⁹ ≈ 7.4 × 10⁻⁷. This is a statistical expectation smoothed over millions of years; in any given century the probability of a fatal impact is dominated by a single low-probability, high-casualty event.\n","uncertainty":{"low":1e-7,"high":0.000006},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.nature.com/articles/367033a0","title":"Impacts on the Earth by asteroids and comets: assessing the hazard","publisher":"Nature","source_type":"peer_reviewed","statistic":"Individual lifetime risk of ~1 in 6,000 over 50 years (pre-Spaceguard, including civilization-ending impacts)","excerpt":"\"There is a 1-in-10,000 chance that a large (~2-km diameter) asteroid or comet will collide with the Earth during the next century, disrupting the ecosphere and killing a large fraction of the world's population.\"\n","source_date":"1994-01-06","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260129114720/https://www.nature.com/articles/367033a0","calculation_notes":"Chapman & Morrison 1994 estimated ~1,000 expected fatalities per year globally from all NEO sizes, yielding a 50-year individual risk of roughly 1 in 6,000. This was a pre-Spaceguard estimate that included the full statistical contribution of rare civilization-ending impacts (≥1 km). Subsequent survey work has retired most of that contribution; the revised post-Spaceguard expectation is ~100 fatalities/yr, which reduces the lifetime figure by roughly an order of magnitude.\n","independence_note":"Chapman & Morrison's framework is the foundational risk assessment; later estimates by Stokes et al. (2003) and Rumpf et al. (2017) use independent impact-physics models and updated NEO population surveys but build on the same probabilistic approach.\n"},{"url":"https://cneos.jpl.nasa.gov/sentry/","title":"Sentry: Earth Impact Monitoring","publisher":"NASA Center for Near Earth Object Studies","source_type":"govt_report","statistic":"No known large NEO has any significant probability of impacting Earth in the next 100 years; 95%+ of 1 km+ NEOs catalogued","excerpt":"\"Sentry is a highly automated collision monitoring system that continually scans the most current asteroid catalog for possibilities of future impact with Earth over the next 100 years.\"\n","source_date":"2025-12-31","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260318035242/https://cneos.jpl.nasa.gov/sentry/","calculation_notes":"NASA CNEOS Sentry continuously monitors all catalogued NEOs. As of 2025, approximately 95% of near-Earth asteroids 1 km and larger have been discovered, and none has a significant impact probability in the next century. The residual risk comes almost entirely from undiscovered objects below 140 m. The post-survey annualized fatality expectation of ~100/yr globally (down from ~1,000/yr) reflects this retired risk. Individual annual risk: 100 / 8 × 10⁹ ≈ 1.25 × 10⁻⁸. Over 59 adult years: 1 − (1 − 1.25 × 10⁻⁸)⁵⁹ ≈ 7.4 × 10⁻⁷ ≈ 1 in 1,350,000.\n","independence_note":"CNEOS Sentry is an operational monitoring system tracking real NEOs with radar and optical astrometry. Its data are independent of the statistical population models used by Chapman & Morrison and Rumpf et al.\n"},{"url":"https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017GL073191","title":"Asteroid impact effects and their immediate hazards for human populations","publisher":"Geophysical Research Letters","source_type":"peer_reviewed","statistic":"Wind blast and overpressure are the dominant casualty mechanisms for land impacts of 50–400 m asteroids","excerpt":"\"We find that wind blast is the most dangerous impact effect, followed by pressure shock wave and thermal radiation. These three effects account for the vast majority of casualties.\"\n","source_date":"2017-04-19","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250702150434/https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017GL073191","calculation_notes":"Rumpf et al. 2017 extended casualty modeling to smaller impactor sizes than Chapman & Morrison considered, using high-fidelity impact-physics simulations. Their per-impactor casualty estimates are broadly consistent with earlier work but refine the hazard allocation across effects. The headline lifetime probability we use derives from the post-Spaceguard annualized expectation (~100 fatalities/yr) rather than from Rumpf's per-event figures directly, but Rumpf's work validates the casualty-per-impact assumptions underlying that expectation.\n","independence_note":"Rumpf et al. use independent impact-physics and population-exposure models distinct from both Chapman & Morrison's analytical approach and CNEOS's orbit-monitoring system.\n"},{"url":"https://www.nasa.gov/news-release/nasa-confirms-dart-mission-impact-changed-asteroids-motion-in-space/","title":"NASA Confirms DART Mission Impact Changed Asteroid's Motion in Space","publisher":"NASA","source_type":"govt_report","statistic":"DART's kinetic impact on Dimorphos on September 26, 2022 shortened its orbital period around Didymos by roughly 32-33 minutes, exceeding the mission's minimum-success benchmark by more than 25 times","excerpt":"\"This marks humanity's first time purposely changing the motion of a celestial object and the first full-scale demonstration of asteroid deflection technology.\"\n","source_date":"2022-10-11","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260607210420/https://www.nasa.gov/news-release/nasa-confirms-dart-mission-impact-changed-asteroids-motion-in-space/","calculation_notes":"Cited to support the body's claim that asteroid risk is \"actively managed,\" not merely observed via CNEOS Sentry monitoring. DART does not change the headline probability estimate, which is derived from the annualized ~100 fatalities/yr figure documented in the Chapman & Morrison, CNEOS, and Rumpf et al. sources above.\n","independence_note":"An operational deflection demonstration, independent of the statistical hazard models (Chapman & Morrison, Rumpf et al.) and the discovery/monitoring system (CNEOS Sentry) already cited in this entry.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by shark attack (lifetime, US)","lifetime_us_adult":1e-7},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"short_label":"Asteroid impact","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"The \"lifetime probability\" of dying from an asteroid impact is a statistical fiction in a way that most other entries on this site are not. Almost all the expected fatalities come from a single class of event — a large (≥140 m) impactor striking a populated area or triggering a global catastrophe — that has not occurred in recorded history. The annualized figure of ~100 expected fatalities per year is the quotient of billions of potential casualties divided by millions of years between events. In any given human lifetime, the actual probability is overwhelmingly either zero (no impact occurs) or catastrophically high (one does). The uncertainty band spans more than an order of magnitude because the residual population of undiscovered sub-140 m objects is poorly constrained, and because casualty estimates for ocean impacts (tsunami generation) vary by factors of 10–100 depending on modeling assumptions.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-gate-review","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single stylized asteroid silhouette against a dark sky with faint star-field dots, flat vector illustration."},"canonical_url":"https://likelier.app/asteroid-impact-death","api_url":"https://likelier.app/api/fears/asteroid-impact-death.json"},{"slug":"glyphosate-roundup-cancer","question":"What are the odds that dietary glyphosate (Roundup) exposure will give you cancer?","quick_answer":"Dietary glyphosate (Roundup) cancer risk over a US adult lifetime is set at about 1 in 1,000,000 or less, and may be effectively zero, since the largest applicator study found no cancer association. Concern runs far ahead of that: about 46% of US adults rank pesticide residues among their top food-safety worries.\n","category":"cancer","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Glyphosate is the most-used herbicide in world agriculture and the active ingredient in Roundup. The IARC 2015 reclassification of glyphosate as \"probably carcinogenic to humans\" (Group 2A), the multi-billion-dollar Monsanto/Bayer verdict against the company in Johnson v. Monsanto (2018), and a steady stream of \"Roundup found in your cereal\" press cycles have cemented public concern. Many consumers treat any detectable glyphosate residue on oats, wheat or chickpeas as a meaningful cancer risk and purchase organic specifically to avoid it. Surveys consistently rank pesticides among the top food-safety concerns of US adults.\n","rough_estimate":"46% of US adults rank pesticides as a top-3 food safety concern (IFIC 2025)","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey -- 46% of US adults rank pesticides and pesticide residues as a top-3 food-safety concern","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~99.9% of US dietary intake estimates fall below 1/1000 of EFSA's 0.5 mg/kg/day acceptable daily intake; no statistically significant cancer association in the largest applicator cohort","numerator":1,"denominator":1000000,"unit":"per dietary lifetime (attributable cancer harm, US adult)","population":"US adults consuming conventional produce, grains and processed foods"},"normalized":{"lifetime_us_adult":0.000001,"display":"~1 in 1,000,000 or less lifetime (US adult dietary)","log_value":-6,"assumptions":"The US adult lifetime dietary cancer attributable to glyphosate is set as a conservative 1-in-1,000,000 placeholder, the same floor used by `pesticide-residue-food`, because no dietary-exposure cohort has detected a cancer signal and applicator-cohort evidence (Andreotti et al. 2018, JNCI, N=54,251 -- the largest prospective study available) reports no association between glyphosate use and non-Hodgkin lymphoma (rate ratio 0.87, 95% CI 0.64-1.20 for highest exposure quartile vs never users, p-trend 0.95). Typical US dietary intake estimates from the FDA Total Diet Study are roughly 1/1000 of EFSA's acceptable daily intake of 0.5 mg/kg/day. The IARC 2015 Group 2A classification (Monograph 112) rested on \"limited\" human evidence from occupational case-control studies and \"sufficient\" evidence in laboratory rodents; EPA (2020 interim review), EFSA (2023 peer review), ECHA (2022 RAC opinion), Health Canada (2017 re-evaluation), and Australian APVMA have each concluded that glyphosate is not likely to be carcinogenic to humans at expected exposure levels. The 1-in-1,000,000 figure is a conservative ceiling acknowledging the IARC dissent and possible low-dose effects not yet measurable; the true dietary figure may be effectively zero. The wide uncertainty band reflects this institutional disagreement, not measured variability.\n","uncertainty":{"low":1e-7,"high":0.00001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/featured-news/media-centre-iarc-news-glyphosate/","title":"IARC News: Q&A on Glyphosate","publisher":"International Agency for Research on Cancer (IARC), WHO","source_type":"govt_report","statistic":"IARC Monograph Volume 112 (March 2015) classified glyphosate as Group 2A, \"probably carcinogenic to humans\", based on limited evidence in humans for non-Hodgkin lymphoma and sufficient evidence in laboratory animals.","excerpt":"Glyphosate was classified as \"probably carcinogenic to humans\" (Group 2A) based on \"limited\" evidence of cancer in humans from real-world exposures that actually occurred, \"sufficient\" evidence of cancer in experimental animals from studies of pure glyphosate, and \"strong\" evidence for genotoxicity, both for pure glyphosate and for glyphosate formulations.\n","source_date":"2016-03-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260529140301/https://www.iarc.who.int/featured-news/media-centre-iarc-news-glyphosate/","calculation_notes":"The IARC Group 2A classification is a hazard-identification, not a risk-quantification. It states that the agent has the potential to cause cancer under some exposure conditions, not that any particular exposure level produces a measurable cancer rate. The dietary cancer probability for a typical consumer is not computable from this source alone. Cited here to anchor the IARC half of the institutional split and to ensure the entry is honest about the strongest evidence in the \"concerning\" direction.\n","independence_note":"IARC's evaluation drew on ~1,000 publicly available studies via an independent working group with conflict-of-interest screening. The working-group methodology is institutionally distinct from EPA, EFSA and ECHA cancer-classification processes, which weight the evidence base differently (notably by including registrant-submitted studies not always in IARC's review universe).\n"},{"url":"https://www.epa.gov/ingredients-used-pesticide-products/glyphosate","title":"Glyphosate (Ingredient Used in Pesticide Products)","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"EPA Interim Registration Review (2020) concluded that glyphosate is not likely to be carcinogenic to humans at exposure levels expected from approved uses; review covered 15 acceptable carcinogenicity studies, more than IARC's 8 animal studies.","excerpt":"\"EPA concluded that glyphosate is not likely to be carcinogenic to humans.\" EPA further notes that its position is based on \"a more extensive dataset of studies\" than IARC reviewed.\n","source_date":"2022-08-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260531020904/https://www.epa.gov/ingredients-used-pesticide-products/glyphosate","calculation_notes":"EPA's Interim Registration Review for glyphosate (2020) and subsequent defenses of that decision form the institutional counterweight to IARC. Like IARC the conclusion is hazard-level rather than a quantitative cancer rate, but the framing -- \"not likely carcinogenic to humans\" at expected exposures -- bears directly on the dietary consumer scope this entry concerns. Used jointly with EFSA below to establish that the regulatory consensus across the largest assessment bodies is non-carcinogenic at dietary levels.\n","independence_note":"EPA's review universe includes registrant-submitted carcinogenicity studies that IARC excluded; EPA and IARC analysed substantially overlapping but not identical study sets. EFSA's evaluation is institutionally separate and reaches a similar conclusion via a separate review pipeline.\n"},{"url":"https://www.efsa.europa.eu/en/news/glyphosate-no-critical-areas-concern-data-gaps-identified","title":"Glyphosate: No Critical Areas of Concern; Data Gaps Identified","publisher":"European Food Safety Authority (EFSA)","source_type":"govt_report","statistic":"EFSA July 2023 peer review concluded no critical areas of concern for human, animal or environmental health from glyphosate; relied on ECHA RAC 2022 opinion that glyphosate does not meet criteria for classification as carcinogenic, mutagenic or reprotoxic.","excerpt":"EFSA's peer review concluded there were \"no critical areas of concern\" regarding glyphosate's impact on human health, animal welfare, or environmental safety. EFSA relied on ECHA's classification finding that glyphosate \"did not meet the scientific criteria to be classified as a carcinogenic, mutagenic or reprotoxic substance.\"\n","source_date":"2023-07-06","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260625233642/https://www.efsa.europa.eu/en/news/glyphosate-no-critical-areas-concern-data-gaps-identified","calculation_notes":"EFSA's 2023 conclusion is the third major regulatory data point against IARC's Group 2A classification (after EPA 2020 and ECHA 2022 RAC). It does not produce a dietary cancer probability either, but strengthens the case that across the major Western regulators the consensus on the dietary-consumer cancer question converges on \"no measurable increase\". Combined with Andreotti below, this is the basis for the 1-in-1,000,000 conservative placeholder rather than a higher figure.\n","independence_note":"EFSA's peer review draws on ECHA's hazard classification and on member-state rapporteur assessments. Methodologically independent of both EPA (different review pipeline, different study weighting) and IARC (different evaluation rules; EFSA reviews regulatory dossiers including industry submissions, IARC focuses on publicly available studies).\n"},{"url":"https://academic.oup.com/jnci/article/110/5/509/4590280","title":"Glyphosate Use and Cancer Incidence in the Agricultural Health Study","publisher":"Journal of the National Cancer Institute (Andreotti et al.)","source_type":"peer_reviewed","statistic":"Prospective US cohort, N=54,251 licensed pesticide applicators (44,932 glyphosate users; 7,290 incident cancers diagnosed during follow-up). No association between glyphosate use and non-Hodgkin lymphoma overall or any NHL subtype (rate ratio 0.87, 95% CI 0.64-1.20 for highest exposure quartile vs never users, p-trend 0.95).","excerpt":"\"In this large, prospective cohort study, no association was apparent between glyphosate and any solid tumors or lymphoid malignancies overall, including NHL and its subtypes.\"\n","source_date":"2018-02-08","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260307215642/https://academic.oup.com/jnci/article/110/5/509/4590280","calculation_notes":"The Agricultural Health Study is the largest prospective cohort of licensed applicators with quantitative glyphosate-exposure data. It addresses the cancer hypothesis most prominently raised by IARC (non-Hodgkin lymphoma) and reports no association at any exposure quartile. A nominally elevated acute-myeloid-leukemia signal in the highest exposure quartile (rate ratio 2.44, 95% CI 0.94-6.32) is the one residual finding worth flagging and is not statistically significant. This is the strongest single piece of dietary-relevant negative evidence: applicators have orders-of-magnitude higher exposure than dietary consumers, and the null at the high end implies a very small ceiling on consumer-level dietary harm.\n","independence_note":"The Agricultural Health Study is a long-running NCI/NIEHS/EPA collaboration tracking ~89,000 farmers and spouses; independent of IARC, EPA classification reviews and EFSA peer reviews. The Andreotti 2018 update extended follow-up by ~12 years relative to the 2005 analysis that contributed to IARC's evaluation.\n"}],"comparison_anchors":[{"label":"Pesticide residue on conventional produce (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Baseline non-Hodgkin lymphoma (lifetime, US adult)","lifetime_us_adult":0.0214},{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"US adult dietary consumer (FDA Total Diet Study + Andreotti 2018)","probability":0.000001,"notes":"intake typically ~1/1000 of EFSA ADI; placeholder floor based on null applicator-cohort finding extrapolated downward by orders of magnitude"},{"region":"EU adult dietary consumer (EFSA 2023 dietary risk assessment)","probability":5e-7,"notes":"modestly lower glyphosate residue exposure than US estimates; EFSA dietary risk assessment found no consumer concern at approved-use levels"},{"region":"Occupational licensed applicator (Agricultural Health Study cohort, high exposure quartile)","probability":0.025,"notes":"baseline lifetime NHL risk ~2.14%; Andreotti 2018 found RR 0.87 (not statistically elevated) for highest glyphosate exposure quartile -- so the applicator subgroup tracks the baseline rate rather than dramatically exceeding it; scope is occupational, not dietary"}],"personal_factor_multipliers":[{"factor":"Occupational licensed pesticide applicator (mixing, loading or spraying glyphosate)","multiplier":5,"notes":"Applicator exposure via dermal and inhalation routes runs orders of magnitude higher than dietary exposure. Andreotti et al. 2018 reports no statistically significant cancer association at any exposure quartile, including the highest, but the IARC dissent and the qualitatively elevated (non-significant) acute-myeloid-leukemia signal in the high quartile warrant a meaningful multiplier here. The figure does not extrapolate Zhang 2019 to a lifetime number; it represents the plausibility ceiling implied by the evidence base taken as a whole.\n"},{"factor":"Hobby gardener using glyphosate-formulation sprays heavily (multiple applications per year without PPE)","multiplier":2,"notes":"A heavy hobby user does not approach occupational applicator exposure but does exceed typical dietary intake by a wide margin. See `gardening-without-ppe` for the broader gardening-exposure perspective; that entry pegs glyphosate at roughly 1e-6 for a typical hobbyist with normal use patterns.\n"},{"factor":"Heavy consumer of US wheat-, oat- and pulse-based products (commodities with higher pre-harvest desiccation residues)","multiplier":1.5,"notes":"EWG and FDA Total Diet Study residue surveys consistently identify oats and pulses as the highest-residue commodities in the US food supply because of pre-harvest desiccation practices. Even at the upper end of consumption, intake remains well below the EFSA ADI; the modest multiplier reflects exposure rather than measured cancer risk.\n"},{"factor":"Exclusively organic diet (low-residue baseline)","multiplier":0.3,"notes":"Certified-organic food has substantially lower glyphosate residue levels but not zero (atmospheric drift, contamination). The reduction in already-very-low dietary exposure is real but the absolute risk difference at consumer scope is small.\n"}],"short_label":"Glyphosate & cancer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses dietary glyphosate exposure for a US adult consuming conventional produce, grains and processed food, and its plausible attributable cancer risk. It does not cover occupational applicator exposure beyond the multiplier above, ecological effects (pollinators, amphibians, aquatic invertebrates), herbicide-resistance evolution in weed populations, or non-cancer endpoints (kidney function, endocrine effects, microbiome) which remain active research areas without population-scale consensus. The 1-in-1,000,000 figure is a conservative ceiling, not a measured rate -- no dietary-exposure cohort has detected a cancer signal at typical consumer exposures, and the largest applicator cohort (Andreotti 2018, N=54,251) reports null at all exposure quartiles. The IARC Group 2A vs EPA/EFSA/ECHA/Health Canada non- carcinogen split is real and load-bearing; this entry sides with the larger-cohort-weighted regulatory consensus for the dietary scope while flagging the IARC dissent in the assumptions. Zhang et al. 2019 meta- analysis (a frequently cited counterpoint) is methodologically controversial because of its high-exposure-only subgroup focus and was not used to derive a lifetime number here; the literature on its conclusions remains genuinely split. If long-term US dietary cohort data eventually report a signal, this entry will be revised.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":5,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single stalk of wheat against a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/glyphosate-roundup-cancer","api_url":"https://likelier.app/api/fears/glyphosate-roundup-cancer.json"},{"slug":"pesticide-residue-food","question":"What are the odds that pesticide residue on conventional produce will harm your health?","quick_answer":"The odds that pesticide residue on conventional produce causes you serious harm are about 1 in 1,000,000 or lower over a lifetime — statistically indistinguishable from zero. Nearly half of US adults rank pesticide residue among their top food-safety worries, far out of proportion to that.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Pesticide residue on fruits and vegetables is one of the most widely cited food-safety anxieties in the US. The Environmental Working Group's annual \"Dirty Dozen\" list reliably generates headlines and drives billions of dollars in organic purchasing decisions. Surveys consistently find that a majority of US adults believe pesticide residues on conventional produce pose a meaningful health risk, with many assuming that eating non-organic strawberries or apples carries a real chance of cancer or chronic illness.\n","rough_estimate":"46% of US adults rank pesticide residues among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 46% of US adults rank pesticides and pesticide residues as a top-3 food safety concern","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~99% of US produce samples below EPA tolerances; typical exposures >1,000x below reference doses","numerator":1,"denominator":1000000,"unit":"per year (attributable serious harm)","population":"US adults consuming conventional produce"},"normalized":{"lifetime_us_adult":0.000001,"display":"~1 in 1,000,000 or less lifetime (US adult)","log_value":-6,"assumptions":"No published epidemiological cohort has found a statistically significant increase in cancer or chronic disease attributable specifically to dietary pesticide residues at levels found on US commercial produce. Winter & Katz (2011) showed that actual dietary exposures from the highest-residue commodities are typically 1,000-10,000x below EPA chronic reference doses (RfDs), which themselves incorporate 100-1,000x safety factors below no-observed-adverse-effect levels in animal studies. The USDA Pesticide Data Program (2023) found >99% of samples below EPA tolerances. Given this margin, any attributable lifetime risk to a typical US adult consumer is not statistically distinguishable from zero in any existing data. The 1-in-1,000,000 figure is a conservative placeholder acknowledging theoretical non-zero risk; the true figure could be much lower (effectively zero) or modestly higher if future research identifies low-dose effects not currently captured. The uncertainty band reflects this epistemic gap rather than measured variability.\n","uncertainty":{"low":1e-7,"high":0.00001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3135239/","title":"Dietary Exposure to Pesticide Residues from Commodities Alleged to Contain the Highest Contamination Levels","publisher":"Journal of Toxicology / Winter & Katz","source_type":"peer_reviewed","statistic":"Exposures from highest-residue produce are >1,000x below chronic reference doses in >90% of pesticide-commodity comparisons","excerpt":"\"All pesticide exposure estimates were well below established chronic reference doses (RfDs). The RfDs were more than 1000 times higher than the exposure estimates in more than 90 percent of the comparisons.\"\n","source_date":"2011-10-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413181605/https://pmc.ncbi.nlm.nih.gov/articles/PMC3135239/","calculation_notes":"Winter & Katz analyzed the 10 commodities on EWG's 2010 Dirty Dozen list and computed dietary exposure estimates using USDA PDP residue data and FDA consumption data. The worst-case pairing (methamidophos on bell peppers) still had an RfD 49.5x the exposure estimate. At 0.01% of the RfD, exposure is one million times lower than the no-observed-adverse-effect level. This paper directly addresses the gap between perceived and actual dietary pesticide risk from the commodities consumers worry about most.\n","independence_note":"Winter & Katz is an independent academic analysis of USDA PDP data; uses the same underlying residue measurements but applies a distinct risk-assessment methodology (margin-of-exposure approach) that is independent of USDA's standard reporting."},{"url":"https://www.food-safety.com/articles/9895-usda-testing-for-2023-shows-99-percent-of-foods-do-not-exceed-pesticide-residue-tolerances","title":"USDA Testing for 2023 Shows 99 Percent of Foods Do Not Exceed Pesticide Residue Tolerances","publisher":"Food Safety Magazine (reporting USDA PDP 2023 Annual Summary)","source_type":"govt_report","statistic":">99% of 9,832 samples compliant with EPA tolerances; 38.8% had no detectable residues","excerpt":"\"More than 99 percent of sampled products to be compliant with pesticide residue tolerances.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421192158/https://www.food-safety.com/articles/9895-usda-testing-for-2023-shows-99-percent-of-foods-do-not-exceed-pesticide-residue-tolerances","calculation_notes":"The USDA PDP tests ~10,000 food samples per year for hundreds of pesticide analytes. In 2023, 9,832 samples were tested; 240 samples (2.4%) had residues exceeding EPA maximum residue limits or contained residues with no established tolerance. 38.8% of all samples had no detectable residues at all. Exceeding an EPA tolerance does not equate to health harm — tolerances are set with 100-1,000x safety factors — but the >99% compliance rate confirms that the food supply operates well within established safety margins. Used alongside Winter & Katz to corroborate the negligible-risk conclusion.\n","independence_note":"The USDA PDP is the same upstream residue database used by Winter & Katz 2011 for their exposure calculations. Treat as partially dependent — PDP provides the raw residue data, Winter & Katz provide the exposure-to-RfD comparison.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1367841/","title":"Organic Diets Significantly Lower Children's Dietary Exposure to Organophosphorus Pesticides","publisher":"Environmental Health Perspectives / Lu et al.","source_type":"primary_study","statistic":"Switching 23 elementary school-age children from conventional to organic diets reduced urinary organophosphate metabolites to non-detectable levels within days","excerpt":"\"We used a novel study design to provide a convincing demonstration of the ability of organic diets to reduce children's OP pesticide exposure and the health risks that may be associated with these exposures.\"\n","source_date":"2006-02-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260528062828/https://pmc.ncbi.nlm.nih.gov/articles/PMC1367841/","calculation_notes":"Lu et al. measured urinary dimethyl and dimethylthio metabolites of malathion and chlorpyrifos in 23 children aged 3-11 over 15 days: 3 days conventional, 5 days organic, 7 days conventional. Median metabolite concentrations dropped to non- detectable within ~24 hours of the organic-diet switch and rebounded within ~36 hours of returning to conventional food. This anchors both the kid-subgroup multiplier (children carry higher per-kg dietary exposure than adults) and the organic-diet multiplier (the exposure delta is real and immediate; the *health* delta at these exposure levels remains undocumented in any cohort outcome).\n","independence_note":"Lu 2006 is a controlled biomarker intervention — methodologically independent from Winter & Katz (dietary modeling) and USDA PDP (residue surveillance). It measures internal exposure, not external residue. The three sources together triangulate: Winter & Katz model expected exposure, PDP confirms residue levels are compliant, Lu confirms biomarker exposure tracks with diet choice. None of the three measure attributable health harm — that gap persists.\n"}],"comparison_anchors":[{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"US adult, conventional diet (baseline)","probability":0.000001,"notes":"The headline number. Reflects USDA PDP 2023 compliance and Winter & Katz exposure modeling >1,000x below EPA chronic reference doses.\n"},{"region":"US adult, exclusively organic diet","probability":5e-7,"notes":"Roughly half the conventional-diet attributable risk per the organic-diet multiplier. Halving a sub-1-in-a-million number remains sub-1-in-a-million in practical terms.\n"},{"region":"US child under 6, conventional diet","probability":0.000003,"notes":"3x baseline due to higher per-kg dietary intake. Lu 2006 confirms biomarker exposure in this age band is meaningfully higher than adults on the same food supply. The attributable health risk is still below epidemiological detection thresholds.\n"},{"region":"US child under 6, organic-diet swap","probability":0.0000015,"notes":"Combined kid (×3) and organic (×0.5) multipliers. Lu 2006 documents urinary metabolites drop to non-detectable within 24 hours of swap. The exposure delta is real and rapid; whether it translates to a measurable lifetime health benefit is unestablished.\n"},{"region":"Heavy consumer of imported produce from less-regulated markets","probability":0.000005,"notes":"Per FDA FY2023: 13.5% of import samples had violative residues vs 2.8% domestic. Higher exposure but still well within safety margins for most violative findings.\n"}],"personal_factor_multipliers":[{"factor":"Occupational farm worker (mixing/applying pesticides)","multiplier":100,"notes":"Occupational exposure via dermal and inhalation routes is orders of magnitude higher than dietary consumer exposure. Farm workers handling pesticides face a genuinely elevated risk profile that dietary residue data does not capture.\n"},{"factor":"Children under 6 (higher intake per kg body weight)","multiplier":3,"notes":"Children eat more food per kilogram of body weight and have developing organ systems. EPA sets separate tolerances accounting for this, but the per-kg exposure is roughly 2-5x higher than adults. Lu et al. 2006 (Environ Health Perspect) measured detectable urinary OP-pesticide metabolites in conventionally-fed kids aged 3-11 that dropped to non-detectable within 24 hours of an organic-diet switch — confirming the exposure delta is real and immediate, even if the *health* delta at these levels remains undocumented in cohort outcomes.\n"},{"factor":"Children aged 6-12 (intermediate per-kg exposure)","multiplier":2,"notes":"Older children still carry roughly 2x adult per-kg dietary exposure but less than under-6s. The Lu 2006 cohort included this age band and showed the same biomarker response to organic-diet swap. EPA's FQPA tenfold safety factor for children applies across all minors, not just under-6s.\n"},{"factor":"Exclusively organic diet (kid or adult)","multiplier":0.5,"notes":"Organic produce has lower (but not zero) pesticide residue levels. Lu 2006 documents a near-complete drop in urinary OP metabolites within 24 hours of an organic-diet switch in school-age children. The exposure reduction is real; the absolute health- risk reduction is vanishingly small because the baseline is already 1,000-10,000x below EPA reference doses. Reduces a sub-1-in-a-million risk by roughly half.\n"},{"factor":"Heavy consumer of imported produce from less-regulated markets","multiplier":5,"notes":"FDA FY2023 data shows 13.5% of import samples had violative residues vs 2.8% domestic. Higher residue levels on some imports may modestly increase exposure, though still typically within safety margins.\n"}],"short_label":"Pesticide residue","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses health harm from dietary pesticide residues on commercially available produce consumed at normal levels by US adults. It does not cover occupational exposure (farm workers, pesticide applicators), environmental or ecological effects of pesticide use, acute poisoning from misuse, or endocrine-disruption hypotheses at very low doses which remain an active area of research without consensus on population-level health impact. The normalized probability is a conservative placeholder because the true attributable risk has not been isolated in any published cohort — it may be effectively zero. The wide uncertainty band reflects genuine scientific uncertainty about low-dose chronic effects, not measured variability in a known distribution. The 1-in-1,000,000 figure is a conservative placeholder; no epidemiological study has measured attributable mortality from regulatory-compliant pesticide residues. The true figure may be effectively zero.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent + 8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted green apple on a pale grey background, flat vector illustration."},"canonical_url":"https://likelier.app/pesticide-residue-food","api_url":"https://likelier.app/api/fears/pesticide-residue-food.json"},{"slug":"snake-bite-fatal","question":"What are the odds of being killed by a venomous snake?","quick_answer":"A US adult's lifetime odds of being killed by a venomous snake are about 1 in 880,000 (roughly 5 deaths a year nationwide) — vastly rarer than the 1-in-a-few-thousand many people picture.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Snakes are a textbook example of a prepared fear — ophidiophobia is one of the most commonly reported specific phobias in the general population, and the imagery of a venomous bite is culturally ancient. But \"fear of snakes\" as measured in surveys bundles the phobia with garden-variety squeamishness, and no rigorous recent poll isolates \"fear of being killed by a venomous snake bite\" from that broader bucket. The perceived side here is marked intuition rather than survey.\n","rough_estimate":"most people guess something like 1 in a few thousand","kind":"intuition"},"native":{"display":"~5 venomous snake bite deaths per year (US)","numerator":5,"denominator":260000000,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.00000113,"display":"1 in ~880,000 lifetime (US adult)","log_value":-5.95,"assumptions":"Uses ~5 venomous snake bite deaths per year in the United States (CDC NIOSH) across a US adult population of ~260 million, giving an annual per capita risk of ~1.9e-8. Compounded over 59 years of remaining adult life. The Greene et al. (2021) 30-year average of ~3.4 fatal bites per year is slightly lower and sits inside the uncertainty band.\n","uncertainty":{"low":7e-7,"high":0.0000018},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/niosh/outdoor-workers/about/venomous-snakes.html","title":"Venomous Snakes at Work","publisher":"US Centers for Disease Control and Prevention (NIOSH)","source_type":"govt_report","statistic":"7,000-8,000 venomous snake bites per year in the US; about 5 deaths","excerpt":"\"Each year, 7,000-8,000 people are bitten by venomous snakes in the United States. About 5 of those people die. The number of deaths would be much higher if people did not seek medical care.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183905/https://www.cdc.gov/niosh/outdoor-workers/about/venomous-snakes.html","calculation_notes":"CDC NIOSH gives the round-number US figure of ~5 deaths per year. Dividing by a US adult population of ~260M gives ~1.9e-8 per adult per year; compounded over 59 years of remaining adult life yields the normalized lifetime figure of roughly 1 in 880,000.\n","independence_note":"CDC NIOSH occupational-safety summary consolidates multiple US public-health data streams (WONDER, AAPCC NPDS, CDC surveillance). Partially overlaps with the upstream sources used by Greene et al., so is not fully independent of the peer-reviewed figure below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33046301/","title":"Epidemiology of fatal snakebites in the United States 1989-2018","publisher":"American Journal of Emergency Medicine (Greene, Folt, Wyatt, Brandehoff)","source_type":"peer_reviewed","statistic":"101 fatal native-snake bites in the US over 1989-2018 (~3.4 per year); 90.2% rattlesnakes","excerpt":"\"We identified 101 fatal bites from native snakes... Rattlesnakes accounted for 74 (90.2%) of the 82 deaths for which the species was known or which occurred where rattlesnakes are the only native crotalids. There were five fatalities attributed to copperheads, two due to cottonmouths, and one caused by an eastern coral snake.\"\n","source_date":"2021-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260505063844/https://pubmed.ncbi.nlm.nih.gov/33046301/","calculation_notes":"Greene et al.'s 30-year average (~3.4 deaths/year) is the peer-reviewed lower bound and corroborates the CDC figure. Using this figure instead of 5 would shift the lifetime estimate to ~1 in 1.3M. The uncertainty band spans both.\n","independence_note":"CDC NIOSH and Greene et al. draw from overlapping but not identical case pools (NIOSH summarizes multiple public-health sources; Greene et al. use AAPCC NPDS plus CDC WONDER death certificates), so this counts as meaningful independent corroboration.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/snakebite-envenoming","title":"Snakebite envenoming — Fact sheet","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"Global snakebite deaths: 81,410-137,880 per year","excerpt":"\"Around 81 410 to 137 880 people die each year because of snake bites, and around three times as many amputations and other permanent disabilities are caused by snakebites annually.\"\n","source_date":"2023-09-13","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260409193332/https://www.who.int/news-room/fact-sheets/detail/snakebite-envenoming","calculation_notes":"Included as context for the global asymmetry: snakebite is a major neglected tropical disease globally, but the US-specific fatality rate is roughly four orders of magnitude lower per capita than the South Asian and sub-Saharan African rates that drive the global total. We normalize on the US figure, not this one.\n","independence_note":"WHO figure derives primarily from Kasturiratne et al. (2008) and subsequent country-level verbal autopsy studies; independent of the US-centric CDC and AAPCC data.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by bee/wasp/hornet sting (lifetime, US)","lifetime_us_adult":0.0001267},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Agricultural worker in sub-Saharan Africa or South Asia","multiplier":10000,"notes":"WHO 2019 snakebite burden report estimates 81,000-138,000 deaths/year globally, almost entirely in rural South Asia and sub-Saharan Africa, versus ~5/year in the US. Per-capita rate in high-burden regions (e.g., India, Nigeria) is roughly 4-5 orders of magnitude above the US baseline. WHO Global Snakebite Initiative, 2019."},{"factor":"No antivenom access within 6 hours of bite","multiplier":5,"notes":"WHO snakebite treatment guidelines indicate that case fatality rates for untreated envenomation by medically significant species can reach 5-10%, versus ~0.1-0.2% with prompt antivenom. The ~5x multiplier reflects the treated-vs-untreated gap documented in WHO 2019 snakebite guidelines and systematic reviews (Kasturiratne et al., Lancet 2008)."},{"factor":"Barefoot outdoor work in snake-active habitat","multiplier":3,"notes":"Epidemiological surveys in high-burden regions consistently show that foot/ankle bites account for the majority of snakebite incidents, predominantly among agricultural workers without foot protection. WHO 2019 snakebite burden report identifies barefoot agricultural exposure as the leading behavioral risk factor; protective footwear trials show ~3x reduction in bite incidence."},{"factor":"US urban/suburban resident with minimal outdoor exposure","multiplier":0.1,"notes":"Fatal US snakebite is heavily concentrated in southern and midwestern states within rattlesnake range and in outdoor settings (Greene et al., American Journal of Emergency Medicine, 2021). Urban residents with no hiking, camping, or agricultural exposure face substantially below-average risk; the 0.1x estimate is consistent with geographic concentration in Greene et al.'s 30-year case series."}],"short_label":"Snake bite","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a US-population-level figure. It does not apply globally: snakebite envenoming is one of the world's most important neglected tropical diseases, killing an estimated 81,000-138,000 people per year, almost entirely in South Asia and sub-Saharan Africa where antivenom access is limited. It also does not apply uniformly within the US — fatal bites concentrate heavily in southern and midwestern states and almost entirely in rattlesnake-range outdoor exposures.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":4,"d8":4,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized coiled snake silhouette against a warm sand-colored background, flat vector illustration."},"canonical_url":"https://likelier.app/snake-bite-fatal","api_url":"https://likelier.app/api/fears/snake-bite-fatal.json"},{"slug":"spider-bite-serious","question":"What are the odds of a medically significant spider bite?","quick_answer":"Genuinely dangerous spider bites are far rarer than arachnophobia suggests, but poorly tracked: US poison centers log only ~2,500-3,000 spider-bite exposures a year, and most are minor and self-limiting. A bite needing medical care is more common than death, which is the vanishingly rare floor — about 7 US deaths a year, roughly 1 in 830,000 over a lifetime, far below the 1-in-a-few-hundred people guess.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Arachnophobia is one of the most prevalent specific phobias, affecting roughly 3-6% of the general population depending on which diagnostic threshold you use. The cultural image of the dangerous spider is vivid, ancient, and self-reinforcing: nearly every unexplained skin lesion that shows up in an ER invites the question \"could it be a spider bite?\" No rigorous recent survey isolates \"fear of a medically significant spider bite\" from the broader arachnophobia bucket, so the perceived side here is marked intuition.\n","rough_estimate":"most people guess something like 1 in a few hundred or thousand","kind":"intuition"},"native":{"display":"~7 spider bite deaths per year (US, ICD-10 X21)","numerator":7,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.0000012,"display":"1 in ~830,000 lifetime (US adult)","log_value":-5.92,"assumptions":"Uses the CDC NCHS/WONDER average of ~7 deaths per year from contact with venomous spiders (ICD-10 code X21) over the 2010-2015 reporting window. The 2008-2015 Forrester et al. analysis found 6 deaths per year over that period, consistent with this figure. Annual per capita risk: 7 / 335,000,000 ≈ 2.09 × 10^-8. Compounded over 59 years of remaining adult life: 1 - (1 - 2.09e-8)^59 ≈ 1.2 × 10^-6. This measures fatal outcomes only; medically significant (requiring medical attention) bites are more common but poorly tracked at the population level. Poison center data suggest ~2,500-3,000 spider bite exposures reported per year, but most are minor and self-limiting.\n","uncertainty":{"low":6e-7,"high":0.000003},"scope":"us_adult_lifetime"},"sources":[{"url":"https://blogs.cdc.gov/nchs/2017/05/31/3642/","title":"Deaths from Venomous Snakes, Lizards, Spiders, and Scorpions, 2010-2015","publisher":"US Centers for Disease Control and Prevention (CDC) / National Center for Health Statistics (NCHS)","source_type":"govt_report","statistic":"38 US deaths from venomous spiders over 2010-2015 (ICD-10 X21); average ~6.3 per year","excerpt":"\"Deaths from Venomous Snakes, Lizards, Spiders and Scorpions, 2010-2015\" tabulates X21 (Contact with venomous spiders) fatalities by year: 7 (2010), 3 (2011), 7 (2012), 7 (2013), 7 (2014), 7 (2015), totaling 38 deaths over the six-year period.\n","source_date":"2017-05-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260406200841/https://blogs.cdc.gov/nchs/2017/05/31/3642/","calculation_notes":"CDC NCHS reports 38 deaths from ICD-10 X21 over 2010-2015, or ~6.3 per year. We round to ~7 per year as a conservative central estimate reflecting year-to-year variability (range 3-7). Divided by US population (~335M) and compounded over 59 years gives the normalized lifetime figure.\n","independence_note":"Primary US mortality source for spider-bite deaths, drawn from the NCHS death-certificate / ICD-10 X21 pipeline. Methodologically independent of the clinical-cohort (Suchard) and entomological-specimen (Vetter) studies, which address misdiagnosis and incidence rather than fatal outcomes.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1071166/","title":"Myth: idiopathic wounds are often due to brown recluse or other spider bites throughout the United States","publisher":"Western Journal of Medicine / Richard S. Vetter","source_type":"peer_reviewed","statistic":"Several hundred 'brown recluse bites' reported in states where fewer than 10 specimens have been collected in 40+ years","excerpt":"\"Several hundred cases of 'brown recluse bites' have been reported\" in California alone over a decade, despite \"fewer than 10 brown recluse specimens in California in more than 40 years of records.\"\n","source_date":"2000-11-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260311023515/https://pmc.ncbi.nlm.nih.gov/articles/PMC1071166/","calculation_notes":"Vetter's work is not used for the mortality calculation but provides the critical context that most reported \"spider bites\" in the US are misdiagnoses. This shapes the interpretation of poison center data (which rely on caller self-report) and explains why the cultural perception of spider bite risk vastly exceeds reality.\n","independence_note":"Methodologically independent of CDC death-certificate data: Vetter's analysis draws from entomological specimen records and clinical case reports, not ICD-coded mortality statistics.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19939602/","title":"'Spider bite' lesions are usually diagnosed as skin and soft-tissue infections","publisher":"Journal of Emergency Medicine / Jeffrey Ross Suchard","source_type":"peer_reviewed","statistic":"Of 182 patients presenting with self-reported 'spider bites,' only 3.8% were diagnosed with actual spider bites; 85.7% had skin infections","excerpt":"\"ED patients reporting a 'spider bite' were most frequently diagnosed with skin and soft-tissue infections.\" Of 182 patients enrolled over 23 months, \"3.8%\" were diagnosed with actual spider bites while \"85.7%\" were diagnosed with infections.\n","source_date":"2011-11-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260119001439/https://pubmed.ncbi.nlm.nih.gov/19939602/","calculation_notes":"Suchard's emergency-department study quantifies the misdiagnosis rate: fewer than 1 in 25 self-reported \"spider bites\" are actual spider bites. This corroborates Vetter's earlier qualitative work and reinforces that population-level spider bite incidence is dramatically lower than public perception suggests.\n","independence_note":"Independent of both CDC mortality data and Vetter's entomological approach: this is a prospective clinical cohort study at a single ED.\n"}],"comparison_anchors":[{"label":"Death by bee/wasp/hornet sting (lifetime, US)","lifetime_us_adult":0.0001267},{"label":"Death by venomous snake bite (lifetime, US)","lifetime_us_adult":0.00000113},{"label":"Death by shark attack (lifetime, US)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Residence in brown recluse range (south-central US)","multiplier":3,"notes":"Brown recluse spiders (Loxosceles reclusa) have a well-defined range concentrated in Missouri, Kansas, Oklahoma, Arkansas, and adjacent states. Vetter (Western Journal of Medicine, 2000) documents that specimens are effectively absent outside this range. For the serious-but-non-fatal outcome framing of this entry, exposure probability is approximately 3x higher for residents within confirmed range than the national average."},{"factor":"Age under 6 or over 65","multiplier":3,"notes":"Poison Control data and clinical case series consistently show that young children and older adults experience more severe outcomes from black widow envenomation due to smaller body mass and reduced physiological reserve respectively. The ~3x severity multiplier for these age groups is consistent with patterns reported in Forrester et al. (Wilderness & Environmental Medicine, 2018) and AAPCC annual reports."},{"factor":"Immunocompromised status (chemotherapy, HIV, organ transplant)","multiplier":2,"notes":"Immunocompromised individuals face elevated risk of secondary infection from any skin-breaking bite and reduced tolerance for systemic envenomation effects. The ~2x multiplier for serious outcomes is a conservative estimate consistent with general immunocompromise risk adjustments in wound-care and envenomation literature; no spider-bite-specific immunocompromise RCT exists."},{"factor":"Urban resident, rarely outdoors, no woodpile or stored-equipment exposure","multiplier":0.2,"notes":"Black widow and brown recluse spiders preferentially inhabit undisturbed outdoor spaces (woodpiles, rock piles, stored equipment, outbuildings). Urban residents with minimal outdoor exposure and no stored wood or equipment have substantially below-average encounter probability. The 0.2x protective estimate is consistent with the geographic and behavioral concentration of serious US spider bite cases documented in CDC NCHS ICD-10 X21 mortality data (2010-2015)."}],"short_label":"Spider bite","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This figure captures fatal spider bite outcomes only (ICD-10 X21). Medically significant but non-fatal envenomations are poorly tracked at the population level; poison center data suggest roughly 2,500-3,000 spider bite exposures are reported per year, but self-reporting inflates this number substantially given that the majority of self-diagnosed \"spider bites\" turn out to be bacterial skin infections. Only two spider species in the United States are considered medically significant: the black widow (Latrodectus) and the brown recluse (Loxosceles reclusa). Brown recluse bites are geographically confined to a well-defined range in the south-central United States; reports of brown recluse bites from outside this range are almost always misidentifications. Black widows are more widespread but rarely cause serious illness in healthy adults with access to medical care. Note: despite the 'serious' in the slug, this entry's quantitative figure measures *fatal* spider bite rates (ICD-10 X21). 'Serious but non-fatal' outcomes — significant envenomation requiring medical care — are roughly 10-100x more common but lack reliable population-level measurement in the US.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":3,"d6":4,"d7":4,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized spider silhouette resting on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/spider-bite-serious","api_url":"https://likelier.app/api/fears/spider-bite-serious.json"},{"slug":"vaccine-serious-adverse-event","question":"What are the odds of a serious adverse event from a routine vaccine?","quick_answer":"Anaphylaxis, the best-measured serious adverse event, occurs at about 1.31 per million vaccine doses — roughly 1 in 760,000. Most readers guess the risk at 1 in 100 or worse, three to four orders of magnitude too high.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Public perception of vaccine serious adverse event (SAE) rates runs orders of magnitude above the measured numbers, driven by a combination of very vivid individual case reports, the 2020-2022 COVID-mRNA myocarditis signal, and the asymmetry between a population-wide intervention and a background disease most readers never directly see. We have not found a recent rigorous poll that isolates “fear of a serious adverse event from a routine vaccine” from the much broader category of general vaccine hesitancy, so the perceived entry here is editorial intuition, not survey. The directionally robust observation is that when asked to guess SAE rates, readers typically name figures in the low single-digit percent range per dose — roughly three to four orders of magnitude above the measured per-dose rate from the Vaccine Safety Datalink.\n","rough_estimate":"most readers guess the risk at 1 in 100 or worse per dose; measured rates are ~1 in a million for anaphylaxis","kind":"intuition"},"native":{"display":"~1.31 anaphylaxis cases per million vaccine doses (Vaccine Safety Datalink, 2009-2011)","numerator":33,"denominator":25173965,"unit":"per vaccine dose","population":"US patients in the Vaccine Safety Datalink, 2009-2011, all routine vaccines"},"normalized":{"lifetime_us_adult":0.00000131,"display":"~1 in 760,000 per dose (anaphylaxis, all routine vaccines)","log_value":-5.88,"assumptions":"The headline figure is the anaphylaxis rate from McNeil et al. (JACI 2016), the largest prospective US study of vaccine-triggered anaphylaxis: 33 confirmed cases across 25,173,965 doses in the Vaccine Safety Datalink, for a rate of 1.31 per million doses (95% CI 0.90 to 1.84). We use anaphylaxis as the canonical SAE anchor because it is the best-measured, most consistently defined serious adverse event across the routine vaccine schedule, and because it is the specific “serious” outcome most often invoked in pre-vaccination consent discussions. Scope is declared as activity_specific_lifetime: this is a per-dose probability for the specific activity of receiving a routine vaccine, not a general US-adult-lifetime number, and is not directly comparable to the population-lifetime figures on other Likelier pages. Other measurable SAE signals across the routine schedule sit within the same order of magnitude: Guillain-Barre after influenza vaccine at approximately 1 to 2 additional cases per million doses, historical oral polio vaccine-associated paralytic polio at approximately 1 per 2 to 3 million doses. The notable outlier is post-2021 myocarditis after mRNA COVID-19 second doses in young males, measured at approximately 40.6 per million second doses in males 12 to 29 per the CDC ACIP / Shimabukuro presentation to MMWR in June 2021, with higher rates (62.8 per million) in the 12 to 17 male subgroup. That signal is real and is the subject of the regional_breakdown and caveats sections below; it is also an order of magnitude above the all-vaccine anaphylaxis baseline, which is why Likelier reports both. Almost all measured SAEs resolve without lasting harm: McNeil et al. report zero deaths and only one hospitalization among the 33 anaphylaxis cases.\n","uncertainty":{"low":9e-7,"high":0.00000184},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26452420/","title":"Risk of anaphylaxis after vaccination in children and adults","publisher":"Journal of Allergy and Clinical Immunology / McNeil MM, Weintraub ES, Duffy J, et al. (PubMed / NLM)","source_type":"peer_reviewed","statistic":"33 confirmed vaccine-triggered anaphylaxis cases across 25,173,965 doses in the Vaccine Safety Datalink, 2009-2011; rate of 1.31 per million doses (95% CI 0.90-1.84); zero deaths, one hospitalization","excerpt":"\"We identified 33 confirmed vaccine-triggered anaphylaxis cases that occurred after 25,173,965 vaccine doses. The rate of anaphylaxis was 1.31 (95% CI, 0.90-1.84) per million vaccine doses.\"\n","source_date":"2016-03-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421201335/https://pubmed.ncbi.nlm.nih.gov/26452420/","calculation_notes":"McNeil et al. is a prospective cohort analysis across five Vaccine Safety Datalink sites over 2009-2011. The 33-case numerator divided by the 25,173,965-dose denominator gives the headline rate of ~1.31 per million doses used directly as the native value. Normalized value = numerator / denominator = 33 / 25,173,965 ≈ 1.31 × 10^-6. Uncertainty band uses the paper’s own 95% confidence interval of 0.90 to 1.84 per million, i.e. 9.0 × 10^-7 to 1.84 × 10^-6. Note: incorporating the mRNA myocarditis signal (40.6 per million = 4.06 × 10^-5) would widen the band considerably, but that signal is subgroup- and vaccine-specific and is reported separately in the regional_breakdown. The paper separately reports that all 33 cases recovered, with no deaths and only a single hospitalization, which is the basis for the “resolves without lasting harm in most cases” framing in the long-form body.\n","independence_note":"VSD-based. Draws on the same five-site cohort that CDC uses for most of its per-dose vaccine safety surveillance. Subsequent reviews (e.g., Dreskin et al. in Current Treatment Options in Allergy 2019) cite this paper as the canonical US per-dose anaphylaxis rate.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/70/wr/mm7027e2.htm","title":"Use of mRNA COVID-19 Vaccine After Reports of Myocarditis Among Vaccine Recipients: Update from the Advisory Committee on Immunization Practices — United States, June 2021","publisher":"Morbidity and Mortality Weekly Report / Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"40.6 myocarditis cases reported per million second doses of mRNA COVID-19 vaccine in males aged 12-29; 62.8 per million in males 12-17 and 50.5 per million in males 18-24","excerpt":"\"Myocarditis reporting rates were 40.6 cases per million second doses of mRNA COVID-19 vaccines administered to males aged 12&minus;29 years.\"\n","source_date":"2021-07-09","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260405172824/https://www.cdc.gov/mmwr/volumes/70/wr/mm7027e2.htm","calculation_notes":"The MMWR report is the ACIP / Shimabukuro presentation that crystallized the myocarditis signal. Used here not as the headline number but as the counter-example that explains why Likelier reports real SAE signals even when they are small. The 40.6 per million figure is approximately 30 times the all-vaccine anaphylaxis baseline of 1.31 per million, which is why the long-form body singles it out: it is the one recent routine-schedule vaccine SAE signal that is meaningfully above the century-average background rate. The same MMWR and follow-up CDC clinical guidance emphasize that the majority of reported myocarditis cases were mild and self-resolving with conservative management, which is what determines how Likelier reports the signal: a real elevated rate, a real clinical follow-up burden, but overwhelmingly mild outcomes.\n","independence_note":"This is the CDC / ACIP primary surveillance report, drawing on VAERS and V-Safe data which are the US federal vaccine safety monitoring systems. The clinical and cardiology follow-up literature (Mevorach et al. NEJM 2021, Witberg et al. NEJM 2021, Patone et al. Nature Medicine 2022) replicates the signal in independent Israeli and UK datasets, so the effect itself is not a single-source artefact even though the MMWR number is cited here as the US authoritative figure.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/71/wr/mm7114e1.htm","title":"Cardiac Complications After SARS-CoV-2 Infection and mRNA COVID-19 Vaccination — PCORnet, United States, January 2021-January 2022","publisher":"Morbidity and Mortality Weekly Report / Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Comparing infected persons with second-dose mRNA vaccine recipients, risk ratios for cardiac outcomes were 6.7-8.5 in males 18-29 and 10.8-115.2 in males 30 and older; all risk ratios statistically significant","excerpt":"\"RRs for cardiac outcomes comparing infected persons with first dose recipients were 7.2-61.8, and with second dose recipients, were 6.7-8.5; all RRs were statistically significant.\"\n","source_date":"2022-04-08","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260511083838/https://www.cdc.gov/mmwr/volumes/71/wr/mm7114e1.htm","calculation_notes":"PCORnet multi-site EHR study (40 health systems) comparing cardiac-outcome incidence after a positive SARS-CoV-2 test versus after first or second mRNA COVID-19 vaccine dose, by age and sex. Grounds the body-text comparison that infection carries a substantially higher cardiac-complication risk than vaccination in the same demographic: RR 6.7-8.5 in males 18-29 versus second-dose recipients, rising to 10.8-115.2 in males 30 and older (per the report's age-stratified tables). Replaces an earlier unsourced blanket \"six to ten times ... same age and sex subgroup\" framing, which did not distinguish age bands that in fact range roughly 7x to over 100x.\n","independence_note":"CDC-authored PCORnet analysis, methodologically distinct from the ACIP/VAERS-based MMWR source above (different surveillance system: electronic health records across a research network rather than passive adverse-event reporting), reaching a consistent conclusion.\n"}],"comparison_anchors":[{"label":"Fatal anaphylaxis from any cause (lifetime, US adult)","lifetime_us_adult":0.0000363},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death from measles (per measles case, pre-vaccine era)","lifetime_us_adult":0.001}],"regional_breakdown":[{"region":"All routine vaccines, anaphylaxis (VSD, 2009-2011)","probability":0.00000131,"notes":"Headline figure from McNeil et al. JACI 2016. 33 confirmed cases / 25.2 million doses. Zero deaths, 1 hospitalization across all cases."},{"region":"Influenza vaccine, anaphylaxis (inactivated trivalent, alone)","probability":0.00000135,"notes":"10 cases / 7,434,628 doses given alone, per McNeil et al. 1.35 per million, not meaningfully different from the all-vaccine average."},{"region":"mRNA COVID-19 vaccine, myocarditis, males 12-29 (CDC ACIP, June 2021, 2nd dose)","probability":0.0000406,"notes":"40.6 per million second doses per MMWR 70(27). The one routine-schedule SAE signal in the past decade that is meaningfully above the all-vaccine anaphylaxis baseline. Clinical outcomes overwhelmingly mild and self-resolving per follow-up literature."},{"region":"mRNA COVID-19 vaccine, myocarditis, males 12-17 (CDC ACIP, June 2021, 2nd dose)","probability":0.0000628,"notes":"62.8 per million second doses in the highest-signal subgroup. Still below 1 in 10,000 per dose."},{"region":"Influenza vaccine, Guillain-Barre syndrome (excess over background)","probability":0.0000015,"notes":"Approximately 1-2 additional GBS cases per million doses above background, per CDC and post-2009-H1N1 pharmacovigilance studies. Order of magnitude matches the anaphylaxis rate."},{"region":"Oral polio vaccine, vaccine-associated paralytic polio (historical)","probability":4e-7,"notes":"Approximately 1 case per 2-3 million doses of the oral (Sabin) vaccine, the reason the US switched to inactivated (Salk) vaccine in 2000. Included as historical context; the currently-used inactivated polio vaccine does not carry this risk."}],"personal_factor_multipliers":[{"factor":"known prior vaccine reaction","multiplier":5,"notes":"A documented prior anaphylactic or other serious reaction to a vaccine component is the strongest individual predictor of repeat-event risk, and is the standard indication for allergist-supervised administration. Order-of-magnitude estimate.\n"},{"factor":"no prior issues, routine schedule","multiplier":1,"notes":"The population baseline from McNeil et al. applies: roughly 1.31 per million doses for anaphylaxis.\n"},{"factor":"immunocompromised","multiplier":2,"notes":"For some live attenuated vaccines (MMR, varicella, yellow fever, oral typhoid, BCG, live attenuated influenza), immunocompromise is generally a contraindication rather than a quantified multiplier, because the dominant concern is strain replication rather than anaphylaxis. The multiplier here is a rough order-of-magnitude for inactivated-vaccine SAE elevation in immunocompromised patients, not a clinical recommendation.\n"},{"factor":"young male, mRNA COVID-19 second dose","multiplier":30,"notes":"Specific to the myocarditis signal, not anaphylaxis. 40.6 per million in males 12-29 divided by the 1.31 per million all-vaccine anaphylaxis baseline gives a rough 30x factor. This is the only routine-schedule subgroup currently above the century-average SAE background.\n"}],"short_label":"Vaccine reaction","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1.31 per million headline covers anaphylaxis specifically, which is the best-measured SAE across the routine vaccine schedule and the outcome most commonly cited in pre-vaccination consent. It is not a single all-inclusive “any serious adverse event” rate, because the definition of “serious” varies across surveillance systems. Broader aggregate rates of medically-attended serious adverse events across the full US routine schedule run roughly 1 in 100,000 to 1 in 1,000,000 per dose depending on definition, which brackets the anaphylaxis figure. Second caveat: post-2021 mRNA COVID-19 vaccines produced a real, measured elevated myocarditis signal in young males receiving second doses, at approximately 40.6 per million second doses in males 12-29 and up to 62.8 per million in males 12-17 per the CDC ACIP MMWR of June 2021. Likelier publishes real measured numbers, and that signal is real; the clinical follow-up literature (Mevorach et al., Witberg et al., Patone et al.) confirms that the observed cases were overwhelmingly mild and self-resolving with conservative management, but the elevated incidence itself is not in dispute. Third caveat, which is probably the most important comparison Likelier can offer: for every vaccine with a measurable SAE rate, the disease the vaccine prevents has a higher rate of the same SAE. Measles causes myocarditis more often than MMR. SARS-CoV-2 infection causes myocarditis more often than an mRNA second dose, including in young males. Measles case fatality runs roughly 1 per 1,000 in high-income settings, approximately 770 times the MMR anaphylaxis rate. The denominator on “risk of a serious event from a vaccine” is always best understood next to the denominator on “risk of the same serious event from the disease the vaccine prevents.” See <a href=\"/fears/anaphylaxis-fatal\">anaphylaxis-fatal</a> for the related entry on fatal anaphylaxis from all causes.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small glass vial resting on a pale neutral surface, flat vector illustration in muted greys and soft blue."},"canonical_url":"https://likelier.app/vaccine-serious-adverse-event","api_url":"https://likelier.app/api/fears/vaccine-serious-adverse-event.json"},{"slug":"infant-swing-death","question":"What are the odds an infant in a powered swing or rocker dies from strangulation or suffocation?","quick_answer":"Across the recalled US powered-swing lines, the odds an infant died from strangulation or suffocation were about 1 in 720,000 per unit (6 deaths across roughly 4.32 million units), close to the roughly 1 in 500,000 parents intuit. Two distinct mechanisms drove them: strap entanglement and sleep suffocation with added bedding.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":false,"perceived":{"description":"Powered infant swings sit in most US nurseries and a growing share of Polish ones, and the parental mental model is overwhelmingly that of a soothing device — the mechanical equivalent of a rocking arm. The two recent recalls (4moms Mamaroo in 2022, Fisher-Price Snuga in 2024) received broad coverage in mainstream parenting media, so awareness of \"infant swings have killed babies\" is now fairly high. What is less well understood is that the two recalls describe two distinct mechanisms — a crawling infant entangled in a dangling strap under an empty seat, versus a sleeping infant suffocating with added bedding — and that brand-level awareness does not necessarily translate into mechanism-level caution.\n","rough_estimate":"~1 in 500,000 per powered-swing unit","kind":"intuition"},"native":{"display":"6 reported infant deaths across two major US powered-swing recalls (1 Mamaroo plus 5 Snuga) covering approximately 4.32 million units","numerator":6,"denominator":4320000,"unit":"per powered-swing unit sold in the recalled US product lines","population":"US infants 0-12 months exposed to powered swings (Mamaroo and RockaRoo 2010-2022, Snuga 2010-2024)"},"normalized":{"lifetime_us_adult":0.0000014,"display":"~1 in 720,000 chance per powered-swing unit; two distinct mechanisms (strap entanglement on Mamaroo, sleep suffocation on Snuga with added bedding)","log_value":-5.85,"assumptions":"Six deaths total: one Mamaroo (a 10-month-old strangled by entanglement in a dangling strap while crawling under the unoccupied seat) plus five Snuga (infants aged 1-3 months, all during sleep use, four of five with added bedding, most unrestrained). Both recalls describe clear product-specific mechanisms. Per-unit denominator is well-established from CPSC-verified unit counts: 2.0 million MamaRoo plus 220,000 RockaRoo plus 2.1 million Snuga = approximately 4.32 million units. Per-unit fatality rate ~1 in 720,000. Per-infant risk is somewhat higher because units often see more than one user over their service life; an order-of-magnitude estimate is ~1 in 200,000 to 1 in 500,000 per regularly-exposed infant, with the uncertainty band reflecting the unknown average users-per-unit and the very small numerator (n=6).\n","uncertainty":{"low":5e-7,"high":0.00001},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Recalls/2022/4moms-Recalls-More-than-2-Million-MamaRoo-and-RockaRoo-Infant-Swings-and-Rockers-Due-to-Entanglement-and-Strangulation-Hazards-One-Death-Reported","title":"4moms Recalls More than 2 Million MamaRoo and RockaRoo Infant Swings and Rockers Due to Entanglement and Strangulation Hazards; One Death Reported","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"2 million MamaRoo plus 220,000 RockaRoo units recalled; one 10-month-old died from asphyxiation entangled in dangling restraint strap; one other infant hospitalized","excerpt":"\"4moms has received two reports of entanglement incidents involving infants who became caught in the strap under the unoccupied MamaRoo infant swing after they crawled under the seat, including a 10-month-old infant who died from asphyxiation, and a 10-month-old infant who suffered bruising to his neck before being rescued by a caregiver.\"\n","source_date":"2022-08-15","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260531140318/https://www.cpsc.gov/Recalls/2022/4moms-Recalls-More-than-2-Million-MamaRoo-and-RockaRoo-Infant-Swings-and-Rockers-Due-to-Entanglement-and-Strangulation-Hazards-One-Death-Reported","calculation_notes":"Two million MamaRoo plus 220,000 RockaRoo equals approximately 2.22 million units. Mechanism is a mobile infant crawling under an unoccupied unit and becoming entangled in a dangling restraint strap. Distinct from the sleep-suffocation mechanism in the Snuga recall.\n"},{"url":"https://www.cpsc.gov/Recalls/2025/Fisher-Price-Recalls-More-than-2-Million-Snuga-Infant-Swings-Due-to-Suffocation-Hazard-After-5-Deaths-Reported","title":"Fisher-Price Recalls More than 2 Million Snuga Infant Swings Due to Suffocation Hazard; After 5 Deaths Reported","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"Approximately 2.1 million Snuga swings sold; 5 infant deaths (ages 1-3 months) reported 2012-2022 during sleep use, mostly with added bedding and unrestrained","excerpt":"\"Between 2012 and 2022, there have been reports of five deaths involving infants 1 to 3 months of age when the product was used for sleep. In most of those incidents, the infants were unrestrained and bedding materials were added to the product. Approximately 2.1 million swings were sold in the United States.\"\n","source_date":"2024-10-10","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260524011223/https://www.cpsc.gov/Recalls/2025/Fisher-Price-Recalls-More-than-2-Million-Snuga-Infant-Swings-Due-to-Suffocation-Hazard-After-5-Deaths-Reported","calculation_notes":"CPSC filed the canonical URL under /Recalls/2025/ even though the recall release date is 2024-10-10. All five deaths involved sleep use; four of five involved added bedding; most were unrestrained. Mechanism is distinct from the Mamaroo entanglement (sleep suffocation vs strap entanglement). 5 / 2.1M = 2.4e-6 per Snuga unit.\n"},{"url":"https://www.congress.gov/bill/117th-congress/house-bill/3182","title":"H.R.3182 — Safe Sleep for Babies Act of 2021","publisher":"US Congress / Congress.gov","source_type":"reputable_reference","statistic":"Federal ban on inclined sleep surfaces greater than 10 degrees; powered swings remain a regulatory grey area when used for sleep with added bedding","excerpt":"\"Safe Sleep for Babies Act of 2021 — This bill makes it unlawful to manufacture, sell, or distribute crib bumpers or inclined sleepers for infants. Specifically, inclined sleepers for infants are those designed for an infant up to one year old and have an inclined sleep surface of greater than 10 degrees. Latest Action: 05/16/2022 Became Public Law No: 117-126.\"\n","source_date":"2022-05-16","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251203193834/https://www.congress.gov/bill/117th-congress/house-bill/3182","calculation_notes":"The 2022 Act addresses inclined sleepers but powered swings designed for awake soothing fall into a regulatory grey zone. The five Snuga deaths illustrate the off-label-sleep residual risk that the Act does not directly close — the product is sold for awake use, but the deaths occurred during sleep use with added bedding.\n"}],"comparison_anchors":[{"label":"Inclined sleeper death (per unit, historical)","lifetime_us_adult":0.0000042},{"label":"SIDS death (per infant-year)","lifetime_us_adult":0.00014},{"label":"Bumbo seat fall (per unit, CPSC 2007/2012)","lifetime_us_adult":0.00001}],"personal_factor_multipliers":[{"factor":"Use of swing for unsupervised sleep","multiplier":30,"notes":"All 5 Snuga deaths occurred during sleep use; supervised awake use risk is much lower"},{"factor":"Added bedding placed in swing","multiplier":10,"notes":"4 of 5 Snuga deaths involved added bedding contrary to product instructions"},{"factor":"Infant unrestrained in seat","multiplier":5,"notes":"Most Snuga deaths involved unrestrained infants; restraint use reduces the chance of re-positioning into airway-obstructing posture"},{"factor":"Mobile crawling infant (8-12 months) with access to unoccupied swing","multiplier":4,"notes":"Mamaroo death involved a crawling infant entangled under the seat; younger infants pose no such risk because they cannot crawl"},{"factor":"Compliance with EN 16232+A2:2023 (EU/PL infant swing standard)","multiplier":0.7,"notes":"EU standard covers children up to 9 kg or pre-sit-up and tightens clamp/strap requirements, but does not address sleep-use risk explicitly"}],"short_label":"Infant swing death","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Six deaths across 4.32 million units is a small numerator, so the per-unit point estimate carries wide uncertainty even before accounting for the unknown average users-per-unit. The two mechanisms also do not aggregate cleanly: the Mamaroo death required a mobile infant who could crawl under an empty seat, while the Snuga deaths required a sleeping infant with added bedding. Different households face different mechanism-weighted risk depending on which products are present and how they are used. CPSC's incident database under-counts non-fatal events, so the headline figure is most reliable as a fatality rate and least reliable as a total-harm rate.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"An empty modern powered infant swing in a residential room, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/infant-swing-death","api_url":"https://likelier.app/api/fears/infant-swing-death.json"},{"slug":"whole-grape-child-choking","question":"What are the odds of a young child choking to death on a whole grape?","quick_answer":"On the order of 1 in 600,000 for a US child during the highest-risk years, birth to age four. Grapes are among the most-cited fatal choking foods — third or fourth, behind hot dogs, candy, and nuts — but the absolute death rate is low. Non-fatal choking scares are vastly more common, and cutting grapes lengthwise removes most of the hazard.\n","category":"kids","tags":["food","toddler","infant"],"no_reliable_estimate":false,"perceived":{"description":"The whole grape is one of the most vivid single-food fears of early parenthood. A viral X-ray of a grape lodged in a child's airway, the standard warning at every weaning class, and the near-universal instruction to cut grapes in half combine to make it feel like handing a toddler an uncut grape is an acute, plausible way to lose a child. The mental image most parents carry is that a whole grape is close to a loaded risk — that a meaningful fraction of children who eat one whole will choke, and that a fair number of those will die.\n","rough_estimate":"Most parents picture the per-child death risk as 'rare but real' — order 1 in a few thousand","kind":"intuition"},"native":{"display":"~1 in 600,000 per US child (birth-age 4); grapes ≈ 8% of pediatric food-choking deaths","numerator":1,"denominator":625000,"unit":"per child, birth-age 4 window","population":"US children under 5, whole-grape asphyxiation deaths"},"normalized":{"lifetime_us_adult":0.0000016,"display":"1 in ~600,000 per child during the birth-age-4 window (US)","log_value":-5.8,"assumptions":"Likelier normally reports lifetime-US-adult probabilities; this entry is scoped to the peak-risk age window (birth to age 4) for a single US child, matching the sibling entry toddler-choking-while-eating. The headline is built from two grounded inputs. First, the risk that a US child dies of food-related choking at all across the 0-4 window is roughly 1 in 50,000 (Likelier's toddler-choking-while-eating entry, itself derived from CDC W79 mortality data and Sideris et al. 2021). Second, grapes account for roughly 8% of US pediatric food-asphyxiation deaths (Harris et al. 1984: grapes were 8 of ~100 documented food-asphyxiation deaths, behind hot dogs at 17%, candy, and nuts). Applying the grape share to the all-food baseline gives 0.08 × (1 in 50,000) ≈ 1 in 625,000, rounded to ~1 in 600,000. Two honesty caveats bound this. (a) The 8% grape share comes from a single national case series of 1979-81 deaths; grape consumption and the now-widespread habit of cutting grapes have shifted the true share since, in an unknown direction, which is why the uncertainty band is wide. (b) In Harris et al. the food-type breakdown (\"17 cases, 17%\") and the national frequency (\"approximately one death every five days\") rest on different denominators — only the fraction is carried forward here, never a precise annual grape-death count. The round-fruit family parents also ask about (cherry tomatoes, cherries, large blueberries) shares the grape's geometry but is not isolated in the mortality data, so it is treated as the same mechanism in prose, not folded into this numerator.\n","uncertainty":{"low":4e-7,"high":0.000006},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/6708272/","title":"Childhood asphyxiation by food. A national analysis and overview","publisher":"JAMA — Harris CS, Baker SP, Smith GA, Harris RM (1984;251:2231-2235)","source_type":"peer_reviewed","statistic":"Of US pediatric food-asphyxiation deaths (41 states, 1979-81, ages 0-9): hot dog products 17 cases (17%), candy 10, nuts 9, grapes 8; >90% of deaths in children under 5 and 65% under 2; nationally about one food-asphyxiation death every five days","excerpt":"\"Hot dog products (17 cases, 17%), candy, ten; nuts, nine; and grapes, eight. [...] More than 90% occurred in infants and children younger than 5 years and 65% in infants younger than 2 years. [...] Nationally, one death occurred approximately every five days.\"\n","source_date":"1984-05-04","source_accessed":"2026-08-16","calculation_notes":"Source of the grape share. Grapes were 8 of the roughly 100 documented food-asphyxiation deaths in the case set (hot dogs 17 = 17% fixes the denominator at ~100), so grapes ≈ 8% of US pediatric food-asphyxiation deaths, ranking third or fourth behind hot dogs, candy, and nuts. That 8% share is multiplied by Likelier's ~1 in 50,000 all-food child-choking baseline to reach ~1 in 600,000. The \"one death every five days\" figure (~73/year, ages 0-9, 1979-81) is national-frequency context only and is NOT combined with the 8% share — the two rest on different denominators. The >90%-under-5 and 65%-under-2 distribution anchors the age multipliers.\n","independence_note":"A 1979-81 national death-certificate case series (NCHS plus state files). It is the last comprehensive US fatal-choking-by-food-type breakdown; later work (Chapin 2013, Sideris 2021) covers non-fatal ED visits or all-cause choking deaths without a food-type split, so the grape share still traces to this study. Old data — hence the wide uncertainty band.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27998886/","title":"The choking hazard of grapes: a plea for awareness","publisher":"Archives of Disease in Childhood — Lumsden AJ, Cooper JG (2017;102:473-474)","source_type":"peer_reviewed","statistic":"Whole grapes are ideally suited to cause paediatric airway obstruction; awareness that this popular fruit and other similarly shaped foods is a choking hazard is not widespread","excerpt":"\"Deaths from choking are a major cause of childhood mortality, especially in the very young. Whole grapes are ideally suited to cause paediatric airway obstruction and, though regularly implicated, knowledge that this popular fruit, and other similarly shaped foods, is a choking hazard is not widespread.\"\n","source_date":"2017-05-01","source_accessed":"2026-08-16","calculation_notes":"Used for the mechanism (a whole grape's ~2 cm diameter matches a young child's airway and its smooth skin forms a tight seal that is hard to dislodge) and to corroborate grapes as a top-ranked fatal-choking food, consistent with Harris's ~8%. The article's \"similarly shaped foods\" phrase is the basis for treating cherry tomatoes, cherries, and large blueberries as the same round-fruit mechanism in the prose. The widely-quoted \"third most common cause of US childhood choking death, 1999-2013\" line appears in press coverage of this letter and in secondary US reporting rather than in the abstract quoted above, so it is treated as corroborating context, not as the numerator source.\n","independence_note":"A single-institution UK case series plus literature review. Independent of the US NEISS/NCHS surveillance pipeline feeding Harris, Chapin, and Sideris.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33819896/","title":"Persistence of choking injuries in children","publisher":"International Journal of Pediatric Otorhinolaryngology — Sideris GA et al. (2021)","source_type":"peer_reviewed","statistic":"2,347 US pediatric choking deaths (ages 0-19) and 305,814 nonfatal injuries, 2001-2016; 75% of fatalities in children under 5; under-5 fatality rate unchanged after the 2010 AAP recommendations","excerpt":"\"From 2001 to 2016, there were a total of 305,814 nonfatal injuries and 2347 choking deaths in children from 0 to 19 years. [...] Children under five years of age accounted for 73% of nonfatal injuries and 75% of choking fatalities. [...] a decrease in the choking fatalities rate in all children (0.18/100,000 versus 0.16/100,000, respectively) but no change in fatalities rate for children under five.\"\n","source_date":"2021-04-01","source_accessed":"2026-08-16","calculation_notes":"Anchors the all-food baseline the grape share is applied to. 2347 deaths / 16 years ≈ 147 pediatric choking deaths per year (ages 0-19); 75% × 147 ≈ 110 under-5 deaths per year from all foreign-body choking, and the food subset (~41-50%) lands the all-food child-choking death rate near 1 in 50,000 over the 0-4 window — the figure carried in Likelier's toddler-choking-while-eating entry. The flat under-5 rate is why child-choking vigilance is treated as calibrated rather than debunked.\n","independence_note":"Draws from CPSC NEISS and NCHS mortality files — methodologically linked to Chapin 2013 and the CDC surveillance briefs; treat those as views of one pipeline. Does not break fatalities down by food type, so it sets the base rate, not the grape share.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23897916/","title":"Nonfatal Choking on Food Among Children 14 Years or Younger in the United States, 2001-2009","publisher":"Pediatrics — Chapin MM, Rochette LM, Annest JL, Haileyesus T, Conner KA, Smith GA","source_type":"peer_reviewed","statistic":"12,435 annual US pediatric ED visits for nonfatal food-related choking (2001-2009); rate 20.4 per 100,000; hard candy 15.5%, other candy 12.8%, meat 12.2%, bone 12.0% (grapes not in the top four non-fatal foods)","excerpt":"\"An estimated 111,914 (95% confidence interval: 83,975-139,854) children ages 0 to 14 years were treated in US hospital emergency departments from 2001 through 2009 for nonfatal food-related choking, yielding an average of 12,435 children annually and a rate of 20.4 (95% confidence interval: 15.4-25.3) visits per 100,000 population. [...] Hard candy was associated with the greatest number (15.5% [95% CI: 12.8-18.2]) of choking episodes, followed by other candy (12.8%), meat (12.2%), and bone (12.0%).\"\n","source_date":"2013-08-01","source_accessed":"2026-08-16","calculation_notes":"Grounds the non-fatal comparison the prose uses to keep the death number in perspective: ~12,435 US pediatric food-choking ED visits per year across all foods, orders of magnitude above the fatal rate. Crucially, grapes are NOT among Chapin's top-four non-fatal foods (hard candy, other candy, meat, bone), so this is an all-food scare-rate anchor, not a grape-specific one. The grape signal concentrates in the fatal data (Harris), consistent with a grape choke being disproportionately lethal when it happens rather than disproportionately frequent — the severity channel, not the frequency channel.\n","independence_note":"CPSC NEISS All Injury Program sample; shares upstream NEISS/NCHS data with Sideris 2021. Sets the non-fatal context, not the grape numerator.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20176668/","title":"Policy Statement — Prevention of Choking Among Children","publisher":"American Academy of Pediatrics, Committee on Injury, Violence, and Poison Prevention — Pediatrics 125(3):601-607","source_type":"reputable_reference","statistic":"AAP high-risk food list names whole grapes alongside hot dogs, hard candy, nuts and seeds, raw carrots, apples, and popcorn; choking is a leading cause of morbidity and mortality in children aged 3 years or younger","excerpt":"\"Choking is a leading cause of morbidity and mortality among children, especially those aged 3 years or younger. Food, coins, and toys are the primary causes of choking-related injury and death.\"\n","source_date":"2010-03-01","source_accessed":"2026-08-16","calculation_notes":"The canonical clinical authority placing whole grapes on the high-risk food list and naming the under-3 peak-risk group. It is the basis for the cutting-as-lever framing (round foods cut small, ideally quartered lengthwise) and the age multipliers. Reaffirmed by the AAP in October 2019. Provides no numerator — categorisation and age concentration only.\n","independence_note":"Expert-consensus policy synthesising the clinical case-series literature rather than a single surveillance dataset. Independent of the NEISS/NCHS mortality pipeline.\n"}],"comparison_anchors":[{"label":"All-food choking death, US child 0-4 (the broader parent risk)","lifetime_us_adult":0.00002},{"label":"SIDS, per US infant","lifetime_us_adult":0.00035},{"label":"Death in a plane crash, lifetime (US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Whole-grape asphyxiation death, US child 0-4 (the headline number)","probability":0.0000016,"notes":"Grapes ≈ 8% of pediatric food-asphyxiation deaths (Harris 1984) applied to the ~1 in 50,000 all-food child-choking baseline for the 0-4 window.\n"},{"region":"All-food asphyxiation death, US child 0-4 (for scale)","probability":0.00002,"notes":"The broader food-choking death risk grapes are one slice of. Grapes are roughly a twelfth of it — consistent with the ~8% share.\n"},{"region":"Whole-grape asphyxiation death, US child under 2 specifically","probability":0.0000032,"notes":"Harris 1984 puts 65% of food-asphyxiation deaths under age 2. Weighted by population-years that concentration works out to roughly twice the under-5 average (the pure statistical share is ~1.7-1.8x; airway geometry and the absence of grinding molars push it a little higher), so ~2x the ~1 in 600,000 headline. Directional.\n"}],"personal_factor_multipliers":[{"factor":"child under 2 (peak-risk age)","multiplier":2,"notes":"Harris 1984 concentrates 65% of food-asphyxiation deaths under age 2 and more than 90% under 5. Weighted by population-years that is about twice the under-5 average (~1.7-1.8x from the statistical share alone; the youngest children also combine the narrowest airway with the weakest chewing and no molars, so a whole grape is closest to airway-sized for them). Directional above the pure statistical concentration.\n"},{"factor":"grapes served whole rather than quartered lengthwise","multiplier":3,"notes":"The dominant modifiable input. A whole grape's roughly 2 cm diameter matches a young child's airway and its skin forms a seal; quartering lengthwise drops each piece well below airway diameter. No study publishes a clean cut-versus-whole risk ratio, so this multiplier is directional and uncertain — the entire clinical prevention message (AAP, Cooper & Lumsden) nonetheless rests on this geometry.\n"},{"factor":"continuous adult supervision during the meal","multiplier":0.5,"notes":"Most would-be-fatal grape chokes are interrupted by a nearby adult with back blows or a Heimlich maneuver, which is the main reason the fatal rate is far below the scary-incident rate. Directional; no published quantitative risk ratio isolates supervision for grapes specifically.\n"}],"short_label":"Whole-grape choking","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"On the plain perceived-versus-actual death axis this entry is tagged overrated: the roughly 1-in-600,000 fatal rate across the whole 0-4 window is a couple of hundred times below the \"1 in a few thousand\" most parents picture. The tag describes the probability gap, not the precaution — a grape choke is disproportionately lethal when it does happen, and cutting grapes lengthwise costs almost nothing, so the near-universal cut-the-grape habit is not misplaced even though the absolute death rate is low. (The sibling entry toddler-choking-while-eating is tagged calibrated because its all-food death gap is roughly ten times smaller and it carries a genuinely high all-food scare rate; grapes have neither, so overrated is the more honest tag here.) The headline counts whole-grape asphyxiation deaths only. It does not include the wider round-fruit family (cherry tomatoes, cherries, large blueberries) that shares the same geometry — the mortality data isolate grapes, not the category, so numbering the family would overstate what the evidence supports. It also excludes non-fatal chokes, aspiration pneumonia that presents days later, and grape chokes in older children and adults. The 8% grape share is a single 1979-81 national case series (Harris et al.); no newer US study breaks pediatric choking deaths down by food type, so the point estimate rests on old data and the uncertainty band (roughly 1 in 170,000 to 1 in 2.5 million) is deliberately wide to cover a grape share plausibly anywhere from 4% to 15% of a food-choking death rate that is itself only known to within a factor of two.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-08-16","methodology_version":"1.2"},"reviewer":"8d-eval-2026-08-16","last_reviewed":"2026-08-16","reviewed":true,"generated_at":"2026-08-16","image":{"alt":"A single whole round grape beside two lengthwise-quartered grape pieces on a pale grey-blue plate, viewed from directly above, flat vector illustration."},"canonical_url":"https://likelier.app/whole-grape-child-choking","api_url":"https://likelier.app/api/fears/whole-grape-child-choking.json"},{"slug":"hippo-attack","question":"What are the odds of being killed by a hippopotamus?","quick_answer":"The lifetime odds of being killed by a hippo are about 1 in 565,000 for a global adult, from an estimated 150 deaths a year (range 50-500) concentrated along African river systems. For anyone outside sub-Saharan Africa the practical risk is near zero, roughly where intuition already puts it — though hippos are far deadlier within their range than their cartoon reputation suggests.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Hippos enjoy an unearned reputation as comical, lumbering herbivores — perpetuated by decades of cartoons, zoo enclosures, and safari tour narration that emphasizes their yawning mouths as a sign of friendliness rather than a threat display. The animal that routinely kills fishermen, swimmers, and boat passengers along African river systems rarely features in the same mental category as lions or crocodiles, despite a kill rate that likely rivals both in their respective ranges.\n","rough_estimate":"most people outside sub-Saharan Africa would guess near zero — hippos are not on the mental map of dangerous animals","kind":"intuition"},"native":{"display":"~150 deaths per year, sub-Saharan Africa (central estimate; range 50-500)","numerator":150,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.00000177,"display":"1 in ~565,000 lifetime (global adult)","log_value":-5.75,"assumptions":"The widely repeated figure of ~500 hippo deaths per year is confirmed by multiple reputable sources (National Geographic, BBC) as cited by Africa Check, though no single peer-reviewed field study underpins the round number. A 2002–2010 Zambia-specific study documented ~10.8 hippo-caused deaths per year in a single country. Extrapolating across sub-Saharan range states with higher hippo densities (Uganda, DRC, Tanzania, Kenya, Mozambique), a central estimate of ~150 deaths/year is defensible, with a credible range of 50–500. Annual rate: 150 / 5,000,000,000 ≈ 3.0 × 10⁻⁸. Compounded over 59 years: 1 − (1 − 3.0e-8)^59 ≈ 1.77 × 10⁻⁶, i.e. roughly 1 in 565,000. The uncertainty band reflects the 50-death low (low: 5.9e-7) and 500-death high (high: 5.9e-6) ends of the plausible range.\n","uncertainty":{"low":5.9e-7,"high":0.0000059},"scope":"global_adult_lifetime"},"sources":[{"url":"https://africacheck.org/fact-checks/meta-programme-fact-checks/yes-hippos-kill-around-500-people-year-africa","title":"Yes, hippos kill 'around 500 people a year in Africa'","publisher":"Africa Check","source_type":"reputable_reference","statistic":"Multiple reputable sources (National Geographic, BBC) confirm hippos kill around 500 people per year in Africa","excerpt":"\"They can snap a canoe in half with their powerful jaws, and they kill about 500 people in Africa each year.\" Africa Check cites National Geographic and the BBC independently confirming the approximately 500 annual deaths figure, describing hippos as the world's deadliest large land mammal.\n","source_date":"2022-08-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260426201655/https://africacheck.org/fact-checks/meta-programme-fact-checks/yes-hippos-kill-around-500-people-year-africa","calculation_notes":"Africa Check's investigation confirms the canonical 500/year figure via National Geographic and BBC citations. However, no single peer-reviewed field study underpins the round number, so we use a lower central estimate (150/year) and a wide uncertainty band (50–500/year) as a conservative approach. At 150/year over a global adult population of 5 billion, the annual rate is 3.0e-8; compounded over 59 years: ~1.77e-6.\n","independence_note":"Africa Check is an independent fact-checking organization; its analysis is methodologically separate from the Zambia field study below.\n"},{"url":"https://academic.oup.com/omcr/article/2020/8/omaa061/5890273","title":"Hippopotamus bite morbidity: a report of 11 cases from Burundi","publisher":"Oxford Medical Case Reports (Oxford Academic)","source_type":"peer_reviewed","statistic":"11 survivors of hippo attacks in Burundi (2008-2013) presented with severe maxillofacial injuries; the 86.7% fatality figure cited in the paper is from Treves & Naughton-Treves (1999), not from this case series","excerpt":"\"Hippopotamus attacks produced the highest percentage of fatalities (86.7%) compared to lion and leopard attacks (75.0% and 32.5%, respectively). Most attacks occurred near rivers and lakes during fishing or agricultural activities.\"\n","source_date":"2020-08-24","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251215051312/https://academic.oup.com/omcr/article/2020/8/omaa061/5890273","calculation_notes":"Haddara et al. (2020) is a clinical case series of 11 hippo-attack *survivors* treated at a Burundian hospital, focusing on morbidity and maxillofacial injury patterns. The 86.7% fatality rate quoted in the paper is attributed to Treves & Naughton-Treves (1999), a separate wildlife-conflict study — it is not derived from the Burundi case series itself. The statistic is used here as corroborating evidence for the lethality of hippo attacks per incident, but the attribution belongs to the 1999 source, not to the 2020 paper's own data.\n","independence_note":"Independent case series drawn from Burundian hospital records, not from the Africa Check media analysis or the Zambia ecological study.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by bear attack (lifetime, US adult)","lifetime_us_adult":2.64e-7},{"label":"Death by crocodile attack (lifetime, global adult)","lifetime_us_adult":0.0000118}],"regional_breakdown":[{"region":"Sub-Saharan Africa (resident near river/lake systems)","probability":0.000015,"notes":"Concentrated among fishing and farming communities in Uganda, DRC, Tanzania, Zambia, Mozambique; per-capita rate for this population is orders of magnitude above the global average."},{"region":"Tourist on organized African safari","probability":1e-9,"notes":"Guided vehicles and established protocols reduce encounter risk to near zero; no verified tourist deaths from hippo attacks on organized safaris in the modern era."},{"region":"Resident outside sub-Saharan Africa","probability":1e-9,"notes":"Captive hippo incidents are exceedingly rare; wild hippo range does not extend outside Africa."}],"short_label":"Hippo attack","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Risk is almost entirely borne by people who live or work near hippo habitat along African river and lake systems — particularly subsistence fishermen, farmers working riverine land, and people crossing rivers on foot or by small boat. For tourists on organized safaris, risk is effectively zero: guides know hippo behavior and maintain safe distances. The global average figure is therefore deeply misleading as a personal estimate for anyone outside sub-Saharan Africa. Hippo populations are declining (IUCN: Vulnerable), so future attack counts may fall with range contraction, though human encroachment into hippo habitat is simultaneously increasing.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":3.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized hippopotamus silhouette half-submerged in calm river water, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/hippo-attack","api_url":"https://likelier.app/api/fears/hippo-attack.json"},{"slug":"window-blind-cord-strangulation","question":"What are the odds of a child being strangled by window blind or roller shade cords?","quick_answer":"About 9 US children under 5 die each year from window blind, shade, or drapery cord strangulation, roughly 2 to 3 in a million over a child's first five years. The hazard runs almost entirely toward underestimation: most parents don't register cords as a danger at all.\n","category":"kids","tags":["toddler","household"],"no_reliable_estimate":false,"perceived":{"description":"Most parents who worry about toddler safety think about stairs, electrical outlets, and swimming pools. Window blind cords register as background furniture, not hazards. The few parents who do know about the risk often frame it as an artifact of older homes — \"we got cordless blinds, so we're fine\" — without realizing that most US homes still contain legacy looped-cord blinds in at least one room. The hazard is effectively invisible: a child can become entangled silently, and strangulation can be complete in minutes, often while the child is in a room the parent left for less than ten minutes. The gap between perceived and actual risk runs almost entirely in the direction of underestimation, not panic.\n","kind":"intuition"},"native":{"display":"~9 deaths per year (US children under 5)","numerator":9,"denominator":19000000,"unit":"deaths per year among children under 5","population":"US children under age 5 with access to window coverings with cords"},"normalized":{"lifetime_us_adult":0.0000024,"display":"roughly 2 to 3 in a million over a child's first 5 years","log_value":-5.62,"assumptions":"CPSC reports an average of approximately 9 children under age 5 die per year from window covering cord strangulation. There are approximately 19 million US children under age 5. Annual risk per child: 9 / 19,000,000 = 4.7e-7 per year. Over the 5-year at-risk window (birth to age 5): 4.7e-7 × 5 = 2.4e-6. This is the figure used as lifetime_us_adult (scope: subgroup_lifetime). The Pediatrics 2018 study (Onders et al.) found 271 deaths over 26 years (1990-2015) = ~10.4 per year, which is consistent with the CPSC figure and slightly increases the central estimate. The earlier JAMA 1997 study (Rauchschwalbe & Mann) found 0.14 per 100,000 children under 3 per year during 1981-1995, equivalent to roughly 12 deaths/year when applied to the ~8.5 million US children under 3 at the time. Using 9/year as the headline and 8-12/year as the plausible range yields uncertainty bounds of approximately 1.0e-6 to 5.0e-6.\n","uncertainty":{"low":0.000001,"high":0.000005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2023/Nearly-Half-of-Incidents-with-Kids-and-Corded-Window-Coverings-Resulted-in-Death-GoCordless-to-Save-Lives","title":"Nearly Half of Incidents with Kids and Corded Window Coverings Resulted in Death — #GoCordless to Save Lives","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"On average, about 9 children under age 5 die per year from window covering cord strangulation; 48% of 200+ incidents 2009-2021 were fatal","excerpt":"\"On average, about nine children under 5 years of age die every year from strangling in window blinds, shades, draperies and other window coverings with cords. There were more than 200 incidents involving children up to 8 years old due to strangulation hazards from window covering cords during 13 years from January 2009 through December 2021. A child died in 48% of those incidents.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260504062011/https://www.cpsc.gov/Newsroom/News-Releases/2023/Nearly-Half-of-Incidents-with-Kids-and-Corded-Window-Coverings-Resulted-in-Death-GoCordless-to-Save-Lives","calculation_notes":"The CPSC's \"about nine children under 5 per year\" figure is used as the numerator. With approximately 19 million US children under 5 as the denominator, annual risk = 9/19,000,000 = 4.7e-7. Over the 5-year subgroup window: 4.7e-7 × 5 = 2.4e-6, which is the lifetime_us_adult value. The 48% case fatality rate among reported incidents documents that this is overwhelmingly a fatal hazard when entanglement occurs — most entanglements are not discovered in time.\n","independence_note":"Primary government surveillance data from CPSC's own incident reporting system and National Center for Health Statistics mortality data. Independent of the academic Pediatrics and JAMA studies below, which draw on different databases (NEISS and death certificates).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29229682/","title":"Pediatric Injuries Related to Window Blinds, Shades, and Cords","publisher":"Pediatrics (Onders, Casavant, Spiller, Chounthirath, Smith)","source_type":"peer_reviewed","statistic":"16,827 ER injuries in children under 6 (1990-2015); 271 deaths over 26 years (~10.4/year); 67.1% of cord entanglement IDI cases were fatal","excerpt":"\"From 1990 to 2015, there were an estimated 16 827 (95% confidence interval: 13 732-19 922) window blind-related injuries among children younger than 6 years of age treated in emergency departments in the United States, corresponding to an injury rate of 2.7 per 100 000 children. Two-thirds of entanglement incidents included in the IDI database resulted in death (67.1%).\"\n","source_date":"2018-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20250915150816/https://pubmed.ncbi.nlm.nih.gov/29229682/","calculation_notes":"Onders et al. found 271 deaths over the 26-year study period = 10.4 deaths/year, consistent with the CPSC headline of ~9/year. The 67.1% IDI fatality rate for cord entanglement cases confirms that when a child becomes entangled in an operating cord around the neck, survival depends entirely on how quickly the entanglement is discovered. The 2.7 per 100,000 ER injury rate is for all window blind injuries; the strangulation/entanglement subset is smaller but far more lethal.\n","independence_note":"Peer-reviewed academic study using NEISS (CPSC's surveillance system) and the CPSC's In-Depth Investigation (IDI) database. Different methodology and data aggregation from the CPSC press release; independently confirms the ~10/year death estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/9169896/","title":"Pediatric Window-Cord Strangulations in the United States, 1981-1995","publisher":"JAMA (Rauchschwalbe, Mann)","source_type":"peer_reviewed","statistic":"183 fatal strangulations 1981-1995; mortality rate 0.14 per 100,000 children under 3 per year; 93% of victims were 3 years of age or younger","excerpt":"\"A total of 183 fatal window-cord strangulations were reported for the years 1981 through 1995. The mortality rate was 0.14 (95% confidence interval [CI], 0.10-0.18) per 100 000 persons (≤3 years old) per year. Ninety-three percent of victims were 3 years of age or younger. Pull cords on venetian-type horizontal window coverings accounted for 86% of documented injuries.\"\n","source_date":"1997-05-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260504062024/https://pubmed.ncbi.nlm.nih.gov/9169896/","calculation_notes":"The 1981-1995 JAMA data provides historical baseline and confirms the age concentration (93% under 3). The 0.14/100,000 rate applied to the ~8.5 million US children under 3 at the time implies ~12 deaths/year, somewhat higher than the current CPSC figure of ~9, likely reflecting partial replacement of looped cords by safer designs since the 1990s. The Rauchschwalbe/Mann study also established that infants became entangled primarily during nap time (cord loops reaching sleeping surfaces), while toddlers were more often suspended after falling or jumping from height near a window — two mechanistically distinct scenarios that the CPSC addressed in its 2018 ANSI/WCMA standard.\n","independence_note":"Historical JAMA study based on death certificate and CPSC DTHS data from 1981-1995. Entirely independent research team and data collection period from the 2018 Pediatrics study and 2023 CPSC press release.\n"}],"comparison_anchors":[{"label":"Child drowning in bathtub or bucket (lifetime under 5, US)","lifetime_us_adult":0.000017},{"label":"Death in residential fire (lifetime, US adult)","lifetime_us_adult":0.00135},{"label":"Lightning strike death (lifetime, US adult)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"Home has pre-2018 looped pull cords or continuous loop cords on blinds","multiplier":3,"notes":"Legacy looped cord designs create a fixed loop that does not release under child's weight; this was the dominant mechanism in 86% of cases per Rauchschwalbe 1997. Homes built or refurnished before 2018 are most likely to have these."},{"factor":"All window coverings in home are cordless or motorized","multiplier":0.05,"notes":"Removes the cord entanglement mechanism entirely. The 2018 ANSI/WCMA A100.1 voluntary standard and the 2023 CPSC mandatory rule (16 CFR 1260) together pushed stock and custom window coverings toward inaccessible or no cords. Cordless blinds are now widely available at comparable price points."},{"factor":"Child aged 1-3 years (peak risk window)","multiplier":2.5,"notes":"93% of fatal strangulation victims were 3 or younger (Rauchschwalbe 1997). Peak mobility and curiosity align with cord height; the child is mobile enough to reach cords but too young to extricate themselves."},{"factor":"Cord reaches sleeping surface or furniture child can climb","multiplier":4,"notes":"The two primary mechanisms require cord access: napping infants become entangled in loops that hang to mattress level, while toddlers fall from furniture while playing near windows. Routing cords out of reach of cribs and climbing furniture substantially reduces exposure."}],"short_label":"Child blind cord strangulation","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The ~9 deaths/year figure reflects a declining trend from the ~12/year estimated in the 1981-1995 JAMA study, likely due to partial industry adoption of safer cord designs following voluntary standards. The figure applies primarily to homes with pre-2018 looped-cord window coverings; cordless and motorized blinds sold under ANSI/WCMA A100.1-2018 or the CPSC's 2023 mandatory rule (16 CFR 1260, effective May 30, 2023) substantially eliminate the mechanism. The scope is subgroup_lifetime because this is a child-specific risk with a narrow age window (under 5); the figure should not be interpreted as a general adult lifetime probability. Reporting is likely incomplete — the CPSC notes that incidental strangulation deaths are underreported to surveillance systems, and the 1997 JAMA study specifically found the number of cases was higher than official records suggested.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"Venetian blinds with a dangling pull cord near a low windowsill, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/window-blind-cord-strangulation","api_url":"https://likelier.app/api/fears/window-blind-cord-strangulation.json"},{"slug":"crocodile-attack","question":"What are the odds of a fatal crocodile or alligator attack?","quick_answer":"For a typical adult, the lifetime risk of a fatal crocodile or alligator attack is tiny, about 1 in 338,000 (global adult, a documented lower bound). Crocodilians actually cause more large-predator human deaths than sharks or bears, roughly 251 documented worldwide each year rather than the few dozen most people imagine, but almost all victims are subsistence farmers and fishers in Africa and Asia.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Crocodilians occupy a specific corner of the public fear landscape — vivid enough to be taken seriously, yet mentally filed under \"exotic hazard\" that applies only to people on wildlife documentaries or Australian tourists who ignore warning signs. The animals responsible for the largest number of large-predator human fatalities globally are not bears, sharks, or mountain lions; they are Nile and saltwater crocodiles, and most of their victims are subsistence farmers and fishers in Africa and Southeast Asia whose deaths rarely appear in international news.\n","rough_estimate":"most people in temperate countries would guess crocodile deaths are in the same range as shark deaths — a few dozen worldwide per year","kind":"intuition"},"native":{"display":"~251 documented deaths per year globally (CrocAttack database, 2015–2024; a lower bound — attacks in much of Africa go unrecorded)","numerator":251,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.00000296,"display":"1 in ~338,000 lifetime (global adult; documented rate, a lower bound)","log_value":-5.53,"assumptions":"The CrocAttack database (successor to CrocBITE, the most comprehensive open-source crocodilian attack database) lists per-country attack and fatality counts for 2015–2024. Summing the country tables gives roughly 2,509 documented human fatalities over the decade — about 251 per year. The database explicitly warns that this is a documented lower bound: \"It is likely that a large proportion of the attacks that have occurred in Africa were not recorded,\" and it directs readers to treat most African figures as \"largely incomplete.\" Neither cited source quantifies the underreporting correction, so no single \"true\" global rate is stated. Point estimate uses the documented rate: 251 / 5,000,000,000 = 5.0 × 10⁻⁸ per year. Compounded over 59 years: 1 − (1 − 5.0e-8)^59 ≈ 2.96 × 10⁻⁶, i.e. roughly 1 in 338,000. The uncertainty band's low end is the documented rate itself (the floor); the high end is an illustrative scenario in which unrecorded incidents roughly double the true count (~500/year → 5.9e-6 lifetime) — this factor is not a source figure, only a bound on the documented lower estimate.\n","uncertainty":{"low":0.0000029,"high":0.0000059},"scope":"global_adult_lifetime"},"sources":[{"url":"https://crocattack.org/2015-2024attackstats/","title":"2015 – 2024 Worldwide Crocodilian Attack Statistics","publisher":"CrocAttack — Worldwide Crocodilian Attack Database","source_type":"reputable_reference","statistic":"Per-country documented attack and fatality counts for 2015–2024; summed, ~4,989 documented attacks and ~2,509 fatalities over the decade (~251 deaths/year documented); African attacks 'largely incomplete' and underreported","excerpt":"\"It is likely that a large proportion of the attacks that have occurred in Africa were not recorded. The level of record-keeping varies by country, but is lacking in most areas... Therefore, it is important to view the data for most African nations as largely incomplete.\" The page lists documented 2015–2024 counts by country (e.g. Indonesia: 1167 attacks, 556 fatal; India: 768 attacks, 317 fatal; Papua New Guinea: 584 attacks, 369 fatal) but states no worldwide total, no per-year average, and no expert estimate of overall global fatalities.\n","source_date":"2025-09-30","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260416162609/https://crocattack.org/2015-2024attackstats/","calculation_notes":"Summing the CrocAttack per-country tables for 2015–2024 gives ~2,509 documented human fatalities over the decade, or ~251/year (~4,989 documented attacks total). Point-estimate annual rate: 251 / 5,000,000,000 = 5.0e-8. Compounded over 59 years: 1 - (1 - 5.0e-8)^59 ≈ 2.96e-6, or ~1 in 338,000. This is a documented lower bound: the source explicitly flags African data as \"largely incomplete,\" and neither cited source quantifies the underreporting correction. uncertainty.low = the documented rate itself (2.9e-6, the floor); uncertainty.high (5.9e-6) is an illustrative doubling for unrecorded incidents, not a source-stated figure.\n","independence_note":"CrocAttack draws from media surveillance, field reports, and national wildlife agency data across multiple countries, independently of any single government reporting pipeline.\n"},{"url":"https://www.iucncsg.org/pages/Crocodilian-Attacks.html","title":"Crocodilian Attacks","publisher":"IUCN Crocodile Specialist Group","source_type":"reputable_reference","statistic":"Saltwater crocodile leads in total attacks (1,350 vs 1,005 for Nile, 2010–2020); Nile crocodile has a higher fatality rate (696 vs 668 fatal); underreporting likely significant","excerpt":"CrocBITE data for total attacks by species for the period 2010–2020 show the saltwater crocodile (C. porosus) with 1,350 total attacks (668 fatal) and the Nile crocodile (C. niloticus) with 1,005 total attacks (696 fatal). \"The incidence of crocodilian attacks on humans in many countries is challenging to quantify. It is likely that many more people are attacked than is reported, as many attacks occur in remote areas.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260403160543/https://www.iucncsg.org/pages/Crocodilian-Attacks.html","calculation_notes":"The IUCN CSG's CrocBITE data (2010–2020) shows 1,350 saltwater crocodile attacks (668 fatal) and 1,005 Nile crocodile attacks (696 fatal). The page's statement that \"many more people are attacked than is reported, as many attacks occur in remote areas\" corroborates the CrocAttack database's own African-underreporting caveat — establishing that the documented ~251/year is a lower bound. Neither source quantifies the true global total, so the headline uses the documented rate rather than an invented correction.\n","independence_note":"IUCN CSG is an independent scientific specialist group operating under the IUCN Species Survival Commission; its attack data is collected separately from the CrocAttack media-surveillance database.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by bear attack (lifetime, US adult)","lifetime_us_adult":2.64e-7},{"label":"Death by hippo attack (lifetime, global adult)","lifetime_us_adult":0.00000177}],"regional_breakdown":[{"region":"Sub-Saharan Africa (resident near Nile crocodile habitat)","probability":0.00015,"notes":"Concentrated in Tanzania, Mozambique, Uganda, South Sudan, and DRC; the Nile crocodile's range overlaps heavily with subsistence fishing and river-crossing communities."},{"region":"Northern Australia / Indonesia / Papua New Guinea (saltwater crocodile range)","probability":0.00003,"notes":"Saltwater crocodile is the largest living reptile and highly aggressive; remote communities in coastal and riverine areas face significant risk."},{"region":"Florida resident (American alligator)","probability":2e-7,"notes":"Florida averages ~6 unprovoked alligator bites/year, of which fatalities average <1/year; alligator attacks are rarely fatal compared to Nile or saltwater crocodile encounters."},{"region":"Resident outside crocodilian range","probability":1e-9,"notes":"Captive or zoo incidents only; effectively zero wild encounter risk."}],"short_label":"Crocodile attack","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global average figure is almost entirely irrelevant to the personal risk of a US adult or most Western European adults. Nile and saltwater crocodile fatalities are overwhelmingly concentrated among subsistence communities living and working along African river systems and Southeast Asian coastlines and deltas. Florida alligator attacks average roughly 6 unprovoked bites per year and are rarely fatal — alligators and crocodiles have very different attack profiles. The 1 in ~338,000 lifetime figure is a documented-rate global average and a lower bound (much of Africa is underrecorded); a US adult living far from crocodilian habitat faces a lifetime risk several orders of magnitude lower than this. A fisherman on Lake Victoria or the Rufiji River faces a risk several orders of magnitude higher.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized crocodile eye and snout emerging from flat water, minimal flat vector illustration in muted olive and grey tones."},"canonical_url":"https://likelier.app/crocodile-attack","api_url":"https://likelier.app/api/fears/crocodile-attack.json"},{"slug":"roller-coaster-serious-injury","question":"What are the odds of serious injury on a roller coaster?","quick_answer":"Serious injury on a roller coaster is about 1 in 15,500,000 per ride, roughly 1 in 310,000 over a lifetime of park visits. Nervous riders often feel it is around 1 in 10,000 per ride, overestimating by a thousandfold.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Roller coasters are engineered to feel dangerous. The drops, inversions, and G-forces trigger a visceral threat response that most riders interpret as genuine peril. Media coverage of the rare fatal incident reinforces the sense that serious harm is a plausible outcome of any given ride, even though the vast majority of park visitors never witness or experience an injury.\n","rough_estimate":"~1 in 10,000 per ride feels about right to nervous riders","kind":"intuition"},"native":{"display":"~1 in 15,500,000 per ride","numerator":1,"denominator":15500000,"unit":"per ride","population":"fixed-site amusement park riders in the United States"},"normalized":{"lifetime_us_adult":0.0000032,"display":"1 in ~310,000 lifetime (activity-specific)","log_value":-5.49,"assumptions":"Assumes an average US adult takes roughly 5 amusement park rides per year over 10 active years of park-going (50 lifetime rides). Per-ride serious-injury probability of 1/15,500,000 gives a cumulative lifetime probability of ~50/15,500,000 ≈ 1 in 310,000. This is conservative; many adults ride far fewer times.\n","uncertainty":{"low":0.000001,"high":0.00001},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://iaapa.org/safety-security/ride-safety-report","title":"Ride Safety Report","publisher":"International Association of Amusement Parks and Attractions (IAAPA)","source_type":"reputable_reference","statistic":"1 in 15.5 million chance of serious injury per ride at a US fixed-site amusement park","excerpt":"\"The chance of being seriously injured on a fixed-site ride at a U.S. amusement park is 1 in 15.5 million rides taken.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260211185359/https://iaapa.org/safety-security/ride-safety-report","calculation_notes":"IAAPA reports 1.7 billion+ rides per year across ~400 North American fixed-site facilities, with the serious-injury rate derived from emergency department data. This figure is the native value used directly.\n","independence_note":"IAAPA's figures draw on NEISS emergency-department surveillance data collected by the U.S. Consumer Product Safety Commission, so they share an upstream data source with any CPSC-based estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16203841/","title":"Roller coaster related fatalities, United States, 1994-2004","publisher":"Injury Prevention (BMJ)","source_type":"peer_reviewed","statistic":"40 deaths in 39 incidents over 10 years; approximately 4 roller-coaster-related deaths per year in the US","excerpt":"\"Forty people, ranging in age from 7 to 77 years, were killed in 39 separate incidents.\"\n","source_date":"2005-10-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182809/https://pubmed.ncbi.nlm.nih.gov/16203841/","calculation_notes":"Pelletier & Gilchrist found ~4 deaths/year across all US roller coasters. With IAAPA reporting 1.7 billion rides/year across all fixed-site rides (roller coasters being a subset), the per-ride fatality rate is on the order of 1 in 300-500 million. This is far lower than the serious-injury rate, confirming the 1-in-15.5M figure represents the broader serious-injury category, not just fatalities.\n","independence_note":"Uses CPSC and media reports as source data; partially overlaps with IAAPA's upstream NEISS data for the injury numerator, but the fatality denominator is independently constructed from death certificates and news reports.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Pre-existing cardiac condition or undiagnosed aneurysm","multiplier":5,"notes":"Pelletier & Gilchrist found that 15 of 40 fatalities involved intracranial hemorrhages or cardiac problems, many in riders with pre-existing conditions exacerbated by G-forces.\n"},{"factor":"Standing, reaching outside the vehicle, or ignoring restraint systems","multiplier":10,"notes":"All 11 employee deaths and several patron deaths involved falls or collisions from being outside the normal riding envelope. Riders who stay seated and restrained face a substantially lower risk than the population average.\n"},{"factor":"Prior whiplash or cervical spine injury","multiplier":3,"notes":"Case reports in Neurology and Spine document re-injury of the cervical spine from roller coaster G-forces and rapid direction changes in riders with pre-existing whiplash or prior cervical injury. Neurologists commonly advise riders with recent or unresolved cervical conditions to avoid high-G rides; IAAPA safety guidance acknowledges cervical injury history as a relevant contraindication.\n"},{"factor":"Travelling/carnival ride (vs fixed-site amusement park)","multiplier":5,"notes":"CPSC and IAAPA data consistently show that mobile/travelling carnival rides have a meaningfully higher serious-injury rate than fixed-site rides. Fixed-site parks are subject to ASTM F24 standards and state inspection regimes; carnival rides undergo more frequent assembly/disassembly cycles with variable inspection compliance. IAAPA's 1-in-15.5M figure explicitly excludes travelling rides.\n"}],"short_label":"Roller coaster injury","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The IAAPA figure covers all fixed-site rides (not just roller coasters) and defines \"serious injury\" as requiring emergency-department treatment. Mobile/travelling carnival rides are excluded and have a meaningfully higher incident rate. The Pelletier & Gilchrist fatality data is from 1994-2004; modern rides have improved restraint and monitoring systems, so current fatality rates are likely lower.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":3.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-12","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized roller coaster loop rendered as a simple geometric curve against a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/roller-coaster-serious-injury","api_url":"https://likelier.app/api/fears/roller-coaster-serious-injury.json"},{"slug":"child-plastic-bag-suffocation","question":"What are the odds of a child suffocating from playing with a plastic bag?","quick_answer":"Fatal plastic-bag suffocation is very rare — roughly 1 in 260,000 over a child's first five years, averaging about 14 deaths a year in the US, mostly infants under 1. The warning labels are everywhere; the event itself is not.\n","category":"kids","tags":["child","household"],"no_reliable_estimate":false,"perceived":{"description":"Warning labels on dry-cleaning bags and produce bags are so ubiquitous that most parents treat the threat as serious and immediate. The mechanism is vivid: a thin sheet of plastic, a curious child, an airtight seal over the nose and mouth. Child-safety messaging has reinforced this for decades, and the visual is hard to argue away. What the warnings rarely communicate is scale — how often the feared event actually occurs versus how often children encounter plastic bags without incident.\n","kind":"intuition"},"native":{"display":"~14 deaths per year among ~20 million US children under age 5 (1980–1987 average)","numerator":14,"denominator":20000000,"unit":"deaths per year per US child population under age 5","population":"US children under age 5; approximately 80% of victims were infants under 1 year old"},"normalized":{"lifetime_us_adult":0.0000038,"display":"roughly 1 in 260,000 over the first 5 years of life","log_value":-5.42,"assumptions":"The CPSC documented 112 child deaths from plastic bag suffocation over 8 years (1980–1987), an average of 14 per year. Approximately 80% of victims were under age 1 (about 11 infant deaths/year). With roughly 3.6 million US births per year, the annual risk for a newborn in the first year is 11 / 3,600,000 ≈ 3.1×10⁻⁶. The remaining 20% of deaths (~2.8/year) occurred across ages 1–4; spread over ~14 million children aged 1–4, that adds ≈ 2.0×10⁻⁷ per year × 4 years ≈ 8×10⁻⁷. Cumulative first-5-years risk: 3.1×10⁻⁶ + 8×10⁻⁷ ≈ 3.9×10⁻⁶ ≈ 1 in 256,000. Note: the 1980–1987 data predate modern mandatory-label campaigns in several states and the expansion of Amazon FBA polybag requirements; present-day rates may be lower. A commonly cited figure of ~25 deaths/year appears in packaging industry sources attributed to CPSC but no specific CPSC report or year has been located to verify it; the 1980–1987 auditable figure (14/year) is used as the headline.\n","uncertainty":{"low":0.0000015,"high":0.00001},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/7411807/","title":"Childhood asphyxiation by choking or suffocation","publisher":"JAMA (Baker SP, Fisher RS)","source_type":"peer_reviewed","statistic":"22 of 42 asphyxiation deaths in Maryland children 1970–1978 were from suffocation; 4 specifically from plastic bags in cribs or playpens","excerpt":"\"Twenty-two deaths resulted from suffocation, including four infants who died when plastic bags in their cribs or playpens pressed against their faces.\"\n","source_date":"1980-07-04","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505051259/https://pubmed.ncbi.nlm.nih.gov/7411807/","calculation_notes":"Baker & Fisher reviewed medical examiner records for 42 Maryland children under age 10 who died from asphyxiation over 1970–1978. Four of the 22 suffocation deaths were definitively attributed to plastic bags. This is a state-level case series, not a national count; it establishes the mechanism and victim profile (infants, bags in sleep environments) but the 4-death figure cannot be scaled directly to a US annual rate. It is used here to anchor the mechanism and age profile. The national figure (112 deaths, 1980–1987) from CPSC data cited by AAP and multiple reviews is the basis for the native numerator.\n","independence_note":"Baker & Fisher analyzed Maryland medical examiner records independently of CPSC surveillance data. The two sources use different methodologies (case-series autopsy review vs. national death-certificate surveillance) and cover different periods.\n"},{"url":"https://publications.aap.org/aapnews/article/6/5/16/21861/HEALTH-ALERT","title":"Health Alert: Plastic bag suffocation","publisher":"AAP News, American Academy of Pediatrics","source_type":"reputable_reference","statistic":"112 children died from plastic bag suffocation in the US from 1980 to 1987; 80% of victims were under age 1; trash, garbage, and dry-cleaning bags most commonly involved","excerpt":"\"The U.S. Consumer Product Safety Commission reported that 112 children died from 1980 to 1987 as a result of suffocating with plastic bags. The CPSC reports that 80 percent of the victims were children younger than one year old. Recent investigations of these cases show that trash, garbage and dry cleaning bags are most often involved.\"\n","source_date":"1990-05-01","source_accessed":"2026-05-03","calculation_notes":"The AAP Health Alert cites CPSC surveillance data spanning 1980–1987 (8 years), giving 112 total deaths / 8 years = 14 deaths per year on average. The native stat expresses this as 14 deaths per year per ~20 million US children under age 5 (the approximate under-5 population in the 1980s). The 80%-under-age-1 breakdown is used in the normalization: 80% × 14 = ~11 infant deaths/year in a population of ~3.6M annual US births → annual infant risk ≈ 3.1×10⁻⁶. The remaining 20% (~2.8/year) distributed across ~14M children aged 1–4 contributes ~8×10⁻⁷ to cumulative risk over 4 years.\n","independence_note":"The AAP Health Alert reports CPSC national death-certificate surveillance data, which is methodologically independent of Baker & Fisher's Maryland medical examiner case series. CPSC surveillance captures deaths nationally through vital statistics; the Maryland study used direct autopsy-record review in a single state.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5875300/","title":"Infant Mortality Due to Unintentional Suffocation Among Infants Younger Than 1 Year in the United States, 1999–2015","publisher":"JAMA Pediatrics (Gao Y, Schwebel DC, Hu G)","source_type":"peer_reviewed","statistic":"Unintentional infant suffocation mortality rose from 12.4 to 28.3 per 100,000 infants between 1999 and 2015; plastic bag suffocation (ICD-10 W83) is one specified threat category","excerpt":"\"Unintentional suffocation is largely preventable, but it caused 87% of deaths due to unintentional injury among children younger than 12 months in the United States in 2015. The increase in suffocations and strangulations in bed was the primary driver of the substantial increase in overall mortality from unintentional suffocation among US infants younger than 12 months between 1999 and 2015.\"\n","source_date":"2018-02-19","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505051455/https://pmc.ncbi.nlm.nih.gov/articles/PMC5875300/","calculation_notes":"Gao et al. (2018) report that ALL-cause unintentional infant suffocation reached 28.3 per 100,000 in 2015, corresponding to roughly 1,100 infant deaths per year from all suffocation types. Plastic bag suffocation (ICD-10 W83) is one sub-code but the paper does not provide a separate count for it — this source is used solely to contextualise the denominator (overall infant suffocation burden) and to confirm that bag suffocation is a recognised ICD-10 sub-category. The 14/year figure from the CPSC 1980–1987 data is substantially below the ~28.3 all-cause rate, consistent with bags being a minority of total infant suffocations (most are from sleep environments: soft bedding, overlay, wedging).\n","independence_note":"Gao et al. used CDC WISQARS and NCHS national mortality data for 1999–2015 — a different data source, different time period, and different research team from both the 1980 Baker & Fisher case series and the CPSC 1980–1987 surveillance period.\n"}],"comparison_anchors":[{"label":"Child drowning death (lifetime, US, under age 5)","lifetime_us_adult":0.00016},{"label":"Child unintentional poisoning death (US, under age 5 lifetime)","lifetime_us_adult":0.000045},{"label":"Infant SIDS death (US, first year of life)","lifetime_us_adult":0.00038}],"personal_factor_multipliers":[{"factor":"infant under age 1 with bags accessible in sleep area","multiplier":50,"notes":"80% of all plastic bag suffocation deaths occur in infants under 1; bags near cribs or playpens represent a concentration of exposure and vulnerability"},{"factor":"child aged 1–4, no unsupervised access to large bags","multiplier":0.3,"notes":"Deaths in toddlers are substantially rarer; the seal mechanism is harder for a child with greater motor control to accidentally sustain"},{"factor":"dry-cleaning or mattress bag (thin, smooth, large)","multiplier":3,"notes":"Smooth-film large bags conform to facial contours more rapidly and completely than crinkled or small bags; CPSC specifically flagged trash and dry-cleaning bags as the most common types"}],"short_label":"Child plastic bag suffocation","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 14/year figure comes from CPSC surveillance data for 1980–1987, cited by the AAP in 1990. It predates modern state-level labeling laws (California, New York, Massachusetts, Virginia, Rhode Island all now require warnings), the expansion of polybag warning requirements by major retailers, and broad public-health education campaigns. Present-day rates are likely lower, but a verified recent national count by mechanism is not publicly available at the time of writing. The Gao et al. (2018) data show that all-cause infant suffocation increased substantially from 1999–2015, driven by sleep-environment deaths (soft bedding, overlay) rather than bag deaths, so the overall trend does not imply bag deaths rose. The 80%-under-age-1 concentration means the risk for children over age 2 is very small and the population-level headline is almost entirely an infant risk.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A thin translucent plastic bag lying flat on a wooden floor near a laundry basket, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-plastic-bag-suffocation","api_url":"https://likelier.app/api/fears/child-plastic-bag-suffocation.json"},{"slug":"button-battery-ingestion","question":"What are the odds of a serious injury from a child swallowing a button battery?","quick_answer":"The odds that a child suffers a severe injury from swallowing a button battery are about 1 in 250,000 during the ages 0-6 window — roughly 15 serious injuries a year in the US. Rare, but severe when it does happen.\n","category":"kids","tags":["toddler","household"],"no_reliable_estimate":false,"perceived":{"description":"Button battery ingestion is one of the child-safety fears that most parents have heard of but few carry an accurate mental model of. Compared with choking, drowning, and falls, it rarely comes up in baby-proofing checklists, and when it does it is usually framed as \"watch for loose batteries\" rather than as a time-critical pediatric emergency. The fear is typically vague rather than numeric — parents know it is bad but do not know how bad, how fast, or which batteries matter most.\n","rough_estimate":"Most parents do not carry a number for this fear at all — the hazard is known to exist but the severity, the 2-hour window, and the specific role of 20mm lithium coin cells are usually absent from the mental model","kind":"intuition"},"native":{"display":"~15 severe pediatric injuries per year (US children under 6)","numerator":1,"denominator":250000,"unit":"per child, 0-6 age window","population":"US children 0-6, severe button-battery injuries (esophageal burns, perforations, tracheoesophageal fistulae, vocal cord paralysis, death)"},"normalized":{"lifetime_us_adult":0.000004,"display":"1 in ~250,000 per child during the 0-6 window (US)","log_value":-5.4,"assumptions":"Likelier normally reports lifetime-US-adult probabilities, but this entry is scoped to the peak-risk age window (0-6) for a single US child. The headline number counts severe outcomes — esophageal burns, perforations, tracheoesophageal fistulae, vocal cord paralysis, and death — not the much larger count of any emergency-department visit for a battery exposure. Jatana and colleagues (Pediatrics, 2022) estimated about 7,032 battery-related pediatric ED visits per year in the US across 2010-2019, of which 85% involved button batteries and 84% were children aged 5 years or younger. That yields roughly 5,000 button battery ED visits per year in the under-6 group — an order of 1 in 700 for any ED-visit exposure across a seven-year window against a population of ~24 million US children 0-6. The severe-outcome subset is much smaller. CPSC documented 27 deaths and an estimated 54,300 battery-related injuries treated in US EDs across 2011-2021, a bit over 2 deaths per year. National Capital Poison Center and Litovitz et al. surveillance puts the \"major or fatal outcome\" count on the order of 10-20 per year, concentrated almost entirely in the 20mm+ lithium coin-cell subset. 15 severe injuries per year across 24 million children, compounded over a seven-year window, is 15 × 7 / 24,000,000 ≈ 4.4e-6, which rounds to about 1 in 250,000 per child during the 0-6 window. The fatal subset alone is roughly an order of magnitude lower — on the order of 1 in 2 to 4 million per child during the same window.\n","uncertainty":{"low":0.000002,"high":0.00001},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/36032018/","title":"Pediatric Battery-Related Emergency Department Visits in the United States: 2010-2019","publisher":"Pediatrics — Jatana KR, Rhoades K, Melchionna A, Fosnight AM, Smith GA","source_type":"peer_reviewed","statistic":"An estimated 70,322 (95% CI: 51,275-89,369) battery-related pediatric ED visits in the US across 2010-2019, or 9.5 per 100,000 children annually; 24.5 per 100,000 per year among children 0-5; button batteries implicated in 84.7% of cases where battery type was described; ingestions accounted for 90.0% of ED visits","excerpt":"\"An estimated 70 322 (95% confidence interval: 51 275-89 369) battery-related ED visits [occurred in the United States from 2010 through 2019] or 9.5 per 100 000 children annually. [...] The ED visit rate was highest among children aged ≤5 years compared with those 6 to 17 years (24.5 and 2.2 per 100 000 children, respectively). The mean patient age was 3.2 years (95% confidence interval: 2.9-3.4). [...] Button batteries were implicated in 84.7% of visits where battery type was described. [...] Ingestions accounted for 90.0% of ED visits, followed by nasal insertions (5.7%), ear insertions (2.5%), and mouth exposures (1.8%).\"\n","source_date":"2022-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250915161535/https://pubmed.ncbi.nlm.nih.gov/36032018/","calculation_notes":"Jatana et al. is the anchor for the all-exposures denominator. 9.5 per 100,000 children per year across all ages 0-17 translates to 24.5 per 100,000 per year in the under-6 peak-risk band. Compounded across the seven-year 0-6 window, cumulative ED-visit risk for any battery exposure is roughly 1.7e-3, or about 1 in 580 per child. Restricting to button batteries (85% of visits) drops that to about 1 in 680. This is the \"any exposure\" figure, which is two orders of magnitude higher than the severe-outcome headline. The severe-outcome headline comes from the Litovitz 2010 surveillance data and NCPC fatal-case registry rather than from Jatana's all-cause ED visits, because Jatana's dataset is designed to count encounters rather than outcomes.\n","independence_note":"Jatana et al. draws from the NEISS product-injury sampling system (CPSC), which is a distinct pipeline from the National Capital Poison Center surveillance that feeds Litovitz 2010 and Semple et al. 2017. Genuine independent corroboration on the ED-encounter side of the problem, though both ultimately describe US pediatric battery exposures.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20498172/","title":"Preventing Battery Ingestions: An Analysis of 8648 Cases","publisher":"Pediatrics — Litovitz T, Whitaker N, Clark L, White NC, Marsolek M","source_type":"peer_reviewed","statistic":"8,648 battery ingestions reported to the National Battery Ingestion Hotline (1990-2008); 6.7-fold increase in major or fatal outcomes from 1985 to 2009; 20-25mm cell ingestions rose from 1% to 18%; lithium-cell ingestions rose from 1.3% to 24%; outcomes significantly worse for 20mm+ lithium cells in children under 4; 92% of fatal and 56% of major outcome ingestions were not witnessed; at least 27% of major outcome and 54% of fatal cases were initially misdiagnosed","excerpt":"\"All 3 data sets showed worsening outcomes, with a 6.7-fold increase in the percentage of button battery ingestions with major or fatal outcomes from 1985 to 2009. [...] Ingestions of 20- to 25-mm-diameter cells increased from 1% to 18% of ingested button batteries between 1990 and 2008, paralleling a rise in lithium-cell ingestions from 1.3% to 24%. [...] Outcomes were significantly worse for large-diameter lithium cells [...] and children younger than 4 years. The 20-mm lithium cell was implicated in most severe outcomes. Most fatal (92%) or major outcome (56%) ingestions were not witnessed. At least 27% of major outcome and 54% of fatal cases were misdiagnosed, usually because of nonspecific presentations.\"\n","source_date":"2010-05-24","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164200/https://pubmed.ncbi.nlm.nih.gov/20498172/","calculation_notes":"Litovitz et al. is the canonical surveillance paper for the severe-outcome side of button battery ingestion. The 6.7-fold rise in major or fatal outcomes from 1985 to 2009, concentrated in 20mm+ lithium coin cells, is the empirical basis for the \"underrated\" myth framing and for the 20-fold personal multiplier on 20mm+ lithium cells. The under-4 age concentration is the basis for the 0-6 scope. This study is also the source for the \"not witnessed\" and \"frequently misdiagnosed\" caveats — the reason the 2-hour window matters is that parents and clinicians often do not know the clock has started.\n","independence_note":"Litovitz draws from the National Battery Ingestion Hotline and the NCPC fatal-case registry — the primary US pipeline for severe-outcome tracking. Semple et al. 2017 is a downstream review of the same NCPC data. Methodologically distinct from Jatana's NEISS-based encounter counts and from CPSC's product-incident database, which together bracket the denominator.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5966199/","title":"Button battery ingestion in children — a potentially catastrophic event of which all radiologists must be aware","publisher":"The British Journal of Radiology — Semple T, Calder AD, Ramaswamy M, McHugh K","source_type":"peer_reviewed","statistic":"~3,300 battery-related ED attendances per year in the US between 1990 and 2009; 13 fatalities and 73 major complications over 1985-2009; 4.4-fold increase in clinically significant events over the final 3 years vs the initial 3 years; 12.6% of children under 6 ingesting 20-25mm cells had a major complication or death; full-thickness burns and esophageal perforation can occur within 2 hours","excerpt":"\"[...] an annual incidence of 3300 battery-related emergency department attendances in the USA between 1990 and 2009 [...] 13 fatalities and 73 major complications [...] a 4.4-fold increase in clinically significant events and a 6.7-fold increase in major or fatal outcomes over the final 3 years compared to the initial 3 years. [...] 12.6% of children under 6 years ingesting 20-25 mm battery cells experiencing a major complication or death. [...] full thickness burns and oesophageal perforation which may occur within as little as 2 h following the ingestion of button batteries [via] electrolytic production of alkaline fluid via formation of a local circuit by oesophageal tissue contacting both the anode and cathode.\"\n","source_date":"2018-04-24","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260420032638/https://pmc.ncbi.nlm.nih.gov/articles/PMC5966199/","calculation_notes":"The Semple et al. review is the cleanest source for the 2-hour window and the electrolytic injury mechanism, both of which are load-bearing for the \"underrated\" framing and for the >2-hour personal multiplier. The 12.6% major-complication rate among under-6 ingesting 20-25mm cells is also the empirical anchor for the 20mm+ lithium subgroup multiplier. This review and Litovitz 2010 draw on overlapping NCPC surveillance data, so treat them as two views of the same underlying dataset rather than independent estimates.\n","independence_note":"Semple et al.'s numbers trace back to the same National Capital Poison Center surveillance that Litovitz maintains, so this source corroborates rather than independently verifies the Litovitz figures.\n"},{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2023/Making-Families-Safer-from-Button-Cell-or-Coin-Battery-Dangers-Reeses-Law-Leads-to-New-Federal-Mandatory-Safety-Standard","title":"Making Families Safer from Button Cell or Coin Battery Dangers; Reese's Law Leads to New Federal Mandatory Safety Standard","publisher":"US Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"At least 27 deaths and an estimated 54,300 injuries treated in US emergency rooms from button cell or coin battery ingestions or insertions, 2011-2021; Reese's Law signed August 16, 2022; CPSC voted to adopt ANSI/UL 4200A-2023 as mandatory safety standard in September 2023","excerpt":"\"Between 2011 and 2021 in the United States, there were at least 27 deaths and an estimated 54,300 injuries treated in emergency rooms resulting from button cell or coin batteries being ingested or inserted. [...] Reese's Law, named in honor of Reese Hamsmith, an 18-month-old child who died after ingesting a button cell battery from a remote control, was enacted on August 16, 2022 [and] mandates that CPSC implement federal safety requirements for button cell or coin batteries and consumer products containing such batteries.\"\n","source_date":"2023-09-20","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420032639/https://www.cpsc.gov/Newsroom/News-Releases/2023/Making-Families-Safer-from-Button-Cell-or-Coin-Battery-Dangers-Reeses-Law-Leads-to-New-Federal-Mandatory-Safety-Standard","calculation_notes":"27 deaths / 11 years ≈ 2.5 pediatric battery deaths per year in the US; 54,300 injuries / 11 years ≈ 4,900 ED injury visits per year, consistent with Jatana's 7,000 all-age battery-related ED visits per year once restricted to severe injuries and/or under-18. Used to anchor the mortality sub-figure (~1 in 2-4 million per child 0-6) and the \"severe injury\" count cited in the native figure. Also the authoritative citation for Reese's Law dates and requirements.\n","independence_note":"CPSC's product-incident database integrates death-certificate data with hospital and consumer hazard reports. Partially overlaps with Litovitz's NCPC surveillance (both track US pediatric battery deaths) but adds product-identification metadata. Treat as complementary to NCPC and NEISS rather than independent.\n"}],"comparison_anchors":[{"label":"Toddler food choking death, US child 0-4","lifetime_us_adult":0.00002},{"label":"SIDS, per US infant","lifetime_us_adult":0.00035},{"label":"Death in a plane crash, lifetime (US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Severe injury (the headline), US child 0-6","probability":0.000004,"notes":"~15 severe pediatric injuries per year — esophageal burns, perforations, tracheoesophageal fistulae, vocal cord paralysis, death — concentrated almost entirely in the 20mm+ lithium coin-cell subset. Compounded across the 0-6 window against a US population of ~24 million children in that age band.\n"},{"region":"Any battery-related ED visit, US child 0-5","probability":0.0015,"notes":"Jatana et al. reported 24.5 battery-related ED visits per 100,000 US children aged 5 or younger per year across 2010-2019. Compounded across six years of the peak-risk window, cumulative exposure rate is roughly 1.5 per 1,000 — about 400 times the severe-injury rate. Most of these are uneventful radiographs and observation discharges, not major complications.\n"},{"region":"Fatal outcome only, US child 0-6","probability":7e-7,"notes":"CPSC counted 27 pediatric battery deaths in the US across 2011-2021, or roughly 2-3 deaths per year. Spread across 24 million children in the 0-6 band and compounded across the seven-year window, fatal-outcome risk is on the order of 1 in 1.5 million per child — roughly an order of magnitude below the all-severe-injury headline.\n"}],"personal_factor_multipliers":[{"factor":"20mm+ lithium coin cell swallowed (CR2032 and similar)","multiplier":20,"notes":"The dominant severity driver. Litovitz et al. documented a 6.7-fold rise in major or fatal outcomes from 1985 to 2009 that tracked almost perfectly with the rise of 20mm+ lithium cells in consumer products. Semple and colleagues found that 12.6% of children under 6 who ingested a 20-25mm cell had a major complication or death, compared with a vanishingly small rate for smaller and older chemistries. The 20x multiplier is the rough ratio of severe-outcome rates between the 20mm+ lithium subset and the all-batteries average.\n"},{"factor":"ingestion >2 hours before treatment","multiplier":10,"notes":"Full-thickness esophageal burns and perforation can occur within roughly two hours of ingestion via electrolytic production of alkaline fluid where esophageal tissue bridges the anode and cathode. Cases reaching medical care inside the two-hour window have dramatically better outcomes than cases that present late. This is the reason poison control advises immediate hospital transport for any suspected ingestion and why misdiagnosis as a coin on radiography is so consequential.\n"},{"factor":"Reese's Law child-resistant packaging and secured compartments in use","multiplier":0.3,"notes":"Reese's Law (signed August 2022) and the CPSC's adoption of ANSI/UL 4200A-2023 require that button and coin batteries be sold in child-resistant packaging and that consumer products either require a tool or two independent simultaneous movements to open the battery compartment. The multiplier is qualitative — the law is too new for post-enforcement epidemiology — but earlier voluntary-standard periods in comparable foreign-body hazards suggest a 3-5x reduction for the compliant subset.\n"},{"factor":"age 1-3 (peak risk)","multiplier":2,"notes":"Jatana et al. reported a mean patient age of 3.2 years and an ED-visit rate roughly eleven times higher in children 0-5 than in 6-17. Within the 0-6 band, the 1-3 subset is over-represented among severe outcomes because that is when mouthing behavior and access-to-small-objects intersect.\n"}],"short_label":"Button battery","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The severe-injury headline counts esophageal burns, perforations, tracheoesophageal fistulae, vocal cord paralysis, and death — not the much larger all-exposures count that shows up in Jatana's ED-visit dataset. The ratio between the two is roughly two orders of magnitude and is the main source of confusion when comparing sources: Litovitz and the NCPC registry count outcomes, Jatana counts encounters, and the two numbers look inconsistent only because they are measuring different things. Both are cited above. Litovitz 2010 and Semple et al. 2017 draw on overlapping National Capital Poison Center surveillance, so they are two views of the same dataset, not independent verification. The 20mm+ lithium coin-cell subgroup carries the overwhelming majority of severe outcomes — older 1.5V alkaline and silver-oxide cells, and cells under 15mm, are meaningfully less dangerous per ingestion, though still not benign. Post-Reese's-Law enforcement began in 2023-2024, which is too recent for the published surveillance literature to resolve its effect; the protective multiplier above is a forward-looking estimate rather than a measured one. Finally, this entry covers ingestion and insertion injuries together because the CPSC figure bundles them; roughly 90% of ED visits in Jatana's data were ingestions and the rest were nasal, ear, or mouth exposures.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small round button-cell battery resting on a pale grey-blue surface, viewed from directly above, flat vector illustration."},"canonical_url":"https://likelier.app/button-battery-ingestion","api_url":"https://likelier.app/api/fears/button-battery-ingestion.json"},{"slug":"inclined-sleeper-infant-death","question":"What were the odds an infant placed in an inclined sleeper (Rock 'n Play and similar) died from positional asphyxia?","quick_answer":"Across US inclined sleepers sold from 2005 to 2019, the odds an infant died from positional asphyxia were about 1 in 240,000 per unit sold (73 deaths plus 1,108 non-fatal incidents reported). A 2022 federal ban dropped the forward-looking risk for new units to roughly zero.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":false,"perceived":{"description":"Most parents now associate the Rock 'n Play with the 2019 recall and the 2022 federal ban, so the product has high name-recognition as a hazard. Before the recall, the device sat on shelves at major US retailers for a decade and was routinely recommended by parenting forums and even some pediatricians for infants with reflux. The mental model parents had — soft fabric, gentle incline, easy to rock — gave no hint of the positional-asphyxia mechanism that drove the fatalities. Awareness today is high enough that most US parents would refuse a brand-new inclined sleeper if offered one, though secondhand units still circulate and the European market has no equivalent product ban.\n","rough_estimate":"~1 in 200,000 to 1 in 500,000 per regularly-exposed infant","kind":"intuition"},"native":{"display":"73 infant deaths plus 1,108 non-fatal incidents reported to CPSC across all US inclined-sleep products, January 2005–June 2019","numerator":73,"denominator":17500000,"unit":"per inclined-sleep unit sold in the US 2005–2019","population":"US infants 0-12 months exposed to inclined-sleep products in homes, 2005–2019"},"normalized":{"lifetime_us_adult":0.0000042,"display":"~1 in 240,000 chance per inclined-sleep unit sold — historical figure; the Safe Sleep for Babies Act (Pub.L. 117-126) banned these products in the US effective late 2022","log_value":-5.38,"assumptions":"73 deaths across all inclined-sleep products reported to CPSC January 2005 through June 2019 (CPSC, 2019-10-31). Denominator estimated from recall units: the Fisher-Price Rock 'n Play recall covered 4.7 million units alone, and competitor units (Kids II, Graco, others) bring total US inclined-sleep unit sales over the period to roughly 15-20 million. Per-unit risk works out to ~1 in 240,000. Per-infant risk is higher because many units are used by more than one child (hand-me-down, secondhand sale), giving an order-of- magnitude estimate of ~1 in 50,000 to 1 in 100,000 per regularly-exposed infant. The uncertainty band reflects this denominator ambiguity. Crucial historical caveat: the Safe Sleep for Babies Act (signed 2022-05-16, taking effect roughly 180 days later) made it unlawful to manufacture, sell, or distribute inclined sleepers with surfaces above 10 degrees. The forward- looking risk for a newly-purchased US unit is approximately zero from late 2022 onward; the residual risk is concentrated in secondhand units still in homes and in the EU and Polish markets, where no equivalent ban exists and the EN 12790-1/-2:2023 reclined-cradle standard governs the product class without prohibiting it.\n","uncertainty":{"low":0.000001,"high":0.00002},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Recalls/2019/Fisher-Price-Recalls-Rock-n-Play-Sleepers-Due-to-Reports-of-Deaths","title":"Fisher-Price Recalls Rock 'n Play Sleepers Due to Reports of Deaths","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"Over 30 infant fatalities in Rock 'n Play Sleepers since 2009 product introduction; 4.7 million units recalled","excerpt":"\"Since the 2009 product introduction, over 30 infant fatalities have occurred in Rock 'n Play Sleepers, after the infants rolled over while unrestrained, or under other circumstances.\"\n","source_date":"2019-04-12","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260317215300/https://www.cpsc.gov/Recalls/2019/Fisher-Price-Recalls-Rock-n-Play-Sleepers-Due-to-Reports-of-Deaths","calculation_notes":"Single-product recall covering 4.7 million Rock 'n Play units. The \"over 30 deaths\" figure here predates the broader CPSC aggregate of 73 deaths across all inclined-sleep products published six months later (Source B). Used here to anchor the recall denominator on a verified manufacturer-reported unit count.\n"},{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2020/CPSC-Cautions-Consumers-Not-to-Use-Inclined-Infant-Sleep-Products","title":"CPSC Cautions Consumers Not to Use Inclined Infant Sleep Products","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"73 infant deaths plus 1,108 non-fatal incidents across all inclined infant sleep products reported to CPSC January 2005 through June 2019","excerpt":"\"CPSC received reports of 1,108 incidents, including 73 infant deaths, related to infant inclined sleep products that occurred from January 2005 through June 2019. Dr. Mannen's report was conclusive that products with inclines 10 degrees or less, with flat and rigid surfaces, are likely safe for infant sleep.\"\n","source_date":"2019-10-31","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260625020202/https://www.cpsc.gov/Newsroom/News-Releases/2020/CPSC-Cautions-Consumers-Not-to-Use-Inclined-Infant-Sleep-Products","calculation_notes":"Canonical aggregate figure across all manufacturers. Despite the URL slug showing /2020/, the release date is 2019-10-31. The 10-degree cutoff Mannen identified in the cited biomechanical assessment became the threshold codified into federal law via the Safe Sleep for Babies Act in 2022 (Source C). 73 / 17.5M = 4.17e-6 per unit sold ≈ 1 in 240,000.\n"},{"url":"https://www.congress.gov/bill/117th-congress/house-bill/3182","title":"H.R.3182 — Safe Sleep for Babies Act of 2021","publisher":"US Congress / Congress.gov","source_type":"reputable_reference","statistic":"Banned manufacture, sale, and distribution of US infant inclined sleepers with surfaces greater than 10 degrees and crib bumpers; signed 2022-05-16 as Public Law 117-126","excerpt":"\"Safe Sleep for Babies Act of 2021 — This bill makes it unlawful to manufacture, sell, or distribute crib bumpers or inclined sleepers for infants. Specifically, inclined sleepers for infants are those designed for an infant up to one year old and have an inclined sleep surface of greater than 10 degrees. Latest Action: 05/16/2022 Became Public Law No: 117-126.\"\n","source_date":"2022-05-16","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251203193834/https://www.congress.gov/bill/117th-congress/house-bill/3182","calculation_notes":"Public Law 117-126 codifies the CPSC 10-degree threshold from Mannen's assessment. Took effect approximately 180 days after signing, i.e. late 2022. No EU member-state equivalent exists; the Polish and EU market is governed by EN 12790-1/-2:2023 (reclined cradles, revised March 2023), which addresses powered motion, electrical safety, entrapment, and cord hazards but does not ban the product class.\n"}],"comparison_anchors":[{"label":"SIDS death (per infant-year)","lifetime_us_adult":0.00014},{"label":"Co-sleeping infant death (per shared night)","lifetime_us_adult":0.000001},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Use of product for unsupervised sleep","multiplier":20,"notes":"Nearly all 73 documented deaths occurred during sleep use; supervised awake play in the same product presents minimal asphyxia risk"},{"factor":"Infant unrestrained in the sleeper","multiplier":5,"notes":"Most fatalities involved infants who rolled while unrestrained; CPSC noted unrestrained-roll as the dominant mechanism in the Rock 'n Play series"},{"factor":"Infant aged 0-3 months vs 3-6 months","multiplier":3,"notes":"Younger infants are over-represented in fatality data; neck strength is insufficient to clear the airway when rolled into the soft sided seat"},{"factor":"Use post-2022-ban (newly purchased US unit)","multiplier":0.01,"notes":"The federal ban took new units off the US market; risk approaches zero for first-time post-ban US buyers because the products are no longer sold"},{"factor":"Secondhand unit purchased pre-ban (still in EU/PL or US homes)","multiplier":1,"notes":"The ban does not address units already in homes; the EU has no equivalent prohibition; recall return rates were low, leaving many units in circulation"}],"short_label":"Inclined sleeper death","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The per-unit-sold figure is the most defensible aggregate, but it understates per-infant risk by an unknown factor because each unit is often used by more than one child. CPSC's incident-reporting database also under-counts: it captures fatalities reliably but only a fraction of near-misses and minor positional events. The 73-death total is bounded above only by the surveillance window (January 2005 through June 2019) — additional deaths after the recall cutoff are not in this count. The headline rate is now historical for the US market; what matters in 2026 is whether a given household has a secondhand or pre-ban unit still in use, and whether they are in a jurisdiction (EU, PL, much of the world) where the product class remains legal.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"An empty inclined infant sleeper on a carpeted floor, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/inclined-sleeper-infant-death","api_url":"https://likelier.app/api/fears/inclined-sleeper-infant-death.json"},{"slug":"avalanche-death","question":"What are the odds of dying in an avalanche?","quick_answer":"Across the whole US population, the lifetime risk of dying in an avalanche is about 1 in 210,000, from roughly 27 deaths a winter. That figure dilutes the risk across the vast majority who never enter avalanche terrain; for active backcountry users it is orders of magnitude higher.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Avalanche fear is sharply bimodal. Most Americans live nowhere near avalanche terrain and give it no thought at all. Among backcountry skiers, snowmobilers, and mountaineers, however, avalanche anxiety is pervasive and often well-calibrated — participants routinely check avalanche forecasts, carry beacons, and take courses. Media coverage of avalanche fatalities tends to spike around dramatic multi-burial incidents, reinforcing the sense that backcountry winter travel is inherently deadly, even as the per-trip risk for trained users remains low.\n","rough_estimate":"Backcountry users often guess ~1 in 1,000 to 1 in 10,000 per season","kind":"intuition"},"native":{"display":"~27 deaths per winter season in the US (10-year average)","numerator":27,"denominator":335000000,"unit":"per year","population":"US residents, all ages, including non-participants in avalanche-terrain activities"},"normalized":{"lifetime_us_adult":0.00000475,"display":"~1 in 210,000 lifetime (US adult, population-level)","log_value":-5.32,"assumptions":"CAIC reports an average of 27 avalanche deaths per winter in the US over the last 10 seasons. Annual rate: 27 / 335,000,000 ≈ 8.06 × 10⁻⁸. Compounded over 59 remaining adult years: 1 − (1 − 8.06 × 10⁻⁸)⁵⁹ ≈ 4.75 × 10⁻⁶ ≈ 1 in 210,000. This is a population-level figure that dilutes the risk across the vast majority of Americans who never enter avalanche terrain. For active backcountry recreationists, the per-participant risk is orders of magnitude higher.\n","uncertainty":{"low":0.0000013,"high":0.000009},"scope":"us_adult_lifetime"},"sources":[{"url":"https://avalanche.state.co.us/accidents/statistics-and-reporting","title":"Statistics and Reporting","publisher":"Colorado Avalanche Information Center (CAIC)","source_type":"govt_report","statistic":"Average of 27 avalanche deaths per winter in the US over the last 10 seasons; over 1,000 fatalities since 1950","excerpt":"\"Over the last 10 winters, an average of 27 people died in avalanches each winter in the United States.\"\n","source_date":"2025-09-30","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260305022420/https://avalanche.state.co.us/accidents/statistics-and-reporting","calculation_notes":"CAIC maintains the database of record for US avalanche fatalities. The 10-year average of 27 deaths per winter is the central estimate. The range over the past decade spans roughly 15 to 37 deaths per season, which drives the uncertainty band. Annual population-level rate: 27 / 335 × 10⁶ ≈ 8.06 × 10⁻⁸. Lifetime over 59 adult years: 1 − (1 − 8.06 × 10⁻⁸)⁵⁹ ≈ 4.75 × 10⁻⁶. Low bound uses ~15 deaths/yr; high bound uses ~37 deaths/yr plus upward trend in backcountry participation.\n","independence_note":"CAIC compiles fatality data from search-and-rescue reports, coroner records, and media accounts. Their database is the primary source cited by avalanche.org and by academic researchers; there is no fully independent parallel US avalanche fatality database.\n"},{"url":"https://avalanche.org/wp-content/uploads/2019/10/19_JORT_Peitzsch_etal.pdf","title":"How old are the people who die in avalanches? A demographic analysis of avalanche fatalities in the United States, 1950–2018","publisher":"Journal of Outdoor Recreation and Tourism","source_type":"peer_reviewed","statistic":"1,084 avalanche fatalities in the US from 1950–51 through 2017–18; victim demographics shifting toward older age groups","excerpt":"\"Between 1950–51 and 2017–18, 1,084 people were killed in avalanches in the United States. The age groups where fatalities are increasing are the ages of 30 to 39 and 40 to 49.\"\n","source_date":"2019-10-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20250330132044/https://avalanche.org/wp-content/uploads/2019/10/19_JORT_Peitzsch_etal.pdf","calculation_notes":"Peitzsch et al. analyzed 68 years of CAIC data and documented 1,084 fatalities. Over that span the annual average was roughly 16, reflecting lower backcountry participation in earlier decades. The upward trend to ~27/yr in recent decades tracks increased backcountry recreation participation rather than increased avalanche frequency. This paper provides the demographic granularity; the CAIC 10-year average provides the more current central estimate for normalization.\n","independence_note":"Peitzsch et al. used the same CAIC database but applied independent demographic and trend analysis. The paper was peer-reviewed and published in a recreation-research journal, providing an academic check on the raw CAIC counts.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"backcountry skier/snowboarder (50+ days/season)","multiplier":100,"notes":"CAIC: nearly all avalanche fatalities involve backcountry recreationists; a frequent backcountry user's per-season risk is orders of magnitude above the population average"},{"factor":"carries avalanche beacon, probe, and shovel","multiplier":0.5,"notes":"companion rescue within 15 minutes improves survival from ~30% to ~90%; gear enables that window"},{"factor":"non-mountain-state resident, no backcountry activity","multiplier":0.01,"notes":"avalanche deaths cluster in CO, UT, MT, WA, AK; non-participants in mountain states have effectively zero risk"},{"factor":"solo backcountry travel (no partner)","multiplier":3,"notes":"Brugger et al. 2007 (Resuscitation): companion rescue within 15 minutes is the primary survival determinant in burial; solo travelers have no companion to perform that rescue, roughly tripling burial fatality rates relative to group travel"},{"factor":"terrain trap below (cliff, creek bed, trees)","multiplier":2.5,"notes":"CAIC and IKAR-MEDCOM avalanche investigation data: terrain traps — cliff bands, creek bottoms, gullies, dense timber below the slide path — dramatically increase burial depth and trauma severity; recognized as a primary aggravating factor in avalanche fatality investigations"}],"short_label":"Avalanche","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The population-level figure of 1 in 210,000 is almost comically misleading as a personal risk estimate, because the denominator includes roughly 300 million Americans who never set foot in avalanche terrain. The relevant risk population is backcountry winter recreationists — estimated at 3–5 million participants in the US — for whom the per-participant annual risk is on the order of 1 in 100,000 to 1 in 200,000 per season, and higher still for those who spend many days in consequential terrain. Snowmobilers and backcountry skiers account for the majority of fatalities. Peitzsch et al. documented a demographic shift toward older victims (30–49), likely reflecting both increased participation by that cohort and the possibility that experience breeds overconfidence. Avalanche risk is also strikingly geographic: Colorado, Alaska, Utah, Montana, and Washington account for the vast majority of US fatalities.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-gate-review","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single snow-covered mountain slope with a subtle crack line across the snowpack, flat vector illustration."},"canonical_url":"https://likelier.app/avalanche-death","api_url":"https://likelier.app/api/fears/avalanche-death.json"},{"slug":"lightning-strike","question":"What are the odds of being killed by lightning?","quick_answer":"Being killed by lightning is rare — about 1 in 209,000 over a lifetime (~27 US deaths a year). And death is the exception even among the people lightning actually hits: the NWS reports that only about 10% of those struck are killed, so roughly 90% survive, often with lasting effects. Death is the rare exception, even among strikes.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"\"Struck by lightning\" is the archetype humans reach for to describe \"almost impossible\" — which itself suggests we intuitively understand it’s rare. We haven’t yet found a rigorous recent survey that isolates \"fear of being killed by lightning\" as a standalone question, so the perceived side here is marked as editorial intuition rather than polled data.\n","rough_estimate":"lightning is a classic 'overestimated' hazard in psychometric risk perception research — its high dread and media salience cause people to place it well above its actual frequency, even as the phrase 'struck by lightning' is colloquially used to mean 'effectively never'","kind":"survey","survey_source":{"title":"Facts and Fears: Understanding Perceived Risk","publisher":"Paul Slovic, Decision Research","url":"https://www.ldeo.columbia.edu/chrr/documents/meetings/roundtable/white_papers/slovic_wp.pdf","year":2002}},"native":{"display":"~27 lightning fatalities per year in the US (10-year average, 2009-2018)","numerator":27,"denominator":333000000,"unit":"per year","population":"US residents"},"normalized":{"lifetime_us_adult":0.00000478,"display":"1 in ~209,000 lifetime (US adult)","log_value":-5.32,"assumptions":"Uses the 10-year average of 27 lightning fatalities per year in the US (2009-2018, NOAA data) divided by the US population (~333M), compounded over 59 years of remaining adult life. Annual rate: 27/333,000,000 ≈ 8.11e-8. Lifetime: 1-(1-8.11e-8)^59 ≈ 4.78e-6. The 30-year average (1989-2018) is higher at 43/year; the 10-year figure reflects the continued decline. Lightning risk is strongly seasonal and strongly correlated with outdoor recreation.\n","uncertainty":{"low":0.000003,"high":0.0000075},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.weather.gov/safety/lightning-odds","title":"Lightning Safety — Odds","publisher":"US National Weather Service (NOAA)","source_type":"govt_report","statistic":"Annual lightning deaths in the US: 27 (10-year average, 2009-2018), 43 (30-year average, 1989-2018); lifetime odds of being struck ~1 in 15,300; only ~10% of those struck are killed, leaving ~90% with varying disability","excerpt":"\"Over the last 30 years (1989-2018) the U.S. has averaged 43 reported lightning fatalities per year. [...] In the last 10 years (2009-2018), the U.S. has averaged 27 lightning fatalities. [...] Odds of being struck in your lifetime (Est. 80 years): 1/15,300. [...] Only about 10% of people who are struck by lightning are killed, leaving 90% with various degrees of disability.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260325083021/https://www.weather.gov/safety/lightning-odds","calculation_notes":"NWS publishes both 10-year (27/yr) and 30-year (43/yr) averages. We use the 10-year figure for the headline as it reflects the continued secular decline. NWS lifetime-odds figures are for *being struck*, not *being killed*. Our normalized figure uses the annual-death count (27) ÷ US population (~333M), compounded over adult remaining years, which focuses specifically on fatality.\n","independence_note":"Primary US lightning-fatality source. NOAA/NWS counts come from Storm Data field reports compiled by local NWS offices, methodologically distinct from CDC's ICD-10 death-certificate tabulation, though the two pipelines describe overlapping events.\n"},{"url":"https://www.cdc.gov/lightning/data-research/","title":"Lightning Strike Victim Data","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"444 lightning strike deaths in the US from 2006-2021 (~28 deaths/year average)","excerpt":"\"From 2006 through 2021, there were 444 lightning strike deaths in the United States. Males are four times more likely than females to be struck by lightning, and leisure activities account for almost two-thirds of lightning deaths.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260411093631/https://www.cdc.gov/lightning/data-research/","calculation_notes":"CDC's 16-year average (~28/yr) is consistent with NOAA's 10-year average (27/yr, 2009-2018). Both sit well below the 30-year NOAA average (43/yr, 1989-2018), confirming the long-term decline. We use NOAA's 10-year average for the headline and cite CDC as independent corroboration of the order of magnitude.\n","independence_note":"NOAA and CDC use different primary data sources (NWS storm reports vs death certificates); counts as meaningfully independent verification.\n"},{"url":"https://journals.ametsoc.org/view/journals/wcas/8/1/wcas-d-15-0032_1.xml","title":"A Summary of Recent National-Scale Lightning Fatality Studies","publisher":"Weather, Climate, and Society / American Meteorological Society","source_type":"peer_reviewed","statistic":"US lightning fatality rate declined from ~6.3 per million in 1901 (peak) to below 0.2 per million by the 2001-2010 decade — a ~95% reduction driven by urbanization and indoor work","excerpt":"[Paraphrase — figures per million, not a verbatim quote] Holle reports the US annual lightning-fatality rate per million inhabitants peaking around 6.3 in 1901 (roughly 4.8 per million as a first-decade average) and falling to below 0.2 per million per year by the 2001-2010 decade — a reduction of more than 90 percent that reflects urbanization, improved weather warnings, and the shift from agricultural to indoor occupations.\n","source_date":"2016-02-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20251123045151/https://journals.ametsoc.org/view/journals/wcas/8/1/wcas-d-15-0032_1.xml","calculation_notes":"Holle provides the long-run rate context (rates expressed per million inhabitants per year). The below-0.2-per-million 2001-2010 rate is consistent with NOAA's modern 27-43 annual deaths: 27/333M ≈ 0.08 per million and 43/333M ≈ 0.13 per million, both under 0.2 (the further reduction from 2010-2018 reflects the continued trend). This independent peer-reviewed analysis corroborates the NOAA per-capita figure and strengthens the historical-decline framing.\n","independence_note":"Holle uses NOAA Storm Data as the underlying death record source, so shares the upstream with NOAA. However, the analytical layer (time-series decomposition, demographic attribution) is methodologically independent of NOAA's raw counts.\n"}],"comparison_anchors":[{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by bee/wasp sting (lifetime, US)","lifetime_us_adult":0.0001267},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Florida / Gulf Coast","probability":0.00001,"notes":"NOAA: Florida has the highest lightning density in the US; fatality rate is roughly 3x the national average"},{"region":"Mountain West (CO, UT)","probability":0.000007,"notes":"afternoon thunderstorm pattern over high terrain elevates exposure for hikers and outdoor workers"},{"region":"Pacific Northwest / Northeast urban","probability":0.000001,"notes":"lower thunderstorm frequency and less outdoor work exposure"}],"personal_factor_multipliers":[{"factor":"outdoor worker (agriculture, construction, roofing)","multiplier":5,"notes":"NOAA: outdoor workers account for a disproportionate share of lightning fatalities"},{"factor":"frequent golfer (100+ rounds/year)","multiplier":3,"notes":"golf courses are open, exposed terrain; golfers are a perennial category in NOAA fatality reports"},{"factor":"primarily indoor worker, urban resident","multiplier":0.2,"notes":"minimal outdoor exposure during thunderstorms"},{"factor":"male sex","multiplier":4,"notes":"CDC lightning data (2006–2021): approximately 80% of US lightning fatalities are male; males are ~4× more likely than females to die from a lightning strike, driven by higher rates of outdoor recreational and occupational exposure"},{"factor":"Florida or Gulf Coast resident","multiplier":3,"notes":"NOAA: Florida has the highest lightning strike density in the continental US; Gulf Coast states collectively account for a disproportionate share of national lightning fatalities, at roughly 3× the national per-capita rate"}],"short_label":"Lightning","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The risk is highly non-uniform: outdoor workers, boaters, golfers, and residents of Florida and the Gulf Coast face a meaningfully higher per-year rate than the national average. Indoor risk is essentially zero.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"injury-reframe-8d-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized lightning bolt rendered as a flat geometric zigzag against a muted sky, vector illustration."},"canonical_url":"https://likelier.app/lightning-strike","api_url":"https://likelier.app/api/fears/lightning-strike.json"},{"slug":"dirty-can-illness","question":"What are the odds of getting sick from drinking out of an unwashed soda can?","quick_answer":"Effectively zero: no illness has ever been documented from drinking out of an unwashed beverage can, against the roughly 50 billion consumed in the US each year. Many people imagine a 1–5% per-can risk that the evidence does not support.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The \"rat urine on soda cans\" email chain has circulated since at least 2002, warning that warehouse-stored cans are coated in rodent urine, pesticide residue, and lethal bacteria. Every version features a named victim who died of leptospirosis after drinking from an unwashed can. The story is vivid, plausible-sounding, and has been translated into dozens of languages. It taps into a deep disgust response -- the idea that an invisible film of contamination sits between your lips and every can you open. Many people ritually wipe or rinse can lids before drinking as a result.\n","rough_estimate":"~1-5% chance of illness per unwashed can","kind":"intuition"},"native":{"display":"~0 documented cases of illness specifically from drinking from an unwashed beverage can","numerator":1,"denominator":10000000,"unit":"estimated illness probability per can consumed unwashed (no documented cases)","population":"canned beverages consumed in the US (~50 billion per year), with no confirmed illness attribution"},"normalized":{"lifetime_us_adult":0.000005,"display":"~0.0005% lifetime probability of illness from unwashed cans over 40 years (effectively zero)","log_value":-5.3,"assumptions":"No documented case of clinically significant illness has been traced to drinking from an unwashed beverage can in any published medical literature, CDC case report, or food safety investigation. Americans consume roughly 50 billion canned beverages per year; the vast majority are consumed without washing the lid. If the per-can illness rate were even 1 in 1 million, that would produce 50,000 identifiable cases per year -- a signal that would be impossible to miss epidemiologically. The absence of any signal places the true rate well below 1 in 10 million per can. We assign a nominal 1-in-10-million figure to represent \"too rare to measure but not physically impossible.\" Over 40 years at ~1,000 cans/year: 40,000 exposures × 1e-7 ≈ 0.004, or ~0.4%. However, this overstates the risk because the 1-in-10-million per-can figure is itself an arbitrary ceiling, not a measured rate. We use 0.000005 (0.0005%) as the lifetime estimate, reflecting a more conservative per-can rate of ~1e-10 for certified cold-chain retail conditions (vs the 1e-7 nominal ceiling). The leptospirosis route (the specific claim in viral emails) is biologically near-impossible: Leptospira bacteria lack a waterproof membrane and die within minutes on dry surfaces. CDC reports ~150 US leptospirosis cases per year, almost all from floodwater or freshwater exposure in Puerto Rico and Hawaii. Zero have been attributed to beverage containers. Bacteria ARE found on can surfaces (Staphylococcus, E. coli, Bacillus in studies from Pakistan and Nigeria), but these organisms are also found on doorknobs, phones, and shopping carts at similar concentrations, and are handled by normal immune function without clinical illness.\n","uncertainty":{"low":1e-7,"high":0.0001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.snopes.com/fact-check/rat-urine-soda-cans/","title":"Can You Catch Leptospirosis from Rat Urine on Soda Cans?","publisher":"Snopes","source_type":"reputable_reference","statistic":"Rated 'Mixture': leptospirosis CAN be contracted from diseased-animal urine, but the specific viral-email deaths (Hawaii stock clerk, Texas/Belgian boater, family picnic) are fabrications with no matching news record; a search of news items turned up no one dying in the manner described","excerpt":"\"In general, urine-encrusted soda cans are not the most likely purveyors of these diseases. Most cans of soda are packaged into cardboard boxes while still on the production line and thus aren't at any risk of contamination... a search of news items fails to turn up anything about anyone's having died in the fashion described. Nonetheless, it's still always a good idea to wipe off cans from exposed six-packs or those dispensed from drink machines, if for no other reason than to avoid picking up something passed on by the human handlers of the product.\"\n","source_date":"2000-07-20","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260505052900/https://www.snopes.com/fact-check/rat-urine-soda-cans/","calculation_notes":"Snopes (Barbara Mikkelson) rates the \"deadly rat urine on soda cans\" claim \"Mixture\": What's True is that leptospirosis is a real, potentially deadly illness spread via the urine of diseased animals; What's False is the specific viral-email death narratives (a Hawaii/Maui stock clerk, a Texas or Belgian woman who went boating, a Sunday family picnic). For each named version Snopes notes \"a search of news items fails to turn up anything about anyone's having died in the fashion described.\" Snopes also relays the CDC/National Center for Infectious Diseases statement that the related hantavirus stock-clerk email \"is untrue.\" The article concludes cans are not a likely disease vector because most are boxed on the production line, but that wiping machine-dispensed or loose cans remains a reasonable precaution against germs from human handlers. This establishes the baseline: no documented death from drinking an unwashed can, while the underlying disease is real.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12413079/","title":"Epidemiological trends of leptospirosis in the United States, 2014-2020","publisher":"PLOS Neglected Tropical Diseases","source_type":"peer_reviewed","statistic":"1,053 US leptospirosis cases over 7 years (~150/year); 54% from Puerto Rico, 15% Hawaii; national incidence 0.48 per 100,000; zero cases attributed to beverage containers","excerpt":"\"A total of 1,053 leptospirosis case reports were received from 34 jurisdictions between 2014 and 2020, with a national incidence rate of 0.48 per 100,000 population. Puerto Rico accounted for 54 percent and Hawaii for 15 percent of all cases. Transmission was associated with recreational water exposure, flooding, and occupational contact.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426195719/https://pmc.ncbi.nlm.nih.gov/articles/PMC12413079/","calculation_notes":"This PLOS NTD study provides the authoritative US leptospirosis epidemiology. 150 cases/year across 330M people = 0.48 per 100,000. Almost all cases are from Puerto Rico and Hawaii, with transmission via floodwater or freshwater exposure. The mainland US sees roughly 30-50 confirmed cases per year. Beverage container surfaces are not listed as a transmission route in the CDC case definition, the epidemiological literature, or this study. Leptospira require constant immersion in water to survive; they die within minutes on dry surfaces.\n"},{"url":"https://jmmg.wum.edu.pk/index.php/ojs/article/view/61","title":"Tops of Beverage Cans Are a Potential Source of Infection: A Study of Bacterial Load Present on the Lids of Beverage Cans","publisher":"The Journal of Microbiology and Molecular Genetics (Wah Medical University)","source_type":"peer_reviewed","statistic":"180 non-refrigerated beverage cans sampled in Pakistan; isolated bacteria belonged to Bacillus, Staphylococcus, Corynebacterium, Streptococcus, Klebsiella, and Escherichia species; tap water + dry tissue wiping removed a maximum of 76.6% of bacterial load","excerpt":"\"One hundred and eighty apparently clean and non-refrigerated beverage cans were collected from different shops. Different types of bacteria were isolated which belonged to Bacillus spp, Staphylococcus spp, Corynebacterium spp, Streptococcus spp, Klebsiella spp and Escherichia spp. Cleaning with tap water plus wiping with dry tissue resulted in maximum removal of bacterial load (76.6%) as compared to other methods (68.8% with dry tissue and 47.3% with tap water).\"\n","source_date":"2022-12-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426195759/https://jmmg.wum.edu.pk/index.php/ojs/article/view/61","calculation_notes":"This is the largest peer-reviewed microbiological study of can lid contamination. It confirms that bacteria ARE present on can surfaces -- this part of the fear is factually correct. However, the study was conducted in Pakistan with open-air retail, non-refrigerated display, and higher ambient temperatures. No equivalent peer-reviewed study exists for US/EU cold-chain retail conditions. The organisms found (Staph, E. coli, Bacillus) are ubiquitous environmental bacteria also found on phones, doorknobs, and kitchen surfaces. Their presence on a can lid does not equate to a meaningful illness pathway for an immunocompetent person.\n"}],"comparison_anchors":[{"label":"Charger left plugged in causing fire (lifetime)","lifetime_us_adult":0.000005},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013},{"label":"Food poisoning requiring hospitalization (lifetime, US)","lifetime_us_adult":0.001}],"personal_factor_multipliers":[{"factor":"immunocompromised individual","multiplier":100,"notes":"Immunosuppressed individuals (chemotherapy, HIV, organ transplant) have reduced ability to fight environmental bacteria that healthy immune systems clear without symptoms"},{"factor":"can stored in warm, humid, open-air conditions (e.g. tropical outdoor market)","multiplier":50,"notes":"The Pakistan/Nigeria studies found high bacterial loads in warm, non-refrigerated retail; cold-chain US/EU retail produces dramatically lower contamination"},{"factor":"can visibly contaminated with animal droppings","multiplier":500,"notes":"Visible contamination is a qualitatively different scenario from normal warehouse storage; this crosses from theoretical to plausible, though still far below food-handling risks"},{"factor":"wiping the lid with a clean tissue before drinking","multiplier":0.25,"notes":"The Pakistan study found tap water + dry tissue wiping removed 76.6% of bacterial load; a simple wipe is effective at reducing even the minimal baseline risk"}],"short_label":"Dirty can illness","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The \"effectively zero\" risk assessment applies to canned beverages from regulated supply chains (US, EU, Japan, etc.) stored in climate-controlled environments. The Pakistani and Nigerian studies finding high contamination rates are real science, but they describe conditions (open-air tropical retail, no refrigeration, no shrink-wrap) that differ fundamentally from US/EU cold-chain distribution. No peer-reviewed microbiological study of can lid contamination in US/EU retail settings has been published, so the low-risk assessment for Western consumers rests on epidemiological absence (no cases) rather than direct measurement (no contamination). The bacteria found on cans in developing-country studies are the same organisms found on every public surface; their presence does not constitute a unique risk pathway. The leptospirosis-specific claim is biologically incoherent: Leptospira cannot survive desiccation, making dry aluminum surfaces an impossible transmission vector.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-24","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A closed aluminum soda can on a clean surface with a small tissue beside it, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/dirty-can-illness","api_url":"https://likelier.app/api/fears/dirty-can-illness.json"},{"slug":"school-shooting-student","question":"What are the odds of a US student being killed in a school shooting?","quick_answer":"A US student's odds of being killed in a school shooting are about 1 in 2,500,000 per year, or roughly 1 in 192,000 across a 13-year K-12 career. About 45% of parents are worried, and many imagine odds of 1 in a few thousand, far above the base rate.\n","category":"crime","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"School shootings dominate parental fear discourse in the US. A 2023 Pew Research Center poll (fielded fall 2022) found 22% of parents saying they are \"extremely or very worried\" that any of their children might get shot at some point, and another 23% \"somewhat worried\" — about 45% combined. That ranks below worry about mental health (40% extremely/very worried) and bullying (35%). The worry level bears little relationship to the base rate: parents in low-crime suburbs report similar levels of concern to those in high-crime urban areas. The fear is driven by saturation media coverage and the particular horror of children as victims, not by frequency.\n","rough_estimate":"~1 in a few thousand lifetime feels about right to many parents","kind":"poll","survey_source":{"title":"What's It Like To Be a Parent in America Today?","publisher":"Pew Research Center","url":"https://www.pewresearch.org/social-trends/2023/01/24/parenting-in-america-today/","year":2023}},"native":{"display":"~1 in 2,500,000 per student per year","numerator":1,"denominator":2500000,"unit":"per student per year","population":"US K-12 students (~50 million enrolled)"},"normalized":{"lifetime_us_adult":0.0000052,"display":"1 in ~192,000 over a 13-year K-12 career","log_value":-5.28,"assumptions":"The K-12 School Shooting Database (CHDS/Naval Postgraduate School) and Everytown for Gun Safety both track school shooting fatalities with slightly different inclusion criteria. CHDS records incidents where a firearm is discharged inside or on school property, K-12, regardless of time or motive; Everytown restricts to incidents during school hours or school-sponsored events. Annual K-12 student fatalities from school shootings average roughly 20 per year over the 2010-2024 window (ranging from single digits in quiet years to 40+ in years with a major incident like Uvalde 2022 or Sandy Hook 2012). Using 20 deaths per year against ~50 million K-12 students gives an annual per-student hazard of ~4.0e-7, or roughly 1 in 2.5 million per year. Compounded over a 13-year K-12 career: 1 - (1 - 4.0e-7)^13 ≈ 5.2e-6, or ~1 in 192,000. The uncertainty band reflects the spread between different databases, the year-to-year volatility driven by outlier events, and the difference between narrow (active-shooter-only) and broader (any discharge on campus) definitions.\n","uncertainty":{"low":0.0000026,"high":0.000013},"scope":"us_adult_lifetime"},"sources":[{"url":"https://k12ssdb.org/methodology-1","title":"K-12 School Shooting Database","publisher":"K-12 School Shooting Database (David Riedman; formerly Center for Homeland Defense and Security, Naval Postgraduate School)","source_type":"primary_study","statistic":"More than 2,380 documented K-12 school shooting incidents in the US from 1966 to present, with an annual average of roughly 15-25 fatalities in the 2010-2024 window","excerpt":"\"The K-12 School Shooting Database is a widely inclusive, open-source research project that documents when a gun is fired, brandished (pointed at a person with intent), or bullet hits school property, regardless of the number of victims, time, day, or reason.\"\n","source_date":"2026-04-10","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260410171521/https://k12ssdb.org/methodology-1","calculation_notes":"CHDS uses the broadest inclusion criteria of the major trackers — any firearm discharge inside or on school property, K-12. This captures accidental discharges, suicides, after-hours incidents, and stray bullets alongside targeted attacks. For the native figure, I filtered to fatalities from targeted attacks during school operations, which narrows the annual count to roughly 15-25 deaths per year in the 2010-2024 window. The central estimate of ~20 deaths/year against ~50 million enrolled K-12 students yields a per-student annual rate of ~4.0e-7, or 1 in 2.5 million. Compounded over 13 years of K-12 enrollment: 1 - (1 - 4.0e-7)^13 ≈ 5.2e-6, or ~1 in 192,000.\n"},{"url":"https://everytownresearch.org/maps/gunfire-on-school-grounds/","title":"Gunfire on School Grounds in the United States","publisher":"Everytown for Gun Safety Support Fund","source_type":"reputable_reference","statistic":"Over 400 incidents of gunfire on school grounds per year (2023-2024); subset with fatalities during school hours averages roughly 20 student deaths per year","excerpt":"\"Since 2013, there have been over 3,500 incidents of gunfire on school grounds. Everytown tracks every time a firearm discharges a live round inside or into a school building, or on or onto a school campus or grounds.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260505062545/https://everytownresearch.org/maps/gunfire-on-school-grounds/","calculation_notes":"Everytown's tracker is broader than \"school shooting\" in the public imagination — it includes accidental discharges, after-hours incidents, and suicides on campus. For the normalized figure, I used the subset of fatal incidents occurring during school hours or school-sponsored events, which aligns with what parents typically fear. This subset yields roughly 15-25 student deaths per year over the past decade, consistent with the CHDS data when filtered to comparable criteria.\n","independence_note":"Everytown maintains its own incident database, independently collected from media reports, law-enforcement records, and school district disclosures. It overlaps with CHDS on most high-profile incidents but uses different inclusion and classification criteria. Treat as an independent source.\n"},{"url":"https://nces.ed.gov/programs/coe/indicator/a01","title":"Indicator of School Crime and Safety: Violent Deaths at School","publisher":"National Center for Education Statistics / Bureau of Justice Statistics","source_type":"govt_report","statistic":"41 school-associated violent deaths in 2020-21; of 2,436 total youth homicides that year, only a small fraction occurred at school","excerpt":"\"From July 2020 through June 2021, there were 41 school-associated violent deaths in the United States, comprising 20 homicides, 17 suicides, 3 legal intervention deaths, and 1 undetermined violent death.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260405215007/https://nces.ed.gov/programs/coe/indicator/a01/","calculation_notes":"NCES/BJS Indicators of School Crime and Safety is the federal government's primary statistical publication on school violence. The \"violent deaths at school\" series counts homicides and suicides of youth ages 5-18 at school or on the way to/from school. This is broader than shooting deaths alone but provides the authoritative federal baseline. In 2020-21, only 11 of 2,436 total youth homicides were school-associated — well under 1% — underscoring how rare school-based lethal violence is relative to the total.\n"},{"url":"https://www.pewresearch.org/social-trends/2023/01/24/parenting-in-america-today/","title":"Parenting in America Today","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"22% of US parents extremely/very worried their child might get shot, 23% somewhat worried (≈45% combined), as of fall 2022; ranks below mental health (40%) and bullying (35%) among parental worries","excerpt":"\"In a fall 2022 Pew Research Center survey, 22% of parents with children under 18 said they were extremely or very worried about any of their children getting shot at some point, while another 23% said they were somewhat worried.\"\n","source_date":"2023-01-24","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260402175334/https://www.pewresearch.org/social-trends/2023/01/24/parenting-in-america-today/","calculation_notes":"Used only for the perceived-risk side. The ≈45% combined worry figure (22% extremely/very + 23% somewhat) is not an elicited probability — it measures the share of parents who report worry about the scenario, not what probability they assign to it. Note Pew measures worry that a child \"might be shot at some point\" generally, not school shootings specifically, and the concern ranks below mental health (40%) and bullying (35%) in the same survey. This is the most recent large-scale national survey tracking parental worry about children being shot.\n","independence_note":"Pew survey methodology is entirely independent from the school shooting incident databases (CHDS, Everytown, NCES). Measures public perception, not incidence.\n"}],"comparison_anchors":[{"label":"Dying in a mass shooting (lifetime, US adult)","lifetime_us_adult":0.0000088},{"label":"Drowning death (lifetime, US)","lifetime_us_adult":0.000565},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"K-12 student, any grade (13-year career)","probability":0.0000052,"notes":"baseline figure; pooled across all grades, school types, and geographies"},{"region":"High school student (grades 9-12)","probability":0.0000035,"notes":"most school shooting fatalities involve high school students, but the 4-year exposure window is shorter than the full K-12 career"},{"region":"Elementary school student (grades K-5)","probability":0.0000015,"notes":"fewer incidents target elementary schools; Sandy Hook (2012) is an outlier in the data"},{"region":"Urban school district","probability":0.0000078,"notes":"higher absolute incident count but also larger student populations; per-student rate is modestly elevated"},{"region":"Suburban/rural school district","probability":0.0000039,"notes":"lower absolute count; several high-profile incidents (Columbine, Sandy Hook, Uvalde) occurred in suburban or small-town settings"}],"personal_factor_multipliers":[{"factor":"High school student (grades 9–12) vs. K-12 average","multiplier":3,"notes":"CHDS K-12 School Shooting Database 2024 and NCES Indicators of School Crime and Safety 2024: fatal shooting incidents are concentrated in high schools and middle schools; CHDS data shows high-school-age students account for a disproportionate share of targeted-attack fatalities, with an estimated 3× per-student annual hazard for high schoolers compared with the overall K-12 pooled average."},{"factor":"Southern US region","multiplier":2,"notes":"CHDS K-12 School Shooting Database regional tabulations: Southern states (TX, FL, GA, and peers) account for a higher absolute number of K-12 school shooting incidents than their student-population share would predict, yielding approximately 2× the per-student incident rate relative to Northeastern and Midwestern states over the 2010–2024 window."},{"factor":"Urban school district","multiplier":2,"notes":"NCES Indicators of School Crime and Safety 2024 and Everytown Gunfire on School Grounds database: urban school districts record higher absolute incident counts than suburban or rural districts; while per-student rates are modulated by larger enrollment denominators, urban students face approximately 2× the per-student incident rate compared with suburban and rural peers based on CHDS city-level tallies."}],"short_label":"School shooting","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Every number here depends heavily on definitions. \"School shooting\" can mean anything from a targeted mass-casualty attack during class (the scenario parents fear) to an accidental discharge in a parking lot after hours. The CHDS database includes all of these; the NCES/BJS series uses \"school-associated violent death\" which adds stabbings and other non-firearm homicides. When filtered to the scenario parents actually worry about — a targeted shooting during school hours — the annual fatality count drops to roughly 10-20 per year, and the per-student risk drops further. The year-to-year variance is extreme: a single Uvalde- or Sandy Hook-scale event can triple the annual death toll. The 13-year compounding assumption treats each year as independent and equally risky, which is a simplification — the trend may be rising, flat, or noisy depending on the window chosen. Finally, the \"lifetime\" framing here is a 13-year K-12 career, not the 59-year adult horizon used for most entries on this site; the number is not directly comparable to other entries without adjusting for the different exposure window.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":2,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An empty school hallway with rows of closed lockers, flat vector illustration, muted tones, no people."},"canonical_url":"https://likelier.app/school-shooting-student","api_url":"https://likelier.app/api/fears/school-shooting-student.json"},{"slug":"elevator-escalator-death","question":"What are the odds of dying in an elevator or escalator accident?","quick_answer":"A US adult's lifetime odds of dying in an elevator or escalator accident are about 1 in 188,000, from roughly 30 deaths a year. Most people would guess hundreds of deaths annually.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"The elevator is a reliable anxiety generator. A metal box suspended by cables over a shaft, entered voluntarily multiple times a day — the setup practically writes its own horror scenario. Escalators add a moving-machinery dimension. No national survey isolates \"fear of dying in an elevator or escalator\" as a standalone item, but pop-culture representation (final-destination scenarios, news coverage of rare entrapment deaths) keeps the fear salient far out of proportion to the actual toll. The perceived risk here is editorial intuition: most people, if asked, would guess the annual death count is substantially higher than it actually is.\n","rough_estimate":"most people would guess hundreds of deaths per year; the real number is ~30","kind":"intuition"},"native":{"display":"~30 elevator/escalator deaths per year in the US","numerator":9,"denominator":100000000,"unit":"per year","population":"US residents and workers, all ages"},"normalized":{"lifetime_us_adult":0.00000531,"display":"1 in ~188,000 lifetime (US adult)","log_value":-5.28,"assumptions":"CPSC and BLS data report approximately 30 deaths per year from elevator and escalator incidents in the US (average over 1992–2016 reporting periods). With a US population of ~333 million, the annual rate is roughly 30/333,000,000 ≈ 9.0 × 10⁻⁸. Compounded over 59 years of remaining adult life at constant hazard: 1 − (1 − 0.00000009)^59 ≈ 0.00000531 ≈ 1 in 188,000. Approximately half of the deaths are occupational (installation, repair, maintenance workers), so the general-public-only rate is roughly half this figure.\n","uncertainty":{"low":0.0000035,"high":0.0000071},"scope":"us_adult_lifetime"},"sources":[{"url":"https://stacks.cdc.gov/view/cdc/221302","title":"Deaths and Injuries Involving Elevators or Escalators","publisher":"CDC / NIOSH (Center to Protect Workers' Rights)","source_type":"govt_report","statistic":"~30 deaths and ~17,000 injuries per year from elevator and escalator incidents in the US","excerpt":"\"Incidents involving elevators and escalators kill about 30 and seriously injure about 17,000 people each year in the United States. Elevators cause almost 90% of the deaths and 60% of serious injuries.\"\n","source_date":"2023-06-15","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413171607/https://stacks.cdc.gov/view/cdc/221302","calculation_notes":"The CPWR/NIOSH report synthesizes BLS Census of Fatal Occupational Injuries (1992–2009) and CPSC National Electronic Injury Surveillance System (1997–2010) data. The ~30 deaths/year figure has been stable across multiple reporting periods. Annual rate: 30 / 333,000,000 ≈ 9.0 × 10⁻⁸ per person-year. Lifetime over 59 adult years: 1 − (1 − 9.0 × 10⁻⁸)^59 ≈ 5.31 × 10⁻⁶. Uncertainty band reflects the range of annual counts across the ~20-year reporting window (roughly 20–40 deaths/year).\n","independence_note":"This report uses BLS occupational-death data and CPSC consumer-injury data, two independent federal surveillance systems with different coverage and methodologies. BLS counts workplace deaths; CPSC counts consumer-product-related emergency-department visits and deaths. Together they capture both occupational and public incidents.\n"},{"url":"https://www.cpwr.com/wp-content/uploads/elevator_escalator_BLSapproved_1.pdf","title":"Deaths and Injuries Involving Elevators and Escalators — A Report of the Center To Protect Workers' Rights","publisher":"CPWR — The Center for Construction Research and Training","source_type":"reputable_reference","statistic":"Half of the ~30 annual deaths are occupational; elevators cause ~90% of deaths; falls into shafts account for 56% of worker deaths","excerpt":"\"Incidents involving elevators and escalators kill about 30 and seriously injure about 17,000 people each year in the United States. Half of the deaths are to people working in or near elevators. Elevators cause almost 90% of the deaths.\"\n","source_date":"2013-06-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250805142847/https://www.cpwr.com/wp-content/uploads/elevator_escalator_BLSapproved_1.pdf","calculation_notes":"CPWR's analysis provides the occupational breakdown that matters for understanding who is actually at risk. Of the ~30 annual deaths, ~15 are workers (installers, repairers, maintenance). Among worker deaths near elevator shafts, 56% are falls into the shaft. For escalators specifically, ~2 passenger deaths per year were recorded, with 75% due to falls. The public (non-occupational) death rate is roughly half the pooled figure, or about 1 in ~376,000 lifetime.\n","independence_note":"CPWR repackages Bureau of Labor Statistics Census of Fatal Occupational Injuries and CPSC data; shares the same upstream as CDC/NIOSH but adds construction-sector context.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"elevator/escalator maintenance worker","multiplier":50,"notes":"CDC/NIOSH: roughly half of elevator deaths are maintenance workers; occupational exposure is the dominant risk factor"},{"factor":"general public, occasional elevator use","multiplier":0.5,"notes":"public fatalities are roughly half the total and spread across the full population of elevator users"},{"factor":"age 65+ (elderly rider)","multiplier":3,"notes":"CPSC NEISS data and CDC/NIOSH reports indicate that escalator deaths among passengers are disproportionately concentrated in adults aged 65 and older, who account for the majority of fall-related escalator fatalities. Reduced balance, slower gait adaptation, and greater frailty on impact all elevate fatal-outcome risk compared to younger adults.\n"},{"factor":"loose clothing, untied footwear, or long scarves on escalator","multiplier":4,"notes":"CPSC incident reports identify clothing and footwear entrapment in escalator machinery (step gaps, side panels, comb plates) as a recurring mechanism for serious injury. Loose laces, flip-flops, and long garments are explicitly cited in CPSC safety advisories as primary entrapment precursors. The ANSI A17.1 escalator safety standard lists entrapment guards as required equipment precisely because clothing contact is a documented hazard.\n"}],"short_label":"Elevator/escalator death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The ~30 deaths/year figure combines two quite different risk populations. About half are occupational deaths — elevator installers, repairers, and maintenance workers who are exposed to open shafts, moving machinery, and fall hazards as part of their job. The other half are members of the public: passengers, building residents, and bystanders. The public-only death rate is roughly half the pooled figure shown above. Escalator deaths among passengers are extremely rare (~2 per year) and overwhelmingly involve falls rather than entrapment or mechanical failure, disproportionately affecting elderly riders. The injury toll (~17,000/year) is orders of magnitude larger than the death toll, and most injuries are non-fatal falls on escalators.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":4,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-opus-4-6-research","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"An abstract set of ascending parallel lines suggesting steps, muted grey and blue tones, flat vector illustration."},"canonical_url":"https://likelier.app/elevator-escalator-death","api_url":"https://likelier.app/api/fears/elevator-escalator-death.json"},{"slug":"aluminum-cookware-alzheimers","question":"What are the odds of getting Alzheimer's disease from cooking with aluminum pots or foil?","quick_answer":"There is no measured Alzheimer's risk from normal aluminum cookware or foil use; the entry carries an upper-bound placeholder of about 1 in 170,000 over a lifetime, and the true attributable risk may be zero. Many consumers still treat aluminum pots and foil as a measurable dementia risk factor, which the evidence does not support.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"The intuition that aluminum cookware and foil contribute to Alzheimer's disease persists decades after the original hypothesis emerged. It traces to two real findings that were overgeneralized: dialysis encephalopathy in patients exposed to aluminum-contaminated dialysate in the 1970s, and elevated aluminum levels found in some Alzheimer's brain tissue. Both observations are real; neither establishes that dietary or cookware aluminum exposure causes Alzheimer's in the general population. Consumer behavior surveys consistently show that a meaningful fraction of households actively avoid aluminum cookware and foil specifically to lower perceived dementia risk, and \"aluminum-free\" labeling commands a price premium in cookware marketing.\n","rough_estimate":"Many consumers treat aluminum cookware as a measurable Alzheimer's risk factor","kind":"intuition"},"native":{"display":"~0 attributable Alzheimer's cases per 100,000 consumer-years of normal aluminum cookware use","numerator":1,"denominator":10000000,"unit":"per year of normal consumer aluminum cookware / foil exposure","population":"US adults using uncoated aluminum pots, anodized aluminum cookware, or aluminum foil at normal cooking conditions"},"normalized":{"lifetime_us_adult":0.0000059,"display":"~1 in 170,000 lifetime (US adult)","log_value":-5.23,"assumptions":"No epidemiological study has measured Alzheimer's disease incidence attributable to consumer aluminum cookware or foil use. The Wang 2016 meta-analysis (OR 1.71, 95% CI 1.35-2.18) covers chronic occupational and high-contamination exposure, not cookware. The Virk & Eslick 2015 meta-analysis of antacid use (a much higher dietary aluminum exposure than cookware) found OR 1.0 (95% CI 0.8-1.2) -- no association. WHO's Joint Expert Committee on Food Additives (JECFA) sets a Provisional Tolerable Weekly Intake of 2 mg Al/kg bodyweight; typical dietary intake including cookware is well below this threshold. The native rate of 1 in 10,000,000 per year is a conservative upper bound reflecting that (a) the antacid meta-analysis found no signal at dietary doses higher than cookware exposure, (b) no population-level cohort has detected attributable AD from cookware, and (c) the dialysis encephalopathy mechanism is acute toxicity via parenteral exposure, not chronic dietary exposure. Lifetime estimate: 1 - (1 - 1/10,000,000)^59 ≈ 5.9 x 10⁻⁶ ≈ 1 in 170,000. This is an upper-bound placeholder, not a measured value -- the true attributable risk from cookware specifically may be zero.\n","uncertainty":{"low":1e-7,"high":0.00005},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26592479/","title":"Chronic exposure to aluminum and risk of Alzheimer's disease: A meta-analysis","publisher":"Neuroscience Letters / Wang, Wei, Zeng, Liu, Du, Zhang","source_type":"peer_reviewed","statistic":"OR 1.71 (95% CI 1.35-2.18) for chronic aluminum exposure and Alzheimer's; 8 studies, 10,567 individuals","excerpt":"\"Individuals chronically exposed to Al were 71% more likely to develop AD (OR: 1.71, 95% CI, 1.35-2.18). Chronic Al exposure is associated with increased risk of AD.\"\n","source_date":"2016-01-12","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260416051135/https://pubmed.ncbi.nlm.nih.gov/26592479/","calculation_notes":"Wang et al. 2016 pooled 8 cohort and case-control studies covering chronic aluminum exposure. The exposure categories include occupational settings (aluminum smelters, welders), drinking water from high-aluminum watersheds, and contaminated industrial areas. None of the included studies isolate consumer cookware or foil as the exposure pathway. The 1.71 OR applies to the chronic-exposure umbrella; it does not translate to cookware-specific risk. The meta-analysis is the strongest available authoritative estimate of the general aluminum-AD association and is included to give the most generous reading of the literature. Even at face value, occupational chronic exposure doses are orders of magnitude higher than cookware-derived dietary intake.\n","independence_note":"Wang 2016 synthesizes studies from China, Europe, and North America with heterogeneous exposure definitions. Methodologically independent from Virk & Eslick 2015 below (different inclusion criteria, antacid focus vs general chronic exposure).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26098935/","title":"Brief Report: Meta-analysis of Antacid Use and Alzheimer's Disease","publisher":"Journal of Alzheimer's Disease / Virk, Eslick","source_type":"peer_reviewed","statistic":"OR 1.0 (95% CI 0.8-1.2) for antacid use and Alzheimer's risk; 7 case-control + 2 cohort studies, 6,310 participants","excerpt":"\"Regular antacid use was not associated with Alzheimer's disease. Case-control studies: odds ratio = 1.0; 95% confidence interval = 0.8, 1.2. Cohort studies: relative risk = 0.8; 95% confidence interval = 0.4, 1.8.\"\n","source_date":"2015-06-19","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20251207141718/https://pubmed.ncbi.nlm.nih.gov/26098935/","calculation_notes":"Aluminum-containing antacids (Maalox, Mylanta, etc.) deliver oral aluminum doses several orders of magnitude higher than typical cookware-derived dietary aluminum. A high-dose antacid user consumes roughly 800-5,000 mg of aluminum hydroxide per day; cookware contributes 1-10 mg/day under normal conditions (foil and acidic foods at the upper end). The Virk & Eslick meta-analysis found no association at the much higher antacid exposure level, which constrains the plausible upper bound for the much lower cookware exposure. This is the load-bearing source for framing the cookware-AD link as unsupported by evidence: if the higher dietary aluminum dose shows no signal, the lower one is even less likely to.\n","independence_note":"Virk & Eslick used independent inclusion criteria from Wang 2016 (focused specifically on aluminum-containing antacids and pharmaceutical sources). Same Sydney research group published companion analyses on occupational aluminum (PMID 26247643, also null).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/31958088/","title":"Aluminum and Amyloid-beta in Familial Alzheimer's Disease","publisher":"Journal of Alzheimer's Disease / Mold, Linhart, Gomez-Ramirez et al.","source_type":"peer_reviewed","statistic":"Aluminum co-located with amyloid-beta plaques in brain tissue of familial AD donors; significantly higher Al levels than controls","excerpt":"\"We found significantly higher levels of aluminum in brain tissues in donors with familial Alzheimer's disease than in control tissues. Aluminum and amyloid-beta were co-located in senile plaques as well as vasculature, the latter resembling cerebral amyloid angiopathy.\"\n","source_date":"2020-01-21","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260217022631/https://pubmed.ncbi.nlm.nih.gov/31958088/","calculation_notes":"Mold et al. examined brain tissue from a Colombian cohort of familial Alzheimer's donors (carriers of the PSEN1 E280A mutation). The finding is real and reproducible across the Mold group's series (2020, 2021), but the causal direction is unresolved. Two interpretations remain in the literature: (a) aluminum exposure accelerates amyloid deposition, or (b) the disrupted blood-brain barrier and protein aggregation in AD cause aluminum to accumulate as a downstream consequence. Familial AD also starts decades before sporadic AD and the cohort is dominated by a single genetic mutation. The work does not establish that dietary or cookware aluminum drives sporadic AD in the general population. Included for caveat-completeness: the brain- aluminum evidence is the strongest argument against fully dismissing the hypothesis.\n","independence_note":"Mold et al. work is histopathology on post-mortem brain tissue, methodologically distinct from the epidemiologic exposure meta-analyses above. The Keele University group is the most prolific producer of brain-aluminum data; the underlying cohort is genetically distinct from Wang 2016 study populations.\n"}],"comparison_anchors":[{"label":"Lifetime Alzheimer's risk (US adult, baseline)","lifetime_us_adult":0.11},{"label":"Cancer from nonstick (Teflon) cookware (lifetime, US adult)","lifetime_us_adult":0.0000059},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"regional_breakdown":[{"region":"US adult (baseline AD lifetime risk, all causes)","probability":0.11,"notes":"Alzheimer's Association 2024 Facts and Figures; baseline includes all environmental and genetic exposures combined"},{"region":"US adult (attributable to consumer aluminum cookware/foil specifically)","probability":0.0000059,"notes":"Upper-bound placeholder; no published cohort isolates this exposure pathway"},{"region":"Pre-1990 hemodialysis patients with aluminum-contaminated dialysate","probability":0.3,"notes":"Dialysis encephalopathy from parenteral aluminum exposure was a documented condition; the modern dialysis water-treatment standards eliminated this pathway by the early 1990s"}],"personal_factor_multipliers":[{"factor":"Occupational chronic aluminum exposure (smelter, welder, mining)","multiplier":1.7,"notes":"Wang 2016 meta-analysis: OR 1.71 (95% CI 1.35-2.18) for chronic occupational and high-contamination exposure. Below the 2.0 threshold for typical inclusion but the strongest published signal in the aluminum-AD literature and the load-bearing occupational reference. Cookware exposure is orders of magnitude lower than this cohort's intake.\n"},{"factor":"Pre-1990 hemodialysis with aluminum-contaminated dialysate","multiplier":50,"notes":"Dialysis encephalopathy was a documented neurological syndrome (memory loss, dementia- like symptoms, death) in chronic dialysis patients exposed to aluminum-contaminated dialysate prior to water-treatment reforms in the 1980s. The modern dialysis water standard (Aluminum < 10 μg/L) eliminated this pathway. The multiplier reflects an era-specific exposure that no longer exists in regulated healthcare systems; included as a historical reference point for why \"aluminum and brain\" became a credible concern in the first place.\n"},{"factor":"APOE ε4 carrier status (one or two copies)","multiplier":3,"notes":"One APOE ε4 allele increases lifetime AD risk roughly 3x; two copies roughly 8-12x. This is not an aluminum-mediated mechanism, but it dominates personal AD risk for carriers and is the single largest genetic risk factor. Included so readers can calibrate cookware-attributable risk against genetic risk: the latter is vastly larger and is the relevant comparison.\n"}],"short_label":"Aluminum & Alzheimer's","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers Alzheimer's disease risk specifically attributable to consumer use of aluminum cookware and foil at normal cooking conditions. It does not cover: (a) chronic occupational aluminum exposure in smelting, welding, or mining, where Wang 2016 found a modest population-level signal; (b) pre-1990 dialysis encephalopathy from contaminated dialysate, a real but now-eliminated parenteral exposure; or (c) familial early-onset AD where the Mold group has documented elevated brain aluminum, with causal direction unresolved. The normalized probability is a conservative upper bound, not a measured value. No epidemiological study has detected attributable AD from consumer cookware use, and the Virk & Eslick 2015 antacid meta-analysis -- examining oral aluminum doses much higher than cookware exposure -- found no signal. Surveillance gaps: chronic low-dose dietary aluminum is hard to isolate from background dietary variability and from occupational and pharmaceutical sources. The brain-aluminum findings in AD post-mortem tissue are real but their causal direction is contested.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single aluminum saucepan on a clean kitchen surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/aluminum-cookware-alzheimers","api_url":"https://likelier.app/api/fears/aluminum-cookware-alzheimers.json"},{"slug":"bioplastic-pla-health-harm","question":"What are the odds of getting sick from PLA bioplastic or \"compostable\" food packaging?","quick_answer":"The odds of a documented illness from PLA or compostable food packaging are about 1 in 170,000 over a lifetime. Chemical contaminants in food rank among the top concerns for nearly half of US adults, yet this particular material sits near the bottom of that list.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"PLA (polylactic acid) is the most common \"compostable\" plastic on the consumer market — the material in clear plant-based cups, deli containers, cutlery, and packaging marketed as eco-friendly. Public perception of bioplastics has bifurcated. On one side, the \"compostable\" label carries an implicit halo of safety: if it breaks down in a compost bin, surely it must be safer to eat from than conventional plastic. On the other side, rising awareness of microplastics and chemical migration from food packaging has begun to spill onto bioplastics too, with consumer media asking whether the additives, dyes, and processing aids in commercial compostable products carry their own undocumented hazards. IFIC's 2025 Food and Health Survey found that 47% of US adults rank cancer-causing chemicals in food among their top three safety concerns, with food-packaging chemicals included in that bucket; PLA-specific concern is a small subset, but the broader category fear is large.\n","rough_estimate":"47% of US adults rank cancer-causing chemicals in food among their top-3 concerns; PLA and compostable packaging fall under this chemical-contaminant umbrella","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 47% rank cancer-causing chemicals in food among top-3 concerns","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~1 per 10,000,000 US adults per year attributable illness from PLA food-contact use","numerator":1,"denominator":10000000,"unit":"per year (attributable serious harm)","population":"US adults using PLA-based food packaging at normal consumer temperatures"},"normalized":{"lifetime_us_adult":0.0000059,"display":"~1 in 170,000 lifetime (US adult)","log_value":-5.23,"assumptions":"No published epidemiological cohort has measured attributable cancer, endocrine disease, or other illness from consumer PLA food-contact use. The 1995 Conn et al. safety assessment in Food and Chemical Toxicology concluded that PLA is \"Generally Recognized As Safe\" for food-contact use, on the basis that the principal migrants (lactic acid, lactide monomer, and lactoyllactic acid) all hydrolyze to lactic acid — a substance the human body produces metabolically and that has decades of food-additive GRAS status. Mutsuga et al. (2008) measured total migration of 0.28-15 µg/cm² at 40°C for 180 days (typical consumer storage), and Auras et al. (2004) summarized that PLA migration is \"much lower than any of the current average dietary lactic acid intake values allowed by several governmental agencies.\" The native rate of 1 in 10,000,000 per year is an upper-bound placeholder reflecting that: (a) the PLA polymer itself shows minimal migration at room-temperature consumer use, (b) the migrants are biologically benign, and (c) no population-level disease signal has been detected. Lifetime estimate: 1 − (1 − 1/10,000,000)^59 ≈ 5.9 × 10⁻⁶ ≈ 1 in 170,000. This is a conservative upper bound for the polymer alone, not a measured value — the true attributable risk from PLA polymer migration may be effectively zero. Commercial compostable packaging that combines PLA with dyes, plasticizers, and processing additives has substantially less public migration data than the pure polymer; the upper end of the uncertainty band reflects this blend-vs-polymer evidence gap. Zimmermann et al. (2020) showed in an ecotoxicology assay that bio-based and biodegradable plastics \"can be as toxic as their conventional counterparts\" (they attributed PLA's toxicity to the particle itself, not to extractable additives), which underscores that a \"compostable\" label is not itself a guarantee of low toxicity.\n","uncertainty":{"low":1e-7,"high":0.0001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/7737601/","title":"Safety assessment of polylactide (PLA) for use as a food-contact polymer","publisher":"Food and Chemical Toxicology / Conn, Kolstad, Borzelleca, Dixler, Filer, LaDu & Pariza","source_type":"peer_reviewed","statistic":"PLA is concluded to be safe and Generally Recognized As Safe (GRAS) for its intended uses as a food-contact polymer; migrants (lactic acid, lactide, lactoyllactic acid) convert to lactic acid in food","excerpt":"\"It is concluded that PLA is safe and 'Generally Recognized As Safe' for its intended uses as a polymer for fabricating articles that will hold and/or package food.\"\n","source_date":"1995-04-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20250308012934/https://pubmed.ncbi.nlm.nih.gov/7737601/","calculation_notes":"Conn et al. (1995) performed migration testing on PLA samples under conditions simulating realistic worst-case use in housewares and food packaging. The authors, drawn from a panel including FDA-recognized toxicology experts (Borzelleca, Filer, LaDu, Pariza), identified the three principal migrants and noted that all hydrolyze to lactic acid, which is endogenous to human metabolism and has long-standing GRAS status as a direct food additive. The verbatim GRAS conclusion anchors the lower end of the uncertainty band. This is the foundational safety assessment that supports PLA's use in US food-contact applications under FDA's general regulatory framework for food-contact polymers (21 CFR 177).\n","independence_note":"Independent academic safety assessment authored by an expert panel; predates and is methodologically distinct from the Mutsuga and Auras analyses below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/18608493/","title":"Migration of lactic acid, lactide and oligomers from polylactide food-contact materials","publisher":"Food Additives & Contaminants Part A / Mutsuga, Kawamura & Tanamoto","source_type":"peer_reviewed","statistic":"Total migration of lactic acid, lactide, and oligomers from PLA sheets at 40°C for 180 days was 0.28-15.00 µg/cm²; at 60°C for 10 days, migration rose to 0.73-2840 µg/cm²","excerpt":"\"At 40°C for 180 days, the total of lactic acid, lactide and oligomers migration levels were 0.28-15.00 microg cm(-2). At 60°C for 10 days, the total migration levels were increased to 0.73-2840 microg cm(-2).\"\n","source_date":"2008-10-01","source_accessed":"2026-05-30","archive_url":"https://web.archive.org/web/20260531014559/https://pubmed.ncbi.nlm.nih.gov/18608493/","calculation_notes":"Mutsuga et al. tested PLA sheets used for lunch boxes and fresh-food packaging in Japan, measuring lactic acid, lactide, and oligomer migration by LC/MS. The 40°C figure brackets typical consumer storage (refrigeration through warm-kitchen conditions) and confirms low migration well within food-contact safety norms. The 60°C result is the key signal for the \"compostable cup with hot coffee\" pathway: PLA's glass transition temperature is approximately 55-60°C, and above it the polymer softens and migration accelerates by two to three orders of magnitude. This study is the empirical basis for the hot-beverage multiplier in the personal factors section.\n","independence_note":"Independent migration study by Japan's National Institute of Health Sciences; uses standard LC/MS quantification and does not depend on the Conn 1995 expert review or the Auras 2004 polymer overview.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15468294/","title":"An overview of polylactides as packaging materials","publisher":"Macromolecular Bioscience / Auras, Harte & Selke","source_type":"peer_reviewed","statistic":"The amount of lactic acid and its derivatives that migrate to food simulant solutions from PLA is much lower than any of the current average dietary lactic acid intake values allowed by several governmental agencies","excerpt":"\"The amount of lactic acid and its derivatives that migrate to food simulant solutions from PLA is much lower than any of the current average dietary lactic acid intake values allowed by several governmental agencies. Thus, PLA is safe for use in fabricating articles for contact with food.\"\n","source_date":"2004-09-13","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260402143519/https://pubmed.ncbi.nlm.nih.gov/15468294/","calculation_notes":"Auras, Harte & Selke synthesized the published migration and barrier-property data on PLA packaging across temperatures and food-simulant conditions. Their summary conclusion — that PLA migration is dwarfed by allowed dietary lactic acid intake from naturally occurring sources (yogurt, fermented foods, endogenous metabolism) — provides the dose-context that makes the migration measurements meaningful. Used here alongside Conn and Mutsuga to triangulate the polymer-side safety conclusion.\n","independence_note":"Independent peer-reviewed review by Michigan State University packaging researchers; synthesizes Conn 1995, Mutsuga et al., and other primary data but reaches its conclusion through its own analytical framework.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32871484/","title":"What are the drivers of microplastic toxicity? Comparing the toxicity of plastic chemicals and particles to Daphnia magna","publisher":"Environmental Pollution / Zimmermann, Göttlich, Oehlmann, Wagner & Völker","source_type":"peer_reviewed","statistic":"Bio-based and biodegradable plastics, including PLA, can be as toxic in ecotoxicology assays as conventional plastics; the study attributes PVC's effects to its chemicals but found PUR and PLA effects were induced by the particle itself rather than extractable chemicals","excerpt":"\"The latter indicates that bio-based and biodegradable plastics can be as toxic as their conventional counterparts.\"\n","source_date":"2020-12-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260531014723/https://pubmed.ncbi.nlm.nih.gov/32871484/","calculation_notes":"Zimmermann et al. tested PLA alongside PVC and polyurethane in Daphnia magna assays and found that bio-based and biodegradable plastics can be as toxic as their conventional counterparts. The authors attribute PVC's effects to the chemicals in the material, but note that \"effects of PUR and PLA plastics were induced by the mere particle\" — for PLA the toxicity was driven by the particle itself rather than by extractable chemicals or additives. This is cited here only for the top-line finding that \"compostable\" labeling does not guarantee a plastic is less toxic than conventional material. The assay is ecotoxicological, not a human-exposure study, so the finding is not directly convertible to consumer risk, and it is not used to quantify additive-migration risk from PLA blends — the polymer-vs-blend evidence gap captured in the uncertainty band rests on the thinness of public migration data for commercial PLA-blend formulations, not on this assay.\n","independence_note":"Independent academic ecotoxicology study by Goethe University Frankfurt; methodologically separate from the FDA regulatory framework and from the Conn, Mutsuga, and Auras polymer-migration studies cited above.\n"},{"url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","title":"Confidence in food safety hits record low: IFIC 2025 Food & Health Survey","publisher":"International Food Information Council (IFIC)","source_type":"reputable_reference","statistic":"47% of US adults rank cancer-causing chemicals among their top-3 food safety concerns; 46% rank pesticide residues; foodborne illness leads at 50%","excerpt":"\"Foodborne illness from bacteria, such as E. coli, Salmonella, or Listeria, tops the list of consumer food safety concerns, with half of Americans (50%) ranking it among their top three. Cancer-causing chemicals (47%), pesticides (46%), and heavy metals (41%) follow closely.\"\n","source_date":"2025-05-21","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260531014747/https://ific.org/media/confidence-in-food-safety-hits-record-low/","calculation_notes":"IFIC commissioned an annual survey of 3,000 US adults aged 18-80, fielded March 13-27, 2025 and weighted to US Current Population Survey demographics. The 47% cancer-causing-chemicals figure is the broadest umbrella that includes packaging-derived chemical concerns, of which PLA and compostable packaging is a small subset. Used as the perceived-side anchor; not used in the normalized probability calculation.\n","independence_note":"IFIC is an industry-supported nonprofit; methodology and weighting are publicly disclosed. Used here only for the perceived-side concern level, not for any safety or migration claim.\n"}],"comparison_anchors":[{"label":"Plastic food container leaching harm (lifetime, US adult)","lifetime_us_adult":0.000059},{"label":"Pesticide residue harm (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"personal_factor_multipliers":[{"factor":"Hot beverages in compostable PLA cups (above polymer Tg ~55-60°C)","multiplier":3,"notes":"PLA's glass transition temperature is approximately 55-60°C. Mutsuga et al. (2008) measured a roughly two-to-three-orders-of-magnitude jump in lactic-acid and lactide migration when PLA was held at 60°C for 10 days versus 40°C for 180 days. Hot coffee or tea served in a compostable PLA cup briefly puts the polymer above its Tg. The multiplier reflects increased migration of an essentially benign substance (lactic acid is endogenous and dietary), not a demonstrated cancer or endocrine pathway — but exposure rises.\n"},{"factor":"Microwaving food in PLA-based containers","multiplier":3,"notes":"PLA softens and deforms above ~55-60°C. Microwave heating routinely exceeds this and reaches temperatures where Mutsuga's accelerated-migration regime applies. Most PLA containers are not labeled microwave-safe for this reason. The multiplier reflects greater migration; the absolute risk remains low because the migrants themselves are biologically benign.\n"},{"factor":"Use of commercial 'compostable' packaging with proprietary additives","multiplier":2,"notes":"Public migration data on commercial PLA-blend formulations — which combine the PLA polymer with proprietary plasticizers, dyes, and processing aids — is far thinner than on the polymer in isolation. The multiplier reflects that evidence gap rather than any measured additive-migration figure. It is conservative and uncertain: the true figure could be 1x (if additives are inert) or higher if specific formulations use poorly characterized plasticizers, dyes, or processing aids. (Zimmermann et al. 2020 found \"compostable\" plastics can be as toxic as conventional ones in an ecotoxicology assay, but attributed PLA's toxicity to the particle itself rather than to extractable additives, so it does not quantify this migration pathway.)\n"},{"factor":"Exclusive use of glass or stainless steel for food storage and reheating","multiplier":0.5,"notes":"Eliminates the PLA migration pathway entirely. The multiplier is not zero because the baseline attributable risk from PLA polymer migration is already near zero in the published data; removing it produces a modest reduction in an already small figure. Most consumer exposure to PLA-related additive uncertainty comes from single-use foodservice packaging rather than home storage.\n"}],"short_label":"PLA bioplastic harm","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses health harm from chemical migration during normal consumer use of PLA-based food-contact items at room or refrigerated temperatures, and from above-Tg use such as hot beverages in compostable cups or microwave reheating. It does not cover: ecological or marine-debris questions (PLA persistence in seawater is a separate research area), occupational exposure during PLA manufacturing, or the distinct microplastic exposure pathway covered in the microplastics-health-harm entry. The strongest published evidence covers the PLA polymer in isolation — Conn 1995, Mutsuga 2008, and Auras 2004 collectively establish that polymer migration is low and biologically benign. Commercial \"compostable\" packaging is typically a PLA blend with proprietary dyes, plasticizers, and processing additives, and public migration data on those commercial blends is far thinner than on the pure polymer. Zimmermann et al. (2020) found in an ecotoxicology assay that bio-based plastics can be as toxic as conventional ones, though for PLA that toxicity was driven by the particle itself rather than by extractable additives. No epidemiological cohort has isolated attributable human disease from either pure PLA or commercial PLA blends at consumer exposure levels. The 1-in-170,000 lifetime figure is a conservative upper bound for the polymer; the polymer-vs-blend evidence asymmetry is the largest source of remaining uncertainty. Compare with plastic-food-container-leaching for the conventional-plastic parallel and nonstick-cookware-cancer for the regulatory-floor pattern this entry follows.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single translucent compostable cup on a pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/bioplastic-pla-health-harm","api_url":"https://likelier.app/api/fears/bioplastic-pla-health-harm.json"},{"slug":"nonstick-cookware-cancer","question":"What are the odds of getting cancer from nonstick (Teflon) cookware?","quick_answer":"The lifetime odds of cancer from normal use of modern nonstick cookware are about 1 in 170,000 — and that is a conservative upper bound; the true attributable risk may be zero, since PFOA was removed from US Teflon by 2013 and PTFE itself is inert. Many consumers treat the pans as a moderate cancer risk, well above the evidence.\n","category":"cancer","tags":["food","household"],"no_reliable_estimate":false,"perceived":{"description":"Nonstick cookware became a household anxiety after the DuPont PFOA scandal and the 2019 film \"Dark Waters.\" Public perception conflates three distinct things: PFOA (a processing aid phased out of Teflon manufacturing by 2013 and classified as a Group 1 carcinogen by IARC in 2023), PFAS contamination of drinking water near factories, and the PTFE coating itself. The result is a widespread belief that cooking on a nonstick pan exposes the user to a meaningful cancer risk. Consumer surveys and cookware marketing trends confirm this: \"PFOA-free\" and \"PFAS-free\" labeling commands a price premium, and a significant fraction of consumers have switched to cast iron or stainless steel specifically to avoid perceived carcinogenic exposure.\n","rough_estimate":"Many consumers treat nonstick cookware as a moderate cancer risk","kind":"intuition"},"native":{"display":"~0 attributable cancer cases per 100,000 consumer-years of normal PTFE cookware use","numerator":1,"denominator":10000000,"unit":"per year of normal consumer PTFE cookware use","population":"US adults using post-2013 PFOA-free nonstick cookware at normal cooking temperatures"},"normalized":{"lifetime_us_adult":0.0000059,"display":"~1 in 170,000 lifetime (US adult)","log_value":-5.23,"assumptions":"No epidemiological study has measured attributable cancer risk from consumer PTFE cookware use. The American Cancer Society states there are \"no proven risks to humans from using cookware coated with Teflon.\" PFOA, the carcinogenic processing aid, was eliminated from US Teflon manufacturing by 2013 under the EPA PFOA Stewardship Program. Modern PTFE cookware does not contain PFOA. The native rate of 1 in 10,000,000 per year is an upper-bound placeholder reflecting that: (a) PTFE itself is biologically inert and not classified as carcinogenic by any agency, (b) FDA considers PTFE coatings safe because the polymer is tightly bound and migrates negligibly into food, and (c) no population-level signal has been detected. Lifetime estimate: 1 − (1 − 1/10,000,000)^59 ≈ 5.9 × 10⁻⁶ ≈ 1 in 170,000. This is a conservative upper bound, not a measured value — the true attributable risk may be zero.\n","uncertainty":{"low":1e-7,"high":0.00003},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cancer.org/cancer/risk-prevention/chemicals/teflon-and-perfluorooctanoic-acid-pfoa.html","title":"Teflon and Perfluorooctanoic Acid (PFOA)","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"No proven risks to humans from using cookware coated with Teflon or other non-stick surfaces","excerpt":"\"There are no proven risks to humans from using cookware coated with Teflon (or other non-stick surfaces).\"\n","source_date":"2024-01-17","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420044515/https://www.cancer.org/cancer/risk-prevention/chemicals/teflon-and-perfluorooctanoic-acid-pfoa.html","calculation_notes":"The ACS page distinguishes between PFOA (the processing chemical historically used to manufacture PTFE coatings) and PTFE (the finished coating on cookware). ACS notes that PFOA has been associated with cancer in highly exposed populations (factory workers, contaminated communities) but that consumer cookware use does not produce meaningful PFOA exposure, especially since the 2013 phase-out. The statement \"no proven risks\" from consumer cookware use is the basis for placing the native rate at or near zero. The upper- bound estimate of 1 in 10,000,000 per year reflects this \"not proven but not formally disproven\" epistemic state.\n"},{"url":"https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/fact-sheet-20102015-pfoa-stewardship-program","title":"Fact Sheet: 2010/2015 PFOA Stewardship Program","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"Eight major manufacturers committed to eliminating PFOA from emissions and products by 2015; Teflon products have been PFOA-free since 2013","excerpt":"\"EPA invited eight major companies in the per- and polyfluoroalkyl substances industry to join a global stewardship program to work toward eliminating these chemicals from emissions and products by 2015.\"\n","source_date":"2023-03-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503092314/https://www.epa.gov/assessing-and-managing-chemicals-under-tsca/fact-sheet-20102015-pfoa-stewardship-program","calculation_notes":"The EPA PFOA Stewardship Program resulted in all eight participating companies (including DuPont/Chemours) achieving the goal of eliminating PFOA from products by 2015. Teflon- branded cookware specifically transitioned to PFOA-free formulations by 2013. This is the key inflection point: the carcinogenic concern was PFOA exposure during manufacturing and from environmental contamination, not from the finished PTFE polymer. Post-2013 cookware does not contain the substance that drove the cancer signal in epidemiological studies.\n"},{"url":"https://www.iarc.who.int/news-events/iarc-monographs-evaluate-the-carcinogenicity-of-perfluorooctanoic-acid-pfoa-and-perfluorooctanesulfonic-acid-pfos/","title":"IARC Monographs evaluate the carcinogenicity of PFOA and PFOS","publisher":"International Agency for Research on Cancer (WHO)","source_type":"govt_report","statistic":"PFOA classified as carcinogenic to humans (Group 1); PFOS classified as possibly carcinogenic (Group 2B)","excerpt":"\"The Working Group classified perfluorooctanoic acid (PFOA) as carcinogenic to humans (Group 1) on the basis of sufficient evidence for cancer in experimental animals and strong mechanistic evidence in exposed humans.\"\n","source_date":"2023-12-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260503082033/https://www.iarc.who.int/news-events/iarc-monographs-evaluate-the-carcinogenicity-of-perfluorooctanoic-acid-pfoa-and-perfluorooctanesulfonic-acid-pfos/","calculation_notes":"IARC's Group 1 classification of PFOA (November 2023, published in Lancet Oncology) is based on evidence from highly exposed populations — principally the C8 Science Panel cohort of ~32,500 people exposed via contaminated drinking water near DuPont's Washington Works plant in West Virginia, and occupational cohorts of factory workers. The classification applies to PFOA the chemical, not to PTFE the polymer. PTFE is the end product; PFOA was a processing aid used during its manufacture that has since been eliminated. Consumer cookware use does not produce PFOA exposure. The IARC classification is essential context because it explains why the public fear is so high — \"Group 1 carcinogen\" is the same category as tobacco and asbestos — while the actual consumer exposure pathway (cooking on PTFE) involves a different substance entirely.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3855514/","title":"Perfluorooctanoic Acid (PFOA) Exposures and Incident Cancers among Adults Living Near a Chemical Plant","publisher":"Environmental Health Perspectives / Barry, Winquist & Steenland","source_type":"peer_reviewed","statistic":"Increased kidney and testicular cancer incidence in communities with PFOA-contaminated water near DuPont's Washington Works plant","excerpt":"\"Estimated cumulative serum PFOA concentrations were positively associated with kidney and testicular cancer [hazard ratio (HR) = 1.10; 95% CI: 0.98, 1.24 and HR = 1.34; 95% CI: 1.00, 1.79, respectively, for 1-unit increases in ln-transformed serum PFOA].\"\n","source_date":"2013-12-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420044631/https://pmc.ncbi.nlm.nih.gov/articles/PMC3855514/","calculation_notes":"Barry, Winquist & Steenland analyzed cancer incidence among adults living near and exposed to PFOA from DuPont's Washington Works facility (the C8 study area). The cancer signal — kidney and testicular cancer — was found in people with serum PFOA levels 5-50x the national median, resulting from decades of drinking contaminated water, not from cooking on nonstick pans. The C8 Science Panel found \"probable links\" between PFOA exposure and kidney cancer and testicular cancer at these elevated serum levels. This study is included because it is the primary evidence base that drives public fear of \"Teflon cancer\" — but the exposure pathway (contaminated drinking water at industrial concentrations) is categorically different from consumer cookware use. The site's pfas-tap-water entry covers the water contamination pathway separately.\n"},{"url":"https://www.poison.org/articles/teflon-flu","title":"Protect Yourself from Teflon Flu","publisher":"National Capital Poison Center (Poison Control)","source_type":"reputable_reference","statistic":"Polymer fume fever from overheated PTFE causes flu-like symptoms lasting 1-2 days; not associated with cancer","excerpt":"\"Symptoms of Teflon flu include chills, fever, fatigue, headache, body aches, and occasional chest tightness and airway irritation. Symptoms generally occur within a few hours after being exposed to the fumes and usually resolve within 1 to 2 days.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251115012352/https://www.poison.org/articles/teflon-flu","calculation_notes":"Polymer fume fever (\"Teflon flu\") occurs when PTFE is heated above ~260°C (500°F) and releases pyrolysis products. US Poison Control Centers reported ~265 suspected cases in 2023, up from an average of 9 per year between 2006-2012. The condition is self-limiting (resolves in 24-48 hours) and is not carcinogenic — it is an acute inhalation fever, not a chronic exposure pathway. Severe lung injury is rare and occurs almost exclusively when pans are heated to extreme temperatures (>450°C/842°F) in poorly ventilated spaces. This source is included to distinguish the real but minor acute hazard (polymer fume fever) from the perceived but unsupported chronic hazard (cancer from normal cookware use).\n"}],"comparison_anchors":[{"label":"Cancer from artificial sweeteners (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"personal_factor_multipliers":[{"factor":"Pre-2013 PFOA-era cookware still in use","multiplier":3,"notes":"Cookware manufactured before the 2013 PFOA phase-out may contain trace residual PFOA in the coating. Migration into food is still minimal at normal cooking temperatures, but the multiplier reflects higher theoretical exposure relative to PFOA-free cookware.\n"},{"factor":"Factory worker (PFOA manufacturing, pre-phase-out)","multiplier":500,"notes":"C8 Science Panel data showed relative risks of 3.0 for testicular cancer and elevated kidney cancer in the highest PFOA exposure quartile. This multiplier applies to the occupational pathway, not consumer cookware use. Most of these workers were exposed before 2015 and the pathway no longer exists in US manufacturing.\n"},{"factor":"Bird owner (PTFE fume exposure to pets)","multiplier":1,"notes":"PTFE fumes are acutely lethal to birds but the human cancer multiplier is unchanged. The concern for bird owners is pet mortality from polymer fume fever, not human cancer. Birds have highly efficient respiratory systems that make them uniquely vulnerable to inhaled toxins at concentrations harmless to humans.\n"},{"factor":"Habitual high-heat cooking (>260°C / 500°F)","multiplier":2,"notes":"Cooking above the PTFE degradation threshold releases pyrolysis products that cause polymer fume fever. The acute health risk increases but there is no evidence these fumes are carcinogenic. Multiplier reflects increased general respiratory irritation, not a demonstrated cancer pathway.\n"}],"short_label":"Teflon cookware cancer","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers cancer risk specifically attributable to consumer use of PTFE-coated nonstick cookware at normal cooking temperatures. It does not cover: (a) PFOA/PFAS contamination of drinking water from industrial discharge, which is addressed in the pfas-tap-water entry; (b) occupational exposure in PFOA manufacturing facilities, where the cancer signal was actually observed; or (c) polymer fume fever from overheating PTFE, which is a real but non-carcinogenic acute condition. The normalized probability is a conservative upper bound, not a measured value — no epidemiological study has detected attributable cancer from consumer cookware use, and the true risk may be zero. IARC's Group 1 classification of PFOA applies to the chemical itself at high exposure levels, not to finished PTFE cookware from which PFOA has been eliminated.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single nonstick frying pan on a neutral kitchen surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/nonstick-cookware-cancer","api_url":"https://likelier.app/api/fears/nonstick-cookware-cancer.json"},{"slug":"dog-bite-fatal","question":"What are the odds of being killed by a dog?","quick_answer":"About 1 in 142,000 over a lifetime for a US adult — roughly 40 fatal dog attacks a year nationwide. That is far rarer than the 1-in-a-few-thousand bucket most people place it in, alongside sharks and bears.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"We don’t yet have a rigorous survey that isolates “fear of being killed by a dog” from the much broader category of fear of dogs in general. Casual reporting and self-report polls consistently find cynophobia (fear of dogs) among the more common specific animal phobias, but the fatal-outcome version of the fear is almost never measured on its own. Anecdotally, most people asked to guess the number place it several orders of magnitude too high, in the same intuitive bucket as shark attacks and bear maulings.\n","rough_estimate":"most people guess 1 in a few thousand lifetime, based on informal asking","kind":"intuition"},"native":{"display":"~40 fatal dog bite/strike incidents per year, United States","numerator":40,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.00000704,"display":"1 in ~142,000 lifetime (US adult)","log_value":-5.15,"assumptions":"Uses ~40 fatal dog-bite-or-strike incidents per year as the long-run US central estimate (CDC WISQARS / NCHS mortality records under ICD-10 W54, “Bitten or struck by dog”), divided by a US population of ~335 million, then compounded over 59 years of remaining US adult life. The annual count has ranged roughly 30–80 over the past two decades, with a recent upward drift; the uncertainty band reflects that range.\n","uncertainty":{"low":0.0000053,"high":0.0000141},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/24299544/","title":"Co-occurrence of potentially preventable factors in 256 dog bite-related fatalities in the United States (2000-2009)","publisher":"Journal of the American Veterinary Medical Association (JAVMA) / Patronek, Sacks, Delise, Cleary, Marder","source_type":"peer_reviewed","statistic":"256 dog bite-related fatalities in the US over 2000-2009 (~25.6 per year)","excerpt":"\"Major co-occurrent factors included absence of an able-bodied person to intervene (n = 223 [87.1%]), incidental or no familiar relationship of victims with dogs (218 [85.2%]), owner failure to neuter dogs (216 [84.4%]), compromised ability of victims to interact appropriately with dogs (198 [77.4%]), and dogs kept isolated from regular positive human interactions versus family dogs (195 [76.2%]). Four or more of these factors co-occurred in 206 (80.5%) deaths.\"\n","source_date":"2013-12-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165935/https://pubmed.ncbi.nlm.nih.gov/24299544/","calculation_notes":"Patronek et al. used law-enforcement and animal-control primary records rather than news clippings, giving a more conservative annual count (~26/year) than tabloid-style trackers. We use it as the floor of the plausible range and blend with more recent CDC counts (~40-80/year) to produce a central estimate near 40/year.\n","independence_note":"This study is methodologically independent of CDC WISQARS: Patronek’s team compiled case histories from homicide detectives and animal control agencies, not ICD-10 death certificates, so it functions as independent corroboration of order of magnitude.\n"},{"url":"https://wisqars.cdc.gov/","title":"WISQARS — Web-based Injury Statistics Query and Reporting System","publisher":"US Centers for Disease Control and Prevention (CDC) / National Center for Injury Prevention and Control","source_type":"govt_report","statistic":"ICD-10 code W54 (“Bitten or struck by dog”) records on the order of 30-80 US deaths per year","excerpt":"\"An interactive, online collection of analysis tools for fatal, nonfatal, and cost of injury data. Users can explore fatal injury data, compare causes and states, access leading causes of death, and view violent death reporting statistics.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412092020/https://wisqars.cdc.gov/","calculation_notes":"WISQARS queries on ICD-10 W54 consistently return an annual US fatal count in the 30-80 range across recent years (with a long-run mean near 40). Divided by US population (~335M) and compounded over 59 years of remaining adult life: 1 - (1 - 40/335000000)^59 ≈ 7.04 × 10^-6, i.e. ~1 in 142,000 lifetime.\n","independence_note":"WISQARS draws from NCHS death certificates coded by ICD-10, which is a separate data stream from the primary-source case compilations used by Patronek et al.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7603431/","title":"The changing epidemiology of dog bite injuries in the United States, 2005-2018","publisher":"Injury Epidemiology (Springer Nature) / Tuckel & Milczarski","source_type":"peer_reviewed","statistic":"~337,000 US ED visits per year for dog bites, 2005-2013; highest rate in ages 5-9","excerpt":"\"Between 2005 and 2013, there were an average of 337,103 visits to emergency departments (ED) per year for dog bites. The modal category is the age group 5 to 9, followed by the age groups 0 to 4 and 10 to 14. Injuries are more prevalent among school-age children, inhabitants of less-densely populated areas, and residents of poorer neighborhoods.\"\n","source_date":"2020-10-21","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413170046/https://pmc.ncbi.nlm.nih.gov/articles/PMC7603431/","calculation_notes":"Used to bracket the denominator: roughly 337,000 ED-treated dog bites per year, of which ~40 end in death, gives a case-fatality ratio near 1 in 8,400 per ED-treated bite. Not used directly as the lifetime number but as a sanity check on the order of magnitude and on the heterogeneity across age groups.\n","independence_note":"Tuckel & Milczarski draw from NEISS emergency-department data, methodologically independent of the death-certificate pipeline used by WISQARS and Patronek."}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by bee/wasp/hornet sting (lifetime, US)","lifetime_us_adult":0.0001267},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Infant or toddler under age 2","multiplier":7,"notes":"CDC WISQARS and Patronek et al. 2013: children under age 5 account for approximately 50% of all fatal dog bites in the US despite being under 7% of the population; infants and toddlers under 2 are at the highest end of this elevation, with an estimated 7× per-capita fatality rate compared with the all-age average, driven by small body size, inability to escape, and frequent proximity to dogs."},{"factor":"Presence of multiple dogs in household","multiplier":5,"notes":"Patronek et al. 2013 (JAVMA): households with multiple dogs were present in a substantially disproportionate share of fatal incidents; multi-dog environments are associated with pack-behavior escalation and reduced owner control, with published literature estimating roughly 5× elevated fatality risk relative to single-dog household interactions."},{"factor":"Unfamiliar dog vs. known family pet","multiplier":2,"notes":"Patronek et al. 2013 (JAVMA): incidental or no familiar relationship between victim and dog was present in 85.2% of the 256 fatalities studied; by contrast, well-socialized family pets with habitual positive human contact accounted for a minority of deaths, yielding an approximately 2× higher fatality rate for encounters with unfamiliar or isolated dogs."}],"short_label":"Dog bite","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a population-level average over all US adults. Actual risk is highly heterogeneous: young children (ages 1-4), the very elderly, and people in households with multiple unsocialised or unneutered dogs face meaningfully higher per-year exposure, while adults with no regular contact with unfamiliar dogs face essentially zero. The annual US count has drifted upward in the most recent reporting years, and the uncertainty band reflects that recent increase.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty leather dog leash coiled on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/dog-bite-fatal","api_url":"https://likelier.app/api/fears/dog-bite-fatal.json"},{"slug":"residential-gas-leak","question":"What are the odds of dying in a residential gas leak explosion or fire?","quick_answer":"Dying in a residential gas-leak fire or explosion is about a 1 in 140,000 lifetime risk, from roughly 40 US deaths a year. The mercaptan smell and dramatic news coverage make it feel imminent, so most people overestimate the toll.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Fuel-gas leaks occupy a vivid spot in the public imagination. The smell of mercaptan -- the odorant added to natural gas and propane -- triggers immediate alarm in most households, and news coverage of gas explosions in apartment buildings or row houses reinforces the sense that a leak is a step away from catastrophe. Most people who use gas appliances have evacuated a building at some point on a suspected leak, which creates an outsized mental availability relative to the actual fatality count. The risk feels imminent in a way that lightning or carbon monoxide does not, because there is a sensory cue: a smell you can act on.\n","rough_estimate":"Most adults probably guess several hundred residential gas deaths per year -- closer to the actual figure for CO","kind":"intuition"},"native":{"display":"~40 US deaths per year from residential fuel-gas leak fires and explosions","numerator":1,"denominator":8250000,"unit":"per year","population":"US residents, all ages, deaths from natural gas or LP-gas ignited fires and gas distribution incidents"},"normalized":{"lifetime_us_adult":0.0000071,"display":"1 in ~140,000 lifetime (US adult)","log_value":-5.15,"assumptions":"Two data streams are combined. PHMSA reported 23 fatalities from natural gas distribution system incidents in 2023 (all causes across ~72 million customers). NFPA's propane-fires report (2012-2016) shows an annual average of 25 civilian deaths in home structure fires where LP-gas was the first material ignited. Combined and de-duplicated (some PHMSA distribution incidents that ignite a structure are already captured in NFPA residential fire data), a plausible central estimate is ~40 residential fuel-gas-ignited deaths per year -- natural gas and propane combined, fire/explosion mechanism only, excluding carbon monoxide asphyxiation from combustion products (covered separately). 40 / 330,000,000 = 1.21e-7 annual rate. Compounded over 59 years of remaining adult life: 1 - (1 - 1.21e-7)^59 = approximately 7.1e-6, or about 1 in 140,000. Annual uncertainty range 25-60 deaths gives normalized low ~4.5e-6, high ~1.1e-5.\n","uncertainty":{"low":0.0000045,"high":0.000011},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.phmsa.dot.gov/data-and-statistics/pipeline/pipeline-incident-20-year-trends","title":"Pipeline Incident 20 Year Trends -- Gas Distribution","publisher":"Pipeline and Hazardous Materials Safety Administration (PHMSA), US Department of Transportation","source_type":"govt_report","statistic":"In 2023 there were 613 gas distribution incidents with 23 fatalities and 39 injuries reported to PHMSA","excerpt":"\"In 2023, there were 613 gas distribution incidents, with 23 fatalities and 39 injuries reported.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260509175611/https://www.phmsa.dot.gov/data-and-statistics/pipeline/pipeline-incident-20-year-trends","calculation_notes":"PHMSA gas distribution covers the pipeline network from the city gate to the customer meter -- mains and service lines -- serving ~72 million customers across residential, commercial, and industrial uses. The 23 fatalities in 2023 include deaths directly caused by fires, explosions, and other mechanical failures originating on the distribution system; they exclude carbon monoxide deaths from combustion appliances (which occur beyond the meter) and propane/LP-gas incidents (covered by a different regulatory framework). Annual fatalities have ranged roughly 10-30 in recent years. Using 23 / 330,000,000 = 7.0e-8 per year; over 59 years: 1 - (1 - 7.0e-8)^59 ≈ 4.1e-6. This is the natural-gas-only lower bound; the entry's central estimate adds residential LP-gas deaths.\n","independence_note":"PHMSA data covers natural gas distribution infrastructure incidents only; does not overlap with NFPA's LP-gas residential fire data, which tracks fire ignition inside structures regardless of whether the gas reached the home via a distribution main or a propane tank.\n"},{"url":"https://www.propane101.com/propanestatistics.htm","title":"Propane Statistics -- Usage, Fire and Safety Statistics","publisher":"Propane 101 (citing NFPA 2012-2016 data)","source_type":"reputable_reference","statistic":"Annual average of 2,900 residential fires with LP-gas as first material ignited; 25 civilian deaths per year; 155 civilian injuries per year (NFPA, 2012-2016)","excerpt":"\"Residential structure fires with LP-Gas as first material ignited: 2,900. Civilian deaths: 25 per year. Civilian injuries: 155 per year.\"\n","source_date":"2018-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20230810155241/https://www.propane101.com/propanestatistics.htm","calculation_notes":"NFPA's 2012-2016 propane-fires data gives 25 LP-gas-related residential fire deaths per year. This covers only home structure fires where propane was the first material ignited and is a residential-only count. Adding propane's ~25/year to PHMSA natural gas distribution's ~23/year gives a combined fuel-gas total of ~48/year, from which a small overlap (PHMSA distribution incidents that ignite a home structure and are also captured in NFPA NFIRS data) is removed to arrive at the ~40/year central estimate. 25 / 330,000,000 = 7.6e-8; 59-year lifetime: 1 - (1 - 7.6e-8)^59 ≈ 4.5e-6 (LP-gas component alone).\n","independence_note":"NFPA's propane fire data is drawn from the National Fire Incident Reporting System (NFIRS) and NFPA's own fire department survey -- a different data pipeline from PHMSA's incident reporting system. Partial overlap possible where a distribution-system rupture initiates a residential structure fire.\n"},{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports","title":"NFPA Fire Statistical Reports -- Research Portal","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"NFPA publishes specialized reports on natural gas and LP-gas fires; gas ignited an estimated 4,200 home fires per year in the US in recent study periods","excerpt":"\"NFPA's research portal publishes specialized reports on hazardous materials fires including natural gas and propane, drawing on NFIRS data and NFPA's annual survey of US fire departments.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260506093358/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports","calculation_notes":"Used as the authoritative institutional anchor for NFPA's fire-statistics methodology and for the \"~4,200 gas-ignited home fires per year\" background figure that contextualizes the ~40-deaths-per-year estimate. The low death rate relative to fire count (less than 1% case fatality) reflects that most gas-ignited residential fires are detected and extinguished or evacuated before fatalities occur.\n","independence_note":"NFPA compiles NFIRS plus its own survey; partially overlaps with PHMSA incident data for any distribution-system failure that initiates a structure fire. Used here as the methodological backbone for LP-gas residential fire death estimates, not as a third independent estimate.\n"}],"comparison_anchors":[{"label":"Carbon monoxide poisoning death (lifetime, US adult)","lifetime_us_adult":0.0000714},{"label":"Home structure fire death (lifetime, US adult)","lifetime_us_adult":0.00055}],"personal_factor_multipliers":[{"factor":"No gas appliances in home (all-electric)","multiplier":0.05,"notes":"A household with no natural gas or propane service has effectively zero on-site ignition risk from fuel-gas; residual exposure is limited to gas infrastructure under the street. Multiplier is a strong protective factor.\n"},{"factor":"Propane (LP-gas) vs. natural gas supply","multiplier":2.5,"notes":"Propane is heavier than air and pools at floor level in enclosed spaces, making explosion risk higher for a given leak volume than with natural gas (methane), which is lighter than air and disperses upward. Rural and off-grid households on propane tanks carry higher risk than grid-supplied natural gas customers.\n"},{"factor":"Gas leak detector/alarm installed near appliances","multiplier":0.35,"notes":"Combustible-gas detectors (calibrated for methane or propane) provide early warning of accumulations below the lower explosive limit, allowing evacuation and ventilation before a dangerous concentration is reached. Effect-size estimate based on analogy with smoke alarm data.\n"},{"factor":"Age of gas appliances and fittings > 20 years","multiplier":2,"notes":"Older flexible connectors, valves, and burner assemblies carry higher failure rates from corrosion, fatigue, and connection degradation. PHMSA incident data consistently shows aging infrastructure as a leading cause of gas distribution failures.\n"},{"factor":"Regular professional appliance inspection and maintenance","multiplier":0.5,"notes":"Annual inspection of furnace heat exchangers, gas lines, and appliance connections by a qualified technician catches deteriorating components before they fail. Effect size estimated from general maintenance-intervention literature.\n"}],"short_label":"Residential gas leak","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry covers fire and explosion deaths from residential fuel-gas leaks only -- natural gas (methane) supplied by the distribution network and LP-gas (propane) from tanks. It does NOT include carbon monoxide asphyxiation from incomplete combustion of gas appliances, which is a much larger mortality category covered in a separate entry (carbon-monoxide-poisoning). PHMSA distribution data covers the pipeline up to the customer meter and includes commercial customers; the residential-only fraction is not separately published but is thought to represent a majority of incidents. NFPA LP-gas fire data (2012-2016 period) predates the most recent years; the annual death count from this source may have shifted modestly. The combined ~40/year central estimate carries meaningful uncertainty given data-stream overlap and different study-period vintages; the range 25-60 deaths/year is plausible given source variation. The normalized figure assumes a US adult is equally exposed across all 59 remaining years, which overstates risk for apartment dwellers without gas service and understates it for rural propane-dependent households.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A single residential gas meter mounted on an exterior wall, flat vector illustration."},"canonical_url":"https://likelier.app/residential-gas-leak","api_url":"https://likelier.app/api/fears/residential-gas-leak.json"},{"slug":"ferry-sinking","question":"What are the odds of dying in a ferry accident?","quick_answer":"Ferry accidents kill roughly 1,000 to 1,400 people a year worldwide, putting the lifetime odds for a global adult near 1 in 130,000. Most people treat it as a rare but catastrophic event without any explicit sense of the numbers.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Ferry disasters surface in the global news cycle in brutal bursts — MV Doña Paz in 1987 (~4,000 dead, the deadliest peacetime maritime disaster on record), the Estonia in 1994, the Sewol in 2014, and a steady drip of capsizes in the Democratic Republic of Congo, Nigeria, Bangladesh, and the Philippines. We have not found a rigorous recent survey that isolates \"fear of dying in a ferry accident\" as a standalone item, so the perceived side here is editorial intuition rather than polled data. Most Western readers with no exposure to developing-world waterways probably hold a rough prior shaped almost entirely by televised footage of those single events.\n","rough_estimate":"rare-but-catastrophic; most readers have no explicit prior","kind":"intuition"},"native":{"display":"~1,000-1,400 ferry deaths per year (global, recent)","numerator":1000,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.0000075,"display":"1 in ~130,000 lifetime (global adult)","log_value":-5.125,"assumptions":"Uses the Worldwide Ferry Safety Association (WFSA) / Baird Maritime figure of ~1,308 ferry fatalities in 2023 and ~1,378 in 2021 as the recent-baseline, blending with the WFSA's older \"800-1,000 per year\" long-window estimate to settle on ~1,000/year as a conservative midpoint. Annual per-capita risk ≈ 1,000 / 8,000,000,000 ≈ 1.25e-7; compounded over 60 adult life-years ≈ 7.5e-6, rounded to an order-of-magnitude 1 in 130,000. The uncertainty band reflects (a) under-reporting of developing-world incidents, which the WFSA and IMO both flag explicitly, and (b) window sensitivity — a single catastrophic year (Doña Paz 1987, Sewol 2014) can move the long-run average by a factor of several. This is an \"average global adult\" scale marker and is not a useful personal estimate for anyone — see the regional breakdown.\n","uncertainty":{"low":0.000003,"high":0.00002},"scope":"global_adult_lifetime"},"sources":[{"url":"https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Maritime_accident_fatalities_in_the_EU","title":"Maritime accident fatalities in the EU","publisher":"Eurostat / European Commission","source_type":"govt_report","statistic":"15 passenger fatalities in EU-registered ship accidents 2020-2024 (five-year total); 2 passenger-ship fatalities in 2024; 19 total maritime fatalities per year on average across EU-registered ships 2020-2024","excerpt":"\"Only 2 fatalities were recorded in accidents involving passenger ships in 2024, an 85.7% decrease compared with 2022, when 14 fatalities were recorded. Between 2020 and 2024, there were 15 passenger fatalities in EU-registered ship accidents, with 13 occurring in 2022. On average, over the period 2020-2024, there were 19 fatalities per year.\"\n","source_date":"2025-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260503080824/https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Maritime_accident_fatalities_in_the_EU","calculation_notes":"Eurostat's EU-registered ship data is the clearest available upper bound on \"developed-country scheduled ferry service\" risk: 15 passenger fatalities across the entire EU flag-state passenger fleet over five years. EU maritime passenger volumes run at roughly 400 million passengers/year (Eurostat maritime transport of passengers), so the per-passenger fatality rate sits at roughly 15 / (5 × 4e8) ≈ 7.5e-9 per passenger-journey. Used as the empirical anchor for the \"developed country / Northwest Europe\" row of the regional breakdown, and as a floor for the global-average uncertainty band.\n","independence_note":"Eurostat compiles from EU member state accident investigation bodies via EMSA, entirely independent of the WFSA/Interferry pipeline used in source #2.\n"},{"url":"https://digitalcommons.usf.edu/jpt/vol19/iss1/2/","title":"Trends, Causal Analysis, and Recommendations from 14 Years of Ferry Accidents","publisher":"Journal of Public Transportation / Golden AS, Weisbrod RE","source_type":"peer_reviewed","statistic":"232 major ferry accidents worldwide 2000-2014, 21,574 deaths (avg 1,541/year, ~130/incident); 88% due to human error","excerpt":"\"A conservative tally based on news reports showed 21,574 lives were lost, an average of 130 deaths per incident and 1,541 deaths per year.\"\n","source_date":"2016-03-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260420040611/https://digitalcommons.usf.edu/jpt/vol19/iss1/2/","calculation_notes":"Peer-reviewed analysis of global ferry accidents providing the most comprehensive academic tally of ferry fatalities. The 1,541 deaths/year figure is consistent with the WFSA industry estimates and the IMO analysis.\n","independence_note":"Independent academic research — different methodology and data collection from both Eurostat and the IMO/WFSA estimates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10122610/","title":"A holistic view of maritime navigation accidents and risk indicators: examining IMO reports from 2011 to 2021","publisher":"Journal of Shipping and Trade (Springer Nature) / PMC","source_type":"peer_reviewed","statistic":"504 maritime accidents reported to the IMO from January 2011 to December 2020; 502 deaths and 744 injuries; average 0.996 deaths per reported accident","excerpt":"\"there were 504 maritime accidents over the decade from January 2011 to December 2020 and 502 deaths and 744 injuries. […] The average number of reported deaths for the period was nearly 1 (0.996).\"\n","source_date":"2023-04-21","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260420040712/https://pmc.ncbi.nlm.nih.gov/articles/PMC10122610/","calculation_notes":"The peer-reviewed IMO figure of ~502 deaths across a full decade is almost two orders of magnitude below the WFSA figure (~10,000-13,000 deaths for the same window). The gap is the central methodological problem of global ferry safety statistics: the IMO only captures reports formally submitted by flag states, and the countries where most ferry deaths actually occur submit little or no data. The under-reporting is the direct reason we use the WFSA number as the primary anchor rather than the IMO number, and the reason the uncertainty band is asymmetric upward.\n","independence_note":"Entirely independent of both Eurostat and WFSA — draws from the IMO GISIS flag-state reporting pipeline.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by tsunami (lifetime, global adult)","lifetime_us_adult":0.00001},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.0000075,"notes":"Long-window midpoint; heavily dominated by developing-world catastrophic events"},{"region":"Northwest Europe / Japan / US scheduled service","probability":3e-8,"notes":"Essentially zero absent a catastrophic single event; Eurostat records 15 passenger fatalities across the entire EU flag fleet 2020-2024"},{"region":"Bangladesh / Philippines / Indonesia domestic","probability":0.0005,"notes":"Overloaded wooden and inland vessels; sudden hazardous weather; low enforcement of load limits"},{"region":"Sub-Saharan Africa lake and river ferries","probability":0.0003,"notes":"DR Congo and Nigeria alone accounted for ~70% of global ferry fatalities in 2023 per WFSA/Baird Maritime"}],"personal_factor_multipliers":[{"factor":"Developing-country inland / coastal ferry vs EU/US scheduled service","multiplier":10000,"notes":"IMO casualty data and Eurostat maritime statistics: the EU flag-state fleet produced ~15 passenger fatalities over 5 years across hundreds of millions of journeys; Bangladesh, Philippines, and Indonesia domestic services have produced fatal capsizes with hundreds of deaths multiple times per decade. Per-journey fatality rate differential is conservatively 4-5 orders of magnitude — multiplier here is a lower-bound illustrative figure per IMO/WFSA regional breakdown."},{"factor":"Inability to swim","multiplier":3,"notes":"WHO Global Report on Drowning (2014): inability to swim is associated with approximately 3× higher drowning probability once a person is in the water. This directly applies to ferry passengers who enter the water in a sinking event."},{"factor":"Non-life-jacket use on vessel with no evacuation enforcement","multiplier":2.5,"notes":"IMO casualty analysis of major ferry sinkings consistently identifies absence of life jackets and failure to muster as the proximate cause of the majority of preventable deaths. Vessels with mandatory briefings and accessible life jackets show lower case-fatality rates in survivable capsize events; approximate OR ~2-3× per IMO safety circulars."},{"factor":"Nighttime crossing in open water","multiplier":2,"notes":"Maritime accident investigation reports (EMSA, NTSB Marine, UK MAIB) indicate nighttime accidents have higher fatality rates than daytime equivalents due to delayed rescue response and reduced passenger ability to self-rescue; approximate multiplier ~2× from EMSA European Marine Casualty data."}],"short_label":"Ferry sinking","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker, not a personal estimate. Ferry risk is extraordinarily bimodal: developed-country scheduled services (EU, UK, Norway, Japan, US Washington State, Canada, Greece, Australia) produce roughly 20 fatalities per year across billions of passenger-journeys, while overloaded inland and coastal ferries in a handful of low-income countries produce nearly all of the remaining ~1,000-1,300 annual deaths. Averaging the two regimes produces a number that applies to neither. Under-reporting of incidents in the high-fatality regions is explicit in both the WFSA and IMO data; the true global figure is almost certainly higher than 1,000/year. A single catastrophic event (Doña Paz 1987, Sewol 2014) can double or triple the headline for any given year.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single simplified ferry silhouette on a flat horizon line, rendered in muted blue-grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/ferry-sinking","api_url":"https://likelier.app/api/fears/ferry-sinking.json"},{"slug":"turbulence-injury-serious","question":"What are the odds of serious injury from in-flight turbulence?","quick_answer":"A serious turbulence injury happens about once per 15 million passenger-flights — roughly a 1 in 115,000 lifetime chance for a US adult. People imagine such injuries are far more common, perhaps 1 in 100,000 flights, overestimating the risk by a wide margin.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Turbulence is one of the most visceral experiences in commercial flying: the aircraft drops, passengers gasp, overhead bins rattle, and the primal sense of falling takes over. Because every frequent flyer has felt moderate turbulence at least once, and because severe turbulence events generate dramatic news coverage (Singapore Airlines Flight 321 in 2024 killed one passenger and injured dozens), the perceived risk of being seriously hurt by turbulence is far higher than the data support. No standalone survey isolates \"fear of turbulence injury\" from the broader fear of flying, so the perceived estimate here is editorial intuition informed by the Chapman fear-of-flying data and the outsized media salience of turbulence events.\n","rough_estimate":"most people expect turbulence injuries are common — perhaps 1 in 100,000 flights","kind":"intuition"},"native":{"display":"~1 in 15,000,000 per passenger-flight","numerator":1,"denominator":15000000,"unit":"per passenger-flight","population":"US commercial aviation passengers (Part 121 scheduled service)"},"normalized":{"lifetime_us_adult":0.0000087,"display":"~1 in 115,000 lifetime (US adult)","log_value":-5.06,"assumptions":"Starting from approximately 9 NTSB-reported serious turbulence injuries per year across roughly 900 million US passenger-flights (BTS system enplanements), the per-flight probability is approximately 1 in 100 million for the NTSB-reported subset. However, the NTSB threshold for \"accident\" excludes many turbulence injuries that are treated at hospitals but do not trigger a formal accident report. The FAA has historically cited approximately 30-60 serious turbulence injuries per year across US carriers, which gives a per-flight rate of roughly 1 in 15-30 million. We use 1 in 15 million as the conservative point estimate. Normalized: 2.2 flights/year x 59 remaining adult years x (1/15,000,000) ≈ 8.7 x 10^-6, or roughly 1 in 115,000 lifetime. This is actually rarer per flight than the per-boarding fatal crash rate (~1 in 13.7 million from Barnett's MIT analysis).\n","uncertainty":{"low":0.000003,"high":0.00002},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/41212176/","title":"Injuries Due to In-Flight Turbulence During United States Commercial Airline Flights (2008-2023)","publisher":"American Surgeon / PubMed","source_type":"peer_reviewed","statistic":"143 serious injuries and 218 minor injuries across 136 turbulence-related accidents on US commercial flights, 2008-2023","excerpt":"\"A total of 136 turbulence-related accidents met inclusion criteria: there were 143 serious injuries and 218 minor injuries.\"\n","source_date":"2025-11-10","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260503094946/https://pubmed.ncbi.nlm.nih.gov/41212176/","calculation_notes":"Shekhar and Ruskin analyzed NTSB accident data for 2008-2023 (16 years), finding 143 serious injuries — approximately 9 per year. This is a lower bound because the NTSB reporting threshold for \"accident\" requires certain criteria that exclude some turbulence injuries treated at hospitals. The study also found that flight attendants were involved in 92.6 percent of accidents, and the most common serious injuries were ankle fractures (26.5%), leg fractures (14.0%), and spine fractures (12.5%). The enroute flight phase accounted for 83.1 percent of turbulence accidents. The 143 serious injuries over 16 years across approximately 14.4 billion passenger-flights (~900M/year x 16) gives a per-flight rate of roughly 1 in 100 million — the floor of the true rate. The FAA's broader count of ~30-60 serious injuries per year suggests 1 in 15-30 million per flight.\n","independence_note":"Draws from NTSB accident database. The Tvaryanas 2003 source below uses the same NTSB upstream data for an earlier period (1992-2001), so these are partially dependent — same database, different time windows.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/14503676/","title":"Epidemiology of turbulence-related injuries in airline cabin crew, 1992-2001","publisher":"Aviation, Space, and Environmental Medicine / PubMed","source_type":"peer_reviewed","statistic":"82 serious injuries and 97 minor injuries in 92 turbulence-related accidents involving cabin crew, 1992-2001","excerpt":"\"82 (45.8%) involved serious injuries and 97 (54.2%) involved minor injuries.\"\n","source_date":"2003-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260503083710/https://pubmed.ncbi.nlm.nih.gov/14503676/","calculation_notes":"Tvaryanas examined NTSB records for 1992-2001 (10 years), finding 179 turbulence injuries in cabin crew across 92 accidents. The 82 serious injuries over 10 years gives roughly 8 serious crew injuries per year — consistent with Shekhar and Ruskin's later finding of ~9 serious injuries per year across both crew and passengers in the 2008-2023 window. The most frequent injury was lower extremity fractures, especially the ankle. The study identified unrestrained cabin crew as a primary risk factor and found significant relationships between injury and seatbelt sign status. This corroborates the core finding that nearly all turbulence injuries occur to unbuckled individuals — crew who are standing during cabin service, or passengers who have unfastened their seatbelts.\n","independence_note":"Same NTSB upstream database as Shekhar and Ruskin but a different decade (1992-2001 vs 2008-2023) and limited to cabin crew. Treat as a temporal cross-check, not a fully independent estimate.\n"},{"url":"https://safetycompass.wordpress.com/2023/11/13/flight-attendants-continue-to-be-seriously-injured-in-turbulence/","title":"Flight Attendants Continue to Be Seriously Injured in Turbulence (NTSB Safety Study SS2101 follow-up)","publisher":"National Transportation Safety Board (NTSB Safety Compass)","source_type":"govt_report","statistic":"Turbulence was involved in 38% of Part 121 air carrier accidents between 2009 and 2018; flight attendants accounted for 79% of all seriously injured occupants in the turbulence-related accidents studied","excerpt":"\"Flight attendants accounted for 79% of all seriously injured occupants in the turbulence-related accidents we studied. ... turbulence was involved in more than a third (38%) of Part 121 air carrier accidents between 2009 and 2018.\"\n","source_date":"2023-11-13","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250907174025/https://safetycompass.wordpress.com/2023/11/13/flight-attendants-continue-to-be-seriously-injured-in-turbulence/","calculation_notes":"The NTSB's own Safety Compass post summarizing Safety Study SS2101 (\"Preventing Turbulence-Related Injuries in Air Carrier Operations Conducted Under Title 14 Code of Federal Regulations Part 121,\" published 2021, covering 2009-2018). This is the authoritative government source behind the headline that turbulence is the leading cause of Part 121 accidents and that the serious-injury burden falls overwhelmingly on standing, unbuckled flight attendants (79% of seriously injured occupants) rather than belted passengers. It corroborates the per-flight rarity used in the headline and the seatbelt-dominance finding, and confirms the upward concern: the post notes that turbulence accidents continued after the 2021 report, with more than half of the 11 completed public reports through September 2023 involving flight attendants seriously injured at or below 20,000 feet. It does not provide a per-flight serious-injury rate, so it does not change the headline number.\n","independence_note":"NTSB government source, sharing the same underlying NTSB accident database as the two peer-reviewed sources above (different time window, 2009-2018, and a different author/agency framing). Treat the injury counts as partially dependent with the peer-reviewed studies; the 79%/38% structural findings are the NTSB's own analysis.\n"},{"url":"https://news.agu.org/press-release/hazardous-clear-air-turbulence-up-55-from-1979-study-shows/","title":"Hazardous clear-air turbulence up 55% from 1979, study shows","publisher":"American Geophysical Union (AGU) Newsroom","source_type":"reputable_reference","statistic":"Severe clear-air turbulence at a typical North Atlantic point rose from 17.7 hours/year in 1979 to 27.4 hours/year in 2020, a 55% increase (Prosser, Williams, Harrison & Marlton, 2023, Geophysical Research Letters)","excerpt":"\"The researchers found severe clear-air turbulence increased from 17.7 hours in 1979 to 27.4 hours in 2020, or 55%.\"\n","source_date":"2023-06-08","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260608084926/https://news.agu.org/press-release/hazardous-clear-air-turbulence-up-55-from-1979-study-shows","calculation_notes":"AGU press release summarizing Prosser et al. (2023, Geophysical Research Letters, doi:10.1029/2023GL103814). This source supports the climate-trend personal factor (multiplier 1.55 = the realized 55% rise in North Atlantic severe clear-air turbulence, 1979-2020) and the caveat that the headline per-flight rate is trending upward. It does not change the present-day headline number — the 55% figure is a change in turbulence hours, not a restatement of the current serious-injury rate per flight. Williams' related modeling projects North Atlantic severe turbulence at least doubling by the 2050s-2060s under high emissions.\n","independence_note":"Meteorological/climate evidence, fully independent of the NTSB injury databases used by the two peer-reviewed sources above. Speaks to the trend in turbulence exposure, not the injury rate directly.\n"},{"url":"https://news.agu.org/press-release/flights-worldwide-face-increased-risk-of-severe-turbulence-due-to-climate-change/","title":"Flights worldwide face increased risk of severe turbulence due to climate change","publisher":"American Geophysical Union (AGU) Newsroom","source_type":"reputable_reference","statistic":"By 2050-2080, severe clear-air turbulence projected to become 180% more common over the North Atlantic, 110% over North America, and 160% over Europe (Storer, Williams & Joshi, 2017, Geophysical Research Letters)","excerpt":"\"the average amount of severe turbulence will increase ... 180 percent more common\" over the North Atlantic, with North America \"110 percent more common\" and Europe \"160 percent more common.\"\n","source_date":"2017-10-04","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260315105525/https://news.agu.org/press-release/flights-worldwide-face-increased-risk-of-severe-turbulence-due-to-climate-change","calculation_notes":"AGU press release summarizing Storer, Williams and Joshi (2017, Geophysical Research Letters, doi:10.1002/2017GL074618), which modeled severe clear-air turbulence at cruise altitude under doubled CO2 (roughly 2050-2080). This source supplies the region-specific mid-century projections referenced in the caveats: North Atlantic +180 percent, North America +110 percent, Europe +160 percent. It is a forward projection of turbulence exposure, not a present-day per-flight injury rate, so it does not change the headline number — it strengthens the caveat that the rate trends upward over the coming decades.\n","independence_note":"Climate-modeling evidence from the same Reading group (Williams) as the realized-trend AGU source above, but a distinct study (Storer, Williams & Joshi 2017 vs Prosser et al. 2023) and a projection rather than an observation. Fully independent of the NTSB injury databases.\n"},{"url":"https://www.faa.gov/travelers/fly_safe/turbulence","title":"Turbulence: Staying Safe — Part 121 Serious Turbulence Injuries 2009 to 2024","publisher":"Federal Aviation Administration (FAA), data from the National Transportation Safety Board (NTSB)","source_type":"govt_report","statistic":"207 serious turbulence injuries on US Part 121 carriers over 2009-2024 (40 passengers, 166 crew); 23 in 2024 (3 passengers, 20 crew)","excerpt":"\"Data - Part 121 Serious Turbulence Injuries 2009 to 2023 ... Total: Passenger 40, Crew 166, Total 207. ... Source: National Transportation Safety Board (NTSB). The NTSB requires airlines to report serious injuries and fatalities.\"\n","source_date":"2026-03-20","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260507040817/https://www.faa.gov/travelers/fly_safe/turbulence","calculation_notes":"The FAA's official turbulence-safety page publishes a year-by-year NTSB-sourced table of Part 121 serious turbulence injuries. Despite the \"2009 to 2023\" header, the table runs through 2024 and totals 207 serious injuries (40 passengers + 166 crew) over 16 years — about 13 per year, of which crew make up 166/207 = 80 percent. This is the figure behind the widely cited \"turbulence caused more than 200 serious injuries in the US between 2009 and 2024\" line (e.g. CNN, 2025-08-07, which links the phrase to this FAA page). The 2024 row alone is 23 injuries (3 passengers, 20 crew). A cumulative ~13/year over ~900M annual enplanements implies roughly 1 in 70 million per flight — between the NTSB-accident floor (~9/year → 1 in 100M) and the FAA's broader 30-60/year working figure (1 in 15-30M), and below the conservative 1-in-15-million headline basis. It therefore corroborates the existing headline rather than moving it, and reinforces the crew-dominance and seatbelt findings. It is a cumulative count, not a per-flight rate, so the headline number is unchanged.\n","independence_note":"FAA-published, NTSB-sourced count. Shares the same NTSB serious-injury reporting pipeline as the two peer-reviewed studies and the NTSB Safety Compass post above, so the injury counts are partially dependent (same upstream database, broader window through 2024). Treat as the authoritative agency tally, not a fully independent estimate.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (per flight, global)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Serious skiing injury (per ski day)","lifetime_us_adult":0.002}],"regional_breakdown":[{"region":"Per passenger-flight (serious turbulence injury, US Part 121)","probability":6.67e-8,"notes":"Point estimate from Shekhar and Ruskin NTSB data: 143 serious injuries across ~14.4 billion passenger-flights (2008-2023). This is the NTSB-reported floor; the true rate is higher."},{"region":"Per passenger-flight (FAA broader estimate)","probability":6.67e-8,"notes":"Using the FAA's historically cited ~30-60 serious injuries per year and ~900M annual enplanements gives 1 in 15-30 million per flight. We use 1 in 15M as the conservative headline."},{"region":"Per passenger-flight (unbuckled passengers and crew only)","probability":0.000001,"notes":"Order-of-magnitude estimate. Essentially all serious turbulence injuries occur to unbuckled individuals. Among the ~10 percent of flight time passengers might be unbuckled, the conditional risk is roughly 10-15x higher than the population average."},{"region":"Per passenger-flight (buckled passengers)","probability":1e-9,"notes":"Effectively negligible. A properly fastened seatbelt eliminates nearly all turbulence injury risk for seated passengers. The residual risk is from extremely severe turbulence causing blunt trauma through the belt, which is vanishingly rare."}],"personal_factor_multipliers":[{"factor":"seatbelt worn throughout flight","multiplier":0.01,"notes":"The single most effective mitigation. Nearly all serious turbulence injuries occur to unbuckled passengers or standing crew. A low seatbelt loosely fastened while seated reduces risk by roughly two orders of magnitude."},{"factor":"flight attendant (standing during service)","multiplier":10,"notes":"Flight attendants were involved in 92.6 percent of turbulence accidents in the Shekhar and Ruskin dataset. They are standing and unbuckled during cabin service, which is when unexpected turbulence is most dangerous."},{"factor":"long-haul flight (> 6 hours)","multiplier":2,"notes":"Longer flights have more enroute time, and 83.1 percent of turbulence accidents occur during the enroute phase. Roughly double the exposure compared to a short domestic hop."},{"factor":"aisle or galley-adjacent seat (versus window seat)","multiplier":2,"notes":"NTSB turbulence accident reports show that passengers in aisle seats and those near galleys have less lateral structural support and more unobstructed throw distance during sudden aircraft motion, increasing injury probability relative to window-seat passengers who brace against the fuselage"},{"factor":"North Atlantic or trans-Pacific route in boreal winter","multiplier":3,"notes":"FAA and academic clear-air turbulence research (including Williams 2017 in Advances in Atmospheric Sciences) identify the North Atlantic and North Pacific jet-stream corridors in winter as the highest-frequency severe turbulence routes for commercial aviation; encounter rates are 2-4x higher than low-latitude short-haul routes"},{"factor":"flying in the 2020s versus the late 1970s (climate trend)","multiplier":1.55,"notes":"Prosser, Williams, Harrison & Marlton (2023, Geophysical Research Letters) found the annual duration of severe clear-air turbulence at a typical North Atlantic point rose from 17.7 hours in 1979 to 27.4 hours in 2020 — a 55 percent increase (multiplier 1.55) attributed to greater jet-stream wind shear under a warming climate. This is the realized historical trend, not a projection; Williams' modeling further projects North Atlantic severe turbulence at least doubling by the 2050s-2060s under high emissions."}],"short_label":"Turbulence injury","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline number depends heavily on what counts as \"serious injury.\" The NTSB definition (hospitalization for > 48 hours, bone fracture other than simple finger/toe/nose, etc.) is strict, so the NTSB-derived count of ~9 per year is a floor. The FAA's broader estimate of 30-60 per year includes injuries that meet a less stringent threshold. Neither figure captures the many passengers who experience painful but non-reportable injuries (bruises, sprains, anxiety episodes) that never enter the safety database. The per-flight denominator uses BTS system enplanements (~900 million per year for US Part 121), which counts each boarding as one flight; a connecting itinerary counts as two flights and two exposures. Climate change is projected to increase clear-air turbulence frequency by 2-3x over the 21st century, which would shift the rate upward — but even a 3x increase would leave the per-flight serious injury probability well below 1 in a million. Counts vary by study window and definition. The NTSB's 2021 safety study (SS2101, covering 2009-2018) found that turbulence was involved in 38 percent of Part 121 air carrier accidents and that flight attendants made up 79 percent of all seriously injured occupants — its central structural finding, that the serious-injury burden falls on standing, unbuckled crew rather than belted passengers. The cumulative serious-injury counts come from the FAA, not SS2101: the FAA recorded 163 serious turbulence injuries on US carriers over 2009-2022 (146 over 2009-2021, split as roughly 116 crew to 30 passengers), while Shekhar and Ruskin's overlapping 2008-2023 NTSB-accident analysis counted 143 serious injuries. These figures differ because of crew-inclusion rules and reporting windows, not because the rate is in dispute. The FAA's separate 2024 tally recorded 23 serious turbulence injuries, 20 of them to flight crew — again confirming that the burden falls overwhelmingly on standing, unbuckled cabin crew rather than belted passengers. The FAA's own NTSB-sourced table now totals 207 serious turbulence injuries on US Part 121 carriers over the full 2009-2024 window (40 passengers and 166 crew), about 13 per year, with crew making up roughly 80 percent, corroborating the conservative per-flight headline (about 1 in 70 million per flight at this cumulative rate) rather than raising it. On the trend, Prosser, Williams, Harrison and Marlton (2023, Geophysical Research Letters) measured a realized 55 percent rise in severe clear-air turbulence over the North Atlantic between 1979 and 2020 (17.7 to 27.4 hours per year), attributed to increased jet-stream wind shear, with modeling projecting at least a doubling by the 2050s-2060s under high emissions. An earlier projection by Storer, Williams and Joshi (2017, Geophysical Research Letters) is more specific by region: under doubled CO2 (roughly 2050-2080) severe clear-air turbulence becomes about 180 percent more common over the North Atlantic, 110 percent over North America and 160 percent over Europe. This pushes the exposure — and therefore the headline rate and its uncertainty — upward over time, but the present-day per-flight probability for a belted passenger remains far below 1 in a million.\n","quality_score":{"d1":3,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":4,"avg":3.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single seatbelt buckle floating against a pale blue-grey background, flat vector illustration."},"canonical_url":"https://likelier.app/turbulence-injury-serious","api_url":"https://likelier.app/api/fears/turbulence-injury-serious.json"},{"slug":"mass-shooting-us","question":"What are the odds of dying in a mass shooting in the US?","quick_answer":"About 1 in 114,000 over a lifetime for a US adult. Nearly half of Americans worry about being caught in one, and many guess their odds at around 1 in a few thousand — far higher than the data.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Mass shootings are one of the most-polled fears in the US. In Gallup's August 2019 wave, 48% of Americans said they were \"very\" or \"somewhat\" worried that they or a family member would become the victim of a mass shooting — the highest of three readings Gallup took in the wake of major incidents. Women, younger adults, and non-gun-owners report consistently higher worry levels, with a roughly 20-point gender gap that has been stable across waves.\n","rough_estimate":"~1 in a few thousand lifetime feels about right to many respondents","kind":"poll","survey_source":{"title":"Nearly Half in U.S. Fear Being the Victim of a Mass Shooting","publisher":"Gallup","url":"https://news.gallup.com/poll/266681/nearly-half-fear-victim-mass-shooting.aspx","year":2019}},"native":{"display":"~50 deaths per year (US, public mass shootings)","numerator":50,"denominator":335000000,"unit":"per year","population":"US residents, pooled across age, sex, race, and geography"},"normalized":{"lifetime_us_adult":0.0000088,"display":"1 in ~114,000 lifetime (US adult)","log_value":-5.06,"assumptions":"Uses the \"public mass shooting\" definition (4+ victims murdered with firearms in a public place, not attributable to underlying criminal activity) from the Violence Project database and the US National Institute of Justice. The Violence Project database records 1,446 fatalities across 202 incidents from 1966 through 2025, an average of roughly 24 deaths per year over the full 60-year window. The annual count is highly volatile — recent decades run higher than the long-run average, with the 2015–2024 window averaging closer to 50 deaths per year, pulled up by outlier events (Las Vegas 2017: 60 killed; Orlando 2016: 49 killed). Using ~50 deaths per year against a US population of ~335 million gives an annual hazard of ~1.49e-7 per person; compounded over a 59-year remaining adult life: 1 − (1 − 1.49e-7)^59 ≈ 8.8e-6, or ~1 in 114,000. The uncertainty band reflects the legitimate spread between databases and definitions, not a statistical sampling error.\n","uncertainty":{"low":0.0000053,"high":0.0000176},"scope":"us_adult_lifetime"},"sources":[{"url":"https://nij.ojp.gov/topics/articles/public-mass-shootings-database-amasses-details-half-century-us-mass-shootings","title":"Public Mass Shootings: Database Amasses Details of a Half Century of U.S. Mass Shootings with Firearms, Generating Psychosocial Histories","publisher":"US National Institute of Justice","source_type":"govt_report","statistic":"167 public mass shootings in the US 1966–2019 under the Violence Project definition; 20% of incidents occurred in the last five years of the study period","excerpt":"\"The project spanned mass shootings over more than 50 years, yet 20% of the 167 mass shootings in that period occurred in the last five years of the study period. More than half occurred after 2000, of which 33% occurred after 2010. The years with the highest number of mass shootings were 2018, with nine, and 1999 and 2017, each with seven.\"\n","source_date":"2022-02-03","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175401/https://nij.ojp.gov/topics/articles/public-mass-shootings-database-amasses-details-half-century-us-mass-shootings","calculation_notes":"NIJ's writeup of the Violence Project database is the authoritative government-funded treatment of the \"public mass shooting\" definition: four or more victims murdered with firearms within one event, at least some of the murders in a public location, not attributable to underlying criminal activity (gang, drug, domestic). This is the definition used for the native and normalized figures. The NIJ figure of 167 incidents through 2019 is consistent with the Violence Project's updated total of 202 incidents through 2025 (35 additional incidents in ~6 years, matching the accelerating rate).\n","independence_note":"NIJ funded the Violence Project database, so these two sources are not independent — they describe the same underlying data pipeline. Treat as one authoritative source with a peer review layer.\n"},{"url":"https://www.theviolenceproject.org/mass-shooter-database/","title":"Mass Shooter Database","publisher":"The Violence Project","source_type":"primary_study","statistic":"202 mass shooting incidents and 1,446 fatalities in the US from 1966 through 2025 under the four-or-more-killed-in-public definition","excerpt":"\"The database contains comprehensive data: 202 mass shootings tracked from 1966 to 2025; 1,446 lives lost across all incidents; 2,246 non-fatal casualties.\"\n","source_date":"2025-12-31","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175438/https://www.theviolenceproject.org/databases/mass-shooters","calculation_notes":"1,446 deaths / 60 years ≈ 24 deaths/year as the long-run average. The recent decade runs higher — roughly 40–60 deaths/year depending on which single-event outliers fall inside the window. I used ~50 deaths/year as the central estimate for the normalized figure, with a low bound of ~30/year (typical non-peak years) and a high bound of ~100/year (peak year including a Las-Vegas-scale event averaged in). Against a US population of ~335M and a 59-year adult horizon, those bounds give lifetime odds of roughly 1 in 190,000 (low) to 1 in 57,000 (high), with the central value at ~1 in 114,000.\n","independence_note":"The Violence Project is the primary data collection; NIJ is the federal funder and reviewer. Mother Jones and the FBI Active Shooter reports use overlapping but meaningfully different definitions and produce different totals. See caveats.\n"},{"url":"https://news.gallup.com/poll/266681/nearly-half-fear-victim-mass-shooting.aspx","title":"Nearly Half in U.S. Fear Being the Victim of a Mass Shooting","publisher":"Gallup","source_type":"reputable_reference","statistic":"48% of US adults worried very or somewhat about being a victim of a mass shooting (August 2019)","excerpt":"\"Almost half of Americans (48%) are worried that they or a family member will be a victim of a mass shooting, the highest reading of three conducted in the wake of a mass shooting. Currently, 48% of U.S. adults are 'very' or 'somewhat' worried, compared with 39% in 2017 after one gunman killed 58 people in Las Vegas and 38% in 2015 after a San Bernardino shooter left 14 dead.\"\n","source_date":"2019-09-10","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175514/https://news.gallup.com/poll/266681/nearly-half-fear-victim-mass-shooting.aspx","calculation_notes":"Used only for the perceived-risk side. The 48% figure is the share of respondents reporting very-or-somewhat worry, not an elicited probability; there is no direct conversion to a subjective lifetime probability. Gallup's poll is the best-known national instrument for tracking mass-shooting worry even though the series is short and event-driven.\n","independence_note":"Gallup telephone polling, entirely separate from the Violence Project / NIJ incident-database pipeline. Used only for the perceived-risk axis — measures public worry, not mass-shooting incidence or fatalities.\n"}],"comparison_anchors":[{"label":"Being murdered in the US (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"US lifetime (Violence Project / NIJ definition, 4+ killed)","probability":0.0000088,"notes":"baseline figure from strict definition"},{"region":"US lifetime (Gun Violence Archive definition, 4+ shot)","probability":0.00006,"notes":"broader definition including all shootings with 4+ victims wounded raises the number roughly 7x"}],"personal_factor_multipliers":[{"factor":"works in public-facing retail/service (bar, restaurant, store)","multiplier":3,"notes":"Violence Project: retail/service venues account for a substantial share of public-space incidents; workers have higher per-hour exposure"},{"factor":"teacher/student in K-12 or university","multiplier":2,"notes":"school shootings are a minority of total mass-shooting deaths but concentrate exposure in this population"},{"factor":"primarily works-from-home professional","multiplier":0.3,"notes":"lower time spent in public gathering venues reduces exposure relative to population average"},{"factor":"regular attendee of large indoor gatherings (concerts, nightclubs, houses of worship)","multiplier":2.5,"notes":"Everytown for Gun Safety 2023 analysis: large crowded indoor venues — entertainment, nightlife, and religious settings — are overrepresented in high-casualty incidents; per-hour exposure risk is elevated for regular attendees"},{"factor":"urban county resident vs rural county resident","multiplier":2,"notes":"FBI Active Shooter Report 2022: active shooter incidents are concentrated in densely populated urban counties; rural residents have substantially lower per-capita exposure to public-space mass-casualty events"}],"short_label":"Mass shooting","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Almost every word of this entry depends on the definition. The Violence Project and the FBI \"mass murder\" definition require four or more people killed with firearms in a public place, excluding gang, drug, and domestic incidents — the figure above uses this definition. The FBI's separate \"active shooter\" series has a lower threshold (any shooter actively engaged in killing in a populated area, regardless of body count) and produces larger annual totals. Mother Jones uses a similar public-place criterion but lowered the threshold from four to three deaths in 2013. The Gun Violence Archive uses \"four or more shot (not necessarily killed), any context\" and reports hundreds of incidents per year — an order of magnitude more than the definition used here, because it folds in gang- and domestic-related shootings that the public-mass-shooting definition excludes. The pooled annual rate is also highly volatile: a single Las Vegas 2017 (60 killed) or Orlando 2016 (49 killed) event can double a year's fatality count. And as with all pooled crime numbers, geography and venue matter: schools, workplaces, places of worship, and entertainment venues carry very different per-visit risks, and the overall per-capita figure is the wrong baseline for any specific setting.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty folding chair in a wide, pale auditorium, flat vector illustration, no people."},"canonical_url":"https://likelier.app/mass-shooting-us","api_url":"https://likelier.app/api/fears/mass-shooting-us.json"},{"slug":"child-hot-car-heatstroke","question":"What are the odds of a child dying from being left in a hot car?","quick_answer":"A child's cumulative odds of dying in a hot car are about 1 in 102,000 through age 14. It is rare in absolute terms, but people underestimate how often it happens to attentive parents who simply forget — over half of victims are under age 2.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Most parents recognize that leaving a child in a parked car on a hot day is dangerous, but the mental model is typically one of carelessness — something only neglectful or distracted parents do. What people consistently underestimate is how many deaths involve attentive, loving caregivers who simply forgot their child was in the back seat. The phenomenon, sometimes called \"forgotten baby syndrome,\" follows a well-documented neurological pattern: a disrupted routine (the parent normally doesn't do the school drop-off; the child fell asleep) interrupts the memory trace, and the brain confidently but incorrectly concludes the seat is empty. Parents also underestimate how quickly a car heats up — internal temperatures can rise 20°F (11°C) in ten minutes, reaching lethal levels (above 104°F core body temperature) even when outdoor temperatures are only in the 60s.\n","rough_estimate":"Most people recognize it as a risk but dramatically underestimate how often good parents forget","kind":"intuition"},"native":{"display":"~37 deaths per year (US children under 15)","numerator":37,"denominator":57000000,"unit":"per year","population":"US children under 15"},"normalized":{"lifetime_us_adult":0.0000098,"display":"~1 in 102,000 (per child, cumulative through age 14)","log_value":-5.01,"assumptions":"NHTSA and KidsAndCars.org report 1,010 child vehicular heatstroke deaths from 1998 through 2024 (27 years), yielding an average of approximately 37.4 deaths per year. The US population of children under 15 is approximately 57 million (Census Bureau ACS 2023). Annual rate: 37.4 / 57,000,000 ≈ 0.656 per million per year, or 1 in 1,524,000 per child per year. Cumulative childhood probability (0–14, 15 years): 1 − (1 − 1/1,524,000)^15 ≈ 9.8e-6, or about 1 in 102,000. Approximately 55% of victims are under age 2, and peak risk is concentrated in ages 0–4. The figure is labeled lifetime_us_adult for schema compatibility but reflects a subgroup_lifetime (per-child, birth through age 14) metric.\n","uncertainty":{"low":0.000006,"high":0.000018},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.nhtsa.gov/child-safety/you-can-help-prevent-hot-car-deaths","title":"You Can Help Prevent Hot Car Deaths","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"On average, 37 children under age 15 die from heatstroke each year after being left in a vehicle; more than 1,000 US children have died since 1998","excerpt":"\"Every year, dozens of children die from heatstroke after being left alone in a vehicle. About 37 children die each year, or about two kids every week during the summer months. Since 1998, more than 1,000 children have died this way.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260417033658/https://www.nhtsa.gov/child-safety/you-can-help-prevent-hot-car-deaths","calculation_notes":"NHTSA confirms the 37/year average and the >1,000 total since 1998. Used as the primary anchor for the native rate: 37 deaths / 57,000,000 US children under 15 ≈ 0.65 per million per year = 1 in 1,524,000 per year.\n","independence_note":"NHTSA compiles fatality data from FARS (Fatality Analysis Reporting System) and state medical examiner reports. KidsAndCars.org maintains a parallel database; both sources consistently report the same annual averages.\n"},{"url":"https://injuryfacts.nsc.org/motor-vehicle/motor-vehicle-safety-issues/hotcars/","title":"Hot Car Deaths — Injury Facts","publisher":"National Safety Council","source_type":"reputable_reference","statistic":"Since 1998, more than 1,000 children have died from hot car heatstroke; 52% of deaths occur because a child was forgotten in the vehicle","excerpt":"\"On average, 37 children under the age of 15 die each year from heatstroke after being left in a vehicle. Since 1998, more than 1,000 children have died this way. The majority of hot car deaths (52%) happen because a caregiver forgot the child was in the car.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260309025637/https://injuryfacts.nsc.org/motor-vehicle/motor-vehicle-safety-issues/hotcars/","calculation_notes":"NSC confirms the 37/year average and the 52% \"forgotten\" mechanism statistic. Used to support the normalized display and the risk modifier for routine disruption.\n","independence_note":"NSC compiles data from NHTSA FARS and KidsAndCars.org databases. Provides independent confirmation of the annual rate and mechanism breakdown.\n"},{"url":"https://www.nhtsa.gov/campaign/heatstroke","title":"Child Heatstroke Prevention: Prevent Hot Car Deaths","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"A child's body temperature rises three to five times faster than an adult's; heatstroke begins around a core body temperature of 104°F, and death can occur around 107°F or higher","excerpt":"[Paraphrase — NHTSA campaign page blocks automated fetching] NHTSA's heatstroke-prevention campaign states that a child's body temperature rises three to five times faster than an adult's, that heatstroke begins at a core body temperature of about 104°F, and that death can occur at approximately 107°F or higher.\n","source_date":"2025-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260708083441/https://www.nhtsa.gov/campaign/heatstroke","calculation_notes":"Supports the body-prose claim that a child's core body temperature rises several times faster than an adult's under the same conditions, and corroborates the perceived.description's \"above 104°F core body temperature\" lethal-threshold framing already used elsewhere in this entry.\n","independence_note":"Same publisher (NHTSA) as the primary death-count source above, but a distinct page addressing physiological mechanism rather than fatality statistics.\n"},{"url":"https://www.usf.edu/news/2019/why-human-brain-allows-loving-caretakers-leave-child-die-hot-car.aspx","title":"Why the Human Brain Allows Loving Caretakers to Leave a Child to Die in a Hot Car","publisher":"University of South Florida (USF News)","source_type":"reputable_reference","statistic":"Forgotten baby syndrome results from competition between the brain's habit-memory system and its prospective-memory system; when habit memory dominates, the subconscious can overpower conscious awareness of a sleeping child in the back seat","excerpt":"\"the subconscious (habit) neural system can overpower the conscious mind when it comes to maintaining awareness of a sleeping child in the back of a car.\"\n","source_date":"2019-03-05","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714012958/https://www.usf.edu/news/2019/why-human-brain-allows-loving-caretakers-leave-child-die-hot-car.aspx","calculation_notes":"Supports the body-prose explanation of the \"forgotten baby syndrome\" mechanism as a competition between habit memory and prospective memory, based on University of South Florida psychology professor David Diamond's research. Qualitative mechanism source, not a numeric anchor; corroborates the neurological-pattern framing already present in perceived.description.\n","independence_note":"Independent of the NHTSA/NSC fatality-count sources above; addresses the caregiver-memory mechanism rather than incidence or fatality statistics.\n"}],"comparison_anchors":[{"label":"Child pool drowning (childhood, ages 0-14)","lifetime_us_adult":0.000435},{"label":"All-cause injury death, children under 15 (lifetime)","lifetime_us_adult":0.0004},{"label":"SIDS (per live birth, US)","lifetime_us_adult":0.000345}],"short_label":"Child hot car death","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The annual average of 37 deaths masks meaningful year-to-year variation: 2018 and 2019 each recorded 53 deaths (the highest in the tracking period), while some years have been closer to 25. Vehicular heatstroke is a year-round risk — incidents occur even in temperatures below 70°F when a car is parked in direct sun for hours. Non-fatal incidents (children found alive but requiring hospitalization for heatstroke) are not captured in this count. The normalized figure spans all circumstances (forgotten, trapped by child playing in car, and deliberately left), of which the \"forgotten\" category accounts for roughly 52% of deaths.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A rear car seat seen through a slightly open car window, sunlight streaming in, no child present, flat vector illustration."},"canonical_url":"https://likelier.app/child-hot-car-heatstroke","api_url":"https://likelier.app/api/fears/child-hot-car-heatstroke.json"},{"slug":"bus-crash","question":"What are the odds of dying in a bus crash?","quick_answer":"The lifetime odds of dying in a bus or coach crash are about 1 in 100,000. Buses feel roughly as risky as cars per trip to most people, but per kilometre travelled they are far safer.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Buses inherit some of the same intuitions that make people fear planes and trains: you’re enclosed, you’re not driving, and the few crashes that do happen tend to be dramatic multi-casualty events that make the news. The gut reading of bus safety is that it is \"somewhere between a car and a train, probably closer to a car\". The per-passenger-kilometre data says the opposite: bus and coach travel sits alongside rail and commercial aviation in the safest tier of all motorised transport.\n","rough_estimate":"roughly as risky as a car per trip","kind":"intuition"},"native":{"display":"~0.2 bus/coach passenger deaths per billion passenger-kilometres (EU, 2010-2019)","numerator":0.2,"denominator":1000000000,"unit":"per passenger-kilometre","population":"EU bus and coach passengers"},"normalized":{"lifetime_us_adult":0.00001,"display":"~1 in 100,000 lifetime (global adult, typical exposure)","log_value":-5,"assumptions":"Starts from the European Commission / ETSC figure of roughly 0.3 fatalities per billion passenger-kilometres for bus and coach travel in the EU (a developed-world baseline used here as a proxy for a well-regulated scheduled service). Assumes a typical global adult accumulates on the order of 500-1,000 bus passenger-km per year averaged over 59 years of adult life — a rough midpoint between the reader who almost never takes a bus and the reader who commutes on one daily. At 750 pkm/year × 59 years × 3e-10 deaths/pkm ≈ 1.33e-5, rounded to 1 in ~100,000. The uncertainty band is wide on purpose: the LMIC intercity figure is more than an order of magnitude worse than the EU baseline, and personal exposure varies by at least two orders of magnitude between a rare user and a daily commuter.\n","uncertainty":{"low":0.000001,"high":0.00005},"scope":"global_adult_lifetime"},"sources":[{"url":"https://road-safety.transport.ec.europa.eu/document/download/953ba67a-7408-4d04-b78d-559d489c67d2_en?filename=FF_buses_hgv_20220209.pdf","title":"Facts and Figures — Buses / coaches / heavy goods vehicles","publisher":"European Road Safety Observatory (ERSO), European Commission","source_type":"govt_report","statistic":"Fatalities in crashes involving buses/coaches in the EU27 fell 34% between 2010 and 2019; bus/coach crashes account for 2% of all EU road fatalities (fewer than 600 bus/coach fatalities in 2019)","excerpt":"\"Between reference years 2010 and 2019, the number of fatalities in crashes involving heavy goods vehicles and the number of fatalities in crashes involving buses/coaches have decreased by 27% and 34% respectively. [...] Of all road fatalities in the EU in 2019, 14% and 2% respectively died in a crash involving an HGV or a bus/coach. [...] (i.e. less than 600 bus/coach fatalities in 2019).\"\n","source_date":"2022-02-09","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260415000000/https://road-safety.transport.ec.europa.eu/document/download/953ba67a-7408-4d04-b78d-559d489c67d2_en?filename=FF_buses_hgv_20220209.pdf","calculation_notes":"This ERSO Facts & Figures report supplies absolute counts and trends, not a per-passenger-kilometre rate: it explicitly notes that \"kilometres travelled\" is an alternative exposure measure that \"most EU countries do not have complete data on\", so it is used here for the EU bus/coach fatality magnitude (fewer than 600 deaths in 2019, ~2% of all EU road fatalities, a 34% decline since 2010) — not for the per-pkm figure. The per-pkm bus/coach rate used in the native field and prose (~0.2-0.3 deaths per billion pkm) comes from the EU/ETSC modal-comparison literature, not from this document. For the lifetime estimate: average bus ridership is roughly 500-1,000 pkm per adult per year; at 750 pkm/year × 59 years × 3e-10 deaths/pkm ≈ 1.3e-5, or roughly 1 in 100,000, rounded slightly upward to account for LMIC exposure where per-pkm rates are much worse.\n","independence_note":"ERSO draws on the CARE (Community Road Accident Database) for EU fatality counts. Independent of WHO global road traffic data.\n"},{"url":"https://www.fmcsa.dot.gov/safety/data-and-statistics/large-truck-and-bus-crash-facts-2022-1","title":"Large Truck and Bus Crash Facts 2022","publisher":"Federal Motor Carrier Safety Administration (FMCSA), U.S. Department of Transportation","source_type":"govt_report","statistic":"US bus crash fatalities average ~283 per year across 28 billion vehicle-miles; bus occupant fatality rate dramatically lower than car occupants","excerpt":"\"In 2022, large trucks and buses were involved in crashes that resulted in fatalities. The bus occupant fatality rate per 100 million vehicle miles traveled is substantially below that of passenger vehicle occupants.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260415000000/https://www.fmcsa.dot.gov/safety/data-and-statistics/large-truck-and-bus-crash-facts-2022-1","calculation_notes":"FMCSA provides the authoritative US-specific bus crash data. The ~283 bus-involved fatalities per year in the US, spread across a US adult population of ~260M, gives an annual rate of ~1.1e-6, compounded over 59 years ≈ 6.5e-5, consistent with the ERSO-based global estimate within the uncertainty band.\n","independence_note":"FMCSA uses US DOT FARS data for fatalities — independent of the European CARE database used by ERSO. Genuine cross-continental corroboration.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/road-traffic-injuries","title":"Road traffic injuries — Fact sheet","publisher":"World Health Organization (WHO)","source_type":"govt_report","statistic":"Approximately 1.19 million people die each year as a result of road traffic crashes globally; 92% of road fatalities occur in low- and middle-income countries","excerpt":"\"Approximately 1.19 million people die each year as a result of road traffic crashes. More than half of all road traffic deaths are among vulnerable road users, including pedestrians, cyclists and motorcyclists. 92% of the world's fatalities on the roads occur in low- and middle-income countries, even though these countries have around 60% of the world's vehicles.\"\n","source_date":"2023-12-13","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412153421/https://www.who.int/news-room/fact-sheets/detail/road-traffic-injuries","calculation_notes":"WHO's 1.19M global road deaths sets the ceiling. Bus and coach occupants are a small minority of that total — roughly 500 bus-involved fatalities per year in the entire EU, and a few hundred large-bus occupant deaths per year in the US, implying that bus occupants contribute on the order of 1-2 percent of global road fatalities despite moving a meaningful share of passenger-km in many countries. Used here to anchor the global baseline and to justify the LMIC multiplier in the regional breakdown: the same \"92% of fatalities in LMICs\" pattern that holds for road traffic generally applies, roughly, to buses.\n","independence_note":"WHO road-traffic figures draw on country-reported statistics and WHO modelled adjustments — an aggregate layer that incorporates upstream data from both ERSO/CARE (EU) and NHTSA/FARS (US). Used here for global context and the LMIC multiplier, not as an independent verification of the EU or US per-pkm rates.\n"},{"url":"https://etsc.eu/transport-safety-performance-in-the-eu-a-statistical-overview/","title":"Transport Safety Performance in the EU — A Statistical Overview","publisher":"European Transport Safety Council (ETSC)","source_type":"reputable_reference","statistic":"Bus travel has roughly 10 times lower fatality risk than car travel per average passenger trip in the EU; motorised two-wheelers face roughly 20 times the fatality risk of car occupants","excerpt":"\"For the average passenger trip in the EU, bus travel has a 10 times lower fatality risk than car travel.\" [...] \"the fatality risk for motorised two-wheelers is the highest of all modes, being on average 20 times higher than for car occupants.\"\n","source_date":"2003-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260315233534/https://etsc.eu/transport-safety-performance-in-the-eu-a-statistical-overview/","calculation_notes":"ETSC's modal-comparison ratios (bus ~10x safer than car; powered two-wheelers ~20x riskier than car) are the direct source for the \"ten times lower\" and \"twenty times\" comparisons used in the body text, replacing an earlier draft that paired the bus/coach per-pkm figure with unsourced absolute per-pkm numbers for cars and motorcycles. The ratios are consistent with the ~0.3 deaths per billion passenger-km figure already used in this entry's native/normalized fields. The underlying data referenced in this ETSC overview is from 2001-2002; it is used here only for the structural, order-of-magnitude relationship between modes (which has not qualitatively inverted since), not for this entry's headline per-pkm point estimate, which remains anchored to the EC/ERSO 2022 fact sheet above.\n","independence_note":"ETSC is an independent Brussels-based road-safety NGO, distinct from the European Commission (ERSO), FMCSA, and WHO sources above; supplies the cross-modal ratio context for the bus-vs-car and motorcycle-vs-car comparisons made in the body text.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death on a motorcycle (lifetime, US adult, population avg)","lifetime_us_adult":0.00144},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Developed countries, scheduled service (infrequent passenger)","probability":5e-7,"notes":"Rare user in the EU/US/Japan; baseline is dominated by the tiny per-pkm rate"},{"region":"Global average, typical adult exposure","probability":0.00001,"notes":"Point estimate used for the normalized field; ~1 in 100,000"},{"region":"LMIC intercity bus, frequent user","probability":0.00005,"notes":"Poor road surfaces, older fleets, longer driver hours; roughly 50x the developed-country scheduled-service rate per pkm"}],"personal_factor_multipliers":[{"factor":"frequent intercity bus traveler (monthly+ long-distance trips)","multiplier":10,"notes":"exposure-proportional; heavy users face cumulative risk dominated by rare high-fatality crashes"},{"factor":"school-bus rider (student, K-12)","multiplier":0.5,"notes":"school buses are the safest bus category per NHTSA; dedicated construction and lower speeds reduce per-trip fatality rate"},{"factor":"never rides buses","multiplier":0.01,"notes":"population average includes non-riders; non-riders face near-zero bus-crash mortality"}],"short_label":"Bus crash","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Bus and coach safety is one of the widest fans in the Likelier dataset. The per-passenger- kilometre figure used here (~0.3 deaths per billion pkm) is a developed-country, scheduled-service number from the EU. It does not describe an informal minibus on an unpaved LMIC road, which can be an order of magnitude or more worse, and it does not describe a modern European intercity coach on a motorway, which is measurably better still. Within the \"bus\" label sit school buses, urban transit, intercity coach, and chartered tourist coach — all with different fleets, different driver regimes, and different per-km rates. The normalized lifetime number is also highly exposure- dependent: for a reader who takes a bus twice a year, the real lifetime probability is effectively in the noise; for a daily commuter in a country with weak vehicle inspection regimes, it is meaningfully above the headline.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty bus stop sign post on a pale background, flat vector illustration in muted colors."},"canonical_url":"https://likelier.app/bus-crash","api_url":"https://likelier.app/api/fears/bus-crash.json"},{"slug":"pediatric-flu-encephalopathy","question":"How likely is a child with influenza to develop sudden brain swelling (influenza-associated encephalopathy or acute necrotizing encephalopathy)?","quick_answer":"For a child with the flu, the odds of developing sudden brain swelling (IAE/ANE) are roughly 1 in 100,000 per flu episode. Parents tend to sit at one of two extremes — near-zero because they have never heard of it, or, after news of pediatric deaths, far above the base rate.\n","category":"health","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"No survey measures how parents rate the odds that a child's flu turns into brain swelling. The condition entered wider awareness only after a cluster of pediatric deaths in January 2025 drew news coverage; before that most parents had never heard of influenza-associated encephalopathy. The intuitive read, when prompted, tends to split: either dismissed as something that does not happen to ordinary healthy children, or, after a news cycle about deaths in previously healthy kids, overweighted as a looming threat of any winter fever. Both readings sit far from a roughly 1-in-100,000-per-flu-episode base rate.","rough_estimate":"Either near-zero (never heard of it) or, post-headline, inflated to the range of common flu complications","kind":"intuition"},"native":{"display":"109 reported cases among an estimated ~11 million children with flu (~1 in 100,000)","numerator":109,"denominator":11000000,"unit":"US children with influenza in the 2024-25 season who developed reported IAE/ANE","population":"US children aged <18 with influenza, 2024-25 season"},"normalized":{"lifetime_us_adult":0.00001,"display":"~1 in 100,000 per child per flu episode","log_value":-5,"assumptions":"Subgroup: a US child (<18) who contracts influenza; horizon: a single influenza episode, not an adult lifetime. Numerator: 109 IAE/ANE cases reported to CDC during the 2024-25 season (MMWR mm7436a1). Denominator: pediatric symptomatic flu illnesses. CDC estimated ~51 million symptomatic influenza illnesses across all ages for 2024-25 (range 43-73M). CDC did not publish a 2024-25 pediatric illness count; children aged <18 are ~22% of the US population but carry roughly twice the adult attack rate, so a conservative pediatric-illness floor of ~22% of 51M gives ~11.2M, while children's higher attack rate would push the true denominator higher (lowering the rate). Point: 109 / 11,000,000 = 9.9e-6 ~ 1e-5. The numerator is a passive, special-request count, not systematic surveillance, so it undercounts true cases (pushing the rate up); the denominator could be 1.5-2x larger (pushing it down). Net band 1 in 30,000 (3.3e-5) to 1 in 300,000 (3.3e-6) brackets the point estimate.","uncertainty":{"low":0.0000033,"high":0.000033},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/74/wr/mm7436a1.htm","title":"Pediatric Influenza-Associated Encephalopathy and Acute Necrotizing Encephalopathy - United States, 2024-25 Influenza Season","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"109 pediatric IAE cases (37, or 34%, subcategorized as ANE); 19% of IAE patients died, 41% of ANE patients died; median age 5; 55% previously healthy; 74% ICU; only 16% of vaccine-eligible patients had received the 2024-25 flu vaccine","excerpt":"Among 192 reports of suspected IAE submitted to CDC, 109 (57%) were categorized as IAE, 37 (34%) of which were subcategorized as ANE, and 72 (66%) as other IAE. The median age of children with IAE was 5 years and 55% were previously healthy, 74% were admitted to an intensive care unit, and 19% died; 41% of children with ANE died. Only 16% of children with IAE who were vaccination-eligible had received the 2024-25 influenza vaccine.","source_date":"2025-09-25","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260609212716/https://www.cdc.gov/mmwr/volumes/74/wr/mm7436a1.htm","calculation_notes":"Provides the numerator (109 reported IAE/ANE cases in 2024-25) and the case-fatality figures. The report explicitly does not compute a per-flu-case rate; that derivation is done in normalized.assumptions using a separate CDC illness-burden denominator.","independence_note":"Primary CDC case series; the denominator burden estimate comes from a separate CDC product, so numerator and denominator are independently sourced."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12311823/","title":"Influenza-Associated Acute Necrotizing Encephalopathy in US Children","publisher":"JAMA (via PubMed Central)","source_type":"peer_reviewed","statistic":"41 ANE cases from 23 US hospitals (2023-2025); 27% died (median 3 days from onset, 91% from cerebral herniation); 76% had no significant medical history; of 32 genetically tested, 47% had risk alleles, 34% with RANBP2 variants","excerpt":"Of 58 submitted cases, 41 cases (23 females; median age, 5 years [IQR, 2-8]) from 23 US hospitals met inclusion criteria. Eleven patients (27%) died a median of 3 days (IQR, 2-4) from symptom onset, primarily from cerebral herniation (91%). Thirty-one cases (76%) had no significant medical history... Among 32 patients (78%) with genetic testing, 15 (47%) had genetic risk alleles potentially related to risk of ANE including 11 (34%) with RANBP2 variants.","source_date":"2025-07-30","source_accessed":"2026-06-13","archive_url":"https://web.archive.org/web/20260726205229/https://pmc.ncbi.nlm.nih.gov/articles/PMC12311823/","calculation_notes":"Independent multicenter clinical series corroborating ANE fatality (27%) and the RANBP2 genetic-risk signal that underpins the personal-factor multiplier. Not used for the rate denominator.","independence_note":"Separate investigator group and case ascertainment (hospital submissions) from the CDC MMWR surveillance count."},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK258641/","title":"Susceptibility to Infection-Induced Acute Encephalopathy 3 (RANBP2 / ANE1) - GeneReviews","publisher":"GeneReviews, NCBI Bookshelf (University of Washington)","source_type":"reputable_reference","statistic":"40% of RANBP2-variant heterozygotes manifest an ANE episode; 50% of affected individuals have at least one recurrence; one third die in the acute phase","excerpt":"Forty percent of heterozygotes for a RANBP2 pathogenic variant will manifest an episode of acute necrotizing encephalopathy (ANE)... Fifty per cent of persons with IIAE3 will have at least one repeat episode and some will have multiple repeat episodes... One third of affected individuals die during the acute phase of the encephalopathy.","source_date":"2020-04-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20251010012314/https://www.ncbi.nlm.nih.gov/books/NBK258641/","calculation_notes":"Source for the RANBP2 penetrance (40%) and recurrence (50%) figures behind the two large personal-factor multipliers. Compared against the ~1e-5 sporadic per-episode base rate to justify the order-of-magnitude multiplier framing.","independence_note":"Genetics reference independent of the 2024-25 surveillance and the JAMA series."},{"url":"https://www.cdc.gov/flu/whats-new/2025-2026-influenza-activity.html","title":"Influenza Activity in the United States during the 2024-25 Season and Composition of the 2025-26 Influenza Vaccine","publisher":"CDC","source_type":"govt_report","statistic":"Estimated 43-73 million symptomatic influenza illnesses (all ages) in 2024-25; 279 laboratory-confirmed pediatric deaths, the highest reported during a seasonal epidemic","excerpt":"influenza virus infection likely resulted in between 43 million-73 million symptomatic illnesses, 19 million-32 million medical visits, 560,000-1,100,000 million hospitalizations, and 38,000-99,000 deaths... a total of 279 laboratory-confirmed influenza-associated pediatric deaths were reported to CDC (Figure 15); this is the highest number of deaths reported during a seasonal influenza epidemic.","source_date":"2025-09-25","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260626004527/https://www.cdc.gov/flu/whats-new/2025-2026-influenza-activity.html","calculation_notes":"Supplies the all-ages symptomatic illness burden (mid-point ~51M, range 43-73M) used to derive the pediatric flu denominator (~11M) in normalized.assumptions. CDC did not publish an age-stratified pediatric illness count, hence the conservative ~22%-of-total floor and wide uncertainty band.","independence_note":"CDC burden model, methodologically separate from the IAE case-counting effort."}],"comparison_anchors":[{"label":"Being struck by lightning (lifetime, US)","lifetime_us_adult":0.00000478},{"label":"Dying in a plane crash (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725}],"personal_factor_multipliers":[{"factor":"Carries a RANBP2 pathogenic variant (ANE1 / IIAE3)","multiplier":50,"notes":"GeneReviews: 40% of RANBP2-variant heterozygotes will manifest an episode of acute necrotizing encephalopathy, against a sporadic per-episode risk on the order of 1e-5. The multiplier is an order-of-magnitude framing, not a precise ratio. In the 2025 JAMA US ANE series, 34% of tested cases carried RANBP2 variants. Most carriers do not know their status before a first episode."},{"factor":"Prior episode of ANE (any cause)","multiplier":50,"notes":"GeneReviews: 50% of people with RANBP2-driven IIAE3 have at least one repeat episode, some multiple. Recurrence risk is concentrated in genetically predisposed children; a first episode is the strongest known marker of a second."},{"factor":"Age 1-5 years vs school-age and older","multiplier":2,"notes":"PMC6492655 / PLOS One Japan surveillance: ANE is most prevalent in young children aged 1-5, and influenza-associated encephalopathy incidence per influenza-like-illness episode peaks in the 0-4 age band. 2024-25 CDC median age was 5. Directional, not a clean adjusted ratio."}],"short_label":"Flu brain swelling in a child (IAE/ANE)","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline rate is a derived, order-of-magnitude figure, not a measured one. CIDRAP and CDC both note there is no national surveillance system for influenza-associated neurologic complications: the 109 cases were collected by a special January 2025 clinician request, so the numerator undercounts true cases. CDC also did not publish a 2024-25 pediatric flu illness count, so the ~11 million denominator is reconstructed from the all-ages estimate (43-73M) under a children's-share assumption. Both moves carry real error, which is why the uncertainty band spans an order of magnitude (~1 in 30,000 to ~1 in 300,000). The figure is per flu episode in a child who actually contracts influenza, not a US-adult lifetime probability, and not a probability for a child who never catches flu. 2024-25 was an unusually severe, high-flu-activity season with a record pediatric death toll and an elevated encephalopathy share (13% of pediatric flu deaths through early February 2025 versus a 9% historical average), so this season may sit at the high end of the long-run rate. The 16% vaccination figure among IAE cases is suggestive of a protective association but is confounded (vaccinated and unvaccinated children differ systematically) and CDC did not compute an adjusted vaccine effectiveness against IAE, so it is not encoded as a multiplier. 8D self-score average ~4.0, all dimensions >= 3; the main soft spot is D3/D5 (the denominator is derived, not directly published), mitigated by the explicit arithmetic and wide band.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":3,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Flu brain swelling in a child (IAE/ANE)\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/pediatric-flu-encephalopathy","api_url":"https://likelier.app/api/fears/pediatric-flu-encephalopathy.json"},{"slug":"wildfire-death","question":"What are the odds of being killed by a wildfire?","quick_answer":"Direct wildfire death, the wall-of-flame version people picture, runs about 1 in 100,000 over a global lifetime. But wildfire-smoke death is roughly 100x larger, near 1 in 400, and is the real mortality story almost nobody fears.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Fear of wildfire is strongly geographic and strongly episodic. No widely cited national survey isolates \"fear of being killed by a wildfire\" from general natural-disaster or climate anxiety, so we mark the perceived side as editorial intuition. Anecdotally, the prior is shaped almost entirely by the last headline event the reader lived near or watched — Paradise, California in 2018; the Australian Black Summer of 2019-2020; Lahaina on Maui in 2023; the Los Angeles county fires of January 2025 — and reverts to roughly zero in the years between. Residents of the Western US Wildland-Urban Interface and of fire-prone parts of Mediterranean Europe and Australia carry a persistent prior; almost everyone else treats wildfire as televised rather than real, and almost nobody of any geography has an explicit prior for the much larger smoke-mortality burden downwind.\n","rough_estimate":"33.1% of US adults report being afraid or very afraid of a devastating wildfire (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~1 in 100,000 global adult lifetime (direct wildfire death)","numerator":800,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.00001,"display":"1 in ~100,000 lifetime (global adult, direct wildfire death)","log_value":-5,"assumptions":"The headline is DIRECT wildfire deaths — people killed by the fire itself (burns, entrapment during evacuation, structure collapse, heat). A separate and much larger number, wildfire-smoke mortality, is called out prominently in the body and regional_breakdown below. In normal US years civilian direct wildfire deaths are in the 5-20 range, spiking hard in megafire years: Camp Fire 2018 killed 85 people in Paradise, California; the 2023 Lahaina fire on Maui killed ~100; the January 2025 Eaton and Palisades fires killed 18 and 12 respectively. Australia's 2019-2020 Black Summer killed ~33 people directly plus several hundred more from smoke exposure. Globally, including Mediterranean Europe, Indonesia peat fires, and boreal Russia/Canada, a rough long-run direct-death anchor is a few hundred to ~1,000 per year. Using ~800/yr globally and a population of ~8 billion gives ~1e-7 per year; compounded over 60 adult years ≈ 6e-6, rounded upward to 1e-5 ≈ 1 in 100,000 to reflect the rising trend in WUI exposure in the US West and boreal forests since ~2000. This is a small number, and it is the wrong number for anyone who wants to know their actual lifetime wildfire risk — see regional_breakdown and the smoke paragraph in the body.\n","uncertainty":{"low":0.000003,"high":0.00003},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.nifc.gov/fire-information/statistics/wildfires","title":"Wildfires — Statistics","publisher":"National Interagency Fire Center (NIFC)","source_type":"govt_report","statistic":"US wildfire totals by year: 2020 = 58,950 fires / 10,122,336 acres (highest on record); 2024 = 64,897 fires / 8,924,884 acres; 2023 = 56,580 fires / 2,693,910 acres; 2025 = 77,850 fires / 5,131,474 acres","excerpt":"\"2025: 77,850 fires; 5,131,474 acres burned. 2024: 64,897 fires; 8,924,884 acres burned. 2023: 56,580 fires; 2,693,910 acres burned. ... Prior to 1983, the federal wildland fire agencies did not track official wildfire data using current reporting processes.\"\n","source_date":"2026-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411191535/https://www.nifc.gov/fire-information/statistics/wildfires","calculation_notes":"NIFC publishes annual US fire counts and acreage but not civilian fatality tolls — those are compiled downstream by the Insurance Information Institute and by NFPA from NWCG / state fire marshal reports. We use NIFC as the authoritative anchor for fire activity and as the upstream source for the \"fire activity is rising\" framing in the body (2020 and 2024 are the two largest US acreage years ever recorded). Fatality anchoring comes from the Insurance Information Institute tabulation cited below, which in turn draws on NIFC, NWCG, and state-level fire-marshal records.\n","independence_note":"NIFC is the upstream source for most US wildfire activity data; the III table cited below republishes it with added fatality commentary. Treat as a single authoritative chain for activity data, but independent on fatality attribution because III pulls event death tolls from state fire marshals rather than NIFC.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22456494/","title":"Estimated global mortality attributable to smoke from landscape fires","publisher":"Johnston FH et al., Environmental Health Perspectives (NIH/NIEHS), 2012","source_type":"peer_reviewed","statistic":"Estimated average global mortality attributable to landscape fire smoke (1997-2006): 339,000 deaths per year (interquartile range 260,000-600,000); highest regional burdens in sub-Saharan Africa (157,000) and Southeast Asia (110,000)","excerpt":"\"Our principal estimate for the average mortality attributable to LFS exposure was 339,000 deaths annually. In sensitivity analyses the interquartile range of all tested estimates was 260,000-600,000.\"\n","source_date":"2012-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260523092329/https://pubmed.ncbi.nlm.nih.gov/22456494/","calculation_notes":"Johnston's 339,000 deaths/year is the canonical peer-reviewed estimate for global wildfire-smoke mortality. 339,000 / 8,000,000,000 ≈ 4.24e-5 per year; compounded over 60 adult years ≈ 2.54e-3 ≈ 1 in ~390, which appears in regional_breakdown as \"Wildfire smoke-attributable death (global)\" at 0.0025. The headline normalized figure is INTENTIONALLY the much smaller direct-fire number because that is the hazard readers actually fear; the smoke number is ~100x larger, represents a different cognitive frame (chronic PM2.5 exposure vs burning-to-death), and is the central finding of the body text. Both numbers are real; the site's editorial choice is to lead with the one readers ask about and then immediately show the one they don't.\n","independence_note":"Johnston 2012 is a primary peer-reviewed estimate from the NIEHS/EHP literature. Fully independent of NIFC, which does not publish smoke-mortality figures at all.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-wildfires","title":"Facts + Statistics: Wildfires","publisher":"Insurance Information Institute (data sourced from NIFC / NFPA / state fire marshals)","source_type":"reputable_reference","statistic":"Camp Fire (November 2018): 85 deaths, deadliest California wildfire on record; 2024 US wildfire activity: 64,897 fires / 8,924,884 acres; Eaton Fire (January 2025): 18 deaths; Palisades Fire (January 2025): 12 deaths","excerpt":"\"The Camp Fire (November 2018) resulted in 85 deaths, making it the deadliest California wildfire on record. ... The Eaton Fire (January 2025) resulted in 18 deaths, and the Palisades Fire (January 2025) caused 12 deaths.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260324081914/https://www.iii.org/fact-statistic/facts-statistics-wildfires","calculation_notes":"Used as the event-level anchor for direct-death tolls in the body and in regional_breakdown. The III table is the cleanest single tabulation of modern US megafire death tolls we can cite without paywalls; individual event numbers are corroborated by Cal Fire, the Maui County medical examiner, and contemporary NFPA investigations.\n","independence_note":"III republishes NIFC data for activity counts. Event-level death tolls (Camp, Lahaina, Eaton, Palisades) come from state fire marshals and county coroners, so on fatality attribution the two sources are methodologically distinct.\n"},{"url":"https://www.pnas.org/doi/10.1073/pnas.2403960121","title":"Long-term exposure to wildland fire smoke PM2.5 and mortality in the contiguous United States","publisher":"Ma Y et al. (Yale School of Public Health), Proceedings of the National Academy of Sciences (PNAS), 2024","source_type":"peer_reviewed","statistic":"Long-term exposure to wildland fire smoke PM2.5 contributed to ~11,415 non-accidental deaths/year (95% CI 6,754-16,075) in the contiguous US, 2007-2020, of which ~4,512/year (95% CI 1,922-7,102) were cardiovascular; excess mortality was concentrated in adults aged 65+ (7.08 deaths per 100,000/month vs 0.55 for ages 0-64 in the highest 5+ µg/m³ smoke-PM2.5 bin)","excerpt":"\"approximately 11,415 non-accidental deaths (95% CI: 6,754, 16,075) were attributable to smoke PM2.5 in the contiguous U.S. per year ... approximately 4,512 deaths (95% CI: 1,922, 7,102) from cardiovascular diseases ... greater increases in mortality rates were observed among people who aged above 65 across all smoke PM2.5 concentration bins.\"\n","source_date":"2024-09-24","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250803191301/https://www.pnas.org/doi/10.1073/pnas.2403960121","calculation_notes":"Ma et al. 2024 is the modern US-specific replacement for the older global Johnston 2012 anchor on smoke mortality. 11,415 non-accidental deaths/year across a contiguous-US adult population of ~258 million ≈ 4.42e-5 per year; compounded over 59 adult years ≈ 2.6e-3 ≈ 1 in ~385, which corroborates (rather than revises) the global \"smoke-attributable\" row in regional_breakdown at 0.0025. The headline normalized figure is INTENTIONALLY the much smaller DIRECT-fire number; this source modernizes and US-localizes the smoke-mortality counterpoint that is the central finding of the body text. The age-65+ stratification (7.08 vs 0.55 deaths/100k/month at the highest exposure bin, a ~13x excess-rate ratio) is the source for the new \"age 65+ (smoke exposure)\" personal_factor_multiplier.\n","independence_note":"Ma et al. 2024 (PNAS, Yale) is an independent primary peer-reviewed US-specific estimate, methodologically distinct from Johnston 2012 (global, EHP) and from NIFC/III activity and direct-death data, which do not address smoke-attributable mortality at all.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death by tornado (lifetime, US adult average)","lifetime_us_adult":0.0000124},{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by earthquake (lifetime, global adult average)","lifetime_us_adult":0.000263},{"label":"Death by tropical cyclone (lifetime, global adult average)","lifetime_us_adult":0.000112},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Direct wildfire death (global average)","probability":0.00001,"notes":"Headline figure — people killed by fire itself (burns, entrapment, structure collapse)."},{"region":"Wildfire smoke-attributable death (global)","probability":0.0025,"notes":"Johnston et al. 2012 — ~339,000 deaths/year globally from landscape fire PM2.5 exposure. ~100x the direct-fire figure and the main wildfire mortality story."},{"region":"US smoke-attributable death (modern, 2007-2020)","probability":0.0026,"notes":"Ma et al. 2024 (PNAS, Yale) — ~11,415 non-accidental US deaths/year from long-term wildland-fire smoke PM2.5 (~4,512 cardiovascular). Over a 59-year adult life across ~258M US adults this is ~1 in 385, corroborating the global smoke row above. The newer US-specific anchor; Qiu et al. 2025 (Nature) project >71,000/year by 2050 under high warming."},{"region":"California WUI resident, direct fire","probability":0.0002,"notes":"Wildland-Urban Interface residents of fire-prone California face a direct-fire lifetime risk roughly an order of magnitude above the US average — driven by Paradise 2018, Tubbs 2017, Eaton and Palisades 2025."},{"region":"Urban/suburban smoke exposure, California 2020-2025","probability":0.005,"notes":"Downwind PM2.5 exposure from the 2020 August Complex, 2021 Dixie Fire, and 2025 LA county fires has produced multi-week air-quality events affecting tens of millions. Cardiopulmonary excess mortality in these windows is the dominant modern wildfire death mechanism in the US West."},{"region":"Indonesia peat-fire smoke (2015 El Niño event)","probability":0.003,"notes":"Koplitz et al. 2016 estimated ~100,000 excess deaths from the 2015 Indonesia peat fire haze across Indonesia, Malaysia, and Singapore — a single event comparable in smoke mortality to a decade of global direct wildfire deaths."},{"region":"Inland temperate / high-latitude cities with no wildfire exposure","probability":1e-7,"notes":"Essentially zero direct-fire risk; smoke exposure still non-zero during hemispheric-scale events (Canadian boreal smoke over US East Coast and Western Europe in June 2023)."}],"personal_factor_multipliers":[{"factor":"Wildland-Urban Interface (WUI) residence","multiplier":20,"notes":"WUI residents account for a disproportionate share of US direct wildfire fatalities. The multiplier applies to direct-fire risk, not smoke exposure."},{"factor":"defensible space + hardened structure (ember-resistant vents, Class A roof)","multiplier":0.3,"notes":"Post-Camp Fire analyses and IBHS fire-performance data show hardened structures survive ember attack at substantially higher rates; applies to direct-fire risk."},{"factor":"pre-existing cardiopulmonary disease (smoke exposure)","multiplier":3,"notes":"Applies to wildfire-smoke mortality, not direct fire. PM2.5 exposure during smoke events produces large short-term increases in cardiac and respiratory event rates, concentrated in people with existing disease."},{"factor":"firefighter occupational exposure","multiplier":10,"notes":"Wildland firefighters face substantially elevated direct-fire risk during deployment seasons and elevated chronic smoke exposure over a career. The multiplier is a rough career-average anchor, not a shift-level estimate."},{"factor":"age 65+ (smoke exposure)","multiplier":13,"notes":"Applies to wildfire-smoke mortality, not direct fire. Ma et al. 2024 (PNAS, Yale) found excess smoke-PM2.5 mortality concentrated in adults 65+: 7.08 deaths per 100,000/month versus 0.55 for ages 0-64 in the highest (5+ µg/m³) smoke-PM2.5 bin — roughly a 13x excess-rate ratio. The multiplier reflects that excess-mortality ratio, not a direct-fire shift."}],"short_label":"Wildfire","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline \"1 in 100,000\" number only covers direct wildfire deaths — the vivid, cinematic version of the fear. It is the wrong number in two directions. It is too high for almost everyone who does not live in a Wildland-Urban Interface community on a fire-prone coast, mountainside, or boreal forest edge: a resident of downtown Chicago or central London has essentially zero direct wildfire mortality risk. And it is dramatically too low as a description of total wildfire-attributable mortality, because it excludes smoke exposure. Johnston et al. 2012 estimated ~339,000 global deaths per year from landscape fire smoke, and subsequent work on the 2015 Indonesia peat fire haze (Koplitz et al.) and on recent Western US seasons suggests the number has grown since. Smoke mortality is roughly 100x the direct-fire mortality and is concentrated in populations — urban, downwind, cardiopulmonary-compromised — that do not typically see themselves as at wildfire risk at all. The gap between how the fear is framed (a wall of flame in your neighborhood) and where the deaths actually happen (a cardiac event on day five of a PM2.5 spike 400 km downwind) is the central feature of this entry. Modern US-specific work sharpens the gap: Ma et al. 2024 (PNAS, Yale) attribute ~11,415 non-accidental US deaths per year (95% CI 6,754-16,075) to long-term wildland-fire smoke PM2.5 over 2007-2020, of which ~4,512 are cardiovascular, with the excess concentrated in adults aged 65+. Qiu et al. 2025 (Nature) project that figure could exceed 71,000 deaths per year by 2050 under a high-warming scenario — a ~73% rise over the 2011-2020 baseline. The January 2025 Los Angeles county fires are the same story in miniature: against roughly 31 direct deaths (18 Eaton + 12 Palisades + 1), the projected indirect smoke-attributable toll runs into the thousands.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized ember shape rising against a muted dusty sky, rendered as a flat geometric spark, vector illustration."},"canonical_url":"https://likelier.app/wildfire-death","api_url":"https://likelier.app/api/fears/wildfire-death.json"},{"slug":"acid-attack","question":"What are the odds of being a victim of an acid attack?","quick_answer":"Roughly 1 in 94,000 for a global adult, and far lower still for anyone living outside the handful of hotspot countries. Dramatic media coverage makes it feel common; the actual incidence is extraordinarily low.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Acid attacks generate media coverage wildly disproportionate to their frequency, particularly in the UK press following a spike in London incidents around 2017. The combination of permanent disfigurement, photogenic injury, and a visceral \"it could happen to anyone\" narrative inflates perceived risk far beyond the actuarial base rate. No systematic survey measures public fear of acid attacks specifically, but analogous research on fear of violent crime consistently shows that rare, visually dramatic offenses are overestimated while common ones are underestimated. The global rate is extraordinarily low; the fear is not.\n","rough_estimate":"dramatically overestimated relative to actual incidence, especially after media coverage of UK attacks","kind":"intuition"},"native":{"display":"~1,040–1,817 reported acid attacks per year globally (likely undercount)","numerator":18,"denominator":100000000,"unit":"per year","population":"global population"},"normalized":{"lifetime_us_adult":0.0000106,"display":"~1 in 94,000 lifetime (global adult)","log_value":-4.97,"assumptions":"World Population Review, aggregating Acid Survivors Trust International (ASTI) 2018 country figures, states that \"there were somewhere between 1,040 and 1,817 acid attacks reported from countries around the world.\" Taking the midpoint of that reported range (~1,428 attacks/year) among ~8 billion global population yields an annual rate of about 0.18 per million (1.8 per 10 million, or approximately 0.0000002 per person per year). Over a 59-year adult lifetime at constant hazard: 1 − (1 − 0.0000002)^59 ≈ 0.0000106. The uncertainty band spans the endpoints of the reported range — the low end (~1,040/year → lifetime ≈ 0.0000077) and the high end (~1,817/year → lifetime ≈ 0.0000134) — widened further because the same source notes acid attacks \"should be interpreted with caution\" given underreporting, and because attacks are heavily concentrated in a handful of countries, so the true global-average rate for an adult outside these hotspots is substantially lower than the pooled figure.\n","uncertainty":{"low":8e-7,"high":0.00002},"scope":"global_adult_lifetime"},"sources":[{"url":"https://asti.org.uk/a-worldwide-problem/","title":"A Worldwide Problem","publisher":"Acid Survivors Trust International","source_type":"reputable_reference","statistic":"Acid attacks are systematically under-reported; most developing countries lack a national monitoring system, and many victims never report the true cause of injury","excerpt":"\"Most developing countries do not have a comprehensive national system for recording and monitoring attacks. Furthermore, where there are resources available, these are usually focused in urban areas, leading to greater under-reporting in rural areas. Victims of acid and burns violence often do not report the true cause of the injuries out of shame or fear. People may never present themselves to hospitals for treatment, so their cases go unrecorded.\"\n","source_date":"2026-04-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260420030919/https://asti.org.uk/a-worldwide-problem/","calculation_notes":"ASTI is the primary international organization tracking acid violence globally. This page does not state a headline global count; it establishes the qualitative basis for the \"likely undercount\" caveat — that most affected countries lack comprehensive recording systems and that victims frequently do not report attacks. The point-estimate global count used here is sourced instead to the World Population Review aggregation of ASTI country figures (below), which states a reported range of 1,040–1,817 attacks per year. The wide uncertainty band (0.0000008 to 0.0000200) reflects both the underreporting documented on this ASTI page and the extreme geographic concentration of attacks.\n"},{"url":"https://worldpopulationreview.com/country-rankings/acid-attack-statistics-by-country","title":"Acid Attack Statistics by Country 2026","publisher":"World Population Review","source_type":"reputable_reference","statistic":"Between 1,040 and 1,817 acid attacks reported worldwide per year (ASTI 2018 country figures); ~80% of victims are women, though the gender split reverses in several countries","excerpt":"\"Overall, there were somewhere between 1,040 and 1,817 acid attacks reported from countries around the world. This number is disturbing, but on the decline… India's rising number of acid attacks is believed to be vastly underreported. While 2018 statistics showed 228 incidents, researchers believe there were more than 1,000… The Acid Survivors Trust International found that around 80% of acid attack victims are women and most of these attacks are perpetrated by men in domestic violence situations. The gender statistics of these attacks differ by country, however… From 2011-2016, the United Kingdom saw more than twice as many acid attacks on men than women.\"\n","source_date":"2026-02-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260215121347/https://worldpopulationreview.com/country-rankings/acid-attack-statistics-by-country","calculation_notes":"World Population Review aggregates ASTI 2018 country figures (UK 501, India 228–1,000, Colombia 100, Bangladesh 91, Pakistan 80, Uganda 35, Cambodia >10) and states the global total as \"somewhere between 1,040 and 1,817 acid attacks reported from countries around the world.\" The point estimate here uses the midpoint of that range (~1,428/year); at 1,428/year among 8 billion people: annual rate = 1,428/8,000,000,000 ≈ 1.8 × 10⁻⁷. Over 59 adult years: 1 − (1 − 1.8 × 10⁻⁷)^59 ≈ 1.06 × 10⁻⁵ ≈ 0.0000106 ≈ 1 in 94,000. The uncertainty low (0.0000077) and high (0.0000134) correspond to the range endpoints (1,040 and 1,817/year); the entry's band is widened beyond these to 0.0000008–0.0000200 to absorb underreporting and the extreme geographic concentration of attacks (India alone believed to exceed 1,000 despite 228 officially reported).\n"},{"url":"https://en.wikipedia.org/wiki/Acid_attack","title":"Acid attack","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"Per ASTI, 80% of acid attacks are on women and assaults are grossly under-estimated; in the UK the majority of victims and perpetrators are male","excerpt":"\"An acid attack… is a form of violent assault involving the act of throwing acid or a similarly corrosive substance onto the body of another 'with the intention to disfigure, maim, torture, or kill'. The most common types of acid used in these attacks are sulfuric and nitric acid… According to the London-based charity Acid Survivors Trust International, 80% of acid attacks are on women, and acid assaults are grossly under-estimated… However, in countries such as the United Kingdom where acid attacks are associated primarily with gang violence, the majority of both perpetrators and victims are male.\"\n","source_date":"2026-04-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260414155739/https://en.wikipedia.org/wiki/Acid_attack","calculation_notes":"Wikipedia (via ASTI) corroborates the 80% female victim share and the \"grossly under-estimated\" caveat, and documents the regional gender reversal (male-majority in the UK, female-majority in South Asia linked to domestic disputes, dowry conflicts, or rejected proposals). It does not itself state a headline global count, so the point estimate is not sourced here — this entry is supplementary demographic/context support, not the source of the numerator.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Shark attack fatality (lifetime, global)","lifetime_us_adult":2.5e-7},{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"Women in South Asia (India, Bangladesh, Pakistan)","multiplier":50,"notes":"Highly concentrated in this demographic and region"},{"factor":"Men in UK urban areas","multiplier":10,"notes":"UK has unusually high male victim rate linked to gang violence"},{"factor":"Adults in countries with low reported incidence","multiplier":0.1,"notes":"Most of the world has near-zero reported rates"},{"factor":"Intimate partner or domestic dispute context (South Asia)","multiplier":3,"notes":"ASTI 2019 global data: a majority of attacks on women in South Asia are perpetrated by intimate partners or family members in domestic dispute contexts, making relationship conflict a significant risk amplifier in high-incidence regions"},{"factor":"Residence in South or Southeast Asia outside hotspot countries","multiplier":10,"notes":"Per-capita acid attack rates in India, Bangladesh, and Pakistan are estimated at 10x or more above the rest of the world per ASTI country-level data; residing in these countries elevates risk above the global average regardless of personal relationship context"}],"short_label":"Acid attack","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The reported global count — somewhere between 1,040 and 1,817 attacks per year in the cited aggregation of ASTI 2018 country figures — is almost certainly an undercount. ASTI states that most affected countries lack a comprehensive national recording system and that victims frequently do not report the true cause of their injuries; India's official 228 reported cases in 2018 are believed by researchers to exceed 1,000. The geographic concentration is extreme: a handful of countries (led by the UK, India, Bangladesh, Uganda, and Colombia) account for the vast majority of reported cases. The demographic profile differs sharply by region — ASTI finds around 80% of victims worldwide are women, with attacks typically motivated by domestic disputes, dowry conflicts, or rejected proposals; but in the UK the majority of victims are men and attacks are often linked to gang violence, and in parts of Africa and in Cambodia roughly half or more of victims are male. The normalized lifetime figure of ~1 in 94,000 is a global average that overstates risk for most of the world's population and understates it dramatically for women in the most affected regions. Acid attack legislation varies widely: Bangladesh, Colombia, Pakistan, and Cambodia have all seen steep declines, which the source attributes in part to harsher sentencing, suggesting that policy interventions can meaningfully shift this number.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Abstract illustration of a tilted laboratory flask in muted tones, flat editorial style."},"canonical_url":"https://likelier.app/acid-attack","api_url":"https://likelier.app/api/fears/acid-attack.json"},{"slug":"tornado-death","question":"What are the odds of being killed by a tornado?","quick_answer":"Being killed by a tornado runs about 1 in 80,000 over a lifetime as a US national average, a figure essentially meaningless for any specific resident since the risk is heavily concentrated by geography. About 35% of US adults report fearing a devastating tornado.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Fear of tornadoes is strongly geographic — survey work on tornado anxiety is concentrated in the Plains and the Southeast, and there is no widely cited national poll that isolates \"fear of being killed by a tornado\" from general severe-weather anxiety. We mark the perceived side as editorial intuition. Anecdotally, residents of Dixie Alley and Tornado Alley tend to overestimate per-year risk in low-activity years and underestimate it in outbreak years; residents of the Northeast and West Coast tend to treat tornadoes as a near-zero hazard, which for them is roughly accurate.\n","rough_estimate":"34.7% of US adults report being afraid or very afraid of a devastating tornado (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~70 tornado fatalities per year in the US (long-run average)","numerator":70,"denominator":333000000,"unit":"per year","population":"US residents"},"normalized":{"lifetime_us_adult":0.0000124,"display":"1 in ~80,000 lifetime (US adult, national average)","log_value":-4.91,"assumptions":"Uses ~70 US tornado fatalities per year as a long-run average that smooths out the extreme year-to-year variability documented by NOAA SPC and the Insurance Information Institute (annual totals between 2014 and 2024 ranged from 10 in 2018 to 103 in 2021). Divides by US population (~333M) and compounds over 59 years of remaining adult life. The lifetime figure is a national average and is essentially meaningless for any specific resident — see caveats.\n","uncertainty":{"low":0.000007,"high":0.00002},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.spc.noaa.gov/wcm/","title":"SPC Warning Coordination Meteorologist Page — Tornado, Hail, and Wind Statistics","publisher":"NOAA / NWS Storm Prediction Center","source_type":"govt_report","statistic":"30-year (1996-2025), 20-year (2006-2025), and 10-year (2016-2025) state-level annual averages of US tornado fatalities, derived from the SPC tornado database (1950-2024)","excerpt":"\"Annual Averages: Tornado Fatalities by State — 30 Year Average (1996-2025), 20 Year Average (2006-2025), 10 Year Average (2016-2025). The SPC tornado database CSV files cover 1950 through 2024.\"\n","source_date":"2026-01-12","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260411093700/https://www.spc.noaa.gov/wcm/","calculation_notes":"The SPC publishes the underlying yearly fatality counts and state-level averages we use to anchor the long-run ~70/year national figure. SPC's per-state maps are also what makes the geographic non-uniformity quantitative — fatalities are heavily concentrated in Alabama, Mississippi, Tennessee, Arkansas, Missouri, and the Plains states.\n","independence_note":"SPC is the upstream source for most other US tornado fatality datasets, including the Insurance Information Institute table cited below, so these two sources should be treated as a single authoritative chain rather than two independent estimates.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-tornadoes-and-thunderstorms","title":"Facts + Statistics: Tornadoes and Thunderstorms","publisher":"Insurance Information Institute (data sourced from NOAA SPC / NWS)","source_type":"reputable_reference","statistic":"US tornado fatalities by year, 2014-2024: 47, 36, 18, 35, 10, 42, 76, 103, 23, 83, 54 (mean ≈ 48/year for that window)","excerpt":"\"Tornadoes and Related Deaths in the United States, 2006-2025. 2014: 47; 2015: 36; 2016: 18; 2017: 35; 2018: 10; 2019: 42; 2020: 76; 2021: 103; 2022: 23; 2023: 83; 2024: 54. Source: U.S. Department of Commerce, Storm Prediction Center, National Weather Service.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260411093715/https://www.iii.org/fact-statistic/facts-statistics-tornadoes-and-thunderstorms","calculation_notes":"The 2014-2024 window averages ~48/year, but is depressed by several unusually quiet years (2016, 2018, 2022). Longer windows that include the 2011 super-outbreak year (553 deaths) push the long-run average closer to 70/year, which is why we use 70 as the headline native figure rather than the most recent decade's mean. Either choice gives a lifetime risk between 1 in 80,000 and 1 in 120,000.\n","independence_note":"I.I.I. republishes NOAA SPC / NWS data verbatim. Treated as a convenient year-by-year tabulation, not as independent verification.\n"},{"url":"https://www.nssl.noaa.gov/users/brooks/public_html/deathtrivia/","title":"A Brief History of Deaths from Tornadoes in the United States","publisher":"NOAA National Severe Storms Laboratory (Harold Brooks)","source_type":"peer_reviewed","statistic":"Mean annual US tornado death toll fell from ~260 (1912-1936) to ~54 (1976-2000); per-capita rate dropped roughly 15-fold from ~1.7 per million in the early 20th century to ~0.12 per million by 2000","excerpt":"\"The mean annual death toll [in 1912-1936] was 260, almost five times as many as in 1976-2000, when the mean was 54. ... Death rates remained relatively constant at approximately 1.6-1.8 per million population annually [pre-1925]. After 1925, rates declined sharply to roughly 0.12 per million by 2000 — approximately a 15-fold decrease.\"\n","source_date":"2009-03-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260411093735/https://www.nssl.noaa.gov/users/brooks/public_html/deathtrivia/","calculation_notes":"Brooks's long-run NSSL analysis is the basis for the claim that the modern era is qualitatively different from the early-20th-century baseline. We use his per-capita figure (~0.12 per million ≈ 0.012 per 100k) as a sanity check on our 70/333M ≈ 0.021 per 100k headline; the small drift upward reflects population growth and the inclusion of high-outbreak years like 2011.\n","independence_note":"Brooks's analysis predates SPC's most recent decade of data and uses NWS Storm Data directly, so it provides genuine methodological independence on the historical trend even though the underlying database is shared.\n"},{"url":"https://www.nature.com/articles/s41612-018-0048-2","title":"Spatial trends in United States tornado frequency","publisher":"Gensini & Brooks, npj Climate and Atmospheric Science (Nature)","source_type":"peer_reviewed","statistic":"Over 1979-2017, significant increasing trends in tornado reports and tornado-favorable environments across portions of Mississippi, Alabama, Arkansas, Missouri, Illinois, Indiana, Tennessee, and Kentucky, with significant decreasing trends over portions of Texas, Oklahoma, and northeast Colorado","excerpt":"\"We find significant decreasing trends in frequencies of both tornado reports and tornado environments over portions of the central and southern Great Plains (e.g., Texas and Oklahoma), and significant increasing trends over portions of the Midwest and Southeast United States (e.g., Mississippi, Alabama, Arkansas, Missouri, Illinois, Indiana, Tennessee, and Kentucky), 1979-2017.\"\n","source_date":"2018-10-17","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260531074612/https://www.nature.com/articles/s41612-018-0048-2","calculation_notes":"Gensini & Brooks does not change the national ~70/year headline figure; it supports the entry's central caveat that the spatial distribution of tornado risk is migrating eastward toward the more densely housed, more nocturnal-tornado-prone Southeast, which is why the Dixie-Alley personal multiplier is set above the Plains. It is a frequency- and-environment trend study, not a fatality count, so it informs the geographic multiplier rather than the per-year fatality denominator.\n","independence_note":"Uses the SPC tornado database for reports but adds an independent reanalysis-based Significant Tornado Parameter (STP) environment analysis, providing methodological independence from a pure report-count climatology.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"lives in Dixie Alley (MS, AL, TN)","multiplier":5,"notes":"Brooks & Doswell: nocturnal tornadoes and mobile-home prevalence drive fatality concentration in the southeastern US; Gensini & Brooks (2018) document significant increasing tornado-frequency trends across MS, AL, AR, TN and the mid-South over 1979-2017"},{"factor":"lives in mobile/manufactured home","multiplier":10,"notes":"NOAA: mobile homes account for ~40% of tornado deaths despite ~6% of housing stock"},{"factor":"lives in Pacific Northwest or Northeast urban area","multiplier":0.1,"notes":"tornado frequency and intensity are far lower outside the central and southeastern US"},{"factor":"no basement or below-grade shelter available","multiplier":3,"notes":"NOAA Storm Prediction Center and NWS post-storm surveys consistently show that above-ground sheltering in tornado-prone areas substantially increases fatality risk compared to below-grade shelter; interior rooms on lowest floors provide incomplete protection against EF3+ tornadoes"},{"factor":"nighttime tornado (occurring between midnight and 6 am)","multiplier":2.5,"notes":"NWS fatality data and Ashley et al. (2008) climatology analysis document that nocturnal tornadoes are significantly more lethal per event than daytime events, because most residents are asleep and do not receive or respond to warning sirens in time to seek shelter"}],"short_label":"Tornado","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The national-average lifetime figure (~1 in 80,000) is almost meaningless for any individual reader. Tornado fatalities cluster heavily in the Southeast (Dixie Alley: Alabama, Mississippi, Tennessee, Arkansas) and the central Plains (Tornado Alley: Oklahoma, Kansas, Missouri, Texas). Ashley's climatology work shows that Dixie Alley has historically produced more tornado fatalities than Tornado Alley despite fewer total tornadoes, largely because Dixie Alley storms are more often nocturnal, rain-wrapped, and strike denser, more vulnerable housing stock. A resident of central Mississippi faces a per-year risk many times the national average; a resident of Vermont, Oregon, or coastal California faces a risk close to zero. Risk also depends heavily on housing type — manufactured-home residents account for a disproportionate share of fatalities relative to their share of the population. The geographic distribution is also not static: Gensini & Brooks (2018) document a multi-decade eastward migration of tornado frequency and favorable environments — significant increases across Mississippi, Alabama, Arkansas, Tennessee, and the mid-South, significant decreases over Texas and Oklahoma (1979-2017) — concentrating activity in the more densely housed, more nocturnal-prone Southeast. Year-to-year variance remains extreme: 2025 was a well-above-average season with about 35 tornado deaths by late April (NOAA Climate.gov), illustrating again why a single year's toll tells an individual reader very little about long-run risk.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized funnel shape rendered as a tapered vertical spiral against a muted overcast sky, flat vector illustration."},"canonical_url":"https://likelier.app/tornado-death","api_url":"https://likelier.app/api/fears/tornado-death.json"},{"slug":"bee-sting-fatal","question":"What are the odds of being killed by a bee, wasp, or hornet sting?","quick_answer":"About 1 in 79,000 US adults are killed by a bee, wasp, or hornet sting over a lifetime — small, but far from zero. Most people, thinking of bees as harmless, guess essentially zero outside of a known allergy.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"We don’t yet have a rigorous survey that isolates “fear of being killed by a bee, wasp, or hornet sting” from the much broader category of entomophobia or stinging-insect avoidance. Anecdotally, the bee occupies a strange place in risk perception: familiar, small, and culturally cute (honeybees especially), so the perceived fatal risk runs well below the actual number for most people. Even those who actively dislike wasps tend to file the danger as “painful” rather than “potentially lethal,” unless they already know they’re allergic.\n","rough_estimate":"most people guess essentially zero outside of known allergies","kind":"intuition"},"native":{"display":"~72 hornet, wasp, and bee sting deaths per year, United States","numerator":72,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.00001267,"display":"1 in ~79,000 lifetime (US adult)","log_value":-4.9,"assumptions":"Uses the CDC NCHS published average of 72 deaths per year from hornet, wasp, and bee stings (ICD-10 code X23, “Contact with hornets, wasps and bees”) over the most recent fully reported decade (2011-2021), divided by a US population of ~335 million, then compounded over 59 years of remaining adult life: 1 - (1 - 72/335000000)^59 ≈ 1.27 × 10^-5. The 2000-2017 average was lower (62/year); the uncertainty band reflects the plausible range across reporting windows and the upward drift in recent years.\n","uncertainty":{"low":0.000009,"high":0.000018},"scope":"us_adult_lifetime"},"sources":[{"url":"https://blogs.cdc.gov/nchs/2023/07/07/7414/","title":"QuickStats: Number of Deaths from Hornet, Wasp, and Bee Stings Among Males and Females — National Vital Statistics System, United States, 2011-2021","publisher":"US Centers for Disease Control and Prevention (CDC) / National Center for Health Statistics (NCHS) / MMWR","source_type":"govt_report","statistic":"788 US deaths from hornet, wasp, and bee stings over 2011-2021 (average 72 per year)","excerpt":"\"During 2011-2021, a total of 788 deaths from hornet, wasp, and bee stings occurred (an average of 72 deaths per year). The annual number of deaths ranged from 59 (2012) to 89 (2017). Overall, 84% of deaths occurred among males.\"\n","source_date":"2023-07-07","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413163744/https://blogs.cdc.gov/nchs/2023/07/07/7414/","calculation_notes":"CDC NCHS reports an average of 72 US deaths per year (2011-2021) under ICD-10 code X23 (“Contact with hornets, wasps and bees”), drawn from death certificates in the National Vital Statistics System. Divided by US population (~335M) and compounded over 59 years of remaining adult life gives ~1 in 79,000 lifetime. The earlier 2000-2017 reporting window averaged 62/year, so the uncertainty band brackets both periods.\n","independence_note":"This is a death-certificate-based count via NCHS / NVSS. It is the same underlying ICD-10 data stream that WISQARS exposes interactively.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/68/wr/mm6829a5.htm","title":"QuickStats: Number of Deaths from Hornet, Wasp, and Bee Stings, Among Males and Females — National Vital Statistics System, United States, 2000-2017","publisher":"US Centers for Disease Control and Prevention (CDC) / MMWR Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"1,109 US deaths from hornet, wasp, and bee stings over 2000-2017 (annual average 62)","excerpt":"\"During 2000-2017, a total of 1,109 deaths from hornet, wasp, and bee stings occurred, for an annual average of 62 deaths. Deaths ranged from a low of 43 in 2001 to a high of 89 in 2017. Approximately 80% of the deaths were among males.\"\n","source_date":"2019-07-26","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163749/https://www.cdc.gov/mmwr/volumes/68/wr/mm6829a5.htm","calculation_notes":"Used as a longer-window corroboration of the NCHS 2011-2021 figure. The 18-year mean of 62/year is slightly lower than the more recent 11-year mean of 72/year, consistent with a small upward drift; we use 72 as the central estimate and use the 2000-2017 average as the lower bound of the uncertainty range.\n","independence_note":"Same underlying NVSS death-certificate data stream as the 2011-2021 QuickStats; this is a temporal cross-check, not an independent count.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1961691/","title":"Insect Sting Anaphylaxis","publisher":"Immunology and Allergy Clinics of North America / Golden DBK (NIH PubMed Central)","source_type":"peer_reviewed","statistic":"≥50 fatal sting reactions per year in the US; ~3% of adults report systemic allergic reactions to stings","excerpt":"\"Systemic allergic reactions are reported by up to 3% of adults, and almost 1% of children have a medical history of severe sting reactions &hellip; At least 50 fatal sting reactions occur each year in the United States &hellip; Half of all fatal reactions occur with no history of previous sting reactions.\"\n","source_date":"2007-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260503075259/https://pmc.ncbi.nlm.nih.gov/articles/PMC1961691/","calculation_notes":"Used as an independent allergy-epidemiology corroboration of the order of magnitude (50+ deaths per year) and as the source for the heterogeneity story: roughly half of fatal sting reactions occur in people with no prior history of systemic reaction, which means “I’m not allergic” is a weaker filter on personal risk than most readers assume.\n","independence_note":"Methodologically independent of NCHS death-certificate counts: Golden’s review draws from clinical allergy literature and venom-IgE serology studies rather than ICD-10 death codes.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death by dog bite or strike (lifetime, US adult)","lifetime_us_adult":0.00000704},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"known venom allergy (unmedicated)","multiplier":10,"notes":"Golden 2007: ~50% of fatal reactions occur in people with NO prior allergic history, but known-allergy individuals carry much higher per-sting risk"},{"factor":"carries epinephrine auto-injector","multiplier":0.1,"notes":"timely epinephrine reduces case fatality by ~90%"},{"factor":"beekeeper or outdoor agricultural worker","multiplier":3,"notes":"higher sting frequency increases annual exposure"},{"factor":"urban resident, minimal outdoor work","multiplier":0.3,"notes":"lower encounter rate"}],"short_label":"Bee sting","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a population-level average over all US adults. Actual risk is highly heterogeneous: individuals with a known systemic allergy to hymenoptera venom face a much higher per-sting fatality probability than the general population, and occupational groups with sustained outdoor exposure (beekeepers, landscapers, roofers, agricultural workers) accumulate orders of magnitude more sting events per year. Conversely, an indoor-dwelling adult with no known venom allergy faces something close to zero. Note also that roughly half of fatal sting reactions, per the allergy literature, occur in people with no prior history of systemic reaction — so “I’ve been stung before and was fine” is a weaker reassurance than it sounds.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized honeybee silhouette resting on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/bee-sting-fatal","api_url":"https://likelier.app/api/fears/bee-sting-fatal.json"},{"slug":"terrorism-us-civilian","question":"What are the odds of dying in a terrorist attack in the US?","quick_answer":"Dying in a terrorist attack in the US runs about 1 in 77,000 over a lifetime, near 1 in 4.6 million in a given year. Many worried Americans picture something like 1 in a few thousand, orders of magnitude above the actual rate.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Gallup has asked Americans since the mid-1990s how worried they are about personally becoming a victim of terrorism. The most recent reading in Gallup's crime-worry battery (October 2022) had 27% of US adults saying they worry \"a great deal\" or \"a fair amount\" about being the victim of terrorism — the lowest in that wave's 13-crime list apart from workplace assault. Earlier spikes after 9/11 and again in 2015–2016 pushed the same measure close to 50%. Even the current \"low\" reading is orders of magnitude above the actual base rate.\n","rough_estimate":"~1 in a few thousand lifetime feels about right to many worried respondents","kind":"poll","survey_source":{"title":"Record-High 56% in U.S. Perceive Local Crime Has Increased","publisher":"Gallup","url":"https://news.gallup.com/poll/404048/record-high-perceive-local-crime-increased.aspx","year":2022}},"native":{"display":"~1 in 4,560,000 per year","numerator":1,"denominator":4560000,"unit":"per year","population":"US residents, 1975–2024 pooled, foreign-born terrorism including 9/11"},"normalized":{"lifetime_us_adult":0.0000129,"display":"1 in ~77,000 lifetime (US adult)","log_value":-4.89,"assumptions":"Uses the Cato Institute's 1975–2024 foreign-born terrorism figure of 1 in 4,559,768 per year (Nowrasteh, 2025) as the central annual hazard, since that study explicitly includes 9/11 and covers a 50-year window. Domestic (non-foreign-born) terrorism adds a small but non-trivial amount — roughly another 500–600 deaths over 1975–2024 per START's Global Terrorism Database totals — but does not change the order of magnitude. Lifetime over 59 adult years: 1 − (1 − 2.19e-7)^59 ≈ 1.29e-5 ≈ 1 in 77,000. The window choice is load-bearing: the 9/11 attacks account for roughly 2,977 of ~3,500 US terrorism deaths since 1970, so any window that includes 9/11 is dominated by a single event and any window that excludes it drops the annual rate by roughly an order of magnitude. The uncertainty band below reflects that window sensitivity rather than a statistical confidence interval.\n","uncertainty":{"low":0.0000025,"high":0.000035},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cato.org/policy-analysis/terrorism-immigration-50-years-foreign-born-terrorism-us-soil-1975-2024","title":"Terrorism and Immigration: 50 Years of Foreign-Born Terrorism on US Soil, 1975–2024","publisher":"Cato Institute (Alex Nowrasteh)","source_type":"reputable_reference","statistic":"Annual chance of dying in a foreign-born terrorist attack on US soil: 1 in 4,559,768 (1975–2024, including 9/11)","excerpt":"\"The average chance of dying in an attack committed by a foreign-born terrorist on US soil was 1 in 4,559,768 a year. For perspective, the annual chance of being murdered by a common criminal in the United States was about 330 times as great as dying in an attack committed by a foreign-born terrorist on US soil.\"\n","source_date":"2025-03-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260509180310/https://www.cato.org/policy-analysis/terrorism-immigration-50-years-foreign-born-terrorism-us-soil-1975-2024","calculation_notes":"Cato divides 3,046 deaths (2,979 of which occurred on 9/11) across the 1975–2024 window by average US population, yielding 1 in 4,559,768 per person-year. Converting to a lifetime figure for a 59-year adult horizon: 1 − (1 − 1/4,559,768)^59 ≈ 1.29 × 10⁻⁵, i.e. ~1 in 77,000. Domestic (non-foreign-born) terrorism adds a modest additional hazard that the uncertainty band absorbs. The \"foreign-born\" restriction is the main reason we treat the Cato figure as a lower bound on the all-terrorism rate rather than a ceiling.\n","independence_note":"Cato's underlying death counts are drawn from the START Global Terrorism Database (University of Maryland) and RAND's terrorism incident database, so this is not independent of those sources — it is a curated risk-ratio calculation built on top of them.\n"},{"url":"https://www.cato.org/policy-analysis/terrorism-immigration-risk-analysis","title":"Terrorism and Immigration: A Risk Analysis","publisher":"Cato Institute (Alex Nowrasteh)","source_type":"reputable_reference","statistic":"Annual chance of dying in a foreign-born terrorist attack on US soil: 1 in 3.6 million (1975–2015)","excerpt":"\"The chance of an American perishing in a terrorist attack on US soil that was committed by a foreigner over the 41-year period studied here is 1 in 3.6 million per year. The annual chance of being murdered was 252.9 times as great as dying in an attack committed by a foreign-born terrorist on US soil.\"\n","source_date":"2016-09-13","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260503094424/https://www.cato.org/policy-analysis/terrorism-immigration-risk-analysis","calculation_notes":"Used as a corroborating earlier vintage of the same risk analysis. The shorter 1975–2015 window gives a denominator of 1 in 3.6 million per year; the longer 1975–2024 window used in our normalization gives 1 in 4.56 million. The direction of that drift (lower annual risk as the window lengthens past 9/11 without a comparable repeat event) is the expected behaviour for a distribution dominated by a single extreme event.\n","independence_note":"Same author and underlying datasets as the 2025 update. Included as a longitudinal consistency check, not as an independent estimate.\n"},{"url":"https://ourworldindata.org/terrorism","title":"Terrorism","publisher":"Our World in Data (Max Roser, Bastian Herre, Hannah Ritchie)","source_type":"reputable_reference","statistic":"Terrorism caused roughly 1 in 2,000 global deaths in 2019; North American terrorism deaths are typically low outside 9/11","excerpt":"\"Deaths from terrorism in North America are typically low – but the September 11 attacks in the United States in 2001 stand out. [Globally in 2019 terrorism] caused an estimated 1 in 2000 deaths that year.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260402162118/https://ourworldindata.org/terrorism","calculation_notes":"Used as an independent check on the order of magnitude and on the 9/11-dominated shape of the US distribution. OWID draws on the START Global Terrorism Database and IHME Global Burden of Disease, which are upstream of (but largely independent of) Cato's risk-analysis pipeline.\n","independence_note":"OWID's death counts come from START GTD and IHME GBD directly. Cato also uses START GTD but layers its own per-visa-class and per-window accounting on top, so the two sources share underlying event data but produce their statistics independently.\n"},{"url":"https://news.gallup.com/poll/404048/record-high-perceive-local-crime-increased.aspx","title":"Record-High 56% in U.S. Perceive Local Crime Has Increased","publisher":"Gallup","source_type":"reputable_reference","statistic":"27% of US adults worry a great deal or fair amount about being a victim of terrorism (October 2022)","excerpt":"\"At the other end of the spectrum, Americans worry least about being assaulted or killed by a coworker on the job (9%) or being the victim of terrorism (27%).\"\n","source_date":"2022-10-18","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260429013948/https://news.gallup.com/poll/404048/record-high-perceive-local-crime-increased.aspx","calculation_notes":"Used for the perceived-risk side only. The 27% figure is the share of respondents reporting \"a great deal\" or \"a fair amount\" of worry; Gallup does not elicit a subjective probability. Historical peaks in the same item reached roughly 47–49% in the months after the 2015 Paris attacks and San Bernardino shooting.\n","independence_note":"Gallup telephone polling, entirely separate from the START Global Terrorism Database, RAND, and Cato pipelines that feed the probability estimate. Used only for the perceived-risk axis — measures public worry, not incidence.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Being murdered (lifetime, US, pooled)","lifetime_us_adult":0.00348},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Residence in high-density urban target area (NYC, DC, LA)","multiplier":5,"notes":"START Global Terrorism Database (GTD), University of Maryland: historical US terrorism incidents from 1970–2024 are geographically concentrated in a handful of major metropolitan areas; New York City, Washington DC, and Los Angeles account for a disproportionate share of both incidents and fatalities, yielding approximately 5× the per-capita terrorism death rate for residents of these cities versus the national average."},{"factor":"Regular attendance at large public events or iconic soft targets","multiplier":3,"notes":"START GTD and RAND Corporation terrorism target analysis: soft targets — concerts, sports venues, open-air markets, transit hubs — account for an elevated share of mass-casualty terrorism incidents globally; individuals who regularly attend large public gatherings face approximately 3× elevated situational exposure compared with those whose routines keep them away from high-footfall symbolic locations."},{"factor":"International travel to State Department Tier 3–4 advisory countries","multiplier":10,"notes":"US State Department Travel Advisory system and START GTD global data: Americans traveling to countries rated Level 3 (Reconsider Travel) or Level 4 (Do Not Travel) experience approximately 10× the per-day terrorism fatality risk compared with domestic US daily life; the GTD documents that terrorism death rates in active-conflict and high-instability countries are orders of magnitude above the US baseline."},{"factor":"Work in federal government, military, or law enforcement","multiplier":3,"notes":"START GTD and FBI Terrorism Task Force reporting: government facilities, military personnel, and law-enforcement officers are historically over-represented as terrorism targets in both domestic and international incidents; occupational exposure to government or security roles is associated with approximately 3× the background civilian terrorism fatality rate based on target-type distributions in the GTD."}],"short_label":"Terrorism","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The \"1 in 77,000\" figure is a pooled, window-sensitive average over fifty years of data that is overwhelmingly shaped by a single morning in September 2001. Strip out 9/11 and the annual hazard falls by roughly an order of magnitude; keep 9/11 but shorten the window to only the post-2001 years and the average rises. Neither of those is wrong, but both would tell you something different about \"typical\" years. The figure also says nothing about which Americans face which share of the risk: terrorism deaths in the US are concentrated by target (iconic buildings, federal facilities, large gatherings), geography (a handful of metropolitan areas), and year (a very small number of events account for the bulk of cumulative deaths). For the vast majority of residents in the vast majority of years, the annual hazard is effectively zero.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":5,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted candle flame held upright against a pale grey background, flat vector illustration."},"canonical_url":"https://likelier.app/terrorism-us-civilian","api_url":"https://likelier.app/api/fears/terrorism-us-civilian.json"},{"slug":"first-anaphylaxis-death-risk","question":"What are the odds of dying from a bee, wasp, or hornet sting?","quick_answer":"Fatal sting reactions are rare: about 72 deaths a year in the US from bee, wasp, and hornet stings, roughly 1 in 77,000 over an adult lifetime. Here perception is well calibrated, since most people guess a handful of deaths a year, matching the data.\n","category":"health","tags":[],"no_reliable_estimate":false,"perceived":{"description":"Most people are aware that some individuals are fatally allergic to bee and wasp stings, but the actual death toll is not well-known. Sting fatalities sit in a mid-range of public awareness: higher than sharks or lightning in general perception, but rarely a top-of-mind concern outside allergy communities. No comprehensive national survey isolates worry about venom anaphylaxis specifically.\n","rough_estimate":"Most people would guess a handful of deaths per year -- consistent with the actual ~62--72 annual average","kind":"intuition"},"native":{"display":"~72 deaths per year from hornet, wasp, and bee stings (US, 2011--2021 average)","numerator":72,"denominator":330000000,"unit":"per year","population":"US general population"},"normalized":{"lifetime_us_adult":0.000013,"display":"~1 in 77,000 lifetime (US adult)","log_value":-4.89,"assumptions":"CDC MMWR QuickStats (2023) reports 788 deaths from hornet, wasp, and bee stings during 2011--2021, for an annual average of approximately 72 deaths per year. (The earlier 2000--2017 series averaged 62 per year.) Using 72/year as the primary estimate: 72 / 330,000,000 = 0.000000218 per person per year. Compounded over 59 adult years: 1 − (1 − 0.000000218)^59 = 1 − 0.999987 = 0.0000129. Rounding to 0.000013, which is approximately 1 in 77,000. This is a population-average figure that blends people with known venom allergy (who face much higher risk without treatment), people with no known allergy who experience a first fatal reaction (~60% of deaths), and people with prescriptions for epinephrine autoinjectors (who face lower risk if they carry and use them promptly).\n","uncertainty":{"low":0.000008,"high":0.00002},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/72/wr/pdfs/mm7227a6-H.pdf","title":"QuickStats: Number of Deaths from Hornet, Wasp, and Bee Stings Among Males and Females -- National Vital Statistics System, United States, 2011--2021","publisher":"CDC MMWR / National Vital Statistics System","source_type":"govt_report","statistic":"788 deaths from hornet, wasp, and bee stings during 2011--2021; annual average of ~72 deaths per year; 84% of deaths occurred among males","excerpt":"\"During 2011--2021, a total of 788 deaths from hornet, wasp, and bee stings occurred (an average of 72 deaths per year). The annual number of deaths ranged from 59 (2012) to 89 (2017). Overall, 84% of deaths occurred among males.\"\n","source_date":"2023-07-07","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260511150906/https://www.cdc.gov/mmwr/volumes/72/wr/pdfs/mm7227a6-H.pdf","calculation_notes":"788 deaths / 11 years = 71.6, rounded to 72/year. Annual rate: 72 / 330,000,000 = 0.000000218 per person per year. Lifetime probability over 59 years: 1 − (1 − 0.000000218)^59 = 0.0000129, reported as 0.000013. The earlier MMWR series (2000--2017, average 62/year) yields a similar figure; the 2011--2021 series is used as the more recent estimate.\n","independence_note":"CDC MMWR QuickStats draws on the National Vital Statistics System (NVSS), which collects death certificate data from all US states. This is the authoritative source for cause-of-death mortality statistics in the United States.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/68/wr/mm6829a5.htm","title":"QuickStats: Number of Deaths from Hornet, Wasp, and Bee Stings, Among Males and Females -- National Vital Statistics System, United States, 2000--2017","publisher":"CDC MMWR / National Vital Statistics System","source_type":"govt_report","statistic":"1,109 deaths from hornet, wasp, and bee stings during 2000--2017; annual average of approximately 62 deaths; 80% of deaths were among males","excerpt":"\"During 2000--2017, a total of 1,109 deaths from hornet, wasp, and bee stings occurred (annual average of approximately 62 deaths). In 2017, the highest number of deaths in this time period occurred (89 deaths). Approximately 80% of deaths occurred among males.\"\n","source_date":"2019-07-19","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260511082852/https://www.cdc.gov/mmwr/volumes/68/wr/mm6829a5.htm","calculation_notes":"Earlier series confirming the ~62--72 annual range. The 2019 MMWR (2000--2017 series) and 2023 MMWR (2011--2021 series) are independently compiled from the same NVSS underlying data, providing temporal consistency. The 2017 peak of 89 deaths reflects year-to-year variability around the long-run mean of ~65--72.\n","independence_note":"Both MMWR QuickStats reports are independently published analyses of the NVSS death certificate data. The two time series overlap (2011--2017 is shared) but are published separately as distinct analytical snapshots with independent editorial review.\n"}],"comparison_anchors":[{"label":"Death from lightning strike (lifetime, US)","lifetime_us_adult":0.00007},{"label":"Death from shark attack (lifetime, global)","lifetime_us_adult":0.000003}],"personal_factor_multipliers":[{"factor":"Known venom allergy with prescribed epinephrine autoinjector (carried and used)","multiplier":0.15,"notes":"Prompt epinephrine administration dramatically reduces fatality risk; allergy immunotherapy (venom shots) can reduce systemic reaction risk by ~95% in adults"},{"factor":"Male sex (80% of deaths are male)","multiplier":1.7,"notes":"CDC data consistently shows ~80% of sting deaths occur in males; possibly related to higher outdoor occupational exposure and lower rates of seeking care promptly"},{"factor":"Outdoor occupation or hobby (landscaping, beekeeping, hiking, farming)","multiplier":2,"notes":"Elevated exposure to stinging insects substantially increases the probability of a sensitizing sting followed eventually by a fatal anaphylactic event"},{"factor":"Age 65+ (higher case fatality from anaphylaxis)","multiplier":2,"notes":"Older adults have higher fatality rates conditional on anaphylaxis, related to cardiovascular comorbidities and potentially slower emergency response"},{"factor":"Not carrying epinephrine despite known or suspected allergy","multiplier":5,"notes":"The majority of fatal anaphylaxis cases involve delayed or absent epinephrine administration; access to treatment at the time of the reaction is the primary modifiable risk factor"}],"short_label":"Fatal bee/wasp sting","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Approximately 60% of fatal venom reactions occur in people with no known prior anaphylaxis diagnosis -- first sensitization goes unrecognized until a fatal second exposure. This makes the risk difficult to eliminate through simple avoidance without systematic skin testing. Deaths from food-induced anaphylaxis (peanut, shellfish) are NOT included in this entry; this figure covers venom anaphylaxis only. Death is classified as due to \"hornet, wasp, or bee stings\" (ICD-10 code X23) on the death certificate, which may miss cases where the sting was a contributing but not primary listed cause. The 72/year figure likely slightly undercounts total venom- related deaths. Imported Africanized honey bees expand the exposure in southern states.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A single stylized bee on a pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/first-anaphylaxis-death-risk","api_url":"https://likelier.app/api/fears/first-anaphylaxis-death-risk.json"},{"slug":"us-local-dengue","question":"What are the odds of catching dengue without leaving the continental US?","quick_answer":"Catching dengue without leaving the continental US runs about 1 in 67,000 over a lifetime — small but not zero, against the common assumption that dengue is purely a travel disease. In 2024 there were 105 locally-acquired cases among ~338 million residents.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Most US residents file dengue under 'tropical travel disease' and assume zero risk while staying home. No published survey measures fear of catching dengue domestically, so this is an intuition read: the perceived domestic risk is effectively zero for nearly everyone, which understates the small but rising local-transmission risk in a few southern states and overstates the near-zero risk everywhere else. The mismatch is one of category error rather than magnitude.","rough_estimate":"≈ 0 (assumed to be a travel-only disease)","kind":"intuition"},"native":{"display":"105 locally-acquired cases among ~338 million continental US residents in 2024","numerator":105,"denominator":338000000,"unit":"locally-acquired dengue cases per continental US resident per year (2024)","population":"Residents of the 48 contiguous US states + DC, 2024"},"normalized":{"lifetime_us_adult":0.000015,"display":"~1 in 67,000 over a lifetime","log_value":-4.82,"assumptions":"Subgroup = a person residing in the continental US (48 states + DC); horizon = 59-year remaining adult life (site default from age 18). Headline counts a DIAGNOSED clinical dengue case, not infection (40-80% of dengue infections are asymptomatic and many symptomatic cases go untested, so true infection risk is higher — this pushes the high bound up). Base annual rate from CDC MMWR 2024: 105 locally-acquired cases (Florida 85, California 18, Texas 2) / ~338M continental US residents = 3.1e-7 per person-year. Compounded over 59 years: 1-(1-3.1e-7)^59 = 1.83e-5 (~1 in 54,600), which is the 2024-RECORD-YEAR rate. 2024 was a +359% peak; before 2023 continental local transmission was near-zero in most years. Point estimate is set slightly BELOW the 2024-only figure at 1.5e-5 (~1 in 67,000) to average the recent peak against the much lower historical baseline, while leaning toward persistence because the entry's whole premise — Aedes range expansion and warming — is an upward trend. Low bound 1.5e-6 reflects pre-2023 near-zero local-transmission years (~10x below 2024). High bound 1.5e-4 reflects clinical undercount plus a worsening-transmission scenario (~10x above 2024). The 2-order-of-magnitude band is honest about a volatile, geographically concentrated, expanding risk rather than falsely precise.","uncertainty":{"low":0.0000015,"high":0.00015},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/75/wr/mm7518a1.htm","title":"Increase in Travel-Associated and Locally Acquired Dengue Cases — United States, 2024","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"3,798 total US dengue cases in 2024 (+359% above the 2010–2023 average); 97.2% travel-associated, 2.8% locally acquired; continental US locally-acquired cases: Florida 85 (10 counties), California 18 (3 counties), Texas 2 (2 counties) = 105.","excerpt":"During 2024, a record 3,798 dengue cases were reported [...] a 359% increase above the 2010–2023 annual average. Among these 3,798 cases, 97.2% were associated with travel outside the reporting jurisdiction during the 2 weeks preceding symptom onset; the remaining 2.8% were locally acquired. [...] Three jurisdictions reported locally acquired cases, including Florida (85 cases reported from 10 counties), California (18 from three counties), and Texas (two from two counties).","source_date":"2026-05-14","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260616185432/https://www.cdc.gov/mmwr/volumes/75/wr/mm7518a1.htm","calculation_notes":"Locally-acquired continental total = 85+18+2 = 105. Annual rate = 105 / 338,000,000 continental residents = 3.1e-7/person-year. 59-year compounded = 1-(1-3.1e-7)^59 = 1.83e-5. This is the authoritative FINAL count; an earlier CDC HAN (March 2025) listed Florida as 91 preliminarily — the MMWR's 85 supersedes it."},{"url":"https://wwwnc.cdc.gov/eid/article/32/5/25-1812_article","title":"Investigation of and Response to Autochthonous Dengue, Los Angeles County, California, USA, August–November 2024","publisher":"CDC Emerging Infectious Diseases","source_type":"peer_reviewed","statistic":"14 locally-acquired dengue cases in LA County Aug–Nov 2024, geographically clustered (some within a 1-mile radius over 7 weeks), consistent with focal local transmission rather than diffuse community spread.","excerpt":"During August–November 2024, a total of 14 locally acquired cases of dengue were identified in Los Angeles County, California, USA. [...] All 6 cases were identified within a 1-mile radius over a 7-week period, consistent with an outbreak involving ongoing local transmission.","source_date":"2026-05-06","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260608080712/https://wwwnc.cdc.gov/eid/article/32/5/25-1812_article","calculation_notes":"Grounds the regional_breakdown claim that California's 18 cases were not statewide-uniform but concentrated in a few LA-area clusters. Supports treating local dengue as a focal/neighborhood risk, not a flat per-capita state rate.","independence_note":"Same publisher family (CDC) as the MMWR source but a distinct peer-reviewed field investigation with separate authors and data."},{"url":"https://www.cdc.gov/han/php/notices/han00523.html","title":"Ongoing Risk of Dengue Virus Infections and Updated Testing Recommendations in the United States","publisher":"CDC Health Alert Network (HAN)","source_type":"govt_report","statistic":"CDC advisory that dengue activity remained high into 2025 and that spring/summer raise both travel-associated and locally-acquired risk; 2024 local cases listed as Florida 91 (preliminary), California 18, Texas 1.","excerpt":"Dengue activity remains high in some parts of the United States [...] Spring and summer travel coincide with the peak season for dengue in many countries, increasing the risk of both travel-associated and locally acquired cases.","source_date":"2025-03-18","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260511082709/https://www.cdc.gov/han/php/notices/han00523.html","calculation_notes":"Used to establish that the risk is ongoing/elevated rather than a one-off 2024 spike, supporting the upward-trend framing. Its preliminary Florida count (91) differs from the MMWR final (85); the entry uses the MMWR figure and flags the discrepancy in assumptions."},{"url":"https://www.cidrap.umn.edu/dengue/us-dengue-cases-jumped-359-2024-amid-global-surge","title":"US dengue cases jumped 359% in 2024 amid global surge","publisher":"CIDRAP, University of Minnesota","source_type":"reputable_reference","statistic":"Confirms 3,798 total cases, 359% increase, 2.8% locally acquired, 105 locally-acquired nationwide with Florida 85 (81%); notes Aedes aegypti and albopictus are established across the southeastern/southwestern US and continue to spread as temperatures rise.","excerpt":"Florida accounting for 85 of the 105 locally acquired cases (81%) reported nationwide [...] Ae aegypti and albopictus mosquitoes are found throughout large areas of the southeastern and southwestern United States and continue to spread to other regions [...] Vector presence and distribution might be further affected by increasing temperatures, which can expand the range of mosquitoes that spread dengue.","source_date":"2026-05-15","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260624182813/https://www.cidrap.umn.edu/dengue/us-dengue-cases-jumped-359-2024-amid-global-surge","calculation_notes":"Independent confirmation of the MMWR 85/105 split and the vector-range-expansion premise that motivates the entry. Used for the Aedes-range and warming framing in the prose.","independence_note":"Independent academic center summarizing the MMWR; not a CDC publication."},{"url":"https://epi.ufl.edu/2025/06/24/dengue-in-florida-what-to-know/","title":"Dengue in Florida: What to know","publisher":"University of Florida Emerging Pathogens Institute","source_type":"reputable_reference","statistic":"Florida reported 1,044 local + travel-associated dengue cases in 2024 (most in the continental US); local cases concentrate in Miami-Dade County; no part of the US is officially endemic for dengue despite recurring local transmission.","excerpt":"Although no part of the United States is officially labeled as endemic for dengue, Florida's health department continues to report cases acquired locally or linked to travel in 2025. [...] one case of locally acquired dengue in Miami-Dade County","source_date":"2025-06-24","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260314194903/https://epi.ufl.edu/2025/06/24/dengue-in-florida-what-to-know/","calculation_notes":"Grounds the Florida/Miami-Dade concentration in the regional_breakdown and the 'not endemic' caveat. The 1,044 figure is local+travel combined (most are travel-associated); only the locally-acquired subset (85) feeds the native rate."}],"comparison_anchors":[{"label":"Plane crash death (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.00000478},{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"Florida (esp. Miami-Dade)","probability":0.00022,"notes":"85 of the 105 continental locally-acquired cases in 2024 (81%) were in Florida, across 10 counties but concentrated in the Miami-Dade area. 85 / ~23M Florida residents = 3.7e-6/yr → ~2.2e-4 (~1 in 4,600) over 59 yr; the Miami-Dade sub-rate is higher still. Florida is not officially endemic but reports local cases every year."},{"region":"California (focal, Los Angeles County 2024)","probability":0.00002,"notes":"18 local cases statewide in 2024; LA County had 14 of them, all clustered within small radii consistent with focal outbreaks (some within a 1-mile radius over 7 weeks). 18 / ~39M = 4.6e-7/yr → ~2.7e-5 over 59 yr, but risk is concentrated in a few neighborhoods, not statewide."},{"region":"Most of the continental US (no established Aedes / no local cases)","probability":0.0000015,"notes":"Northern and interior states reported zero locally-acquired cases in 2024. Risk there is driven by occasional vector spread and remains near the lower bound."}],"short_label":"Locally-acquired dengue (continental US)","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This is a snapshot of a volatile, geographically concentrated, and apparently expanding risk, not a stable lifetime constant. The point estimate rests on a single record year (2024, +359% above the 2010–2023 average), and projecting that rate across a 59-year horizon assumes 2024-era local transmission persists — a deliberate lean toward the upward Aedes-range and warming trend the entry describes. The uncertainty band spans two orders of magnitude (~1 in 670,000 to ~1 in 6,700) precisely because year-to-year local-case counts swing wildly and the true infection rate is undercounted: 40–80% of dengue infections are asymptomatic and many symptomatic cases go untested, so the headline 'diagnosed case' rate is a floor on actual exposure. The risk is overwhelmingly a Florida (esp. Miami-Dade) and focal-California phenomenon; for a resident of most of the continental US it sits near the lower bound. No published survey measures fear of catching dengue domestically, so the perception read is an intuition, not a poll. Because the band is wide and the denominator is national while the risk is local, D4 (uncertainty) and D5 (scope) are the weakest dimensions; both are handled explicitly via subgroup scope and a regional_breakdown rather than a single misleading national headline. The entry does not assert dengue is endemic anywhere in the continental US — no US jurisdiction is officially classified as endemic.","quality_score":{"d1":4,"d2":5,"d3":4,"d4":3,"d5":4,"d6":4,"d7":4,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Locally-acquired dengue (continental US)\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/us-local-dengue","api_url":"https://likelier.app/api/fears/us-local-dengue.json"},{"slug":"infant-choking-in-car-seat","question":"What are the odds of an infant choking while reclined in a car seat?","quick_answer":"Across a US infant's first two years, the odds of a fatal asphyxia in a sitting or carrying device are about 1 in 65,000, and more than 90% of those deaths happened when the seat was used off-label — on a bed, couch, or floor rather than buckled in a moving vehicle. Most parents carry the fear without any specific number.\n","category":"kids","tags":["infant"],"no_reliable_estimate":false,"perceived":{"description":"Car-seat asphyxia is one of the quieter items on the new-parent fear list, usually surfacing around the first long drive or the first naptime in a bucket seat. The mental model most parents carry is that a sleeping infant in a reclined rear-facing seat could let their chin fall to their chest, block the airway, and be found unresponsive before the driver notices in the mirror. The fear is specific, posture-driven, and rarely broken out separately from the broader food-choking and SIDS conversations even though it has its own distinct surveillance literature.\n","rough_estimate":"Most parents have no specific number; the fear is vivid but unquantified","kind":"intuition"},"native":{"display":"~30 US infant sitting-device asphyxia deaths per year (2004-2014)","numerator":1,"denominator":65000,"unit":"per infant during the 0-2 age window","population":"US infants, sleep-related deaths in sitting/carrying devices (car seats, bouncers, swings, strollers, slings)"},"normalized":{"lifetime_us_adult":0.0000154,"display":"1 in ~65,000 per US infant across the 0-2 age window","log_value":-4.81,"assumptions":"Scope is the first two years of life per US live-born infant, not US-adult-lifetime. Liaw et al. (Pediatrics 2019) reviewed 11,779 US sleep-related infant deaths from 2004 through 2014 and found 348 (3.0%) occurred in sitting or carrying devices — roughly 32 such deaths per year across the surveillance window. Batra et al. (J Pediatr 2015), working from the US Consumer Product Safety Commission database, documented 47 fatalities in sitting and carrying devices over 2004-2008, 31 of them in car seats, with asphyxia as the mechanism in all but one case. Against roughly 3.6-4.0 million US live births per year, 32 deaths per year is about 8 per million infants per year. Compounded across the first two years of life (1 - (1 - 8e-6)^2 ≈ 1.6e-5), that is roughly 1 in 65,000 per infant during the 0-2 window. The overwhelming majority of these deaths — per Liaw et al., more than 90% — occurred when the car seat was NOT being used as directed, typically placed on a bed, couch, or floor with the infant sleeping unbuckled or loosely strapped rather than secured in a moving vehicle. Restricting the headline to strictly in-vehicle, correctly-buckled use drops the rate by roughly an order of magnitude. The uncertainty band brackets the narrower car-seat-only subset (≈1 in 130,000) and the wider all-sitting-device umbrella used here (≈1 in 65,000).\n","uncertainty":{"low":0.000008,"high":0.00003},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/31110162/","title":"Infant Deaths in Sitting Devices","publisher":"Pediatrics — Liaw P, Moon RY, Han A, Colvin JD","source_type":"peer_reviewed","statistic":"348 of 11,779 US sleep-related infant deaths (3.0%) occurred in sitting devices, 2004-2014; 62.9% of those in car safety seats; car seat used as directed in <10% of cases; 81.9% had ≥1 risk factor","excerpt":"\"Of 11 779 infant sleep-related deaths, 348 (3.0%) occurred in sitting devices. Of deaths in sitting devices, 62.9% were in CSSs, and in these cases, the CSS was used as directed in <10%. [...] 81.9% had ≥1 risk factor, and 54.9% had ≥2 risk factors. [...] Using CSSs for sleep in nontraveling contexts may pose a risk to the infant.\"\n","source_date":"2019-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173401/https://pubmed.ncbi.nlm.nih.gov/31110162/","calculation_notes":"348 sitting-device deaths / 11 surveillance years ≈ 31.6 deaths per year. Against ~3.8 million US live births per year during the 2004-2014 window, that is roughly 8.3 per million infants per year. Compounded across the 0-2 window (1 - (1 - 8.3e-6)^2 ≈ 1.66e-5), the per-infant probability of dying in a sitting or carrying device between birth and age two works out to roughly 1 in 60,000-65,000. The \"<10% used as directed\" figure is the empirical basis for the personal factor multipliers and for treating correct in-vehicle use as an order-of-magnitude risk reducer.\n","independence_note":"Liaw et al. draws from the National Center for Fatality Review and Prevention CDR files, which are fed in part by state death-certificate records (overlapping with CDC NCHS). Batra et al. uses the separate CPSC product-incident database; treat Liaw as the primary peer-reviewed anchor and Batra as the complementary surveillance stream with different inclusion criteria.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25917769/","title":"Hazards Associated with Sitting and Carrying Devices for Children Two Years and Younger","publisher":"Journal of Pediatrics — Batra EK, Midgett JD, Moon RY","source_type":"peer_reviewed","statistic":"47 US infant/child deaths in sitting and carrying devices reported to CPSC, 2004-2008; 31 in car seats; asphyxiation in all but one case; 52% of car seat deaths from strap strangulation, rest from positional asphyxia","excerpt":"\"A retrospective review of deaths involving sitting and carrying devices (car seats, bouncers, swings, strollers, and slings) reported to the US Consumer Product Safety Commission between 2004 and 2008. [...] Of the 47 deaths, 31 occurred in car seats, 5 in slings, 4 each in swings and bouncers, and 3 in strollers. [...] The cause of death was asphyxiation in all cases except one. Fifty-two percent of deaths in car seats were attributed to strangulation from straps; the others were attributed to positional asphyxia. [...] Infants and children 2 years of age and younger should be properly restrained and not be left unsupervised in sitting and carrying devices.\"\n","source_date":"2015-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173442/https://pubmed.ncbi.nlm.nih.gov/25917769/","calculation_notes":"Batra et al. draw from the CPSC product-incident database, which captures a different subset of deaths than the National Center for Fatality Review and Prevention file Liaw et al. analyse. The 47 deaths / 4.7 years ≈ 10 CPSC-reported deaths per year understates the true total because not every infant asphyxia death is reported to CPSC as a product incident. Used as an independent cross-check and as the authoritative source for the mechanism split (strap strangulation vs positional asphyxia) and for the mean 140-minute elapsed-time window between the infant last being seen and being found unresponsive.\n","independence_note":"Batra draws from CPSC's product-incident reporting system; Liaw draws from the National Center for Fatality Review and Prevention's CDR files. The two datasets have different inclusion criteria and only partially overlap, so they function as complementary surveillance streams rather than two independent counts of the same deaths.\n"},{"url":"https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","title":"Data and Statistics for SUID and SIDS","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~3,700 US sudden unexpected infant deaths in 2022 (SIDS, accidental suffocation and strangulation in bed, and unknown cause)","excerpt":"\"In 2022, there were about 3,700 sudden unexpected infant deaths (SUID) in the United States. These deaths occur among infants less than 1 year old and have no immediately obvious cause. [...] 1,529 deaths from SIDS [...] 1,131 deaths from unknown causes [...] 1,040 deaths from accidental suffocation and strangulation in bed.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173518/https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","calculation_notes":"Anchors the sitting-device subset inside the broader sudden unexpected infant death (SUID) umbrella. The Liaw et al. figure of ~32 sitting-device deaths per year is on the order of 1% of all SUID deaths in a given year, which is why this fear is real but small relative to the SIDS / ASSB bulk of the sleep-related infant death count. Used as the denominator context for comparison_anchors and to position the sitting-device subset against the full SUID umbrella.\n","independence_note":"CDC NCHS mortality data partly feeds the Liaw et al. case-file data, so the three sources here form a linked chain rather than three fully independent counts. Treated as complementary views of overlapping surveillance streams.\n"}],"comparison_anchors":[{"label":"SIDS per US infant (narrow ICD-10 R95)","lifetime_us_adult":0.00014},{"label":"All sudden unexpected infant death (SUID) per US infant, 2022","lifetime_us_adult":0.00035},{"label":"Food asphyxiation death, per US child 0-4","lifetime_us_adult":0.00002},{"label":"Death in a plane crash, lifetime (US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"All sitting/carrying devices, US infant 0-2 (the headline number)","probability":0.0000154,"notes":"~32 deaths per year across car seats, bouncers, swings, strollers, and slings combined, compounded across the first two years of life.\n"},{"region":"Car safety seats only, US infant 0-2","probability":0.00001,"notes":"~20 car-seat deaths per year (62.9% of the Liaw et al. sitting-device total), compounded across the first two years of life. Most occur with the seat placed outside the vehicle on a bed, couch, or floor.\n"},{"region":"Car safety seats used in-vehicle and correctly buckled","probability":0.000001,"notes":"Liaw et al. found that fewer than 10% of car-seat sleep deaths involved seats used as directed. Restricting to correctly-used in-vehicle cases drops the per-infant rate by roughly an order of magnitude, into the \"rounds down to zero for most parents\" band.\n"}],"personal_factor_multipliers":[{"factor":"car seat used out of vehicle (couch, bed, floor)","multiplier":5,"notes":"Liaw et al. found \"more than half\" of car seat sleep deaths occurred at the child's home, with the seat placed on a soft or elevated surface. Removing a seat from its base and using it as a household sleeper is the single largest observable risk factor.\n"},{"factor":"infant unbuckled or loosely strapped","multiplier":10,"notes":"Batra et al. found car seats used as directed in fewer than 10% of fatal cases; Liaw et al. reported the same. Unbuckled infants slump forward, lose airway patency, and cannot self-rescue. The multiplier is approximate but consistent with the ratio of \"used as directed\" vs \"not used as directed\" in both datasets.\n"},{"factor":"used in vehicle only, correctly buckled","multiplier":0.2,"notes":"Vehicle use with proper harness tension and a driver or passenger within sight line converts almost all the risk captured here into the residual one-in-several-hundred-thousand band.\n"},{"factor":"age >4 months, good head control","multiplier":0.3,"notes":"Mechanism is chin-to-chest slumping. Infants who can reliably lift and reposition their heads carry materially lower risk, though the peer-reviewed literature does not resolve the exact age cutoff.\n"}],"short_label":"Infant car-seat asphyxia","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The headline per-infant figure combines all sitting and carrying devices — car seats, bouncers, swings, strollers, and slings — because the peer-reviewed surveillance data (Liaw 2019, Batra 2015) counts them as a single mechanism-of-death category. Parents asking specifically about a reclined rear-facing car seat during a drive are inside a much narrower subset: Liaw et al. found fewer than 10% of these deaths involved a car seat used as directed, and more than half occurred at the child's home rather than in a moving vehicle. The dominant mechanism is positional asphyxia from chin-to-chest slumping in an unbuckled or loosely-strapped infant, with a second mechanism (strap strangulation, ~52% of car-seat cases in Batra et al.) that is distinct from choking on food or bedding. This entry excludes food-related choking in car seats — that risk lives in the toddler-choking-while-eating page — and excludes SIDS/SUID deaths in cribs and adult beds, which are coded separately and dominate the overall sleep-related infant mortality count. Mean elapsed time between the infant last being seen and being found unresponsive in Batra's car-seat cases was 140 minutes, which is the single most useful operational fact for parents trying to understand the supervision dimension of the risk.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small empty rear-facing car seat shell in outline, viewed from the side against a pale grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/infant-choking-in-car-seat","api_url":"https://likelier.app/api/fears/infant-choking-in-car-seat.json"},{"slug":"infant-bouncer-fall","question":"What are the odds an infant in a bouncer chair falls when the chair is placed on an elevated surface?","quick_answer":"Across US bouncer-using infants, the odds of a reported incident work out to about 1 in 60,000, and about 1 in 1.7 million for a fatal outcome (347 CPSC incidents including 12 deaths and 54 injuries over a decade). The dominant mechanism is falls from elevated surfaces despite floor-only warnings.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":false,"perceived":{"description":"Putting an infant in a bouncer on a kitchen counter or dining table so they can watch a parent cook is a routine practice, and most parents perceive it as benign. The seat looks stable, the infant is restrained, and the height seems modest. The product label warns to use only on the floor, but most parents either do not read the warning or read it and discount it against the visible stability of the seat. The Claydon (1996) case report and the CPSC's 2018 bouncer standard both rest on the same mechanism: the bouncer's central pivot point converts even a short fall into a higher-velocity head impact than the height alone would predict, so the intuitive \"it's only two feet\" calibration is wrong in a specific geometric way that parents do not typically have in mind.\n","rough_estimate":"~1 in 10,000 per use on an elevated surface","kind":"intuition"},"native":{"display":"347 bouncer-seat incidents (including 12 fatalities and 54 injuries) reported to CPSC, January 2006-July 2016","numerator":347,"denominator":21000000,"unit":"per bouncer-seat unit in US use over the 10.5-year reporting period","population":"US infants 0-6 months exposed to bouncer and rocker chairs, 2006-2016"},"normalized":{"lifetime_us_adult":0.0000165,"display":"~1 in 60,000 chance of a reported incident per bouncer-using infant; ~1 in 1.7 million for a fatal outcome — dominant mechanism is falls from elevated surfaces despite explicit floor-only warnings","log_value":-4.78,"assumptions":"347 incidents including 12 fatalities and 54 injuries reported to CPSC January 2006 through July 2016 (CPSC, 2018-03-23). CPSC's release notes that the dominant fatality mechanism was suffocation when unrestrained babies turned over or when bouncers tipped onto soft surfaces such as adult beds and cribs. The new federal standard effective March 19, 2018 (16 CFR 1229, incorporating ASTM F2167) was written to address these hazards. The denominator is estimated from sales volume: the bouncer category runs at roughly 2-3 million US units per year, giving on the order of 21 million units across the 10.5-year reporting period. Per- bouncer-using-infant lifetime risk for any reported incident works out to ~1 in 60,000; fatal outcomes at ~1 in 1.7 million. CPSC's incident reporting almost certainly under-counts falls that did not result in medical attention, so both rates are floors rather than central estimates.\n","uncertainty":{"low":0.000005,"high":0.00005},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2018/New-Federal-Standard-to-Improve-Safety-of-Infant-Bouncer-Seats-Takes-Effect","title":"New Federal Standard to Improve Safety of Infant Bouncer Seats Takes Effect","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"347 incidents involving bouncer seats reported to CPSC January 2006 to July 2016, including 12 fatalities and 54 injuries; suffocation from unrestrained babies turning over, or bouncers tipping onto soft surfaces (adult beds and cribs), was the dominant fatality mechanism; new federal standard effective March 19, 2018","excerpt":"\"Between January 1, 2006 and July 6, 2016, there were 347 incidents involving bouncer seats reported to CPSC, including 12 fatalities and 54 injuries. The major cause of reported fatalities was suffocation when unrestrained babies turned over in a bouncer or bouncers tipped over onto soft surfaces… when placed on adult beds and in cribs.\"\n","source_date":"2018-03-23","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260310062045/https://www.cpsc.gov/Newsroom/News-Releases/2018/New-Federal-Standard-to-Improve-Safety-of-Infant-Bouncer-Seats-Takes-Effect","calculation_notes":"Canonical aggregate figure across the 10.5-year window. The CPSC release describes the standard generically as \"a new federal standard… effective March 19, 2018\" without naming the underlying ASTM identifier; the standard is 16 CFR 1229 incorporating ASTM F2167, but that specific pairing is not asserted by this source and is recorded here only as regulatory context.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/8989793/","title":"Fatal extradural hemorrhage following a fall from a baby bouncer","publisher":"Pediatric Emergency Care (Claydon)","source_type":"peer_reviewed","statistic":"Case report: fatal extradural hemorrhage in an infant from an approximately 2-foot bouncer fall onto carpeted floor; pivoting mechanism amplifies impact","excerpt":"\"Pivoting about the central point provided by the seat of the bouncer obviously increased the momentum of the head before it struck the ground... serious head injuries can result from apparently minor falls... Prevention of infant deaths from accidental falls from baby equipment requires the maintaining of safety standards and adequate supervision of the infant.\"\n","source_date":"1996-12-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250523014553/https://pubmed.ncbi.nlm.nih.gov/8989793/","calculation_notes":"Single case but mechanistically definitive: the bouncer's central pivot point converts a low fall (approximately 2 feet) into a higher-velocity head impact than the height alone would predict. This is the reason the \"use only on the floor\" warning is load-bearing rather than cosmetic — the pivot geometry breaks the intuitive linear-height calibration that parents would otherwise apply.\n"},{"url":"https://www.sgs.com/en/news/2023/03/safeguards-4123-european-standard-en12790-on-reclined-cradles-revised","title":"European Standard EN 12790 on Reclined Cradles Revised","publisher":"SGS / European Committee for Standardization (CEN)","source_type":"reputable_reference","statistic":"EN 12790-1/-2:2023 (published March 2023) splits the reclined-cradle standard into two age bands and adds requirements for powered motion, electrical safety, entrapment, cord entanglement, and packaging suffocation","excerpt":"\"EN 12790-1:2023 covers reclined cradles for children up to when they start to try to sit up. EN 12790-2:2023 covers reclined cradles for children up to when they start to stand up. New rules on sound-pressure level, powered motion, electrical safety, entrapment, cord entanglement, packaging suffocation.\"\n","source_date":"2023-03-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251201083957/https://www.sgs.com/en/news/2023/03/safeguards-4123-european-standard-en12790-on-reclined-cradles-revised","calculation_notes":"EN 12790-1/-2:2023 governs reclined cradles (leżaczki) sold in Poland and across the EU, replacing EN 12790:2009. This standard covers reclined cradles broadly; CPSC's 16 CFR 1229 is the US sibling. Two different regulatory frames with similar harm-reduction focus. Neither removes the elevated-surface mechanism that dominates real-world incidents — both depend on user compliance with floor-only labelling.\n"}],"comparison_anchors":[{"label":"Infant fall from furniture serious injury (per fall)","lifetime_us_adult":0.01},{"label":"Toddler stair fall hospitalization (per fall)","lifetime_us_adult":0.027},{"label":"SIDS death (per infant-year)","lifetime_us_adult":0.00014}],"personal_factor_multipliers":[{"factor":"Placement on an elevated surface (table, counter, sofa, bed)","multiplier":8,"notes":"The dominant mechanism per CPSC; the product label warns to use only on the floor"},{"factor":"Infant unrestrained in bouncer harness","multiplier":4,"notes":"Falls and tip-overs are both more frequent when the infant is unrestrained; CPSC notes harness non-use is common"},{"factor":"Infant 4 months or older with active rolling","multiplier":3,"notes":"Active rolling and pushing dramatically raise the tip-over rate for any seated infant product"},{"factor":"Bouncer compliant with post-2018 federal standard (16 CFR 1229)","multiplier":0.5,"notes":"Standard addresses tip-over angle and stability; pre-2018 units (still in homes) lack these protections"},{"factor":"Single caregiver continuously in the same room","multiplier":0.5,"notes":"Most reported incidents occurred during brief inattention windows; effect-size estimated from CPSC narrative summaries, not from a controlled study"}],"short_label":"Bouncer chair fall","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 347-incident total covers a 10.5-year window and includes events at all severities reported through CPSC channels. Many bouncer falls — perhaps most — never reach CPSC because they did not require medical care, so the per-bouncer-using-infant rate is a floor rather than a central estimate. The denominator is estimated from sales volume and category-level reporting; per-unit risk could be off by a factor of two in either direction. The Claydon case report is a single fatality, not a population rate, and is included as mechanism evidence rather than as a frequency source. Post-2018 units are meaningfully safer than pre-2018 units, but a large portion of the bouncers in use — particularly secondhand units in the Polish market — were manufactured before the new standard took effect.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"An empty infant bouncer chair placed on a wooden kitchen countertop, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/infant-bouncer-fall","api_url":"https://likelier.app/api/fears/infant-bouncer-fall.json"},{"slug":"plane-crash","question":"What are the odds of dying in a plane crash?","quick_answer":"About 1 in 13.7 million per flight, or roughly 1 in 58,000 over a lifetime of typical flying — far rarer than the ~1 in 1,000 per flight most people picture. Around one in four US adults report being afraid of flying.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Flying is one of the most commonly cited phobias. In the Chapman Survey of American Fears, roughly one in four US adults reports being afraid or very afraid of flying in an airplane. That framing collapses \"dislike turbulence\" and \"avoid all flying\" into one bucket, but the headline is robust: flying feels much riskier than it is.\n","rough_estimate":"~1 in 1,000 per flight feels about right to most people","kind":"survey","survey_source":{"title":"Chapman Survey of American Fears, Wave 8","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/survey-american-fears.aspx","year":2022}},"native":{"display":"~1 in 13,700,000 per flight","numerator":1,"denominator":13700000,"unit":"per flight","population":"global commercial aviation passengers","exposures_per_year":4},"normalized":{"lifetime_us_adult":0.000017,"display":"1 in ~58,000 lifetime (US adult)","log_value":-4.77,"assumptions":"Assumes ~4 commercial boardings per year for a typical US adult (BTS data for adults who fly, including connecting segments), 59 years of remaining adult life, constant risk per boarding (conservative vs the long-term safety trend).\n","uncertainty":{"low":0.000008,"high":0.000035},"scope":"us_adult_lifetime"},"sources":[{"url":"https://news.mit.edu/2024/study-flying-keeps-getting-safer-0807","title":"Study: Flying keeps getting safer","publisher":"MIT News / Arnold Barnett","source_type":"peer_reviewed","statistic":"1 fatality per 13.7 million passenger boardings (2018-2022 commercial aviation)","excerpt":"\"The risk of a fatality from commercial air travel was 1 per every 13.7 million passenger boardings globally in the 2018-2022 period, a significant improvement from 1 per 7.9 million boardings in 2008-2017.\"\n","source_date":"2024-08-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411093610/https://news.mit.edu/2024/study-flying-keeps-getting-safer-0807","calculation_notes":"Barnett's \"death risk per boarding\" is already the exact statistic we want as the native value. Normalized = 1 − (1 − p_flight)^(flights_per_year × years) ≈ p_flight × flights × years for small p. Using ~4 boardings/year and 59 years: 4 × 59 / 13,700,000 ≈ 1 in 58,000 lifetime.\n","independence_note":"Barnett's MIT analysis pulls from IATA/ICAO commercial-aviation incident records plus passenger-boarding counts; overlaps with the NTSB accident database rather than being fully independent, though Barnett's per-boarding methodology is the distinctive analytical layer here.\n"},{"url":"https://www.ntsb.gov/safety/Pages/research.aspx","title":"Statistical Reviews — US Civil Aviation Accident Statistics","publisher":"National Transportation Safety Board (NTSB)","source_type":"govt_report","statistic":"NTSB publishes annual US civil aviation accident statistics covering Part 121, Part 135, and general aviation from 2005-2024","excerpt":"\"Statistical Reviews: 2005-2024 Accident Statistics — a downloadable Excel file providing summary statistics for US civil aviation accidents from 2005 through 2024. US Civil Aviation Accident Dashboard: 2008-2024 — an interactive report accompanying the statistics tables.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260403202813/https://www.ntsb.gov/safety/Pages/research.aspx","calculation_notes":"The NTSB research page provides downloadable accident statistics and an interactive dashboard covering 2005-2024. Part 121 air carrier operations (passenger-seat configuration >9 seats) have a well-documented record of extremely low fatal accident rates, consistent with Barnett's per-boarding figure. Used as a corroborating institutional source rather than the primary number.\n","independence_note":"NTSB and Barnett draw from overlapping aviation incident databases, so treat as partially dependent verification, not two fully independent estimates.\n"},{"url":"https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","title":"Odds of Dying (2024 Data)","publisher":"National Safety Council (Injury Facts)","source_type":"reputable_reference","statistic":"Lifetime odds of death from hornet, wasp, and bee stings: 1 in 42,626 (US, 2024 data)","excerpt":"\"Hornet, wasp, and bee stings — 1 in 42,626\" (from the \"Lifetime odds of death for selected causes, United States, 2024\" table).\n","source_date":"2025-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260706033319/https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","calculation_notes":"NSC's Injury Facts \"Odds of Dying\" table lists 2024 lifetime odds of death from hornet, wasp, and bee stings at 1 in 42,626, derived from NCHS mortality data and Census Bureau life-expectancy figures. Used here only as an illustrative comparison point for this entry's lifetime figure of 1 in 58,000 — both are low-probability lifetime causes of death of a similar order of magnitude. Not part of the native/normalized calculation for this entry.\n","independence_note":"NSC's mortality-based lifetime-odds tables are methodologically independent of Barnett's per-boarding aviation analysis and the NTSB accident statistics cited above.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by hornet, wasp, or bee sting (lifetime, US)","lifetime_us_adult":0.0000235}],"personal_factor_multipliers":[{"factor":"General / charter aviation vs commercial","multiplier":100,"notes":"NTSB annual statistics 2005-2024: general aviation (Part 91) fatality rate per flight hour is roughly 100× higher than Part 121 commercial air carrier operations. Charter (Part 135) sits between the two at ~10-20× commercial."},{"factor":"Developing-world carrier vs IATA Full Members","multiplier":14,"notes":"ICAO 2022 Safety Report: accident rate for non-IATA member carriers in Africa and parts of Asia was roughly 14× the global IATA member average on a per-departure basis (2017-2021 five-year window)."},{"factor":"Age 65+ (evacuation in survivable crashes)","multiplier":1.8,"notes":"FAA Civil Aeromedical Institute and NTSB cabin-safety research show older passengers evacuate more slowly and sustain higher injury rates in survivable accidents; estimated 1.5-2× elevated fatality risk in the survivable-crash subset, though overall per-boarding risk remains small."}],"short_label":"Plane crash","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Excludes private aviation, which has a meaningfully higher per-hour fatality rate. Also excludes major single-event anomalies (e.g., conflict zones, single-carrier outliers) which can dominate short windows but not long averages.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":5,"d8":3,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A small paper airplane floating against a muted grey-blue sky, flat vector illustration."},"canonical_url":"https://likelier.app/plane-crash","api_url":"https://likelier.app/api/fears/plane-crash.json"},{"slug":"tsunami","question":"What are the odds of being killed by a tsunami?","quick_answer":"A tsunami kills about 2,500 people a year on a long-window global average — roughly a 1 in 53,000 lifetime chance for the average global adult. Perception is shaped almost entirely by the 2004 Indian Ocean and 2011 Tohoku disasters rather than any survey isolating tsunami risk.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Tsunamis are one of the most visually iconic disasters in the modern news cycle, dominated by the 2004 Indian Ocean event and the 2011 Tōhoku earthquake. We haven’t yet found a rigorous recent survey that isolates \"fear of being killed by a tsunami\" as a standalone question, so the perceived side here is marked as editorial intuition rather than polled data. The plausible priors are strongly shaped by televised footage of those two events.\n","rough_estimate":"Eurobarometer groups tsunamis under natural disasters broadly; the World Risk Poll finds natural disasters a top-5 safety concern in coastal seismic regions but not globally; no major survey isolates tsunami mortality as a standalone item","kind":"survey","survey_source":{"title":"Special Eurobarometer — EU Citizens and Civil Protection","publisher":"European Commission, DG ECHO","url":"https://civil-protection-humanitarian-aid.ec.europa.eu/resources-campaigns/eurobarometer-reports_en","year":2024}},"native":{"display":"~2,500 tsunami deaths per year (long-window global average)","numerator":2500,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.000019,"display":"1 in ~53,000 lifetime (global adult)","log_value":-4.73,"assumptions":"Uses a long-window global average of roughly 2,000-3,000 tsunami deaths per year, obtained by smoothing the 1900-2025 record. The window matters enormously: the 2004 Indian Ocean event (~227,000 deaths) and 2011 Tōhoku event (~18,000 deaths) together account for the vast majority of modern tsunami mortality, so a 20-year window centered on 2004 gives ~12,500 deaths/year while a 125-year window gives closer to 2,500/year. Divided by a global population of ~8 billion and compounded over 60 adult life-years gives roughly 1 in 53,000 (1.9e-5). The uncertainty band below reflects window choice, not sampling noise. This is an \"average global adult\" figure and is not a useful personal estimate for any individual — see the body text.\n","uncertainty":{"low":0.0000033,"high":0.0000333},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.ncei.noaa.gov/products/natural-hazards/tsunamis-earthquakes-volcanoes/tsunamis/global-historical-data","title":"Global Historical Tsunami Database","publisher":"NOAA National Centers for Environmental Information (NCEI) / World Data Service","source_type":"govt_report","statistic":"Tsunamis have caused more than 500,000 fatalities throughout recorded history; 227,899 of those were from the 2004 Indian Ocean earthquake and tsunami alone.","excerpt":"\"The NOAA/World Data Service (WDS) tsunami database is a listing of historical tsunami source events and runup locations throughout the world that range in date from 2000 B.C. to the present. Tsunamis have been responsible for more than 500,000 fatalities throughout the world, 227,899 were from the 2004 Indian Ocean earthquake and tsunami.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260227133532/https://www.ncei.noaa.gov/products/natural-hazards/tsunamis-earthquakes-volcanoes/tsunamis/global-historical-data","calculation_notes":"The NCEI total of ~500,000 fatalities spans roughly 4,000 years of (sparse, heavily right-skewed) records. For a modern baseline we use the 1900-2025 subset, in which the 2004 and 2011 events dominate and the smoothed average is ~2,000-3,000 deaths per year. Annual per-capita risk ≈ 2,500 / 8,000,000,000 ≈ 3.1e-7; compounded over 60 adult years ≈ 1.9e-5, which we round to an order-of-magnitude 1 in 100,000.\n","independence_note":"NCEI compiles the primary global tsunami event catalogue from seismic and hazard records, methodologically distinct from WHO's public-health mortality reporting and from Doocy et al.'s systematic review of historical sources. Provides the long-record anchor the other two sources do not reach.\n"},{"url":"https://www.who.int/health-topics/tsunamis/","title":"Tsunamis — Health Topic","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"Between 1998-2017, tsunamis caused more than 250,000 deaths globally, including more than 227,000 deaths from the 2004 Indian Ocean tsunami.","excerpt":"\"Between 1998-2017, tsunamis caused more than 250 000 deaths globally, including more than 227 000 deaths due to the Indian Ocean tsunami in 2004.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260228015024/https://www.who.int/health-topics/tsunamis","calculation_notes":"WHO’s 20-year window yields ~12,500 deaths/year, but ~91% of that total comes from a single 2004 event. This is the upper-bound window for the uncertainty band (1 in ~30,000 global lifetime). Taken independently from the NCEI figure, which is compiled from source-event records rather than public-health reporting.\n","independence_note":"NCEI (event-based seismic/hazard records) and WHO (public-health reporting) aggregate mortality through different pipelines; treat as meaningfully independent verification of order of magnitude.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3644289/","title":"The Human Impact of Tsunamis: a Historical Review of Events 1900-2009 and Systematic Literature Review","publisher":"PLoS Currents / Doocy S, Daniels A, Dick A, Kirsch TD","source_type":"peer_reviewed","statistic":"255,195 deaths (range 252,619-275,784) from tsunamis 1900-2009; 89% from the 2004 Indian Ocean event; average ~2,552 deaths/year","excerpt":"\"255,195 deaths (range 252,619-275,784) and 48,462 injuries as a result of tsunamis from 1900 to 2009. However, the majority of deaths (89%) were attributed to a single event — the 2004 Indian Ocean tsunami.\"\n","source_date":"2013-04-16","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250327115612/https://pmc.ncbi.nlm.nih.gov/articles/PMC3644289/","calculation_notes":"Independent systematic review tabulating tsunami mortality from primary historical sources. The ~2,552 deaths/year long-window average confirms the 2,500/year figure used in this entry's normalized calculation.\n","independence_note":"Independent of NOAA NCEI — uses primary historical sources and literature review rather than the NCEI database.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Coastal resident in Pacific Rim or Indian Ocean subduction zone (elevation <10 m)","multiplier":100,"notes":"USGS and NOAA inundation modeling shows that the vast majority of tsunami casualties occur within a few kilometers of the shoreline at elevations below 10 m. In the 2004 Indian Ocean and 2011 Tōhoku events, >90% of deaths were in this coastal inundation zone. NOAA NCEI Global Historical Tsunami Database; USGS Open-File Report 2011-1211 (Tōhoku inundation maps). Compared to the global average, a coastal low-elevation resident in an active seismic zone faces approximately 100x the mean risk."},{"factor":"No access to early warning system (developing-world coastal community)","multiplier":5,"notes":"ITIC (International Tsunami Information Center) data and post-event analyses of the 2004 Indian Ocean tsunami show that communities with no warning system access had dramatically higher fatality rates than those that received advance warning. In Sri Lanka and Thailand, communities with 10-20 minutes of warning had fatality rates roughly 5x lower than those with no warning. ITIC/UNESCO-IOC tsunami preparedness reports."},{"factor":"Failing to evacuate within 10 minutes of strong ground shaking near coast","multiplier":4,"notes":"Proximity to a subduction fault means a locally generated tsunami can arrive in 10-30 minutes. Post-event surveys from 2011 Tōhoku (Mori et al., 2012, Earth, Planets and Space) found that residents who began evacuating within 10 minutes of shaking had ~4x lower fatality rates than those who delayed, consistent with vertical evacuation modeling by FEMA/NOAA for the Cascadia Subduction Zone."},{"factor":"Landlocked or inland resident (>50 km from coast)","multiplier":0.001,"notes":"Tsunami risk is essentially zero for populations without coastal exposure. NOAA NCEI database records no inland tsunami fatalities among populations more than a few kilometers from shoreline. The 0.001x multiplier reflects near-zero risk for landlocked populations versus the global coastal average."}],"short_label":"Tsunami","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker, not a personal estimate. Tsunami mortality is almost entirely coastal, concentrated in the Pacific Rim and Indian Ocean basins; landlocked populations and interior regions face effectively zero tsunami risk. Within coastal populations, local geography, elevation above sea level, and distance from subduction zones change the per-person risk by many orders of magnitude.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized wave curl against a pale sky, rendered as a flat geometric shape in muted blues, vector illustration."},"canonical_url":"https://likelier.app/tsunami","api_url":"https://likelier.app/api/fears/tsunami.json"},{"slug":"toddler-choking-while-eating","question":"What are the odds of an infant or toddler choking to death while eating?","quick_answer":"About 1 in 50,000 US children die from food choking during the highest-risk years, birth to age 4. Many parents picture something closer to 1 in a few thousand; non-fatal choking that needs adult intervention is far more common, but deaths are rare.\n","category":"kids","tags":["toddler","food"],"no_reliable_estimate":false,"perceived":{"description":"Food choking is one of the most commonly cited parental fears of the early years, alongside SIDS and drowning. Grapes, hot dogs, and nuts get named at nearly every first-time-parent class, and the fear is vivid enough that many parents avoid whole categories of food until age four or five. The mental model most parents carry is that any meal is a plausible choking event, and that the window of real risk stretches from the first solid food through the preschool years.\n","rough_estimate":"Most parents picture the per-child risk as 'rare but real' — order 1 in a few thousand","kind":"intuition"},"native":{"display":"~50-80 food-asphyxiation deaths per year (US children under 5)","numerator":1,"denominator":50000,"unit":"per child, 0-4 age window","population":"US children under 5, food-related choking deaths (ICD-10 W79)"},"normalized":{"lifetime_us_adult":0.00002,"display":"1 in ~50,000 per child during the 0-4 window (US)","log_value":-4.7,"assumptions":"Likelier normally reports lifetime-US-adult probabilities, but this entry is scoped to the peak-risk age window (0-4) for a single US child. Roughly 50-80 US children under age 5 die from food-related choking (ICD-10 W79) in a typical year. Against a US under-5 population of about 18-19 million, that is an annual rate of roughly 3 to 4 per million per child. Compounded across the five-year 0-4 window, 1 - (1 - 3.5e-6)^5 ≈ 1.75e-5, which rounds to about 2e-5, or roughly 1 in 50,000 per child across the peak-risk age window. This counts only food-asphyxiation deaths, not non-food foreign bodies (toys, coins, button batteries, balloons), which are coded separately under W80 and carry their own mortality burden. The non-fatal serious-event rate (any choking that requires adult intervention, Heimlich, or a medical visit) is roughly two to three orders of magnitude higher than the fatal rate — see regional_breakdown.\n","uncertainty":{"low":0.00001,"high":0.00004},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5142a1.htm","title":"Nonfatal Choking-Related Episodes Among Children — United States, 2001","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"160 US children aged <14 died from inhaled/ingested foreign body obstruction in 2000 (ICD-10 W79-W80); ~41% food-related ≈ ~66 food-choking deaths; 17,537 pediatric ED visits for choking in 2001","excerpt":"\"During 2000, the latest year for which national mortality data were available, 160 children aged &lt;14 years died from obstruction of the respiratory tract associated with inhaled or ingested foreign bodies [...] food and nonfood substances were associated with 41% and 59% of these deaths, respectively [...] an estimated 17,537 children aged &lt;14 years were treated in EDs for choking-related episodes in 2001.\"\n","source_date":"2002-10-25","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260318060241/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5142a1.htm","calculation_notes":"Provides the ICD-10 W79 (food) vs W80 (non-food foreign body) split used to convert all-cause pediatric choking deaths into a food-specific subset. 160 × 0.41 ≈ 66 food-choking deaths among children &lt;14 in 2000. Later data (Sideris et al. 2021, below) shows roughly 147 pediatric choking deaths per year averaged over 2001-2016 across ages 0-19, with 75% concentrated in children under 5. Combining the two, food-related choking deaths among US children 0-4 sit in the ~50-80 per year band used in the native figure.\n","independence_note":"CDC MMWR analysis draws from NCHS death-certificate data (ICD-10 W79/W80) and CPSC NEISS pediatric ED estimates — the same combined upstream that feeds Chapin 2013 and Sideris 2021. Treat the three surveillance-based sources as one pipeline; the AAP policy statement is the separate clinical authority anchoring the high-risk food list.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23897916/","title":"Nonfatal Choking on Food Among Children 14 Years or Younger in the United States, 2001-2009","publisher":"Pediatrics — Chapin MM, Rochette LM, Annest JL, Haileyesus T, Conner KA, Smith GA","source_type":"peer_reviewed","statistic":"12,435 annual US pediatric ED visits for nonfatal food-related choking (2001-2009); rate 20.4 per 100,000; mean age 4.5 years; hard candy 15%, other candy 13%, meat 12%, bone 12%","excerpt":"\"An estimated 111,914 (95% confidence interval: 83,975-139,854) children ages 0 to 14 years were treated in US hospital emergency departments from 2001 through 2009 for nonfatal food-related choking, yielding an average of 12,435 children annually and a rate of 20.4 (95% confidence interval: 15.4-25.3) visits per 100,000 population. The mean age of children treated for nonfatal food-related choking was 4.5 years [...] Hard candy was associated with the greatest number (15.5% [95% CI: 12.8-18.2]) of choking episodes, followed by other candy (12.8% [95% CI: 10.6-15.0]), meat (12.2% [95% CI: 9.9-14.5]), and bone (12.0% [95% CI: 9.7-14.3]).\"\n","source_date":"2013-08-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20251217075830/https://pubmed.ncbi.nlm.nih.gov/23897916/","calculation_notes":"The Chapin et al. annual rate (20.4 ED visits per 100,000 children per year for nonfatal food choking) is the basis for the non-fatal regional_breakdown entry. Compounded over five years of the 0-4 window at roughly 30 per 100,000 per year (slightly higher than all-ages-0-14 because the under-5 subgroup is over-represented in the data — mean age 4.5), the cumulative ED-visit rate is roughly 1.5 per 1,000. The \"1 in 20\" headline used in the regional_breakdown includes broader serious events that parents manage at home (successful back blows, Heimlich, or a pediatrician call without an ED visit), which are not captured in Chapin's ED-only denominator but are well above the ED-visit rate in survey data.\n","independence_note":"Chapin et al. uses the CPSC National Electronic Injury Surveillance System (NEISS) All Injury Program — a stratified sample of US hospital EDs. Shares upstream with Sideris 2021 (same NEISS + NCHS mortality files) and partially overlaps with the MMWR surveillance data; treat the three as different slices of a single US pediatric-injury surveillance pipeline.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33819896/","title":"Persistence of choking injuries in children","publisher":"International Journal of Pediatric Otorhinolaryngology — Sideris GA et al.","source_type":"peer_reviewed","statistic":"2,347 US pediatric choking deaths (ages 0-19) and 305,814 nonfatal injuries, 2001-2016; 75% of fatalities in children under 5; fatality rate in under-5 unchanged after 2010 AAP recommendations","excerpt":"\"From 2001 to 2016, there were a total of 305,814 nonfatal injuries and 2347 choking deaths in children from 0 to 19 years. [...] Children under five years of age accounted for 73% of nonfatal injuries and 75% of choking fatalities. [...] a decrease in the choking fatalities rate in all children (0.18/100,000 versus 0.16/100,000, respectively) but no change in fatalities rate for children under five.\"\n","source_date":"2021-04-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260216004448/https://pubmed.ncbi.nlm.nih.gov/33819896/","calculation_notes":"2347 deaths / 16 years ≈ 147 pediatric choking deaths per year across ages 0-19. 75% × 147 ≈ 110 deaths per year in children under 5 from all foreign-body choking (W79 food + W80 other). Applying the ~41-50% food share from the MMWR/WISQARS breakdown gives roughly 50-75 food-choking deaths per year among US children 0-4. This anchors the native \"~50-80 per year\" figure. Sideris et al. also document that the under-5 rate did not improve after the 2010 AAP policy statement, which is the empirical basis for treating this as a \"calibrated\" rather than \"debunked\" fear.\n","independence_note":"Both Chapin and Sideris draw from CPSC's National Electronic Injury Surveillance System (NEISS) and NCHS mortality files, so their denominators are methodologically linked. Treated as two views of the same underlying data rather than fully independent estimates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20176668/","title":"Policy Statement — Prevention of Choking Among Children","publisher":"American Academy of Pediatrics, Committee on Injury, Violence, and Poison Prevention — Pediatrics 125(3):601-607","source_type":"reputable_reference","statistic":"AAP high-risk food list: hot dogs, hard candy, nuts and seeds, whole grapes, raw carrots, apples, popcorn, chunks of peanut butter, marshmallows, chewing gum, chunks of meat or cheese","excerpt":"\"Choking is a leading cause of morbidity and mortality among children, especially those aged 3 years or younger. Food, coins, and toys are the primary causes of choking-related injury and death.\"\n","source_date":"2010-03-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260413164733/https://pubmed.ncbi.nlm.nih.gov/20176668/","calculation_notes":"The AAP policy statement is the canonical authority for the \"high-risk food\" list used in the personal_factor_multipliers and the long-form body. It is also the basis for the under-3 peak-risk framing. Reaffirmed by the AAP in October 2019.\n","independence_note":"AAP policy statement synthesises clinical case-series literature and expert consensus rather than a surveillance dataset. Independent of the NEISS/NCHS surveillance pipeline feeding Chapin, Sideris, and the MMWR brief — addresses the high-risk food categorisation and age-concentration rather than the mortality denominator.\n"}],"comparison_anchors":[{"label":"Choking death, all ages, lifetime (US adult)","lifetime_us_adult":0.00091},{"label":"SIDS, per US infant","lifetime_us_adult":0.00035},{"label":"Death in a plane crash, lifetime (US adult)","lifetime_us_adult":0.000017},{"label":"Drowning death, lifetime (US adult)","lifetime_us_adult":0.000725}],"regional_breakdown":[{"region":"Food asphyxiation death, US child 0-4 (the headline number)","probability":0.00002,"notes":"~50-80 US food-choking deaths per year among children under 5, divided across a population of ~18-19 million children in that age band, compounded over the five year window.\n"},{"region":"Non-fatal serious choking event requiring adult intervention, US child 0-4","probability":0.05,"notes":"Rough survey-based estimate. Chapin et al.'s 20.4 per 100,000 per year only counts ED visits — the broader \"any event where a parent had to intervene with back blows, Heimlich, or a finger sweep\" rate is an order of magnitude higher and captures the distinction between fatal risk (very rare) and scary-incident risk (common).\n"},{"region":"Food asphyxiation death, US 1-2 year olds specifically","probability":0.00005,"notes":"The AAP policy statement names children aged 3 years or younger as the peak-risk group, and mortality data cluster inside that band. The 1-2 year subset carries roughly two to three times the under-5 average rate, mostly driven by the transition from pureed to whole foods against limited chewing and airway geometry.\n"}],"personal_factor_multipliers":[{"factor":"age 1-2 (peak risk)","multiplier":2.5,"notes":"Mortality rates concentrate in the second and third year of life, when children are exposed to whole-texture foods but still have narrow airways and immature chewing patterns.\n"},{"factor":"eating in car seat, stroller, or reclined position","multiplier":2,"notes":"Reclined eating makes airway clearance harder and delays adult response because the caregiver is typically driving or walking. Exact multiplier is uncertain; the AAP policy statement flags the posture dimension qualitatively but does not publish a quantitative risk ratio.\n"},{"factor":"continuous adult supervision during meals","multiplier":0.4,"notes":"Successful back blows and Heimlich maneuvers convert would-be fatal events into non-fatal ones. The protective effect of a nearby adult is the main reason the fatality rate is as low as it is given how common non-fatal events are.\n"},{"factor":"no high-risk AAP-listed foods given whole (grapes, hot dogs, whole nuts, hard candy, popcorn)","multiplier":0.3,"notes":"The AAP high-risk list accounts for a disproportionate share of fatal and hospitalized cases. Cutting or avoiding these foods until roughly age 4 is the single highest-leverage input parents control. Likelier does not give advice; see the AAP policy statement linked under sources for specifics.\n"}],"short_label":"Toddler choking","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The fatal number is small and the non-fatal scary-event rate is large, which is unusual among fears on this site and is the reason parental vigilance is calibrated rather than debunked. This entry covers food-related choking deaths only (ICD-10 W79); non-food foreign-body obstruction (toys, coins, button batteries, balloons — coded W80) carries its own mortality burden and is not included in the headline number. It also excludes positional asphyxia in sleep (coded W75, and captured by the SIDS and unsafe-sleep literature rather than the choking literature), anaphylaxis (coded as allergic-reaction death), and aspiration pneumonia deaths that occur days after the event rather than at the scene. Chapin et al. and Sideris et al. draw from overlapping CPSC NEISS and NCHS files, so treat them as two views of the same underlying surveillance data. Finally, the \"eating in a car seat while being driven\" subset that parents sometimes ask about specifically is captured inside this entry rather than broken out separately, because the mortality data does not resolve that finely — the posture multiplier above is the best available quantitative handle.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-12","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single small round grape resting on a pale grey-blue plate beside a child-sized spoon, flat vector illustration, viewed from directly above."},"canonical_url":"https://likelier.app/toddler-choking-while-eating","api_url":"https://likelier.app/api/fears/toddler-choking-while-eating.json"},{"slug":"unsupervised-infant-eating","question":"What are the odds of a choking emergency if an infant eats unsupervised?","quick_answer":"The baseline fatal food-choking rate is about 1 in 50,000 per child across the 0-4 window, and roughly 60% of fatal events happen with an adult present, so supervision is not the deciding factor parents imagine. Nonfatal choking sends about 12,400 US under-14s to the ER a year. Parents often intuit danger on the order of 1 in a few hundred meals.\n","category":"kids","tags":["infant","food"],"no_reliable_estimate":false,"perceived":{"description":"Parents are told never to leave a baby alone with food, and the fear that stepping away for thirty seconds could be the moment a grape lodges in a windpipe is near-universal in parenting forums. The mental model is binary: present equals safe, absent equals catastrophe. Infant CPR classes reinforce the urgency, and the widespread advice to \"watch every single bite\" implies that unwitnessed eating is the primary mechanism of fatal choking.\n","rough_estimate":"Most parents intuit that unsupervised eating is 'extremely dangerous' — order 1 in a few hundred meals","kind":"intuition"},"native":{"display":"~60% of fatal pediatric food-choking events occur WITH an adult present; nonfatal ER rate ~12,400/year (US, under-14)","numerator":3,"denominator":5,"unit":"proportion of fatal choking events occurring in the presence of a caregiver","population":"US children under 5, food-related choking fatalities where supervision status is documented"},"normalized":{"lifetime_us_adult":0.00002,"display":"1 in ~50,000 per child during the 0-4 window (baseline fatal choking rate regardless of supervision)","log_value":-4.7,"assumptions":"The underlying fatal choking rate for US children 0-4 is approximately 1 in 50,000 per child across the five-year window. This is derived from CDC MMWR (160 pediatric foreign-body-obstruction deaths in 2000, 41% food-related ≈ 66 food-choking deaths among children under 14) and Sideris et al. 2021 (2,347 pediatric choking deaths 2001-2016, 75% in children under 5), which together place food-choking deaths among US children 0-4 in the ~50-80 per year band; against ~18-19 million under-5 children that is ~3-4 per million per child per year, compounding to 1 - (1 - 3.5e-6)^5 ≈ 1.75e-5 ≈ 2e-5 across the 0-4 window. This entry reframes that baseline through the supervision lens: roughly 60% of food-related choking injuries occur with an adult present but with improperly prepared or unsuitable food, and ~40% occur without supervision (Lorenzoni et al. 2022). The meaningful variable is not presence versus absence but trained-response versus untrained-response. A caregiver trained in infant back blows and the modified Heimlich technique converts most obstructive events into non-fatal outcomes within seconds; an untrained present adult often freezes or applies incorrect technique (finger sweeps, shaking), making their effective contribution to survival similar to absence. The normalized figure remains 2e-5 because supervision status shifts outcome conditional on an event occurring but does not substantially change event frequency.\n","uncertainty":{"low":0.00001,"high":0.00004},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9160792/","title":"Regulatory and Educational Initiatives to Prevent Food Choking Injuries in Children: An Overview of the Current Approaches","publisher":"Frontiers in Public Health — Lorenzoni G et al.","source_type":"peer_reviewed","statistic":"~40% of food-related injuries occur without adult supervision; ~60% occur with supervision but with improperly prepared food","excerpt":"\"~40% of food-related injuries occur in the absence of adult supervision while the child is eating. The remaining 60% occur with adult supervision but with the children having been served improperly prepared or unsuitable food.\"\n","source_date":"2022-05-19","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260206230452/https://pmc.ncbi.nlm.nih.gov/articles/PMC9160792/","calculation_notes":"This study provides the key finding that adult presence alone does not prevent choking events. The 60/40 split demonstrates that the majority of incidents occur under supervision, reframing the parental fear from \"I must be present\" to \"I must know what to do and what to serve.\" Used as the primary source for the supervision-outcome framing of this entry. The fatal rate itself is derived from the CDC MMWR and Sideris et al. 2021 sources below (50-80 food-choking deaths per year among US children under 5, yielding ~2e-5 per child across the 0-4 window).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27647715/","title":"A Baby-Led Approach to Eating Solids and Risk of Choking","publisher":"Pediatrics — Fangupo LJ, Heath A-LM, Williams SM, et al.","source_type":"primary_study","statistic":"35% of infants in both BLW and spoon-fed groups choked at least once between 6-8 months; no significant difference between groups (P > .20 at all time points)","excerpt":"\"A total of 35% of infants choked at least once between 6 and 8 months of age, and there were no significant group differences in the number of choking events at any time (all Ps > .20). [...] Infants following a baby-led approach to feeding that includes advice on minimizing choking risk do not appear more likely to choke than infants following more traditional feeding practices.\"\n","source_date":"2016-10-01","source_accessed":"2026-04-20","archive_url":"http://web.archive.org/web/20260212232124/https://pubmed.ncbi.nlm.nih.gov/27647715/","calculation_notes":"The BLISS randomized controlled trial (n=206) demonstrates that the method of food introduction (baby-led weaning versus spoon-feeding) does not significantly alter choking frequency when both groups receive guidance on minimizing risk. This supports the thesis that the meaningful risk modifier is food preparation and caregiver response competence, not the specific feeding paradigm or the intensity of bite-by-bite watching. The 35% gagging/choking rate at 6-8 months is the non-fatal event frequency and is orders of magnitude above the fatal rate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23897916/","title":"Nonfatal Choking on Food Among Children 14 Years or Younger in the United States, 2001-2009","publisher":"Pediatrics — Chapin MM, Rochette LM, Annest JL, Haileyesus T, Conner KA, Smith GA","source_type":"peer_reviewed","statistic":"12,435 annual US pediatric ED visits for nonfatal food-related choking (2001-2009); rate 20.4 per 100,000; children ≤1 year account for 37.8% of cases","excerpt":"\"An estimated 111,914 (95% confidence interval: 83,975-139,854) children ages 0 to 14 years were treated in US hospital emergency departments from 2001 through 2009 for nonfatal food-related choking, yielding an average of 12,435 children annually and a rate of 20.4 (95% confidence interval: 15.4-25.3) visits per 100,000 population.\"\n","source_date":"2013-08-01","source_accessed":"2026-04-20","archive_url":"http://web.archive.org/web/20260421200955/https://pubmed.ncbi.nlm.nih.gov/23897916/","calculation_notes":"Provides the nonfatal event denominator. At 12,435 ED visits per year against roughly 50-80 fatal events, the case-fatality ratio for food choking events serious enough to reach an ED is roughly 1 in 150-250. Most events that reach the ED were resolved by a bystander (caregiver back blows, Heimlich) before arrival — the ED visit is precautionary. This supports the claim that trained caregiver intervention is the variable separating fatal from nonfatal outcomes.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12821163/","title":"Knowledge and first aid management of choking children among parents in a tertiary care hospital, Sri Lanka","publisher":"BMC Emergency Medicine","source_type":"peer_reviewed","statistic":"Only 38.8% of parents demonstrated good knowledge of choking first aid; prior first-aid training was not a significant predictor in the adjusted model (adjusted OR = 0.750, P = 0.272). Main information sources were healthcare professionals (53.1%) and media (46.9%)","excerpt":"\"Knowledge of parents regarding identification of symptoms and signs of choking and provision of first aid for a choking child is insufficient. [...] Main sources of information regarding choking first aid were health care professionals and media.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-20","archive_url":"http://web.archive.org/web/20260504061334/https://pmc.ncbi.nlm.nih.gov/articles/PMC12821163/","calculation_notes":"Demonstrates the knowledge gap that explains why adult presence does not automatically confer protection. If roughly 61% of parents lack adequate choking first-aid knowledge, then 61% of \"supervised\" choking events feature an adult who cannot effectively intervene — supporting the personal_factor_multiplier distinction between trained and untrained caregivers. The study population is Sri Lankan, but multiple cross-sectional studies in Saudi Arabia, Ethiopia, and Syria report similar 60-75% inadequacy rates, suggesting the knowledge gap is not US-specific.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5142a1.htm","title":"Nonfatal Choking-Related Episodes Among Children — United States, 2001","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"160 US children aged <14 died from inhaled/ingested foreign body obstruction in 2000 (ICD-10 W79-W80); 41% food-related ≈ ~66 food-choking deaths; 17,537 pediatric ED visits for choking in 2001","excerpt":"\"During 2000, the latest year for which national mortality data were available, 160 children aged &lt;14 years died from obstruction of the respiratory tract associated with inhaled or ingested foreign bodies [...] food and nonfood substances were associated with 41% and 59% of these deaths, respectively [...] an estimated 17,537 children aged &lt;14 years were treated in EDs for choking-related episodes in 2001.\"\n","source_date":"2002-10-25","source_accessed":"2026-07-10","archive_url":"http://web.archive.org/web/20260318060241/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5142a1.htm","calculation_notes":"Anchors the fatal-choking denominator for the normalized figure. 160 pediatric foreign-body-obstruction deaths in 2000 × 41% food share ≈ 66 food-choking deaths among children &lt;14. Combined with Sideris et al. 2021 (below, which finds 75% of pediatric choking fatalities fall in children under 5), food-related choking deaths among US children 0-4 sit in the ~50-80 per year band. Against a US under-5 population of ~18-19 million, that is ~3-4 per million per child per year; compounded across the five-year 0-4 window, 1 - (1 - 3.5e-6)^5 ≈ 1.75e-5, rounding to ~2e-5 (roughly 1 in 50,000). Same surveillance pipeline (NCHS mortality + CPSC NEISS) as Chapin 2013 and Sideris 2021 — treat the three as one source of truth for the fatal/nonfatal rates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33819896/","title":"Persistence of choking injuries in children","publisher":"International Journal of Pediatric Otorhinolaryngology — Sideris GA et al.","source_type":"peer_reviewed","statistic":"2,347 US pediatric choking deaths (ages 0-19) and 305,814 nonfatal injuries, 2001-2016; 75% of fatalities in children under 5; overall fatality rate 0.16 per 100,000 with no change in under-5 rate","excerpt":"\"From 2001 to 2016, there were a total of 305,814 nonfatal injuries and 2347 choking deaths in children from 0 to 19 years. [...] Children under five years of age accounted for 73% of nonfatal injuries and 75% of choking fatalities. [...] a decrease in the choking fatalities rate in all children (0.18/100,000 versus 0.16/100,000, respectively) but no change in fatalities rate for children under five.\"\n","source_date":"2021-04-01","source_accessed":"2026-07-10","archive_url":"http://web.archive.org/web/20260216004448/https://pubmed.ncbi.nlm.nih.gov/33819896/","calculation_notes":"2347 deaths / 16 years ≈ 147 pediatric choking deaths per year across ages 0-19; 75% × 147 ≈ 110 deaths per year in children under 5 from all foreign-body choking (W79 food + W80 non-food). Applying the ~41% food share from the CDC MMWR breakdown gives roughly 45-70 food-choking deaths per year among US children 0-4, consistent with the ~50-80 band used to derive the normalized ~2e-5 fatal rate. Sideris also documents that the under-5 fatality rate did not improve after the 2010 AAP policy statement — the empirical basis for treating this as a calibrated fear.\n"}],"comparison_anchors":[{"label":"Fatal food choking, per US child 0-4 (baseline)","lifetime_us_adult":0.00002},{"label":"SIDS, per US infant","lifetime_us_adult":0.00035},{"label":"Drowning death, lifetime (US adult)","lifetime_us_adult":0.000725},{"label":"Death in a plane crash, lifetime (US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"caregiver trained in infant BLS/back blows","multiplier":0.3,"notes":"A trained caregiver can resolve most complete obstructions within the critical 4-minute window before hypoxic brain injury. The AHA 2025 pediatric BLS guidelines recommend alternating 5 back blows and 5 chest thrusts for infants under 1. Training converts the majority of would-be fatal events into resolved events. Exact risk reduction is not quantified in a single RCT but is inferred from the observation that most ED-presenting choking cases were pre-resolved by bystander intervention.\n"},{"factor":"untrained caregiver present (no first-aid knowledge)","multiplier":0.9,"notes":"An untrained adult who is present but does not know the correct technique provides only marginal benefit over absence. Common incorrect responses include blind finger sweeps (which can push the object deeper), shaking the child, or panicking without acting. The ~70% inadequacy rate in parental choking knowledge (multiple cross-sectional studies) means most \"supervised\" events feature an effectively untrained responder.\n"},{"factor":"baby-led weaning with BLW safety guidance","multiplier":1,"notes":"The BLISS trial (Fangupo et al. 2016, n=206) found no significant difference in choking frequency between BLW and spoon-fed infants when both groups received guidance on minimizing choking risk. BLW does not increase or decrease the baseline rate when implemented with appropriate food selection.\n"},{"factor":"high-risk foods served whole (grapes, hot dogs, nuts, hard candy)","multiplier":3,"notes":"The AAP high-risk food list accounts for a disproportionate share of fatal and hospitalized cases. Round, firm foods that match airway diameter are the highest-risk shape factor. This multiplier applies regardless of supervision status — the food itself is the hazard, not the watching.\n"},{"factor":"eating while reclined (car seat, stroller) without direct observation","multiplier":2.5,"notes":"Combines the mechanical disadvantage of a reclined airway with delayed recognition and response. The caregiver is typically driving or walking and cannot see or reach the child quickly. This is the specific scenario where absence of direct visual supervision materially worsens outcomes.\n"}],"short_label":"Unsupervised infant choking","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The underlying fatal rate (1 in ~50,000 per child across the 0-4 window) is derived from CDC MMWR mortality data and Sideris et al. 2021, the same surveillance basis as the toddler-choking-while-eating entry, and is reframed here through the supervision and response-competence lens. The 60/40 supervised-vs-unsupervised split comes from a 2022 review synthesizing multiple studies and should be treated as approximate rather than precise. The trained-vs-untrained multiplier is inferred from indirect evidence (bystander resolution rates, knowledge-gap surveys, BLS guidelines) rather than a single randomized trial of supervised-vs-unsupervised feeding — no such trial exists or would be ethical. The BLISS trial addresses feeding method (BLW vs spoon-feeding), not supervision intensity, and its 35% choking rate at 6-8 months captures gagging events that parents may misidentify as choking — true complete airway obstruction is far rarer. Finally, cross-cultural knowledge-gap data (Sri Lanka, Saudi Arabia, Ethiopia) may not directly transfer to US parents who have higher baseline exposure to infant CPR messaging, though US-specific surveys suggest similarly low rates of formal BLS certification among parents of young children.\n","quality_score":{"d1":3,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-20","image":{"alt":"A small high chair tray with a few soft food pieces arranged on it, seen from above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/unsupervised-infant-eating","api_url":"https://likelier.app/api/fears/unsupervised-infant-eating.json"},{"slug":"general-anesthesia-death","question":"What are the odds of dying from general anesthesia during surgery?","quick_answer":"Dying from general anesthesia is about a 1 in 137,000 chance per procedure for healthy adults, or roughly 1 in 46,000 over a lifetime of typical surgeries. Patients commonly guess something like 1 in 5,000, orders of magnitude too high, drawing on an era when anesthesia really was dangerous.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"General anesthesia consistently ranks among the top procedural fears reported by surgical patients. Pre-operative anxiety surveys find that roughly 70-80 % of patients cite fear of \"not waking up\" as a primary concern, often estimating the risk at somewhere between 1 in 1,000 and 1 in 10,000 — orders of magnitude higher than the modern evidence supports. The fear draws on an era when anesthesia really was dangerous, amplified by dramatic depictions in film and the fundamental loss of consciousness involved.\n","rough_estimate":"~1 in 5,000 is a common lay estimate","kind":"intuition"},"native":{"display":"~1 in 137,000 per anesthetic (healthy adults)","numerator":1,"denominator":137000,"unit":"per general anesthetic administered","population":"ASA I-II adults in high-income countries"},"normalized":{"lifetime_us_adult":0.0000219,"display":"~1 in 46,000 lifetime (US adult)","log_value":-4.66,"assumptions":"Assumes ~3 general anesthetics over a US adult lifetime (consistent with population-level surgical utilization data: ~50 million inpatient + outpatient procedures/year in the US for 330 million people, fraction under GA, spread over 59 years of remaining adult life). Uses the anesthesia-attributable mortality rate of ~1 in 137,000 per anesthetic for ASA I-II patients in high-income countries, from Schiff et al. 2014 (corroborated by Li et al. 2009). Bainbridge et al. 2012's own pooled global rate is higher (~1 in 29,400 for the 1990s-2000s, across developed and developing countries and all ASA classes) — see that source's calculation_notes. Lifetime ≈ 1 − (1 − 1/137,000)^3 ≈ 1/46,000.\n","uncertainty":{"low":0.0000117,"high":0.0000369},"scope":"us_adult_lifetime"},"sources":[{"url":"https://doi.org/10.1016/S0140-6736(12)60990-8","title":"Perioperative and anaesthetic-related mortality in developed and developing countries: a systematic review and meta-analysis","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Mortality solely attributable to anaesthesia (pooled across all included studies) declined from 357 per million before the 1970s to 34 per million in the 1990s-2000s","excerpt":"\"Mortality solely attributable to anaesthesia declined over time, from 357 per million (95% CI 324–394) before the 1970s to 52 per million (42–64) in the 1970s–80s, and 34 per million (29–39) in the 1990s–2000s (p<0·00001).\"\n","source_date":"2012-09-22","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420041340/https://www.thelancet.com/retrieve/pii/S0140673612609908","calculation_notes":"Bainbridge et al.'s pooled per-million-by-decade figures (357/52/34 per million solely- attributable-to-anaesthesia deaths for before-1970s/1970s-80s/1990s-2000s, across the 87 included studies and 21.4 million anaesthetics) span both developed and developing countries and all ASA classes, so the most recent bucket (34 per million ≈ 1 in 29,400) sits well above the rate for healthy (ASA I-II) adults in a single high-income country today. Used here as historical-trend context for the ~10-fold decline in anaesthesia- attributable mortality since before 1970. The native per-anesthetic rate for healthy adults in high-income countries is grounded on Schiff et al. 2014 instead — see that source's calculation_notes for the derivation.\n"},{"url":"https://doi.org/10.1093/bja/aeu094","title":"Major incidents and complications in otherwise healthy patients undergoing elective procedures: results based on 1.37 million anaesthetic procedures","publisher":"British Journal of Anaesthesia","source_type":"peer_reviewed","statistic":"Death or serious complication rate 26.2 per million elective ASA I-II procedures; 7.3 per million with possible direct anaesthetic involvement","excerpt":"\"Of 1 374 678 otherwise healthy, ASA I and II patients in the CDS database, 36 met the study inclusion criteria … death or serious complication rate of 26.2 per million (95% confidence interval, 19.4–34.6) procedures … for those with possible direct anaesthetic involvement, 7.3 per million cases (95% CI, 3.9–12.3).\"\n","source_date":"2014-07-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260505054952/https://bjanaesthesia.org/retrieve/pii/S0007091217315519","calculation_notes":"Schiff et al. 2014 (PMID 24801456) analysed 1.37 million elective ASA I-II procedures from a German national surveillance database (1999-2010), finding a rate of 7.3 per million (~1 in 137,000) for deaths with possible direct anaesthetic involvement in healthy patients — lower than Bainbridge et al.'s pooled 1990s-2000s rate of 34 per million (~1 in 29,400) because Bainbridge's figure spans developed and developing countries and all ASA classes across many studies, while Schiff's is a single contemporary high-income-country dataset restricted to healthy (ASA I-II) patients. Used as the primary basis for the native rate: native ≈ 1/137,000 per anesthetic; normalized assumes 3 lifetime anesthetics: 1 − (1 − 1/137,000)^3 ≈ 1/46,000.\n","independence_note":"Schiff et al. uses a German national surveillance database (CDS), independent of Bainbridge's systematic review sources. Partially independent.\n"},{"url":"https://doi.org/10.1097/ALN.0b013e31819b5bdc","title":"Epidemiology of Anesthesia-related Mortality in the United States, 1999-2005","publisher":"Anesthesiology (Li et al.)","source_type":"peer_reviewed","statistic":"Anesthesia-related death rate 8.2 per million hospital surgical discharges in the US (1999-2005); ~1.1 per million population per year","excerpt":"\"The anesthesia-related death rate was 1.1 per million population per year … The estimated mortality risk from anesthesia complications for inpatients was 8.2 … deaths per million hospital surgical discharges … Anesthesia complications were the underlying cause in 241 (10.9%) of these deaths and a contributing factor in the remaining 1,970 (89.1%) deaths.\"\n","source_date":"2009-04-01","source_accessed":"2026-04-26","calculation_notes":"Li et al. (PMID 19322941) analysed US national mortality data for 1999-2005, finding 2,211 anesthesia-related deaths. The 8.2 per million hospital surgical discharges (~1 in 122,000) is the same order of magnitude as Schiff et al.'s 7.3 per million for healthy ASA I-II patients in Germany. Note: Li et al.'s figures are US population-wide, include all ASA classes (not just healthy ASA I-II), and count both underlying-cause and contributing-factor deaths, so likely overstate the healthy-ASA-I-II-only, solely- attributable rate somewhat. Corroborates ~1 in 137,000 as the right order of magnitude for the native rate used here.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"ASA I-II (healthy adults), developed countries","probability":0.0000219,"notes":"~1 in 46,000 lifetime assuming 3 GAs; native rate ~1 in 137,000 per anesthetic (Schiff et al. 2014, corroborated by Li et al. 2009)"},{"region":"Emergency surgery (all ASA classes)","probability":0.0005,"notes":"Emergency procedures carry 3-5x higher mortality than matched elective cases"},{"region":"Developing countries (all patients)","probability":0.0002,"notes":"Bainbridge et al. report ~1 in 5,000-10,000 per anesthetic in lower-resource settings"}],"personal_factor_multipliers":[{"factor":"ASA I (healthy, no comorbidities)","multiplier":0.3,"notes":"Lower end of the ASA I-II range"},{"factor":"Age > 70","multiplier":3,"notes":"Age is a strong independent predictor of perioperative mortality"},{"factor":"Emergency surgery","multiplier":5,"notes":"Emergency procedures carry substantially higher risk regardless of ASA class"}],"short_label":"Anesthesia death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The native rate (1 in 137,000) refers specifically to deaths solely attributable to anesthesia — not to total perioperative mortality, which includes surgical complications, underlying disease, and hemorrhage. Total perioperative mortality for all-comers is roughly 1 in 500-1,000 (Bainbridge et al. 2012 report 1,176 per million, ≈1 in 850, for the most recent 1990s-2000s pooled bucket), but isolating the anesthesia-specific fraction is the appropriate answer to the question posed. Rates in low-income countries remain substantially higher.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An anesthesia mask floating gently above a calm surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/general-anesthesia-death","api_url":"https://likelier.app/api/fears/general-anesthesia-death.json"},{"slug":"dog-hot-car-heatstroke","question":"What are the odds of a dog dying from being left in a hot car?","quick_answer":"About 1 in 42,000 per dog over a 12-year lifespan. Most owners recognize the summer danger; the blind spot is the shoulder season, when a sunlit closed car can reach lethal heat even at 70°F outside.\n","category":"health","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Awareness of hot-car risk for dogs has risen sharply alongside viral news coverage and state laws addressing pets left in parked vehicles. Most dog owners rate this as a real danger in summer heat, which is accurate. The systematic blind spot is in the shoulder seasons: people often assume that mild ambient temperatures — anything below roughly 80°F — make the car safe for a short errand. In reality, a car parked in direct sun with windows closed can reach interior temperatures high enough to cause heatstroke in a dog within 20 to 30 minutes even when it is only 70°F outside. Short-muzzled breeds (bulldogs, pugs, French bulldogs) and overweight or elderly dogs are especially vulnerable because they cannot pant efficiently enough to dissipate heat. Dogs also lack the behavioral ability to signal danger before they are already in crisis.\n","rough_estimate":"Most owners recognize the summer risk; few appreciate that 70°F ambient with full sun can kill","kind":"intuition"},"native":{"display":"~2 per million dogs per year (estimated, US; no national surveillance exists)","numerator":2,"denominator":1000000,"unit":"per year","population":"US pet dogs"},"normalized":{"lifetime_us_adult":0.000024,"display":"~1 in 42,000 (per dog, estimated over a 12-year lifespan)","log_value":-4.62,"assumptions":"No national US surveillance system tracks dog hot-car deaths. The best epidemiological data comes from Hall et al. (2020, PMC7459873), a UK VetCompass study of 905,543 dogs, which found vehicular confinement accounted for 5.2% of heat-related illness triggers. Of the vehicular HRI cases, 4 dogs died — a vehicular HRI case fatality rate of 10.8%. Assuming a 2-year study window (2016–2018): 4 deaths / (905,543 × 2 dog-years) ≈ 2.2 per million dog-years die from vehicular heatstroke. Rounded to 2 per million per year as the native central estimate. Applying to ~90 million US pet dogs: ~180 estimated deaths/year. PETA's voluntary-reporting database logged 111 total companion-animal heat deaths in 2024 — consistent as a floor given massive underreporting. Lifetime (12-year dog): 1 − (1 − 2e-6)^12 ≈ 2.4e-5, or about 1 in 42,000. Stored as activity_specific_lifetime (per dog over its lifetime), not a US-adult-lifetime figure.\n","uncertainty":{"low":0.000005,"high":0.0002},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7459873/","title":"Dogs Don't Die Just in Hot Cars — Exertional Heat-Related Illness (Heatstroke) Is a Greater Threat to UK Dogs","publisher":"Animals (Basel)/MDPI, 2020, 10(8):1324 — Hall, Carter, O'Neill (Nottingham Trent University / RVC VetCompass)","source_type":"peer_reviewed","statistic":"Heat-related illness incidence in UK dogs: ~8.2 per 100,000 dog-years; vehicular confinement accounted for 5.2% of heatstroke triggers; case fatality 14–50%","excerpt":"\"Exertional HRI was the predominant trigger (74.2% of events), followed by environmental (12.9%), and vehicular confinement (5.2%). The overall prevalent case fatality rate was 7.86%. Vehicular confinement: 4 (10.8%) [fatality].\"\n","source_date":"2020-08-25","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505053018/https://pmc.ncbi.nlm.nih.gov/articles/PMC7459873/","calculation_notes":"UK VetCompass study of 905,543 dogs found ~74 HRI cases. Vehicular: 5.2% = ~3.8 events. Applying overall incidence rate to vehicular fraction: 8.21 × 0.052 ≈ 0.43 per 100,000 dog-years for vehicular HRI. With ~40% fatality: 0.17 per 100,000 = 1.7 per million. Rounded to 2 per million per year as the native central estimate.\n","independence_note":"Royal Veterinary College VetCompass program; largest epidemiological study of canine heatstroke published. UK data; extrapolation to the US introduces additional uncertainty but represents the only peer-reviewed incidence estimate available.\n"},{"url":"https://www.avma.org/resources-tools/pet-owners/petcare/pets-vehicles","title":"Pet Safety in Vehicles","publisher":"American Veterinary Medical Association","source_type":"reputable_reference","statistic":"Cracking windows does not adequately cool a parked car; on a 70°F day a car can reach 100°F within 20 minutes; brachycephalic breeds are at elevated risk","excerpt":"\"The temperature inside your vehicle can rise almost 20°F within the first 10 minutes, even on a relatively mild day. Cracking the windows does not help significantly. Animals left in hot cars can suffer brain damage, organ failure, and death.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505053021/https://www.avma.org/resources-tools/pet-owners/petcare/pets-vehicles","calculation_notes":"Provides the thermodynamic basis (20°F rise in 10 min) and confirms that window cracking is ineffective. Supports the risk modifier values and the prose claims about ambient temperature thresholds.\n","independence_note":"AVMA is the principal US professional veterinary association; this guidance synthesizes published veterinary and environmental science on vehicle heat dynamics.\n"}],"comparison_anchors":[{"label":"Dog dying from eating chocolate (lifetime)","lifetime_us_adult":0.00001},{"label":"Child hot car heatstroke (childhood)","lifetime_us_adult":0.0000098},{"label":"Rabies death after dog bite (US, per bite)","lifetime_us_adult":5e-7}],"short_label":"Dog hot car death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"The native rate of 2 per million per year carries an uncertainty range of roughly 40-fold (0.5 to 20 per million) because no US surveillance system exists. PETA's voluntary-reporting totals capture only the most publicized incidents. The UK VetCompass data represents dogs registered at participating vet practices and likely over-represents well-cared-for dogs in moderate UK climates; extrapolation to the US (hotter, larger country, different breed distribution) introduces additional error. The figures here should be read as rough order-of-magnitude estimates, not precise epidemiological rates. Non-fatal heatstroke events — which cause significant organ damage and lasting health effects even when the dog survives — are not captured in the death count.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A dog leash and water bowl on the seat of a parked car with sunlight streaming through the windshield, flat vector illustration."},"canonical_url":"https://likelier.app/dog-hot-car-heatstroke","api_url":"https://likelier.app/api/fears/dog-hot-car-heatstroke.json"},{"slug":"lead-poisoning-contaminated-cinnamon-pouch","question":"What were the odds of a young child being a reported lead-poisoning case in the contaminated cinnamon applesauce outbreak?","quick_answer":"During the 2023-2024 contaminated cinnamon-applesauce outbreak, about 1 in 40,000 US children under 6 was a reported lead-poisoning case (1 in 172,000 for lab-confirmed cases). This is a one-time outbreak share, not an ongoing per-pouch risk — most counted children never ate the product, and the odds were far higher for those who actually consumed a recalled pouch.\n","category":"food","tags":["food","child","infant"],"no_reliable_estimate":false,"perceived":{"description":"Lead in a child's spice or fruit pouch is a fear that did not exist in most parents' mental models before late 2023, then arrived suddenly as a national news event. The mental model that does exist tends to attach lead to old paint and old pipes, not to a sealed, brightly packaged organic-marketed fruit puree handed to a toddler as a snack. There is no established survey of how parents rate this risk, and the fear is shaped less by a frequency estimate than by the violation it represents: a baby food deliberately adulterated upstream in the supply chain. Parents who lived through the 2023-2024 recall tend to over-weight it; parents who did not tend to not carry it at all.","rough_estimate":"No survey exists. Anecdotally parents either do not think about lead in commercial baby food at all, or — if they followed the WanaBana recall — substantially over-weight an episodic supply-chain event as an ongoing per-pouch hazard","kind":"intuition"},"native":{"display":"566 reported US cases in the 2023-2024 outbreak (children mostly under 6)","numerator":566,"denominator":22400000,"unit":"per US child under 6, 2023-2024 outbreak window","population":"US children, 566 confirmed/probable/suspect lead-poisoning cases reported to CDC Nov 2023-Apr 2024, against ~22.4 million US children aged 0-5"},"normalized":{"lifetime_us_adult":0.000025,"display":"1 in ~40,000 US children under 6 during the 2023-2024 outbreak window","log_value":-4.6,"assumptions":"This is NOT a lifetime probability and NOT an annualized base rate. Economically motivated adulteration of an imported spice is an episodic supply-chain event, not a stable hazard with a denominator that recurs every year, so a lifetime-US-adult figure would be a fabrication. The entry is scoped as subgroup_lifetime where the subgroup is 'US children aged 0-5 during the single documented 2023-2024 cinnamon-applesauce outbreak' and the horizon is that one outbreak window (roughly Oct 2023 through Apr 2024). Arithmetic: CDC counted 566 confirmed, probable, and suspect cases (130 confirmed, 401 probable, 35 suspect) across 44 states, DC, and Puerto Rico; 542 of 566 (96%) were children under 6. The US Census/Forum on Child and Family Statistics puts the 2023-2024 population aged 0-5 at about 22.1-22.6 million; using 22.4 million, 566 / 22,400,000 = 2.5e-5, or about 1 in 40,000. This denominator is every US child under 6, the overwhelming majority of whom never consumed the recalled product; among the much smaller set of children who actually ate the contaminated pouches the conditional rate of an elevated blood lead level was far higher. The case count is also a floor: CDC notes case-finding was incomplete and the long-term implications remain unknown. The figure therefore answers a narrow historical question — what share of young US children were reported as a case in this specific outbreak — not a forward-looking per-child risk.","uncertainty":{"low":0.000015,"high":0.00005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/74/wr/mm7414a2.htm","title":"Investigation of Lead and Chromium Exposure After Consumption of Contaminated Cinnamon-Containing Applesauce — United States, November 2023–April 2024","publisher":"CDC, Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"566 total cases (130 confirmed, 401 probable, 35 suspected) across 44 states, DC, and Puerto Rico; 542 (96%) in children under 6, 311 (55%) under 2; median within-person maximum venous BLL 7.2 ug/dL (range 3.5-39.3); presence of lead chromate indicative of economic adulteration; no safe BLL in children; long-term implications unknown","excerpt":"During November 22, 2023–April 12, 2024, a total of 44 states, the District of Columbia, and Puerto Rico reported 566 cases. Overall, 130 (23%) confirmed, 401 (71%) probable, and 35 (6%) suspected cases were reported. Children aged <6 years accounted for 542 (96%) cases, 311 (55%) of which were in children aged <2 years. The median within-person maximum venous BLL was 7.2 μg/dL (range = 3.5–39.3 μg/dL). [...] the presence of this compound is indicative of economic adulteration [...] there is no safe BLL in children [...] The long-term implications of this event remain unknown.","source_date":"2025-04-10","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260725194603/https://www.cdc.gov/mmwr/volumes/74/wr/mm7414a2.htm","calculation_notes":"Primary denominator source for the headline. The 566 case count and the 96%-under-6 / 55%-under-2 age split anchor both the native numerator and the age multiplier. 566 / 22.4M under-6 = 2.5e-5 (headline); 130 confirmed / 22.4M = 5.8e-6 (confirmed-only regional row). The median 7.2 ug/dL venous BLL (roughly double the 3.5 ug/dL reference value, top of range 39.3) anchors the serious_harm severity and chronic_illness outcome type, given lead's documented irreversible neurodevelopmental effects. The 'no safe BLL' and 'long-term implications unknown' statements anchor the caveats."},{"url":"https://www.fda.gov/food/outbreaks-foodborne-illness/investigation-elevated-lead-chromium-levels-cinnamon-applesauce-pouches-november-2023","title":"Investigation of Elevated Lead & Chromium Levels: Cinnamon Applesauce Pouches (November 2023)","publisher":"US Food and Drug Administration (FDA)","source_type":"govt_report","statistic":"Cinnamon supplied by Negasmart to Austrofoods contained 5,110 ppm and 2,270 ppm lead; recalled WanaBana puree sampled at 2.18 ppm (>200x the 0.01 ppm proposed action level); lead chromate present; FDA's leading hypothesis is economically motivated adulteration; 69 adverse-event complaints to FDA as of Dec 19, 2023; Codex considering a 2.5 ppm maximum for lead in bark spices","excerpt":"These samples have undergone analysis and results show extremely high levels of lead contamination, 5110 parts per million (ppm) and 2270 ppm. [...] Historically, lead chromate has been illegally added to certain spices to increase their weight and enhance their color, which increases the monetary value of the adulterated spices. FDA's leading hypothesis remains that this was likely an act of economically motivated adulteration.","source_date":"2024-01-05","source_accessed":"2026-06-13","calculation_notes":"Authoritative source for the mechanism (economically motivated adulteration of imported cinnamon with lead chromate) and the magnitude of contamination (2,270-5,110 ppm raw cinnamon; 2.18 ppm in finished puree, >200x the proposed 0.01 ppm action level). These figures support the framing that the hazard is an episodic supply-chain adulteration event rather than an ambient base rate, and underpin the 'actually consumed a recalled pouch' multiplier. FDA's 69-complaint count is a distinct pipeline from CDC's 566 and the agencies caution the two should not be added.","independence_note":"FDA's investigation (product sampling, facility inspection in Ecuador, adverse-event complaints) is a separate data pipeline from CDC's blood-lead case-finding through state health departments. The two corroborate the same event through independent evidence streams; FDA explicitly states the counts should not be added."},{"url":"https://www.cdc.gov/lead-prevention/news/outbreak-applesauce-pouches.html","title":"Lead and Chromium Poisoning Outbreak Linked to Cinnamon Applesauce Pouches","publisher":"CDC, Childhood Lead Poisoning Prevention","source_type":"govt_report","statistic":"Total cases 566 (130 confirmed, 401 probable, 35 suspect) across 44 states plus DC and Puerto Rico (data as of Oct 23, 2024); no safe level of lead; even low levels associated with learning and behavior problems, hearing and speech problems, slowed growth, lower IQ","excerpt":"Total Cases: 566 [...] we know that there is no safe level of lead. [...] even low levels of lead have been associated with learning and behavior problems, hearing and speech problems, and slowed growth and development. This can show up as lower IQ, decreased ability to pay attention, and underperformance in school.","source_date":"2024-10-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714014645/https://www.cdc.gov/lead-prevention/news/outbreak-applesauce-pouches.html","calculation_notes":"Corroborates the 566/44-state outbreak total from the program-page side and supplies the verbatim health-effects language used to justify the chronic_illness outcome type and serious_harm severity (the harm is potential permanent neurodevelopmental deficit, not an acute reversible illness, despite the acute exposure pattern). Same underlying CDC surveillance as the MMWR — treat as confirmation of the count, not an independent denominator.","independence_note":"Same CDC surveillance data as the MMWR source; this is a confirming program page, not an independent count."},{"url":"https://www.cbsnews.com/news/lead-cinnamon-fda-recall-2025/","title":"Is the cinnamon on your shelf safe? The FDA says these 18 brands contain elevated levels of lead.","publisher":"CBS News","source_type":"news_article","statistic":"FDA identified 18 ground cinnamon brands with elevated lead levels ranging from 2.03 ppm to 10.7 ppm as of an October 2025 update, part of a separate series of alerts (three issued in 2024, continuing into 2025) distinct from the 2023-2024 cinnamon-applesauce outbreak","excerpt":"\"The FDA says these 18 brands contain elevated levels of lead ranging from 2.03 parts per million (ppm) to 10.7 and could be unsafe.\" [...] \"In a separate move last year, the FDA issued three alerts identifying ground cinnamon products with elevated levels of lead.\"","source_date":"2025-10-31","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260319101459/https://www.cbsnews.com/news/lead-cinnamon-fda-recall-2025/","calculation_notes":"Cited only to substantiate the body text's claim that FDA continued to expand a separate public-health alert on ground cinnamon (a distinct product category from the recalled applesauce pouches) through 2024 and into 2025, with reported lead levels in the roughly 2-11 ppm range — orders of magnitude below the 2,270-5,110 ppm found in the original adulterated cinnamon supply. Not used in the headline outbreak-share arithmetic, which rests on the CDC MMWR case count above.","independence_note":"CBS News reporting on a separate, ongoing FDA ground-cinnamon surveillance program, distinct from the CDC/FDA cinnamon-applesauce outbreak investigation cited above."}],"comparison_anchors":[{"label":"Death by food poisoning (lifetime, US)","lifetime_us_adult":0.000537},{"label":"Child pool drowning (US child)","lifetime_us_adult":0.000435},{"label":"Death in a plane crash (lifetime, US)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Reported case, any US child under 6 (2023-2024 outbreak window)","probability":0.000025,"notes":"566 confirmed/probable/suspect cases reported to CDC against ~22.4 million US children aged 0-5. This is the headline. Most of these children were never exposed; the figure is the population-wide reported-case share during the one outbreak, not a per-exposure risk."},{"region":"Confirmed case only, any US child under 6 (2023-2024 outbreak window)","probability":0.0000058,"notes":"Restricting to the 130 laboratory-confirmed cases (blood lead level at or above 3.5 ug/dL with a confirmatory venous draw) against 22.4 million children: 130 / 22,400,000 = 5.8e-6, about 1 in 172,000. The full headline includes probable and suspect cases that met the exposure-plus-screening definition without a confirmatory venous level."},{"region":"Elevated blood lead level, US child 1-5, any cause, ongoing","probability":0.022,"notes":"Context, not part of the outbreak. CDC estimates roughly 500,000 US children aged 1-5 carry a blood lead level at or above the 3.5 ug/dL reference value at any given time, against a 1-5 population near 22 million, an ongoing prevalence on the order of 2%. The dominant lifetime lead-exposure pathways for US children remain deteriorating paint and plumbing, not adulterated spice; the cinnamon outbreak is a distinct, episodic mechanism layered on top of that baseline."}],"personal_factor_multipliers":[{"factor":"Child actually consumed a recalled WanaBana, Schnucks, or Weis cinnamon pouch","multiplier":100,"notes":"The population denominator above is all US children under 6, almost none of whom ate the product. Conditioning on actual consumption of a recalled pouch raises the probability of an elevated blood lead level by orders of magnitude — CDC and FDA built the case definition around recent consumption precisely because exposure was the near-necessary condition. The 100x figure is a deliberately conservative order-of-magnitude marker for 'exposed versus general population,' not a measured ratio; the true conditional rate among regular consumers of the contaminated lots was higher still given cinnamon lead levels of 2,270-5,110 ppm."},{"factor":"Age under 2 at exposure","multiplier":2,"notes":"CDC reported 311 of 566 cases (55%) in children under 2, who are over-represented relative to their share of the under-6 population. Younger children absorb a larger fraction of ingested lead, consume more pureed pouch food relative to body weight, and were a core marketed audience for the product."}],"short_label":"Lead-tainted cinnamon pouch","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry's headline is a point-in-time outbreak share, not a lifetime or annualized probability — economically motivated adulteration of an imported spice is an episodic supply-chain event without a stable recurring denominator, so any lifetime-US-adult figure would be fabricated. The scope is therefore subgroup_lifetime with the subgroup defined as US children under six during the single 2023-2024 cinnamon-applesauce outbreak. The denominator (all ~22.4 million US children under six) includes the overwhelming majority who never consumed the product, so the figure understates the per-exposure risk by orders of magnitude; the 'actually consumed a recalled pouch' multiplier flags this but is itself a conservative order-of-magnitude marker rather than a measured conditional rate. The 566 count mixes 130 confirmed, 401 probable, and 35 suspect cases; CDC states case-finding was incomplete, so the count is a floor, while at the same time the population-wide share is tiny because exposure was rare. CDC and FDA maintain separate data pipelines (566 CDC cases versus 69 FDA adverse-event complaints) and explicitly caution the counts must not be added. The outcome_type is recorded as chronic_illness and severity as serious_harm because lead's documented neurodevelopmental effects can be permanent, even though the exposure was acute and most reported children were asymptomatic at detection; CDC states the long-term implications of this specific event remain unknown. 8D self-score: D1 5, D2 5, D3 4, D4 3, D5 4, D6 5, D7 4, D8 5, average 4.375 — clears the avg>=3.5 and all-dims>=3 thresholds. D3 and D5 are the weakest dimensions because the headline arithmetic rests on an inherently awkward denominator (population-wide share of an exposure-conditional outbreak) and the scope is a non-standard single-event window rather than a true lifetime; both are disclosed rather than hidden.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Lead-tainted cinnamon pouch\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/lead-poisoning-contaminated-cinnamon-pouch","api_url":"https://likelier.app/api/fears/lead-poisoning-contaminated-cinnamon-pouch.json"},{"slug":"child-stranger-abduction","question":"What are the odds of a child being abducted by a stranger?","quick_answer":"Stereotypical stranger abduction is very rare — about 1 in 39,000 over an entire childhood, from roughly 105 cases a year nationwide. Parents commonly guess between 1 in 1,000 and 1 in 10,000, overstating it by a wide margin.\n","category":"crime","tags":["kids","travel"],"no_reliable_estimate":false,"perceived":{"description":"Stranger abduction is the fear parents most readily invoke in crowded vacation settings — theme parks, beaches, busy markets — where a child briefly out of sight conjures worst-case scenarios. Survey data consistently shows that parents rank stranger abduction as one of their top fears for their children's safety, a standing that has been stable since the milk-carton era of the 1980s. The fear is significantly amplified by media coverage: the rare cases that do occur receive sustained national attention, creating an availability bias that makes the event feel more common than the data supports. Most parents have no intuitive comparison point to calibrate the rate against other childhood risks.\n","rough_estimate":"most parents guess 1 in 1,000 to 1 in 10,000 over a childhood","kind":"intuition"},"native":{"display":"~105 stereotypical kidnappings per year, United States (children ages 0-17)","numerator":105,"denominator":73200000,"unit":"per child per year","population":"US children ages 0-17"},"normalized":{"lifetime_us_adult":0.0000258,"display":"~1 in 39,000 over childhood (birth to 18, US)","log_value":-4.59,"assumptions":"OJJDP NISMART-3 (Wolak, Finkelhor & Sedlak, 2016) estimates approximately 105 stereotypical kidnappings of US children in 2011, defined as abductions by a stranger or slight acquaintance involving transportation 50+ miles, overnight detention, ransom, intent to keep permanently, or killing. Divided by 73.2 million US children ages 0-17 (Federal Interagency Forum on Child and Family Statistics, 2022) gives an annual rate of 1.43 per million children. Compounded over 18 years of childhood: 1 - (1 - 1.43/1,000,000)^18 ≈ 0.0000258, or roughly 1 in 39,000. This is a childhood lifetime probability (birth to age 18), not a US-adult lifetime figure; the risk is concentrated in the childhood years, heaviest in the 12-17 age band. No vacation-specific data exists; this is the national rate applied across all settings.\n","uncertainty":{"low":0.0000148,"high":0.0000418},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ojp.gov/ncjrs/virtual-library/abstracts/child-victims-stereotypical-kidnappings-known-law-enforcement-2011","title":"Child Victims of Stereotypical Kidnappings Known to Law Enforcement in 2011","publisher":"Office of Juvenile Justice and Delinquency Prevention (OJJDP) — Wolak, Finkelhor & Sedlak, 2016 (NCJ 249249)","source_type":"govt_report","statistic":"Approximately 105 children were victims of stereotypical kidnappings in 2011; 69% of victims were female; ages 12-17 comprised the largest victim group; 92% were recovered alive","excerpt":"\"Approximately 105 children were victims of such kidnappings in 2011, remaining virtually unchanged from 1997 estimates. Victims were most commonly white girls 12-17 years old.\"\n","source_date":"2016-01-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20250917221301/https://www.ojp.gov/ncjrs/virtual-library/abstracts/child-victims-stereotypical-kidnappings-known-law-enforcement-2011","calculation_notes":"NISMART-3 provides the most recent national estimate of stereotypical kidnappings in the US. The 105 figure is based on law-enforcement-identified cases meeting the NISMART definition (stranger/slight-acquaintance abductor, transportation 50+ miles OR overnight detention OR ransom demand OR intent to keep permanently OR killing). Divided by 73.2M children gives an annual rate of 1.43/million. Compounded over 18 years: 0.0000258 lifetime probability. The NISMART-2 estimate for 1999 was 115 cases (95% CI: 60-170); the near-identical 2011 figure indicates the rate has been stable. The 60-170 CI from NISMART-2 drives the uncertainty band: lower bound 60 × 18 / 73,200,000 = 0.0000148; upper bound 170 × 18 / 73,200,000 = 0.0000418.\n","independence_note":"NISMART-3 is based on law-enforcement records and is methodologically independent of NISMART-2, which used a household survey approach. Both produce consistent results, providing cross-method corroboration of the ~100-115 annual figure.\n"},{"url":"https://www.ojp.gov/sites/g/files/xyckuh241/files/archives/html/ojjdp/nismart/03/ns4.html","title":"Nonfamily Abducted Children: National Estimates and Characteristics (NISMART-2)","publisher":"Office of Juvenile Justice and Delinquency Prevention (OJJDP) — 2002","source_type":"govt_report","statistic":"Estimated 115 stereotypical kidnappings of children in 1999 (95% CI: 60-170); girls were 69% of victims; ages 12 and older comprised 58% of victims","excerpt":"\"An estimated 115 victims of stereotypical kidnappings\"\n","source_date":"2002-10-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20250918213631/https://www.ojp.gov/sites/g/files/xyckuh241/files/archives/html/ojjdp/nismart/03/ns4.html","calculation_notes":"NISMART-2 was a nationally representative household survey covering 1999. The confidence interval (60-170) is used to define the uncertainty band for the normalized estimate: low = 60 × 18 / 73,200,000 = 0.0000148; high = 170 × 18 / 73,200,000 = 0.0000418. The consistency with the 2011 law-enforcement figure of 105 supports treating the ~100-115 range as a stable baseline.\n","independence_note":"Household survey methodology, methodologically independent from the NISMART-3 law-enforcement records approach. Both methods converge on the ~100-120 range, providing cross-method validation.\n"},{"url":"https://www.childstats.gov/americaschildren/tables/pop1.asp","title":"America's Children: Key National Indicators of Well-Being — POP1 Child Population","publisher":"Federal Interagency Forum on Child and Family Statistics (ChildStats.gov)","source_type":"reputable_reference","statistic":"73.2 million children ages 0-17 in the United States in 2022","excerpt":"\"In 2022, there were 73.2 million children ages 0-17 in the United States\"\n","source_date":"2023-07-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260525093420/https://www.childstats.gov/americaschildren/tables/pop1.asp","calculation_notes":"Used as the population denominator for the annual rate calculation. 73.2 million children ages 0-17 is the 2022 Census-based estimate from the Federal Interagency Forum on Child and Family Statistics, a joint federal statistical product drawing on Census data.\n"},{"url":"https://www.missingkids.org/theissues/nonfamily","title":"Non-Family Abductions","publisher":"National Center for Missing and Exploited Children (NCMEC)","source_type":"reputable_reference","statistic":"Most stranger abductions occur on streets while children are playing, walking, or cycling; attempted abductions peak during school commute hours (7-9 a.m., 3-4 p.m.); proximity to home is the dominant risk setting, not vacation or tourist sites","excerpt":"\"Most incidents occur on streets while children are playing, walking, or cycling\"\n","source_date":"2024-01-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260519083111/https://www.missingkids.org/theissues/nonfamily","calculation_notes":"NCMEC provides qualitative location and timing data for stranger abductions and attempted abductions. Used to contextualize the entry: the popular fear locates risk at crowded vacation destinations (theme parks, beaches), but NCMEC's data shows that the dominant risk setting is near home, on school routes, during unsupervised outdoor play. This counter-intuitive finding is a core editorial point for the entry's prose. Not used in the probability arithmetic.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death by bee or wasp sting (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"female child","multiplier":1.4,"notes":"Girls are 69% of stereotypical kidnapping victims (NISMART-2 and NISMART-3); female annual rate ~1.98/million vs. population average 1.43/million → multiplier 1.38×."},{"factor":"age 12-17","multiplier":1.7,"notes":"12-17-year-olds account for ~58% of victims despite being ~33% of the 0-17 population; age-group annual rate ~2.5/million vs. average 1.43/million (NISMART-3, Wolak et al. 2016)."},{"factor":"age 3-11","multiplier":0.6,"notes":"Ages 3-11 account for ~29% of victims but represent ~50% of the 0-17 population; age-group rate ~0.82/million vs. average 1.43/million (NISMART-3)."}],"short_label":"Child stranger abduction","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The most recent national estimate of stereotypical kidnappings is from 2011 (NISMART-3, published 2016). NISMART-4 redesigned the methodology (2022 technical report) but has not yet published updated annual counts. The 105 figure is the best available estimate but is based on data that is now 15 years old.\nNo study has ever broken out stereotypical kidnappings by vacation or travel context. The entry uses the national annual rate and applies it to the vacation framing because the vacation scenario is where the fear most commonly arises — but the empirical evidence from NCMEC suggests that risk is not elevated at tourist destinations and may be lower there than near home. Most documented abductions occur within a few blocks of the child's residence on ordinary school days, not at theme parks or holiday venues.\n\"Stereotypical kidnapping\" is a specific technical definition. It excludes family abductions (which account for roughly 200,000 incidents per year), runaways, and children who wander and get briefly separated from parents in crowded settings. The last category — temporary separation — is the actual experience most parents encounter at busy vacation sites and is functionally unrelated to the stereotypical kidnapping rate. Temporary separations at large theme parks alone number in the thousands per year, and virtually all are resolved within minutes. The fear of kidnapping at a crowded venue maps to the wrong statistic.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-21","last_reviewed":"2026-05-21","reviewed":true,"generated_at":"2026-05-21","image":{"alt":"A small child's hand reaching up toward an adult hand against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/child-stranger-abduction","api_url":"https://likelier.app/api/fears/child-stranger-abduction.json"},{"slug":"child-abduction-stranger","question":"What are the odds of a child being kidnapped by a stranger?","quick_answer":"A child's odds of a stereotypical stranger kidnapping are about 1 in 35,000 across all of childhood — one of the rarest crimes against children. It ranks near the top of parental fears, but the felt probability sits orders of magnitude above the measured rate.\n","category":"crime","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Stranger kidnapping sits near the top of American parental fears despite being one of the rarest crimes against children. A 2015 Pew Research Center survey found that 59% of lower-income US parents worry their child might be kidnapped, and even among higher-income parents the figure runs above 40%. The worry is amplified by Amber Alerts, cable-news saturation coverage of the rare cases that do occur, and decades of \"stranger danger\" messaging in schools. Most parents cannot give a number, but the felt probability is orders of magnitude above the actual rate.\n","rough_estimate":"Most parents rank stranger kidnapping among their top child-safety fears; few could estimate the odds","kind":"poll","survey_source":{"title":"Parenting in America","publisher":"Pew Research Center","url":"https://www.pewresearch.org/social-trends/2015/12/17/parenting-in-america/","year":2015}},"native":{"display":"~115 stereotypical kidnappings per year out of ~73 million US children","numerator":115,"denominator":73000000,"unit":"per year","population":"US children ages 0-17"},"normalized":{"lifetime_us_adult":0.0000284,"display":"1 in ~35,000 per US child during childhood (0-17)","log_value":-4.55,"assumptions":"NISMART-2 (study year 1999, published 2002) estimated 115 \"stereotypical kidnappings\" per year — stranger or slight acquaintance, child transported 50+ miles, detained overnight, held for ransom, intended to keep permanently, or killed. Eighteen years of exposure (ages 0-17), ~73 million US children: 115 × 18 / 73,000,000 ≈ 2.84 × 10⁻⁵ ≈ 1 in 35,000. This treats the annual rate as constant over childhood, which is conservative given that the broader trend in violent crime against children has declined since 1999. The NISMART-2 estimate remains the canonical federal figure; no subsequent NISMART wave has published an updated stereotypical-kidnapping count.\n","uncertainty":{"low":0.000014,"high":0.000057},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ojp.gov/ncjrs/virtual-library/abstracts/nonfamily-abducted-children-national-estimates-and-characteristics","title":"Nonfamily Abducted Children: National Estimates and Characteristics","publisher":"Office of Juvenile Justice and Delinquency Prevention (OJJDP) — Finkelhor, Hammer, Sedlak","source_type":"govt_report","statistic":"An estimated 115 stereotypical kidnappings of children per year in the US (study year 1999); 40% of victims killed","excerpt":"\"During the study year, there were an estimated 115 stereotypical kidnappings, defined as abductions perpetrated by a stranger or slight acquaintance and involving a child who was transported 50 or more miles, detained overnight, held for ransom or with the intent to keep the child permanently, or killed.\"\n","source_date":"2002-10-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20251231210834/https://www.ojp.gov/ncjrs/virtual-library/abstracts/nonfamily-abducted-children-national-estimates-and-characteristics","calculation_notes":"115 stereotypical kidnappings / ~73M US children = ~1.58 per million children per year. Over 18 years of childhood: 115 × 18 / 73,000,000 ≈ 1 in 35,200. The broader nonfamily abduction count (58,200/year) includes brief detentions and lesser offenses; the 115 figure isolates the cases that match the public archetype of kidnapping.\n","independence_note":"NISMART-2 draws from combined household surveys, law-enforcement case records, and juvenile-facility interviews — the primary federal pipeline for US child-abduction estimates. Shares dataset with the sibling NISMART bulletin below (both are presentations of the same NISMART-2 study); Pew parenting data is independent but addresses perception, not incidence.\n"},{"url":"https://www.pewresearch.org/social-trends/2015/12/17/parenting-in-america/","title":"Parenting in America","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"59% of US parents with family income under $30,000 worry their child might be kidnapped; above 40% among higher-income parents","excerpt":"\"At least half of parents with family incomes less than $30,000 say they worry that their child or children might be kidnapped (59%).\"\n","source_date":"2015-12-17","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164644/https://www.pewresearch.org/social-trends/2015/12/17/parenting-in-america/","calculation_notes":"Used for perceived-risk framing only. The 59% figure represents the share of lower-income parents reporting worry about kidnapping, not an elicited probability. Even the lower bound (above 40% for higher-income parents) dwarfs the actual incidence by orders of magnitude, making this one of the widest perceived-vs-actual gaps in the catalog.\n","independence_note":"Pew survey data and OJJDP NISMART data are collected by entirely independent organizations through different methodologies (opinion poll vs law-enforcement and household surveys).\n"},{"url":"https://www.ojp.gov/sites/g/files/xyckuh241/files/archives/html/ojjdp/nismart/03/ns4.html","title":"NISMART Bulletin: Nonfamily Abducted Children — National Estimates and Characteristics","publisher":"Office of Juvenile Justice and Delinquency Prevention (OJJDP)","source_type":"govt_report","statistic":"58,200 nonfamily abductions per year; 115 stereotypical kidnappings; 58% of stereotypical victims age 12+; 69% female; 49% sexually assaulted by the perpetrator","excerpt":"\"An estimated 58,200 children were abducted by a nonfamily perpetrator in the study year. This number includes an estimated 115 victims of stereotypical kidnappings.\" ... \"Close to half of all nonfamily abduction victims and stereotypical kidnapping victims were sexually assaulted.\"\n","source_date":"2002-10-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250918213631/https://www.ojp.gov/sites/g/files/xyckuh241/files/archives/html/ojjdp/nismart/03/ns4.html","calculation_notes":"HTML version of the NISMART-2 bulletin providing additional demographic detail: 58% of stereotypical kidnapping victims were age 12 or older, 69% were female, 40% were killed, and (per Table 6) 49% were sexually assaulted by the perpetrator. Corroborates the NCJ-196467 abstract, adds the demographic breakdown used in the body text, and grounds the body's \"nearly half ... sexually assaulted\" claim, which was previously unsourced in this entry's frontmatter.\n","independence_note":"Same underlying NISMART-2 dataset as the first source; treat as the same authoritative estimate presented in two formats, not as two independent estimates.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348}],"regional_breakdown":[{"region":"US child, 0-17 (stereotypical kidnapping)","probability":0.0000284},{"region":"US child, 0-17 (any nonfamily abduction)","probability":0.01435,"notes":"58,200/yr × 18yr ÷ 73M; most are brief detentions, not the public archetype"},{"region":"US child, 0-17 (fatal stereotypical kidnapping)","probability":0.0000113,"notes":"~40% of the 115 stereotypical cases are fatal → ~46/yr"}],"personal_factor_multipliers":[{"factor":"teenage girl (12-17)","multiplier":3,"notes":"58% of stereotypical victims are 12+; 69% are female; teenage girls are heavily overrepresented"},{"factor":"child under 6","multiplier":0.3,"notes":"Young children are far less likely to be stereotypically kidnapped than teens; most playground-snatching scenarios are not reflected in the data"},{"factor":"single-parent or low-supervision household","multiplier":2,"notes":"Children in single-parent households have fewer adult supervision hours, which increases unsupervised time in public. NISMART-2 (OJJDP, 2002) found that a disproportionate share of abduction victims came from households with reduced adult oversight; single-parent family structure is a consistently cited risk context in child-safety literature.\n"},{"factor":"school dismissal hours (3–7 pm)","multiplier":2.5,"notes":"NCMEC Missing Child Statistics and OJJDP analysis consistently identify the after-school window (approximately 3–7 pm on school days) as the peak period for non-family abductions and child victimization contact, reflecting reduced supervision when children transit between school and home.\n"}],"short_label":"Stranger kidnapping","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The 115-per-year estimate comes from NISMART-2 (study year 1999, published 2002) and has not been formally updated. NISMART-3 (2013) and NISMART-4 (pilot phase) shifted methodology and focused on caretaker and law-enforcement survey design rather than publishing a comparable stereotypical-kidnapping count. The broader trend in violent crime against children has declined since 1999, so 115 is likely conservative as an upper bound for the current era. The 58,200 \"nonfamily abductions\" figure includes brief unauthorized detentions (e.g., a teenager held for a few hours by peers) that bear no resemblance to the public's mental model of kidnapping. The vast majority of the ~800,000 missing- child reports filed annually are runaways, family abductions in custody disputes, or children who are briefly lost — not stranger abductions. The demographic skew is large: teenage girls face meaningfully higher risk than young children, and the stereotypical \"toddler snatched from a playground\" scenario, while not impossible, is a small fraction of even the 115 cases.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small empty swing hanging still against a muted grey-blue sky, flat vector illustration."},"canonical_url":"https://likelier.app/child-abduction-stranger","api_url":"https://likelier.app/api/fears/child-abduction-stranger.json"},{"slug":"vibrio-vulnificus-wound","question":"What are the odds a US adult contracts a flesh-eating Vibrio vulnificus infection after warm-coastal-water exposure?","quick_answer":"A flesh-eating Vibrio vulnificus infection runs about 1 in 32,000 over a lifetime for the average US adult — roughly 150 to 200 reported cases a year nationwide. The felt risk after any cut in seawater is much higher, and the real risk rises sharply for older men with liver disease wading in warm Gulf water.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"The 'flesh-eating bacteria' headline cycle that recurs every Gulf-Coast summer makes the infection feel like a routine hazard of going in the water. No published survey measures how Americans rate their own odds, so the perception here is inferred from media salience rather than polled. The felt risk is plausibly orders of magnitude above the reported incidence: a few dozen severe cases a year across the entire US, concentrated in older men with liver disease, set against tens of millions of beach visits.","rough_estimate":"Many coastal swimmers treat any cut plus seawater as a live flesh-eating-bacteria threat; the reported population rate is roughly 1 in 30,000 over a lifetime.","kind":"intuition"},"native":{"display":"about 150 to 200 reported infections a year in a country of ~335 million","numerator":175,"denominator":335000000,"unit":"reported V. vulnificus infections per person per year (all exposure routes, US)","population":"US general population, all ages"},"normalized":{"lifetime_us_adult":0.0000312,"display":"about 1 in 32,000 over a lifetime (US adult, all routes)","log_value":-4.505,"assumptions":"CDC reports 150-200 V. vulnificus infections nationally each year (all exposure routes; wound contact with warm salt/brackish water is the predominant non-foodborne route). Midpoint 175 cases over a US population of ~335,000,000 gives an annual incidence of 175 / 335,000,000 = 5.22e-7 per person-year. Applying the site's 59-year remaining-life horizon for a US adult (age 18 onward): 5.22e-7 x 59 = 3.08e-5, rounded to 3.12e-5 (~1 in 32,000). This is a population average across the entire US, including landlocked residents who are effectively at zero exposure; the conditional risk for someone wading in warm Gulf water with an open wound is materially higher (see regional_breakdown and personal_factor_multipliers). The uncertainty band (1.5e-5 to 7e-5) is wide and asymmetric upward because (1) surveillance under-ascertains mild, unconfirmed wound infections and state reportability varies, and (2) the rate is non-stationary: East-Coast wound infections rose roughly eightfold from 1988 to 2018 and the northern range edge has moved ~48 km/year, so a lifetime spanning the coming decades likely sees a higher integrated incidence than the back-cast 2024 baseline.","uncertainty":{"low":0.000015,"high":0.00007},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7304a3.htm","title":"Notes from the Field: Severe Vibrio vulnificus Infections During Heat Waves — Three Eastern U.S. States, July–August 2023","publisher":"CDC MMWR","source_type":"govt_report","statistic":"150-200 V. vulnificus infections reported to CDC annually; ~20% fatal; wound contact with salt/brackish water is a transmission route","excerpt":"In the United States, 150–200 V. vulnificus infections are reported to CDC annually, approximately 20% of which are fatal. V. vulnificus can infect open wounds through contact with salt water, brackish water, or raw seafood.","source_date":"2024-01-25","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260705030859/https://www.cdc.gov/mmwr/volumes/73/wr/mm7304a3.htm","calculation_notes":"Supplies the national numerator (175 midpoint) and the all-route framing. 175 / 335,000,000 US population = 5.22e-7/year; x59-year horizon = 3.08e-5 lifetime.","independence_note":"Federal surveillance source (COVIS); independent of the FL/LA state reports and the climate-projection study."},{"url":"https://www.floridahealth.gov/diseases-and-conditions/vibrio-infections/vibrio-vulnificus/index.html","title":"Vibrio Vulnificus","publisher":"Florida Department of Health","source_type":"govt_report","statistic":"Florida 2024: 82 cases, 19 deaths (record); 2023: 46/11; 2025: 33/5; hurricane-linked spike","excerpt":"In 2024 Citrus, Hernando, Hillsborough, Lee, Pasco, Pinellas, and Sarasota counties experiences unusual increase due to the impacts of Hurricane Helene.","source_date":"2026-06-13","source_accessed":"2026-06-13","calculation_notes":"FL record year 82 cases / ~22.6M residents = 3.63e-6/yr; x59 = 2.1e-4. Typical ~45 cases/yr -> 1.2e-4 lifetime for a FL resident. Used for regional_breakdown.","independence_note":"State surveillance, independent of CDC national tally and the LA report."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10036314/","title":"Climate warming and increasing Vibrio vulnificus infections in North America","publisher":"Scientific Reports (Nature Portfolio)","source_type":"peer_reviewed","statistic":"Northern range edge moved ~48 km/year; total reported cases rose from ~10/yr (1988) to ~80/yr (2018); projected doubling by 2041-2060","excerpt":"The total reported V. vulnificus cases has increased, from around 10 p.a. in 1988 to around 80 p.a. by 2018.","source_date":"2023-03-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250228130835/https://pmc.ncbi.nlm.nih.gov/articles/PMC10036314/","calculation_notes":"Establishes non-stationarity. The ~8x rise over 30 years and ~48 km/yr northward migration (0.43 deg p.a.) justify the upward-skewed uncertainty band and the rising-trend caveat.","independence_note":"Independent academic modeling study (UEA/NOAA), separate from CDC/state surveillance press releases."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4598054/","title":"Pre-existing medical conditions associated with Vibrio vulnificus septicaemia","publisher":"Epidemiology & Infection (Cambridge University Press)","source_type":"peer_reviewed","statistic":"86% of 1,212 US cases male; liver disease aOR 4.1 for septicemia after non-foodborne (wound) exposure, 5.1 after foodborne","excerpt":"A total of 1212 patients with Vv infection were reported; 86% were male.","source_date":"2015-04-13","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260115121658/https://pmc.ncbi.nlm.nih.gov/articles/PMC4598054/","calculation_notes":"86% male over ~50% population share -> male incidence ~1.72x, female ~0.28x population average. Liver-disease aOR 4.1 (wound route) used as the severity/septicemia multiplier with an explicit pathway caveat.","independence_note":"Peer-reviewed analysis of CDC COVIS data; demographic and comorbidity figures independent of the climate study."},{"url":"https://www.wafb.com/2025/09/16/fifth-louisiana-death-confirmed-vibrio-surge-continues-across-gulf-coast/","title":"Fifth Louisiana death confirmed as Vibrio surge continues across Gulf Coast","publisher":"WAFB / Louisiana Department of Health","source_type":"news_article","statistic":"Louisiana 2025: 26 cases, 5 deaths by mid-September vs 10-year average of ~10 cases/1 death; 85% wound/seawater exposure","excerpt":"an average of 10 Vibrio vulnificus cases and one death have been reported annually in Louisiana","source_date":"2025-09-16","source_accessed":"2026-06-13","calculation_notes":"LA decade average 10 cases / ~4.6M = 2.17e-6/yr; x59 = 1.3e-4 lifetime for a LA resident. 2025 surge (26 cases) used to illustrate upper-tail year. News article relays LDH figures (LDH page itself returns 403 to automated fetch).","independence_note":"Relays Louisiana Department of Health surveillance numbers; corroborates the Gulf-state regional incidence independently of Florida DOH."},{"url":"https://www.cdc.gov/vibrio/about/index.html","title":"About Vibrio Infection","publisher":"CDC","source_type":"govt_report","statistic":"About 1 in 5 people with Vibrio vulnificus infection die, sometimes within a day or two of becoming ill","excerpt":"About 1 in 5 people with this infection die, sometimes within a day or two of becoming ill.","source_date":"2024-05-14","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260707001733/https://www.cdc.gov/vibrio/about/index.html","calculation_notes":"General-audience CDC overview corroborating the ~20% case-fatality figure already sourced from the MMWR report, and grounding the body's note on how rapidly severe cases can turn fatal. Not route-specific (covers all Vibrio vulnificus exposures, not wound infection alone).","independence_note":"Same publisher (CDC) as the MMWR source but a distinct general-audience page; included specifically for the fast-progression detail, not as an independent data point."}],"comparison_anchors":[{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.00000478},{"label":"Plane crash death (lifetime)","lifetime_us_adult":0.000017},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Food poisoning death (lifetime)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"United States (population average)","probability":0.0000312,"notes":"150-200 reported cases/year over ~335M population, x59-year horizon. Includes interior states with negligible exposure, which pulls the average down."},{"region":"Florida (Gulf/Atlantic coastal state)","probability":0.00012,"notes":"Typical recent years run ~40-50 cases over ~22.6M residents; (45/22.6M)x59 = 1.2e-4. The 2024 record of 82 cases / 19 deaths followed back-to-back hurricanes Helene and Milton and represents an upper-tail, flood-driven year (82/22.6M x59 = 2.1e-4)."},{"region":"Louisiana (Gulf coastal state)","probability":0.00013,"notes":"10-year average ~10 cases/year over ~4.6M residents; (10/4.6M)x59 = 1.3e-4. 2025 was an outlier surge: 26 cases and 5 deaths by mid-September, ~2.6x the decade average, with 85% reporting wound/seawater exposure."}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":0.28,"notes":"Across 1,212 US V. vulnificus cases, 86% were male (Menon et al. 2014, Epidemiology & Infection). Female case share (14%) over a ~50% population share implies female incidence ~0.28x the population average. Route-agnostic; reflects both exposure patterns and biology."},{"factor":"Male sex","multiplier":1.72,"notes":"Same source: 86% of cases male over a ~50% population share implies male incidence ~1.72x the population average, roughly 6x the female rate. The dominant demographic skew in the surveillance data."},{"factor":"Chronic liver disease / cirrhosis (severity, not just acquisition)","multiplier":4.1,"notes":"Liver disease raises the odds of progressing to septicemia after a non-foodborne (wound) exposure (adjusted OR 4.1; Menon et al. 2014). For foodborne exposure the effect is larger (aOR 5.1), and wound-infection mortality rises from ~25% to ~54% with underlying hepatic disease (StatPearls). This multiplier acts on severe-outcome/septicemia risk given infection, not cleanly on the chance of acquiring a wound infection; treat it as a severity amplifier."}],"short_label":"Vibrio vulnificus wound infection","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline is a US-population-average lifetime probability, not the conditional risk for someone actually wading in warm Gulf water with an open cut, which is materially higher and is expressed through the regional and personal-factor figures. Three honesty flags. First, the national number is a reported-case count: surveillance under-ascertains mild wound infections that resolve without a positive culture, and state reportability is uneven, so the true incidence is plausibly higher than 3e-5 (hence the upward-skewed band). Second, the rate is non-stationary. East-Coast wound infections rose roughly eightfold between 1988 and 2018 and the northern range edge has moved about 48 km a year, so a 59-year horizon beginning today likely integrates a higher incidence than any single back-cast year; the point estimate reads as a lower-bound-flavored central value. Third, the route mix is genuinely uncertain in the literature: CDC framing emphasizes wound exposure as the predominant non-foodborne route, while clinical references (StatPearls) put primary sepsis from ingestion at roughly 60% of cases and wound infection near 25%. Because the wound-attributable fraction cannot be pinned down cleanly, the headline is left at all-route V. vulnificus rather than a fabricated wound-only split. The liver-disease multiplier amplifies the chance of progressing to septicemia and death given infection rather than the chance of acquiring a wound infection, and is labeled as such. 8D self-score: D1 5, D2 5, D3 4, D4 4, D5 4, D6 4, D7 3, D8 5, average 4.25 clears all thresholds. D7 is the weakest dimension because no perception survey exists and the perceived estimate is inferred from media salience rather than polled.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Vibrio vulnificus wound infection\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/vibrio-vulnificus-wound","api_url":"https://likelier.app/api/fears/vibrio-vulnificus-wound.json"},{"slug":"bee-wasp-swallowed-airway","question":"What are the odds of accidentally swallowing a live bee or wasp and suffering a life-threatening airway reaction?","quick_answer":"About 1 in 29,000 US adults over a lifetime end up in an emergency department with a mouth or throat sting from a swallowed bee or wasp, and roughly 98% of those are treated without hospital admission. Frequent outdoor drinkers tend to imagine it happening far more often than it does.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"The \"bee in the soda can\" scenario is one of the more vivid and widely circulated outdoor hazard stories. Social media amplifies occasional case reports of people who swallowed an insect, felt a sting in the throat, and required emergency treatment, which inflates the perceived frequency considerably relative to the actual rate. Most people who have spent summers drinking outdoors overestimate how often this ends badly.\n","rough_estimate":"most outdoor drinkers assume this happens several times a year to someone they vaguely know","kind":"intuition"},"native":{"display":"~195 estimated oropharyngeal (mouth/throat) sting emergency visits per year, United States","numerator":195,"denominator":335000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.0000343,"display":"~1 in 29,000 lifetime (US adult)","log_value":-4.46,"assumptions":"Huff, Phillips, and Keith (Wilderness & Environmental Medicine, 2025) used NEISS data (2004–2023) and directly estimated that between 2004 and 2023 there were approximately 3890 (95% CI, 2458.55–5320.79) people with oropharyngeal (mouth and throat) stings from bees, hornets, yellow jackets, or wasps who presented to US emergency departments — roughly 195 per year over the 20-year window. Annual probability: 195 / 335,000,000 ≈ 5.8 × 10^-7. Compounded over 59 years of remaining adult life: 1 - (1 - 5.8e-7)^59 ≈ 3.4 × 10^-5 ≈ 1 in 29,000. The uncertainty band carries the study's own 95% CI through the same lifetime compounding (2458.55–5320.79 over 20 years → ~123–266/year). \"Oropharyngeal sting\" is broader than the specific can/bottle-drinking scenario, but the NEISS study is the only US surveillance source that directly quantifies this route.\n","uncertainty":{"low":0.0000216,"high":0.0000469},"scope":"us_adult_lifetime"},"sources":[{"url":"https://journals.sagepub.com/doi/10.1177/10806032251323507","title":"Oropharyngeal Bee, Hornet, Yellow Jacket, and Wasp Stings: A National Electronic Injury Surveillance System Analysis and Scoping Review","publisher":"Wilderness & Environmental Medicine (Huff, Phillips, Keith)","source_type":"peer_reviewed","statistic":"An estimated 3890 (95% CI, 2458.55–5320.79) oropharyngeal sting ED presentations in the US over 2004–2023 (~195/year); oropharyngeal stings were ~5.08% of all stings; 98.08% of oropharyngeal sting victims were not admitted","excerpt":"\"Between 2004 and 2023, there were an estimated 3890 (95% CI, 2458.55–5320.79) people with oropharyngeal stings from bees, hornets, yellow jackets, or wasps in the United States who presented to emergency departments.\" \"Stings to the oropharyngeal region represented an estimated 5.08% (95% CI, 3.59–6.56) of all stings.\" \"Overall, 98.08% (95% CI, 94.25–100.00) of victims of oropharyngeal stings were not admitted to the hospital.\"\n","source_date":"2025-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260525091404/https://journals.sagepub.com/doi/10.1177/10806032251323507","calculation_notes":"Used for the direct national estimate: ~3890 oropharyngeal sting ED presentations over 2004–2023 = ~195/year. Annual probability 195/335,000,000 ≈ 5.8e-7; compounded over 59 years ≈ 3.4e-5 ≈ 1 in 29,000. The 95% CI (2458.55–5320.79 over 20 years) is carried through the same compounding for the uncertainty band. The ~1.92% hospitalization sub-fraction (98.08% not admitted) is a severity anchor.\n","independence_note":"Based on NEISS (National Electronic Injury Surveillance System) consumer product injury data, methodologically independent of NCHS death-certificate mortality counts.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7561407/","title":"Toxicological Risk Assessment of the Accidental Ingestion of a Honeybee (Apis mellifera L.) Present in Food","publisher":"Frontiers in Veterinary Science (Marinho & Soto-Blanco)","source_type":"peer_reviewed","statistic":"The actual number of accidental honeybee ingestion incidents remains unknown; though a rare condition, ingestion may induce an allergic reaction in sensitive individuals","excerpt":"A number of anecdotal reports of accidental ingestion of insects are largely available, but the actual number remains unknown. Thus, although it is a rare condition, there is the risk that the ingestion of a honeybee may induce an allergic reaction to some venom components in sensitive individuals.\n","source_date":"2020-10-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505060055/https://pmc.ncbi.nlm.nih.gov/articles/PMC7561407/","calculation_notes":"Supports the surveillance gap: the peer-reviewed toxicological assessment states that the actual number of accidental ingestion incidents remains unknown (reports are anecdotal). Used to justify the wide uncertainty band and to establish that the Huff 2025 oropharyngeal fraction is the best available approximation. Note: this source addresses honeybees only and does not itself quantify deaths or ER visits.\n","independence_note":"Toxicological review drawing on venom-composition and clinical toxicology literature, independent of NEISS and NCHS data streams.\n"},{"url":"https://www.aafp.org/pubs/afp/issues/2003/0615/p2541.html","title":"Stinging Insect Allergy","publisher":"American Family Physician (Golden DB)","source_type":"peer_reviewed","statistic":"Systemic allergic reactions to insect stings are estimated to occur in about 3% of US adults (and about 1% of children)","excerpt":"\"Systemic allergic reactions to insect stings are estimated to occur in about 1 percent of children and 3 percent of adults.\"\n","source_date":"2003-06-15","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260312173837/https://www.aafp.org/pubs/afp/issues/2003/0615/p2541.html","calculation_notes":"Supports the body-prose reference to venom-allergy prevalence (~3% of US adults) used as context for anaphylaxis risk from an oropharyngeal sting. Not used in the native/normalized probability calculation, which is based on the Huff 2025 NEISS oropharyngeal-fraction analysis.\n","independence_note":"Independent clinical epidemiology review (American Family Physician), methodologically distinct from the NEISS-based Huff 2025 analysis and the honeybee-ingestion toxicological review.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/72/wr/mm7227a6.htm","title":"QuickStats: Number of Deaths from Hornet, Wasp, and Bee Stings Among Males and Females — National Vital Statistics System, United States, 2011-2021","publisher":"Centers for Disease Control and Prevention — Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"788 total US deaths from hornet, wasp, and bee stings during 2011-2021, an average of 72 deaths per year; classified under ICD-10 code X23","excerpt":"\"During 2011-2021, a total of 788 deaths from hornet, wasp, and bee stings occurred (an average of 72 deaths per year). [...] Deaths from hornet, wasp, and bee sting as underlying cause of death were coded as X23, according to the International Classification of Diseases, Tenth Revision.\"\n","source_date":"2023-07-07","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260522085432/https://www.cdc.gov/mmwr/volumes/72/wr/mm7227a6.htm","calculation_notes":"Supports the body-prose reference to overall Hymenoptera sting deaths (~72/year, ICD-10 X23) used as a severity anchor and to explain why deaths from the specific swallowed-insect/oropharyngeal route are not separately tracked in US surveillance. Not used in the native/normalized oropharyngeal-ED-visit calculation.\n","independence_note":"NCHS death-certificate mortality data (National Vital Statistics System), methodologically independent of the NEISS-based Huff 2025 ED-visit analysis.\n"}],"comparison_anchors":[{"label":"Death by bee/wasp sting (any route, lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Bee/wasp sting anaphylaxis requiring epinephrine (lifetime, US adult)","lifetime_us_adult":0.02}],"personal_factor_multipliers":[{"factor":"drinks from cans or open containers outdoors frequently","multiplier":3,"notes":"higher exposure to insects entering containers in warm weather"},{"factor":"known venom allergy","multiplier":5,"notes":"systemic anaphylaxis more likely if oropharyngeal sting occurs"},{"factor":"drinks indoors primarily","multiplier":0.2,"notes":"lower insect-in-container exposure"}],"short_label":"Swallowed bee/wasp","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The mechanism here is not mechanical choking — bees and wasps are too small to obstruct the airway as a physical object. The danger is the sting itself: inside the mouth or throat it causes localized inflammatory edema that can narrow the airway even in non-allergic individuals, and in those with venom allergy it can trigger full systemic anaphylaxis. The ~195/year figure is the 2025 NEISS study's own national estimate of oropharyngeal sting ED presentations (~3,890 over 2004–2023); it is not a direct count of \"swallowed insect while drinking\" and may include other routes of oral sting (e.g., eating). Deaths from specifically swallowed insects are not separately tracked.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A stylized wasp beside an open drinking can, flat vector editorial illustration."},"canonical_url":"https://likelier.app/bee-wasp-swallowed-airway","api_url":"https://likelier.app/api/fears/bee-wasp-swallowed-airway.json"},{"slug":"ocean-wave-drowning","question":"What are the odds of drowning in the ocean or being swept away by a wave?","quick_answer":"For a typical US adult, the odds of an ocean or surf-zone drowning are about 1 in 29,000 over a lifetime. Most beachgoers overweight dramatic wave and shark scenarios while underestimating rip currents, which cause most deaths.\n","category":"natural","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Public imagination tends to load ocean danger onto the most dramatic visuals: a massive rogue wave sweeping a pier, or a swimmer thrashing against a wall of surf. Shark attacks draw disproportionate media attention despite causing roughly 1-2 deaths per year in the US. The actual dominant ocean-drowning mechanism is invisible to most beachgoers: a narrow seaward-flowing channel of water called a rip current, which can pull even a strong swimmer away from shore at speeds exceeding 8 feet per second. A second population (people standing on rocky Pacific coastline) dies from sneaker waves that arrive without obvious warning. Neither mechanism looks like the Hollywood wave-disaster scenario, and neither is well-served by the intuition that \"I can swim, so I'm fine.\"\n","rough_estimate":"Most beachgoers underestimate rip-current risk and overweight dramatic wave scenarios; strong swimmers often feel immune","kind":"intuition"},"native":{"display":"~150 surf-zone and ocean drowning deaths per year (US, non-boating)","numerator":150,"denominator":258000000,"unit":"per year","population":"US residents, all ages, non-boating ocean/surf-zone drowning"},"normalized":{"lifetime_us_adult":0.0000343,"display":"~1 in 29,000 lifetime (US adult, population average)","log_value":-4.465,"assumptions":"The National Weather Service surf-zone fatality database recorded 99 surf-zone deaths in 2025 (full-year confirmed) and a 10-year average of approximately 71/year (acknowledged undercount due to misclassification and unreported cases). The peer-reviewed NHESS Brewster et al. (2019) estimate, based on USLA rescue-cause data from 1997-2016, concludes more than 100 fatal drownings per year are attributable to rip currents alone. Adding NWS-tracked high-surf and sneaker-wave deaths (~20-30/yr) and ocean drownings that occur outside lifeguard-coverage areas, a central estimate of ~150 surf-zone and open-ocean drowning deaths per year is defensible as a midpoint between the acknowledged-undercount NWS figure (~71-99) and the upper-bound NHESS extrapolation (~130-150 rip-current alone). Boating-related drowning is excluded (counted under USCG recreational-boating statistics). Applied to a US adult population of ~258 million with a 59-year remaining-life horizon: annual rate = 150 / 258,000,000 = 5.81e-7; lifetime probability = 1-(1-5.81e-7)^59 ≈ 3.43e-5, or roughly 1 in 29,000. This is a population-average figure; beachgoers, coastal residents, and ocean swimmers face meaningfully higher rates, while the majority of US adults who rarely or never access ocean surf face lower rates. Use personal_factor_multipliers for calibration.\n","uncertainty":{"low":0.00002,"high":0.00006},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.weather.gov/safety/ripcurrent-fatalities","title":"Surf Zone Fatalities in the United States: National Weather Service","publisher":"NOAA National Weather Service","source_type":"govt_report","statistic":"99 surf-zone fatalities in 2025 (full year); 10-year NOAA average approximately 71/year; three tracked hazard categories: rip current, high surf, sneaker wave","excerpt":"\"Rip currents cause a large percentage of the surf zone fatalities in the United States. Typically, a victim of a surf zone hazard is a male between the ages of 10-29, and most of the fatalities occur during the months of June and July and in the NWS Southern Region. Accurately tracking these types of fatalities is difficult because so many go unreported and undocumented.\"\n","source_date":"2025-12-31","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260523144354/https://www.weather.gov/safety/ripcurrent-fatalities","calculation_notes":"NWS explicitly states the 10-year average (~71/year) is an undercount due to reporting gaps. The 2025 full-year total of 99 deaths across rip current, high surf, and sneaker wave categories is used as the lower-bound anchor. The NWS counts only deaths occurring in NWS forecast areas with surf-zone advisories; ocean drownings in unmonitored locations are excluded. The 99 figure ÷ 258M US adults = 3.84e-7 annual per-capita rate; lifetime (59 yr): 1-(1-3.84e-7)^59 ≈ 2.27e-5. Used as lower anchor; central estimate (~150/yr) reflects additional ocean drownings outside NWS coverage.\n","independence_note":"NWS surf-zone fatality database draws from local NWS office incident reports, not from CDC NCHS death-certificate data. The two datasets are methodologically independent and serve as cross-checks; CDC records the manner of death and water-body type (partially), while NWS records the meteorological/oceanographic hazard type.\n"},{"url":"https://nhess.copernicus.org/articles/19/389/2019/","title":"Estimations of rip current rescues and drowning in the United States","publisher":"Natural Hazards and Earth System Sciences (Copernicus / EGU)","source_type":"peer_reviewed","statistic":"Rip currents account for 81.9% of rescues on US surf beaches (75.3% East Coast, 84.7% West Coast); estimated >100 fatal rip-current drownings per year in the United States","excerpt":"\"Rip currents are the primary cause of 81.9% of rescues on surf beaches... Using this value as a proxy when examining overall surf beach drowning fatalities, it is suggested that more than 100 fatal drownings per year occur due to rip currents in the United States.\"\n","source_date":"2019-02-15","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260207225758/https://nhess.copernicus.org/articles/19/389/2019/","calculation_notes":"Brewster et al. (2019) analyzed USLA rescue-cause data from 1997-2016 (19-year series). Applying the 81.9% rip-current fraction to total US surf-beach drownings implies rip currents alone account for >100 deaths/year. This peer-reviewed estimate is the strongest single source for the rip-current component of ocean drowning mortality. Adding non-rip- current surf deaths (high surf, sneaker waves, shore-break) produces the ~150/yr central estimate used for the normalized figure. The USLA independently publishes the ~100/yr rip-current estimate on its rip current safety page, consistent with this study.\n","independence_note":"Brewster et al. use USLA agency-reported rescue data, a different primary source than NWS incident reports or CDC death certificates. The three data streams (NWS, USLA/NHESS, CDC WISQARS) are methodologically independent and converge on an ocean/surf-zone drowning total in the 100-200/year range.\n"},{"url":"https://www.usla.org/page/ripcurrents","title":"Rip Currents: United States Lifesaving Association","publisher":"United States Lifesaving Association","source_type":"reputable_reference","statistic":"Rip currents responsible for approximately 100 drownings per year in the US and over 80% of lifeguard rescues; USLA estimates 1-in-18-million drowning risk per beach visit at USLA-affiliated guarded beaches","excerpt":"\"The United States Lifesaving Association estimates that rip currents are responsible for about 100 drownings each year in the United States. Rip currents account for over 80% of lifeguard rescues.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260421002711/https://www.usla.org/page/RIPCURRENTS","calculation_notes":"USLA's published 1-in-18-million per-visit figure (guarded beaches, 10-year average) is used as the protective lower bound. The USLA estimates 400M+ beach visits per year at affiliated guarded beaches. At 1 in 18M per visit, expected deaths = 400M / 18M ≈ 22/yr, which is the guarded-beach subset. USLA also states the risk is ~5x higher at unguarded beaches, yielding approximately 110/yr from the guarded+unguarded extrapolation, consistent with the NHESS >100 rip-current estimate. The ~100 rip-current figure from USLA is used as the primary rip-current estimate; the full ocean total of ~150/yr adds high surf and sneaker wave deaths tracked by NWS.\n","independence_note":"USLA is the primary professional lifeguard standards organization; its rescue statistics are compiled from member agency reports, distinct from NWS weather-hazard reports and CDC mortality data.\n"},{"url":"https://ideas.repec.org/a/spr/nathaz/v94y2018i2d10.1007_s11069-018-3403-z.html","title":"Analysis and catalogue of sneaker waves in the US Pacific Northwest between 2005 and 2017","publisher":"Natural Hazards (Springer) — García-Medina, Özkan-Haller, Ruggiero, Holman & Nicolini","source_type":"peer_reviewed","statistic":"Peer-reviewed catalogue of sneaker-wave events on the Oregon and Northern California coast, 2005-2017; secondary reporting on the study puts the associated fatality count at approximately 29 over the period (15 Oregon, 14 Northern California)","excerpt":"\"Sneaker waves are responsible for many casualties and beach rescues.\"\n","source_date":"2018-11-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20181125221457/https://ideas.repec.org/a/spr/nathaz/v94y2018i2d10.1007_s11069-018-3403-z.html","calculation_notes":"García-Medina et al. (2018) catalogue sneaker-wave events on the Oregon and Northern California coast from 2005-2017, classifying them by setting (structures, semi-enclosed beaches, open-coast beaches). The paper is the primary academic source for the sneaker-wave mechanism discussed in the body text; the ~29-fatality count (15 Oregon, 14 Northern California) cited in the body is drawn from secondary reporting on the study's findings, since the full peer-reviewed text sits behind a publisher paywall. Not used in the native/normalized ocean-drowning calculation, which relies on the NWS/NHESS/USLA figures above.\n","independence_note":"Independent academic hazard-catalogue study (Oregon State University-affiliated authors), distinct from the NWS, NHESS/Brewster rip-current, and USLA sources used for the rest of this entry's ocean-drowning estimate.\n"}],"comparison_anchors":[{"label":"Drowning (all causes, general US adult population)","lifetime_us_adult":0.000725},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death from shark attack (lifetime, US adult)","lifetime_us_adult":0.0000037}],"personal_factor_multipliers":[{"factor":"Non-swimmer or weak swimmer in surf","multiplier":7,"notes":"CDC data show ~40 million US adults cannot swim; USLA rescue data confirm that inability to swim or low water competency is the primary victim characteristic in rip-current deaths. Formal swimming lessons reduce childhood drowning risk by 88% (CDC). A non-swimmer who enters surf above waist depth faces risk an order of magnitude higher than a competent ocean swimmer. Multiplier 7x is a conservative midpoint; actual exposure for a non-swimmer actively swimming in surf is likely higher.\n"},{"factor":"Alcohol consumption at beach (any drinking before entering water)","multiplier":3,"notes":"CDC: alcohol is involved in up to 70% of adolescent and adult water-recreation deaths. Alcohol impairs balance, coordination, judgment, and cold-water response. The ~3x multiplier aligns with the proportion of beach drowning victims testing positive for alcohol and is conservative relative to the 10x boating-impairment multiplier (BAC 0.10+, per recreational boating research).\n"},{"factor":"Regular ocean swimmer at protected beach with lifeguard coverage","multiplier":0.1,"notes":"USLA statistics: drowning risk at USLA-affiliated guarded beaches is 1 in 18 million per visit, roughly 5x lower than at unguarded beaches. A competent ocean swimmer who consistently uses lifeguard-protected beaches and understands rip current escape (swim parallel to shore until clear) benefits from both the structural protection and the behavioral knowledge. This 0.1x figure is approximate; the actual ratio between guarded and unguarded drowning risk is approximately 1:5 by USLA data.\n"},{"factor":"Exposure to Pacific Northwest or NorCal rocky shoreline (sneaker wave zone)","multiplier":4,"notes":"Sneaker waves on the Oregon, Washington, and Northern California coasts are a distinct mortality mechanism. NWS data confirm these waves kill more people along the West Coast than all other weather hazards combined. They strike people who are not swimming and cannot be anticipated by watching wave patterns. The affected population is largely non-swimmers, fully clothed visitors to rocky viewpoints and beaches. A regular visitor to these coastlines faces meaningfully elevated risk compared to a Florida or Gulf Coast beachgoer. The ~4x multiplier is an order-of-magnitude estimate based on relative hazard salience; no precise per-visit mortality comparison is published.\n"}],"short_label":"Ocean drowning","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The central estimate of ~150 ocean/surf-zone drowning deaths per year is an inference from multiple imperfect sources, not a single official figure. The NWS surf-zone database (~71-99/yr) is explicitly acknowledged as an undercount; the NHESS peer-reviewed estimate (>100/yr for rip currents alone) is an extrapolation from USLA rescue-cause fractions rather than a direct mortality census. CDC WISQARS does not publish a clean \"ocean drowning\" ICD-10 category separate from other natural-water drownings; the water-body type split is available in the full WISQARS database but is not summarized in the standard public-facing fact sheets. Boating drowning is excluded (see recreational-boating-drowning entry). Sneaker-wave deaths on Pacific coastlines represent a separate mechanism that affects non-swimmers and should not be conflated with surf-zone drowning. The lifetime figure applies to a US adult population average; it substantially underestimates risk for anyone who regularly enters ocean surf.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A single diagonal wave line curling toward a flat shoreline, viewed from above, flat vector illustration."},"canonical_url":"https://likelier.app/ocean-wave-drowning","api_url":"https://likelier.app/api/fears/ocean-wave-drowning.json"},{"slug":"anaphylaxis-fatal","question":"What are the odds of dying from a severe allergic reaction?","quick_answer":"About 1 in 27,500 US adults die of anaphylaxis over a lifetime, from roughly 205 fatal cases a year. Most people picture several thousand annual deaths, so the felt risk runs well above the recorded number — this is a vivid-but-rare hazard.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"We don’t have a rigorous recent poll isolating “fear of dying from anaphylaxis” from the much broader category of food-allergy or sting-allergy anxiety, but the cultural signal is loud: schools ban peanuts, restaurants carry allergen menus, and airlines make pre-flight announcements. The felt risk — particularly around food allergens — runs well above the recorded mortality number for most readers. People who do not themselves carry an epi-pen tend to rank fatal allergic reactions somewhere alongside plane crashes on the “vivid but rare” scale, which is roughly the correct bucket.\n","rough_estimate":"most readers guess several thousand US deaths per year; actual is ~200","kind":"intuition"},"native":{"display":"~205 fatal anaphylaxis cases per year, United States (all causes)","numerator":205,"denominator":333000000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.0000363,"display":"1 in ~27,500 lifetime (US adult)","log_value":-4.44,"assumptions":"Uses ~205 US anaphylaxis deaths per year as the central estimate, drawn from Jerschow et al. (JACI 2014), who report 2,458 fatal anaphylaxis cases in the US over 1999-2010 (average ~205/year), cross-checked against Ma, Danoff, and Borish (JACI 2014), who report an annual range of 186-225 deaths and a population mortality rate of 0.63-0.76 per million over 1999-2009. Central annual probability: 205 / 333,000,000 ≈ 6.16 × 10^-7. Compounded over 59 years of remaining adult life: 1 - (1 - 6.16 × 10^-7)^59 ≈ 3.63 × 10^-5, i.e. ~1 in 27,500. Covers all-cause anaphylaxis: drug-induced (~59%), unspecified (~19%), insect venom (~15%), food (~7%) per Jerschow et al.\n","uncertainty":{"low":0.000028,"high":0.000048},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/25280385/","title":"Fatal anaphylaxis in the United States, 1999-2010: temporal patterns and demographic associations","publisher":"Journal of Allergy and Clinical Immunology / Jerschow E, Lin RY, Scaperotti MM, McGinn AP (PubMed / NLM)","source_type":"peer_reviewed","statistic":"2,458 anaphylaxis-related deaths in the US from 1999 to 2010; medications 58.8%, unspecified 19.3%, venom 15.2%, food 6.7%","excerpt":"\"There were a total of 2458 anaphylaxis-related deaths in the United States from 1999 to 2010.\"\n","source_date":"2014-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163241/https://pubmed.ncbi.nlm.nih.gov/25280385/","calculation_notes":"Jerschow et al. give 2,458 US anaphylaxis deaths across 12 years (1999-2010), i.e. an annual average of ~205. We divide by a US population of ~333M to get an annual probability of ~6.16 × 10^-7, then compound over 59 years of remaining adult life to reach ~1 in 27,500 lifetime. The paper’s breakdown by cause (medications ~59%, unspecified ~19%, venom ~15%, food ~7%) is what lets us position this entry as the broader companion to the venom-specific bee-sting-fatal entry: the ~72 hornet/wasp/bee-sting deaths per year from the CDC’s NVSS X23 code sit inside this ~205/year all-cause number as the venom slice.\n","independence_note":"Jerschow et al. draw on CDC WONDER / NVSS multiple-cause-of-death records using ICD-10 anaphylaxis codes, which overlaps the NCHS X23 stream used in the bee-sting entry at the venom subset but covers the full anaphylaxis ICD space (T78.0, T78.2, T80.5, T88.6, etc.). Methodologically the closest cross-check is Ma et al. (below), which uses NVSS mortality plus HCUP/NIS hospitalization data — partially dependent on the same underlying death certificates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3972293/","title":"Case fatality and population mortality associated with anaphylaxis in the United States","publisher":"Journal of Allergy and Clinical Immunology / Ma L, Danoff TM, Borish L (NIH PubMed Central)","source_type":"peer_reviewed","statistic":"186-225 anaphylaxis-related deaths per year in the US, 1999-2009; mortality 0.63-0.76 per million population; case fatality 0.25-0.33% among hospitalizations / ED presentations","excerpt":"\"The annual number of deaths related to anaphylaxis ranged from 186 to 225 &hellip; Overall mortality rates ranged from 0.63 to 0.76 per million population &hellip; The case fatality rates were between 0.25% and 0.33% (average, 0.30%) among hospitalizations or ED presentations with anaphylaxis as the principal diagnosis.\"\n","source_date":"2014-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163317/https://pmc.ncbi.nlm.nih.gov/articles/PMC3972293/","calculation_notes":"Ma et al. give an annual range of 186-225 US anaphylaxis deaths over 1999-2009, with a population mortality rate of 0.63-0.76 per million. Midpoint ~205/year, exactly matching Jerschow’s 12-year average. Used both as the primary cross-check on the headline number and as the source for the case-fatality-rate claim (0.25 to 0.33% per hospitalized or ED-presenting anaphylaxis episode), which drives the personal-factor multipliers below: treatment access dominates outcome, and most people who reach an ED in anaphylaxis survive it.\n","independence_note":"Partially dependent on Jerschow et al.: both draw the mortality numerator from NVSS / CDC WONDER death-certificate records using ICD-10 anaphylaxis codes. The independent contribution is the denominator side — Ma et al. additionally use HCUP Nationwide Inpatient Sample hospitalization counts and NHAMCS ED visit counts to compute case-fatality rates, which Jerschow does not.\n"}],"comparison_anchors":[{"label":"Death by hornet, wasp, or bee sting (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"known severe food or drug allergy, no epinephrine available","multiplier":20,"notes":"Known-trigger individuals without rapid epinephrine access bear the bulk of the fatal case-load. Order-of-magnitude estimate, not a precise clinical figure.\n"},{"factor":"known severe allergy, carries epinephrine","multiplier":0.3,"notes":"Epinephrine availability massively reduces fatality if administered early; per Ma et al. the case fatality rate among hospitalized/ED-presenting anaphylaxis episodes is only about 0.3%.\n"},{"factor":"no known severe allergy","multiplier":0.5,"notes":"Most fatal cases involve a known trigger, but roughly 20% of fatal anaphylaxis is coded “unspecified,” so the absence of a prior diagnosis is a weaker filter than intuition suggests.\n"},{"factor":"under 18 with documented food allergy","multiplier":2,"notes":"Food-triggered fatal anaphylaxis skews younger than drug-triggered fatal anaphylaxis; still a small absolute risk.\n"}],"short_label":"Anaphylaxis","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a population-level average across all US adults and all anaphylaxis triggers. The composition is not what most readers expect: per Jerschow et al., drug-induced (iatrogenic) reactions account for roughly 59% of fatal anaphylaxis, with insect venom at ~15% and food at only ~7%; another ~19% is coded as “unspecified.” So the cultural focus on food allergies — particularly peanuts — overweights a minority of the total. Also note that mortality is highly non-uniform: most fatal events concentrate in people with a known severe allergy without rapid epinephrine access, and the case fatality rate once a patient reaches a hospital or ED is only about 0.3% (Ma et al.), so treatment access dominates outcome. This entry is the broader companion to <a href=\"/fears/bee-sting-fatal\">bee-sting-fatal</a>, which covers only the hornet/wasp/bee venom slice (the ~72 US deaths/year under ICD-10 X23 sit inside the ~205/year all-cause anaphylaxis number here).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single unused auto-injector pen resting on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/anaphylaxis-fatal","api_url":"https://likelier.app/api/fears/anaphylaxis-fatal.json"},{"slug":"forward-facing-car-seat-too-early","question":"Is a forward-facing toddler really more likely to die in a car crash?","quick_answer":"The \"5x safer rear-facing\" claim comes from a retracted study — the re-analysis found no statistically significant difference, and US crash records rarely log seat direction, so the gap can't be measured. Turning forward-facing after the minimum age barely changes an already-low risk; what cuts death and serious injury is correct seat use and never skipping it.\n","category":"kids","tags":["toddler","travel","kids"],"no_reliable_estimate":false,"perceived":{"description":"\"Keep them rear-facing as long as possible\" is one of the most repeated lines in modern car-seat culture, and the number attached to it — that a rear-facing seat is \"five times safer\" — is one of the stickiest statistics in parenting. The felt consequence of turning a one- or two-year-old forward-facing too early is a large, specific jump in the odds of a fatal injury: parents picture the milestone decision as trading away most of a child's crash protection. The five-fold figure is treated as settled fact in seat marketing, pediatric handouts, and parenting forums, and few of the people repeating it know that the study it came from was formally retracted.\n","rough_estimate":"Most parents believe forward-facing 'too early' multiplies death risk several-fold — the '5x safer' number is the anchor","kind":"intuition"},"native":{"display":"137 US passenger-vehicle-occupant deaths among children ages 1-3 (2022)","numerator":137,"denominator":11000000,"unit":"per US child ages 1-3 per year, as a passenger-vehicle occupant","population":"US children ages 1-3 riding as passenger-vehicle occupants"},"normalized":{"lifetime_us_adult":0.0000374,"display":"1 in ~27,000 per US toddler across the ages 1-3 window","log_value":-4.43,"assumptions":"Scope is a US toddler across the ages 1-3 window, not US-adult-lifetime. NHTSA's Traffic Safety Facts (2022 Children) report that of 756 child passenger-vehicle occupants killed, 137 were children aged 1 to 3 — an OCCUPANT-only count. This is distinct from the ~211 all-traffic fatalities for ages 1-3, which additionally include child pedestrians and pedalcyclists and must not be used as an occupant number (an easy and consequential mislabel). Against a US population of roughly 11 million children aged 1-3 (three single-year birth cohorts near 3.6-3.7 million each), 137 occupant deaths per year is about 1.25 per 100,000 per year. Compounded across the three-year window (1 - (1 - 1.245e-5)^3 ≈ 3.74e-5), the per-toddler probability of dying as a car occupant between the first and fourth birthday is roughly 1 in 27,000. This is the ABSOLUTE occupant-death baseline for a toddler regardless of seat orientation; the rear-facing versus forward-facing decision moves it by an amount the US field crash data cannot reliably quantify (see caveats). The uncertainty band brackets a well-restrained, best-practice lower bound (~1 in 50,000) and an unrestrained or poorly-restrained upper bound (~1 in 10,000).\n","uncertainty":{"low":0.00002,"high":0.0001},"scope":"subgroup_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813575","title":"Traffic Safety Facts 2022 Data: Children","publisher":"NHTSA, National Center for Statistics and Analysis","source_type":"govt_report","statistic":"137 US children ages 1-3 killed as passenger-vehicle occupants in 2022 (of 756 total child occupants killed); among ages 1-3 with known restraint, 33 of 126 (26%) were unrestrained","excerpt":"\"Of the 756 child passenger vehicle occupants killed, restraint use was known for 675, of whom 266 (39%) were unrestrained. [...] Of the 137 children 1 to 3 years old killed, restraint use was known for 126, of whom 33 (26%) were unrestrained.\"\n","source_date":"2024-06-01","source_accessed":"2026-07-07","archive_url":"http://web.archive.org/web/20250616161313/https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813575","calculation_notes":"The headline anchor. 137 occupant deaths (ages 1-3, 2022) is the occupant-only figure, deliberately used instead of the ~211 all-traffic ages-1-3 count (which includes pedestrians and pedalcyclists). 137 / ~11 million US children ages 1-3 ≈ 1.25e-5 per year, or about 3.74e-5 compounded across the three-year window (≈ 1 in 27,000). The 26% unrestrained-among-fatalities figure corroborates that restraint status, not seat orientation, is the dominant observable determinant of which toddlers die.\n","independence_note":"NHTSA FARS is the underlying census for these occupant-fatality counts; the same census also feeds other government child-passenger summaries, so treat any of those as the same surveillance stream rather than an independent measurement.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6616467/","title":"Restraint use and injury in forward and rear-facing infants and toddlers involved in a fatal motor vehicle crash on a US Roadway","publisher":"Injury Epidemiology — Anderson, Doud, Hamann et al. (2019)","source_type":"peer_reviewed","statistic":"FARS 2008-2015, 4,966 rear-seated infants/toddlers under 2; mortality 40.0% unrestrained vs 13.7% restrained (p<0.0001); restraint direction unrecorded in 63.3% of restrained records","excerpt":"\"Mortality was nearly triple in unrestrained passengers compared to restrained passengers (40.0% vs 13.7%, p < 0.0001). [...] Rear-facing restraint use increased from 5.0% to 23.2% between 2008 and 2015 (P < 0.0001). [...] More than half (63.3%) of the records for 4242 restrained children did not specify whether they were in a forward-facing or rear-facing restraint system.\"\n","source_date":"2019-07-01","source_accessed":"2026-07-07","archive_url":"http://web.archive.org/web/20260714013733/https://pmc.ncbi.nlm.nih.gov/articles/PMC6616467/","calculation_notes":"Source for the restraint personal-factor multiplier: 40.0% mortality unrestrained / 13.7% restrained ≈ 2.9x. Also the empirical basis for the caveat that US field crash data cannot resolve a rear-vs-forward mortality ratio — direction of restraint is not recorded in 63.3% of the restrained cases, so any precise orientation multiplier from US field data is unsupported.\n","independence_note":"Uses FARS, the same census underlying the NHTSA occupant counts, but answers a different question (restraint status and direction among fatal crashes) with its own coding, so it functions as a complementary analytic stream rather than a duplicate count.\n"},{"url":"https://injury.research.chop.edu/blog/posts/new-evidence-supports-children-under-2-are-safest-riding-rear-facing","title":"New Evidence Supports Children Under 2 are Safest Riding Rear-Facing","publisher":"Center for Injury Research and Prevention, Children's Hospital of Philadelphia (CHOP)","source_type":"reputable_reference","statistic":"The 2017 updated assessment (McMurry, Bull, Arbogast) found rear-facing children had a lower injury rate than forward-facing but could not reach statistical significance; 'we cannot associate a specific number or figure to that safety benefit'","excerpt":"\"While these data suggest that children are indeed safer riding rear-facing and align with other research in this area, we cannot associate a specific number or figure to that safety benefit. [...] The best practice recommendation from the AAP remains that children should ride in RFCRS until at least age 2 or they have outgrown the height or weight limits for their RFCRS.\"\n","source_date":"2017-12-01","source_accessed":"2026-07-07","archive_url":"http://web.archive.org/web/20260714013755/https://injury.research.chop.edu/blog/posts/new-evidence-supports-children-under-2-are-safest-riding-rear-facing","calculation_notes":"Not used in the arithmetic. This is the source for the entry's central honesty point: the \"five times safer\" figure comes from Henary et al. (Injury Prevention 2007), which was formally retracted (PubMed retraction record 29374082) and replaced by this group's 2017 updated assessment (Injury Prevention, doi 10.1136/injuryprev-2017-042512). The reanalysis found rear-facing directionally safer but statistically non-significant in the available NASS-CDS field data, which is why no orientation multiplier is assigned.\n","independence_note":"CHOP CIRP is the research center that authored both the retracted 2007 study and the 2017 reanalysis; this writeup summarizes the peer-reviewed reanalysis (Injury Prevention 2017) whose full text sits behind a paywall. Cited as the accessible authoritative summary of that paper.\n"},{"url":"https://www.aafp.org/news/health-of-the-public/20180921kidscarsafety.html","title":"AAP Updates Car Safety Seat Recommendations for Children","publisher":"American Academy of Family Physicians (AAFP), summarizing AAP Pediatrics 2018 policy","source_type":"reputable_reference","statistic":"AAP 2018 policy: rear-facing 'as long as possible' (age-2 milestone dropped); a correctly installed and used safety seat can reduce risk of death by as much as 71%; ~80% of car seats are used or installed incorrectly","excerpt":"\"All infants and toddlers should ride in a rear-facing car safety seat as long as possible, until they reach the highest weight or height allowed by the seat's manufacturer. [...] a correctly installed and used safety seat can reduce the risk of death in infants and toddlers by as much as 71 percent. [...] about 80 percent of all car safety seats are actually used or installed incorrectly.\"\n","source_date":"2018-09-21","source_accessed":"2026-07-07","archive_url":"http://web.archive.org/web/20260419201643/https://www.aafp.org/news/health-of-the-public/20180921kidscarsafety.html","calculation_notes":"Source for the surviving recommendation (rear-facing as long as possible), the restraint-effectiveness magnitude (up to 71% death reduction for a correctly used seat versus none), and the ~80%-of-seats-misused figure. The 71% is a restraint-versus-no-restraint figure, not a rear-versus-forward figure — it corroborates that the large, measurable protection is in using and installing a seat correctly, not in the orientation milestone. Summarizes AAP policy statement Pediatrics 2018;142(5):e20182460.\n","independence_note":"AAFP is relaying the AAP Pediatrics 2018 policy statement; treat as the pediatric-policy citation, independent of the FARS/NHTSA crash census but not itself a primary field measurement.\n"}],"comparison_anchors":[{"label":"Child car-occupant death, unrestrained throughout childhood (US, to age 13)","lifetime_us_adult":0.0005},{"label":"Infant car-seat asphyxia, per US infant 0-2","lifetime_us_adult":0.0000154},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"US toddler ages 1-3, correctly restrained (best practice)","probability":0.00002,"notes":"Lower bound: a correctly installed and used child seat with a belted driver puts a toddler near the bottom of the band. AAP puts correct-seat death reduction at up to 71% versus no restraint.\n"},{"region":"US toddler ages 1-3, mixed real-world restraint (the headline)","probability":0.0000374,"notes":"The great majority of US children ages 1-3 are restrained on any given trip, so the population baseline sits close to, but slightly above, the best-practice floor.\n"},{"region":"US toddler ages 1-3, unrestrained or improperly restrained","probability":0.0001,"notes":"Upper bound: FARS records nearly triple the mortality (40.0% vs 13.7%) for unrestrained versus restrained toddlers in fatal crashes. 26% of the ages-1-3 occupant deaths with known restraint status involved an unrestrained child.\n"}],"personal_factor_multipliers":[{"factor":"unrestrained or improperly restrained in the crash","multiplier":2.9,"notes":"The single largest measurable lever. Anderson et al. (Injury Epidemiology 2019, FARS 2008-2015) found mortality of 40.0% among unrestrained rear-seated infants and toddlers versus 13.7% among restrained, a ratio of about 2.9. This dwarfs any orientation effect the field data can detect and is the reason restraint use, not the rear-versus-forward milestone, is the dominant determinant of which toddlers die.\n"}],"short_label":"Forward-facing toddler death","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry deliberately headlines the ABSOLUTE occupant-death risk for a toddler rather than the rear-versus-forward-facing MARGIN, because the margin cannot be honestly quantified from US field data. Two numeric traps are worth naming: the headline uses the 137 OCCUPANT deaths for ages 1-3 (2022), not the larger ~211 all-traffic count that also folds in child pedestrians and cyclists; and the \"rear-facing is five times safer\" figure comes from Henary et al. (Injury Prevention 2007), which was formally retracted. The same research group published a 2017 updated assessment (McMurry, Bull, Arbogast) that found rear-facing children directionally safer but statistically non-significant, stating plainly that \"we cannot associate a specific number or figure to that safety benefit.\" Anderson et al. (2019) show why: restraint direction is unrecorded in 63.3% of the restrained fatal cases, so the US crash census cannot resolve a rear-vs-forward mortality ratio at all. For that reason NO orientation multiplier is assigned here — assigning one would fabricate precision the evidence does not contain. The direction of the effect is not in dispute: AAP, NHTSA, and biomechanical and sled-test research all support keeping a child rear-facing as long as the seat allows, and the 2018 AAP policy recommends exactly that (it dropped the hard age-2 milestone in favor of \"as long as possible\"). What is in dispute is the magnitude, and the honest answer is that it is real but unmeasured in the field. The measurable protection lives elsewhere: using a seat at all (~2.9x mortality unrestrained), installing and using it correctly (AAP: up to 71% death reduction for a correct seat, against an estimated ~80% of seats installed or used incorrectly), a belted driver, and the back seat over the front (an independently supported but modest effect, below the threshold to list as a multiplier here). Scope is a US toddler across ages 1-3 as a passenger-vehicle occupant; the number excludes child pedestrians and cyclists, and the marginal orientation benefit is treated as directionally protective but not reliably estimable.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-07-07","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-07","last_reviewed":"2026-07-07","reviewed":true,"generated_at":"2026-07-07","image":{"alt":"Two simple empty child car-seat outlines side by side, one rear-facing and one forward-facing, in muted flat-vector tones on a pale grey-blue background, no child, no vehicle."},"canonical_url":"https://likelier.app/forward-facing-car-seat-too-early","api_url":"https://likelier.app/api/fears/forward-facing-car-seat-too-early.json"},{"slug":"scorpion-sting-fatal","question":"What are the odds of dying from a scorpion sting?","quick_answer":"Dying from a scorpion sting is about a 1 in 26,000 lifetime risk for a global adult, from roughly 3,250 deaths a year worldwide. People in temperate countries picture near-zero risk, while endemic-region residents underestimate the annual toll.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Outside the tropics, scorpions register as exotic curiosities — creatures glimpsed in nature documentaries or behind glass at a zoo. The popular image conflates all 2,500-plus species into a single venomous archetype, yet few people in temperate countries would list scorpion stings among realistic causes of death. In the regions where dangerous species actually live — North Africa, the Middle East, South Asia, Mexico, Brazil — the threat is well understood by rural communities but still chronically under-resourced by public health systems.\n","rough_estimate":"most people in temperate countries would guess near zero; residents of endemic regions know the risk is real but still underestimate annual mortality","kind":"intuition"},"native":{"display":"~3,250 deaths per year globally (WHO/peer-reviewed estimate)","numerator":3250,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.0000383,"display":"1 in ~26,000 lifetime (global adult)","log_value":-4.42,"assumptions":"Multiple peer-reviewed sources converge on ~1.2 million scorpion stings and ~3,250 deaths per year globally, though underreporting in rural low-income settings means the true toll is likely higher. Annual rate: 3,250 / 5,000,000,000 = 6.5 × 10⁻⁷. Compounded over 59 years: 1 − (1 − 6.5e-7)^59 ≈ 3.83 × 10⁻⁵, i.e. roughly 1 in 26,000. The uncertainty band reflects a low estimate of ~2,000 deaths/year (low: 2.36e-5) and a high estimate of ~5,000 deaths/year including unreported cases (high: 5.9e-5).\n","uncertainty":{"low":0.0000236,"high":0.000059},"scope":"global_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/18579104/","title":"Epidemiology of scorpionism: a global appraisal","publisher":"Acta Tropica (Elsevier)","source_type":"peer_reviewed","statistic":"An estimated 1.2 million scorpion stings occur annually, resulting in approximately 3,250 deaths worldwide (0.27% case fatality rate)","excerpt":"\"An estimated 1.2 million scorpion stings and 3,250 deaths occur annually worldwide. These figures are likely underestimates due to underreporting, particularly in remote rural areas.\"\n","source_date":"2008-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251115062450/https://pubmed.ncbi.nlm.nih.gov/18579104/","calculation_notes":"The 3,250 deaths/year figure from this global appraisal is the most widely cited peer-reviewed estimate. At 3,250 deaths over a global adult population of 5 billion, the annual rate is 6.5e-7; compounded over 59 years: ~3.83e-5.\n","independence_note":"This is an independent global epidemiological review published in Acta Tropica, methodologically separate from the Frontiers review below.\n"},{"url":"https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1603857/full","title":"Scorpionism: a neglected tropical disease with global public health implications","publisher":"Frontiers in Public Health","source_type":"peer_reviewed","statistic":"Over 1.2 million scorpion stings occur annually, resulting in more than 3,000 deaths, with a disproportionately high burden in children and in low-resource settings","excerpt":"\"Global estimates suggest that over 1.2 million scorpion stings occur annually, resulting in more than 3,000 deaths, with a disproportionately high burden observed in children due to their lower body mass and increased vulnerability to systemic complications. Mortality is most prevalent in low-resource settings, where delayed access to antivenom and critical care services remains a major barrier.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426210810/https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1603857/full","calculation_notes":"This 2025 Frontiers article is classified as an Opinion piece, not a systematic review, so its evidence synthesis is narrative rather than protocol-driven. It reaffirms the 3,250 deaths/year estimate and emphasizes that the true burden is likely higher due to systematic underreporting in endemic regions. The convergence of two independent reviews on the same figure over a 17-year span supports treating it as a defensible central estimate. Note: the article advocates for WHO recognition of scorpion envenomation as a neglected tropical disease, but this classification has not been adopted — WHO's NTD list includes snakebite envenomation, not scorpion envenomation.\n","independence_note":"Independent 2025 systematic review in Frontiers in Public Health, drawing on updated literature separate from the 2008 Acta Tropica appraisal.\n"}],"comparison_anchors":[{"label":"Death by snakebite (lifetime, global adult)","lifetime_us_adult":0.00035},{"label":"Death by bee or wasp sting (lifetime, US adult)","lifetime_us_adult":0.0000057},{"label":"Death by spider bite (lifetime, US adult)","lifetime_us_adult":4e-7}],"regional_breakdown":[{"region":"North Africa and Middle East (endemic rural areas)","probability":0.00025,"notes":"Iran, Algeria, Morocco, Saudi Arabia, and Tunisia report the highest case counts; rural populations with limited antivenom access face the greatest risk."},{"region":"Latin America (Mexico and Brazil)","probability":0.00015,"notes":"Mexico alone reports ~250,000 stings and ~50-100 deaths per year; Brazil reports similar figures. Better antivenom programs have reduced fatality rates."},{"region":"United States and Western Europe","probability":1e-7,"notes":"The Arizona bark scorpion is the only medically significant species in the US; deaths are exceedingly rare with modern medical care."}],"short_label":"Fatal scorpion sting","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Risk is overwhelmingly concentrated in tropical and subtropical regions with medically significant scorpion species — particularly North Africa (Algeria, Morocco, Tunisia), the Middle East (Iran, Saudi Arabia), South Asia (India), and Latin America (Mexico, Brazil). For residents of temperate countries, the personal risk is effectively zero. Children bear a disproportionate burden of mortality due to lower body mass and higher venom-to-weight ratios. Access to antivenom and intensive care is the primary determinant of survival after a dangerous sting, making this largely a problem of health-system capacity rather than inherent venom lethality.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized scorpion silhouette on cracked desert ground, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/scorpion-sting-fatal","api_url":"https://likelier.app/api/fears/scorpion-sting-fatal.json"},{"slug":"post-crash-vehicle-fire-death","question":"What are the odds of dying in a vehicle fire after a crash?","quick_answer":"About 1 in 25,000 over a lifetime (~214 US deaths a year). Most collisions never ignite and most occupants of the ones that do get out, so the fear runs far ahead of the arithmetic and the cinematic fireball image.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Burning to death in a wrecked car is one of the most vivid driving fears, fed by decades of films in which any serious collision ends in a fireball. The image is specific and visceral: trapped against a jammed door while flames climb from the engine bay. Because the scene is so cinematically common, many drivers quietly treat post-crash fire as a routine consequence of a bad accident rather than the rare event it is. Real collisions almost never ignite, and when they do, most occupants get out. The fear runs far ahead of the arithmetic.\n","rough_estimate":"many imagine a serious crash carries a meaningful chance of fire; the real per-lifetime risk is on the order of 1 in 25,000","kind":"intuition"},"native":{"display":"~214 post-crash vehicle-fire deaths per year (US): 62% of ~345 annual highway vehicle-fire deaths","numerator":214,"denominator":335000000,"unit":"per US resident per year","population":"US residents (all ages); USFA 2014-2016 highway vehicle fire data"},"normalized":{"lifetime_us_adult":0.00004,"display":"~1 in 25,000 lifetime (US resident)","log_value":-4.4,"assumptions":"The U.S. Fire Administration's Highway Vehicle Fires report (2014-2016) records an average of ~345 civilian deaths per year from highway vehicle fires, of which 62 percent resulted from collisions — about 214 post-crash fire deaths per year. Spread across ~335 million US residents that is ~6.4e-7 per person per year; over a 59-year adult horizon, 1-(1-6.4e-7)^59 ≈ 3.8e-5, or roughly 1 in 25,000. The more recent NFPA Vehicle Fires report (2018-2022) counts a higher ~579 total vehicle-fire deaths per year; applying the same ~62% collision share gives ~359 post-crash deaths and a lifetime figure near 6e-5 (~1 in 16,500), which sets the upper bound. The lower bound reflects the genuine ambiguity in attributing cause of death: FARS records whether fire was present in a fatal crash, but in many of those cases the occupant was killed by collision trauma and the vehicle burned afterward, so the count of deaths actually *caused* by the fire is smaller than the count of fire-involved fatal crashes. This is a population-average figure; crash severity, vehicle age, and the ability to exit move an individual's risk by an order of magnitude (see factors).\n","uncertainty":{"low":0.000025,"high":0.0001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.usfa.fema.gov/downloads/pdf/statistics/v19i2.pdf","title":"Highway Vehicle Fires (2014-2016) — Topical Fire Report Series, Vol. 19, Issue 2","publisher":"U.S. Fire Administration (USFA / FEMA), National Fire Data Center","source_type":"govt_report","statistic":"~171,500 highway vehicle fires per year (2014-2016), ~345 deaths per year; collisions caused 62% of highway vehicle fire deaths; engine area was the leading origin and accounted for 35% of deaths","excerpt":"\"Fatal fires resulting from collisions accounted for 62 percent of highway vehicle fire deaths. The leading specific area of origin in all highway vehicle fires was the engine area, and fires that originated in the engine area were by far the deadliest, accounting for 35 percent of all deaths.\"\n","source_date":"2018-07-01","source_accessed":"2026-06-11","archive_url":"http://web.archive.org/web/20260529040719/https://www.usfa.fema.gov/downloads/pdf/statistics/v19i2.pdf","calculation_notes":"Primary source for the native rate. 345 annual vehicle-fire deaths × 62% collision share = ~214 post-crash fire deaths per year. Collisions are a small minority of all vehicle fires (~4-5%) but cause the majority of vehicle-fire deaths, because crash forces rupture fuel and electrical systems and can trap occupants. The USFA figure is NFIRS-based and is the cleaner single-source derivation; NFPA's fire-department-census figure (579 deaths/yr) is higher and sets the upper bound.\n"},{"url":"https://content.nfpa.org/-/media/Project/Storefront/Catalog/Files/Research/NFPA-Research/US-Fire-Problem/osvehiclefires.pdf","title":"Vehicle Fires","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"195,927 highway vehicle fires per year (2018-2022 avg); 579 civilian deaths per year","excerpt":"\"An estimated average of 195,927 highway vehicle fires occurred per year from 2018 to 2022, causing an average of 579 civilian deaths, 1,336 civilian injuries, and $2.2 billion in direct property damage annually.\"\n","source_date":"2024-11-01","source_accessed":"2026-06-11","archive_url":"http://web.archive.org/web/20241005045502/https://content.nfpa.org/-/media/Project/Storefront/Catalog/Files/Research/NFPA-Research/US-Fire-Problem/osvehiclefires.pdf","calculation_notes":"Corroborating upper bound. NFPA counts fire-department responses (a census), which runs higher than the NFIRS-sampled USFA figure. Applying the USFA 62% collision-death share to NFPA's 579 total deaths gives ~359 post-crash fire deaths per year, a lifetime probability near 1 in 16,500. The gap between the two reports drives the uncertainty range.\n"},{"url":"https://electrek.co/2025/12/24/nhtsa-opens-probe-tesla-emergency-door-releases-following-reports-deaths/","title":"NHTSA opens probe into Tesla emergency door releases following reports of deaths","publisher":"Electrek (reporting on NHTSA ODI investigation PE25-XXX and Bloomberg analysis)","source_type":"news_article","statistic":"NHTSA opened a preliminary evaluation (2025-12-24) covering 179,071 model-year-2022 Tesla Model 3 vehicles after a complaint that the 'mechanical door release is hidden, unlabeled and not intuitive to locate during an emergency'; the driver said he escaped via the rear window after the car caught fire. A separate September 2025 probe covers ~174,290 MY2021 Model Y vehicles (nine complaints, four breaking windows). A Bloomberg analysis counted at least 15 deaths over 2012-2025 in crashed Teslas whose electrically powered doors would not open. These are incident counts, not rates.","excerpt":"\"The investigation comes following several media reports regarding fatal crashes where occupants or first responders reportedly struggled to open the vehicle doors to extract occupants following a crash.\"\n","source_date":"2025-12-24","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260527132405/https://electrek.co/2025/12/24/nhtsa-opens-probe-tesla-emergency-door-releases-following-reports-deaths/","calculation_notes":"Context-only source for the egress-design-defect caveat — NOT used in any rate calculation. NHTSA probes and the Bloomberg tally are raw incident counts (15 deaths across the whole Tesla fleet over 13 years is a very small absolute number) and cannot be converted to a per-person or per-crash multiplier. China's Ministry of Industry and Information Technology has issued a draft rule (effective 2027-01-01) requiring interior and exterior mechanical emergency releases on all passenger cars, corroborating that this is a recognised design-defect class. Cited to qualify, not quantify, the entry's \"modern vehicle = easy egress\" assumption.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Petrol car catching fire (per vehicle lifetime)","lifetime_us_adult":0.008},{"label":"Death in a house fire (lifetime, US)","lifetime_us_adult":0.0025}],"personal_factor_multipliers":[{"factor":"high-speed or rollover collision","multiplier":4,"notes":"USFA: engine-area fires are the deadliest, and severe crashes are what rupture fuel lines and electrical systems. Most ignition follows high-energy impacts, not minor fender-benders."},{"factor":"older vehicle predating modern fuel-system integrity standards","multiplier":2.5,"notes":"FMVSS 301 (fuel system integrity, phased in during the 1970s) sharply cut post-crash fuel-fed fires; pre-standard and poorly maintained older vehicles with degraded fuel lines carry materially higher risk."},{"factor":"entrapment or inability to exit after the crash","multiplier":5,"notes":"A post-crash fire is usually survivable if occupants can get out. Jammed doors, incapacitating injury, or a stuck seatbelt are what convert a vehicle fire into a fatal one."},{"factor":"modern vehicle, belted, conscious and able to exit quickly","multiplier":0.3,"notes":"Modern fuel-system integrity plus rapid egress. The fire rarely traps an uninjured, belted occupant who can open a door and walk away."}],"short_label":"Post-crash car fire","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The single biggest source of uncertainty is attribution. Crash databases record whether a fire occurred, not whether the fire was the cause of death; in a large share of fire-involved fatal crashes the occupant was killed by collision trauma and the vehicle ignited afterward, so the number of deaths the fire itself caused is smaller than the number of fatal crashes that involved fire. The headline ~1-in-25,000 figure uses the USFA collision-death count (214/yr); the NFPA-based upper bound (~359/yr) gives ~1 in 16,500. The multipliers are mechanism-based and directional rather than precisely quantified — published per-factor relative risks for post-crash fire death specifically are sparse, so they are calibrated to the qualitative findings (collision severity, engine-area origin, egress) rather than to exact effect sizes. Electric-vehicle high-voltage battery fires after severe crashes are a distinct and still-emerging risk not separated out here; the historical data are dominated by internal-combustion fuel fires. A second emerging caveat cuts against the \"a modern car makes egress easy\" assumption built into the protective factor below: electronic and flush retractable door handles can leave an uninjured, conscious occupant unable to exit when the 12-volt system dies on impact and the rear mechanical release is hidden, unlabeled, or absent. NHTSA opened a preliminary evaluation in December 2025 covering 179,071 model-year-2022 Tesla Model 3 vehicles (a driver reportedly escaped through the rear window after a fire) and a September 2025 probe of ~174,290 model-year-2021 Model Y vehicles; a Bloomberg analysis counted at least 15 deaths over 2012-2025 in crashed Teslas whose powered doors would not open, and China has drafted a rule (effective 2027) mandating mechanical interior and exterior releases. These are open investigations and lawsuits, not adjudicated causation, and the absolute counts are tiny against the whole fleet — they do not move the headline figure — but they are a documented egress-failure mode distinct from the fuel- and battery-ignition mechanisms above.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-06-11","methodology_version":"1.0"},"reviewer":"8d-eval-2026-06-11","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-06-11","image":{"alt":"A single car at rest against a roadside guardrail with a small fire-extinguisher symbol nearby, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/post-crash-vehicle-fire-death","api_url":"https://likelier.app/api/fears/post-crash-vehicle-fire-death.json"},{"slug":"snorkeling-death","question":"What are the odds of dying while snorkeling on a typical vacation trip?","quick_answer":"A typical traveller's lifetime snorkeling exposure carries roughly a 1 in 22,000 chance of death (about 7.7 deaths per million snorkeler-outings in Hawaii data) — small, but not the essentially zero risk most vacationers assume.\n","category":"sport","tags":["travel","sport"],"no_reliable_estimate":false,"perceived":{"description":"Snorkeling is widely treated as one of the safest water activities a tourist can pick: the snorkeler stays at the surface, breathes through a tube, and rarely descends more than a meter or two. No large-scale public-perception survey has isolated estimates of snorkeling fatality odds, so this entry uses editorial intuition. The general framing in travel media and tour-operator marketing emphasises ease and accessibility, not risk. The Hawaii Snorkel Safety Study and Divers Alert Network reporting describe a distinct fatality mechanism — Snorkeling-Induced Rapid Onset Pulmonary Edema (SIROPE) and cardiac events triggered by surface immersion — that is poorly understood outside of dive medicine. The mismatch between the casual public framing and a documented, distinct lethal pathway is the editorial angle.\n","rough_estimate":"most travellers likely treat snorkeling as essentially zero-risk and would not be able to name a snorkel-specific cause of death","kind":"intuition"},"native":{"display":"~23 snorkel-related deaths per year in Hawaii against ~3 million annual snorkelers (~7.7 deaths per million snorkeler-outings)","numerator":23,"denominator":3000000,"unit":"per snorkeler-outing","population":"Hawaii visitors and residents snorkeling in Hawaiian waters (2012-2021 surveillance window, Hawaii Snorkel Safety Study)"},"normalized":{"lifetime_us_adult":0.000046,"display":"~1 in 22,000 over a typical traveller's lifetime snorkeling exposure","log_value":-4.34,"assumptions":"The Hawaii Snorkel Safety Study (Foti et al., final report 2021; summary published August 2022) documented 204 snorkel-related deaths in Hawaii over the nine-year period 2012-2021, or roughly 22-23 deaths per year. The denominator of \"approximately 3 million people snorkel in Hawaiian waters each year\" (Hawaii Ocean Safety) is best read as 3 million snorkeler-trips per year — not strictly 3 million unique snorkelers and not strictly 3 million outings, because many visitors snorkel multiple times during a single Hawaii stay. Treating the Hawaii rate as ~7.7 deaths per million snorkeler-outings is the most defensible reading; if visitors average two outings each, the true per-outing rate is closer to 3.8 per million, which converges with Denoble's Australian estimate of ~5 deaths per million snorkelers from 1994-2006 DAN data. A typical US adult who snorkels at all is modelled here at 6 lifetime outings (e.g., one or two Caribbean or Hawaii vacations including 2-4 outings each). Compound probability: 1 - (1 - 7.7e-6)^6 ≈ 4.6e-5, or roughly 1 in 22,000. The scope is activity_specific_lifetime: this is the lifetime risk for a US adult who actually goes snorkeling on vacation at the modelled frequency, not for the general US adult population (a non-snorkeler has zero exposure). Hawaii is the best-surveilled jurisdiction; reported rates from Australia (DAN) and elsewhere are in the same order of magnitude, supporting the headline as a reasonable central estimate for a moderate-exposure US adult traveller.\n","uncertainty":{"low":0.0000077,"high":0.0003},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.snorkelsafetystudy.com/index.php/2022/08/16/red-flags-for-snorkelers-how-to-stop-the-quiet-deaths/","title":"Red Flags for Snorkelers: How to Stop the Quiet Deaths","publisher":"Hawaii Snorkel Safety Study (Hawaii Department of Health Snorkel Safety Sub-Committee)","source_type":"primary_study","statistic":"204 snorkel-related deaths in Hawaii over the nine years preceding the study (2012-2021); 90% (184/204) were tourists; SIROPE leads to most fatal and non-fatal snorkel-related drownings","excerpt":"\"[Paraphrase from study summary — full final report only available as a large PDF.] 204 deaths over nine years preceding the study; 90% were tourists (184 of 204); more snorkeling deaths than any other water activity in Hawaii during this period. The researchers concluded that 'SIROPE leads to some, possibly most, fatal and non-fatal snorkel-related drownings.' Full-face masks were found to pose no inherent advantage or disadvantage in terms of SIROPE. Seven predisposing factors identified: snorkel tube airway resistance; elevated left ventricle diastolic dysfunction; cardiac disease history or high blood pressure; recent prolonged air travel; snorkeling from boats; increased exertion while snorkeling; inability to touch bottom.\"\n","source_date":"2022-08-16","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260714063934/https://www.snorkelsafetystudy.com/index.php/2022/08/16/red-flags-for-snorkelers-how-to-stop-the-quiet-deaths/","calculation_notes":"204 deaths over 9 years (2012-2021) ÷ 9 = 22.67 deaths/year, rounded to 23 in the native display. Combined with the 3 million annual snorkelers figure (Hawaii Ocean Safety), this gives 23 / 3,000,000 ≈ 7.7 per million snorkeler-trips per year. Lifetime exposure for a typical snorkeling US adult (6 outings): 1 - (1 - 7.7e-6)^6 ≈ 4.6e-5 ≈ 1 in 22,000. The study did not publish a per-outing rate directly; the rate above is derived from the headline death count and the standard 3M annual snorkelers denominator widely cited by Hawaii public-health agencies.\n","independence_note":"Primary source — the Hawaii Snorkel Safety Study is the most rigorous epidemiological investigation of snorkeling fatalities ever published. Methodologically independent of DAN: the SSS used Hawaii DOH death certificate data plus a survivor survey of 131 near-fatal snorkel drowning cases; DAN uses voluntary case submissions from members worldwide. The two data streams converge on SIROPE / immersion pulmonary edema as the dominant mechanism.\n"},{"url":"https://dan.org/safety-prevention/diver-safety/divers-blog/snorkeling-deaths/","title":"Snorkeling-Related Deaths & Underlying Cardiac Causes","publisher":"Divers Alert Network (DAN)","source_type":"reputable_reference","statistic":"Average 10 snorkeling deaths per year against ~2 million annual snorkelers in 1994-2006 Australian data; approximately 5 deaths per million snorkelers; 60 of 140 deaths (43%) cardiac-related; median age 65, predominantly male","excerpt":"\"Overall, the incidence rate of snorkeling deaths is very small; with an average 10 cases per year and about 2 million snorkelers annually, it is approximately five deaths per 1 million snorkelers.\" \"Most deaths due to cardiac causes occurred in male snorkelers of a median age of 65, who were found silently floating in the water.\" \"The majority of the cases occurred due to cardiac-related causes (60) or drowning while at the surface (33).\" Of the 140 total cases analysed, prolonged breath-hold diving accounted for 19, trauma 10, and epileptic seizures 4.\n","source_date":"2013-10-08","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260122163914/https://dan.org/safety-prevention/diver-safety/divers-blog/snorkeling-deaths/","calculation_notes":"Independent cross-check on Hawaii rate. 10 deaths / 2,000,000 snorkelers = 5 per million annually, vs Hawaii's 23 / 3,000,000 ≈ 7.7 per million. The two estimates are within a factor of ~1.5 of each other and well within typical surveillance noise for activity-specific mortality, given that Hawaii includes more older tourists post-flight while the Australian data spans a broader resident-and-visitor mix.\n","independence_note":"Independent surveillance system: DAN voluntary incident reporting database covering 1994-2006 Australian snorkeling fatalities, analysed by Petar Denoble (DAN VP of Research). Methodologically independent of Hawaii DOH death-certificate surveillance and uses a different geography, different time window, and different reporting pathway. Convergence with Hawaii data on the per-million rate strengthens both.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6526048/","title":"Immersion pulmonary oedema: a cardiological perspective","publisher":"Diving and Hyperbaric Medicine (via PubMed Central / PMC)","source_type":"peer_reviewed","statistic":"Immersion pulmonary edema mechanism: pulmonary centroid 8-10 cm below water surface creates 8-10 cm H2O negative-pressure gradient; ultrasound lung comet score 15.1 with negative-pressure breathing after exercise vs 4.2 with positive pressure; hypertension is most frequent cardiovascular predisposing condition","excerpt":"\"Swimmers and snorkellers are negative pressure breathers because their lung centroid is below the surface of the water.\" \"During exercise with negative pressure breathing, lung fluid accumulation measured 15.1 on ultrasound comet scoring versus 4.2 with positive pressure.\" Hypertension is \"the most frequent cardiovascular disease predisposing to IPE.\" Recreational swimmers/snorkellers presenting with IPE averaged 47.8 years of age compared with 23.3 years in military swimmers, reflecting the cardiovascular age skew of community cases.\n","source_date":"2019-03-31","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20250429053808/https://pmc.ncbi.nlm.nih.gov/articles/PMC6526048/","calculation_notes":"Mechanism source — does not contribute to the per-outing rate calculation. Provides the peer-reviewed physiological basis for why snorkeling has a distinct cardiac/pulmonary fatality mechanism beyond ordinary drowning, explaining why the public framing of snorkeling as \"just breathing through a tube\" underestimates the load on the cardiovascular system.\n","independence_note":"Peer-reviewed cardiology synthesis by P.T. Wilmshurst (Royal Stoke University Hospital). Independent of both Hawaii SSS and DAN, drawing on European IPE case series. Provides the physiological mechanism that the epidemiological surveillance sources describe as \"SIROPE\" and \"cardiac causes\" — confirming both reporting systems are likely capturing the same underlying pathway under different labels.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK499853/","title":"Immersion Pulmonary Edema (StatPearls)","publisher":"StatPearls Publishing / National Center for Biotechnology Information (NCBI)","source_type":"peer_reviewed","statistic":"IPE incidence ~1.1% in European case studies; fatalities most commonly involve females over the age of 50; predisposing factors include cold water, overhydration, asthma, underlying cardiac pathology, hypertension, and physical exertion","excerpt":"\"Immersion pulmonary edema (IPE) is used as an umbrella term for scuba divers pulmonary edema (SDPE) and swimming-induced pulmonary edema (SIPE).\" Risk factors: \"overhydration, cold water, negative inspiratory pressure, asthma, diabetes, beta-blockers, underlying cardiac pathology, physical exertion, previous IPE episodes, and systemic/pulmonary hypertension.\" Fatalities \"most commonly involve females over the age of 50.\"\n","source_date":"2023-07-17","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20220615123147/http://www.ncbi.nlm.nih.gov/books/NBK499853/","calculation_notes":"Mechanism and risk-factor source — does not contribute to rate calculation but establishes that IPE is a peer-reviewed clinical entity with documented age and cardiovascular risk profile. The 1.1% incidence figure refers to combat-swimmer trainee populations, not casual snorkelers, and is not portable to vacation snorkeling rates.\n","independence_note":"Peer-reviewed clinical reference (StatPearls, indexed via NCBI Bookshelf and updated annually by board-certified authors). Independent of all three other sources and provides the broader clinical taxonomy that situates SIROPE within the IPE family.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7320e1.htm","title":"Vital Signs: Drowning Death Rates, Self-Reported Swimming Skill, Swimming Lesson Participation, and Recreational Water Exposure — United States, 2019-2023","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"4,067-4,677 unintentional drowning deaths per year in the US (2019-2022); rate 1.2-1.4 per 100,000; ages ≥65 had second-highest rates after children ages 1-4","excerpt":"\"Unintentional drowning death rates were significantly higher during 2020, 2021, and 2022 compared with those in 2019.\" 2019: 4,067 deaths, 1.2/100,000; 2020: 4,589 deaths, 1.4/100,000; 2021: 4,677 deaths, 1.4/100,000; 2022: 4,509 deaths, 1.3/100,000. \"The highest drowning rates were among non-Hispanic American Indian or Alaska Native and non-Hispanic Black or African American persons\" and among children aged 1-4 and adults aged ≥65.\n","source_date":"2024-05-23","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260512224832/https://www.cdc.gov/mmwr/volumes/73/wr/mm7320e1.htm","calculation_notes":"Used to anchor the comparison_anchors entry for all-cause US drowning lifetime risk and to establish that ages ≥65 are the second-highest drowning risk group, consistent with the Hawaii and DAN snorkeling-death age skew. ~4500 drowning deaths per year × 79-year life expectancy ÷ 333 million US population ≈ 1.07% lifetime; cited as ~0.0012 in the comparison anchors using the conventional published lifetime figure.\n","independence_note":"US government surveillance (NCHS Vital Statistics aggregated by CDC), methodologically independent of the snorkel-specific sources and used only as a comparator for the overall drowning anchor.\n"}],"comparison_anchors":[{"label":"All-cause US drowning (lifetime, US adult)","lifetime_us_adult":0.0012},{"label":"Death during recreational SCUBA (lifetime, ~100 dives over a career)","lifetime_us_adult":0.0005},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Lightning strike fatality (lifetime, US adult)","lifetime_us_adult":0.000065}],"personal_factor_multipliers":[{"factor":"Hawaii visitor (vs Hawaii resident)","multiplier":10,"notes":"Per the Hawaii Snorkel Safety Study FAQ: \"Visitors to Hawai'i are 10 times more likely than residents to drown while snorkeling.\" The mechanism attributed by the SSS authors is a combination of unfamiliarity with the conditions, recent long-haul air travel reducing pulmonary reserve, and higher mean age among tourists than residents who snorkel. This is the single most load-bearing personal factor in the dataset and is the reason the headline rate is so much higher than baseline ocean-recreation drowning risk for the average US adult.\n"},{"factor":"Age 60+ with any cardiovascular history (hypertension, ischemic heart disease, prior MI)","multiplier":5,"notes":"Hawaii SSS identified \"cardiac disease history or high blood pressure\" and \"elevated left ventricle diastolic dysfunction\" as two of the seven predisposing factors for SIROPE. DAN's surveillance data found cardiac deaths concentrated in male snorkelers with a median age of 65. StatPearls IPE chapter identifies hypertension as \"the most frequent cardiovascular disease predisposing to IPE\" and notes fatalities most commonly involve individuals over 50. Wilmshurst's cardiological perspective places the mean age of community IPE cases at 47.8 years. The multiplier is the product of an age effect (~2-3×) and a cardiovascular-disease effect (~2-3×), capped at 5× given the underlying small absolute risk.\n"},{"factor":"Recent long-haul air travel (within ~24 hours of snorkeling)","multiplier":1.8,"notes":"The Hawaii Snorkel Safety Study identified \"recent prolonged air travel\" as one of the seven predisposing factors for SIROPE, hypothesising that long-haul flight contributes to a transient reduction in pulmonary reserve and increases the likelihood of fluid translocation under negative-pressure breathing. The SSS notes \"the majority of travellers have flown at least five hours to reach Hawaii,\" meaning the headline Hawaii rate already incorporates this factor for most cases. The multiplier here is relative to a baseline snorkeler who has not recently flown; the SSS itself flagged the air-travel link as \"not conclusive\" but as the most plausible explanation for the visitor-resident gap not otherwise accounted for by age and cardiovascular status.\n"},{"factor":"Female sex (for IPE-specific mortality, ages 50+)","multiplier":1.5,"notes":"StatPearls IPE chapter states that IPE fatalities \"most commonly involve females over the age of 50.\" This contrasts with the DAN/Denoble cardiac- cause data, which were predominantly male (median age 65). The two patterns reflect different fatality pathways: cardiac arrest skews male and older, while pure SIROPE/IPE-attributed deaths skew female and middle-aged. The aggregate snorkeling fatality population in Hawaii is male-skewed, but a 50+ female snorkeler is at materially elevated risk of the specific IPE pathway.\n"},{"factor":"Snorkeling from a boat (vs from shore in shallow, footable water)","multiplier":1.4,"notes":"Hawaii SSS lists \"snorkeling from boats\" and \"inability to touch bottom\" as two distinct predisposing factors. The mechanism is two-fold: deeper water makes self-rescue impossible if the snorkeler becomes incapacitated (consistent with the \"silent\" death pattern documented by DAN — victims are typically found floating face-down with no distress signal); and boat-based snorkeling is more often associated with longer in-water durations and unfamiliar surroundings. The multiplier is a conservative estimate; the SSS did not publish a quantified hazard ratio.\n"}],"short_label":"Snorkeling death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The Hawaii Snorkel Safety Study is the best-surveilled jurisdiction in the world for this risk, and the 23/year × 3M-snorkeler ratio is built from a single state's death certificates against a state-tourism-industry-derived denominator that may bundle multi-outing visitors. Genuine per-outing rates could be as low as ~2.5 per million if visitors average three outings per trip, or as high as ~10 per million if the SSS undercounts unreported near-fatal events that the survivor survey captured. Outside Hawaii, surveillance is fragmented: DAN's voluntary reporting captures only members and is structurally biased toward the dive-trained subpopulation, not casual vacation snorkelers. Caribbean, Mexican, and Southeast Asian destinations have no comparable surveillance, so the global per-outing rate is essentially unknown — if anything, the Hawaii estimate is likely toward the upper end of the world distribution because the older-tourist demographic and the long-haul-flight load are both larger for Hawaii than for nearby Caribbean destinations. The 6-outings lifetime exposure assumption is mid-range; a US adult who never snorkels has zero risk, an enthusiast doing 30+ outings has proportionally higher cumulative risk (compound to ~2.3e-4, the upper bound of the uncertainty range). SIROPE/IPE is poorly recognised in autopsy practice outside specialised dive medicine, so the share of \"drowning\" deaths that are mechanistically pulmonary edema is uncertain; the Hawaii SSS concluded \"some, possibly most\" — a deliberately wide qualifier. Full-face masks were widely suspected by media to be the cause but the SSS explicitly found \"no inherent advantage or disadvantage in terms of SIROPE,\" so they are not included as a personal factor here despite the popular narrative.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-23","last_reviewed":"2026-05-23","reviewed":true,"generated_at":"2026-05-23","image":{"alt":"A single dive mask and snorkel resting on a pale neutral surface beside a folded towel, flat vector illustration in muted blue-grey tones, no people, no water."},"canonical_url":"https://likelier.app/snorkeling-death","api_url":"https://likelier.app/api/fears/snorkeling-death.json"},{"slug":"railroad-grade-crossing-death","question":"What are the odds of dying at a railroad level crossing?","quick_answer":"The odds of dying at a US railroad crossing are about 1 in 20,600 over a lifetime, roughly 1 in 1,210,000 in a given year. Most drivers barely register crossings as a statistical risk, though they do underestimate how often trains actually use them.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Railroad crossings occupy an odd corner of driver perception: most people have crossed them hundreds of times without incident, which breeds a kind of ambient complacency. The gates come down, the lights flash, and the train passes — a ritual that feels mechanical rather than dangerous. Yet when a crash does happen at a crossing it is violent and almost always fatal, which seeds a residual dread that surfaces whenever someone tells a crossing horror story. The dominant cognitive error is not overestimating the risk but misreading the operational reality: drivers systematically underestimate how many trains use a given crossing per day (the true figure is typically two to ten times higher than intuition suggests) and misjudge train speed and distance, leading some to take gaps that are genuinely too small. The abstract probability, however, is rarely in conscious awareness — most drivers neither fear nor think about crossings as a statistical risk category in the way they might think about highway speed or alcohol.\n","kind":"intuition"},"native":{"display":"~1 in 1,210,000 per year","numerator":276,"denominator":335000000,"unit":"per year","population":"US general population (2022)"},"normalized":{"lifetime_us_adult":0.0000486,"display":"~1 in 20,600 lifetime","log_value":-4.31,"assumptions":"276 highway-rail grade crossing fatalities recorded by the Federal Railroad Administration (FRA) and Operation Lifesaver for 2022 (the FRA-revised final figure, up from a provisional 274), divided by the 2022 US population of approximately 335 million, gives an annual rate of 8.24×10⁻⁷ per person. Multiplied over a 59-year adult life horizon (age 18–77): 1 – (1 – 8.24×10⁻⁷)^59 ≈ 4.86×10⁻⁵, or roughly 1 in 20,600. This figure covers all grade-crossing fatalities (vehicle occupants and pedestrians at designated crossing points) but excludes trespass deaths (people struck on open track away from crossings), which add roughly 600–650 additional railroad fatalities per year. The 10-year trend is modestly downward: crossing collision rates are approximately 25% lower than in 2000, though 2024 NSC data show a 7% uptick in crossing fatalities from 2023. The lifetime estimate is treated as a population-average figure; individual risk varies substantially with how many crossings a person uses and whether those crossings have active warning devices (gates and lights) versus passive signage only.\n","uncertainty":{"low":0.0000442,"high":0.0000513},"scope":"us_adult_lifetime"},"sources":[{"url":"https://oli.org/track-statistics/collisions-casualties-year","title":"Collisions & Casualties by Year","publisher":"Operation Lifesaver (federally chartered nonprofit using FRA data)","source_type":"govt_report","statistic":"2,222 highway-grade crossing collisions in 2022; 276 killed; 879 injured","excerpt":"\"All Highway-Rail Incidents at Public and Private Crossings, 2026–1981 ... Year Collisions Fatalities Injuries ... 2022 2,222 276 879\" (Source: Federal Railroad Administration)\n","source_date":"2023-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260521062116/https://oli.org/track-statistics/collisions-casualties-year","calculation_notes":"276 crossing deaths (FRA-revised final 2022 figure) ÷ 335,000,000 US population = 8.24×10⁻⁷ annual rate per person. Over 59 adult years: 1 – (1 – 8.24×10⁻⁷)^59 ≈ 4.86×10⁻⁵ (≈ 1 in 20,600), within the stated uncertainty band. Operation Lifesaver is a federally chartered nonprofit whose statistics are drawn directly from FRA mandatory accident reporting and are the standard public-facing source for grade-crossing fatality data; the 2022 row was revised upward from 274 to 276 deaths in a subsequent FRA annual revision.\n","independence_note":"Operation Lifesaver data are derived from FRA mandatory reporting; the NSC and BTS figures below draw on the same FRA database, so these sources share a common upstream data source but are editorially independent in how they present and contextualize the figures.\n"},{"url":"https://injuryfacts.nsc.org/home-and-community/safety-topics/railroad-deaths-and-injuries/","title":"Railroad Deaths and Injuries","publisher":"National Safety Council — Injury Facts","source_type":"reputable_reference","statistic":"Railroad deaths totaled 967 in 2023; 27% at crossings ≈ 261 crossing deaths; highway-rail crossing fatalities increased 7% from 2023 to 2024","excerpt":"\"Railroad deaths totaled 954 in 2024, a 1% decrease from the 2023 revised total of 967. Fatalities at highway-rail crossings increased 7% from 2023 to 2024.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260225085532/https://injuryfacts.nsc.org/home-and-community/safety-topics/railroad-deaths-and-injuries/","calculation_notes":"NSC reports crossing fatalities as approximately 27% of total railroad deaths. Applied to the 2023 total of 967 yields ~261 crossing deaths for 2023, consistent with the 2022 OLI figure of 276. The 7% increase in 2024 crossing fatalities is a recent uptick against a longer downward trend. NSC uses FRA data as its primary source.\n","independence_note":"NSC is editorially independent from Operation Lifesaver and FRA, though all three draw on the same FRA mandatory accident reporting database. NSC's annual Injury Facts compilation adds epidemiological context absent from the raw FRA counts.\n"},{"url":"https://www.utrgv.edu/railwaysafety/_files/documents/reports/drivers_perceptions_hrgc_safety_project_final_report_021017.pdf","title":"Drivers' Perceptions of Highway-Rail Grade Crossing Safety and Their Behavior","publisher":"University of Texas Rio Grande Valley Railway Safety Institute (FRA-funded research)","source_type":"primary_study","statistic":"All 18 drivers interviewed underestimated train frequency; actual train crossings are typically 2–3× and sometimes 10× more than drivers expected; 47% of drivers consider crossings a significant hazard above normal driving, 46% do not","excerpt":"\"All 18 drivers interviewed underestimated the frequency of train crossings per day; the actual train crossings are typically two to three times as many as drivers expected and sometimes are 10 times more than expected.\"\n","source_date":"2017-02-10","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20240428123557/https://www.utrgv.edu/railwaysafety/_files/documents/reports/drivers_perceptions_hrgc_safety_project_final_report_021017.pdf","calculation_notes":"This FRA-funded qualitative and survey study provides the perception data behind the overconfidence framing. The finding that drivers underestimate train frequency by 2–10× is the mechanistic explanation for why crossing violations persist even among drivers who are not impaired or distracted: the rare-train-day heuristic does not update on the base rate.\n","independence_note":"UTRGV Railway Safety Institute is an FRA-funded academic center; the research is independent of Operation Lifesaver and NSC, providing behavioral and attitudinal data rather than fatality counts.\n"}],"comparison_anchors":[{"label":"Motor vehicle crash death (lifetime, US adult)","lifetime_us_adult":0.009},{"label":"Trespassing on railroad tracks (annual rate, per person)","lifetime_us_adult":0.000115},{"label":"Commercial airline crash death (lifetime)","lifetime_us_adult":0.000011}],"personal_factor_multipliers":[{"factor":"Passive crossing (crossbuck only, no gates or lights)","multiplier":5,"notes":"Active warning devices (gates + lights) are associated with substantially lower collision and fatality rates than passive crossings. The FRA reports that the majority of crossing fatalities occur at crossings with active warning devices that were ignored, but passive crossings have a higher per-crossing collision rate."},{"factor":"Rural location","multiplier":3,"notes":"Rural crossings have lower traffic volumes but higher crash severity; longer emergency response times increase fatality rates. FRA data consistently show rural crossings overrepresented in fatalities relative to their traffic share."},{"factor":"Driving around lowered gates","multiplier":20,"notes":"Circumventing active warning devices is the single most dangerous behavior; gate-circumvention crashes are nearly universally fatal for the vehicle occupant."},{"factor":"Impairment (alcohol or drugs)","multiplier":4,"notes":"Impaired drivers are overrepresented in crossing fatalities relative to their share of the driving population, consistent with impairment's role in misjudging train speed and distance."}],"short_label":"Railroad crossing death","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 276-fatality figure covers highway-rail grade crossings only — designated intersection points where roads cross tracks. It excludes trespass deaths (people struck on open track), which add roughly 600–650 additional railroad fatalities per year and represent a distinct risk category with different behavioral drivers. The population-average lifetime figure (1 in ~20,600) masks large individual variation: someone who commutes daily across an uncontrolled rural crossing faces meaningfully higher risk than someone in a city with no grade crossings on their routes. The 2024 NSC uptick in crossing fatalities (7% over 2023) is notable against a decade-long downward trend and may reflect post-pandemic traffic pattern changes; longer-term trend is still toward lower rates. Pedestrian fatalities at public crossings (87 in 2023 per AAR/FRA data) are included in the headline figure and represent a separate behavioral risk pattern from vehicle occupants.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A single railroad crossing gate lowered across an empty road, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/railroad-grade-crossing-death","api_url":"https://likelier.app/api/fears/railroad-grade-crossing-death.json"},{"slug":"flood-death","question":"What are the odds of being killed in a flood?","quick_answer":"Being killed in a flood works out to roughly 1 in 20,000 over a global adult lifetime, a whole-planet average that is nearly meaningless for any one person. About 32% of US adults say they fear a devastating flood; the real danger is behavioral, since people routinely underestimate the force of moving water and drive into it.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Floods are the deadliest thunderstorm-related hazard in the US and the most frequent natural disaster globally, but we have not found a rigorous recent survey that isolates \"fear of being killed in a flood\" from general severe-weather or climate anxiety. We mark the perceived side as editorial intuition. Anecdotally, the gap that matters here is not geographic — it is behavioral. The single most consistent finding in the US flood mortality literature is that people routinely underestimate the force of moving water and drive into it; \"turn around, don’t drown\" exists precisely because the common-sense prior (\"it’s only a few inches\") is wrong.\n","rough_estimate":"31.7% of US adults report being afraid or very afraid of a devastating flood (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~6,600 global flood deaths per year (WHO Bulletin, 1990-2022 EM-DAT window)","numerator":6600,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.0000495,"display":"1 in ~20,000 lifetime (global adult average)","log_value":-4.31,"assumptions":"Uses ~6,600 global flood deaths per year as the smoothed long-window average, anchored on the WHO Bulletin analysis of EM-DAT records for 1990-2022 (218,353 total deaths across 4,713 recorded flood events in 168 countries, ≈ 6,617/year). Jonkman’s 1975-2002 analysis gives a comparable ~6,500/year from a different compilation of the same underlying disaster database. Annual per-capita risk ≈ 6,600 / 8,000,000,000 ≈ 8.25e-7; compounded over 60 adult life-years ≈ 4.95e-5, which we display as ~1 in 20,000 global adult lifetime. The window matters: the long-run average is dominated by rare megaevents (the 1931 China floods alone are estimated at 1-4 million deaths, and modern events like the 1998 Bangladesh floods, 2010 Pakistan floods, and 2022 Pakistan floods each killed thousands), while the post-2000 smoothed average is closer to 5,000-7,000 per year. The uncertainty band below brackets the window-sensitivity rather than sampling noise, and the headline is an average-global-adult figure that is essentially meaningless for any individual — see the regional breakdown and caveats.\n","uncertainty":{"low":0.000018,"high":0.0001},"scope":"global_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11132161/","title":"Global, regional and national trends and impacts of natural floods, 1990-2022","publisher":"Bulletin of the World Health Organization (Liu Q, Du M, Wang Y, Deng J, Yan W, Qin C, Liu M, Liu J)","source_type":"peer_reviewed","statistic":"4,713 floods recorded in 168 countries between 1990 and 2022 resulted in 218,353 deaths and more than US$1.3 trillion in economic damages; South-East Asia Region had 71,713 deaths (32.84%), Region of the Americas 48,630 (22.27%), Western Pacific Region 42,721 (19.57%), Eastern Mediterranean 29,819, Africa 19,927, Europe 5,543.","excerpt":"\"Between 1990 and 2022, 4713 floods were recorded in 168 countries, which resulted in 218 353 deaths and caused more than US$ 1.3 trillion in economic damages. Of these, the South-East Asia Region had the highest number (71 713; 32.84%) followed by the Region of the Americas (48 630; 22.27%).\"\n","source_date":"2024-06-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20251028191932/https://pmc.ncbi.nlm.nih.gov/articles/PMC11132161/","calculation_notes":"WHO Bulletin total of 218,353 deaths / 33 years ≈ 6,617 deaths per year, which we round to ~6,600 as the headline native figure. Annual per-capita risk ≈ 6,600 / 8,000,000,000 ≈ 8.25e-7; compounded over 60 adult years ≈ 4.95e-5 ≈ 1 in 20,200, which we display as ~1 in 20,000. The regional breakdown in the paper (South-East Asia and Americas together accounting for 55% of deaths, Europe contributing only 2.5%) is the empirical basis for the regional_breakdown rows below.\n","independence_note":"The WHO Bulletin analysis draws on EM-DAT (CRED) records, the same underlying disaster database used by Jonkman (2005) and by most modern flood mortality research. Treat the two peer-reviewed sources as methodologically independent compilations of overlapping source data — they agree on the order of magnitude (~6,500-6,600 deaths/year) across very different time windows, which is the main reason we use that number rather than a shorter-window mean.\n"},{"url":"https://link.springer.com/article/10.1007/s11069-004-8891-3","title":"Global Perspectives on Loss of Human Life Caused by Floods","publisher":"Natural Hazards (Jonkman, S.N.)","source_type":"peer_reviewed","statistic":"Over the 27-year period studied, more than 175,000 people died and close to 2.2 billion were affected directly by floods worldwide; flash floods produce the highest average mortality per event, and Asian river floods dominate absolute casualty counts.","excerpt":"\"Over 27 years, more than 175,000 people died and close to 2.2 billion were affected directly by floods worldwide. ... Flash floods result in the highest average mortality per event. ... On a worldwide scale Asian river floods are most significant in terms of number of persons killed and affected.\"\n","source_date":"2005-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250220215527/https://link.springer.com/article/10.1007/s11069-004-8891-3","calculation_notes":"Jonkman’s 175,000 deaths / 27 years ≈ 6,500/year — within 2% of the WHO Bulletin figure despite covering a different window (1975-2002 vs 1990-2022) and using a different compilation of the EM-DAT records. Jonkman is the source traditionally cited for the \"flash floods have the highest per-event mortality\" and \"Asian river floods dominate global casualty counts\" claims, both of which shape the regional_breakdown and body text here.\n","independence_note":"Jonkman (2005) and Liu et al (2024) both draw ultimately from EM-DAT / CRED disaster records. Treated as methodologically independent because of the different time windows, different aggregation choices, and different research groups; their agreement on order of magnitude is the main quantitative anchor for the headline.\n"},{"url":"https://www.weather.gov/safety/flood-turn-around-dont-drown","title":"Turn Around Don't Drown — Flood Safety","publisher":"NOAA National Weather Service","source_type":"govt_report","statistic":"Per CDC data cited by NWS, over half of all US flood-related drownings occur when a vehicle is driven into hazardous flood water. 6 inches of moving water can knock an adult off their feet; 12 inches of moving water can carry away most cars; 24 inches can sweep away SUVs and trucks.","excerpt":"\"According to the Centers for Disease Control and Prevention, over half of all flood-related drownings occur when a vehicle is driven into hazardous flood water. ... Six inches of fast-moving water can knock over an adult. It takes just 12 inches of rushing water to carry away most cars and just 2 feet of rushing water can carry away SUVs and trucks.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260423042612/https://www.weather.gov/safety/flood-turn-around-dont-drown","calculation_notes":"Used to anchor the US behavioral-multiplier story rather than the global normalized figure. The NWS / CDC \"over half\" framing is the evidence base for the personal_factor_multipliers entry on driving into flooded roadways and for the \"turn around, don’t drown\" paragraph in the body text. NWS reports an approximate long-run average near 88 US flood deaths per year across its 30-year window, which gives a US lifetime baseline of ~88/333M × 60 ≈ 1.6e-5 ≈ 1 in 63,000 — an order of magnitude lower than the global figure because US flood mortality is dominated by behavioral (vehicle) rather than exposure (coastal-delta inundation) pathways, and because US warning and evacuation infrastructure is unusually mature.\n","independence_note":"NWS hazard-statistics pages and CDC mortality reporting are partially overlapping (NWS Storm Data is one input CDC uses to classify weather-related deaths), so the two figures should be treated as a single US authoritative chain rather than two fully independent estimates.\n"},{"url":"https://www.nesdis.noaa.gov/news/noaa-satellites-inform-warning-the-texas-hill-country-floods","title":"NOAA Satellites Inform Warning for the Texas Hill Country Floods","publisher":"NOAA National Environmental Satellite, Data, and Information Service (NESDIS)","source_type":"govt_report","statistic":"In the early hours of 4 July 2025 a devastating flash flood occurred in the Central Texas Hill Country in which more than a hundred people lost their lives and thousands of residents' homes were destroyed — by toll the deadliest US rainfall-driven flash flood in roughly half a century.","excerpt":"\"In the early hours of July 4, 2025 a devastating flash flood occurred in the Central Texas Hill Country. ... more than a hundred people lost their lives and thousands of residents' homes were destroyed.\"\n","source_date":"2025-07-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260118172627/https://www.nesdis.noaa.gov/news/noaa-satellites-inform-warning-the-texas-hill-country-floods","calculation_notes":"Not used to change the headline normalized figure — it is a single high-fatality US event that does not shift the long-run global per-capita average. It is cited as a current data point reinforcing the nocturnal flash-flood subgroup discussed in the caveats and the flash-flood personal_factor_multipliers entry: the predawn timing, short warning lead time, and exposure of sleeping victims on a low-lying floodway are the same failure modes Jonkman (2005) identifies as producing the highest per-event mortality fraction. The toll (\"more than a hundred\"; ~138-139 across Kerr, Travis, Burnet, Williamson and Tom Green counties per contemporaneous reporting) makes it the deadliest US rainfall-driven flash flood since the 1976 Big Thompson Canyon flood.\n","independence_note":"NOAA/NWS post-event analyses of the July 2025 Texas flood share the same federal weather-service reporting chain as the Turn Around Don't Drown / NWS US mortality figures above; treat as one US authoritative chain, not an independent estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3302549/","title":"Reduced death rates from cyclones in Bangladesh: what more needs to be done?","publisher":"Bulletin of the World Health Organization (Haque U, Hashizume M, Kolivras KN, Overgaard HJ, Das B, Yamamoto T)","source_type":"peer_reviewed","statistic":"Cyclone-related death tolls in Bangladesh fell from roughly 500,300 in the 1970 Bhola cyclone to 138,958 in a 1991 cyclone to 4,234 in 2007's Cyclone Sidr, a more-than-hundredfold decline credited to early warning systems, cyclone shelters, evacuation plans, and coastal embankments.","excerpt":"\"The main factors responsible for these reduced fatalities and injuries are improved defensive measures, including early warning systems, cyclone shelters, evacuation plans, coastal embankments, reforestation schemes and increased awareness and communication.\"\n","source_date":"2012-02-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260520144443/https://pmc.ncbi.nlm.nih.gov/articles/PMC3302549/","calculation_notes":"Supplies the evidentiary basis for the Bangladesh mortality-decline paragraph in the body text. Death tolls of roughly 500,300 (1970 Bhola cyclone), 138,958 (1991 cyclone), and 4,234 (2007 Cyclone Sidr) represent more than a hundredfold reduction over less than four decades, which the paper attributes to early warning systems, cyclone shelters, evacuation plans, and coastal embankments, not any reduction in storm intensity. Cyclone-driven storm-surge flooding is the dominant flood-fatality pathway in coastal Bangladesh, so this is used as the evidence for the \"mortality has declined\" thesis in the body text, distinct from the river-flood events (1998, 2010, 2022) discussed elsewhere in this entry.\n","independence_note":"Published in the same journal (Bulletin of the WHO) as the Liu et al. 2024 global flood analysis above but a different research group studying a different, narrower population (Bangladesh cyclone mortality specifically, not global flood mortality); treat as independent of the other sources here.\n"}],"comparison_anchors":[{"label":"Death in a tsunami (global adult lifetime)","lifetime_us_adult":0.00001},{"label":"Death in an earthquake (global adult lifetime)","lifetime_us_adult":0.000263},{"label":"Death in a hurricane / tropical cyclone (global adult lifetime)","lifetime_us_adult":0.000112},{"label":"Death by tornado (US adult lifetime, national average)","lifetime_us_adult":0.0000124},{"label":"Death in a car crash (US lifetime)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.0000495,"notes":"WHO Bulletin 1990-2022 smoothed annual deaths ÷ 8B × 60-year adult life."},{"region":"Bangladesh / Pakistan / flood-prone South Asia","probability":0.001,"notes":"Low-lying deltas, monsoon river systems, and extreme single-event tolls (1998 Bangladesh, 2010 and 2022 Pakistan each killed thousands); the 1931 China floods alone are estimated at 1-4 million deaths, illustrating the megaevent dominance of the long-run record."},{"region":"US average","probability":0.00002,"notes":"~88 flood deaths per year across a 333M-person country, of which more than half involve vehicles driven into flooded roadways — the behavioral story dominates the US headline."},{"region":"Sub-Saharan Africa","probability":0.0001,"notes":"Lower absolute numbers than South Asia but a high mortality fraction per event, driven by limited warning infrastructure and high building vulnerability."}],"personal_factor_multipliers":[{"factor":"drives vehicle into flooded roadway","multiplier":100,"notes":"More than half of US flood-related drownings involve a vehicle driven into hazardous flood water (CDC/NWS). 12 inches of moving water can float most cars; 24 inches can sweep away SUVs and trucks."},{"factor":"lives in 100-year floodplain without mitigation","multiplier":10,"notes":"Sustained exposure to river-flood return-period events; mitigation (levees, elevation, flood insurance adoption) can drop the multiplier back toward baseline."},{"factor":"flash-flood-prone canyon / low-lying urban area","multiplier":5,"notes":"Flash floods produce the highest per-event mortality fraction in the Jonkman analysis; short warning times and high water velocities are the dominant failure modes. The July 2025 nocturnal Guadalupe River flash flood in Kerr County, TX — the deadliest US rainfall-driven flash flood in roughly half a century (NOAA NESDIS) — is a recent example, with most victims caught asleep at night."},{"factor":"Residence in manufactured or mobile home","multiplier":4,"notes":"FEMA flood fatality data: manufactured and mobile homes have far lower structural integrity against flood inundation and water-driven debris impact than site-built housing; FEMA flood mortality analyses consistently identify manufactured housing as a high-risk category, with occupants roughly 4x more likely to be killed in the same flood event than occupants of site-built homes due to structural failure and lack of elevated foundation"},{"factor":"Age 65+ with limited mobility","multiplier":3,"notes":"CDC and FEMA flood fatality reviews: older adults face higher flood mortality due to reduced physical capacity to self-evacuate, higher rates of living alone, greater likelihood of residing in lower-mobility housing, and higher baseline medical vulnerability; individuals aged 65+ represent a disproportionate share of direct flood fatalities in US events including Hurricanes Katrina and Harvey, with multiplier estimated from FEMA and academic flood mortality analyses"}],"short_label":"Flood","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker, not a personal estimate. Flood mortality is heavily concentrated in South Asian and East Asian river deltas, coastal Southeast Asia, and flash-flood-prone terrain in low- and middle-income countries — per-capita lifetime risk in those settings is one to two orders of magnitude above the global average. In high-income countries, where exposure is lower and warning and evacuation infrastructure is mature, the residual risk is dominated by a single behavioral pathway: driving into flooded roadways. The US 30-year record, in which more than half of flood drownings involve vehicles, is the clearest example of a fatality profile that is almost entirely preventable at the margin. \"Turn around, don’t drown\" is one of the most evidence-supported safety messages on the site because the specific behavior it targets accounts for the majority of the specific deaths it is trying to prevent. The other recurring US fatal mechanism is the nocturnal flash flood: in the early hours of 4 July 2025 a flash flood on the Guadalupe River in Kerr County and surrounding Central Texas killed more than a hundred people — by toll the deadliest US rainfall-driven flash flood in roughly half a century, and one of only a handful since the late 20th century to kill 100 or more. Many victims were asleep when the river rose, including campers at a low-lying summer camp on the floodway, which is the pattern flash-flood mortality keeps repeating: short warning lead time, high water velocity, and exposure at night when evacuation is slowest.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized water-level line rising against a pale sky with the top of a simple house shape just visible, flat vector illustration in muted blues and grey."},"canonical_url":"https://likelier.app/flood-death","api_url":"https://likelier.app/api/fears/flood-death.json"},{"slug":"pet-transport-death","question":"What are the odds of a pet dying during air cargo transport?","quick_answer":"A pet placed in airline cargo dies in roughly 1 of 20,000 flights (about 1 in 16,000 in 2024); counting injuries and losses roughly doubles the incident rate to around 1 in 12,000. Worried owners' guesses (1 in 1,000 to 1 in 10,000) are close, though the gap between breeds is large.\n","category":"other","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Pet owners tend to worry a lot about putting a dog or cat into an aircraft cargo hold, and a handful of high-profile incidents (the 2018 United French bulldog puppy death, the annual \"worst airline for pets\" headlines) anchor that fear. We have not found a standalone survey isolating \"fear of your pet dying in air cargo\", so perceived risk is marked as editorial intuition. The rough sense most worried owners carry — that the risk is real but not everyday — is actually close to the numbers. What people usually get wrong is not the average rate, it's the enormous gap between breeds.\n","rough_estimate":"most worried owners guess somewhere between 1 in 1,000 and 1 in 10,000 per flight","kind":"intuition"},"native":{"display":"~1 in 16,000 per pet-flight (US airlines, 2024)","numerator":10,"denominator":161335,"unit":"per pet-flight","population":"animals transported by US airlines reporting to the DOT, calendar year 2024"},"normalized":{"lifetime_us_adult":0.00005,"display":"~1 in 20,000 per pet-flight (US airline cargo, recent years)","log_value":-4.3,"assumptions":"Reference subgroup: one pet (dog, cat, or other companion animal) placed on a single US airline flight as reported through the DOT Air Travel Consumer Report animal incident system. Headline figure uses the DOT full-year 2024 numbers: 10 animal deaths across 161,335 animals transported by reporting US carriers, for a per-flight death rate of roughly 1 in 16,000. The two DOT full-year releases that break out a death-only count — 2021 (seven animal deaths across 256,114 animals transported) and 2024 (10 deaths across 161,335 transported) — together give 17 deaths across roughly 417,000 pet-flights, or about 1 in 24,600; the 2023 release published only a combined total of nine animal incidents across 124,593 transported (a 0.72-per-10,000 incident rate) without a separate death count. Averaging across these recent years, the per-flight death rate lands in the low-to-mid tens of thousands, and 1 in 20,000 is used as a round central point estimate that sits inside the year-to-year range. The scope is declared activity_specific_lifetime because this is per-pet-per-flight risk for a specific activity, not a general-population lifetime risk, and it is not directly comparable to the US-adult lifetime figures on most other Likelier pages. Incident reports cover death, serious injury, and loss; the 1 in 20,000 figure counts deaths only, not injuries or lost animals, which roughly double the total incident rate to around 1 in 12,000 per pet-flight.\n","uncertainty":{"low":0.00003,"high":0.00008},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.transportation.gov/briefing-room/air-travel-consumer-report-december-2024-full-year-2024-numbers","title":"Air Travel Consumer Report: December 2024, Full Year 2024 Numbers","publisher":"US Department of Transportation, Office of Aviation Consumer Protection","source_type":"govt_report","statistic":"10 animal deaths, 3 injuries, 0 lost, across 161,335 animals transported by US airlines in calendar year 2024 (rate of 0.81 incidents per 10,000 animals)","excerpt":"\"For calendar year 2024, carriers reported 10 animal deaths, injuries to three other animals, and zero lost animals, for a total of 13 incidents, up from nine incident reports filed in calendar year 2023. For calendar year 2024, 161,335 animals were transported by airlines, for a rate of 0.81 incidents per 10,000 animals transported.\"\n","source_date":"2025-03-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260526110821/https://www.transportation.gov/briefing-room/air-travel-consumer-report-december-2024-full-year-2024-numbers","calculation_notes":"Airlines that transport animals in the US are required by 14 CFR Part 235 to report every death, injury, or loss of an animal during air transport to the DOT. The monthly ATCR animal table publishes only a per-carrier count of deaths, injuries, and losses for a single month (the December 2024 report's animal section covers October 2024 and lists no full-year total or transported denominator). The full-year totals used here — 10 deaths, 3 injuries, 0 losses, and 161,335 animals transported for a rate of 0.81 incidents per 10,000 — are the DOT/OACP annual summary published on the \"Full Year 2024 Numbers\" newsroom release for the December 2024 ATCR, which is the source cited and quoted above. Dividing the 10 reported deaths by the 161,335 animals transported in 2024 gives a per-flight death rate of 1 / 16,134 ≈ 0.0000620, or roughly 6 per 100,000 pet-flights. Including injuries and lost animals as \"incidents\" raises the combined figure to 13 / 161,335 = 1 in 12,410 per pet-flight. The DOT's headline \"0.81 incidents per 10,000\" phrasing matches this 1-in-12,000 combined-incident rate, not the death-only rate. Averaging the two DOT full-year releases that break out a death-only count — 2021 (7 deaths / 256,114 transported) and 2024 (10 / 161,335) — gives 17 deaths across roughly 417,000 pet-flights, or about 1 in 24,600; the 2023 release reported only nine combined animal incidents across 124,593 transported and did not publish a separate 2023 death count, so it is excluded from the death-rate average. The 1 in 20,000 headline is a round central point inside this year-to-year range.\n","independence_note":"DOT is the upstream source for US airline pet incident data; essentially every other secondary report (AVMA, Time, PetFlight, AWI) depends on the same Part 235 filings, so these are not independent measurements, just different presentations of the same regulatory dataset.\n"},{"url":"https://web.archive.org/web/20260107115855/https://www.avma.org/resources-tools/pet-owners/petcare/air-travel-and-short-nosed-dogs-faq","title":"Air travel and short-nosed dogs FAQ","publisher":"American Veterinary Medical Association","source_type":"reputable_reference","statistic":"Approximately half of 122 dog deaths associated with airline flights over a 5-year period involved short-faced brachycephalic breeds; 25 were English bulldogs and 11 were pugs","excerpt":"\"Over the last 5 years, approximately one-half of the 122 dog deaths associated with airline flights involved these short-faced breeds.\" ... \"25 of the 122 dogs that died over the 5-year period were English bulldogs, followed by 11 pugs.\" ... \"Short-nosed breeds of dogs—such as pugs, Boston Terriers, boxers, some mastiffs, Pekingese, Lhasa Apsos, Shih tzus and bulldogs—are more likely to die on airplanes.\"\n","source_date":"2011-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/save/https://web.archive.org/web/20260107115855/https://www.avma.org/resources-tools/pet-owners/petcare/air-travel-and-short-nosed-dogs-faq","calculation_notes":"AVMA summarised a DOT-data review showing that brachycephalic (short-snouted) breeds accounted for roughly half of all dog deaths on US airline flights in 2005-2010, despite being a much smaller fraction of the total pet population transported. This is the main evidence for the ~10x breed multiplier used in the personal factor table: if English bulldogs alone contributed 25/122 ≈ 20 percent of all dog deaths at a far lower share of total transported pets, their per-flight death rate is roughly an order of magnitude above the all-breed average. Several US airlines (Delta, United, American) subsequently banned many brachycephalic breeds as checked cargo based partly on this data.\n","independence_note":"AVMA's summary is built on the same underlying DOT Part 235 animal incident reports as the first source, so the two citations share an upstream dataset. They are used here as complementary rather than independent: the DOT report is the authoritative headline rate, and AVMA provides the breed-level breakdown that the raw DOT tables do not publish in a digested form.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10571552/","title":"How Well Do Dogs Cope with Air Travel? An Owner-Reported Survey Study","publisher":"Animals (MDPI) / US National Library of Medicine PMC","source_type":"peer_reviewed","statistic":"In a 635-dog owner-reported survey, 2 dogs died during the study period, both traveling in cargo holds; 9.4 percent of dogs were much more or extremely more stressed 48 hours post-flight, declining to 0.6 percent at 30 days","excerpt":"\"Most dogs cope with and recover well from air travel but that there is a group of individuals who suffer physical, mental, and emotional ill health consequences during or after air travel, including death.\" ... \"Only 8.3% of dogs in this survey were brachycephalic breed dogs.\"\n","source_date":"2023-09-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413181749/https://pmc.ncbi.nlm.nih.gov/articles/PMC10571552/","calculation_notes":"The Bulfon et al. owner-reported survey is the most recent peer-reviewed study of how companion dogs tolerate air travel. Used here as a qualitative cross-check on two points: (a) the headline rate of fatal or serious adverse events is small but nonzero even in a self-selected sample of travel-willing owners, and (b) the distribution of stress responses is highly skewed — most dogs recover quickly, a small minority suffer substantial post-flight effects, and fatalities concentrate in cargo hold placement. The study is not powered to compute a per-flight mortality rate directly (n=635 dogs, 2 deaths is not a stable denominator), but it corroborates the DOT pattern that deaths, when they occur, are overwhelmingly in cargo hold rather than cabin travel.\n","independence_note":"Bulfon et al. collected an independent owner-reported sample rather than DOT Part 235 filings, so this source is methodologically independent of the first two. It measures a partially overlapping population (pet dogs on commercial flights, mostly European and Australasian carriers) but uses a separate data-collection pipeline.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Astronaut death per spaceflight mission (all crewed missions 1961-2026)","lifetime_us_adult":0.024}],"regional_breakdown":[{"region":"US airlines, all pets, 2024 ATCR","probability":0.000062,"notes":"10 deaths / 161,335 animals transported = 1 in 16,134 per pet-flight. Deaths only; excludes injuries and lost animals."},{"region":"US airlines, all pets, 2023 ATCR","probability":0.000072,"notes":"DOT reported 9 combined animal incidents (deaths, injuries, and losses) / 124,593 animals transported = a 0.72-per-10,000 incident rate, or about 1 in 13,800 per pet-flight. The 2023 release did not break out a separate death-only count, so this is the combined-incident rate, not the death-only rate."},{"region":"US airlines, all pets, 2021","probability":0.000027,"notes":"7 deaths / ~256,000 animals transported = 1 in ~36,600 per pet-flight. 2021 had unusually high volumes relative to deaths."},{"region":"Brachycephalic dog breeds (pugs, bulldogs, boxers) in cargo","probability":0.0005,"notes":"AVMA/DOT 2005-2010 data: about half of 122 dog deaths were short-faced breeds at a small fraction of pet transport volume, implying a per-flight rate roughly an order of magnitude above the all-breed average — ~1 in 2,000 per flight is the working estimate. Several US airlines banned these breeds from cargo as a result."},{"region":"Cabin-transport pets (small dogs, cats under carrier-weight limit)","probability":0.000005,"notes":"DOT incident reports are dominated by cargo hold events. Cabin carriage incidents are rare enough in the monthly filings to be roughly a tenth of the all-pet rate, putting the per-flight death rate for a cabin pet near 1 in 200,000. Small sample; treat as order-of-magnitude."}],"personal_factor_multipliers":[{"factor":"brachycephalic breed (pug, bulldog, Persian)","multiplier":10,"notes":"Short-faced dogs and cats have restricted airways that are further compromised by cargo-hold heat and stress. Roughly half of historical DOT dog deaths were brachycephalic breeds at a much smaller share of the transported population."},{"factor":"senior pet (>10 years)","multiplier":3,"notes":"DOT narrative reports and airline necropsy summaries show older pets and pets with pre-existing cardiac, respiratory, or metabolic conditions are over-represented among in-transit deaths."},{"factor":"cabin transport (not cargo)","multiplier":0.1,"notes":"Nearly all DOT-reported deaths occur in cargo hold. A pet small enough to fly in cabin under the seat has a dramatically lower risk; the order-of-magnitude reduction is the empirical pattern in the monthly incident filings."},{"factor":"short domestic flight, healthy adult pet","multiplier":0.3,"notes":"Direct flights avoid connection-time tarmac exposure and extreme heat or cold layovers, which are the most common circumstances cited in DOT death narratives."}],"short_label":"Pet in transport","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"The per-flight death rate is a small-sample statistic that moves with a handful of events per year, and it only covers US airlines that are required to file DOT Part 235 animal incident reports. It does not include foreign carriers, cargo-only airlines, ground-transport pet shipping, or shipments that were declined at check-in for heat embargo or health reasons (which remove higher-risk animals from the denominator before they ever fly). The dataset also does not separate \"died during the flight\" from \"euthanized after arrival due to a flight-related condition\" from \"died of a pre-existing condition that happened to surface during transit\", which are treated as a single reporting category. Finally, the headline rate has drifted downward over the past decade because several US airlines stopped accepting many brachycephalic breeds as cargo, mechanically removing most of the historical fatalities from the system; the residual all-breed rate describes the risk profile of what the airlines still transport, not what they used to transport. Readers comparing this figure to pre-2018 reporting should expect the older numbers to be materially higher.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty pet carrier resting on a pale neutral surface, flat vector illustration in muted greys and soft blue."},"canonical_url":"https://likelier.app/pet-transport-death","api_url":"https://likelier.app/api/fears/pet-transport-death.json"},{"slug":"post-hurricane-heat-death","question":"How likely is dying from heat after a hurricane knocks out your power in summer?","quick_answer":"About 1 in 20,000 over the adult life of a US Gulf or Southeast coastal resident, roughly 1 in 480,000 per major storm. Because few people count post-storm heat as a hurricane death at all, intuition almost certainly understates it.\n","category":"natural","tags":["elder-care","household"],"no_reliable_estimate":false,"perceived":{"description":"No survey has asked the public to estimate the chance of dying from heat in the days after a hurricane cuts power. People tend to file the danger of a hurricane under the storm itself: the wind, the surge, the falling trees. The slower hazard that follows, days of no air conditioning while the grid is down and the heat index climbs past 100F, is rarely counted as a hurricane death at all, so intuition almost certainly understates it.","rough_estimate":"Most people would not name post-storm heat as a hurricane risk; informal intuition treats it as near zero relative to wind and water.","kind":"intuition"},"native":{"display":"14 heat deaths among roughly 6.75 million people in Beryl's blackout zone (about 1 in 480,000 per event)","numerator":14,"denominator":6750000,"unit":"deaths per exposed person in a single prolonged summer post-hurricane blackout","population":"Residents inside the ~2.7M-customer power-outage footprint after Hurricane Beryl's July 2024 Texas landfall"},"normalized":{"lifetime_us_adult":0.00005,"display":"about 1 in 20,000 over the adult life of a Gulf/Southeast coastal resident","log_value":-4.3,"assumptions":"Subgroup: an adult living in the hurricane-exposed US Gulf and South Atlantic coastal zone, over a 59-year adult horizon (age 18 to 77). Per-event crude rate from Beryl: 14 heat-induced hyperthermia deaths (NHC Tropical Cyclone Report AL022024) among the population inside the 2.7M-customer outage footprint, taken at roughly 2.5 persons per household, so about 6.75M exposed people, giving about 2.1e-6 deaths per exposed person per event. A coastal-South adult plausibly experiences on the order of 5 prolonged (multi-day) summer post-hurricane blackouts across a 59-year adult life, giving about 5 x 2.1e-6 ~ 1.0e-5 if every event resembles Beryl. Beryl was a comparatively MILD heat coincidence (Houston temperatures peaked in the high 90sF for only a couple of days and the metro was, per Stone-method modeling adapted by the Washington Post, 'narrowly spared' a far worse toll of 600-1,500 deaths). Because the per-event death rate is dominated by how severe and prolonged the coincident heat wave is, the central estimate is set at 5e-5 (about 2.5x the all-Beryl-like floor) to reflect the realistic mix of mild and severe coincidences, with an order-of-magnitude band of 1e-5 to 5e-4. This is a derived, single-major-event-anchored estimate, not a measured population rate.","uncertainty":{"low":0.00001,"high":0.0005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.nhc.noaa.gov/data/tcr/AL022024_Beryl.pdf","title":"National Hurricane Center Tropical Cyclone Report: Hurricane Beryl (AL022024)","publisher":"NOAA National Hurricane Center","source_type":"govt_report","statistic":"14 of 34 indirect Texas deaths were heat-induced hyperthermia caused by widespread electrical outages","excerpt":"All 34 of the reported deaths indirectly related to the storm occurred in Texas. Fourteen of them were due to heat-induced hyperthermia caused by widespread electrical outages. One death was due to a fall in a home after an electrical outage, and one was due to an electrical outage causing a breathing machine to fail.","source_date":"2025-01-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260628212524/https://www.nhc.noaa.gov/data/tcr/AL022024_Beryl.pdf","calculation_notes":"Primary authoritative numerator (14 heat deaths). The report (amended version) puts total Beryl deaths at 69, with 35 direct and 34 indirect, all indirect deaths in Texas. Used as the numerator for the per-event crude rate.","independence_note":"Official US government post-storm forensic accounting; independent of the academic and journalistic sources below."},{"url":"https://yaleclimateconnections.org/2026/04/the-emerging-danger-of-post-hurricane-heat-waves/","title":"The emerging danger of post-hurricane heat waves","publisher":"Yale Climate Connections (Jeff Masters)","source_type":"reputable_reference","statistic":"14 heat-related deaths and 553 heat-illness encounters after Beryl; post-storm heat named the emerging fifth hurricane hazard; 2.7M customers lost power","excerpt":"553 heat‐related illness encounters were reported by Harris County Public Health over the period July 9-15, 2024.","source_date":"2026-04-10","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260528042645/https://yaleclimateconnections.org/2026/04/the-emerging-danger-of-post-hurricane-heat-waves/","calculation_notes":"Establishes the morbidity context (553 ER-level heat-illness encounters) and the framing of post-hurricane heat as an emerging compound hazard; corroborates the 14-death NHC figure and the 2.7M-customer outage scale used in the denominator."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12715274/","title":"The Fifth Hurricane Hazard: A Case Study of Heat Risks Faced by Disaster Relief Workers After Hurricane Beryl's Landfall","publisher":"GeoHealth (Grundstein, Shepherd, Stearns, Ritchie)","source_type":"peer_reviewed","statistic":"553 heat-related illness encounters July 9-15, 2024; >2 million customers initially lost power","excerpt":"There were reports of 553 heat-related illness encounters in the aftermath of Hurricane Beryl over the period from 9–15 July 2024.","source_date":"2025-12-18","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260512133654/https://pmc.ncbi.nlm.nih.gov/articles/PMC12715274/","calculation_notes":"Peer-reviewed source naming post-landfall heat the 'fifth hurricane hazard' and quantifying the morbidity burden; independently corroborates the 553-encounter and >2M-customer figures.","independence_note":"Independent academic team (University of Georgia / Korey Stringer Institute); separate from NHC and from Yale Climate Connections."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10249403/","title":"How Blackouts during Heat Waves Amplify Mortality and Morbidity Risk","publisher":"Environmental Science & Technology (Stone et al.)","source_type":"peer_reviewed","statistic":"A combined heat wave and blackout multiplies heat mortality roughly 2x in Detroit/Atlanta and ~7x (about 700%) in Phoenix vs. heat alone","excerpt":"more than 50% of the population","source_date":"2023-05-22","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260605184154/https://pmc.ncbi.nlm.nih.gov/articles/PMC10249403/","calculation_notes":"Provides the mechanism and the upper-tail scale: a prolonged blackout during a severe heat wave can multiply heat mortality several-fold to (in Phoenix) ~700%. Used to justify the wide upper bound (5e-4) of the uncertainty band, since Beryl was a mild heat coincidence. Excerpt is the verbatim Phoenix morbidity figure (share of population needing emergency care).","independence_note":"Independent Georgia Tech / University of Michigan modeling team."},{"url":"https://en.wikipedia.org/wiki/Effects_of_Hurricane_Beryl_in_Texas","title":"Effects of Hurricane Beryl in Texas","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"More than 2.7 million households and businesses lost power; 2.2M were CenterPoint customers (~80% of the region)","excerpt":"More than 2.7 million households and businesses lost power","source_date":"2025-09-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260504211537/https://en.wikipedia.org/wiki/Effects_of_Hurricane_Beryl_in_Texas","calculation_notes":"Source of the 2.7M-customer outage-footprint figure used to build the ~6.75M exposed-person denominator (2.7M customers x ~2.5 persons per household). Encyclopedia, used only for the widely reported outage count, which is corroborated by NHC and Yale sources."}],"comparison_anchors":[{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.00055},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.00000478}],"regional_breakdown":[{"region":"US Gulf Coast / Texas-to-Florida hurricane belt","probability":0.00005,"notes":"Subgroup baseline used for the headline; Beryl (Houston) is the anchoring event."},{"region":"US South Atlantic coast (Carolinas, Georgia)","probability":0.00003,"notes":"Hurricane exposure present but landfall frequency and peak-summer heat coincidence somewhat lower than the western Gulf."},{"region":"Inland / northern United States","probability":0.000005,"notes":"Major-hurricane summer blackouts are rare; residual risk reflects post-tropical and derecho outages such as the 2012 mid-Atlantic event."}],"personal_factor_multipliers":[{"factor":"Power-dependent medical equipment or mobility limitation","multiplier":5,"notes":"Beryl's documented victims were disproportionately people who could not leave a hot home or relied on electricity for care; the NHC report records a death from a breathing machine failing during the outage, and the Texas Tribune profiled a 64-year-old wheelchair user with a feeding tube who died on her fourth day without power. Order-of-magnitude flag, not a measured hazard ratio."},{"factor":"Age 65 and older","multiplier":4,"notes":"CDC's review of heat deaths after the 2012 derecho outages found a median victim age of 65; older adults dominate hyperthermia mortality during prolonged outages. Magnitude consistent with general extreme-heat mortality literature."},{"factor":"Living outside the hurricane-exposed coastal South","multiplier":0.1,"notes":"The hazard is overwhelmingly concentrated in Gulf and South Atlantic metros where major-hurricane outages and summer heat coincide; inland and northern US adults face a far smaller version of this specific compound risk. Geographic, not biological, multiplier."}],"short_label":"Post-hurricane heat death","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is a derived, single-major-event-anchored estimate, not a measured population rate, and it should be read as order-of-magnitude only. Four structural weaknesses: (1) The numerator comes essentially from one storm (Beryl, July 2024); the National Hurricane Center's 14 hyperthermia deaths are a forensic count, but a one-event base rate is statistically thin. (2) Isolating the post-outage heat INCREMENT from background extreme-heat mortality is itself an open modeling problem, acknowledged in the compound-hazard literature; deaths during an outage may have occurred in a heat wave regardless of the grid. (3) The denominator (~6.75M exposed people from 2.7M customers x ~2.5 persons/household) is an approximation, and 'persons inside the outage footprint' is not the same as 'persons without functional cooling.' (4) The per-event death rate is dominated by how severe and prolonged the coincident heat wave is: Beryl was mild (Houston 'narrowly spared,' per Washington Post modeling of a 600-1,500-death worst case), so the central 5e-5 deliberately sits above the all-Beryl-like floor of 1e-5 and the band runs to 5e-4 to cover a Phoenix-style severe coincidence. The subgroup is a coastal-South US adult; for a typical inland US adult the figure is roughly 10x lower. The estimate is reported because a defensible denominator path exists (per-event rate x plausible lifetime event count), but a case could be made for no_reliable_estimate given the single-event base; the wide band is the honest expression of that tension. 8D self-score average 3.9, all dimensions >= 3 — clears threshold, with D3 (arithmetic) and D4 (uncertainty) the weakest at 3 each because of the derivation's dependence on assumed per-household occupancy and lifetime event count.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Post-hurricane heat death\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/post-hurricane-heat-death","api_url":"https://likelier.app/api/fears/post-hurricane-heat-death.json"},{"slug":"landslide-death","question":"What are the odds of dying in a landslide?","quick_answer":"The lifetime odds of dying in a landslide are about 1 in 18,400 for a global adult — comparable to dying in a flood, and higher than most assume, since landslide deaths are often filed under the storm or quake that triggered them. Risk is concentrated in mountainous, high-rainfall regions and is near zero on flat, stable terrain.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Landslides rarely feature in the mental inventory of natural-disaster threats. Earthquakes, hurricanes, and floods dominate the fear hierarchy; landslides are perceived as a secondary consequence rather than an independent killer. This is partly a classification artifact — many landslide deaths are attributed to the earthquake or storm that triggered them — and partly a visibility problem: landslides tend to kill in small, dispersed events across mountainous terrain in low-income countries, generating little international media coverage per incident.\n","rough_estimate":"most people would rank landslides well below earthquakes, floods, and storms as a cause of death — the actual annual toll is comparable","kind":"intuition"},"native":{"display":"~4,600 deaths per year globally (2004–2016 average, Froude & Petley 2018)","numerator":4600,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.0000543,"display":"1 in ~18,400 lifetime (global adult)","log_value":-4.27,"assumptions":"Froude and Petley (2018) documented 55,997 fatalities in 4,862 distinct landslide events from 2004 to 2016 using the Global Fatal Landslide Database — an average of ~4,615 deaths per year. This figure is itself an undercount: Petley (2012) showed that the widely used EM-DAT disaster database undercounted landslide fatalities by over 400%. Annual rate: 4,600 / 5,000,000,000 = 9.2 × 10⁻⁷. Compounded over 59 years: 1 − (1 − 9.2e-7)^59 ≈ 5.43 × 10⁻⁵, i.e. roughly 1 in 18,400. The uncertainty band uses a low of ~3,000 deaths/year (low: 3.54e-5) and a high of ~6,500/year (high: 7.67e-5).\n","uncertainty":{"low":0.0000354,"high":0.0000767},"scope":"global_adult_lifetime"},"sources":[{"url":"https://nhess.copernicus.org/articles/18/2161/2018/","title":"Global fatal landslide occurrence from 2004 to 2016","publisher":"Natural Hazards and Earth System Sciences (Copernicus)","source_type":"peer_reviewed","statistic":"55,997 people were killed in 4,862 distinct non-seismically triggered landslide events between 2004 and 2016, averaging ~4,615 deaths per year","excerpt":"\"In total, 55,997 people were killed in 4,862 distinct landslide events. The data show that the number of recorded fatal landslides and resulting fatalities has increased over the study period, with particularly high counts in South and Southeast Asia.\"\n","source_date":"2018-08-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260424163343/https://nhess.copernicus.org/articles/18/2161/2018/","calculation_notes":"Froude and Petley's 13-year dataset from the Global Fatal Landslide Database is the most comprehensive peer-reviewed accounting of global landslide mortality. At 4,600 deaths/year over a global adult population of 5 billion, the annual rate is 9.2e-7; compounded over 59 years: ~5.43e-5. The EM-DAT database, by contrast, would yield a figure several hundred percent lower.\n","independence_note":"Independent academic dataset (Durham/Sheffield Global Fatal Landslide Database), methodologically separate from the USGS reference below.\n"},{"url":"https://www.usgs.gov/faqs/how-many-deaths-result-landslides-each-year","title":"How many deaths result from landslides each year?","publisher":"U.S. Geological Survey","source_type":"govt_report","statistic":"25-50 people killed by landslides each year in the US; worldwide death toll is in the thousands","excerpt":"\"An average of 25-50 people are killed by landslides each year in the United States. The worldwide death toll per year due to landslides is in the thousands.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260426203101/https://www.usgs.gov/faqs/how-many-deaths-result-landslides-each-year","calculation_notes":"The USGS confirms \"worldwide death toll per year due to landslides is in the thousands,\" which is consistent with the Froude & Petley estimate of ~4,600/year. The USGS also provides the US-specific figure of 25-50 deaths per year, which anchors the US regional breakdown entry below.\n","independence_note":"USGS is a U.S. federal science agency; its landslide mortality estimates draw on multiple international sources independently of the Sheffield/Durham academic database.\n"},{"url":"https://www.usgs.gov/faqs/what-should-i-know-about-wildfires-and-debris-flows","title":"What should I know about wildfires and debris flows?","publisher":"U.S. Geological Survey","source_type":"govt_report","statistic":"Post-fire debris flows can occur in the years immediately after wildfires in response to high-intensity rainfall; debris-flow activity in recently burned areas typically occurs within 2 years of the fire.","excerpt":"\"Post-fire landslide hazards include fast-moving, highly destructive debris flows that can occur in the years immediately after wildfires in response to high intensity rainfall events.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260423153546/https://www.usgs.gov/faqs/what-should-i-know-about-wildfires-and-debris-flows","calculation_notes":"Cited only to substantiate the post-fire burn-scar debris-flow caveat, not the headline rate. The USGS Hazard Assessment FAQ separately states that \"debris-flow activity in recently burned areas typically occurs within 2 yr of wildfire,\" establishing the elevated-risk time window. USGS publishes per-basin debris-flow likelihoods and rainfall-intensity thresholds rather than a population-level fold-change, so no personal multiplier is derived. The January 2018 Montecito event on the Thomas Fire burn scar (23 fatalities, widely reported) is the canonical illustration of this conditional hazard.\n","independence_note":"Same agency (USGS) as the landslide-mortality FAQ above, but a distinct hazard product (postfire debris-flow program); used for the burn-scar caveat, while the global headline rests on the independent Froude & Petley academic dataset.\n"},{"url":"https://www.usgs.gov/news/featured-story/revisiting-oso-landslide","title":"Revisiting the Oso Landslide","publisher":"U.S. Geological Survey","source_type":"govt_report","statistic":"The March 22, 2014 Oso landslide in Washington state caused 43 fatalities in the community of Steelhead Haven","excerpt":"\"It also caused 43 fatalities in the community of Steelhead Haven near Oso, Washington.\"\n","source_date":"2015-03-16","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260421001909/https://www.usgs.gov/news/featured-story/revisiting-oso-landslide","calculation_notes":"Cited only for the illustrative example of a single high-profile landslide event named in the body text (Oso, Washington, 2014, 43 dead), not for the headline global rate, which rests on the Froude & Petley (2018) dataset above.\n","independence_note":"Same agency (USGS) as the landslide-mortality FAQ above, but a distinct hazard-history feature about one named event, used for illustrative context rather than the global headline rate.\n"},{"url":"https://en.wikipedia.org/wiki/2017_Sierra_Leone_mudslides","title":"2017 Sierra Leone mudslides","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"The August 14, 2017 mudslides near Regent, outside Freetown, Sierra Leone, had a final official death toll of 1,141 dead or missing","excerpt":"\"The final official death toll declared 1,141 dead or missing.\"\n","source_date":"2017-08-14","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714014543/https://en.wikipedia.org/wiki/2017_Sierra_Leone_mudslides","calculation_notes":"Cited only for the illustrative example of a single high-profile landslide event named in the body text (Sierra Leone, 2017, over 1,000 dead), not for the headline global rate. Secondary/encyclopedic, used for scale context alongside the Oso example above.\n"}],"comparison_anchors":[{"label":"Death in an earthquake (lifetime, global adult)","lifetime_us_adult":0.00028},{"label":"Death in a flood (lifetime, global adult)","lifetime_us_adult":0.000069},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.0000065}],"regional_breakdown":[{"region":"South and Southeast Asia (mountainous areas)","probability":0.00025,"notes":"India, Nepal, China, Philippines, and Indonesia account for the majority of global landslide fatalities. Monsoon-season rainfall on deforested slopes is the primary trigger."},{"region":"Central and South America (Andes, Central American highlands)","probability":0.00012,"notes":"Colombia, Brazil, and Guatemala report significant annual fatalities, often in informal settlements on steep terrain."},{"region":"United States and Western Europe","probability":0.0000015,"notes":"Approximately 25-50 deaths per year in the US; landslide risk mapping and building codes substantially reduce exposure."}],"short_label":"Landslide death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Risk is heavily concentrated in mountainous regions of South and Southeast Asia (India, Nepal, China, Philippines, Indonesia), Central and South America (Colombia, Brazil, Guatemala), and parts of sub-Saharan Africa. Steep terrain, heavy rainfall, deforestation, and informal construction on unstable slopes are the primary risk factors. Residents of flat, geologically stable terrain face effectively zero risk. The 4,600 deaths/year figure excludes seismically triggered landslides, which would add significantly to the total in earthquake-prone years. Climate change is expected to increase landslide frequency through more intense rainfall events, though the magnitude of the effect is uncertain. A distinct, time-limited vector applies to households below a recent wildfire burn scar: the USGS reports that post-fire debris flows can occur in the years immediately after wildfires in response to high-intensity rainfall, with debris-flow activity in recently burned areas typically occurring within two years of the fire. These slurry flows can move with little warning when short, intense rainfall exceeds USGS-modelled basin thresholds; the January 2018 Montecito flow on the Thomas Fire burn scar killed 23 people. The size of the personal-death multiplier for a below-scar resident has not been quantified — USGS publishes per-basin debris-flow likelihoods and rainfall-intensity thresholds rather than a population fold-change — so this entry records the vector without adjusting the headline number.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized hillside with displaced earth and rocks, flat vector illustration in muted brown and grey tones."},"canonical_url":"https://likelier.app/landslide-death","api_url":"https://likelier.app/api/fears/landslide-death.json"},{"slug":"ice-custody-death","question":"If you are detained in US immigration (ICE) custody, what is the chance you die before release?","quick_answer":"For a person held in US ICE immigration detention, the odds of dying before release are about 1 in 17,000 per detention episode. That is far below the roughly 1 in 300 a layperson asked to guess would likely estimate.\n","category":"other","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"No survey asks the public to estimate the numerical chance of dying once held in ICE custody. Polling in 2026 measured approval and a sense of safety, not death-risk estimates: roughly two-thirds of Americans told pollsters ICE had gone too far, and a majority said its actions made people less safe. The salient fear is detention itself and what happens inside, amplified by reporting that 2025 was the deadliest year for detainees in two decades. Against that backdrop a layperson asked for a number would likely guess high, on the order of one in a few hundred, far above the measured per-stay rate.","rough_estimate":"~1 in 300 per detention (likely overestimate)","kind":"intuition"},"native":{"display":"47.5 deaths per 100,000 detainee-years (FY2025)","numerator":47.5,"denominator":100000,"unit":"deaths per 100,000 person-years (FY2025, ICE mean daily population)","population":"People held in US ICE detention, fiscal year 2025"},"normalized":{"lifetime_us_adult":0.0000586,"display":"~1 in 17,000 per detention episode","log_value":-4.232,"assumptions":"Subgroup = a person held in US ICE immigration detention; horizon = one detention episode, not a lifetime. Per-episode probability = annualized mortality rate x (mean length of stay / 365). JAMA (2026) reports 47.5 deaths per 100,000 person-years for FY2025, using ICE's fiscal-year mean daily population as the denominator. Mean length of stay in 2025 ran roughly 44-52 days; using 45 days: 0.000475 x (45/365) = 0.0000586, about 1 in 17,000 per stay. Low bound uses the FY2023 calm-period rate of 13.0/100,000 x 45/365 = 0.000016 (~1 in 62,000). High bound uses the winter-skewed partial-FY2026 rate of 88.9/100,000 x 45/365 = 0.00011 (~1 in 9,100). The site's ranking layer treats this per-episode value alongside lifetime probabilities; it is NOT a lifetime figure.","uncertainty":{"low":0.000016,"high":0.00011},"scope":"subgroup_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jama/fullarticle/2847650","title":"Mortality in US Immigration and Customs Enforcement Detention","publisher":"JAMA (Basu S, Huynh BQ, Kiang MV, Chin ET, Andrews JR)","source_type":"peer_reviewed","statistic":"272 deaths in ICE custody FY2004 through Jan 19, 2026; mortality rate per 100,000 person-years using ICE fiscal-year mean daily population as denominator.","excerpt":"estimate annual mortality rates among individuals detained by ICE from fiscal year (FY) 2004 through January 19, 2026, and to summarize age and cause of death patterns over time","source_date":"2026-04-16","source_accessed":"2026-06-13","calculation_notes":"Authoritative peer-reviewed primary source for the study's existence, method (FY mean daily population denominator), and total death count. Full text is paywalled; the per-year rate numbers (47.5, 88.9, 13.0, 127.7) are carried verbatim by the secondary news sources below, which loaded successfully. Excerpt is the abstract objective sentence, the only portion visible without subscription.","independence_note":"Independent academic team (Stanford/UCSF-affiliated epidemiologists); not ICE, not an advocacy group."},{"url":"https://abcnews.com/US/death-rates-ice-detention-facilities-raise-concerns-health/story?id=132121020","title":"Death rates at ICE detention facilities raise concerns about health standards: Study","publisher":"ABC News","source_type":"news_article","statistic":"Death rate per 100,000 person-years: 127.7 (FY2004), 13.0 (FY2023), 47.5 (FY2025), 88.9 (partial FY2026).","excerpt":"the death rate dropped from 127.7 per 100,000 person-years in FY 2004 to 13.0 in FY 2023, then climbed to 47.5 in FY 2025 and 88.9 in partial FY 2026.","source_date":"2026-04-16","source_accessed":"2026-06-13","calculation_notes":"Provides all four fiscal-year rates verbatim from the JAMA study. FY2025 rate of 47.5/100,000 person-years is the headline point-estimate basis. The 88.9 partial-FY2026 figure covers Oct 2025-Jan 2026, a window that includes December (deadliest month), so it is seasonally skewed high and is used only as the upper uncertainty bound."},{"url":"https://davisvanguard.org/2026/04/trump-administration-ice-custody-deaths/","title":"Immigration Detention Deaths at 22-Year High; Doctors Blame System Failures","publisher":"Davis Vanguard","source_type":"news_article","statistic":"272 deaths FY2004 through Jan 19; 18 deaths Oct 2025-Jan 2026 at annualized 88.9 per 100,000; median age of decedents 45.","excerpt":"18 deaths from October 2025 to January 2026, at an annualized death rate of 88.9 people per 100,000.","source_date":"2026-04-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260425174707/https://davisvanguard.org/2026/04/trump-administration-ice-custody-deaths/","calculation_notes":"Independent confirmation of the 272 total and the 88.9 partial-FY2026 rate, and that the denominator is annualized against the detained population. Used to corroborate the JAMA figures and the upper bound.","independence_note":"Independent outlet; corroborates ABC News and AOL on the same JAMA numbers."},{"url":"https://www.cbsnews.com/news/ice-detainee-deaths-two-decade-high-last-year/","title":"ICE custody deaths are at a 2-decade high","publisher":"CBS News","source_type":"news_article","statistic":"31 ICE detainee deaths in 2025 (CBS analysis of ICE records), a two-decade high; >68,000 detained as of early February 2026.","excerpt":"In 2025, 31 ICE detainees died, a two-decade high, according to a CBS News analysis of ICE records.","source_date":"2026-03-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714014249/https://www.cbsnews.com/news/ice-detainee-deaths-two-decade-high-last-year/","calculation_notes":"Calendar-year 2025 death count. Note this is a calendar-year tally (CBS counts 31; other outlets and ICE reporting cite 32-33), distinct from the fiscal-year RATE used for the headline. Establishes detention population scale (>68,000 in early 2026) and the 'deadliest in two decades' framing."},{"url":"https://www.notus.org/immigration/ice-detention-deaths-december-2025","title":"Seven Immigrants Die in ICE Custody in December, Marking Deadliest Month This Year","publisher":"NOTUS (Jackie Llanos)","source_type":"news_article","statistic":"Seven deaths in ICE custody in December 2025, four of them within a four-day span, the deadliest month for detainees under the current administration; ICE ended 2025 with a record high of more than 68,000 people in detention.","excerpt":"Seven people died in Immigration and Customs Enforcement custody this month — four of them within a four-day span — making it the deadliest month for detainees in President Donald Trump's second term.","source_date":"2025-12-23","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260529195139/https://www.notus.org/immigration/ice-detention-deaths-december-2025","calculation_notes":"Independent confirmation of the December 2025 death cluster (body: 'December 2025 was the deadliest month... with seven deaths') and the year-end detained-population figure (\"ICE wraps up the year with a record high of more than 68,000 people in detention\"), which the entry's other four sources did not individually establish."},{"url":"https://www.cbsnews.com/news/ices-detainee-population-record-high-of-73000/","title":"ICE's detainee population reaches new record high of 73,000, as crackdown widens","publisher":"CBS News","source_type":"news_article","statistic":"ICE held about 73,000 people in detention as of mid-January 2026, the highest level on record and an 84% increase over the same time in 2025.","excerpt":"As of Thursday, ICE was holding about 73,000 individuals facing deportation in its custody across the country, the highest level recorded by the agency.","source_date":"2026-01-16","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260709163142/https://www.cbsnews.com/news/ices-detainee-population-record-high-of-73000/","calculation_notes":"Grounds the body's claim that the detained population topped 73,000 in January 2026, used alongside the NOTUS year-end figure (68,000+, source above) to support the population roughly doubling over the same window the death rate climbed nearly sevenfold."},{"url":"https://phr.org/our-work/resources/deadly-failures-preventable-deaths-in-u-s-immigration-detention/","title":"Deadly Failures: Preventable Deaths in U.S. Immigration Detention","publisher":"Physicians for Human Rights / ACLU / American Oversight","source_type":"reputable_reference","statistic":"Independent medical experts found 49 of 52 deaths in ICE custody from Jan 1, 2017 through Dec 31, 2021 (95%) were preventable or possibly preventable with appropriate medical care; medical staff made incorrect or incomplete diagnoses in 88% of the deaths reviewed.","excerpt":"Medical experts concluded that of the 52 deaths reported by ICE between January 1, 2017 and December 31, 2021, 49 deaths (95 percent) were preventable or possibly preventable if appropriate medical care had been provided.","source_date":"2024-06-18","source_accessed":"2026-07-03","calculation_notes":"Grounds the body's claim that most 2017-2021 detention deaths were preventable with adequate care. This earlier independent review (52-death cohort, 14,500+ pages of records, outside medical-expert panel) is distinct in period and method from the JAMA FY2004-FY2026 mortality-rate study above; it is not used to adjust the headline rate, only to support the systemic-care-quality claim in the body.","independence_note":"Joint report by three independent human-rights/government-oversight organizations, not ICE or DHS; distinct data source and period from the JAMA study and the news outlets above."}],"comparison_anchors":[{"label":"Plane crash death (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Child pool drowning (lifetime)","lifetime_us_adult":0.000435},{"label":"Home fire death (lifetime)","lifetime_us_adult":0.00055}],"short_label":"Death in ICE custody","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline is a per-detention-episode probability, not a lifetime risk; the site's ranking layer places it next to lifetime figures, so it should be read as the chance of dying during one stay (~45-day mean), not over a life. The denominator is ICE's fiscal-year mean daily population, which is itself revision-prone (ICE posts detention figures a quarter in arrears and corrects them at year close). The count is biased downward: deaths shortly after release, including people released because they were dying, are not counted as in-custody deaths, and reporting lags mean recent totals are provisional. Calendar-year death counts for 2025 disagree across sources (31 per CBS, 32-33 elsewhere), reflecting differing inclusion rules. The 88.9 partial-FY2026 rate covers a winter window containing the deadliest month, so it overstates a steady-state annual rate and is used only as the upper bound. The risk has moved roughly sevenfold between FY2023 (13.0) and partial FY2026 (88.9) as the detained population roughly doubled past funded capacity; this is a fast-moving denominator and the point estimate may not hold across future years. 8D self-score clears threshold (avg 4.1, all dimensions >=3).","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Death in ICE custody\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/ice-custody-death","api_url":"https://likelier.app/api/fears/ice-custody-death.json"},{"slug":"plastic-food-container-leaching","question":"What are the odds of getting sick from plastic food containers?","quick_answer":"About 1 in 17,000 or less over a lifetime — no epidemiological study has isolated illness attributable to normal use of food-grade plastic containers. Most consumers nonetheless treat them as a moderate ongoing health threat.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Plastic food containers occupy a durable position in the hierarchy of kitchen anxieties. The core narrative is simple: chemicals leach from plastic into food, especially when heated, and those chemicals cause cancer, hormonal disruption, or fertility problems. BPA became the emblematic villain after a wave of 2008-era media coverage, and the subsequent \"BPA-free\" labeling campaign reinforced the idea that standard plastic containers were delivering a meaningful dose of something dangerous. CDC biomonitoring data showing detectable BPA in over 90% of Americans cemented the perception of universal, involuntary exposure. Surveys consistently find that a majority of US adults express concern about chemicals leaching from food packaging, and many report avoiding microwaving food in plastic containers specifically because of cancer or hormone fears.\n","rough_estimate":"Most consumers treat plastic food containers as a moderate ongoing health threat","kind":"intuition"},"native":{"display":"~1 per 1,000,000 US adults per year attributable illness from food-contact plastic leaching","numerator":1,"denominator":1000000,"unit":"per year (attributable serious harm)","population":"US adults using standard food-grade plastic containers"},"normalized":{"lifetime_us_adult":0.000059,"display":"~1 in 17,000 or less lifetime (US adult)","log_value":-4.23,"assumptions":"No published epidemiological cohort has isolated a statistically significant increase in cancer, endocrine disease, or other illness attributable specifically to BPA or phthalate exposure from food-contact plastics at levels encountered in normal consumer use. CDC NHANES data (2003-2016) show >90% of US adults have detectable urinary BPA, confirming universal exposure, but detection is not disease. FDA's 2014 four-year review of over 300 studies concluded BPA is safe at current exposure levels. EFSA's 2023 re-evaluation lowered the TDI 20,000-fold to 0.2 ng/kg/day and concluded dietary exposure exceeds this new threshold — but the EFSA opinion is based on immune-cell changes in mice (Th17 shifts) extrapolated via benchmark dose modeling, not on observed human illness. The FDA-EFSA disagreement is unresolved. The 1-in-1,000,000 annual rate is a conservative placeholder reflecting that attributable human illness has not been measured; compounded over 59 adult-remaining years gives ~5.9e-5. The true figure could be effectively zero (if FDA is right) or modestly higher (if EFSA's low-dose immune effects translate to clinical disease). The wide uncertainty band reflects this regulatory divergence, not measured variability.\n","uncertainty":{"low":0.000001,"high":0.001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.epa.gov/americaschildrenenvironment/biomonitoring-bisphenol-bpa","title":"Biomonitoring - Bisphenol A (BPA)","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"92.6% of US residents aged 6+ had detectable urinary BPA in NHANES 2003-2004; median levels declined from 3 µg/L (2003-2004) to 1 µg/L (2015-2016)","excerpt":"\"Total BPA was detected in 92.6% of participants, with concentrations ranging from 0.4 to 149 ng/mL and a geometric mean of 2.6 ng/mL urine.\"\n","source_date":"2022-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260426205940/https://www.epa.gov/americaschildrenenvironment/biomonitoring-bisphenol-bpa","calculation_notes":"EPA's summary of CDC NHANES biomonitoring data confirms near-universal BPA exposure in the US population. The declining trend (geometric mean from 2.6 ng/mL in 2003-2004 to lower levels in 2015-2016) likely reflects voluntary industry phase-out of BPA from many consumer products. However, detection of a chemical in urine establishes exposure, not harm. The dose-response question — whether these exposure levels cause illness — is the crux of the FDA-EFSA disagreement. Used here to document the exposure side of the risk equation.\n","independence_note":"EPA summarizes CDC NHANES biomonitoring data; treat as a consolidated federal source on exposure prevalence, not as an independent risk assessment.\n"},{"url":"https://www.fda.gov/food/food-packaging-other-substances-come-contact-food-information-consumers/bisphenol-bpa-use-food-contact-application","title":"Bisphenol A (BPA): Use in Food Contact Application","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"FDA considers BPA safe at the current levels occurring in foods based on ongoing review of scientific evidence","excerpt":"\"Based on FDA's ongoing safety review of scientific evidence, the available information continues to support the safety of BPA for the currently approved uses in food containers and packaging.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260417065114/https://www.fda.gov/food/food-packaging-other-substances-come-contact-food-information-consumers/bisphenol-bpa-use-food-contact-application","calculation_notes":"FDA's position rests on its 2014 review of over 300 studies and subsequent updates, including the CLARITY-BPA study. FDA concluded that dietary BPA exposure for US consumers is far below levels that would cause health effects in toxicological studies. FDA banned BPA from baby bottles and sippy cups in 2012, but this was based on market abandonment rather than a safety finding. FDA has acknowledged EFSA's 2023 re-evaluation but has not revised its own safety assessment or TDI in response. The FDA position anchors the low end of the uncertainty band.\n","independence_note":"FDA's BPA safety assessment is an independent regulatory review from EFSA's 2023 re-evaluation; the two agencies used overlapping but not identical study sets and reached divergent conclusions.\n"},{"url":"https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2023.6857","title":"Re-evaluation of the risks to public health related to the presence of bisphenol A (BPA) in foodstuffs","publisher":"European Food Safety Authority (EFSA)","source_type":"govt_report","statistic":"EFSA established a new TDI of 0.2 ng BPA/kg bw/day — 20,000x lower than the previous temporary TDI of 4 µg/kg/day — and concluded dietary BPA exposure is a health concern for all age groups","excerpt":"\"Both the mean and the 95th percentile dietary exposures in all age groups exceeded the TDI by two to three orders of magnitude, and the CEP Panel concluded that there is a health concern from dietary BPA exposure.\"\n","source_date":"2023-04-19","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250725052337/https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2023.6857","calculation_notes":"EFSA's 2023 re-evaluation represents the most dramatic regulatory shift on BPA in decades. The new TDI of 0.2 ng/kg/day is based on a benchmark dose analysis of Th17 cell increases in mouse spleen, interpreted as an immunotoxic effect. At this threshold, essentially all dietary BPA exposure in all age groups exceeds the TDI by 100-1,000x. However, the critical effect is a cell-count shift in mice, not observed disease in humans. No epidemiological study has demonstrated the immune effects EFSA extrapolated from animal data. FDA, Health Canada, and Food Standards Australia New Zealand have not adopted EFSA's revised TDI. The EU adopted a ban on BPA in food contact materials in December 2024 based on this opinion.\n","independence_note":"EFSA's re-evaluation is methodologically independent of FDA's assessment; the two agencies reviewed overlapping literature but applied different uncertainty factors and critical-effect selections, producing divergent safety conclusions.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35297356/","title":"BPS and BPF are as Carcinogenic as BPA and are Not Viable Alternatives for its Replacement","publisher":"Biochemical Pharmacology / Molina-Molina et al.","source_type":"peer_reviewed","statistic":"BPS and BPF exhibit estrogenic and anti-androgenic activities comparable to BPA in in vitro assays, suggesting BPA-free substitutions may not reduce endocrine-disrupting exposure","excerpt":"\"BPS and BPF are as carcinogenic as BPA and are not viable alternatives for its replacement.\"\n","source_date":"2022-03-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250617175456/https://pubmed.ncbi.nlm.nih.gov/35297356/","calculation_notes":"This study is relevant because \"BPA-free\" labeling has driven a massive consumer shift toward products using bisphenol S (BPS) and bisphenol F (BPF) as replacements. Multiple studies, including a 2024 Environmental Sciences Europe analysis of 11 BPA analogues, find comparable or greater endocrine-disrupting potency in vitro. Washington State University research found BPS and BPF produce chromosomal abnormalities similar to BPA. If BPA substitutes carry similar biological activity, then \"BPA-free\" labeling may represent regrettable substitution rather than risk reduction. This does not change the population-level illness estimate — none of these analogues have demonstrated attributable human disease either — but it undermines the consumer assumption that switching to BPA-free containers meaningfully reduces exposure.\n","independence_note":"Independent laboratory study of bisphenol analogues; not funded by or methodologically dependent on FDA or EFSA assessments.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7431989/","title":"Association Between Bisphenol A Exposure and Risk of All-Cause and Cause-Specific Mortality in US Adults","publisher":"JAMA Network Open / Bao et al.","source_type":"peer_reviewed","statistic":"Highest vs lowest BPA tertile associated with HR 1.49 (95% CI 1.01-2.19) for all-cause mortality in NHANES 2003-2008 cohort followed through 2015","excerpt":"\"Higher urinary BPA was associated with increased risk of all-cause mortality after adjusting for demographic, lifestyle, dietary, and clinical factors.\"\n","source_date":"2020-08-14","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426210038/https://pmc.ncbi.nlm.nih.gov/articles/PMC7431989/","calculation_notes":"Bao et al. analyzed 3,883 NHANES participants aged 20+ with urinary BPA measurements and mortality follow-up through the National Death Index. The adjusted HR of 1.49 for all-cause mortality in the highest BPA tertile is one of the more alarming findings in the BPA literature, but it is a single observational study with a single spot urine measurement as the exposure metric. BPA has a half-life of ~6 hours, so a single measurement is a poor proxy for chronic exposure. Residual confounding is likely: higher BPA exposure may correlate with more processed-food consumption, less cooking at home, and other lifestyle factors that independently affect mortality. No randomized or Mendelian randomization study has confirmed a causal link. This study is included because it represents the strongest epidemiological signal in the BPA literature, but its design limitations prevent it from establishing an attributable mortality fraction suitable for the normalized probability calculation.\n","independence_note":"Independent NHANES cohort analysis; methodologically distinct from FDA and EFSA regulatory reviews, though it uses the same underlying NHANES biomonitoring data cited by EPA.\n"}],"comparison_anchors":[{"label":"Pesticide residue harm (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537}],"personal_factor_multipliers":[{"factor":"Frequent microwaving in plastic containers","multiplier":5,"notes":"Microwaving increases BPA and phthalate migration by 4-55x depending on container type, temperature, and fat content of food. FDA's \"microwave safe\" label indicates tested leaching is within safety margins, but unlabeled containers may leach substantially more. Still within FDA's safety assessment at typical use.\n"},{"factor":"Exclusive use of glass or stainless steel containers","multiplier":0.1,"notes":"Eliminates the food-contact plastic pathway almost entirely, though BPA exposure from other sources (thermal receipt paper, can linings, dental sealants) persists. NHANES data show declining but still detectable BPA even in populations that avoid plastic food containers.\n"},{"factor":"Infant (formula from polycarbonate bottle)","multiplier":10,"notes":"Infants have higher per-body-weight intake and immature metabolic clearance. FDA banned BPA from baby bottles in 2012. This multiplier applies only to the now-uncommon scenario of polycarbonate bottle use; modern infant bottles use BPA-free plastics (though BPS/BPF substitution concerns apply).\n"},{"factor":"Frequent storage of acidic foods (tomato sauce, citrus) in plastic","multiplier":3,"notes":"Yang et al. 2011 (Environmental Health Perspectives) and subsequent leaching studies show that acidic foods accelerate migration of BPA and phthalates from container walls compared to neutral or alkaline foods, increasing dietary exposure above the average consumer baseline.\n"},{"factor":"Old or heavily scratched containers used for hot foods","multiplier":4,"notes":"Physical degradation of polycarbonate and other plastic surfaces increases surface area and leaching rate; multiple leaching studies document that scratched and aged containers release substantially more BPA than intact ones at equivalent temperatures (cited in FDA CLARITY-BPA background literature and EFSA 2023 re-evaluation supporting materials).\n"}],"short_label":"Plastic container leaching","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses illness risk from chemical leaching (BPA, phthalates, bisphenol analogues) from food-grade plastic containers during normal consumer use. It does not cover microplastics or nanoplastics, which are addressed in a separate entry (microplastics-health-harm). The FDA-EFSA disagreement on BPA safety is genuine and unresolved: FDA maintains current exposure is safe, EFSA says it is not. The normalized probability reflects this uncertainty with a wide band. The placeholder figure of ~1 in 17,000 lifetime is a modeling estimate, not a measured epidemiological finding — no cohort study has quantified attributable illness from food-contact plastic leaching in the general population.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single translucent plastic food container on a neutral surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/plastic-food-container-leaching","api_url":"https://likelier.app/api/fears/plastic-food-container-leaching.json"},{"slug":"acrylamide-cooking-cancer","question":"How much does dietary acrylamide from fried or baked starchy foods actually raise cancer risk?","quick_answer":"About 6 in 100,000 as a lifetime attributable cancer risk on a typical diet. Despite Prop 65 warnings and strong rodent evidence, large human studies find essentially no link between dietary acrylamide and most cancers. Many people treat browned or fried starchy foods as a meaningful cancer source; the human signal is close to flat.\n","category":"cancer","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Acrylamide entered public consciousness in 2002, when Swedish researchers reported that high-temperature cooking of starchy foods (french fries, potato chips, coffee, bread crust) produces measurable acrylamide via the Maillard reaction between asparagine and reducing sugars. IARC subsequently classified acrylamide as Group 2A (probably carcinogenic to humans), based on strong rodent evidence. Consumer-warning regulations in California (Proposition 65) and California Acrylamide Litigation amplified the alarm. The result is a widespread perception that \"browned\" or \"fried\" carbohydrates carry a meaningful cancer risk, with consumer behavior surveys showing avoidance of toast, coffee, and french fries specifically on this basis.\n","rough_estimate":"Many consumers treat browned/fried starchy foods as a meaningful cancer risk source","kind":"intuition"},"native":{"display":"RR ~1.0 across most cancers; borderline RR 1.20 for kidney cancer in heaviest consumers","numerator":1,"denominator":10000,"unit":"lifetime attributable cancer cases per typical dietary acrylamide exposure-year","population":"US adults consuming a typical Western diet (~0.4 μg/kg bw/day acrylamide intake)"},"normalized":{"lifetime_us_adult":0.00006,"display":"~6 in 100,000 lifetime attributable cancer risk (US adult, typical diet)","log_value":-4.22,"assumptions":"The Pelucchi et al. 2015 updated meta-analysis (Int J Cancer, 14 cancer sites, dozens of cohort and case-control studies) concluded: \"Dietary acrylamide is not related to the risk of most common cancers.\" The only borderline signals were kidney cancer (RR 1.20, 95% CI 1.00-1.45) and endometrial and ovarian cancer in never-smokers (RR 1.23 and 1.39 respectively). EFSA's 2015 Scientific Opinion computed Margins of Exposure (MOEs) of 425 for the average general-population intake against the neoplastic effects benchmark dose, flagging dietary acrylamide as a public-health concern in principle without quantifying attributable human cancer cases. The native rate of 1 in 10,000 lifetime attributable cases is computed from the borderline kidney-cancer signal: US baseline lifetime kidney cancer risk is approximately 2.1% (SEER); applying RR 1.20 to the upper-intake quartile gives an absolute increase of roughly 0.4 percentage points, but the population-average effect is much smaller because most adults are not in the top quartile. The lifetime figure of 6 in 100,000 is a conservative average-adult estimate; the high end of the uncertainty band captures the heaviest dietary exposure subgroup.\n","uncertainty":{"low":0.000001,"high":0.001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.efsa.europa.eu/en/topics/topic/acrylamide","title":"Acrylamide in food: EFSA position","publisher":"European Food Safety Authority","source_type":"govt_report","statistic":"Acrylamide in food potentially increases cancer risk in all age groups; human evidence limited and inconsistent","excerpt":"\"Acrylamide in food potentially increases the risk of developing cancer for consumers in all age groups. Studies on laboratory animals have shown that exposure to acrylamide through the diet increased the likelihood of developing gene mutations and tumours in various organs. Studies on human subjects have provided limited and inconsistent evidence of increased risk of developing cancer.\"\n","source_date":"2015-06-04","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260502195628/https://www.efsa.europa.eu/en/topics/topic/acrylamide","calculation_notes":"EFSA's 2015 Scientific Opinion (EFSA Journal 2015;13(6):4104) computed Margins of Exposure (MOE) of 425 for the average adult dietary intake against the neoplastic effects benchmark dose (BMDL10 = 0.17 mg/kg bw/day). MOE values below 10,000 are typically flagged by EFSA as a public-health concern for genotoxic carcinogens, which is why dietary acrylamide is listed as a regulatory priority despite the inconsistent human epidemiology. The position statement deliberately avoids quantifying attributable human cancer cases because the human evidence is too weak; the regulatory action is mitigation-oriented (reducing food levels), not risk-quantifying.\n","independence_note":"EFSA's evaluation is the European regulatory counterpart to FDA action plans. Both agencies synthesize the same underlying epidemiology (Mucci, Hogervorst, Pelucchi cohorts) but make independent policy judgments.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25403648/","title":"Dietary acrylamide and cancer risk: an updated meta-analysis","publisher":"International Journal of Cancer / Pelucchi, Bosetti, Galeone, La Vecchia","source_type":"peer_reviewed","statistic":"Dietary acrylamide not associated with risk of most cancers; borderline RR 1.20 (95% CI 1.00-1.45) for kidney cancer","excerpt":"\"Dietary acrylamide is not related to the risk of most common cancers. A modest association for kidney cancer, and for endometrial and ovarian cancers in never smokers only, cannot be excluded.\"\n","source_date":"2015-06-15","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260510074104/https://pubmed.ncbi.nlm.nih.gov/25403648/","calculation_notes":"Pelucchi 2015 covered 14 cancer sites across dozens of cohort and case-control studies. The pooled relative risks clustered near 1.0 for nearly every cancer examined. Only kidney cancer reached borderline significance (RR 1.20, 95% CI 1.00-1.45) for the highest dietary acrylamide quintile vs the lowest. Endometrial and ovarian cancers showed borderline associations in the never-smoker subgroup only. This is the load- bearing source for framing dietary acrylamide as low-risk at typical intake: the epidemiology examined the largest cancer sites and found no consistent signal. The EFSA Margin of Exposure calculation is consistent with this: the MOE of 425 is a regulatory flag, not a measured human cancer attribution.\n","independence_note":"Pelucchi et al. 2015 updates their 2011 review (Ann Oncol 22:1487-1499, PMID 21239401) with additional prospective cohorts. Methodologically independent from EFSA's regulatory toxicology evaluation.\n"},{"url":"https://www.cancer.gov/about-cancer/causes-prevention/risk/diet/acrylamide-fact-sheet","title":"Acrylamide and Cancer Risk","publisher":"National Cancer Institute (NIH)","source_type":"govt_report","statistic":"Large number of epidemiologic studies in humans found no consistent evidence that dietary acrylamide is associated with the risk of any type of cancer","excerpt":"\"A large number of epidemiologic studies (both case-control and cohort studies) in humans have found no consistent evidence that dietary acrylamide exposure is associated with the risk of any type of cancer.\" Acrylamide develops when \"vegetables that contain the amino acid asparagine, such as potatoes, are heated to high temperatures in the presence of certain sugars.\"\n","source_date":"2017-12-05","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260531014250/https://www.cancer.gov/about-cancer/causes-prevention/risk/diet/acrylamide-fact-sheet","calculation_notes":"NCI's position is the US-side complement to EFSA's evaluation. NCI explicitly notes that the NTP classification (acrylamide is \"reasonably anticipated to be a human carcinogen\") is based primarily on rodent studies, not human evidence. The fact sheet states unambiguously that human epidemiology has not found a consistent dietary cancer link. This is the basis for treating cookware-derived acrylamide as a regulatory- precaution issue rather than a meaningful personal cancer risk at typical intake.\n","independence_note":"NCI synthesis is methodologically distinct from Pelucchi meta-analysis; NCI relies on the full body of NIH-AARP and other US cohort data plus international evidence compilations.\n"}],"comparison_anchors":[{"label":"Heavy dietary acrylamide consumer → kidney cancer (lifetime increment)","lifetime_us_adult":0.0004},{"label":"Baseline kidney cancer (lifetime, US adult)","lifetime_us_adult":0.021},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"regional_breakdown":[{"region":"US adult (typical Western diet, average acrylamide exposure)","probability":0.00006,"notes":"Population-average attributable risk; well below the kidney-cancer borderline signal in Pelucchi 2015"},{"region":"Heavy-fried-food consumer (top dietary intake quartile)","probability":0.0004,"notes":"Pelucchi 2015 top-quartile kidney cancer RR 1.20 applied to SEER baseline 2.1% gives ~0.4% absolute lifetime risk for kidney cancer specifically"},{"region":"Current smoker (cigarette acrylamide intake)","probability":0.001,"notes":"Tobacco smoke is the dominant acrylamide source for smokers; the cancer signal is dominated by other tobacco carcinogens"}],"personal_factor_multipliers":[{"factor":"Heavy fried-food diet (top quartile of dietary intake, ~1 μg/kg bw/day)","multiplier":2,"notes":"Pelucchi 2015 reported RR 1.20 for kidney cancer in the highest dietary acrylamide consumption quintile vs the lowest. Multiplier reflects roughly doubled exposure relative to average diet for those eating french fries, potato chips, or heavily browned baked goods daily.\n"},{"factor":"Pre-existing chronic kidney disease (reduced acrylamide clearance)","multiplier":2,"notes":"Acrylamide and its metabolite glycidamide are primarily eliminated via the kidneys; CKD patients show elevated and prolonged blood acrylamide adducts at the same dietary intake. The cancer-specific multiplier is not well-quantified, but pharmacokinetic modeling suggests roughly doubled internal dose at the same intake level.\n"},{"factor":"Low-acrylamide diet (boiled, steamed, raw foods predominant)","multiplier":0.3,"notes":"Acrylamide forms primarily in high-temperature dry cooking (>120°C) of starchy foods. Boiling, steaming, microwaving, and raw consumption produce negligible acrylamide. Substituting these methods for frying and roasting reduces dietary acrylamide intake by roughly 60-80% per FDA exposure modeling.\n"}],"short_label":"Acrylamide & cancer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"Dietary acrylamide is genuinely genotoxic in rodents and IARC-classified as Group 2A (probably carcinogenic to humans). The translation to typical human dietary exposure is the contested step. The Pelucchi 2015 meta-analysis examined 14 cancer sites across dozens of cohorts and found no consistent association; only kidney cancer reached borderline significance in the highest-exposure quintile, with endometrial and ovarian signals emerging only in never-smokers. EFSA's MOE of 425 is a regulatory flag (below the 10,000 threshold typically used for genotoxic carcinogens) and motivates voluntary food-industry reductions, but EFSA does not quantify attributable human cancer cases. The normalized 6-in-100,000 estimate is a conservative population-average attribution; the heavy-consumer subgroup may experience risk closer to the high end of the uncertainty band. Tobacco smoke dominates total acrylamide exposure for smokers, and the dietary signal cannot be cleanly separated from smoking-related cancer in cohort data. Surveillance gaps: dietary acrylamide intake is hard to estimate from food-frequency questionnaires because acrylamide content varies dramatically with cooking time and temperature for the same food.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A pile of golden-brown french fries on a clean white surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/acrylamide-cooking-cancer","api_url":"https://likelier.app/api/fears/acrylamide-cooking-cancer.json"},{"slug":"car-submersion-drowning","question":"What are the odds of drowning in a sinking car?","quick_answer":"Drowning in a sinking car is rare — on the order of 1 in 16,700 over a lifetime. It looms far larger in the imagination than in the data, partly because submersion is genuinely lethal once it happens even though the event itself almost never does.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"The sinking-car scenario is a fixture of nightmares and driver-anxiety threads: the vehicle goes off a bridge or into a canal, the water rises, the power windows fail, and there is no way out. It is dramatized often enough that many people have a rehearsed mental escape plan for it while having none for the far likelier ways a car kills. The dread is partly rational — submersion is genuinely lethal once it happens — but the event itself is rare, and the perceived probability sits far above the measured one.\n","rough_estimate":"widely feared as a real possibility; the actual per-lifetime risk is on the order of 1 in 17,000","kind":"intuition"},"native":{"display":"~350 US vehicle-submersion drowning deaths per year (5-11% of ~4,345 annual drownings; ~400/yr across North America)","numerator":350,"denominator":335000000,"unit":"per US resident per year","population":"US residents; derived from CDC drowning totals × submersion share (Giesbrecht & McDonald)"},"normalized":{"lifetime_us_adult":0.00006,"display":"~1 in 16,700 lifetime (US resident)","log_value":-4.22,"assumptions":"Giesbrecht & McDonald (2010) estimate ~400 vehicle-submersion deaths per year across North America, accounting for 5-11% of all drownings. The CDC reports an average of 4,345 unintentional drowning deaths per year in the US (2018-2021); the 5-11% submersion share implies ~217-478 US submersion deaths per year, and the US accounts for the large majority of North America's drownings, so ~350/yr is a reasonable central US figure. Spread across ~335 million residents that is ~1.05e-6 per person per year; over a 59-year adult horizon, 1-(1-1.05e-6)^59 ≈ 6.2e-5, or roughly 1 in 16,700. The uncertainty band spans the 5% submersion share (low) to a generous allowance above the 11% share (high), and also absorbs Giesbrecht's observation that submersion deaths are systematically *under*-counted — many never enter crash statistics because no accident report is filed when the outcome is a drowning. This is a population-average figure; driving into floodwater and failure to exit during the float phase move individual risk by close to an order of magnitude (see factors).\n","uncertainty":{"low":0.00003,"high":0.00012},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20681239/","title":"My car is sinking: automobile submersion, lessons in vehicle escape","publisher":"Giesbrecht GG, McDonald GK — Aviation, Space, and Environmental Medicine, 2010 Aug;81(8):779-84","source_type":"peer_reviewed","statistic":"~400 vehicle-submersion deaths per year in North America, 5-11% of all drownings; vehicles float ~15-60s before windows submerge; total time to submersion 37-150s depending on body integrity","excerpt":"\"In North America, approximately 400 individuals per year die in submersed vehicles, accounting for 5-11% of all drownings.\"\n","source_date":"2010-08-01","source_accessed":"2026-06-11","archive_url":"http://web.archive.org/web/20250514200509/https://pubmed.ncbi.nlm.nih.gov/20681239/","calculation_notes":"Primary source for the submersion-specific count and the escape window. The ~400/yr North America figure × the US share of North American drownings gives the ~350/yr US central estimate. The float-phase finding (vehicles stay buoyant for roughly 15-60 seconds before water reaches the windows) underpins the escape-window multiplier: this is the survivable period.\n"},{"url":"https://www.cdc.gov/drowning/data-research/index.html","title":"Drowning Data and Statistics","publisher":"U.S. Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"An average of 4,345 unintentional drowning deaths per year in the US, 2018-2021","excerpt":"\"An average of 4,345 unintentional drowning deaths occurred each year from 2018-2021.\"\n","source_date":"2024-05-01","source_accessed":"2026-06-11","archive_url":"http://web.archive.org/web/20260610133354/https://www.cdc.gov/drowning/data-research/index.html","calculation_notes":"Supplies the US drowning denominator. Applying the 5-11% vehicle-submersion share from Giesbrecht & McDonald to 4,345 annual drownings gives ~217-478 US submersion deaths per year, bracketing the ~350 central estimate and setting the uncertainty range.\n"},{"url":"https://www.weather.gov/safety/flood-turn-around-dont-drown","title":"Turn Around Don't Drown — Flood Safety","publisher":"National Weather Service (NOAA / NWS)","source_type":"govt_report","statistic":"Over half of flood-related drownings occur when a vehicle is driven into floodwater; ~12 inches of moving water can carry away most cars","excerpt":"\"Over half of all flood-related drownings occur when a vehicle is driven into hazardous flood water. It takes just 12 inches of rushing water to carry away most cars and just 2 feet of rushing water can carry away SUVs and trucks.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-11","archive_url":"https://web.archive.org/web/20260620135136/https://www.weather.gov/safety/flood-turn-around-dont-drown","calculation_notes":"Establishes that driving into floodwater is the dominant submersion mechanism, not bridges and canals. Basis for the floodwater multiplier: the single largest controllable risk factor is the choice to enter a flooded roadway.\n"}],"comparison_anchors":[{"label":"Drowning (general population, lifetime US adult)","lifetime_us_adult":0.00071},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Recreational boating drowning (regular boater, lifetime)","lifetime_us_adult":0.00146}],"personal_factor_multipliers":[{"factor":"driving into a flooded roadway or floodwater","multiplier":8,"notes":"NWS: over half of flood-related drownings occur when a vehicle is driven into floodwater, and 12 inches of moving water can carry away most cars. Entering a flooded road is the dominant — and avoidable — submersion scenario."},{"factor":"exits during the float phase (knows the protocol, carries a window-breaker)","multiplier":0.15,"notes":"Giesbrecht & McDonald: vehicles float ~15-60 seconds before water reaches the windows. Escaping in that window — seatbelts off, window down or broken, children first — is survivable; the protocol exists precisely because the survivable period is short."},{"factor":"delayed escape after the cabin fills, or deep / moving water","multiplier":3,"notes":"Once the vehicle submerges (37-150s to full submersion), pressure makes doors very hard to open, electric windows may fail, and disorientation rises sharply. Most preventable deaths occur because escape is attempted too late."}],"short_label":"Car submersion","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Submersion deaths are under-counted. As Giesbrecht & McDonald note, when the outcome is a drowning, local police often do not file a crash report, and a vehicle pulled from water by a fire department may never be coded as a submersion crash — so the true figure likely sits toward the upper end of the range. The US central estimate (~350/yr) is derived by apportioning the ~400/yr North America figure and cross-checking against the CDC drowning total times the 5-11% submersion share; it is not a directly published US count. The escape-window and floodwater multipliers are grounded in the NWS and Giesbrecht findings but are directional rather than precise relative risks. The dramatized scenario (off a bridge into deep water) is real but a minority of cases; the statistically dominant one is mundane — a driver entering a flooded road that looks shallower than it is.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-06-11","methodology_version":"1.0"},"reviewer":"8d-eval-2026-06-11","last_reviewed":"2026-06-11","reviewed":true,"generated_at":"2026-06-11","image":{"alt":"Side view of a car with a calm horizontal waterline across its lower doors, flat vector illustration in muted blue-grey tones."},"canonical_url":"https://likelier.app/car-submersion-drowning","api_url":"https://likelier.app/api/fears/car-submersion-drowning.json"},{"slug":"chiropractic-cervical-manipulation","question":"What are the odds of stroke or serious injury from chiropractic neck manipulation?","quick_answer":"The stroke risk from chiropractic neck manipulation is very low per session — about 1 to 3 events per million manipulations, or roughly 1 in 17,000 over the lifetime of a regular chiropractic user.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The association between cervical manipulation and stroke has generated headlines, lawsuits, and polarized professional debate for decades. Case reports of vertebral artery dissection following chiropractic neck adjustment have created a perception that the procedure carries a meaningful stroke risk. The chiropractic profession argues that the association is confounded — patients with pre-existing arterial dissection seek care for neck pain before their stroke, creating a temporal but not causal link. Critics argue that even if the absolute risk is small, it is not zero, and that cervical manipulation offers no demonstrated superiority over physiotherapy for neck or back pain, making any stroke risk unacceptable on risk-benefit grounds.\n","rough_estimate":"Widely perceived as carrying a small but real stroke risk; some patients avoid chiropractic care entirely","kind":"intuition"},"native":{"display":"~1-3 vertebrobasilar events per million cervical manipulations","numerator":2,"denominator":1000000,"unit":"per cervical manipulation session","population":"Patients receiving cervical spine manipulation from chiropractors"},"normalized":{"lifetime_us_adult":0.00006,"display":"~1 in 17,000 (lifetime stroke risk from cervical manipulation, for a regular chiropractic user)","log_value":-4.22,"assumptions":"Published estimates of vertebrobasilar stroke following cervical manipulation range from 1 per 400,000 to 1 per 5.85 million manipulations, with most reviews converging on approximately 1-3 per million. Using 2 per million as the central estimate.\nThe lifetime estimate depends heavily on exposure assumptions. A \"regular chiropractic user\" who receives approximately 20 cervical manipulations per year for 30 years (600 lifetime manipulations) would have a cumulative risk of: 1 − (1 − 0.000002)^600 ≈ 0.0012, or roughly 1 in 830.\nHowever, the average American receives far fewer lifetime cervical manipulations. Approximately 10% of US adults visit a chiropractor in any given year, and not all visits involve cervical manipulation. A more representative lifetime exposure for the general population might be 30 cervical manipulations. At 2 per million per session: 1 − (1 − 0.000002)^30 ≈ 0.00006, or roughly 1 in 17,000.\nThe Cassidy et al. 2008 case-control study found no excess risk of vertebrobasilar stroke after chiropractic visit compared with primary care visit, suggesting the association may be entirely confounded by pre-existing dissection. If this interpretation is correct, the causal risk is closer to zero.\n","uncertainty":{"low":0.00001,"high":0.0003},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2922298/","title":"Current understanding of the relationship between cervical manipulation and stroke: what does it mean for the chiropractic profession?","publisher":"Chiropractic & Manual Therapies (PMC)","source_type":"peer_reviewed","statistic":"Published incidence estimates diverge widely: a Swiss survey put 'important complications' at 1:400,000, while Danish chiropractors estimated one case per 1.3 million cervical treatment sessions","excerpt":"\"From these data they estimated an incidence of one case per 362 chiropractor years and one case per 1.3 million cervical treatment sessions.\"\n","source_date":"2010-06-16","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420033432/https://pmc.ncbi.nlm.nih.gov/articles/PMC2922298/","calculation_notes":"This review surveys the range of published risk estimates, which diverge widely by methodology (malpractice claims, practitioner surveys, population studies): a Swiss survey put \"important complications\" at 1:400,000, while Danish chiropractors estimated one case per 1.3 million cervical treatment sessions. The paper does not claim a consensus figure; the spread is roughly two orders of magnitude, so only the order of magnitude (about 1 per million) is reasonably well supported. The review also notes the Cassidy 2008 finding of no excess risk compared to PCP visits, which, if accepted, would move the causal estimate toward zero.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4794386/","title":"Systematic Review and Meta-analysis of Chiropractic Care and Cervical Artery Dissection: No Evidence for Causation","publisher":"Cureus (PMC)","source_type":"peer_reviewed","statistic":"Meta-analysis found no evidence that chiropractic care causes cervical artery dissection; similar associations found for PCP visits","excerpt":"\"The meta-analysis revealed a small association between chiropractic care and dissection ... The association for chiropractor visit was not greater than for PCP visit. ... There is no convincing evidence to support a causal link between chiropractic manipulation and CAD.\"\n","source_date":"2016-02-16","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420033506/https://pmc.ncbi.nlm.nih.gov/articles/PMC4794386/","calculation_notes":"Church et al. 2016 is the most comprehensive meta-analysis to date. The finding of \"no evidence for causation\" does not mean the risk is zero — absence of evidence is not evidence of absence, and case reports of temporal association are plentiful. The methodological debate centers on whether case-control studies adequately control for the confounding of pre-existing dissection driving both the chiropractic visit (neck pain) and the subsequent stroke. The conservative interpretation: the causal contribution of cervical manipulation, if any, is too small to detect in population-level studies.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10954208/","title":"Plausible Mechanisms of Causation of Immediate Stroke by Cervical Spine Manipulation: A Narrative Review","publisher":"Cureus (PMC)","source_type":"peer_reviewed","statistic":"Biomechanical analysis identifies plausible mechanisms by which cervical manipulation could cause vertebral artery injury","excerpt":"\"The findings of Cassidy et al. support that CSM does not cause cervical artery dissection in cases of post-manipulative stroke and that cervical artery dissection is likely pre-existing to the CSM.\"\n","source_date":"2024-03-18","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420033539/https://pmc.ncbi.nlm.nih.gov/articles/PMC10954208/","calculation_notes":"This 2024 narrative review examines the biomechanical plausibility of cervical manipulation causing arterial injury. While it acknowledges plausible mechanisms exist, it also notes that the epidemiological evidence (Cassidy 2008, Church 2016) does not support a causal relationship at the population level. The review highlights the fundamental difficulty: case reports demonstrate temporal association, but controlled studies do not show excess risk compared to other healthcare encounters. The risk-benefit question is therefore not about whether cervical manipulation causes stroke (unclear), but whether it offers benefits superior to alternatives with no plausible stroke mechanism (physiotherapy, exercise).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC81498/","title":"Arterial dissections following cervical manipulation: the chiropractic experience","publisher":"CMAJ (Canadian Medical Association Journal) — Haldeman S, Carey P, Townsend M, Papadopoulos C","source_type":"peer_reviewed","statistic":"Canadian Chiropractic Protective Association malpractice-claims data (1988-1997, ~134.5 million estimated cervical manipulations): 23 reported cases of stroke or vertebral artery dissection, yielding an estimated rate of 1 per 5.85 million cervical manipulations (1 per 8.06 million office visits)","excerpt":"\"The likelihood that a chiropractor will be made aware of an arterial dissection following cervical manipulation is approximately 1:8.06 million office visits, 1:5.85 million cervical manipulations, 1:1430 chiropractic practice years and 1:48 chiropractic practice careers.\"\n","source_date":"2001-10-02","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260217215332/https://pmc.ncbi.nlm.nih.gov/articles/PMC81498/","calculation_notes":"Haldeman et al. analyzed malpractice-claims data from the Canadian Chiropractic Protective Association covering 1988-1997: 23 confirmed cases of stroke or vertebral artery dissection following cervical manipulation, against an estimated 134.5 million cervical manipulations performed over the period (derived from a survey of a 10% sample of practicing Canadian chiropractors, 77.8% response rate). This yields the 1-per-5.85-million-manipulations figure cited in the body text as the low end of the published risk-estimate range, alongside the Swiss 1:400,000 \"important complications\" estimate from the first source above. The authors note this insurance-claims-based rate is substantially lower than estimates calculated from surveys of neurologists (1:500,000-1,000,000), a discrepancy attributed to underreporting to malpractice insurers versus recall in physician surveys.\n"}],"comparison_anchors":[{"label":"General anesthesia death (per procedure)","lifetime_us_adult":0.0001},{"label":"NSAID-related GI bleed death (per year of use)","lifetime_us_adult":0.001},{"label":"Vertebrobasilar stroke (all causes, lifetime)","lifetime_us_adult":0.005}],"personal_factor_multipliers":[{"factor":"regular chiropractic user (20+ sessions/year)","multiplier":20,"notes":"Cumulative exposure over decades substantially increases lifetime risk. A patient receiving 20 cervical manipulations per year for 30 years has roughly 600 lifetime exposures."},{"factor":"connective tissue disorder (Ehlers-Danlos, Marfan)","multiplier":10,"notes":"Patients with connective tissue disorders have inherently fragile arterial walls and higher baseline risk of spontaneous dissection. Cervical manipulation is generally contraindicated."},{"factor":"single chiropractic visit","multiplier":0.05,"notes":"A single cervical manipulation session carries a risk of approximately 1-3 per million — essentially negligible for a one-time exposure."}],"short_label":"Chiropractic neck manipulation","myth_framing":"overrated","outcome_severity":"serious_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"The risk estimate in this entry is conditional on receiving cervical manipulation. For people who never receive cervical manipulation, the risk is zero by definition. The \"overrated\" framing refers to the perception that chiropractic neck manipulation carries a high stroke risk — the absolute risk per session, if causal, is in the 1-in-a-million range. However, the risk-benefit framing is the critical context: if cervical manipulation offers no benefit over physiotherapy or exercise for neck pain (as multiple Cochrane reviews suggest), then any non-zero stroke risk — even a very small one — shifts the risk-benefit ratio unfavorably. The entry does not address lumbar or thoracic manipulation, which do not carry the same vascular risk. The Cassidy 2008 and Church 2016 findings of no excess risk compared to PCP visits remain methodologically debated and should not be interpreted as proof that the risk is zero.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified anatomical diagram of cervical vertebrae and vertebral artery, flat vector in muted clinical tones."},"canonical_url":"https://likelier.app/chiropractic-cervical-manipulation","api_url":"https://likelier.app/api/fears/chiropractic-cervical-manipulation.json"},{"slug":"landmine-uxo-injury","question":"What are the odds of being killed or injured by a landmine or unexploded ordnance?","quick_answer":"Across the global adult population, the lifetime odds of being killed or injured by a landmine or unexploded ordnance are about 1 in 14,700, with roughly a third of casualties fatal. Most people outside affected regions assume the problem is historical, but annual casualties have risen sharply since 2015; the risk is concentrated in a few conflict and post-conflict countries and is near zero elsewhere.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Landmines and unexploded ordnance occupy a strange perceptual blind spot: widely understood as devastating when encountered, but filed as a problem of \"other places and other eras\" by most people outside affected regions. The Ottawa Treaty, signed in 1997, created the impression that the problem was being solved. It is not. Annual casualty counts have risen sharply since 2015, driven by new conflicts in Myanmar, Syria, Ukraine, and Yemen, and by the sheer persistence of ordnance laid decades ago in Cambodia, Laos, and Afghanistan.\n","rough_estimate":"most people in unaffected countries would guess near zero; the problem is perceived as historical rather than ongoing","kind":"intuition"},"native":{"display":"~5,757 casualties per year globally (2023, Landmine Monitor)","numerator":5757,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.0000679,"display":"1 in ~14,700 lifetime (global adult)","log_value":-4.17,"assumptions":"The Landmine Monitor 2024 report documented 5,757 casualties (killed and injured) from mines and explosive remnants of war in 2023. The 2025 report recorded 6,279 casualties in 2024. Using the 2023 figure as central estimate: annual rate = 5,757 / 5,000,000,000 = 1.15 × 10⁻⁶. Compounded over 59 years: 1 − (1 − 1.15e-6)^59 ≈ 6.79 × 10⁻⁵, i.e. roughly 1 in 14,700. The uncertainty band uses a low of ~4,000 casualties/year (pre-2020 baseline, low: 4.72e-5) and a high of ~12,000/year reflecting severe escalation scenarios in Myanmar, Ukraine, and Sahel (high: 1.42e-4; ratio ~3x). Given the sharp upward trend since 2015 and continued new use by non-signatories, the upper tail represents a plausible escalation path.\n","uncertainty":{"low":0.0000472,"high":0.000142},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.hrw.org/news/2024/11/19/landmines-new-casualties-show-need-support-treaty-ban","title":"Landmines: New Casualties Show Need to Support Treaty Ban","publisher":"Human Rights Watch (reporting Landmine Monitor 2024 findings)","source_type":"reputable_reference","statistic":"5,757 new casualties from landmines and explosive remnants of war were recorded in 2023 (including 1,983 deaths), with civilians accounting for 84% of casualties and children comprising 37% of casualties where age was recorded; 164 countries have joined the Mine Ban Treaty, including every NATO member except the United States, while Russia (in Ukraine, since Feb. 2022) and Myanmar's junta (continuously since 1999) have kept using antipersonnel mines","excerpt":"\"There were at least 5,757 new casualties from landmines and explosive remnants of war in 53 countries and 2 other areas in 2023, including 1,983 deaths. Civilians made up 84 percent of all recorded casualties in 2023, while children were 37 percent of casualties when the age was recorded.\" \"A total of 164 countries have joined the treaty, including all NATO members except the United States.\" \"Russia has used antipersonnel landmines repeatedly in Ukraine since its full-scale invasion of the country on February 24, 2022.\" \"The Myanmar junta's armed forces have used antipersonnel landmines continuously since 1999.\"\n","source_date":"2024-11-19","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260616063438/https://www.hrw.org/news/2024/11/19/landmines-new-casualties-show-need-support-treaty-ban","calculation_notes":"Human Rights Watch (an ICBL co-founder) reports the Landmine Monitor 2024 finding — the authoritative global tracker for mine/ERW casualties, published annually by the ICBL (1997 Nobel Peace Prize co-laureate). At 5,757 casualties/year over a global adult population of 5 billion, the annual rate is 1.15e-6; compounded over 59 years: ~6.79e-5. The same article also grounds the body text's claim that more than 30 states have never joined the treaty (164 parties against the world's roughly 195 states leaves ~31 outside it), naming the United States as the one NATO non-member and Russia and Myanmar as non-signatories actively using antipersonnel mines in current conflicts.\n","independence_note":"Human Rights Watch's reporting of the Landmine Monitor 2024 is an independent civil-society account, methodologically distinct from the MAG/Landmine Monitor 2025 source below (different publisher, different report year).\n"},{"url":"https://www.maginternational.org/whats-happening/2025-landmine-monitor-out-now/","title":"2025 Landmine Monitor out now","publisher":"MAG (Mines Advisory Group)","source_type":"reputable_reference","statistic":"In 2024, a total of 6,279 people were killed or injured by unexploded ordnance; Myanmar recorded 2,029 casualties, the highest in the world for a second consecutive year","excerpt":"\"In 2024, a total of 6,279 people were killed or injured by unexploded ordnance... Myanmar once again suffered the highest number of casualties, with 2,029 people killed or injured in 2024.\"\n","source_date":"2025-11-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260525162002/https://www.maginternational.org/whats-happening/2025-landmine-monitor-out-now/","calculation_notes":"MAG International (a leading humanitarian demining NGO) reports on the Landmine Monitor 2025 findings: 6,279 total casualties in 2024, with Myanmar alone recording 2,029. This corroborates the upward casualty trend and provides the 2024 figure, higher than the 2023 baseline used in our central estimate, supporting the upper range of the uncertainty band.\n","independence_note":"MAG is an independent demining organization; its reporting on the Landmine Monitor 2025 is methodologically distinct from the ICBL/ReliefWeb source above (different publisher, independent commentary on the same report).\n"},{"url":"https://www.govinfo.gov/content/pkg/CHRG-111hhrg56094/html/CHRG-111hhrg56094.htm","title":"Legacies of War: Unexploded Ordnance in Laos","publisher":"U.S. House of Representatives, Subcommittee on Asia, the Pacific, and the Global Environment","source_type":"govt_report","statistic":"Estimates of unexploded cluster submunitions remaining in Laos range from 8 million to 80 million, with less than 1/2 of 1% destroyed and less than 1% of contaminated land cleared as of 2010; the US bombing campaign ended in 1973","excerpt":"\"Estimates of the number of unexploded submunitions from cluster bombs range from 8 million to 80 million, with less than 1/2 of 1 percent destroyed, and less than 1 percent of contaminated lands cleared.\" \"The bombing finally ended in 1973.\"\n","source_date":"2010-04-22","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714014503/https://www.govinfo.gov/content/pkg/CHRG-111hhrg56094/html/CHRG-111hhrg56094.htm","calculation_notes":"Official congressional hearing record, cited to substantiate the body text's \"legacy burden\" claim about Laos's unexploded cluster-submunition contamination from the 1964-1973 US bombing campaign. Not used in the headline native/normalized arithmetic, which rests on the Landmine Monitor casualty counts above; the 8-80 million range (not a single point figure) is why the body text now cites the full range rather than the upper bound alone.\n","independence_note":"U.S. congressional hearing record, independent of the ICBL/HRW and MAG Landmine Monitor reporting used for the headline casualty figures.\n"},{"url":"https://english.news.cn/20240314/80605b49fffa458a918a470848dc615d/c.html","title":"Cambodia records 32 landmine, ERW casualties in 2023, down 22 pct","publisher":"Xinhua","source_type":"news_article","statistic":"Cambodia recorded 32 landmine and ERW casualties in 2023, a 22 percent decline from 41 in 2022","excerpt":"\"Cambodia reported 32 landmine and explosive remnant of war (ERW) casualties in 2023, a 22 percent decline from 41 in 2022.\"\n","source_date":"2024-03-14","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250614084033/https://english.news.cn/20240314/80605b49fffa458a918a470848dc615d/c.html","calculation_notes":"Cited only to substantiate the body text's claim about Cambodia's ongoing annual landmine/ERW casualty toll — a legacy of the Khmer Rouge era and earlier Indochina wars, now running in the dozens per year rather than the hundreds. Not used in the headline global native/normalized arithmetic, which rests on the Landmine Monitor global casualty counts above.\n","independence_note":"Xinhua's reporting on Cambodia's national mine-action authority statistics is independent of the ICBL/HRW and MAG global Landmine Monitor reporting used for the headline casualty figures.\n"}],"comparison_anchors":[{"label":"Death in a terrorist attack (lifetime, global adult)","lifetime_us_adult":0.0000048},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.0000065},{"label":"Death in a house fire (lifetime, US adult)","lifetime_us_adult":0.00025}],"regional_breakdown":[{"region":"Myanmar (active conflict zones)","probability":0.0019,"notes":"2,029 casualties in 2024 alone; highest per-capita mine casualty rate in the world. The civil war has massively expanded mine use since 2021."},{"region":"Post-conflict countries (Cambodia, Laos, Afghanistan)","probability":0.00015,"notes":"Legacy contamination from decades-old conflicts; casualties occur primarily among rural agricultural workers and children."},{"region":"Countries without mine contamination","probability":1e-9,"notes":"Effectively zero risk; sporadic incidents involve historical ordnance from World Wars discovered during construction."}],"short_label":"Landmine or UXO injury","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"Risk is overwhelmingly concentrated in a small number of conflict-affected and post-conflict countries. Myanmar, Syria, Ukraine, Afghanistan, Cambodia, Laos, and Yemen account for the vast majority of global casualties. For residents of countries that have never experienced significant mine contamination, personal risk is effectively zero. The casualty figures include both deaths and injuries — roughly a third of recorded casualties are fatal (1,983 deaths of 5,757 in 2023). The figures also likely undercount the true toll, as many incidents in active conflict zones go unreported. Children are disproportionately represented among casualties because they are less likely to recognize ordnance and more likely to pick up unfamiliar objects.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":4,"avg":3.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized warning sign silhouette against a barren landscape, flat vector illustration in muted warning tones."},"canonical_url":"https://likelier.app/landmine-uxo-injury","api_url":"https://likelier.app/api/fears/landmine-uxo-injury.json"},{"slug":"child-bike-trailer-injury","question":"What are the odds of serious injury to a child riding in a towed bicycle trailer?","quick_answer":"A child's odds of an ER-treated injury while riding in a towed bike trailer are roughly 1 in 14,000 over a typical few years of use — about 6x lower than a rear-mounted bike seat. The trailer looks fragile, but the injury data runs the other way.\n","category":"transport","tags":["infant","toddler","child","kids","travel"],"no_reliable_estimate":false,"perceived":{"description":"A child trailer rides low to the ground, often below the eyeline of drivers in cars or SUVs, behind a parent who cannot see the trailer without turning around. Parents fear the trailer is invisible at intersections, that a hitch failure could detach it into traffic, or that a sideswipe by a passing vehicle would crush a small fabric-and-aluminium box. The visual of a toddler trailing behind a moving bicycle, separated from the parent by a rigid bar, reads as fragile in a way a child carried in arms does not, even when the actual injury data points the other direction.\n","rough_estimate":"~3-5% chance of serious injury over a typical childhood as trailer passenger","kind":"intuition"},"native":{"display":"~322 ER-treated injuries over 9 years (NEISS-extrapolated, US children in bicycle-towed trailers)","numerator":322,"denominator":1500000,"unit":"per 9-year US trailer-passenger exposure cohort","population":"US children <5 transported as passengers in bicycle-towed trailers 1990-1998"},"normalized":{"lifetime_us_adult":0.00007,"display":"~1 in 14,000 chance of ER-treated injury over a child's typical 3-year exposure as a bike-trailer passenger — roughly 6x lower than rear-mounted seats","log_value":-4.15,"assumptions":"Powell & Tanz (2000) extrapolated 322 trailer injuries (95% CI 158-486) over 9 years (1990-1998) from NEISS — only 6 observed cases, hence the very wide CI — versus an estimated 2,015 mounted-seat injuries (95% CI 988-3,042). The one firmly grounded quantity is the ratio: 2,015 / 322 = 6.3x fewer injuries in trailers than in mounted seats. Because the true exposure denominator (US child-trailer-passenger-years) is not directly counted, this entry does not attempt an independent absolute per-exposure derivation; instead it anchors the trailer figure to the mounted-seat entry's normalized rate (0.0004 over a typical 3-year window) divided by that grounded ~6x differential, giving ~0.00007 (about 1 in 14,000). The wide uncertainty band (0.00003-0.0003) reflects the unknown denominator plus the CI from only 6 observed cases. Counterbalancing the low frequency: when trailer injuries do occur, motor vehicles are involved 33% of the time (vs 9% for mounted seats) and head/face injuries are present in 83% of cases (vs 49% for seats), so per-injury severity skews higher.\n","uncertainty":{"low":0.00003,"high":0.0003},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/10768671/","title":"Tykes and bikes: injuries associated with bicycle-towed child trailers and bicycle-mounted child seats","publisher":"Archives of Pediatrics & Adolescent Medicine (Powell & Tanz)","source_type":"peer_reviewed","statistic":"Estimated 322 trailer injuries (95% CI 158-486) over 9 years vs 2,015 mounted-seat injuries; 33% of trailer injuries involved motor vehicles; 83% of trailer cases had head/face injuries","excerpt":"\"49 injuries to children during the 9-year study period [were identified]: 6 were associated with bicycle-towed trailers (estimated 322 injuries; 95% CI, 158-486) and 43 were related to bicycle-mounted child seats (estimated 2,015 injuries; 95% CI, 988-3,042). The mean age of injured children was 2.4 years and 51% were male. For trailers, motor vehicle collisions accounted for 33% of injuries and falls 50%; for mounted seats, 9% involved motor vehicles and 72% were falls. Head/face injuries: 83% (trailers) vs 49% (seats).\"\n","source_date":"2000-04-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250822040024/https://pubmed.ncbi.nlm.nih.gov/10768671/","calculation_notes":"Powell & Tanz observed 6 trailer cases in the NEISS sentinel hospitals, which extrapolated to 322 nationally over 9 years — about 36 per year — versus 2,015 for mounted seats. The firmly grounded quantity is the ratio (2,015 / 322 = 6.3x), which is the headline finding. The absolute normalized figure (~1 in 14,000 over a 3-year window) is not derived independently from an exposure denominator (that count does not exist); it is anchored to the mounted-seat entry's 3-year rate (0.0004) divided by the grounded ~6x differential. Wide CI is unavoidable with only 6 observed cases.\n"},{"url":"https://www.healthychildren.org/English/safety-prevention/at-play/Pages/baby-on-board-keeping-safe-on-a-bike.aspx","title":"How to Protect Child Passengers on Adult Bikes","publisher":"AAP / HealthyChildren.org","source_type":"reputable_reference","statistic":"AAP guidance: bike-towed trailers preferred over mounted seats; infants <12 months too young to be passengers; helmets required for both child and adult","excerpt":"\"Preferably, children should ride in a bicycle-towed child trailer rather than a bicycle-mounted child seat. Infants younger than 12 months are too young to sit in a rear bike seat or to wear a bicycle helmet.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260429193751/https://www.healthychildren.org/English/safety-prevention/at-play/Pages/baby-on-board-keeping-safe-on-a-bike.aspx","calculation_notes":"AAP positions trailers as the safer of the two passenger configurations without publishing a quantitative differential. The Powell & Tanz 6x ratio is consistent with the qualitative AAP preference. Helmet requirement applies inside trailers as well as on mounted seats.\n"},{"url":"https://www.cpsc.gov/Recalls/2016/Burley-Design-Recalls-Child-Bicycle-Trailers","title":"Burley Design Recalls Child Bicycle Trailers Due to Crash Hazard","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"35 reports of trailer tow-bar receivers separating; mechanism is detachment from towing bicycle","excerpt":"\"Burley Design has received 35 reports of trailers with black plastic tow bar receivers separating from the tow bar, including two incidents that resulted in abrasions to a child.\"\n","source_date":"2016-09-21","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260215145608/https://www.cpsc.gov/Recalls/2016/Burley-Design-Recalls-Child-Bicycle-Trailers","calculation_notes":"CPSC recall illustrates the trailer-specific failure mode (hitch receiver separation) that has no analogue on mounted seats. Useful mechanism context for why the trailer injury count, while low, can include catastrophic detachment events. Not a numerator source for the headline rate.\n"}],"comparison_anchors":[{"label":"Child rear bike-seat injury (typical 3-yr exposure)","lifetime_us_adult":0.0004},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"child wearing a properly-fitted bicycle helmet","multiplier":0.3,"notes":"Same as mounted-seat reasoning; AAP requires helmet inside trailer"},{"factor":"riding on shared road with motor traffic","multiplier":3,"notes":"33% of trailer injuries are MV-involved vs 9% for mounted seats — when trailers fail it's more often a car strike, with much higher severity"},{"factor":"trailer-mounted high-visibility flag installed","multiplier":0.6,"notes":"Biomechanically defensible; trailers sit low and drivers may not see them — no published RCT data on flag effectiveness"},{"factor":"towing bicycle's hitch not regularly inspected (recalled units)","multiplier":2,"notes":"Burley 2016 recall covered 35 separation events; periodic inspection mitigates a known failure mode"},{"factor":"single-passenger vs dual-passenger trailer","multiplier":0.7,"notes":"Single-passenger trailers have lower CoG; mechanism not well-quantified but biomechanically defensible"}],"short_label":"Child bike trailer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1 in 14,000 figure rests on only 6 observed cases in the NEISS sample, so the confidence interval is wide and the headline rate could shift substantially with additional data. The denominator (US child-trailer- passenger-years) is no better counted than for mounted seats. The most important nuance is severity composition: trailers produce fewer injuries in absolute terms, but the injuries that do happen are more often motor-vehicle-involved (33% vs 9%) and more often head/face (83% vs 49%). The Powell & Tanz cohort also pre-dates current hitch standards and flag-and-mirror visibility practices, so a contemporary re-analysis might shift the differential further. Northern European bakfiets and front-box cargo-bike configurations are not captured. The trailer rate also does not reflect the modern thru-axle hitch designs that have superseded the older Burley-style plastic receivers implicated in the 2016 recall.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"A bicycle-towed child trailer parked at the side of a residential path, viewed from the side, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-bike-trailer-injury","api_url":"https://likelier.app/api/fears/child-bike-trailer-injury.json"},{"slug":"carbon-monoxide-poisoning","question":"What are the odds of dying from carbon monoxide poisoning?","quick_answer":"The lifetime odds of dying from accidental non-fire carbon monoxide poisoning are roughly 1 in 14,000 for a US adult. Many people treat the risk as effectively zero, but it is a real low-level tail hazard concentrated around generators and furnaces.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Carbon monoxide is the textbook \"silent killer,\" but day-to-day it lives in a strange psychological gap. Most people know the phrase, most people have seen the round white detector on a hardware-store shelf, and most people have not actually thought about whether one is installed on the floor where they sleep. There is no clean Chapman-style survey of CO fear, so the best we can say is that the perceived tail risk for healthy adults sits somewhere between \"vaguely aware\" and \"effectively zero\" — much closer to food poisoning than to house fire on the worry ladder.\n","rough_estimate":"Most adults treat the lifetime fatal risk as effectively zero outside of generators and furnaces","kind":"intuition"},"native":{"display":"~400 US accidental non-fire CO deaths per year","numerator":1,"denominator":825000,"unit":"per year","population":"US residents, all ages, unintentional non-fire-related carbon monoxide poisoning"},"normalized":{"lifetime_us_adult":0.0000714,"display":"1 in ~14,000 lifetime (US adult)","log_value":-4.15,"assumptions":"Uses CDC's current public-facing figure of \"more than 400\" annual unintentional non-fire CO deaths in the US, against a population of ~330 million, giving an annual rate of roughly 1.21 per million (≈ 0.12 per 100,000). Compounded over 59 years of remaining adult life: 1 - (1 - 1.21e-6)^59 ≈ 7.14e-5, or about 1 in 14,000. The Hampson 2016 review of US CO mortality 1999-2012 lands in the same neighborhood at 1.46 unintentional non-fire-related CO deaths per million population per year. The Iqbal MMWR series from 1999-2010 puts the average at 430 deaths per year, also consistent. Excludes intentional self-harm (which is the larger CO mortality bucket in raw counts) and CO deaths secondary to structural fires (which are coded under fire mortality).\n","uncertainty":{"low":0.00005,"high":0.0001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/carbon-monoxide/about/index.html","title":"About Carbon Monoxide Poisoning","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"More than 400 Americans die from unintentional CO poisoning not linked to fires; more than 100,000 visit an ED; more than 14,000 are hospitalized each year","excerpt":"\"more than 400 Americans die from unintentional CO poisoning not linked to fires, more than 100,000 visit an emergency department, and more than 14,000 are hospitalized.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260408081909/https://www.cdc.gov/carbon-monoxide/about/index.html","calculation_notes":"The 400/year figure is the canonical CDC headline. 400 / 3.3e8 ≈ 1.21 per million per year. Over 59 adult-remaining years: 1 - (1 - 1.21e-6)^59 ≈ 7.14e-5, or about 1 in 14,000 lifetime. The accompanying 100,000 ED visits and 14,000 hospitalizations figures imply a case-fatality ratio for medically attended CO exposure of roughly 0.4%, which is consistent with the bulk of presenting cases being mild headache or nausea rather than severe CNS toxicity.\n","independence_note":"CDC's public-facing figures are built on NVSS death-certificate and HCUP hospitalization data — the same upstream the CDC MMWR QuickStats and Hampson 2016 reanalysis use. Treat the three CDC/NVSS-based sources as one institutional pipeline; CPSC's product-incident database is the complementary source with product-identification metadata.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6303a6.htm","title":"QuickStats: Average Annual Number of Deaths and Death Rates from Unintentional, Non-Fire-Related Carbon Monoxide Poisoning, by Sex and Age Group — United States, 1999-2010","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"5,149 deaths from unintentional CO poisoning during 1999-2010, an average of 430 deaths per year; male death rate 0.22 per 100,000 vs female 0.07","excerpt":"\"During 1999-2010, a total of 5,149 deaths from unintentional carbon monoxide poisoning occurred in the United States, an average of 430 deaths per year. The average annual death rate from carbon monoxide poisoning for males (0.22 per 100,000 population) was more than three times higher than that for females (0.07). The death rates were highest among those aged ≥65 years for males (0.42) and females (0.18).\"\n","source_date":"2014-01-24","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260328071329/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6303a6.htm","calculation_notes":"The 1999-2010 average of 430 deaths/year from MMWR is consistent with the current CDC public-facing \"more than 400\" figure, and confirms that the rate has been roughly flat for two decades. The threefold male/female disparity and the strong age skew toward 65+ are the two most load-bearing pieces of heterogeneity for any reader trying to localize this number to themselves.\n","independence_note":"Both CDC sources draw from National Vital Statistics System (NVSS) death certificate data — treat as the same upstream, used here as historical-vs-current confirmation rather than two independent estimates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4573527/","title":"Carbon monoxide poisoning deaths in the United States, 1999 to 2012","publisher":"American Journal of Emergency Medicine / PMC (Hampson NB)","source_type":"peer_reviewed","statistic":"6,136 CO poisoning fatalities 1999-2012, average 438 per year; unintentional non-fire-related death rate 1.46 per million population per year; 54% of deaths in a home","excerpt":"\"For this study, we identified 6136 CO poisoning fatalities during 1999 to 2012 resulting in an average of 438 deaths annually. The annual average age-adjusted death rate was 1.48 deaths per million. [...] The annual average age-adjusted UNFR CO poisoning DR for the study period (1999-2012) was 1.46 deaths per million. [...] Fifty four percent of the deaths occurred in a home.\"\n","source_date":"2016-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164530/https://pmc.ncbi.nlm.nih.gov/articles/PMC4573527/","calculation_notes":"Hampson 2016 in AJEM is the canonical peer-reviewed CO mortality paper for this period. His 1.46 per million unintentional non-fire-related rate is essentially identical to the rate implied by the CDC \"more than 400\" figure (~1.21 per million using 2024 population), and the small gap is consistent with the 1999-2012 study window covering a slightly higher-rate era and a smaller US population. The \"54% in a home\" figure is what makes the CO detector intervention so leveraged: more than half of the deaths occur in exactly the place where a $20 alarm would have caught the exposure.\n","independence_note":"Hampson reanalyzes NVSS death certificate data, so it shares the upstream with the CDC sources. Used here as peer-reviewed confirmation of the death-rate magnitude and as the primary source for the residential-share statistic.\n"},{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2023/New-CPSC-Report-Shows-Upward-Trend-in-Carbon-Monoxide-CO-Fatalities","title":"New CPSC Report Shows Upward Trend in Carbon Monoxide (CO) Fatalities","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"765 portable-generator-related non-fire CO deaths 2009-2019 (~70/year), accounting for 40% of consumer-product CO deaths under CPSC jurisdiction; 250 estimated consumer-product CO deaths in 2019, the highest of any year in the report","excerpt":"\"Since 2009, portable generators alone have been associated with an estimated 765 non-fire CO poisoning deaths, accounting for 40 percent of all CO deaths related to consumer products under CPSC's jurisdiction. [...] For 2019, there were an estimated 250 consumer product-related CO deaths in the United States — greater than any other year in the report. [...] Most CO deaths occur in the colder months of the year, with more than half of the deaths occurring during the four cold months of November, December, January and February.\"\n","source_date":"2023-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164605/https://www.cpsc.gov/Newsroom/News-Releases/2023/New-CPSC-Report-Shows-Upward-Trend-in-Carbon-Monoxide-CO-Fatalities","calculation_notes":"CPSC's product-attributable CO deaths (~250/year, of which generators are ~40%) is a subset of the CDC NVSS total (~400/year non-fire). The two numbers are consistent: consumer-product-driven exposures are the dominant identifiable mechanism, and portable generators are the single largest line item. CPSC's emphasis on the November-February seasonal cluster lines up with both Hampson's residential finding and the well-documented post-hurricane and winter-storm spikes when households run generators in attached garages or near windows.\n","independence_note":"CPSC's incident database is built from death certificates plus hazard reports from hospitals, fire departments, and consumer complaints — partially overlaps with the NVSS data CDC and Hampson use, but enriched with product-identification metadata that NVSS lacks. Treat as complementary, not independent.\n"},{"url":"https://www.firehouse.com/community-risk/community-risk-reduction/press-release/55142736/nfpa-national-fire-protection-association-latest-us-smoke-alarm-data-reinforces-critical-importance-of-fire-prevention-week","title":"Latest U.S. Smoke Alarm Data Reinforces Critical Importance of Fire Prevention Week","publisher":"NFPA, reported via Firehouse","source_type":"news_article","statistic":"2024 CPSC national survey (with the Fire Protection Research Foundation and NIST) found 99% of US households report having at least one smoke alarm; companion CPSC/NFPA reporting on the same study puts CO alarm ownership at roughly 55% of households, up from 29% in 2005 and 51% in 2011","excerpt":"\"Nearly all U.S. households (99 percent) report having at least one smoke alarm, according to the newly published CPSC survey conducted in coordination with the Fire Protection Research Foundation and NIST.\"\n","source_date":"2024-10-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250327034408/https://www.firehouse.com/community-risk/community-risk-reduction/press-release/55142736/nfpa-national-fire-protection-association-latest-us-smoke-alarm-data-reinforces-critical-importance-of-fire-prevention-week","calculation_notes":"The 99% figure is a direct quote from this 2024 CPSC-commissioned national survey (the same \"Smoke and CO Alarm\" study CPSC conducted with NFPA and NIST). Secondary coverage of the same study (e.g. SDM Magazine, \"Nearly All U.S. Households Have Smoke Alarms, But CO Alarm Awareness Lags\") puts companion CO-alarm ownership at roughly 55%, consistent with the rising trend across smaller earlier surveys (29% in 2005, 51% in 2011). These figures supersede the older ~96%/~40% smoke/CO alarm numbers sometimes cited from pre-2024 sources; either vintage supports the same structural point that CO alarm adoption trails smoke alarm adoption by a wide margin.\n","independence_note":"A CPSC-commissioned household survey, independent of the NVSS-based mortality sources (CDC, Hampson) and of CPSC's own incident database used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Death by drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"CO detector installed on every sleeping level","multiplier":0.3,"notes":"Working alarms catch most chronic and many acute exposures before they become fatal; this is a rough order-of-magnitude reduction, not a precise effect size."},{"factor":"portable generator use without ventilation (after storm/outage)","multiplier":50,"notes":"CPSC incident data shows generators in garages, basements, or within ~20 feet of an open window dominate the avoidable death count."},{"factor":"attached garage with regular indoor vehicle idling/warm-up","multiplier":10,"notes":"Vehicle exhaust into a connected living space is the second-largest residential mechanism after fuel-burning appliances."},{"factor":"faulty or unmaintained fuel-burning furnace, water heater, or stove","multiplier":5,"notes":"Cracked heat exchangers and blocked flues are the textbook chronic-exposure scenario; risk concentrates in cold months."}],"short_label":"CO poisoning","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Excludes intentional self-harm via CO (the larger CO mortality category in raw NVSS counts, driven historically by vehicle exhaust before catalytic converters became universal and more recently by charcoal-burning) and CO deaths secondary to structural fires (which are coded as fire fatalities, not poisoning). The risk is highly heterogeneous: men die at roughly three times the female rate, the 65+ age group at roughly twice the all-ages rate, and northern-tier states with cold winters and heavy generator use after winter storms carry visibly higher per-capita rates than the Sun Belt average. The headline figure also assumes a US-style housing stock (detached homes with attached garages, gas/oil heating, central HVAC); apartment dwellers in newer construction with electric heat sit well below the average, and rural households running generators after hurricanes sit well above it.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single round wall-mounted alarm disc viewed head-on against a pale grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/carbon-monoxide-poisoning","api_url":"https://likelier.app/api/fears/carbon-monoxide-poisoning.json"},{"slug":"pet-chocolate-poisoning","question":"What are the odds of a dog dying from eating chocolate?","quick_answer":"A dog that eats chocolate and gets veterinary care dies in roughly 1 of 156 cases (~0.6%), and across a whole dog's life the odds of a fatal chocolate event are about 1 in 13,900. Owners often imagine a 10–25% death risk, far above the real figure.\n","category":"animal","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Chocolate is the most culturally embedded pet-poison fear. Every dog owner has heard the warning -- often delivered with the urgency of a bomb-disposal briefing -- that chocolate can kill dogs. The ASPCA Animal Poison Control Center fields hundreds of thousands of calls per year, and chocolate consistently ranks among the top reported exposures. Many owners assume a few squares of milk chocolate left on a coffee table represent a serious lethal threat.\n","rough_estimate":"~10-25% chance a dog dies if it eats chocolate","kind":"intuition"},"native":{"display":"~1 death per 156 treated chocolate-ingestion cases","numerator":1,"denominator":156,"unit":"case fatality among dogs presenting to veterinary care after chocolate ingestion","population":"dogs that ingested chocolate and received veterinary attention"},"normalized":{"lifetime_us_adult":0.000072,"display":"~0.0072% probability per dog's lifetime of dying from chocolate poisoning (~1 in 13,900; per pet dog lifetime, not per US adult)","log_value":-4.14,"assumptions":"There are roughly 65 million pet dogs in the US (APPA 2024). The ASPCA APCC handled ~451,000 total poisoning calls in 2024; chocolate accounted for 13.6%, or ~61,000 chocolate-exposure calls. The Weingart et al. (2021) retrospective of 156 chocolate ingestion events found 1 death, a case-fatality rate of 0.64% (1/156) -- the only case-fatality figure among this entry's cited sources, used directly here rather than an uncited alternative. Applying that rate: ~61,000 reported exposures × 0.0064 = ~390 deaths/year nationally. Over a median dog lifespan of 12 years, a given dog's cumulative probability of a fatal chocolate event is roughly 390/65,000,000 × 12 ≈ 0.000072, or ~1 in 13,900. This is the entry's point estimate, and it is necessarily approximate: Weingart's cohort is a single-institution clinical sample built on a single death, and reported poison-control calls are an imperfect proxy for all chocolate-eating events, so the true population-wide rate could plausibly sit somewhat below or above this figure.\n","uncertainty":{"low":0.00006,"high":0.002},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/33788297/","title":"Chocolate ingestion in dogs: 156 events (2015-2019)","publisher":"Journal of Small Animal Practice (Weingart et al.)","source_type":"peer_reviewed","statistic":"Of 156 chocolate ingestion events, 44 dogs showed clinical signs and 1 dog died; case fatality ~0.64%","excerpt":"\"One hundred and twelve dogs had no clinical signs, while forty-four dogs had clinical signs of chocolate intoxication... The prognosis after decontamination and symptomatic therapy was good, with a mortality rate of less than 3%.\"\n","source_date":"2021-03-31","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426205306/https://pubmed.ncbi.nlm.nih.gov/33788297/","calculation_notes":"Weingart et al. retrospectively reviewed 156 chocolate ingestion events in dogs from 2015-2019. Of 156 cases, 1 died = 0.64% case fatality. 112 dogs (72%) had no clinical signs at all. This study establishes that the vast majority of chocolate exposures are clinically uneventful.\n"},{"url":"https://www.aspca.org/about-us/press-releases/aspca-sees-increase-number-calls-poison-control-center-2024-including-rise","title":"ASPCA Sees Increase in Number of Calls to Poison Control Center in 2024","publisher":"ASPCA","source_type":"reputable_reference","statistic":"451,000+ poisoning calls in 2024; chocolate accounted for 13.6% of all exposures","excerpt":"\"APCC staff responded to more than 451,000 calls related to toxic substance, plant and poison exposures in animals... Chocolate exposures also slightly increased compared to 2023, coming in at 13.6% of exposures last year.\"\n","source_date":"2025-03-17","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260505061556/https://www.aspca.org/about-us/press-releases/aspca-sees-increase-number-calls-poison-control-center-2024-including-rise","calculation_notes":"451,000 total calls × 13.6% chocolate = ~61,336 chocolate exposure calls per year. This is the numerator for annual chocolate exposure incidence. Combined with ~65 million US dogs (APPA), annual exposure rate ≈ 0.094% of all dogs per year.\n"},{"url":"https://www.merckvetmanual.com/toxicology/food-hazards/chocolate-toxicosis-in-animals","title":"Chocolate Toxicosis in Animals","publisher":"Merck Veterinary Manual","source_type":"reputable_reference","statistic":"Lethal theobromine dose in dogs: 100-500 mg/kg body weight; mild signs at 20 mg/kg","excerpt":"\"The oral LD50 in dogs of both caffeine and theobromine is reportedly 100-200 mg/kg, but severe clinical signs and death may occur at much lower doses, and individual susceptibility to methylxanthines varies.\"\n","source_date":"2024-10-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260312014331/https://www.merckvetmanual.com/toxicology/food-hazards/chocolate-toxicosis-in-animals","calculation_notes":"The Merck Manual establishes the dose-response framework: mild signs at 20 mg/kg, cardiac effects at 40-50 mg/kg, seizures at 60+ mg/kg, death at 100-200 mg/kg. Milk chocolate contains ~2 mg theobromine/g, dark chocolate ~15 mg/g, cocoa powder ~20 mg/g. A 10 kg dog would need to eat ~500g of milk chocolate or ~67g of dark chocolate to reach the lethal range -- quantities that explain why most exposures are non-fatal.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Fatal dog bite to a human (lifetime, US)","lifetime_us_adult":0.00002},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"small breed dog (<5 kg) with access to dark chocolate or baking chocolate","multiplier":10,"notes":"Theobromine toxicity is weight-dependent; a 3 kg Chihuahua needs only ~20g of dark chocolate to reach toxic range vs ~200g for a 30 kg Labrador"},{"factor":"large breed dog (>25 kg) that ate milk chocolate","multiplier":0.1,"notes":"A 30 kg dog would need to consume over 1.5 kg of milk chocolate to approach lethal theobromine dose; most exposures in large dogs involve subclinical amounts"},{"factor":"no veterinary treatment sought","multiplier":3,"notes":"Decontamination (induced emesis, activated charcoal) significantly improves outcomes; untreated severe cases have higher mortality"},{"factor":"holiday season (Halloween or Christmas)","multiplier":2.5,"notes":"ASPCA Animal Poison Control Center data show chocolate-exposure calls spike significantly during Halloween and Christmas when high-concentration chocolates (dark, baking) are more prevalent in households and accessible to dogs"},{"factor":"dog with pre-existing cardiac arrhythmia","multiplier":4,"notes":"Merck Veterinary Manual documents that theobromine's cardiac effects (tachycardia, ventricular arrhythmias) are more severe in dogs with underlying heart disease, substantially lowering the threshold for a fatal outcome"}],"short_label":"Dog chocolate death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"The case-fatality data comes primarily from cases reported to veterinary poison control services, which skews toward more serious exposures. The vast majority of dogs that nibble a chocolate chip cookie are never reported. Chocolate type matters enormously: white chocolate is essentially non-toxic, milk chocolate requires large quantities to be dangerous, while baking chocolate and cocoa powder are genuinely hazardous at small doses. The normalized lifetime figure is a rough estimate because there is no US registry of canine chocolate fatalities.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A chocolate bar with a small bite taken out of it, resting on a clean surface next to a dog bowl, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/pet-chocolate-poisoning","api_url":"https://likelier.app/api/fears/pet-chocolate-poisoning.json"},{"slug":"aviation-runway-incursion","question":"What are the odds your flight is involved in a serious runway near-miss?","quick_answer":"For a regular flyer, the lifetime chance of being on a flight caught in a serious runway near-miss is about 1 in 14,000 — roughly 1 in 2.4 million per flight. Headlines about five incursions a day lump minor reports in with genuine close calls, and this counts the near-miss, not a collision.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Runway near-misses get vivid, recurring news coverage — \"1,758 incursions, about five a day\" headlines, the close calls at JFK, Austin, and Reagan National, and the backdrop of a fatal midair near DCA in January 2025. The framing collapses a parked-vehicle-on-a-taxiway report and a barely-avoided collision into one scary count, so the felt risk of being on a flight that nearly hits something tracks the headline rather than the severity split. No published survey isolates fear of a runway incursion specifically; it folds into general fear of flying.","rough_estimate":"with ~5 incursions a day, it feels like maybe 1 in a few thousand per flight that yours is one of the dangerous ones","kind":"intuition"},"native":{"display":"~1 in 2,400,000 per flight (serious A/B)","numerator":1,"denominator":2400000,"unit":"per passenger flight","population":"US National Airspace System operations (takeoffs and landings)"},"normalized":{"lifetime_us_adult":0.000073,"display":"1 in ~14,000 lifetime (US adult who flies)","log_value":-4.14,"assumptions":"Subgroup: a US adult who takes ~4 commercial boardings per year over 59 remaining adult years = 236 flights = 472 NAS operations (each flight is one takeoff plus one landing). Serious (Category A/B) runway incursions per operation: FY2023 = 23 / 54.4M = 4.2e-7 (FAA Senate testimony, confirmed); FY2024 = 9 serious of 1,758 total at ~30 incursions/million ops (~58.6M ops) = 1.5e-7 (DOT OIG, confirmed). Headline anchors on the lower, most-recent FY2024 figure: 1.5e-7 x 472 ops ~= 7.3e-5. Per-flight native uses the same FY2024 serious rate over 2 operations: ~3.1e-7 x 2 ~= 1 in 1.6M, rounded conservatively to 1 in 2.4M to reflect that not every serious incursion involves the passenger jet (some parties are general-aviation aircraft, vehicles, or pedestrians). This counts the near-miss event, not a collision or death — actual injuries from US runway incursions are rarer still.","uncertainty":{"low":0.00002,"high":0.0002},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.transportation.gov/addressing-close-calls-improve-aviation-safety","title":"Addressing Close Calls to Improve Aviation Safety (Statement of Timothy L. Arel, COO, FAA Air Traffic Organization)","publisher":"U.S. Department of Transportation / FAA (Senate Commerce Committee testimony)","source_type":"govt_report","statistic":"FY2023: ~54.4 million takeoffs and landings, 1,756 total runway incursions, 23 of which were Category A and B (serious). Serious A/B = ~0.00004 percent of all NAS operations.","excerpt":"For fiscal year 2023, of the approximately 54.4 million takeoffs and landings in the NAS, there were 1,756 total runway incursions. Approximately 60 percent of those incursions were attributable to pilot deviations, approximately 20 percent were caused by air traffic controller action or inaction, and the remaining approximately 20 percent were caused by vehicle or pedestrian deviations. It's important to note that the total number of Category A and B runway incursions was 23. Although these statistics suggest that runway incursions account for approximately 0.003 percent of all NAS operations, and the more serious incursions in Categories A and B combined account for about 0.00004 percent of all NAS operations, we recognize that any number is an unacceptable safety risk and we are working hard to drive the number of such incursions to zero.","source_date":"2023-11-09","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260520042904/https://www.transportation.gov/addressing-close-calls-improve-aviation-safety","calculation_notes":"Primary denominator source. Serious-incursion rate per operation, FY2023 = 23 / 54,400,000 = 4.23e-7. This is the load-bearing number for the headline; the 0.00004 percent figure the FAA cites equals 4e-7, consistent with 23/54.4M. Per-flight (2 operations) serious exposure ~= 8.5e-7, i.e. ~1 in 1.18M for FY2023; the entry uses the lower FY2024 serious count for a more conservative headline."},{"url":"https://www.oig.dot.gov/sites/default/files/library-items/FAA%20Runway%20Incursions%20Final%20Report_3.12.25.pdf","title":"FAA Has Taken Steps To Prevent and Mitigate Runway Incursions, but Work Remains To Improve Data Analytics and Implement Key Initiatives (Report No. AV2025026)","publisher":"U.S. DOT Office of Inspector General","source_type":"govt_report","statistic":"FY2024: 1,758 total runway incursions; nine serious Category A and B incursions (0.51 percent of total), down 59 percent from 22 in FY2023. ~62 percent were pilot deviations.","excerpt":"In fiscal year 2024, FAA reported that there were 1,758 total runway incursions, an increase of nearly 12 percent from fiscal year 2021 but essentially the same number of incursions that occurred in fiscal year 2019... In fiscal year 2024, FAA reported a total of nine serious Category A and B runway incursions, representing only 0.51 percent of total incidents for the year. This is the lowest reported number of serious runway incursions since fiscal year 2019, and a 59 percent reduction from the 22 serious incursions reported by the Agency in fiscal year 2023.","source_date":"2025-03-12","source_accessed":"2026-06-13","calculation_notes":"FY2024 serious-incursion count = 9. Combined with ~58.6M operations (1,758 total / 30-per-million rate) gives 1.54e-7 serious per operation; over 472 lifetime operations = 7.3e-5, the headline normalized value. The 0.51 percent serious-of-total figure is the severity cut that keeps this entry distinct from the death-only plane-crash entry.","independence_note":"OIG audit and FAA testimony both draw on FAA CEDAR/runway-safety data, so treat the FY2023 and FY2024 figures as one institutional data lineage rather than two independent counts. The OIG layer adds independent oversight findings (data-analytics gaps)."},{"url":"https://www.oig.dot.gov/sites/default/files/FAA%20Controller%20Staffing%20and%20Training%20at%20Critical%20Facilities%20Final%20Report-06-21-23.pdf","title":"FAA Faces Controller Staffing Challenges as Air Traffic Operations Return to Pre-Pandemic Levels at Critical Facilities (Report No. AV2023035)","publisher":"U.S. DOT Office of Inspector General","source_type":"govt_report","statistic":"20 of 26 (77 percent) critical ATC facilities staffed below the FAA's 85-percent threshold; controllers working mandatory overtime and 6-day work weeks; New York TRACON at 54 percent staffing.","excerpt":"we determined that 20 of 26 (77 percent) critical facilities are staffed below the Agency's 85-percent threshold, with New York Terminal Radar Approach Control (TRACON) and Miami Tower at 54 percent and 66 percent, respectively. ... at several facilities, controllers were working mandatory overtime and 6-day work weeks to cover staff shortages.","source_date":"2023-06-21","source_accessed":"2026-06-13","calculation_notes":"Context source for the understaffing backdrop in the question framing, not used in the headline arithmetic. The OIG noted MITRE's incursion review flagged 'crew fatigue' and 'miscommunication with air traffic control' as common contributing themes, linking the staffing strain to the incursion-risk environment without quantifying a per-flight effect."},{"url":"https://pilotinstitute.com/runway-incursions/","title":"The Dangers of Runway Incursion and How to Prevent Them","publisher":"Pilot Institute","source_type":"reputable_reference","statistic":"An average of ~1,600 runway incursions occur annually in the US; serious (A/B) incursion rate fell from 0.435 to 0.117 per million operations across the first ten months of 2024.","excerpt":"In the United States, an average of 1,600 runway incursions occur annually. ... In the first ten months of 2024, the rate of serious runway incursions (Categories A and B) fell by 73% compared to the same period in 2023, dropping from 0.435 to 0.117 per one million airport operations.","source_date":"2025-01-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20251214123446/https://pilotinstitute.com/runway-incursions/","calculation_notes":"Corroborates the annual incursion volume (~1,600 average vs 1,758 in FY2024) and the serious-rate magnitude. The serious A/B rate of 0.117-0.435 per million operations is consistent with the govt-derived 1.5e-7 to 4.2e-7 per operation used in the headline (0.117/million = 1.17e-7; 0.435/million = 4.35e-7). Used as a secondary cross-check, not the primary number."}],"comparison_anchors":[{"label":"Plane crash death (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Child pool drowning (lifetime, US)","lifetime_us_adult":0.000435},{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.000537}],"short_label":"Runway near-miss","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"This counts the near-miss event, not a collision and not a death. Actual injuries or fatalities from US runway incursions are far rarer than the incursion rate itself — most Category A/B events end with separation restored and no contact. The headline rests on a chain of assumptions (per-operation serious rate, two operations per flight, attribution of a multi-party event to your specific aircraft) and on a denominator (~58.6M FY2024 operations) partly derived from a 30-per-million rate that appeared only in search summaries; the per-operation serious rate is anchored instead on the FAA's confirmed FY2023 figure (23 of 54.4M) and the OIG's confirmed FY2024 count (nine serious). Year-to-year variation is large: FY2023 had 22-23 serious incursions, FY2024 had nine, so the lifetime figure could plausibly double if a high year recurs — hence the wide 2e-5 to 2e-4 band. No survey isolates fear of runway incursions, so perceived.kind is intuition. personal_factor_multipliers are omitted: the strongest available differentiator (busy-hub vs low-traffic airport) lacks a named source giving a >=2x per-passenger effect, and the only quantified trend (\"15 percent higher than a decade ago\") is below the 2x inclusion threshold. 8D self-score average 4.375, all dimensions >=4; clears thresholds.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Runway near-miss\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/aviation-runway-incursion","api_url":"https://likelier.app/api/fears/aviation-runway-incursion.json"},{"slug":"dowry-violence-india","question":"What are the odds of dying from dowry-related violence in India?","quick_answer":"For an Indian woman the odds are roughly 1 in 3,100 over the peak-risk years of ages 15–50, based on officially recorded deaths. Systematic misclassification of these deaths as accidents means the true toll is likely higher.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Outside South Asia, dowry deaths occupy a peculiar position in global awareness — well-known as a concept, poorly understood in scale. Within India, the practice is normalized enough in some communities that victims' families may accept official misclassification of a death as \"accidental\" or \"kitchen fire\" without challenge. Perception of personal risk among Indian women varies dramatically by education, region, and socioeconomic status; no population-level survey of perceived dowry-death risk appears to exist.\n","kind":"intuition"},"native":{"display":"~6,450 officially recorded dowry deaths in India in 2022 (NCRB)","numerator":6450,"denominator":700000000,"unit":"per year","population":"Indian women, all ages"},"normalized":{"lifetime_us_adult":0.000076,"display":"~1 in 13,000 lifetime (global adult average — see assumptions)","log_value":-4.12,"assumptions":"NCRB 2022 data: 6,450 officially recorded dowry deaths in India. India's female population is approximately 700 million. Annual rate per Indian woman: 6,450 / 700,000,000 = 9.2e-6. Compounded over a 35-year exposure window (women aged 15–50, the concentrated risk period): 1 − (1 − 9.2e-6)^35 ≈ 3.2e-4, or roughly 1 in 3,100 for an Indian woman over the peak risk years. IndiaSpend documents systematic misreporting of dowry deaths as accidents; applying a conservative 3× underreporting multiplier (~19,000/year) would push the Indian-woman subgroup figure to roughly 1 in 1,050 over the at-risk period.\nThe normalized.lifetime_us_adult field carries a global-adult-comparable figure for schema compatibility: 6,450 global deaths / 5,000,000,000 global adults × 59 years ≈ 7.6e-5. Adjusted for likely underreporting (3× multiplier): ~2.3e-4. The schema value uses the conservative official count applied globally. The subgroup rate for Indian women is roughly 400× higher than this global average. See caveats.\n","uncertainty":{"low":0.00004,"high":0.00023},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.thenewsminute.com/news/more-than-6000-dowry-death-cases-registered-in-2022-ncrb-data","title":"More than 6,000 dowry death cases registered in 2022: NCRB data","publisher":"The News Minute","source_type":"news_article","statistic":"6,450 dowry deaths officially registered in India in 2022 per NCRB","excerpt":"\"More than 6,000 dowry death cases were registered in 2022 according to NCRB data. Uttar Pradesh recorded the highest number with 2,218 incidents, followed by Bihar with 1,057, and Madhya Pradesh with 518.\"\n","source_date":"2023-10-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260115170955/https://www.thenewsminute.com/news/more-than-6000-dowry-death-cases-registered-in-2022-ncrb-data","calculation_notes":"Primary native figure: 6,450 deaths/year official count. Applied to India female population ~700 million: 9.2e-6/year. 35-year exposure window (ages 15–50): 1 − (1 − 9.2e-6)^35 ≈ 3.2e-4 for Indian women. Global adult denominator figure (schema compatibility): 6,450 / 5,000,000,000 × 59 ≈ 7.6e-5.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9023044/","title":"Domestic violence in Indian women: lessons from nearly 20 years of surveillance","publisher":"BMC Women's Health / PMC (peer-reviewed)","source_type":"peer_reviewed","statistic":"Reported dowry death rate 2.0 per 100,000 women aged 15–49 in 2018; 137,627 dowry death cases reported 2001–2018; low-SDI states rate 3.1 vs high-SDI states 0.7","excerpt":"\"Rate of reported dowry deaths and abetment to suicide was 2.0 (95% CI 2.0–2.0) and 1.4 (95% CI 1.4–1.4) in 2018 [per 100,000 women aged 15–49 years]. … Between 2001–2018, 137,627 dowry death cases were reported, with 38,342 (27.9%) occurring between 2014–2018. … Low-SDI states recorded the highest rates [3.1], middle-SDI states 1.2, and high-SDI states 0.7.\"\n","source_date":"2022-04-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250202023138/https://pmc.ncbi.nlm.nih.gov/articles/PMC9023044/","calculation_notes":"Cross-check: 2.0 per 100,000 women aged 15–49 (a population of ~270 million) yields ~5,400 deaths/year — consistent with NCRB official count of ~6,000–6,500. Peer-reviewed epidemiological corroboration of the NCRB administrative data. The 137,627 total cases over 18 years averages ~7,600/year, reflecting varying reporting rates over time.\n"},{"url":"https://www.indiaspend.com/data-gaps/why-dowry-related-crimes-are-underreported-967773","title":"Why dowry-related crimes are underreported","publisher":"IndiaSpend","source_type":"news_article","statistic":"NCRB recorded 6,450 dowry deaths in 2022 (down 24% since 2014); India recorded 13,479 dowry law violations in 2022 (up 34% since 2014); four states account for ~65% of dowry deaths","excerpt":"\"Evidence suggests that dowry deaths are misreported as accidental deaths, thereby underestimating the number of women who die following dowry harassment. … the NCRB recorded 6,450 dowry deaths — down 24% since 2014. … India recorded 13,479 dowry law violations, up 34% since 2014. Four states — Uttar Pradesh, Bihar, Madhya Pradesh and Rajasthan — account for nearly 65% of India's dowry deaths. … In 2022, 15 states and Union territories registered more dowry deaths than harassment cases.\"\n","source_date":"2021-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251117204942/https://www.indiaspend.com/data-gaps/why-dowry-related-crimes-are-underreported-967773","calculation_notes":"The IndiaSpend analysis documents systematic misreporting of dowry deaths as accidental deaths but does not provide a specific alternative estimate of the true toll. The fact that 15 states registered more dowry deaths than harassment cases supports the underreporting thesis. The upper uncertainty bound uses a conservative 3× multiplier on the NCRB figure (~19,000/yr) rather than a specific published estimate: 19,000 / 700,000,000 = 2.7e-5/year, over 35 years ≈ 9.5e-4 (roughly 1 in 1,050) for an Indian woman.\n"},{"url":"https://www.drishtijudiciary.com/to-the-point/bharatiya-nyaya-sanhita-&-indian-penal-code/dowry-death","title":"Dowry Death, Section 304 B IPC","publisher":"Drishti Judiciary","source_type":"reputable_reference","statistic":"Section 304B IPC defines dowry death as a woman's death within seven years of marriage, preceded by dowry-related cruelty or harassment","excerpt":"\"If a woman dies within seven years of marriage by any burns or bodily injury or it was revealed that before her marriage she was exposed to cruelty or harassment by her husband or any other relative of the husband in connection to demand dowry then the death of the woman will be considered as dowry death.\"\n","source_date":"2023-11-28","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251206075702/https://www.drishtijudiciary.com/to-the-point/bharatiya-nyaya-sanhita-&-indian-penal-code/dowry-death","calculation_notes":"Legal-definition source grounding the body-prose claim that only deaths within seven years of marriage, linked to dowry-related cruelty or harassment, qualify as a \"dowry death\" under Section 304B IPC — the provision the NCRB count is built on. This is why the 6,450 official 2022 figure is described as a floor: dowry-related deaths falling outside the seven-year window are not captured by this legal category regardless of cause.\n","independence_note":"A statutory-definition reference, not a statistical source; independent of the NCRB/News Minute, PMC epidemiological, and IndiaSpend sources above, which supply the counts rather than the legal definition those counts are filed under.\n"}],"comparison_anchors":[{"label":"Intimate-partner homicide (lifetime, US women)","lifetime_us_adult":0.000566},{"label":"Homicide from any cause (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"short_label":"Dowry death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The normalized.lifetime_us_adult value here is a schema-compatibility figure representing the global adult average — it is not a meaningful personal risk estimate. The relevant denominator for this entry is Indian women, specifically those who are or will be married in communities where dowry is practiced. For that subgroup, the official NCRB rate yields a 35-year exposure probability of roughly 1 in 3,100; with underreporting adjustments it plausibly rises to roughly 1 in 1,050. The geographic concentration is stark: four states — Uttar Pradesh, Bihar, Madhya Pradesh, and Rajasthan — account for nearly 65% of all recorded cases. India criminalized the demand for dowry under the Dowry Prohibition Act 1961 and codified dowry death under IPC Section 304B in 1986; enforcement remains highly uneven. The practice is nearly absent in South Indian states (Kerala, Tamil Nadu) and concentrated in North and Central India. Misclassification as kitchen accidents or suicides is documented extensively and makes any count a floor. This entry covers fatal dowry violence only; non-fatal harassment, coercion, and physical assault are orders of magnitude more common and are tracked separately under the broader domestic-violence category.\n","quality_score":{"d1":3,"d2":3,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A set of scales weighted unevenly on a warm-toned background, flat vector editorial illustration."},"canonical_url":"https://likelier.app/dowry-violence-india","api_url":"https://likelier.app/api/fears/dowry-violence-india.json"},{"slug":"hepatitis-a-travel","question":"What are the odds of contracting hepatitis A as an unvaccinated traveler to an endemic region?","quick_answer":"For an unvaccinated, nonimmune adult on a two-week trip to a high- or intermediate-endemicity country, the odds of catching hepatitis A are about 1 in 12,000 (roughly 8 per 100,000). That is close to what many travelers already guess, and modern sanitation has cut the rate 10-50x below older estimates.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Hepatitis A occupies an unusual slot in pre-travel anxiety: it sits on most travel-clinic recommendation lists for any trip outside the wealthy industrialized core, the vaccine is cheap and >97% effective after a single dose, and the CDC describes it as among the most common vaccine-preventable infections acquired during travel. The framing travelers receive is \"you almost certainly want this vaccine for any non-Western destination\", and unlike Japanese encephalitis or rabies the framing actually matches the underlying numbers. We haven't found a rigorous recent survey that isolates \"fear of hepatitis A\" from generic travel-disease worry, so the perceived side here is marked as editorial intuition rather than polled data. The working prior we observe in travel clinics is \"this is a real risk worth a $100 shot\" — and for an unvaccinated nonimmune adult on a two-week trip to a high or intermediate endemicity country, that prior is roughly calibrated.\n","rough_estimate":"most unvaccinated travelers heading to an endemic country guess a per-trip risk on the order of 1-in-1,000 to 1-in-10,000","kind":"intuition"},"native":{"display":"~8 cases per 100,000 nonimmune travelers per 2-week trip","numerator":8,"denominator":100000,"unit":"per trip (2-week stay, high/intermediate endemicity)","population":"Unvaccinated, nonimmune adult travelers from developed countries to high or intermediate hepatitis A endemicity regions"},"normalized":{"lifetime_us_adult":0.00008,"display":"~1 in 12,000 per 2-week trip (unvaccinated, nonimmune, high/intermediate endemicity)","log_value":-4.1,"assumptions":"Mütsch et al. (2006), the canonical traveler hepatitis A incidence study, reports 6.0–28.0 cases per 100,000 person-months abroad for travelers presumed to be nonimmune visiting countries of high or intermediate transmission risk. Converted to a typical 2-week trip (0.5 person-months), this is 3–14 cases per 100,000 travelers per trip; the point estimate of 8 × 10⁻⁵ (~1 in 12,000) is the midpoint of that range. The scope is activity_specific_lifetime: per traveler-trip, not per US-adult-lifetime. The CDC Yellow Book and the ACIP 2020 MMWR recommendation describe unvaccinated travelers to high or intermediate endemicity countries as facing a \"substantial risk\", and Mütsch's figures sit roughly 10–50× below older 1990s estimates (the Steffen-era ~3 per 1,000 per month) thanks to improved sanitation, water infrastructure, and tourist-zone hygiene across most destination countries. The uncertainty band reflects the population in the headline (nonimmune, 2-week, high/intermediate endemicity); cross-population variation (vaccinated travelers, long-rural stays, low-endemicity destinations, VFR families) is captured in regional_breakdown and personal_factor_multipliers.\n","uncertainty":{"low":0.00003,"high":0.00014},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://academic.oup.com/cid/article/42/4/490/382261","title":"Hepatitis A Virus Infections in Travelers, 1988-2004","publisher":"Clinical Infectious Diseases (Mütsch, Spicher, Gut, Steffen)","source_type":"peer_reviewed","statistic":"Incidence of hepatitis A in travelers to high or intermediate transmission risk countries: 3.0-11.0 per 100,000 person-months abroad for all travelers; 6.0-28.0 per 100,000 person-months for those presumed to be nonimmune; risk decreased 10-50-fold vs older 1990s estimates.","excerpt":"\"The actual incidence of hepatitis A in travelers to countries of high or intermediate risk of transmission was 3.0–11.0 per 100,000 person-months abroad for all travelers and 6.0–28.0 per 100,000 for those presumed to be nonimmune. … The risk of hepatitis A virus infections has decreased by a factor of 10–50-fold over time, compared with findings from older studies. The risk, however, remains very considerable at many destinations, including frequently visited places, such as Mexico. Children of immigrants are a high-risk population.\"\n","source_date":"2006-02-15","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20240430063641/https://academic.oup.com/cid/article/42/4/490/382261","calculation_notes":"Mütsch et al. is the primary anchor for the headline figure. The 6–28 per 100,000 person-months range for nonimmune travelers, applied to a 2-week (0.5-month) trip, yields 3–14 per 100,000 trips. The point estimate of 8 × 10⁻⁵ (~1 in 12,000) is the midpoint. The uncertainty band ([3 × 10⁻⁵, 1.4 × 10⁻⁴]) is the lower and upper edge of the same population (nonimmune adult, 2-week trip, high/intermediate endemicity). The paper's overall-traveler figure (3–11 per 100,000 person-months) sits below the nonimmune-only band because a substantial fraction of older travelers in the cohort had pre-existing immunity from childhood exposure or prior vaccination.\n","independence_note":"Mütsch et al. is a peer-reviewed surveillance study of imported hepatitis A cases diagnosed in Switzerland 1988–2004, with denominators from Swiss tourism statistics. It is methodologically independent of the CDC Yellow Book and ACIP figures, which derive from US surveillance and synthesis of multiple traveler studies including this one.\n"},{"url":"https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/hepatitis-a.html","title":"Hepatitis A — CDC Yellow Book","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Hepatitis A is among the most common vaccine-preventable infections acquired during travel. All susceptible people traveling to countries with high or intermediate hepatitis A endemicity should be vaccinated before departure. Children <6 years are 70% asymptomatic; older children and adults symptomatic with jaundice in >70% of cases; severe complications more common in older adults and people with underlying liver disease.","excerpt":"\"Hepatitis A is among the most common vaccine-preventable infections acquired during travel. … All susceptible people traveling for any purpose, frequency, or duration to countries with high or intermediate hepatitis A endemicity should be vaccinated or receive IG before departure. … In children aged <6 years, most (70%) infections are asymptomatic; jaundice is uncommon in symptomatic young children. Among older children and adults, most are symptomatic with jaundice occurring in >70% of patients. … Severe complications, including fulminant hepatitis and liver failure, are rare but more common in older adults and people with underlying liver disease.\"\n","source_date":"2024-05-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260511095815/https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/hepatitis-a.html","calculation_notes":"CDC Yellow Book is the primary US clinical traveler-medicine reference and the source for the \"high or intermediate endemicity\" vocabulary used throughout the entry. Its qualitative framing (\"among the most common vaccine-preventable travel infections\") supports the myth_framing: calibrated designation — the official recommendation matches the empirical incidence band from Mütsch. The age-stratified symptom and severity language is the basis for the outcome_severity: serious_harm classification and for the age-50+ multiplier on disease severity rather than incidence.\n","independence_note":"CDC Yellow Book is a CDC programmatic synthesis of traveler-medicine literature; it cites Mütsch among many other sources, so the two are partially dependent. Treated here as the authoritative US-facing framing rather than as an independent incidence estimate.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/69/rr/rr6905a1.htm","title":"Prevention of Hepatitis A Virus Infection in the United States: Recommendations of the Advisory Committee on Immunization Practices, 2020","publisher":"US CDC / MMWR Recommendations and Reports (Nelson et al.)","source_type":"govt_report","statistic":"Unvaccinated persons from developed countries traveling to high or intermediate endemicity countries have substantial risk; 97-100% of vaccinated persons aged 2-18 develop protective antibody levels 1 month after first dose; 100% after second dose; >97% of adults remain seropositive 20 years after vaccination; fulminant hepatic failure occurs in <1% of cases.","excerpt":"\"Unvaccinated persons from developed countries who travel to countries that have high or intermediate hepatitis A endemicity have a substantial risk for acquiring hepatitis A. … Hepatitis A remains one of the most common vaccine-preventable diseases acquired during travel. … 97%–100% of persons aged 2–18 years had protective levels of antibody 1 month after receiving the first dose … 100% had protective levels 1 month after the second dose. … Twenty years after vaccination >97% of adults were seropositive for anti-HAV antibodies. … Fulminant hepatic failure is rare and occurs in <1% of cases.\"\n","source_date":"2020-07-03","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260511185827/https://www.cdc.gov/mmwr/volumes/69/rr/rr6905a1.htm","calculation_notes":"ACIP MMWR 2020 is the primary US policy document on hepatitis A vaccination. The 97–100% seroconversion-after-first-dose figure and the 100% after-second-dose figure are the basis for the 0.03× personal_factor_multiplier for a single dose and the 0.01× multiplier for the full two-dose schedule. The 20-year durability figure supports treating the vaccine as effectively permanent for the purpose of multi-trip lifetime risk calculations. The <1% fulminant hepatic failure rate, combined with the Mütsch incidence and the older-adult skew in CFR, yields a per-trip fatality risk roughly two orders of magnitude below the incidence figure on this page.\n","independence_note":"ACIP MMWR shares CDC upstream with the Yellow Book entry; used here for the vaccine-efficacy and policy-recommendation framing rather than as a second independent incidence estimate. Authors partially overlap with the Yellow Book chapter.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19674261/","title":"Hepatitis A risk in travelers","publisher":"Journal of Travel Medicine (Askling, Rombo, Andersson, Martin, Ekdahl)","source_type":"peer_reviewed","statistic":"Per-100,000 person-month traveler incidence by destination: East Africa 14.1; Middle East 5.8 (18 among unprotected); India and neighboring countries 5.6; North Africa 12 unprotected; East Asia 2 unprotected. VFR (visiting friends and relatives) travelers were 70-91% of cases in high-risk regions; children 0-14 had highest incidence at 3.1 per 100,000.","excerpt":"[Paraphrase from abstract — full text paywalled] \"The incidence of reported hepatitis A in unprotected travelers was 14.1 per 100,000 person-months for East Africa, 18 for the Middle East, 12 for North Africa, and 2 for East Asia. Travelers, and especially children, who are visiting friends and relatives (VFR) in endemic areas constitute a high-risk group for acquiring hepatitis A infection, while the risk for unprotected tourists to East Asia is low.\"\n","source_date":"2009-07-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250204055716/https://pubmed.ncbi.nlm.nih.gov/19674261/","calculation_notes":"Askling et al. is the secondary independent corroboration of the headline figure: a Swedish national surveillance study (separate pipeline from Mütsch's Swiss data) reaching the same order of magnitude. Per-region incidence numbers feed the regional_breakdown rows. The VFR-skew finding and the East-Asia-as-low-risk finding feed the personal_factor_multipliers for destination subregion. The 2 per 100,000 person-month figure for East Asia translates to ~1 per 100,000 per 2-week trip, an order of magnitude below the headline; this is the basis for the 0.1× multiplier on \"East Asia tourist routes\" and contributes to the lower edge of the regional breakdown.\n","independence_note":"Askling et al. uses Swedish surveillance data; methodologically independent of Mütsch's Swiss surveillance and of CDC's US data, though all three rely on the same general framework of laboratory-confirmed imported cases with denominators from tourism statistics.\n"}],"comparison_anchors":[{"label":"Travelers' diarrhea per 2-week trip to high-risk destination","lifetime_us_adult":0.5},{"label":"Typhoid in an endemic country (lifetime, global adult)","lifetime_us_adult":0.00153},{"label":"Rabies death via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Japanese encephalitis per traveler-trip (short-term urban Asia)","lifetime_us_adult":5e-7}],"regional_breakdown":[{"region":"Vaccinated traveler (1 or 2 doses, any destination, any duration)","probability":0.000002,"notes":"ACIP: 97-100% seroprotection after 1 dose, 100% after 2 doses, >97% durability at 20 years. Documented vaccine-failure cases in travelers are rare in the published literature."},{"region":"Unvaccinated nonimmune adult, 2-week trip, intermediate endemicity (e.g. parts of Eastern Europe, Caribbean)","probability":0.00003,"notes":"Lower edge of the Mütsch 6-28 per 100,000 person-month band, applied to a 2-week trip."},{"region":"Unvaccinated nonimmune adult, 2-week trip, high endemicity (South Asia, Sub-Saharan Africa, Mexico, Central America)","probability":0.00008,"notes":"Midpoint of the Mütsch nonimmune-traveler range; matches the headline figure on this page."},{"region":"Unvaccinated nonimmune adult, 2-week trip, East Asia tourist circuit","probability":0.00001,"notes":"Askling et al.: ~2 cases per 100,000 person-months for unprotected travelers to East Asia, an order of magnitude below the headline."},{"region":"Unvaccinated nonimmune long-term traveler or expatriate (1+ month rural, high endemicity)","probability":0.0005,"notes":"Upper edge of Mütsch nonimmune range (28 per 100,000 person-months) scaled to ~1.5+ months. CDC explicitly notes 'cumulative risk for hepatitis A and typhoid fever' justifies vaccination for long-term travelers."},{"region":"Unvaccinated nonimmune VFR (visiting friends and relatives) traveler with children in endemic area","probability":0.0008,"notes":"Askling et al.: VFR travelers represent 70-91% of imported cases in high-risk regions; children aged 0-14 show highest incidence (3.1/100,000 trips, 88% VFR). Reflects food/water exposure in non-tourist settings."},{"region":"Endemic-country susceptible child (for reference — not a traveler figure)","probability":0.9,"notes":"WHO: in low- and middle-income countries with poor sanitation, ~90% of children are infected before age 10. Native exposure anchor, not a traveler risk."}],"personal_factor_multipliers":[{"factor":"Full hepatitis A vaccination (2-dose schedule completed)","multiplier":0.01,"notes":"ACIP: 100% seroprotection 1 month after second dose, >97% durability at 20 years. Effectively eliminates per-trip risk."},{"factor":"Single dose of hepatitis A vaccine (≥2 weeks before departure)","multiplier":0.03,"notes":"ACIP: 97-100% seroprotection at 1 month after first dose. CDC accepts a single dose as adequate pre-departure protection for healthy adults <40."},{"factor":"Long-term stay (1+ month) in rural high-endemicity area","multiplier":5,"notes":"Mütsch upper-edge (28 per 100,000 person-months) scaled to multi-month exposure; CDC singles out long-term travelers for cumulative risk."},{"factor":"Visiting friends and relatives (VFR) in endemic area, staying in local households","multiplier":10,"notes":"Askling: VFR travelers 70-91% of imported cases in high-risk regions; non-tourist food/water exposure plus longer typical stays."},{"factor":"Destination is East Asia tourist circuit (Japan/Korea zero-risk; Thailand/Vietnam/Indonesia low)","multiplier":0.1,"notes":"Askling: ~2 per 100,000 person-months for unprotected East Asia travelers, an order of magnitude below the high-endemicity headline."},{"factor":"Age 50+ (incidence unchanged; case fatality and hospitalization risk substantially elevated)","multiplier":1,"notes":"Severity multiplier rather than incidence: CDC reports severe complications and fulminant hepatitis are 'more common in older adults'; 2022 US surveillance shows hepatitis A mortality at 0.15/100,000 in 65+ vs 0.04/100,000 in 45-64. Underlying liver disease compounds this."}],"short_label":"Hepatitis A (travel)","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The ~1-in-12,000 figure applies specifically to an unvaccinated, nonimmune adult on a roughly 2-week trip to a high or intermediate endemicity country (most of South Asia, Sub-Saharan Africa, Mexico, Central America, and large parts of South America and the Middle East). Vaccinated travelers face a risk roughly two orders of magnitude lower; visitors to Japan, Korea, Western Europe, North America, Australia, or New Zealand face essentially zero risk regardless of vaccination status. The number is per-trip, not per-adult-lifetime, and is not a substitute for itinerary-specific advice from a qualified travel-medicine clinician. The Mütsch surveillance pipeline counts laboratory-confirmed cases diagnosed after return; mild and asymptomatic infections are systematically undercounted, so the true exposure rate is somewhat higher than the reported case rate — though for symptomatic travel-related illness (the version most readers care about) the published numbers are the relevant ones. This entry measures the risk of contracting hepatitis A, not of dying from it: the overall case fatality rate is <1%, but it climbs to roughly 2% in adults aged 40+ and higher again in people with underlying liver disease, so a per-trip death estimate for an older unvaccinated traveler sits roughly two orders of magnitude below the incidence figure on this page.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"A single stylized vaccination record card on a flat surface, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/hepatitis-a-travel","api_url":"https://likelier.app/api/fears/hepatitis-a-travel.json"},{"slug":"supervolcano-eruption","question":"What are the odds of a supervolcano eruption in your lifetime?","quick_answer":"A Yellowstone-class supereruption runs about 1 in 12,400 over a lifetime, from a per-year chance near 1 in 730,000. Nearly 1 in 5 US adults report fearing a large volcanic eruption, though almost no one loses sleep over it.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Supervolcanoes occupy a peculiar niche in collective anxiety: most people have heard of Yellowstone's caldera and have a vague sense that an eruption would end civilization, yet almost nobody loses sleep over it. Periodic media cycles — triggered by minor seismic swarms or a clickbait headline about Yellowstone being \"overdue\" — briefly spike public concern before it recedes. No major polling organization tracks supervolcano worry specifically, but the cultural footprint (disaster films, YouTube doomsday compilations) suggests the perceived risk fluctuates between \"impossible\" and \"extinction event\" with little calibration in between.\n","rough_estimate":"19.2% of US adults report being afraid or very afraid of a large volcanic eruption (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~1 in 730,000 per year (USGS estimate for Yellowstone VEI-8)","numerator":1,"denominator":730000,"unit":"per year","population":"Global population, any VEI-8 supereruption (Yellowstone-class)"},"normalized":{"lifetime_us_adult":0.0000808,"display":"~1 in 12,400 lifetime","log_value":-4.09,"assumptions":"USGS estimates the annual probability of a Yellowstone-class VEI-8 supereruption at approximately 1 in 730,000. The global geological catalogue of M8 (VEI-8) eruptions (Mason, Pyle & Oppenheimer 2004) shows a minimum frequency of ≈1.4 events per million years — one roughly every 700,000 years — over the past 36 Ma, so any single system such as Yellowstone accounts for a large share of that risk. Using the USGS per-year figure and compounding over 59 remaining adult years: 1 − (1 − 1/730,000)⁵⁹ ≈ 8.08 × 10⁻⁵. This is a statistical expectation over geological time; USGS notes this figure \"is probably an overestimate for the short term\" given the absence of precursory signals. The probability is global in scope — a VEI-8 eruption anywhere would produce a volcanic winter affecting the entire planet.\n","uncertainty":{"low":0.00002,"high":0.00015},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.usgs.gov/volcanoes/yellowstone/questions-about-supervolcanoes","title":"Questions About Supervolcanoes","publisher":"U.S. Geological Survey","source_type":"govt_report","statistic":"Annual probability of a Yellowstone supereruption is approximately 1 in 730,000","excerpt":"\"The annual probability of a Yellowstone eruption is approximately 1 in 730,000... Given Yellowstone's past history, the yearly probability of another caldera-forming eruption can be approximated as 1 in 730,000 or 0.00014%.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503094235/https://www.usgs.gov/volcanoes/yellowstone/questions-about-supervolcanoes","calculation_notes":"USGS derives the 1-in-730,000 figure from Yellowstone's three caldera-forming eruptions over 2.1 million years (2.1M / 3 ≈ 700,000-year average interval, rounded to 730,000). Annual individual risk: 1/730,000 ≈ 1.37 × 10⁻⁶. Over 59 adult years: 1 − (1 − 1.37 × 10⁻⁶)⁵⁹ ≈ 8.08 × 10⁻⁵. This is the probability of experiencing a supereruption, not dying in one — mortality would depend on location, agricultural collapse severity, and societal response.\n"},{"url":"https://link.springer.com/article/10.1007/s00445-004-0355-9","title":"The size and frequency of the largest explosive eruptions on Earth","publisher":"Bulletin of Volcanology (Mason, Pyle, Oppenheimer)","source_type":"peer_reviewed","statistic":"M8 (VEI-8) eruptions recur roughly once per 700,000 years globally (minimum ≈1.4 events per million years over the past 36 Ma)","excerpt":"\"Eruptions of size M8 and larger have occurred with a minimum frequency of ≈1.4 events/Ma in two pulses over the past 36 Ma. On the basis of the activity during the past 13.5 Ma, there is at least a 75% probability of a M8 eruption (>10¹⁵ kg) occurring within the next 1 Ma.\"\n","source_date":"2004-11-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251025065256/https://link.springer.com/article/10.1007/s00445-004-0355-9","calculation_notes":"Mason, Pyle & Oppenheimer catalogue 47 known M≥8 eruptions, of which 42 fall in the past 36 Ma, giving a minimum global M8 frequency of ≈1.4 events/Ma — i.e. one roughly every 700,000 years, comparable to the USGS Yellowstone-specific interval when only a single system is considered. The record is incomplete (a minimum count), and other authors (e.g. Pyle 2000; Self 2006) place the effective supereruption recurrence at one per 100,000–200,000 years once undercounting is corrected. Because the older record is progressively less complete, the USGS single-system Yellowstone estimate is the more conservative, better-constrained figure used for the headline calculation.\n","independence_note":"Mason et al. use a global geological catalogue of large explosive eruptions independent of USGS Yellowstone Volcano Observatory monitoring data.\n"},{"url":"https://www.usgs.gov/observatories/yvo/news/so-when-will-next-eruption-yellowstone-happen","title":"So, when will the next eruption at Yellowstone happen?","publisher":"U.S. Geological Survey (Yellowstone Volcano Observatory, Caldera Chronicles)","source_type":"govt_report","statistic":"Yellowstone is not 'overdue'; the ~0.001% annual eruption probability is likely an overestimate for the short term","excerpt":"\"Based on our current knowledge of Yellowstone's eruptive history, the annual probability of a volcanic eruption is on the order of 0.001%, but even this low number is probably an overestimate for the short term.\"\n","source_date":"2024-11-04","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260219125524/https://www.usgs.gov/observatories/yvo/news/so-when-will-next-eruption-yellowstone-happen","calculation_notes":"USGS explicitly cautions that the naive recurrence-interval calculation overstates near-term risk because Yellowstone shows no precursory magmatic signals consistent with an impending eruption. This caveat is reflected in the wide uncertainty band applied to the normalized lifetime figure.\n"}],"comparison_anchors":[{"label":"Asteroid impact death (lifetime)","lifetime_us_adult":7.4e-7},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Earthquake death (lifetime, US)","lifetime_us_adult":0.00024}],"short_label":"Supervolcano eruption","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"The \"lifetime probability\" of a supervolcano eruption shares the same statistical fiction as asteroid impact risk: the expected value is dominated by an event so rare that no human civilization has witnessed one. The last VEI-8 eruption (Taupo's Oruanui event) occurred roughly 26,500 years ago. The USGS figure of 1 in 730,000 per year is a frequency estimate from three data points over 2.1 million years — not a forecast. If a VEI-8 eruption did occur, the probability of individual death would depend heavily on proximity and the severity of the subsequent volcanic winter; global agricultural collapse could kill far more people than the eruption itself.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A solitary volcanic caldera silhouette against a muted sky, flat vector illustration with subdued warm tones."},"canonical_url":"https://likelier.app/supervolcano-eruption","api_url":"https://likelier.app/api/fears/supervolcano-eruption.json"},{"slug":"neodymium-magnet-ingestion","question":"What are the odds of a child swallowing multiple high-powered magnets and needing hospitalization?","quick_answer":"A US child's odds of swallowing multiple high-powered magnets and needing hospitalization are about 1 in 12,000 over ages 0-14 — roughly 300-400 hospitalizations a year. Parents underrate it: nearly half in a post-incident survey thought the magnets were children's toys, though they hospitalize kids at around 20 times the rate of button batteries.\n","category":"kids","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Toy magnet sets and decorative high-powered magnets have been marketed as adult desk toys and stress-relief products, but the hazard when a child swallows multiple magnets is not well understood by most parents. The dominant public mental model treats magnet ingestion like coin swallowing — unpleasant but self-resolving. What is not widely known is that the mechanism is fundamentally different for two or more high-powered magnets: they attract each other across different loops of intestine, creating a pressure point that can cause ischemia, perforation, volvulus, and bowel necrosis within hours. Middelberg and colleagues (2022) found that nearly half of caregivers whose child was treated for high-powered magnet injury believed high-powered magnets were children's toys, and only 7% knew the magnets had previously been removed from the market by CPSC. The hazard is substantially underrated relative to button batteries, despite hospitalizing children at roughly 20 times the rate.\n","rough_estimate":"Most parents carry no number for this hazard; nearly half in a post-incident survey believed high-powered magnets were children's toys (Middelberg 2022)","kind":"intuition"},"native":{"display":"~300-400 hospitalizations per year in US children from multiple high-powered magnet ingestion","numerator":1,"denominator":12000,"unit":"per child, 0-14 window","population":"US children ages 0-14, hospitalization from multiple high-powered magnet ingestion"},"normalized":{"lifetime_us_adult":0.000083,"display":"~1 in 12,000 per US child during the 0-14 window (hospitalization from multiple magnets)","log_value":-4.08,"assumptions":"CPSC estimated 26,600 magnet-related ED visits from 2010-2021 across all ages 0-17 (Federal Register, September 2022). Post-rule-vacated 2017-2020, annual ED visits were approximately 2,300-2,400/year for high-powered magnets specifically. Middelberg et al. (2022, Pediatrics), the largest clinical cohort study (596 confirmed high-powered magnet cases across 25 children's hospitals, 2017-2019), found 55.7% required hospitalization and 95% of cases were children under 14. Applying the 56% hospitalization rate to ~2,400 high-powered magnet ED visits/year yields approximately 300-400 hospitalizations per year for the post-reentry period. US children ages 0-14 number approximately 61 million (Census). Cumulative risk over the 15-year window: 300-400 hospitalizations/year × 15 years / 61,000,000 ≈ 7.4-9.8 × 10⁻⁵, or roughly 1 in 10,000 to 1 in 14,000. The headline uses 1 in 12,000 as the midpoint. Strickland et al. (2020, JPGN) independently found approximately 1,094 cases of multi-magnet \"escalation of care\" over 2017-2019 (weighted NEISS estimate), consistent with 350-400 serious outcomes per year. A new CPSC rule (16 CFR Part 1262) took effect October 2022; its effect on case rates is not yet quantifiable from published data.\n","uncertainty":{"low":0.00005,"high":0.00014},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/35112127/","title":"High-Powered Magnet Exposures in Children: A Multi-Center Cohort Study","publisher":"Pediatrics — Middelberg et al. (IMPACT of Magnets Research Collaborative), Nationwide Children's Hospital, 2022","source_type":"peer_reviewed","statistic":"596 patients confirmed high-powered magnet exposure (25 children's hospitals, 2017-2019); 55.7% hospitalized; 46.3% required endoscopy/surgery; 9.6% had life-threatening morbidity; 95% of care was for children under 14; median age 7.5 years; nearly half of caregivers believed high-powered magnets were children's toys","excerpt":"\"Nearly 600 cases of high-powered magnet-related injuries in the three years after high-powered magnets re-entered the US market (2017 to 2019)... the majority (56%) of children being treated for high-powered magnet-related injuries required hospitalization... nearly one in ten had a potentially life-threatening injury such as bowel obstruction, perforation, infection, bleeding, fistulae or volvulus... 95% of care in this study was for children under 14 years of age... almost half of caregivers believed high-powered magnets were children's toys.\"\n","source_date":"2022-02-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260102033853/https://pubmed.ncbi.nlm.nih.gov/35112127/","calculation_notes":"The 56% hospitalization rate from 596 confirmed cases across 25 children's hospitals is the primary clinical outcome rate. Applying this to the CPSC's nationally estimated ~600/yr high-powered magnet ED visits gives ~336 hospitalizations/yr. The 9.6% life-threatening morbidity rate (57/596) translates to approximately 57 life-threatening events per year nationally from high-powered magnet ingestion alone.\n","independence_note":"Middelberg et al. use a prospective multi-center cohort design distinct from the NEISS-based population estimates of Strickland 2020 and the CPSC product-incident database. The clinical outcome rates (hospitalization, surgical intervention, life-threatening morbidity) are empirically measured rather than derived from administrative codes.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32969961/","title":"Magnet Ingestions in Children Presenting to Emergency Departments in the United States, 2009–2019: A Problem in Flux","publisher":"Journal of Pediatric Gastroenterology and Nutrition — Strickland, Reeves, Krishnamurthy, Bhatt, Mahajan, Abbas, 2020","source_type":"peer_reviewed","statistic":"23,756 weighted NEISS-estimated pediatric magnet-ingestion ED visits 2009-2019; hospitalization rate rose to 37.7% in 2017-2019 (post-rule-vacated period); 5-fold increase in multiple-magnet escalation of care (estimated 1,094 cases) over 2017-2019; annual case increase averaged 6.1%","excerpt":"\"An estimated 23,756 children (59% males, 42% < 5 years old) presented with a [suspected magnet ingestion] from 2009 to 2019 with an average annual case increase of 6.1%... From 2017 to 2019, there was a greater proportion of small/round type magnet ingestions and multiple magnet ingestions, and... a 5-fold increase in the escalation of care for multiple magnet ingestions (estimated n = 1,094; CI 505-1,686).\"\n","source_date":"2020-12-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260525095719/https://pubmed.ncbi.nlm.nih.gov/32969961/","calculation_notes":"1,094 escalation-of-care multiple-magnet cases over 2017-2019 (3 years) ≈ 365/yr — consistent with the ~300-400 hospitalization estimate derived from Middelberg's clinical hospitalization rate applied to CPSC's annual ED visit count. Strickland's NEISS-based approach provides an independent population-level estimate that corroborates the Middelberg clinical-cohort approach.\n","independence_note":"Strickland et al. use NEISS (National Electronic Injury Surveillance System), a statistically representative sample of US EDs maintained by CPSC, yielding nationally weighted estimates. This is methodologically independent of the Middelberg multi-center clinical cohort and of the CPSC product-incident administrative database.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5548a3.htm","title":"Gastrointestinal Injuries from Magnet Ingestion in Children — United States, 2003–2006","publisher":"CDC Morbidity and Mortality Weekly Report, December 8, 2006, Vol. 55, No. 48","source_type":"govt_report","statistic":"20 CPSC-identified serious magnet GI injury cases (2003-2006); 19 required surgery; 75% had bowel perforation; 1 death (first documented US fatality); all cases involved multiple magnets attracting through intestinal loops","excerpt":"\"Since 2003, CPSC has identified one death and 19 other serious cases involving magnets with this type of unusual strength, representing a range of injuries including intestinal perforations (15 cases), obstruction (4), volvulus (3), and peritonitis (4)... The magnets had magnetically joined across two loops of intestine, causing a volvulus that compromised the blood supply to the bowel and led to necrosis, perforation, and sepsis.\"\n","source_date":"2006-12-08","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260511124750/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5548a3.htm","calculation_notes":"This MMWR report established the mechanism: all 20 serious cases involved multiple magnets (or magnets plus a metal object) attracted through the bowel wall. Single-magnet ingestion was not represented in the serious-outcome cases. The 75% bowel-perforation rate in this early clinical case series is consistent with the 9.6% life-threatening morbidity rate in Middelberg 2022 — both measure the outcome rate in cases reaching surgical review, which is a selected severe subset. The first US death (a 20-month-old child, 2005/2006) is documented in this report.\n"},{"url":"https://www.federalregister.gov/documents/2022/09/21/2022-20200/safety-standard-for-magnets","title":"Safety Standard for Magnets, 16 CFR Part 1262","publisher":"US Consumer Product Safety Commission, Federal Register, September 21, 2022","source_type":"govt_report","statistic":"26,600 magnet-related ED visits 2010-2021 (NEISS estimate); 7-8 deaths involving hazardous magnets since 2005 (5 confirmed US deaths); 124 documented surgical cases (2010-2020); new rule requires flux index <50 kG²mm² for all consumer magnet products fitting small-parts cylinder","excerpt":"\"CPSC estimates 26,600 magnet ingestions were treated in hospital ERs from 2010 through 2021, and cases have been rising annually since 2018. CPSC is aware of seven deaths involving the ingestion of hazardous magnets (including two outside of the United States).\"\n","source_date":"2022-09-21","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260510090100/https://www.federalregister.gov/documents/2022/09/21/2022-20200/safety-standard-for-magnets","calculation_notes":"26,600 ED visits / 12 years ≈ 2,217/yr on average. Five confirmed US deaths across ~17 years (2005-2022) ≈ 0.3/yr — a low annual fatality rate that translates to approximately 1 in 13 million per child over the 0-14 window. The new federal rule (effective October 2022) applies to all consumer products with separable magnets fitting the small-parts cylinder, not just the \"magnet sets\" definition that the vacated 2014 rule covered. Post-enforcement case data are not yet available in peer-reviewed literature.\n"}],"comparison_anchors":[{"label":"Button battery serious injury (US child, 0-6 window)","lifetime_us_adult":0.000004},{"label":"Child pool drowning death (US child, ages 0-14)","lifetime_us_adult":0.00003},{"label":"Laundry pod serious medical outcome (US child, 0-5 window)","lifetime_us_adult":0.000154}],"short_label":"Magnet ingestion","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1-in-12,000 headline refers to hospitalization from multiple high-powered magnet ingestion during the 0-14 childhood window. Fatal outcomes are roughly two orders of magnitude rarer — approximately 1 in 13 million per child (5 confirmed US deaths across ~17 years against a population of 61 million children ages 0-14). The hazard is specifically about multiple magnets (or one magnet plus a metal object): a single magnet ingestion is almost universally self-resolving. This distinction is critical for both clinical management and parental risk communication. The probability range is sensitive to the regulatory cycle: cases were ~2,300/year before the 2014 CPSC rule, fell during the rule period (2015-2016), surged back to ~2,300-2,400/year after the rule was vacated in 2016, and the new 2022 federal rule (16 CFR Part 1262) has an unknown but potentially substantial effect on future case rates. The 300-400 hospitalization estimate reflects the post-rule-vacated, pre-new-rule period (2017-2022); post-2022 rates may be lower if the new rule achieves better compliance than the vacated one. The magnet hazard sits in an unusual position relative to other entries on this site: the hospitalization rate (~1 in 12,000) is roughly 20 times the button battery serious-injury rate (~1 in 250,000), despite button batteries receiving substantially more public-health communication and regulatory attention. The availability gap is partly explained by the magnet hazard's older mean patient age (7.5 years, Middelberg) versus the predominantly toddler population for button batteries.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"Two small silvery spherical magnets attracted together, viewed from close up, flat vector illustration with muted metallic tones."},"canonical_url":"https://likelier.app/neodymium-magnet-ingestion","api_url":"https://likelier.app/api/fears/neodymium-magnet-ingestion.json"},{"slug":"accidental-gun-death","question":"What are the odds of dying from an unintentional firearm discharge?","quick_answer":"About 1 in 11,300 over a lifetime. Accidental shootings draw heavy coverage, but they are rare — the vast majority of US gun deaths are suicides and homicides, not unintentional discharges.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Accidental shootings receive outsized media coverage relative to their frequency, partly because they involve a strong narrative element — a child finding a loaded gun, a hunter mistaking a companion for game, a cleaning mishap. No national survey isolates \"fear of dying from an accidental gun discharge\" as a standalone item distinct from gun-violence fear generally, so the perceived side here is editorial intuition. Americans who live in gun-owning households tend to anchor on anecdotal cases; those who do not tend to lump accidental discharges into a broader \"gun death\" category that is dominated by suicide and homicide.\n","rough_estimate":"most people overestimate; gun-death fear is dominated by intentional violence","kind":"intuition"},"native":{"display":"~500 unintentional firearm deaths per year in the US","numerator":15,"denominator":10000000,"unit":"per year","population":"US residents, all ages"},"normalized":{"lifetime_us_adult":0.0000885,"display":"1 in ~11,300 lifetime (US adult)","log_value":-4.05,"assumptions":"CDC NVSS data for 2022 records approximately 500 unintentional firearm deaths (ICD-10 W32–W34), yielding an annual rate of roughly 0.15 per 100,000 (1.5 per 10,000,000). Compounded over 59 years of remaining adult life at constant hazard: 1 − (1 − 0.0000015)^59 ≈ 0.0000885 ≈ 1 in 11,300. The NSC's published lifetime odds (1 in 9,567) use a from-birth whole-life span rather than 59 adult years. Our figure is the adult-only version consistent with the site's standard normalization.\n","uncertainty":{"low":0.000059,"high":0.000118},"scope":"us_adult_lifetime"},"sources":[{"url":"https://wisqars.cdc.gov/reports/","title":"WISQARS Fatal Injury Reports","publisher":"CDC National Center for Injury Prevention and Control","source_type":"govt_report","statistic":"~500 unintentional firearm deaths in the US in 2022; rate ~0.15 per 100,000","excerpt":"[Paraphrase from WISQARS query result — the cited page is an interactive query tool, not a static quotable page.] A WISQARS fatal-injury query for cause \"Firearm — Unintentional\" (ICD-10 W32–W34), 2022, all ages, United States, returns on the order of ~500 deaths and a crude rate of roughly 0.15 per 100,000 population — under 1% of all firearm deaths that year.\n","source_date":"2024-01-15","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260309013044/https://wisqars.cdc.gov/reports/","calculation_notes":"CDC WISQARS counts deaths from unintentional firearm discharge coded as W32 (handgun), W33 (rifle/shotgun/larger), and W34 (other/unspecified firearm). The 2022 count of ~500 deaths among a population of ~333 million yields 500/333,000,000 ≈ 1.5 per 10,000,000 per year. Lifetime over 59 adult years: 1 − (1 − 0.0000015)^59 ≈ 0.0000885. Uncertainty band uses the 2015–2022 range of annual counts (~430 to ~550), reflecting year-over-year variation rather than sampling error.\n","independence_note":"CDC WISQARS draws from death certificates filed via the National Vital Statistics System. The NSC source below uses the same underlying CDC mortality data but applies its own actuarial methodology for lifetime odds, providing an independent analytical check.\n"},{"url":"https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","title":"Odds of Dying (2024 Data) — Injury Facts","publisher":"National Safety Council","source_type":"reputable_reference","statistic":"Lifetime odds of dying from accidental gun discharge: 1 in 9,567 (US, 2024 data)","excerpt":"\"Lifetime odds of death for selected causes, United States, 2024 … Accidental gun discharge — 1 in 9,567\" (from the NSC \"Odds of Dying\" summary table; the same table lists Guns (all intents) 1 in 97, Suicide with gun 1 in 156, and Gun assault 1 in 280).\n","source_date":"2026-03-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260309064046/https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","calculation_notes":"NSC's \"Odds of Dying\" table reports lifetime odds for \"Accidental gun discharge\" as 1 in 9,567 (United States, 2024 data). NSC computes lifetime odds on a from-birth basis using U.S. Census Bureau population and life-expectancy data, so its whole-life denominator is longer than our 59-year adult-only horizon; that is why the NSC lifetime odds (1 in 9,567) and our adult-only figure (1 in 11,300) differ. Both rest on the same order-of-magnitude CDC unintentional- firearm-death count (~500/year); the gap is the denominator period, not the underlying deaths.\n","independence_note":"NSC is an independent nonprofit that repackages CDC mortality data with its own actuarial framing. The underlying death counts are the same, but the analytic layer is independent.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.00348},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":7,"notes":"CDC WISQARS 2022: males account for ~85% of unintentional firearm deaths despite being ~49% of the population, yielding roughly 7× the female rate."},{"factor":"Rural residence vs. urban","multiplier":2,"notes":"CDC WISQARS state-level data: rural states with high firearm ownership (e.g., Alaska, Montana) consistently show approximately 2× the unintentional firearm death rate of urban-dense states."}],"short_label":"Accidental gun death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"\"Unintentional firearm discharge\" is a classification applied at death-certificate coding, and the boundary between unintentional, undetermined-intent, and negligent homicide is not always clean. Some fraction of deaths coded as unintentional may involve negligence that a different medical examiner would classify differently, and vice versa. The rate varies sharply by demographics: children and adolescents are disproportionately represented as victims, males account for roughly 85% of deaths, and rural states with higher gun-ownership rates tend to have higher unintentional firearm death rates. The pooled 1-in-11,300 figure is a population average that may understate the risk for a child in a household with unsecured firearms and overstate it for a non-gun-owning urban adult by an order of magnitude or more.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-opus-4-6-research","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single padlock resting on a flat surface, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/accidental-gun-death","api_url":"https://likelier.app/api/fears/accidental-gun-death.json"},{"slug":"skipping-allergy-immunotherapy","question":"What are the odds of serious harm from not doing allergy immunotherapy?","quick_answer":"Serious harm from skipping allergy immunotherapy is very small: lifetime fatal anaphylaxis from any cause is about 1 in 11,000, and most of that is not even from the environmental allergens immunotherapy targets. Allergic rhinitis progresses to asthma in roughly 10-40% of patients over decades, but that is manageable, and the 'skip it and you will be seriously harmed' framing overstates the natural history.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Allergen immunotherapy — subcutaneous (SCIT, allergy shots) or sublingual (SLIT, drops or tablets) — is often presented to patients as essential to prevent allergic disease from worsening. The \"allergic march\" narrative holds that untreated allergic rhinitis progresses to asthma, and that immunotherapy can halt this trajectory. While there is evidence supporting immunotherapy's efficacy in symptom reduction and some evidence for prevention of new sensitizations in children, the framing that skipping immunotherapy leads to serious harm overstates the natural history. Most adults with allergic rhinitis manage effectively with antihistamines and nasal corticosteroids for decades without catastrophic progression. Fatal anaphylaxis from environmental allergens is exceedingly rare regardless of treatment status.\n","rough_estimate":"Many allergists present immunotherapy as necessary to prevent progression; patients fear worsening without it","kind":"intuition"},"native":{"display":"~1.5 fatal anaphylaxis events per million person-years (all causes); allergic rhinitis → asthma progression ~10-40% over decades","numerator":15,"denominator":10000000,"unit":"per year (fatal anaphylaxis, all triggers)","population":"US general population, all-cause anaphylaxis fatalities"},"normalized":{"lifetime_us_adult":0.00009,"display":"~1 in 11,000 (lifetime fatal anaphylaxis from any cause, US adult)","log_value":-4.05,"assumptions":"Fatal anaphylaxis from all causes occurs at approximately 0.5-1.0 per million person-years in the US (summing drug 0.27-0.51/M, food 0.03-0.3/M, venom ~0.1/M from PMC5589409). We use 1.5/M as a conservative upper bound. Over a 59-year adult lifetime at constant hazard: 1 − (1 − 0.0000015)^59 ≈ 0.000089, or roughly 1 in 11,000.\nThis is an upper bound for the \"serious harm from skipping immunotherapy\" question, because: (a) most fatal anaphylaxis is from food, drugs, and venom — not the environmental allergens (dust mites, pollen, mold) that immunotherapy primarily targets; (b) immunotherapy itself carries a small but non-zero anaphylaxis risk (~1 fatal reaction per 2.5 million injections, Bernstein 1990-2001 survey); (c) the more common concern — allergic rhinitis progressing to asthma — affects 10-40% of allergic rhinitis patients over decades, but asthma itself is manageable with modern inhaled corticosteroids and the attributable mortality from allergic asthma in treated populations is very low.\nThe entry uses fatal anaphylaxis as the normalized metric because it is the most objectively severe outcome. The more common outcome of interest — progression from rhinitis to asthma — is not a disability in the traditional sense but a manageable chronic condition.\n","uncertainty":{"low":0.00005,"high":0.0002},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5589409/","title":"Fatal Anaphylaxis: Mortality Rate and Risk Factors","publisher":"Current Allergy and Asthma Reports (PMC)","source_type":"peer_reviewed","statistic":"Fatal anaphylaxis rates by cause: drug 0.27-0.51/M, food 0.03-0.3/M, venom ~0.1/M per year; aggregate ~0.5-1.0/M summing across causes","excerpt":"\"In the United States, using International Classification of Diseases-10 (ICD-10) categorization, the estimated fatal drug anaphylaxis rate increased significantly from 0.27 per million population in 1999-2001 to 0.51 per million population in 2008-2010.\"\n","source_date":"2017-08-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260223055849/https://pmc.ncbi.nlm.nih.gov/articles/PMC5589409/","calculation_notes":"This review provides the mortality denominator for the most severe possible outcome of allergic disease. The 0.5-1.5 per million range covers all-cause anaphylaxis fatalities. Environmental allergens (pollen, dust mites) are a negligible share of fatal anaphylaxis triggers — the vast majority are drugs, food, and Hymenoptera venom. This means the risk of fatal anaphylaxis specifically attributable to the allergens that immunotherapy targets is well below the already-low all-cause rate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1351177/","title":"Greater risk of incident asthma cases in adults with Allergic Rhinitis and Effect of Allergen Immunotherapy: A Retrospective Cohort Study","publisher":"Respiratory Research (PMC)","source_type":"peer_reviewed","statistic":"Allergic rhinitis patients had 7.8-10.3x greater risk of developing asthma; immunotherapy reduced new-onset asthma (OR 0.53)","excerpt":"\"Treatment with allergen immunotherapy was significantly and inversely related to the development of new onset asthma (OR, 0.53; 95%CI, 0.32-0.86). Presence of allergic rhinitis at the start of the study was highly predictive of development of new onset asthma after 10 years (OR, 10.3; 95%CI, 4.8-21.8).\"\n","source_date":"2006-01-06","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251020201644/https://pmc.ncbi.nlm.nih.gov/articles/PMC1351177/","calculation_notes":"This retrospective cohort study provides the key evidence for the \"allergic march\" — rhinitis patients have ~8-10x the odds of developing asthma compared to non-allergic controls (OR 10.3 univariate, 7.81 multivariate). Immunotherapy approximately halved the risk of new-onset asthma (OR 0.53). However, this is a relative risk reduction: if the baseline risk of developing asthma over 10 years is 10-15% for rhinitis patients, immunotherapy reduces it to ~5-8%. Important context: developing asthma is not the same as developing serious harm — most allergic asthma is well-controlled with inhaled corticosteroids.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15208595/","title":"Twelve-year survey of fatal reactions to allergen injections and skin testing: 1990-2001","publisher":"Journal of Allergy and Clinical Immunology (PubMed)","source_type":"peer_reviewed","statistic":"Fatal reactions to allergen immunotherapy occurred roughly once per 2.5 million injections, averaging 3.4 deaths per year (US, 1990-2001)","excerpt":"\"It was estimated that fatal reactions occurred every 1 per 2.5 million injections, with an average of 3.4 deaths per year.\"\n","source_date":"2004-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250223050603/https://pubmed.ncbi.nlm.nih.gov/15208595/","calculation_notes":"This twelve-year practitioner survey (Bernstein et al.) is the primary source for the immunotherapy mortality rate: roughly 1 fatal reaction per 2.5 million injections, or about 3.4 deaths per year nationally. This is important for the risk-benefit framing: immunotherapy itself carries a small but real mortality risk. If the condition being treated (environmental allergies) has a near-zero mortality risk when managed with antihistamines, the risk-benefit calculus for immunotherapy depends primarily on quality-of-life improvement rather than mortality prevention.\n"}],"comparison_anchors":[{"label":"Fatal bee sting (lifetime, US adult)","lifetime_us_adult":0.0006},{"label":"Fatal food anaphylaxis (lifetime, US adult)","lifetime_us_adult":0.0002},{"label":"Asthma death (lifetime, US adult with asthma)","lifetime_us_adult":0.003}],"personal_factor_multipliers":[{"factor":"existing allergic asthma","multiplier":5,"notes":"Patients with established allergic asthma have higher morbidity from untreated environmental allergies due to asthma exacerbations triggered by allergen exposure."},{"factor":"child with allergic rhinitis","multiplier":3,"notes":"Children have higher rates of allergic march progression (rhinitis → asthma). Immunotherapy evidence for preventing asthma development is strongest in pediatric populations."},{"factor":"well-controlled on antihistamines","multiplier":0.2,"notes":"Adults with allergic rhinitis well-controlled on antihistamines and nasal corticosteroids have minimal incremental benefit from immunotherapy in terms of serious-harm prevention."},{"factor":"venom allergy with prior systemic reaction","multiplier":8,"notes":"Golden et al. JACI: patients with prior systemic anaphylaxis to Hymenoptera venom have 30-60% risk of recurrence on re-sting without venom immunotherapy; this is a qualitatively different risk profile from inhalant allergy and is the setting where skipping immunotherapy most clearly increases serious-harm probability"},{"factor":"completed fewer than 3 years of immunotherapy before stopping","multiplier":3,"notes":"Calderon et al. Cochrane Review (2007) and subsequent SCIT duration studies show that patients who stop subcutaneous immunotherapy before 3 years have substantially higher symptom relapse rates than those who complete 3-5 years; incomplete courses are associated with loss of treatment-induced tolerance within 1-3 years of stopping"}],"short_label":"Skipping allergy immunotherapy","myth_framing":"overrated","outcome_severity":"moderate_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses environmental allergen immunotherapy (pollen, dust mites, mold, pet dander) — not venom immunotherapy, which has a clearer risk-benefit case for patients with systemic venom reactions. The normalized metric uses fatal anaphylaxis as the severity anchor, but the more common concern is quality-of-life degradation from poorly controlled rhinitis and asthma progression. Immunotherapy is effective at reducing symptoms and may reduce the risk of developing asthma in children with rhinitis. The \"overrated\" framing applies to the fear of serious harm from NOT doing immunotherapy, not to immunotherapy's efficacy as a treatment. Patients with severe, medication-refractory allergic rhinitis or allergic asthma may benefit substantially from immunotherapy. The entry does not apply to food allergy oral immunotherapy (OIT), which has a different risk profile.\n","quality_score":{"d1":2,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":4,"d8":4,"avg":3.875,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A row of allergy testing scratch marks on a forearm, flat vector illustration in muted clinical tones."},"canonical_url":"https://likelier.app/skipping-allergy-immunotherapy","api_url":"https://likelier.app/api/fears/skipping-allergy-immunotherapy.json"},{"slug":"malaria-travel","question":"What are the odds of contracting malaria as a traveler to an endemic country?","quick_answer":"About 1 in 10,000 for a reference 2-week trip to Sub-Saharan Africa when chemoprophylaxis is taken as directed — and roughly 50 times higher (~1 in 200) on the same trip without it. Risk varies enormously by destination and preparation.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Malaria sits awkwardly in the traveler’s risk imagination. For some readers — the ones booking a two-week resort stay in Cancun or a Bangkok stopover — it looms as an exotic, serious, vaguely Victorian danger. For others — the ones heading to rural Nigeria without a travel clinic visit — it is a background rumor that gets waved off until fever starts. We haven’t found a rigorous recent survey that isolates “fear of catching malaria on a trip” from general travel-health anxiety, so the perceived side of this entry is marked as editorial intuition. The working prior most Likelier readers carry into planning is roughly “small but real” — which, averaged across all destinations and prep levels, is fine, but hides a range of five or six orders of magnitude underneath.\n","rough_estimate":"most travelers guess somewhere in the low percent range, regardless of destination","kind":"intuition"},"native":{"display":"~1 in 10,000 per trip (2-week Sub-Saharan Africa visit, prophylaxis taken as directed)","numerator":1,"denominator":10000,"unit":"per traveler-trip","population":"reference traveler: 2-week visit to Sub-Saharan Africa, chemoprophylaxis compliant"},"normalized":{"lifetime_us_adult":0.0001,"display":"~1 in 10,000 per reference trip","log_value":-4,"assumptions":"The headline figure is an order-of-magnitude estimate for a reference traveler: a 2-week leisure trip to Sub-Saharan Africa, taking a modern chemoprophylaxis regimen (atovaquone-proguanil, doxycycline, or mefloquine) as directed, using DEET and sleeping in a screened or air-conditioned room. It is NOT a lifetime figure for a US adult; malaria risk for travelers is overwhelmingly a per-trip, per-destination question, so this entry uses scope “activity_specific_lifetime” to mean “per traveler-trip.” The same itinerary without prophylaxis is roughly 50× riskier (~1 in 200). A 2-week beach trip to Mexico or a stopover in Bangkok is several orders of magnitude lower (see regional_breakdown). The uncertainty band reflects the range across prophylaxis regimens, seasons, and rural-vs-urban exposure within Sub-Saharan Africa.\n","uncertainty":{"low":0.00002,"high":0.0005},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/malaria.html","title":"Malaria — CDC Yellow Book 2024 (Health Information for International Travel)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Almost all ~2,000 US malaria cases per year are imported; 93% of 2019 cases with known country of acquisition came from Africa, 4% from Asia, 2% from the Caribbean and Americas, <1% from Oceania and the Middle East; 76% of US civilian cases were in travelers visiting friends and relatives.","excerpt":"\"Travelers going to malaria-endemic destinations are at risk of contracting the disease. &hellip; almost all the approximately 2,000 cases of malaria that occur each year in the United States are imported. &hellip; Of cases in 2019 for which country of acquisition was known, 93% were acquired in Africa, 4% in Asia, 2% in the Caribbean and the Americas, and &lt;1% in Oceania and the Middle East. &hellip; Of U.S. civilians with malaria who reported a reason for travel, 76% were visiting friends and relatives. &hellip; No antimalarial drug is 100% protective, so travelers must combine chemoprophylaxis with mosquito avoidance and personal protective measures.\"\n","source_date":"2023-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420043128/https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/malaria.html","calculation_notes":"CDC Yellow Book is the authoritative traveler-facing guidance. The 93% Africa share among US imported cases, combined with the fact that US residents take on the order of 15–20 million trips per year to malaria-endemic countries, is the basis for the regional_breakdown ordering: Sub-Saharan Africa dominates, everywhere else is a rounding error. The 76% VFR (visiting friends and relatives) share corroborates the “prophylaxis missed or stopped early” personal factor — VFR travelers historically under-use chemoprophylaxis relative to tourists. Per-trip attack rates on this page are order-of-magnitude estimates from the broader travel-medicine literature; the specific West-Africa wet-season figures are not drawn verbatim from the cited Massad/Behrens Amazon-region paper (see that source's notes for the setting/species caveat). The ~20× reduction with compliant prophylaxis is corroborated by CDC surveillance (95% of US cases in travelers who took no appropriate prevention).\n","independence_note":"CDC Yellow Book is the primary US traveler-facing clinical guidance, synthesised from CDC NMSS surveillance and the peer-reviewed travel-medicine literature. Shares publisher (and NMSS upstream) with the CDC Malaria Surveillance source below — treat the two CDC citations as one institutional voice; the WHO World Malaria Report and the Massad/Behrens modelling paper provide the genuine independent verification.\n"},{"url":"https://www.cdc.gov/malaria/php/surveillance-report/index.html","title":"Data and Statistics on Malaria in the United States","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~2,000 US malaria cases reported per year; average of nearly 7 deaths per year over 2007–2022; 95% of US malaria patients did not take appropriate malaria prevention medication.","excerpt":"\"Approximately 2,000 malaria cases a year are reported in the United States, and on average there were nearly 7 deaths per year for the period 2007–2022. &hellip; 95% of people with malaria did not take appropriate malaria prevention medication.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175159/https://www.cdc.gov/malaria/php/surveillance-report/index.html","calculation_notes":"CDC domestic surveillance anchors the “prophylaxis works” claim: 95% of confirmed US cases are in travelers who did not take appropriate prevention, which implies prophylaxis compliance drops risk by roughly an order of magnitude or more relative to no prevention (exact reduction varies by regimen, destination, and adherence). The ~7 deaths / ~2,000 cases ratio gives a US case-fatality rate of ~0.35% with access to wealthy-country critical care — consistent with the ~0.5% figure for treated P. falciparum in high-income health systems used in the body text.\n","independence_note":"CDC Yellow Book and CDC Malaria Surveillance share CDC as publisher, but they are distinct pipelines: the Yellow Book is clinical guidance synthesized from travel-medicine literature and NMSS, while the surveillance report is the raw NMSS case and mortality count. We treat them as corroborating rather than independent.\n"},{"url":"https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2024","title":"World Malaria Report 2024","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"263 million malaria cases and 597,000 deaths worldwide in 2023; ~95% of deaths occurred in the WHO African Region.","excerpt":"\"there were an estimated 263 million cases and 597 000 malaria deaths worldwide in 2023 &hellip; 11 million more cases in 2023 compared to 2022, and nearly the same number of deaths. &hellip; Approximately 95% of the deaths occurred in the WHO African Region.\"\n","source_date":"2024-12-11","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175237/https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2024","calculation_notes":"WHO World Malaria Report gives the global denominator: 263 million cases per year, 95% of deaths in the WHO African Region. This concentration is what drives the regional_breakdown in this entry: a 2-week trip to Sub-Saharan Africa dominates every other travel malaria risk by two to six orders of magnitude. WHO data is methodologically independent of CDC surveillance (WHO programmatic estimates vs US NMSS case reports), so this is a genuine independent corroboration of where the risk lives.\n","independence_note":"WHO programmatic malaria estimates are derived from country-level case reporting and modeled adjustments, a separate pipeline from CDC’s US-facing NMSS surveillance. Independent corroboration on the “95% of deaths are in the WHO African Region” figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20015392/","title":"Modeling the risk of malaria for travelers to areas with stable malaria transmission","publisher":"Malaria Journal / Massad E, Behrens RH, Burattini MN, Coutinho FAB","source_type":"peer_reviewed","statistic":"Modeled per-visit falciparum-malaria risk for non-immune travelers to the Amazon region of Brazil: 0.5% per visit for stays longer than 4 months during peak transmission. The abstract quantifies no figure for shorter/typical visits.","excerpt":"\"The calculated risk for visitors staying longer than 4 months during peak transmission was 0.5% per visit.\"\n","source_date":"2009-12-16","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413175320/https://pubmed.ncbi.nlm.nih.gov/20015392/","calculation_notes":"Massad/Behrens model transmission risk for non-immune travelers to the Amazon region of Brazil (P. falciparum). The only per-visit figure the abstract quantifies is 0.5% per visit for stays longer than 4 months during peak transmission — a long-stay figure, not a 2-week figure, and for a South American setting rather than Sub-Saharan Africa. It is cited here as evidence that even modeled, compliant-free per-visit malaria risk for a long endemic-region stay is on the order of a fraction of one percent, and that duration and season dominate. It does NOT directly substantiate this entry's Sub-Saharan Africa per-trip anchors, which remain order-of-magnitude extrapolations; the species (predominantly P. falciparum in SSA vs the Amazon P. vivax/falciparum mix) and setting differ, so no numeric transfer is made from this paper to the headline.\n","independence_note":"Independent of CDC Yellow Book and WHO — uses mathematical transmission modelling rather than surveillance data. Note the setting/species mismatch flagged in calculation_notes: this paper models the Amazon region of Brazil, not Sub-Saharan Africa.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (per flight)","lifetime_us_adult":7.3e-8},{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Mosquito-borne disease death (lifetime, global average)","lifetime_us_adult":0.00525}],"regional_breakdown":[{"region":"Sub-Saharan Africa, rural, no prophylaxis, 1 month wet season","probability":0.03,"notes":"Order-of-magnitude estimate from travel-medicine literature. West Africa without prevention is the highest-risk itinerary a civilian traveler can take."},{"region":"Sub-Saharan Africa, 2 weeks, compliant prophylaxis + DEET + screens","probability":0.0001,"notes":"The headline reference case. Prophylaxis is not 100% protective but drops risk roughly 20×."},{"region":"South Asia (India, Pakistan, Bangladesh), urban, with prophylaxis","probability":0.000001,"notes":"Mostly P. vivax; urban transmission is low. 4% of US imported cases come from Asia, across a vastly larger traveler volume than Africa."},{"region":"SE Asia urban (Bangkok, Kuala Lumpur, Singapore, Ho Chi Minh City)","probability":1e-8,"notes":"Essentially zero. Major SE Asian cities are malaria-free; risk is limited to specific forested border zones."},{"region":"Caribbean and Mexican tourist zones","probability":1e-8,"notes":"Essentially zero in standard resort areas. Residual risk in parts of rural Haiti and a few Mexican states only."}],"personal_factor_multipliers":[{"factor":"Chemoprophylaxis taken as directed (atovaquone-proguanil, doxycycline, or mefloquine)","multiplier":0.05,"notes":"CDC: 95% of US malaria patients did not take appropriate prevention. Roughly a 20× risk reduction."},{"factor":"Prophylaxis missed doses or stopped before the end of the post-travel course","multiplier":3,"notes":"Post-travel course matters: P. falciparum has a ~1-month incubation, vivax/ovale can relapse months later."},{"factor":"Long-stay expat, year-round exposure in endemic area","multiplier":50,"notes":"Duration of exposure is the single biggest driver. A year in rural SSA without prevention is percent-level annual risk."},{"factor":"DEET + long-lasting insecticidal net (LLIN) + screened accommodation, on top of prophylaxis","multiplier":0.3,"notes":"Mosquito avoidance and prophylaxis are complementary; each adds roughly the same order of protection alone."},{"factor":"Visiting friends and relatives (VFR) vs leisure tourist","multiplier":5,"notes":"VFR travelers account for 76% of US civilian malaria cases per CDC Yellow Book, driven by lower prophylaxis uptake and longer stays in rural areas."}],"short_label":"Malaria (travel)","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry is deliberately scoped per trip, not per lifetime, because travel malaria risk is dominated by destination and duration rather than by who you are. The headline number — ~1 in 10,000 for a 2-week Sub-Saharan Africa trip with compliant prophylaxis — is an order-of-magnitude estimate, not a precise rate. True attack rates vary by country, rural vs urban setting, season, local transmission intensity, drug resistance patterns, and prophylaxis regimen. Off-prophylaxis risk in wet-season rural West Africa is percent-level per month and genuinely dangerous; risk in tourist Mexico or urban SE Asia is effectively zero. P. falciparum, dominant in SSA, has a case-fatality rate of roughly 5–10% untreated and ~0.5% with prompt treatment in a wealthy-country hospital; diagnostic delay in non-endemic emergency rooms is a documented problem, so post-travel fever within the weeks or months after an endemic trip should be evaluated for malaria even if symptoms look flu-like. This entry covers probability of infection; the sibling entry `mosquito-borne-disease` covers mortality aggregated across all mosquito-borne diseases globally.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small stylized pill capsule resting beside a folded paper passport silhouette on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/malaria-travel","api_url":"https://likelier.app/api/fears/malaria-travel.json"},{"slug":"shared-drink-backwash-infection","question":"What are the odds of catching a meaningful infection from sharing a drink bottle or cup?","quick_answer":"The lifetime risk of a meaningful infection from sharing a drink is under 1 in 10,000 — CDC calls it 'so small it is undetectable,' with no documented meningococcal case ever traced to drink-sharing. The instinct runs deep, but only about 0.4% of students even carry that pathogen in their saliva.\n","category":"health","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The instinct that sharing a bottle is a disease vector runs deep — reinforced by the knowledge that saliva contains bacteria and viruses, that backwash is visibly real, and that oral herpes, mononucleosis, and meningitis are all popularly associated with saliva contact. The common framing is: the first person to drink from the bottle leaves pathogens behind, and the second person who pours from it or drinks after carries those pathogens away. Meningitis among university students is particularly salient in this mental model, as is the recurring advice to \"not share drinks\" from school health campaigns. Actual estimates of per-event transmission probability are not widely available to the public, so this is flagged as intuition.\n","kind":"intuition"},"native":{"display":"~0.4% of university students carry Neisseria meningitidis in their saliva (vs. 32% nasopharyngeal carriage)","numerator":4,"denominator":1000,"unit":"students with meningococcal carriage in saliva swabs","population":"UK university students (Orr et al. 2003, Emerging Infectious Diseases, n=258)"},"normalized":{"lifetime_us_adult":0.0001,"display":"< 1 in 10,000 lifetime (risk 'so small it is undetectable' per CDC, no case-control association for drink-sharing and meningococcal disease)","log_value":-4,"assumptions":"No study has directly quantified per-event transmission probability for any meaningful infection via shared cup or bottle. The two strongest anchors are: (1) Orr et al. 2003 (Emerging Infectious Diseases / CDC), which tested saliva carriage of meningococcus — the pathogen most commonly cited in \"don't share drinks\" campaigns — and found carriage in only 0.4% of students' saliva swabs vs. 32% in nasopharyngeal swabs, and which also reviewed a case-control study among university students that \"found no association between meningococcal acquisition and sharing of glasses or cigarettes\"; (2) Manangan et al. 1998 (CDC, AJIC), which reviewed decades of surveillance around the common communion cup scenario (hundreds of thousands of people sharing a cup weekly for years) and concluded \"no documented transmission of any infectious disease has ever been traced to the use of a common communion cup\" and that \"the risk is so small that it is undetectable.\" The 0.0001 lifetime estimate is a derived upper bound consistent with zero detected transmissions across millions of observed cup-sharing events. For the other candidate pathogens: HSV-1 is already present in ~63.5% of US adults (iScience 2024 meta-analysis), making most participants in any sharing event immune; for the ~36% who are seronegative, cup sharing is far less efficient than direct mucosal contact for primary transmission. EBV/mononucleosis is already present in ~95% of adults globally (StatPearls) and is documented to transmit primarily via kissing or sexual contact at university age; household secondary attack rates are low. Strep throat transmission controlled human infection studies found no evidence of fomite transmission (PMC 2024). The 1 in 10,000 lifetime figure represents the upper bound on meaningful clinical infection (hospitalisation, lasting pathogen acquisition) specifically attributable to drink-sharing as a route, as distinct from other household contact routes that would occur anyway.\n","uncertainty":{"low":0,"high":0.001},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3033078/","title":"Saliva and Meningococcal Transmission","publisher":"Emerging Infectious Diseases (CDC journal) — Orr HJ, Gray SJ, Macdonald M, Stuart JM","source_type":"govt_report","statistic":"Meningococcal saliva carriage: 0.4% (1 of 258 students) vs. nasopharyngeal carriage 32.2%; case-control study among university students found no association between meningococcal acquisition and sharing of glasses or cigarettes","excerpt":"\"Low prevalence of carriage in saliva swabs (one swab [0.4%]) suggests that low levels of salivary contact are unlikely to transmit meningococci. On the basis of this evidence, we propose that guidelines for public health management of meningococcal disease should not include low-level salivary contact (e.g., sharing drinks) with a case-patient as an indication for chemoprophylaxis.\"\n","source_date":"2003-10-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505063257/https://pmc.ncbi.nlm.nih.gov/articles/PMC3033078/","calculation_notes":"Orr et al. is the direct empirical test of saliva as a meningococcal transmission route — the specific pathogen most feared in the \"don't share drinks\" narrative. The result is striking: despite 32% nasopharyngeal carriage, only 0.4% of saliva swabs were positive, meaning the organism almost never reaches the saliva in sufficient quantity to transmit. The case-control data (no association with drink-sharing and disease acquisition) adds epidemiological confirmation on top of the biological finding. The authors explicitly recommend removing drink-sharing from the chemoprophylaxis eligibility criteria — a public-health-facing conclusion that directly addresses the fear motivating this entry. Used as the primary evidence anchor for the meningococcal component of the native stat.\n","independence_note":"Independent UK epidemiological study published in CDC's own journal; distinct from the AJIC communion cup review below and the HSV-1 and EBV sources.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/9795685/","title":"Risk of infectious disease transmission from a common communion cup","publisher":"American Journal of Infection Control (Manangan LP, Sehulster LM, Chiarello L, Simonds DN, Jarvis WR — CDC Hospital Infections Program)","source_type":"govt_report","statistic":"No documented transmission of any infectious disease traced to use of a common communion cup; CDC conclusion: risk 'so small that it is undetectable'","excerpt":"\"Although no documented transmission of any infectious disease has ever been traced to the use of a common communion cup, a theoretical risk of transmitting infectious diseases exists. The consensus of the CDC is that such risk is so small that it is undetectable.\"\n","source_date":"1998-10-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505063345/https://pubmed.ncbi.nlm.nih.gov/9795685/","calculation_notes":"The communion cup scenario is the highest-frequency shared-cup event in the epidemiological literature: hundreds of thousands of participants, weekly exposure across decades, a large unselected population. Despite this scale of observation, the CDC-authored AJIC review identified zero documented disease transmissions. If we conservatively assume 100 million person-cup-sharing events observed and zero infections detected, the per-event upper bound on meaningful infection probability is on the order of 1 in 10 million. The lifetime 0.0001 estimate allows for substantially more risk than this (treating it as a conservative upper bound), since the communion cup scenario may involve less backwash saliva exchange than a shared personal bottle. Used as the primary anchor for the \"undetectable signal\" framing.\n","independence_note":"Independent CDC epidemiological review, distinct from the Orr meningococcal study above and the virology sources below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11367537/","title":"Epidemiology of herpes simplex virus type 1 in the United States: Systematic review, meta-analyses, and meta-regressions","publisher":"iScience (Cell Press)","source_type":"peer_reviewed","statistic":"Pooled mean HSV-1 seroprevalence among US adults: 63.5% (95% CI 61.3–65.7%); seroprevalence declining ~1% per year; ~36% of US adults are seronegative and theoretically susceptible to primary HSV-1","excerpt":"\"The pooled mean HSV-1 seroprevalence was 63.5% (95% CI: 61.3–65.7) among general-population adults... Seroprevalence declined by 0.99-fold (95% CI: 0.99–0.99) per year.\"\n","source_date":"2024-08-05","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260225081021/https://pmc.ncbi.nlm.nih.gov/articles/PMC11367537/","calculation_notes":"The 63.5% US adult seroprevalence figure establishes that the majority of participants in any drink-sharing event already carry HSV-1 and are immune to primary acquisition. For the ~36% who are seronegative, HSV-1 can in principle be acquired via contact with infected saliva; however, transmission efficiency via fomite (a shared bottle) is substantially lower than via direct mucosal contact, and no study has quantified per-event cup-sharing transmission probability. The high background immunity means the expected proportion of sharing partners who are both (a) HSV-1 positive and (b) actively shedding virus in detectable quantities and (c) sharing a bottle with a seronegative partner is low. Used to establish the HSV-1 context and immunity floor.\n","independence_note":"Independent of the meningococcal CDC sources; distinct pathogen, distinct methodology (meta-analysis of seroprevalence studies).\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK559285/","title":"Epstein-Barr Virus — StatPearls","publisher":"StatPearls Publishing / NCBI Bookshelf","source_type":"reputable_reference","statistic":"Global adult EBV seroprevalence: ~95%; US ages 18–19: 82.9% already infected; EBV 'is not considered a highly contagious disease'; kissing is the major route of primary EBV transmission among adolescents and young adults","excerpt":"\"Nearly 95% of the world's adult population has EBV exposure... The transmission of the Epstein Barr virus occurs in several ways, such as deep kissing or food-sharing... EBV is not considered a highly contagious disease.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505063417/https://www.ncbi.nlm.nih.gov/books/NBK559285/","calculation_notes":"The ~95% global adult seroprevalence makes EBV acquisition via drink-sharing negligible for the vast majority of adults — only the ~5% globally (higher among younger US adults) who are seronegative are at any risk of primary EBV infection. Even for this seronegative minority, kissing is the documented primary transmission route (Crawford et al. 2006, CID: 46% seroconversion over 3 university years, with penetrative sexual intercourse the primary identified risk factor), not cup-sharing. StatPearls explicitly classifies EBV as \"not highly contagious.\" Used to establish that mononucleosis from shared drinks is an extremely unlikely scenario given background immunity and preferential transmission routes.\n","independence_note":"Independent reference source (NIH/NCBI StatPearls), distinct from the meningococcal CDC studies and the HSV-1 meta-analysis.\n"}],"comparison_anchors":[{"label":"Infection from shared toilet seat (lifetime)","lifetime_us_adult":0.0001},{"label":"Infection from shared towel (lifetime)","lifetime_us_adult":0.002},{"label":"Infection from sharing food with a child — CMV (seronegative parent, annual)","lifetime_us_adult":0.24}],"short_label":"Infection from shared drink","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses the risk of a meaningful infection — one involving a pathogen with lasting clinical consequences — from sharing a drink bottle or cup where one person's saliva enters the liquid. The risk of a self-limiting cold or mild flu from sharing a bottle with someone who is actively symptomatic is not measured here and is not meaningfully separated from the general household droplet and contact-transmission risk that exists regardless of whether you shared a bottle. All the pathogens most commonly named in this concern (meningococcus, HSV-1, EBV, strep) have low or no documented transmission probability via this route for distinct mechanistic reasons: meningococcus almost never reaches saliva in detectable concentrations (0.4% saliva carriage); HSV-1 is already carried by 63.5% of US adults; EBV is already carried by ~95% of global adults and transmits mainly via intimate contact; strep is transmitted primarily via respiratory droplets, not fomites. The CDC communion cup analysis, which represents decades of surveillance across millions of cup-sharing events, found the risk \"so small it is undetectable.\" None of this means backwash is aesthetically neutral; it means the infection biology does not match the cultural fear. The caveat that does apply: sharing a bottle with someone who is visibly symptomatic (actively coughing, with an obvious cold sore, or known to have strep throat) raises even the already-low per-event risk meaningfully — the \"don't share drinks when sick\" advice is rational and is not what this entry challenges.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A single water bottle with two drinking straws beside it, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/shared-drink-backwash-infection","api_url":"https://likelier.app/api/fears/shared-drink-backwash-infection.json"},{"slug":"hurricane-death","question":"What are the odds of being killed by a hurricane (tropical cyclone)?","quick_answer":"Averaged across the world's adults, the lifetime odds of being killed by a hurricane are about 1 in 8,900, though this global figure is dominated by rare megastorms and is not a personal estimate. Nearly 30% of US adults say they fear a devastating hurricane.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"\"Hurricane\" in the Atlantic and East Pacific, \"typhoon\" in the West Pacific, and \"cyclone\" in the South Pacific and Indian Ocean all refer to the same weather system — a tropical cyclone. Fear of these storms is strongly geographic, and no widely cited national survey isolates \"fear of being killed by a hurricane\" from general severe-weather or climate anxiety, so we mark the perceived side as editorial intuition. Anecdotally, coastal residents of the US Gulf and Southeast, the Philippines, and the Bay of Bengal tend to carry an explicit prior shaped by the last storm they lived through; inland readers and high-latitude readers tend to treat the hazard as televised rather than real.\n","rough_estimate":"29.8% of US adults report being afraid or very afraid of a devastating hurricane (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~15,700 tropical cyclone deaths per year (50-year global average, WMO/CRED 1970-2021)","numerator":15700,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.000112,"display":"1 in ~8,900 lifetime (global adult average)","log_value":-3.95,"assumptions":"Uses the WMO figure of 779,324 tropical cyclone deaths over the 50-year 1970-2021 window — an average of 43 deaths per day, or ~15,700 per year globally — as the headline native figure. Divided by a global population of ~8 billion and compounded over 60 adult life-years gives ~1.1e-4, which we round to an order-of-magnitude ~1 in 8,900. The window matters enormously: the 1970 Bhola cyclone alone (~300,000-500,000 deaths in what is now Bangladesh) and Cyclone Nargis in 2008 (~138,000 deaths in Myanmar) together account for a large fraction of the 50-year total, and the smoothed post-Nargis average is several times lower. The uncertainty band below reflects window choice and the dominance of rare megaevents, not sampling noise. This is an \"average global adult\" figure and is not a useful personal estimate for any individual — see the regional breakdown and caveats.\n","uncertainty":{"low":0.00003,"high":0.0003},"scope":"global_adult_lifetime"},"sources":[{"url":"https://wmo.int/topics/tropical-cyclone","title":"Tropical Cyclone — Topic Page","publisher":"World Meteorological Organization (WMO)","source_type":"govt_report","statistic":"1,945 disasters attributed to tropical cyclones over the 50 years 1970-2021, killing 779,324 people — an average of 43 deaths per day, or roughly 15,700 per year.","excerpt":"\"Over the past 50 years, 1,945 disasters have been attributed to tropical cyclones, which killed 779,324 people and caused US$ 1.4 trillion in economic losses – an average of 43 deaths and US$ 78 million in damages daily.\"\n","source_date":"2023-05-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260415000000/https://wmo.int/topics/tropical-cyclone","calculation_notes":"WMO's 50-year total of 779,324 deaths / 50 ≈ 15,587 deaths per year, which we round to ~15,700. Annual per-capita risk ≈ 15,700 / 8,000,000,000 ≈ 1.96e-6; compounded over 60 adult years ≈ 1.18e-4 ≈ 1 in 8,500, which we round to ~1 in 8,900 to stay conservative against the fact that most recent decades have had lower death tolls than the 50-year average because the average is pulled hard by the 1970 Bhola cyclone and 2008 Nargis. The uncertainty band brackets the post-1990 smoothed average (~5,000/year, giving ~1 in 27,000) on the optimistic side and a window centered on 1970-2008 (~25,000/year, giving ~1 in 5,300) on the pessimistic side.\n","independence_note":"WMO's topic page and WMO's 2023 press release (cited below) both draw from the same underlying WMO/CRED Atlas of Mortality and Economic Losses from Weather, Climate and Water Extremes, so treat them as a single authoritative source chain with two presentation layers rather than two fully independent estimates.\n"},{"url":"https://wmo.int/news/media-centre/economic-costs-of-weather-related-disasters-soars-early-warnings-save-lives","title":"Economic costs of weather-related disasters soars but early warnings save lives","publisher":"World Meteorological Organization (WMO) press release, drawing on WMO/CRED EM-DAT data","source_type":"govt_report","statistic":"Over 1970-2021, Bangladesh recorded 520,758 deaths from 281 weather/climate/water-related events (the highest national toll in Asia, overwhelmingly driven by tropical cyclones); Cyclone Nargis in Myanmar (2008) killed 138,366; total weather-related disasters caused >2 million deaths and US$4.3 trillion in economic losses globally.","excerpt":"\"Extreme weather, climate and water-related events caused 11 778 reported disasters between 1970 and 2021, with just over 2 million deaths and US$ 4.3 trillion in economic losses. ... Bangladesh has the highest death toll in Asia with 520 758 deaths due to 281 events. ... Tropical cyclone Nargis in 2008 led to 138 366 deaths.\"\n","source_date":"2023-05-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173134/https://wmo.int/news/media-centre/economic-costs-of-weather-related-disasters-soars-early-warnings-save-lives","calculation_notes":"Used as the geographic breakdown anchor. Bangladesh's 520,758 figure spans all weather-climate-water hazards, not only cyclones, but historically the Bay of Bengal cyclone record (1970 Bhola, 1991 Bangladesh, 2008 Nargis across the bay in Myanmar) dominates that toll. The regional_breakdown row for Bangladesh / Bay of Bengal coastal populations uses ~0.003 lifetime as an order-of-magnitude estimate consistent with a 520,758 cumulative toll, a coastal population of ~30-50 million, and a 50-year window — heavily front-loaded in 1970 and declining sharply since.\n","independence_note":"Same WMO/CRED underlying dataset as the topic page above. Cited separately because it adds the country-level breakdown (Bangladesh, Myanmar) and the early-warning policy framing, which the topic page does not.\n"},{"url":"https://www.thelancet.com/journals/lanplh/article/PIIS2542-5196(23)00143-2/fulltext","title":"Global short-term mortality risk and burden associated with tropical cyclones from 1980 to 2019","publisher":"The Lancet Planetary Health (2023)","source_type":"peer_reviewed","statistic":"~97,430 excess deaths per decade globally from tropical cyclone exposure (1980-2019), ~9,700/year; 6% mortality increase in 2 weeks post-exposure","excerpt":"\"Tropical cyclone exposure was associated with a 6% increase in mortality in the first 2 weeks following exposure.\"\n","source_date":"2023-08-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20241008065054/https://www.thelancet.com/journals/lanplh/article/PIIS2542-5196(23)00143-2/fulltext","calculation_notes":"Uses excess-mortality time-series methodology across 341 locations in 14 countries — genuinely independent of the WMO/CRED disaster-database approach. The ~9,700/year excess-death figure is lower than WMO/CRED direct counts because it captures a different signal (short-term excess mortality vs. reported disaster deaths).\n","independence_note":"Fully independent of WMO/CRED — different methodology (epidemiological time-series vs disaster database), different authorship, different data sources.\n"},{"url":"https://www.nhc.noaa.gov/data/tcr/AL092024_Helene.pdf","title":"National Hurricane Center Tropical Cyclone Report — Hurricane Helene (AL092024)","publisher":"NOAA National Hurricane Center","source_type":"govt_report","statistic":"Hurricane Helene caused at least 250 fatalities in the United States — 175 direct, 72 indirect, and 3 of unknown cause — making it the deadliest hurricane in the contiguous U.S. since Katrina (2005). Of the 175 direct deaths, 94 were from freshwater (inland rainfall/flash/river) flooding, the single largest category, clustered far inland in western North Carolina, eastern Tennessee and northwestern South Carolina; 65 were wind deaths.","excerpt":"\"Helene is responsible for at least 250 fatalities in the United States (including at least 175 direct deaths), making it the deadliest hurricane in the contiguous U.S. since Katrina in 2005. ... In addition to the 175 direct and 72 indirect fatalities, 3 people died of unknown causes related to Helene, which brings the total number of Helene-related fatalities in the United States to 250. ... The freshwater flooding fatalities were clustered in western North Carolina, eastern Tennessee and northwestern South Carolina.\"\n","source_date":"2025-03-19","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260712104359/https://www.nhc.noaa.gov/data/tcr/AL092024_Helene.pdf","calculation_notes":"US-scope single-season evidence, not used to revise the global headline (which is a 50-year global average on the global_adult_lifetime axis). Cited to document the 2024 Atlantic season's exceptional US toll and to strengthen two mechanisms already in this entry: (1) the direct-vs-indirect split — NHC counts 175 direct (drowning, surge, wind, structural collapse) vs 72 indirect (heart attacks, vehicle accidents, post-storm cleanup) per the Rappaport (2014) classification; (2) the inland-flooding mechanism — 94 of the 175 direct deaths (~54%) were from freshwater rainfall/river/ flash flooding hundreds of miles inland in the southern Appalachians, outpacing both wind (65) and coastal storm surge, overturning the common assumption that surge is always the dominant US hurricane killer. The Helene direct-death composition is a cause breakdown, not a per-person relative-risk multiplier, so it is not converted into a personal_factor_multiplier.\n","independence_note":"Independent primary US-agency post-event analysis (NOAA NHC), distinct from the WMO/CRED global disaster database and the Lancet excess-mortality study. US-scope only — does not bear on the global headline figure.\n"}],"comparison_anchors":[{"label":"Death in an earthquake (global adult lifetime)","lifetime_us_adult":0.000263},{"label":"Death in a tsunami (global adult lifetime)","lifetime_us_adult":0.00001},{"label":"Death in a tornado (US adult lifetime, national average)","lifetime_us_adult":0.0000124},{"label":"Death in a plane crash (US adult lifetime, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (US lifetime)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.00011},{"region":"Bangladesh / Bay of Bengal coast","probability":0.003,"notes":"Most catastrophic cyclones historically, densely populated low-lying deltas; the 1970 Bhola cyclone alone is estimated to have killed 300,000-500,000 people."},{"region":"Philippines / coastal Southeast Asia","probability":0.0005,"notes":"High storm frequency (Haiyan 2013 ~7,300 deaths) and exposed coastal populations."},{"region":"US Gulf Coast / Southeast","probability":0.00005,"notes":"Katrina (~1,800) and Maria in Puerto Rico (~3,000) are the dominant modern events; strong early-warning and evacuation infrastructure keeps the per-capita rate low."},{"region":"Inland continents / high latitudes","probability":1e-7,"notes":"Essentially zero — no tropical cyclone exposure."}],"personal_factor_multipliers":[{"factor":"resident of Caribbean / Central America coast","multiplier":50,"notes":"CRED/EM-DAT: small island nations and low-capacity coastal states account for the vast majority of hurricane fatalities globally"},{"factor":"resident of US Gulf Coast in mobile home","multiplier":10,"notes":"storm-surge flooding and structural vulnerability concentrate fatalities in vulnerable coastal housing"},{"factor":"inland US resident (>100 miles from coast)","multiplier":0.1,"notes":"inland tornado/flooding deaths from remnants are a small fraction of total hurricane mortality"},{"factor":"failure to evacuate a mandatory evacuation zone","multiplier":10,"notes":"Blake & Zelinsky 2018 (NHC official Katrina report) and post-Katrina and Maria mortality analyses: residents who remain in mandatory evacuation zones during Category 3+ landfalls face an order-of-magnitude higher direct mortality risk than those who evacuate; storm surge — the leading cause of direct hurricane death — is largely survivable by evacuees and nearly inescapable for those who shelter in place in surge zones"},{"factor":"age 65+ in coastal surge zone","multiplier":5,"notes":"Blake & Zelinsky 2018 NHC report and CDC hurricane mortality reviews: adults aged 65+ account for roughly 60% of direct hurricane deaths in the US, driven by higher rates of functional limitation, lower evacuation compliance, greater dependence on medical infrastructure, and higher baseline vulnerability to storm-surge drowning and heat/cold exposure during power outages"}],"short_label":"Hurricane","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker, not a personal estimate. Tropical cyclone mortality is heavily concentrated in a small set of low-lying, densely populated coastlines: the Bay of Bengal (Bangladesh, eastern India, Myanmar), the Philippines and coastal Southeast Asia, and the Caribbean. Within those regions, the per-capita lifetime risk is one to two orders of magnitude higher than the global average. Conversely, inland and high-latitude populations have essentially no direct exposure. The US Gulf and Southeast, despite receiving heavy media coverage, contribute only a small fraction of global cyclone deaths because of early-warning systems, evacuation infrastructure, and building codes — the dominant exception being storm-surge events where evacuation is incomplete (Katrina) and post-storm cascading failures where the official death toll is contested (Maria). The long-run average is also dominated by pre-1990 megaevents; a 30-year window centered on 2000 gives a considerably lower annual number than the 50-year WMO window, which is why the uncertainty band is wide. The 2024 Atlantic season is a reminder that the modern US death toll is not monotonically falling: Hurricane Helene killed at least 250 people in the United States (175 direct, 72 indirect, 3 of unknown cause per the NHC's Tropical Cyclone Report), the deadliest contiguous-US hurricane since Katrina. Notably, most of Helene's direct deaths — 94 of 175 — came from freshwater rainfall and river/flash flooding hundreds of miles inland in the southern Appalachians, more than wind (65) and more than coastal storm surge, which cuts against the common assumption that storm surge is always the dominant US hurricane killer and shows that inland exposure is not zero where remnants stall over terrain.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized spiral shape representing a tropical cyclone from above, rendered in muted blues and grey against a pale sky, flat vector illustration."},"canonical_url":"https://likelier.app/hurricane-death","api_url":"https://likelier.app/api/fears/hurricane-death.json"},{"slug":"chagas-disease","question":"What are the odds of contracting Chagas disease in Latin America?","quick_answer":"Averaged across the global adult population, the lifetime odds of dying from Chagas disease are about 1 in 8,500. The risk is highly concentrated, though — in endemic rural Latin America it runs dozens of times higher, while elsewhere it is negligible.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Outside specialist medicine, Chagas disease barely registers as a named threat. Most people in Latin America treat the triatomine \"kissing bug\" as a minor nuisance rather than a disease vector, and the 10-to-20-year lag between infection and visible cardiac damage means the connection between a childhood insect bite and adult heart failure is rarely made. Global awareness is even lower: in wealthy countries, Chagas is categorized as a \"neglected tropical disease,\" a bureaucratic label that accurately predicts how little attention it receives.\n","kind":"intuition"},"native":{"display":"~10,000 deaths per year, global population","numerator":10000,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.000118,"display":"~1 in 8,500 lifetime (global adult average)","log_value":-3.93,"assumptions":"WHO estimates more than 10,000 Chagas-attributable deaths per year globally, almost entirely in Latin America. Global adult population ~5 billion. Annual mortality rate: 10,000 / 5,000,000,000 = 0.000002 per adult per year. Compounded over 59 years (US adult remaining-life horizon from age 18): 1 - (1 - 0.000002)^59 ≈ 0.000118, or roughly 1 in 8,500. This is a global average; the risk is highly concentrated. In endemic rural Latin America (~100 million people at high risk per WHO), with ~9,000 deaths attributable to that population, the subgroup annual rate is ~9,000 / 100,000,000 ≈ 0.00009, compounding to ~1 in 190 over a lifetime — roughly 45× the global figure. The lifetime risk of ever becoming infected (not dying) in endemic areas is substantially higher: WHO estimates about 8 million currently infected from a population of more than 100 million at risk.\n","uncertainty":{"low":0.000075,"high":0.0003},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/chagas-disease","title":"Chagas disease (also known as American trypanosomiasis) — WHO Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"~8 million people infected worldwide; more than 10,000 deaths per year; more than 100 million at risk","excerpt":"\"About 8 million people worldwide, mostly in Latin America, are estimated to be infected with Trypanosoma cruzi … more than 100 million people are considered at risk of infection … leading to more than 10 000 deaths every year.\"\n","source_date":"2026-04-08","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250430004105/https://www.who.int/news-room/fact-sheets/detail/chagas-disease","calculation_notes":"The 10,000+ annual deaths figure is the WHO headline for the mortality denominator. Divided by 5 billion global adults gives an annual rate of 2.0 × 10⁻⁶. The 100+ million at-risk figure is used to contextualize the subgroup lifetime risk calculation in the assumptions field.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S2667193X24002084","title":"The epidemiology of Chagas disease in the Americas","publisher":"The Lancet Regional Health – Americas","source_type":"peer_reviewed","statistic":"~8 million infected in the Americas; Bolivia, Paraguay, and the Chaco region carry the highest transmission rates; cardiac involvement in 20–30% of chronically infected individuals","excerpt":"\"An estimated 8 million people are infected with Trypanosoma cruzi in the Americas. The highest-burden countries are Bolivia, Argentina, and Brazil. Cardiac Chagas disease develops in approximately 20–30% of chronically infected individuals, typically emerging 10–20 years after initial infection, and is the leading cause of infectious cardiomyopathy in the region.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-24","calculation_notes":"The 20–30% cardiac involvement rate informs the severity framing: most people infected with T. cruzi do not die quickly — the disease progresses silently, which is why the gap between infection prevalence and recognized mortality is so wide. The endemic-country concentration supports the subgroup lifetime-risk estimate in the assumptions.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9414802/","title":"Global, Regional, and National Trends of Chagas Disease from 1990 to 2019 (GBD Study)","publisher":"PLOS Neglected Tropical Diseases / PMC","source_type":"primary_study","statistic":"GBD 2019: Chagas responsible for 9,487 deaths globally; 275,377 DALYs; age-standardized death rate 0.12 per 100,000","excerpt":"\"Chagas disease accounted for 9,487 deaths in 2019, compared to 11,235 in 1990, representing a 15.6% decrease. Total DALYs were 275,377 (95% UI = 184,453–459,354) in 2019. The age-standardized DALY rate was 3.34 (95% UI 2.25–5.57) per 100,000 population in 2019, a substantial decrease from 8.51 in 1990.\"\n","source_date":"2022-08-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260102043631/https://pmc.ncbi.nlm.nih.gov/articles/PMC9414802/","calculation_notes":"GBD estimates (9,487/yr) are broadly consistent with WHO's figure of more than 10,000/yr. The WHO figure is used as the primary numerator since it aligns with surveillance data; the GBD figure cross-validates the order of magnitude (annual rate 9,487 / 5B ≈ 1.90 × 10⁻⁶, lifetime ≈ 0.000112).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3731239/","title":"Epidemiology of and Impact of Insecticide Spraying on Chagas Disease in Communities in the Bolivian Chaco","publisher":"PLOS Neglected Tropical Diseases","source_type":"peer_reviewed","statistic":"Overall T. cruzi seroprevalence of 51.7% in surveyed Bolivian Chaco communities; age-specific prevalence 19.8% (ages 2-15), 72.7% (ages 15-30), and 97.1% (over 30)","excerpt":"\"The T. cruzi seroprevalence among serosurvey participants was 51.7%. The prevalence increased from 19.8% among children 2–15 years old to 72.7% among those 15–30 years old and 97.1% among participants older than 30 years.\"\n","source_date":"2013-08-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250328063525/https://pmc.ncbi.nlm.nih.gov/articles/PMC3731239/","calculation_notes":"Provides a concrete, peer-reviewed seroprevalence figure for the specific high-burden Bolivian Chaco communities referenced in the body text. This is an infection-prevalence figure, distinct from the mortality-based headline lifetime risk; it grounds the geographic-concentration claim without altering the WHO/GBD-derived mortality numerator.\n","independence_note":"Independent academic field survey (Samuels et al., 2013) of specific Bolivian Chaco communities, methodologically distinct from the population-level WHO and GBD mortality estimates used for the headline figure.\n"}],"comparison_anchors":[{"label":"Malaria death (global adult lifetime)","lifetime_us_adult":0.0042},{"label":"Typhoid death (global adult lifetime)","lifetime_us_adult":0.00129},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"short_label":"Chagas disease","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The global average (1 in 8,500 lifetime) obscures extreme geographic concentration: rural Bolivia, the Gran Chaco, and parts of Central America have infection prevalence 10–40× higher than the regional mean. Vector control programs have reduced transmission substantially in the Southern Cone since the 1990s, so people born after 1995 in many countries face lower transmission risk than older adults. Conversely, urbanization, climate change, and sylvatic spillover are expanding the triatomine range into previously unaffected areas. Blood-transfusion and vertical (mother-to-child) transmission now account for a growing share of new cases outside traditional endemic zones, including in North America and Europe via migration.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"Flat vector illustration of a triatomine bug silhouette on a cracked earth background, muted ochre and slate tones."},"canonical_url":"https://likelier.app/chagas-disease","api_url":"https://likelier.app/api/fears/chagas-disease.json"},{"slug":"wild-berry-fox-tapeworm","question":"What are the odds of getting a fatal fox tapeworm infection from eating wild berries?","quick_answer":"For an adult in an endemic central European region, the lifetime odds of a fatal fox-tapeworm infection are roughly 1 in 8,500, and the berry-eating route specifically is not even statistically confirmed. Many regular foragers imagine something closer to 1 in 1,000; US baseline risk is lower still.\n","category":"health","tags":["food","travel"],"no_reliable_estimate":false,"perceived":{"description":"The fear circulates widely in central Europe: a single unwashed raspberry or blueberry picked at fox-height could deliver Echinococcus multilocularis eggs that silently grow a parasitic tumor in the liver for years, often undetected until the disease is advanced. Health warnings routinely name wild berries and mushrooms as vectors, and the word \"fox tapeworm\" carries an outsized dread relative to how rare the actual disease is.\n","rough_estimate":"many central Europeans estimate at least 1 in 1,000 risk for regular foragers","kind":"intuition"},"native":{"display":"~0.20 per 100,000 per year (southwest Germany, 2012–2019)","numerator":2,"denominator":1000000,"unit":"per year","population":"adults in an endemic region of central Europe (southwest Germany)"},"normalized":{"lifetime_us_adult":0.000118,"display":"~1 in 8,500 lifetime (endemic central European adult)","log_value":-3.928,"assumptions":"Uses the most recent peer-reviewed incidence figure for an endemic European region: 0.20 cases per 100,000 per year in southwest Germany (2012–2019), from Graeter et al. (Infection 2020). That annual rate of 2.0 × 10^-6 per person, compounded over 59 adult years, yields 1 − (1 − 0.000002)^59 ≈ 0.000118, or about 1 in 8,500. This is the risk for an adult living in an actively endemic European region; the EU mean from the KNOW-PATH systematic review (1997–2023) is 0.063/100,000, yielding a lower-bound lifetime estimate of about 1 in 27,000. The Doubs département in France and Swiss Jura regularly record rates of 1.0–1.4/100,000, giving an upper-bound lifetime estimate of about 1 in 1,200. US adults face meaningfully lower baseline risk: Echinococcus multilocularis has limited US distribution (north-central states, Alaska), human cases are not nationally reportable, and recorded cases are sporadic — reliable US incidence data does not exist.\n","uncertainty":{"low":0.0000372,"high":0.000826},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7674360/","title":"Spatial distribution and incidence trend of human alveolar echinococcosis in southwest Germany: increased incidence and urbanization of the disease?","publisher":"Infection (Springer); Graeter et al.","source_type":"peer_reviewed","statistic":"Annual AE incidence per 100,000: 0.12 (2004–2011), rising to 0.20 (2012–2019) in southwest Germany","excerpt":"\"The data from our regional referral center for AE in southwest Germany suggest rising regional incidence for AE (annual incidence per 100,000 population 2004–2011: 0.12; 2012–2019: 0.20).\"\n","source_date":"2020-11-20","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20250215061249/https://pmc.ncbi.nlm.nih.gov/articles/PMC7674360/","calculation_notes":"Anchor for native rate. 0.20 per 100,000 per year = 2.0 × 10^-6 annual hazard per person. Lifetime probability over 59 adult years: 1 − (1 − 0.000002)^59 = 0.000118 ≈ 1 in 8,475. Chosen as native because it is the most cited recent peer-reviewed figure for a well-defined endemic region with active surveillance.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5531747/","title":"Potential risk factors associated with human alveolar echinococcosis: Systematic review and meta-analysis","publisher":"PLOS Neglected Tropical Diseases; Piarroux et al.","source_type":"peer_reviewed","statistic":"Meta-analysis OR for 'ate unwashed strawberries': 1.39 (95% CI 0.87–2.23, p=0.17, not significant); 'ate wild vegetables and fruit': OR 1.38 (95% CI 0.90–2.10, p=0.14, not significant). Strongest risk factors: dog ownership including dogs that hunt, farming, rural residence in endemic areas.\n","excerpt":"\"Results were not statistically significant for 'ate unwashed strawberries' (OR 1.39; 95% CI 0.87–2.23; p = 0.17), 'ate wild vegetables and fruit' (OR 1.38; 95% CI 0.90–2.10; p = 0.14) and 'ate mushrooms' (OR 0.72; 95% CI 0.38–1.39; p = 0.33).\"\n","source_date":"2017-08-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260503084420/https://pmc.ncbi.nlm.nih.gov/articles/PMC5531747/","calculation_notes":"Documents the epidemiological evidence (or lack thereof) for the berry transmission pathway. Non-significant ORs for berry and wild-produce consumption mean the published literature does not establish berries as a statistically confirmed transmission route. Strong risk factors (dog that hunts game, farmer status) are behaviorally distinct from foraging. Used to frame the gap between the popular fear narrative and the case-control evidence.\n","independence_note":"Systematic review pooling multiple European case-control studies — captures the dominant epidemiological evidence base for AE risk factors.\n"},{"url":"https://www.sciencedirect.com/science/article/abs/pii/S147330992500283X","title":"Unveiling the incidences and trends of alveolar echinococcosis in Europe: a systematic review from the KNOW-PATH project","publisher":"The Lancet Infectious Diseases","source_type":"peer_reviewed","statistic":"Mean European AE incidence 1997–2023: 0.063 per 100,000; 4,207 cases documented; Germany, France, Austria, Switzerland = 68% of total","excerpt":"\"The mean annual incidence from 1997 to 2023 throughout Europe was 0.063 cases per 100,000 people. Historically endemic Austria, France, Germany, and Switzerland accounted for 2,864 (68.08%) of 4,207 cases documented in Europe.\"\n","source_date":"2025-05-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20250805154821/https://www.sciencedirect.com/science/article/abs/pii/S147330992500283X","calculation_notes":"Provides the EU-wide mean annual incidence used as the lower bound of the uncertainty range. 0.063/100,000 = 6.3 × 10^-7 per year. Lifetime over 59y: 1 − (1 − 0.00000063)^59 ≈ 3.72 × 10^-5, i.e. ~1 in 26,900. Also confirms the 2025 state of European surveillance and the ongoing rising trend.\n","independence_note":"Independent systematic review from the KNOW-PATH project; uses ECDC TESSy surveillance data and national registries rather than single-center data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12052437/","title":"Presence of Echinococcus eggs in the environment and food: a review of current data and future prospects","publisher":"Parasitology (Cambridge University Press)","source_type":"peer_reviewed","statistic":"E. multilocularis DNA detected in 5.4% of strawberries and 7.3% of blueberries in European endemic countries (seven-country study); viability of those eggs explicitly unknown.\n","excerpt":"\"E. multilocularis DNA was detected in 1.2% of lettuces, 5.4% of strawberries, and 7.3% of blueberries in European endemic countries… although eggs are assumed to be the source of the DNA detected in these studies, the viability of such eggs is unknown.\"\n","source_date":"2025-04-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260525101133/https://pmc.ncbi.nlm.nih.gov/articles/PMC12052437/","calculation_notes":"Documents the molecular evidence for berry contamination while underscoring the key limitation: DNA detection ≠ viable infectious eggs. Used to characterise the contamination pathway that underlies public concern, not to derive a probability estimate. Supports the framing that wild berries are a plausible but unconfirmed dominant transmission route.\n","independence_note":"Independent review of environmental contamination literature; distinct from the epidemiological case-control literature cited above.\n"},{"url":"https://wwwnc.cdc.gov/eid/article/31/5/24-1608_article","title":"Comprehensive Survival Analysis of Alveolar Echinococcosis Patients, University Hospital Zurich, Zurich, Switzerland, 1973-2022","publisher":"Emerging Infectious Diseases (CDC)","source_type":"peer_reviewed","statistic":"Without adequate treatment, 90% of AE patients died within 10 years of disease onset in the pre-benzimidazole era; modern benzimidazole therapy and surgery have shifted the main survival-limiting factor to non-AE causes of death","excerpt":"\"Without adequate treatment, 90% of AE patients died within 10 years of disease onset.\"\n","source_date":"2025-05-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250604095140/https://wwwnc.cdc.gov/eid/article/31/5/24-1608_article","calculation_notes":"Zurich cohort survival analysis (1973-2022) is the source for the body's closing severity note: greater-than-90% fatality within 10 years for untreated alveolar echinococcosis, and the substantial improvement in outcomes since benzimidazole drug therapy became available. Descriptive context on disease severity only; not used in the probability calculation, which relies on the incidence sources above.\n","independence_note":"Single-center long-term survival cohort, methodologically distinct from the incidence/epidemiology sources above (Graeter, Piarroux, KNOW-PATH, Parasitology review); addresses disease course after diagnosis rather than incidence.\n"}],"comparison_anchors":[{"label":"Fatal anaphylaxis from bee/wasp sting (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Residence in high-incidence endemic region (e.g. Doubs/France, Swiss Jura)","multiplier":10,"notes":"AE incidence varies more than an order of magnitude across Europe. The Doubs département (France) and Swiss Jura regularly record 1.0–1.4 cases per 100,000 per year, vs the EU mean of 0.063/100,000. Living in the highest-incidence European zones implies roughly 10–20× the average EU risk. Source: Graeter et al., Infection 2020; KNOW-PATH systematic review, Lancet Infectious Diseases 2025."},{"factor":"Dog ownership (dog roams freely or hunts in endemic area)","multiplier":4,"notes":"Owning a dog that hunts or roams outdoors in an endemic region is one of the most consistently replicated AE risk factors across European case-control studies. Dogs can act as transport hosts, depositing E. multilocularis eggs shed from foxes onto domestic surfaces. The Piarroux 2017 meta-analysis identifies dog ownership (including dogs that hunt) as among the strongest risk factors for AE, though it does not report a pooled odds ratio specifically for dogs that kill game. Source: Piarroux et al., PLOS NTD 2017."},{"factor":"Immunocompromised status","multiplier":3,"notes":"Immunocompromised patients (transplant recipients, those on long-term corticosteroids, HIV-positive with low CD4 counts) experience faster and more aggressive alveolar echinococcosis progression. While baseline incidence is not higher, once infected the subclinical-to-symptomatic conversion is accelerated. WHO/OIE guidelines flag immunosuppression as a major modifier of disease course and outcomes. Source: WHO Informal Working Group on Echinococcosis (WHO-IWGE), WHO Technical Report Series 2003; Brunetti E et al., Acta Tropica 2010."}],"myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry normalizes to an adult in an actively endemic European region (southwest Germany, 2012–2019). The figure is not a US adult baseline — Echinococcus multilocularis has a restricted US range, human cases are not nationally reportable, and no reliable US incidence series exists. The number also does not distinguish the berry pathway from other routes: case-control meta-analyses find dog ownership and farming to be the statistically robust risk factors, while berry consumption carries a non-significant pooled OR of ~1.4. No_reliable_estimate is set to false because regional European incidence is well-documented — but the fraction attributable specifically to wild berry ingestion cannot be isolated from published data.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A cluster of wild blueberries on low-growing stems against a muted forest floor, flat vector illustration."},"canonical_url":"https://likelier.app/wild-berry-fox-tapeworm","api_url":"https://likelier.app/api/fears/wild-berry-fox-tapeworm.json"},{"slug":"toddler-nicotine-pouch-poisoning","question":"How likely is a young child to be reported to poison control for swallowing a nicotine pouch?","quick_answer":"About 1 in 7,900 US children are reported to poison control for swallowing a nicotine pouch before age 6, and most of those reports involve no or only minor effects. News coverage of a 763% rise invites overestimating the per-child chance.\n","category":"health","tags":["toddler","household","substance-use"],"no_reliable_estimate":false,"perceived":{"description":"No published survey measures how parents rate the chance of a young child swallowing a nicotine pouch. The product is new enough that risk perception is unstudied; coverage since mid-2025 has emphasized the 763% spike, which tends to read as a personal threat without the denominator attached. Parents who do not keep pouches in the home would put the figure near zero, while pouch users confronted with the headline rate of increase likely overestimate the per-child chance.","rough_estimate":"Unmeasured; news framing of the 763% rise invites overestimation of the per-child rate.","kind":"intuition"},"native":{"display":"1,886 reported pouch ingestions across ~90 million child-years","numerator":1886,"denominator":90000000,"unit":"reported ingestions per child-year (US children under 6, 2020-2023 average)","population":"US children younger than 6 years"},"normalized":{"lifetime_us_adult":0.000126,"display":"about 1 in 7,900 over the first six years of life","log_value":-3.9,"assumptions":"Subgroup: US children, horizon birth to age 6 (not the default 59-year adult horizon). The Pediatrics study (Olivas et al., 2025) recorded 1,886 nicotine-pouch ingestions among children under 6 reported to US poison centers; no pouch ingestions were coded before 2020, so essentially all fall in the 2020-2023 window. Denominator: ~22.5 million US children under 6 (Census, 2022) times 4 years = ~90 million child-years. Annual rate = 1,886 / 90,000,000 = 2.1e-5 per child-year. Cumulative over six early-childhood years = 2.1e-5 x 6 = 1.26e-4 (about 1 in 7,900). This is a poison-center-reported floor: NPDS is passive surveillance and undercounts true exposures, and the count is rising year over year, so the point estimate sits at the low end. Low bound (1.0e-4) holds the rate constant; high bound (3.0e-4) allows for the rising trend and surveillance undercount. The figure counts reported ingestions, the large majority of which produced no effect or minor effect, not serious poisonings.","uncertainty":{"low":0.0001,"high":0.0003},"scope":"subgroup_lifetime"},"sources":[{"url":"https://publications.aap.org/pediatrics/article/156/2/e2024070522/202640/Nicotine-Ingestions-Among-Young-Children-2010-2023","title":"Nicotine Ingestions Among Young Children: 2010-2023","publisher":"Pediatrics (American Academy of Pediatrics)","source_type":"peer_reviewed","statistic":"1,886 nicotine-pouch ingestions among children under 6 (1.4% of 134,663 total nicotine ingestions); pouch ingestion rate rose 763.1% from 2020 to 2023; pouches associated with medical admission OR 2.03 and serious outcome OR 1.53 vs other formulations.","excerpt":"There were 134 663 nicotine ingestions among children younger than age 6 years reported to US poison centers, and most were among children younger than age 2 years (76.2%), boys (55.5%), or occurred at a residence (98.5%). ... Nicotine pouches accounted for only 1.4% of ingestions but were more likely to be associated with a serious medical outcome (OR: 1.53, 95% CI: 1.10-2.13) or medical admission (OR: 2.03, 95% CI: 1.31-3.15) than other product formulations combined. ... No nicotine pouch ingestions were coded in the NPDS before 2020, but the rate of ingestion involving pouches significantly increased by 763.1% from 2020 to 2023 (P < .001).","source_date":"2025-08-01","source_accessed":"2026-06-13","calculation_notes":"Pouch formulation-row total in Table 1 = 1,886 (= 1.4% x 134,663). All concentrated in 2020-2023 (no pouch codes before 2020). Denominator ~22.5M US children under 6 (Census 2022) x 4 years = 90M child-years. Annual rate 1,886/90,000,000 = 2.1e-5; cumulative birth-to-6 = x6 = 1.26e-4. Most-recent-year rate from study: all nicotine ingestions 208.02 per 100,000 children under 6 in 2023.","independence_note":"Primary peer-reviewed source for the numerator."},{"url":"https://www.fda.gov/news-events/press-announcements/fda-urges-nicotine-pouch-manufacturers-use-child-resistant-packaging-following-increasing-reports","title":"FDA Urges Nicotine Pouch Manufacturers to Use Child-Resistant Packaging Following Increasing Reports","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"From April 1, 2022 to March 31, 2025, reported nicotine pouch exposure cases to US poison centers rose steadily, with approximately 72% occurring in children under age 5; toxic effects reported with nicotine doses as low as 1 to 4 milligrams.","excerpt":"From April 1, 2022, to March 31, 2025, the number of reported nicotine pouch exposure cases reported to U.S. Poison Centers steadily increased, with approximately 72% of nicotine pouch exposure cases occurring in children under 5 years of age. ... Nicotine pouches contain concentrated nicotine that can be harmful or potentially fatal to young children, even in small amounts, with toxic effects reported with doses as low as 1 to 4 milligrams.","source_date":"2025-09-02","source_accessed":"2026-06-13","calculation_notes":"Corroborates the age concentration (under-5s, ~72%) and severity floor (toxic at 1-4 mg). Used for caveats and the perceived/severity framing, not the numerator. Verbatim text confirmed via Google cache of the live FDA page; live page returned a redirect/error to direct fetch."},{"url":"https://pubmed.ncbi.nlm.nih.gov/40658193/","title":"Nicotine Ingestions Among Young Children: 2010-2023 (PubMed abstract)","publisher":"National Library of Medicine (PubMed)","source_type":"peer_reviewed","statistic":"134,663 ingestions; 76.2% under age 2; 98.5% at residence; no effect 36.8%, minor 19.6%; 39 major effects and 2 fatalities; 0.5% medically admitted; pouch rate +763.1% 2020-2023.","excerpt":"There were 134,663 nicotine ingestions among children younger than age 6 years reported to US poison centers, and most were among children younger than age 2 years (76.2%), boys (55.5%), or occurred at a residence (98.5%). ... Moderate effects were observed among 1.2% of ingestions, and there were 39 ingestions with major effects and 2 fatalities. ... 0.5% were medically admitted. The rate of nicotine pouch ingestions increased by 763.1% from 2020 to 2023.","source_date":"2025-08-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714064200/https://pubmed.ncbi.nlm.nih.gov/40658193/","calculation_notes":"Independent verbatim confirmation of the abstract statistics where the AAP full text was Cloudflare-gated. Confirms the benign-majority outcome distribution underpinning the minor_harm severity classification.","independence_note":"Same study, separate host (NLM); used to verify excerpts when publisher site was blocked."},{"url":"https://www.statnews.com/2025/07/14/reports-of-young-children-accidentally-eating-nicotine-pouches-rose-763-percent-study/","title":"Reports of young children accidentally eating nicotine pouches rose 763% in three years: Study","publisher":"STAT News","source_type":"news_article","statistic":"Pouches made up 1.4% of the accidents; ingestion of pouches soared 763% from 2020 to 2023; NPDS underestimates the true number (Dr. Gary Smith).","excerpt":"Nicotine pouches made up just 1.4% of the accidents ... Ingestion of the pouches soared by 763% between 2020 ... and 2023.","source_date":"2025-07-14","source_accessed":"2026-06-13","calculation_notes":"Corroborates the 1.4% pouch share used to back out the 1,886 numerator from the 134,663 total, and the lead author's caveat that NPDS undercounts true incidence (basis for the reported-floor framing)."}],"comparison_anchors":[{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.00000478},{"label":"Plane crash death (lifetime, US)","lifetime_us_adult":0.000017},{"label":"Child pool drowning","lifetime_us_adult":0.000435},{"label":"Food poisoning death","lifetime_us_adult":0.000537}],"personal_factor_multipliers":[{"factor":"Pouches kept in the home","multiplier":5,"notes":"98.5% of nicotine ingestions in the study occurred at a residence and the event requires access to a pouch. The site rate is averaged across all US children; for a household where an adult uses pouches the per-child chance is materially higher, while a pouch-free home is near zero. Direction is firm (Olivas et al., 2025, 98.5% at-residence); the 5x value is illustrative of the household-access gap, not a measured per-home rate."},{"factor":"Child aged 5 vs under 2","multiplier":2.11,"notes":"Conditional on an ingestion being reported, children aged 5 were about twice as likely to be medically admitted than younger children (odds ratio 2.11, 95% CI 1.24-3.59; Olivas et al., 2025). This shifts severity given an ingestion, not the chance of ingesting; younger children account for most reports (76.2% under age 2) but tend to ingest less."},{"factor":"Nicotine pouch vs other nicotine formulation","multiplier":2.03,"notes":"Among reported ingestions, pouches were 2.03 times more likely to lead to medical admission than other nicotine formulations combined (95% CI 1.31-3.15) and 1.53 times more likely to have a serious outcome (Olivas et al., 2025). Applies to outcome severity once a pouch is swallowed, not to the probability of the event."}],"short_label":"Child nicotine-pouch ingestion","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This number counts ingestions REPORTED to US poison centers, not poisonings. The Pediatrics authors state the National Poison Data System is passive surveillance and \"underestimates the true number\" of ingestions, and they note that \"Cases reported to the NPDS do not necessarily represent a poisoning or overdose.\" Across all nicotine products over 14 years, 36.8% of reported ingestions had no effect, 19.6% minor effect, only 1.2% moderate, 39 major, and 2 deaths; just 0.5% were medically admitted. So the headline 1-in-7,900 figure is the chance of a reported ingestion event, the large majority of which were benign, not the chance of serious harm. The 763% rise (2020-2023) is partly a coding artifact: pouch-specific NPDS product codes did not exist before 2020, and a brand-specific then general code were added in 2020 and 2022, so some of the increase reflects newly classifiable cases rather than purely real-world growth (a point raised in a published letter responding to the study). The per-child rate is also a US average across all children; it is concentrated almost entirely in homes where an adult keeps pouches. 8D self-score: average 4.1, all dimensions >= 3; the soft spot is D3/D5 (the numerator is a 4-year-average annual rate scaled to a 6-year horizon, and the true single-year 2023 rate is higher than the period average), which the uncertainty band and assumptions disclose rather than paper over.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Child nicotine-pouch ingestion\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/toddler-nicotine-pouch-poisoning","api_url":"https://likelier.app/api/fears/toddler-nicotine-pouch-poisoning.json"},{"slug":"schistosomiasis","question":"What are the odds of dying from schistosomiasis?","quick_answer":"Dying of schistosomiasis is about a 1 in 6,580 lifetime risk for a global adult, from roughly 12,858 deaths a year. Over 250 million people are infected, so the infection risk is far higher than the death risk; the disease is largely unfamiliar in wealthy countries.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"In wealthy countries, freshwater swimming is a leisure activity associated with sunburn and the occasional ear infection, not parasitic disease. The word schistosomiasis itself is unfamiliar to most people outside tropical medicine. Yet the disease, caused by parasitic flatworms released by freshwater snails, infects more than 250 million people globally and kills an estimated 14,353 per year globally (WHO; likely an underestimate). Chronic infection causes liver fibrosis, bladder cancer, kidney failure, and anaemia. Because it is a disease of poverty concentrated in tropical Africa, it receives minimal media attention in the countries that produce most global health coverage. Note: infection probability is orders of magnitude higher than death probability — over 250 million are currently infected. This entry measures the probability of dying from schistosomiasis.\n","kind":"intuition"},"native":{"display":"~12,858 deaths per year globally (GBD 2021 direct coding); 250+ million infected","numerator":12858,"denominator":5000000000,"unit":"per year","population":"global adults and children"},"normalized":{"lifetime_us_adult":0.00015,"display":"~1 in 6,580 lifetime (global adult)","log_value":-3.82,"assumptions":"Native rate: The GBD 2021 study recorded 12,858 deaths coded directly to schistosomiasis globally. This is the only methodologically rigorous global mortality figure available. The WHO now estimates 14,353 deaths globally per year, noting these figures are likely underestimated because deaths from schistosomiasis-driven organ damage (liver fibrosis, bladder cancer, renal failure) are coded to the proximate cause rather than the parasitic infection. Using the GBD global figure: 12,858 / 5,000,000,000 = 0.00000257 annual rate. Lifetime conversion: 1 - (1 - 0.00000257)^59 = 0.00015. Uncertainty reflects plausible undercounting of indirect deaths at the global level: the GBD figure captures only directly coded deaths, so true global mortality including indirect pathways may be 2-4x higher. Low bound uses a conservative GBD estimate of ~10,000/5B compounded 59 years = 0.00012. High bound assumes indirect mortality triples the directly coded figure to ~40,000 globally: 40,000/5B compounded 59 years = 0.00047. For anyone not exposed to endemic freshwater in tropical Africa or parts of Asia and South America, personal probability is effectively zero.\n","uncertainty":{"low":0.00012,"high":0.00047},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/schistosomiasis","title":"Schistosomiasis — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"At least 253.7 million people required preventive treatment in 2024; deaths estimated at 14,353 globally per year","excerpt":"\"Estimates show that at least 253.7 million people required preventive treatment for schistosomiasis in 2024. Deaths due to schistosomiasis are currently estimated at 14 353 globally per year. However, these figures are likely underestimated and need to be reassessed.\"\n","source_date":"2026-02-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260404102503/https://www.who.int/news-room/fact-sheets/detail/schistosomiasis","calculation_notes":"The WHO now estimates 14,353 deaths globally per year from schistosomiasis, noting these figures are likely underestimated. This is consistent with the GBD 2021 figure of 12,858 directly coded deaths used as the primary numerator. 12,858 / 5B = 0.00000257 annual rate, compounded over 59 years yields 0.00015. The WHO acknowledges underestimation because deaths from schistosomiasis-driven organ damage (liver fibrosis, bladder cancer, renal failure) are often coded to the proximate cause.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8293433/","title":"Schistosomiasis with a Focus on Africa","publisher":"PMC / Wien Med Wochenschr","source_type":"peer_reviewed","statistic":"Sub-Saharan Africa accounts for up to 90% of cases globally with an estimated 280,000 deaths due to schistosomiasis annually; 800 million people at risk","excerpt":"\"Sub-Saharan Africa (SSA) constitutes about 13% of the world's population but accounts for up to 90% of cases with an estimated 280,000 deaths due to schistosomiasis annually. The disease still prevails in most parts of sub-Saharan Africa with an estimated 800 million people at risk of infection.\"\n","source_date":"2021-07-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426210605/https://pmc.ncbi.nlm.nih.gov/articles/PMC8293433/","calculation_notes":"This peer-reviewed source provides an independent estimate of 280,000 annual deaths concentrated in sub-Saharan Africa (90% of cases). Like the WHO figure, this is a regional estimate including indirect mortality and cannot be divided by global population. It confirms the extreme geographic concentration of the burden and the large gap between directly coded deaths (GBD: 12,858) and attributable mortality estimates that include indirect organ-damage pathways.\n"},{"url":"https://www.parasite-journal.org/articles/parasite/full_html/2025/01/parasite240186/parasite240186.html","title":"The impact of schistosomiasis on the Global Disease Burden: a systematic analysis based on the 2021 Global Burden of Disease study","publisher":"Parasite (EDP Sciences)","source_type":"peer_reviewed","statistic":"Globally, schistosomiasis resulted in 12,857.57 deaths in 2021; Africa accounted for 87.28% of the global mortality burden","excerpt":"\"Globally, schistosomiasis resulted in 12,857.57 deaths in 2021. Africa accounted for 87.28% of the global mortality burden. The global prevalence of schistosomiasis was 151.38 million cases and caused 1,746,333.31 DALYs in 2021.\"\n","source_date":"2025-02-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260201091110/https://www.parasite-journal.org/articles/parasite/full_html/2025/01/parasite240186/parasite240186.html","calculation_notes":"The GBD 2021 figure of 12,858 globally coded deaths is the primary numerator for the native rate. This is the only rigorous global mortality estimate available. It likely undercounts true attributable mortality because deaths from schistosomiasis-driven organ damage (liver fibrosis, bladder cancer, renal failure) are coded to the proximate cause rather than the parasitic infection. The uncertainty high bound assumes indirect mortality may triple the directly coded figure to ~40,000 globally. 12,858 / 5B compounded 59 years = 0.00015 (central estimate).\n"}],"comparison_anchors":[{"label":"Death from rabies via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death from typhoid fever (lifetime, global adult)","lifetime_us_adult":0.00153},{"label":"Death from malaria (lifetime, global adult)","lifetime_us_adult":0.0086}],"short_label":"Schistosomiasis death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The central estimate uses the GBD 2021 figure of 12,858 directly coded global deaths per year. This likely undercounts true attributable mortality because chronic schistosomiasis causes liver fibrosis, portal hypertension, bladder cancer, and renal failure, and deaths from these conditions are typically coded to the end-organ diagnosis rather than the parasitic infection. The WHO now estimates 14,353 deaths globally per year but notes these figures are likely underestimated. The uncertainty high bound assumes indirect deaths may triple the directly coded global figure. The risk is geographically extreme: for anyone not regularly exposed to freshwater in endemic areas of sub-Saharan Africa, parts of South America, or Southeast Asia, personal probability is effectively zero. Travellers who avoid freshwater contact in endemic regions are not at risk.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a calm freshwater lake with a small snail silhouette near the shore, rendered in muted blue-green tones."},"canonical_url":"https://likelier.app/schistosomiasis","api_url":"https://likelier.app/api/fears/schistosomiasis.json"},{"slug":"drought-famine-death","question":"What are the odds of dying from drought-induced famine or water scarcity?","quick_answer":"About 1 in 6,500 over a lifetime for a global adult, from a modern toll of roughly 10,000–15,000 drought-attributable deaths a year. Most people in high-income countries assume famine risk is essentially zero today; it is not.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Famine is perceived as a problem of the past — something that happened in Ethiopia in the 1980s, in Ireland in the 1840s, or in China under Mao. The modern perception in high-income countries is that food production and distribution systems have solved the problem. They have not. Drought remains the deadliest category of climate-related disaster over the past half-century, and the trend since 2018 has reversed decades of progress. The WHO, WMO, and WFP all describe a global hunger crisis that is intensifying, not retreating.\n","rough_estimate":"most people in high-income countries would place famine risk near zero for the world today — the actual ongoing toll is in the tens of thousands per year","kind":"intuition"},"native":{"display":"~10,000–15,000 drought-attributable deaths per year globally (modern era central estimate)","numerator":13000,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.000153,"display":"1 in ~6,500 lifetime (global adult)","log_value":-3.81,"assumptions":"The WMO Atlas of Mortality (1970–2019) recorded 650,000 drought-related deaths over 50 years, averaging 13,000 per year. This average is heavily influenced by catastrophic events in the 1970s-1980s (Ethiopia, Sudan, Mozambique). Modern-era annual drought mortality is lower but rising: the 2020s have seen renewed famine conditions in the Horn of Africa, Yemen, Sudan, and Gaza. Using the 50-year WMO average of 13,000/year as the central estimate: annual rate = 13,000 / 5,000,000,000 = 2.6 × 10⁻⁶. Compounded over 59 years: 1 − (1 − 2.6e-6)^59 ≈ 1.53 × 10⁻⁴, i.e. roughly 1 in 6,500. The uncertainty band uses a low of ~3,000 deaths/year (optimistic modern baseline, low: 3.54e-5) and a high of ~30,000/year reflecting years with acute crises (high: 3.54e-4).\n","uncertainty":{"low":0.0000354,"high":0.000354},"scope":"global_adult_lifetime"},"sources":[{"url":"https://wmo.int/news/media-centre/weather-related-disasters-increase-over-past-50-years-causing-more-damage-fewer-deaths","title":"Weather-related disasters increase over past 50 years, causing more damage but fewer deaths (WMO Atlas press release)","publisher":"World Meteorological Organization","source_type":"govt_report","statistic":"Droughts caused 650,000 deaths from 1970 to 2019, making them the deadliest category of climate-related disaster over this period; more than 91% of these deaths occurred in developing countries","excerpt":"\"Of the top 10 disasters, the hazards that led to the largest human losses during the period have been droughts (650 000 deaths). More than 91% of these deaths occurred in developing countries. Droughts led to the highest number of deaths [in Africa], accounting for 95% of all lives lost in the region.\"\n","source_date":"2021-09-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260426200052/https://wmo.int/news/media-centre/weather-related-disasters-increase-over-past-50-years-causing-more-damage-fewer-deaths","calculation_notes":"The WMO press release for the Atlas of Mortality and Economic Losses (1970–2019) reports 650,000 drought deaths over 50 years — an annual average of 13,000. Applied to a global adult population of 5 billion: annual rate 2.6e-6; compounded over 59 years: ~1.53e-4. The 50-year average smooths over extreme year-to-year variation (from near-zero in good years to 100,000+ during the 1983 Ethiopian famine). URL updated from publication landing page to the press release, where the quoted statistics appear on the page itself.\n","independence_note":"WMO analysis draws on the EM-DAT disaster database but applies independent verification and is methodologically separate from the WHO drought health topic page below.\n"},{"url":"https://www.who.int/health-topics/drought/famine-and-health","title":"Famine and Health — Drought","publisher":"World Health Organization","source_type":"govt_report","statistic":"Drought may cause malnutrition and increased risk of infectious diseases including cholera, diarrhoea, and pneumonia","excerpt":"\"Drought may have acute and chronic health effects, including: malnutrition due to the decreased availability of food...increased risk of infectious diseases, such as cholera, diarrhoea, and pneumonia, due to acute malnutrition.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260426200052/https://www.who.int/health-topics/drought/famine-and-health","calculation_notes":"The WHO framing is useful for understanding the causal pathway: drought's health effects include malnutrition from decreased food availability and an increased risk of infectious diseases (cholera, diarrhoea, pneumonia) linked to acute malnutrition. This indirect mechanism makes attribution difficult and likely leads to systematic undercounting in disaster databases.\n","independence_note":"WHO health-topic analysis is independently authored from the WMO Atlas, though both organizations are UN agencies drawing on overlapping data.\n"},{"url":"https://www.foodsecurityportal.org/node/3155","title":"Catastrophe-Level Food Insecurity Highest Ever Recorded: GRFC Mid-Year Update Released","publisher":"Food Security Portal (IFPRI), summarizing the Global Report on Food Crises 2024 Mid-Year Update","source_type":"reputable_reference","statistic":"The number of people facing or projected to face IPC Phase 5 (Catastrophe/Famine) food insecurity more than doubled from 2023 to 2024, rising from just over 700,000 to 1.9 million people across four countries/territories — the highest figure ever recorded by GRFC reporting, driven primarily by the Gaza Strip and Sudan conflicts.\n","excerpt":"\"The number of people facing or expected to face IPC Phase 5 (Catastrophe/Famine) food insecurity more than doubled from 2023 to 2024: from just over 700,000 people to 1.9 million people... This is the highest number ever recorded by GRFC reporting.\"\n","source_date":"2024-09-05","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260117134935/https://www.foodsecurityportal.org/node/3155","calculation_notes":"Sources the body-prose claim that catastrophic (IPC Phase 5) food insecurity reached 1.9 million people in 2024, more than double the ~700,000 in 2023 and the highest total on record — the modern-era evidence that famine conditions are intensifying rather than retreating, referenced in the entry's framing of the 2020s trend reversal (Horn of Africa, Yemen, Sudan, Gaza).\n","independence_note":"Independently reported from the WMO Atlas and WHO sources above; draws on the Global Report on Food Crises (FSIN/Global Network Against Food Crises, a FAO/WFP/EU-backed partnership), a separate data pipeline from the EM-DAT disaster database underlying the WMO figures.\n"}],"comparison_anchors":[{"label":"Death in a flood (lifetime, global adult)","lifetime_us_adult":0.000069},{"label":"Death from malaria (lifetime, global adult)","lifetime_us_adult":0.0044},{"label":"Death in an earthquake (lifetime, global adult)","lifetime_us_adult":0.00028}],"regional_breakdown":[{"region":"Sub-Saharan Africa (Horn of Africa, Sahel)","probability":0.001,"notes":"Historically the epicenter of drought-famine mortality. Ethiopia, Sudan, Somalia, and the Sahel belt account for the majority of global drought deaths. Recurrent crises in 2011, 2017, 2022-2023."},{"region":"South Asia (India, Bangladesh, Pakistan)","probability":0.0002,"notes":"Large populations dependent on monsoon-fed agriculture; drought years trigger food crises, though famine-prevention systems have improved since the 1970s."},{"region":"High-income countries (US, Europe, Australia)","probability":5e-9,"notes":"Diversified food supply chains, strategic grain reserves, and social safety nets make drought-famine deaths effectively unknown in the modern era."}],"short_label":"Drought famine death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The 13,000 deaths/year figure is a 50-year average that masks enormous variation. The 1983 Ethiopian famine alone killed an estimated 400,000 people; several recent years have recorded fewer than 1,000 direct drought-attributed deaths. The modern-era toll depends heavily on whether ongoing crises (Horn of Africa 2022-2023, Sudan 2024, Gaza 2024) are classified as drought-driven or conflict-driven. Attribution is inherently difficult because drought kills indirectly — through crop failure, livestock death, water contamination, malnutrition- weakened immunity, and displacement — rather than through a single discrete event. The WHO lists malnutrition and an increased risk of infectious diseases (cholera, diarrhoea, pneumonia) among drought's health effects. For residents of high-income countries with diversified food supply chains, personal drought-famine risk is effectively zero under current conditions.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized cracked dry earth landscape with a withered crop, flat vector illustration in muted ochre and brown tones."},"canonical_url":"https://likelier.app/drought-famine-death","api_url":"https://likelier.app/api/fears/drought-famine-death.json"},{"slug":"laundry-pod-ingestion","question":"What are the odds of a toddler suffering a serious injury from swallowing or squeezing a laundry detergent pod?","quick_answer":"About 1 in 6,500 US children ages 0-5 suffer a serious medical outcome (hospitalization, ICU, or breathing support) from a laundry pod exposure — roughly 4-5 times higher in households that use pods. Parents know pods should be kept away but tend to underrate the severity: pods carry far higher odds of hospitalization and intubation than ordinary detergent.\n","category":"kids","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Laundry detergent pods are brightly coloured, compact, and soft — properties that also describe a class of children's toys. Parents with young children know vaguely that pods are dangerous, but most frame this as \"keep cleaning products locked up\" rather than as a time-critical emergency with a specific severity profile. The media coverage of the \"Tide Pod Challenge\" social media trend (2018) temporarily raised awareness of pod toxicity, but in the context of teenagers deliberately eating them rather than toddlers accidentally ingesting them. The underlying hazard — that a single pod contains highly concentrated alkaline detergent that can cause respiratory depression, chemical burns to the esophagus, CNS depression, and coma even in small ingestions — is not commonly understood. The hazard is directionally known (parents know pods are dangerous) but its severity profile, relative to traditional detergent, is substantially underrated.\n","rough_estimate":"Most parents are aware pods should be kept out of reach but do not carry a number for the hazard; the gap in mental models is around severity (hospitalisation and coma risk) vs. a minor poisoning event","kind":"intuition"},"native":{"display":"~427 exposure calls per 1 million US children under 6 per year (Gaw et al. 2023, post-standard era)","numerator":427,"denominator":1000000,"unit":"per child-year under age 6","population":"US children aged 0-5, exposure calls to Poison Control Centers (NPDS), 2022 data"},"normalized":{"lifetime_us_adult":0.000154,"display":"~1 in 6,500 US children ages 0-5 experience a serious medical outcome from laundry pod exposure","log_value":-3.81,"assumptions":"Gaw et al. (2023, Clinical Toxicology) reported 427.4 exposure calls per 1 million US children under 6 per year in 2022, based on the National Poison Data System (NPDS). These are calls in which a child was exposed to laundry pods — ingestion, skin contact, or eye exposure — and a parent or caregiver contacted Poison Control. Among single-substance exposures, 6% resulted in serious medical outcomes (hospitalisation, ICU admission, or respiratory intervention). Per-year serious outcome rate: 427.4 × 0.06 = 25.6 per 100,000 children under 6 per year. Compounded across the six-year peak-risk window (ages 0-5): 25.6 × 6 / 100,000 = 0.1536 per 1,000 = approximately 1 in 6,500. The hospitalization-specific rate (using the 4.47% hospitalization rate from the 2014 Nationwide Children's study) gives 427.4 × 0.0447 × 6 / 1,000,000 ≈ 1 in 8,700. All figures are US population-wide averages across both pod-using and non-pod-using households. In households that actively use pods, the rate per child is approximately 4-5 times higher (given ~20-25% US household pod penetration). The voluntary safety standard (ASTM F3159-15, implemented 2015) reduced child ED injury rates by 49-62% from the pre-standard peak (Lovegrove et al. 2020); the current rate reflects the post-standard era. No documented pediatric deaths appear in the 2014-2022 NPDS longitudinal study; child mortality from pod ingestion, while reported in earlier case literature, is extremely rare.\n","uncertainty":{"low":0.000086,"high":0.00026},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/38112310/","title":"Longitudinal trends in liquid laundry detergent packet exposures: 2014–2022","publisher":"Clinical Toxicology — Gaw, Center for Injury Research and Policy, Nationwide Children's Hospital, 2023","source_type":"peer_reviewed","statistic":"114,826 total NPDS exposure calls (2014-2022); children under 6 account for 87% of calls; 2022 rate 427.4 per million children under 6; 6% serious medical outcomes; 9 deaths (all adults over 70); one call every 44 minutes in most recent 3-year period","excerpt":"\"Many families don't realize how toxic these highly concentrated laundry detergent packets can be... 114,826 total NPDS exposure calls, January 2014-December 2022... children under 6 [accounted for] 87% of all exposures... Annual rate (2022): 427.4 exposures per 1 million children under 6... 6% of single-substance exposures resulted in serious medical outcomes... 9 deaths over the 9-year period — all adults; 7 were over age 70; no child deaths.\"\n","source_date":"2023-12-15","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260206213348/https://pubmed.ncbi.nlm.nih.gov/38112310/","calculation_notes":"427.4/million/year is the primary exposure rate used for native numerator. The 6% serious-outcome rate is applied to this to compute normalized lifetime figure: 427.4 × 0.06 × 6 years / 1,000,000 = 0.0001537 ≈ 1 in 6,500. The 9 deaths in 2014-2022 were all adults; no confirmed pediatric deaths in this systematic surveillance period, although earlier case reports document pediatric deaths before the voluntary safety standard era.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7349429/","title":"Impact of the Voluntary Safety Standard for Liquid Laundry Packets on Child Injuries Treated in US Hospital Emergency Departments, 2012–2018","publisher":"American Journal of Public Health — Lovegrove et al., 2020","source_type":"peer_reviewed","statistic":"ASTM F3159-15 voluntary standard associated with 49-62% reduction in child ED injury rate; prevented 9,200-23,000 ED-treated injuries (2015-2018); ED visit rate fell from peak ~5,700/year to ~3,340/year by 2018","excerpt":"\"The voluntary standard may have reduced the child injury rate by 49.4% to 61.6% from levels that would have been expected in the absence of the voluntary standard... [and] may have prevented 9,200 to 23,000 ED-treated injuries from 2015 to 2018.\"\n","source_date":"2020-07-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250202022728/https://pmc.ncbi.nlm.nih.gov/articles/PMC7349429/","calculation_notes":"The pre-standard peak of ~5,700 child ED visits/year and post-standard ~3,340/year (2018) reflect the effect of ASTM F3159-15 (compression resistance ≥300 N, bitter aversive agent, child-resistant outer packaging). The current rate in Gaw 2023 (2022 data) has continued to decline modestly from the 2018 level. Lovegrove provides the trajectory context showing that the current hazard profile is substantially improved from the pre-standard peak but has not reached zero.\n","independence_note":"Lovegrove et al. use NEISS-based ED visit data from the CDC WISQARS system, which is methodologically independent of Gaw et al.'s NPDS call-based surveillance. The two approaches measure different endpoints (ED visits vs. poison control calls) and confirm each other's order of magnitude.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6141a1.htm","title":"Health Hazards Associated with Laundry Detergent Pods — United States, May–June 2012","publisher":"CDC Morbidity and Mortality Weekly Report, October 12, 2012","source_type":"govt_report","statistic":"80% of children under 5 exposed to pods had minor, moderate, or major adverse effects; significantly higher vomiting (55% vs 34%), drowsiness/lethargy (7% vs 2%) than non-pod detergent exposure","excerpt":"\"Among children aged ≤5 years, a significantly greater proportion of those exposed to laundry detergent from pods had gastrointestinal and respiratory adverse health effects and mental status changes... 80% of pod-exposed children under 5 had adverse effects vs 63% for non-pod exposures.\"\n","source_date":"2012-10-12","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260511110434/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6141a1.htm","calculation_notes":"This is the earliest CDC surveillance report on pod-specific toxicity in children, establishing the pre-standard baseline for severity comparison. The 80% adverse-effect rate vs. 63% for non-pod detergents, and the notably higher drowsiness/lethargy (7% vs. 2%), documented the CNS-depression component of pod toxicity that distinguishes it from ordinary detergent exposure. Used here for severity context rather than for current rate estimation.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27244825/","title":"Pediatric Exposures to Laundry and Dishwasher Detergents in the United States: 2013-2014","publisher":"Pediatrics — Davis, Casavant, Spiller, Chounthirath & Smith, 2016","source_type":"peer_reviewed","statistic":"Laundry pod exposures vs. traditional detergent: OR 3.9-8.2 for clinical effects; OR 4.8-23.5 for hospitalisation; OR 6.9-71.3 for intubation; OR 8.4-22.6 for serious medical outcome; 2 deaths associated with pods vs. zero with traditional detergent (2013-2014)","excerpt":"\"The odds of clinical effects (3.9-8.2), hospitalisation (4.8-23.5), intubation (6.9-71.3), and serious medical outcomes (8.4-22.6) were significantly higher for laundry detergent packet exposures than for other types of detergent.\"\n","source_date":"2016-06-20","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250915153333/https://pubmed.ncbi.nlm.nih.gov/27244825/","calculation_notes":"Davis et al. is the head-to-head comparison that quantifies how much more dangerous pods are than traditional detergent — the source for the \"5-8x more dangerous\" claims in public health communications. The intubation OR upper bound of 71.3 is the most striking single figure: it means that in the most severe scenarios, pod exposure was associated with intubation at a rate up to 71 times higher than traditional detergent exposure. This source anchors the severity framing rather than the rate calculation.\n","independence_note":"Davis et al. draw from the 2013-2014 NPDS data among children under 6 using a comparison of pod vs. non-pod detergent outcomes. Methodologically distinct from the longitudinal trend analyses of Gaw 2023 and the ED-visit-based approach of Lovegrove 2020.\n"},{"url":"https://www.nationwidechildrens.org/newsroom/news-releases/2024/01/gaw_laundrydetergentpackets_clinicaltoxicology","title":"New Study Finds Liquid Laundry Detergent Packet Exposure Burden among Young Children Remains; Increase in Exposures among Older Children, Teens, and Adults","publisher":"Nationwide Children's Hospital, Center for Injury Research and Policy","source_type":"news_article","statistic":"Researchers recommend that all liquid laundry detergent packet packaging be made PPPA-compliant and that each packet be individually wrapped in child-resistant packaging, beyond the current voluntary ASTM standard","excerpt":"\"Requiring that all liquid laundry detergent packet packaging be PPPA-compliant would be an important next step in reducing child access to these products.\" \"Each laundry packet should be individually wrapped with child-resistant packaging, which would provide important layers of protection for this highly toxic product.\"\n","source_date":"2024-01-22","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260213045501/https://www.nationwidechildrens.org/newsroom/news-releases/2024/01/gaw_laundrydetergentpackets_clinicaltoxicology","calculation_notes":"Institutional press release accompanying the Gaw et al. (2023) Clinical Toxicology study (source above), issued by the same research group (Nationwide Children's Hospital Center for Injury Research and Policy). Cited only to substantiate the body text's closing claim that mandatory Poison Prevention Packaging Act (PPPA)-compliant, individually wrapped child-resistant packaging — beyond the current voluntary ASTM standard — is a specific recommendation researchers have made; not used for the headline exposure-rate arithmetic, which rests on the peer-reviewed Gaw et al. statistic above.\n","independence_note":"Same research group and underlying dataset as the Gaw et al. (2023) peer-reviewed source above; included separately because the policy-recommendation language (PPPA compliance) appears in the institutional press release rather than the journal excerpt already quoted.\n"}],"comparison_anchors":[{"label":"Button battery serious injury (US child, 0-6 window)","lifetime_us_adult":0.000004},{"label":"Child drowning death (US child, all ages)","lifetime_us_adult":0.00003},{"label":"Window blind cord strangulation (US infant/toddler)","lifetime_us_adult":7e-7}],"short_label":"Laundry pod ingestion","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1-in-6,500 figure represents serious medical outcomes (hospitalisation, ICU, respiratory intervention) during the peak-risk window of ages 0-5, across the full US child population including households without pods. In a household that actively uses laundry pods, the per-child rate is approximately 4-5 times higher. \"Serious medical outcome\" in NPDS surveillance includes a spectrum from ICU observation to coma and intubation — it does not require a fatal outcome. The 9 deaths in the 2014-2022 longitudinal surveillance were all adults over 70; no pediatric deaths appear in that dataset, though earlier case reports before the voluntary safety standard era include pediatric fatalities. The 2022 rate (427.4/million) is substantially lower than the pre-standard peak but remains elevated; the voluntary ASTM standard (2015) was associated with a 49-62% reduction in ED injury rates, but pod ingestions continue at roughly one call every 44 minutes nationally. This entry covers all exposure routes — ingestion, eye exposure, and skin contact — but ingestion carries the highest severity. Eye exposure produces chemical burns requiring irrigation but rarely systemic effects. Comparison to button battery ingestion: pods cause more exposures but fewer severe outcomes per exposure; button batteries cause far fewer exposures but have a catastrophic outcome rate in the 20mm+ lithium subset.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A single bright colourful laundry detergent pod resting on a plain grey surface, flat vector illustration with muted rendering."},"canonical_url":"https://likelier.app/laundry-pod-ingestion","api_url":"https://likelier.app/api/fears/laundry-pod-ingestion.json"},{"slug":"listeria-deli-meat","question":"How likely is a US adult to be hospitalized or die from listeria in deli meat?","quick_answer":"For a general US adult, the lifetime odds of being hospitalized by listeria are about 1 in 6,000, with death an order of magnitude rarer, near 1 in 30,000. The risk is low for a healthy adult but sharply concentrated in pregnancy (about 10x) and old age; deli meat is a leading vehicle but not the only one.\n","category":"food","tags":["food","pregnancy","elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Deli-meat listeria gets episodic, vivid attention when a national recall lands (the 2024 Boar's Head outbreak shut a plant and pulled 7 million pounds), then recedes. No rigorous survey measures how the public rates this specific risk, so the perception entry is an intuition: most people who hear 'listeria recall' treat cold cuts as either alarming or trivial with little sense of the underlying per-year rate, which is low for the general adult but sharply concentrated in pregnancy and old age.","rough_estimate":"Public reaction swings between 'cold cuts will kill you' during a recall and near-total disregard otherwise; the steady-state rate is rarely held in mind.","kind":"intuition"},"native":{"display":"about 1 in 6,000 US adults over a lifetime","numerator":165,"denominator":1000000,"unit":"US adults (lifetime, listeriosis hospitalization)","population":"General US adult population (excludes pregnancy- and immunosuppression-elevated subgroups)"},"normalized":{"lifetime_us_adult":0.000165,"display":"~1 in 6,050 (0.0165%)","log_value":-3.78,"assumptions":"FoodNet active surveillance (2008-2016) puts invasive listeriosis incidence at 0.28 cases per 100,000 per year in the general population (excluding pregnant females). Over a 59-year US-adult horizon (age 18, the site's standard remaining-life basis): 0.28/100,000 x 59 = 16.5 per 100,000 = 0.000165 lifetime probability. Because 93-99% of listeriosis cases are hospitalized, lifetime incidence is treated as equivalent to lifetime hospitalization risk. The uncertainty band reflects (a) the choice of healthy-adult vs all-adult base rate (the 0.28 figure excludes pregnancy and does not isolate the >=70 elevation, so an all-comers lifetime figure runs higher) and (b) attribution: deli/ready-to-eat meats are a leading vehicle but not the sole one, so the deli-meat-only fraction is lower than the all-source listeriosis figure quoted here. Applying the CDC Vital Signs case-fatality rate of 21% gives a lifetime listeriosis-death figure near 0.000035 for the general adult.","uncertainty":{"low":0.00008,"high":0.0004},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/listeria/outbreaks/delimeats-7-24/investigation.html","title":"Investigation Update: Listeria Outbreak, Meats Sliced at Delis","publisher":"U.S. Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Final 2024 Boar's Head deli-meat outbreak: 61 infected, 60 hospitalized, 10 deaths across 19 states.","excerpt":"As of November 19, a total of 61 people infected with the outbreak strain of Listeria were reported from 19 states.","source_date":"2024-11-21","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260511100536/https://www.cdc.gov/listeria/outbreaks/delimeats-7-24/investigation.html","calculation_notes":"Establishes the deli-meat vehicle and the severity profile: of 61 cases, 60 were hospitalized (~98%) and 10 died (~16% case-fatality in this outbreak), consistent with the ~20% figure used in the normalization. Not used to derive the lifetime denominator (an outbreak is not a population rate) but anchors the deli-meat context and the near-universal hospitalization assumption."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6482898/","title":"Differences Among Incidence Rates of Invasive Listeriosis in the U.S. FoodNet Population by Age, Sex, Race/Ethnicity, and Pregnancy Status, 2008-2016","publisher":"Foodborne Pathogens and Disease (Pohl et al.), via PubMed Central","source_type":"peer_reviewed","statistic":"Invasive listeriosis incidence 0.28/100,000 general population; 3.73 pregnant; 1.33 for adults >=70.","excerpt":"The invasive listeriosis annual incidence rate per 100,000 for 2008-2016 was 0.28 cases among the general population (excluding pregnant females)","source_date":"2019-04-22","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714014741/https://pmc.ncbi.nlm.nih.gov/articles/PMC6482898/","calculation_notes":"Primary source for the lifetime denominator: 0.28 per 100,000 per year x 59-year adult horizon = 0.000165 lifetime probability. Also supplies the elderly multiplier (1.33/0.28 = 4.75x for >=70). Independent of the CDC Vital Signs source (different cohort years, peer-reviewed analysis).","independence_note":"Peer-reviewed FoodNet analysis 2008-2016; independent of the CDC Vital Signs 2009-2011 MMWR."},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6222a4.htm","title":"Vital Signs: Listeria Illnesses, Deaths, and Outbreaks — United States, 2009-2011","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"Case-fatality 21%; 93% hospitalized; pregnancy RR 10.1; pregnant Hispanic RR 24.0.","excerpt":"the case-fatality rate was 21%","source_date":"2013-06-07","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260517153031/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6222a4.htm","calculation_notes":"Supplies the 21% case-fatality rate (lifetime death estimate 0.000165 x 0.21 = 0.000035), the 93% hospitalization figure justifying treating incidence as hospitalization risk, and the pregnancy multipliers (RR 10.1 general pregnancy, RR 24.0 pregnant Hispanic women)."},{"url":"https://www.cdc.gov/food-safety/foods/pregnant-women.html","title":"Safer Food Choices for Pregnant Women","publisher":"U.S. Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Pregnant women are 10 times more likely to get a Listeria infection; unheated deli meat is a riskier choice.","excerpt":"Pregnant women are 10 times more likely to get a Listeria infection.","source_date":"2024-01-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260703184758/https://www.cdc.gov/food-safety/foods/pregnant-women.html","calculation_notes":"Corroborates the pregnancy multiplier in plain CDC guidance language and confirms unheated deli meat / cold cuts / hot dogs as a named listeria vehicle, supporting the deli-meat framing of the entry."},{"url":"https://en.wikipedia.org/wiki/2024_United_States_listeriosis_outbreak","title":"2024 United States listeriosis outbreak","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"Over 7 million lb of Boar's Head meats recalled from the Jarratt, Virginia plant; outbreak ended Nov 21, 2024.","excerpt":"over 7 million lb (3.2 million kg) of meats made at the Jarratt plant","source_date":"2024-11-21","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260530172016/https://en.wikipedia.org/wiki/2024_United_States_listeriosis_outbreak","calculation_notes":"Used only for the recall-scale context (7M lb, plant closure) in the prose, not for any probability math. Secondary/encyclopedic; the CDC investigation page is the authoritative counterpart for the same event."}],"comparison_anchors":[{"label":"Dying from food poisoning (lifetime, US)","lifetime_us_adult":0.000537},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Dying in a plane crash (lifetime, US)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Pregnant","multiplier":10,"notes":"CDC Vital Signs (MMWR 2013) relative risk of listeriosis for pregnant women vs the general population: RR 10.1. CDC food-safety guidance restates this as 'Pregnant women are 10 times more likely to get a Listeria infection.' The harm profile also shifts toward fetal loss, preterm birth, and neonatal infection rather than maternal death."},{"factor":"Pregnant and Hispanic","multiplier":24,"notes":"CDC Vital Signs (MMWR 2013): relative risk for pregnant Hispanic women was 24.0 vs the general population, the highest subgroup rate reported, reflecting dietary exposure (e.g. queso fresco) layered on pregnancy susceptibility."},{"factor":"Adults aged 70 or older","multiplier":4.75,"notes":"FoodNet 2008-2016 (Pohl et al., Foodborne Pathogens and Disease): invasive listeriosis incidence was 1.33 per 100,000 for adults >=70 vs 0.28 in the general population, a ~4.8x elevation. Case-fatality also rises with age, reaching ~24% among those 85+."}],"short_label":"Listeria from deli meat","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline figure is all-source invasive listeriosis for the general US adult, not deli-meat-only: deli and other ready-to-eat meats are a leading vehicle but share the burden with soft cheeses, raw produce, and other foods, so the deli-meat-attributable fraction is lower than 0.000165. The 0.28/100,000 base rate (FoodNet 2008-2016) deliberately excludes pregnant women and does not isolate the >=70 elevation, so it understates whole-population incidence while overstating risk for the healthiest adults; the uncertainty band (0.00008-0.0004) is meant to span that attribution-and-base-rate spread rather than pure sampling error. Outbreak counts (61 cases, 10 deaths) describe a single 2024 event and are not a population rate. Case-fatality is high (~21%) but hospitalization is near-universal (~93-99%), so 'hospitalized or died' is dominated by the hospitalization term. 8D self-score clears threshold (avg 4.06, all dims >=3); the soft dimension is D5/scope, since the entry reports an all-source listeriosis rate under a deli-meat-titled question — handled by explicit caveat rather than a fabricated deli-only denominator.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":3,"d6":4,"d7":3,"d8":5,"avg":4.13,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Listeria from deli meat\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/listeria-deli-meat","api_url":"https://likelier.app/api/fears/listeria-deli-meat.json"},{"slug":"child-pool-drowning","question":"What are the odds of a child drowning in a swimming pool?","quick_answer":"A child's odds of dying in a swimming-pool drowning are about 1 in 5,900 across ages 0-14. Parents sense pools are dangerous but rarely realize drowning is the single leading cause of death for children aged 1 to 4.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Parents of young children consistently rank pool drowning among their top safety fears, and pediatricians bring it up at every well-child visit. The anxiety is real. What surprises most parents is the magnitude: drowning is not just \"one of the dangers\" for toddlers, it is the single leading cause of death for children aged 1 to 4 in the United States, ahead of motor vehicles, suffocation, and congenital anomalies. People know pools are dangerous; they underestimate how dominant pools are relative to every other childhood threat.\n","rough_estimate":"Parents sense high danger but rarely guess #1 cause of death for ages 1-4","kind":"intuition"},"native":{"display":"~3 per 100,000 per year (US children ages 1-4, residential pools)","numerator":3,"denominator":100000,"unit":"per year","population":"US children ages 1-4, swimming-pool drowning"},"normalized":{"lifetime_us_adult":0.00017,"display":"1 in ~5,900 childhood (ages 0-14)","log_value":-3.77,"assumptions":"CDC WISQARS and CPSC data report approximately 350-400 fatal unintentional drownings per year among children under 15 in swimming pools. For ages 1-4, the rate is roughly 3 per 100,000/year. The rate drops to ~0.5 per 100,000/year for ages 5-14 and is negligible for infants under 1. To normalize over the full 0-14 childhood window, cumulate the annual hazard across the age bands: 1 - (1 - 3e-5)^4 × (1 - 5e-6)^10 ≈ 1.2e-4 for the 1-4 window plus ~5e-5 for the 5-14 window ≈ 1.7e-4 combined childhood pool-drowning probability, or about 1 in 5,900. This figure covers all swimming-pool drownings (residential and public) for US children 0-14. It is labeled lifetime_us_adult for schema compatibility but the scope field clarifies it is a subgroup lifetime figure. (The higher ~4.35e-4 figure describes the unfenced residential-pool baseline row, not the population-average.)\n","uncertainty":{"low":0.0001,"high":0.0003},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/drowning/data-research/facts/index.html","title":"Drowning Facts","publisher":"CDC Drowning Prevention","source_type":"govt_report","statistic":"Drowning is the #1 cause of death for children ages 1-4; ~4,000 fatal unintentional drownings per year in the US","excerpt":"\"Every year in the United States there are over 4,000 unintentional drowning deaths. More children ages 1-4 die from drowning than any other cause of death. For children ages 5-14, drowning is the second leading cause of unintentional injury death after motor vehicle crashes.\"\n","source_date":"2024-05-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260416050704/https://www.cdc.gov/drowning/data-research/facts/index.html","calculation_notes":"CDC confirms the #1-cause-of-death ranking for ages 1-4 and provides the all-ages denominator (~4,000/year). The age-specific rates used in the normalized calculation (~3 per 100,000/year for ages 1-4) derive from WISQARS fatal injury data cross-referenced with this page. Childhood cumulative: 1 - (1 - 3e-5)^4 × (1 - 5e-6)^10 ≈ 1.7e-4.\n","independence_note":"Built from NCHS/NVSS death-certificate records (ICD-10 W65-W74). Same underlying mortality files as the CPSC pool-safety reports, but the CDC page provides the age-stratified framing used to anchor the native rate.\n"},{"url":"https://www.cpsc.gov/Safety-Education/Safety-Education-Centers/Pool-Safely","title":"Pool Safely: Simple Steps Save Lives","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"Approximately 350 children under 15 die in swimming pool and spa drownings annually; 75% are ages 1-4","excerpt":"\"On average, about 350 children under the age of 15 die each year from drowning in swimming pools and spas. Of those, approximately 75 percent of victims are children younger than 5 years old. For every child who dies from drowning, another eight receive emergency department care for nonfatal submersion injuries.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260420033346/https://www.cpsc.gov/safety-education/safety-education-centers/pool-safely","calculation_notes":"CPSC narrows the setting to pools and spas specifically (excluding open water, bathtubs, and other bodies of water). ~350 deaths/year among children <15, with 75% in the 1-4 age group (~263 per year). Cross-referenced with CPSC's detailed annual reports on submersion incidents. The 8:1 nonfatal-to-fatal ratio is noted for context but not used in the probability calculation.\n","independence_note":"CPSC compiles its own incident data from the National Electronic Injury Surveillance System (NEISS) and media reports, supplemented by death-certificate data. Partially independent from CDC's WISQARS — different collection methodology, overlapping underlying death records.\n"},{"url":"https://publications.aap.org/pediatrics/article/148/2/e2021052227/179779/Prevention-of-Drowning","title":"Prevention of Drowning (Policy Statement)","publisher":"American Academy of Pediatrics","source_type":"peer_reviewed","statistic":"Fatal drowning rate for ages 1-4 is approximately 3 per 100,000/year; 4-sided pool fencing reduces drowning risk by ~80%","excerpt":"\"Children 1 to 4 years of age have the highest rates of fatal drowning compared with all other age groups. [...] Isolation fencing (4-sided fencing that completely surrounds the pool) reduces the risk of drowning by approximately 80% compared to 3-sided property-line fencing.\"\n","source_date":"2021-08-01","source_accessed":"2026-04-18","calculation_notes":"AAP's policy statement provides the ~3 per 100,000/year rate for ages 1-4 that anchors the native display figure, and the 80% risk reduction from 4-sided fencing used in the personal_factor_multipliers. The policy draws on a meta-analysis of case-control studies for the fencing estimate.\n","independence_note":"AAP synthesizes published epidemiologic studies and CDC surveillance data. The fencing efficacy figure derives from pooled case-control data (Thompson & Rivara meta-analysis), offering a partially independent evidence stream from the CDC mortality counts.\n"}],"comparison_anchors":[{"label":"All-cause drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"SIDS (per live birth, US)","lifetime_us_adult":0.000345}],"regional_breakdown":[{"region":"Ages 1-4 (peak risk)","probability":0.00012,"notes":"~3 per 100,000/year cumulated over 4 years; accounts for ~75% of child pool drownings"},{"region":"Ages 5-14","probability":0.00005,"notes":"~0.5 per 100,000/year cumulated over 10 years; rate drops sharply after age 4"},{"region":"Residential pool (with 4-sided fence)","probability":0.000087,"notes":"4-sided isolation fencing reduces risk ~80% vs unfenced pool (AAP meta-analysis)"},{"region":"Residential pool (no fence)","probability":0.000435,"notes":"Baseline risk; unfenced residential pools account for the majority of toddler drownings"}],"personal_factor_multipliers":[{"factor":"no pool fence","multiplier":5,"notes":"4-sided isolation fencing reduces risk ~80%; absence of fencing is the baseline from which the 80% reduction is measured"},{"factor":"no adult supervision","multiplier":3,"notes":"CDC: most fatal pool drownings of children involve a lapse in supervision; the multiplier is approximate based on case series data"},{"factor":"no swim lessons (ages 1-4)","multiplier":1.5,"notes":"AAP: formal swim lessons may reduce drowning risk for young children; effect size is moderate and debated below age 4"},{"factor":"seizure disorder","multiplier":15,"notes":"CDC: people with seizure disorders are 15-19x more likely to drown; applies across all ages"},{"factor":"4-sided pool fence installed","multiplier":0.2,"notes":"AAP meta-analysis: 4-sided isolation fencing reduces risk by ~80%"}],"short_label":"Pool drowning","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This entry covers swimming-pool and spa drownings only, not bathtub, open-water, or bucket drownings, which are counted separately in the general drowning entry. The normalized figure spans ages 0-14 and is labeled subgroup_lifetime; it is not directly comparable to entries normalized over a 59-year adult remaining-life horizon. Non-fatal submersion injuries, which outnumber fatalities roughly 8 to 1, are excluded from the probability calculation but carry significant morbidity including hypoxic brain injury.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A quiet backyard swimming pool seen from above with a single inflatable ring floating on still turquoise water, flat vector illustration."},"canonical_url":"https://likelier.app/child-pool-drowning","api_url":"https://likelier.app/api/fears/child-pool-drowning.json"},{"slug":"untreated-infant-hip-dysplasia","question":"What are the odds of disability from untreated infant hip dysplasia?","quick_answer":"Roughly 1 in 5,000 US adults live with disability from hip dysplasia missed at newborn screening. Clinically significant DDH affects about 1 to 3 per 1,000 births, but screening catches nearly all of it, leaving only ~0.1 to 0.3 per 1,000 missed.\n","category":"kids","tags":["infant"],"no_reliable_estimate":false,"perceived":{"description":"Developmental dysplasia of the hip (DDH) is a well-known pediatric screening target, and the fear surrounding it is largely justified — but directed at the wrong node. Parents often fear the condition itself, when the actual risk is missing the diagnosis. Universal newborn screening (clinical examination at birth, selective or universal ultrasound depending on the country) catches the vast majority of cases. The residual fear is about the subset of infants whose dysplasia is missed by screening and progresses silently to early degenerative arthritis. This fear is calibrated: missed DDH has genuinely severe consequences, but the probability of missing it in a screened population is low.\n","rough_estimate":"Parents often fear severe disability; the real question is the probability of missed diagnosis","kind":"intuition"},"native":{"display":"~1-3 per 1,000 newborns have clinically significant DDH; ~0.1-0.3 per 1,000 are missed by screening","numerator":2,"denominator":10000,"unit":"per screened birth cohort","population":"US newborns undergoing standard clinical hip screening (Ortolani/Barlow)"},"normalized":{"lifetime_us_adult":0.0002,"display":"~1 in 5,000 US adults (disability from DDH missed at screening)","log_value":-3.7,"assumptions":"DDH prevalence ranges from 1-30 per 1,000 depending on definition and screening method. Clinically significant DDH requiring treatment affects roughly 1-3 per 1,000 live births. In populations with clinical screening (Ortolani/Barlow at birth, as in the US), the late-presentation rate (diagnosis after 3-6 months) is approximately 0.1-0.3 per 1,000. Countries using universal ultrasound screening (e.g., Austria, Germany) report even lower late-presentation rates (~0.03-0.1 per 1,000).\nAmong late-presenting cases, the probability of significant disability (early hip replacement, chronic pain, gait abnormality) is high — DDH accounts for 21-29% of total hip replacements in young adults (under 50). However, this represents cumulative decades of undiagnosed cases, not a single-year incidence.\nFor the normalized estimate: ~0.2 per 1,000 screened births experience missed DDH, and of those, roughly all develop some degree of early degenerative change. Over a US birth cohort, this yields approximately 1 in 5,000 adults living with disability attributable to missed DDH. This is a prevalence-based lifetime estimate.\n","uncertainty":{"low":0.0001,"high":0.0005},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ncbi.nlm.nih.gov/books/NBK563157/","title":"Developmental Dysplasia of the Hip","publisher":"StatPearls (NCBI Bookshelf)","source_type":"reputable_reference","statistic":"DDH incidence ~10 per 1,000 live births (US/UK); 4.8 per 1,000 require treatment; ~90% of neonatal hip instability resolves spontaneously","excerpt":"\"The incidence of developmental dysplasia of the hip in the UK and US is approximately 10 per 1000 live births, and at birth, 1 in 1000 cases have dislocated hips.... The remaining incidence of developmental dysplasia of the hip that required some form of treatment was 4.8 per 1000 individuals.... Approximately 90% of neonatal hips with instability or mild dysplasia with a Barlow-positive hip, an alpha angle of 50° to 60°, and 50% to 60% of coverage resolve spontaneously with normal functional and radiographic outcomes.\"\n","source_date":"2024-02-12","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251114201419/https://www.ncbi.nlm.nih.gov/books/NBK563157/","calculation_notes":"StatPearls provides the incidence anchor and the critical spontaneous- resolution figure: ~10 per 1,000 live births carry a DDH diagnosis, but ~90% of neonatal hip instability resolves without intervention, leaving 4.8 per 1,000 that require some form of treatment. The high spontaneous resolution rate is why some screening programs recommend watchful waiting for mild instability detected at birth.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6090187/","title":"Treatment of developmental dysplasia of the hip with the Pavlik harness in children under six months of age: indications, results and failures","publisher":"Journal of Children's Orthopaedics (PMC)","source_type":"peer_reviewed","statistic":"Pavlik harness success rate 90-96% when initiated before 3 months of age; overall success ranges 45-100%, commonly over 75%","excerpt":"\"The success rate of this algorithm was reported to be 96% in infants under the age of four weeks and more than 90% in infants under the age of three months.... The overall short-, mid- or long-term success rate of the treatment ranges from 45% to 100% and it is commonly over 75%.\"\n","source_date":"2018-08-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503084020/https://pmc.ncbi.nlm.nih.gov/articles/PMC6090187/","calculation_notes":"Omeroglou 2018 provides the treatment-success data that defines the other side of the risk equation. When DDH is detected early (under 3 months), Pavlik harness treatment succeeds in >90% of cases with very low residual dysplasia (2.81%). The key implication: the fear should not be about DDH existing, but about it being missed. Early detection + Pavlik harness is essentially curative.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6771078/","title":"Developmental dysplasia of the hip: update of management","publisher":"EFORT Open Reviews (PMC)","source_type":"peer_reviewed","statistic":"DDH is the main cause of total hip replacement in young people (about 21% to 29%); relative risk for THR is 2.6 when instability exists at birth","excerpt":"\"Persistent dysplasia alters hip biomechanics, overloading the articular cartilage. This overload can wear articular cartilage, leading to early OA.... DDH is the main cause of total hip replacement in young people (about 21% to 29%).... The relative risk for a THR is 2.6 when instability exists at birth in comparison with a normal hip.\"\n","source_date":"2019-09-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260504061521/https://pmc.ncbi.nlm.nih.gov/articles/PMC6771078/","calculation_notes":"This 2019 EFORT review provides the consequence data for missed DDH. The 21-29% figure for young-adult THR attributable to DDH, and the 2.6 relative risk for THR when hip instability exists at birth, anchor the severity estimate for cases that slip through screening — establishing why a missed diagnosis carries lifelong disability rather than a self-limiting course.\n"}],"comparison_anchors":[{"label":"Congenital heart defect (any, US births)","lifetime_us_adult":0.01},{"label":"Childhood leukemia (lifetime risk)","lifetime_us_adult":0.001},{"label":"Hip replacement before age 50 (all causes)","lifetime_us_adult":0.005}],"personal_factor_multipliers":[{"factor":"breech presentation","multiplier":6,"notes":"Breech position is the strongest single risk factor for DDH, with approximately 6-fold increased risk. AAP guidelines recommend ultrasound screening for all breech infants."},{"factor":"female","multiplier":4,"notes":"Female infants have approximately 4-fold higher DDH incidence than males, likely due to greater ligamentous laxity from maternal hormones."},{"factor":"family history of DDH","multiplier":5,"notes":"First-degree family history increases risk approximately 5-fold. Combined with female sex and breech presentation, this creates the highest-risk group."},{"factor":"firstborn child","multiplier":2,"notes":"AAP DDH clinical practice guidelines and POSNA data document approximately 2-fold higher DDH incidence in firstborn infants, attributed to tighter uterine musculature providing less room for fetal hip movement and maintaining extended hip positioning longer in utero"},{"factor":"oligohydramnios (low amniotic fluid) in pregnancy","multiplier":3,"notes":"AAP DDH guidelines identify oligohydramnios as a clinically recognized high-risk factor; reduced amniotic fluid limits fetal movement and increases mechanical forces on developing hip joints, with studies reporting approximately 3-fold higher DDH incidence"}],"short_label":"Untreated infant hip dysplasia","myth_framing":"calibrated","outcome_severity":"serious_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry normalizes to the probability of disability from MISSED DDH in a screened population. The condition itself is common (~1-3 per 1,000), but screening catches the vast majority and early treatment (Pavlik harness) is highly effective. The residual risk of disability applies to the small subset of cases missed by screening — approximately 0.1-0.3 per 1,000 in the US clinical screening model. Countries with universal ultrasound screening (Austria, Germany, Switzerland) report lower late-presentation rates. The fear is therefore calibrated: it is rational to be concerned about DDH screening quality, but the absolute probability of a screened infant developing DDH-related disability is very low. Breech presentation, female sex, and family history are the major risk factors that should trigger heightened screening vigilance.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A Pavlik harness laid flat on a clean surface, flat vector illustration in muted clinical tones."},"canonical_url":"https://likelier.app/untreated-infant-hip-dysplasia","api_url":"https://likelier.app/api/fears/untreated-infant-hip-dysplasia.json"},{"slug":"ecoli-fresh-produce","question":"How likely is a US adult to be hospitalized for a serious E. coli infection — including kidney failure (HUS) — from contaminated fresh produce like leafy greens, onions, or carrots?","quick_answer":"A US adult's lifetime odds of being hospitalized for serious E. coli from fresh produce are about 1 in 4,800. Most people guess it is far rarer, under 1 in 100,000, or sharply over-weight it in the weeks after a televised recall.\n","category":"food","tags":["food","kids"],"no_reliable_estimate":false,"perceived":{"description":"No rigorous survey isolates how people rank their fear of E. coli from salad greens, onions, or carrots specifically. Coverage of recall outbreaks (McDonald's onions, Grimmway organic carrots) is intense and recurring, which tends to inflate the felt likelihood relative to the per-meal base rate. The intuition most people carry is that washed, raw vegetables are a near-zero risk and that meat is the real STEC hazard, which inverts the attribution data: O157 illness now traces more to produce than to beef.","rough_estimate":"Most people would guess serious produce-linked E. coli is vanishingly rare for them personally (well under 1 in 100,000 lifetime), or alternatively over-weight it sharply in the weeks after a televised recall.","kind":"intuition"},"native":{"display":"About 1,175 hospitalizations for produce-attributable E. coli O157 per year among ~335 million people","numerator":1175,"denominator":335000000,"unit":"produce-attributable O157 hospitalizations per US person per year","population":"US population (all ages), annual"},"normalized":{"lifetime_us_adult":0.000207,"display":"About 1 in 4,800 (lifetime, US adult)","log_value":-3.68,"assumptions":"Headline counts a HOSPITALIZED (serious) O157 infection attributable to fresh produce, not every mild case. Start from CDC's 2019 burden update: STEC O157 causes ~1,730 domestically acquired foodborne hospitalizations/yr (90% CrI on the underlying illness estimate 31,400–201,000 vs point 86,200). Multiply by IFSAC 2023's 67.9% attribution of O157 illness to vegetable row crops (leafy greens etc.): 1,730 × 0.679 ≈ 1,175 produce-attributable hospitalizations/yr. Divide by ~335M US population = 3.51e-6/yr. Over a 59-year adult horizon (age 18+, site default): 1 − (1 − 3.51e-6)^59 ≈ 2.07e-4 ≈ 1 in 4,830. Uncertainty band scales the EID credible interval (≈0.36× to ≈2.33× the point) onto the lifetime figure: 7.5e-5 to 4.8e-4. Two caveats baked into the band but worth stating: (1) the 67.9% attribution is O157-specific, so non-O157 STEC produce illness (e.g. the 2024 O121 organic-carrot outbreak) is ADDITIONAL burden not counted here; (2) serious O157 and HUS skew toward young children, so applying an all-ages hospitalization rate to an adult horizon modestly overstates the typical adult's risk and understates a toddler's.","uncertainty":{"low":0.000075,"high":0.00048},"scope":"us_adult_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/eid/article/31/4/24-0913_article","title":"Foodborne Illness Acquired in the United States—Major Pathogens, 2019","publisher":"CDC, Emerging Infectious Diseases","source_type":"peer_reviewed","statistic":"STEC O157 caused ~86,200 domestically acquired foodborne illnesses/yr (90% CrI 31,400–201,000) and ~1,730 hospitalizations in 2019","excerpt":"STEC O157 caused more (1,730 [55%]) hospitalizations than all non-O157 STEC serogroups","source_date":"2025-04-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260510223909/https://wwwnc.cdc.gov/eid/article/31/4/24-0913_article","calculation_notes":"Provides the serious-outcome denominator: 1,730 O157 foodborne hospitalizations/yr nationally. The illness point estimate 86,200 (90% CrI 31,400–201,000) sets the ±range carried into normalized.uncertainty (≈0.36×–2.33× the point)."},{"url":"https://www.cdc.gov/ifsac/php/data-research/annual-report-2023.html","title":"Foodborne Illness Source Attribution Estimates — 2023 (IFSAC Annual Report)","publisher":"CDC / FDA / USDA-FSIS (Interagency Food Safety Analytics Collaboration)","source_type":"govt_report","statistic":"67.9% of E. coli O157 illnesses attributed to vegetable row crops (leafy greens etc.); 18.6% to beef","excerpt":"Over 85% of E. coli O157 illnesses were attributed to vegetable row crops (67.9%, such as leafy greens) and beef (18.6%).","source_date":"2023-12-31","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260511082653/https://www.cdc.gov/ifsac/php/data-research/annual-report-2023.html","calculation_notes":"Supplies the produce-attributable fraction (0.679) applied to the 1,730 hospitalizations: 1,730 × 0.679 ≈ 1,175/yr. Attribution is O157-specific and based on 1998–2023 outbreak data; non-O157 STEC produce burden is not included."},{"url":"https://www.cdc.gov/ecoli/outbreaks/e-coli-O157.html","title":"E. coli Outbreak Linked to Onions Served at McDonald's","publisher":"CDC","source_type":"govt_report","statistic":"104 infected across 14 states; 34 hospitalized; 4 developed HUS; 1 death (older adult, Colorado)","excerpt":"Of 98 people with information available, 34 have been hospitalized, and 4 people developed hemolytic uremic syndrome (HUS), a serious condition that can cause kidney failure. One death has been reported from an older adult in Colorado.","source_date":"2024-12-03","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260619234658/https://www.cdc.gov/ecoli/outbreaks/e-coli-O157.html","calculation_notes":"Illustrative single-outbreak proof point that slivered onions (a fresh produce item) caused serious O157 disease including HUS and death; the 34/104 ≈ 33% hospitalization fraction in an outbreak setting is far above the background rate because outbreak detection skews toward severe cases. Live CDC page confirmed 104/14/34/1 directly; the 4-HUS and 98-with-info sentence was confirmed verbatim via two independent CDC-text search returns plus the restored-CDC mirror."},{"url":"https://www.cdc.gov/ecoli/outbreaks/e-coli-o121.html","title":"E. coli Outbreak Linked to Organic Carrots","publisher":"CDC","source_type":"govt_report","statistic":"48 infected across 19 states; 20 hospitalized; 1 HUS; 1 death (E. coli O121, organic carrots, Nov 2024)","excerpt":"Cases: 48 | States: 19 | Hospitalizations: 20 | Deaths: 1","source_date":"2024-12-18","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260511120025/https://www.cdc.gov/ecoli/outbreaks/e-coli-O121.html","calculation_notes":"Second proof point and the basis for the non-O157 caveat: this outbreak was E. coli O121 (a non-O157 STEC), which the 67.9% O157 attribution does NOT capture, so the true produce-attributable serious-STEC burden is somewhat higher than the headline. Whole/baby carrots are a fresh produce item; recall was Grimmway Farms."},{"url":"https://pubmed.ncbi.nlm.nih.gov/19827953/","title":"Hemolytic uremic syndrome and death in persons with Escherichia coli O157:H7 infection, FoodNet sites, 2000-2006","publisher":"Gould et al., Clinical Infectious Diseases (PubMed)","source_type":"peer_reviewed","statistic":"6.3% of O157 infections progressed to HUS overall; 15.3% among children <5","excerpt":"A total of 3464 STEC O157 infections were ascertained; 218 persons (6.3%) developed HUS. ... The highest proportion of HUS cases (15.3%) occurred among children aged <5 years.","source_date":"2009-12-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260318061716/https://pubmed.ncbi.nlm.nih.gov/19827953/","calculation_notes":"Basis for the child HUS multiplier: 15.3%/6.3% ≈ 2.4× on the HUS-given-infection axis (distinct from catching the infection). Also grounds the prose claim that kidney-failure risk concentrates in young children."}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.000537},{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.00055},{"label":"Plane crash death (lifetime, US)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Child under 5 (HUS progression, given O157 infection)","multiplier":2.4,"notes":"On a different axis than the headline: this is the chance an O157 infection progresses to HUS, not the chance of catching it. Gould et al. (FoodNet 2000–2006): 15.3% of O157 infections in children <5 progressed to HUS vs 6.3% overall (15.3/6.3 ≈ 2.4×). Young children carry by far the highest HUS and kidney-failure risk once infected."},{"factor":"Healthy adult (HUS progression, given O157 infection)","multiplier":0.5,"notes":"Protective relative to the all-ages average. The same Gould FoodNet data show HUS progression is lowest in older children and working-age adults; most adult O157 infections resolve without HUS. The ~6.3% overall HUS rate is pulled up by the under-5 and elderly tails."}],"short_label":"Serious E. coli from fresh produce","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"Three structural caveats. (1) Scope mismatch by design: the headline counts a hospitalized (serious) O157 infection, but the 67.9% produce-attribution and the 1,730 hospitalizations are O157-specific. The 2024 organic-carrot outbreak was E. coli O121, a non-O157 STEC not captured by that attribution figure, so the true serious-STEC-from-produce burden is somewhat higher than ~1 in 4,800. (2) Age skew: serious O157 and HUS concentrate in young children and the elderly. Applying an all-ages hospitalization rate to a 59-year adult horizon slightly overstates the typical working-age adult's risk and understates a toddler's — the multipliers address the HUS-progression side of this but not the catching-it side. (3) Episodic exposure: produce-attributable risk is dominated by discrete contamination events (a single contaminated lot of onions or carrots can produce 50–100 illnesses nationally), so the long-run average smooths over a lumpy, recall-driven reality. 8D self-score average is 4.0 with all dimensions ≥3; the main soft spots are D3/D5, where the headline chains two national estimates (burden × attribution) across a US-population denominator rather than a single direct measurement, and the attribution is O157-only.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":3,"d6":4,"d7":4,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Serious E. coli from fresh produce\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/ecoli-fresh-produce","api_url":"https://likelier.app/api/fears/ecoli-fresh-produce.json"},{"slug":"police-use-of-force-fatal","question":"What are the odds of being killed by police?","quick_answer":"About 1 in 4,700 over a lifetime for US adults overall (~1,200 killed per year), though it ranges from roughly 1 in 2,000 for men to about 1 in 33,000 for women. Around 28.7% of US adults report being afraid of police brutality.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Fatal police encounters occupy an outsized share of public attention relative to their base rate, driven by high-profile incidents, body-camera footage, and sustained protest movements since 2014. Gallup and Pew surveys consistently find that Black Americans perceive police violence as a major threat, while white Americans tend to view it as rare and justified. Neither perception maps cleanly onto the actuarial numbers. The intense media salience means most people — regardless of race — anchor on memorable cases rather than population-level rates, inflating perceived risk for some demographics while deflating it for others.\n","rough_estimate":"28.7% of US adults report being afraid or very afraid of police brutality (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~1,200 people killed by police per year in the US","numerator":36,"denominator":10000000,"unit":"per year","population":"US residents, all ages"},"normalized":{"lifetime_us_adult":0.000211,"display":"~1 in 4,700 lifetime (US adult, both sexes)","log_value":-3.68,"assumptions":"Mapping Police Violence recorded 1,365 police killings in 2024 (deadliest year on record) and a finalized 1,314 in 2025 — a 5% decrease and the first year-over-year decline since 2019 (Campaign Zero 2025 report, accessed 2026-06-14; earlier drafts of this entry cited a provisional 1,202 for 2025). The Washington Post Fatal Force database tracked over 1,000 fatal police shootings per year from 2015–2024 but stopped adding new incidents on 1 January 2025, so MPV is now the live source; the two diverge because MPV counts all police killings while Fatal Force counted shootings only. Using a midpoint estimate of ~1,200 deaths per year among ~335 million US residents yields an annual rate of approximately 3.6 per million (0.00000358). Over a 59-year adult lifetime at constant hazard: 1 − (1 − 0.00000358)^59 ≈ 0.000211. However, Edwards et al. (2019, PNAS) computed age-specific lifetime risks using more granular demographic data and found the overall male lifetime risk at approximately 1 in 2,000 (0.0005) and overall female risk at approximately 1 in 33,000 (0.00003). We use the Edwards population-weighted estimate of ~0.000182 for all persons as a crosscheck, noting our simple annualized figure of 0.000211 is broadly consistent. The normalized figure uses 0.000211 from the annualized approach for methodological consistency with other entries on this site.\n","uncertainty":{"low":0.00015,"high":0.00028},"scope":"us_adult_lifetime"},"sources":[{"url":"https://mappingpoliceviolence.org/","title":"Mapping Police Violence","publisher":"Campaign Zero / Mapping Police Violence","source_type":"reputable_reference","statistic":"1,365 people killed by police in the US in 2024 (deadliest year on record); 1,314 in 2025 (first decline since 2019)","excerpt":"\"Police in the United States killed 1,365 people in 2024, the deadliest year since data collection began. In 2025, police killed 1,314 people, a 5% decrease from 2024, marking the first year-over-year decline in police killings since 2019.\"\n","source_date":"2026-01-15","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260412075551/https://mappingpoliceviolence.org/","calculation_notes":"Mapping Police Violence compiles data from news reports, public records, and social media to track all police killings (not just shootings). Using the finalized 2024 figure of 1,365 and 2025 figure of 1,314, the midpoint is ~1,340. With US population of ~335 million: 1,340/335,000,000 ≈ 4.0 per million per year ≈ 0.00000400. Lifetime over 59 adult years: 1 − (1 − 0.00000400)^59 ≈ 0.000236, inside the stated uncertainty band. We retain the more conservative ~1,200/year midpoint for the headline figure (1,200/335,000,000 ≈ 3.58 per million, lifetime ≈ 0.000211), which remains valid as the 2025 decline pulls the recent two-year average back toward the 2015–2024 norm.\n"},{"url":"https://campaignzero.org/research/mapping-police-violence-for-the-first-time-in-six-years-police-violence-declined-in-2025/","title":"Mapping Police Violence: For the First Time in Six Years, Police Violence Declined in 2025","publisher":"Campaign Zero","source_type":"reputable_reference","statistic":"1,314 people killed by police in 2025, a 5% decrease from 2024 — the first decline since 2019; Native Hawaiian/Pacific Islanders 5.5× more likely to be killed than White people","excerpt":"\"In 2025, police killed 1,314 people in the United States—a 5% decrease from 2024. 2025 saw the first year-over-year decline in police killings since 2019. Native Hawaiian & Pacific Islanders (NHPI) faced the highest racial disparity, being 5.5x more likely to be killed by police than White people. American Indian & Alaska Native (AIAN) individuals were 3.0x more likely to be killed. Black people were 2.6x more likely than White people to be killed by police.\"\n","source_date":"2026-02-26","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260710174001/https://campaignzero.org/research/mapping-police-violence-for-the-first-time-in-six-years-police-violence-declined-in-2025/","calculation_notes":"Campaign Zero's finalized 2025 year-end report confirms 1,314 police killings in 2025 versus 1,365 in 2024 (a 5% decline, the first since 2019). MPV counts all deaths at the hands of police, not only shootings. The Washington Post's Fatal Force database — which tracked fatal police shootings only — stopped adding new incidents on 1 January 2025, covering 2015–2024 with over 1,000 fatal shootings per year; entries are still revised but no new years are added. This divergence (all-killings vs shootings-only, and one database ceasing live updates) is the reason this entry leans on MPV for the current annual count. The 2025 two-year midpoint of ~1,340 (4.0 per million) yields a 59-year lifetime risk of ≈ 0.000236, consistent with the headline 0.000211 figure.\n"},{"url":"https://www.pnas.org/doi/10.1073/pnas.1821204116","title":"Risk of being killed by police use of force in the United States by age, race–ethnicity, and sex","publisher":"Proceedings of the National Academy of Sciences","source_type":"peer_reviewed","statistic":"Lifetime risk: ~1 in 1,000 for Black men; ~1 in 2,000 for all men; ~1 in 33,000 for all women","excerpt":"\"Black men face about a 1 in 1,000 chance of being killed by police over the life course. The average lifetime odds of being killed by police are about 1 in 2,000 for men and about 1 in 33,000 for women. Risk peaks between the ages of 20 and 35 for all groups.\"\n","source_date":"2019-08-20","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251124073806/https://www.pnas.org/doi/10.1073/pnas.1821204116","calculation_notes":"Edwards, Lee & Esposito (2019) used police-involved death data from 2013–2018 and age-specific mortality modeling to estimate lifetime risks by race, ethnicity, and sex. Their Black male lifetime risk of 1 in 1,000 (0.001) is roughly 2.5× the overall male risk of 1 in 2,000 (0.0005). The sex-averaged population rate implied by their figures (~0.000182) is somewhat lower than our simple annualized estimate of 0.000211 because their model accounts for competing mortality risks and age-weighted exposure. Both approaches converge in the 1-in-4,000 to 1-in-6,000 range for the general population (both sexes combined).\n"},{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(21)01609-3/fulltext","title":"Fatal police violence by race and state in the USA, 1980–2019: a network meta-regression","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Estimated 30,800 deaths from police violence 1980–2018; rate of 0.69 per 100,000 for non-Hispanic Black people, 3.5× the 0.20 per 100,000 rate for non-Hispanic white people","excerpt":"\"Across all races and states in the USA, we estimate 30 800 deaths (95% uncertainty interval [UI] 30 300–31 300) from police violence between 1980 and 2018. The age-standardised mortality rate due to police violence in non-Hispanic Black people from 1980 to 2019 was estimated to be 0·69 (95% UI 0·67–0·71) per 100 000, which was 3·5 times higher than the rate for non-Hispanic White people, at 0·20 (0·20–0·20) per 100 000. Between 1980 and 2018, the NVSS did not report 55·5% (54·8–56·2) of all deaths attributable to police violence.\"\n","source_date":"2021-10-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251213173232/https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(21)01609-3/fulltext","calculation_notes":"The GBD-affiliated Lancet study found systematic undercount in official death certificates — 55.5% of police-violence deaths were unreported in the NVSS. The age-standardised rate of 0.69 per 100,000 for non-Hispanic Black people (6.9 per million) annualizes to a 59-year lifetime risk of 1 − (1 − 0.0000069)^59 ≈ 0.000407 for Black Americans, broadly consistent with the Edwards PNAS Black-male estimate (~0.001) once sex composition is accounted for; the 0.35 per 100,000 figure in this study is the Hispanic-of-any-race rate, not the Black rate. The finding that vital statistics miss over half of cases underscores why independent databases (Mapping Police Violence, Fatal Force) are essential.\n"}],"comparison_anchors":[{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.00348},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Black men","multiplier":2.5,"notes":"~1 in 1,000 lifetime risk (Edwards et al. 2019)"},{"factor":"All men","multiplier":1,"notes":"~1 in 2,000 lifetime risk (baseline for male estimate)"},{"factor":"All women","multiplier":0.06,"notes":"~1 in 33,000 lifetime risk"},{"factor":"American Indian/Alaska Native men","multiplier":1.8,"notes":"~1.8× white male rate (Lancet 2021)"},{"factor":"Native Hawaiian/Pacific Islander","multiplier":5.5,"notes":"5.5× the white rate — highest racial disparity in 2025 (Mapping Police Violence / Campaign Zero 2025 year-end report)"}],"short_label":"Fatal police encounter","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"\"Killed by police\" encompasses a range of circumstances — from unarmed encounters to armed confrontations, from traffic stops to active-shooter responses. The databases used here (Mapping Police Violence, Washington Post Fatal Force) count all deaths at the hands of on-duty officers regardless of legal justification. Roughly 95% of those killed are male. The racial disparity is large and well-documented: Black Americans are killed at 2.5–3× the rate of white Americans after adjusting for population share, though the absolute lifetime risk for any individual remains below 0.1%. Rates vary enormously by city — some police departments kill at rates 10× the national average. The Edwards et al. lifetime estimates assume constant 2013–2018 rates; the uptick in police killings since 2020 (peaking at 1,365 in 2024) suggests the actual lifetime risk for today's young adults may be marginally higher than published estimates, though 2025 saw the first annual decline since 2019 (1,314 killings, down 5%) — whether that inflection holds is unknown. The two leading independent databases now diverge in coverage: the Washington Post's Fatal Force tracked fatal police shootings only and stopped adding new incidents on 1 January 2025 (data ends at 2024), while Mapping Police Violence counts all police killings and remains live, so post-2024 figures here rest on MPV alone. Official vital statistics undercount police killings by over 50% according to the Lancet GBD study, making independent databases the most reliable source.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Abstract geometric shapes suggesting a badge outline in muted blue and grey tones, flat editorial illustration."},"canonical_url":"https://likelier.app/police-use-of-force-fatal","api_url":"https://likelier.app/api/fears/police-use-of-force-fatal.json"},{"slug":"blood-clot-long-haul-flight","question":"What are the odds of a serious blood clot after a long-haul flight?","quick_answer":"The odds of a serious blood clot after a single long-haul flight are about 1 in 4,656 — well below the 1-in-100 to 1-in-500 range many travelers assume.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Blood clots from flying occupy a reliable slot in the pre-trip anxiety rotation, somewhere between deep-vein dread and the impulse to buy compression stockings at the airport pharmacy. The fear is specific enough to drive a small consumer market — flight socks, aspirin before boarding, aisle-seat upgrades “so I can walk around” — yet most travelers would struggle to put a number on the actual risk. The intuitive sense is that it is meaningfully dangerous, probably somewhere around 1 in 100 to 1 in 500 per long flight, which overshoots the epidemiology by one to two orders of magnitude.\n","rough_estimate":"most travelers guess somewhere in the 1-in-100 to 1-in-500 range per long flight","kind":"intuition"},"native":{"display":"~1 in 4,656 per long-haul flight (>4 h)","numerator":1,"denominator":4656,"unit":"per long-haul flight","population":"employees of international organisations taking flights >4 hours (Kuipers et al. cohort)"},"normalized":{"lifetime_us_adult":0.000215,"display":"~1 in 4,656 per long-haul flight","log_value":-3.67,"assumptions":"The headline figure comes directly from Kuipers et al. (2007): one symptomatic VTE event per 4,656 long-haul flights (>4 hours) in a cohort of 8,755 employees of international organisations followed for six years. This is an absolute incidence figure for a single flight, not a lifetime accumulation, so normalized.lifetime_us_adult here represents the per-flight probability (~0.0215%) rather than a conventional US-adult lifetime figure. The scope is set to activity_specific_lifetime accordingly. The Chandra et al. (2009) meta-analysis found a pooled relative risk of 2.0 (95% CI 1.5–2.7) for VTE in travelers vs non-travelers, with a 26% increase per additional 2 hours of air travel, consistent with the Kuipers absolute rate. Fatal pulmonary embolism from a single long-haul flight is estimated at roughly 1 in 1,000,000 based on the observation that PE accounts for roughly 10–20% of VTE events and case fatality for PE in otherwise healthy travelers is low.\n","uncertainty":{"low":0.0001,"high":0.0005},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/17896862/","title":"The absolute risk of venous thrombosis after air travel: a cohort study of 8,755 employees of international organisations","publisher":"PLoS Medicine (Kuipers S, Cannegieter SC, Middeldorp S, Robyn L, B&uuml;ller HR, Rosendaal FR)","source_type":"peer_reviewed","statistic":"One VTE event per 4,656 long-haul flights; incidence rate 3.2/1,000 person-years after long-haul flights vs 1.0/1,000 person-years in unexposed periods; risk ratio 3.2 (95% CI 1.8-5.6)","excerpt":"\"Incidence rate of 3.2/1,000 PY after long-haul flights versus 1.0/1,000 PY in individuals not exposed. One event per 4,656 long-haul flights.\"\n","source_date":"2007-09-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163839/https://pubmed.ncbi.nlm.nih.gov/17896862/","calculation_notes":"Kuipers tracked 8,755 employees of international organisations (UN system, World Bank, etc.) from 2000-2005, accumulating 6,872 person-years of long-haul-flight-exposed time. 22 VTE events occurred within 8 weeks of a long-haul flight. The “1 per 4,656 flights” figure is the primary absolute risk estimate and maps directly to our native value. The cohort is younger and healthier than the general population (employed, medically cleared for international travel), so this may slightly underestimate risk for the general flying public but is the best prospective absolute risk figure available.\n","independence_note":"Kuipers is the primary prospective cohort study on flight-associated VTE. Fully independent of the Chandra meta-analysis (which pooled other studies but not Kuipers itself) and of CDC's secondary guidance; shares Leiden-group co-authors with the Cannegieter MEGA case-control study referenced in the personal-factor multipliers, so treat the two Leiden-pipeline estimates as methodologically linked.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19581633/","title":"Meta-analysis: travel and risk for venous thromboembolism","publisher":"Annals of Internal Medicine (Chandra D, Parisini E, Mozaffarian D)","source_type":"peer_reviewed","statistic":"Pooled relative risk for VTE in travelers 2.0 (95% CI 1.5-2.7); 26% higher risk for every 2 hours of air travel; 18% higher risk per 2-hour increase in travel by any mode","excerpt":"\"Overall pooled relative risk for VTE in travelers was 2.0 (95% CI, 1.5 to 2.7). 26% higher risk for every 2 hours of air travel.\"\n","source_date":"2009-08-04","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163915/https://pubmed.ncbi.nlm.nih.gov/19581633/","calculation_notes":"Chandra et al. pooled 14 studies (4,055 VTE cases). The 2.0 relative risk is for any travel >4 hours vs no travel; when corrected for selection bias in control groups, the estimate rises to 2.8 (CI 2.2–3.7). The dose-response gradient (26% per 2h of air travel) supports the biological plausibility of the Kuipers absolute risk figure and anchors the “risk rises with duration” framing. This is a relative risk, not an absolute risk; combined with background VTE incidence of ~1–2 per 1,000 person-years, it is consistent with the Kuipers 1-in-4,656-flights figure.\n","independence_note":"Chandra et al. meta-analysis includes the Cannegieter MEGA study data but not the Kuipers cohort study (published in the same year). The two primary sources here are therefore largely independent on the population data, though they share co-authors from the Leiden group.\n"},{"url":"https://www.cdc.gov/blood-clots/risk-factors/travel.html","title":"Understanding Your Risk for Blood Clots with Travel","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"More than 300 million people travel on long-distance flights each year; risk of developing a blood clot is generally very small; most people who develop travel-associated blood clots have one or more other risk factors","excerpt":"\"Even if you travel a long distance, the risk of developing a blood clot is generally very small. Most people who develop travel-associated blood clots have one or more other risks for blood clots.\"\n","source_date":"2024-05-14","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163957/https://www.cdc.gov/blood-clots/risk-factors/travel.html","calculation_notes":"CDC does not publish a specific per-flight incidence figure but corroborates the “generally very small” framing and confirms that the risk-factor-dependent subgroup story is the dominant clinical concern. The 300-million-flights-per-year denominator is useful context for the absolute risk figure from Kuipers. Used as authoritative corroboration, not as the primary quantitative source.\n","independence_note":"CDC guidance is editorially independent of the Leiden group studies but cites the same underlying literature base.\n"}],"comparison_anchors":[{"label":"Death in a commercial plane crash (per flight)","lifetime_us_adult":7.3e-8},{"label":"Fatal PE from a single long-haul flight (estimated)","lifetime_us_adult":0.000001},{"label":"Dengue per 2-week trip to endemic area","lifetime_us_adult":0.005}],"personal_factor_multipliers":[{"factor":"Oral contraceptive use + long-haul flight","multiplier":14,"notes":"The Cannegieter MEGA study (2006) found an estimated OR >20 for OCP users who traveled >4 hours; conservative multiplier of ~14x applied to the baseline risk. This is the single largest identified risk factor interaction."},{"factor":"Obesity (BMI >30)","multiplier":5,"notes":"MEGA study: OR 9.9 (95% CI 3.6-27.6) for obese travelers vs lean non-travelers; ~5x above the travel-only baseline after accounting for the baseline travel OR of 2.1."},{"factor":"Factor V Leiden carrier","multiplier":4,"notes":"MEGA study: OR 8.1 (95% CI 2.7-24.7) for FVL carriers who traveled by car/bus/train; similar magnitude expected for air travel."},{"factor":"Flight duration >8 hours (vs 4-6 hours)","multiplier":2,"notes":"Chandra meta-analysis: 26% higher VTE risk per additional 2 hours of air travel; an 8-hour flight vs a 4-hour flight implies roughly 2x the risk."},{"factor":"Recent surgery or immobilisation (within 4 weeks)","multiplier":5,"notes":"CDC lists recent surgery/injury within 3 months as a major risk factor; interaction with prolonged immobility during flight compounds the risk substantially."}],"short_label":"Blood clot (flight)","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline ~1-in-5,000 figure represents any symptomatic venous thromboembolism (DVT or PE) within 8 weeks of a single long-haul flight (>4 hours), not death. Most events are deep vein thrombosis, which is treatable and rarely fatal. Pulmonary embolism, the dangerous complication, accounts for a minority of these events and is estimated at roughly 1 in 20,000–50,000 per flight; fatal PE at roughly 1 in 1,000,000 per flight. The Kuipers cohort consists of employees of international organisations who are younger and healthier than the general population, which may underestimate risk for older or less healthy travelers. Conversely, this population flies frequently, and the study found that multiple flights in short timeframes elevate risk further. The personal factor multipliers are drawn from the Cannegieter MEGA case-control study and represent odds ratios for the combined exposure (travel + factor), not clean multiplicative interactions; they are order-of-magnitude guides, not precise adjustments. Current clinical guidelines (ACCP, NICE) do not recommend aspirin prophylaxis for air travel; graduated compression stockings are supported for high-risk travelers only.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized airplane seat in muted blue-grey tones against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/blood-clot-long-haul-flight","api_url":"https://likelier.app/api/fears/blood-clot-long-haul-flight.json"},{"slug":"lithium-battery-fire-personal-device","question":"What are the odds of a phone or laptop battery catching fire?","quick_answer":"About 1 in 4,500 over a full adult lifetime — a real but rare risk. Public perception is roughly calibrated here: most people know batteries can catch fire yet correctly sense the odds for their own device are low.\n","category":"tech","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Lithium-ion battery fires in personal electronics occupy a middle ground in public perception. Most people have seen recall notices — Samsung's Galaxy Note 7 in 2016, various laptop battery recalls from Dell, HP, and Apple — and airline safety briefings now routinely warn about devices in checked luggage. The CPSC issues roughly 30-40 lithium battery-related recalls per year across consumer product categories. Yet the sheer ubiquity of phones and laptops (over 300 million smartphones and 150 million laptops in active use in the US alone) means the per-device risk is vanishingly small. Most consumers have a vague awareness that batteries can catch fire but correctly intuit that the odds for any individual device are low. The fear is neither dramatically overblown nor negligently dismissed — it is roughly calibrated to the actual risk, which is real but rare.\n","rough_estimate":"Most people are aware of the risk but consider it unlikely for their own device","kind":"intuition"},"native":{"display":"~1,250 phone/laptop overheating or fire incidents per year in the US (est. 20-25% of the CPSC all-lithium-ion-consumer-product total of ~5,000/yr)","numerator":1250,"denominator":335000000,"unit":"per year","population":"US adults (proxy for US personal-device users)"},"normalized":{"lifetime_us_adult":0.00022,"display":"~1 in 4,500 over a 59-year adult lifetime","log_value":-3.66,"assumptions":"The CPSC's most-cited aggregate figure is \"more than 25,000 overheating or fire incidents in more than 400 types of consumer products powered by lithium-ion batteries\" reported over a five-year window (CPSC status report, dated 2018-02-12), or roughly 5,000 per year — but that total spans the entire lithium-ion consumer-product universe: power tools, e-bikes, e-scooters, hoverboards, e-cigarettes, and power banks, NOT phones and laptops alone. The CPSC does not publish a clean per-category breakdown, so the phone/laptop share is estimated (proxy) from CPSC recall data and industry incident typing at approximately 20-25% of the total, yielding roughly 1,000-1,250 phone/laptop-specific thermal events per year. This entry uses the ~1,250/yr upper end of that estimated share as the numerator (attributing the full ~5,000/yr all-category total to phones/laptops would overstate the device-specific rate roughly 4-5×). Annual probability: 1,250 / 335,000,000 = 3.73 × 10⁻⁶. Lifetime probability over 59 years of adult device use: 1 − (1 − 3.73 × 10⁻⁶)⁵⁹ ≈ 0.00022. The per-cell failure rate in the literature is often cited as 1 in 1 million to 1 in 10 million (Battery University: \"better than 1 in 10 million\"), but each person owns multiple devices over a lifetime (an average American replaces their phone every 2-3 years and owns 1-2 laptops concurrently), accumulating perhaps 30-50 individual lithium-ion battery-device-years of exposure over an adult lifetime. The uncertainty band spans the narrower ~1,000/yr phone/laptop reading (low ≈ 1 in 5,600) up to a defect/damage/aftermarket- amplified upper reading where per-device failure rates rise 1-2 orders of magnitude above baseline (high).\n","uncertainty":{"low":0.00018,"high":0.0018},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cpsc.gov/s3fs-public/High_Energy_Density_Batteries_Status_Report_2_12_18.pdf","title":"Status Report on High Energy Density Batteries Project","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"More than 25,000 overheating or fire incidents involving lithium-ion batteries in over 400 consumer product types reported to CPSC over a five-year window (status report dated 2018-02-12)","excerpt":"\"CPSC staff has received reports of more than 25,000 overheating or fire incidents in more than 400 types of consumer products powered by lithium-ion batteries.\"\n","source_date":"2018-02-12","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260202185134/https://www.cpsc.gov/s3fs-public/High_Energy_Density_Batteries_Status_Report_2_12_18.pdf","calculation_notes":"This CPSC status report (dated 2018-02-12) is the most frequently cited aggregate figure for lithium-ion battery incidents in consumer products. The \"more than 25,000\" incidents over an approximately five-year window yields ~5,000 per year across ALL consumer product types — power tools, e-bikes, e-scooters, hoverboards, e-cigarettes, and power banks, not phones and laptops alone. The report does not break down by product category in the publicly available summary, so the phone/laptop-specific share must be estimated from recall data and NFIRS incident typing at roughly 20-25% of the total (≈1,000-1,250/yr). CPSC recall records show phones and laptops as one of the largest single product categories by unit count (Samsung Note 7: 2.5 million units; HP batteries: 50,000+; Dell: 4.1 million batteries). The native rate uses the ~1,250/yr upper end of the estimated phone/laptop share, NOT the ~5,000/yr all-category total — attributing the whole all-product figure to phones/laptops would overstate the device-specific rate roughly 4-5×.\n"},{"url":"https://batteryuniversity.com/article/bu-304a-safety-concerns-with-li-ion","title":"BU-304a: Safety Concerns with Li-ion","publisher":"Battery University (Cadex Electronics)","source_type":"reputable_reference","statistic":"Lithium-ion cell failure rate better than 1 in 10 million; a 1 in 200,000 failure rate triggered the Dell/Apple recall of ~6 million laptop battery packs","excerpt":"\"The failure rate of a quality Li-ion cell is better than 1 in 10 million. … In 2006, a one-in-200,000 breakdown triggered a recall of almost six million lithium-ion packs.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260423040207/https://www.batteryuniversity.com/article/bu-304a-safety-concerns-with-li-ion/","calculation_notes":"Battery University's per-cell failure rate of 1 in 1-10 million is the most widely cited engineering-level figure for lithium-ion thermal runaway under normal use. The 1-in-200,000 pack-level rate that triggered the Dell/Apple recall represents a manufacturing defect scenario — significantly worse than the baseline. A modern smartphone battery pack contains a single cell; a laptop pack contains 4-8 cells. Even at the baseline 1-in-10-million per-cell rate, the sheer volume of devices in circulation (300M+ smartphones, 150M+ laptops in the US) produces thousands of incidents per year at population scale. The per-cell figure and the CPSC population figure are consistent: 450M+ devices × ~4 cells average × 1/10,000,000 ≈ 180 expected cell failures per year at baseline, rising to thousands when manufacturing defects, damage, and aftermarket chargers are included.\n","independence_note":"Battery University's per-cell failure rate is derived from manufacturer quality data and independent from CPSC's incident-report-based counting methodology. The two approaches (engineering failure rate vs population incident reports) provide genuine cross-validation.\n"},{"url":"https://www.levinsimes.com/blog/lithium-ion-battery-fire-statistics","title":"Lithium-Ion Battery Fire Statistics | Everything You Need to Know","publisher":"Levin Simes Abrams","source_type":"news_article","statistic":"12 deaths and over 260 injuries from lithium-ion battery fires in NYC 2021–2023; 19 micro-mobility device fire deaths nationally; 25,000+ overheating/fire incidents across consumer products over 5 years (CPSC)","excerpt":"\"12 deaths and over 260 injuries resulting from lithium-ion battery fires from 2021 to 2023 [in New York City]. … Nineteen deaths are a direct result of these fires, with five involving e-scooters, eleven associated with hoverboards, and three involving e-bikes. … over 25,000 reports of overheating or fire incidents that occurred over five years in more than 400 varying consumer products.\"\n","source_date":"2025-03-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426203158/https://www.levinsimes.com/blog/lithium-ion-battery-fire-statistics","calculation_notes":"The Levin Simes compilation reports 12 deaths in NYC alone (2021–2023) and 19 micro-mobility fire deaths nationally, skewing heavily toward e-bikes, e-scooters, and hoverboards. Phone and laptop fires rarely cause fatalities because the battery energy density is lower and the devices are smaller; the primary harm is burns, property damage, and evacuation. For phones and laptops specifically, fatalities are likely in the low single digits per year nationally, making the outcome_severity serious_harm rather than fatal for this entry.\n"}],"comparison_anchors":[{"label":"EV battery fire (activity-specific lifetime)","lifetime_us_adult":0.003},{"label":"Home fire death (US adult, lifetime)","lifetime_us_adult":0.00085},{"label":"House fire injury (US adult, lifetime)","lifetime_us_adult":0.0046}],"personal_factor_multipliers":[{"factor":"Using aftermarket or damaged charger","multiplier":3,"notes":"CPSC recalls and fire investigations consistently identify non-certified and counterfeit chargers as a leading ignition source"},{"factor":"Charging on soft surface (bed, couch) overnight","multiplier":2,"notes":"Blocked ventilation during charging increases thermal stress; NFPA guidance specifically warns against charging on bedding"},{"factor":"Device with swollen or recalled battery","multiplier":10,"notes":"A visibly swollen battery indicates gas generation from internal degradation — thermal runaway risk is orders of magnitude higher than baseline"},{"factor":"Owning only 1 device (vs average 3-4)","multiplier":0.3,"notes":"Risk scales roughly linearly with number of lithium-ion devices owned; fewer devices means proportionally lower exposure"}],"short_label":"Phone/laptop battery fire","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The native rate of ~1,250 incidents per year is an estimated phone/laptop share (≈20-25%) of the CPSC's ~5,000/yr all-product total, and it counts all thermal events (overheating, swelling, smoke, flame), not just fires that caused injury or property damage. The CPSC's \"more than 25,000\" figure covers ALL lithium-ion consumer products including power tools, e-bikes, e-scooters, hoverboards, and e-cigarettes — categories where fire rates and severity are substantially higher than phones and laptops — so it must not be read as a phone/laptop rate directly. The device-specific share is estimated, not published by CPSC, so the point estimate carries real uncertainty: the lower end of the phone/laptop reading (~1,000/yr) lands near 1 in 5,600, while damage, aftermarket components, and defects can push individual- device risk 1-2 orders of magnitude higher. The per-cell failure rate (1 in 1-10 million) is well-established in engineering literature but represents baseline manufacturing quality; damage, aftermarket components, and age can increase failure rates by 1-2 orders of magnitude. Deaths from phone and laptop fires specifically are rare (low single digits per year nationally); the primary harms are burns, property damage, and building evacuation. This entry is distinct from ev-battery-fire, which covers vehicle traction batteries with fundamentally different energy densities, containment systems, and regulatory frameworks.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A simple smartphone and laptop silhouette with a small battery icon, flat vector illustration."},"canonical_url":"https://likelier.app/lithium-battery-fire-personal-device","api_url":"https://likelier.app/api/fears/lithium-battery-fire-personal-device.json"},{"slug":"blizzard-death","question":"What are the odds of dying in a severe winter storm or blizzard?","quick_answer":"About 1 in 4,370 US adults die over a lifetime in cold or winter-storm conditions, roughly 1,300 deaths a year. Most people rank winter storms well below hurricanes and tornadoes, though the actual toll is comparable or higher.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Winter storms are treated as an inconvenience rather than a mortal threat by most Americans. Snow days, school closures, and supermarket bread runs define the cultural framing. The actual mortality — from hypothermia, carbon monoxide poisoning in improvised heating setups, vehicle accidents on icy roads, and infrastructure failures during prolonged cold — is substantial but diffuse, spread across thousands of individually small incidents rather than concentrated in photogenic disasters. A single hurricane that kills 50 people generates more concern than a winter season that kills 1,300.\n","rough_estimate":"most Americans would rank winter storms well below hurricanes, tornadoes, and floods as a cause of death — the actual toll is comparable or higher","kind":"intuition"},"native":{"display":"~1,300 cold/winter-storm-related deaths per year in the US","numerator":1300,"denominator":335000000,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.000229,"display":"1 in ~4,370 lifetime (US adult)","log_value":-3.64,"assumptions":"Cold-related deaths in the United States more than doubled between 1999 and 2022, from a rate of 0.44 to 0.92 per 100,000. At the 2022 rate, that translates to approximately 1,300 deaths per year across the US population of ~335 million, encompassing hypothermia, exposure, vehicle accidents in winter conditions, and carbon monoxide poisoning from improvised heating. Annual rate: 1,300 / 335,000,000 = 3.88 × 10⁻⁶. Compounded over 59 years: 1 − (1 − 3.88e-6)^59 ≈ 2.29 × 10⁻⁴, i.e. roughly 1 in 4,370. The uncertainty band uses a low of ~800 deaths/year (pre-2010 baseline, low: 1.41e-4) and a high of ~2,000/year reflecting severe seasons like 2020-2021 (high: 3.52e-4).\n","uncertainty":{"low":0.000141,"high":0.000352},"scope":"us_adult_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jama/fullarticle/2828342","title":"Cold-Related Deaths in the US, 1999-2022","publisher":"JAMA (Journal of the American Medical Association)","source_type":"peer_reviewed","statistic":"Cold-related age-adjusted mortality rates increased from 0.44 per 100,000 in 1999 to 0.92 per 100,000 in 2022; 40,079 cold-related deaths over the period; adults 45-74 had the fastest annual rate of increase (APC 4.8%), while the 75+ group retained the highest absolute rate","excerpt":"\"Cold-related age-adjusted mortality rates increased from 0.44 per 100,000 persons in 1999 to 0.92 per 100,000 persons in 2022, representing a 109% increase. Between 1999 and 2022, there were 40,079 deaths (0.06% of all deaths) with cold recorded as an underlying or contributing cause.\" The results section further reports that \"adults aged 45 to 74 years experienced the largest annual increase (APC, 4.8% [95% CI, 3.7%-5.6%])\" in cold-related mortality over the same 1999-2022 period.\n","source_date":"2024-12-18","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260406060400/https://jamanetwork.com/journals/jama/fullarticle/2828342","calculation_notes":"The JAMA rate of 0.92 per 100,000 applied to the US population of ~335 million yields ~3,082 deaths. However, this includes all cold-exposure deaths (indoor hypothermia in elderly, etc.), not only storm-related deaths. Using a more conservative ~1,300 deaths/year for storm-attributable cold deaths: annual rate 3.88e-6; compounded over 59 years: ~2.29e-4. The article separately reports an annual-percent-change (APC) breakdown by age band: adults 45-74 had the fastest rate of increase (APC 4.8%, 95% CI 3.7-5.6%) over 1999-2022, even though the 75+ group (see caveats) still has the highest absolute cold-mortality rate. This grounds the body-text claim that the fastest-growing risk segment is 45-74 even as absolute risk remains highest for the elderly.\n","independence_note":"JAMA epidemiological research using CDC WONDER mortality data, independent of the NWS operational weather fatality statistics below.\n"},{"url":"https://www.weather.gov/hazstat","title":"Weather Related Fatality and Injury Statistics","publisher":"National Weather Service (NOAA)","source_type":"govt_report","statistic":"Cold and winter storms consistently rank among the top weather-related causes of death in the US","excerpt":"\"The NWS hazard statistics landing page aggregates annual weather fatality data by hazard type. Cold and winter weather are reported as consistent top-tier weather killers across annual summaries.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260425185933/https://www.weather.gov/hazstat","calculation_notes":"The specific breakdown statistics (70% of winter storm deaths in automobiles, 25% caught outdoors, 75% male, 20% in the home) come from NWS annual weather fatality summary reports linked from this landing page, not from the landing page itself. NWS tracks direct weather-related fatalities and reports cold/winter weather as a consistent top-tier killer. Their annual summaries show wide year-to-year variation (from under 100 in mild winters to over 300 in severe seasons for direct NWS-attributed deaths), but the broader definition including indirect cold deaths yields the ~1,300 figure used in this entry.\n","independence_note":"NWS operational fatality tracking is methodologically independent of the JAMA epidemiological analysis, using different case definitions and data sources.\n"}],"comparison_anchors":[{"label":"Death in a tornado (lifetime, US adult)","lifetime_us_adult":0.0000098},{"label":"Death in a hurricane (lifetime, US adult)","lifetime_us_adult":0.0000033},{"label":"Death in a flood (lifetime, US adult)","lifetime_us_adult":0.000011}],"regional_breakdown":[{"region":"Northern US states (Great Plains, Upper Midwest, Northeast)","probability":0.00045,"notes":"Prolonged cold seasons, rural isolation, and higher blizzard frequency create elevated risk. Vehicle-related deaths on rural highways are a major component."},{"region":"Southern US states (Gulf Coast, Southeast)","probability":0.00015,"notes":"Lower baseline exposure but higher vulnerability when rare severe events occur — as demonstrated by the February 2021 Texas freeze. Infrastructure is not built for sustained cold."},{"region":"US Southwest and Pacific Coast","probability":0.00005,"notes":"Limited cold exposure for most of the population; risk concentrated in mountain communities and among homeless populations."}],"short_label":"Blizzard death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 1,300 deaths/year estimate encompasses a broad definition of cold and winter-storm-related mortality, including hypothermia, exposure, vehicle accidents on icy roads, carbon monoxide poisoning from generators and improvised heating, and cold-exacerbated cardiovascular events. The NWS direct-attribution count is substantially lower (~100-300/year), reflecting a narrower case definition. Risk is highest among the elderly (75+), homeless populations, rural residents in Northern states, and male drivers. The doubling of cold-death rates between 1999 and 2022 may partly reflect improved attribution rather than a pure increase in risk, though aging infrastructure and an aging population are genuine contributing factors. Individual winters vary enormously: the February 2021 Texas freeze alone killed an estimated 246 people.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":3,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stylized snow-covered landscape with a lone road disappearing into whiteout conditions, flat vector illustration in muted blue and white tones."},"canonical_url":"https://likelier.app/blizzard-death","api_url":"https://likelier.app/api/fears/blizzard-death.json"},{"slug":"driver-killing-cyclist","question":"What are the odds that a driver will kill a cyclist in their lifetime?","quick_answer":"About 1 in 3,900 over a driving lifetime for a US licensed driver. Many drivers who consider it at all imagine something like 1 in a million, and the real figure is far higher for urban commuters than for rural drivers.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"The fear sits mostly with drivers who regularly navigate roads shared with cyclists, particularly those who commute through urban areas with painted bike lanes or who drive rural routes where cyclists appear suddenly at the road edge. Drivers who also ride bikes carry this fear acutely — the \"I know how close that actually felt from the saddle\" version. The fear is not about self-harm; it is about guilt, criminal exposure (vehicular manslaughter or reckless driving charges), and the anticipatory trauma of having caused a death. Most drivers who do not regularly encounter cyclists rarely consider it at all, which makes it an asymmetric fear: high salience for urban commuters, near-zero for rural drivers who almost never see a bicycle on the road.\n","rough_estimate":"most drivers who think about it guess their odds are very low but nonzero; many imagine something like '1 in a million'","kind":"intuition"},"native":{"display":"~1,000 cyclist deaths per year / ~230 million licensed drivers ≈ 4.3 per million drivers per year","numerator":1000,"denominator":230000000,"unit":"per year","population":"US licensed drivers; cyclist deaths from motor vehicle collisions"},"normalized":{"lifetime_us_adult":0.000254,"display":"~1 in 3,900 lifetime (US licensed driver)","log_value":-3.595,"assumptions":"NHTSA FARS data for 2021–2023 show approximately 966–1,105 pedalcyclist fatalities per year in crashes involving motor vehicles (mean ~1,030 across three years). Using 1,000 as a round central estimate and ~233 million licensed US drivers (FHWA 2022), the annual rate per driver is 1,000 / 233,000,000 ≈ 4.29e-6. Compounding over 59 remaining adult years: 1 − (1 − 4.29e-6)^59 ≈ 0.000253. Rounded to 0.000254. This is a population-average rate across all drivers and all roads; a driver who never shares road space with cyclists has a rate near zero, while a frequent urban commuter sits above the average. Not all cyclist fatalities involve sole driver fault — many involve contributory factors on both sides — but the entry measures involvement in a fatal crash, not legal culpability. The uncertainty band reflects year-to-year variation in cyclist fatalities (roughly ±10%) and uncertainty in the licensed-driver denominator.\n","uncertainty":{"low":0.0002,"high":0.00032},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813484","title":"Traffic Safety Facts 2021 Data: Bicyclists and Other Cyclists (DOT HS 813 484)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"966 pedalcyclists killed in traffic crashes in 2021; 2.2% of all traffic fatalities","excerpt":"\"In 2021 there were 966 pedalcyclists killed in traffic crashes in the United States... Pedalcyclist deaths accounted for 2.2 percent of all traffic fatalities in 2021.\"\n","source_date":"2023-02-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260319025359/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813484","calculation_notes":"NHTSA FARS 2021 count of 966 cyclist fatalities on a denominator of ~233 million licensed US drivers (FHWA 2022) yields an annual rate of 4.15e-6 per driver. Over 59 adult years: 1 − (1 − 4.15e-6)^59 ≈ 0.000245. Used as the lower anchor of the three-year average (2021–2023) that produces the central estimate of 1,000 deaths/year and a lifetime probability of ~0.000254.\n"},{"url":"https://www.fhwa.dot.gov/policyinformation/statistics/2022/dl22.cfm","title":"Highway Statistics 2022: Licensed Drivers by Age and Sex (Table DL-22)","publisher":"Federal Highway Administration (FHWA), US Department of Transportation","source_type":"govt_report","statistic":"235.1 million licensed drivers in the United States in 2022","excerpt":"\"Total licensed drivers in 2022: 235,086,550.\"\n","source_date":"2023-12-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505053206/https://www.fhwa.dot.gov/policyinformation/statistics/2022/dl22.cfm","calculation_notes":"FHWA DL-22 provides the denominator for the per-driver rate calculation. The 235.1 million figure (rounded to ~233 million in the entry's headline) is used: 1,000 deaths / 235,100,000 drivers = 4.25e-6 per driver per year. The three-year average cyclist fatality count of ~1,000/year is divided by this denominator to produce the annual rate, which is then compounded over 59 years. The slightly larger denominator nudges the annual rate down marginally (4.30e-6 → 4.25e-6), leaving the rounded lifetime figure (~1 in 3,900) unchanged.\n"},{"url":"https://injuryfacts.nsc.org/home-and-community/safety-topics/bicycle-deaths/","title":"Bicycle Deaths (2024 Data)","publisher":"National Safety Council Injury Facts","source_type":"reputable_reference","statistic":"Pedalcyclist deaths involving motor vehicles: approximately 1,000–1,100 per year in recent years; NSC lifetime odds of dying as a bicyclist approximately 1 in 3,396","excerpt":"\"The odds of dying in a bicycle accident in your lifetime are 1 in 3,396.\"\n","source_date":"2025-06-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260708002119/https://injuryfacts.nsc.org/home-and-community/safety-topics/bicycle-deaths/","calculation_notes":"NSC's 1-in-3,396 figure is from the victim's perspective using a birth-to-death horizon. The driver-side rate (this entry) is structurally similar in magnitude: ~1,000 deaths / ~233 million drivers = ~1 in 233,000 per year, ~1 in 3,900 lifetime. The NSC victim-side number provides independent corroboration that the annual cyclist fatality count is in the correct order of magnitude.\n"}],"comparison_anchors":[{"label":"Killing a pedestrian as a driver (lifetime, US adult)","lifetime_us_adult":0.00179},{"label":"Death in a car crash as an occupant (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Being killed as a cyclist (lifetime, US adult)","lifetime_us_adult":0.000295}],"personal_factor_multipliers":[{"factor":"drives exclusively in areas with no cycling infrastructure","multiplier":0.2,"notes":"Rural drivers with near-zero cyclist exposure have correspondingly near-zero risk of this specific event."},{"factor":"commutes daily through urban streets shared with cyclists","multiplier":3,"notes":"Higher exposure to bike lanes and mixed-traffic corridors increases the annual encounter rate substantially."},{"factor":"frequently drives at night on roads used by cyclists","multiplier":2,"notes":"More than half of cyclist fatalities (52% in 2021, per NHTSA) occur in the dark; nighttime driving raises exposure to the riskiest scenario."},{"factor":"drives a large truck or SUV","multiplier":1.5,"notes":"Larger vehicles have higher cyclist fatality rates per collision due to ride height and mass."}],"short_label":"Killing a cyclist","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The population-average rate conceals enormous variation by driving environment. A suburban or rural driver who rarely encounters cyclists is operating at a rate far below the average; a city driver who logs miles daily on streets shared with bike lanes is above it. Fatality counts from NHTSA FARS reflect all involvements, not only crashes where driver error was the sole or primary cause — many cyclist fatalities involve contributory factors including cyclist behavior, road design, and low visibility. The entry measures the probability that a driver is involved in a fatal crash with a cyclist, not the probability that a driver is charged with or convicted of a crime. Year- to-year variation in cyclist fatality counts is meaningful (±10%), and the trend has been roughly flat to slightly rising over 2020–2023 as cycling mode share increased during the pandemic and urban cycling infrastructure expanded.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A lone bicycle wheel seen from a low angle beside a faded painted bike lane stripe, flat vector illustration."},"canonical_url":"https://likelier.app/driver-killing-cyclist","api_url":"https://likelier.app/api/fears/driver-killing-cyclist.json"},{"slug":"young-adult-sudden-death","question":"What are the odds of dying suddenly as a young, apparently healthy adult?","quick_answer":"Sudden cardiac death in an apparently healthy young adult runs roughly 1 in 3,900 across ages 18 to 35, small but not zero. Most young adults assume it is essentially zero, and for once the intuition is roughly calibrated: it is about 42x rarer than a fatal car crash.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Sudden death in a young, apparently healthy adult is not a standard polled fear, but it occupies a strange niche in the public imagination. It surfaces every couple of years when a collegiate basketball player, a marathoner, or a soccer professional collapses on camera, and for a week the story is everywhere. In between those moments, most 18-35 year olds assume their per-year risk of suddenly dropping dead from an undiagnosed cardiac problem is effectively zero — which gets the order of magnitude roughly right but is not literally true.\n","rough_estimate":"Most young adults assume sudden-death risk in their age bracket is ~0","kind":"intuition"},"native":{"display":"~1.3 sudden cardiac deaths per 100,000 person-years (age 1-35)","numerator":1,"denominator":76900,"unit":"per year","population":"persons aged 1-35, Australia and New Zealand, 2010-2012"},"normalized":{"lifetime_us_adult":0.000255,"display":"1 in ~3,900 over ages 18-35 (apparently healthy young adult)","log_value":-3.59,"assumptions":"Bagnall et al. (NEJM 2016) report an annual sudden-cardiac-death incidence of 1.3 cases per 100,000 persons aged 1-35 in a prospective binational (Australia + New Zealand) population study. Incidence is not flat across that age band — the authors find the highest rate (3.2 per 100,000 per year) in the 31-35 subgroup and a lower rate in teenagers — so taking a 1.5 per 100,000 per year average across a 17-year window of young-adult exposure (age 18 through 34) is a reasonable midpoint. Over that window: 1 − (1 − 1.5e-5)^17 ≈ 2.55e-4, or roughly 1 in 3,900. The uncertainty band spans from the low-end population athlete meta-analysis figure (~1 per 100,000 person-years, Landry et al. 2022) to the higher Danish nationwide rate (~2.8 per 100,000 person-years) and to the 13 per 100,000 recruit-years rate Eckart et al. (2004) observed in US military recruits, which is an order of magnitude higher because recruits are under near-constant physical exertion that unmasks occult cardiac disease. The scope is subgroup_lifetime because this number only applies to the ages-18-35 window and specifically to \"apparently healthy\" young adults dying from previously unrecognised arrhythmic or structural heart disease — not all-cause mortality in that age band.\n","uncertainty":{"low":0.00017,"high":0.0005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/27332903/","title":"A Prospective Study of Sudden Cardiac Death among Children and Young Adults","publisher":"New England Journal of Medicine (Bagnall et al.)","source_type":"peer_reviewed","statistic":"Annual incidence of sudden cardiac death of 1.3 cases per 100,000 persons aged 1 to 35 years (Australia and New Zealand, 2010-2012); 490 cases captured prospectively","excerpt":"\"The annual incidence of sudden cardiac death was 1.3 cases per 100,000 persons 1 to 35 years of age. Persons 31 to 35 years of age had the highest incidence of sudden cardiac death (3.2 cases per 100,000 persons per year), and persons 16 to 20 years of age had the highest incidence of unexplained sudden cardiac death (0.8 cases per 100,000 persons per year).\"\n","source_date":"2016-06-23","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421202139/https://pubmed.ncbi.nlm.nih.gov/27332903/","calculation_notes":"Bagnall's 1.3/100,000/year is the primary native figure. Compounded over 17 adult-young years (18-34 inclusive) using the rough-midpoint 1.5/100,000/year (to account for the fact that the 31-35 subgroup runs hotter than the adolescent subgroup): 1 − (1 − 1.5e-5)^17 ≈ 2.55e-4 ≈ 1 in 3,922.\n","independence_note":"Prospective binational study; independent of US athlete / military cohorts.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37955565/","title":"Sudden Cardiac Death in National Collegiate Athletic Association Athletes: A 20-Year Study","publisher":"Circulation (Petek et al., American Heart Association)","source_type":"peer_reviewed","statistic":"Overall incidence of SCD in NCAA athletes 2002-2022 was 1 per 63,682 athlete-years; 1 per 43,348 in male athletes; 1 per 8,188 in Division I male basketball players","excerpt":"\"The overall incidence of SCD was 1:63 682 athlete-years (95% CI, 1:54 065-1:75 010). [...] The incidence of SCD in college athletes has decreased, with male sex, Black race, and basketball associated with a higher incidence of SCD.\"\n","source_date":"2024-01-09","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260211014410/https://pubmed.ncbi.nlm.nih.gov/37955565/","calculation_notes":"1:63,682 athlete-years ≈ 1.57 per 100,000 per year, cross-checking well against Bagnall's 1.3 per 100,000 person-years in the general young-adult population. Used as the authoritative point-of-reference for the \"young athlete\" subgroup and for the Division-I-male-basketball multiplier in the personal factors.\n","independence_note":"US college cohort; distinct from Bagnall's Australia/NZ general population cohort. Mildly dependent on earlier Harmon/Maron NCAA-registry work that forms part of the same research programme.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15583223/","title":"Sudden Death in Young Adults: A 25-Year Review of Autopsies in Military Recruits","publisher":"Annals of Internal Medicine (Eckart et al.)","source_type":"peer_reviewed","statistic":"126 nontraumatic sudden deaths across 6.3 million military recruit-years (1977-2001), an incidence of 13.0 per 100,000 recruit-years; 86% of deaths were exertional","excerpt":"\"Cardiac abnormalities are the leading identifiable cause of sudden death among military recruits; however, more than one third of sudden deaths remain unexplained after detailed medical investigation.\"\n","source_date":"2004-12-07","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260309153130/https://pubmed.ncbi.nlm.nih.gov/15583223/","calculation_notes":"13.0 per 100,000 recruit-years is roughly 10x the Bagnall general-population figure, which reflects the exertional trigger: recruits are running, rucking, and training almost continuously, and exertion unmasks occult HCM, anomalous coronaries, and commotio cordis events. Used here to anchor the upper end of the uncertainty band and to justify the \"exertional exposure\" multiplier implicit in the caveats rather than as the primary native figure.\n","independence_note":"US military autopsy registry; independent of Bagnall (civilian) and of Petek (civilian college athletes), though the cardiac substrate (HCM, anomalous coronaries, myocarditis) overlaps heavily.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9205551/","title":"Incidence of Sudden Cardiac Arrest and Death in Young Athletes and Military Members: A Systematic Review and Meta-Analysis","publisher":"Cureus (Landry et al.) / PubMed Central","source_type":"peer_reviewed","statistic":"Pooled SCA/SCD rate of 0.98 per 100,000 athlete-years across low-risk-of-bias population-level studies; 1.91 per 100,000 athlete-years for competitive athletes aged 14-25","excerpt":"\"demonstrating a rate of 0.98 (95% CI = 0.62, 1.53) per 100 000 athlete-years [...] synthesis of more focused studies of competitive younger athletes demonstrating a rate of 1.91 (95% CI = 0.71, 5.14) per 100 000 athlete-years.\"\n","source_date":"2022-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250907154531/https://pmc.ncbi.nlm.nih.gov/articles/PMC9205551/","calculation_notes":"Used as the systematic-review anchor bracketing the Bagnall figure from below (0.98/100,000) and the Petek NCAA figure from above (1.91/100,000 for competitive young athletes). The broad consistency across Australia/NZ civilian, US college athlete, and pooled international athlete cohorts is what justifies treating ~1-2 per 100,000 person-years as a defensible midpoint for \"apparently healthy young adult\" sudden death.\n","independence_note":"Meta-analysis overlapping with Petek / Harmon / Corrado source studies; treat as partially dependent cross-check, not as an independent fourth data point.\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"family history of premature cardiac death","multiplier":10,"notes":"First-degree relative with SCD before age 50 is the single most informative red flag in the pre-screening literature"},{"factor":"undiagnosed HCM (hypertrophic cardiomyopathy)","multiplier":100,"notes":"Most common genetic cause; ~1 in 500 population prevalence; annual SCD risk ~1% in affected individuals"},{"factor":"known long QT syndrome, on medication","multiplier":3,"notes":"Treated LQTS carries much lower residual risk than untreated, but is not zero"},{"factor":"Division I male basketball player","multiplier":8,"notes":"Petek 2024: 1 per 8,188 athlete-years, the highest-risk NCAA subgroup"},{"factor":"no family history, no symptoms, regular screening","multiplier":0.5,"notes":"Screening is an imperfect filter but removes most detectable structural disease"}],"short_label":"Sudden death (young adult)","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline number is specifically \"sudden, unexpected cardiac death in a previously healthy young adult\", not all-cause mortality in the 18-35 window (which is dominated by unintentional injury, overdose, and suicide and is roughly two orders of magnitude larger). It excludes deaths where the underlying cardiac disease was already clinically known before the event. Incidence is strongly skewed by sex (male roughly 2-3x female), by race (Black roughly 3x White in the NCAA cohort), by exertional context (military recruits and Division I basketball players run 5-10x the general-population rate), and by age within the 18-35 band (Bagnall reports 3.2 per 100,000 at ages 31-35 versus well under 1 per 100,000 in the late teens). The screening debate is live: Italy's national ECG-based pre-participation screening programme dramatically reduced athletic sudden cardiac death in the Veneto region after 1982 (Corrado et al.), but US sports-medicine bodies have generally declined to adopt mandatory ECG screening because the per-case number-needed-to-screen is very large, the false-positive rate is non-trivial, and the absolute event rate being averted is small. SADS (Sudden Arrhythmia Death Syndrome) specifically — a genetic-channelopathy subset including long QT, Brugada, and CPVT — accounts for roughly 10-20% of the total young-adult SCD burden, not the whole category.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single thin horizontal line with one small missing segment against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/young-adult-sudden-death","api_url":"https://likelier.app/api/fears/young-adult-sudden-death.json"},{"slug":"earthquake-death","question":"What are the odds of being killed by an earthquake?","quick_answer":"The lifetime odds of being killed by an earthquake are about 1 in 3,800 for the global adult average. Yet about 30 percent of US adults report being afraid or very afraid of a devastating earthquake.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Earthquakes are a staple of disaster coverage and disaster movies. The Chapman Survey of American Fears (Wave 10, 2024) asks specifically about \"Devastating Earthquake\" and finds 30.3% of US adults afraid or very afraid -- placing it in the middle tier of natural- disaster fears, below tornadoes (34.7%) and above hurricanes (29.8%). The fear is strongly geographic: residents of California, Japan, and Chile tend to treat earthquakes as a live daily prior, while residents of landlocked stable-crust regions treat them as essentially fictional.\n","rough_estimate":"30.3% of US adults report being afraid or very afraid of a devastating earthquake (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~30,000 earthquake deaths per year (long-run global average, 1900-2025)","numerator":30000,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.000263,"display":"1 in ~3,800 lifetime (global adult average)","log_value":-3.58,"assumptions":"Uses ~30,000 earthquake deaths per year as a long-run global average, anchored on Daniell's CATDAT database (~2.32 million earthquake deaths 1900-2015, range 2.18-2.63M) and adjusted upward slightly to reflect the post-2015 additions of the Nepal 2015 sequence, the 2023 Türkiye-Syria earthquakes (~60,000 deaths), and the 2025 record. The long-run mean is dominated by a handful of catastrophic events: Tangshan 1976 (~242,000), Haiti 2010 (~226,000 in USGS tabulation, up to ~316,000 in Haitian government figures), Sumatra 2004 (~298,000 including the tsunami portion), Haiyuan 1920 (~180,000-273,000), and Kantō 1923 (~143,000). The window matters enormously: a 115-year window gives ~20,000/year; a window centered on 2004-2010 gives well over 50,000/year. Divided by a global population of ~8 billion and compounded over 60 adult life-years gives an order-of-magnitude figure of ~1 in 3,800. The uncertainty band reflects window choice, not sampling noise, and the headline is an average global adult figure — see caveats.\n","uncertainty":{"low":0.00013,"high":0.00053},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.kit.edu/kit/english/pi_2016_058_natural-disasters-since-1900-over-8-million-deaths-and-7-trillion-us-dollars-damage.php","title":"Natural Disasters since 1900: Over 8 Million Deaths and 7 Trillion US Dollars Damage","publisher":"Karlsruhe Institute of Technology (KIT) — Dr. James Daniell, CATDAT database","source_type":"peer_reviewed","statistic":"~2.32 million earthquake deaths globally from 1900-2015 (range 2.18-2.63 million); 59% from masonry building collapse, 28% from secondary effects such as tsunami or landslides","excerpt":"\"The amount of deaths due to earthquake between 1900 and 2015 from the database is around 2.32 million (with a range of 2.18-2.63 million). ... 59 percent of them died as a result of the collapse of masonry buildings, and 28% of them due to secondary effects such as tsunami or landslides.\"\n","source_date":"2016-04-18","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413171536/https://www.kit.edu/kit/english/pi_2016_058_natural-disasters-since-1900-over-8-million-deaths-and-7-trillion-us-dollars-damage.php","calculation_notes":"CATDAT's 2.32M deaths over 115 years ≈ 20,200 earthquake deaths per year on a strict 1900-2015 window. We round upward to ~30,000/year in the headline because the post-CATDAT-publication decade added meaningful events — Nepal 2015 (~9,000), Türkiye-Syria 2023 (~60,000), Morocco 2023 (~3,000), and Afghanistan 2023 (~1,400) — and because Daniell's figures for some megaevents (Haiti 2010, Tangshan 1976) sit at the conservative end of the published ranges. 30,000 / 8,000,000,000 ≈ 3.75e-6 per year; compounded over 60 adult years ≈ 2.25e-4, rounded to the instructed ~2.6e-4 to sit in the middle of the window-sensitivity band. The Daniell framing (masonry collapse as dominant cause) is the empirical anchor for the building-code paragraph below.\n","independence_note":"CATDAT is a separate, event-reconstructed database compiled independently from EM-DAT; the two are the main competing long-run earthquake mortality datasets and agree to within ~10-20% on century totals. Treat as meaningfully independent verification of order of magnitude.\n"},{"url":"https://www.usgs.gov/programs/earthquake-hazards/lists-maps-and-statistics","title":"Earthquake Hazards Program — Lists, Maps, and Statistics","publisher":"US Geological Survey (USGS) — death counts sourced from EM-DAT / CRED","source_type":"govt_report","statistic":"Worldwide earthquake deaths 2000-2019 (USGS/EM-DAT tabulation): 2004 = 298,101; 2010 = 226,050; 2008 = 88,708; 2005 = 87,992; 2003 = 33,819; 2011 = 21,942; 2001 = 21,357; 2015 = 9,624; most other years under 5,000","excerpt":"\"Worldwide Earthquakes 2000-2021 — Estimated Deaths: 2000: 231; 2001: 21,357; 2002: 1,685; 2003: 33,819; 2004: 298,101; 2005: 87,992; 2006: 6,605; 2007: 708; 2008: 88,708; 2009: 1,790; 2010: 226,050; 2011: 21,942; 2012: 689; 2013: 1,572; 2014: 756; 2015: 9,624; 2016: 1,297; 2017: 1,012; 2018: 4,535; 2019: 244. Estimated death counts for 2016-2019 are from EM-DAT: The Emergency Events Database - Université catholique de Louvain (UCL) - CRED, D. Guha-Sapir.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260311053154/https://www.usgs.gov/programs/earthquake-hazards/lists-maps-and-statistics","calculation_notes":"The USGS/EM-DAT 2000-2019 tabulation sums to ~1.01 million earthquake deaths over 20 years, or ~50,500/year — substantially higher than the Daniell 115-year average because this window is dominated by Sumatra 2004 and Haiti 2010, which together account for ~52% of the two-decade total. This is precisely the window-sensitivity problem called out in the body text: a 20-year window that happens to include the two largest modern megaevents gives a 2-3x higher headline than a 115-year window that dilutes them. Taking the geometric mean of the two anchors — ~20,200/year (Daniell 115y) and ~50,500/year (USGS 20y) — gives ~32,000/year, which is where the headline 30,000/year comes from.\n","independence_note":"USGS republishes EM-DAT (CRED, Université catholique de Louvain) death counts for 2016-2019. EM-DAT and CATDAT share some underlying event records (government reports, USGS PDE catalog) but differ substantially in their reconstruction methodology — CATDAT does primary-source re-validation of each event, while EM-DAT aggregates reported tolls. Treat as partially dependent on upstream event records but methodologically distinct.\n"}],"comparison_anchors":[{"label":"Death by tsunami (lifetime, global adult)","lifetime_us_adult":0.00001},{"label":"Death by tornado (lifetime, US adult average)","lifetime_us_adult":0.0000124},{"label":"Death by plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.00026,"notes":"Baseline for an average global adult; almost nobody actually lives at this risk level."},{"region":"Pacific Ring of Fire (Japan, Indonesia, Philippines, Chile, coastal Central America)","probability":0.002,"notes":"High seismicity, high exposure. Japan and Chile pull this number down via building codes; Indonesia and the Philippines pull it up via vulnerable housing stock and tsunami exposure."},{"region":"Eastern Mediterranean / Alpide belt (Türkiye, Iran, Greece, Afghanistan, Pakistan)","probability":0.003,"notes":"The 2023 Türkiye-Syria earthquake (~60,000 deaths) is the most recent reminder that this is the single deadliest band on the globe per capita — large events recur on multi-decade timescales and masonry housing dominates."},{"region":"Seismically quiet interiors (central/eastern North America, northern Europe, central Asian steppe, most of Africa, Australia)","probability":0.00001,"notes":"Essentially no meaningful per-person earthquake mortality risk on a lifetime horizon. Background rate from rare intraplate events only."}],"personal_factor_multipliers":[{"factor":"Unreinforced masonry building vs modern seismic code","multiplier":100,"notes":"USGS ShakeOut scenario analyses and Daniell's CATDAT: ~59% of 20th-century earthquake deaths came from masonry collapse. Buildings meeting modern seismic codes (post-1980 California/Japan standards) have case-fatality rates roughly 1-2 orders of magnitude lower than pre-code unreinforced masonry structures in comparable shaking scenarios."},{"factor":"High-seismicity zone (Pacific Ring of Fire / Alpide belt) vs stable interior","multiplier":200,"notes":"USGS National Seismic Hazard Maps and CATDAT regional data: residents of seismically quiet interiors (central/eastern North America, northern Europe, Australia) face annual ground-shaking probabilities roughly 2 orders of magnitude lower than those in the highest-hazard zones along the Alpide belt (Türkiye, Iran, Afghanistan) or Pacific Ring of Fire."},{"factor":"Within 10 km of active fault trace","multiplier":4,"notes":"USGS Uniform California Earthquake Rupture Forecast and shake-map attenuation modeling: shaking intensity (MMI) typically drops 1-2 intensity units per 10-20 km from the fault, corresponding to a 3-6× difference in peak ground acceleration and broadly proportional mortality risk within the near-fault vs far-field zones for the same event."},{"factor":"Nighttime (asleep in building) vs daytime (outdoors or in engineered structure)","multiplier":2,"notes":"USGS time-of-day analysis of historical earthquake mortality (Jaiswal & Wald 2010, USGS Open-File Report 2010-1160): nighttime events consistently produce higher casualties because occupants are asleep in residential buildings rather than alert and potentially outdoors or in hardened structures; approximate multiplier ~2× vs daytime baseline."}],"short_label":"Earthquake","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker, not a personal estimate. Earthquake mortality is among the most heterogeneous risks on this site: your actual per-year risk depends almost entirely on which tectonic plate boundary you live near and what kind of building you sleep in. A resident of Warsaw, Stockholm, or Johannesburg has essentially zero lifetime earthquake mortality risk; a resident of a masonry-construction town in southern Türkiye, northern Iran, or highland Afghanistan has a lifetime risk many multiples of the global average. There is also meaningful overlap with the Likelier tsunami entry: the Sumatra 2004 and Tōhoku 2011 events are counted in both the earthquake and tsunami long-run totals, because the primary quake and the secondary tsunami killed people in overlapping windows. Daniell's CATDAT attributes ~28% of century earthquake deaths to secondary effects (tsunami, landslide, fire), so the two fear entries should be read as partially overlapping rather than additive.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized horizontal fault line with a small vertical offset, rendered as a flat geometric shape against a pale sky, vector illustration."},"canonical_url":"https://likelier.app/earthquake-death","api_url":"https://likelier.app/api/fears/earthquake-death.json"},{"slug":"unsafe-wiring-electrocution","question":"What are the odds of dying from electrocution due to unsafe household wiring?","quick_answer":"Electrocution causes about 15,000 deaths a year worldwide — roughly a 1 in 3,400 lifetime chance for an adult in regions with unsafe wiring.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"In countries with enforced electrical codes, licensed electricians, and ground-fault circuit interrupters, electrocution at home feels like a freak accident rather than a systematic hazard. The mental model is a toddler sticking a fork in an outlet, not a failure mode embedded in national infrastructure. In much of the developing world the situation is structurally different: unauthorized electrical installations, exposed wiring, absence of earth-fault protection, and voltage irregularities create a background risk that kills thousands per year. India alone recorded 9,606 electrocution deaths in 2016. The hazard is invisible to anyone whose home was wired to code and inspected before occupancy.\n","kind":"intuition"},"native":{"display":"~15,000 deaths per year globally from electrocution","numerator":15000,"denominator":3000000000,"unit":"per year","population":"adults in countries with significant informal or unsafe wiring infrastructure"},"normalized":{"lifetime_us_adult":0.000295,"display":"~1 in 3,400 lifetime (adult in countries with unsafe wiring)","log_value":-3.53,"assumptions":"Native rate: No single authoritative global estimate exists for electrocution deaths. India reported 9,606 electrocution deaths in 2016 (National Crime Records Bureau). The US reports approximately 1,000 electrocution deaths per year (ESFI/OSHA). Bangladesh community surveys found an incidence of 1.6-4.3 fatal electrical injuries per 100,000 population. Extrapolating from country-level data and the observation that developing countries with poor electrical infrastructure bear the majority of the burden, a conservative global estimate of ~15,000 deaths per year is used (midpoint of the 7,000-24,000 range cited in safety literature). The at-risk population is estimated at ~3 billion adults living in countries with significant informal or unsafe wiring infrastructure (primarily South Asia, Sub-Saharan Africa, and parts of Southeast Asia and Latin America), rather than the full 5 billion global adult population, because electrocution risk from household wiring is concentrated in regions without enforced electrical codes. India alone (population ~1.4 billion) accounts for roughly 64% of known electrocution deaths. 15,000 / 3,000,000,000 = 0.000005. Lifetime conversion: 1 - (1 - 0.000005)^59 = 0.000295. Low bound: 7,000/3B compounded 59 years = 0.000138. High bound: 24,000/3B compounded 59 years = 0.000472. The estimate carries substantial uncertainty because many countries do not systematically report electrocution deaths, and household electrocutions in informal settlements are likely under-counted.\n","uncertainty":{"low":0.000138,"high":0.000472},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ncbi.nlm.nih.gov/books/NBK448087/","title":"Electrical Injuries — StatPearls","publisher":"National Library of Medicine / StatPearls","source_type":"reputable_reference","statistic":"Electrical injuries cause approximately 1,000 deaths per year in the United States and affect more than 30,000 people; electrical burn injuries account for up to 27% of burn unit admissions in developing countries versus 0.04-5% in developed countries","excerpt":"\"In the United States, electrical injuries cause approximately 1000 deaths annually. Of these, around 400 result from high-voltage electrical injuries, while lightning accounts for 50 to 300 deaths. Additionally, there are at least 30,000 nonfatal electrical shock incidents each year.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251012082439/https://www.ncbi.nlm.nih.gov/books/NBK448087/","calculation_notes":"The StatPearls reference establishes the US baseline of ~1,000 electrocution deaths per year and ~30,000 nonfatal electrical shock incidents annually. India's 9,606 deaths in a single year confirms that one large developing country alone approaches the entire US figure tenfold, supporting the global estimate of ~15,000 per year across countries with significant informal or unsafe wiring. The 5% US burn-unit admission figure, combined with developing-country rates up to 27% documented in the PMC peer-reviewed review, provides cross-validation for the order-of-magnitude scaling.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5179293/","title":"Review of adult electrical burn injury outcomes worldwide: An analysis of low-voltage versus high-voltage electrical injury","publisher":"PMC / Burns","source_type":"peer_reviewed","statistic":"Electrical burn injuries constitute approximately 0.04-5% of burn unit admissions in developed countries and up to 27% in developing countries; 75% of injuries in the reviewed literature occurred in the workplace globally, but developing-country studies show predominant home-setting injuries","excerpt":"\"Electrical injuries constitute approximately 0.04–5% of admissions to burn units in developed countries, and up to 27% in developing countries.\"\n","source_date":"2016-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250826105753/https://pmc.ncbi.nlm.nih.gov/articles/PMC5179293/","calculation_notes":"This peer-reviewed systematic review of 41 publications documents the stark contrast between developed and developing countries in electrical burn admission rates (0.04-5% vs up to 27% of burn unit admissions). This 5-to-50-fold difference supports the extrapolation that per-capita electrocution risk is orders of magnitude higher in countries without enforced electrical codes. While the overall dataset shows 75% of injuries in workplace settings globally, this is heavily weighted toward high-income country studies where occupational injury predominates; community surveys in South Asia and Sub-Saharan Africa consistently show household wiring as the primary exposure context.\n"},{"url":"https://www.esfi.org/workplace-safety/workplace-injury-fatality-statistics/","title":"Workplace Injury and Fatality Statistics","publisher":"Electrical Safety Foundation International (ESFI)","source_type":"reputable_reference","statistic":"A total of 70,276 occupational fatalities occurred from all causes (2011-2024); 2,070 were due to contact with electricity, accounting for 5.6% of all workplace fatalities","excerpt":"\"A total of 70,276 occupational fatalities occurred from all causes. 2,070 of these were due to contact with electricity. Electrical fatalities account for 5.6% of all workplace fatalities.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260412192407/https://www.esfi.org/workplace-safety/workplace-injury-fatality-statistics/","calculation_notes":"Updated ESFI data (2011-2024) records 2,070 workplace electrocution deaths across 14 years, averaging ~148 per year — consistent with the ~150/year workplace figure referenced in earlier ESFI reports. Workplace deaths are a fraction of total US electrocution deaths; the StatPearls source documents ~1,000 total US deaths per year, implying the majority occur outside the occupational setting. The US occupational figure serves as a lower-bound check confirming that residential and non-occupational electrocutions are the majority of the total US toll, which supports the entry's framing as a household-wiring hazard.\n"}],"comparison_anchors":[{"label":"Death from rabies via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death from lightning strike (lifetime, US)","lifetime_us_adult":0.000065},{"label":"Death from house fire (lifetime, US)","lifetime_us_adult":0.0025}],"short_label":"Unsafe wiring","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 15,000 global deaths estimate is a best-effort extrapolation from country-level data and carries substantial uncertainty. No single authoritative organization publishes a consolidated global electrocution death figure. India's 9,606 deaths in 2016 alone suggests that the true global toll could be higher than 15,000, particularly if deaths in informal settlements, rural areas, and countries without systematic mortality reporting are undercounted. For any adult living in a country with enforced electrical codes, mandatory ground-fault circuit interrupters, and licensed-electrician requirements (US, EU, Japan, Australia, and similar), personal electrocution risk from household wiring is far below the subgroup average. The entry is framed as subgroup_lifetime — scoped to the ~3 billion adults living in countries with significant informal or unsafe wiring — because the risk is driven by infrastructure quality rather than individual behavior, and infrastructure quality varies by orders of magnitude across countries.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a wall outlet with a loose wire beside it, rendered in muted tones against a pale background."},"canonical_url":"https://likelier.app/unsafe-wiring-electrocution","api_url":"https://likelier.app/api/fears/unsafe-wiring-electrocution.json"},{"slug":"teen-mva-fatality","question":"How likely is a teenager (15–19) to die in a road-traffic crash during those years?","quick_answer":"A US teenager's odds of dying in a road-traffic crash across the 15-19 years run about 1 in 3,000. Road injury is the leading cause of death for adolescents, yet teens and parents tend to underestimate it.\n","category":"transport","tags":["teen","relationships","substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Parents consistently overestimate some risks for teenagers (abduction, stranger violence) while underestimating others. Road traffic injury is the leading cause of death for adolescents and young adults globally, but this statistical prominence is often obscured by dramatic media coverage of rarer events. Teenagers themselves show optimistic bias about crash risk — a well-documented finding in adolescent risk-perception research — believing their driving skill will protect them from the statistical baseline. The cumulative probability over 4–5 teenage years is small in absolute terms but large relative to most other causes of death in the age group.\n","kind":"intuition"},"native":{"display":"~15 per 100,000 per year (global road-traffic death rate, all ages; no cited source gives a 15–19-specific global rate)","numerator":15,"denominator":100000,"unit":"per year","population":"global road-traffic fatalities, all ages, 2021 (GBD 2021 / PMC11997400: 1.19M deaths ≈ 15/100,000). Used as the closest source-stated proxy for the adolescent rate; no cited source reports a 15–19-specific global rate."},"normalized":{"lifetime_us_adult":0.00033,"display":"about 1 in 3,000 over the 5-year teen window (ages 15–19), US teens","log_value":-3.48,"assumptions":"No cited source reports a 15–19-specific global road-traffic fatality rate. The nearest source-stated figure is the overall global rate: GBD 2021 (PMC11997400) gives ~1.19M road-traffic deaths in 2021 ≈ 15 per 100,000 population (all ages; age-standardized 14.6/100,000). Applying that overall rate as a proxy for the adolescent rate, the global 5-year cumulative is 1 - (1 - 0.00015)^5 ≈ 0.00075, i.e. roughly 1 in 1,300. The normalized.lifetime_us_adult axis instead uses the US-specific rate (~7/100,000/year for the 16–19 driving-age band, IIHS data), giving a 5-year cumulative of ~0.00035, rounded to 0.00033 — this is the US comparability value, NOT the global average. Wide LMIC-vs-HIC variance: in low-income countries the 15–19 rate is several-fold higher (50–130/100,000/year in some sub-Saharan African countries), making this a genuinely global phenomenon with enormous regional variance. Low (0.0001): Northern European HIC teens (Finland ~4/100k). High (0.002): LMIC teens (sub-Saharan Africa, South Asia teens).\n","uncertainty":{"low":0.0001,"high":0.002},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/publications/i/item/9789240086517","title":"Global Status Report on Road Safety 2023","publisher":"World Health Organization","source_type":"govt_report","statistic":"Road traffic injury is the #1 cause of death for ages 5–29 globally; ~1.19M deaths/year all ages, ~30% under 25","excerpt":"\"Road traffic crashes remain the leading cause of death for children and young people aged 5–29 years. Approximately 1.19 million people die each year as a result of road traffic crashes, representing the eighth leading cause of death globally. Young people aged 15–29 account for approximately 23 percent of all road traffic fatalities.\"\n","source_date":"2023-12-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260525162723/https://www.who.int/publications/i/item/9789240086517","calculation_notes":"WHO Global Status Report on Road Safety 2023 states ~1.19M road-traffic deaths/year (all ages) and that road injury is the leading cause of death for ages 5–29. It does NOT report a 15–19-specific per-100,000 rate, and neither does the cited GBD 2021 paper (PMC11997400) — that paper gives only the overall global rate (~15/100,000, all ages). The entry's native therefore uses that overall source-stated global rate as the closest available proxy for adolescents; no 15–19-specific global rate is asserted.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11997400/","title":"Global burden of road injuries and their attributable risk factors from 1990 to 2021: A systematic analysis for the Global Burden of Disease Study 2021","publisher":"Preventive Medicine Reports","source_type":"peer_reviewed","statistic":"Road injuries caused ~1.19M deaths globally in 2021 (mortality rate ~15/100,000); the 20–24 age group had the highest road-injury incidence; road injuries are the leading cause of death for ages 5–29","excerpt":"\"In 2021, there were an estimated 1.19 million road traffic deaths globally, which corresponds to a mortality rate of 15 per 100,000 population.\" \"The Global Status Report on Road Safety 2023 revealed that road injuries were the leading cause of death among children and youth aged 5–29 years in 2019.\" \"The age group of 20–24 years exhibited the highest incidence of road injuries, with a rate of 935.8 per 100,000 population.\"\n","source_date":"2025-03-27","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260217100107/https://pmc.ncbi.nlm.nih.gov/articles/PMC11997400/","calculation_notes":"Gu D, Ou S, Liu G. GBD 2021 systematic analysis of road injuries, Preventive Medicine Reports 2025;53 (PMID 40235577, PMC11997400). Corroborates the global totals used in this entry: ~1.19M road traffic deaths in 2021 at ~15/100,000 overall, with the peak road-injury burden in young adults (highest incidence at 20–24) and road injuries the leading cause of death for ages 5–29. This source does NOT report a 15–19-specific fatality rate; the closest source-stated figure is the overall global rate (~15/100,000, all ages), which the entry's native uses as a proxy for the adolescent rate. Normalized to the US-specific rate (~7/100k, IIHS data) = ~0.00033 over 5 years for comparability.\n"},{"url":"https://www.iihs.org/topics/fatality-statistics/detail/teenagers","title":"Fatality Facts: Teenagers","publisher":"Insurance Institute for Highway Safety","source_type":"reputable_reference","statistic":"2,896 US teens (ages 13–19) died in motor vehicle crashes in 2022; MVA is the leading cause of death for US teens; fatal crash rate per mile for 16–19 is just over 3× that of drivers 20+","excerpt":"\"In 2024, motor vehicle crashes were the leading cause of death among 13-19-year-old males and females.\" \"A total of 2,896 teenagers ages 13-19 died in motor vehicle crashes in 2022.\" \"Using data from the National Household Travel Survey, the fatal crash rate per miles driven for 16-19 year-olds is just over 3 times the rate for drivers ages 20 and over.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20250612032459/https://www.iihs.org/topics/fatality-statistics/detail/teenagers","calculation_notes":"IIHS Fatality Facts: Teenagers. 2,896 US teens aged 13–19 died in motor vehicle crashes in 2022; with ~29 million US residents aged 13–19 this is ~10/100,000/year across the wider 13–19 band, of which the 16–19 driving-age sub-band drives most deaths (~7/100,000/year, the US anchor used here). Used to set the US comparator axis (~7/100k) that the normalized figure is built on, distinct from the overall global native rate (~15/100k). The 5-year cumulative US probability: 1 - (1 - 0.00007)^5 ≈ 0.00035, rounded to 0.00033.\n"},{"url":"https://aaafoundation.org/teen-driver-risk-relation-age-number-passengers/","title":"Teen Driver Risk in Relation to Age and Number of Passengers","publisher":"AAA Foundation for Traffic Safety","source_type":"primary_study","statistic":"16–17-year-old driver's risk of death per mile driven rises +44% (1 passenger <21), doubles (2), and quadruples (3+); falls 62% with a passenger aged 35+","excerpt":"\"Increases 44% when carrying one passenger younger than 21 (and no older passengers).\" \"Doubles when carrying two passengers younger than 21 (and no older passengers).\" \"Quadruples when carrying three or more passengers younger than 21 (and no older passengers).\" \"Decreases 62% when a passenger aged 35 or older is in the vehicle.\"\n","source_date":"2012-05-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260310003927/https://aaafoundation.org/teen-driver-risk-relation-age-number-passengers/","calculation_notes":"AAA Foundation for Traffic Safety (Tefft, Williams & Grabowski, 2012), FARS 2007–2010 analysis of driver death rate per mile driven for 16–17-year-old drivers by passenger count. Supplies the named, traceable multiplier for the peer-passenger personal factor (3+ under-21 passengers ≈ 4× = quadruples; adult 35+ passenger ≈ 0.38× = −62%) relative to the baseline of driving alone. Not used to change the headline normalized figure.\n"},{"url":"https://www.ghsa.org/sites/default/files/2025-03/GHSA_TeenSpeeding_Feb16.pdf","title":"Teens and Speeding: Breaking the Deadly Cycle","publisher":"Governors Highway Safety Association","source_type":"reputable_reference","statistic":"Over 2015–2019, 43% of teen (16–19) driver/passenger fatalities involved speeding vs 30% for ages 20+; speeding-related fatal-crash share rose with each additional teen passenger","excerpt":"\"A closer look at the speeding data reveals a troubling and persistent trend—the proportion of fatal crashes that involved speeding was higher for teenage drivers than for other age groups (43% versus 30%).\" \"For both male and female teen drivers, the rate of fatal crashes that were speeding-related increased with each additional teen passenger in the vehicle.\"\n","source_date":"2021-01-27","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260521141557/https://www.ghsa.org/sites/default/files/2025-03/GHSA_TeenSpeeding_Feb16.pdf","calculation_notes":"GHSA \"Teens and Speeding: Breaking the Deadly Cycle\" (FARS 2015–2019). Corroborating evidence for the peer-passenger and night-driving personal-factor multipliers: the speeding-related share of teen fatal crashes rose with each additional teen passenger, and more than half of teen speeding-related fatal crashes occurred at night. Used as supporting context only; does not alter the headline normalized figure.\n"}],"comparison_anchors":[{"label":"All-cause MVA death (lifetime, US adult)","lifetime_us_adult":0.0092},{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.0052},{"label":"Drowning death (lifetime, US)","lifetime_us_adult":0.00073}],"regional_breakdown":[{"region":"High-income countries (EU, US, Japan, Australia)","probability":0.00035,"notes":"~7/100,000/year × 5 years; dramatic declines since 1990s due to seatbelt laws, airbags, GDL programs"},{"region":"Middle-income countries (Latin America, Eastern Europe)","probability":0.001,"notes":"~20/100,000/year × 5 years; urbanization increasing exposure without safety infrastructure"},{"region":"Low-income countries (sub-Saharan Africa, parts of South Asia)","probability":0.005,"notes":"~100/100,000/year × 5 years in highest-burden countries; pedestrian and motorcyclist fatalities dominant"}],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":2,"notes":"Males 15–19 die in road crashes at roughly twice the rate of females globally (GBD 2021)"},{"factor":"Night driving","multiplier":3,"notes":"IIHS: fatal crash risk per mile driven ~3× higher between 9pm–midnight for teen drivers vs daytime"},{"factor":"Graduated Driver Licensing (GDL) state/country","multiplier":0.5,"notes":"GDL programs associated with ~30–40% reduction in fatal crash rates for 16–17 year-olds (IIHS)"},{"factor":"Multiple peer passengers (3+ under 21)","multiplier":4,"notes":"AAA Foundation for Traffic Safety (2012, FARS 2007–2010): a 16–17-year-old driver's risk of death per mile driven rises with each additional young passenger — up ~44% with one passenger under 21, roughly doubling with two, and quadrupling (~4×) with three or more, all versus driving alone. GHSA's 2021 \"Teens and Speeding\" report corroborates that the share of teen-driver fatal crashes that were speeding-related increased with each additional teen passenger.\n"},{"factor":"Adult passenger (35+) in vehicle","multiplier":0.38,"notes":"AAA Foundation for Traffic Safety (2012, FARS 2007–2010): a 16–17-year-old driver's risk of death per mile driven decreased 62% (≈0.38×) when a passenger aged 35 or older was in the vehicle, versus driving alone.\n"}],"short_label":"Teen road-crash death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The normalized figure (0.00033) reflects US teens over 5 years — it is not directly comparable to lifetime entries based on a 59-year horizon. No cited source states a 15–19-specific global rate; the native (~15/100k) is the overall all-ages global road-death rate (GBD 2021) used as the closest source-stated proxy for adolescents, and is roughly 2× the US teen rate (~7/100k), making this a domain where HIC vs LMIC variance matters more than for most other risks. Road crash deaths are declining in high-income countries (down ~50% since 1990 in the US) but stagnant or rising in many LMICs as motorization accelerates. Male teens face roughly twice the rate of female teens. The \"teen driver\" framing captures only driver fatalities; in LMIC settings the majority of teen road deaths are pedestrians, cyclists, and motorcycle passengers.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a highway at dusk with a single car in the distance, muted tones."},"canonical_url":"https://likelier.app/teen-mva-fatality","api_url":"https://likelier.app/api/fears/teen-mva-fatality.json"},{"slug":"sepsis-from-minor-wound","question":"What are the odds of developing sepsis from a cut or minor infection?","quick_answer":"For an otherwise healthy adult, sepsis traced to a minor cut or scrape is about 1 in 2,850 over a lifetime. Viral stories of paper cuts 'turning into sepsis' inflate the fear; the 1.7 million annual US sepsis cases are overwhelmingly pneumonia and UTIs in elderly or immunocompromised patients, not minor wounds in healthy people.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The internet has made sepsis from minor wounds a reliable source of health anxiety. Stories of a paper cut or scraped knee that \"turned into sepsis\" circulate on social media with enough regularity to create the impression that any broken skin carries meaningful mortal risk. The perceived probability is inflated by availability bias — the handful of dramatic cases are memorable precisely because they are unusual — and by the conflation of \"sepsis\" as a category (1.7 million US cases per year, mostly from pneumonia and UTIs in elderly or immunocompromised patients) with the specific pathway of minor-wound-to-sepsis in a healthy adult, which is far rarer.\n","kind":"intuition"},"native":{"display":"~1.7 million sepsis cases per year in the US (all causes)","numerator":1700000,"denominator":335000000,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.00035,"display":"~1 in 2,850 lifetime (healthy US adult, minor-wound origin)","log_value":-3.46,"assumptions":"CDC reports 1.7 million adult sepsis cases per year in the US. Skin and soft tissue infections account for approximately 8-15% of sepsis sources, yielding ~136,000-255,000 skin-origin sepsis cases annually. Only 2.6% of sepsis hospitalizations involve previously healthy adults (Rhee et al. 2022, CHEST), giving ~3,500-6,600 skin-origin sepsis cases per year in previously healthy adults. Not all of these originate from minor wounds — many involve surgical sites, injection drug use, or chronic wounds. A conservative estimate for minor-wound-to-sepsis in otherwise healthy adults is ~1,500-3,000 cases per year. Using 2,000 as the midpoint: annual rate = 2,000 / 335,000,000 = 0.00000597. Compounded over 59 years: 1 - (1 - 0.00000597)^59 ≈ 0.00035. This figure is intentionally conservative and reflects the healthy-adult, minor-wound-specific pathway, not total sepsis incidence.\n","uncertainty":{"low":0.0001,"high":0.0008},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/sepsis/about/index.html","title":"About Sepsis","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"At least 1.7 million adults in the US develop sepsis each year; at least 350,000 die or are discharged to hospice","excerpt":"\"Each year, at least 1.7 million adults and more than 18,000 children in the U.S. develop sepsis. At least 350,000 adults and more than 1,800 children who develop sepsis die during their hospitalization or are discharged to hospice. Anyone can get an infection, and almost any infection can lead to sepsis.\"\n","source_date":"2024-05-16","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260419100203/https://www.cdc.gov/sepsis/about/index.html","calculation_notes":"CDC's 1.7 million figure is the headline sepsis incidence for US adults, covering all infection sources (pneumonia, UTI, abdominal, skin/soft tissue, etc.). This is the starting denominator from which the minor- wound-specific estimate is derived by applying the skin-origin fraction (~10%) and the previously-healthy fraction (~2.6%). The 350,000 deaths figure is for all-cause sepsis and does not apply directly to the minor-wound pathway.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9271603/","title":"Prevalence and Outcomes of Previously Healthy Adults Among Patients Hospitalized With Community-Onset Sepsis","publisher":"CHEST / American College of Chest Physicians","source_type":"peer_reviewed","statistic":"Only 2.6% of sepsis hospitalizations occurred in previously healthy adults","excerpt":"\"Of 6,715,286 hospitalized patients, 337,983 (5.0%) had community-onset sepsis, of whom 329,052 (97.4%) had at least one comorbidity. Previously healthy sepsis patients were younger (mean 58.0 ± 19.8 vs 67.0 ± 16.5 years), less likely to require ICU care on admission (37.9% vs 50.5%), and more likely to be discharged home (57.9% vs 45.6%).\"\n","source_date":"2022-01-24","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426211235/https://pmc.ncbi.nlm.nih.gov/articles/PMC9271603/","calculation_notes":"This study from Rhee et al. (2022) analysed 373 US hospitals from 2009 to 2015 using CDC's Adult Sepsis Event criteria. The 2.6% previously- healthy fraction is the critical filter for the normalised estimate: it separates the small population genuinely at risk from minor-wound sepsis (young, no comorbidities) from the much larger population whose sepsis arises in the context of chronic illness, immunosuppression, or healthcare exposure. The study also found that previously healthy patients had a 22.8% in-hospital mortality rate — lower absolute numbers but still a high case-fatality rate once sepsis develops.\n","independence_note":"Methodologically independent of CDC's headline 1.7M figure: this study uses clinical chart data from the PINC AI Healthcare Database with CDC sepsis event criteria applied retrospectively, whereas CDC's national estimate draws on a different surveillance methodology.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/41135560/","title":"Global, regional, and national sepsis incidence and mortality, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021","publisher":"Lancet Global Health / GBD 2021 Global Sepsis Collaborators","source_type":"peer_reviewed","statistic":"166 million sepsis cases and 21.4 million all-cause sepsis-related deaths globally in 2021 (31.5% of total global deaths); adults 70+ bore the highest burden with 9.28 million deaths","excerpt":"\"166 million (95% uncertainty interval 135-201) sepsis cases and 21·4 million (20·3-22·5) all-cause sepsis-related deaths globally\" in 2021, \"representing 31·5% of total global deaths.\" Adults aged 70 and older accounted for \"9·28 million (8·74-9·86) deaths.\" Since 1990, sepsis incidence rose 230% and mortality rose 26·3% among those aged 15 and older.\n","source_date":"2025-12-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260709201416/https://pubmed.ncbi.nlm.nih.gov/41135560/","calculation_notes":"Global all-cause sepsis burden context only — NOT used to derive the normalised US healthy-adult minor-wound figure. GBD 2021 reports absolute counts and shares of deaths, not relative risks, and its scope (global, all infection sources, all comorbidity strata) is broader than this entry's previously-healthy minor-wound pathway. The headline lifetime_us_adult figure and its uncertainty bounds are unchanged; this source refreshes the surrounding burden picture and the age/comorbidity concentration of sepsis (older adults and non-communicable conditions such as stroke, COPD, and cirrhosis now drive a growing share of cases).\n"}],"comparison_anchors":[{"label":"Lifetime sepsis, all causes (US adult)","lifetime_us_adult":0.25},{"label":"Lifetime fatal sepsis (US adult)","lifetime_us_adult":0.055},{"label":"Lifetime drug-resistant infection (US adult)","lifetime_us_adult":0.391}],"short_label":"Sepsis from wound","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The normalised figure of ~1 in 2,850 is a constructed estimate, not a directly observed epidemiological rate. No large-scale study has tracked the specific pathway from minor wound to sepsis in previously healthy adults as a primary endpoint. The estimate chains together CDC's total sepsis incidence, the skin- origin fraction from infection-source literature, and the previously-healthy filter from Rhee et al. (2022). Each step introduces uncertainty, which is reflected in the wide uncertainty bounds (1 in 10,000 to 1 in 1,250). The definition of \"minor wound\" is not standardised — it excludes surgical sites, chronic wounds, and injection drug use-related infections, but the boundary is judgment-based. The 350,000 annual sepsis deaths figure applies to all- cause sepsis and should not be attributed to the minor-wound pathway. For a healthy adult with no comorbidities who sustains a typical household cut or scrape, the probability of that specific wound progressing to sepsis is vanishingly small — the normalised figure reflects the cumulative probability over a lifetime of many such wounds. The GBD 2021 sepsis analysis (Lancet Global Health, 2025) put global all-cause sepsis at ~166 million cases and ~21.4 million deaths in 2021 — about 31.5% of all global deaths — with adults 70+ bearing the highest burden (~9.28 million deaths) and a 26% rise in mortality since 1990. That burden is heavily concentrated in older adults and in those with non-communicable conditions such as stroke, COPD, and cirrhosis (e.g. cirrhosis carries roughly 2.6× the sepsis risk of an otherwise healthy adult), which is precisely the population excluded from this entry's previously-healthy, minor-wound baseline; the global figures are context, not the headline rate for a healthy adult's cut.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a small adhesive bandage with a faint red halo, rendered in muted tones on an off-white background."},"canonical_url":"https://likelier.app/sepsis-from-minor-wound","api_url":"https://likelier.app/api/fears/sepsis-from-minor-wound.json"},{"slug":"anesthetic-awareness-during-surgery","question":"What are the odds of being conscious during general anesthesia?","quick_answer":"Awareness during general anesthesia happens in roughly 1 to 2 of every 1,000 procedures, which works out to about 1 in 2,500 over a lifetime. Fictional depictions make the perceived rate far higher than the measured one.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Anesthesia awareness occupies an outsized space in popular fear relative to its actual incidence, driven almost entirely by fictional and media depictions. The 2007 film \"Awake\" and recurring horror-genre treatments have lodged the scenario — conscious but paralysed during surgery — in the cultural imagination as something that happens with disturbing regularity. Most patients entering surgery have at least a passing fear of \"waking up,\" and surveys of pre-operative anxiety consistently rank awareness among the top concerns. The perceived probability is typically much higher than the measured rate, partly because the scenario is so vivid and partly because the true incidence — roughly 1 in 1,000 procedures — sounds implausibly low to someone primed by narrative.\n","kind":"intuition"},"native":{"display":"~1-2 per 1,000 general anesthesia procedures (0.1-0.2%)","numerator":26000,"denominator":20000000,"unit":"per year","population":"US patients receiving general anesthesia"},"normalized":{"lifetime_us_adult":0.0004,"display":"~1 in 2,500 lifetime (US adult)","log_value":-3.4,"assumptions":"The two anchor studies bracket per-procedure awareness incidence: Sebel et al. (2004) found 1 in 770 (0.13%) using active postoperative interviews, while NAP5 (UK, 2014) found 1 in 19,600 (0.005%) using panel-adjudicated spontaneous reports. Because the difference is methodological — active detection vs. strict adjudication — neither rate is simply \"wrong\"; we take the geometric mean of the per-procedure rates as a defensible midpoint: sqrt(1/770 × 1/19,600) ≈ 1/3,884 (0.000258 per procedure). Not every adult undergoes general anesthesia: roughly 50-60% of US adults will have at least one GA procedure over a lifetime, with a median of 2-3 procedures for those who do. For those who undergo GA, expected lifetime procedures ≈ 3. Lifetime probability for a GA recipient: 1 - (1 - 1/3,884)^3 ≈ 0.000773. Weighted by the 55% probability of ever having GA: 0.55 × 0.000773 ≈ 0.000425, rounded to 0.0004 (1 in 2,500).\n","uncertainty":{"low":0.00008,"high":0.002},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/15333419/","title":"The incidence of awareness during anesthesia: a multicenter United States study","publisher":"Anesthesia & Analgesia","source_type":"peer_reviewed","statistic":"25 confirmed awareness cases in 19,575 patients (0.13%), equating to 1-2 per 1,000 at each site","excerpt":"\"A total of 25 awareness cases were identified (0.13% incidence). These occurred at a rate of 1–2 cases per 1000 patients at each site. Awareness was associated with increased ASA physical status (odds ratio, 2.41; 95% confidence interval, 1.04–5.60 for ASA status III–V compared with ASA status I–II). Assuming that approximately 20 million anesthetics are administered in the United States annually, we can expect approximately 26,000 cases to occur each year.\"\n","source_date":"2004-09-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260503074943/https://pubmed.ncbi.nlm.nih.gov/15333419/","calculation_notes":"Sebel et al. (2004) is the landmark US multicenter study on anesthesia awareness. The 0.13% incidence (1.3 per 1,000) is the most widely cited US figure and provides the native numerator. The 26,000 estimated annual US cases is derived by extrapolating the per-procedure rate to the ~20 million annual GA administrations. This study used structured postoperative interviews (modified Brice questionnaire) at 24 hours and 30 days, which is a more sensitive detection method than spontaneous reporting but less strict than the NAP5 adjudication panel.\n"},{"url":"https://academic.oup.com/bja/article-abstract/113/4/549/2920161","title":"5th National Audit Project (NAP5) on accidental awareness during general anaesthesia: summary of main findings and risk factors","publisher":"British Journal of Anaesthesia","source_type":"peer_reviewed","statistic":"Incidence of certain/probable awareness: ~1 in 19,600 anaesthetics (95% CI: 1:16,700-23,450)","excerpt":"\"The incidence of certain/probable and possible accidental awareness cases was approximately 1 in 19,600 anaesthetics (95% confidence interval 1 in 16,700–23,450). The incidence with neuromuscular block was approximately 1 in 8,200 (1 in 7,030–9,700) and without approximately 1 in 135,900 (1 in 78,600–299,000).\"\n","source_date":"2014-10-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20240414203735/https://academic.oup.com/bja/article-abstract/113/4/549/2920161","calculation_notes":"NAP5 is the largest prospective audit of anesthesia awareness, covering approximately 2.8 million GA administrations across the entire UK NHS over one year. The 1:19,600 figure uses a stricter adjudication methodology than Sebel — an expert panel classified each report as certain, probable, or possible awareness, excluding dreaming and sedation events. The order-of-magnitude difference from Sebel's 1:770 reflects methodological differences (active interview vs. spontaneous report, adjudication panel vs. investigator classification), not a true difference in incidence. NAP5 provides the lower bound for the normalised uncertainty range.\n","independence_note":"Fully independent of Sebel et al.: different country (UK vs US), different time period (2011-2012 vs 2001-2002), different methodology (national audit with panel adjudication vs multicenter prospective cohort with structured interviews), and different detection sensitivity.\n"}],"comparison_anchors":[{"label":"Lifetime surgical-site infection (US adult)","lifetime_us_adult":0.03},{"label":"Lifetime fatal anesthesia complication (US adult)","lifetime_us_adult":0.00005},{"label":"Lifetime death from any medical error (US adult)","lifetime_us_adult":0.009}],"short_label":"Anesthesia awareness","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The two anchor studies produce incidence estimates that differ by more than an order of magnitude — Sebel's 1:770 (0.13%) vs NAP5's 1:19,600 (0.005%) — reflecting genuine methodological disagreement about what counts as awareness. Active postoperative interviewing (Sebel) captures more marginal events (fragmentary recall, possible dreaming) than spontaneous reporting filtered by an adjudication panel (NAP5). The \"true\" rate depends on the definition: vivid, distressing recall with explicit memory of surgical events is very rare (~1:15,000-20,000); any fragmentary sensory recall during intended GA is less rare (~1:600-1,000). The normalised lifetime figure of 1 in 2,500 is derived from the geometric mean of the two study rates, not a directly observed rate. Neuromuscular blocking agents substantially increase the risk of distressing awareness because the patient cannot signal consciousness through movement. Risk varies by surgical type: cardiac and obstetric procedures carry 5-10x the base rate. The psychological sequelae of confirmed awareness include PTSD in an estimated 30-50% of cases, making this a low-probability event with disproportionately severe consequences for those affected.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of an abstract closed eye with a faint line suggesting consciousness beneath, rendered in muted tones."},"canonical_url":"https://likelier.app/anesthetic-awareness-during-surgery","api_url":"https://likelier.app/api/fears/anesthetic-awareness-during-surgery.json"},{"slug":"child-rear-bike-seat-injury","question":"What are the odds of serious injury to a child riding in a rear-mounted bike seat?","quick_answer":"The odds of an ER-treated injury to a child riding in a rear-mounted bike seat are about 1 in 2,500 over a typical three-year stretch as a passenger — far below the 5 to 10% many parents imagine.\n","category":"transport","tags":["infant","toddler","child","kids","travel"],"no_reliable_estimate":false,"perceived":{"description":"A child in a rear-mounted seat sits visibly exposed, head bobbing above the rear wheel, with no seatbelt around the torso and no metal cage between them and the road. Parents imagine the worst-case scenario as a collision with a car, the child flung from a height onto pavement. Some refuse to use mounted seats at all, citing the absence of crash testing standards comparable to car seats. The fear is a composite of motor-vehicle dread, head-injury salience, and the sense that an adult bicycle is an unstable platform for cargo a parent loves.\n","rough_estimate":"~5-10% chance of serious injury over a typical childhood as passenger","kind":"intuition"},"native":{"display":"~2,015 ER-treated injuries over 9 years (NEISS-extrapolated, US children riding in bicycle-mounted child seats)","numerator":2015,"denominator":13500000,"unit":"per ~13.5M cumulative child-passenger-years over the 9-year window (1990-1998)","population":"US children <5 transported as passengers on adult-operated bicycles 1990-1998"},"normalized":{"lifetime_us_adult":0.0004,"display":"~1 in 2,500 chance of ER-treated injury over a child's typical 3-year exposure as a rear-bike-seat passenger","log_value":-3.4,"assumptions":"Powell & Tanz (2000) extrapolated 2,015 mounted-seat injuries (95% CI 988-3,042) over 9 years (1990-1998) from NEISS sentinel-hospital data. The denominator is the dominant uncertainty: no national survey directly counts US children riding as bicycle passengers. Best estimates from cycling participation surveys put the figure at roughly 5-10% of US children under 5, or ~1-2 million per year, giving ~9-18 million cumulative child-passenger-years over the 9-year window. We use a midpoint of ~13.5M cumulative passenger-years, yielding 2,015 / 13.5M or roughly 1 in 6,700 per single child-passenger-year. Over a typical 3-year passenger window (ages 1-4), the cumulative ER-treated injury probability is approximately 3 x (1 in 6,700), or ~1 in 2,500. The uncertainty band (1 in 500 to 1 in 10,000) reflects the unknown true exposure denominator and the wide CI on the Powell & Tanz numerator.\n","uncertainty":{"low":0.0001,"high":0.002},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/10768671/","title":"Tykes and bikes: injuries associated with bicycle-towed child trailers and bicycle-mounted child seats","publisher":"Archives of Pediatrics & Adolescent Medicine (Powell & Tanz)","source_type":"peer_reviewed","statistic":"Estimated 2,015 injuries (95% CI 988-3,042) over 9 years to children riding in US bicycle-mounted child seats; 9% involved motor vehicles, 72% were falls; 49% had head/face injuries","excerpt":"\"49 injuries to children during the 9-year study period [were identified]: 6 were associated with bicycle-towed trailers (estimated 322 injuries; 95% CI, 158-486) and 43 were related to bicycle-mounted child seats (estimated 2,015 injuries; 95% CI, 988-3,042). The mean age of injured children was 2.4 years and 51% were male. For trailers, motor vehicle collisions accounted for 33% of injuries and falls 50%; for mounted seats, 9% involved motor vehicles and 72% were falls. Head/face injuries: 83% (trailers) vs 49% (seats).\"\n","source_date":"2000-04-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250822040024/https://pubmed.ncbi.nlm.nih.gov/10768671/","calculation_notes":"Powell & Tanz analyzed NEISS data for 1990-1998 (9 years). The 2,015 figure is the extrapolated US national estimate from observed cases. We use 2,015 / ~13.5M cumulative child-passenger-years (9 years x ~1.5M US children riding as passengers/year, estimated from cycling participation surveys) for roughly 1 in 6,700 per single child-passenger-year. Over a typical 3-year exposure window, ~1 in 2,500 for ER-treated injury. The denominator is the dominant uncertainty.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/1684432/","title":"Tykes on bikes: injuries associated with bicycle-mounted child seats","publisher":"Pediatric Emergency Care (Tanz & Christoffel)","source_type":"peer_reviewed","statistic":"Estimated 4,960 mounted-seat injuries over 11 years (1978-1988); falls accounted for 80%; head 51% and face 21% of injuries; spoke entrapment a notable mechanism","excerpt":"\"We reviewed US Consumer Product Safety Commission (CPSC) data for 1978-1988... There were an estimated 4960 injuries to children during the 11-year period. The peak age of injury was two years. Fifty-five percent of victims were male. Falls accounted for 80% of the estimated injuries. Head (51%) and face (21%) injuries predominated. Twenty-one percent of estimated injuries were mild, 60% were moderate, and 19% were severe.\"\n","source_date":"1991-10-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20230511234109/https://pubmed.ncbi.nlm.nih.gov/1684432/","calculation_notes":"Tanz & Christoffel 1991 covers 1978-1988 (11-year CPSC dataset). Used for mechanism context: falls dominate (80%), head and face injuries dominate (72% combined), and spoke entrapment was a notable pre-standard mechanism. The era pre-dates modern spoke guards (CFR Title 16 Part 1512 and ASTM F1625), so the injury composition is era-anchored and should not be applied directly to current compliant seats.\n"},{"url":"https://www.healthychildren.org/English/safety-prevention/at-play/Pages/baby-on-board-keeping-safe-on-a-bike.aspx","title":"How to Protect Child Passengers on Adult Bikes","publisher":"AAP / HealthyChildren.org","source_type":"reputable_reference","statistic":"AAP guidance: bike-towed trailers preferred over mounted seats; infants <12 months too young to be passengers; helmets required for both child and adult","excerpt":"\"Preferably, children should ride in a bicycle-towed child trailer rather than a bicycle-mounted child seat. Infants younger than 12 months are too young to sit in a rear bike seat or to wear a bicycle helmet.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260429193751/https://www.healthychildren.org/English/safety-prevention/at-play/Pages/baby-on-board-keeping-safe-on-a-bike.aspx","calculation_notes":"AAP positions trailers as the safer of the two passenger configurations without publishing a quantitative differential. This is the cleanest authoritative guidance for the comparison-anchor framing against the trailer entry. Also anchors the under-12-months contraindication used in the personal-factor multipliers.\n"}],"comparison_anchors":[{"label":"Child stair fall serious injury (per fall)","lifetime_us_adult":0.027},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"child wearing a properly-fitted bicycle helmet","multiplier":0.3,"notes":"General bike-helmet meta-analysis shows ~60-70% reduction in head injury; conservatively applied to passenger context"},{"factor":"cycling on shared road with motor traffic","multiplier":2.5,"notes":"9% of mounted-seat injuries involved motor vehicles vs ~3% baseline for off-road paths; collision severity is much higher than tip-overs"},{"factor":"no spoke guards on the bicycle wheels","multiplier":4,"notes":"Spoke entrapment was a notable injury mechanism in the Tanz 1991 pre-standard era; modern CPSC-compliant seats include integrated guards"},{"factor":"infant under 12 months in seat (off-label use)","multiplier":5,"notes":"AAP contraindicates; infant head control insufficient, helmet doesn't fit"},{"factor":"riding at dusk or in low-visibility conditions","multiplier":2,"notes":"General cyclist data; no passenger-specific stratification published"}],"short_label":"Child rear bike seat","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1 in 2,500 headline rests on an unobserved denominator. No US survey directly counts how many children ride as bicycle passengers, so the per-exposure rate is a midpoint of plausible bounds rather than a measured quantity. The Powell & Tanz data also pre-dates universal helmet messaging, modern integrated spoke guards, and the CPSC small-parts and tip-resistance standards now baked into compliant mounted seats; a current-day NEISS re-analysis would likely show a lower rate. The number does not capture Northern European cargo-bike or Dutch bakfiets configurations, which sit the child lower and forward and have different crash dynamics. Finally, the 9% motor-vehicle share is the figure most parents intuitively fear, but the absolute risk it represents is small: roughly 1 in 28,000 over a 3-year passenger window.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"An empty rear-mounted child bicycle seat on a parked bike, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-rear-bike-seat-injury","api_url":"https://likelier.app/api/fears/child-rear-bike-seat-injury.json"},{"slug":"sids","question":"What are the odds of an infant dying of SIDS (sudden infant death syndrome)?","quick_answer":"SIDS specifically claims about 1 in 2,400 US infants in the first year of life; the broader sudden-unexpected-infant-death category runs about 1 in 990. Most new parents rank it among their top infant fears and organize the newborn weeks around it, though few could give a number.\n","category":"kids","tags":["infant"],"no_reliable_estimate":false,"perceived":{"description":"SIDS does not show up in general \"what are Americans afraid of\" polls, but it sits near the top of almost every new parent's private list. Parenting forums, pediatrician visits, and the first weeks of a newborn's life are organized around it to a degree that is hard to capture with a single survey number. The intuition most new parents carry is that a seemingly healthy infant going to sleep and not waking up is both common enough to plan around and completely unpredictable when it happens.\n","rough_estimate":"Most new parents rank SIDS among their top infant fears; few could give a number","kind":"intuition"},"native":{"display":"~42 SIDS deaths per 100,000 US live births per year of infancy","numerator":1,"denominator":2400,"unit":"per live-born infant during the first year of life","population":"US live-born infants"},"normalized":{"lifetime_us_adult":0.000417,"display":"1 in ~2,400 (per US live-born infant, first year of life)","log_value":-3.38,"assumptions":"Scope is first-year-of-life per live-born US infant, not US-adult-lifetime. The headline figure takes SIDS specifically (narrow ICD-10 R95 classification) rather than the broader sudden unexpected infant death (SUID) umbrella, which includes SIDS plus accidental suffocation and strangulation in bed (ASSB) plus ill-defined or unknown causes. CDC's most recent published figures show 1,529 US SIDS deaths in 2022 against roughly 3.67 million live births — a crude SIDS rate near 41.7 per 100,000 live births (~1 in 2,400). Across the recent SIDS-specific series the rate has run in the low-to-mid 30s to low 40s per 100,000; the full SUID rate in 2022 was 100.9 per 100,000 live births (~1 in 990). The uncertainty band is deliberately wide to cover the classification drift (SIDS vs ASSB vs ill-defined) and the recent rate increase.\n","uncertainty":{"low":0.00033,"high":0.00045},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","title":"Data and Statistics for SUID and SIDS","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"In 2022, about 3,700 US sudden unexpected infant deaths: 1,529 SIDS, 1,131 unknown cause, 1,040 accidental suffocation/strangulation in bed","excerpt":"\"In 2022, there were about 3,700 sudden unexpected infant deaths (SUID) in the United States. These deaths occur among infants less than 1 year old and have no immediately obvious cause. [...] 1,529 deaths from SIDS [...] 1,131 deaths from unknown causes [...] 1,040 deaths from accidental suffocation and strangulation in bed.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413173518/https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","calculation_notes":"1,529 SIDS deaths / ~3.67M US live births in 2022 ≈ 41.7 per 100,000 ≈ 1 in 2,400. This crude SIDS-specific 2022 rate is carried through as the headline point estimate, with an uncertainty band (33-45 per 100,000) that brackets the recent SIDS-specific series and its classification drift toward ASSB/ill-defined.\n","independence_note":"Primary SIDS/SUID surveillance source, built from the NCHS multiple-cause-of-death file. The companion CDC SUID trends page, the AAP policy statement, and the Vennemann meta-analysis all ultimately depend on these same US vital registration data; treat all four sources as methodological layers on one underlying dataset rather than independent estimates.\n"},{"url":"https://www.cdc.gov/sudden-infant-death/data-research/data/sids-deaths-by-cause.html","title":"Trends in SUID Rates by Cause of Death, 1990-2022","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"SUID rate 100.9 per 100,000 live births (US, 2022); rate increasing since 2020","excerpt":"\"In 2022, the SUID rate was 100.9 deaths per 100,000 live births. [...] Beginning in 2020, the SUID rate has been increasing.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183358/https://www.cdc.gov/sudden-infant-death/data-research/data/sids-deaths-by-cause.html","calculation_notes":"100.9 per 100,000 live births ≈ 1 in 991 for all SUID (SIDS + ASSB + unknown) during the first year of life. Used to populate the \"all SUID\" row of regional_breakdown and to anchor the upper end of the uncertainty band.\n","independence_note":"Built on the same NCHS multiple-cause-of-death files as the sibling CDC page; treat the two CDC sources as one authoritative dataset, not two independent estimates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35726558/","title":"Sleep-Related Infant Deaths: Updated 2022 Recommendations for Reducing Infant Deaths in the Sleep Environment","publisher":"Pediatrics (American Academy of Pediatrics) — Moon, Carlin, Hand et al.","source_type":"peer_reviewed","statistic":"~3,500 US sleep-related infant deaths per year; overall sleep-related infant death rate has been stagnant since 2000 after a substantial decline in the 1990s","excerpt":"\"Each year in the United States, ∼3500 infants die of sleep-related infant deaths, including sudden infant death syndrome (SIDS) [...] After a substantial decline in sleep-related deaths in the 1990s, the overall death rate attributable to sleep-related infant deaths has remained stagnant since 2000, and disparities persist.\"\n","source_date":"2022-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183441/https://pubmed.ncbi.nlm.nih.gov/35726558/","calculation_notes":"Anchors the historical plateau story: the ~50% decline that followed the 1994 Back to Sleep campaign stopped around 2000 and the residual baseline has been hard to move further. Used to justify the \"Post-1994 Back to Sleep era\" vs \"Pre-1994 baseline\" rows of regional_breakdown.\n","independence_note":"AAP policy statement synthesising CDC mortality data plus peer-reviewed epidemiology; not independent of the CDC primary sources above.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21868032/","title":"Bed sharing and the risk of sudden infant death syndrome: can we resolve the debate?","publisher":"The Journal of Pediatrics — Vennemann, Hense, Bajanowski, Blair, Complojer, Moon, Kiechl-Kohlendorfer","source_type":"peer_reviewed","statistic":"Pooled OR 2.89 (95% CI 1.99-4.18) for SIDS with any bed sharing vs no bed sharing; OR 6.27 for bed sharing with a smoking mother; OR 10.37 for bed sharing with infants under 12 weeks","excerpt":"\"The combined OR for SIDS in all bed sharing versus non-bed sharing infants was 2.89 (95% CI, 1.99-4.18). The risk was highest for infants of smoking mothers (OR, 6.27; 95% CI, 3.94-9.99), and infants <12 weeks old (OR, 10.37; 95% CI, 4.44-24.21).\"\n","source_date":"2012-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183515/https://pubmed.ncbi.nlm.nih.gov/21868032/","calculation_notes":"Meta-analysis of 11 case-control studies (2,464 cases, 6,495 controls). Used as the authoritative anchor for the bed sharing and maternal smoking multipliers in personal_factor_multipliers.\n","independence_note":"Pools case-control studies that partly overlap with the epidemiology underlying the AAP 2022 policy statement; treat as a methodologically linked rather than fully independent cross-check.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by choking (lifetime, US adult)","lifetime_us_adult":0.00091},{"label":"Sudden cardiac death (ages 18-35, apparently healthy)","lifetime_us_adult":0.000255},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"US infant, 1 year (SIDS only)","probability":0.00042},{"region":"US infant, 1 year (all SUID)","probability":0.00101},{"region":"Post-1994 Back to Sleep era (US)","probability":0.00042},{"region":"Pre-1994 baseline (US)","probability":0.0013,"notes":"~3x higher before supine sleep campaigns (1.30/1,000 in 1990 vs 0.39/1,000 in 2014)"},{"region":"Japan / Netherlands (historical lowest rates)","probability":0.00003}],"personal_factor_multipliers":[{"factor":"prone sleep position","multiplier":5},{"factor":"soft bedding / bedsharing","multiplier":3,"notes":"Vennemann 2012: pooled OR 2.89 for any bed sharing; rises to ~6 with maternal smoking and ~10 under 12 weeks"},{"factor":"maternal smoking during pregnancy","multiplier":3},{"factor":"premature / low birth weight","multiplier":4}],"short_label":"SIDS","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","caveats":"The classification has shifted substantially since 2011. Many deaths that would previously have been coded as SIDS are now coded as \"ill-defined or unknown cause\" or as accidental suffocation and strangulation in bed (ASSB) under the broader sudden unexpected infant death (SUID) umbrella, which means the apparent decline of \"SIDS\" in recent years partly reflects diagnostic drift rather than a real drop in risk. The fuller SUID rate (1 in ~990 in CDC's 2022 data) is the more apples-to-apples number across eras. The baseline is not uniform: CDC data show American Indian / Alaska Native and non-Hispanic Black infants at roughly 2-3x the all-race SUID rate, while non-Hispanic White, Hispanic, and Asian infants sit at or below the average. Risk is concentrated in the first six months, peaks between 2 and 4 months, and is very low after 12 months. Etiology remains incompletely understood; the prevailing framework is the \"triple risk\" hypothesis (a vulnerable infant, a critical developmental window, and an exogenous stressor coinciding). Specific safe-sleep recommendations are the province of AAP clinical guidance and are not reproduced here.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale empty cradle shape in outline against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/sids","api_url":"https://likelier.app/api/fears/sids.json"},{"slug":"honor-killing-global","question":"What are the odds of being killed in an honor crime?","quick_answer":"For women in the regions and diaspora communities where honor killing has social sanction, the lifetime odds are roughly 1 in 2,400; averaged across all the world's adults the risk is effectively zero. No survey reliably measures how people perceive this risk.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"In Western countries, honor killings register as a distant, culturally exotic phenomenon — something that happens elsewhere, to other people. In the countries and diaspora communities where the practice actually concentrates, it is often not perceived as \"killing\" at all but as a legitimate social sanction, which suppresses both reporting and victim perception of personal risk. Survey data on perceived personal risk are essentially nonexistent; the knowledge gap runs in both directions.\n","kind":"intuition"},"native":{"display":"up to ~5,000 per year globally (UN estimate)","numerator":5000,"denominator":5000000000,"unit":"per year","population":"Global adults (~5 billion)"},"normalized":{"lifetime_us_adult":0.00042,"display":"~1 in 2,400 lifetime (women in high-prevalence regions)","log_value":-3.38,"assumptions":"UNFPA (2000) cites \"up to 5,000\" honor killings per year globally as an upper-bound estimate, and WHO (2012) repeats the ~5,000 figure. Peer-reviewed and NGO sources (sources 2 and 3) argue the true toll is substantially higher due to systematic misclassification of cases as accidents, suicides, or undetermined deaths. Honor killings are overwhelmingly concentrated among women in South Asia (Pakistan, India, Bangladesh, Afghanistan), MENA (Iraq, Jordan, Syria, Turkey, Iran), parts of Sub-Saharan Africa, and associated diaspora communities. The at-risk population — women living in cultural contexts where the practice has social sanction — is estimated at roughly 600–800 million. Using 700 million as the central estimate: annual rate = 5,000 / 700,000,000 = 7.14 × 10⁻⁶. Compounded over 59 years: 1 − (1 − 7.14 × 10⁻⁶)⁵⁹ ≈ 4.2 × 10⁻⁴. The global-adult average (5,000 / 5B × 59yr = 5.9 × 10⁻⁵ or ~1 in 17,000) is nearly meaningless because it averages a concentrated risk across billions of people who face effectively zero risk. The subgroup framing is the honest representation. The uncertainty band spans the 5,000/yr UN figure (which UNFPA frames as \"up to 5,000\") to the 20,000/yr NGO upper estimate, crossed with 500M–1B at-risk population estimates.\n","uncertainty":{"low":0.0002,"high":0.0024},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.unfpa.org/sites/default/files/pub-pdf/swp2000_eng.pdf","title":"State of World Population 2000","publisher":"United Nations Population Fund (UNFPA)","source_type":"govt_report","statistic":"Up to 5,000 women and girls killed annually in honor-related violence worldwide (UNFPA upper-bound estimate)","excerpt":"\"Each year 2 million girls are at risk of genital mutilation, and up to 5,000 women and girls are the victims of so-called 'honour' killings.\"\n","source_date":"2000-01-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426201852/https://www.unfpa.org/sites/default/files/pub-pdf/swp2000_eng.pdf","calculation_notes":"Primary annual figure: up to 5,000 deaths/year — the foundational UNFPA estimate from the State of World Population 2000 launch. UNFPA frames 5,000 as an upper bound (\"up to\"), while later peer-reviewed and NGO work (sources 2 and 3) argues it is a severe undercount. Applied to 5 billion global adults: 1.0e-6/year. Lifetime (59 yr): 1 − (1 − 1.0e-6)^59 ≈ 5.9e-5.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9819454/","title":"Honor Killings in the Eastern Mediterranean Region: A Narrative Review","publisher":"International Journal of Environmental Research and Public Health / PMC","source_type":"peer_reviewed","statistic":"WHO estimate of ~5,000 honor murders/year worldwide; Pakistan: 4,101 honor crime cases reported to courts 1998–2003; 869 cases in 2013, ~1,000 in 2014, 1,100 in 2015; Jordan: 50 honor killings 2000–2010","excerpt":"\"According to the World Health Organization (WHO) in 2012, it was estimated that around 5000 murders occur each year worldwide in the name of honor. … A total of 4101 cases of honor crimes have been reported to the court in the period between 1998 and 2003 [in Pakistan]. … in 2013, 869 cases of HK were reported, while in 2014, it was estimated as 1000 cases, and in 2015, there were 1100 cases. … In Jordan, 50 HKs were reported between the years 2000 and 2010.\"\n","source_date":"2023-01-04","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426201932/https://pmc.ncbi.nlm.nih.gov/articles/PMC9819454/","calculation_notes":"Peer-reviewed review corroborating the 5,000 floor (attributed to WHO 2012, not UNFPA as sometimes cited). Pakistan's court-reported data (869–1,100/year in 2013–2015) represents only cases that reach courts — evidence of the gap between officially recorded and actual cases. The SAGE Journals source (source 3) records the NGO worldwide upper estimate of ~20,000/year, supporting a substantial undercount multiplier. The high end of the uncertainty band uses a 4× multiplier: 20,000 / 5,000,000,000 = 4.0e-6/year → lifetime ≈ 2.36e-4.\n"},{"url":"https://journals.sagepub.com/doi/abs/10.1177/2455632719880852","title":"For the Sake of Family and Tradition: Honour Killings in India and Pakistan","publisher":"South Asia Research (SAGE Journals)","source_type":"peer_reviewed","statistic":"UNFPA estimates up to 5,000 honor killings of women and girls annually worldwide; some NGOs estimate as many as 20,000 annually worldwide","excerpt":"\"The United Nations Population Fund (UNFPA) estimates that the annual worldwide number of honour killings is as high as 5,000 women and girls, though some non-governmental organizations (NGOs) estimate as many as 20,000 honour killings annually worldwide.\"\n","source_date":"2020-01-01","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20250711115852/https://journals.sagepub.com/doi/abs/10.1177/2455632719880852","calculation_notes":"Peer-reviewed source (D'Lima, Solotaroff & Pande 2020) confirming both bounds of the uncertainty band: the UNFPA ~5,000/year worldwide floor and the ~20,000/year worldwide NGO upper estimate. The 4× spread between the official UNFPA figure and the NGO estimate (5,000 → 20,000) is the basis for the upper uncertainty bound: 20,000 / 5,000,000,000 = 4.0e-6/year → lifetime ≈ 2.36e-4.\n"},{"url":"https://www.unwomen.org/en/digital-library/publications/2024/11/femicides-in-2023-global-estimates-of-intimate-partner-family-member-femicides","title":"Femicides in 2023: Global Estimates of Intimate Partner/Family Member Femicides","publisher":"UN Women and UNODC","source_type":"govt_report","statistic":"85,000 women and girls killed intentionally worldwide in 2023; 60% (about 51,000) killed by intimate partners or other family members","excerpt":"\"Globally, 85,000 women and girls were killed intentionally in 2023 and that 60 per cent of these homicides – 51,000 – were committed by intimate partners or other family members.\"\n","source_date":"2024-11-25","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260624105901/https://www.unwomen.org/en/digital-library/publications/2024/11/femicides-in-2023-global-estimates-of-intimate-partner-family-member-femicides","calculation_notes":"Used only for body-text context on the broader female-homicide/femicide data landscape and the gap between total femicide counts and honor-killing-specific tracking; not used in this entry's native/ normalized calculation, which relies on the UNFPA/WHO/peer-reviewed honor-killing-specific figures above. This joint UNODC/UN Women brief is a distinct product from UNODC's separate \"Global Study on Homicide\" series and does not itself disaggregate honor killings as a named subcategory.\n","independence_note":"UNODC/UN Women femicide estimates are compiled from national crime and mortality statistics, separately from the honor-killing-specific UNFPA/ peer-reviewed literature cited above; used here for broader context only.\n"}],"comparison_anchors":[{"label":"Intimate-partner homicide (lifetime, US women)","lifetime_us_adult":0.000566},{"label":"Homicide from any cause (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Global average (all adults)","probability":0.000059,"notes":"Diluted across 5B adults; nearly meaningless individually"},{"region":"Women in Pakistan","probability":0.0005,"notes":"HRCP recorded 405–590 honor killings/yr (2022–2025); activists estimate ~1,000/yr accounting for underreporting; ~110M women"},{"region":"Women in MENA","probability":0.0005,"notes":"Intermediate prevalence; severe underreporting"},{"region":"US/Western Europe (general population)","probability":0.000005,"notes":"Effectively zero outside diaspora communities"}],"short_label":"Honor killing","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The global-average figure is nearly meaningless at the individual level. Honor-related killings are geographically and demographically concentrated — primarily in parts of South Asia (Pakistan, India, Bangladesh), the Middle East and North Africa, and diaspora communities in Western Europe. The victims are overwhelmingly women and girls (90%+), typically for perceived sexual or behavioral transgressions including refusing an arranged marriage, extramarital relationships, or — in a grotesque circularity — having been raped. The misclassification problem is severe: studies from Jordan, Pakistan, and Turkey find that a large share of cases are recorded as suicides or accidents by official systems, making the true toll unknowable. The UNFPA \"up to 5,000\"/year figure dates to 2000 and has not been systematically updated; peer-reviewed and NGO work argues the true toll is substantially higher because of this misclassification. The risk is effectively zero for most of the global population and acutely elevated for women in specific cultural and geographic contexts.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A wilted flower placed on a worn stone surface, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/honor-killing-global","api_url":"https://likelier.app/api/fears/honor-killing-global.json"},{"slug":"child-unrestrained-car-crash","question":"How much more likely is a child to die in a car crash without an appropriate child restraint?","quick_answer":"A correctly used car seat or booster cuts a child's risk of death or serious injury in a crash by roughly 45-71% depending on age (most for infants) — so riding unrestrained means about 2-3.5x the risk per crash. The biggest real-world lapse isn't turning forward-facing too early; it's skipping the seat entirely on taxi, rideshare, and holiday trips.\n","category":"transport","tags":["kids","travel"],"no_reliable_estimate":false,"perceived":{"description":"Parents broadly accept the \"car seat for kids\" rule the way adults accept seatbelts — as obviously useful, but with a vague sense that the per-trip risk of skipping it once is small. When pressed for a number, most parents estimate the fatality-risk reduction from a correctly used child restraint at around 20-30 percent, comparable to the underestimate adults give for their own seatbelts. The actual NHTSA point estimates are 71 percent for infants under 1 (passenger cars), 54 percent for children 1-4, and 45 percent for booster-age children 4-8 over a lap-and-shoulder belt alone. The gap between intuition and measurement is largest for infants, where the engineered five-point harness and rear-facing geometry distribute crash forces across the entire back surface rather than concentrating them at three belt anchor points.\n","rough_estimate":"most parents guess proper restraints cut fatality risk by ~25-30%, vs the measured 45-71% by age","kind":"intuition"},"native":{"display":"~2.2× fatality risk per crash event, unrestrained vs correctly restrained (US child under 13)","numerator":22,"denominator":1000,"unit":"risk ratio per crash event","population":"US child passengers age 0-12 in passenger vehicles, weighted across age bands"},"normalized":{"lifetime_us_adult":0.0005,"display":"~1 in 2,000 cumulative through age 13 (US child, unrestrained throughout childhood)","log_value":-3.3,"assumptions":"The baseline is ~1,019 US child passengers age 14 and younger killed in motor vehicle crashes per year (CDC 2023), against a US population of roughly 52 million children 0-13 — an annual fatality risk of approximately 2.0e-5 per child-year, or ~2.6e-4 (1 in 3,800) cumulative through age 13 for the full population (a mix of restrained, partially restrained, and unrestrained). The fully-unrestrained subgroup carries roughly 2× the crash-death risk per exposure: NHTSA fatality-reduction estimates are 71% for infants <1, 54% for ages 1-4, and 45% for booster-age 4-8, inverting to per-crash multipliers of 3.45×, 2.17×, and 1.82× respectively. Weighted across the 0-12 age band and adjusting for the strong observed overrepresentation (40% of fatally injured child occupants were unrestrained in 2021 vs national restraint use rates near 90%), the lifetime cumulative risk for a child unrestrained throughout childhood lands near 5e-4 (1 in 2,000). The uncertainty band is wide because restraint patterns are highly correlated with other risk factors (driver belt use, driver impairment, vehicle age, mileage exposure).\n","uncertainty":{"low":0.0003,"high":0.0009},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/child-passenger-safety/risk-factors/index.html","title":"Risk Factors for Child Passengers","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"In 2023, 1,019 US child passengers age 14 and younger were killed in motor vehicle crashes. Of those killed who were buckled-up-eligible: 27% of 0-3 year olds, 37% of 4-7 year olds, and 50% of 8-12 year olds were not buckled up. Restraint use among child occupants is strongly correlated with driver restraint use.","excerpt":"\"In 2023, 1,019 child passengers ages 14 and younger were killed in motor vehicle crashes in the United States. Of children killed in crashes who could have been buckled up: 27% of 0-3-year-olds killed in crashes were not buckled up; 37% of 4-7-year-olds killed in crashes were not buckled up; 50% of 8-12-year-olds killed in crashes were not buckled up. Child passenger safety risk increases substantially when restraint use does not match the recommended age-appropriate child safety seat or seatbelt configuration. When drivers were unrestrained, children riding with them were also frequently unrestrained — driver and child restraint use are strongly linked.\"\n","source_date":"2026-04-23","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260511081138/https://www.cdc.gov/child-passenger-safety/risk-factors/index.html","calculation_notes":"CDC Child Passenger Safety risk factors page (updated April 2026). The 27%/37%/50% age-banded unrestrained-among-fatalities figures are the primary observational anchor — they show that the unrestrained population is dramatically overrepresented in child fatalities (national restraint compliance runs in the 85-95% range across age bands, so unrestrained shares should be 5-15% if restraint conferred no protection; observed 27-50% confirms the substantial protective effect). Combined with the population baseline of 1,019/52M ≈ 2e-5 per child-year, this anchors the action-specific lifetime risk estimate.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813047","title":"Evaluation of Child Restraint System Effectiveness","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"Correctly used child restraints reduce fatality risk by 71% for infants under 1 year old in passenger cars (58% in light trucks); 54% for children 1-4 in passenger cars (59% in light trucks); booster seats reduce serious injury risk by 45% for children 4-8 vs lap-and-shoulder belt alone","excerpt":"\"Child restraint systems, when used correctly, reduce the risk of fatal injury for infants under 1 year old by 71 percent in passenger cars and 58 percent in light trucks. For children aged 1 to 4 years, child restraints reduce fatal injury risk by 54 percent in passenger cars and 59 percent in light trucks. Booster seats, used as designed for children aged 4 to 8, reduce the risk of serious injury by 45 percent compared with the use of seat belts alone. Rear-facing child restraints provide additional protection by distributing crash forces across the entire back of the seat and supporting the child's head, neck, and spine; they are recommended for as long as the child fits within the manufacturer's height and weight limits.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260525093633/https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813047","calculation_notes":"NHTSA technical evaluation of child restraint system effectiveness. The 71%/54%/45% effectiveness rates are the canonical figures cited by AAP, IIHS, and state highway safety offices. Inverting to per-crash multipliers for unrestrained children: infants 1/(1-0.71)=3.45×, toddlers 1/(1-0.54)=2.17×, booster-age 1/(1-0.45)=1.82×. The 2.2× weighted-average risk-ratio in the native field reflects a population-weighted blend across age bands, lightly downweighted from the simple arithmetic mean (2.48×) because of the confounding overlap with other risk factors (driver impairment, vehicle age, nighttime driving).\n","independence_note":"NHTSA fatality-reduction estimates derive from NHTSA's own FARS crash database and the Pediatric Crashworthiness Research evaluations. CDC's observational data on unrestrained shares (in the first source) draws from the same FARS upstream feed but presents a different analytical layer (descriptive fatality profile vs causal effectiveness).\n"},{"url":"https://www.sciencedaily.com/releases/2016/04/160430100359.htm","title":"Exempt from passenger restraint laws, taxis pose risky rides for small children","publisher":"Pediatric Academic Societies / Cohen Children's Medical Center (via ScienceDaily)","source_type":"peer_reviewed","statistic":"Observational study of 116 children across 69 New York metropolitan taxis: only 11% of small children were properly restrained. Survey of 97 taxi companies found 39% reported car safety seat availability, with most imposing limitations (reservations, extra fees, quantity caps). 70% increased risk of death or injury for 7-8 year olds not properly restrained","excerpt":"\"Researchers observed 116 children across 69 taxis in the New York metropolitan area and found that only 11 percent of small children were properly restrained. When 97 taxi companies were surveyed, 39 percent reported car safety seat availability, with many imposing limitations like reservations, extra fees, or quantity restrictions. There were more than 40,000 motor vehicle collisions involving taxis, limousines, and car services in 2015 alone, and exemptions to car seat laws put unrestrained children at risk. The study found a 70 percent increased risk of death or injury for 7-to-8-year-olds not properly restrained compared with their appropriately restrained peers.\"\n","source_date":"2016-04-30","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20251012182932/https://www.sciencedaily.com/releases/2016/04/160430100359.htm","calculation_notes":"Koffsky & Milanaik 2016 (Pediatric Academic Societies meeting, Cohen Children's Medical Center). The 11% proper-restraint rate in taxis is the strongest direct measurement of the taxi/holiday exception — the population this entry exists to address. Holiday/vacation/airport transit periods concentrate this exposure: parents who normally use a child restraint at home often go without in a taxi, rideshare, hotel shuttle, or rental car. Most US states exempt taxis from child restraint laws (notable exceptions: NYC mandated 2017, California requires rideshare but exempts taxis). The 70% incremental injury/death risk for unrestrained 7-8 year olds is consistent with the NHTSA effectiveness data for that age band.\n"}],"comparison_anchors":[{"label":"Child traffic crash death — restrained baseline (lifetime through age 13)","lifetime_us_adult":0.00018},{"label":"Adult unbelted crash death (lifetime, US adult)","lifetime_us_adult":0.019},{"label":"Child drowning (lifetime, US through age 14)","lifetime_us_adult":0.0005}],"personal_factor_multipliers":[{"factor":"Infant (<1) unrestrained vs rear-facing seat","multiplier":3.45,"notes":"NHTSA: car seats reduce infant fatality risk by 71% in passenger cars; the per-crash multiplier without restraint is 1/(1-0.71) = 3.45×. Rear-facing geometry is critical — it distributes crash forces across the entire back of the seat rather than the three belt-anchor points"},{"factor":"Toddler (1-4) unrestrained vs forward-facing car seat","multiplier":2.17,"notes":"NHTSA: 54% fatality reduction for restrained toddlers in passenger cars; per-crash multiplier without restraint is 1/(1-0.54) = 2.17×"},{"factor":"Booster-age (4-8) seatbelt only vs proper booster","multiplier":1.82,"notes":"NHTSA: booster seats reduce serious injury risk by 45% vs lap-and-shoulder belt alone; multiplier 1/(1-0.45) = 1.82×. Lap belt alone (no shoulder belt) can cause submarining and abdominal injury in this age group"},{"factor":"Taxi or rideshare without child restraint (occasional/holiday exposure)","multiplier":2,"notes":"Koffsky 2016 observed 11% proper restraint in NYC taxis; most US states exempt taxis from child restraint laws. The per-trip multiplier reflects the same 1.8-2.5× per-crash risk as for any other unrestrained exposure — the law's exemption does not change the physics. Increased relevance during airport transit, vacation, hotel shuttle, and rental-car days when parents do not have their normal seat installed"},{"factor":"Driver also unrestrained","multiplier":1.6,"notes":"CDC: when drivers are unrestrained, children riding with them are also frequently unrestrained, and the driver-behavior correlate (higher speeds, impairment, less defensive driving) further elevates crash severity even controlling for child restraint status"},{"factor":"Rear-facing past age 2 (vs forward-facing)","multiplier":0.2,"notes":"Rear-facing seats are recommended as long as the child fits the manufacturer's height/weight limits (typically through age 2-4). Rear-facing reduces severe-injury risk roughly 5× vs forward-facing for toddlers in frontal crashes, the dominant crash mode"}],"short_label":"Child without restraint","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 2.2× per-crash risk ratio is a weighted average across infant, toddler, and booster-age groups; the per-event ratio is largest for infants (3.5×) and smallest for booster-age children (1.8×). Lifetime estimates assume a child remains unrestrained throughout childhood, which is the worst case; most US children are restrained on most trips, so the realistic personal estimate depends on the share of trips taken unrestrained. Taxi, rideshare, vacation, and hotel-shuttle exposure is the dominant \"occasional unrestrained\" pattern for otherwise-careful families — Koffsky 2016 found only 11% of small children in NYC taxis were properly restrained, and most US states exempt taxis from child restraint laws despite the underlying crash physics being unchanged. NHTSA fatality-reduction figures assume correct installation and correct use; misuse (loose harness, twisted straps, forward-facing too early, no top tether) substantially reduces protective effect, and roughly half of US child restraints in observational studies show some form of misuse. The entry does not address airbag-related risks for children improperly placed in the front seat (covered in [[kid-front-seat-airbag]]) or in-seat misuse such as bulky winter coats (covered in [[puffer-jacket-car-seat]]). For unrestrained adults, see [[unbelted-crash-death]], which provides a parallel adult-lifetime baseline.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"An empty child car seat secured by a harness in the back seat of a car, flat vector illustration, no people."},"canonical_url":"https://likelier.app/child-unrestrained-car-crash","api_url":"https://likelier.app/api/fears/child-unrestrained-car-crash.json"},{"slug":"war-civilian-casualty","question":"What are the odds of dying as a civilian in war?","quick_answer":"Averaged across the whole world, dying as a civilian in war runs about 1 in 2,000 over a lifetime — roughly 70,000 civilian conflict deaths a year against 8 billion people. That global average hides an enormous spread: near-zero for most readers, vastly higher inside active conflict zones.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"We have not yet found a rigorous global poll that isolates \"personal fear of being killed as a civilian in war\" as a standalone item, and national polls on war-related fear are heavily confounded by whether the respondent’s own country is currently at peace, at war, or adjacent to one. The perceived side here is therefore marked as editorial intuition rather than polled data. The plausible prior most readers carry is shaped less by a probability estimate than by whichever conflict happens to dominate their news feed on a given week.\n","rough_estimate":"distribution-dependent: near zero for most readers, very large for residents of active conflict zones","kind":"intuition"},"native":{"display":"~70,000 civilian conflict deaths per year (recent global average)","numerator":70000,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.0005,"display":"1 in ~2,000 lifetime (global adult)","log_value":-3.3,"assumptions":"Uses a recent-window global average of roughly 60,000–100,000 civilian deaths per year from organized violence, obtained by taking UCDP’s total organized-violence fatality counts for 2019–2024 (~80,000–160,000 per year) and applying a civilian share that the UCDP / PRIO literature places in the 40–60% range over this window, with recent conflicts in Gaza and Ukraine pushing the civilian share substantially higher in specific theatres (94% civilian in the UCDP classification of Middle East conflicts in 2024). Dividing a central estimate of ~70,000 civilian deaths per year by a global population of ~8 billion gives an annual per-capita hazard of ~8.75e-6, which compounded over 60 adult years gives ~5.25e-4, i.e. roughly 1 in ~1,900. The uncertainty band below reflects window choice and the civilian-share assumption, not sampling noise. Crucially, this is a global-average scale marker and not a personal estimate for any individual — see the regional_breakdown and the body text for the several-orders-of-magnitude spread across populations.\n","uncertainty":{"low":0.0002,"high":0.002},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.uu.se/en/press/press-releases/2025/2025-06-11-ucdp-sharp-increase-in-conflicts-and-wars","title":"UCDP: Sharp increase in conflicts and wars","publisher":"Uppsala University / Uppsala Conflict Data Program","source_type":"primary_study","statistic":"Nearly 160,000 people died in organized violence globally in 2024; 13,900 civilian deaths from targeted one-sided violence (+31% year-on-year); 94% of the ~26,000 deaths in two Middle East conflicts were civilians.","excerpt":"\"nearly 160,000 people died in organised violence during the year. [...] 13,900 civilian deaths in this type of targeted attack, an increase of 31 per cent compared with the previous year. [...] around 26,000 deaths in these two conflicts, 94 per cent of which were civilians. [...] 2024 was the fourth most violent year since the 1994 Rwandan genocide.\"\n","source_date":"2025-06-11","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260114035101/https://www.uu.se/en/press/press-releases/2025/2025-06-11-ucdp-sharp-increase-in-conflicts-and-wars","calculation_notes":"UCDP’s 2024 total of ~160,000 organized-violence deaths is the upper-window anchor. The 13,900 figure captures only targeted one-sided violence and understates total civilian deaths, because many civilian deaths occur as battle-related collateral in state-based and non-state conflicts (the 94% civilian share quoted for two Middle East conflicts implies tens of thousands of additional civilian deaths inside the state-based battle-related deaths total). A conservative central estimate of ~70,000 civilian deaths per year for the 2019–2024 window is obtained by applying a 40–50% civilian share to the ~160,000 total for 2022–2024 and a lower civilian share to the ~80,000 total UCDP reported for 2019.\n","independence_note":"UCDP is the upstream source for most other conflict-deaths aggregations including Our World in Data and the World Bank’s battle-related deaths indicator. Treat this as the primary anchor, not as an independent cross-check.\n"},{"url":"https://www.prio.org/publications/14006","title":"Conflict Trends: A Global Overview, 1946–2023","publisher":"Peace Research Institute Oslo (PRIO), Siri Aas Rustad","source_type":"primary_study","statistic":"More than 122,000 battle-related deaths in state-based conflicts in 2023; approximately 21,000 additional deaths from 75 non-state conflicts; one-sided violence against civilians recorded in 35 countries.","excerpt":"\"In 2023, 59 state-based conflicts were recorded in 34 countries, the highest number of conflicts registered since 1946. The wars in Ukraine and Gaza were the primary contributors to the more than 122,000 battle-related deaths in 2023. [...] In 2023, 75 non-state conflicts were recorded resulting in approximately 21,000 battle-related deaths. [...] One-sided violence against civilians was recorded in 35 countries in 2023.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260215091122/https://www.prio.org/publications/14006","calculation_notes":"PRIO’s 2023 total of ~143,000 battle-related deaths from state-based plus non-state conflicts, plus a separate one-sided-violence category, corroborates the UCDP order of magnitude for the recent-conflict-intensity window. PRIO reports this as the \"third most violent year since the end of the Cold War\", confirming that the 2022–2024 period sits at the high end of the post-1989 distribution.\n","independence_note":"PRIO and UCDP collaborate closely and share underlying event-coding conventions, but PRIO publishes its own annual trends report with independent aggregation choices. Treat as a methodologically aligned second opinion rather than a fully independent estimate.\n"},{"url":"https://ourworldindata.org/war-and-peace","title":"War and Peace","publisher":"Our World in Data (Bastian Herre, Lucas Rodés-Guirao, Max Roser)","source_type":"reputable_reference","statistic":"Close to 80,000 people died due to fighting in armed conflicts globally in 2019, roughly 1 in 700 deaths that year; fewer than 20,000 in 2005.","excerpt":"\"Globally, close to 80,000 people died due to fighting in armed conflicts in 2019. [...] This means conflicts caused around 1 in 700 deaths. [...] fewer than 20,000 people died in armed conflicts [in 2005]. [...] Since 1800, more than 37 million people worldwide have died while actively fighting in wars. [...] fewer people died in conflicts in recent decades than in most of the 20th century.\"\n","source_date":"2024-03-20","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260409003122/https://ourworldindata.org/war-and-peace","calculation_notes":"OWID’s 2019 figure of ~80,000 total armed-conflict deaths sits at the low-window end of our range and is used as the lower bound of the uncertainty band. The long-run statement that recent decades have lower absolute conflict deaths than most of the 20th century is the basis for treating the uncertainty band as wide rather than tight — pick a 20-year window inside a peaceful decade and the central estimate moves by a factor of several.\n","independence_note":"Our World in Data’s conflict-deaths series is built on UCDP data. Cited here for its longer historical framing and for independent editorial interpretation, not as a numerically independent estimate.\n"},{"url":"https://www.smallarmssurvey.org/sites/default/files/resources/GBAV2015-Full-Publication-ENG.pdf","title":"Global Burden of Armed Violence 2015: Every Body Counts","publisher":"Geneva Declaration Secretariat / Small Arms Survey / Cambridge University Press","source_type":"reputable_reference","statistic":"508,000 violent deaths/year (2007-2012) in conflict and non-conflict; direct conflict deaths rising from 55,000 to 70,000/year between 2004-09 and 2007-12 periods","excerpt":"\"An estimated 508,000 people died violently each year during the period 2007-12, in conflict and non-conflict settings combined.\"\n","source_date":"2015-05-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20240719192935/https://smallarmssurvey.org/sites/default/files/resources/GBAV2015-Full-Publication-ENG.pdf","calculation_notes":"Geneva Declaration uses WHO vital registration, national surveys, and its own estimation methodology in addition to UCDP data. The 70,000 direct conflict deaths/year figure is consistent with the UCDP 2024 estimate used in this entry's headline. Provides cross-validation from a genuinely different pipeline.\n","independence_note":"Partially independent of UCDP — uses UCDP as one input but combines it with WHO vital registration and its own survey methodology, producing independently derived totals.\n"}],"comparison_anchors":[{"label":"Death in a terrorist attack in the US (lifetime, US adult)","lifetime_us_adult":0.0000129},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Being murdered (lifetime, US, pooled)","lifetime_us_adult":0.00348}],"regional_breakdown":[{"region":"Global average (recent 10-year window)","probability":0.0005,"notes":"The headline number. A pooled average across 8 billion people, the vast majority of whom live outside any active conflict."},{"region":"Residents of high-intensity active conflict zones","probability":0.01,"notes":"Order-of-magnitude estimate for civilian populations inside the most lethal current theatres. Actual local rates vary by several orders of magnitude within a single country depending on frontline distance, siege conditions, and targeting."},{"region":"Residents of peaceful industrialized countries","probability":0.00001,"notes":"Essentially zero absent a major interstate war. This is the modal experience for most readers of an English-language website."},{"region":"Global average (smoothed across 2005–2015 low-conflict decade)","probability":0.00015,"notes":"The same calculation done over a quieter window produces a central estimate several times lower than the 2019–2024 window. Window choice is load-bearing."}],"personal_factor_multipliers":[{"factor":"resident of active conflict zone (Ukraine, Sudan, Gaza, Yemen)","multiplier":500,"notes":"UCDP/PRIO: civilian mortality in active conflict zones is orders of magnitude above global average; typical figures range from 0.1% to 5% annual"},{"factor":"resident of peaceful high-income country (no active conflict)","multiplier":0.01,"notes":"global average is dominated by conflict-zone mortality; a resident of a peaceful country faces near-zero war-civilian risk unless conflict emerges"},{"factor":"resident of fragile state near active conflict (Lebanon, DRC, Colombia)","multiplier":10,"notes":"spillover effects and intermittent conflict raise risk meaningfully without reaching active-conflict levels"},{"factor":"civilian living in or near urban combat zone (within 5 km of active front)","multiplier":20,"notes":"ACLED and OHCHR conflict event data show civilian casualty rates are dramatically higher in urban combat environments due to blast radius, shrapnel, and building collapse; OHCHR Ukraine monitoring reports document fatality rates per district that are 10-30x higher within 5 km of active front lines than in rear areas of the same conflict"},{"factor":"civilian infrastructure worker in conflict zone (utility, medical, aid)","multiplier":3,"notes":"ICRC reports and OHCHR conflict documentation consistently show that medical personnel, utility workers, and humanitarian aid workers in active conflict zones face elevated targeting risk and incidental harm; UN Office for the Coordination of Humanitarian Affairs tracks aid worker fatalities as a separate elevated-risk category"}],"short_label":"War (civilian)","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The \"1 in 2,000\" figure is a scale marker, not a personal estimate, and the dispersion around it is larger than for almost any other fear on this site. Residents of active conflict zones face risks that are orders of magnitude higher than this number, while residents of countries at peace face risks that are orders of magnitude lower. This entry covers civilian deaths in war. A separate entry covers deaths of serving military personnel, which is a distinct population with its own base rates. Counting civilian war deaths is also methodologically contested: UCDP’s coding requires identified events and is widely regarded as a conservative floor, while estimates that include excess mortality from war-induced disease, famine, and displaced-population health effects can be several times higher for the same conflicts. The figure used here is the UCDP-style direct-violence count.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single folded paper crane in muted grey against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/war-civilian-casualty","api_url":"https://likelier.app/api/fears/war-civilian-casualty.json"},{"slug":"heat-stroke-outdoor","question":"What are the odds of dying from heat stroke or severe heat illness during outdoor activity?","quick_answer":"The lifetime odds of dying from heat are about 1 in 1,900 for a US adult, and those deaths are dominated by the elderly without air conditioning and outdoor workers, not recreational hikers or athletes. That means the typical healthy adult overstates their own heat-death risk during summer activity by one to two orders of magnitude.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Heat stroke and severe dehydration occupy a reliable slot in the pre-summer anxiety calendar, particularly among parents sending children to outdoor camps, hikers planning desert trips, and anyone who has read a news story about a marathon runner collapsing. The fear is vivid and concrete enough that it drives a large market in electrolyte drinks, misting fans, and public-health campaigns, yet most healthy recreational adults would struggle to locate themselves on any actual risk distribution. The intuitive estimate — that serious heat illness is a meaningful personal risk during outdoor summer activity — overstates the typical healthy adult's exposure by at least one to two orders of magnitude while simultaneously underweighting the subgroups that carry the real mortality burden.\n","rough_estimate":"most recreational adults would guess 1-in-500 to 1-in-5,000 odds of serious harm during an active outdoor summer","kind":"intuition"},"native":{"display":"~2,325 heat-related deaths per year (age-adjusted rate 0.62 per 100,000; US, 2023)","numerator":2325,"denominator":335000000,"unit":"per year","population":"US total population, all ages (2023 NVSS; JAMA 2024)"},"normalized":{"lifetime_us_adult":0.000525,"display":"~1 in 1,900 lifetime (US adult)","log_value":-3.28,"assumptions":"Uses 2,325 US heat-related deaths in 2023 from JAMA/NVSS (Ramirez et al. 2024), the most recent year with published final counts, against the AAMR of 0.62 per 100,000 person-years. Applying the crude annual hazard (2,325 / ~261 million US adults aged 18+) yields ~8.9 per million adults per year. Compounded over 59 years of remaining adult life: 1 − (1 − 8.9e-6)^59 ≈ 0.000525, or roughly 1 in 1,900. The headline entry is framed as the general-population risk, which is dominated by non-recreational deaths (elderly at home without air conditioning, outdoor workers). Recreational and outdoor-activity heat deaths represent a small fraction of that total; most hiker/athlete heat fatalities appear in exertional heat stroke counts, not the broader NVSS tally. The uncertainty band spans: low end uses the 2004-2018 annual average of ~702 deaths (MMWR 2020) as a plausible floor under conservative cause-of-death coding; high end applies broader contributing-cause coding (approximately 5,000+ deaths/year) per EPA technical documentation.\n","uncertainty":{"low":0.00015,"high":0.0011},"scope":"us_adult_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jama/fullarticle/2822854","title":"Trends of Heat-Related Deaths in the US, 1999–2023","publisher":"JAMA (Ramirez et al., including CDC/NCHS co-authors)","source_type":"peer_reviewed","statistic":"2,325 heat-related deaths in 2023; age-adjusted mortality rate 0.62 per 100,000 person-years; 117% increase in count from 1999 (1,069) to 2023 (2,325); AAPC +16.8% per year from 2016–2023","excerpt":"\"The number of heat-related deaths increased from 1069 in 1999 to 2325 in 2023, a 117% increase in the number of heat-related deaths and a 63% increase in the age-adjusted mortality rate (AAMR). From 1999 to 2023, a total of 21,518 deaths were recorded as heat-related in the US. The AAMR increased from 0.38 per 100,000 person-years in 1999 to 0.62 per 100,000 person-years in 2023.\"\n","source_date":"2024-08-27","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260305025125/https://jamanetwork.com/journals/jama/fullarticle/2822854","calculation_notes":"Primary annual count and AAMR source. 2,325 deaths in 2023 against a US adult population of ~261 million (18+) yields an approximate annual adult hazard of 8.9 per million. Lifetime calculation: 1 − (1 − 8.9e-6)^59 ≈ 0.000525 ≈ 1 in 1,900. The JAMA paper uses NCHS NVSS ICD-10 codes X30 (exposure to excessive natural heat), W92 (exposure to excessive heat of man-made origin), and T67 (effects of heat and light) — the same coding base as the CDC MMWR 2004-2018 report.\n","independence_note":"Ramirez et al. draws directly from CDC WONDER / NVSS death-certificate data. The second source (CDC MMWR 2020) uses the same underlying NVSS pipeline for the earlier 2004-2018 period; treat the two as methodologically linked but covering non-overlapping date ranges.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/69/wr/mm6924a1.htm","title":"Heat-Related Deaths — United States, 2004–2018","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR), Vol. 69, No. 24","source_type":"govt_report","statistic":"Average 702 heat-related deaths per year, 2004–2018; 415 with heat as underlying cause, 287 as contributing cause; persons aged ≥65 years accounted for ~39% of deaths at a rate of 0.7 per 100,000","excerpt":"\"During 2004–2018, a total of 10,527 heat-related deaths occurred in the United States, an average of 702 per year. Of these, 6,221 (59.1%) had heat as the underlying cause of death, and 4,306 (40.9%) had heat as a contributing cause. Persons aged ≥65 years accounted for 4,019 (38.2%) of decedents; the rate of heat-related deaths among persons aged ≥65 years was 0.7 per 100,000.\"\n","source_date":"2020-06-18","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260513190247/https://www.cdc.gov/mmwr/volumes/69/wr/mm6924a1.htm","calculation_notes":"The 2004-2018 average of 702/year provides the low-end anchor for the uncertainty band. Using 702 deaths / ~245 million US adults circa 2011 midpoint ≈ 2.86 per million per year. Lifetime: 1 − (1 − 2.86e-6)^59 ≈ 0.000169 ≈ 1 in 5,900. Rounded to 0.00015 for the uncertainty low. The age distribution data (39% of deaths among persons ≥65) is used for the caveats heterogeneity section; it anchors the claim that mortality concentrates in elderly non-recreational decedents.\n","independence_note":"CDC MMWR report using NVSS/NCHS death-certificate data, same pipeline as JAMA 2024 Ramirez et al. but covering an earlier, lower-mortality period. Used here as a floor anchor for the uncertainty band and for the age-distribution breakdown.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6029a1.htm","title":"Nonfatal Sports and Recreation Heat Illness Treated in Hospital Emergency Departments — United States, 2001–2009","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR), Vol. 60, No. 29","source_type":"govt_report","statistic":"~5,946 persons treated annually in US EDs for sports/recreation heat illness, rate of 2.0 per 100,000 population; 7.1% of patients hospitalized; 72.5% male, 35.6% aged 15–19","excerpt":"\"An estimated 5,946 persons were treated in U.S. emergency departments (EDs) each year for a heat illness sustained while participating in a sport or recreational activity, for an estimated annual rate of 2.0 ED visits per 100,000 population. Incidence was highest among males (72.5%) and among those aged 15–19 years (35.6%), and 7.1% of patients were hospitalized.\"\n","source_date":"2011-07-29","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260503090854/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6029a1.htm","calculation_notes":"The sports/recreation-specific data reframes the broader mortality statistic: ~5,946 ED visits/year for recreational heat illness, of which ~7.1% = ~422 are hospitalized. Recreational heat illness hospitalizations are thus roughly one order of magnitude less common than heat illness requiring any ER visit. Fatal recreational heat stroke is a subset of hospitalizations and is not separately enumerated in this dataset; it represents a small fraction of the ~702-2,325 total annual heat deaths. This source is used to support the claim that the headline mortality figure is not primarily a recreational risk.\n","independence_note":"Uses National Electronic Injury Surveillance System-All Injury Program (NEISS-AIP), a probability sample of US hospital EDs — methodologically independent from the NVSS death-certificate pipeline used in the mortality sources.\n"},{"url":"https://www.osha.gov/heat-exposure/rulemaking","title":"Heat Injury and Illness Prevention in Outdoor and Indoor Work Settings Rulemaking","publisher":"US Department of Labor, Occupational Safety and Health Administration (OSHA)","source_type":"govt_report","statistic":"OSHA's first nationwide proposed heat standard, published in the Federal Register on August 30, 2024 (89 FR 70698; doc 2024-14824), covering both outdoor and indoor work settings across general industry, construction, maritime, and agriculture. Workplace heat-fatality counts are tracked by the BLS Census of Fatal Occupational Injuries: 48 US workers died from exposure to environmental heat in 2024 (NSC Injury Facts / BLS CFOI), an event type that 'can involve either indoor or outdoor exposure'.","excerpt":"\"The Heat Injury and Illness Prevention in Outdoor and Indoor Work Settings proposed rule was published in the Federal Register on August 30, 2024 ... Workers in outdoor and indoor work settings without adequate climate controls are at risk of hazardous heat exposure. Certain heat-generating processes, machinery, and equipment (e.g., hot tar ovens, furnaces, etc.) can also cause hazardous heat when cooling measures are not in place.\" (OSHA). \"Exposure to environmental heat resulted in 48 work-related deaths in 2024 ... This event type can involve either indoor or outdoor exposure.\" (NSC Injury Facts, citing the BLS Census of Fatal Occupational Injuries.)\n","source_date":"2024-08-30","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260703165115/https://www.osha.gov/heat-exposure/rulemaking","calculation_notes":"Supports the occupational/indoor caveat only; it does not change the headline number. On-the-job heat deaths are a small, separate slice of the all-adult NVSS tally that drives this entry: 48 CFOI worker deaths in 2024 against a US employed labor force of roughly 161 million is about 0.3 per million worker-years — far below this entry's ~8.9 per million general-adult annual hazard, so occupational heat death does not constitute a ≥2× personal multiplier on the population baseline. OSHA's proposed rule is cited because it establishes that hazardous workplace heat — and the deaths it causes — extends to indoor settings (warehouses, kitchens, foundries, hot-process industries), which this recreational/outdoor entry does not otherwise capture.\n","independence_note":"OSHA rulemaking page (regulatory/programmatic source) plus BLS CFOI worker-fatality counts surfaced via NSC Injury Facts. The CFOI worker-death pipeline is occupational (employer-reported / investigated workplace fatalities) and is distinct from the NVSS death-certificate pipeline used for the headline general-population figures, so it adds an independent occupational dimension rather than restating the mortality sources.\n"}],"comparison_anchors":[{"label":"Dying in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Fatal drowning (lifetime, US adult)","lifetime_us_adult":0.0009},{"label":"Lightning strike death (lifetime, US adult)","lifetime_us_adult":0.000083}],"personal_factor_multipliers":[{"factor":"outdoor manual worker (agriculture, construction, landscaping) in summer","multiplier":8,"notes":"OSHA and BLS data indicate outdoor workers account for a disproportionate share of heat fatalities. CDC analysis of occupational heat deaths 1992-2021 found ~33 worker heat deaths per year vs. baseline; agriculture and construction workers face the highest burden. An 8x multiplier relative to the general adult average is conservative.\n"},{"factor":"age 65 or older","multiplier":4,"notes":"CDC MMWR 2020: persons aged ≥65 had a heat-related mortality rate of 0.7 per 100,000 vs. the all-ages rate of ~0.5 per 100,000 circa 2004-2018 — approximately 1.4× the average, but concentrated in sedentary/indoor settings. In the 2023 elevated-mortality year, older adults remain disproportionately represented. The 4x estimate applies to elderly adults specifically engaged in outdoor exertion without adequate hydration or shade.\n"},{"factor":"no access to air conditioning in high heat-index region","multiplier":3,"notes":"The CDC MMWR noted higher heat death rates in large central metropolitan and non-metropolitan counties and in lower-income populations. Lack of home AC during heat waves is one of the strongest environmental risk amplifiers for heat mortality, particularly for elderly and chronically ill individuals.\n"},{"factor":"healthy adult, moderate outdoor recreation (hiking, cycling), acclimatized","multiplier":0.1,"notes":"The ~5,946 annual sports/recreation ED visits represent about 2 per 100,000 population — predominantly non-fatal. A healthy acclimatized adult engaging in standard recreational outdoor activity during summer faces substantially below-average fatal heat illness risk compared to the population mean, which is pulled up by elderly and occupational deaths.\n"}],"short_label":"Heat stroke (outdoor)","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline ~1-in-1,900 lifetime figure represents all heat-related deaths across the US population — including elderly people dying at home without air conditioning, outdoor laborers, and the unhoused — not specifically recreational or vacation-related heat illness. Recreational and sports-specific heat deaths account for a small fraction of the total; the CDC's sports/recreation data captures roughly 6,000 ER visits per year, with only ~7% requiring hospitalization and very few resulting in death. The 2023 total of 2,325 heat deaths reflects a sharply upward trend since 2016 (AAPC +16.8%/year) compared with the 2004-2018 average of ~702/year; the uncertainty band spans this historical range. Risk is highly heterogeneous: persons aged ≥65, outdoor workers, people without home air conditioning, and those with cardiovascular or renal disease carry a disproportionate share of mortality. A healthy, acclimatized recreational adult with access to water and shade faces substantially lower odds than the population average. Heat illness that requires ER evaluation is far more common than heat death but is rarely life-threatening in otherwise healthy people; the serious fear — fatal heat stroke — is concentrated in the high-risk subgroups named above. A distinct slice of heat mortality is occupational and on-the-job rather than recreational, and it is not confined to people working outdoors. OSHA's first nationwide heat standard, proposed in the Federal Register on August 30, 2024, deliberately covers both outdoor and indoor work settings — warehouses, commercial kitchens, foundries, and other hot-process industries where machinery and equipment (\"hot tar ovens, furnaces\") generate hazardous heat without adequate cooling. The BLS Census of Fatal Occupational Injuries recorded 48 US workers dead from environmental-heat exposure in 2024, a count BLS notes \"can involve either indoor or outdoor exposure.\" These on-the-job deaths are a small fraction (~0.3 per million worker-years) of the broader all-adult tally that drives the headline figure, and a weekend hiker carries none of this occupational exposure — but for paid work in un-cooled indoor or outdoor environments the risk is real and is the subject of active federal rulemaking.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A single water bottle casting a long shadow on a sunlit dusty trail, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/heat-stroke-outdoor","api_url":"https://likelier.app/api/fears/heat-stroke-outdoor.json"},{"slug":"food-poisoning-death","question":"What are the odds of dying from food poisoning?","quick_answer":"Dying from food poisoning is about a 1 in 1,860 chance over a US adult lifetime, from roughly 3,000 US deaths a year. Most people treat the tail risk as effectively zero for healthy adults, which understates it slightly.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Food poisoning is one of those fears that lives almost entirely in the aftermath of the last bad meal. Most people have had a memorable bout of it, most assume \"worst case I spend a night on the bathroom floor,\" and very few associate the words with a death certificate. There is no good survey of what fraction of US adults believe foodborne illness could kill them, so the best we can say is that the perceived tail risk is usually treated as effectively zero for healthy adults.\n","rough_estimate":"50% of US adults rank foodborne illness among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 50% of US adults rank foodborne illness (E. coli, Salmonella, Listeria) as a top-3 food safety concern","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~3,000 US foodborne illness deaths per year","numerator":1,"denominator":110000,"unit":"per year","population":"US residents, all ages, domestically acquired foodborne illness"},"normalized":{"lifetime_us_adult":0.000537,"display":"1 in ~1,860 lifetime (US adult)","log_value":-3.27,"assumptions":"Uses the CDC / Scallan et al. 2011 central estimate of ~3,000 US deaths per year from domestically acquired foodborne illness (known pathogens plus unspecified agents), against a US population of ~330 million, giving an annual rate of roughly 9.1 per million (≈ 0.91 per 100,000). Compounded over 59 years of remaining adult life: 1 - (1 - 9.1e-6)^59 ≈ 5.37e-4, or about 1 in 1,860. The uncertainty band is wide because Scallan's methodology imputes heavily from under-reported case data — the published 90% credible interval on the combined death estimate spans roughly 1,000-5,600 deaths per year. Excludes allergic reactions, deliberate poisoning, and non-foodborne gastrointestinal infections.\n","uncertainty":{"low":0.00018,"high":0.001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","title":"Foodborne Illness Acquired in the United States — Major Pathogens","publisher":"CDC Emerging Infectious Diseases / Scallan et al.","source_type":"peer_reviewed","statistic":"31 major pathogens cause ~9.4 million illnesses, ~56,000 hospitalizations, and ~1,351 deaths per year in the US","excerpt":"\"We estimated that 31 pathogens acquired in the United States caused 9.4 million episodes of foodborne illness (90% credible interval [CrI] 6.6–12.7 million), 55,961 hospitalizations (90% CrI 39,534–75,741), and 1,351 deaths (90% CrI 712–2,268) each year.\"\n","source_date":"2011-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420041128/https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","calculation_notes":"This is the \"known pathogens\" half of Scallan 2011. The companion paper on unspecified agents adds another ~1,686 deaths/year, bringing the combined total to ~3,037 deaths/year — the figure CDC cites publicly as \"~3,000 US deaths per year from foodborne illness.\" We use the combined total for the normalized lifetime calculation.\n","independence_note":"Scallan et al. (2011a) is the companion paper to Scallan et al. (2011b); they partition foodborne illness into known-pathogen vs unspecified-agent components from the same CDC FoodNet and outbreak surveillance data."},{"url":"https://wwwnc.cdc.gov/eid/article/17/1/p2-1101_article","title":"Foodborne Illness Acquired in the United States — Unspecified Agents","publisher":"CDC Emerging Infectious Diseases / Scallan et al.","source_type":"peer_reviewed","statistic":"Unspecified agents add ~38.4 million illnesses, ~71,878 hospitalizations, and ~1,686 deaths per year; combined total ~47.8 million illnesses, ~127,839 hospitalizations, ~3,037 deaths","excerpt":"\"We estimated that acute gastroenteritis caused 179 million episodes annually. After adjusting for non-foodborne transmission [...] and for acute, nongastroenteritis illness, an estimated 38.4 million (90% CrI 19.8–61.2 million) episodes of domestically acquired foodborne illness from unspecified agents occurred annually, resulting in 71,878 hospitalizations (90% CrI 9,924–157,340) and 1,686 deaths (90% CrI 369–3,338). Overall, we estimated that each year 47.8 million episodes of domestically acquired foodborne illness occur, resulting in 127,839 hospitalizations and 3,037 deaths.\"\n","source_date":"2011-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172029/https://wwwnc.cdc.gov/eid/article/17/1/p2-1101_article","calculation_notes":"The combined ~3,037 deaths/year is the basis for the normalized figure. 3,037 / 3.3e8 ≈ 9.2 per million per year. Over 59 adult-remaining years: 1 - (1 - 9.2e-6)^59 ≈ 5.4e-4, or about 1 in 1,850. Rounded to 5.37e-4 (1 in 1,860). The 90% CrI on the unspecified-agents death estimate alone (369–3,338) is nearly an order of magnitude wide, which is why our normalized uncertainty band stretches from roughly 1 in 5,500 to 1 in 1,000 lifetime.\n","independence_note":"The two Scallan 2011 papers are companion pieces from the same author team and methodology — they are not independent estimates of the same quantity, they are complementary partitions of the total burden (known pathogens + unspecified agents).\n"},{"url":"https://www.cdc.gov/food-safety/about/index.html","title":"About Food Safety","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC estimates 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths per year from foodborne illness in the US","excerpt":"\"CDC estimates that each year 48 million people get sick from a foodborne illness, 128,000 are hospitalized, and 3,000 die.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260329223244/https://www.cdc.gov/food-safety/about/index.html","calculation_notes":"CDC's public-facing page restates the Scallan 2011 totals as round numbers. This is the canonical figure cited in policy and press. Used here to confirm that the ~3,000/year estimate remains the current CDC headline fifteen years after publication, not as an independent estimate.\n","independence_note":"Derived directly from Scallan 2011 — treat as confirmation of continuing use, not as an independent data point.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/food-safety","title":"Food Safety — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Globally, an estimated 600 million foodborne illnesses and 420,000 deaths annually; children under 5 account for 125,000 deaths","excerpt":"\"An estimated 600 million — almost 1 in 10 people in the world — fall ill after eating contaminated food and 420 000 die every year [...] Children under 5 years of age carry 40% of the foodborne disease burden, with 125 000 deaths every year.\"\n","source_date":"2024-05-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420041201/https://www.who.int/news-room/fact-sheets/detail/food-safety","calculation_notes":"WHO's 420,000 global deaths / ~8 billion population ≈ 5.3 per 100,000 per year, roughly 6x the US per-capita rate. The gap reflects the concentration of foodborne mortality in low- and middle-income countries with less developed cold chains, sanitation, and clinical care. Used as an order-of-magnitude cross-check and to frame that the US number is an optimistic baseline relative to the global picture.\n","independence_note":"WHO's Foodborne Disease Burden Epidemiology Reference Group (FERG) estimates are methodologically adjacent to the Scallan approach and share some input data — treat as related-but-not-identical rather than fully independent.\n"},{"url":"https://pirg.org/media-center/new-report-hospitalizations-deaths-from-contaminated-food-doubled-in-2024/","title":"NEW REPORT: Hospitalizations, deaths from contaminated food doubled in 2024","publisher":"US PIRG Education Fund","source_type":"reputable_reference","statistic":"Recall-linked foodborne outbreak deaths rose from 8 (2023) to 19 (2024) and hospitalizations from 230 to 487, even as total food recalls fell ~5% to 296","excerpt":"\"Nearly 1,400 people became ill from food they ate in 2024 – 98% of them from just 13 outbreaks. [...] All but one of the outbreaks involved Listeria, Salmonella or E. coli. [...] More than 500 people were hospitalized or died in 2024 from contaminated food, up from about 240 in 2023. The number of recalls because of Listeria, Salmonella or E. coli increased significantly and comprised 39% of all recalls.\"\n","source_date":"2025-02-18","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250323111554/https://pirg.org/media-center/new-report-hospitalizations-deaths-from-contaminated-food-doubled-in-2024/","calculation_notes":"The PIRG Food for Thought 2025 report (covering 2024) tallies only recall-linked outbreaks with confirmed illnesses, not the population-level burden. Per the underlying report (as summarized by CIDRAP, 2025-02-18), confirmed outbreak deaths rose from 8 in 2023 to 19 in 2024 and hospitalizations from 230 to 487, while total illnesses rose from 1,118 to 1,392 — even though overall recalls fell ~5% to 296. The 2024 surge was driven by two large events: the Boar's Head Listeria outbreak (61 sickened, 60 hospitalized, ~10 deaths) and the McDonald's Quarter Pounder E. coli O121:H19 outbreak (104 sickened, 34 hospitalized, 1 death). These outbreak counts are on a completely different (recall-traced) denominator than the ~3,000 deaths/year population estimate behind the normalized figure, so they do NOT move the ~1-in-1,860 lifetime number — they are cited only as a recent-trend caveat that severe and fatal outcomes ticked up year-over-year in 2024.\n","independence_note":"Independent of the Scallan/CDC estimates — PIRG compiles FDA and USDA recall and outbreak announcements directly, a recall-traced tally rather than a modeled population-burden estimate. CIDRAP (Univ. of Minnesota) provided secondary coverage of the same report.\n"}],"comparison_anchors":[{"label":"Death by drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Age 65+","multiplier":5,"notes":"CDC FoodNet active surveillance data show adults 65+ have roughly 5× the foodborne illness hospitalization rate of 18–64 year-olds for priority pathogens (Salmonella, Listeria). Listeria case-fatality rates for seniors approach 20–30% vs ~2% for healthy young adults. Source: CDC FoodNet Surveillance Report (2021); Scallan et al. Emerging Infectious Diseases 2011."},{"factor":"Immunocompromised (transplant, chemotherapy, advanced HIV)","multiplier":100,"notes":"CDC estimates immunocompromised individuals face ~100× higher risk of invasive Listeriosis compared to the general population. Listeria monocytogenes is the leading foodborne killer in this group, with a case-fatality rate of ~20%. Source: CDC, 'Listeria (Listeriosis) — People at Risk' (2024); Scallan et al. (2011)."},{"factor":"Pregnancy","multiplier":10,"notes":"Pregnant women are approximately 10× more likely to develop Listeriosis than the general population, due to pregnancy-related immune suppression. Listeria infection during pregnancy carries ~22% fetal or neonatal mortality. Source: ACOG Practice Bulletin; CDC Listeria surveillance data (2024)."},{"factor":"Residence in or frequent exposure to care home / institutional setting","multiplier":3,"notes":"Institutional settings (nursing homes, long-term care facilities) account for a disproportionate share of multi-person foodborne illness outbreaks — CDC NORS data (2009–2018) show healthcare facilities contributed roughly 3× the per-person outbreak incidence vs households, driven by Norovirus and Salmonella. Source: CDC National Outbreak Reporting System (NORS) Annual Reports."}],"short_label":"Food poisoning (US)","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"Excludes allergic reactions (anaphylaxis to food allergens is coded separately), deliberate poisoning, and gastrointestinal infections transmitted person-to-person or via water rather than food. The Scallan estimates are modeled from passive surveillance data with substantial under-reporting, so the true number could plausibly be anywhere in the low thousands to mid-five-thousands per year — the uncertainty band on the normalized figure reflects that. Risk is also highly heterogeneous: immunocompromised adults, pregnant women, adults over 65, and infants carry per-capita risks several times higher than the all-ages average, driven largely by Listeria, invasive Salmonella, and Toxoplasma. Recent-trend note: severe and fatal outcomes from recall-linked outbreaks roughly doubled year-over-year in 2024 even as overall food recalls fell ~5% — confirmed outbreak deaths rose from 8 (2023) to 19 (2024) and hospitalizations from 230 to 487 (US PIRG Food for Thought 2025; CIDRAP), driven mainly by the Boar's Head Listeria (~10 deaths) and McDonald's Quarter Pounder E. coli events. These are recall-traced outbreak counts in the low tens, not a change to the ~3,000-deaths-per-year population estimate, so they do not move the headline 1-in-1,860 lifetime figure; they signal that when contamination does occur it has recently involved larger, more severe clusters.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted ceramic plate viewed from above on a pale grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/food-poisoning-death","api_url":"https://likelier.app/api/fears/food-poisoning-death.json"},{"slug":"home-fire-death","question":"What are the odds of dying in a house fire?","quick_answer":"The lifetime odds of dying in a home fire are about 1 in 1,800 for a US adult, from roughly 2,600 civilian home-fire deaths a year. Most people know a working smoke alarm matters but cannot name the actual number, so this risk is calibrated in direction if not in magnitude.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"House fires occupy a peculiar spot in the American risk imagination. Almost every reader has stood next to a smoke alarm, read a hotel evacuation card, or rehearsed a family escape plan, and most understand intuitively that a working smoke alarm matters a lot. What almost nobody can name is the actual number: the annual US home fire death toll, the per-capita rate, or the lifetime odds. The direction of belief is roughly right — fire is real, smoke alarms help — but the magnitude is essentially unknown to the general public, which is why this entry is tagged as calibrated rather than debunked or underrated.\n","rough_estimate":"Most adults know smoke alarms matter but cannot name the lifetime number","kind":"intuition"},"native":{"display":"~7.8 home fire deaths per million US population per year","numerator":1,"denominator":128000,"unit":"per year","population":"US residents, all ages, civilian deaths in home structure fires"},"normalized":{"lifetime_us_adult":0.00055,"display":"1 in ~1,800 lifetime (US adult)","log_value":-3.26,"assumptions":"NFPA reports an annual average of roughly 2,600 civilian deaths in home structure fires across 2019-2023, against a US population of ~333 million, giving a crude rate of about 7.8 per million per year (1 in ~128,000). USFA's 2023 per-capita series gives 13.1 deaths per million for all structure fires combined; the home-only subset is smaller. Taking 7.8 per million per year as the home-fire input and compounding across 59 years of remaining adult life gives 1 - (1 - 7.8e-6)^59 ≈ 4.6e-4, or about 1 in 2,170. Rounded up to 5.5e-4 (≈ 1 in 1,820) to account for the fact that adults 65+ experience fire death rates roughly 2-3× the all-ages baseline and most cohorts will pass through that band. Excludes wildfire fatalities, which are coded separately and contribute a small fraction of the total.\n","uncertainty":{"low":0.0004,"high":0.0007},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-structure-fires","title":"Home Structure Fires","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"US fire departments responded to an estimated average of 328,590 home structure fires per year during 2019-2023, causing an annual average of 2,600 civilian deaths and 10,770 civilian injuries","excerpt":"\"Local fire departments responded to an estimated average of 328,590 home structure fires per year in 2019-2023. These fires caused an annual average of 2,600 civilian deaths; 10,770 civilian fire injuries; and $8.9 billion in direct property damage.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172622/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-structure-fires","calculation_notes":"NFPA's 2,600 deaths/year over a US population of ~333 million yields a crude rate of 7.81 per million per year (1 in ~128,000 per year). Compounded over 59 adult-remaining years at constant rate: 1 - (1 - 7.81e-6)^59 ≈ 4.6e-4. Adjusted to 5.5e-4 (≈ 1 in 1,820) to reflect that home fire fatality rates are elevated in adults 65+, a band most adult readers will pass through. NFPA also reports that \"the death rate per 1,000 home structure fires is approximately 60 percent lower in homes with working smoke alarms than in homes with no alarms or none that operated,\" which anchors the personal-factor multipliers below.\n","independence_note":"NFPA aggregates NFIRS (National Fire Incident Reporting System) data with its own fire-department survey; shares the NFIRS upstream with USFA but adds independent analytical methodology and multi-year trend analysis."},{"url":"https://www.usfa.fema.gov/statistics/residential-fires/","title":"Residential Building Fire Estimates","publisher":"US Fire Administration (USFA), FEMA","source_type":"govt_report","statistic":"2,890 deaths from residential building fires in 2023; 344,600 fires; 10,400 injuries; $11.27 billion in dollar loss","excerpt":"\"In 2023, an estimated 344,600 residential building fires were reported to fire departments in the United States. These fires caused an estimated 2,890 deaths, 10,400 injuries and $11,266,200,000 in dollar loss.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172658/https://www.usfa.fema.gov/statistics/residential-fires/","calculation_notes":"USFA's 2023 single-year figure of 2,890 deaths runs slightly above NFPA's five-year average of 2,600, consistent with the USFA observation that the 10-year fire death rate per million population rose 27% through 2023. Used as the corroborating single-year anchor; the normalized number uses the NFPA five-year average as the more stable input to avoid being dominated by any single year.\n","independence_note":"USFA and NFPA draw on overlapping incident reporting pipelines — chiefly the National Fire Incident Reporting System (NFIRS) — plus NFPA's own survey of US fire departments. Treat the two figures as methodologically linked rather than fully independent; they agree on order of magnitude and on the direction of the 10-year trend.\n"},{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/smoking-materials","title":"Home Fires Started by Smoking Materials","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"Smoking materials (cigarettes, pipes, cigars) started an estimated 15,209 home structure fires per year during 2019-2023, causing an annual average of 590 civilian deaths","excerpt":"\"During 2019-2023, there was an estimated annual average of 15,209 reported home structure fires that were started by smoking materials. These fires caused an average of 590 civilian deaths and 1,048 civilian injuries per year.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172733/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/smoking-materials","calculation_notes":"Used as the authoritative anchor for the \"smoking materials are the leading cause of home fire deaths\" framing and for the smoker-in-household multiplier. 590 smoking-related deaths / 2,600 total home fire deaths ≈ 23% of all US home fire fatalities trace to a smoking-material ignition, even though smoking causes well under 10% of all home fires — the disproportion is what drives the 1.7× household multiplier used below.\n","independence_note":"Shares the NFIRS-plus-NFPA-survey pipeline with the top-level NFPA home structure fires report; not independent, used for cause-specific decomposition.\n"},{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-cooking-fires","title":"Home Cooking Fires","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"42% of reported home fires started in the kitchen or cooking area; cooking fires are the leading cause of home fire injuries (37%)","excerpt":"\"Forty-two percent of reported home fires started in the kitchen or cooking area. These fires were the leading cause of home fire injuries (37 percent).\"\n","source_date":"2023-09-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260709061912/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-cooking-fires","calculation_notes":"Corroborates the body-text framing that cooking is the leading cause of home fires by volume (42% of fires start in the kitchen/cooking area) while causing a comparatively small share of home fire deaths. Not used in the native/normalized calculation; cited for the cause-distribution framing only.\n","independence_note":"Same NFIRS-plus-NFPA-survey pipeline as the other NFPA sources in this entry; cited here for cause-specific decomposition only.\n"}],"comparison_anchors":[{"label":"Choking death (lifetime, US adult)","lifetime_us_adult":0.00091},{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Accidental fall death (lifetime, US adult)","lifetime_us_adult":0.0074},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"resident in home with working smoke alarms","probability":0.00027,"notes":"NFPA: fatality rate in homes with working smoke alarms is roughly half the overall average"},{"region":"resident in home without smoke alarms","probability":0.0011,"notes":"NFPA: ~40% of fire deaths occur in homes without smoke alarms, despite only ~3% of homes lacking them"},{"region":"resident age 65+, limited mobility","probability":0.0015,"notes":"elderly mobility-limited residents face ~3x population-average fatality rate"}],"personal_factor_multipliers":[{"factor":"working smoke alarm present","multiplier":0.5,"notes":"NFPA: death rate per 1,000 home fires is ~60% lower in homes with working alarms vs none or non-operational"},{"factor":"no working smoke alarm","multiplier":2.5,"notes":"Roughly three of every five US home fire deaths occur in homes where no alarm was present or the alarm failed to operate"},{"factor":"age 65+","multiplier":2.5,"notes":"Older adults carry fire death rates several times the all-ages baseline"},{"factor":"smoker in household","multiplier":1.7,"notes":"Cigarettes and other smoking materials are the leading cause of home fire deaths, accounting for ~23% of all home fire fatalities"},{"factor":"sprinkler system in home","multiplier":0.2,"notes":"Smoke alarms combined with automatic extinguishing systems are associated with a ~90% lower death rate vs homes with neither"}],"short_label":"House fire","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"property","valence":"negative","caveats":"Excludes deaths in wildfires and vehicle fires, which are coded separately. The normalized figure is an all-ages average applied to a generic US adult lifetime; per-capita home fire death rates are meaningfully higher in adults 65+, in the very young, and in households without working smoke alarms, and meaningfully lower in households with smoke alarms, sprinklers, and no smoking materials. NFPA's home-structure-fire total (~2,600/year, 2019-2023) and USFA's residential-building-fire total (~2,890 in 2023) overlap but are not identical definitions: USFA's residential-building category includes dormitories, hotels, and other residential structures beyond one- and two-family dwellings and apartments, while NFPA's \"home\" category is narrower. The two agree on order of magnitude but differ by ~10% in any given year.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small white smoke alarm disc mounted on a pale grey ceiling, viewed from below, flat vector illustration."},"canonical_url":"https://likelier.app/home-fire-death","api_url":"https://likelier.app/api/fears/home-fire-death.json"},{"slug":"femicide-intimate-partner","question":"What are the odds of being killed by an intimate partner?","quick_answer":"A US woman's odds of being killed by an intimate partner are about 0.96 per 100,000 per year, or roughly 1 in 1,770 over an adult lifetime. Most women hold no quantified estimate, and advocacy framing tends to convey it as disturbingly common.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Intimate-partner homicide occupies a peculiar position in public risk perception. It is simultaneously the subject of intense advocacy attention and strikingly absent from standard crime-fear polls, which tend to ask about stranger violence. Most women, when asked informally, place their risk of being killed by a partner somewhere between \"negligible\" and \"something that happens to other people.\" Domestic-violence awareness campaigns, meanwhile, emphasise frequency and severity in ways that can make the risk feel omnipresent. Neither framing tracks the epidemiology particularly well.\n","rough_estimate":"Most women have no quantified estimate; advocacy framing suggests 'disturbingly common'","kind":"intuition"},"native":{"display":"~0.96 per 100,000 women per year","numerator":96,"denominator":10000000,"unit":"per year","population":"US women, all ages, 2018-2021 average (CDC NVDRS)"},"normalized":{"lifetime_us_adult":0.000566,"display":"~1 in 1,770 over an adult lifetime (US woman)","log_value":-3.25,"assumptions":"CDC MMWR (August 2024) reports intimate-partner homicide rates among US women of 0.97 per 100,000 (2018-2019) and 0.95 per 100,000 (2020-2021), averaging ~0.96 per 100,000 per year. BJS separately finds that 34% of the ~4,970 female murder victims in 2021 were killed by intimate partners, yielding ~1,690 IPH deaths that year — consistent with the NVDRS rate applied to the ~175 million US female population. Compounded over 59 years of adult life: 1 − (1 − 9.6e-6)^59 ≈ 5.66e-4, or roughly 1 in 1,770. This is a population average across all US women regardless of relationship status. The uncertainty band reflects both measurement variation (NVDRS vs. BJS figures differ slightly) and year-to-year fluctuation. For US men, the intimate- partner homicide rate is roughly one-fifth the female rate.\n","uncertainty":{"low":0.0004,"high":0.00085},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7334a4.htm","title":"Notes from the Field: Intimate Partner Homicide Among Women — United States, 2018–2021","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"3,991 female victims of intimate partner homicide, 2018-2021; rates of 0.97 per 100,000 (2018-2019) and 0.95 per 100,000 (2020-2021)","excerpt":"\"During 2018–2021, a total of 3,991 female victims of intimate partner homicide were identified [...] Rates of intimate partner homicide during 2018–2019 (0.97 per 100,000) and 2020–2021 (0.95) were not significantly different.\"\n","source_date":"2024-08-29","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260420040315/https://www.cdc.gov/mmwr/volumes/73/wr/mm7334a4.htm","calculation_notes":"Primary native figure: ~0.96 per 100,000 women per year (average of 2018-2019 and 2020-2021 periods). 3,991 victims over 4 years ≈ 998 per year. Compounded: 1 − (1 − 9.6e-6)^59 ≈ 5.66e-4. The NVDRS covered 48 states plus DC by 2021, providing near-national coverage.\n"},{"url":"https://bjs.ojp.gov/female-murder-victims-and-victim-offender-relationship-2021","title":"Female Murder Victims and Victim-Offender Relationship, 2021","publisher":"Bureau of Justice Statistics (BJS)","source_type":"govt_report","statistic":"34% of ~4,970 female murder victims in 2021 were killed by an intimate partner; 76% were killed by someone known to them","excerpt":"\"Of the estimated 4,970 female victims of murder and nonnegligent manslaughter in 2021, 34% were killed by an intimate partner [...] About 6% of the 17,970 males murdered that year were victims of intimate partner homicide.\"\n","source_date":"2022-11-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420040343/https://bjs.ojp.gov/female-murder-victims-and-victim-offender-relationship-2021","calculation_notes":"34% of 4,970 ≈ 1,690 female IPH victims in 2021. Applied to ~175 million US females: 1,690 / 175,000,000 ≈ 9.66 per million ≈ 0.97 per 100,000. Cross-checks the NVDRS rate to within rounding. The 6% figure for males (≈1,078 male IPH victims) confirms the strong gender disparity: women are ~5x more likely to be killed by an intimate partner per unit of population.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10933122/","title":"Femicide in the United States: a call for legal codification and national surveillance","publisher":"Frontiers in Public Health","source_type":"peer_reviewed","statistic":"Multiple-source estimates suggest about half of female homicide victims in the US are killed by intimate partners (vs. one-third in the single-year FBI count)","excerpt":"\"According to the U.S. Federal Bureau of Investigation (FBI), an estimated 4,970 female victims were murdered in 2021, one third of whom were documented to have been killed by an intimate partner [...] Other estimates utilizing multiple data sources suggest that half of female victims of homicide in the U.S. are killed by intimate partners.\"\n","source_date":"2024-02-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260420040522/https://pmc.ncbi.nlm.nih.gov/articles/PMC10933122/","calculation_notes":"Provides broader framing and confirms that IPV-related homicide constitutes a major category of female homicide in the US. The article reports the single-year FBI count (one third of 4,970 female murder victims in 2021 killed by an intimate partner) and notes that \"other estimates utilizing multiple data sources\" put the share at about half — bracketing the BJS 34% figure. Used for contextual cross-check, not for the primary rate calculation. This source does not address firearm involvement.\n"}],"comparison_anchors":[{"label":"Homicide from any cause (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Black women","multiplier":3,"notes":"CDC MMWR: Black women constituted ~13% of the population but ~30% of IPH victims during 2018-2021; the disparity widened during 2020-2021"},{"factor":"women aged 18-44 vs. all ages","multiplier":1.8,"notes":"Reproductive-age women have higher IPH rates than older women; risk concentrates during peak relationship-formation years"},{"factor":"recent separation from partner","multiplier":5,"notes":"Multiple studies identify the separation period as the highest-risk window for IPH; risk is elevated for months to a year after leaving"},{"factor":"male victim","multiplier":0.2,"notes":"BJS 2021: ~6% of male murders were by intimate partners vs 34% of female murders; per-capita male IPH rate is roughly one-fifth the female rate"},{"factor":"household with firearms","multiplier":2.5,"notes":"Firearms were used in ~67% of IPH cases (CDC MMWR); access to firearms in the home is a well-documented risk factor for IPH"}],"short_label":"Intimate-partner homicide","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The headline number is a population average across all US women regardless of relationship status, age, or demographics. It therefore understates the risk for women in abusive relationships and overstates it for women who are not in relationships or who have low-risk partnerships. The strongest individual predictor of IPH is a prior history of intimate-partner violence in the relationship. The gender disparity is stark — roughly 5:1 female-to-male — but male victims of IPH do exist and are likely undercounted due to reporting barriers. Racial disparities are substantial and persistent: Black women face approximately 3x the IPH rate of White women. The period immediately following separation from an abusive partner is the single highest-risk window, which complicates simplistic \"just leave\" advice. Firearms are present in roughly two-thirds of IPH cases, and state-level firearm access laws correlate with IPH rates in ecological analyses. The NVDRS data used here covered 48 states plus DC by 2021 but was not fully national in the earlier years of the 2018-2021 window. This entry complements the general homicide-us entry by focusing on the intimate-partner subset and its distinctive risk profile.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":2,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A closed door with a thin shadow across the threshold against a muted background, flat vector illustration."},"canonical_url":"https://likelier.app/femicide-intimate-partner","api_url":"https://likelier.app/api/fears/femicide-intimate-partner.json"},{"slug":"mercury-from-fish","question":"What are the odds that eating fish regularly will harm you from mercury exposure?","quick_answer":"For the low-mercury species that make up most US seafood, the lifetime risk of clinical harm from mercury is about 1 in 1,700 for a typical consumer — even though 41% of US adults rank heavy metals among their top food-safety concerns. Caution is genuinely warranted mainly for pregnant women eating high-mercury predator species.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Mercury in fish is one of the most durable food-safety anxieties in the US. The 2004 FDA/EPA joint advisory warning pregnant women about methylmercury landed hard in public consciousness and never fully left. Surveys find that a substantial share of US adults — particularly women of childbearing age — avoid or limit fish consumption specifically because of mercury fears. The irony is well documented: the advisory itself noted that most commercial fish species are low-mercury, but the takeaway that stuck was \"fish = mercury = danger.\" Consumers routinely overestimate the risk from salmon, shrimp, and canned light tuna while underestimating the cardiovascular and neurodevelopmental benefits of eating more fish.\n","rough_estimate":"41% of US adults rank heavy metals in food among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 41% of US adults rank heavy metals in food as a top-3 food safety concern; mercury in fish is the most prominent heavy-metal food-safety issue","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~1 in 100,000 per year (attributable clinical harm, typical US fish consumer)","numerator":1,"denominator":100000,"unit":"per year (attributable clinical harm)","population":"US adults consuming typical commercial fish species"},"normalized":{"lifetime_us_adult":0.00059,"display":"~1 in 1,700 lifetime (US adult, typical fish consumer)","log_value":-3.23,"assumptions":"FDA mercury monitoring data shows that the species comprising >90% of US seafood consumption (shrimp, salmon, canned light tuna, tilapia, pollock, cod, catfish, crab, clams, pangasius) have mean mercury concentrations of 0.01-0.12 ppm, well below the EPA reference dose of 0.1 ug/kg/day for a 70 kg adult eating two servings per week. Clinical methylmercury toxicity at dietary exposure levels is essentially undocumented in the general US adult population eating commercial seafood. The 1-in-100,000 per-year native figure is a conservative upper bound acknowledging theoretical risk from cumulative low-level exposure; the 59-year lifetime conversion yields ~1 in 1,700. The wide uncertainty band reflects the gap between \"no observed clinical harm at typical exposures\" and the precautionary possibility that subtle neurocognitive effects exist below current detection thresholds. Mozaffarian & Rimm (2006) concluded that the net health effect of fish consumption is overwhelmingly positive — avoiding fish to dodge mercury is, for most people, the riskier choice.\n","uncertainty":{"low":1e-7,"high":0.005},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fda.gov/food/environmental-contaminants-food/technical-information-development-fdaepa-advice-about-eating-fish-those-who-might-become-or-are","title":"Technical Information on Development of FDA/EPA Advice about Eating Fish","publisher":"US Food and Drug Administration / US Environmental Protection Agency","source_type":"govt_report","statistic":"FDA/EPA classify commercial fish into Best Choices (≤0.15 µg/g mercury, 2-3 servings/week), Good Choices (0.15-0.46 µg/g, 1 serving/week), and Choices to Avoid (>0.46 µg/g)","excerpt":"\"Fish with an average mercury concentration less than or equal to 0.15 µg/g was placed in the 'Best Choices – eat 2 to 3 servings a week' category … Fish with an average mercury concentration greater than 0.15 µg/g up to 0.23 µg/g was placed in the 'Good Choices – eat 1 serving a week' category … Fish with an average mercury concentration greater than 0.46 µg/g was placed in the 'Choices to Avoid' category.\"\n","source_date":"2021-10-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260513103731/https://www.fda.gov/food/environmental-contaminants-food/technical-information-development-fdaepa-advice-about-eating-fish-those-who-might-become-or-are","calculation_notes":"The original consumer advice URL (fda.gov/food/consumers/advice-about-eating-fish) now returns 404; the technical information page remains available and contains the underlying mercury thresholds. The FDA/EPA advisory classifies commercial fish into three tiers by mercury concentration: Best Choices (≤0.15 µg/g, 2-3 servings/week), Good Choices (0.15-0.46 µg/g, 1 serving/week), and Choices to Avoid (>0.46 µg/g — shark, swordfish, king mackerel, tilefish, bigeye tuna). The species in \"Best Choices\" account for >90% of US seafood consumption by volume. At 2-3 servings/week of Best Choices fish, methylmercury intake stays well below the EPA reference dose (0.1 µg/kg/day), which itself incorporates a 10x safety factor below the no-observed-adverse-effect level.\n","independence_note":"FDA/EPA advisory is the primary US regulatory guidance on fish-mercury exposure, based on independent federal risk assessments and monitoring data, not derived from the Mozaffarian & Rimm or Oken & Bellinger academic analyses.\n"},{"url":"https://jamanetwork.com/journals/jama/fullarticle/203640","title":"Fish Intake, Contaminants, and Human Health: Evaluating the Risks and the Benefits","publisher":"JAMA / Mozaffarian & Rimm","source_type":"peer_reviewed","statistic":"Modest fish consumption (1-2 servings/week) reduces coronary death risk by 36%; benefits far exceed methylmercury risks for all populations except possibly women of childbearing age consuming high-mercury species","excerpt":"\"For major health outcomes among adults, based on both the strength of the evidence and the potential magnitudes of effect, the benefits of fish intake exceed the potential risks. For women of childbearing age, the benefits of modest fish intake, excepting a few selected species, also outweigh risks.\"\n","source_date":"2006-10-18","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260204141259/https://jamanetwork.com/journals/jama/fullarticle/203640","calculation_notes":"Mozaffarian & Rimm conducted a systematic review of fish consumption, omega-3 fatty acids, and contaminant exposure. They found that 1-2 servings/week of fish reduced coronary heart disease mortality by 36% (RR 0.64, 95% CI 0.46-0.89) and total mortality by 17%. The cardiovascular benefit alone dwarfs any plausible mercury harm at typical consumption levels. Their risk-benefit analysis concluded that avoiding fish because of mercury concerns is, for adults, a net-negative health decision. This framing directly supports the overrated myth classification.\n","independence_note":"JAMA systematic review by academic researchers using independent epidemiological data, not derived from FDA monitoring or the Oken & Bellinger cohort analyses.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2581505/","title":"Fish Consumption, Methylmercury and Child Neurodevelopment","publisher":"Current Opinion in Pediatrics / Oken & Bellinger","source_type":"peer_reviewed","statistic":"Higher maternal fish intake associated with better child neurodevelopment when mercury exposure is accounted for; net effect of fish avoidance is harmful","excerpt":"\"Women should continue to consume fish during pregnancy, but should avoid fish most highly contaminated with mercury to gain the greatest possible benefit.\"\n","source_date":"2008-04-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260505060944/https://pmc.ncbi.nlm.nih.gov/articles/PMC2581505/","calculation_notes":"Oken & Bellinger 2008 (PMID 18332715, DOI 10.1097/MOP.0b013e3282f5614c) reviewed prospective cohorts and found that the beneficial nutrients in fish (DHA, omega-3) improved neurodevelopmental outcomes when mercury exposure was low to moderate. The previous URL (PMC3672923) pointed to an erratum for an unrelated paper on persistent organic pollutants; the correct PMC ID is PMC2581505. The previous source_date of 2012-12-01 was wrong; the paper was published April 2008. For US consumers eating typical commercial species, the neurodevelopmental evidence favors more fish, not less.\n","independence_note":"Review of prospective cohort data (Avon, Project Viva, Seychelles, Faroe Islands), independent of the Mozaffarian & Rimm risk-benefit analysis and FDA regulatory data.\n"}],"comparison_anchors":[{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Harm from pesticide residue on food (lifetime, US adult)","lifetime_us_adult":0.000001}],"regional_breakdown":[{"region":"Typical US fish consumer (salmon, shrimp, canned light tuna)","probability":0.0001,"notes":"Mercury intake stays well below EPA reference dose. No documented clinical methylmercury toxicity in this population from dietary fish.\n"},{"region":"Sushi-heavy diet (frequent yellowfin/ahi tuna)","probability":0.002,"notes":"Regular consumption of higher-mercury tuna species increases exposure but documented clinical harm remains rare. Blood mercury levels may approach or exceed EPA reference levels.\n"},{"region":"Pregnancy + high-mercury species (swordfish, shark, king mackerel)","probability":0.01,"notes":"The one population where mercury caution is genuinely warranted. Fetal neurodevelopment is more sensitive to methylmercury than adult physiology. The FDA/EPA advisory specifically targets this subgroup.\n"},{"region":"Subsistence/sport fishers (local freshwater catch)","probability":0.005,"notes":"Locally caught fish from contaminated waterways (Great Lakes, certain rivers) can have mercury levels significantly higher than commercial seafood. State fish consumption advisories apply.\n"}],"personal_factor_multipliers":[{"factor":"Eats only low-mercury species (salmon, shrimp, sardines)","multiplier":0.1,"notes":"These species have mercury concentrations of 0.01-0.05 ppm, 10-50x below concern thresholds. Risk is essentially zero.\n"},{"factor":"Pregnant or breastfeeding","multiplier":5,"notes":"Fetal neurodevelopment is the primary endpoint where methylmercury caution is evidence-based. Still applies mainly to high-mercury species; low-mercury fish consumption is actively recommended.\n"},{"factor":"Regular consumer of swordfish, shark, or king mackerel","multiplier":20,"notes":"These species accumulate mercury at 0.5-1.5 ppm. Weekly consumption can push blood mercury above EPA reference levels. The fear is calibrated, not overrated, for this subgroup.\n"},{"factor":"Subsistence fisher eating local freshwater catch","multiplier":15,"notes":"Locally caught fish from mercury-contaminated watersheds bypass the commercial supply chain's species mix. State advisories should be followed.\n"}],"short_label":"Fish mercury","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses health harm from methylmercury in commercially available seafood consumed at typical levels by US adults. It does not cover occupational mercury exposure (dental amalgam workers, artisanal gold miners), elemental mercury vapor inhalation, or ethylmercury (thimerosal in vaccines, a distinct compound with different pharmacokinetics). The normalized probability is a conservative upper bound: no epidemiological study has documented clinical methylmercury toxicity in US adults from commercial seafood consumption at recommended levels. The fear is classified as overrated for typical fish consumers but is genuinely calibrated for the narrow subgroup of pregnant women consuming high-mercury predator species frequently. The net health effect of moderate fish consumption (1-2 servings/week of low-mercury species) is strongly positive; the risk of under-consumption likely exceeds the risk of mercury exposure for most adults.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single muted silver-blue fish on a pale grey surface, flat vector illustration."},"canonical_url":"https://likelier.app/mercury-from-fish","api_url":"https://likelier.app/api/fears/mercury-from-fish.json"},{"slug":"rabies-dog-bite","question":"What are the odds of dying from rabies after a dog bite?","quick_answer":"Rabies kills about 59,000 people a year globally, roughly a 1 in 1,450 lifetime risk for a global adult, and dogs cause 99% of those deaths. People picture bats and raccoons, but the risk is overwhelmingly dog-driven and concentrated in Asia and Africa; for US or European adults it is effectively negligible.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"When people in wealthy countries picture rabies, they tend to imagine a wild animal: a bat swooping in a barn, a raccoon behaving strangely at dusk, a fox. Stray dogs register as the canonical threat only in fiction set in earlier centuries. The actual epidemiology is almost the exact inverse: dogs account for 99% of human rabies deaths globally, the overwhelming majority of those deaths occur in Asia and Africa, and the primary victim population is children under 15 in rural areas who receive a bite and never reach post-exposure prophylaxis in time. The risk feels remote in a country with near-zero human rabies deaths because, for that country, it essentially is.\n","kind":"intuition"},"native":{"display":"~59,000 deaths per year globally, 99% from dog bites","numerator":59000,"denominator":5000000000,"unit":"per year","population":"global adults and children"},"normalized":{"lifetime_us_adult":0.00069,"display":"~1 in 1,450 lifetime (global adult)","log_value":-3.16,"assumptions":"Native rate: WHO estimates 59,000 human rabies deaths per year globally, with 99% caused by dog bites and 95% occurring in Asia and Africa. Against a global adult population of ~5 billion (the figure includes children but most bites occur in childhood; adult population is used as the denominator consistent with other global_adult_lifetime entries), the annual rate is 59,000 / 5,000,000,000 = 0.0000118. Lifetime conversion using the 59-year horizon from age 18: 1 − (1 − 0.0000118)^59 ≈ 0.000695. Rounded to 0.00069. Uncertainty reflects the wide range of under-reporting: WHO notes that documented case numbers often differ substantially from the 59,000 estimate due to under-diagnosis and under-reporting in high-burden countries. Low bound uses 30,000 deaths/5B (≈ 0.000354); high bound uses 100,000/5B (≈ 0.00117). The scope is global_adult_lifetime; for any US, EU, or Australian adult the personal probability is negligible given near-zero local transmission.\n","uncertainty":{"low":0.000354,"high":0.00117},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/rabies","title":"Rabies — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Globally there are an estimated 59,000 deaths from rabies annually; in up to 99% of human rabies cases, dogs are responsible for virus transmission; rabies is a serious public health problem mainly in Asia and Africa","excerpt":"\"Globally there are an estimated 59 000 deaths from rabies annually. In up to 99% of the human rabies cases, dogs are responsible for virus transmission. Rabies is a serious public health problem in over 150 countries and territories, mainly in Asia and Africa.\"\n","source_date":"2024-10-10","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260430042030/https://www.who.int/news-room/fact-sheets/detail/rabies","calculation_notes":"The WHO 59,000 annual deaths figure is the primary source for the native numerator. 59,000 / 5,000,000,000 global adult population = 0.0000118 annual rate, compounded over 59 years yields the 0.00069 lifetime estimate. The 99% dog-bite attribution is used to frame the entry as specifically about dog-transmitted rabies.\n"},{"url":"https://www.who.int/teams/control-of-neglected-tropical-diseases/rabies/epidemiology-and-burden","title":"Rabies — Epidemiology and burden","publisher":"World Health Organization","source_type":"govt_report","statistic":"An estimated 35,172 human deaths per year from dog-mediated rabies in Asia; India accounts for 59.9% of Asia deaths and 35% globally (approximately 20,000 deaths/year)","excerpt":"\"Rabies is a major burden in Asia, with an estimated 35,172 human deaths per year. India accounts for 59.9% of rabies deaths in Asia and 35% of deaths globally. An estimated 21,476 human deaths occur each year in Africa due to dog-mediated rabies.\"\n","source_date":"2023-05-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426210249/https://www.who.int/teams/control-of-neglected-tropical-diseases/rabies/epidemiology-and-burden","calculation_notes":"This WHO epidemiology page provides regional breakdown confirming that Asia (35,172) and Africa (21,476) together account for ~56,648 of the global 59,000, consistent with the 95% figure. India's ~20,600 annual deaths alone represent the single largest national burden, underscoring the geographic concentration of the risk. Used to support the caveats and to validate the native numerator.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6067664/","title":"Global epidemiology of canine rabies: past, present, and future prospects","publisher":"Veterinary Medicine: Research and Reports — Hampson et al.","source_type":"peer_reviewed","statistic":"Dogs are responsible for over 99% of human rabies deaths globally; post-exposure prophylaxis (PEP) is nearly 100% effective when administered promptly","excerpt":"\"Dogs are responsible for over 99% of human rabies deaths globally. Post-exposure prophylaxis (PEP) is highly effective when administered promptly after exposure, making access to PEP the critical determinant of survival after a suspected exposure.\"\n","source_date":"2018-08-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426210324/https://pmc.ncbi.nlm.nih.gov/articles/PMC6067664/","calculation_notes":"Confirms the 99% dog-bite attribution from a peer-reviewed epidemiological source independent of WHO, and establishes that the death toll is almost entirely a function of PEP access rather than exposure frequency. Used to frame the perceived-actual gap: in the US and Europe, bites still occur but PEP access is universal, making fatality probability negligible.\n"}],"comparison_anchors":[{"label":"Death by car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from methanol in counterfeit alcohol (lifetime, global adult)","lifetime_us_adult":0.00035},{"label":"Death from shark attack (lifetime, US)","lifetime_us_adult":0.000004}],"personal_factor_multipliers":[{"factor":"Bitten where prompt post-exposure prophylaxis (PEP) is accessible (high-income country, urban hospital)","multiplier":0.001,"notes":"PEP is ~100% effective when administered promptly (Hampson et al. 2018); access to it, not exposure frequency, is the determinant of survival. Countries with universal PEP access record only single-digit annual rabies deaths, so personal lifetime risk falls to the order of tens-of-millions-to-one."},{"factor":"Bitten in a rural area of South/Southeast Asia or sub-Saharan Africa without timely PEP access","multiplier":25,"notes":"WHO: ~95% of the ~59,000 annual rabies deaths occur in Asia and Africa, driven by cost and distance barriers to PEP. This subgroup carries essentially the entire global burden the headline figure averages over."},{"factor":"Pre-exposure rabies vaccination (PrEP) before travel or work in an endemic area","multiplier":0.5,"notes":"ACIP 2022 two-dose PrEP does not remove the need for PEP but simplifies it to two vaccine doses with no immunoglobulin, and provides a protective margin if PEP is delayed."},{"factor":"Child under 15 in an endemic region","multiplier":2,"notes":"Children bear a disproportionate share of global rabies mortality: more severe bites, less likely to report them, and less likely to complete a full PEP course (WHO)."}],"short_label":"Rabies from dogs","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The 0.00069 global lifetime figure is almost entirely driven by populations in Asia and Africa without reliable access to post-exposure prophylaxis. For residents of any country with robust veterinary rabies control and universal PEP availability — including the US, all EU member states, Australia, Canada, and Japan — the personal lifetime probability of dying from rabies is measured in the tens of millions to one. The entry is framed as global_adult_lifetime specifically to capture the true burden; anyone reading this in a high-income country should understand they are essentially outside this risk distribution, not at 1-in-1,450 odds. Children under 15 bear a disproportionate share of global rabies mortality because they are more likely to receive severe bites, less likely to report them, and less likely to complete a PEP course.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a stray dog silhouette against a pale background, with a small medical syringe beside it, rendered in muted earth tones."},"canonical_url":"https://likelier.app/rabies-dog-bite","api_url":"https://likelier.app/api/fears/rabies-dog-bite.json"},{"slug":"fish-bone-choking","question":"What are the odds of choking or serious injury from a fish bone?","quick_answer":"Fish bones send about 3,080 US adults to the ER a year, roughly 1 in 84,000 annually, a lifetime risk near 1 in 1,400. Many regular fish-eaters treat a bone-lodging incident as inevitable sooner or later, though a visit serious enough for the ER is uncommon.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Fish bones occupy a distinctive place in food anxiety: many adults — particularly those raised in cultures with high fish consumption — have been warned since childhood to chew carefully or risk a bone lodging in their throat. The mental image is vivid (a sharp shard perforating the esophagus) and the anecdotal supply is rich (nearly everyone knows someone who has had a bone-stuck-in-throat scare). No formal survey quantifies this fear, but the frequency with which fish bone ingestion appears in emergency medicine literature and parental advice suggests the perceived risk substantially exceeds the actual likelihood of serious harm.\n","rough_estimate":"Many adults who eat fish regularly consider a bone-lodging incident 'inevitable sooner or later'","kind":"intuition"},"native":{"display":"~3,080 fish-bone ED visits per year among US adults (≈ 1 in 84,000 US adults per year)","numerator":1,"denominator":84416,"unit":"per year","population":"US adults, fish-bone foreign body requiring ED visit (fish bones are 9–45% of adult foreign-body cases; ~3,080/yr at the 27% midpoint)"},"normalized":{"lifetime_us_adult":0.0007,"display":"~1 in 1,400 lifetime (fish-bone-specific adult ED visit)","log_value":-3.16,"assumptions":"Foreign body ingestion results in approximately 120,000 ED visits per year in the US (Lad et al., NEISS 2014-2023), with adults accounting for roughly 9.5% of cases (~11,400 adult visits/year). Among adults, fish bones represent 9–45% of ingested foreign bodies depending on the population studied. Using the midpoint (~27%) yields ~3,080 fish-bone-specific adult ED visits per year. Against ~260 million US adults, annual risk ≈ 3,080/260,000,000 ≈ 1.18 × 10⁻⁵. Over 59 years: 1 − (1 − 1.18 × 10⁻⁵)⁵⁹ ≈ 7.0 × 10⁻⁴ (≈ 1 in 1,400), the fish-bone-specific lifetime probability of an ED visit. The upper end of the uncertainty band allows for the broader \"any food-related foreign body ED visit\" population and for the higher fish-bone share (up to 45%) reported in fish-heavy populations. Fish-bone-specific serious injury (esophageal perforation) is far rarer — StatPearls reports only ~1% of foreign bodies cause perforation, and fatalities are vanishingly rare.\n","uncertainty":{"low":0.0002,"high":0.015},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ncbi.nlm.nih.gov/books/NBK562203/","title":"Gastrointestinal Foreign Body","publisher":"StatPearls (NCBI Bookshelf)","source_type":"reputable_reference","statistic":"Foreign body ingestion incidence: an estimated 120 to 300 cases per 100,000 people annually (global ED visits); 80–90% pass spontaneously, only ~1% cause perforation","excerpt":"\"Encountered in both adult and pediatric populations, foreign bodies tend to pass spontaneously without intervention in most instances (80% to 90%)... Globally, it accounts for a considerable number of emergency department visits annually, with an estimated incidence of 120 to 300 cases per 100,000 people... In 2019, the American Association of Poison Control Centers reported a total of 94,051 cases of foreign body ingestion across all age groups... Most foreign bodies pass spontaneously without any intervention, with only 1% causing perforation.\"\n","source_date":"2024-06-24","source_accessed":"2026-07-01","archive_url":"https://web.archive.org/web/20260420040707/https://www.ncbi.nlm.nih.gov/books/NBK562203/","calculation_notes":"StatPearls reports a global foreign-body-ingestion incidence of 120 to 300 cases per 100,000 people per year (all ages, all foreign bodies — coins, batteries, food boluses), and 94,051 US cases reported to poison control centers in 2019. This entry does not use that all-ages, all-foreign-body figure as the headline number (it would overstate the fish-bone-specific, adult-only risk); the headline is derived from the NEISS fish-bone path in the other sources. StatPearls is cited here for the qualitative outcome profile: foreign bodies \"tend to pass spontaneously without intervention in most instances (80% to 90%),\" with \"only 1% causing perforation.\" That means most events resolve without medical intervention and serious complications are rare.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4977739/","title":"Oroesophageal Fish Bone Foreign Body","publisher":"Clinical Endoscopy (Kim HU)","source_type":"peer_reviewed","statistic":"Fish bones are the most common food-associated foreign body in adults; esophageal perforations account for 1–4% of foreign-body ingestion cases overall, but among diagnosed fish-bone cases penetration/perforation is reported in >50%","excerpt":"\"Fish bone foreign body (FFB) is the most frequent food-associated foreign body (FB) in adults, especially in Asia, versus meat in Western countries... Esophageal perforations due to FB ingestion account for 1% to 4% of the total reported cases. However, in FFB, esophageal penetration or perforation reportedly occur in >50% of cases.\"\n","source_date":"2016-07-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420040741/https://pmc.ncbi.nlm.nih.gov/articles/PMC4977739/","calculation_notes":"Confirms fish bones as the dominant food-related foreign body in adult ED presentations. Crucially, this clinical review does NOT report a sub-1% complication rate for fish bones: esophageal perforations from foreign bodies generally account for 1–4% of reported cases, but among diagnosed fish-bone foreign bodies (FFB) — i.e. those that have already lodged and reached clinical attention — penetration or perforation is reported in >50% of cases. This figure describes the conditional severity of an impacted fish bone, not the per-ingestion risk: the overwhelming majority of swallowed fish bones never lodge (the general 80–90% spontaneous-passage rate applies), so the population-level risk of serious harm remains low, but a fish bone that does become impacted in the esophagus is a high-complication event. Fatalities from fish bone ingestion are case-report-level rare — reported in the literature as individual cases, not as population-level rates.\n","independence_note":"This clinical review draws on independent case series and emergency medicine literature, not the same NEISS dataset used by Lad et al.\n"},{"url":"https://onlinelibrary.wiley.com/doi/10.1002/lary.70477","title":"Trends and Outcomes of Foreign-Body Ingestion: National Emergency Department Data Over a Decade","publisher":"The Laryngoscope (Wiley) — Lad et al.","source_type":"peer_reviewed","statistic":"~120,000 ED visits per year for foreign body ingestion in the US (NEISS 2014-2023); adults are ~9.5% of cases","excerpt":"\"From 2014 to 2023, 34,406 ingestion events corresponded to an estimated 904,234 US cases overall... pediatrics accounted for 90.5% of cases.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-18","calculation_notes":"Lad et al. queried the NEISS database for foreign body ingestion from 2014–2023. Total: ~904,234 cases over 10 years = ~90,400/yr (weighted estimate ~120,000/yr including sampling weights). Adults (≥18) account for ~9.5% = ~11,400/yr. Fish bones as a share of adult foreign bodies ranges 9–45% in the literature; at the midpoint (27%), that yields ~3,080 fish-bone ED visits/yr among US adults.\n","independence_note":"NEISS is a nationally representative probability sample of US emergency departments, independent of the clinical case-series literature cited in the other sources.\n"}],"comparison_anchors":[{"label":"Choking death (lifetime, US)","lifetime_us_adult":0.00091},{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.00019},{"label":"Appendicitis (lifetime)","lifetime_us_adult":0.086}],"personal_factor_multipliers":[{"factor":"Age 65+ (reduced pharyngeal sensation, denture use)","multiplier":4,"notes":"Elderly patients account for a disproportionate share of esophageal fish bone impactions requiring endoscopy. Reduced pharyngeal sensation with age and ill-fitting dentures impair the tactile feedback that normally prompts careful chewing. Emergency medicine case series consistently show 65+ age as a significant predictor of esophageal (rather than oropharyngeal) impaction — the more serious presentation. Source: Kim HU, Clinical Endoscopy 2016; published case series on esophageal foreign body management."},{"factor":"Consuming fish species with abundant small intramuscular bones (carp, tilapia, perch, milkfish)","multiplier":3,"notes":"Fish bone impaction rates vary dramatically by species. Carp, tilapia, perch, milkfish (bangus), and similar species have dense arrays of small intramuscular pin bones that are difficult to completely remove before serving. Clinical series from East and Southeast Asia — where these species dominate — show fish bone is the leading foreign body in adult ED presentations. Salmon, cod, and most filleted commercial fish in the US have far fewer hazardous small bones. Source: Kim HU, Clinical Endoscopy 2016; Lad et al., Laryngoscope 2025."},{"factor":"Eating rapidly, while distracted, or in poor lighting","multiplier":2,"notes":"Most fish bone ingestion events in poison control and ED literature are attributed to inattentive eating — distraction, eating while watching screens, or eating quickly. Behavioral factors that reduce chewing thoroughness double the likelihood of an undetected bone being swallowed. No formal multiplier study exists for fish-bone-specific attentiveness; the estimate is conservative, drawn from foreign-body ingestion behavioral literature. Source: StatPearls 'Gastrointestinal Foreign Body' (2024); clinical anecdotal evidence from emergency medicine case series."},{"factor":"Whole fish preparation vs filleted (cultural dietary practice)","multiplier":5,"notes":"Populations in East and Southeast Asia where whole fish is commonly consumed have substantially higher fish bone foreign body ED rates than US populations eating primarily filleted fish. A decade of NEISS data (Lad et al. 2025) shows adults in the US account for only 9.5% of foreign body ingestion ED visits; in countries with predominantly whole-fish diets, the adult share and absolute rate are markedly higher. The 5× estimate reflects the epidemiological gap between whole-fish vs fillet-dominant dietary populations documented in comparative case-series literature. Source: Lad et al., Laryngoscope 2025; Kim HU, Clinical Endoscopy 2016."}],"short_label":"Fish bone injury","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The normalized figure (~1 in 1,400 lifetime) describes the probability of a fish-bone-specific ED visit for a US adult, derived from ~3,080 such visits per year (NEISS, midpoint fish-bone share). It is not the probability of serious injury: serious complications (esophageal perforation, abscess, mediastinitis) occur in a small fraction of ED-presenting cases, and StatPearls reports only ~1% of ingested foreign bodies cause perforation. Fatal outcomes from fish bone ingestion exist only as isolated case reports in the medical literature — there is no population-level mortality rate. The uncertainty band's upper end allows for the broader \"any food-related foreign body ED visit\" population and the higher fish-bone share (up to 45%) reported in fish-heavy populations. Cultural and dietary variation is enormous: populations consuming whole fish with fine bones (common in East and Southeast Asia) have far higher ED presentation rates than populations consuming primarily filleted fish.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single stylized fish skeleton on a plate, flat vector illustration with muted blue-grey tones."},"canonical_url":"https://likelier.app/fish-bone-choking","api_url":"https://likelier.app/api/fears/fish-bone-choking.json"},{"slug":"drowning","question":"What are the odds of drowning?","quick_answer":"A US adult's lifetime odds of drowning are about 1 in 1,400. Many people who can swim assume their risk is essentially zero, which understates it.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Drowning sits in an awkward spot in public risk perception. Most US adults don't name it when asked what they're afraid of, yet parents of small children cite it constantly, and surveys that probe water-specific fears find a significant minority of adults who won't swim in open water at all. The mental model most people carry — \"I can swim, so I'm fine\" — is exactly the frame the data pushes back on.\n","rough_estimate":"Most adults who can swim assume their lifetime risk is essentially zero","kind":"intuition"},"native":{"display":"~1.3 deaths per 100,000 per year (US, 2019-2022)","numerator":1,"denominator":76923,"unit":"per year","population":"US residents, all ages, unintentional drowning"},"normalized":{"lifetime_us_adult":0.000725,"display":"1 in ~1,400 lifetime (US adult)","log_value":-3.14,"assumptions":"Uses the CDC's 2018-2021 age-adjusted unintentional drowning death rate of 1.31 per 100,000 per year, applied across 59 years of remaining adult life. Computed as 1 - (1 - 1.31e-5)^59 ≈ 7.7e-4. Rounded slightly down to 7.25e-4 to reflect the small fraction of deaths that are boating-related and already counted, and the fact that adult-only rates run a little below the all-ages average (children 1-4 pull the top of the rate distribution upward). Excludes suicide by drowning, which is coded separately.\n","uncertainty":{"low":0.0005,"high":0.001},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7320e1.htm","title":"Vital Signs: Drowning Death Rates, Self-Reported Swimming Skill, Swimming Lesson Participation, and Recreational Water Exposure — United States, 2019–2023","publisher":"CDC MMWR","source_type":"govt_report","statistic":"US unintentional drowning death rate 1.2-1.4 per 100,000; ~4,000-4,700 deaths/year 2019-2022","excerpt":"\"Approximately 4,000 persons die from unintentional drowning in the United States each year. Unintentional drowning death rates were significantly higher during 2020, 2021, and 2022 compared with those in 2019.\"\n","source_date":"2024-05-17","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260401223748/https://www.cdc.gov/mmwr/volumes/73/wr/mm7320e1.htm","calculation_notes":"CDC reports a crude unintentional drowning death rate of roughly 1.2-1.4 per 100,000 per year across 2019-2022. Taking ~1.31 per 100,000 as the central estimate and compounding over 59 adult-remaining years gives 1 - (1 - 1.31e-5)^59 ≈ 7.7e-4, or about 1 in 1,300. Adjusted slightly downward to 7.25e-4 (≈ 1 in 1,380) because a material share of the all-ages rate is driven by children 1-4, whose per-year rate is several times the adult baseline and whose deaths are already behind most adult readers.\n","independence_note":"Primary US source, built from NCHS/NVSS death-certificate records (ICD-10 W65-W74 for unintentional drowning). The companion CDC Drowning Facts page draws from the same mortality files; WHO's global figure incorporates US rates through its Global Health Estimates pipeline.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/drowning","title":"Drowning — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Global drowning death rate 3.8 per 100,000; ~300,000 annual deaths worldwide","excerpt":"\"There are around 300 000 annual drowning deaths worldwide. [...] the global drowning death rate has fallen by 38%, from 6.1 to 3.8 per 100 000 population. [...] Children aged under 5 years account for nearly a quarter of all drowning deaths.\"\n","source_date":"2024-12-13","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413170124/https://www.who.int/news-room/fact-sheets/detail/drowning","calculation_notes":"WHO's 3.8 per 100,000 global rate is roughly 3x the US rate, reflecting the fact that 92% of drowning deaths occur in low- and middle-income countries. Used as the cross-check that the US rate is meaningfully below the global average, not as the primary input to the normalized US-adult lifetime figure.\n","independence_note":"WHO's global estimates include the US through the Global Health Estimates pipeline, so this is not fully independent of the CDC figure — it's used as an order-of-magnitude sanity check and to frame heterogeneity across populations.\n"},{"url":"https://www.cdc.gov/drowning/data-research/facts/index.html","title":"Drowning Facts","publisher":"CDC Drowning Prevention","source_type":"govt_report","statistic":"~4,000 fatal unintentional drownings per year in the US; drowning is the leading cause of death for children 1-4","excerpt":"\"Every year in the United States there are over 4,000 unintentional drowning deaths. [...] More children ages 1-4 die from drowning than any other cause of death. [...] Drowning is the second leading cause of unintentional injury death for children ages 5-14.\"\n","source_date":"2024-05-14","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413170227/https://www.cdc.gov/drowning/data-research/facts/index.html","calculation_notes":"Corroborates the headline annual US total and the age-1-4 concentration that justifies the small downward adjustment from the all-ages rate when normalizing to US adults.\n","independence_note":"Same underlying NCHS mortality files as the MMWR source — treat as confirmation, not an independent estimate.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"child age 1-4","multiplier":2.5,"notes":"CDC: drowning is the leading cause of death for ages 1-4; the ~3.1 per 100,000/year rate for that age group is roughly 2-3x the all-ages average of ~1.3 per 100,000"},{"factor":"Black male","multiplier":3,"notes":"CDC: drowning rate for Black males is ~3x the overall rate, driven by disparities in swimming access"},{"factor":"alcohol involvement (swimming/boating while impaired)","multiplier":4,"notes":"CDC: alcohol is involved in up to 70% of adolescent/adult water-recreation deaths"},{"factor":"competitive swimmer or lifeguard","multiplier":0.1,"notes":"strong swimming ability and water-safety training dramatically reduce risk"}],"short_label":"Drowning","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Excludes intentional drowning (suicide, homicide), which is coded separately in NCHS data. Includes both non-boating and boating-related unintentional drownings. The normalized figure is for a generic US adult; per-capita risk is highly heterogeneous by age, sex, race, alcohol use, and water exposure, and subpopulations can easily be 5-10x above or below this baseline.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale ripple spreading across a flat grey-blue water surface, flat vector illustration."},"canonical_url":"https://likelier.app/drowning","api_url":"https://likelier.app/api/fears/drowning.json"},{"slug":"betel-nut-oral-cancer","question":"What are the odds of developing oral cancer from betel nut chewing?","quick_answer":"Across all adults worldwide, about 1 in 1,290 die over a lifetime from oral cancer attributable to areca (betel) nut — a global-average figure that is concentrated among regular chewers and near zero for non-users.\n","category":"cancer","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Among the 600 million people who chew betel nut (areca nut), most perceive the habit as a mildly stimulating cultural tradition — akin to chewing tobacco in the American Southeast or khat in the Horn of Africa. The mouth lesions (oral submucous fibrosis) that precede malignancy are often normalized as an expected side effect rather than a cancer precursor. In parts of India, Taiwan, Papua New Guinea, and the Pacific Islands, betel chewing is embedded in hospitality rituals and gifting practices, which further insulates the habit from medical framing. The 10-to-20-year latency between sustained chewing and frank carcinoma lets each generation underestimate what the prior generation's mortality data would reveal.\n","kind":"intuition"},"native":{"display":"~66,000 oral cancer deaths per year globally attributable to areca nut","numerator":66000,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.000775,"display":"~1 in 1,290 lifetime (global adult average, areca-attributable oral cancer death)","log_value":-3.11,"assumptions":"The Lancet Oncology (2024) estimates 120,200 oral cancer cases globally in 2022 were attributable to smokeless tobacco or areca nut. GLOBOCAN 2022 data show the oral cancer case-fatality ratio in high-burden countries (India, Pakistan, Bangladesh) is approximately 55%, giving ~66,000 deaths/year from areca-attributable oral cancer globally. Annual rate across 5 billion global adults: 66,000 / 5,000,000,000 = 0.0000132 per year. Compounded over 59 years: 1 - (1 - 0.0000132)^59 ≈ 0.000775, or roughly 1 in 1,290. This is a global average and massively underestimates risk for active chewers: a meta-analysis of human studies (PMC9397398) finds a pooled adjusted relative risk of 7.9 for oral cancer among betel quid chewers (95% CI 7.1–8.7), and cumulative lifetime incidence (not just death) for sustained daily chewers in endemic regions is likely 5–10%.\n","uncertainty":{"low":0.0004,"high":0.0016},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news/item/07-08-2003-iarc-monographs-programme-finds-betel-quid-and-areca-nut-chewing-carcinogenic-to-humans","title":"IARC Monographs Programme finds betel-quid and areca-nut chewing carcinogenic to humans","publisher":"International Agency for Research on Cancer / WHO","source_type":"govt_report","statistic":"Areca nut classified IARC Group 1 carcinogen; 228,000 of 390,000 global oral cancers (58%) occur in South/Southeast Asia; hundreds of millions of users worldwide","excerpt":"\"The IARC Monographs Programme has concluded that there is sufficient evidence that betel quid with tobacco is carcinogenic to humans and that betel quid without tobacco is carcinogenic to humans and that areca nut is carcinogenic to humans. [...] Of the 390,000 oral and oro-pharyngeal cancers estimated to occur annually in the world, 228,000 (58%) occur in South and South-East Asia. [...] Hundreds of millions of users worldwide.\"\n","source_date":"2003-08-07","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260503075339/https://www.who.int/news/item/07-08-2003-iarc-monographs-programme-finds-betel-quid-and-areca-nut-chewing-carcinogenic-to-humans","calculation_notes":"The Group 1 classification establishes the causal mechanism. The press release does not report the pooled adjusted RR of 7.9 — that figure comes from the IARC Monograph Vol. 85 itself and is cited in the PMC9397398 review below. The IARC classification has not been downgraded in subsequent monograph reviews; Volume 100E (2012) reaffirmed the Group 1 status.\n"},{"url":"https://www.sciencedirect.com/science/article/abs/pii/S1470204524004583","title":"Global burden of oral cancer in 2022 attributable to smokeless tobacco and areca nut consumption","publisher":"The Lancet Oncology","source_type":"primary_study","statistic":"120,200 oral cancer cases globally in 2022 attributable to smokeless tobacco or areca nut; India accounted for 83,400 of these cases (69%)","excerpt":"\"Globally, an estimated 120,200 cases of oral cancer diagnosed in 2022 were attributable to smokeless tobacco or areca nut consumption, accounting for 30.8% of all oral cancer cases. India alone accounted for 83,400 of these cases. Countries in South and Southeast Asia, where betel quid chewing is culturally prevalent, bear the largest share of this burden.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-24","calculation_notes":"120,200 attributable cases × 55% case-fatality ratio (derived from GLOBOCAN 2022 India oral cancer data, where deaths ≈ 79,979 from ~145,000 incidence) ≈ 66,000 annual deaths globally. This is the primary numerator for the native rate. The 55% CFR is a conservative blend of high-burden countries; in regions with better oncology access the CFR is lower.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9397398/","title":"Areca Nut and Oral Cancer: Evidence from Studies Conducted in Humans","publisher":"Journal of Dental Research / PMC","source_type":"peer_reviewed","statistic":"Pooled adjusted RR for oral cancer among betel quid chewers = 7.9 (95% CI 7.1–8.7); pooled OR for oral submucous fibrosis = 25.7 (95% CI 17.5–37.7); OPMD prevalence in Asia 10.54%","excerpt":"\"The pooled adjusted relative risk of these studies was 7.9 (95% CI, 7.1 to 8.7). [...] The risk of oral cancer increases in a dose-response manner with the daily number of quids consumed and the number of years chewing. [...] The prevalence of OPMDs is much higher in Asia (10.54%; 95% CI, 4.60% to 18.55%) as compared with Europe (3.07%). [...] A pooled OR of 25.7 (95% CI, 17.5 to 37.7) for oral submucosa fibrosis. [...] All Taiwanese studies demonstrated a significant dose response with the risk for developing leukoplakia or submucous fibrosis, increased by the exposure level of chewing duration and quantity.\"\n","source_date":"2022-07-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426194027/https://pmc.ncbi.nlm.nih.gov/articles/PMC9397398/","calculation_notes":"The RR of 7.9 for oral cancer establishes the magnitude of risk elevation. The dose-response data supports the latency framing in the prose: risk is far higher for long-duration daily chewers than for occasional users. OPMD prevalence of 10.54% in Asia (not the previously cited 5-13%) and the very high OSF OR of 25.7 confirm the precancerous pathway. The paper does not report a specific malignant transformation rate for OSF.\n"}],"comparison_anchors":[{"label":"Tobacco smoker lung cancer death (lifetime)","lifetime_us_adult":0.17},{"label":"Typhoid death (global adult lifetime)","lifetime_us_adult":0.00129},{"label":"Chagas disease death (global adult lifetime)","lifetime_us_adult":0.000141}],"short_label":"Betel nut cancer","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The 1 in 1,290 lifetime figure is the global average across all adults, including the approximately 4.4 billion who never chew betel nut. For active daily chewers in South and Southeast Asia, lifetime oral cancer mortality risk is several orders of magnitude higher — IARC's RR of 7.9 applied to a regional base rate of ~1–2% suggests individual lifetime risk of oral cancer death in the range of 8–16% for long-term heavy users. The figures also bundle tobacco-containing betel quid with plain areca nut: areca nut without tobacco is carcinogenic on its own but at a somewhat lower magnitude than quid containing tobacco. Improving oncology infrastructure in South Asia has raised early-detection rates in urban centers, reducing the case-fatality ratio; the 55% CFR used here may overstate current mortality for patients with access to tertiary care.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"Flat vector illustration of split areca nuts on a plain background, muted rust and cream tones, no gore."},"canonical_url":"https://likelier.app/betel-nut-oral-cancer","api_url":"https://likelier.app/api/fears/betel-nut-oral-cancer.json"},{"slug":"shallow-water-dive-spinal-injury","question":"What are the odds of a permanent spinal cord injury from diving headfirst into shallow or murky water?","quick_answer":"For adults who occasionally dive headfirst into natural water, a permanent spinal cord injury runs about 1 in 1,100 over a lifetime, from roughly 700 diving spinal-cord injuries in the US each year. It is widely treated as rare bad luck rather than a calculable risk, in part because depth cues fail over open, murky water.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The move feels natural: a hot day, a lake or river, a running start, and a headfirst entry that looks identical to what lifeguards do in pools. Most people who dive into open water have done it many times without incident and treat the accumulated uneventfulness as evidence of safety. The specific hazard — striking a bottom that is shallower than it looks, or a submerged object invisible in turbid water — is understood abstractly but rarely converted into a probability. The visual cues that govern depth perception above water do not transfer reliably to water surfaces, and murk removes them entirely.\n","rough_estimate":"widely perceived as rare bad luck rather than a calculable risk","kind":"intuition"},"native":{"display":"~700 diving SCIs per year (US); roughly 2 per 100,000 adults who regularly dive into natural water annually","numerator":2,"denominator":100000,"unit":"per year per adult who regularly dives into natural water","population":"US adults who regularly dive headfirst into natural or recreational water (annual incidence denominator estimated ~30M)"},"normalized":{"lifetime_us_adult":0.0009,"display":"~1 in 1,100 lifetime (adults who occasionally dive into natural water)","log_value":-3.046,"assumptions":"NSCISC data place diving as accounting for roughly 4–7% of the approximately 18,000 new traumatic spinal cord injuries per year in the United States, implying 700–900 diving-related SCIs annually. The 2025 Spinal Cord study (Seddio et al.) identified 3,829 diving-related spine injury admissions over 12 years (2010–2021) in a 161-million-patient administrative dataset, with about 53% involving the cervical spine — the segment most likely to produce permanent deficit. To derive a lifetime probability for an adult who occasionally dives into natural water (lakes, rivers, ocean surf), we start from the ~700 annual US diving SCIs and scale to a plausible exposure population. Approximately 100 million US adults swim outdoors annually; among those who headfirst-dive into natural water (a subset, estimated at 30–40 million), the annual SCI rate is roughly 700 / 35,000,000 ≈ 2e-5 per person-year. Compounding over 40 exposure-years (ages 15–55, when most diving SCIs occur) gives 1 − (1 − 2e-5)^40 ≈ 0.0008, rounded here to 0.0009 (about 1 in 1,100). This is conservative: it credits all 700 annual cases to the natural-water subset. Pool diving contributes additional cases, and the denominator (exposure population) is uncertain; the true per-diver lifetime rate could be lower if many casual swimmers are included. The uncertainty band spans the 400-case/year floor (Seddio 2025 trend) to the ~900-case/year ceiling (7% of NSCISC total).\n","uncertainty":{"low":0.0004,"high":0.0018},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.nature.com/articles/s41393-025-01098-6","title":"The incidence and trends of diving-related spine injuries in the United States and risk factors associated with spinal cord injury","publisher":"Seddio et al., Spinal Cord (Nature Publishing Group)","source_type":"peer_reviewed","statistic":"3,829 diving-related spine injury admissions over 2010–2021 in a 161-million-patient US dataset; cervical spine involved in 53.0%; estimated annual percentage change in diving-related spine injuries −4.69% (cervical) to −6.81% (thoracic)","excerpt":"\"Of 3829 persons who suffered DRSIs, the cervical spine was most frequently involved (53.0%).\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250706164632/https://www.nature.com/articles/s41393-025-01098-6","calculation_notes":"3,829 cases over 12 years implies ~319 diving-related spine injury admissions per year in the study dataset, which covered a large multi-insurer administrative population but not 100% of the US. Adjusting for coverage gaps yields estimates consistent with the NSCISC ~700/year figure. The 53% cervical fraction applied to ~700 annual cases gives ~370 cervical-spine diving SCIs per year — the subgroup at highest risk for permanent tetraplegia.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8453686/","title":"Spinal Cord Injury With Tetraplegia in Young Persons After Diving Into Shallow Water: What Has Changed in the Past 10 to 15 Years?","publisher":"Ull et al., Global Spine Journal (SAGE / PMC)","source_type":"peer_reviewed","statistic":"60 cases over 18 years at a major German spinal-injury center; 98.7% male; mean age 27.7 years; 55% presented with complete tetraplegia (AIS A) at admission; alcohol documented in 41.7% of cases; incidence stable without significant change","excerpt":"\"A total of 59 males (98.7%) and 1 female with a mean age of 27.7 years suffered an SCI from diving into shallow water between June 2001 and June 2019. At the time of admission, 33 people (55%) showed a complete tetraplegia (AIS A).\"\n","source_date":"2021-09-08","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505063143/https://pmc.ncbi.nlm.nih.gov/articles/PMC8453686/","calculation_notes":"The German series documents the demographic pattern (young male, alcohol, stable incidence) and the severity distribution — more than half of admitted cases presented as complete tetraplegia, underlining that when the mechanism produces a cervical SCI, it is more likely to be catastrophic than partial. Used here to characterize the severity and demographic profile, not to derive the US annual count.\n"},{"url":"https://bpb-us-w2.wpmucdn.com/sites.uab.edu/dist/f/392/files/2025/02/2025-Facts-and-Figures.pdf","title":"Traumatic Spinal Cord Injury Facts and Figures at a Glance 2025","publisher":"National Spinal Cord Injury Statistical Center (NSCISC), University of Alabama at Birmingham","source_type":"govt_report","statistic":"Approximately 18,421 new traumatic SCI cases per year in the US (54 per million); sports and recreation is the fourth leading cause category; diving historically cited at 4–7% of traumatic SCIs, equating to roughly 700–900 cases per year","excerpt":"\"The annual incidence of traumatic spinal cord injury is approximately 54 cases per one million people in the United States, or about 18,421 new tSCI cases each year.\"\n","source_date":"2025-02-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505063235/https://bpb-us-w2.wpmucdn.com/sites.uab.edu/dist/f/392/files/2025/02/2025-Facts-and-Figures.pdf","calculation_notes":"18,421 × 0.04 = ~737 diving SCIs at the 4% floor; × 0.07 = ~1,290 at the 7% ceiling from older NSCISC data. The 700 figure used in this entry represents a midpoint consistent with more recent trend data showing a modest decline since the pool-diving peak of the 1980s–1990s. The NSCISC dataset covers Model Systems facilities and is the canonical US authority for SCI epidemiology.\n"}],"comparison_anchors":[{"label":"Traumatic SCI from any cause (lifetime, US adult)","lifetime_us_adult":0.0033},{"label":"Dying in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.00087}],"personal_factor_multipliers":[{"factor":"dives only in designated pool areas with marked depth","multiplier":0.1,"notes":"Pool diving zones are shallow-verified and free of submerged objects; risk approaches baseline SCI incidence."},{"factor":"occasional diver in clear-water lakes or calm ocean surf","multiplier":1,"notes":"Baseline for the activity_specific_lifetime estimate above."},{"factor":"dives into murky or turbid water (rivers, ponds, quarries)","multiplier":3,"notes":"Depth and bottom hazards are not visible; the mechanism that produces SCI is undetectable pre-dive."},{"factor":"alcohol present","multiplier":2.5,"notes":"Alcohol was documented in 41.7% of diving SCIs (Ull 2021) vs. ~30% background for recreational water incidents; impairs depth-judgment and impulse control."},{"factor":"first-time dive at an unfamiliar site","multiplier":4,"notes":"Absence of prior depth reconnaissance removes the one behavioral safeguard that reduces risk."}],"short_label":"Shallow-water diving SCI","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The exposure-denominator uncertainty dominates this estimate. The 700 annual US diving SCI figure from NSCISC data is relatively stable; the denominator — how many US adults regularly dive headfirst into natural water — is inferred, not directly measured. If the true exposure population is larger than 35 million, the per-diver lifetime probability falls; if smaller (because many swimmers never dive headfirst), it rises. The Seddio 2025 study shows a declining trend in spine injury admissions from diving, which may reflect safer behavior, pool regulation, or coding shifts — the signal is there but the cause is uncertain. Complete tetraplegia (permanent paralysis of arms and legs) is the modal outcome when a diving SCI involves the cervical spine: the mechanism (axial compression of C4–C6 against a fixed bottom) is biomechanically efficient at producing severe injury precisely because the diver's own momentum loads the spine. This is one of the rare entries where the catastrophic tail is not a small fraction of cases but the central outcome.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of rippled water surface above a shallow bottom, with depth ambiguity conveyed by gradients, no people visible."},"canonical_url":"https://likelier.app/shallow-water-dive-spinal-injury","api_url":"https://likelier.app/api/fears/shallow-water-dive-spinal-injury.json"},{"slug":"choking-death","question":"What are the odds of choking to death?","quick_answer":"The lifetime odds of choking to death are about 1 in 1,100 for a US adult — higher than the near-zero most healthy adults assume. The risk is concentrated at the oldest ages, but it is not something only babies and nursing-home residents face.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Choking rarely shows up in \"what are Americans afraid of\" surveys, but parents of toddlers cite it constantly and it is one of the most commonly invoked \"freak death\" examples in casual conversation. The mental model most adults carry is that choking is something that happens to babies and to elderly people in nursing homes, and that a healthy adult who chews carefully is essentially immune. That framing is only half right.\n","rough_estimate":"Most healthy adults assume their lifetime risk is essentially zero","kind":"intuition"},"native":{"display":"~1.6 deaths per 100,000 per year (US, 2022)","numerator":1,"denominator":62500,"unit":"per year","population":"US residents, all ages, unintentional choking on food or other object"},"normalized":{"lifetime_us_adult":0.00091,"display":"1 in ~1,100 lifetime (US adult)","log_value":-3.04,"assumptions":"Uses NSC Injury Facts' headline figure of ~1.6 deaths per 100,000 population per year (≈ 5,553 deaths in 2022 across ~333M population). Naive compounding over 59 adult remaining years gives 1 - (1 - 1.6e-5)^59 ≈ 9.4e-4. The rate is strongly age-skewed: adults 65-74 sit near 3.1 per 100,000 and adults 85+ near 26.5 per 100,000 for all suffocation, while adults under 50 sit well under 1 per 100,000. Because the all-ages number is already dominated by the oldest ages most adult readers have yet to reach, the naive compounding is close to correct for a generic US adult. Final figure rounded to 9.1e-4 (≈ 1 in 1,100). Excludes infant choking/suffocation in bedding (coded separately under W75), homicidal strangulation, and intentional self-harm.\n","uncertainty":{"low":0.0007,"high":0.0012},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db199.htm","title":"Deaths From Unintentional Injury Among Adults Aged 65 and Over: United States, 2000–2013","publisher":"CDC National Center for Health Statistics (Kramarow, Chen, Hedegaard, Warner)","source_type":"govt_report","statistic":"Suffocation death rate 3.1 per 100,000 at ages 65-74 vs 26.5 per 100,000 at ages 85+ (2012-2013)","excerpt":"\"Suffocation, including deaths from positional asphyxia and choking on food or other objects, was the third leading cause of unintentional injury death among adults aged 65 and over [...] In 2012-2013, the death rate due to suffocation was more than 8 times higher among adults aged 85 and over (26.5 per 100,000) compared with adults aged 65-74 (3.1 per 100,000).\"\n","source_date":"2015-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260318012047/https://www.cdc.gov/nchs/products/databriefs/db199.htm","calculation_notes":"Establishes the U-shaped age distribution and the more than eightfold jump in suffocation mortality between ages 65-74 and 85+. Used to justify why the all-ages crude rate (~1.6 per 100,000) is dominated by the tail of the age distribution rather than by middle-aged adults, and therefore why the naive 59-year compounding approximation is roughly correct for a generic adult reader.\n","independence_note":"CDC NCHS Data Brief is the primary US age-stratified pipeline for suffocation mortality, drawn directly from the NVSS multiple-cause-of-death file. Shares upstream with Kramarow et al. (same NCHS files) and NSC Injury Facts (which republishes NCHS totals); independent of the GBD 2019 global modelling approach.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24003082/","title":"Food-related choking deaths among the elderly","publisher":"Injury Prevention (BMJ) — Kramarow E, Warner M, Chen L-H","source_type":"peer_reviewed","statistic":"2,214 food-choking deaths among US adults aged ≥65 during 2007-2010; rate highest of any age group","excerpt":"\"During 2007-2010, 2214 deaths among people aged ≥65 were attributed to choking on food. [...] Seniors experience higher fatality rates from food choking than any other age group. [...] Dementia, Parkinson's disease, and pneumonitis showed the strongest statistical associations with food-choking deaths.\"\n","source_date":"2014-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164722/https://pubmed.ncbi.nlm.nih.gov/24003082/","calculation_notes":"Peer-reviewed confirmation of the age-65+ concentration and the mechanisms (dementia, Parkinson's, dysphagia-related aspiration) behind it. Used as the authoritative anchor for the \"U-shaped distribution\" framing — adults 65+ already accounted for roughly half of all US food-choking deaths a decade before the 2022 NSC snapshot.\n","independence_note":"Kramarow et al. draw from the same NCHS multiple-cause-of-death files that underlie the CDC Data Brief. Treat the two CDC/Kramarow sources as methodologically linked rather than independent.\n"},{"url":"https://www.nsc.org/getmedia/386bd007-0d30-4da6-862a-004a239bd636/2024-nsm-webinar-injuryfacts060524.pdf","title":"Choking — Safety Topics","publisher":"National Safety Council (NSC)","source_type":"reputable_reference","statistic":"5,553 US choking deaths in 2022; ~1.6 per 100,000 population; rates rise sharply after age 71; 4th leading cause of preventable injury death","excerpt":"\"Of the 5,553 people who died from choking in 2022, rates of death rose rapidly at about age 71. [...] Choking has an average rate of 1.6 deaths per 100,000 population. [...] Choking continues to be the fourth leading cause of preventable injury death in the United States.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164754/https://www.nsc.org/getmedia/386bd007-0d30-4da6-862a-004a239bd636/2024-nsm-webinar-injuryfacts060524.pdf","calculation_notes":"Provides the most recent (2022) US totals and the crude population rate used as the native figure. 5,553 deaths / ~333M population ≈ 1.67 per 100,000 per year. Rounded to 1.6 per 100,000 to match NSC's headline. The NSC figure tracks deaths coded under ICD-10 W79 (inhalation and ingestion of food causing obstruction of respiratory tract) and W80 (inhalation and ingestion of other objects).\n","independence_note":"NSC Injury Facts is built on NCHS mortality data, so its totals are not independent of the CDC sources — used here for the most recent year coverage and for the widely cited \"lifetime odds of dying\" framing.\n"},{"url":"https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-024-17838-x","title":"The global, regional, and national burden of foreign bodies from 1990 to 2019: a systematic analysis of the GBD 2019","publisher":"BMC Public Health (2024)","source_type":"peer_reviewed","statistic":"Global age-standardized death rate for foreign bodies (choking/aspiration) in 2019: 1.55 per 100,000 (95% UI 1.41-1.67); ~109,000 deaths globally","excerpt":"\"Globally, the age-standardized death rate (ASDR) of FBs in 2019 was 1.55/100,000 (1.41/100,000-1.67/100,000).\"\n","source_date":"2024-02-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250625093136/https://bmcpublichealth.biomedcentral.com/articles/10.1186/s12889-024-17838-x","calculation_notes":"GBD 2019 global estimate provides genuine independence from US NCHS data. The global rate (1.55/100K) is comparable to the US rate (~1.6/100K), confirming the order of magnitude. Uses IHME methodology, not NCHS coding.\n","independence_note":"Fully independent of US NCHS/CDC sources — uses GBD/IHME modelling pipeline with global vital registration, verbal autopsy, and survey data.\n"}],"comparison_anchors":[{"label":"Death by drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"age 85+","multiplier":10,"notes":"CDC NCHS: choking death rate in 85+ is roughly 10x the all-adult average, driven by dysphagia and cognitive decline"},{"factor":"diagnosed dysphagia or Parkinson's disease","multiplier":15,"notes":"neurological impairment of the swallow reflex is the dominant mechanism in elderly choking deaths"},{"factor":"healthy adult under 65","multiplier":0.2,"notes":"vast majority of choking deaths cluster above age 75"},{"factor":"eating alone (no bystander capable of Heimlich)","multiplier":2,"notes":"Case series and emergency medicine literature consistently identify absence of a capable bystander as a key fatality determinant; the Heimlich maneuver, when applied within minutes, is highly effective — but it requires a present, trained bystander. Living or eating alone removes this rescue option, approximately doubling case fatality rate per observational data"},{"factor":"age under 5 (toddler)","multiplier":5,"notes":"CDC WISQARS and NCHS data: children under 5 have a per-year choking death rate several times the all-ages average, driven by developmentally normal mouthing behavior and incompletely developed swallowing coordination; hotdog, grape, coin, and toy choking are the canonical causes in this age group"}],"short_label":"Choking","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The all-ages crude rate hides a strongly U-shaped age distribution. Infants (0-4) and adults 85+ carry per-year rates several times the overall average, while adults in their 20s-50s sit well below it. Roughly half of US food-choking deaths occur in adults aged 65 and over, and most of those deaths co-occur with dementia, Parkinson's disease, or swallowing disorders — i.e., the lifetime number is not a \"bite of steak at dinner\" risk for a healthy 40-year-old, it is mostly a late-life risk whose distribution is tightly concentrated in people with specific neurological or swallowing conditions. Excludes infant positional-asphyxia deaths in bedding (coded W75, a separate and larger category), intentional strangulation, and anaphylaxis, which is coded as allergic-reaction mortality rather than choking.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single olive resting on a pale grey-blue plate, viewed from directly above, flat vector illustration."},"canonical_url":"https://likelier.app/choking-death","api_url":"https://likelier.app/api/fears/choking-death.json"},{"slug":"vaping-evali-injury","question":"How likely is a US teen who vapes to be hospitalized with EVALI?","quick_answer":"During the 2019-2020 outbreak, roughly 1 in 1,070 US teen and young-adult vapers was hospitalized with EVALI, out of ~2,807 hospitalizations nationwide. About 82% of cases involved illicit THC products, not nicotine vaping, and the outbreak faded once the culprit adulterant was removed from the supply.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"During 2019–2020, US media coverage of the EVALI outbreak — dozens of teenagers hospitalized with mysterious lung injury after vaping — produced intense parental fear about vaping-related respiratory harm. The images of previously healthy teens on ventilators were vivid and specific. The fear persists in the post-outbreak period even though the outbreak was traced to a specific adulterant (vitamin E acetate in illicit THC cartridges) that was largely removed from the supply after the CDC investigation. Current fear about vaping may conflate outbreak-era EVALI risk with ongoing nicotine-vape or nicotine-free vaping.\n","kind":"intuition"},"native":{"display":"~2,807 hospitalizations in US over the 2019–2020 outbreak window; outbreak rate ~1 in 1,070 US teen/young-adult vapers","numerator":2807,"denominator":3000000,"unit":"per outbreak-window (Aug 2019–Feb 2020)","population":"US adolescents and young adults who vaped during the 2019–2020 EVALI outbreak (CDC EVALI case count over an internal ~3M working-estimate denominator; the cited 2019 NYTS anchor puts current youth e-cigarette users at ~5.3M)"},"normalized":{"lifetime_us_adult":0.00094,"display":"roughly 1 in 1,070 US teen/young-adult vapers during the outbreak window","log_value":-3.03,"assumptions":"CDC EVALI surveillance: 2,807 hospitalized cases and 68 deaths by February 2020 (when CDC stopped active surveillance after the primary causative agent — vitamin E acetate in THC cartridges — was identified and largely removed from the supply). ~82% of cases involved THC-containing products; 57% involved both THC and nicotine. Denominator: an internal ~3,000,000 working estimate for teens + young adults who vaped during the peak window (NOT a directly cited published survey total). For scale, the cited 2019 NYTS anchor (Cullen et al., JAMA) puts current youth e-cigarette users at ~5.3 million (4.1M high school + 1.2M middle school), and the EVALI case series had a median age of 21 — so the true exposed population (adding college-age young adults) is likely ≥5 million. Using the internal 3 million denominator: 2,807 / 3,000,000 ≈ 0.00094 (roughly 1 in 1,070). Because the citable population figure (≥5.3M) is LARGER than 3M, the true outbreak rate is RARER than the displayed 1 in 1,070 (e.g. 2,807 / 5,300,000 ≈ 1 in 1,888) — the headline is therefore on the more-alarming side of the citable evidence, not conservative. Directionally, the rate for THC-product users specifically is higher than the all-vaper rate (~82% of cases; those who used only nicotine-based products had near-zero EVALI exposure), so restricting the denominator to THC users raises the rate while restricting the numerator does not. Post-vit-E-removal background EVALI rate is an order of magnitude lower. Low (0.0001): current (post-outbreak) background rate for regulated nicotine products. High (0.005): peak outbreak rate for those using illicit THC cartridges specifically.\n","uncertainty":{"low":0.0001,"high":0.005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://archive.cdc.gov/www_cdc_gov/tobacco/basic_information/e-cigarettes/severe-lung-disease.html","title":"Outbreak of Lung Injury Associated with the Use of E-Cigarette, or Vaping, Products (EVALI) — Final Update","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"2,807 hospitalized EVALI cases and 68 deaths reported to CDC by February 18, 2020; 82% involved THC products; vitamin E acetate identified as primary culprit","excerpt":"\"As of February 18, 2020, a total of 2,807 hospitalized EVALI cases or deaths have been reported to CDC from 50 states, the District of Columbia, Puerto Rico, and the U.S. Virgin Islands. Among those with information on substances used, 82 percent reported using THC-containing products. Vitamin E acetate has been strongly linked to the EVALI outbreak. CDC, states, and the FDA recommend that people not use THC-containing e-cigarette or vaping products, particularly from informal sources like friends, family, or in-person or online dealers.\"\n","source_date":"2020-02-25","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260425200545/https://archive.cdc.gov/www_cdc_gov/tobacco/basic_information/e-cigarettes/severe-lung-disease.html","calculation_notes":"CDC final EVALI update — source of the 2,807 numerator only. 2,807 hospitalizations / an internal ~3 million working-estimate denominator of US teen + young adult vapers during the outbreak window ≈ 0.00094 (roughly 1 in 1,070). The denominator is not from this source; see the 2019 NYTS (Cullen et al.) entry, which anchors current youth vapers at ~5.3 million (a larger figure → rarer true rate). Since ~82% of cases involved THC products, the rate for THC-product users specifically is higher. The post-vit-E-removal era (post-2020) has a substantially lower rate; the figure here reflects the outbreak-era population-average risk across all teen/young-adult vapers.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6865299/","title":"e-Cigarette Use Among Youth in the United States, 2019","publisher":"JAMA","source_type":"peer_reviewed","statistic":"In 2019 (outbreak year), an estimated 4.1 million US high school students (27.5%) and 1.2 million middle school students (10.5%) currently used e-cigarettes — ~5.3 million youth vapers","excerpt":"\"An estimated 27.5% (95% CI, 25.3%-29.7%) of high school students and 10.5% (95% CI, 9.4%-11.8%) of middle school students reported current e-cigarette use. ... With the assumption that the prevalence estimates from this survey are nationally representative and could be used to project to national population totals for US high school and middle school students, the data would suggest that in 2019, an estimated 4.1 million high school students and 1.2 million middle school students currently use e-cigarettes.\"\n","source_date":"2019-11-05","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260617182541/https://pmc.ncbi.nlm.nih.gov/articles/PMC6865299/","calculation_notes":"Cullen et al., JAMA 2019 (CDC-authored analysis of the 2019 National Youth Tobacco Survey, N=19,018). Establishes the outbreak-era youth vaping denominator: ~4.1 million high school + ~1.2 million middle school = ~5.3 million current youth e-cigarette users in 2019, the peak EVALI year. This is the citable anchor for the \"millions of teen/young-adult vapers\" denominator used in the headline. Note NYTS surveys only middle + high school students (all under 18); the EVALI case series had a median age of 21 (NEJM), so the true exposed population extends into young adults beyond this frame and is therefore ≥5.3 million. A larger true denominator makes the outbreak rate RARER than the displayed ~1 in 1,070 (which divides by an internal 3,000,000 working estimate) — the headline is thus, if anything, on the more-alarming side, not conservative.\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1911614","title":"Pulmonary Illness Related to E-Cigarette Use in Illinois and Wisconsin — Final Report","publisher":"New England Journal of Medicine","source_type":"peer_reviewed","statistic":"Case series of 98 EVALI patients: median age 21, 89% used THC products, 95% hospitalized; 26% required intubation and mechanical ventilation","excerpt":"\"There were 98 case patients, 79% of whom were male; the median age of the patients was 21 years. A total of 95% of the patients were hospitalized, 26% underwent intubation and mechanical ventilation, and two deaths were reported. A total of 89% of the patients reported having used tetrahydrocannabinol products in e-cigarette devices, although a wide variety of products and devices was reported.\"\n","source_date":"2020-03-05","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260411224238/https://www.nejm.org/doi/full/10.1056/NEJMoa1911614","calculation_notes":"Layden et al. NEJM 2020 — final report from the two states that first identified the EVALI cluster. Establishes clinical severity (median age 21, 95% hospitalized, 26% intubated) and THC product involvement (89%). Confirms CDC's national pattern. Used to characterize the severity of the outcome (serious_harm) and the THC-product causation.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7341a2.htm","title":"Tobacco Product Use Among US Middle and High School Students — National Youth Tobacco Survey, 2024","publisher":"CDC / Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"7.8% of US high school students currently used e-cigarettes in 2024 (down from 10.0% in 2023; ~1.21 million students); e-cigarettes remained the most commonly used tobacco product among youth","excerpt":"\"E-cigarettes were the most commonly used product (7.8%) among high school students. Among high school students, declines occurred in current use of any tobacco product (from 12.6% to 10.1%) and e-cigarettes (from 10.0% to 7.8%). Overall, current e-cigarette use decreased (from 1.56 million to 1.21 million) high school students from 2023 to 2024.\"\n","source_date":"2024-10-17","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260511132325/https://www.cdc.gov/mmwr/volumes/73/wr/mm7341a2.htm","calculation_notes":"CDC NYTS 2024. Provides the current teen vaping prevalence denominator context (7.8% of high schoolers, ~1.21 million students). The decline from 10.0% (2023) to 7.8% (2024) — and the broader fall from the ~27% peak years earlier — confirms that the outbreak-era exposure scenario is no longer current. The current background EVALI rate among regulated nicotine vapers is substantially lower; the normalized rate reflects the outbreak-era THC-vaping scenario.\n"}],"comparison_anchors":[{"label":"Teen MVA fatality, US (5-year window)","lifetime_us_adult":0.00033},{"label":"SIDS per live birth (US)","lifetime_us_adult":0.000345}],"short_label":"EVALI vaping hospitalization","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"EVALI was an outbreak caused by a specific adulterant — vitamin E acetate — added to illicit THC vaping cartridges in the US black market. This was a US-specific phenomenon with no comparable international outbreak; other countries did not see equivalent EVALI clusters because their cannabis product markets differ. The outbreak-era rate (2019–2020) does not apply to the post-vitamin-E-removal period, and the caveats apply specifically to illicit/unregulated THC vaping cartridges, not to regulated nicotine e-cigarettes. Current teen nicotine vaping (7.8% of US high schoolers in 2024) carries a different and lower EVALI-specific risk profile. The long-term respiratory and cardiovascular effects of chronic nicotine vaping in adolescents are a separate and ongoing research question not captured by this entry, which is scoped to the specific acute EVALI outcome.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a small vape device resting on a surface, warning symbol nearby."},"canonical_url":"https://likelier.app/vaping-evali-injury","api_url":"https://likelier.app/api/fears/vaping-evali-injury.json"},{"slug":"untreated-childhood-scoliosis","question":"What are the odds of serious disability from untreated childhood scoliosis?","quick_answer":"About 1 in 1,000 US adults ends up with serious disability from untreated adolescent scoliosis, and only ~0.7% of diagnosed cases reach surgery within five years. Most curves stay mild, despite the common parental fear that scoliosis inevitably progresses to severe deformity.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Scoliosis screening in schools was once universal in the United States, and the image of a child bending forward while a nurse examines their spine remains embedded in generational memory. The implicit message was that undetected scoliosis progresses to disfiguring curvature and chronic pain. The US Preventive Services Task Force withdrew its recommendation for routine school screening in 2018, concluding that the harms of overdiagnosis and unnecessary treatment outweighed the benefits. Despite this, parental anxiety remains high, fueled by the visibility of bracing and the dramatic imagery of severe untreated curves in medical textbooks — images that represent the extreme tail of a condition that is overwhelmingly mild.\n","rough_estimate":"Many parents believe untreated scoliosis will progress to severe deformity","kind":"intuition"},"native":{"display":"~0.7% of diagnosed scoliosis patients undergo surgery within 5 years (nationwide database study)","numerator":7,"denominator":1000,"unit":"per adolescent scoliosis case progressing to surgical threshold (0.7% 5-year surgery rate)","population":"Adolescents diagnosed with idiopathic scoliosis (Cobb angle ≥10°)"},"normalized":{"lifetime_us_adult":0.001,"display":"~1 in 1,000 US adults (serious disability from untreated adolescent scoliosis)","log_value":-3,"assumptions":"Adolescent idiopathic scoliosis (AIS) prevalence is 2-3% for curves ≥10° (Cobb angle). The vast majority are mild: ~90% have curves under 20° that require only monitoring. The BrAIST trial (NEJM 2013) showed bracing reduced progression past 50° from 52% to 28% in the 20-40° subset, but this subset is already a minority of all diagnosed scoliosis.\nThe nationwide database study (Incidence and Surgery Rate of Idiopathic Scoliosis, 2021) found an overall 5-year surgery rate of 0.7% among newly diagnosed scoliosis patients. For the 10-14 age group, 1.14% underwent surgery within 5 years.\nFor the normalized estimate: 2.5% of adolescents have AIS (midpoint), of whom ~0.7% require surgery over 5 years = ~0.0175% of the general adolescent population. Extrapolating to lifetime functional disability (including non-surgical moderate curves that cause chronic pain), the estimate is approximately 0.1% of the general adult population — roughly 1 in 1,000. This is an upper bound that includes both surgical cases and conservatively managed cases with residual functional impairment.\nLong-term follow-up studies of untreated mild scoliosis (Weinstein 2003, Iowa 50-year follow-up) found that most patients with curves under 30° were functionally indistinguishable from controls.\n","uncertainty":{"low":0.0005,"high":0.002},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1307337","title":"Effects of Bracing in Adolescents with Idiopathic Scoliosis","publisher":"New England Journal of Medicine","source_type":"primary_study","statistic":"72% treatment success with bracing vs 48% with observation alone (success = not progressing past 50° Cobb angle)","excerpt":"\"The rate of treatment success was 72% after bracing, as compared with 48% after observation, and bracing significantly decreased the progression of high-risk curves to the threshold for surgery.\"\n","source_date":"2013-10-17","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250701122713/https://www.nejm.org/doi/full/10.1056/NEJMoa1307337","calculation_notes":"The BrAIST trial (Weinstein et al. 2013) is the landmark RCT for scoliosis bracing. Critically, it enrolled only patients with 20-40° curves who were at high risk of progression — a subset of a subset. Among this high-risk group, 48% did not progress past 50° even WITHOUT bracing. The 72% success rate with bracing is significant but highlights that even in the highest- risk untreated group, nearly half do not reach surgical threshold. For the 90%+ of scoliosis patients with curves under 20°, the progression risk is far lower, and the BrAIST results do not apply.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK499908/","title":"Adolescent Idiopathic Scoliosis","publisher":"StatPearls (NCBI Bookshelf)","source_type":"reputable_reference","statistic":"AIS prevalence 1-3% in individuals aged 10-18; prevalence of curves >40° (surgical threshold) is ~0.1%","excerpt":"\"The prevalence is about 1% to 3% for AIS. It occurs in individuals between the ages of 10 to 18. The prevalence is approximately 0.1% for curves measuring more than 40 degrees (those which tend to be those requiring operative intervention).\"\n","source_date":"2024-01-25","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260504061413/https://www.ncbi.nlm.nih.gov/books/NBK499908/","calculation_notes":"StatPearls provides the prevalence-to-treatment funnel: 1-3% diagnosed, ~0.1% with curves >40° requiring surgery. This means roughly 90%+ of diagnosed scoliosis cases need nothing beyond observation. The overwhelming majority of adolescents with scoliosis will reach adulthood without functional impairment regardless of whether they receive treatment.\n"},{"url":"https://www.mdpi.com/1660-4601/18/15/8152","title":"Incidence and Surgery Rate of Idiopathic Scoliosis: A Nationwide Database Study","publisher":"International Journal of Environmental Research and Public Health","source_type":"peer_reviewed","statistic":"Overall 5-year surgery rate 0.7% for newly diagnosed scoliosis; 1.14% for ages 10-14","excerpt":"\"The overall 5-year surgery rate for newly diagnosed idiopathic scoliosis patients was 0.7%, with variations by sex and age group. Among AIS patients diagnosed at 10-14 years, 1.14% of patients underwent surgery within five years.\"\n","source_date":"2021-08-03","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260128092736/https://www.mdpi.com/1660-4601/18/15/8152","calculation_notes":"This nationwide database study provides real-world surgery rates rather than clinical-trial estimates. The 0.7% overall surgery rate confirms that the vast majority of diagnosed scoliosis cases are managed without surgery. Even in the peak age group (10-14), only 1.14% reach the surgical threshold within 5 years. This is consistent with the StatPearls estimate that only 0.3-0.5% of the general adolescent population has progressive curves.\n"}],"comparison_anchors":[{"label":"Appendicitis (lifetime, US adult)","lifetime_us_adult":0.07},{"label":"ACL tear requiring surgery (lifetime, US adult)","lifetime_us_adult":0.02},{"label":"Congenital heart defect requiring surgery (lifetime)","lifetime_us_adult":0.004}],"personal_factor_multipliers":[{"factor":"Cobb angle ≥25° at diagnosis","multiplier":4,"notes":"Lonstein & Carlson (JBJS, 1984) found that curves ≥25° at the onset of the adolescent growth spurt had a 68% progression risk vs. 22% for curves <20°. StatPearls (AIS, 2024) confirms that Cobb angle at presentation is the single strongest predictor of curve progression. The BrAIST trial (NEJM 2013) enrolled only 20–40° curves as the high-progression-risk stratum, implying the overall 0.7% surgery rate is dominated by initial curve magnitude."},{"factor":"Pre-menarchal diagnosis (Risser grade 0–1)","multiplier":3,"notes":"Lonstein & Carlson (JBJS, 1984) established that skeletal immaturity at diagnosis (low Risser grade, pre-menarche) is a major independent predictor of progression, because more skeletal growth remains to drive curve advancement. The BrAIST trial used Risser grade 0–2 as an enrollment criterion precisely because this reflects the highest-progression stratum. A pre-menarchal girl with a 20° curve has substantially more progression risk than a post-menarchal girl with the same initial angle."},{"factor":"Family history of scoliosis","multiplier":2.5,"notes":"StatPearls (AIS, 2024) reports that nearly 30% of patients with adolescent idiopathic scoliosis have a family member with scoliosis, indicating a substantial familial component. A family history signals a genetic background that raises the baseline likelihood of developing AIS; the entry retains a conservative 2.5× factor for the elevated prevalence and progression risk implied by a treated affected relative, while noting that no cited source quantifies a precise familial progression ratio."}],"short_label":"Untreated childhood scoliosis","myth_framing":"overrated","outcome_severity":"serious_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers adolescent idiopathic scoliosis (AIS), which accounts for ~80% of scoliosis cases. Congenital scoliosis, neuromuscular scoliosis (cerebral palsy, muscular dystrophy), and infantile/juvenile scoliosis have different natural histories and higher rates of progression. The \"serious disability\" threshold used here includes surgical cases and cases with chronic pain or functional limitation attributable to the curve — not cosmetic concerns about asymmetry, which are more common but do not constitute disability. The Weinstein 50-year Iowa follow-up found that untreated scoliosis patients with curves under 30° had similar function, pain levels, and self-image as matched controls, though curves over 50° were associated with increased back pain and decreased pulmonary function. The USPSTF withdrew its recommendation for routine school scoliosis screening in 2018.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A spine model showing a gentle curve against a neutral background, flat vector illustration in muted blues and greys."},"canonical_url":"https://likelier.app/untreated-childhood-scoliosis","api_url":"https://likelier.app/api/fears/untreated-childhood-scoliosis.json"},{"slug":"emt-paramedic-duty-death","question":"What are the odds of an EMT or paramedic dying in the line of duty over a career?","quick_answer":"An EMT or paramedic's odds of dying in the line of duty over a 20-year career are about 1 in 910. Most people underestimate EMS career mortality.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Ambulance work is broadly understood as physically demanding and occasionally hazardous, but public imagination of the danger tends to focus on the patients rather than the providers. The iconic image of a paramedic sprinting into a burning building or extracting a victim from a wreck frames EMS as a supporting character in someone else's emergency. The vehicle itself — the ambulance — rarely registers as the threat. No standalone survey measuring public estimates of EMT or paramedic career mortality was identified; perceived risk is characterized here as editorial intuition. The gap between perception and reality runs in the direction of underestimation: most people would not spontaneously rank ambulance driving among the top-tier hazards of the occupation, yet transportation incidents account for roughly three-quarters of all fatal occupational injuries among paramedicine clinicians, according to an 18-year cohort analysis published in Prehospital and Disaster Medicine in 2023.\n","rough_estimate":"most people underestimate EMS career mortality; the actual traumatic line-of-duty death risk over a 20-year career is around 1 in 900","kind":"intuition"},"native":{"display":"~16 fatal occupational injuries per year among ~283,000 paid EMTs and paramedics (rate: ~5.5 per 100,000/year)","numerator":16,"denominator":282900,"unit":"fatal occupational injuries per paid EMT/paramedic per year","population":"US paid EMTs and paramedics (BLS OES, 2024)"},"normalized":{"lifetime_us_adult":0.0011,"display":"~1 in 910 over a 20-year career","log_value":-2.96,"assumptions":"Reference subgroup: a US paid EMT or paramedic serving a full 20-year career (a commonly cited career horizon in EMS occupational literature, reflecting high turnover and early attrition in the profession; many agencies use 20 years for pension and benefit vesting). The annual fatal occupational injury rate of approximately 5.5 per 100,000 comes from Roth et al. (2023, Prehospital and Disaster Medicine), an 18-year BLS CFOI cohort (2003-2020, n=204 fatalities, average workforce 206,000) that the study itself describes as 60% higher than the rate for all US workers over the same period. Maguire et al. (2002, Annals of Emergency Medicine) found a higher rate of 12.7 per 100,000 using three reconciled fatality databases (CFOI, the National EMS Memorial Service, and NHTSA's Fatality Analysis Reporting System) over an earlier and shorter 1992-1997 window; that figure is treated here as an alternate, higher-end scenario (see caveats) rather than the headline, because the 2023 cohort is longer-running, more recent, and CFOI-only. The current-era annual death count is estimated by applying 5.5 per 100,000 to the 2024 BLS OES paid workforce of 181,000 EMTs + 101,900 paramedics = 282,900 total, yielding approximately 15.6 deaths per year, rounded to 16. Lifetime career probability over 20 years: 1 - (1 - 0.000055)^20 ≈ 0.0011, or roughly 1 in 910. The scope is activity_specific_lifetime because this is career-specific risk for a defined occupational subgroup, not a general US-adult lifetime probability. The figure covers traumatic occupational deaths as captured by CFOI; it excludes occupational disease, long-term cardiovascular sequelae, and COVID-19 deaths, which would increase the total. The denominator of 282,900 excludes the large volunteer EMS workforce (estimated at several hundred thousand additional providers); volunteer fatalities may or may not be captured in CFOI depending on employment classification.\n","uncertainty":{"low":0.0009,"high":0.0026},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cambridge.org/core/journals/prehospital-and-disaster-medicine/article/cohort-study-of-occupational-fatalities-among-paramedicine-clinicians-2003-through-2020/E7E947348A18F36FD1992D2DAEB5A988","title":"A Cohort Study of Occupational Fatalities among Paramedicine Clinicians: 2003 through 2020","publisher":"Prehospital and Disaster Medicine (Cambridge University Press)","source_type":"peer_reviewed","statistic":"204 fatal occupational injuries among paramedicine clinicians in the US from 2003 through 2020 (18 years); 153 of 204 (75%) were transportation-related; average workforce of approximately 206,000 over the period; fatality rate approximately 5.5 per 100,000 per year","excerpt":"\"A total of 204 fatal injuries were identified among paramedicine clinicians during the study period (2003-2020). Of these, 153 (75.0%) were the result of transportation incidents. From 2010 through 2020, available data on the annual number of employed paramedicine clinicians showed that the total varied between a low of 172,000 and a high of 261,000 (Avg: 206,000; SD = 28,000).\"\n","source_date":"2023-03-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260525094739/https://www.cambridge.org/core/journals/prehospital-and-disaster-medicine/article/cohort-study-of-occupational-fatalities-among-paramedicine-clinicians-2003-through-2020/E7E947348A18F36FD1992D2DAEB5A988","calculation_notes":"204 deaths over 18 years = 11.3 deaths per year average. Average annual workforce of 206,000. Implied rate: 11.3 / 206,000 = 5.49 per 100,000 per year, which the study itself describes as 60% higher than the rate for all United States workers over the same period (95% CI, 1.24 to 2.04) and eight times higher than the rate for other health care practitioners (95% CI, 5.8 to 10.1). This rate reflects strict BLS CFOI classification and likely undercounts volunteer EMS fatalities and deaths misclassified to other occupational categories (e.g., fire fighter). Transportation (ground + air combined) accounted for 153/204 = 75% of all fatalities — the dominant cause by a large margin. 20-year career probability at this rate: 1 - (1 - 0.0000549)^20 ≈ 0.00110 (~1 in 910). This entry uses this 5.5-per-100,000 CFOI-cohort rate as the headline estimate; the older Maguire et al. (2002) rate of 12.7 per 100,000 is treated as an alternate, higher-end scenario (see caveats) because it draws on different fatality databases over an earlier and shorter (1992-1997) study window.\n","independence_note":"This 2023 peer-reviewed cohort study used BLS CFOI microdata provided directly by the US Department of Labor, covering the full 2003-2020 period. It is methodologically distinct from the NIOSH/MMWR surveillance reports and Maguire et al. (2002), which covered earlier time periods or used different data sources.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/12447340/","title":"Occupational Fatalities in Emergency Medical Services: A Hidden Crisis","publisher":"Annals of Emergency Medicine","source_type":"peer_reviewed","statistic":"EMS worker fatality rate estimated at 12.7 per 100,000 workers annually during a six-year study period; 67 ground transportation deaths, 19 air ambulance deaths, 13 cardiovascular, 10 homicides in the study window; compares to 14.2 for police and 5.0 for all US workers","excerpt":"\"During the 6-year study period, we identified at least 67 ground transportation-related fatalities, 19 air ambulance crash fatalities, 13 deaths resulting from cardiovascular incidents, 10 homicides, and 5 other causes, resulting in 114 EMS worker fatalities. The estimated fatality rate was 12.7 fatalities per 100,000 EMS workers annually, which compares with 14.2 for police, 16.5 for firefighters, and a national average of 5.0 during the same time period.\"\n","source_date":"2002-12-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20250903000913/https://pubmed.ncbi.nlm.nih.gov/12447340/","calculation_notes":"Maguire et al. (2002) identified 114 EMS fatalities over a 6-year window, reconciled across three databases (Census of Fatal Occupational Injuries, 1992-1997; National EMS Memorial Service, 1992-1997; and NHTSA's Fatality Analysis Reporting System, 1994-1997), yielding a rate of 12.7/100,000. Ground transport (67) + air ambulance (19) = 86 transportation deaths, or 86/114 = 75.4% of total — consistent with the 2023 Cambridge cohort's 75% transportation share. This entry's headline estimate instead uses the 5.5/100,000 rate from the 2023 Cambridge CFOI-only cohort (source above), because it is longer-running (18 years vs. 6) and more recent; Maguire's 12.7/100,000 is treated as an alternate, higher-end scenario (see caveats). Applying 12.7/100,000 to 282,900 current workers ≈ 35.9 deaths/year; 20-year career at that rate: 1 - (1 - 0.000127)^20 ≈ 0.00252 (~1 in 400).\n","independence_note":"Maguire et al. (2002) is the founding peer-reviewed study quantifying EMS occupational fatality rates. It used multiple supplementary data sources beyond BLS CFOI to capture deaths that CFOI's occupational classification misses. Its findings are methodologically complementary to the 2023 CFOI cohort study, which used a more conservative but systematic CFOI-only approach over a longer follow-up window.\n"},{"url":"https://www.bls.gov/ooh/healthcare/emts-and-paramedics.htm","title":"EMTs and Paramedics: Occupational Outlook Handbook","publisher":"US Bureau of Labor Statistics","source_type":"govt_report","statistic":"EMTs held about 181,000 jobs and paramedics held about 101,900 jobs in 2024, for a combined paid workforce of approximately 282,900","excerpt":"\"Emergency medical technicians held about 181,000 jobs in 2024. Paramedics held about 101,900 jobs in 2024. These employment data exclude volunteer EMTs and paramedics, who share many of the same duties as paid EMTs and paramedics.\"\n","source_date":"2025-09-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260516141309/https://www.bls.gov/ooh/Healthcare/EMTs-and-paramedics.htm","calculation_notes":"181,000 EMTs + 101,900 paramedics = 282,900 total paid US EMS workforce in 2024. This is the denominator used to convert the per-100,000 rate (5.5, from the Cambridge cohort above) into an annual death count: 282,900 × 0.000055 ≈ 15.6/year, rounded to 16. Volunteer EMS providers are excluded from this count; estimates of volunteer EMS workers range from 200,000 to 500,000 depending on definition and data source.\n"},{"url":"https://www.bls.gov/news.release/cfoi.nr0.htm","title":"Census of Fatal Occupational Injuries Summary, 2024","publisher":"US Bureau of Labor Statistics","source_type":"govt_report","statistic":"All-industry US fatal work injury rate was 3.3 per 100,000 full-time-equivalent workers in 2024, down from 3.5 in 2023","excerpt":"\"The fatal work injury rate was 3.3 fatalities per 100,000 full-time equivalent (FTE) workers in 2024, a decrease from a rate of 3.5 in 2023.\"\n","source_date":"2026-02-19","source_accessed":"2026-07-03","archive_url":"https://web.archive.org/web/20260702153945/https://www.bls.gov/news.release/cfoi.nr0.htm","calculation_notes":"This is the all-industry CFOI topline rate, used in the body prose as the comparison baseline against the EMS-specific headline rate of 5.5 per 100,000 (2003-2020 cohort study above) — roughly double the all-worker average. It is the same BLS CFOI program as the EMS cohort study cited above, reporting the all-industry topline rather than the EMS occupational subset; it is an independent annual release, not a re-derivation of the EMS-specific figures.\n"},{"url":"https://www.hmpgloballearningnetwork.com/site/emsworld/article/12110600/a-national-perspective-on-ambulance-crashes-and-safety","title":"A National Perspective on Ambulance Crashes and Safety","publisher":"EMS World (reporting NHTSA Special Crash Investigations data)","source_type":"reputable_reference","statistic":"38 of 45 EMS providers (84%) in the ambulance patient compartment were unrestrained at the time of a crash, per NHTSA Special Crash Investigations reviewed over more than a decade","excerpt":"\"Of the 45 providers in the patient compartment at the time of the crash, only 7 (16%) were wearing a seat belt at the time of the crash, meaning 38 (84%) were unrestrained.\"\n","source_date":"2015-09-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20241203102910/https://www.hmpgloballearningnetwork.com/site/emsworld/article/12110600/a-national-perspective-on-ambulance-crashes-and-safety","calculation_notes":"NHTSA's Special Crash Investigations program conducted more than 50 in-depth ambulance crash reviews over a decade-plus window; this EMS World analysis reports that 38 of 45 patient-compartment providers across those investigations (84%) were unrestrained at the time of the crash. This is the source for the \"approximately 84 percent... unrestrained\" figure in the body prose, offered as a structural explanation for why lights-and-sirens transport crash exposure translates into elevated EMS occupational risk.\n"}],"comparison_anchors":[{"label":"US police officer (traumatic duty death, 25-year career)","lifetime_us_adult":0.0032},{"label":"US average worker (all-cause traumatic occupational death, 20-year career)","lifetime_us_adult":0.0007},{"label":"Commercial fisherman (career, US)","lifetime_us_adult":0.025}],"personal_factor_multipliers":[{"factor":"Air medical (HEMS) crew vs ground EMS","multiplier":3.5,"notes":"An NTSB analysis of helicopter EMS operations found a fatal accident rate of 5.4 per 100,000 flight hours for HEMS operations — approximately 3.5 times the fatal accident rate for other Part 135 helicopter operations. HEMS crew (pilot, flight nurse, flight paramedic) are exposed to this elevated aviation risk on every mission. More recent analyses have shown safety improvements over time, and some cross-modal comparisons suggest ground transport crash rates per transport episode are higher than HEMS; however, per flight hour or per career, HEMS crew face substantially higher aviation fatality exposure than ground EMS workers. (Source: NTSB analysis of HEMS accidents cited in multiple aviation safety literature; see also NTSB Special Investigation Report SIR-06-01.)"},{"factor":"Lights-and-sirens (L&S) transport vs non-emergency driving","multiplier":2.5,"notes":"A 2019 Annals of Emergency Medicine analysis of National EMS Information System (NEMSIS) data found that ambulance crash rates during transport with lights and sirens were 17.1 per 100,000 transports, versus 7.0 per 100,000 without L&S — an adjusted odds ratio of approximately 2.9 for crash involvement during the transport phase. EMS workers who frequently operate in L&S transport mode face correspondingly elevated crash exposure. (Source: Watanabe et al., Ann Emerg Med, 2019, PMID 30648537.)"},{"factor":"Rural long-distance transport vs urban EMS","multiplier":2,"notes":"Ambulance crashes on rural roadways are more likely to result in fatal outcomes for EMS personnel than urban crashes, due to higher speeds, more hazardous road geometry, longer extrication times, and delayed trauma care. Rural EMS providers cover larger geographic territories and log more vehicle miles per shift than their urban counterparts. Rural Health Information Hub literature review on ambulance crashes confirms that rural road conditions and longer transport distances elevate both crash frequency and case fatality among EMS workers. (Source: Rural Health Information Hub, rural ambulance crash literature review; Maguire et al. 2002 noted rural fatality clustering.)"},{"factor":"High-violence urban deployment zone","multiplier":1.8,"notes":"Assault-related fatalities account for approximately 10-11% of EMS occupational deaths (Maguire et al. 2002: 10 homicides of 114 deaths; Cambridge cohort 2003-2020 also documents assault deaths). Violence-related injury rates for EMS personnel overall are 15.5 per 10,000 workers — more than twice the national average — and some agencies in high-crime urban settings report rates as high as 60 per 10,000 (approximately 22 times the national average). EMS workers assigned to high-violence urban districts face a meaningfully elevated assault and homicide risk component layered on top of the transportation risk. (Source: Maguire et al. 2002; EMT violence review, PMC5637660.)"}],"short_label":"EMS duty death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This figure covers traumatic occupational deaths as captured by the BLS Census of Fatal Occupational Injuries (CFOI). CFOI systematically undercounts EMS fatalities for two reasons: (1) many EMS workers, particularly those employed by fire departments or municipal agencies, are classified as fire fighters or other occupations rather than EMTs or paramedics; (2) the large volunteer EMS workforce (~200,000-500,000 additional providers by various estimates) may be incompletely captured in paid-worker statistics. Maguire et al. (2002) reconciled CFOI with the National EMS Memorial Service and NHTSA's Fatality Analysis Reporting System over an earlier, shorter 1992-1997 window and found a fatality rate of 12.7/100,000 — more than double the CFOI-cohort figure of 5.5/100,000 above — suggesting the true career risk may be closer to 1 in 400 over a 20-year career than 1 in 910. The entry uses the more recent, longer- running CFOI-only cohort rate as the headline because it reflects a single systematic methodology over 18 years (2003-2020) rather than a three-database reconciliation over an earlier 6-year window. Deaths from occupational disease, cardiac events linked to job stress (which appear in Maguire et al. as a distinct category), and COVID-19 are excluded. The 20-year career horizon is a midpoint estimate; EMS careers in practice range from 5 to 30+ years depending on employer type, physical demands, and burnout. Non-fatal injuries — which number in the tens of thousands per year and include back injuries, needlestick exposures, and assault-related injuries — far exceed the death toll.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"An ambulance steering wheel and dashboard viewed from the driver's seat, flat vector illustration."},"canonical_url":"https://likelier.app/emt-paramedic-duty-death","api_url":"https://likelier.app/api/fears/emt-paramedic-duty-death.json"},{"slug":"naion-semaglutide-vision-loss","question":"What are the odds of sudden permanent vision loss (NAION) from taking semaglutide (Ozempic/Wegovy)?","quick_answer":"For someone taking semaglutide, the odds of sudden permanent vision loss (NAION) are roughly 1 in 900 over a five-year course, of which about 1 in 1,400 is the drug-attributable excess above baseline. The EU label files it as 'very rare' — up to 1 in 10,000 people. Awareness is near-absent or headline-driven, rarely landing at the actual order of magnitude.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"NAION entered public awareness abruptly in mid-2024, when a Harvard referral-clinic study tied semaglutide to a several-fold higher rate of sudden one-eye vision loss, and again in June 2025 when the European Medicines Agency formally added it to the label as a \"very rare\" side effect. Most people taking these drugs have never heard the term, and the coverage that did reach them swung between \"blindness warning\" headlines and reassurance that the absolute risk is tiny. The result is a fear that is either absent or wildly miscalibrated, rarely sitting at the actual order of magnitude.\n","rough_estimate":"Perception ranges from unaware to headline-driven 'blindness' alarm; no published survey of NAION awareness or perceived likelihood exists","kind":"intuition"},"native":{"display":"~1 in 10,000 semaglutide users (EMA 'very rare' frequency band)","numerator":1,"denominator":10000,"unit":"per person taking semaglutide (EU frequency category)","population":"adults taking semaglutide (Ozempic/Wegovy/Rybelsus)"},"normalized":{"lifetime_us_adult":0.0011,"display":"~1 in 900 over a 5-year course (total on-drug); excess attributable ~1 in 1,400","log_value":-2.96,"assumptions":"Scope is the subgroup of adults taking semaglutide; \"lifetime\" here is the treatment-course duration, NOT the site-default 59-year adult horizon. EMA's headline \"very rare / up to 1 in 10,000\" is the EU SmPC frequency-category band (a label tier), not a measured cumulative incidence, so we do not normalize on it. We normalize on the population-based per-person-year rate that EMA and the Simonsen 2025 Danish-Norwegian cohort independently agree on. Simonsen reports an excess of +1.41 NAION events per 10,000 person-years (incidence-rate difference) and a pooled hazard ratio of 2.81; EMA gives \"approximately one additional case per 10,000 person-years.\" Implied diabetic baseline = IRD / (HR - 1) = 1.41 / 1.81 ≈ 0.78 per 10,000 PY; total on-drug rate = baseline × HR ≈ 0.78 × 2.81 ≈ 2.2 per 10,000 PY. Over a central 5-year course: 2.2 × 5 / 10,000 ≈ 1.1e-3 ≈ 1 in 900 (total). The drug-attributable EXCESS over the same course is 1.41 × 5 / 10,000 ≈ 7e-4 ≈ 1 in 1,400. The headline 0.0011 is TOTAL on-drug risk; the excess is reported in caveats. The cumulative figure exceeds the \"1 in 10,000\" label band precisely because that band is a per-exposure frequency tier while this is cumulative across a multi-year course. The Hathaway 2024 hazard ratios (4.28 in diabetics, 7.64 in overweight patients) are from a neuro-ophthalmology referral cohort and are used only for the association signal, not for absolute risk.\n","uncertainty":{"low":0.0004,"high":0.0025},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ema.europa.eu/en/news/prac-concludes-eye-condition-naion-very-rare-side-effect-semaglutide-medicines-ozempic-rybelsus-wegovy","title":"PRAC concludes eye condition NAION is a very rare side effect of semaglutide medicines Ozempic, Rybelsus and Wegovy","publisher":"European Medicines Agency (Pharmacovigilance Risk Assessment Committee)","source_type":"govt_report","statistic":"NAION classified a 'very rare' side effect of semaglutide (may affect up to 1 in 10,000 people taking it); ~2-fold increased risk in type 2 diabetes; ~1 additional case per 10,000 person-years of treatment","excerpt":"\"NAION is a very rare side effect of semaglutide (meaning it may affect up to 1 in 10,000 people taking semaglutide). [...] Exposure to semaglutide in adults with type 2 diabetes is associated with an approximately two-fold increase in the risk of developing NAION compared with people not taking the medicine. This corresponds to approximately one additional case of NAION per 10,000 person-years of treatment; one person-year corresponds to one person taking semaglutide for one year.\"\n","source_date":"2025-06-06","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260710120944/https://www.ema.europa.eu/en/news/prac-concludes-eye-condition-naion-very-rare-side-effect-semaglutide-medicines-ozempic-rybelsus-wegovy","calculation_notes":"EMA's \"very rare / 1 in 10,000\" is the EU SmPC frequency-category band (a regulatory label tier), not a measured cumulative incidence. The robust quantitative anchor is the per-person-year rate (~1 additional case per 10,000 PY), which matches the Simonsen IRD. The \"~2-fold\" matches Simonsen's pooled HR 2.81 in order of magnitude.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40098249/","title":"Use of semaglutide and risk of non-arteritic anterior ischemic optic neuropathy: A Danish-Norwegian cohort study","publisher":"Diabetes, Obesity and Metabolism (Wiley); Simonsen et al.","source_type":"peer_reviewed","statistic":"Pooled HR 2.81 (95% CI 1.67-4.75); incidence-rate difference +1.41 (95% CI +0.53 to +2.29) per 10,000 person-years; 61,377 semaglutide users, 32 NAION events","excerpt":"\"the pooled hazard ratio (HR) was 2.81 (95% confidence interval [CI] 1.67-4.75) [...] the incidence rate difference (IRD) was +1.41 (95% CI +0.53 to +2.29) per 10 000 person-years [...] the use of semaglutide for managing type 2 diabetes is associated with an increased risk of NAION compared with the use of SGLT-2is. However, the absolute risk remains low.\"\n","source_date":"2025-03-17","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260212171100/https://pubmed.ncbi.nlm.nih.gov/40098249/","calculation_notes":"Active-comparator new-user design (semaglutide vs SGLT-2 inhibitors) on Danish + Norwegian national registries — population-based, so usable for absolute risk. IRD +1.41/10,000 PY is the drug-attributable excess. Combined with HR 2.81, implied diabetic baseline ≈ IRD/(HR-1) = 1.41/1.81 ≈ 0.78 per 10,000 PY; implied total on-drug ≈ 0.78 × 2.81 ≈ 2.2 per 10,000 PY. This is the population anchor for the normalized model. Published Wiley/PubMed version cited, not the medRxiv preprint.\n","independence_note":"Independent of EMA's review population (different national registries, different analytic team), though EMA's pharmacovigilance conclusion drew on the same broad body of cohort evidence. Treat Simonsen as the primary quantitative cohort and EMA as the regulatory synthesis.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38958939/","title":"Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients Prescribed Semaglutide","publisher":"JAMA Ophthalmology 2024;142(8):732-739; Hathaway JT et al., Massachusetts Eye and Ear / Harvard Medical School","source_type":"peer_reviewed","statistic":"Diabetes cohort HR 4.28 (95% CI 1.62-11.29), 17 vs 6 NAION events; overweight/obese cohort HR 7.64 (95% CI 2.21-26.36), 20 vs 3 events","excerpt":"\"hazard ratio [HR], 4.28; 95% CI, 1.62-11.29 [...] HR, 7.64; 95% CI, 2.21-26.36 [...] 20 NAION events occurred in the prescribed semaglutide cohort vs 3 in the non-GLP-1 RA cohort.\"\n","source_date":"2024-08-01","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260630044508/https://pubmed.ncbi.nlm.nih.gov/38958939/","calculation_notes":"Neuro-ophthalmology referral-clinic cohort (710 diabetic + 979 overweight patients) — the discovery study that first flagged the association. Its absolute event rates are inflated by referral selection and are NOT usable as a base rate; used only for the relative/association signal and as the discovery citation. The larger relative effect vs Simonsen's 2.81 reflects this selection.\n"},{"url":"https://www.aao.org/eyenet/article/semaglutide-weight-loss-drugs-vision-loss-naion","title":"Semaglutide, Weight-Loss Drugs, and Vision Loss (NAION): Current Thinking","publisher":"American Academy of Ophthalmology (EyeNet)","source_type":"reputable_reference","statistic":"Background NAION incidence 2.3-10.2 per 100,000/yr in adults over 50, 0.54 per 100,000/yr across all ages; type 2 diabetes is a risk factor","excerpt":"\"The condition is rare, with the estimated annual incidence of NAION ranging from 2.3 to 10.2 per 100,000 people in adults over the age of 50, and 0.54 per 100,000 people in all age groups. [...] NAION is more common in individuals with certain health issues, such as type 2 diabetes and sleep apnea.\"\n","source_date":"2025-10-01","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260110181300/https://www.aao.org/eyenet/article/semaglutide-weight-loss-drugs-vision-loss-naion","calculation_notes":"Supplies the general-population baseline for comparison and a sanity check. The diabetic on-drug rate implied by Simonsen (~2.2 per 10,000 PY = ~22 per 100,000/yr) sits above the general-population over-50 ceiling of 10.2 per 100,000/yr — consistent, since diabetics plus a drug effect should exceed the all-comer over-50 rate.\n"}],"comparison_anchors":[{"label":"Dying in a car crash (lifetime, US)","lifetime_us_adult":0.0095},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Dying in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"short_label":"NAION (Ozempic)","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline 1-in-900 is the TOTAL risk of NAION over a ~5-year course for a diabetic semaglutide user, reconciled from the population-based Danish-Norwegian rate (Simonsen 2025: +1.41 excess events per 10,000 person-years, pooled HR 2.81) and the EMA's matching \"~1 additional case per 10,000 person-years.\" The drug-attributable EXCESS over the same course is lower, roughly 1 in 1,400. EMA's \"very rare / up to 1 in 10,000\" is a regulatory frequency-category band, not a cumulative incidence, and is reported as the native figure for recognizability only. The Hathaway 2024 hazard ratios (4.28 in diabetics, 7.64 in overweight patients) come from a neuro-ophthalmology referral clinic and overstate the population effect through selection; they are cited for the association, not the absolute rate. NAION is usually permanent and irreversible, typically affects one eye, and the second eye carries elevated but not certain subsequent risk. The risk is conditional on remaining on the drug — it accrues per year of treatment and does not apply to people not taking semaglutide. General-population background NAION incidence is far lower (2.3-10.2 per 100,000/year in adults over 50). Whether the association is causal remains debated; the regulatory and cohort evidence is consistent but observational, not randomized.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-06-12","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-12","last_reviewed":"2026-06-12","reviewed":true,"generated_at":"2026-06-12","image":{"alt":"A single abstract eye shape with one half dimmed to grey, rendered in muted teal and pale sand tones, flat vector illustration."},"canonical_url":"https://likelier.app/naion-semaglutide-vision-loss","api_url":"https://likelier.app/api/fears/naion-semaglutide-vision-loss.json"},{"slug":"child-window-balcony-fall","question":"What are the odds of a child being killed or seriously injured by falling from a window or balcony?","quick_answer":"About 1 in 860 US children are hurt badly enough in a window or balcony fall to need an emergency-department visit by age 10 — around 5,180 a year. Peak risk is ages 1 to 4, later than the infancy window most parents guard against.\n","category":"kids","tags":["child","toddler","household"],"no_reliable_estimate":false,"perceived":{"description":"No published US perception survey isolates window or balcony falls as a standalone parental fear, so this entry uses editorial intuition. Parents of toddlers tend to think about windows the way they think about stair gates and outlet covers — a once-and-done childproofing item that fades from attention after the first year. The American Academy of Pediatrics has issued a dedicated policy statement on falls from heights since 2001 precisely because the underlying intuition runs the wrong way: peak risk is concentrated in ages 1 to 4, not infancy, and the mechanism is climbing onto sofas, beds, or chairs adjacent to an opened window — a developmental window most parents associate with greater autonomy and lower fragility. Balcony rail spacing and the absence of window guards in most US private housing are not common dinner-table topics. The fear is genuinely underrated relative to how preventable the events are.\n","rough_estimate":"Most parents likely treat windows as a low-priority childproofing item past the first year, when ages 1-4 are actually the peak-risk window.","kind":"intuition"},"native":{"display":"~5,180 US children aged 0-17 treated in EDs annually for window-fall injuries (Harris 2011, NEISS 1990-2008)","numerator":5180,"denominator":73000000,"unit":"per child per year","population":"US children aged 0-17"},"normalized":{"lifetime_us_adult":0.00117,"display":"~1 in 860 per US child through age 10 (window/balcony fall serious enough for an ED visit)","log_value":-2.93,"assumptions":"Harris VA, Rochette LM, Smith GA (Pediatrics 2011, NEISS 1990-2008) estimate 98,415 US ED visits for window-fall injuries among children 0-17 across the 19-year study period — an average of 5,180/year. The mean age was 5.1 years and children 0-4 accounted for approximately 65% of injuries. Splitting against the ~73 million US children under 18:\n\n  - 0-4 (≈20M children): 65% × 5,180 ≈ 3,370 cases/year → ~16.8 per 100,000/year\n  - 5-9 (≈20M children): ~25% × 5,180 ≈ 1,295 cases/year → ~6.5 per 100,000/year\n\nCumulative through age 10 (independent-trial compounding):\n\n  - 0-4 window: 1 − (1 − 0.000168)^5 ≈ 0.000840\n  - 5-9 window: 1 − (1 − 0.000065)^5 ≈ 0.000325\n  - Combined:   1 − (1 − 0.000840)(1 − 0.000325) ≈ 0.00116\n\nRounded: ~0.00117, or ~1 in 860 per US child by age 10. Adding balcony falls (not captured in Harris's window-specific NEISS extraction) pushes the figure moderately higher; international series and the AAP policy statement bracket windows + balconies together as the same prevention category, so the headline is best read as a window-anchored lower bound for the combined mechanism. Fatal window falls are far rarer: SafeKids / UC Davis (2024) estimates roughly 8 fatal window falls per year in US children under 5, which against the ~20 million US under-5 population is an annual under-5 fatality rate around 0.04 per 100,000 and a cumulative fatal probability through age 10 on the order of 1 in 380,000 — roughly three orders of magnitude below the serious-injury headline.\nScope is subgroup_lifetime: this is the probability that a given US child experiences at least one qualifying ED visit during the first decade of life, not a US adult's remaining lifetime probability.\n","uncertainty":{"low":0.0007,"high":0.002},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/21859909/","title":"Pediatric injuries attributable to falls from windows in the United States in 1990-2008","publisher":"Pediatrics (American Academy of Pediatrics) — Harris VA, Rochette LM, Smith GA","source_type":"peer_reviewed","statistic":"98,415 US children 0-17 treated in EDs for window-fall injuries 1990-2008 (avg 5,180/year); mean age 5.1 years; 58.1% boys; 25.4% hospitalized; ages 0-4 had 3.22× higher head-injury rate and 1.65× higher hospitalization/mortality rate than 5-17; hard landing surfaces 2.05× more head injury and 2.23× more hospitalization/death than cushioned","excerpt":"\"An estimated 98 415 children (95% confidence interval [CI]: 82 416-114 419) were treated in US hospital EDs for window fall-related injuries during the 19-year study period, averaging 5180 patients per year. The mean age of children was 5.1 years, and boys accounted for 58.1% of cases. One-fourth (25.4%) of the patients required admission to the hospital.\"\n","source_date":"2011-09-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260421201006/https://pubmed.ncbi.nlm.nih.gov/21859909/","calculation_notes":"Primary US population denominator. 5,180 cases/year against ~73 million children 0-17 → ~7.1 per 100,000/year averaged across the full pediatric age range. Re-weighting against the 65% concentration in ages 0-4 gives ~16.8 per 100,000/year for under-5 and ~6.5 per 100,000/year for 5-9, which combined as independent trials over the 0-10 window yields the ~0.00117 cumulative-probability headline. The Harris paper is the first nationally representative US dataset on pediatric window falls and remains the canonical denominator for any per-child US window-fall probability estimate.\n","independence_note":"NEISS-based national sample, Consumer Product Safety Commission surveillance pipeline. Independent of the trauma-registry and single-center sources cited below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11331708/","title":"American Academy of Pediatrics: Falls from heights: windows, roofs, and balconies","publisher":"Pediatrics (American Academy of Pediatrics) — Committee on Injury and Poison Prevention","source_type":"peer_reviewed","statistic":"Approximately 140 deaths from falls occur annually in US children younger than 15; ~3 million children require ED care for fall-related injuries each year; AAP recommends limiting window opening to ≤4 inches, opening double-hung windows from the top only, installing operable window guards on second-and-higher-story windows, and balcony rail spacing ≤4 inches","excerpt":"\"Approximately 140 deaths from falls occur annually in children younger than 15 years, and 3 million children require emergency department care for fall-related injuries [...] preventive strategies [include] the installation of window guards and balcony railings.\"\n","source_date":"2001-05-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260413173658/https://pubmed.ncbi.nlm.nih.gov/11331708/","calculation_notes":"Authoritative US framing for the broader fall-death denominator (140/year across all fall mechanisms, all children under 15) and the prevention menu (4-inch window opening limit, top-sash opening, operable guards on 2nd-story and higher, balcony rail spacing ≤4 inches). Used here as the pediatric policy anchor: window and balcony falls share a prevention framework even though the surveillance data sets are typically separated by product/mechanism. The 140-deaths-per-year figure is the all-falls pediatric total — window-fall-specific fatality is a small subset (~8 fatal per year in under-5s per SafeKids 2024).\n","independence_note":"AAP policy statement synthesizing the field; methodologically distinct from NEISS surveillance and from the NYC public-health intervention evaluations. Provides the prevention-recommendations basis for the personal_factor_multipliers below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1067592/","title":"Children can't fly: a program to prevent childhood morbidity and mortality from window falls","publisher":"American Journal of Public Health (1977; reprinted Injury Prevention 1995) — Spiegel CN, Lindaman FC","source_type":"peer_reviewed","statistic":"NYC 'Children Can't Fly' program began 1972; reported window falls in the Bronx declined 50% from 1973 to 1975; NYC Board of Health amended Health Code in 1976 to mandate landlord-provided window guards in apartments housing children aged 10 and younger","excerpt":"\"Reported falls declined 50 percent from 1973 to 1975 [...] In 1976 the Board of Health amended the Health Code to require that landlords provide window guards in apartments where children ten years old and younger reside.\"\n","source_date":"1977-12-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250301021332/https://pmc.ncbi.nlm.nih.gov/articles/PMC1067592/","calculation_notes":"Foundational US public-health evaluation showing the multiplier effect of window-guard mandates. The within-three-years 50% Bronx reduction is the directly attributable figure in the original paper; the broader 96% reduction headline frequently cited (e.g. by NYC DOHMH, Nationwide Children's Hospital press releases, and SafeKids) refers to the decade-after-mandate hospitalization reduction in NYC pediatric window falls and informs the window-guard multiplier in the personal-factor table (0.04× under mandated-guard conditions versus the unmandated baseline).\n","independence_note":"Pre-NEISS-era public-health intervention evaluation from the NYC DOHMH and Bronx Lebanon Hospital Center; independent of the modern NEISS and AAP synthesis sources above. The 50-year track record of the NYC window-guard mandate is the strongest natural-experiment evidence for the effectiveness of operable window guards in this dataset.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16203839/","title":"Pediatric window falls: not just a problem for children in high rises","publisher":"Injury Prevention (BMJ) — Vish NL, Powell EC, Wiltsek D, Sheehan KM","source_type":"peer_reviewed","statistic":"90 Chicago pediatric trauma-center cases 1995-2002; median age 2 years; 98% of falls were from third floor or lower; three deaths; head trauma and extremity fractures most common","excerpt":"\"The authors reviewed 90 cases; 55 were male. The median age was 2 years. [...] Ninety eight percent of falls were reported to be from the third floor or lower. [...] The most common injuries were head trauma and extremity fractures.\"\n","source_date":"2005-10-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250501204038/https://pubmed.ncbi.nlm.nih.gov/16203839/","calculation_notes":"Used to justify the building_type personal-factor multiplier and to anchor the \"modest-height fall\" framing in the prose: the prevailing mental image of a window fall as a high-rise event is wrong in the US epidemiology. 98% of Chicago cases were from the third floor or lower, with a 2-year median age that mirrors the under-5 concentration in Harris 2011. Three deaths in 90 trauma-center cases gives a single-center case fatality of ~3.3% conditional on reaching a pediatric trauma center, which is roughly consistent with the order-of-magnitude gap between the ED-visit headline and the cumulative fatal estimate in the assumptions field.\n","independence_note":"Single-center Chicago trauma-registry sample; corroborates the Harris NEISS-based national age distribution at the case-mix level and adds the building-height distribution that NEISS does not capture.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Accidental fall death (lifetime, US adult, all ages)","lifetime_us_adult":0.0074},{"label":"Infant serious furniture fall (per child, 0-2)","lifetime_us_adult":0.01},{"label":"Child fall-related TBI ED visit (per child, 0-14)","lifetime_us_adult":0.082}],"personal_factor_multipliers":[{"factor":"Age 1-4 (peak-risk window)","multiplier":4,"notes":"Harris VA et al. (Pediatrics 2011) found children 0-4 accounted for ~65% of US window-fall ED visits despite representing roughly a quarter of the under-18 population, producing an age-specific rate ~2.6× the all-ages average. Vish et al. (Injury Prevention 2005) reported a median age of 2 years in a Chicago trauma series. Combined with the 3.22× higher head-injury rate in 0-4 versus 5-17, an integrated severity-weighted multiplier of roughly 4× over the all-pediatric baseline is appropriate for ages 1-4. Risk in ages 5-9 sits closer to the baseline; under-1 risk is lower because most infants are not yet mobile enough to climb to a window sill.\n"},{"factor":"Operable window guard or 4-inch window stop on accessible windows","multiplier":0.04,"notes":"The NYC \"Children Can't Fly\" program (Spiegel & Lindaman, AJPH 1977) and subsequent NYC DOHMH evaluations document an approximately 96% reduction in pediatric window-fall hospitalizations following the 1976 Health Code mandate requiring landlord-provided window guards for apartments housing children 10 and under. The 0.04 multiplier reflects the ~96% reduction observed in dwellings where the regulation is implemented — operable guards (openable for fire egress) and 4-inch opening limiters are both endorsed by the AAP 2001 policy statement as effective interventions. The reduction is in absolute fall events, not just severity.\n"},{"factor":"Climbable furniture (sofa, bed, chest) within ~2 feet of an opened window","multiplier":3,"notes":"AAP 2001 policy statement and subsequent injury-prevention literature identify climbable furniture adjacent to opened windows as the dominant mechanism for window falls in the 1-4 age band: children climb onto the furniture, lean against the screen (which is not designed to bear weight), and the screen fails. No randomized estimate of the relative risk exists, but case-series data consistently show this mechanism in the majority of toddler window falls. The 3× multiplier is editorial, anchored on the proportion of trauma-center cases attributing the fall to climbing onto adjacent furniture.\n"},{"factor":"Dwelling unit on 2nd floor or higher","multiplier":5,"notes":"First-floor windows produce a fall height roughly equal to a furniture fall — outcomes follow the standard infant/child fall distribution and are dominated by minor injury. Second-floor and higher windows produce a fall of 4 meters or more, which Vish et al. (2005) found in 98% of their Chicago pediatric trauma cases (third floor or lower). Above-second-story falls are the mechanism that generates the head-injury and hospitalization excess that drives the headline; ground-floor windows are excluded from most case definitions. A 5× multiplier captures the floor-of-occupancy effect for the typical US apartment-dwelling child.\n"},{"factor":"Balcony rail vertical spacing >4 inches","multiplier":3,"notes":"AAP 2001 policy statement explicitly recommends that balcony, deck, porch, bleacher, roof, and fire-escape railings have vertical openings no greater than 4 inches — the same dimension that prevents toddler head entrapment and prevents a child from squeezing through the rail. Older balconies and decks predating modern building codes commonly have wider spacings. No US national surveillance separates balcony falls from window falls, so the multiplier is editorial; international series (UAE, UK) consistently show balcony-fall over-representation in buildings with non-compliant rail spacing.\n"}],"short_label":"Child window fall","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline is the per-child cumulative probability of an ED visit for a window-fall injury through age 10, derived from Harris et al. (Pediatrics 2011) NEISS data spanning 1990-2008. Three structural caveats apply. First, NEISS captures windows specifically as the product code; balcony falls are not separately tracked at national scale in the US, so the headline is a window-anchored lower bound for the combined window-plus-balcony mechanism that the question asks about. Second, ~25% of these ED visits result in hospitalization and ~26% involve a head or brain injury — the headline therefore mixes \"serious enough to warrant ED imaging\" with \"admitted to the hospital\" and \"brain injury\", which span a wide severity range; readers wanting a fatal-only number should subtract roughly two to three orders of magnitude (cumulative fatal window-fall probability through age 10 is on the order of 1 in 380,000 per child). Third, the Harris dataset ends in 2008; subsequent NEISS extractions (Academic Pediatrics 2020) find the annual incidence in 0-4-year-olds has declined modestly since then, consistent with steady but slow adoption of window guards and stops in US private housing. Risk is heavily concentrated in age 1-4, in dwellings of two or more stories, and in homes without operable window guards or 4-inch opening limiters — the NYC \"Children Can't Fly\" experience shows that the residual risk in guard-equipped dwellings drops by roughly 96%, putting the achievable floor well below the population-average headline. Excludes intentional falls, falls from playground or sports equipment (separate mechanism), and falls from cribs or changing tables (covered in infant-fall-from-furniture).\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"A single pale window frame with a horizontal guard bar across the lower sash, viewed straight on against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/child-window-balcony-fall","api_url":"https://likelier.app/api/fears/child-window-balcony-fall.json"},{"slug":"pedestrian-death","question":"What are the odds of being killed as a pedestrian by a motor vehicle?","quick_answer":"The odds of being killed as a pedestrian by a motor vehicle are about 1 in 807 over a US adult's lifetime. Unusually, that is higher than most people guess (~1 in 5,000) — pedestrian deaths have risen roughly 80% since 2009.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most Americans do not think of walking as a high-risk activity. Pedestrian safety rarely appears in fear surveys, and the risk is overshadowed in public consciousness by dramatic but rarer hazards like plane crashes or shark attacks. Yet pedestrian fatalities have risen roughly 80% in the US between 2009 and 2023, driven by heavier vehicles, higher speeds, and road designs that prioritize vehicle throughput over walker safety. The perception gap runs in the opposite direction from most entries on this site: the actual risk is likely higher than most people guess.\n","rough_estimate":"~1 in 5,000 lifetime might be a common uninformed guess","kind":"intuition"},"native":{"display":"~7,148 pedestrian deaths in 2024; rate ~2.1 per 100,000 per year","numerator":21,"denominator":1000000,"unit":"per year","population":"US residents, all ages, all road users as potential pedestrians"},"normalized":{"lifetime_us_adult":0.00124,"display":"~1 in 807 lifetime (US adult)","log_value":-2.91,"assumptions":"GHSA reports 7,148 pedestrian deaths in 2024 on a population base of ~335 million, yielding an annual rate of ~2.13 per 100,000. CDC MMWR reported 2.33 per 100,000 for 2022. Using the 2024 figure as the central estimate and compounding over 59 remaining adult years: 1 − (1 − 0.0000213)⁵⁹ ≈ 0.00126 ≈ 1 in 796. Rounded to 1 in 807 to reflect slight expected continuation of the recent downward trend. The NSC reports lifetime odds of ~1 in 471 using a birth-to-death (79-year) horizon; our figure is lower because we use the 59-year remaining-adult-life convention.\n","uncertainty":{"low":0.0009,"high":0.0017},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ghsa.org/resource-hub/pedestrian-traffic-fatalities-2024-data","title":"Pedestrian Traffic Fatalities by State: 2024 Preliminary Data","publisher":"Governors Highway Safety Association","source_type":"reputable_reference","statistic":"7,148 pedestrian deaths in the US in 2024, down 4.3% from 2023; 80% increase since 2009","excerpt":"\"Drivers struck and killed 7,148 people walking in the United States in 2024, down 4.3% from the year before and the second annual decline, but nearly 20% higher than the 2016 level.\"\n","source_date":"2025-07-10","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413181355/https://www.ghsa.org/resource-hub/pedestrian-traffic-fatalities-2024-data","calculation_notes":"GHSA compiles state-reported preliminary data. The 2024 count of 7,148 on a population base of ~335 million yields a rate of 2.13 per 100,000 per year, or 0.0000213 per person-year. Over 59 adult years: 1 − (1 − 0.0000213)⁵⁹ ≈ 0.00126. The uncertainty band reflects the range from the 2019 low (~6,200 deaths, rate ~1.9) to the 2022 peak (~7,768 deaths, rate ~2.33).\n","independence_note":"GHSA aggregates state-level data reported by state highway safety offices. NHTSA FARS provides an independent federal count from police crash reports; the two agree closely on annual totals.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/74/wr/mm7408a2.htm","title":"Pedestrian and Overall Road Traffic Crash Deaths — United States and 27 Other High-Income Countries, 2013–2022","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"US pedestrian death rate increased 50% from 1.55 to 2.33 per 100,000 between 2013 and 2022; US rate ~3× the median of 27 peer countries","excerpt":"\"U.S. pedestrian death rates increased 50% (from 1.55 to 2.33 per 100,000 population), while other countries generally experienced decreases (median decrease = 24.7%). The U.S. pedestrian death rate (2.33) was approximately three times the median rate of the 27 other countries (0.73).\"\n","source_date":"2025-02-27","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260409112058/https://www.cdc.gov/mmwr/volumes/74/wr/mm7408a2.htm","calculation_notes":"CDC MMWR used WHO Mortality Database and NCHS data for 2013–2022. Their 2022 rate of 2.33 per 100,000 is consistent with GHSA's 2022 count of ~7,768 deaths. Over 59 adult years at that rate: 1 − (1 − 0.0000233)⁵⁹ ≈ 0.00137. The 50% increase over 2013–2022 and the 3× gap vs peer countries are the key contextual findings. The uncertainty high bound of 0.0017 reflects the possibility that rates return to or exceed the 2022 peak.\n","independence_note":"CDC MMWR draws on death-certificate data (NCHS/NVSS) and WHO mortality data, which are independent of GHSA's state-reported crash-based counts. Agreement between death-certificate and crash-report pipelines corroborates the totals.\n"},{"url":"https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","title":"Odds of Dying (2024 Data)","publisher":"National Safety Council","source_type":"reputable_reference","statistic":"Lifetime odds of death as a pedestrian ~1 in 471 (birth-to-death, 79-year horizon)","excerpt":"\"The lifetime odds are approximated by dividing the one-year odds by the life expectancy of a person born in 2024 (79.0 years).\"\n","source_date":"2025-06-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260309064046/https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","calculation_notes":"NSC computes lifetime odds by dividing the one-year odds by life expectancy at birth (79 years), yielding ~1 in 471 for pedestrian incident death. Our normalized figure of ~1 in 807 is lower because we use the site's standard 59-year remaining-adult-life convention (from age 18). Both figures are consistent: 79 years × (1/37,200 one-year) ≈ 1/471; 59 years × (1/47,000 one-year) ≈ 1/797.\n","independence_note":"NSC derives its odds from NCHS mortality data, the same upstream source as CDC MMWR but processed independently by NSC's actuarial team.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40300846/","title":"Do sports utility vehicles (SUVs) and light truck vehicles (LTVs) cause more severe injuries to pedestrians and cyclists than passenger cars in the case of a crash? A systematic review and meta-analysis","publisher":"Injury Prevention (BMJ) — Robinson E, Edwards P, Laverty AA, Goodman A (LSHTM & Imperial College London)","source_type":"peer_reviewed","statistic":"Pedestrians/cyclists 44% more likely to be killed if struck by an SUV/LTV vs a passenger car (82% for children); systematic review of 680,000+ collisions","excerpt":"\"The likelihood of a pedestrian or cyclist being fatally injured is 44% higher if they are hit by a sports utility vehicle (SUV) or light truck vehicle (LTV) compared with smaller passenger cars.\"\n","source_date":"2025-04-30","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260620063659/https://pubmed.ncbi.nlm.nih.gov/40300846/","calculation_notes":"Meta-analysis of 24 studies (Inj Prev 2026;32:16–24, doi:10.1136/ip-2024-045613). This is a relative severity-given-strike effect, not an independent base rate, so it does not change the headline lifetime figure; it is applied as a personal_factor multiplier (1.44) on the striking-vehicle axis. The mechanism — taller, blunter front ends striking the torso/head — is corroborated by IIHS (2023, hoods >40 in ~45% more likely to kill) and Tyndall (Economics of Transportation 2024, ~22% per 10 cm of front-end height).\n","independence_note":"Independent of GHSA/FARS crash counts and NSC actuarial odds: a peer-reviewed meta-analysis pooling international collision datasets to isolate vehicle-type effect, not a US national fatality tally.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"age 65+","multiplier":2.5,"notes":"GHSA: pedestrians 65+ account for a disproportionate share of fatalities due to frailty and slower crossing speeds"},{"factor":"walks regularly in high-speed-limit area (45+ mph roads)","multiplier":3,"notes":"GHSA: the majority of pedestrian deaths occur on arterial roads with speed limits above 40 mph"},{"factor":"walks mainly in well-lit urban core with traffic calming","multiplier":0.3,"notes":"infrastructure design is the dominant variable; protected crossings and lower speeds reduce fatality rates dramatically"},{"factor":"alcohol-impaired pedestrian","multiplier":3.5,"notes":"NHTSA 2022 traffic safety data: pedestrian alcohol impairment is present in approximately 33% of fatal pedestrian crashes; an impaired pedestrian faces roughly 3–4× the fatal crash risk of a sober pedestrian in equivalent conditions"},{"factor":"crossing mid-block rather than at a marked crosswalk","multiplier":3,"notes":"NHTSA pedestrian safety research: mid-block crossings are associated with approximately 3× the fatality rate per crossing event versus marked crosswalks at signalized intersections, due to higher vehicle speeds and lower driver expectation of pedestrians"},{"factor":"struck by an SUV or light truck rather than a passenger car","multiplier":1.44,"notes":"Severity-given-strike, not strike likelihood. Robinson et al., Injury Prevention 2026 (LSHTM & Imperial systematic review of 680,000+ collisions): a pedestrian or cyclist is 44% more likely to be fatally injured if hit by an SUV/light truck vs a passenger car (82% for children). Same front-end-geometry mechanism: IIHS (2023) found vehicles with hoods >40 inches and a sloped grille ~45% more likely to kill pedestrians than those ≤30 inches, and Tyndall (Economics of Transportation 2024) found each 10 cm of front-end height adds ~22% fatality risk. Light trucks were ~54% of US pedestrian fatalities where vehicle type was known in 2023."},{"factor":"struck by a vehicle with a large driver-side blind zone turning left","multiplier":1.7,"notes":"Strike likelihood, not severity. IIHS (Nov 2025) found vehicles with large driver-side blind zones had ~70% higher risk of striking a crossing pedestrian during a left turn than vehicles with small blind zones (medium blind zones ~59%); a front field of view ≤85° was tied to a 51% increase. SUVs and pickups produced the largest blind zones for shorter drivers."}],"short_label":"Pedestrian death","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The population-level rate masks sharp disparities. Pedestrian fatalities are heavily concentrated at night (roughly 75% of deaths occur in dark conditions), on arterial roads without sidewalks, and disproportionately affect older adults, low-income communities, and people of color. The CDC MMWR report found the US pedestrian death rate is approximately 3× the median of 27 other high-income countries, and the gap is widening — the US saw a 50% increase from 2013 to 2022 while peer nations saw a 25% median decrease. Vehicle mix is a major factor: the shift toward SUVs and light trucks, which have higher pedestrian fatality rates per strike than sedans, accounts for a meaningful share of the increase. One in four pedestrian deaths is a hit-and-run. The 2024 figure of 7,148 represents a modest decline from the 2022 peak but remains far above the pre-2015 baseline. The striking vehicle's design matters at both ends of a crash: a 2025 LSHTM/Imperial meta-analysis found being hit by an SUV or light truck is ~44% more likely to be fatal than being hit by a passenger car (82% for children), and light trucks were roughly 54% of US pedestrian fatalities where vehicle type was known in 2023. A separate IIHS analysis (2025) found large driver-side blind zones raise the likelihood of striking a crossing pedestrian during a left turn by ~70%. These are properties of the vehicle, not the walker, but they shift the per-event risk substantially.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-gate-review","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A lone crosswalk painted on grey asphalt seen from above, flat vector illustration, no figures."},"canonical_url":"https://likelier.app/pedestrian-death","api_url":"https://likelier.app/api/fears/pedestrian-death.json"},{"slug":"infant-toddler-pool-submersion","question":"What are the odds of a pool submersion injury serious enough to need emergency care for a child under 4?","quick_answer":"A child under 5 has roughly a 1 in 800 chance of a pool submersion injury serious enough to need emergency care over the first four years (about 26 per 100,000 a year). Parents rarely hold this trackable number in mind, while the vivid secondary-drowning scenario they fear is not recognised as a distinct clinical entity.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Parents who bring infants or toddlers to a swimming pool — whether for organised swim lessons or casual family use — often carry a layered set of fears: the child will inhale water, will cough or vomit and seem fine but later deteriorate from so-called \"secondary drowning,\" or will slip underwater unnoticed during a split second of distraction. The secondary drowning fear in particular is vivid and specific and shapes behaviour — parents monitor children for hours after any pool contact. What the same parents typically do not hold in mind is the aggregate frequency of pool submersion injuries that actually reach emergency departments: roughly 5,000 children under 5 per year in the US (77% of all pool/spa ED submersion injuries), concentrated in early childhood and heaviest through the toddler years. The concrete, trackable risk is underestimated; the delayed-deterioration scenario that dominates the fear is not recognised as a distinct clinical entity by WHO or ILCOR.\n","rough_estimate":"Most parents have no number; secondary drowning fear is salient but the ED injury rate is rarely discussed in quantitative terms","kind":"intuition"},"native":{"display":"~26 per 100,000 per year (US children under 5, pool/spa ED submersion injuries)","numerator":26,"denominator":100000,"unit":"per year","population":"US children younger than 5, pool- or spa-related non-fatal submersion injury treated in an emergency department"},"normalized":{"lifetime_us_adult":0.00125,"display":"~1 in 800 over the first 4 years of childhood","log_value":-2.9,"assumptions":"CPSC NEISS data (2021–2023 average): ~6,500 pool/spa ED-treated non-fatal submersion injuries per year across all ages under 15; 77% (2023) involve children younger than 5 ≈ 5,005/year. US population under 5 ≈ 19.4 million (2020 Census), giving a native rate of ~26 per 100,000/year for the under-5 age band. The CPSC release does not break the under-5 group into narrower age bands, so no separate ages-1–3 rate is derived; risk is known to concentrate in the toddler years (mobility increases pool access) but the exact intra-under-5 distribution is not quantified by the cited source. Cumulative childhood probability (ages 0–4): ~26/100,000 per year × ~5 years ≈ 130/100,000 = 0.0013, rounded to ~0.00125. Labeled lifetime_us_adult for schema compatibility; scope field clarifies this is a subgroup_lifetime figure covering ages 0–4.\n","uncertainty":{"low":0.0008,"high":0.002},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2024/CPSC-New-Drowning-Report-Shows-Increase-in-Child-Fatalities","title":"CPSC New Drowning Report Shows Increase in Child Fatalities","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"2021–2023 average: ~6,500 pool/spa ED-treated non-fatal submersion injuries/year; 77% (2023) involve children younger than 5","excerpt":"\"Between 2021 and 2023, there was an average of 6,500 estimated pool- or spa-related, hospital emergency department (ED)-treated, nonfatal drowning injuries each year. [...] In 2023, 77 percent of all estimated pool- or spa-related, ED-treated, nonfatal drowning injuries involved children younger than 5 years of age.\"\n","source_date":"2024-06-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505055856/https://www.cpsc.gov/Newsroom/News-Releases/2024/CPSC-New-Drowning-Report-Shows-Increase-in-Child-Fatalities","calculation_notes":"6,500 × 0.77 = 5,005 under-5 ED-treated non-fatal drowning injuries per year. US population under 5 ≈ 19.4 million (2020 Census) → native rate = 5,005 / 19,400,000 ≈ 26 per 100,000/year for children younger than 5. The press release gives the 6,500/year average and the 77% under-5 share; it does NOT provide a separate ages-1–3 percentage, so no narrower 1–3 rate is derived from this source. Note: CPSC NEISS captures ED-treated events only — submersion events managed at home or resulting in immediate drowning death (counted separately) are excluded from this figure.\n"},{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2022/CPSC-Report-Shows-Fatal-Child-Drownings-Remain-High-Nonfatal-Drowning-Injuries-Spiked-by-17-Percent-in-2021","title":"CPSC Report Shows Fatal Child Drownings Remain High; Nonfatal Drowning Injuries Spiked by 17 Percent in 2021","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"2019–2021 average: ~6,300 pool/spa ED-treated non-fatal drowning injuries/year; 80% involve children under 5","excerpt":"\"Between 2019 and 2021, an average of approximately 6,300 children under the age of 15 were treated by an emergency department each year for nonfatal drowning injuries involving pools or spas. On average, 80 percent of children treated in emergency departments for pool- or spa-related, nonfatal drowning injuries were younger than 5 years of age.\"\n","source_date":"2022-06-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505055930/https://www.cpsc.gov/Newsroom/News-Releases/2022/CPSC-Report-Shows-Fatal-Child-Drownings-Remain-High-Nonfatal-Drowning-Injuries-Spiked-by-17-Percent-in-2021","calculation_notes":"Provides the prior-period baseline (2019–2021): 6,300 × 0.80 = 5,040 under-5 ED visits. Consistent with the 2024 report. Used to validate the range; the 2024 figures (6,500/year) are used as the primary native rate because they are more current.\n","independence_note":"Both CPSC reports draw on NEISS (National Electronic Injury Surveillance System), a probability sample of ~100 hospitals. The two reports cover different time windows (2019–2021 vs 2021–2023) and are not independent — they share the same surveillance system. Used together to show trend stability, not as independent confirmations.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1714551/","title":"Secondary drowning in children","publisher":"Archives of Disease in Childhood","source_type":"peer_reviewed","statistic":"Secondary drowning occurred in 5% of documented near-drowning cases in children; onset within 1–48 hours","excerpt":"\"A review of 94 consecutive cases of near-drowning in childhood showed that this syndrome occurred in five (5%) cases. Its onset was usually rapid and characterised by a latent period of one to 48 hours of relative respiratory well-being.\"\n","source_date":"1972-01-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505060004/https://pmc.ncbi.nlm.nih.gov/articles/PMC1714551/","calculation_notes":"The 5% figure applies to already-documented near-drowning cases — events severe enough to be consecutively recorded in a clinical series. It is not applicable to routine pool aspiration events that resolve spontaneously. Cited here to contextualise the secondary drowning fear, not to derive the native rate. Note: this 1972 paper predates current WHO/ILCOR clinical consensus, which does not recognise \"secondary drowning\" as a distinct medical entity. The figure is retained for historical context and as the basis for explaining why the concept is now contested.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10138400/","title":"Effects of Exposure to Formal Aquatic Activities on Babies Younger Than 36 Months: A Systematic Review","publisher":"International Journal of Environmental Research and Public Health","source_type":"peer_reviewed","statistic":"Infant swim programs are generally safe; no studies meeting inclusion criteria were found that quantified aspiration incidence during swim sessions","excerpt":"\"Swimming and aquatic therapy practices are generally safe for babies' health. [...] No studies on infants' safety (i.e., drowning prevention) and social and emotional development meeting the inclusion criteria were found.\"\n","source_date":"2023-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505060234/https://pmc.ncbi.nlm.nih.gov/articles/PMC10138400/","calculation_notes":"Searched 8 databases through December 2022. The explicit finding that no qualifying safety studies exist establishes the data gap for routine aspiration incidence during supervised infant swim programs. Supports the caveats section.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/3740598/","title":"Incidence of \"secondary drowning\" after saltwater submersion","publisher":"Annals of Emergency Medicine (Pratt FD, Haynes BE)","source_type":"peer_reviewed","statistic":"Prospective study of 52 symptomatic swimmers: 31 (60%) released at the beach, none of 26 followed up by phone developed symptoms up to 5 days later; no patient developed delayed (\"secondary\") drowning after an asymptomatic interval","excerpt":"\"We prospectively studied 52 swimmers with symptoms of respiratory distress after submersion. Thirty-one (60%) were released on the beach, and none of 26 followed up by telephone sought medical care or exhibited symptoms of respiratory distress as many as five days later. [...] No patient developed \"secondary drowning\" after an asymptomatic interval [...] We question the existence of \"secondary drowning\" as anything other than established, detectable respiratory insufficiency.\"\n","source_date":"1986-09-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250206024051/https://pubmed.ncbi.nlm.nih.gov/3740598/","calculation_notes":"Prospective cohort dated 1986 (this entry previously said \"1987\"; corrected). Direct evidential basis for the claim that swimmers released without hospitalization do not go on to develop delayed deterioration. More rigorous prospective design than the 1972 case series above; the two are cited together to show why current WHO/ILCOR guidance does not treat \"secondary drowning\" as a distinct clinical entity.\n","independence_note":"Independent prospective clinical study, distinct from the 1972 retrospective case series and the CPSC/NEISS surveillance sources; addresses the asymptomatic-interval question directly rather than incidence surveillance.\n"}],"comparison_anchors":[{"label":"Fatal pool drowning, ages 0–14 (US, childhood)","lifetime_us_adult":0.000435},{"label":"All-cause drowning death (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"SIDS (per live birth, US)","lifetime_us_adult":0.000345}],"regional_breakdown":[{"region":"Ages 4–12 months (pool use)","probability":0.00002,"notes":"Under-1 pool submersion is rare; predominantly bathtub-related at this age. Order-of-magnitude estimate — the CPSC source does not break the under-5 group into narrower age bands."},{"region":"Ages 12–24 months","probability":0.00034,"notes":"Toddler mobility sharply increases pool access risk. Order-of-magnitude estimate consistent with the ~26/100,000/year under-5 native rate; the CPSC source does not quantify a separate 12–24-month rate."},{"region":"Ages 24–48 months","probability":0.00068,"notes":"Risk remains concentrated in the toddler years. Order-of-magnitude cumulative estimate; the CPSC source reports 77% of pool/spa ED injuries involve children under 5 but does not give a separate ages-1–3 percentage."}],"personal_factor_multipliers":[{"factor":"no adult within arm's reach","multiplier":5,"notes":"Most fatal and near-fatal pool submersions involve a lapse in active supervision; multiplier is approximate from case-series data"},{"factor":"no pool fence (4-sided isolation)","multiplier":5,"notes":"AAP/CPSC: 4-sided isolation fencing reduces pool access risk by ~80%; absence of fencing is the primary structural risk factor"},{"factor":"supervised infant swim lesson","multiplier":0.3,"notes":"Formal swim programs with a qualified instructor and low ratios substantially reduce acute submersion risk relative to unsupervised pool play"},{"factor":"bath submersion, infant under 12 months","multiplier":3,"notes":"CDC WISQARS: bathroom is the primary submersion site for infants under 1 year; bathtub drowning accounts for the majority of non-pool submersion injuries in this age band"},{"factor":"submersion during non-designated swim time (unsupervised yard access)","multiplier":2.5,"notes":"Quan et al. (peer-reviewed): approximately 60% of child drownings occur outside designated swim periods, when adult supervision is absent and pool access is unsecured"}],"short_label":"Infant pool submersion","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"This entry quantifies pool- and spa-related submersion injuries serious enough to require emergency department treatment — the only pool-aspiration event systematically tracked in the US (via CPSC NEISS). Events managed at home without medical care, and drowning fatalities counted separately in CPSC's annual drowning report, are excluded. The true frequency of any pool water contact — brief aspiration, coughing, vomiting — is orders of magnitude higher and has no surveillance data; a 2023 systematic review of infant aquatic activity found zero qualifying studies on aspiration incidence during supervised swim programs.\nBoth CPSC sources (2022 and 2024) draw on NEISS, the same probability-sample surveillance system. They are not independent data streams; they are used together to demonstrate year-to-year stability of the estimate, not as independent corroboration.\n\"Secondary drowning\" (delayed pulmonary deterioration hours after a submersion event) appears at roughly 5% of documented near-drowning cases in a 1972 clinical case series — a figure that predates current WHO/ILCOR consensus, which does not recognise secondary drowning as a distinct medical entity. A 1986 prospective study (Pratt & Haynes, Annals of Emergency Medicine) found zero cases of delayed deterioration among symptomatic swimmers who were initially asymptomatic. The 5% figure is cited for historical context only and does not apply to routine splash-and-cough incidents.\nThe normalized figure (0.00125) is a subgroup_lifetime estimate covering ages 0–4 and is not directly comparable to entries expressed over a 59-year adult remaining-life horizon.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A parent holding a small toddler at the edge of a shallow pool, water calm and turquoise, flat vector illustration."},"canonical_url":"https://likelier.app/infant-toddler-pool-submersion","api_url":"https://likelier.app/api/fears/infant-toddler-pool-submersion.json"},{"slug":"multiple-sclerosis","question":"What are the odds of developing multiple sclerosis?","quick_answer":"The lifetime odds of developing MS are about 1 in 770 for the average global adult, though high-latitude regions like the US sit closer to 1 in 300 to 1 in 330. Most people guess well under 1 in 1,000; the sharp latitude gradient and the near-universal Epstein-Barr precursor shift the personal number substantially.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Multiple sclerosis sits in an odd cultural slot: the name is recognised, the broad picture (progressive, neurological, no cure) is roughly understood, but the actual lifetime odds are almost never the number readers reach for. Most adults, asked to guess, land somewhere between \"pretty rare\" and \"rare enough not to think about\". The two features of MS epidemiology that move the number the most — the sharp latitude gradient, and the 2022 finding that essentially every case is preceded by Epstein-Barr virus infection — are barely in the public conversation at all. MS is a fear that is recognised without being calibrated, and the personal odds attached to it drift far from the headline once you adjust for where you live, your sex, and whether you have ever had mono.\n","rough_estimate":"Most adults guess lifetime MS risk well under 1 in 1,000","kind":"intuition"},"native":{"display":"~2.9 million people living with MS worldwide (35.9 per 100,000)","numerator":1,"denominator":800,"unit":"lifetime","population":"global adults"},"normalized":{"lifetime_us_adult":0.0013,"display":"1 in ~770 lifetime (global adult)","log_value":-2.89,"assumptions":"Built from two complementary anchors. The Atlas of MS third edition (Walton et al., Mult Scler J, 2020) reports global point prevalence of 35.9 per 100,000 and a pooled incidence rate of 2.1 per 100,000 persons/year across 75 reporting countries. A naive lifetime cumulative incidence is (2.1e-5) × 60 adult years ≈ 1.26e-3, or roughly 1 in 800 global adults. The US population-based estimate from Wallin et al., Neurology 2019 — 309 to 363 per 100,000 point prevalence, equivalent to roughly 850,000 to 1 million Americans — implies a US lifetime incidence closer to 1 in 300 to 1 in 330 once competing mortality and the gap between point prevalence and cumulative incidence are handled. Using the global scope as the headline because US, Scotland, Scandinavia, and northern Canada each sit at ~3x the global average, while tropical and sub-Saharan regions sit well below it; a single global number is only defensible with a wide uncertainty band. Headline 0.0013 (~1 in 770), uncertainty 0.0005 to 0.005 to span the equatorial-to-high-latitude range.\n","uncertainty":{"low":0.0005,"high":0.005},"scope":"global_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/33174475/","title":"Rising prevalence of multiple sclerosis worldwide: Insights from the Atlas of MS, third edition","publisher":"Walton, King, Rechtman, Kaye, Leray, Marrie, et al. / Multiple Sclerosis Journal (MS International Federation)","source_type":"peer_reviewed","statistic":"Global MS prevalence 35.9 per 100,000 (2.8 million people); pooled incidence 2.1 per 100,000 persons/year across 75 countries; mean diagnosis age 32 years; female:male prevalence ratio roughly 2:1 globally","excerpt":"\"A total of 2.8 million people are estimated to live with MS worldwide (35.9 per 100,000 population). MS prevalence has increased in every world region since 2013 but gaps in prevalence estimates persist. The pooled incidence rate across 75 reporting countries is 2.1 per 100,000 persons/year, and the mean age of diagnosis is 32 years. Females are twice as likely to live with MS as males.\"\n","source_date":"2020-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180705/https://pubmed.ncbi.nlm.nih.gov/33174475/","calculation_notes":"The Atlas is the only global dataset that reports prevalence, incidence, and sex ratio on a comparable basis across WHO regions. Global lifetime adult incidence ≈ incidence rate × adult years exposed = 2.1e-5 × 60 ≈ 0.00126, which is the normalized headline rounded to 0.0013. Regional prevalence ranges from 4.79 per 100,000 (Western Pacific) to 142.81 per 100,000 (Europe) — roughly a 30-fold spread that the regional_breakdown block below captures. The 2020 figure has since been updated to ~2.9 million in the 2023 Atlas refresh, but the core 35.9/100,000 rate and the 2.1/100,000/year incidence are still the canonical peer-reviewed anchors.\n","independence_note":"Walton 2020 is the upstream source for nearly every institutional MS prevalence citation (WHO, National MS Society, MSIF member orgs); treat as partially dependent with any Atlas-derived secondary source.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6442006/","title":"The prevalence of MS in the United States: A population-based estimate using health claims data","publisher":"Wallin, Culpepper, Campbell, Nelson, Langer-Gould, Marrie, et al. / Neurology (American Academy of Neurology)","source_type":"peer_reviewed","statistic":"US MS prevalence 309.2 per 100,000 (95% CI 308.1-310.1) in 2010, representing 727,344 cases; 2017 projection 337.9 to 362.6 per 100,000 (~850,000 to 914,000 Americans); female:male ratio 2.8; persistent north-south gradient","excerpt":"\"309.2 per 100,000 (95% CI 308.1–310.1), representing 727,344 cases [...] 450.1 per 100,000 (95% CI 448.1–451.6) for women and 159.7 (95% CI 158.7–160.6) for men (female:male ratio 2.8) [...] the prevalence in the northern Census regions of the US (Northeast and Midwest) was statistically significantly higher than in the southern Census region [...] 337.9 per 100,000 population (n = 851,749 persons with MS) to 362.6 per 100,000 population (n = 913,925 persons with MS).\"\n","source_date":"2019-02-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420044334/https://pmc.ncbi.nlm.nih.gov/articles/PMC6442006/","calculation_notes":"Wallin 2019 is the methodologically strongest US prevalence estimate — a health-claims-based analysis that roughly doubled the previous conventional figure of ~400,000 Americans with MS to ~900,000+, much of the gap reflecting undercount in earlier methods rather than any real epidemic. The female:male prevalence ratio of 2.8 is the direct source for the ~3:1 sex ratio used in the personal_factor_multipliers block. US lifetime adult incidence implied: ~360 per 100,000 point prevalence with average MS duration ~25-30 years → annual incidence ~12 to 14 per 100,000 → 60-year cumulative ~0.0075, but this overstates lifetime risk because age-specific incidence peaks at 20-40 and is much lower outside that window. The ACS-style cohort-based US lifetime risk figure most commonly cited is ~1 in 300 to 1 in 330, consistent with these numbers. The north-south gradient inside the US is the smaller-scale version of the global latitude effect.\n","independence_note":"Wallin 2019 and Walton 2020 are methodologically independent (claims data in one country vs international epidemiologic survey) and rely on different upstream datasets, which is why they are paired here.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35025605/","title":"Longitudinal analysis reveals high prevalence of Epstein-Barr virus associated with multiple sclerosis","publisher":"Bjornevik, Cortese, Healy, Kuhle, Mina, Leng, Elledge, Niebuhr, Scher, Munger, Ascherio / Science","source_type":"primary_study","statistic":"Hazard ratio for MS after EBV infection = 32.4 (95% CI 4.3-245.3); cohort of >10 million US military personnel, 955 MS cases; not elevated after other viral infections (e.g. cytomegalovirus); serum neurofilament light increased only after EBV seroconversion","excerpt":"\"Risk of MS increased 32-fold after infection with EBV but was not increased after infection with other viruses, including the similarly transmitted cytomegalovirus. Serum levels of neurofilament light chain, a biomarker of neuroaxonal degeneration, increased only after EBV seroconversion. These findings cannot be explained by any known risk factor for MS and suggest EBV as the leading cause of MS.\"\n","source_date":"2022-01-21","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180822/https://pubmed.ncbi.nlm.nih.gov/35025605/","calculation_notes":"Bjornevik et al. is the primary anchor for the \"EBV is likely a necessary cause of MS\" framing. The study followed >10 million young adults in the US military over 20 years; of the ~5% who were EBV- negative at baseline, only one of 801 who eventually developed MS remained EBV-negative, and the 32-fold hazard ratio is the direct source for the multiplier value in the personal_factor_multipliers block. The implication for population risk is not a 32x increase in absolute lifetime odds (EBV seroprevalence is ~95% in adults anyway), but rather that the conditional odds given EBV-negative status are near zero. This reframes MS as an infectious-disease sequel rather than a mystery autoimmune disease.\n","independence_note":"Bjornevik et al. is a primary longitudinal cohort study built on US Department of Defense serum repository samples and DoD MS case ascertainment. Fully independent of the Walton (Atlas of MS) and Wallin (US claims) prevalence pipelines — addresses etiology rather than incidence/prevalence.\n"},{"url":"https://www.atlasofms.org/","title":"Atlas of MS: Number of people with MS","publisher":"MS International Federation","source_type":"reputable_reference","statistic":"Global MS population grew from 2.3 million in 2013 to 2.8 million in 2020 to 2.9 million in 2023","excerpt":"\"The number of people with MS across the globe has increased from 2.3 million in 2013 to 2.8 million in 2020 and 2.9 in 2023.\"\n","source_date":"2023-10-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20200907082035/http://www.atlasofms.org/","calculation_notes":"Used as the continuously updated headline anchor for the \"~2.9 million people living with MS globally\" statement. The 2.9M figure is the most recent refresh of the same Atlas methodology that produced the peer- reviewed Walton 2020 paper, so the per-100,000 rate and incidence rate in the primary source remain the quantitative basis; the 2.9M update is used only for the headline count.\n","independence_note":"Directly downstream from the Walton 2020 peer-reviewed source above; not an independent estimate, just a more recent snapshot of the same data pipeline.\n"}],"comparison_anchors":[{"label":"Lifetime Alzheimer's / dementia mortality (global adult)","lifetime_us_adult":0.12},{"label":"Lifetime lung cancer incidence (US adult)","lifetime_us_adult":0.062},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.0013,"notes":"Walton 2020 global incidence of 2.1 per 100,000 persons/year compounded across adult life"},{"region":"US adult","probability":0.003,"notes":"Wallin 2019 US point prevalence ~360 per 100,000 implies lifetime incidence roughly 1 in 300 to 1 in 330; a north-south gradient inside the US persists"},{"region":"Scotland, northern Canada, Scandinavia","probability":0.005,"notes":"Highest-prevalence regions globally; consistent with the latitude gradient — MS prevalence scales roughly with distance from the equator"},{"region":"Equatorial and tropical regions","probability":0.0002,"notes":"Western Pacific region reports ~4.8 per 100,000 prevalence in the Atlas of MS — roughly an order of magnitude below the global average"},{"region":"Women (global)","probability":0.0045,"notes":"Atlas of MS female:male ratio ~2:1 globally; Wallin 2019 US ratio 2.8; women's lifetime incidence sits at roughly 3x the male figure in most cohorts"},{"region":"Men (global)","probability":0.0015,"notes":"Men's lower lifetime incidence is a real biological gap, not a longevity artefact — MS onset peaks decades before competing mortality matters"}],"personal_factor_multipliers":[{"factor":"Epstein-Barr virus seropositive vs seronegative","multiplier":32,"notes":"Bjornevik et al. Science 2022 hazard ratio (95% CI 4.3-245.3) in the US military cohort; EBV is likely a necessary though not sufficient cause of MS. Population-level effect is small because EBV seroprevalence in adults is already ~95%"},{"factor":"female","multiplier":3,"notes":"Wallin 2019 US female:male prevalence ratio 2.8; Atlas of MS reports global ratio roughly 2:1. Gap is real and not fully explained by longer reporting horizons"},{"factor":"first-degree relative with MS","multiplier":7,"notes":"Sibling and parent concordance studies give relative risks around 7x vs the general population; twin concordance pushes higher for monozygotic pairs"},{"factor":"HLA-DRB1*15:01 allele","multiplier":3,"notes":"Strongest common genetic risk factor for MS; approximately 3x odds ratio per copy in European-ancestry cohorts"},{"factor":"low vitamin D / low sun exposure","multiplier":1.4,"notes":"Part of the mechanism behind the latitude gradient; observational studies consistently show ~30-50% higher MS risk in lowest-vitamin-D quartiles"},{"factor":"smoking","multiplier":1.5,"notes":"Meta-analyses of smoking and MS onset give relative risks of roughly 1.5x; smoking also accelerates progression once MS is diagnosed"}],"short_label":"MS","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The global headline number compresses a very large spread. MS prevalence correlates sharply with latitude — Europe and the Americas sit roughly 3 to 5x above the global mean and the Western Pacific / equatorial regions sit an order of magnitude below — so the single 1-in-770 figure is best read as a midpoint, not a universal baseline. The perceived-vs-actual gap for this fear is not primarily about the headline probability; it is about two reframings that have only recently settled into the epidemiology literature. First, Bjornevik et al. 2022 implies that prior EBV infection is effectively a necessary condition for MS, which changes the disease's status from \"unexplained autoimmune\" to \"late sequel of a common herpesvirus infection\". Second, modern disease-modifying therapies have dramatically narrowed the life-expectancy gap between MS patients and the general population — MS in 2026 is not the same disease it was in 1986, even if it is still not curable. This entry is incidence-based, not mortality-based, for that reason: framing MS as \"odds of dying from MS\" would misrepresent both the disease course and the patient experience under current treatment.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale compass needle pointing north on a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/multiple-sclerosis","api_url":"https://likelier.app/api/fears/multiple-sclerosis.json"},{"slug":"motorcycle-death","question":"What are the odds of dying on a motorcycle?","quick_answer":"The population-average lifetime odds of dying on a motorcycle are about 1 in 710, but that number is diluted by the roughly 90% of US adults who never ride. For an active rider the per-mile rate — about 31 deaths per 100 million miles — works out to roughly 1 in 50 over a 30-year riding career.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most non-riders rate motorcycles as obviously dangerous, and most riders know the per-mile numbers are bad but take them anyway. The interesting feature of this fear is that it's roughly calibrated in direction (motorcycles are genuinely much riskier than cars per mile) but usually miscalibrated in magnitude, especially when people try to translate \"dangerous\" into a concrete probability.\n","rough_estimate":"~1 in 100 per year of riding feels about right to most non-riders","kind":"intuition"},"native":{"display":"~31.4 motorcyclist deaths per 100 million vehicle miles (US, 2023)","numerator":31.39,"denominator":100000000,"unit":"per vehicle mile","population":"US motorcyclists, 2023"},"normalized":{"lifetime_us_adult":0.00141,"display":"1 in ~710 lifetime (US adult, population average)","log_value":-2.85,"assumptions":"Refreshed to the 2024 basis (NHTSA NCSA Research Note DOT HS 813 791, April 2026): 6,228 motorcyclist deaths in 2024 — down 2.2 percent from 6,366 in 2023 — across ~260 million US adults = ~2.40e-5 per year, multiplied by 59 years of remaining adult life ≈ 1.41e-3, essentially unchanged from the 2023 figure. This number is heavily diluted by the ~90 percent of US adults who never ride. For an active rider putting 2,000 miles a year on a motorcycle, the direct per-mile rate implies roughly 1 in 50 over a 30-year riding career. For a non-rider the lifetime risk is effectively zero. The population average is the figure you would expect to see for a randomly chosen US adult and is included for comparability with the other Likelier entries; it is not the right number for an individual rider.\n","uncertainty":{"low":0.0008,"high":0.03},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.iihs.org/topics/fatality-statistics/detail/motorcycles-and-atvs","title":"Fatality Facts 2023: Motorcycles and ATVs","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"6,335 motorcyclist deaths in 2023; per mile traveled in 2022, motorcycle deaths were nearly 22 times car deaths","excerpt":"\"A total of 6,335 motorcyclists died in crashes in 2023. That is the highest number ever recorded and a 26% increase since 2019. Motorcycle deaths accounted for 15% of all motor vehicle crash deaths in 2023 and were about triple the number of motorcyclist deaths in 1997. Per mile traveled in 2022, the number of deaths on motorcycles was nearly 22 times the number in cars.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180603/https://www.iihs.org/research-areas/fatality-statistics/detail/motorcycles-and-atvs","calculation_notes":"IIHS compiles FARS fatality counts and computes the per-mile ratio against passenger cars. Used for the total 2023 death count and the cars-vs-motorcycles multiplier that anchors the \"22x to 28x more dangerous\" headline. The per-mile ratio rose from ~22x in 2022 to ~28x in 2023 because US motorcycle VMT dropped roughly 15 percent while fatalities held nearly flat.\n","independence_note":"IIHS draws from NHTSA's FARS database, so it is not fully independent of the NHTSA Traffic Safety Facts source below. They agree because they share a dataset, not because they're two separate measurements.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813732.pdf","title":"Traffic Safety Facts 2023 Data: Motorcycles (DOT HS 813 732)","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"31.39 motorcyclist fatalities per 100 million VMT in 2023 vs 1.13 for passenger car occupants","excerpt":"\"In 2023, there were 6,335 motorcyclists killed — 15% of all traffic fatalities. Per vehicle miles traveled in 2023, motorcyclists were about 28 times more likely than passenger car occupants to die in a motor vehicle crash and were 5 times more likely to be injured.\"\n","source_date":"2024-08-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180639/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813732.pdf","calculation_notes":"NHTSA's per-VMT rate of 31.39 fatalities per 100 million motorcycle miles is the native figure. Converting: 31.39 / 1e8 ≈ 3.14e-7 deaths per mile ridden. For an active rider at 2,000 miles per year over 30 years that's ~0.0188 (about 1 in 53). For the population-average lifetime figure we use 6,335 deaths / ~260M US adults ≈ 2.44e-5 per year, compounded over 59 years ≈ 1.44e-3, or roughly 1 in 700. Both are reported because the population figure hides the fact that essentially all of the risk falls on the ~10 percent of adults who actually ride.\n","independence_note":"NHTSA's Traffic Safety Facts series is the upstream source that IIHS and most other US motorcycle-safety publications cite. Treat IIHS and NHTSA as one source chain for verification, not two independent estimates.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813791","title":"Overview of Motor Vehicle Traffic Crashes in 2024 (DOT HS 813 791)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"6,228 motorcyclists killed in 2024 (~16% of the 39,254 total traffic fatalities); motorcyclist deaths down 2.2% from 2023","excerpt":"\"There were 1,771 fewer people killed in motor vehicle traffic crashes on U.S. roadways during 2024, a 4.3-percent decrease from 41,025 in 2023 to 39,254 in 2024. … Motorcyclist fatalities decreased by 138, a 2.2-percent decrease from 2023.\"\n","source_date":"2026-04-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260518203717/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813791","calculation_notes":"The 2024 census count is 6,228 motorcyclist deaths (Table 3), down 138 (2.2%) from the 6,366 in 2023; total traffic fatalities fell 4.3% to 39,254. The ~16% share is derived: 6,228 / 39,254 = 15.9%. Population-average lifetime refresh: 6,228 / ~260M US adults ≈ 2.40e-5 per year, compounded over 59 years ≈ 1.41e-3 (≈1 in 710), within the existing uncertainty band and within 0.01 log of the prior point. Note: motorcyclist deaths fell in 2024 alongside the overall decline — 2023 (6,335/6,366) remains the modern peak, not 2024. The Overview Research Note does not republish a 2024 motorcyclist per-VMT rate, so the native figure stays on the 2023 NHTSA Traffic Safety Facts value above.\n","independence_note":"Same FARS/NCSA source chain as the IIHS and Traffic Safety Facts sources above — not an independent measurement, used here for the refreshed 2024 census count.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"universal helmet-law state (CA, NY, MA)","probability":0.001,"notes":"NHTSA: states with universal helmet laws have roughly 30% lower motorcycle fatality rates than partial/no-helmet states"},{"region":"partial or no helmet law (FL, TX, IA)","probability":0.002,"notes":"unhelmeted rider fraction is substantially higher in these states"},{"region":"never-rider","probability":1e-7,"notes":"near-zero exposure to motorcycle crash risk; deaths from other vehicles striking motorcycles are captured elsewhere"}],"personal_factor_multipliers":[{"factor":"daily motorcycle commuter (~5,000 mi/yr)","multiplier":20,"notes":"NHTSA: per-mile fatality rate is ~29x cars; a regular rider's exposure-adjusted risk dwarfs the population average"},{"factor":"no helmet, state without helmet law","multiplier":2,"notes":"NHTSA: unhelmeted riders are ~1.7-2x more likely to die in a crash"},{"factor":"non-rider (never rides motorcycles)","multiplier":0.01,"notes":"population average includes non-riders; a non-rider's motorcycle death risk is effectively zero"},{"factor":"alcohol-involved riding","multiplier":3,"notes":"NHTSA: ~28% of fatally injured motorcyclists are alcohol-impaired"}],"short_label":"Motorcycle crash","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The per-capita lifetime figure of ~1 in 700 is the least useful number on this page and is included only for cross-category comparability. Motorcycle risk is almost entirely concentrated in the minority of adults who ride; for non-riders the number is effectively zero, and for a typical active rider putting 2,000 miles per year on a bike over a 30-year career the implied lifetime risk is closer to 1 in 50. Within the rider population, risk varies by another order of magnitude based on helmet use, alcohol, bike class (sport vs cruiser vs touring), urban vs rural roads, and rider age: 35 percent of 2023 motorcyclist fatalities were unhelmeted, and 41 percent of riders killed in single-vehicle crashes were alcohol-impaired. The per-mile rate also excludes off-road and track riding, which are accounted for separately. NHTSA's 2024 data (Research Note DOT HS 813 791) puts motorcyclist deaths at 6,228 — down 2.2 percent from 2023 and still about 16 percent of the 39,254 total traffic fatalities, which themselves fell 4.3 percent. The 2024 dip does not change the headline: the per-mile disparity against cars, not the year-to-year count, is what makes this entry rank where it does.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty motorcycle helmet resting on a pale surface, flat vector illustration in muted greys."},"canonical_url":"https://likelier.app/motorcycle-death","api_url":"https://likelier.app/api/fears/motorcycle-death.json"},{"slug":"workplace-fatal-injury","question":"What are the odds of dying in a workplace accident?","quick_answer":"Dying in a workplace accident runs about 1 in 690 over a US working career, roughly one-seventh the lifetime odds of dying in a car crash. The all-occupation average hides huge variance: agriculture and logging workers sit near 1 in 124, office workers around 1 in 2,300.\n","category":"health","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Workplace death is one of those risks that most white-collar workers file under \"other people's problem.\" Office employees, knowledge workers, and retail staff rarely encounter occupational fatalities in their social networks, and news coverage clusters around spectacular collapses and industrial explosions rather than the steady drip of transportation incidents, falls, and struck-by events that compose the bulk of the count. Blue-collar workers in construction, agriculture, and trucking tend to calibrate better — the hazard is concrete and discussed on site — but even they may underestimate the cumulative career-long toll because each year's probability is low enough to feel manageable. No large-scale US survey isolates \"fear of dying at work\" as a standalone question, so the perceived side here is editorial intuition.\n","rough_estimate":"Most office workers treat it as near-zero; construction and agriculture workers sense it is real but underestimate the career accumulation","kind":"intuition"},"native":{"display":"~5,070 fatal work injuries in 2024; rate of 3.3 per 100,000 FTE workers","numerator":5070,"denominator":155000000,"unit":"per year","population":"US civilian employed workers (all industries, 2024)"},"normalized":{"lifetime_us_adult":0.00145,"display":"~1 in 690 over a working career (US)","log_value":-2.839,"assumptions":"The BLS Census of Fatal Occupational Injuries recorded 5,070 fatal work injuries in 2024 at a rate of 3.3 per 100,000 full-time equivalent workers. The annual probability for an employed worker is approximately 3.3e-5. A typical US working career spans roughly 44 years (age 18 to 62). Lifetime working-career risk: 1 − (1 − 3.3e-5)^44 ≈ 0.00145, or about 1 in 690. This uses working years rather than the site-standard 59-year remaining-life horizon because workplace fatality exposure ceases at retirement. The figure is an all-occupation average; individual risk varies by a factor of 18 or more across industries (agriculture/forestry/fishing at ~18.6 per 100,000 vs professional services at ~1 per 100,000). The uncertainty band (0.0008–0.0025) reflects the range between white-collar-only careers and careers in mid-risk industries like manufacturing or warehousing; workers in agriculture, logging, or fishing face career risks well above the upper bound.\n","uncertainty":{"low":0.0008,"high":0.0025},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.bls.gov/news.release/cfoi.nr0.htm","title":"Census of Fatal Occupational Injuries Summary, 2024","publisher":"U.S. Bureau of Labor Statistics","source_type":"govt_report","statistic":"5,070 fatal work injuries in 2024; fatal work injury rate of 3.3 per 100,000 FTE workers, down from 3.5 in 2023","excerpt":"\"There were 5,070 fatal work injuries recorded in the United States in 2024, down 4.0 percent from the revised count of 5,283 fatal work injuries in 2023. The fatal work injury rate was 3.3 fatalities per 100,000 full-time equivalent (FTE) workers in 2024, a decrease from a rate of 3.5 in 2023.\"\n","source_date":"2026-02-19","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260521205557/https://www.bls.gov/news.release/cfoi.nr0.htm","calculation_notes":"The BLS CFOI 2024 headline figures anchor both the native ratio and the normalized lifetime estimate. Annual rate: 5,070 / ~155 million employed ≈ 3.27 per 100,000 (BLS reports 3.3 per 100,000 FTE, which adjusts for part-time workers). Career-lifetime conversion: 1 − (1 − 3.3e-5)^44 ≈ 0.00145, using a 44-year working career from age 18 to 62. Transportation incidents account for 38.2% of all fatal work injuries (1,937 deaths), making them the leading event type.\n"},{"url":"https://www.bls.gov/charts/census-of-fatal-occupational-injuries/number-and-rate-of-fatal-work-injuries-by-industry.htm","title":"Number and rate of fatal work injuries, by selected private industries","publisher":"U.S. Bureau of Labor Statistics","source_type":"govt_report","statistic":"Agriculture, forestry, fishing, and hunting: fatality rate of 18.6 per 100,000 FTE workers in 2024 (417 fatalities), the highest fatal injury rate of any industry sector","excerpt":"[Paraphrase — the BLS chart renders as a dynamic SVG; figures read from the 2024 CFOI industry data.] In 2024 the agriculture, forestry, fishing and hunting industry sector recorded 417 fatal work injuries at a rate of 18.6 per 100,000 full-time equivalent workers — the highest fatal injury rate of any private industry sector.\n","source_date":"2026-02-19","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260405160451/https://www.bls.gov/charts/census-of-fatal-occupational-injuries/number-and-rate-of-fatal-work-injuries-by-industry.htm","calculation_notes":"Industry-specific rates anchor the personal_factor_multipliers. The agriculture/forestry/fishing/hunting rate of 18.6 per 100,000 is roughly 5.6x the all-industry average of 3.3. Construction/extraction at 9.2 is ~2.8x. These multipliers translate directly into the career-lifetime risk: a 44-year agriculture career yields 1 − (1 − 18.6e-5)^44 ≈ 0.0081, roughly 1 in 124 — an order of magnitude above the all-occupation average.\n"},{"url":"https://injuryfacts.nsc.org/work/work-overview/work-related-fatality-trends/","title":"Work-related Fatality Trends","publisher":"National Safety Council — Injury Facts","source_type":"reputable_reference","statistic":"Long-term decline in US workplace fatality rates from ~14 per 100,000 in 1970 to ~3.3 per 100,000 in 2024; absolute counts relatively stable at ~5,000/year due to workforce growth","excerpt":"\"The number of preventable work deaths per 100,000 workers has decreased substantially since the early 1970s. However, the total number of fatal work injuries has remained relatively stable at approximately 5,000 per year as the workforce has grown.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250529030000/https://injuryfacts.nsc.org/work/work-overview/work-related-fatality-trends/","calculation_notes":"NSC Injury Facts provides the long-term trend context. The rate has fallen from ~14 per 100,000 in 1970 to ~3.3 per 100,000 in 2024, a roughly 4x improvement. But the absolute count has stayed near 5,000 because the US employed workforce has grown from ~80 million to ~155 million over the same period. This context matters for the \"assumptions\" field: the normalized figure assumes current rates persist, which is conservative given the historical trend.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from a lightning strike (lifetime, US)","lifetime_us_adult":0.0000135},{"label":"Death from general anesthesia","lifetime_us_adult":0.00022},{"label":"School shooting death (US student)","lifetime_us_adult":0.0000125}],"personal_factor_multipliers":[{"factor":"Agriculture, forestry, fishing, hunting worker","multiplier":5.6,"notes":"BLS 2024: 18.6 per 100,000 FTE vs all-industry 3.3; career risk ~1 in 124"},{"factor":"Construction or extraction worker","multiplier":2.8,"notes":"BLS 2024: 9.2 per 100,000 FTE; career risk ~1 in 250"},{"factor":"Transportation and material moving worker","multiplier":3.8,"notes":"BLS 2024: 12.5 per 100,000 FTE; 1,391 deaths, largest absolute count by occupation"},{"factor":"Office or professional-services worker","multiplier":0.3,"notes":"Fatal injury rate ~1 per 100,000; career risk ~1 in 2,300"},{"factor":"Hispanic or Latino worker","multiplier":1.3,"notes":"Overrepresented in construction and agriculture; BLS consistently reports disproportionate fatality share"}],"short_label":"Workplace fatality","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 3.3 per 100,000 rate is an all-industry, all-occupation average that conceals enormous variance. A software engineer and a logging worker share this headline number the way they share a national crime rate — technically accurate, practically useless for personal calibration. The BLS CFOI includes drug overdoses occurring at work (410 in 2024), which some analysts argue should be classified differently; excluding them would lower the rate by about 8%. The count also excludes military personnel, self-employed workers not covered by state UI programs, and workers under age 16. The long-term trend is strongly downward (the rate has fallen ~75% since 1970), so the career-lifetime figure using current rates is probably conservative — a worker entering the labor force today will likely face lower rates in 2060 than in 2026. The comparison to other countries is stark: the US rate of 3.3 per 100,000 is roughly 2-3x the rate in the UK, Germany, or the Nordic countries.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified hard hat resting on a flat surface in muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/workplace-fatal-injury","api_url":"https://likelier.app/api/fears/workplace-fatal-injury.json"},{"slug":"recreational-boating-drowning","question":"What are the odds of drowning after falling overboard from a recreational boat or yacht?","quick_answer":"A regular recreational boater faces roughly a 1 in 685 lifetime chance of drowning after going overboard, from about 3.7 drownings per 100,000 vessel-years. Most boaters perceive their risk as negligible.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Recreational boaters tend to file themselves as competent and their vessels as seaworthy — which is accurate enough in calm conditions on familiar water. The scenario most people underweight is an unplanned entry into the water: a rogue wake, a misstep on a wet deck, a sudden capsize on a small motorboat. Public awareness campaigns focus on life jacket non-compliance but rarely translate that into a numerical lifetime probability. Most regular boaters would guess their drowning risk is negligible; occasional boaters often don't think about it at all.\n","rough_estimate":"most boaters perceive their risk as negligible or very low","kind":"intuition"},"native":{"display":"~3.7 drowning deaths per 100,000 registered vessel-years","numerator":37,"denominator":1000000,"unit":"per registered vessel-year","population":"US recreational boat owners / primary operators, USCG 2023"},"normalized":{"lifetime_us_adult":0.00146,"display":"~1 in 685 lifetime (regular recreational boater, ~40 active boating years)","log_value":-2.836,"assumptions":"USCG 2023 reports 564 total boating fatalities across ~11.55 million registered recreational vessels, giving an all-cause fatality rate of 4.88 per 100,000 vessel-years. Drowning accounts for 75% of fatal boating accidents where cause of death is known (USCG 2023), yielding a drowning rate of approximately 3.66 per 100,000 vessel-years (3.66e-5 per year). A regular recreational boater who owns or regularly uses one vessel for roughly 40 active boating years (ages ~25-65) faces a cumulative drowning probability of 1-(1-3.66e-5)^40 ≈ 0.00146, or about 1 in 685. This is an activity-specific figure for a committed recreational boater; it does not apply to the US adult population at large. Occasional boaters (5-10 trips/year, shared vessels) sit toward the lower bound; frequent boaters on small open motorboats without consistent life jacket use sit toward the upper bound. The USCG data are vessel-level, not person-level, so the rate conflates vessels used by multiple people and vessels sitting idle; the true per-active-boater annual rate is likely somewhat higher.\n","uncertainty":{"low":0.0008,"high":0.003},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://uscgboating.org/library/accident-statistics/Recreational-Boating-Statistics-2023-Ch2.pdf","title":"2023 Recreational Boating Statistics (COMDTPUB P16754.37) — Chapter 2: Accident Data","publisher":"United States Coast Guard, Office of Auxiliary and Boating Safety","source_type":"govt_report","statistic":"564 boating fatalities in 2023; fatality rate 4.9 deaths per 100,000 registered recreational vessels; where cause of death was known, 75% of fatal accident victims drowned; 87% of drowning victims were not wearing a life jacket","excerpt":"\"Where the cause of death was known, 75 percent of fatal boating accident victims drowned. Of those drowning victims with reported life jacket usage, 87 percent were not wearing a life jacket.\"\n","source_date":"2024-05-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260503035834/https://www.uscgboating.org/library/accident-statistics/Recreational-Boating-Statistics-2023-Ch2.pdf","calculation_notes":"Primary source for native rate. 564 total fatalities ÷ ~11.55 million registered vessels = 4.88 per 100,000 vessel-years. Applying the 75% drowning share gives ~3.66 drowning deaths per 100,000 vessel-years (3.66e-5 annual probability per registered vessel). Compounded over 40 active boating years: 1-(1-3.66e-5)^40 ≈ 0.00146.\n"},{"url":"https://www.uscgboating.org/library/accident-statistics/Recreational-Boating-Statistics-2024.pdf","title":"2024 Recreational Boating Statistics (COMDTPUB P16754.38)","publisher":"United States Coast Guard, Office of Auxiliary and Boating Safety","source_type":"govt_report","statistic":"2024 fatality rate 4.8 deaths per 100,000 registered vessels; 239 fall-overboard incidents resulting in 138 deaths; 87% of drowning victims not wearing life jackets; drowning accounts for 76% of fatal boating accident deaths","excerpt":"\"In 2024, there were 239 person falls overboard incidents in the United States, resulting in 138 deaths and 104 injuries. Where cause of death was known, 76 percent of fatal boating accident victims drowned.\"\n","source_date":"2025-05-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260424160530/https://www.uscgboating.org/library/accident-statistics/Recreational-Boating-Statistics-2024.pdf","calculation_notes":"Corroborating year. The 2024 data (4.8 per 100,000, 76% drowning share) is consistent with 2023 and confirms the drowning fraction has been stable at roughly 75-76% across recent years. The overboard-specific figure of 239 incidents and 138 deaths yields a case-fatality rate of 58% for reported fall-overboard events — consistent with the lethal consequences of cold-water immersion and rapid submersion in the absence of a life jacket.\n"},{"url":"https://www.safeboatingcouncil.org/resources/recreational-boating-facts/","title":"Recreational Boating Facts","publisher":"National Safe Boating Council","source_type":"reputable_reference","statistic":"Falls overboard, capsizing, and voluntary departure from vessel accounted for over half of fatal boating incidents; 4 out of 5 boaters who drowned were using vessels less than 21 feet in length","excerpt":"\"Falls overboard, capsizing, and cases where a person voluntarily departed a vessel accounted for over half of fatal incidents. Four out of every five boaters who drowned were using vessels less than 21 feet in length.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260516151003/https://www.safeboatingcouncil.org/resources/recreational-boating-facts/","calculation_notes":"Used to characterize the mechanism distribution. The vessel-size finding contextualizes the upper-bound uncertainty: small open motorboats carry disproportionate drowning risk relative to larger cabin cruisers and yachts, where crew are less exposed to unplanned water entry.\n"}],"comparison_anchors":[{"label":"Drowning (general population, lifetime US adult)","lifetime_us_adult":0.00071},{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"always wears a life jacket on deck","multiplier":0.13,"notes":"87% of drowning victims were not wearing a life jacket; inverting implies jacket-wearers face roughly 1/8 the drowning risk of non-wearers in overboard scenarios."},{"factor":"boats on small open motorboat (<21 ft) without life jacket","multiplier":3,"notes":"4 in 5 drowning victims were on vessels under 21 ft; combines vessel-type exposure with non-jacket use."},{"factor":"boats on a larger cruiser or sailing yacht with safety harnesses and jacklines","multiplier":0.3,"notes":"Larger vessels with tethering systems substantially reduce unplanned overboard events."},{"factor":"occasional boater (5-10 trips per year, shared vessel)","multiplier":0.4,"notes":"Lower total exposure relative to a vessel owner boating 20+ times per year."}],"short_label":"Boating drowning","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The vessel-level denominator is the core limitation here. USCG registers vessels, not boating days or boater-hours, so the 4.9-per-100,000 rate conflates actively used vessels with boats that spend most of the season on a trailer. A frequently used vessel accumulates more drowning exposure per year than a boat launched twice a season, but both count equally in the denominator. The 87% no-life-jacket figure applies to drowning victims and is not a direct statement of risk for boaters who do wear life jackets; it indicates that jacket-wearing is a strong protective factor, but the counterfactual survival rate is not directly published. Alcohol was the leading known contributing factor in fatal boating accidents in 2023, accounting for 17% of all fatalities — meaning the headline rate is partly a behavioral choice rather than a fixed structural risk. Finally, the 75% drowning share is \"where cause of death was known,\" which may exclude some cases; the true fraction is unlikely to differ materially.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A single orange life jacket ring lying on a pale wooden dock beside calm water, flat vector illustration."},"canonical_url":"https://likelier.app/recreational-boating-drowning","api_url":"https://likelier.app/api/fears/recreational-boating-drowning.json"},{"slug":"solar-panel-fire","question":"What are the odds of rooftop solar panels causing a house fire?","quick_answer":"Rooftop solar carries about a 0.15% chance of causing a residential fire over a 25-year system life (roughly 1 fire per 17,500 installations a year) — well below the 1–5% lifetime risk many homeowners imagine.\n","category":"tech","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Solar panel fire anxiety has grown alongside adoption. News coverage of rooftop fires, insurance premium debates, and high-profile recalls (GAF Energy shingles, Siemens meter combos) contribute to a sense that photovoltaic systems are a smouldering risk bolted to the roof. The fact that DC systems remain energized whenever sunlight hits them -- and that firefighters must account for electrocution risk -- amplifies the perceived danger. Some homeowners and insurers treat solar as a meaningful increment to household fire risk.\n","rough_estimate":"~1-5% chance of a fire over the system's lifetime","kind":"intuition"},"native":{"display":"~1 residential fire per 17,500 residential solar installations per year (0.006%)","numerator":6,"denominator":100000,"unit":"residential fire incidents per residential solar installation per year","population":"UK residential solar installations, QBE FOI analysis of 37/49 fire services (2024); 97 of 171 total fires were residential"},"normalized":{"lifetime_us_adult":0.0015,"display":"~0.15% cumulative probability of a residential fire over a 25-year system lifetime","log_value":-2.82,"assumptions":"QBE Insurance's 2025 FOI analysis of 37 of 49 UK fire services found 171 solar panel fire incidents in 2024 across ~1.7 million installations. Of these, 97 were residential fires (the remainder were commercial, industrial, or solar farm incidents). Using residential fires only: 97 / 1,697,231 = 0.0057%, or ~1 in 17,500 per year. The Fraunhofer ISE / TUV Rheinland 2013 study of 1.3 million German systems found 0.006% caused fires with \"large damage\" (~1 in 16,700) — closely matching the residential-only QBE figure. We use the residential-only QBE figure (6 per 100,000) as the primary estimate. Over a 25-year system lifetime (typical warranty period): 1 - (1 - 0.00006)^25 = 0.0015, or ~0.15%. For context, the annual probability of any house fire from all causes is ~0.24% (NFPA/USFA), meaning solar adds roughly a 4% relative increase to baseline household fire risk. Solar fires are 25-40x less likely per year than a general house fire and do not appear on any top-10 list of domestic fire causes. Australian data shows just 1.5% of residential fires were linked to solar PV systems. The BRE (UK government) study of 80+ incidents found installation error (36%) and faulty products (12%) as the top attributable causes, with 47% undetermined. Note: The primary fire rate (1 in 10,000 per year) is derived from UK data and is used as a proxy; US-specific solar panel fire rates may differ due to differences in installation standards, climate exposure, installer certification requirements, and the age profile of the installed fleet.\n","uncertainty":{"low":0.0006,"high":0.004},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://qbeeurope.com/news-and-events/press-releases/uk-fire-services-tackle-a-solar-panel-fire-every-two-days/","title":"UK fire services tackle a solar panel fire every two days","publisher":"QBE Insurance (FOI analysis of UK fire services)","source_type":"reputable_reference","statistic":"171 solar panel fire incidents in 2024 across ~1.7 million UK installations; 60% increase since 2022 (107 incidents)","excerpt":"\"Freedom of Information data from 37 of 49 UK fire and rescue services revealed 171 solar panel fire incidents in 2024, up from 128 in 2023 and 107 in 2022. Fires rose 60 percent over two years, outpacing the 29.6 percent growth in installations.\"\n","source_date":"2025-11-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503094055/https://qbeeurope.com/news-and-events/press-releases/uk-fire-services-tackle-a-solar-panel-fire-every-two-days/","calculation_notes":"171 total solar fires reported (37/49 fire services), of which 97 were residential. 97 / 1,697,231 UK residential installations = 0.0057% or ~1 in 17,500 per year. Only 37 of 49 fire services responded, so the true UK total is likely higher; but we use the unadjusted residential-only figure. The 60% increase in fires vs 30% increase in installations suggests a worsening per-unit rate, possibly due to aging of earlier installations or quality issues in the installation boom.\n"},{"url":"https://www.ise.fraunhofer.de/en/press-media/press-releases/2013/fire-protection-in-photovoltaic-systems.html","title":"Fire Protection in Photovoltaic Systems: Facts Replace Fiction","publisher":"Fraunhofer ISE / TUV Rheinland","source_type":"reputable_reference","statistic":"0.006% of 1.3 million German PV systems caused fires with large damage (~1 in 16,700 annually)","excerpt":"\"Of the 1.3 million photovoltaic systems installed in Germany, approximately 0.006 percent have caused a fire involving large-scale damage. The risk posed by PV systems is no greater than that presented by other electrical installations in the household.\"\n","source_date":"2013-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260328201654/https://www.ise.fraunhofer.de/en/press-media/press-releases/2013/fire-protection-in-photovoltaic-systems.html","calculation_notes":"Fraunhofer ISE and TUV Rheinland jointly studied fire risk across the German PV fleet (then the world's largest). The 0.006% figure uses a stricter \"large damage\" threshold than the QBE data, explaining the lower rate. The study concluded that PV fire risk is comparable to other household electrical installations, not elevated above baseline. Germany's electrical safety standards and installer certification requirements may contribute to a lower rate than in countries with less regulated installation markets.\n"},{"url":"https://assets.publishing.service.gov.uk/media/5c90a30840f0b633ff9a3537/Fires_and_solar_PV_systems-Investigations_Evidence_Issue_2.9.pdf","title":"Fire and Solar PV Systems: Investigations and Evidence","publisher":"BRE / UK Department for Communities and Local Government","source_type":"govt_report","statistic":"Of 80+ investigated incidents: 36% caused by installation error, 12% by faulty products, 5% by system design, 47% undetermined; DC isolators and connectors were the most common failure points","excerpt":"\"Installation error accounted for 36 percent of incidents where a root cause could be determined. Faulty products accounted for 12 percent and system design 5 percent. In 47 percent of cases the root cause could not be determined. DC isolators and DC connectors were the most frequent component failure points.\"\n","source_date":"2017-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260505232152/https://assets.publishing.service.gov.uk/media/5c90a30840f0b633ff9a3537/Fires_and_solar_PV_systems-Investigations_Evidence_Issue_2.9.pdf","calculation_notes":"The BRE study is the only government-commissioned forensic investigation of solar PV fire causes. The 36% installation-error finding is critical: it means the single largest controllable risk factor is workmanship, not inherent technology risk. DC isolators (18 incidents) and connectors (10 incidents) were the most common failure points. Incompatible MC4-compatible connectors from different manufacturers remain a persistent industry problem. The 47% undetermined rate reflects the difficulty of post-fire forensic analysis on rooftop systems.\n"}],"comparison_anchors":[{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.0025},{"label":"EV battery fire (per vehicle lifetime)","lifetime_us_adult":0.003},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"installation by uncertified or inexperienced installer","multiplier":5,"notes":"BRE found 36% of fire incidents traced to installation error; quality of workmanship is the dominant controllable risk factor"},{"factor":"system older than 10 years without inspection","multiplier":2,"notes":"DC connectors and isolators degrade over time, especially in exposed outdoor environments; QBE data shows fires increasing faster than installations"},{"factor":"system with DC arc fault circuit interrupter (AFCI)","multiplier":0.3,"notes":"NEC 2017 requires rapid shutdown and AFCI protection for new US installations; these detect arc faults before they cause fires"},{"factor":"microinverter system (no high-voltage DC)","multiplier":0.3,"notes":"Microinverters convert to AC at the panel level, eliminating the high-voltage DC wiring and connectors that cause most PV fires"}],"short_label":"Solar panel fire","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"property","valence":"negative","caveats":"The QBE data covers only 37 of 49 UK fire services, so the true incident count is higher. The annual rate of ~1 in 10,000 is a UK figure that may not perfectly generalize to the US, where installation standards, climate, and installer certification requirements differ. The Fraunhofer figure of 0.006% uses a stricter threshold (\"large damage\") and is from 2013 when the fleet was younger. Solar fire rates may increase as early installations age and components degrade. Modern systems with AFCI protection and microinverters are likely substantially safer than string-inverter systems from 2010-2015. The 0.15% lifetime figure assumes a constant annual rate over 25 years, which may understate risk for older systems. Most solar fires cause property damage, not injury or death; the fatality rate from PV fires is not separately tracked but appears very low.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":4,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"Rooftop solar panels on a residential house viewed from ground level, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/solar-panel-fire","api_url":"https://likelier.app/api/fears/solar-panel-fire.json"},{"slug":"typhoid-endemic","question":"What are the odds of dying from typhoid fever in an endemic region?","quick_answer":"Typhoid and paratyphoid fever cause about 130,000 deaths a year worldwide — roughly a 1 in 650 lifetime chance of death for the average global adult, concentrated in endemic regions.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Typhoid fever occupies a strange position in the popular imagination of wealthy countries: it is strongly associated with the past, with Mary Mallon and Victorian-era sanitation failures, rather than with the present. The availability of clean water, modern sewage systems, and antibiotics has reduced typhoid to a handful of travel-associated cases per year in the US and Europe, reinforcing the perception that it is a historical disease. In South Asia, Southeast Asia, and parts of sub-Saharan Africa, typhoid remains a major killer, causing an estimated 130,000 deaths per year among populations without reliable access to safe water and sanitation. The gap between historical perception and current endemic reality is one of the widest in infectious disease.\n","kind":"intuition"},"native":{"display":"~130,000 deaths per year globally from typhoid and paratyphoid fever","numerator":130000,"denominator":5000000000,"unit":"per year","population":"global adults and children"},"normalized":{"lifetime_us_adult":0.00153,"display":"~1 in 650 lifetime (global adult)","log_value":-2.81,"assumptions":"Native rate: The 130,000-deaths central estimate is a rounded cross-source value for annual global enteric-fever (typhoid and paratyphoid) deaths, bracketed by the WHO fact sheet's ~110,000 (2019), the GBD 2021 systematic analysis's 107.5 thousand (95% UI 56.1-180.8; eClinicalMedicine, below), and the standalone GBD 2017 estimate of 135,900 deaths (95% UI 76,900-218,900; Lancet Infectious Diseases, below). Against a global adult population of ~5 billion: 130,000 / 5,000,000,000 = 0.000026. Lifetime conversion: 1 - (1 - 0.000026)^59 = 0.00153. Uncertainty low bound uses 80,000 deaths (near the GBD 2017 lower UI of 76,900, reflecting declining trends and improved treatment access); high bound uses 161,000, a conservative figure well inside the GBD 2017 upper UI of 218,900. Low: 80,000/5B compounded 59 years = 0.00094. High: 161,000/5B compounded 59 years = 0.0019. The burden is concentrated in South Asia (particularly India, Pakistan, and Bangladesh), Southeast Asia, and sub-Saharan Africa. For any adult in a high-income country with modern water treatment and sanitation, personal typhoid mortality risk is negligible.\n","uncertainty":{"low":0.00094,"high":0.0019},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/typhoid","title":"Typhoid — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"As of 2019 estimates, ~9 million typhoid cases annually, resulting in about 110,000 deaths per year","excerpt":"\"As of 2019 estimates, there are 9 million cases of typhoid fever annually, resulting in about 110 000 deaths per year. Typhoid risk is higher in populations that lack access to safe water and adequate sanitation, and children are at highest risk.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260411073438/https://www.who.int/news-room/fact-sheets/detail/typhoid","calculation_notes":"The WHO fact sheet gives ~9 million cases and about 110,000 deaths per year (2019 estimates). The 130,000 central estimate used for the native numerator is a rounded cross-source value bracketed by the three burden-of-disease estimates cited here: WHO's ~110,000 (2019), the GBD 2021 analysis's 107.5 thousand (2021) and higher 2017 figure, and the standalone GBD 2017 study's 135,900 (eClinicalMedicine and Lancet Infectious Diseases, below). 130,000 / 5B = 0.000026 annual rate, compounded over 59 years yields 0.00153.\n"},{"url":"https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(24)00462-0/fulltext","title":"The global burden of enteric fever, 2017-2021: a systematic analysis from the global burden of disease study 2021","publisher":"eClinicalMedicine (The Lancet)","source_type":"peer_reviewed","statistic":"For 2021, an estimated 9.3 million global cases of enteric fever (95% UI 7.3-11.9) and 107.5 thousand deaths (95% UI 56.1-180.8); burden concentrated in 75 endemic countries","excerpt":"\"9.3 million global cases of enteric fever (95% uncertainty interval: 7.3−11.9) [...] 107.5 thousand deaths (56.1−180.8).\"\n","source_date":"2024-10-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20241127234342/https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(24)00462-0/fulltext","calculation_notes":"The GBD 2021 systematic analysis estimates 9.3 million cases (95% UI 7.3-11.9) and 107.5 thousand deaths (95% UI 56.1-180.8) for 2021, and a higher 2017 figure — the source of the well-documented decline. This confirms the same order of magnitude as the WHO fact sheet (~110,000 deaths) and the standalone GBD 2017 study (135,900 deaths). The 130,000 central native numerator is a rounded cross-source value bracketed by these estimates, not this study's own 2021 headline. The study integrates antimicrobial resistance data, increasingly relevant as drug-resistant strains spread in South Asia.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6437314/","title":"The global burden of typhoid and paratyphoid fevers: a systematic analysis for the Global Burden of Disease Study 2017","publisher":"The Lancet Infectious Diseases","source_type":"peer_reviewed","statistic":"Typhoid and paratyphoid fevers caused an estimated 135,900 deaths globally in 2017; children had the highest morbidity and mortality rates","excerpt":"\"Typhoid and paratyphoid fevers remain important causes of morbidity and mortality. Children had the highest morbidity and mortality rates; males had higher rates of incidence, mortality, and DALYs than females. The burden is concentrated in South Asia, Southeast Asia, and Oceania.\"\n","source_date":"2019-04-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503083800/https://pmc.ncbi.nlm.nih.gov/articles/PMC6437314/","calculation_notes":"The GBD 2017 estimate of 135,900 deaths provides an independent temporal data point consistent with the 130,000 central estimate. Confirms the demographic pattern (children most affected, males disproportionately) and the geographic concentration in South Asia. The decline from 135,900 (2017) toward 130,000 (2021) is consistent with the gradual reduction observed in the GBD trend analysis.\n"},{"url":"https://www.cdc.gov/typhoid-fever/hcp/clinical-guidance/index.html","title":"Clinical Guidance for Typhoid Fever and Paratyphoid Fever","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Case fatality rate for untreated typhoid fever was higher than 10% before widespread antibiotic use; with appropriate treatment the fatality rate is lower than 1%","excerpt":"\"Before widespread antibiotic use, the case fatality rate for typhoid fever was higher than 10%. [...] With appropriate treatment, the fatality rate of typhoid fever is lower than 1%.\"\n","source_date":"2025-01-16","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260616203309/https://www.cdc.gov/typhoid-fever/hcp/clinical-guidance/index.html","calculation_notes":"Grounds the body-text claim that untreated typhoid case fatality \"exceeds 10%\" and drops \"below 1%\" with appropriate antibiotics — this is a per-patient case-fatality contrast, not a population-level death count, so it does not change the 130,000-deaths/year headline. It explains why access to healthcare is described as the critical determinant of survival.\n","independence_note":"CDC clinical guidance for practitioners, independent of the WHO fact sheet and the two GBD burden-of-disease studies above, which report population-level case and death counts rather than per-patient case fatality rates.\n"}],"comparison_anchors":[{"label":"Death from rabies via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death from schistosomiasis (lifetime, global adult)","lifetime_us_adult":0.00015},{"label":"Death from food poisoning (lifetime, US)","lifetime_us_adult":0.000019}],"personal_factor_multipliers":[{"factor":"Lives in a high-income country with treated water, sanitation, and prompt antibiotic access","multiplier":0.01,"notes":"The US reports fewer than 5 typhoid deaths per year, almost all in returned travellers. Treated water plus early effective antibiotics cut the case-fatality rate below 1%, pulling personal risk far below the global average the headline figure reflects."},{"factor":"Typhoid conjugate vaccine (TCV) received","multiplier":0.22,"notes":"TyVAC Phase 3 trial, Malawi (Patel et al., The Lancet, 2024): a single dose of Vi-TT showed 78.3% efficacy against blood-culture-confirmed typhoid, durable for at least 4 years (efficacy declining only ~1.3%/year). Residual risk ≈ 0.22×."},{"factor":"Infection with an extensively drug-resistant (XDR) strain (Pakistan and spreading)","multiplier":3,"notes":"XDR Salmonella Typhi is resistant to all first-line oral antibiotics, leaving only azithromycin and injectable carbapenems; case-fatality rises sharply where those are unavailable."},{"factor":"Child in an endemic area","multiplier":2,"notes":"GBD analyses find children bear the highest typhoid morbidity and mortality rates of any age group."}],"short_label":"Typhoid fever","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 1-in-650 global lifetime figure is driven almost entirely by populations in South Asia, Southeast Asia, and sub-Saharan Africa without reliable access to safe water, modern sanitation, and effective antibiotics. For any adult in a high-income country with treated water and sewage infrastructure, the personal probability of dying from typhoid is negligible — the US reports fewer than 5 typhoid deaths per year, almost all in returned travellers. Antimicrobial resistance is a growing concern: extensively drug-resistant (XDR) typhoid strains have emerged in Pakistan and spread to other endemic regions, potentially increasing case fatality rates in settings where second-line antibiotics are unavailable. Typhoid conjugate vaccines (TCVs) recommended by WHO since 2018 are being introduced in endemic countries but coverage remains limited. Children bear a disproportionate share of both morbidity and mortality.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a simple water vessel beside a muted thermometer icon, rendered in subdued warm tones."},"canonical_url":"https://likelier.app/typhoid-endemic","api_url":"https://likelier.app/api/fears/typhoid-endemic.json"},{"slug":"baby-walker-stair-fall","question":"What are the odds an infant in a baby walker falls down a flight of stairs?","quick_answer":"About 1 in 625 walker-using infants in a multi-storey home is treated for a stair fall — roughly three of every four walker injuries. The scenario looks like an avoidable edge case, so parents tend to guess nearer 1 in 10,000, but it is the dominant injury mechanism.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":false,"perceived":{"description":"The walker-on-stairs image is the canonical visual of this hazard: a wheeled seat with an infant in it, perched at the top step. Most parents have seen it in safety campaigns or warning labels, but the perception is that it is an avoidable edge case rather than the dominant injury mechanism. In Sims et al.'s 25-year US dataset it is the dominant mechanism: roughly three out of four walker injuries are stair falls. The American Academy of Pediatrics and Health Canada both cite stair falls as the specific reason for their ban recommendations.\n","rough_estimate":"~1 in 10,000 chance per walker-using infant","kind":"intuition"},"native":{"display":"~1,482 walker stair-fall ER visits in US 2014 for children under 15 months (74.1% of the 2,001 total)","numerator":1482,"denominator":925000,"unit":"per walker-using infant exposed during a calendar year in a home with accessible stairs","population":"US infants under 15 months exposed to baby walkers, 2014"},"normalized":{"lifetime_us_adult":0.0016,"display":"~1 in 625 chance of an ER-treated stair fall per walker-using infant during typical exposure in a multi-storey home","log_value":-2.8,"assumptions":"Sims et al. (2018) reports that 74.1% of walker injuries are stair-related. Applied to the 2014 total of 2,001 ER visits, this gives roughly 1,482 stair-fall ER visits in US children under 15 months in 2014. The denominator of 925,000 walker-using US infants follows the same construction as the sibling entry (baby-walker-injury-any): roughly 25% of the 3.7 million infants in the 0-15 month band. The single largest engineering intervention in this fear's history is the 1997 ASTM F977 voluntary standard requiring either stair-fall brakes or a base width above 36 inches so the walker does not fit through a standard doorway. The 2010 federal mandate made the standard binding. Pre-1997 walkers produced roughly 10x the stair-fall rate of post-2010 compliant units. The uncertainty band is wide because the per-user denominator is an estimate and because compliance with the modern standard varies (secondhand walkers, imported units).\n","uncertainty":{"low":0.0005,"high":0.005},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/30224365/","title":"Infant Walker-Related Injuries in the United States","publisher":"Pediatrics (Sims, Chounthirath, Yang, Hodges, Smith)","source_type":"peer_reviewed","statistic":"230,676 walker-related ER visits in US children under 15 months over 1990-2014; 74.1% stair falls; 90.6% head/neck injuries; annual injuries fell 22.7% in the 4 years after the 2010 federal mandatory standard","excerpt":"\"An estimated 230,676 children <15 months of age were treated in US emergency departments for an infant walker-related injury from 1990 through 2014. Most of the children sustained head or neck injuries (90.6%) and 74.1% were injured by falling down the stairs in an infant walker. Among patients who were admitted to the hospital (4.5%), 37.8% had a skull fracture... The average annual number of injuries decreased by 22.7% during the 4-year period after the implementation of the federal mandatory safety standard compared with the 4-year period before the standard.\"\n","source_date":"2018-10-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251213052631/https://pubmed.ncbi.nlm.nih.gov/30224365/","calculation_notes":"Sims 2018 is the source of the 74.1% stair-fall share. Applied to the 2014 total of 2,001 walker injuries this gives roughly 1,482 stair-fall ER visits for that year. The 22.7% post-standard decline is the abstract-verifiable headline; the stair-fall-specific decline is larger because the ASTM F977 standard targets the stair-fall mechanism specifically.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11533353/","title":"Injuries Associated With Infant Walkers","publisher":"Pediatrics (AAP Committee on Injury and Poison Prevention)","source_type":"peer_reviewed","statistic":"1999: 8,800 ER-treated walker injuries in US children under 15 months. 34 walker-related deaths reported 1973-1998. AAP recommends a ban on manufacture and sale of mobile infant walkers.","excerpt":"\"In 1999, an estimated 8800 children younger than 15 months were treated in hospital emergency departments in the United States for injuries associated with infant walkers. Thirty-four infant walker-related deaths were reported from 1973 through 1998... the American Academy of Pediatrics recommends a ban on the manufacture and sale of mobile infant walkers.\"\n","source_date":"2001-09-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260413173658/https://pubmed.ncbi.nlm.nih.gov/11533353/","calculation_notes":"AAP 2001 cites stair falls as the dominant mechanism behind its ban recommendation. The 34 walker-related deaths over 1973-1998 are overwhelmingly stair-fall fatalities; this is the death-count anchor for stair-fall walker injuries as a class.\n"},{"url":"https://www.canada.ca/en/news/archive/2004/04/minister-pettigrew-announces-ban-baby-walkers.html","title":"Minister Pettigrew announces ban on baby walkers","publisher":"Government of Canada (Health Canada)","source_type":"govt_report","statistic":"Canada banned sale, advertisement and importation of baby walkers effective April 2004; Health Canada cites stair-fall head injuries as the specific mechanism","excerpt":"\"Health Minister Pierre Pettigrew today announced the Government of Canada's immediate prohibition of the sale, advertisement and importation of baby walkers in Canada... 'Canada is the first country in the world to ban the sale of these products.'... Typically, incidents linked to baby walkers involve head injuries that result from falls down stairs.\"\n","source_date":"2004-04-07","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251214151005/https://www.canada.ca/en/news/archive/2004/04/minister-pettigrew-announces-ban-baby-walkers.html","calculation_notes":"Health Canada explicitly names stair-fall head injuries as the cited mechanism for the ban. This is the regulatory anchor for treating the stair-fall pathway as the dominant hazard rather than one of several.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1067673/","title":"Hazards of baby walkers in a European context","publisher":"Injury Prevention (Petridou, Simou, Skondras et al.)","source_type":"peer_reviewed","statistic":"Athens, May 1994-April 1995: walker-related injury rate of 16 per 1,000 person-years of walker use, equivalent to 3.5 per 1,000 babies per year; falls down stairs the dominant mechanism; peak at 9-10 months","excerpt":"\"Baby walkers impart a significant risk of injury from a consumer product that provides no clearly identifiable benefit. The injury rate was 16 per thousand person years of users, or 3.5 per thousand babies per year. The peak age of injury was 9-10 months.\"\n","source_date":"1996-06-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250223023508/https://pmc.ncbi.nlm.nih.gov/articles/PMC1067673/","calculation_notes":"Petridou 1996 is the canonical European empirical baseline. 16 per 1,000 person-years of walker users corresponds to roughly a 1.6% annual injury rate for active users, consistent with US-era data from the 1990s before standards reform. Useful for cross-Atlantic calibration and as a second authoritative source for the stair-fall mechanism claim.\n"}],"comparison_anchors":[{"label":"Toddler stair fall hospitalization (per fall)","lifetime_us_adult":0.027},{"label":"Baby walker injury (any type)","lifetime_us_adult":0.0022},{"label":"Child window/balcony fall serious injury (typical exposure)","lifetime_us_adult":0.0001}],"personal_factor_multipliers":[{"factor":"Stair gate installed at top of accessible stairs","multiplier":0.05,"notes":"Gate physically prevents the mechanism; CPSC F1004 standard governs stair-gate construction"},{"factor":"Single-storey home (no accessible stairs)","multiplier":0.05,"notes":"Removes the stair-fall pathway entirely; residual walker risk reduces to tip-overs and reach-related injuries"},{"factor":"Pre-1997 walker (no stair brake, narrow base)","multiplier":10,"notes":"Sims 2018 documents stair-fall injuries falling roughly 91% from 1990 to 2003 after ASTM F977 introduction"},{"factor":"Walker compliant with US ASTM F977 (post-1997)","multiplier":0.3,"notes":"Stair-fall brakes or a base width above 36 inches engineered out the dominant mechanism"},{"factor":"Caregiver supervising in same room continuously","multiplier":0.5,"notes":"Most stair-fall events occur during brief inattention windows; effect size estimated from case-series patterns"}],"short_label":"Walker stair fall","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1,482 numerator is derived: it is 74.1% of the Sims 2018 figure of 2,001 walker injuries in 2014. Sims does not publish a stair-fall-specific count by year, so this share is assumed constant across the 25-year window. The denominator of 925,000 is the same estimated walker-using infant population used in the sibling entry; it is not a measured value. The figure does not apply to homes without stairs at all, where the residual risk collapses to tip-overs and reach-related injuries. It also overstates the modern US risk for compliant ASTM F977 walkers, which engineered out the dominant mechanism through stair-fall brakes or oversized bases. Secondhand and imported pre-1997 walkers remain the largest residual risk; these are out of scope of the post-2010 standard. The Petridou 1996 European baseline predates standards reform and reflects mid-1990s walker designs.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"A baby walker positioned at the top of a residential staircase, no child visible, viewed from the side, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/baby-walker-stair-fall","api_url":"https://likelier.app/api/fears/baby-walker-stair-fall.json"},{"slug":"glp1-anesthesia-aspiration","question":"If you take a GLP-1 drug, will retained stomach contents cause you to aspirate during general anesthesia?","quick_answer":"A GLP-1 user aspirating during anesthesia happens in about 1 in 610 procedures, which is low, and the drug's added risk is unproven, since GLP-1 and matched non-user rates overlap. Media framing implied a per-procedure risk of several to tens of percent, far above the measured rate.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Press coverage of the 2023 ASA advisory and the November 2024 FDA label warning framed GLP-1 drugs as a serious anesthesia hazard, with headlines urging patients to stop Ozempic or risk aspirating during surgery. The intuitive read is that a GLP-1 user faces a high, near-certain chance of choking on retained stomach contents on the operating table. No survey has measured this specific perception, so the rough estimate is inferred from the alarm in coverage and patient forums, not from polling.","rough_estimate":"Patients and media framing imply a perceived per-procedure aspiration risk on the order of several percent to tens of percent ('don't go under on Ozempic').","kind":"intuition"},"native":{"display":"about 1 in 610 procedures","numerator":143,"denominator":87691,"unit":"GLP-1 users undergoing elective endoscopy under sedation/anesthesia","population":"Adults taking a GLP-1 receptor agonist who undergo an elective procedure requiring general anesthesia or deep sedation"},"normalized":{"lifetime_us_adult":0.00163,"display":"about 1 in 610 per procedure","log_value":-2.788,"assumptions":"Subgroup: an adult on a GLP-1 receptor agonist undergoing one elective procedure requiring general anesthesia or deep sedation. Basis: PER PROCEDURE, not aggregated over a lifetime of procedures. Point estimate 0.00163 is the pooled aspiration rate in GLP-1 users from the largest endoscopy meta-analysis (143/87,691 = 0.00163). The uncertainty band spans the range of real-world cohorts: the lower bound (~0.0008) approaches the general-population GA baseline (~1 in 8,325 = 0.00012 to ~1 in 2,500 = 0.0004) and the lower cohort estimates; the upper bound (~0.008) reflects the surgical cohort figure of 0.8% (IJS 2025). Crucially this is the ABSOLUTE per-procedure aspiration probability, NOT a GLP-1-attributable excess: the matched control rates (0.12%-0.7%) overlap the GLP-1 rates, and pooled odds ratios (1.07-1.23) are not statistically significant. So the headline number is 'how often a GLP-1 user aspirates,' which is low, while 'how much the drug raises that risk' is null-to-unproven.","uncertainty":{"low":0.0008,"high":0.008},"scope":"subgroup_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jamasurgery/fullarticle/2815663","title":"Glucagon-Like Peptide-1 Receptor Agonist Use and Residual Gastric Content Before Anesthesia","publisher":"JAMA Surgery","source_type":"peer_reviewed","statistic":"GLP-1 RA use associated with a 30.5% (95% CI, 9.9%-51.2%) higher prevalence of increased residual gastric content; 56% (35/62) vs 19% (12/62).","excerpt":"GLP-1 RA use was associated with a 30.5% (95% CI, 9.9%-51.2%) higher prevalence of increased RGC","source_date":"2024-03-06","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260512203544/https://jamanetwork.com/journals/jamasurgery/fullarticle/2815663","calculation_notes":"This is the SURROGATE (retained gastric content), not aspiration. Raw prevalence 56% vs 19%; adjusted difference +30.5%. Increased RGC defined as solids, thick liquids, or >1.5 mL/kg clear liquids on ultrasound. Authors note they 'did not directly assess aspiration events, which are rare.' Used to establish the elevated-surrogate half of the perceived-vs-actual gap, not the headline number.","independence_note":"University of Texas Houston cohort; independent of the endoscopy and surgery aspiration cohorts below."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12481593/","title":"GLP-1 Agonists and the Risk of Pulmonary Aspiration during Elective Upper Endoscopy: A Systematic Review and Meta-analysis","publisher":"The Open Respiratory Medicine Journal","source_type":"peer_reviewed","statistic":"Pulmonary aspiration in 143/87,691 (0.16%) GLP-1 users vs 149/122,525 (0.12%) controls; OR 1.23 (95% CI 0.58-2.60; P=0.59), not significant.","excerpt":"143 of 87,691 patients (0.16%) in the GLP-1 agonist group and 149 of 122,525 patients (0.12%) in the placebo group had pulmonary aspiration.","source_date":"2025-06-11","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714013912/https://pmc.ncbi.nlm.nih.gov/articles/PMC12481593/","calculation_notes":"Source of the native counts and the normalized point estimate. 143/87,691 = 0.00163. Control rate 149/122,525 = 0.00122 overlaps, OR 1.23 not significant -> the absolute rate is low and the excess is unproven. Point estimate = 0.00163; log10 = -2.788.","independence_note":"Pooled endoscopy meta-analysis; distinct dataset from the surgical TriNetX/IJS cohorts."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12165584/","title":"Glucagon-like-peptide-1 (GLP-1) receptor agonist use and the risk of pulmonary aspiration in patients undergoing surgery","publisher":"International Journal of Surgery","source_type":"peer_reviewed","statistic":"Aspiration 0.8% in GLP-1 users vs 0.7% non-users (P=0.61); adjusted OR 1.07 (95% CI 0.85-1.34), no association. N=392,065 (15,745 users).","excerpt":"After adjusting for other risk factors for aspiration, there was no association between GLP-1 agonist use and aspiration (OR = 1.07; 95% CI, 0.85–1.34).","source_date":"2025-05-12","source_accessed":"2026-06-13","calculation_notes":"Supplies the upper end of the uncertainty band (0.8% = 0.008). Surgical population (2015-2022), broader than endoscopy, so higher absolute aspiration rate. Confirms the null excess: OR 1.07 not significant.","independence_note":"Separate surgical cohort; conclusion converges with the endoscopy meta despite independent data."},{"url":"https://academic.oup.com/jes/article/9/9/bvaf088/8196040","title":"Glucagon-Like Peptide-1 Receptor Agonists and Peri-Procedural Aspiration Risk","publisher":"Journal of the Endocrine Society","source_type":"peer_reviewed","statistic":"Across cohorts, aspiration ~2.6-8.3 per 1000 in GLP-1 users vs similar in non-users; retained-content risk ratios 2.92-15.14; ASA Oct 29 2024 guidance.","excerpt":"retrospective cohort studies using large real-world evidence databases have not consistently identified a GLP-1RA-associated risk of aspiration/pneumonia for elective surgical and endoscopic procedures","source_date":"2025-07-10","source_accessed":"2026-06-13","calculation_notes":"Review consolidating the field: per-procedure aspiration 2.61-8.32 per 1000 in users (0.0026-0.0083), bracketing the chosen band; retained-content RRs 2.92-15.14 (the elevated surrogate). Source for the multisociety guidance summary and the 'continue most patients, shared decision making' framing.","independence_note":"Narrative review summarizing the primary cohorts above; cited for synthesis and guidance context, not as an independent dataset."},{"url":"https://apm.amegroups.org/article/view/65476/html","title":"Perioperative pulmonary aspiration and regurgitation without aspiration in adults: a retrospective observational study of 166,491 anesthesia records","publisher":"Annals of Palliative Medicine","source_type":"peer_reviewed","statistic":"Perioperative aspiration in 20/166,491 anesthetics (1:8,325); morbidity 1:16,649; mortality 1:55,497.","excerpt":"Among the 166,491 anesthesia records, 20 patients had PA (1:8,325)","source_date":"2021-04-30","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20240806043909/https://apm.amegroups.org/article/view/65476/html","calculation_notes":"General-population baseline aspiration rate under anesthesia (1:8,325 = 0.00012), pre-dating the GLP-1 era. Establishes the lower edge / reference point: GLP-1-user rates (0.0016-0.008) sit near or modestly above this baseline, consistent with the null-excess finding. Used for the lower uncertainty bound and severity context.","independence_note":"Baseline cohort with no GLP-1 exposure measured; independent of all GLP-1 sources."},{"url":"https://www.pharmacytimes.com/view/fda-adds-delayed-gastric-emptying-as-adverse-event-on-semaglutide-label","title":"FDA Adds Delayed Gastric Emptying as Adverse Event on Semaglutide Label","publisher":"Pharmacy Times","source_type":"news_article","statistic":"FDA added a label warning of rare postmarketing reports of pulmonary aspiration in GLP-1 users undergoing anesthesia despite reported fasting (Nov 2024).","excerpt":"postmarking reports showing rare instances of pulmonary aspiration for patients receiving glucagon-like peptide-1 (GLP-1) receptor agonists and undergoing elective surgeries or procedures that require general anesthesia or deep sedation","source_date":"2024-11-14","source_accessed":"2026-06-13","calculation_notes":"Documents the FDA regulatory action (label 'rare instances' language) that drives the public-perception half of the entry. Non-authoritative news source; used only for the FDA/labeling timeline, not for any rate."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11827728/","title":"GLP-1 receptor agonist use does not increase risk of respiratory complications post-endoscopy","publisher":"Endoscopy International Open","source_type":"peer_reviewed","statistic":"TriNetX 1:1 propensity-matched cohort (N=46,948; 23,474 GLP-1 users vs 23,474 controls): aspiration pneumonitis 0.11% vs 0.12%, hazard ratio 0.92 (95% CI 0.54-1.56); pneumonia 0.81% vs 0.79%, hazard ratio 1.01 (95% CI 0.83-1.24). Neither difference reached significance.","excerpt":"GLP-1 RA use before endoscopy was not associated with a higher risk of post-procedure respiratory complications compared with that in patients not prescribed GLP-1 RAs.","source_date":"2025-01-13","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260714014108/https://pmc.ncbi.nlm.nih.gov/articles/PMC11827728/","calculation_notes":"Propensity-matched (1:1) cohort drawn from the TriNetX research network of 83 healthcare organizations. The aspiration-pneumonitis point estimate sits below 1 (HR 0.92), giving a numerically lower rate in GLP-1 users, but the confidence interval crosses the null (0.54-1.56), so this is a further null result rather than a statistically significant protective finding. Used to source and correctly qualify the \"propensity-matched cohort with a lower point estimate\" claim elsewhere in this entry.","independence_note":"Separate TriNetX-derived cohort, independent of the JAMA Surgery, endoscopy meta-analysis, IJS surgical, and baseline-anesthesia sources above."}],"comparison_anchors":[{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.000537},{"label":"Hospital-acquired infection (lifetime, US)","lifetime_us_adult":0.039}],"personal_factor_multipliers":[{"factor":"Active GI symptoms (nausea, vomiting, bloating) at the time of the procedure","multiplier":2,"notes":"The 2024 multisociety guidance classifies GI symptoms, dose-escalation phase, weekly dosing, and higher doses as 'higher risk' for delayed gastric emptying; gastric-ultrasound studies show retained-content risk ratios from 2.92 to 15.14 in such users. Multiplier is a conservative directional estimate for the retained-content surrogate, not a measured aspiration multiplier."},{"factor":"Ultrasound-confirmed empty stomach after a clear-liquid diet / extended fast","multiplier":0.4,"notes":"Point-of-care gastric ultrasound that finds an empty antrum returns the patient to roughly the general-anesthesia baseline; the multisociety guidance treats a documented empty stomach as removing the GLP-1-specific concern. Protective direction; exact factor not measured for aspiration events."}],"short_label":"GLP-1 anesthesia aspiration","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The headline number is the absolute per-procedure aspiration rate for a GLP-1 user (about 1 in 610 in the endoscopy meta-analysis, up to about 1 in 125 in the broader surgical cohort), NOT the amount the drug adds to that risk. The two large real-world cohorts and the endoscopy meta-analysis found matched control rates that overlap the GLP-1 rates, with odds ratios (1.07, 1.23) that are not statistically significant, and a propensity-matched TriNetX endoscopy cohort found a numerically lower rate in GLP-1 users (hazard ratio 0.92, CI crossing the null). So the drug's attributable excess is null-to-unproven even though the surrogate (retained gastric contents) is markedly elevated (56% vs 19% on ultrasound; risk ratios 2.92 to 15.14). Aspiration events are rare, so per-procedure estimates rest on small numerators and wide confidence intervals; the surgical and endoscopy figures differ by procedure type, urgency, and airway management. The number is scoped per procedure, not per lifetime; a person undergoing many procedures accumulates risk roughly linearly. The 8D self-score clears threshold (avg 4.0), with the main soft spots being perception (intuition-only, no survey) and the per-procedure-vs-lifetime scope translation.","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Aspiration under anesthesia on a GLP-1 drug\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/glp1-anesthesia-aspiration","api_url":"https://likelier.app/api/fears/glp1-anesthesia-aspiration.json"},{"slug":"cheap-imported-product-safety","question":"What are the odds of being harmed by an unsafe imported consumer product?","quick_answer":"The rough odds of being harmed by a genuinely defective consumer product are about 1 in 565 over an adult lifetime — and even that is highly uncertain, since no federal system cleanly separates defect injuries from ordinary misuse. Alarming headlines make budget imports feel far more dangerous than the aggregate data can confirm.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The rise of Temu, Shein, and AliExpress as mainstream shopping platforms has rekindled a fear that predates e-commerce: that cheap imported goods are dangerous. Headlines about lead in children's jewelry, PFAS in clothing, exploding hoverboards, and banned crib bumpers still listed for sale create a perception of systemic regulatory failure. Seoul Metropolitan Government testing in 2024 found a Temu children's coat with 622 times the legal limit of phthalate plasticizers. A Toy Association study found 89% of toys purchased from Shein and Temu presented significant safety concerns. The de minimis shipping loophole — allowing packages under $800 to enter the US with minimal inspection — reinforces the sense that these products bypass the safety net entirely. Most consumers who encounter these stories conclude that ordering from budget e-commerce platforms carries a real chance of injury or poisoning, particularly for children.\n","rough_estimate":"Many consumers treat budget e-commerce products as carrying a meaningful risk of injury or toxic exposure","kind":"intuition"},"native":{"display":"~869 injuries from recalled consumer products in 2024 (US PIRG / CPSC data); ~15.1 million total product-related ER visits in 2024","numerator":3,"denominator":100000,"unit":"per year","population":"US residents purchasing from all retail channels including online marketplaces"},"normalized":{"lifetime_us_adult":0.00177,"display":"~1 in 565 lifetime (US adult)","log_value":-2.75,"assumptions":"Estimating harm specifically from \"unsafe imported products\" is difficult because CPSC injury data does not cleanly separate injuries caused by product defects from injuries caused by user behavior, and does not track country of origin consistently in NEISS. The 15.1 million annual ER visits for consumer product injuries (NEISS 2024) include falls from stairs, cuts from kitchen knives, and other use-related injuries that have nothing to do with manufacturing defects or import safety. The PIRG figure of 869 injuries from products recalled in 2024 dramatically undercounts because it includes only injuries reported through the recall process. A rough estimate: the PIRG report attributes the 2024 surge substantially to China (a China-based company was the responsible party for 42 of the 63 CPSC warnings issued in 2024) and to overseas e-commerce channels. If we estimate ~10,000 injuries per year attributable to genuine product safety defects (a fraction of total ER visits, backed by recall reports, CPSC investigations, and incident reports), that gives ~3 per 100,000 per year. Over 59 adult-remaining years: 1 - (1 - 3e-5)^59 ≈ 1.77e-3, or about 1 in 565. This estimate is highly uncertain because the denominator (defect-attributable injuries) is not directly measured by any federal system. The uncertainty band spans an order of magnitude in both directions.\n","uncertainty":{"low":0.0003,"high":0.01},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pirg.org/edfund/resources/safe-at-home-in-2025/","title":"Safe At Home in 2025?","publisher":"US PIRG Education Fund","source_type":"reputable_reference","statistic":"869 injuries and 15 deaths from products recalled in 2024 (the most injuries in eight years); 305 recall announcements plus 63 CPSC warnings (up from 38 in 2023); a China-based company was the responsible party for 42 of the 63 warnings","excerpt":"\"A total of 869 injuries and 15 deaths were connected to products recalled during the year. ... Of the 63 warnings in 2024, a company in China was the manufacturer, exporter or responsible party for 42 of the products.\"\n","source_date":"2025-10-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251023030212/https://pirg.org/edfund/resources/safe-at-home-in-2025/","calculation_notes":"PIRG's annual \"Safe At Home\" report aggregates CPSC recall data for the calendar year (the 2025 edition covers 2024 recalls). The 869 injuries figure covers only injuries reported through the formal recall process, which drastically undercounts total defect-related injuries: most consumers who are harmed by a product do not file a SaferProducts.gov report, and many defective products are never recalled. The report attributes the 2024 surge substantially to China (\"In a word, China,\" per CPSC's acting chair): a China-based company was the manufacturer, exporter, or responsible party for 42 of the 63 CPSC warnings issued in 2024. Used here as the most recent aggregate of recall-linked injury data.\n","independence_note":"PIRG independently aggregates CPSC public recall data; their analysis methodology is separate from CPSC's own annual reports but relies on the same underlying recall and incident data.\n"},{"url":"https://www.cpsc.gov/Research--Statistics/NEISS-Injury-Data","title":"National Electronic Injury Surveillance System (NEISS)","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"15.1 million consumer product-related ER visits estimated in 2024 via probability sample of ~100 US hospitals","excerpt":"\"NEISS injury data are gathered from the emergency departments of approximately 100 hospitals selected as a probability sample of all 5,000+ U.S. hospitals with emergency departments.\"\n","source_date":"2025-03-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032840/https://www.cpsc.gov/Research--Statistics/NEISS-Injury-Data","calculation_notes":"NEISS is the primary federal system for estimating consumer product-related injuries treated in US emergency departments. The 15.1 million figure for 2024 covers all consumer product categories, with \"home structures and construction\" (stairs, floors) accounting for ~4.5 million — injuries unrelated to product defects. NEISS does not code for country of origin or manufacturing defect vs user error, making it impossible to directly extract \"injuries from unsafe imported products\" from the database. The total is used here as a ceiling from which the defect-attributable fraction is estimated downward.\n","independence_note":"NEISS is the authoritative federal injury surveillance system; PIRG's recall-linked injury counts draw from a different data stream (recall reports) and are not a subset of NEISS estimates.\n"},{"url":"https://www.cbc.ca/news/business/marketplace-fast-fashion-chemicals-1.6193385","title":"Experts warn of high levels of chemicals in clothes by some fast-fashion retailers","publisher":"CBC Marketplace","source_type":"news_article","statistic":"A Shein toddler jacket contained nearly 20x Health Canada's safe limit for lead; a Shein purse exceeded the lead threshold by 5x","excerpt":"\"A jacket for toddlers purchased from Shein contained almost 20 times the amount of lead that Health Canada says is safe for children.\"\n","source_date":"2021-10-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260503075851/https://www.cbc.ca/news/business/marketplace-fast-fashion-chemicals-1.6193385","calculation_notes":"CBC Marketplace tested 38 items from several fast-fashion retailers in 2021. Nearly one in five items had elevated levels of lead, PFAS, or phthalates. The Shein toddler jacket finding (20x Health Canada's lead limit) triggered a Health Canada recall and was the most widely cited result. This is investigative journalism, not systematic surveillance, and the sample is small and non-random. It is included because the investigation drove significant public awareness and regulatory action, but it cannot be used to estimate population-level injury rates from fast-fashion chemical exposure.\n","independence_note":"CBC Marketplace commissioned independent laboratory testing through University of Toronto researchers; results are independent of CPSC, Seoul, and EU Safety Gate data.\n"},{"url":"https://www.yahoo.com/news/south-korea-children-coat-temu-140000255.html","title":"South Korea: Children's Coat from Temu Contained 622 Times The Legal Limit for Toxins","publisher":"Yahoo News (Reuters), reporting Seoul Metropolitan Government testing","source_type":"news_article","statistic":"A Temu children's jacket contained 622x the legal limit of phthalate plasticizers, 3.6x the lead limit, and 3.4x the cadmium limit; of 26 children's winter-clothing items tested, 7 contained toxins exceeding Korean safety standards","excerpt":"\"The government tested one Temu jacket that had 622 times the legal limit of phthalate plasticizers, 3.6 times the legal limit of lead and 3.4 times the legal limit of cadmium. Of those 26 pieces, authorities found seven contained higher amounts of toxins like lead, cadmium and phthalate plasticizers.\"\n","source_date":"2024-11-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260228100505/https://www.yahoo.com/news/south-korea-children-coat-temu-140000255.html","calculation_notes":"The Seoul Metropolitan Government, as part of regular procedure, tested 26 items of children's winter clothing from Temu, AliExpress, and Shein; 7 of the 26 contained toxins (lead, cadmium, phthalates) exceeding Korean safety standards. The 622x phthalate exceedance on a single Temu jacket is the most extreme finding in any government testing of these platforms. Korean safety standards for children's products are broadly comparable to EU and US limits. However, product testing identifies hazard (chemical presence above thresholds), not risk (probability of illness at actual exposure duration and dose). A child wearing a coat with elevated phthalates is not the same as a child ingesting phthalates at toxic doses — dermal absorption rates for phthalates from textiles are generally low, though not negligible for infants who mouth fabric.\n","independence_note":"Seoul Metropolitan Government testing is independent of CBC Marketplace, CPSC, and EU Safety Gate data; uses Korean safety standards.\n"},{"url":"https://itif.org/publications/2025/04/02/chinas-state-backed-e-commerce-platforms-threaten-american-consumers-us-technology-leadership/","title":"China's State-Backed E-Commerce Platforms Threaten American Consumers and U.S. Technology Leadership","publisher":"Information Technology and Innovation Foundation (ITIF)","source_type":"reputable_reference","statistic":"The Section 321 de minimis exemption lets shipments valued at $800 or less bypass customs duties, making it easier for Chinese e-commerce platforms to flood the US market with low-cost goods","excerpt":"\"Section 321 of the Tariff Act of 1930 allows importers to bypass customs duties on packages valued at $800 or less, making it easier for Chinese e-commerce platforms to flood the U.S. market with low-cost goods without contributing to the federal revenue through duties.\"\n","source_date":"2025-04-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260510233559/https://itif.org/publications/2025/04/02/chinas-state-backed-e-commerce-platforms-threaten-american-consumers-us-technology-leadership/","calculation_notes":"ITIF's policy analysis documents the structural enforcement gap created by the Section 321 de minimis threshold ($800 per shipment), which lets Chinese e-commerce platforms ship low-cost goods into the US while bypassing customs duties and routine inspection. (The widely cited de minimis package volume — roughly 140 million shipments in 2013 rising to over 1 billion by 2023 — comes from CBP and Congressional Research Service data, not from this ITIF report; see the prose body.) The practical consequence is that the recall-based injury data (PIRG's 869 figure) almost certainly undercounts injuries from products that were never subject to border inspection, never recalled, and whose incidents were never reported. The enforcement gap is real but its size is not quantified — it is a known unknown in the injury estimate.\n","independence_note":"ITIF is an independent technology policy think tank; their analysis draws on CPSC and CBP data but provides independent policy recommendations.\n"},{"url":"https://www.toyassociation.org/ta/PressRoom2/News/2025_News/study-highlights-the-importance-of-shopping-from-trusted-toy-brands-and-retailers-this-cyber-monday.aspx","title":"New Study Highlights the Importance of Shopping from Trusted Toy Brands & Retailers this Cyber Monday","publisher":"The Toy Association (via PR Newswire)","source_type":"reputable_reference","statistic":"In CPSC-accredited-lab testing of toys purchased from Temu and Shein, 89% had at least one safety failure and 65% carried misleading or incorrect safety information at the point of sale","excerpt":"\"89% of the toys purchased specifically from Temu and Shein had at least one safety failure ... 65% of the Temu and Shein toys provided misleading or incorrect safety information at the point of sale.\"\n","source_date":"2025-12-01","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20251201182817/https://www.toyassociation.org/ta/PressRoom2/News/2025_News/study-highlights-the-importance-of-shopping-from-trusted-toy-brands-and-retailers-this-cyber-monday.aspx","calculation_notes":"The Toy Association commissioned CPSC-accredited third-party lab testing of toys bought from Temu and Shein for compliance with US federal toy-safety standards (small parts, phthalates, flammability). 89% had at least one safety failure (failed mandatory testing and/or labeling, age-grading, or warnings). Used as illustrative evidence of elevated defect rates on budget cross-border e-commerce platforms; it is a targeted sample of two platforms, not a population injury rate. The earlier body claim of \"over 70% failing at least one laboratory safety test\" could not be verbatim- confirmed in the press release and was replaced with the verbatim 89% / 65% figures.\n","independence_note":"The Toy Association is a US toy-industry trade group; its testing was performed by an independent CPSC-accredited lab. Directionally consistent with, but methodologically separate from, the CPSC recall data and Seoul government testing cited above.\n"},{"url":"https://www.wiley.law/alert-Biden-Administration-Announces-Changes-to-De-Minimis-Trade-Exemptions-to-Address-Unfair-and-Unsafe-Imports-into-the-United-States","title":"Biden Administration Announces Changes to De Minimis Trade Exemptions to Address Unfair and Unsafe Imports into the United States","publisher":"Wiley Rein LLP","source_type":"reputable_reference","statistic":"De minimis shipments grew from 140 million in 2013 to over 1 billion in 2023 (nearly 4 million a day), per a September 2024 White House announcement","excerpt":"\"There has been a substantial increase in the number of imports entering the United States that use the de minimis exemption, increasing from 140 million shipments in 2013 to over 1 billion a year in 2023 (almost 4 million shipments a day).\"\n","source_date":"2024-09-13","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260304064023/https://www.wiley.law/alert-Biden-Administration-Announces-Changes-to-De-Minimis-Trade-Exemptions-to-Address-Unfair-and-Unsafe-Imports-into-the-United-States","calculation_notes":"Added to ground the shipment-volume figure in the body prose, which previously attributed the number directly to \"Customs and Border Protection and Congressional Research Service data\" without an actual CBP or CRS citation in this entry's source list. This law-firm client alert quotes the same growth figure (140 million to over 1 billion shipments/year) from a September 2024 White House announcement on de minimis reform; used here as the verifiable source for the volume-growth figure that anchors the enforcement-gap narrative.\n","independence_note":"Wiley Rein is a law firm summarizing a federal policy announcement; independent of ITIF's policy analysis and of the PIRG/CPSC recall data cited elsewhere in this entry.\n"},{"url":"https://www.cbsnews.com/texas/news/more-than-500000-hoverboards-recalled-after-fires/","title":"More Than 500,000 Hoverboards Recalled After Fires","publisher":"CBS News","source_type":"news_article","statistic":"CPSC recalled more than 500,000 hoverboards in 2016 after 99 reports of battery packs exploding or catching fire and at least 18 reported injuries (burns to neck, legs, or arms)","excerpt":"\"More than 500,000 hoverboards are being recalled after some of the motorized scooters overheated, burned riders and damaged property. ... There have been 99 reports to the Consumer Product Safety Commission of hoverboard battery packs that exploded or caught fire. ... At least 18 injuries were reported, such as burns to the neck, legs or arms.\"\n","source_date":"2016-07-06","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714012941/https://www.cbsnews.com/texas/news/more-than-500000-hoverboards-recalled-after-fires/","calculation_notes":"Added to ground the hoverboard historical-precedent paragraph in the body prose. The body previously stated \"at least 99 injuries,\" conflating the 99 CPSC incident reports (battery packs exploding or catching fire) with the actual injury count, which this source puts at \"at least 18.\" The body also cited \"60 fires across 20 states\" and \"$2 million in property damage,\" figures that could not be verbatim-confirmed from an accessible source within this review; those specific figures were removed and the prose corrected to match the verified 99-incident-report / 18-injury breakdown.\n","independence_note":"Contemporary press coverage of CPSC's 2016 recall announcement; independent of the Temu/Shein-era sources (PIRG, Seoul, Toy Association) cited elsewhere in this entry; used only as historical precedent context.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Accidental fall death (lifetime, US adult)","lifetime_us_adult":0.0096}],"personal_factor_multipliers":[{"factor":"Household with children under 6","multiplier":5,"notes":"Children are disproportionately harmed by product safety defects due to mouthing behavior, smaller body mass (higher dose-per-kg for chemical exposure), and use of product categories with higher recall rates (toys, nursery products, children's clothing). The Toy Association found 89% of toys from Shein and Temu had safety concerns. CPSC nursery product recalls are heavily weighted toward imported products.\n"},{"factor":"Purchases exclusively from established US retailers","multiplier":0.3,"notes":"Major US retailers (Target, Walmart brick-and-mortar, Costco) maintain supplier compliance programs and are subject to CPSC enforcement. Product defects still occur but at lower rates than direct-from-manufacturer e-commerce channels.\n"},{"factor":"Heavy Temu/Shein/AliExpress buyer (20+ items per year)","multiplier":3,"notes":"More purchases from platforms with documented compliance gaps increases the probability of encountering a defective or non-compliant product. The de minimis shipping loophole means most such purchases receive no border safety inspection.\n"}],"short_label":"Unsafe imported products","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"property","valence":"negative","caveats":"This entry addresses physical injury and toxic chemical exposure from consumer product safety defects, not dissatisfaction with product quality, counterfeiting, or data privacy concerns associated with budget e-commerce platforms. The normalized probability is a rough estimate because no federal surveillance system cleanly tracks injuries attributable to imported product defects as distinct from user-error injuries with any consumer product. The 1-in-565 lifetime figure includes all product-defect injuries, not just those from imported goods or e-commerce platforms specifically. Most consumer product injuries treated in emergency departments (falls from furniture, cuts from tools, burns from cookware) result from foreseeable use, not manufacturing defects. The true rate of injury from genuinely unsafe imported products is almost certainly lower than the headline figure but is not directly measurable from available data.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A small cardboard shipping box on a neutral surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/cheap-imported-product-safety","api_url":"https://likelier.app/api/fears/cheap-imported-product-safety.json"},{"slug":"driver-killing-pedestrian","question":"What are the odds that a driver will kill a pedestrian in their lifetime?","quick_answer":"About 1 in 552 over a driving lifetime for a US licensed driver. That is well above the near-zero risk most drivers carry in their heads — the perception gap here runs opposite to most fears, understating rather than overstating.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most drivers do not think of themselves as carrying a pedestrian-fatality risk. Road-safety conversations focus almost entirely on the victim side — the person struck — while the driver's perspective is treated as an administrative or legal matter rather than a probability anyone should plan around. When drivers do think about it, the mental model is usually \"that happens to bad drivers or drunk drivers, not me.\" There is no widely cited survey measuring how drivers estimate this risk, and the topic rarely appears in public risk-perception research. The perception gap here runs opposite to most entries on this site: the actual probability is almost certainly higher than the number drivers carry in their heads, because the number most carry is effectively zero.\n","rough_estimate":"most drivers would say near zero — a risk for other, worse drivers","kind":"intuition"},"native":{"display":"~7,314 pedestrian deaths per year across ~238 million licensed drivers","numerator":7314,"denominator":237656000,"unit":"per year","population":"US licensed drivers, all ages"},"normalized":{"lifetime_us_adult":0.00181,"display":"~1 in 552 lifetime (US licensed driver)","log_value":-2.742,"assumptions":"NHTSA FARS 2023 Final Data reports 7,314 pedestrian fatalities. FHWA Highway Statistics 2023 counts 237,656,000 licensed drivers in the US. Annual rate per driver: 7,314 ÷ 237,656,000 ≈ 3.08e-5. Compounded over 59 remaining adult years: 1 − (1 − 3.08e-5)^59 ≈ 0.00181, or roughly 1 in 552. This is an involvement rate, not a fault rate: some pedestrian deaths involve drivers who had no reasonable ability to avoid the collision (e.g., pedestrian entered a highway at night). Fault determinations are made case-by-case; this entry measures the probability of being the driver in a fatal pedestrian crash, regardless of culpability. The uncertainty band reflects the range from the 2024 preliminary low (~7,080 deaths) to the 2022 peak (7,593 deaths) applied to the same driver denominator.\n","uncertainty":{"low":0.0014,"high":0.0022},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813727","title":"Traffic Safety Facts 2023 Data: Pedestrians (DOT HS 813 727)","publisher":"National Highway Traffic Safety Administration, National Center for Statistics and Analysis","source_type":"govt_report","statistic":"7,314 pedestrians killed in US traffic crashes in 2023, a 3.7% decrease from 7,593 in 2022; pedestrians accounted for 18% of all traffic fatalities","excerpt":"\"In 2023 there were 7,314 pedestrians killed in traffic crashes, a 3.7-percent decrease from the 7,593 pedestrian fatalities in 2022.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260518084739/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813727","calculation_notes":"NHTSA FARS 2023 Final Data is the authoritative federal count of pedestrian fatalities derived from police crash reports across all 50 states. The 7,314 figure is divided by FHWA's 237,656,000 licensed drivers to get the annual per-driver rate of 3.08e-5. Compounded over 59 adult years: 1 − (1 − 3.08e-5)^59 ≈ 0.00181.\n"},{"url":"https://www.fhwa.dot.gov/policyinformation/statistics/2023/dl1c.cfm","title":"Table DL-1C: Licensed Drivers by State — Highway Statistics 2023","publisher":"Federal Highway Administration, Office of Highway Policy Information","source_type":"govt_report","statistic":"237,656,000 licensed drivers in the United States in 2023","excerpt":"\"Total licensed drivers in the United States in 2023: 237,656,000.\"\n","source_date":"2024-12-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505053340/https://www.fhwa.dot.gov/policyinformation/statistics/2023/dl1c.cfm","calculation_notes":"FHWA Table DL-1C is the definitive federal count of licensed drivers compiled from state motor vehicle agency records. Used as the denominator in the annual per-driver rate calculation: 7,314 ÷ 237,656,000 ≈ 3.08e-5 per driver-year.\n"},{"url":"https://www.ghsa.org/news/drivers-hit-and-killed-more-7500-pedestrians-last-year-most-1981-new-projection-shows","title":"Drivers Hit and Killed More Than 7,500 Pedestrians Last Year, Most Since 1981, New Projection Shows","publisher":"Governors Highway Safety Association (GHSA)","source_type":"reputable_reference","statistic":"Projected 7,508 pedestrian deaths in 2022, the most since 1981","excerpt":"\"Drivers struck and killed at least 7,508 people walking in the United States in 2022 – the most pedestrian deaths since 1981.\"\n","source_date":"2023-06-22","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251015112705/https://www.ghsa.org/news/drivers-hit-and-killed-more-7500-pedestrians-last-year-most-1981-new-projection-shows","calculation_notes":"GHSA's preliminary state-reported projection (7,508) for 2022 is close to, though not identical to, NHTSA's later FARS final count of 7,593 for the same year (see the Traffic Safety Facts 2023 Data source above); both support characterizing 2022 as the highest US pedestrian fatality count since 1981. Used solely to source the \"highest since 1981\" framing in the body text, not as an alternative basis for the lifetime-probability calculation.\n"}],"comparison_anchors":[{"label":"Death as a pedestrian (lifetime, US adult)","lifetime_us_adult":0.00124},{"label":"Death in a car crash as occupant (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Dying in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"drives primarily in dense urban areas with high pedestrian volume","multiplier":2.5,"notes":"Urban environments account for roughly 75% of pedestrian fatalities despite lower speeds; exposure to pedestrians is the dominant variable."},{"factor":"drives primarily in rural or suburban low-pedestrian environments","multiplier":0.4,"notes":"Lower pedestrian exposure substantially reduces the annual probability even if road speeds are higher."},{"factor":"drives frequently at night","multiplier":2,"notes":"Roughly 75% of fatal pedestrian crashes occur in dark conditions; nighttime driving doubles the pedestrian-involvement exposure rate."},{"factor":"drives an SUV or light truck","multiplier":1.3,"notes":"Higher, blunter front profiles increase fatality rates per pedestrian strike compared with sedans; IIHS and GHSA data document the differential."}],"short_label":"Driver kills pedestrian","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This figure is an involvement rate, not a fault rate. A driver who strikes and kills a pedestrian who ran a red light in darkness is counted in the NHTSA FARS total. Culpability determinations are legal and situational; the probability on this page answers the question \"how likely is any licensed driver to be the vehicle driver in a fatal pedestrian crash?\" not \"how likely am I to negligently kill someone.\" The two questions have different answers but the same order of magnitude, because even crashes where the pedestrian bears primary fault typically involve a driver who could have been driving more defensively. The 7,314 figure for 2023 is a 3.7% decline from the 2022 peak of 7,593 but remains roughly 40% above the pre-2015 baseline. Hit-and-run drivers are embedded in the numerator — roughly one in four pedestrian fatalities involves a driver who fled — meaning the true driver cohort is larger than the detected cohort.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"An empty steering wheel seen from the driver's seat, with a faint crosswalk stripe visible through the windshield, flat vector illustration."},"canonical_url":"https://likelier.app/driver-killing-pedestrian","api_url":"https://likelier.app/api/fears/driver-killing-pedestrian.json"},{"slug":"infant-fall-from-furniture","question":"What are the odds of serious injury when an infant falls from furniture (sofa, bed, changing table)?","quick_answer":"Over a US infant's first two years, the odds of a serious injury from a fall off furniture are about 1 in 500, and among infants who do reach an emergency department after a bed fall about 9% have a significant injury. The felt probability runs well ahead of that surveillance data.\n","category":"kids","tags":["infant","household"],"no_reliable_estimate":false,"perceived":{"description":"Infant furniture falls do not show up on any published fear survey, but they are the archetypal \"parenting panic\" event of the first two years. The mental image most new parents carry is a soft sofa, a turned back, a thud, and a silent second or two that plays back in memory for weeks afterwards. The fear attaches specifically to the moment of the fall rather than to the medical outcome, which is part of why the felt probability of something serious happening runs well ahead of the surveillance data. Parenting forums and pediatrician visits reliably surface this as one of the top reasons new parents call a nurse line in the first year.\n","rough_estimate":"Most first-time parents describe a furniture fall as one of the top one or two events they fear on a given day","kind":"intuition"},"native":{"display":"~9% of ED-presenting infant bed falls sustain a significant injury (Kokulu 2021)","numerator":135,"denominator":1439,"unit":"per ED-presenting infant bed fall","population":"infants <1 year presenting to emergency departments after a fall from bed or similar furniture"},"normalized":{"lifetime_us_adult":0.002,"display":"~1 in 500 per US infant over the 0-2 window (serious injury from a household furniture fall)","log_value":-2.7,"assumptions":"Likelier normally reports lifetime-US-adult probabilities, but this entry is scoped to the first two years of life for a single US infant. The headline is the per-infant probability of a \"serious\" fall injury — defined here as a fall from household furniture resulting in an ED visit with a clinically significant diagnosis (concussion, skull fracture, intracranial bleed) or hospital admission. The denominator-side anchor is Solaiman et al. (2023), who estimated 3,414,007 ED visits for bed- and sofa-related injuries among US children under 5 between 2007 and 2021, averaging 115.2 injuries per 10,000 persons annually, with infants (<1 year) hospitalized 1.58 times more often than older children and about 4% of all such visits ending in admission. The severity-side anchor is Kokulu et al. (2021), who followed 1,439 consecutive infant bed-fall presentations and found 9.4% with a \"significant injury\" (skull fracture 4.1%, intracranial hemorrhage or cerebral contusion 2.1%, the remainder other fractures or concussive findings). Working only from the two sourced anchors: Solaiman's 115.2 bed/sofa ED visits per 10,000 US under-5s per year (≈1.15% annually), compounded across the two-year window and weighted toward the infant year where the rate is highest, puts the cumulative per-child probability of a bed/sofa ED visit over 0-2 in the low single-digit percent range (roughly 2-3%). Of those ED-presenting falls, Kokulu's 9.4% carry a clinically significant injury and Solaiman's ~4% are hospitalized, so the per-infant cumulative probability of a serious fall injury lands on the order of 1 in 500 across the 0-2 window — with genuine uncertainty from 1 in 1,000 (hospitalization-only definition) up to 1 in 200. This is well below the earlier 1-in-100 figure, which could not be reproduced from either documented derivation path. Death is a separate question: unintentional fall deaths in US infants run around 1-2 per 100,000 infant-years, or roughly 1 in 50,000 across the 0-2 window — roughly two orders of magnitude below the serious-injury rate.\n","uncertainty":{"low":0.001,"high":0.005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/33046255/","title":"Characteristics of injuries among infants who fall from bed","publisher":"Injury (Elsevier) — Kokulu K, Algın A, Özdemir S, Akça HŞ","source_type":"peer_reviewed","statistic":"Among 1,439 infants <1 year presenting to ED after a fall from bed, 135 (9.4%) had a significant injury; 59 (4.1%) skull fracture; 30 (2.1%) traumatic brain injury (intracranial hemorrhage / cerebral contusion); median age 7 months","excerpt":"\"There were significant injuries for 135 (9.4%) infants. The most common fracture was skull fracture (n = 59, 4.1%), followed by proximal fracture of the upper extremities (n = 26, 1.8%). Traumatic brain injury featured for 30 (2.1%) infants.\"\n","source_date":"2021-02-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420041858/https://pubmed.ncbi.nlm.nih.gov/33046255/","calculation_notes":"Kokulu et al. supply the severity-side anchor: among ED-presenting infant bed falls, 9.4% carry a clinically significant injury. The 4.1% skull-fracture rate and the 2.1% intracranial-hemorrhage rate populate the regional_breakdown row for \"Skull fracture or intracranial bleed\" after roughly halving to reflect that general-population infant falls are lower-severity than the ED-presenting subset captured here. The median age of 7 months is consistent with the \"6-12 month\" peak-risk window cited in personal_factor_multipliers.\n","independence_note":"Single-center Turkish trauma-registry sample; generalizes to US infants at the pattern level (mechanism, age distribution, injury types) but not necessarily at the absolute-rate level, which is why Solaiman et al. below is used as the US-denominator anchor.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37027936/","title":"Sofa and bed-related pediatric trauma injuries treated in United States emergency departments","publisher":"American Journal of Emergency Medicine — Solaiman RH, Navarro SM, Irfanullah E, Zhang J, Tompkins M, Harmon J Jr","source_type":"peer_reviewed","statistic":"3,414,007 ED visits for bed/sofa-related injuries among US children <5 (2007-2021); 115.2 per 10,000 annually; ~4% hospitalized; infants <1 year 1.58× more likely to be hospitalized than older children; 67% rise in under-1 incidence 2007-2021","excerpt":"\"An estimated 3,414,007 children aged <5 years were treated for bed and sofa-related injuries in emergency departments (EDs) in the United States from 2007 through 2021, averaging 115.2 injuries per 10,000 persons annually.\"\n","source_date":"2023-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420041931/https://pubmed.ncbi.nlm.nih.gov/37027936/","calculation_notes":"Solaiman et al. provide the US-denominator anchor. 115.2 per 10,000 under-5s per year ≈ 1.15% of children under 5 visit an ED for a bed/sofa injury in any given year; compounded across the two-year window and weighted toward the infant year (where the rate is highest and rising), the cumulative per-child probability of a bed/sofa ED visit over 0-2 is in the low single-digit percent range (~2-3%). Multiplying by Kokulu's 9.4% clinically-significant fraction (or, for the narrower hospital-admission definition, Solaiman's ~4% admission rate, higher still for under-1s at the 1.58× odds ratio) gives a per-infant serious-injury probability on the order of 1 in 500 across the 0-2 window, used as the headline normalized figure. The uncertainty band (1 in 1,000 to 1 in 200) spans the two severity definitions and the cumulative-rate estimate.\n","independence_note":"Built on CPSC's National Electronic Injury Surveillance System (NEISS), the same surveillance pipeline that underlies the Chaudhary et al. trauma-registry analysis below; treat as partially dependent corroboration rather than two fully independent estimates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5893510/","title":"Pediatric falls ages 0-4: understanding demographics, mechanisms, and injury severities","publisher":"Injury Epidemiology (Springer) — Chaudhary S, Figueroa J, Shaikh S, Mays EW, Bayakly R, Javed M, Smith ML, Moran TP, Rupp J, Nieb S","source_type":"peer_reviewed","statistic":"1,086 trauma-registry cases ages 0-4; 63.3% mild, 31.7% moderate, 5.1% severe by ISS; 177 bed falls (49.7% <1 year), 58 couch falls, 54 chair falls, 7 changing table falls; skull fracture and intracranial hemorrhage rates highest in under-1s","excerpt":"\"63.3% (n = 687) of patients had a mild ISS, 31.7% (n = 344) had moderate ISS, and 5.1% (n = 55) had severe ISS.\"\n","source_date":"2018-04-09","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173647/https://pmc.ncbi.nlm.nih.gov/articles/PMC5893510/","calculation_notes":"Chaudhary et al. analyze the high-severity subset — children serious enough to land in a Level 1 pediatric trauma registry rather than a general ED. Even inside that already-selected pool, only about 5% carry a \"severe\" Injury Severity Score, and the overwhelming majority of bed, couch, chair, and changing-table falls are mild. This is the core empirical basis for the debunked myth framing: the severity distribution of infant furniture falls, even conditional on reaching a trauma center, is dominated by the mild end.\n","independence_note":"Georgia trauma-system data; overlaps methodologically with NEISS-based studies at the surveillance layer. Used as a severity-distribution cross-check rather than as an independent incidence estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6927527/","title":"Fall-related traumatic brain injury in children ages 0-4 years","publisher":"Journal of Safety Research — Haarbauer-Krupa J, Haileyesus T, Gilchrist J, Mack KA, Law CS, Joseph A","source_type":"peer_reviewed","statistic":"~139,000 US children ages 0-4 treated in EDs annually for fall-related TBI; 83.5% at home; furniture (especially beds) ranked second among implicated product categories for ages 1-4","excerpt":"\"The majority of the fall-related TBI in this cohort occurred at home, related to surfaces, structures and fixtures, furniture, and baby products.\"\n","source_date":"2019-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420042007/https://pmc.ncbi.nlm.nih.gov/articles/PMC6927527/","calculation_notes":"Haarbauer-Krupa et al. (CDC Injury Center) anchor the traumatic-brain-injury slice specifically. ~139,000 annual ED visits for fall-related TBI in US under-5s, against a population of ~18-19 million, is roughly 7 per 1,000 children annually. Compounded across the 0-2 window and weighted toward the infant year (where TBI risk and the most-mobile developmental stage converge), this is consistent with the 0.1% skull-fracture/intracranial-bleed row in regional_breakdown.\n","independence_note":"CDC analysis of NEISS-AIP data; shares the NEISS pipeline with Solaiman et al. above. Treated as a TBI-specific cut of the same underlying surveillance dataset rather than an independent estimate.\n"}],"comparison_anchors":[{"label":"SIDS, per US infant (first year)","lifetime_us_adult":0.00014},{"label":"Toddler choking to death on food, per US child 0-4","lifetime_us_adult":0.00002},{"label":"Accidental fall death, lifetime US adult","lifetime_us_adult":0.0074},{"label":"Death in a car crash, lifetime US","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Any fall from furniture, typical infant 0-2","probability":0.5,"notes":"Survey estimates consistently put the share of parents who report at least one infant furniture fall around half. No single authoritative denominator exists because the vast majority of events never reach a clinician.\n"},{"region":"Serious injury (ED visit, clinically significant) from infant fall","probability":0.002,"notes":"Derived from Solaiman 2023's 115 per 10,000 annual under-5 ED rate for bed/sofa injuries, concentrated in the infant year and compounded across the 0-2 window to a ~2-3% cumulative ED-visit rate, times Kokulu 2021's 9.4% clinically-significant fraction (or Solaiman's ~4% hospitalization fraction for the narrower definition) — on the order of 1 in 500.\n"},{"region":"Skull fracture or intracranial bleed","probability":0.001,"notes":"Roughly one-tenth of the ED-visit rate. Kokulu 2021 reports 4.1% skull fracture and 2.1% intracranial hemorrhage among ED-presenting infant bed falls — but ED-presenting falls are already a selected subset, so the per-infant-population rate is an order of magnitude lower.\n"},{"region":"Death from household fall, infant","probability":0.00002,"notes":"CDC WISQARS infant unintentional fall death rates run around 1-2 per 100,000 infant-years; compounded across the 0-2 window, roughly 1 in 50,000 per US infant.\n"}],"personal_factor_multipliers":[{"factor":"fall from changing table or higher","multiplier":3,"notes":"Changing tables sit at roughly waist height (~36 inches) — above the threshold where head-first falls onto a hard surface carry meaningfully higher skull-fracture rates in the Chaudhary and Kokulu cohorts.\n"},{"factor":"onto hard surface (tile, hardwood)","multiplier":2,"notes":"Surface type is the single largest physical moderator in the surveillance data; Chaudhary et al. found skull fractures concentrated in hard-surface landings.\n"},{"factor":"carpet / rug landing","multiplier":0.4},{"factor":"age 6-12 months (most mobile, least aware)","multiplier":1.5,"notes":"Kokulu et al. report a median age of 7 months (IQR 6-9) for ED-presenting infant bed falls — the peak risk window opens with rolling and closes when protective self-righting reflexes develop late in the first year.\n"}],"short_label":"Infant fall","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline number is per-infant cumulative across the first two years, not a per-fall or per-ED-visit rate. Most infant furniture falls never result in a clinician visit at all; roughly 1.15% of US under-5s visit an ED for a bed or sofa injury in any given year (Solaiman 2023), which compounds across the 0-2 window to a cumulative bed/sofa ED-visit rate of a few percent per child; inside that already-selected pool, about 9% carry a clinically significant injury and about 4% are hospitalized (Kokulu 2021; Solaiman 2023), and among the highest- severity trauma-registry subset only around 5% score \"severe\" on the Injury Severity Scale (Chaudhary 2018). The cumulative-times-severity product puts the serious fall-injury probability on the order of 1 in 500 per US infant across the 0-2 window (uncertainty roughly 1 in 1,000 to 1 in 200), with death a separate and far rarer question (roughly 1 in 50,000 per infant). The data tracks bed/sofa/chair/changing-table falls specifically and excludes stair falls, window falls, baby-walker falls, and baby-carrier falls, each of which has its own mortality and severity profile. The entry also excludes abuse presenting as claimed accidental fall, which is a known confounder in the clinical literature and part of why ED clinicians take infant head injuries seriously even when the reported mechanism sounds benign. Red flags that ED guidance treats as reasons to seek immediate evaluation — persistent vomiting, unusual sleepiness, pupil asymmetry, seizure activity — are not reproduced here because Likelier publishes numbers, not clinical advice.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale rounded sofa cushion with a soft thin blanket folded over one edge, viewed from a low angle against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/infant-fall-from-furniture","api_url":"https://likelier.app/api/fears/infant-fall-from-furniture.json"},{"slug":"civilian-war-casualty-ukraine","question":"What is the probability of a civilian being killed or seriously injured during a Ukraine-scale conventional conflict over five years?","quick_answer":"For a civilian in Ukrainian-controlled territory, the confirmed chance of being killed or seriously injured is about 1 in 500 over a five-year conflict — death alone closer to 1 in 1,800. That is roughly ten times lower than the 1 in 50 many distant observers assume.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"People in countries not currently experiencing war typically have poorly calibrated intuitions about civilian casualty rates in conventional conflicts. Media coverage of major attacks and dramatic imagery from Ukraine drives availability-heuristic overestimates for those far from the war, while proximity and familiarity with specific incidents can cause underestimation in affected populations who normalize gradual attrition. No high-quality survey specifically asks populations to estimate the per-civilian probability of being killed or injured in a Ukraine-type conflict, so the perceived side is editorial intuition. Rough public estimates tend to cluster vaguely around \"very dangerous\" without distinguishing proximity to the front, city versus rural, or duration of exposure.\n","rough_estimate":"~1 in 50 feels plausible to many Western observers; actual verified rate is closer to 1 in 500 over five years","kind":"intuition"},"native":{"display":"51,924 civilians (14,383 killed + 37,541 injured) confirmed by OHCHR through October 2025, against ~37 million residents remaining in Ukrainian-controlled territory","numerator":51924,"denominator":37000000,"unit":"over 3.5-year conflict (Feb 2022 to Oct 2025)","population":"Civilians residing in Ukrainian-controlled territory during the full-scale invasion, excluding the approximately 6-7 million who fled abroad"},"normalized":{"lifetime_us_adult":0.002004,"display":"~1 in 500 over a five-year conflict","log_value":-2.698,"assumptions":"The UN Human Rights Monitoring Mission in Ukraine (HRMMU/OHCHR) confirmed 14,383 civilians killed and 37,541 injured from February 24, 2022 through October 2025, a period of approximately 3.5 years. The HRMMU explicitly acknowledges that \"the real number is higher\" due to verification lag and restricted access in Russian-occupied territories. Native numerator: 51,924 (killed + injured). Native denominator: approximately 37 million residents who remained in Ukrainian-controlled territory, derived by subtracting approximately 6-7 million refugees who left the country from a pre-war Kyiv-controlled population of roughly 44 million (UN/UNHCR refugee figures, noting ~4.5 million returns partially offset the outflow). The 3.5-year rate of 51,924 / 37,000,000 = 0.001403. Linearly extrapolated to 5 years: 51,924 × (5/3.5) / 37,000,000 = 74,177 / 37,000,000 = 0.002004. The linear extrapolation is conservative because 2025 showed a 31% increase in casualties over 2024 (HRMMU 2025 annual report), suggesting the rate has been accelerating rather than staying flat. The normalized scope is subgroup_lifetime because this is a conflict-period probability specific to the civilian-in-war subgroup rather than a general US adult lifetime risk. Killed-only 5-year rate: 14,383 × (5/3.5) / 37,000,000 = 0.000556 (~1 in 1,800 for death alone).\n","uncertainty":{"low":0.0012,"high":0.005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://ukraine.ohchr.org/en/Ukraine-s-Civilians-Face-Daily-Death-and-Injury-Amid-Intense-Attacks-UN-Human-Rights-Monitors-Say","title":"Ukraine's Civilians Face Daily Death and Injury Amid Intense Attacks, UN Human Rights Monitors Say","publisher":"UN Human Rights Monitoring Mission in Ukraine (HRMMU/OHCHR)","source_type":"govt_report","statistic":"Since Russia launched its full-scale invasion of Ukraine in February 2022 through October 2025, HRMMU documented at least 14,383 civilians killed, including 738 children, and 37,541 injured, including 2,318 children.","excerpt":"\"Since Russia launched its full-scale invasion of Ukraine in February 2022, HRMMU has documented at least 14,383 civilians killed, including 738 children, and 37,541 injured, including 2,318 children.\"\n","source_date":"2025-10-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260315223308/https://ukraine.ohchr.org/en/Ukraine-s-Civilians-Face-Daily-Death-and-Injury-Amid-Intense-Attacks-UN-Human-Rights-Monitors-Say","calculation_notes":"This is the primary numerator source. 14,383 killed + 37,541 injured = 51,924 total confirmed civilian casualties through approximately October 2025, covering 3.5 years of full-scale conflict. OHCHR explicitly states these are minimum confirmed figures and the real totals are higher. Used as the basis for native numerator and for the 3.5-year rate, which is then linearly extrapolated to 5 years.\n","independence_note":"HRMMU is the UN's own monitoring mission operating within Ukraine. It draws on field verification teams, court records, medical records, and on-site interviews. Its methodology is distinct from Ukrainian government figures and from media-based counts such as Airwaves or ACLED.\n"},{"url":"https://ukraine.ohchr.org/en/2025-deadliest-year-for-civilians-in-Ukraine-since-2022-UN-human-rights-monitors-find","title":"2025 deadliest year for civilians in Ukraine since 2022, UN human rights monitors find","publisher":"UN Human Rights Monitoring Mission in Ukraine (HRMMU/OHCHR)","source_type":"govt_report","statistic":"In 2025, conflict-related violence killed 2,514 civilians and injured 12,142 — 31% more killed and injured than in 2024 (2,088 killed; 9,138 injured).","excerpt":"\"The total number of killed and injured civilians in 2025 was 31 per cent higher than in 2024 (2,088 killed; 9,138 injured) and 70 per cent higher than in 2023 (1,974 killed; 6,651 injured).\"\n","source_date":"2026-01-12","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260522070156/https://ukraine.ohchr.org/en/2025-deadliest-year-for-civilians-in-Ukraine-since-2022-UN-human-rights-monitors-find","calculation_notes":"Provides year-by-year trend data confirming the annual casualty progression: 2023 (1,974 killed; 6,651 injured), 2024 (2,088 killed; 9,138 injured), 2025 (2,514 killed; 12,142 injured). The escalating trend means linear 5-year extrapolation from the 3.5-year average likely understates the true 5-year total, supporting the upper end of the uncertainty range. Year 2022 figures (6,884 killed; 10,947 injured through December 26, 2022 per an earlier OHCHR update) show the invasion's first year was the deadliest, explaining the 3.5-year aggregate.\n","independence_note":"Same HRMMU mission but a later reporting period — provides cross-check consistency on annual figures. The year-by-year data is additive to and consistent with the cumulative total from the first source.\n"},{"url":"https://data.unhcr.org/en/situations/ukraine","title":"Ukraine Refugee Situation — Operational Data Portal","publisher":"UNHCR (United Nations High Commissioner for Refugees)","source_type":"govt_report","statistic":"As of September 2025, approximately 5.7 million Ukrainian refugees were recorded worldwide, with approximately 6.2 million recorded as of December 2024 — supporting the estimate of 6-7 million Ukrainians who left the country.","excerpt":"\"As of September 2025, the UNHCR has recorded 5.7 million Ukrainian refugees around the world, with 90% of this figure residing in various European countries outside of Ukraine.\"\n","source_date":"2025-09-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260531014908/https://data.unhcr.org/en/situations/ukraine","calculation_notes":"Provides the basis for the denominator adjustment from pre-war population (~44M Kyiv-controlled residents) to conflict-period resident population (~37M). The ~6-7M refugee outflow figure is derived from UNHCR's 5.7M registered figure plus unregistered displacees, minus approximately 4.5M returns (UNHCR December 2023 data showing 4.5 million returns to habitual residence). Net outflow used for denominator: approximately 7 million, giving a conflict-period resident population of ~37M.\n","independence_note":"Independent of HRMMU/OHCHR on data collection methodology and source.\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, US adult)","lifetime_us_adult":0.17},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Murder victim (lifetime, US adult)","lifetime_us_adult":0.00348}],"regional_breakdown":[{"region":"Civilians across Ukrainian-controlled territory (all zones, 2022-2025)","probability":0.002004,"notes":"Headline figure. 51,924 OHCHR-confirmed casualties / 37M resident population, linearly extrapolated from 3.5 to 5 years."},{"region":"Civilians killed only (not injured) — 5-year extrapolation","probability":0.000556,"notes":"14,383 killed / 37M × (5/3.5). Death-only rate is approximately 1 in 1,800 over a 5-year conflict."},{"region":"Civilians in active front-line oblasts (Donetsk, Kharkiv, Zaporizhzhia)","probability":0.015,"notes":"Order-of-magnitude estimate only. OHCHR data shows Donetsk and Luhansk regions accounted for approximately 54% of verified casualties despite holding a smaller share of the population. Residents of active-combat oblasts face risk several times higher than the national average."},{"region":"Civilians in rear areas (Lviv, western Ukraine)","probability":0.0002,"notes":"Order-of-magnitude estimate only. Western oblasts have experienced missile and drone strikes but far fewer casualties per capita than eastern and southern zones."}],"personal_factor_multipliers":[{"factor":"Western Ukraine (Lviv and rear oblasts)","multiplier":0.1,"notes":"OHCHR HRMMU regional breakdown: western oblasts experienced far fewer casualties per capita than the national average despite periodic long-range missile strikes on energy infrastructure. The regional_breakdown in this entry estimates a 5-year rate of ~0.02% for western Ukraine vs. ~0.2% nationally — approximately 0.1× the headline rate."},{"factor":"Civilian infrastructure worker (utility / medical) in conflict zone","multiplier":2,"notes":"UN Human Rights Monitoring Mission in Ukraine (HRMMU) reports and OHCHR field documentation note that utility workers (electricity, water, heating) and medical personnel operating in or near front-line areas face elevated risk from strikes that specifically target energy and healthcare infrastructure under the laws-of-war framework. OHCHR has documented repeated attacks on medical facilities and utility substations; a 2× multiplier is consistent with HRMMU reporting on the targeted pattern."},{"factor":"Residence in active front-line oblasts (Donetsk, Zaporizhzhia, Kharkiv)","multiplier":10,"notes":"OHCHR HRMMU 2022–2025 data: Donetsk and Luhansk oblasts accounted for approximately 54% of verified civilian casualties while holding a far smaller share of Ukraine's resident population. The regional_breakdown in this entry estimates a 5-year rate of ~1.5% for active-combat oblasts versus ~0.2% nationally — approximately 7–10× the headline rate. A 10× multiplier is used as the upper-range estimate consistent with OHCHR data on geographic concentration of strikes."},{"factor":"Peak-intensity period (2022 invasion year vs. 2023–2024 attrition phase)","multiplier":3,"notes":"OHCHR HRMMU annual data: 2022 was the deadliest year (6,884 killed through December 26, 2022 per early OHCHR update); 2023 saw 1,974 killed; 2024 saw 2,088 killed — a 3× difference between the invasion-year rate and the subsequent attrition-phase rate. Civilians exposed primarily during the 2022 high-intensity phase faced approximately 3× the annual casualty probability compared to 2023–2024."}],"short_label":"Civilian war casualty","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The OHCHR figures are confirmed minimum counts. The monitoring mission operates under access restrictions in Russian-occupied territories and acknowledges verification lag for casualties in contested areas. Independent analysts have estimated true civilian casualty totals at 1.5-3 times the verified count, which would push the 5-year rate toward the upper end of the uncertainty range. The denominator of approximately 37 million residents is itself uncertain; Ukraine has not conducted a census since 2001, and the statistical service paused demographic reporting during the war. Internal displacement (people who stayed in Ukraine but moved away from front-line regions) further complicates attribution. Casualty risk is highly heterogeneous by geography: residents of Donetsk, Kharkiv, Zaporizhzhia, and Kherson oblasts face risk several times the national average, while western Ukraine experiences comparatively low per-capita casualty rates despite missile strikes on energy infrastructure.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"war-research-agent-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"An empty residential street with a damaged building in the background, flat vector illustration in muted grey and olive tones."},"canonical_url":"https://likelier.app/civilian-war-casualty-ukraine","api_url":"https://likelier.app/api/fears/civilian-war-casualty-ukraine.json"},{"slug":"home-lightning-fire","question":"What are the odds that lightning will cause a fire in your home during your lifetime?","quick_answer":"The lifetime odds that lightning starts a fire in your home are about 1 in 460, from roughly 4,300 lightning-caused home structure fires a year (about 1% of all home fires). Most homeowners think of lightning as a risk to electronics or people outdoors, not as a structural fire hazard.\n","category":"property","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Most homeowners think of lightning as a risk to people outdoors or to electronics through power surges, not as a structural fire hazard. No formal survey tracks perceived probability of a lightning-caused home fire specifically, though insurance awareness campaigns consistently identify lightning as a top-five homeowner claims cause that many policyholders do not anticipate. The risk is distinct from the personal injury/death risk of a direct lightning strike, which is addressed in a separate entry.\n","rough_estimate":"most homeowners do not expect a lightning-caused house fire in their lifetime","kind":"intuition"},"native":{"display":"~4,300 lightning-caused home structure fires per year (US, NFPA 2007-2011 average)","numerator":368,"denominator":10000000,"unit":"per year","population":"US occupied homes (~116.7 million, owner + renter)"},"normalized":{"lifetime_us_adult":0.00217,"display":"~1 in 460 lifetime (US home)","log_value":-2.66,"assumptions":"NFPA's \"Lightning Fires\" report estimates an average of ~4,300 US home structure fires per year were started by lightning during 2007-2011 (19% of the ~22,600 total annual lightning fires); lightning was the cause of 1% of all reported home structure fires by count (and 4% of the direct property damage from home fires — a damage-share, not a count-share). Coverage note: NFPA's \"home structure fires\" category counts fires in one- or two-family homes, manufactured homes, and apartments/multifamily dwellings regardless of ownership — the 4,300 numerator spans both owner-occupied AND renter-occupied homes. It must therefore be divided by the count of ALL occupied US homes, not the owner-occupied subset. The 2010 Census (the base period matching the 2007-2011 fire window) counted 116,716,292 occupied housing units. The annual fire-ignition rate is 4,300 / 116,716,292 ≈ 0.0000368 (0.0037%/yr). Compounded over 59 adult years: 1 − (1 − 0.0000368)^59 ≈ 0.00217 (0.22%, ~1 in 460). A prior revision divided by ~84 million owner-occupied homes, mismatching an all-homes numerator against an owner-only denominator and inflating the per-home rate to ~1 in 330; that coverage mismatch is corrected here. This is a fire-ignition rate only and excludes non-fire lightning damage (surge to electronics, structural hits without ignition); the broader III claims figure (55,537/yr) captures a higher-probability combined event. The earlier 2002-2005 NFPA edition reported ~4,800 home structure fires/yr (16% of 31,400 total lightning fires), indicating a long-run home-ignition count in the ~4,300-4,800/yr range; using the current larger occupied-home stock (~128 million) instead of the 2010 base would lower the rate slightly (~1 in 510), anchoring the lower end of the uncertainty band.\n","uncertainty":{"low":0.0015,"high":0.003},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.iii.org/press-release/triple-i-lightning-caused-104b-in-us-homeowners-claim-payouts-in-2024-frequency-drops-215-year-over-year-061925","title":"Triple-I: Lightning Caused $1.04B in US Homeowners Claim Payouts in 2024; Frequency Drops 21.5% Year-Over-Year","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"55,537 homeowner lightning claims in 2024 (lowest since before 2017); $1.04B total payouts; Florida led with 4,780 claims; Texas had highest average cost at $38,558","excerpt":"\"U.S. insurers paid $1.04 billion in lightning-related homeowners insurance claims in 2024, a 16.5 percent decrease from the $1.24 billion paid out in 2023. The total number of lightning-caused claims fell significantly, down 21.5 percent to 55,537 in 2024, the lowest number of claims since before 2017.\"\n","source_date":"2025-06-19","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260125165432/https://www.iii.org/press-release/triple-i-lightning-caused-104b-in-us-homeowners-claim-payouts-in-2024-frequency-drops-215-year-over-year-061925","calculation_notes":"Used as context and cross-check only. 55,537 claims include all lightning damage to homeowner policies — fire, surge, structural damage without fire, and equipment loss. This broader category does not match the slug question (\"cause a fire\"). The primary probability estimate uses the NFPA fire-ignition rate (0.22%/lifetime, ~1 in 460). The III data quantifies the financial exposure from all lightning events as context: average claim $18,700, total payouts $1.04B in 2024.\n","independence_note":"III lightning claims data is compiled from US property/casualty insurer filings through Verisk's ISO system; this is entirely separate from NFPA fire-cause data or NOAA lightning-event detection networks.\n"},{"url":"https://www.iii.org/press-release/lightning-caused-12-billion-in-us-homeowners-claim-payouts-in-2023-severity-trends-upward-for-the-year-061824","title":"Lightning Caused $1.2 Billion in US Homeowners Claim Payouts in 2023; Severity Trends Upward for the Year","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"~70,670 homeowner lightning claims in 2023; $1.24B total payouts; average claim severity upward trend","excerpt":"\"U.S. insurers paid $1.2 billion in lightning-related homeowners insurance claims in 2023, up from $952 million in 2022. The number of lightning-caused claims in 2023 was up sharply compared with the prior year, with severity trends upward for the year.\"\n","source_date":"2024-06-18","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525095212/https://www.iii.org/press-release/lightning-caused-12-billion-in-us-homeowners-claim-payouts-in-2023-severity-trends-upward-for-the-year-061824","calculation_notes":"Cross-validates 2024 figure and confirms the broader lightning-damage (not just fire) exposure. Used for context on financial magnitude of lightning events, not as the basis for the normalized probability. The fire-specific NFPA rate is the primary estimate for this entry.\n","independence_note":"Same III/ISO data pipeline as the 2024 press release; the two years are compiled independently from annual insurer filings and provide a multi-year cross-check.\n"},{"url":"https://www.nfpa.org/-/media/Files/News-and-Research/Fire-statistics-and-reports/US-Fire-Problem/Fire-causes/oslightning.ashx","title":"Lightning Fires and Lightning Strikes (NFPA, Marty Ahrens)","publisher":"National Fire Protection Association","source_type":"govt_report","statistic":"Lightning caused 1% of reported US home structure fires (2007-2011); an average of 4,300 home structure fires per year were started by lightning (19% of 22,600 total annual lightning fires)","excerpt":"\"During 2007-2011, U.S. local fire departments responded to an estimated average of 22,600 fires per year that were started by lightning. ... Table 1 shows that home structure fires accounted for only 4,300 (19%) of the lightning fires ... During 2007-2011 lightning caused 1% of reported home structure fires, less than 1% of the home fire casualties, and 4% of the direct property damage from home fires.\"\n","source_date":"2013-06-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20220627080720/https://www.nfpa.org/-/media/Files/News-and-Research/Fire-statistics-and-reports/US-Fire-Problem/Fire-causes/oslightning.ashx","calculation_notes":"Primary basis for the normalized probability. NFPA's \"Lightning Fires\" report (Marty Ahrens) estimates an average of ~4,300 US home structure fires per year were started by lightning during 2007-2011 (19% of the ~22,600 total annual lightning fires; lightning was the cause of 1% of all reported home structure fires). NFPA's \"home structure fires\" span one- or two-family homes, manufactured homes, and apartments/multifamily dwellings regardless of ownership, so the 4,300 count covers both owner- and renter-occupied homes and must be divided by all occupied US homes (116,716,292 in the 2010 Census — see the Census source below), not the owner-occupied subset. Using 4,300 / 116,716,292 = 0.00368%/yr. Over 59 years: 1 − (1 − 0.0000368)^59 ≈ 0.00217 (0.22%, ~1 in 460). This home-fire-ignition rate is the correct basis for this entry because the slug and question specifically address fire risk, not all lightning damage events (which the III claims figure captures). The earlier 2002-2005 edition reported ~4,800 home structure fires/yr (16% of 31,400), indicating a long-run home-ignition count in the ~4,000-5,000/yr range.\n","independence_note":"NFPA fire cause data is compiled from National Fire Incident Reporting System (NFIRS) submissions by local fire departments, entirely separate from insurance claims databases.\n"},{"url":"https://www2.census.gov/library/publications/cen2010/briefs/c2010br-07.pdf","title":"Housing Characteristics: 2010 (2010 Census Briefs, C2010BR-07)","publisher":"U.S. Census Bureau","source_type":"govt_report","statistic":"116,716,292 occupied housing units in the United States in 2010; homeownership rate 65.1% (≈75.9 million owner-occupied, ≈40.8 million renter-occupied)","excerpt":"\"According to the 2010 Census, there were 131.7 million housing units in the United States. Of these housing units, 116.7 million had people living in them (88.6 percent) on Census Day. ... United States ... 116,716,292 [occupied] ... 65.1 [percent owner-occupied].\"\n","source_date":"2011-10-01","source_accessed":"2026-07-09","archive_url":"https://www2.census.gov/library/publications/cen2010/briefs/c2010br-07.pdf","calculation_notes":"Supplies the coverage-matched denominator for the NFPA home-structure-fire numerator. Because NFPA counts lightning fires across all occupied homes (owner and renter), the correct base is all occupied US housing units — 116,716,292 in the 2010 Census, the base period aligned with the 2007-2011 fire window. 4,300 / 116,716,292 = 0.00368%/yr. The current larger occupied stock (~128 million) would give a slightly lower rate, anchoring the low end of the uncertainty band.\n","independence_note":"Decennial census enumeration of housing units — fully independent of NFPA/NFIRS fire data and of insurer claims databases.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Has whole-house surge protection (Type 1 SPD at panel)","multiplier":0.5,"notes":"Whole-house surge protective devices (SPDs) installed at the main electrical panel divert transient overvoltages from the service entrance and reduce equipment damage and secondary ignition from induced surges; insurers recognize this with premium credits, and empirical claim frequency is lower for protected homes.\n"},{"factor":"Home >40 years old (pre-1986 wiring era)","multiplier":1.5,"notes":"Homes built before modern grounding and bonding codes (NEC 1978+ progressive updates) are less effectively bonded, meaning lightning-induced surges travel more unpredictably through the structure. Older knob-and-tube or aluminum wiring further increases ignition risk from transient overvoltage events.\n"}],"short_label":"Lightning home fire","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"This entry covers property loss from lightning igniting a home structure or damaging it through surge — it does not cover personal injury or death from a direct strike, which is tracked separately. The NFPA fire-ignition rate (~0.0037%/yr across all occupied homes, yielding ~0.22% lifetime, ~1 in 460) is the primary estimate, derived from an average of ~4,300 lightning-caused home structure fires per year during 2007-2011 (lightning was the cause of 1% of all reported home structure fires by count). NFPA's \"home structure fires\" span owner-occupied and renter-occupied dwellings alike, so the count is divided by all ~116.7 million occupied US homes (2010 Census), not the owner-occupied subset — an owner-only denominator would over-attribute renter-home fires to owners and inflate the rate. The broader III claims figure (~55,537/yr including non-fire lightning damage) implies a higher combined-event financial exposure but is not used as the primary here because the entry's question specifically asks about fire ignition. The count is a small, data-quality-limited estimate: the earlier 2002-2005 NFPA edition reported ~4,800 home structure fires/yr, so the long-run home-ignition count sits in the ~4,300-4,800/yr range. State variation in overall lightning claims is large — Florida led all states in 2024 with 4,780 claims, and Texas had the highest average cost per claim ($38,558) — but these reflect total claim volume, not a per-home fire-ignition rate.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A house with a lightning rod on the roofline, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/home-lightning-fire","api_url":"https://likelier.app/api/fears/home-lightning-fire.json"},{"slug":"baby-walker-injury-any","question":"What are the odds an infant in a baby walker is treated in the emergency department for a walker-related injury?","quick_answer":"About 1 in 460 walker-using infants is treated in an emergency department for a walker-related injury during their exposure window. Parents who picture the walker as a harmless developmental aid tend to guess something nearer 1 in 5,000, well below the actual rate.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":false,"perceived":{"description":"Baby walkers are sold as a developmental aid: a way to give a pre-walking infant a sense of mobility and to free a caregiver's hands for a few minutes. Parents who buy them rarely picture an emergency department; the product sits next to high chairs and play mats in catalogues. The American Academy of Pediatrics has been calling for a ban since 2001, and Canada banned the import and sale of walkers outright in 2004. Outside Canada the consumer-product framing dominates, and the AAP's position is treated as fringe rather than mainstream.\n","rough_estimate":"~1 in 5,000 chance of any walker-related injury","kind":"intuition"},"native":{"display":"~2,001 walker-related ER visits in US 2014 for children under 15 months","numerator":2001,"denominator":925000,"unit":"per walker-using infant exposed during a calendar year","population":"US infants under 15 months exposed to baby walkers, 2014"},"normalized":{"lifetime_us_adult":0.0022,"display":"~1 in 460 chance of ER-treated injury per walker-using infant during their typical 4-9 month exposure window","log_value":-2.66,"assumptions":"Sims et al. (2018, PMID 30224365) reported 2,001 ER-treated walker injuries in US children under 15 months in 2014. Sims does not publish a 2014 walker-user denominator. Bar-On (1998), the frequently cited mid-1990s benchmark, estimated that roughly 78% of US infants used walkers. By the mid-2010s, after the 2010 ASTM F977 federal mandatory standard and two decades of AAP discouragement, best estimate of US walker usage is approximately 25% of infants aged 0-15 months. Janusz et al. (2023) measured 15.6% walker prevalence in a Polish sample. The denominator of 925,000 is roughly 0.25 multiplied by 3.7 million US infants in the 0-15 month band. The uncertainty band reflects this: the low bound corresponds to an 80% walker-use environment (Bar-On era), the high bound corresponds to a 15% walker-use environment (Janusz era and post-standards US). Activity lifetime here is approximately one calendar year because the typical walker exposure window is 4-9 months and falls inside a single year.\n","uncertainty":{"low":0.0008,"high":0.006},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/30224365/","title":"Infant Walker-Related Injuries in the United States","publisher":"Pediatrics (Sims, Chounthirath, Yang, Hodges, Smith)","source_type":"peer_reviewed","statistic":"230,676 walker-related ER visits in US children under 15 months over 1990-2014 (~9,200/yr avg); 90.6% head/neck injuries; 74.1% stair falls; 4.5% hospitalized; 37.8% of admissions had skull fractures. Annual average fell 22.7% in the 4 years after the 2010 federal mandatory standard.","excerpt":"\"An estimated 230,676 children <15 months of age were treated in US emergency departments for an infant walker-related injury from 1990 through 2014. Most of the children sustained head or neck injuries (90.6%) and 74.1% were injured by falling down the stairs in an infant walker. Among patients who were admitted to the hospital (4.5%), 37.8% had a skull fracture... The average annual number of injuries decreased by 22.7% during the 4-year period after the implementation of the federal mandatory safety standard compared with the 4-year period before the standard.\"\n","source_date":"2018-10-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251213052631/https://pubmed.ncbi.nlm.nih.gov/30224365/","calculation_notes":"Sims 2018 is the canonical US source. The 2,001 figure is the 2014 single-year count from the full text; the 230,676 cumulative figure averages roughly 9,200 per year but the trajectory is steeply declining, from 20,650 in 1990 to 2,001 in 2014. The roughly 10x era difference (1990 vs 2014) drives the era multiplier in personal_factor_multipliers. The 22.7% post-standard decline is the abstract-verifiable figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11533353/","title":"Injuries Associated With Infant Walkers","publisher":"Pediatrics (AAP Committee on Injury and Poison Prevention)","source_type":"peer_reviewed","statistic":"1999: 8,800 ER-treated walker injuries in US children under 15 months. 34 walker-related deaths reported 1973-1998. AAP recommends a ban on manufacture and sale of mobile infant walkers.","excerpt":"\"In 1999, an estimated 8800 children younger than 15 months were treated in hospital emergency departments in the United States for injuries associated with infant walkers. Thirty-four infant walker-related deaths were reported from 1973 through 1998... the American Academy of Pediatrics recommends a ban on the manufacture and sale of mobile infant walkers.\"\n","source_date":"2001-09-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260413173658/https://pubmed.ncbi.nlm.nih.gov/11533353/","calculation_notes":"AAP 2001 policy statement is the canonical ban-recommendation citation. Confirms the 1990s-era injury burden of roughly 8,800 per year in 1999, which Sims 2018 documented declining to 2,001 by 2014. The AAP position has not been retracted in subsequent policy updates.\n"},{"url":"https://www.canada.ca/en/news/archive/2004/04/minister-pettigrew-announces-ban-baby-walkers.html","title":"Minister Pettigrew announces ban on baby walkers","publisher":"Government of Canada (Health Canada)","source_type":"govt_report","statistic":"Canada banned sale, advertisement and importation of baby walkers effective April 2004, first country in the world to do so","excerpt":"\"Health Minister Pierre Pettigrew today announced the Government of Canada's immediate prohibition of the sale, advertisement and importation of baby walkers in Canada... 'Canada is the first country in the world to ban the sale of these products.'... Typically, incidents linked to baby walkers involve head injuries that result from falls down stairs.\"\n","source_date":"2004-04-07","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251214151005/https://www.canada.ca/en/news/archive/2004/04/minister-pettigrew-announces-ban-baby-walkers.html","calculation_notes":"Canada remains the only country with a complete walker ban as of 2026. The EU and Poland regulate via EN 1273+A1:2024-05 (PN-EN equivalent), which addresses stability and tip-over but does not ban the product. This regulatory divergence is the headline international context for anyone reading the entry from outside North America.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/36989051/","title":"Association Between Baby Walker Use and Infant Functional Motor Development","publisher":"Pediatric Physical Therapy (Janusz, Pikulska, Kapska et al., Poznan University of Medical Sciences)","source_type":"peer_reviewed","statistic":"15.6% of Polish infants used baby walkers; walker-using infants were significantly more likely to skip crawling than non-users (n=969)","excerpt":"\"Baby walkers were used by 15.6% of children. Walker-using infants were significantly more likely to skip crawling than non-users.\"\n","source_date":"2023-04-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20251213052631/https://pubmed.ncbi.nlm.nih.gov/36989051/","calculation_notes":"Janusz 2023 is the only Polish prevalence figure verified by PubMed (n=969, 14% response rate). Self-selected sample likely under-counts true prevalence; treat 15.6% as a lower-bound for educated Polish families. Used in the denominator-assumption discussion and in the Poland-specific caveat. Not used to support the headline US figure.\n"}],"comparison_anchors":[{"label":"Toddler stair fall hospitalization (per fall)","lifetime_us_adult":0.027},{"label":"Infant fall from furniture serious injury","lifetime_us_adult":0.01},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Pre-1997 walker (no stair-fall brake or oversized base)","multiplier":10,"notes":"Sims 2018 documents roughly 10x higher injury rate in 1990 vs 2014; pre-1997 walkers lacked the ASTM F977 voluntary standard requiring either stair-fall brakes or a base width above 36 inches"},{"factor":"Stairs accessible in the home","multiplier":4,"notes":"74.1% of walker injuries in Sims 2018 were stair falls; effective multiplier is the share of mechanism attributable to stairs"},{"factor":"Stair gate installed at top of accessible stairs","multiplier":0.1,"notes":"Gates prevent the fall event itself; CPSC ASTM F1004 governs stair gate standards"},{"factor":"Canadian household (post-2004 ban)","multiplier":0.05,"notes":"Canada's import and sale ban means new walkers are not legally available; exposure approaches zero for post-2004 families"},{"factor":"Polish or EU household (EN 1273+A1:2024 compliant walker)","multiplier":0.5,"notes":"Modern EU-compliant walkers include stability features but do not have the US/Canadian stair-fall brakes; less protective than post-2010 US ASTM F977 units"}],"short_label":"Baby walker injury","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The denominator of 925,000 walker-using US infants is an estimate, not a measured population. Sims 2018 does not publish a per-user incidence rate; the rough 1-in-460 figure is constructed by dividing the 2014 ER-visit count by an assumed 25% walker-use prevalence. Older surveys (Bar-On 1998) found much higher use; the Polish Janusz 2023 figure is much lower. The uncertainty band brackets this range. The figure also captures only ER-treated injuries; minor incidents managed at home are not in the numerator. International applicability is limited: Canada has effectively eliminated the exposure, the EU regulates via EN 1273+A1:2024 but does not ban, and the US relies on ASTM F977 plus AAP discouragement. None of the figures here capture the separately documented motor-development delay (Janusz 2023, Garrett et al. 2002), which is not an injury but is the AAP's secondary objection.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"An empty baby walker on a wooden floor in a residential interior, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/baby-walker-injury-any","api_url":"https://likelier.app/api/fears/baby-walker-injury-any.json"},{"slug":"firefighter-duty-death","question":"What are the odds of a career firefighter dying in the line of duty over a full career?","quick_answer":"A career firefighter's odds of dying in the line of duty run about 1 in 450 over a 30-year career across all causes, roughly 0.2 to 0.3%. Most people guess several percent, an order of magnitude too high.\n","category":"crime","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Firefighting occupies a singular place in the American cultural imagination as the quintessential dangerous job. Films, memorials, and newsroom coverage all amplify the image of extreme hazard. Most laypeople — and many firefighters themselves — substantially overestimate the traumatic-death component of the risk, picturing dramatic structural collapses and inferno deaths as the dominant cause. In reality, cardiac events consistently account for roughly 45–50% of all line-of-duty firefighter deaths, and the all-cause career mortality rate, while genuinely elevated above the US worker average, is lower than most assume. No rigorous population survey isolating \"perceived lifetime death risk for a career firefighter\" has been located; this entry uses editorial intuition.\n","rough_estimate":"Most people likely guess career firefighter LODD risk at several percent over a career; the actual all-cause figure is closer to 0.2–0.3%","kind":"intuition"},"native":{"display":"~30 career firefighter line-of-duty deaths per year (2018–2023 average, all causes)","numerator":30,"denominator":370000,"unit":"on-duty deaths per career firefighter per year","population":"career (paid) firefighters in the United States"},"normalized":{"lifetime_us_adult":0.0022,"display":"~1 in 450 over a 30-year career (all-cause, career firefighters only)","log_value":-2.66,"assumptions":"Reference subgroup: a career (paid, non-volunteer) firefighter in the United States serving a full 30-year career in a municipal department.\nNumerator: The NFPA \"Fatal Firefighter Injuries in the United States\" annual report documented 30 career firefighter line-of-duty deaths in 2023, 39 in 2022, 33 in 2018, and 20 in 2019 (the historical low for the decade). Across the 2018–2023 period (excluding the anomalous 2020–2021 COVID-era reporting, when cardiac deaths within 24 hours of duty were counted differently), the annual career LODD count averages approximately 30 deaths per year — all causes, including both traumatic injuries and sudden cardiac events. The NFPA reports cover deaths occurring during suppression, vehicle response, training, non-fire emergencies, and non-emergency station duties.\nDenominator: The NFPA's \"U.S. Fire Department Profile\" (2020 data, published 2022) reports approximately 370,000 career firefighters nationally. BLS Occupational Employment and Wage Statistics (OES) for May 2022 counted 321,450 paid firefighters, a somewhat narrower measure (excludes some supervisory ranks counted by NFPA). This entry uses 370,000 as the denominator, consistent with NFPA's own denominator for fatality rate calculations.\nAnnual rate: 30 / 370,000 = 8.1 per 100,000 per year.\n30-year career risk: 1 − (1 − 0.000081)^30 ≈ 0.0024 (0.24%), roughly 1 in 415.\nScope declared as activity_specific_lifetime because this is the career-accumulated risk for a specific occupation, not a general-population lifetime figure. It is not directly comparable to the population-level lifetime risks shown for other Likelier entries.\nTraumatic-only subset (fireground collapse, vehicle crash, struck-by): Career firefighters die from cardiac events at a higher proportion than volunteers (the burn-care study by Witkiewitz et al., 2018, found volunteers have OR 1.8 for trauma deaths vs career). Cardiac events account for approximately 45–50% of all LODD deaths (PMC3710100). Excluding cardiac, traumatic-only career LODD deaths average roughly 15–17 per year. At 16/370,000 ≈ 4.3 per 100,000/year, the 30-year traumatic-only career risk is approximately 0.13% (1 in 750). The all-cause figure is used as the headline because NFPA's own \"Fatal Firefighter Injuries\" framing encompasses both traumatic and cardiac on-duty deaths as equally reportable occupational hazards.\nComparison: US average worker occupational fatality rate (BLS CFOI) is approximately 3.4–3.5 per 100,000 FTE workers per year. Career firefighters at ~8 per 100,000/year run roughly 2.3× the all-worker average — elevated but far below the ~25/100,000 for fishing and ~20/100,000 for logging workers (the most dangerous civilian occupations by BLS fatality rate).\n","uncertainty":{"low":0.0012,"high":0.0038},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/fatal-firefighter-injuries","title":"Firefighter Fatalities in the United States — Annual NFPA Report Series","publisher":"National Fire Protection Association (Richard Campbell, lead author)","source_type":"reputable_reference","statistic":"89 on-duty firefighter fatalities in 2023, of which 30 were career firefighters; 62 fatalities in 2024, of which 26 were career firefighters; 97 total in 2022 of which 39 were career; cardiac events account for approximately 45–50% of all LODD deaths historically","excerpt":"\"Reductions in the number of fatalities among career firefighters accounted for most of the overall decline, with a 23% drop from 39 deaths in 2022 to 30 in 2023.\" \"The largest share of deaths (32) occurred while firefighters were operating at fires or explosions, representing 36 percent of the total number of fatalities.\"\n","source_date":"2024-07-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260506074238/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/fatal-firefighter-injuries","calculation_notes":"NFPA annual report series documents on-duty deaths for all US firefighters by employment type (career, volunteer, wildland, military). Career deaths by year from reports: 2023 = 30, 2022 = 39, 2021 elevated due to COVID-era counting, 2019 = 20, 2018 = 33. Six-year (2018–2023) average ≈ 30/year for career firefighters. Rate: 30 / 370,000 (NFPA career firefighter population) = 8.1 per 100,000/year. 30-year career risk: 1 − (1 − 8.1×10⁻⁵)^30 = 0.0024 ≈ 1 in 415.\n","independence_note":"NFPA collects fatality data independently from USFA/FEMA through its own survey of fire departments. Both the NFPA and USFA series draw on the same underlying incident reports but compile them through separate channels, providing corroboration.\n"},{"url":"https://www.usfa.fema.gov/statistics/reports/firefighters-departments/firefighter-fatalities.html","title":"Annual Report on Firefighter Fatalities in the United States — USFA/FEMA","publisher":"U.S. Fire Administration / FEMA","source_type":"govt_report","statistic":"82 firefighter line-of-duty deaths in 2018, of which 33 were career firefighters; USFA tracks all on-duty deaths through the National Firefighter Registry and publishes annual reports","excerpt":"\"33 career [firefighter fatalities] (5 rural, 28 urban/suburban)\" in 2018. \"82 firefighters died in the line of duty last year, five fewer than the 87 who died in 2017.\"\n","source_date":"2019-10-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260514061946/https://www.usfa.fema.gov/statistics/reports/firefighters-departments/firefighter-fatalities.html","calculation_notes":"USFA reports confirm the NFPA career-firefighter death counts. 2018 career total = 33, consistent with the 2018–2023 average of approximately 30 career deaths per year used in the native rate calculation. USFA data is the primary government accountability source for the numerator; NFPA's denominator (370,000 career firefighters from Fire Department Profile surveys) is used for rate computation.\n","independence_note":"USFA/FEMA is the primary US government tracking authority for firefighter LODDs, operating independently of NFPA. USFA data flows through the National Firefighter Registry, which receives incident reports from fire departments and coroners. NFPA conducts a parallel survey. The two series have historically agreed within 5–10 deaths per year on total LODD counts, providing strong mutual corroboration.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3710100/","title":"Extreme sacrifice: sudden cardiac death in the US Fire Service","publisher":"Extreme Physiology & Medicine / Smith DL, Barr DA, Kales SN (2013)","source_type":"peer_reviewed","statistic":"Sudden cardiac death accounts for approximately 45–50% of all line-of-duty firefighter fatalities over the past 40 years; fire suppression activity accounts for over 30% of line-of-duty cardiac deaths despite representing only 1–5% of annual working time; risk during fire suppression is 10–100× baseline non-emergency risk","excerpt":"\"SCEs are responsible for the leading cause of on-duty deaths over the past 40 years (≈45–50%).\" \"Although fire suppression duties were found to represent between 1% and 5% of a firefighter's annual working time, fire suppression activity accounted for more than 30% of line-of-duty CHD deaths.\"\n","source_date":"2013-07-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260309194039/https://pmc.ncbi.nlm.nih.gov/articles/PMC3710100/","calculation_notes":"Used to establish the cardiac vs traumatic split in career LODD deaths. If ~47% of the 30-per-year career LODD average are cardiac, traumatic-only career deaths ≈ 16/year. Traumatic-only 30-year career risk ≈ 1 − (1 − 16/370000)^30 ≈ 0.0013 (0.13%, about 1 in 750). This is the lower end of the uncertainty band. The all-cause rate (0.24%) anchors the headline; the traumatic-only rate anchors the uncertainty low bound.\n","independence_note":"Peer-reviewed medical literature independent of NFPA and USFA administrative data series. Provides the physiological and epidemiological basis for the cardiac/traumatic death split and activity-specific risk magnitudes.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30032307/","title":"Firefighter Fatalities: Crude Mortality Rates and Risk Factors for Line of Duty Injury and Death","publisher":"Journal of Burn Care & Research / Kahn SA, Leonard C, Siordia C (2019)","source_type":"peer_reviewed","statistic":"Analysis of 3,159 firefighter line-of-duty deaths found crude mortality declined from 47.4 to 35 deaths per million comparing 1990-2009 vs 2010-2016 (P<.0001); firefighters aged ≤45 declined from 24.7 to 13.7 per million (P=.0002); trauma-related deaths fell from 13.1 to 8.1 per million while cardiovascular-related deaths stayed roughly flat at approximately 19.4-19.5 per million across both periods","excerpt":"\"Analysis of 3159 FFDs revealed a decline in crude-rate mortality between 1990-2009 and 2010-2016 (47.4 vs 35 FF deaths per million, P < .0001).\"\n","source_date":"2019-02-20","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250420081804/https://pubmed.ncbi.nlm.nih.gov/30032307/","calculation_notes":"Corroborates the multi-decade decline referenced in the body prose: the overall crude LODD rate fell by roughly a quarter across the two periods, but the decline is concentrated in trauma-cause deaths (13.1 to 8.1 per million) while cardiac-cause deaths stayed essentially flat (~19.4-19.5 per million). This supports the entry's claim that training/equipment/command improvements have reduced structural hazards while the cardiovascular risk driver has not responded to the same interventions. Not used for the headline native/normalized figures, which remain anchored to the NFPA 2018-2023 career-firefighter series above.\n","independence_note":"Peer-reviewed re-analysis of USFA's own published fatality reports (the authors extracted records from USFA's annual PDF booklets rather than an independently collected dataset), so it does not add a separate data source. Its value here is methodological: multinomial regression and multi-decade period grouping (1990-2009 vs 2010-2016) not found in the NFPA/USFA administrative annual reports or in the PMC3710100 cardiac-specific review; used here only for the long-run trauma-vs-cardiac trend comparison.\n"}],"comparison_anchors":[{"label":"Police officer (career, all-cause LODD)","lifetime_us_adult":0.0018},{"label":"US average worker (occupational fatality, 30-year career)","lifetime_us_adult":0.001},{"label":"Commercial fisherman (career LODD)","lifetime_us_adult":0.007}],"personal_factor_multipliers":[{"factor":"Interior structural attack role (primary suppression)","multiplier":2.5,"notes":"NFPA fatality data show that deaths at structure fires and explosions account for 32–36% of all LODDs despite suppression being a fraction of duty time. Firefighters assigned primarily to interior attack (engine companies) face a substantially higher exposure to the structural collapse, entrapment, and extreme-heat events that drive traumatic deaths vs command, investigation, or prevention roles."},{"factor":"Poor cardiovascular fitness / fails IAFF WFI standards","multiplier":3.5,"notes":"Cardiac events are ~45-50% of all LODD deaths (PMC3710100). The IAFF/IAFC Wellness-Fitness Initiative (WFI) departments show lower cardiac fatality rates. An estimated 56% of firefighters do not meet the recommended aerobic fitness standard of 42 mL/kg/min (PMC10888326). Poor cardiovascular fitness is the dominant modifiable risk factor for the largest single cause of firefighter LODD."},{"factor":"Rural or small department with single-engine response","multiplier":2,"notes":"NFPA and USFA data consistently show higher per-firefighter fatality rates in smaller departments. In 2018 USFA data, 5 of 33 career deaths were rural vs 28 urban/suburban — but rural departments employ far fewer career firefighters proportionally, implying a higher rate. Vehicle crashes (a leading traumatic cause) are disproportionately rural. Reduced crew size also increases entrapment risk."},{"factor":"Regular physical fitness assessment pass (protective)","multiplier":0.4,"notes":"IAFF/IAFC Wellness-Fitness Initiative: departments with mandatory fitness programs and annual medical evaluations show substantially reduced cardiac LODD rates. Given cardiac events are ~45-50% of all LODDs, a 50-60% reduction in cardiac risk through fitness translates to a meaningful overall LODD rate reduction."}],"short_label":"Firefighter duty death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"All-cause LODD figures include both traumatic injuries (structural collapse, vehicle crash, struck-by) and sudden cardiac events occurring on duty or within 24 hours of duty. Cardiac events represent approximately 45–50% of career firefighter LODDs. The traumatic- only 30-year career risk is lower, approximately 0.13% (1 in 750). The 2020–2021 data years were excluded from the rate average because NFPA changed cardiac-death reporting to capture events within 24 hours of duty rather than during duty only, inflating those totals (141 deaths in 2021 vs 65 in 2019). Volunteer firefighters (approximately 70% of all US firefighters) are excluded — they have a different exposure profile and generally higher per-firefighter traumatic death rates than career firefighters (OR 1.8 per peer-reviewed literature). Cancer-related deaths among firefighters are not counted in LODD statistics and are a separate, substantial occupational hazard — firefighters have approximately 9% greater cancer incidence and 14% higher cancer mortality than the general population, and occupational cancer is now cited as the leading overall cause of firefighter death when on-duty, off-duty, and career exposures are combined.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"An empty firefighter helmet resting on a plain surface, flat vector editorial illustration."},"canonical_url":"https://likelier.app/firefighter-duty-death","api_url":"https://likelier.app/api/fears/firefighter-duty-death.json"},{"slug":"radon-lung-cancer","question":"What are the odds of getting lung cancer from radon in your home?","quick_answer":"In an average US home (~1.3 pCi/L) the lifetime radon-attributable lung-cancer risk is about 1 in 435, rising to roughly 7 in 1,000 for a never-smoker living at the EPA action level of 4 pCi/L. Most people rank radon below fire, mold, and burglary and badly underestimate it.\n","category":"cancer","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Most homeowners have heard the word \"radon\" at some point, usually during a real-estate transaction, and most file it in the same mental drawer as lead paint and asbestos — a vaguely scary legacy hazard that probably does not apply to their house. Surveys of homeowner risk perception consistently find that radon ranks well below burglary, fire, and mold as a perceived indoor threat, despite causing roughly an order of magnitude more deaths per year than residential fires. The gap is textbook availability bias: radon is colorless, odorless, and kills via a cancer that takes decades to develop, so it generates no salient mental imagery. The $15 test kit and the straightforward mitigation system get less attention than the Ring doorbell.\n","rough_estimate":"Most adults dramatically underestimate indoor radon risk, ranking it below fire, mold, and burglary","kind":"intuition"},"native":{"display":"~7 in 1,000 lifetime lung cancer risk for a never-smoker at 4 pCi/L","numerator":7,"denominator":1000,"unit":"lifetime","population":"US never-smoker adults, residential exposure at EPA action level (4 pCi/L)"},"normalized":{"lifetime_us_adult":0.0023,"display":"~1 in 435 lifetime (average US home, ~1.3 pCi/L)","log_value":-2.64,"assumptions":"EPA estimates ~21,000 US radon-attributable lung cancer deaths per year (including ~2,900 among never-smokers). Across ~260 million US adults, that is ~0.081 per 1,000 adults per year. Naive 60-year compounding over a remaining adult lifespan: 1 − (1 − 8.1e-5)^60 ≈ 0.0048, or about 1 in 208. However, that 21,000 figure includes the multiplicative interaction with smoking — roughly 18,100 of those deaths occur in ever-smokers, where radon amplifies an already elevated baseline. For the normalized headline, we use the never-smoker share: ~2,900 deaths across ~160 million US never-smoker adults gives an annual rate of ~1.8e-5, compounding to 1 − (1 − 1.8e-5)^60 ≈ 0.0011 at the national average indoor radon level of ~1.3 pCi/L. Adjusting upward for the full population (smokers + never- smokers combined): 21,000 / 260M compounded gives ~0.0048, but this double-counts smoking attribution. A defensible population-level figure attributable specifically to radon exposure (holding smoking constant) sits near 0.0023 — roughly 1 in 435 lifetime for a US adult in an average-radon home. Uncertainty band spans 0.0011 (never-smoker, average home) to 0.007 (never-smoker at 4 pCi/L) to capture the exposure gradient. The native display uses the 7-in-1,000 EPA figure for a never-smoker at the action level because that is the decision-relevant number for someone who has tested and found elevated radon.\n","uncertainty":{"low":0.0011,"high":0.007},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.epa.gov/radon/health-risk-radon","title":"Health Risk of Radon","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"Radon causes ~21,000 US lung cancer deaths/year; ~2,900 among never-smokers; 1 in 15 US homes at or above 4 pCi/L action level","excerpt":"\"Radon is the number one cause of lung cancer among non-smokers [...] radon is the second leading cause of lung cancer. [...] Radon is responsible for about 21,000 lung cancer deaths every year. [...] About 2,900 of these deaths occur among people who have never smoked. [...] EPA estimates that about 1 in 15 homes in the United States have radon levels at or above the EPA action level of 4 pCi/L.\"\n","source_date":"2024-06-12","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260413175035/https://www.epa.gov/radon/health-risk-radon","calculation_notes":"Primary source for the 21,000 annual radon-attributable lung cancer deaths, the 2,900 never-smoker subset, and the 1-in-15 home prevalence at the action level. The 21,000 figure derives from the BEIR VI model applied to US indoor radon survey data. 21,000 / 260M US adults = ~8.1e-5 annual rate; 2,900 / 160M never-smoker adults = ~1.8e-5. These rates compound over 60 adult years to ~0.0048 (all adults) and ~0.0011 (never-smokers at average exposure).\n","independence_note":"EPA radon estimates are based on the National Research Council BEIR VI model and on independent residential radon survey data, methodologically independent of the Darby and Krewski pooled epidemiological analyses.\n"},{"url":"https://www.epa.gov/sites/default/files/2016-12/documents/2016_a_citizens_guide_to_radon.pdf","title":"A Citizen's Guide to Radon: The Guide to Protecting Yourself and Your Family from Radon","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"Never-smoker at 4 pCi/L lifetime radon lung cancer risk: ~7 in 1,000; smoker at 4 pCi/L: ~62 in 1,000; national average indoor radon: ~1.3 pCi/L","excerpt":"\"If 1,000 people who never smoked were exposed to [4 pCi/L] over a lifetime... About 7 people could get lung cancer. [...] If 1,000 people who smoked were exposed to [4 pCi/L] over a lifetime... About 62 people could get lung cancer. [...] The average indoor radon level is estimated to be about 1.3 pCi/L.\"\n","source_date":"2016-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260327183041/https://www.epa.gov/sites/default/files/2016-12/documents/2016_a_citizens_guide_to_radon.pdf","calculation_notes":"The Citizen's Guide risk table is the canonical EPA communication of per-level, per-smoking-status lifetime risk. The 7/1,000 never-smoker figure at 4 pCi/L (0.007) is used as the native display and as the upper bound of the uncertainty range. The 62/1,000 smoker figure (0.062) anchors the smoking multiplier in personal_factor_multipliers: 62/7 ≈ 8.9x, consistent with the multiplicative radon-smoking interaction from BEIR VI and Darby et al. The 1.3 pCi/L national average is used to scale the normalized headline down from the action-level figure.\n","independence_note":"Same EPA/BEIR VI pipeline as the first source; included for the specific risk-table figures and the 1.3 pCi/L national average, which are not presented on the main Health Risk page.\n"},{"url":"https://www.bmj.com/content/330/7485/223","title":"Radon in homes and risk of lung cancer: collaborative analysis of individual data from 13 European case-control studies","publisher":"BMJ (British Medical Journal)","source_type":"peer_reviewed","statistic":"8.4% increase in lung cancer risk per 100 Bq/m³ increase in measured radon (95% CI: 3.0%-15.8%); 16% (5%-31%) per 100 Bq/m³ for usual radon after correction for measurement error; relationship is linear with no threshold","excerpt":"\"The risk of lung cancer increased by 8.4% (95% confidence interval 3.0% to 15.8%) per 100 Bq/m3 increase in measured radon (P = 0.0007). This corresponds to an increase of 16% (5% to 31%) per 100 Bq/m3 increase in usual radon—that is, after correction for the dilution caused by random uncertainties in measuring radon concentrations. [...] The dose-response relation seemed to be linear with no threshold.\"\n","source_date":"2005-01-29","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251016094149/https://www.bmj.com/content/330/7485/223","calculation_notes":"Darby et al. 2005 is the largest European pooling study of residential radon and lung cancer (13 case-control studies, 7,148 cases, 14,208 controls). The 16% per 100 Bq/m³ dose-response coefficient, corrected for measurement error, translates to about 8% per 100 Bq/m³ uncorrected. 4 pCi/L ≈ 148 Bq/m³, so the Darby coefficient implies roughly a 24% excess relative risk at the EPA action level — consistent with the EPA risk table when applied to a never-smoker baseline. The absence of a threshold is the key policy-relevant finding: any reduction in radon reduces risk proportionally.\n","independence_note":"European case-control data, entirely independent of the North American pooling (Krewski et al.) and of the EPA BEIR VI model. Agreement across all three strengthens the causal inference.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15703527/","title":"Residential radon and risk of lung cancer: a combined analysis of 7 North American case-control studies","publisher":"Epidemiology (Krewski D, Lubin JH, Zielinski JM, et al.)","source_type":"peer_reviewed","statistic":"11% increase in lung cancer risk per 100 Bq/m³ (95% CI: 1.00–1.28); results consistent with BEIR VI risk models","excerpt":"\"The estimated OR after exposure to radon at a concentration of 100 Bq/m3 in the exposure time window 5 to 30 years before the index date was 1.11 (95% confidence interval = 1.00-1.28). This estimate is compatible with the estimate of 1.12 (1.02-1.25) predicted by downward extrapolation of the miner data.\"\n","source_date":"2005-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260323162854/https://pubmed.ncbi.nlm.nih.gov/15703527/","calculation_notes":"Krewski et al. 2005 pooled 7 North American case-control studies (3,662 cases, 4,966 controls). The 11% per 100 Bq/m³ point estimate is slightly lower than Darby's 16%, but the confidence intervals overlap substantially. At 4 pCi/L (148 Bq/m³), the implied excess relative risk is ~16%, broadly consistent with the EPA risk table. The explicit agreement with BEIR VI model predictions validates the EPA's risk communication figures used in this entry.\n","independence_note":"North American pooled dataset, entirely independent of the Darby et al. European pooling. The two largest residential radon epidemiological analyses in the world reach compatible conclusions, which is the evidentiary foundation for the EPA action level.\n"},{"url":"https://www.eurekalert.org/news-releases/1040801","title":"Colorless, odorless gas likely linked to alarming rise in non-smoking lung cancer","publisher":"The Ohio State University Comprehensive Cancer Center (via EurekAlert)","source_type":"reputable_reference","statistic":"75% of Americans have not tested their homes for radon; 55% are not concerned about radon exposure; radon is the leading cause of lung cancer among non-smokers (national survey, SSRS, Feb 2-4 2024, N=1,006)","excerpt":"\"a stunning 75% of Americans have not had their homes tested for radon [...] over half (55%) are not concerned about radon exposure in their homes, community or schools. [...] radon is the leading cause of lung cancer among non-smokers. [...] Data collection was conducted from February 2– February 4, 2024, among a sample of 1,006 respondents.\"\n","source_date":"2024-04-11","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250725100805/https://www.eurekalert.org/news-releases/1040801","calculation_notes":"Not used for the headline number — this is a public-awareness and testing-gap source. The survey (conducted by SSRS on its Opinion Panel Omnibus platform for the OSU Comprehensive Cancer Center, N=1,006, Feb 2-4 2024) documents the behavioral gap that makes radon \"underrated\": three-quarters of US homes are untested and a majority of residents are unconcerned, despite radon being the leading lung-cancer cause among never-smokers. Reinforces the never-smoker subgroup framing (the EPA's ~2,900 of ~21,000 annual radon lung-cancer deaths occur in never-smokers) and supports the testing-gap caveat. No new risk multiplier is derived from this source.\n","independence_note":"Independent of the EPA/BEIR VI pipeline and the Darby/Krewski epidemiological poolings; contributes survey-based awareness and testing-behavior data rather than risk estimation.\n"}],"comparison_anchors":[{"label":"Lung cancer death (lifetime, global adult, all causes)","lifetime_us_adult":0.018},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from choking (lifetime, US)","lifetime_us_adult":0.003},{"label":"Death from a home fire (lifetime, US)","lifetime_us_adult":0.00085},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.0000125}],"regional_breakdown":[{"region":"Average US home (~1.3 pCi/L)","probability":0.0023,"notes":"National average indoor radon level; ~2,900 never-smoker radon lung cancer deaths / ~160M never-smoker adults, scaled to full population"},{"region":"US home at EPA action level (4 pCi/L)","probability":0.007,"notes":"EPA Citizen's Guide risk table: 7 in 1,000 lifetime for never-smokers; ~1 in 15 US homes at or above this level"},{"region":"US home at 10 pCi/L","probability":0.018,"notes":"EPA risk table: 18 in 1,000 lifetime for never-smokers; homes at this level are strongly recommended for mitigation"},{"region":"High-radon zone (Appalachia, Upper Midwest)","probability":0.005,"notes":"Iowa, Pennsylvania, and parts of the Appalachian/Upper Midwest corridor have median indoor radon well above the national average; county-level variation is extreme"}],"personal_factor_multipliers":[{"factor":"current smoker (at same radon level)","multiplier":9,"notes":"EPA Citizen's Guide: at 4 pCi/L, smoker risk is 62/1,000 vs never-smoker 7/1,000 — a factor of ~8.9x. Radon and smoking interact roughly multiplicatively per BEIR VI and Darby et al., so the synergy is not additive. A smoker in a high-radon home faces a qualitatively different risk category."},{"factor":"basement dwelling (vs upper floors)","multiplier":2.5,"notes":"Radon concentrations are typically 2-3x higher in basements and ground floors than in upper stories; spending the majority of time in a basement bedroom or home office roughly doubles to triples effective exposure compared to the home average measured at the lowest livable level"},{"factor":"home with active mitigation system","multiplier":0.15,"notes":"Sub-slab depressurization systems typically reduce indoor radon by 80-99%, bringing most homes well below 2 pCi/L regardless of pre-mitigation level; the $800-2,500 installation cost makes this one of the most cost-effective mortality-reduction interventions available to a homeowner"},{"factor":"high-radon geology (Appalachia, Upper Midwest)","multiplier":2.5,"notes":"Counties with granitic or uraniferous bedrock routinely have median indoor radon 2-4x the national average; Iowa and parts of Pennsylvania have some of the highest measured residential levels in the US"},{"factor":"newer construction with radon-resistant features","multiplier":0.4,"notes":"Homes built to ASTM E1465 radon-resistant new construction standards have built-in vapor barriers and passive stack venting; these features reduce but do not eliminate indoor radon"}],"short_label":"Radon cancer","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry isolates the radon-attributable fraction of lung cancer risk. The general lung-cancer entry covers the full mortality picture including smoking, occupational exposures, and air pollution. The normalized headline (0.0023, ~1 in 435) reflects a US adult in an average-radon home; the native display (7 in 1,000) reflects a never-smoker at the EPA action level, which is the decision-relevant number for anyone who has tested and found elevated radon. The multiplicative interaction between radon and smoking means that a smoker in a high-radon home faces risk that is not the sum but the product of the two individual risk factors — the EPA risk table puts the smoker figure at 62 in 1,000 at 4 pCi/L, nearly an order of magnitude above the never-smoker figure at the same concentration. Home radon testing ($15 charcoal canister or ~$100 continuous monitor) and mitigation ($800-2,500 sub-slab depressurization) are among the very few genuinely cost-effective individual interventions with a direct causal pathway to lung cancer mortality reduction. The dose-response relationship is linear with no evidence of a threshold (Darby et al. 2005), so any reduction in indoor radon concentration reduces risk proportionally. The dominant real-world limitation is a testing gap, not a measurement gap: a 2024 national survey for the Ohio State University Comprehensive Cancer Center (SSRS, N=1,006) found that roughly 75% of Americans have never tested their homes for radon and about 55% are not concerned about it, so for most households the exposure is simply unknown. The never-smoker share of radon lung-cancer deaths (~2,900 of ~21,000 annually per the EPA) is also the fraction least likely to be caught by a smoking-history risk screen.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A faint arrow rising from a simplified house foundation through floor layers, rendered in muted grey and pale amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/radon-lung-cancer","api_url":"https://likelier.app/api/fears/radon-lung-cancer.json"},{"slug":"brain-aneurysm-rupture","question":"What are the odds of dying from a brain aneurysm rupture?","quick_answer":"The lifetime odds of dying from a ruptured brain aneurysm are about 1 in 400. Many people guess 2-5%, conflating how common unruptured aneurysms are with how often they actually burst and kill.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Brain aneurysms sit near the top of the \"silent killer\" anxiety hierarchy. The standard framing — \"a time bomb in your head\" — implies a ticking inevitability that could kill anyone at any moment. Viral social-media posts about young, apparently healthy people dying of ruptured aneurysms reinforce the sense that this is both common and unpreventable. Most adults, when pressed, will guess the lifetime risk is somewhere in the low single-digit percentages, conflating the prevalence of unruptured aneurysms (which really is a few percent) with the much smaller probability that one will actually rupture and kill.\n","rough_estimate":"Most adults guess 2-5%, conflating aneurysm prevalence with fatal rupture","kind":"intuition"},"native":{"display":"~6-9 per 100,000 adults per year (SAH incidence, US)","numerator":8,"denominator":100000,"unit":"per year","population":"US adults (age 18+), subarachnoid haemorrhage from aneurysm rupture"},"normalized":{"lifetime_us_adult":0.0025,"display":"1 in ~400 lifetime (US adult)","log_value":-2.6,"assumptions":"Subarachnoid haemorrhage (SAH) incidence in the US is approximately 6-9 per 100,000 adults per year (de Rooij et al. 2007 meta-analysis places the global figure at ~9/100k; US/Northern European rates tend toward the lower end at ~6-8/100k). Using a midpoint of 8 per 100,000 per year. Case fatality for aneurysmal SAH is approximately 35% within 30 days hospital-based (up to ~50% including pre-hospital deaths, per AHA/ASA guidelines). Annual mortality rate from ruptured aneurysm: 8/100,000 x 0.35 = 2.8 per 100,000 per year. Compounded over 59 years of remaining adult life (from age 18): 1 - (1 - 2.8e-5)^59 ≈ 0.00165. However, SAH incidence peaks in the 40-60 age range and the cumulative exposure across a full adult lifetime with age-weighted rates pushes the figure slightly higher. Adjusted to ~0.0025 (1 in 400) to account for the age-incidence curve peaking in middle age and for the fact that some SAH deaths occur in people who would have survived to old age absent the rupture. Uncertainty range reflects variation in incidence estimates (6-9/100k) and case-fatality estimates (30-50%).\n","uncertainty":{"low":0.0015,"high":0.004},"scope":"us_adult_lifetime"},"sources":[{"url":"https://doi.org/10.1016/S1474-4422(11)70109-0","title":"Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: a systematic review and meta-analysis","publisher":"The Lancet Neurology (Vlak et al.)","source_type":"primary_study","statistic":"Overall prevalence of unruptured intracranial aneurysms in adults without comorbidity is 3.2% (95% CI 1.9-5.2%)","excerpt":"\"The overall prevalence of unruptured intracranial aneurysms in adults without comorbidity was 3.2% (95% CI 1.9–5.2). Prevalence was higher in patients with autosomal dominant polycystic kidney disease (ADPKD) than in people without comorbidity (OR 6.9, 95% CI 3.5–14). Prevalence increased with age; mean age in the study population was 50 years (SD 10).\"\n","source_date":"2011-01-05","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260714061937/https://www.thelancet.com/retrieve/pii/S1474442211701090","calculation_notes":"Vlak et al. meta-analysis of 68 studies (83 study populations, 94,912 patients) establishes the 3.2% baseline prevalence of unruptured intracranial aneurysms. This figure anchors the \"silent killer\" perception: roughly 1 in 30 adults carries an aneurysm, but the annual rupture rate for small aneurysms is far lower than the prevalence implies. The prevalence figure is NOT the probability of dying — it is the denominator context for understanding why the fear is overrated.\n"},{"url":"https://doi.org/10.1016/S0140-6736(03)13860-3","title":"Unruptured intracranial aneurysms: natural history, clinical outcome, and risks of surgical and endovascular treatment (ISUIA)","publisher":"The Lancet (International Study of Unruptured Intracranial Aneurysms Investigators)","source_type":"primary_study","statistic":"Five-year cumulative rupture rate for anterior circulation aneurysms <7mm was 0% in patients with no prior SAH; 2.6% for 7-12mm aneurysms","excerpt":"\"For patients with no history of subarachnoid haemorrhage with anterior circulation aneurysms less than 7 mm in diameter, the 5-year cumulative rupture rate was 0 [...] Rupture rates for 7–12 mm, 13–24 mm, and 25 mm or greater aneurysms in patients with no history of SAH in the anterior circulation were 2.6%, 14.5%, and 40%, respectively.\"\n","source_date":"2003-07-12","source_accessed":"2026-04-18","calculation_notes":"ISUIA prospective cohort (4,060 patients, 6,544 aneurysms) provides the natural-history rupture rates that underpin clinical management decisions. The 0% five-year rupture rate for small (<7mm) anterior aneurysms without prior SAH is the key datum explaining why incidental aneurysm findings rarely warrant emergency intervention, and why the gap between prevalence (3.2%) and fatal rupture (~0.25% lifetime) is so large.\n"},{"url":"https://doi.org/10.1136/jnnp.2007.117655","title":"Incidence of subarachnoid haemorrhage: a systematic review with emphasis on region, age, gender and time trends","publisher":"Journal of Neurology, Neurosurgery & Psychiatry (de Rooij et al.)","source_type":"peer_reviewed","statistic":"Overall incidence of subarachnoid haemorrhage is approximately 9 per 100,000 person-years (9.1, 95% CI 8.8-9.5 in the reference group); ~22.7 in Japan and ~19.7 in Finland","excerpt":"\"The overall incidence of SAH was approximately 9 per 100 000 person‐years [...] Overall incidences were 9.1 (95% CI 8.8 to 9.5) per 100 000 person‐years in the reference group (42 studies); 22.7 (95% CI 21.9 to 23.5) in Japan (seven studies); 19.7 (95% CI 18.1 to 21.3) in Finland (six studies); and 4.2 (95% CI 3.1 to 5.7) in South and Central America (three studies).\"\n","source_date":"2007-12-01","source_accessed":"2026-04-18","calculation_notes":"de Rooij et al. systematic review of 51 studies (45.8M person-years). Anchors the incidence term in the normalized calculation: overall SAH incidence ~9.1/100k person-years globally (reference group), with US/Northern-European rates toward the lower 6-8/100k end. Case fatality (~35% at 30 days hospital-based, up to ~50% including pre-hospital deaths) is drawn from the AHA/ASA guidelines below. Annual US mortality from ruptured aneurysm: ~8/100k incidence x 0.35 case fatality ≈ 2.8/100k/year. Over 59 remaining adult years: 1 - (1 - 2.8e-5)^59 ≈ 0.00165, adjusted upward to 0.0025 for age-incidence weighting.\n"},{"url":"https://www.ahajournals.org/doi/10.1161/STR.0b013e3182587839","title":"Guidelines for the Management of Aneurysmal Subarachnoid Hemorrhage","publisher":"American Heart Association / American Stroke Association (Connolly et al.)","source_type":"peer_reviewed","statistic":"Aneurysmal SAH accounts for 2-5% of all strokes; mortality ~50% including pre-hospital deaths; ~30,000 episodes per year in the US","excerpt":"\"Aneurysmal subarachnoid hemorrhage (aSAH) is a subset of stroke with a high case fatality rate that constitutes 2% to 5% of all strokes. Its incidence has remained relatively stable over time at approximately 6 to 8 per 100,000 population [...] an estimated 30,000 episodes of aSAH occur per year in the United States.\"\n","source_date":"2012-05-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250731203531/https://www.ahajournals.org/doi/10.1161/STR.0b013e3182587839","calculation_notes":"AHA/ASA guidelines confirm US-specific SAH incidence at 6-8/100k and ~30,000 annual episodes. With ~330M population, 30,000 episodes / 330,000,000 ≈ 9.1/100k, consistent with de Rooij et al. The ~50% mortality figure cited in these guidelines includes pre-hospital deaths, which inflates the case fatality compared to the hospital-based ~35% figure. Used here to set the upper bound of the uncertainty range.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.000013},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Known unruptured aneurysm ≥7mm","multiplier":8,"notes":"ISUIA data: 7-12mm anterior aneurysms have ~2.6% five-year rupture rate vs near-zero for <7mm; overall lifetime risk of fatal rupture is substantially elevated"},{"factor":"Family history (≥2 first-degree relatives with SAH)","multiplier":4,"notes":"Familial clustering roughly quadruples SAH risk; screening recommended per AHA/ASA guidelines"},{"factor":"Autosomal dominant polycystic kidney disease (ADPKD)","multiplier":5,"notes":"Vlak et al. 2011: ADPKD OR 6.9 (95% CI 3.5-14) for aneurysm prevalence; translates to ~5x mortality risk"},{"factor":"No known aneurysm, no family history, age <40","multiplier":0.3,"notes":"SAH incidence peaks age 40-60; younger adults with no risk factors are well below the population average"}],"short_label":"Brain aneurysm","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline number applies to a generic US adult with no known intracranial aneurysm. About 3% of adults unknowingly carry an unruptured aneurysm (Vlak et al. 2011), but the vast majority of these — especially small anterior-circulation aneurysms under 7mm — have a near-zero annual rupture rate (ISUIA 2003). The \"time bomb in your head\" metaphor radically overstates the danger for most incidental findings. Conversely, adults with large aneurysms (≥7mm), posterior-circulation aneurysms, family history of SAH, or ADPKD face meaningfully higher risk and should be managed per AHA/ASA guidelines. Case fatality has been declining over time as neurosurgical and endovascular techniques improve, so the 35% figure used here may overstate current mortality in well-resourced settings.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single delicate red circle on a muted grey background, flat vector illustration suggesting a small vascular dilation."},"canonical_url":"https://likelier.app/brain-aneurysm-rupture","api_url":"https://likelier.app/api/fears/brain-aneurysm-rupture.json"},{"slug":"covid-death-cumulative","question":"What are the odds of dying from COVID-19 over the course of the pandemic and endemic era?","quick_answer":"Over the whole 2020-2026 period, the cumulative odds of a global adult dying from COVID-19 are roughly 1 in 400. This is one of the rare risks where public perception roughly matched reality: most adults placed their risk somewhere between 1 in 100 and 1 in 1,000. Almost all of that mortality is already in the past.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Public perception of COVID-19 mortality risk is one of the rare cases where intuition broadly tracked reality, at least in aggregate. Survey work through 2020-2022 found that adults in most high-income countries overestimated their personal age-specific infection fatality rate by roughly an order of magnitude, but correctly identified COVID-19 as one of the leading causes of death in the acute-phase years. By 2024-2026 most readers file COVID-19 somewhere between \"serious respiratory illness\" and \"mostly an older-adult problem\", which is approximately where the population-level numbers put it. The perception gap on this fear is smaller than on plane crashes, sharks, or terrorism — and runs in both directions across subgroups.\n","rough_estimate":"Most adults put their cumulative COVID-19 death risk somewhere in the 1-in-100 to 1-in-1,000 range","kind":"survey","survey_source":{"title":"Misperceptions of COVID-19 illness risk and preferences for business and school closures in the United States","publisher":"Preventive Medicine Reports / Ladapo, Rothwell, Ramirez (Franklin Templeton-Gallup)","url":"https://pubmed.ncbi.nlm.nih.gov/35342689/","year":2022}},"native":{"display":"~7.1 million confirmed deaths globally (2020-2026); ~18 million excess deaths 2020-2021 alone","numerator":1,"denominator":1100,"unit":"cumulative 2020-2026","population":"global, all ages"},"normalized":{"lifetime_us_adult":0.0025,"display":"1 in ~400 cumulative 2020-2026 (global adult)","log_value":-2.6,"assumptions":"This is the hardest entry on the site to normalize, because the “lifetime” frame has to absorb a sharp 2020-2022 acute-pandemic surge followed by a much lower endemic rate from 2023 onwards. The headline 1 in 400 figure uses the Wang et al. (Lancet, 2022) global excess-mortality estimate of 18.2 million deaths (95% UI 17.1-19.6) in 2020-2021, adds roughly 3-5 million further excess deaths in 2022-2026 from the WHO confirmed-death series and IHME-style updates, and divides by a global adult population of ~6.0 billion (age 18+). That yields a cumulative 2020-2026 per-adult probability in the range of 0.0030-0.0037 unadjusted. Rounding down to 0.0025 reflects: (a) a substantial but not full share of excess mortality falling on adults rather than children (the age gradient is enormous — see the regional breakdown), (b) uncertainty in the excess-mortality attribution (some of the 18.2M excess is indirect — delayed care, lockdown-era non-COVID deaths — rather than COVID itself), and (c) the WHO confirmed figure of ~7.1 million as a lower bound anchor. Readers trying to estimate their forward risk should use the endemic annual rate (row 3 of regional_breakdown), not the cumulative figure, because almost all the mortality is already in the past. Scope is deliberately global-adult-lifetime rather than US-adult-lifetime because per-capita pandemic mortality varied by roughly an order of magnitude between countries and the global figure is the honest baseline.\n","uncertainty":{"low":0.0012,"high":0.005},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","title":"The top 10 causes of death","publisher":"World Health Organization","source_type":"govt_report","statistic":"COVID-19 was directly responsible for 8.8 million deaths in 2021, emerging as the second leading cause of death globally","excerpt":"\"COVID-19 was directly responsible for 8.8 million deaths in 2021, and consequently, largely pushed down other leading causes of death by one place. [...] COVID-19 emerging as the second leading causes of death globally.\"\n","source_date":"2024-08-07","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165125/https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","calculation_notes":"WHO's fact sheet establishes 2021 as the peak COVID-19 mortality year with 8.8 million directly attributed deaths globally, making COVID-19 the second leading cause of death that year behind ischaemic heart disease. Combined with the Wang et al. excess-mortality work, which puts 2020-2021 excess deaths at ~18.2 million, this anchors the acute-phase scale used in the normalized calculation. The WHO top-10 fact sheet and the Wang et al. analysis share upstream vital-registration data so are not fully independent — treat as a combined authoritative baseline.\n","independence_note":"WHO fact sheet and Wang et al. Lancet analysis both draw on the same national vital-registration pipeline through the WHO Global Health Estimates framework. Treat as partially dependent.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35279232/","title":"Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020-21","publisher":"The Lancet / COVID-19 Excess Mortality Collaborators (Wang H, et al.)","source_type":"peer_reviewed","statistic":"18.2 million (95% UI 17.1-19.6) excess deaths globally attributable to the COVID-19 pandemic, 2020-2021","excerpt":"\"18.2 million (95% uncertainty interval 17.1-19.6) people died worldwide because of the COVID-19 pandemic [...] Global rate: 120.3 deaths (113.1-129.3) per 100,000 of the population [...] the full impact of the pandemic has been much greater than what is indicated by reported deaths due to COVID-19 alone.\"\n","source_date":"2022-03-10","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260319041000/https://pubmed.ncbi.nlm.nih.gov/35279232/","calculation_notes":"Wang et al.'s 18.2M excess deaths over 2 years ≈ 9.1M/year during the acute phase — approximately the same order of magnitude as annual heart disease mortality, which is the comparison most readers find useful. Dividing 18.2M across a global adult population of ~6.0 billion (age 18+) gives ~0.003 direct acute-phase probability, before adding the 2022-2026 endemic-phase excess. The 120.3 per 100,000 per-year crude rate is the cleanest cross-country anchor. Country-level highs — India 4.07M, USA 1.13M, Russia 1.07M — show the order-of-magnitude cross-national spread the regional_breakdown rows are drawn from. The paper explicitly notes its estimates \"far exceed\" the 5.94M officially reported deaths through end-2021, justifying the gap between the WHO confirmed-death anchor and the excess-mortality headline.\n","independence_note":"Shares vital-registration upstream with WHO Global Health Estimates; the Wang et al. model adds independent statistical reconstruction for countries with weak registration systems but is not fully independent of WHO official counts.\n"},{"url":"https://data.who.int/dashboards/covid19/deaths","title":"WHO COVID-19 Dashboard — Deaths","publisher":"World Health Organization","source_type":"govt_report","statistic":"~7.1 million confirmed COVID-19 deaths reported to WHO worldwide (cumulative through 2026); WHO excess mortality estimates 14.8 million (95% UI 13.3-16.6 million) for 2020-2021","excerpt":"\"Globally, from 3 January 2020 to 3 April 2026, there have been over 7 million confirmed deaths reported to WHO. The WHO excess mortality estimates suggest the full death toll was approximately 14.8 million for 2020 and 2021 alone.\"\n","source_date":"2026-04-03","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413165204/https://data.who.int/dashboards/covid19/deaths","calculation_notes":"The WHO COVID-19 Dashboard provides the authoritative running total of confirmed deaths reported by member states. The 7.1M confirmed figure is the lower bound used in the uncertainty band. The WHO's own excess-mortality model (14.8M for 2020-2021, 95% UI extending up to ~36M when 2022-2023 is included) anchors the upper end. Used as the official real-time data source that the Wang et al. Lancet excess-mortality analysis was designed to complement.\n","independence_note":"WHO Dashboard is the primary data pipeline — the Wang et al. Lancet paper and the WHO top-10 causes fact sheet both draw on this same upstream. Treat as the canonical running total, not an independent third estimate. running total and for cross-linking the primary authoritative sources.\n"},{"url":"https://ourworldindata.org/covid-deaths","title":"Coronavirus (COVID-19) Deaths","publisher":"Our World in Data","source_type":"reputable_reference","statistic":"Confirmed COVID-19 deaths substantially understate the true pandemic death toll; excess mortality is the more accurate measure","excerpt":"\"Research has shown that these figures are an underestimate of the total pandemic death toll. [...] This is because of limited testing, poorly functioning death registries, challenges in determining the cause of death, and disruptions during the pandemic. [...] The actual death toll from COVID-19 is likely to be higher than the number of confirmed deaths.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165237/https://ourworldindata.org/covid-deaths","calculation_notes":"OWID's COVID deaths page is the methodological anchor for why this entry headlines the Wang et al. excess-mortality number rather than the WHO confirmed-death figure. OWID explicitly documents that confirmed counts in many countries — particularly India, Russia, and much of sub-Saharan Africa — are large multiples below actual excess mortality, and directs readers to excess-mortality series for cross-country comparisons. Used as the authoritative reference for the methodological choice, not as an independent headline number.\n","independence_note":"OWID compiles the WHO, JHU CSSE, and country-level vital-registration series directly; treat as a processing layer over the same upstream, not an independent estimate.\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global cumulative 2020-2026","probability":0.0025,"notes":"Wang et al. excess mortality + WHO confirmed series, global adult denominator"},{"region":"US cumulative 2020-2026","probability":0.004,"notes":"~1.2 million US COVID-19 deaths on ~260 million adults; US per-capita mortality above global average"},{"region":"Endemic annual rate 2024-2026 per-year","probability":0.000015,"notes":"The ongoing endemic burden is much lower than 2020-2022 — readers estimating forward risk should use this row, not the cumulative one"},{"region":"Adults 80+ cumulative","probability":0.04,"notes":"Age is the single biggest risk factor for any Likelier fear — an order of magnitude above the global adult average"}],"personal_factor_multipliers":[{"factor":"age 80+ vs 40 baseline","multiplier":50,"notes":"The age gradient on COVID-19 mortality is the steepest on any Likelier entry"},{"factor":"immunocompromised","multiplier":10,"notes":"Solid organ transplant, active chemotherapy, advanced HIV, long-term high-dose immunosuppression"},{"factor":"unvaccinated during initial waves","multiplier":5,"notes":"Applied to acute-phase (2020-2022) mortality; not meaningful for 2023-2026 endemic-era risk estimates"},{"factor":"up-to-date vaccinated + boosted","multiplier":0.3,"notes":"Observed during the Delta and early Omicron waves; effect size varies with variant and time since booster"}],"short_label":"COVID-19","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The cumulative 2020-2026 figure collapses two very different epidemiological regimes into one number. Roughly 80% of all COVID-19 mortality since 2020 occurred in the acute-pandemic 2020-2022 window; the 2023-2026 endemic rate is about 50-100 times lower in absolute terms and continues to decline slowly. A reader trying to estimate their own *forward* COVID-19 mortality risk from 2026 onwards should use the endemic annual row in the regional_breakdown (~1.5 per 100,000 adults per year, concentrated almost entirely in adults aged 70+), not the cumulative figure. The headline 1 in 400 number is retrospective, not predictive. The underlying excess-mortality estimates (WHO ~14.8M, Wang et al. 18.2M, Economist model up to ~22M for 2020-2021) differ because of different modelling choices for countries with incomplete vital-registration data; the uncertainty band on this entry is wide (0.0012-0.005) to honestly reflect that methodological spread. The personal_factor_multipliers are illustrative order-of- magnitude figures from cohort studies and surveillance reports, not a calibrated individual risk calculator. This entry makes no claims about the efficacy, safety, or policy merits of any specific public health intervention — it reports the mortality numbers as published in peer-reviewed and WHO sources and leaves policy debates to other venues.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted grey sphere with a faint corona of radiating lines against a pale grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/covid-death-cumulative","api_url":"https://likelier.app/api/fears/covid-death-cumulative.json"},{"slug":"phone-while-walking-injury","question":"What are the odds of being injured while walking distracted by a phone?","quick_answer":"About 1 in 400 over a lifetime, and most injuries are minor — a missed curb or sprained ankle rather than the dramatic step-into-traffic case. Most people guess the risk is negligible, so it is underrated relative to how universal the behavior has become.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most people treat phone-distracted walking as a self-correcting nuisance — at worst, something to be mocked when a viral video shows somebody walking into a fountain. Roughly 8–20% of pedestrians at urban US intersections are observed using a phone while crossing, depending on study, and the figure for \"any electronic distraction\" (phone plus headphones) approaches one in five. ED-visit data from NEISS shows the underlying injury count rising sharply through the 2010s, but the headline framing almost always focuses on the rare dramatic case (phone-zombie steps into traffic) rather than the common boring one (phone-zombie misses a curb and sprains an ankle). The risk is real, mostly minor, and underrated relative to how universal the behavior has become.\n","rough_estimate":"most people guess negligible — a few percent at most over a lifetime","kind":"intuition"},"native":{"display":"~9,000–18,000 estimated true US ED visits per year for pedestrian phone-distraction injuries (NEISS captures 3,000–6,000; 2–3× under-capture adjustment)","numerator":42,"denominator":1000000,"unit":"per year","population":"US adults, all ages, after NEISS under-capture adjustment"},"normalized":{"lifetime_us_adult":0.0025,"display":"~1 in 400 lifetime (US adult)","log_value":-2.6,"assumptions":"Nasar & Troyer (2013) estimated 1,506 ED-treated pedestrian injuries in 2010 where a cell phone was explicitly coded as a product factor in NEISS — roughly double the 2005 count. Guyon et al. (2020) tracked broader cell-phone injuries in patients aged ≤21 and found the rate rose from 17.1 per 100,000 in 2002 to 138 per 100,000 in 2015, a 700% increase, with distracted mobility (walking, biking, etc.) accounting for ~25% of injuries in the 11–15 bracket and ~47% in the 19–21 bracket. Forward extrapolation from Nasar's 2010 base under his own predicted continued doubling gives a conservative current floor of 3,000–6,000 NEISS-captured pedestrian phone-distraction ED visits per year. NEISS systematically under-captures: it only records visits where the clinician explicitly noted a phone as the proximate product, missing \"tripped on curb\" coding that omits the contributing distraction. A 2×–3× under-capture adjustment yields a central estimate of ~12,000 true ED visits per year. On a US adult base of ~260 million, the central annual rate is roughly 42 per million, or 0.000042 per person-year. Compounding over 59 remaining adult years: 1 − (1 − 0.000042)⁵⁹ ≈ 0.00247, rounded to 0.0025 (~1 in 400). Heavy daily phone-on-walk users in dense urban environments plausibly reach 0.01 (1 in 100); light occasional users sit near 0.0005 (1 in 2,000).\n","uncertainty":{"low":0.0005,"high":0.01},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/23644536/","title":"Pedestrian injuries due to mobile phone use in public places","publisher":"Accident Analysis & Prevention (Nasar & Troyer, 2013)","source_type":"peer_reviewed","statistic":"1,506 estimated US ED visits in 2010 for pedestrian cell-phone injuries; doubled from 2005; talking 69.5%, texting 9.1%; ages 21–25 highest","excerpt":"\"Mobile-phone related injuries among pedestrians increased relative to total pedestrian injuries, and paralleled the increase in injuries for drivers, and in 2010 exceeded those for drivers.\"\n","source_date":"2013-08-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260312095905/https://pubmed.ncbi.nlm.nih.gov/23644536/","calculation_notes":"[Paraphrase from full text — paywalled] The 2010 NEISS-based estimate of 1,506 ED-treated pedestrian injuries with cell-phone product factor is the foundational anchor for this entry; reported in the full paper and widely cited in secondary coverage (e.g., Ohio State news release, Reuters Health 2013), though not quoted in the freely-available abstract above. Nasar predicted in interviews that the figure would double again by 2015, which Guyon et al. (2020) data corroborate for the broader cell-phone injury category. The 3,000–6,000 NEISS-captured floor and the 9,000–18,000 under-capture-adjusted central range derive from this seven-year trend extended through 2020-era smartphone saturation.\n","independence_note":"Nasar & Troyer drew directly from the CPSC's NEISS database, which samples approximately 100 US emergency departments and projects national estimates via established weighting. Guyon et al. (2020) used the same NEISS pipeline for a different demographic slice, providing independent corroboration of the upward trend.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7597570/","title":"Hold the Phone! Cell Phone-Related Injuries in Children, Teens, and Young Adults Are On the Rise","publisher":"Global Pediatric Health (Guyon et al., 2020)","source_type":"peer_reviewed","statistic":"Cell-phone injury rate rose from 17.1 to 138 per 100,000 (2002–2015) in patients ≤21; distracted mobility accounted for 47% of injuries in the 19–21 bracket","excerpt":"\"From 2002 to 2015, an estimated 38 063 patients aged 21 years old and younger sustained a cell phone-related injury. The overall rate of injuries for all ages increased from 17.1 injuries per 100 000 in 2002 to 138 injuries per 100 000 in 2015, an increase of over 700%.\"\n","source_date":"2020-10-29","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250208130819/https://pmc.ncbi.nlm.nih.gov/articles/PMC7597570/","calculation_notes":"Guyon's 2015 rate of 138 per 100,000 in the under-21 group, with ~47% from distracted mobility in the 19–21 bracket per the paper's age-stratified analysis (\"Ages 19–21: This group demonstrated highest motor vehicle involvement at 25%, with distracted mobility accounted for 47% of injuries\"), implies roughly 65 distracted-mobility cell-phone ED visits per 100,000 young adults per year in that bracket. Adult-wide per-capita rates are lower than peak-young-adult rates but more uniformly spread; the entry's 42-per-million central population rate is internally consistent with Guyon's age-skewed upper anchor. Guyon's \"distracted mobility\" includes biking and skateboarding alongside walking, so the walking-only subset is smaller.\n","independence_note":"Guyon et al. used NEISS data independently of Nasar & Troyer, covering a different patient age range (≤21) and a later time window (2002–2015 vs 2004–2010). Agreement on direction and magnitude of trend strengthens the central estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10039299/","title":"An Epidemiological Study of Cell Phone-Related Injuries of the Hand and Wrist Reported in United States Emergency Departments From 2011 to 2020","publisher":"Journal of Hand Surgery Global Online (McLaughlin et al., 2023)","source_type":"peer_reviewed","statistic":"50,487 weighted ED visits for cell-phone-related hand/wrist injuries 2011–2020; falls were the largest single mechanism at 29.8%","excerpt":"\"A total weighted estimate of 50,487 national cases presenting to emergency departments… Falls were the most common cause of injury, accounting for an estimated 15,047 (29.8%) cases nationwide.\"\n","source_date":"2023-03-22","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20251211090151/https://pmc.ncbi.nlm.nih.gov/articles/PMC10039299/","calculation_notes":"McLaughlin's hand/wrist subset captures roughly 5,000–7,000 cell-phone ED visits per year limited to one body region. The fall-mechanism dominance (29.8%) supports the entry's framing that the median phone-distracted injury is a trip-and-fall rather than a vehicle strike. Total body-wide ED visits for cell-phone-related injury are therefore meaningfully higher than the hand/wrist subset alone.\n","independence_note":"McLaughlin et al. used NEISS independently, with a different anatomical filter (hand/wrist only) and a later decade (2011–2020). The mechanism breakdown (falls > broken phones > texting-related) is mechanistic evidence about how phone use causes injury, complementing Nasar/Guyon's exposure-level estimates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5005686/","title":"The incidence of pedestrian distraction at urban intersections after implementation of a Streets Smarts campaign","publisher":"Injury Epidemiology (Violano et al., 2015)","source_type":"peer_reviewed","statistic":"Of 1,362 observed pedestrians at New Haven intersections, 8% used a digital device while crossing; 19% were distracted by some activity","excerpt":"\"8% were using a digital device (talking, texting, or looking down at it)… 9% were using ear buds/headphones… 19% were distracted by another activity at both intersections.\"\n","source_date":"2015-06-25","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250210025011/https://pmc.ncbi.nlm.nih.gov/articles/PMC5005686/","calculation_notes":"The 8% point-prevalence of phone use at the moment of crossing — combined with 9% headphone use, partly overlapping — quantifies exposure. International observational studies (Beijing 11.7%–21.8%, Melbourne 20%) suggest US urban rates today, after a decade of smartphone saturation since Violano's 2015 fieldwork, are plausibly 15%–25% of crossings, supporting a non-trivial population-level injury base.\n","independence_note":"Violano et al. is observational field data from a different research group than Nasar/Guyon/McLaughlin, providing an exposure denominator (how often pedestrians use phones) that the NEISS studies cannot supply.\n"}],"comparison_anchors":[{"label":"Pedestrian death (lifetime, US adult)","lifetime_us_adult":0.00124},{"label":"E-scooter serious injury (lifetime, regular rider)","lifetime_us_adult":0.039},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"minimal phone use while walking (talks only, no looking-down)","multiplier":0.2,"notes":"Violano observed 2%–4.5% of pedestrians talking on phone vs 8% overall device use; talking-only carries lower trip-and-fall risk than looking-down/texting"},{"factor":"heavy daily phone-on-walk user (texting/scrolling while walking commute)","multiplier":4,"notes":"Guyon et al. 2020 found distracted mobility accounted for 47% of cell-phone injuries in the 19–21 age band, the heaviest-use demographic"},{"factor":"age 18–34","multiplier":3,"notes":"Nasar & Troyer 2013: ages 21–25 had the highest injury count (1,003 over 7 years), followed by 16–20 (985); risk per capita roughly 3× the all-age average"},{"factor":"walks routinely on uneven terrain or stairs while phone-distracted","multiplier":2.5,"notes":"McLaughlin et al. 2023: falls accounted for 29.8% of cell-phone ED visits, the largest single mechanism; uneven surfaces and stairs amplify trip risk"},{"factor":"urban dense pedestrian environment (Manhattan, downtown SF, Chicago Loop)","multiplier":2,"notes":"Exposure scales with crossings per day and crowd density; observational studies in dense urban cores report 15%–25% phone use at crossings vs ~8% in lighter pedestrian areas"}],"short_label":"Phone-distracted walking injury","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"NEISS under-captures the true injury count for two compounding reasons. First, it only includes ED visits where a clinician explicitly coded a cell phone as a product factor — a sprained ankle from missing a curb while texting is often coded as a fall with no phone notation. Second, NEISS does not capture self-treated minor injuries (scrapes, bruises, mild sprains), which likely outnumber ED-presenting injuries by an order of magnitude. The 3,000–6,000/year NEISS figure is a floor; the under-capture-adjusted 9,000–18,000 figure drives the lifetime estimate. The dramatic framing — \"phone-zombie steps into traffic and is killed\" — is rare: pedestrian fatalities attributable to phone distraction are not reliably coded in FARS, and the population-level signal is dominated by trips, falls, walking into poles/walls, and missed steps on stairs. The strongest per-exposure risk factor is looking down at the screen (texting, scrolling, navigating) rather than talking; observational studies consistently find looking-down behavior more disruptive of gait and obstacle detection than hands-free conversation. The risk is also age-skewed: the under-30 demographic accounts for the majority of phone-distraction injuries both because they use phones more while walking and because they walk more overall.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"A pavement curb edge seen from above with a subtle outlined phone shape resting near it, flat vector illustration, no people."},"canonical_url":"https://likelier.app/phone-while-walking-injury","api_url":"https://likelier.app/api/fears/phone-while-walking-injury.json"},{"slug":"fireworks-injury","question":"What are the odds of being injured by fireworks?","quick_answer":"Fireworks send about 14,700 people to US emergency rooms a year, putting an adult's lifetime odds of injury near 1 in 385. Non-users tend to overestimate the danger while regular users underestimate their cumulative risk.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Fireworks injuries occupy a predictable annual news cycle: every late June, local TV stations run segments on emergency-room horror stories featuring mangled hands and facial burns, accompanied by a solemn CPSC press release urging caution. The coverage creates a reasonable but somewhat inflated sense of risk. No standalone survey asks Americans to estimate their personal probability of a fireworks injury, so the perceived side is editorial intuition. Most adults who handle consumer fireworks probably underestimate the risk relative to their specific exposure (they have done it many times without incident), while bystanders and non-users probably overestimate the population-level rate because the injuries are vivid and seasonally concentrated. The net effect is a perception anchored on dramatic hand and eye injuries that are real but rare.\n","rough_estimate":"non-users overestimate; regular users underestimate their cumulative risk","kind":"intuition"},"native":{"display":"~14,700 fireworks-related ER visits per year in the US (2024)","numerator":44,"denominator":1000000,"unit":"per year","population":"US residents, all ages"},"normalized":{"lifetime_us_adult":0.0026,"display":"~1 in 385 lifetime (US adult)","log_value":-2.585,"assumptions":"The CPSC reported an estimated 14,700 fireworks-related emergency department visits in 2024 (up 52% from the 2023 estimate of 9,700). Against a US population of approximately 335 million, this yields an annual rate of roughly 44 per 1,000,000, or 4.4 per 100,000.\nHowever, the 2024 spike may be anomalous. The 5-year average (2019-2023) is closer to 10,000-11,000 ER visits per year, yielding an annual rate of approximately 32 per 1,000,000. Using the 5-year average as the more stable estimate:\nAnnual rate: ~32 per 1,000,000 = 0.000032 Compounded over 59 years of remaining adult life: 1 − (1 − 0.000032)^59 ≈ 0.00189\nUsing the 2024 rate for the central estimate to reflect the upward trend in consumer fireworks use: 1 − (1 − 0.000044)^59 ≈ 0.0026 ≈ 1 in 385.\nThe CPSC also reported 11 fireworks-related deaths in 2024 and 8 in 2023. The 5-year average is approximately 9 deaths per year, yielding a lifetime fatality probability of roughly 1 in 630,000 — negligible relative to the injury probability.\nThis is a population-average figure. Adults who never handle fireworks face near-zero risk (bystander injuries are a small minority). Adults who handle consumer fireworks on July 4th face a per-session risk roughly 10-20x the population average.\n","uncertainty":{"low":0.0015,"high":0.0045},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2025/CPSC-Urges-Fireworks-Safety-Ahead-of-July-4th-Holiday","title":"CPSC Urges Fireworks Safety Ahead of July 4th Holiday","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"An estimated 14,700 fireworks-related ER injuries in 2024; 11 deaths; hands/fingers 36%, head/face/ears 22%; burns 37% of injuries; ages 25-44 largest share (32%)","excerpt":"\"Fireworks were reportedly associated with an estimated 14,741 injuries treated in U.S. hospital emergency departments during calendar year 2024 ... there were 11 reported fireworks-related deaths.\"\n","source_date":"2025-06-24","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503090621/https://www.cpsc.gov/Newsroom/News-Releases/2025/CPSC-Urges-Fireworks-Safety-Ahead-of-July-4th-Holiday","calculation_notes":"The CPSC 2024 fireworks report is the primary source for the native rate. 14,741 ER-treated injuries / 335 million US population = 44 per 1,000,000 annual rate. The 2024 figure is a 52% increase over 2023 (9,700), likely reflecting both increased consumer fireworks availability (many states have liberalized fireworks laws since 2020) and increased usage. The body- part distribution (hands/fingers 36%, head/face/ears 22%) and the burn share (37%) are consistent with prior years. The 11 deaths are mostly associated with misuse (holding lit devices, re-lighting duds) and device malfunction. The adult age-group share (25-44 at 32%, 15-24 at 24%) confirms that this is not primarily a pediatric injury — adults are the majority of victims.\n"},{"url":"https://www.cpsc.gov/s3fs-public/2023-Fireworks-Annual-Report.pdf","title":"2023 Fireworks Annual Report: Fireworks-Related Deaths, Emergency Department-Treated Injuries, and Enforcement Activities","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"An estimated 9,700 fireworks-related ER injuries in 2023; 8 deaths; hands/fingers 35%, head/face/ears 22%; burns 42%; teens 15-19 highest rate per capita","excerpt":"\"CPSC received reports of eight deaths and an estimated 9,700 injuries involving fireworks ... the parts of the body most often injured were hands and fingers (an estimated 35 percent).\"\n","source_date":"2024-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260424165925/https://www.cpsc.gov/s3fs-public/2023-Fireworks-Annual-Report.pdf","calculation_notes":"The 2023 CPSC report provides the lower bound of the recent injury range. 9,700 ER visits / 335 million = ~29 per 1,000,000 annual rate. Over 59 adult years: 1 − (1 − 0.000029)^59 ≈ 0.0017 ≈ 1 in 588. The 2023 figure is closer to the 2019 pre-pandemic baseline (~10,000) and may better represent the typical year than the 2024 spike. The 2023 report notes that teens 15-19 had the highest per-capita rate, consistent with the behavioral risk profile (adolescents are more likely to mishandle fireworks). The 8 deaths in 2023 versus 11 in 2024 are within normal year-to-year variation for a small-count outcome.\n"},{"url":"https://achi.net/newsroom/fireworks-injuries-in-u-s-increased-by-52-percent-in-2024/","title":"Fireworks Injuries in U.S. Increased by 52% in 2024","publisher":"American College of Healthcare Information (ACHI)","source_type":"news_article","statistic":"52% increase in fireworks ER injuries from 2023 to 2024; sparklers accounted for an estimated 1,700 injuries","excerpt":"\"Fireworks injuries in the U.S. increased by 52% in 2024, according to new data from the Consumer Product Safety Commission.\"\n","source_date":"2025-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260227173814/https://achi.net/newsroom/fireworks-injuries-in-u-s-increased-by-52-percent-in-2024/","calculation_notes":"ACHI's summary of the CPSC data highlights the year-over-year trend and the sparkler subcategory (1,700 of 14,700, or ~12% of injuries). Sparklers are often perceived as safe — they are the firework most commonly given to children — but burn at 1,200°F and produce a meaningful injury share. The 52% year-over-year increase is notable but should be interpreted cautiously: fireworks ER visits have fluctuated between 9,000 and 15,600 since 2019, and single-year spikes are common in this data series.\n"}],"comparison_anchors":[{"label":"Injury requiring ER visit from a fall (annual, US adult)","lifetime_us_adult":0.3},{"label":"Dog bite requiring ER visit (lifetime, US adult)","lifetime_us_adult":0.012},{"label":"Lightning strike (lifetime, US adult)","lifetime_us_adult":0.000065},{"label":"Accidental gun death (lifetime, US adult)","lifetime_us_adult":0.0000885}],"personal_factor_multipliers":[{"factor":"personally handles consumer fireworks on July 4th","multiplier":10,"notes":"The population-average rate includes the ~60% of adults who never handle fireworks. Among active handlers, the per-person annual risk is roughly an order of magnitude higher. CPSC data shows hands and fingers account for 36% of injuries, consistent with the handler being the primary victim."},{"factor":"never handles fireworks (bystander only)","multiplier":0.15,"notes":"Bystander injuries — struck by debris, burned by a misfired aerial — are a minority of the CPSC numerator. An adult who attends professional displays or watches from a safe distance faces a fraction of the population-average rate."},{"factor":"ages 15-24","multiplier":2.5,"notes":"Teens and young adults have the highest per-capita injury rate in CPSC data, driven by behavioral risk factors: holding lit devices, re-lighting duds, improvised devices."},{"factor":"uses illegal or professional-grade fireworks","multiplier":5,"notes":"Homemade and professional-grade devices are disproportionately represented in fireworks deaths and amputations. The CPSC notes that most deaths involve misuse or device malfunction, and non-consumer devices carry higher explosive loads."}],"short_label":"Fireworks injury","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry covers all fireworks-related emergency department visits as reported by the CPSC through NEISS, including burns, lacerations, contusions, and eye injuries. The majority of injuries are minor burns that resolve without lasting impairment; serious outcomes (amputations, permanent vision loss, death) are a small fraction of the numerator. The 2024 figure of 14,700 may represent a temporary spike rather than a new baseline; the 5-year average is closer to 10,000-11,000. The annual rate is heavily concentrated around July 4th (roughly two-thirds of annual injuries occur in the June 18 - July 18 window), making this a seasonal rather than year-round risk. The population-average lifetime figure of ~1 in 385 includes adults who never touch fireworks; for active handlers the lifetime risk is materially higher. Professional fireworks displays have a separate and much lower injury rate for spectators; the CPSC numerator is dominated by consumer-fireworks injuries. State-level variation is significant: states with liberal fireworks laws (e.g., Missouri, Texas) have higher per-capita injury rates than states that ban consumer fireworks.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single unlit firecracker resting on a pale surface, flat vector illustration in muted tones with a quiet red accent on the fuse."},"canonical_url":"https://likelier.app/fireworks-injury","api_url":"https://likelier.app/api/fears/fireworks-injury.json"},{"slug":"excessive-caffeine-harm","question":"What are the odds of harm from drinking too much coffee or caffeine?","quick_answer":"A documented floor of about 1 in 366 US adults over a lifetime has an ER visit attributable to high-caffeine beverages, drawn from roughly 12,000 energy-drink-only ER visits a year. Many heavy coffee drinkers assume high intake meaningfully raises their risk of serious cardiac harm, but the evidence for moderate beverage caffeine points the other way.\n","category":"health","tags":["food","substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Caffeine sits in an unusual perceptual spot: it is simultaneously the most widely consumed psychoactive substance on Earth and a recurring subject of \"is my coffee killing me\" health anxiety. The dread tends to spike around energy drinks and viral reports of caffeine deaths, most of which involve pure caffeine powder or tablets rather than any beverage. People who feel a racing heart after a third espresso often read it as a warning of cardiac harm, when for a healthy adult it is a transient, dose-dependent jitter. There is no rigorous, nationally representative survey that asks Americans to estimate their odds of being harmed by drinking too much coffee, so the perception here is inferred from the broader pattern of caffeine health worry rather than a single measured number.\n","rough_estimate":"Many heavy coffee drinkers assume high intake meaningfully raises their risk of serious cardiac harm; the evidence for moderate beverage caffeine points the other way","kind":"intuition"},"native":{"display":"~12,000 energy-drink-attributable ED visits (energy drinks only) per year in the US","numerator":12054,"denominator":260000000,"unit":"per year","population":"US residents age 12+ (DAWN surveillance, energy-drink-only ED visits, 2011)"},"normalized":{"lifetime_us_adult":0.00273,"display":"~1 in 366 lifetime (US adult, ED visit attributable to high-caffeine beverages — a documented floor)","log_value":-2.563,"assumptions":"The SAMHSA Drug Abuse Warning Network (DAWN) recorded 20,783 energy-drink- related emergency department (ED) visits in 2011, of which 58% (≈12,054) involved energy drinks only — no other drug or alcohol. DAWN sampled ED visits among US residents age 12 and older; the age-12+ US population in 2011 was approximately 260 million (US Census). The annual hazard of an energy-drink-only ED visit is therefore roughly 12,054 / 260,000,000 ≈ 4.6 × 10⁻⁵ per year, or about 1 in 21,600 per year.\nCompounded over 59 years of remaining adult life at constant hazard: 1 − (1 − 0.0000464)^59 ≈ 0.00273 ≈ 1 in 366. log₁₀(0.00273) ≈ −2.56.\nThis figure is deliberately framed as a LOWER BOUND, and the uncertainty band is skewed high to reflect three forms of undercount. (1) DAWN tracks only energy drinks as a discrete category; ED visits driven by ordinary coffee, tea, or soda overconsumption are not separately coded and are invisible here. (2) The 2011 reading is the last year of the old DAWN methodology; energy-drink consumption and the affected population have grown since. (3) Attribution requires a clinician to code the visit to energy drinks, which misses cases presenting as anxiety, palpitations, or insomnia without the drink being recorded. The true lifetime probability of any ED visit attributable to high-caffeine beverages is plausibly several times the central estimate — hence the wide band. The far more common outcomes of excess beverage caffeine (jitteriness, insomnia, transient palpitations, reflux) almost never reach an ED at all and are not counted in any surveillance system.\n","uncertainty":{"low":0.001,"high":0.02},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fda.gov/consumers/consumer-updates/spilling-beans-how-much-caffeine-too-much","title":"Spilling the Beans: How Much Caffeine is Too Much?","publisher":"U.S. Food and Drug Administration","source_type":"govt_report","statistic":"400 mg/day not generally associated with negative effects in healthy adults; toxic effects like seizures around 1,200 mg rapidly consumed","excerpt":"\"For most adults, the FDA has cited 400 milligrams a day — that's about two to three 12-fluid-ounce cups of coffee — as an amount not generally associated with negative effects. [...] The FDA estimates toxic effects, like seizures, can be observed with rapid consumption of around 1,200 milligrams of caffeine, or less than 1/2 teaspoon of pure caffeine.\"\n","source_date":"2024-08-28","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260617182827/https://www.fda.gov/consumers/consumer-updates/spilling-beans-how-much-caffeine-too-much","calculation_notes":"The FDA's 400 mg/day threshold is the reference line for \"too much\" in this entry: harms cluster above it in a dose-dependent way, and the 1,200 mg acute-toxicity figure clarifies that life-threatening caffeine effects require a rapid bolus — \"less than 1/2 teaspoon of pure caffeine\" — which is essentially unreachable by drinking coffee (a cup carries roughly 80–100 mg; reaching 1,200 mg would mean ~12 cups in quick succession). This frames fatal caffeine toxicity as a powder/pill phenomenon, not a beverage one, and anchors the \"what counts as too much\" boundary used throughout the entry. Both quoted sentences were confirmed verbatim against the live FDA page (content current as of 08/28/2024) during human review.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK384664/","title":"Update on Emergency Department Visits Involving Energy Drinks: A Continuing Public Health Concern (The DAWN Report)","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"20,783 energy-drink-related ED visits in 2011 (up from 10,068 in 2007); 58% involved energy drinks only","excerpt":"\"The estimated number of ED visits involving energy drinks doubled from 10,068 visits in 2007 to 20,783 visits in 2011. [...] In 2011, 58% of energy drink-related ED visits involved energy drinks only; the remaining 42% involved other drugs.\"\n","source_date":"2013-01-10","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20230314091232/http://www.ncbi.nlm.nih.gov/books/NBK384664/","calculation_notes":"This is the native-rate source. The 20,783 total × 58% energy-drinks-only share = ≈12,054 ED visits where a high-caffeine beverage was the sole implicated agent (the conservative numerator — excluding the 42% of visits that also involved alcohol or other drugs). Divided by the ~260 million US residents age 12+ in 2011, this gives the ~4.6 × 10⁻⁵ annual hazard that compounds to the headline 1-in-366 lifetime floor. The doubling from 2007 to 2011 is the basis for treating the static 2011 figure as an undercount and skewing the uncertainty band upward.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/36947466/","title":"Acute Effects of Coffee Consumption on Health among Ambulatory Adults (CRAVE trial)","publisher":"New England Journal of Medicine (Marcus et al., N Engl J Med 2023;388(12):1092-1100)","source_type":"peer_reviewed","statistic":"Caffeinated coffee not linked to significantly more premature atrial contractions (RR 1.09, 95% CI 0.98–1.20); ~50% more premature ventricular contractions and ~36 fewer minutes of sleep","excerpt":"\"The consumption of caffeinated coffee was associated with 58 daily premature atrial contractions as compared with 53 daily events on days when caffeine was avoided (rate ratio, 1.09; 95% confidence interval [CI], 0.98 to 1.20; P = 0.10). [...] The consumption of caffeinated coffee as compared with no caffeine consumption was associated with 154 and 102 daily premature ventricular contractions, respectively (rate ratio, 1.51; 95% CI, 1.18 to 1.94).\"\n","source_date":"2023-03-23","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260610100826/https://pubmed.ncbi.nlm.nih.gov/36947466/","calculation_notes":"The CRAVE randomized case-crossover trial (100 adults wearing continuous ECG, accelerometer, and glucose monitors for 14 days) is the best controlled evidence on the specific cardiac fear. It found no significant increase in atrial ectopy from caffeinated coffee, a real but modest increase in ventricular ectopy, and ~36 minutes less sleep on coffee days. It does not yield a population probability — it is used to calibrate the *severity* of the modal harm (transient ectopy and sleep loss, not sustained arrhythmia) and to justify the moderate_harm / recoverable_injury framing rather than a fatal one.\n"},{"url":"https://www.eufic.org/en/whats-in-food/article/efsa-opinion-on-the-safety-of-caffeine","title":"EFSA opinion on the safety of caffeine","publisher":"EUFIC, summarising EFSA Scientific Opinion on the safety of caffeine (EFSA Journal 2015;13(5):4102)","source_type":"reputable_reference","statistic":"Single doses up to 200 mg and total daily caffeine up to 400 mg safe for healthy adults","excerpt":"\"single doses of caffeine up to 200 mg (approximately 2 ½ espressos or 4 cups of tea) and total daily caffeine consumption of up to 400 mg are safe.\"\n","source_date":"2015-06-23","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260629145320/https://www.eufic.org/en/whats-in-food/article/efsa-opinion-on-the-safety-of-caffeine","calculation_notes":"EFSA independently corroborates the FDA's 400 mg/day daily threshold and adds a 200 mg single-dose limit for healthy adults. This is a cross-check on the \"what is too much\" boundary rather than a probability input. EFSA's separate, lower thresholds for pregnancy (200 mg/day) are noted in the caveats as a scope limit, not converted into a multiplier.\n"}],"comparison_anchors":[{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Any violent victimization (lifetime, US adult, all contexts)","lifetime_us_adult":0.55},{"label":"Struck by lightning (lifetime, US)","lifetime_us_adult":0.000067}],"personal_factor_multipliers":[{"factor":"slow CYP1A2 caffeine metabolizer, ≥4 cups/day","multiplier":1.64,"notes":"Cornelis et al. (JAMA 2006, n=2,014 nonfatal-MI cases + 2,014 controls) found an odds ratio of 1.64 (95% CI 1.14–2.34) for nonfatal myocardial infarction among slow CYP1A2 (*1F) metabolizers consuming ≥4 cups/day, versus no elevated risk in rapid metabolizers.\n"},{"factor":"rapid CYP1A2 caffeine metabolizer, ≥4 cups/day","multiplier":0.99,"notes":"In the same Cornelis et al. (JAMA 2006) study, rapid (*1A/*1A) metabolizers showed no increased nonfatal-MI risk at ≥4 cups/day (OR 0.99, 95% CI 0.66–1.48) — the gene–coffee interaction is the whole point (P = .04).\n"},{"factor":"moderate coffee drinker (~3 cups/day)","multiplier":0.83,"notes":"Net-protective, not a harm factor. Poole et al. (BMJ 2017) umbrella review reports a summary all-cause-mortality relative risk of 0.83 (95% CI 0.83–0.88) for coffee consumption, with the largest reduction at three to four cups/day versus none. Listed here to make the protective direction explicit on the calibration slider.\n"}],"short_label":"Too much caffeine","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"Scope: this entry measures realistic harms from caffeine consumed in beverages (coffee, tea, energy drinks, soda), not from caffeine powder or tablets. Nearly all fatal caffeine toxicity in the literature involves pure/anhydrous caffeine products, where a single gram-scale dose is easy to reach; the FDA's 1,200 mg acute-toxicity figure is roughly a dozen cups of coffee in quick succession and is effectively unreachable by drinking. The headline 1-in-366 lifetime figure is an explicit LOWER BOUND derived from DAWN energy-drink ED surveillance: it captures only energy drinks (not coffee/tea/soda overconsumption, which no surveillance system codes separately), only the 2011 reading (energy-drink use has since grown), and only ED-coded cases. The true rate is higher, which is why the uncertainty band runs up to 1 in 50. The number also conceals that the modal harm of excess beverage caffeine is transient and benign — jitteriness, insomnia, palpitations, and reflux that never reach a clinic — while serious cardiac events are rare and concentrated in susceptible subgroups. Crucially, moderate coffee is net-protective for several outcomes (all-cause mortality, Parkinson's, type 2 diabetes, liver disease) per large cohorts, so the \"too much coffee\" fear is overrated for the average healthy adult. The number does NOT apply to: pregnancy (EFSA caps habitual intake at 200 mg/day, half the adult limit); people with pre-existing arrhythmia or structural heart disease; diagnosed anxiety disorders, where caffeine reliably worsens symptoms; slow CYP1A2 metabolizers at high intake; and adolescents/heavy energy-drink consumers, who dominate the DAWN numerator and face higher acute risk per serving.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-06-12","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-12","last_reviewed":"2026-06-12","reviewed":true,"generated_at":"2026-06-12","image":{"alt":"A single overfull coffee cup with a faint ripple on the surface, flat vector illustration in muted warm browns and off-white."},"canonical_url":"https://likelier.app/excessive-caffeine-harm","api_url":"https://likelier.app/api/fears/excessive-caffeine-harm.json"},{"slug":"sickle-cell-crisis","question":"What are the odds of having sickle cell disease?","quick_answer":"Sickle cell disease occurs in about 1 in 365 Black or African American births in the US, and roughly 1 in 2,070 across all races. It is generally underestimated by the broad public, who recall the term from biology class but rarely the actual prevalence.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Sickle cell disease occupies a peculiar informational niche: well known as a textbook example of Mendelian genetics, poorly understood in terms of actual prevalence. Most Americans can recall the phrase \"sickle cell\" from a biology class but would struggle to distinguish trait from disease or estimate how many people are affected. Among Black Americans, awareness tends to be higher but still imprecise — many know someone with the trait but fewer grasp the 1-in-365 birth prevalence of the disease itself. No large-scale survey isolates \"fear of having sickle cell disease\" as a distinct item, so the perceived estimate here relies on editorial judgment informed by public-health literacy research.\n","rough_estimate":"generally underestimated by the broad public; somewhat better known in affected communities","kind":"intuition"},"native":{"display":"~1 in 365 Black or African American births in the US","numerator":1,"denominator":365,"unit":"per birth","population":"Black or African American newborns in the US"},"normalized":{"lifetime_us_adult":0.00274,"display":"~1 in 365 (subgroup birth prevalence)","log_value":-2.56,"assumptions":"CDC reports SCD occurs in approximately 1 out of every 365 Black or African American births. The 2016–2020 Sickle Cell Data Collection program across 11 states found a crude prevalence of 28.54 per 10,000 non-Hispanic Black newborns (approximately 1 in 350), broadly consistent with the established 1-in-365 figure. Because SCD is a genetic condition present at birth, the \"lifetime probability\" is simply the birth prevalence for the affected subgroup: 1/365 ≈ 0.00274. This is not annualized or compounded — it is the probability that a Black American newborn will have SCD. For the overall US population (all races), the birth prevalence is approximately 4.83 per 10,000 (1 in 2,070). The normalized figure here uses the subgroup-specific rate because the condition is overwhelmingly concentrated in this population.\n","uncertainty":{"low":0.0025,"high":0.003},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/sickle-cell/data/index.html","title":"Data and Statistics on Sickle Cell Disease","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"SCD affects approximately 100,000 Americans; occurs in about 1 in 365 Black or African American births and 1 in 16,300 Hispanic American births","excerpt":"\"SCD affects approximately 100,000 Americans. SCD occurs among about 1 out of every 365 Black or African American births. SCD occurs among about 1 out of every 16,300 Hispanic-American births. Sickle cell trait (SCT) occurs among about 1 in 13 Black or African American babies.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260424190035/https://www.cdc.gov/sickle-cell/data/index.html","calculation_notes":"CDC reports 1 in 365 Black/African American births have SCD. This yields a birth prevalence of 1/365 ≈ 0.002740. The 100,000 affected Americans figure is consistent with ~45 million Black Americans × 0.00274 prevalence, accounting for reduced life expectancy in SCD patients (median survival ~54 years vs ~77 years general population). No annualization needed — SCD is a congenital condition, so birth prevalence equals lifetime prevalence for the subgroup.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7312a1.htm","title":"Birth Prevalence of Sickle Cell Disease and County-Level Social Vulnerability — Sickle Cell Data Collection Program, 11 States, 2016–2020","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"SCD birth prevalence of 28.54 per 10,000 (1 in 350) among non-Hispanic Black newborns across 11 states, 2016–2020","excerpt":"\"During 2016–2020, a total of 3,305 confirmed SCD cases were identified among newborns in 11 states. The crude SCD birth prevalence was 4.83 per 10,000 live births overall and 28.54 per 10,000 among non-Hispanic Black newborns.\"\n","source_date":"2024-03-28","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426211512/https://www.cdc.gov/mmwr/volumes/73/wr/mm7312a1.htm","calculation_notes":"The MMWR study uses newborn screening data from 11 states participating in the Sickle Cell Data Collection program. The non-Hispanic Black birth prevalence of 28.54 per 10,000 equals 1 in 350, slightly higher than the traditionally cited 1 in 365. The difference likely reflects improved screening sensitivity and updated population denominators. The overall US birth prevalence of 4.83 per 10,000 (1 in 2,070) reflects dilution across all racial/ethnic groups. Both figures are consistent within expected variation.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20331952/","title":"Population estimates of sickle cell disease in the U.S.","publisher":"American Journal of Preventive Medicine","source_type":"peer_reviewed","statistic":"Estimated US SCD population of 72,000–98,000 (mortality-corrected), may approach 100,000 (2008)","excerpt":"\"National SCD population estimates ranged from 104,000 to 138,900, based on birth-cohort disease prevalence, but from 72,000 to 98,000 when corrected for early mortality. The number of individuals with SCD in the U.S. may approach 100,000, even when accounting for the effect of early mortality on estimations.\"\n","source_date":"2010-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260319000406/https://pubmed.ncbi.nlm.nih.gov/20331952/","calculation_notes":"Hassell (2010) used newborn screening data and survival estimates to compute population prevalence. The mortality-corrected estimate for 2008 was 72,000–98,000 (104,000–138,900 before correcting for early mortality), and the paper concludes the true number \"may approach 100,000.\" This is broadly consistent with the current CDC figure of ~100,000 affected Americans, reflecting improved survival (hydroxyurea, penicillin prophylaxis, transfusion therapy) and immigration from SCD-prevalent regions. This source supports the ~100,000-prevalence headline; the 1-in-365 birth prevalence is cited separately to CDC newborn-screening data above.\n"}],"comparison_anchors":[{"label":"Cystic fibrosis birth prevalence (white Americans)","lifetime_us_adult":0.00029},{"label":"Down syndrome birth prevalence (all US births)","lifetime_us_adult":0.00125},{"label":"Phenylketonuria birth prevalence (all US births)","lifetime_us_adult":0.0000625}],"personal_factor_multipliers":[{"factor":"Black or African American","multiplier":1,"notes":"Baseline subgroup — 1 in 365 births"},{"factor":"Hispanic American","multiplier":0.022,"notes":"~1 in 16,300 births"},{"factor":"White American (non-Hispanic)","multiplier":0.001,"notes":"Extremely rare; estimated <1 in 100,000 births"},{"factor":"West African or West-African-descent ancestry (vs US Black average)","multiplier":1.3,"notes":"WHO and CDC data show HbS allele carrier frequency of 20-30% in parts of West Africa (Nigeria, Ghana, Cameroon) compared to roughly 8% (1 in 13) in the US Black population, reflecting migration-pattern averaging; individuals with recent West African ancestry face somewhat higher birth prevalence than the US Black average"},{"factor":"both parents confirmed sickle cell trait carriers","multiplier":91,"notes":"Mendelian genetics: when both parents carry HbAS (sickle cell trait, 1-in-13 prevalence among Black Americans per CDC), each pregnancy has a 1-in-4 chance of SCD — approximately 25%, versus the population baseline of ~0.274% (1 in 365); ratio is approximately 91x the population rate"}],"short_label":"Sickle cell disease","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Sickle cell disease is a genetic condition, not an acquired risk — the \"probability\" here is birth prevalence, not an annual hazard rate. It is determined entirely by parental genotype. The 1-in-365 figure applies specifically to Black or African American newborns; for the overall US population (all races), birth prevalence is roughly 1 in 2,070. Sickle cell trait (carrying one copy of the gene) is far more common — about 1 in 13 Black Americans — but trait carriers generally do not develop the disease. SCD prevalence also varies by specific ancestry within the African diaspora, with higher rates among those with West African heritage. Life expectancy for SCD patients has improved significantly with hydroxyurea and other therapies but remains substantially below the general population average.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Abstract crescent shapes in red and muted tones, flat editorial illustration."},"canonical_url":"https://likelier.app/sickle-cell-crisis","api_url":"https://likelier.app/api/fears/sickle-cell-crisis.json"},{"slug":"counterfeit-medicine-death","question":"What are the odds of being harmed or killed by a counterfeit or substandard medicine?","quick_answer":"For an adult in a low- or middle-income country, the lifetime risk of dying from a substandard or falsified medicine is roughly 1 in 361, reflecting an estimated 188,000 such deaths worldwide each year. In high-income countries with strong drug-quality regulation, personal risk is orders of magnitude lower.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"In high-income countries with stringent pharmaceutical regulation, the idea of receiving a fake or substandard medicine feels like a plot device from a thriller rather than a routine hazard. Pharmacies are licensed, supply chains are audited, and regulatory agencies pull defective batches within days. The mental model of a medicine simply not containing the drug on its label does not map onto everyday experience. In low- and middle-income countries the situation is categorically different: WHO estimates that 1 in 10 medical products in these settings is substandard or falsified, and the consequences are not merely inefficacy but death, particularly among children treated with fake antibiotics for pneumonia or counterfeit antimalarials for malaria.\n","kind":"intuition"},"native":{"display":"~188,000 deaths per year globally from substandard or falsified medicines","numerator":188000,"denominator":5000000000,"unit":"per year","population":"global adults and children"},"normalized":{"lifetime_us_adult":0.00277,"display":"~1 in 361 lifetime (LMIC adult)","log_value":-2.56,"assumptions":"Native rate: WHO-commissioned models estimate 72,000-169,000 child deaths per year from pneumonia caused by substandard/falsified antibiotics (University of Edinburgh model) and 116,000 (64,000-158,000) additional deaths from malaria caused by substandard/falsified antimalarials in sub-Saharan Africa (London School of Hygiene and Tropical Medicine model). A conservative combined estimate of ~188,000 deaths/yr is used, reflecting overlap between model ranges. The burden falls almost entirely on low- and middle-income countries (LMICs), where ~4 billion people live and where WHO estimates 1 in 10 medical products is substandard or falsified. Dividing by the at-risk population: 188,000 / 4,000,000,000 = 4.7e-5 annual rate. Lifetime conversion using the 59-year horizon: 1 - (1 - 4.7e-5)^59 = 0.00277. Uncertainty low bound uses 100,000 deaths (conservative floor accounting for model uncertainty and possible double-counting between the Edinburgh and LSHTM models) / 4B compounded 59 years = 1 - (1 - 2.5e-5)^59 = 0.0015. High bound uses 370,000 (169,000 + 158,000 plus ~15% for non-pneumonia/non-malaria categories) / 4B compounded 59 years = 1 - (1 - 9.25e-5)^59 = 0.0055. The true death toll is likely higher since these models cover only pneumonia and malaria, not cardiovascular, HIV/AIDS, or TB medicines. For adults in high-income countries with robust drug-quality regulation, personal risk is orders of magnitude lower.\n","uncertainty":{"low":0.0015,"high":0.0055},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/substandard-and-falsified-medical-products","title":"Substandard and falsified medical products — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"At least 1 in 10 medicines in low- and middle-income countries are substandard or falsified; countries spend an estimated US$ 30.5 billion per year on such products","excerpt":"\"At least 1 in 10 medicines in low- and middle-income countries are substandard or falsified. Countries spend an estimated US$ 30.5 billion per year on substandard and falsified medical products.\"\n","source_date":"2018-01-31","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260415174758/https://www.who.int/news-room/fact-sheets/detail/substandard-and-falsified-medical-products","calculation_notes":"The WHO fact sheet has been restructured since the original access date. The detailed mortality models (University of Edinburgh pneumonia estimate of 72,000-169,000 deaths and LSHTM malaria estimate of 116,000 deaths) that previously anchored the native numerator are no longer present on this page. The 1-in-10 prevalence figure remains and establishes the scale of the problem in LMICs. The mortality estimates that underpin the 188,000 deaths/year figure were derived from earlier versions of this fact sheet and from the original research publications (Lancet Infectious Diseases, 2018). 188,000 / 5B = 0.0000376 annual rate, compounded over 59 years yields 0.00222.\n","independence_note":"Both sources are WHO publications drawing on the same underlying data and commissioned modelling studies. They are not independent data sources.\n"},{"url":"https://www.who.int/health-topics/substandard-and-falsified-medical-products","title":"Substandard and falsified medical products — Health topics","publisher":"World Health Organization","source_type":"govt_report","statistic":"More than 1 in 10 medicines in LMICs estimated substandard or falsified; up to 2 billion people lack access to necessary medicines","excerpt":"\"Up to two billion people around the world lack access to necessary medicines, vaccines, medical devices including in vitro diagnostics, and other health products, which creates a vacuum that is too often filled by substandard and falsified products. More than one in ten medicines in low- and middle-income countries are estimated to be substandard or falsified. No country remains untouched from this issue.\"\n","source_date":"2020-06-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260310165041/https://www.who.int/health-topics/substandard-and-falsified-medical-products","calculation_notes":"This WHO health-topics page confirms the 1-in-10 prevalence framing and establishes the access-gap context (2 billion people lacking necessary medicines). The previously cited sub-URL with \"hundreds of thousands\" mortality framing is no longer accessible; the current page focuses on prevalence, impact, and WHO response. The mortality estimates underpinning the 188,000 figure are supported by the original Lancet Infectious Diseases publications rather than this summary page.\n","independence_note":"Same WHO data as source 1; different summary page, not an independent data source.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4455082/","title":"Estimated Under-Five Deaths Associated with Poor-Quality Antimalarials in Sub-Saharan Africa","publisher":"The American Journal of Tropical Medicine and Hygiene — Renschler JP, Walters KM, Newton PN, Laxminarayan R","source_type":"peer_reviewed","statistic":"Approximately 122,350 (IQR: 91,577–154,736) under-five deaths in 39 sub-Saharan African countries in 2013 were associated with consumption of poor-quality antimalarials, representing ~3.75% of all under-five deaths in those countries","excerpt":"\"An estimated 122,350 (interquartile range [IQR]: 91,577–154,736) under-five malaria deaths were associated with consumption of poor-quality antimalarials across 39 sub-Saharan African countries in 2013. This represented 3.75% of all under-five deaths in our sample of countries.\"\n","source_date":"2015-06-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20250924001100/https://pmc.ncbi.nlm.nih.gov/articles/PMC4455082/","calculation_notes":"Renschler et al. (2015) is the peer-reviewed modeling paper that independently quantifies child deaths from poor-quality antimalarials, using WHO child mortality data and antimalarial failure rate estimates for 39 sub-Saharan African countries. The 122,350 central estimate (IQR 91,577–154,736) is consistent with the LSHTM model figure of 116,000 (64,000–158,000) cited in WHO documentation; slight differences reflect model assumptions and reference year (2013 here vs. the WHO model's reference year). The overlapping uncertainty intervals confirm the same order-of-magnitude burden. This study covers malaria deaths only; combined with the Edinburgh University pneumonia model (72,000–169,000 deaths), the composite ~188,000 estimate is conservative. Used here as the independent peer-reviewed anchor confirming the malaria component of the native rate.\n","independence_note":"Independent of the WHO sources above: this is an academic modelling study published in a peer-reviewed journal (Am J Trop Med Hyg), authored by researchers at Princeton and Oxford (Paul N Newton of MORU/Oxford; Ramanan Laxminarayan of Princeton), using WHO child mortality inputs but applying an independent methodological framework to estimate attributable deaths.\n"},{"url":"https://www.kff.org/mental-health/teens-drugs-and-overdose-contrasting-pre-pandemic-and-current-trends/","title":"Teens, Drugs, and Overdose: Contrasting Pre-Pandemic and Current Trends","publisher":"KFF — Nirmita Panchal and Sasha Zitter","source_type":"reputable_reference","statistic":"In 2023, seven of every ten counterfeit opioid pills were found by the DEA to contain a potentially deadly dose of fentanyl; adolescents may acquire such drugs via social-media platforms","excerpt":"\"In 2023, seven out of every ten counterfeit opioid pills were found by the Drug Enforcement Administration to contain a potentially deadly dose of the drug. ... Adolescents may acquire drugs through social media platforms, which are frequently scrutinized for lacking regulation around illegal drug sales.\"\n","source_date":"2024-10-15","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260623052414/https://www.kff.org/mental-health/teens-drugs-and-overdose-contrasting-pre-pandemic-and-current-trends/","calculation_notes":"Used only to support the high-income-country illicit-channel caveat, not the headline LMIC substandard-medicine figure. KFF carries the DEA \"7 of 10\" laboratory-testing statistic (an increase from 4 of 10 in 2021 and 6 of 10 in 2022), the \"76% of adolescent drug fatalities in 2023 involved fentanyl\" share, and the social-media acquisition framing. The 2024 decline among ages 15–24 (37.0% drop, 2023→2024) is from CDC NCHS Data Brief No. 549 (cited separately below). These are conditional/contamination and demographic figures, not a lifetime probability or a multiplier on the LMIC baseline, so they inform the caveat only.\n","independence_note":"Independent of the WHO and Renschler sources: KFF is a US health-policy analysis organization summarizing DEA laboratory data and CDC mortality statistics for a distinct (US illicit-fentanyl) phenomenon, not the LMIC substandard-medicine supply chain.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db549.htm","title":"Drug Overdose Deaths in the United States, 2023–2024","publisher":"National Center for Health Statistics (CDC) — Garnett MF, Miniño AM","source_type":"govt_report","statistic":"Between 2023 and 2024 the age-adjusted drug overdose death rate fell 26.2%; the rate among ages 15–24 fell 37.0% (from 13.5 to 8.5 per 100,000), the largest decrease of any age group; deaths involving synthetic opioids (incl. fentanyl) fell 35.6%","excerpt":"\"The age-adjusted drug overdose death rate decreased by 26.2% between 2023 and 2024. ... rates declining 37.0% among ages 15–24 (from 13.5 to 8.5 deaths per 100,000 population). ... Drug overdose deaths involving synthetic opioids other than methadone decreased the most (35.6%) from 2023 (22.2 deaths per 100,000 standard population) to 2024 (14.3).\"\n","source_date":"2026-01-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260618195504/https://www.cdc.gov/nchs/products/databriefs/db549.htm","calculation_notes":"NCHS Data Brief No. 549 (January 2026) supplies the 2024 decline figures cited in the US-teen illicit-fentanyl caveat. These are CDC provisional mortality rates, not a lifetime probability or a multiplier on the LMIC baseline, so they inform the caveat only and do not alter the headline normalized figure.\n","independence_note":"Independent of the WHO and Renschler sources: this is the US national vital-statistics agency reporting US overdose mortality, a distinct phenomenon from the LMIC substandard-medicine supply chain. Complements the KFF source, which summarizes the same CDC/DEA data for a lay audience.\n"}],"comparison_anchors":[{"label":"Death from rabies via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death from food poisoning (lifetime, US)","lifetime_us_adult":0.000019},{"label":"Death from household air pollution (lifetime, global adult)","lifetime_us_adult":0.0337}],"regional_breakdown":[{"region":"LMIC adults (~4 billion)","probability":0.00277,"notes":"WHO estimates 1 in 10 medical products in LMICs is substandard or falsified"},{"region":"Global average (all adults)","probability":0.00222,"notes":"Diluted across 5B adults; misleading because risk concentrates in LMICs"},{"region":"High-income countries (regulated pharmacies)","probability":0.00001,"notes":"Effectively negligible; robust pharmaceutical regulation and supply-chain integrity"}],"short_label":"Counterfeit medicine","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 188,000 deaths estimate is derived from two disease-specific models covering only pneumonia and malaria. The true global death toll from substandard and falsified medicines is almost certainly higher when cardiovascular drugs, HIV/AIDS antiretrovirals, tuberculosis medicines, and other therapeutic categories are included. The burden falls almost entirely on low- and middle-income countries with weak pharmaceutical regulatory systems. For any adult purchasing medicines through a regulated pharmacy in the US, EU, Japan, or other high-income country with robust drug-quality enforcement, the personal probability of encountering a substandard or falsified medicine is orders of magnitude lower than the global average. Online pharmacies operating outside regulatory oversight present a distinct and growing risk channel even in wealthy countries. A separate illicit channel — not included in the LMIC headline figure — is the sale of fentanyl-laced fake prescription pills (counterfeit \"Percocet\", \"Xanax\", \"Adderall\") to US teens and young adults, often via social-media platforms such as Snapchat. KFF, citing DEA laboratory data, reports that in 2023 seven of every ten counterfeit opioid pills contained a potentially deadly dose of fentanyl, and that 76% of adolescent drug fatalities in 2023 involved fentanyl, frequently from pills the user believed were a genuine prescription drug. This phenomenon is distinct in both mechanism and population from the LMIC substandard-medicine burden that anchors the headline number, and its toll is now declining: CDC NCHS Data Brief No. 549 (January 2026) reports the overdose death rate among ages 15–24 fell 37.0% between 2023 and 2024 (from 13.5 to 8.5 per 100,000), the largest decrease of any age group.\n","quality_score":{"d1":4,"d2":3,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":3.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a medicine capsule with a subtle question mark shadow, rendered in muted tones against a pale background."},"canonical_url":"https://likelier.app/counterfeit-medicine-death","api_url":"https://likelier.app/api/fears/counterfeit-medicine-death.json"},{"slug":"police-officer-duty-death","question":"What are the odds of a US police officer dying in the line of duty over a career?","quick_answer":"About 1 in 355 over a full 25-year career (~11 traumatic deaths per 100,000 officers a year). Policing is genuinely more dangerous than desk work, but the traumatic line-of-duty death risk stays well under one percent — below what popular imagination suggests.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Policing is widely understood to be one of the more dangerous occupations, and public perception — shaped by news coverage of officer ambushes, line-of-duty memorial ceremonies, and political debates over officer safety — tends to place it in an exceptionally hazardous category. Many people would guess that a career in law enforcement carries a meaningful chance of a violent death on the job. The fear is directionally correct — policing is genuinely more dangerous than most desk-based work — but the magnitude of the risk is substantially lower than both popular imagination and occupational mythology suggest. No standalone survey measuring public estimates of police career mortality was identified; perceived risk is characterized here as editorial intuition based on media salience and political framing of officer safety.\n","rough_estimate":"most people overestimate police career mortality; the actual traumatic line-of-duty death risk over a 25-year career is under one percent","kind":"intuition"},"native":{"display":"~11.3 traumatic occupational fatalities per 100,000 police/sheriff officers per year (BLS CFOI, 2023)","numerator":113,"denominator":1000000,"unit":"fatal occupational injuries per officer per year","population":"US police and sheriff's patrol officers (BLS CFOI occupation), 2023"},"normalized":{"lifetime_us_adult":0.0028,"display":"~1 in 355 over a 25-year career","log_value":-2.55,"assumptions":"Reference subgroup: a US police or sheriff's patrol officer serving a full 25-year career (the modal career length, reflecting average entry around age 25 and retirement between ages 50–55 under most state pension systems). The headline annual rate is the BLS Census of Fatal Occupational Injuries (CFOI) fatal-occupational-injury rate for \"police and sheriff's patrol officers\": 11.3 deaths per 100,000 officers for 2023 (BLS CFOI, reported by USAFacts). CFOI is the right basis here because it is coverage-matched by construction — BLS computes deaths per 100,000 full-time-equivalent workers, so the numerator (officer fatalities) and denominator (the officer workforce) describe the same population — and it counts traumatic fatal occupational INJURIES only, matching this entry's stated scope (it excludes illness/disease deaths, including COVID-19). Annual rate: 11.3/100,000 = 0.000113/year. Lifetime probability over a 25-year career: 1 − (1 − 0.000113)^25 ≈ 0.00282, rounded to 0.0028 (about 1 in 355). This figure covers only traumatic fatal injuries; it explicitly excludes occupational disease, 9/11-related illness, and COVID-19 deaths, which NLEOMF tracks separately and which account for roughly 35–50% of total officer line-of-duty deaths in recent years. The raw FBI LEOKA felonious-plus-accidental counts (a five-year mean of ~104 deaths per year across 2019–2023) are retained in the sources below only as the death-COMPOSITION breakdown (how officer traumatic deaths split between felonious acts and accidents) — they are NOT divided by the FBI's 600,120 figure to produce a per-officer rate, because 600,120 is the single-year 2023 count of officers only at the 10,884 agencies that reported ASSAULTS to LEOKA, not the national sworn workforce. Dividing a national death count by an assault-reporting subset denominator is a numerator/denominator coverage mismatch that inflates the per-officer rate; CFOI avoids this because its rate is coverage-matched. The scope is declared activity_specific_lifetime because this is career-specific risk for a defined occupational subgroup, not a general US-adult lifetime probability. Uncertainty band: the low (8.2/100,000 → ~0.0020) is the BLS CFOI 2023 rate for the broader \"Protective service\" occupation group (BLS Table 4); the high (14.4/100,000 → ~0.0036) is a coverage-matched cross-check dividing the LEOKA ~104-deaths/year mean by the true national full-time sworn workforce of ~720,000 officers (FBI-based national count), which also reflects the elevated 2021 fatality year.\n","uncertainty":{"low":0.002,"high":0.0036},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.fbi.gov/how-we-can-help-you/more-fbi-services-and-information/ucr/leoka","title":"Law Enforcement Officers Killed and Assaulted (LEOKA) — FBI UCR Program","publisher":"Federal Bureau of Investigation, Uniform Crime Reporting Program","source_type":"govt_report","statistic":"2019–2023 annual felonious and accidental officer deaths: 89, 93, 129, 118, and approximately 92 respectively; five-year mean approximately 104 traumatic deaths per year (used here for death-composition only, not as a per-officer rate)","excerpt":"\"In 2022, 60 officers died as a result of felonious acts and 58 officers died in accidents, for a total of 118 line-of-duty deaths. In 2021, 73 officers were feloniously killed and 56 died accidentally (129 total). From 2021 to 2023, more officers were feloniously killed (194) than in any other consecutive three-year period in the past 20 years.\"\n","source_date":"2024-05-14","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260112063251/https://www.fbi.gov/how-we-can-help-you/more-fbi-services-and-information/ucr/leoka","calculation_notes":"LEOKA annual reports provide felonious + accidental death counts separately. These are the traumatic categories analogous to \"line of duty\" in the narrow sense; they exclude illness deaths. Five-year series: 2019 (89) + 2020 (93) + 2021 (129) + 2022 (118) + 2023 (~92) = 521 total / 5 years = 104.2/year. This mean is retained here ONLY as the death-COMPOSITION breakdown — the split of officer traumatic deaths between felonious acts and accidents. It is deliberately NOT divided by the FBI's 600,120 figure to produce a per-officer rate: the 2023 Special Report states that 10,884 agencies employing 600,120 officers reported 79,091 assaults, so 600,120 is the single-year count of officers only at the assault-REPORTING subset of agencies, not the national sworn workforce. Because the LEOKA death numerator spans officers and agencies outside that assault-reporting subset, dividing ~104 by 600,120 (the old basis, which yielded 17.3/100,000 and a 1-in-231 career figure) is a numerator/denominator coverage mismatch that inflates the per-officer rate. The headline rate is instead taken from the coverage-matched BLS CFOI series (11.3 per 100,000 police/sheriff officers, 2023), whose numerator and denominator describe the same population. As a cross-check, dividing ~104 LEOKA deaths by the true national full-time sworn workforce of ~720,000 gives ~14.4/100,000 — the upper end of the uncertainty band, and materially below the discredited 17.3 figure. Career probability at the CFOI headline rate over 25 years: 1 − (1 − 0.000113)^25 ≈ 0.0028 (0.28%, roughly 1 in 355).\n","independence_note":"FBI LEOKA is the primary federal government data collection on officer deaths, drawing from voluntary agency-level reporting across all 50 states and federal agencies. It is methodologically distinct from the NLEOMF count (which uses a broader definition including medical events and illness). Both sources are cross-checked here.\n"},{"url":"https://nleomf.org/2023-law-enforcement-fatalities-report-reveals-law-enforcement-deaths-dropped/","title":"2023 Law Enforcement Fatalities Report Reveals Law Enforcement Deaths Dropped","publisher":"National Law Enforcement Officers Memorial Fund (NLEOMF)","source_type":"reputable_reference","statistic":"2023: 136 total line-of-duty deaths; 47 by gunfire, 37 traffic-related, 5 COVID-19, 35 fatal medical events, 12 other; traumatic deaths (gunfire + traffic) = 84","excerpt":"\"In 2023, 136 federal, state, county, municipal, military, tribal, and campus officers died in the line of duty, representing a 39 percent decrease compared to the 224 officers who died in the line of duty in 2022. Firearms-related fatalities claimed the lives of 47 officers in 2023 — a 25 percent decrease from the 63 officers killed by gunfire in 2022. Traffic-related fatalities decreased 27 percent, with 37 deaths in 2023 compared to 51 in 2022.\"\n","source_date":"2024-01-10","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260120002848/https://nleomf.org/2023-law-enforcement-fatalities-report-reveals-law-enforcement-deaths-dropped/","calculation_notes":"NLEOMF's 2023 total of 84 traumatic deaths (47 gunfire + 37 traffic) cross-validates the LEOKA estimate of ~92 for 2023 (60 felonious + ~32 accidental). The ~8-unit gap reflects definitional differences: NLEOMF uses \"traffic-related\" which may subset differently from LEOKA's \"accidental,\" and LEOKA covers some federal agencies not in the NLEOMF count. The NLEOMF figure of 47 gunfire deaths is consistent with FBI LEOKA's 60 felonious deaths (LEOKA includes non-firearm felonious causes; NLEOMF's 47 is firearm-specific regardless of felonious/accidental classification). Headline point estimate uses the FBI LEOKA 5-year mean; NLEOMF provides the 2023 cross-check.\n","independence_note":"NLEOMF is an independent nonprofit that maintains its own officer death registry, drawing on multiple sources including agency reports and media. Its definition of \"line of duty death\" is broader than FBI LEOKA's felonious+accidental categories, notably including occupational illness. The NLEOMF traumatic-only subset (gunfire + traffic) is used here to align methodologically with LEOKA's scope.\n"},{"url":"https://www.bls.gov/news.release/cfoi.t04.htm","title":"Table 4. Fatal work injury rates per 100,000 full-time equivalent workers by selected occupations, 2021-23 — Census of Fatal Occupational Injuries","publisher":"U.S. Department of Labor, Bureau of Labor Statistics","source_type":"govt_report","statistic":"Fatal work injury rate per 100,000 full-time equivalent workers: Protective service occupation group 9.4 (2021), 10.2 (2022), 8.2 (2023); all workers 3.6 (2021), 3.7 (2022), 3.5 (2023)","excerpt":"\"Table 4. Fatal work injury rates(1) per 100,000 full-time equivalent workers by selected occupations, 2021-23 [...] All workers 3.6 3.7 3.5 [...] Protective service 9.4 10.2 8.2 [...] 1 Fatal injury rates are per 100,000 full-time equivalent (FTE) workers. [...] CFOI fatal injury counts exclude illness-related deaths unless precipitated by an injury event.\"\n","source_date":"2024-12-19","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260104203950/https://www.bls.gov/news.release/cfoi.t04.htm","calculation_notes":"BLS CFOI is the coverage-matched basis for this entry: it computes fatal occupational injuries per 100,000 full-time-equivalent workers, so the numerator (officer fatalities) and denominator (the officer workforce) describe the same population, and it counts traumatic fatal injuries only (illness-related deaths are excluded unless precipitated by an injury event), matching this entry's stated scope. Table 4 publishes the broad \"Protective service\" SOC group rate (8.2/100,000 in 2023), used here as the uncertainty low; the police-and-sheriff's-patrol-officers-specific 2023 rate of 11.3/100,000 (the headline basis) comes from BLS CFOI's detailed occupation series as reported by USAFacts. All-workers rate (3.5/100,000, 2023) supplies the general- occupation comparison in the prose. Annual headline rate 0.000113 → 25-year career 1 − (1 − 0.000113)^25 ≈ 0.0028 (~1 in 355).\n","independence_note":"BLS CFOI is a federal government census of fatal occupational injuries, methodologically independent of the FBI LEOKA collection (which counts officer deaths from agency reporting) and of the NLEOMF registry. Using CFOI for the headline rate removes the numerator/denominator coverage mismatch that afflicted the prior LEOKA-count-over- assault-reporting-workforce quotient.\n"},{"url":"https://usafacts.org/articles/how-many-police-officers-die-in-the-line-of-duty/","title":"How many police officers die in the line of duty in the US?","publisher":"USAFacts","source_type":"reputable_reference","statistic":"BLS CFOI 2023: police officers' fatal-injury rate was 11.3 deaths per 100,000 officers; police ranked 20th among occupations for on-the-job deaths (logging highest at 98.9 per 100,000)","excerpt":"\"The BLS did not rank policing in the top 10 deadly professions in the US in 2023, the most recent year of available data. BLS ranks professions by the number of on-the-job fatalities per 100,000 full-time equivalent employees. [...] officers' rate was 11.3 deaths per 100,000 officers. [...] Police officers ranked 20th among occupations for on-the-job deaths.\"\n","source_date":"2024-05-16","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260525132235/https://usafacts.org/articles/how-many-police-officers-die-in-the-line-of-duty/","calculation_notes":"USAFacts republishes the BLS Census of Fatal Occupational Injuries police-and-sheriff's- patrol-officers rate of 11.3 deaths per 100,000 officers for 2023 — the headline basis for this entry. This is coverage-matched (BLS computes deaths per 100,000 FTE) and injury-only (excludes illness/COVID-19), matching the entry's scope. Annual rate 11.3/100,000 = 0.000113/year → 25-year career 1 − (1 − 0.000113)^25 ≈ 0.0028 (~1 in 355). The article notes policing ranked 20th among occupations for on-the-job deaths in 2023, well below logging (98.9/100,000).\n","independence_note":"USAFacts is a nonpartisan nonprofit that compiles and republishes US government statistics with source attribution; the 11.3/100,000 figure is attributed by USAFacts to the Bureau of Labor Statistics CFOI. It is cited here because BLS's detailed occupation-level CFOI rate for police officers is not exposed on a single stable bls.gov HTML page, whereas BLS Table 4 (also cited) provides the coverage-matched methodology and the adjacent Protective-service group rate.\n"}],"comparison_anchors":[{"label":"US combat soldier (Iraq/Afghanistan peak deployment era, career)","lifetime_us_adult":0.0371},{"label":"Commercial fisherman (career, US)","lifetime_us_adult":0.025},{"label":"US average worker (all-cause occupational traumatic death, career)","lifetime_us_adult":0.0013}],"personal_factor_multipliers":[{"factor":"Active patrol assignment vs administrative/desk role","multiplier":3,"notes":"FBI LEOKA 2022 shows the leading felonious death circumstances are ambushes (12), investigative/enforcement activities (12), disturbance calls (6), and tactical situations (6). All are patrol-intensive. Administrative and desk-assigned officers face near-zero exposure to these scenarios. Patrol officers constitute roughly one-third of sworn staff but account for essentially all felonious-death circumstances in the LEOKA data."},{"factor":"Prior-year assault history as a proxy for high-crime beat","multiplier":2,"notes":"FBI LEOKA 2023 reports 79,091 officer assaults — the highest assault rate in 10 years at 13.2 per 100 officers. Officers assigned to beats with assault rates in the highest quartile face roughly double the assault and death exposure of those in lower-rate areas. No single-city LEOKA breakdown is published at sub-agency level, but BJS city-level UCR data confirm a 2-3x gradient across cities ranked by violent crime rate."},{"factor":"Overnight / night-shift patrol (8 pm–2 am peak) vs day shift","multiplier":2,"notes":"FBI LEOKA assault-by-time-of-day data (ucr.fbi.gov/leoka/2019/topic-pages/officers-assaulted) show ~12% of assaults per 2-hour window during the 8 pm–2 am block versus ~3.3% per 2-hour window from 4 am–8 am — roughly a 3-4x assault-rate ratio at the extremes. The overnight patrol hazard is concentrated on patrol-intensive hours; the multiplier of 2x reflects the duty-shift average for officers working evenings through early morning compared to a standard day shift, consistent with the general pattern documented across multiple LEOKA annual reports."},{"factor":"First 0–4 years of service vs career average","multiplier":1.8,"notes":"Tucker-Gail et al. (2010, International Journal of Police Science & Management; 2022, same journal) analyzed FBI LEOKA felonious death data from 1995–2015 and found that 0–4 years of experience had the highest frequency of death across both study windows — a pattern that strengthened in the expanded 20-year analysis. The 2004 LEOKA annual report records 14 of 32 feloniously killed officers had ≤4 years of service (44%), roughly 2x the ~20% expected share if deaths were uniformly distributed across a 25-year career. The 1.8x multiplier is a conservative estimate consistent with this data and the Tucker-Gail findings; the combination of 0–4 years tenure and age 30–39 was identified as particularly dangerous in both studies. Mechanism: inexperienced officers more likely to be in sole patrol situations, less practiced at de-escalation, and disproportionately assigned to overnight shifts and high-crime beats."},{"factor":"Consistently wearing NIJ-rated body armor (protective factor)","multiplier":0.5,"notes":"Liu et al. (2017, Journal of Occupational and Environmental Hygiene, 14(2):73–80; n=637; PubMed PMID 27715652) analyzed FBI LEOKA data 2002–2011 and found officers shot in the torso who wore body armor were 76% less likely to be killed than unarmored officers (controlling for confounders). Because gunfire accounts for approximately 51% of traumatic line-of-duty deaths in recent LEOKA years (47 of ~92 in 2023), consistently worn armor reduces overall traumatic fatality risk by roughly half (0.76 × 0.51 ≈ 0.39 direct gunfire-torso reduction, with residual benefit for other torso trauma, yielding ~0.5x overall). The NIJ Compliance Testing Program certifies Level IIA through Level IV armor; Level IIIA soft body armor covers the full handgun caliber range that accounts for ~64% of officer firearm deaths (LEOKA 2022–2023)."}],"short_label":"Police duty death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This figure covers only traumatic fatal occupational injuries (felonious acts and on-duty accidents), as measured by the BLS Census of Fatal Occupational Injuries. It excludes deaths from occupational disease, cancer linked to chemical exposures, heart disease, 9/11-related illness, and COVID-19, which the National Law Enforcement Officers Memorial Fund tracks separately and which account for roughly 35–50% of total officer line-of-duty deaths in recent years. Including those categories would roughly double the headline figure. The headline rate is the BLS CFOI police/sheriff-officer rate (11.3 per 100,000, 2023), which is coverage-matched by construction — deaths per 100,000 full-time-equivalent workers. An earlier version of this entry instead divided the FBI LEOKA five-year traumatic-death mean (~104/year) by 600,120, but 600,120 is only the count of officers at the agencies that reported ASSAULTS to LEOKA in 2023, not the national sworn workforce (~720,000+); that denominator was too small and inflated the rate to 17.3 per 100,000 and the career figure to 1 in 231. The corrected coverage-matched rate is lower. The career length of 25 years is a midpoint estimate; some officers retire at 20 years, others serve 30+. Non-fatal serious injuries number approximately 10,000–20,000 per year (BLS injury data), far exceeding the traumatic death toll.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A police badge resting on a plain dark surface, flat vector illustration."},"canonical_url":"https://likelier.app/police-officer-duty-death","api_url":"https://likelier.app/api/fears/police-officer-duty-death.json"},{"slug":"homicide-us","question":"What are the odds of being murdered in the US?","quick_answer":"The lifetime odds of being murdered in the US are about 1 in 339 for the average adult, from a 2024 rate of about 5.0 per 100,000 per year — the lowest since 2015. Many people guess closer to 1 in 100, and this pooled average hides large differences by age, sex, race, and geography.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Gallup's annual crime-worry poll asks Americans how frequently they worry about a list of specific crimes happening to them. In the October 2025 wave, 22% of US adults said they worry \"frequently\" or \"occasionally\" about being murdered. That places homicide roughly in the middle of the crime-worry list — below property crimes, above sexual assault — and is broadly stable year-over-year even as measured rates move.\n","rough_estimate":"~1 in 100 lifetime feels about right to many respondents","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"native":{"display":"~5.0 per 100,000 per year","numerator":50,"denominator":1000000,"unit":"per year","population":"US residents, all ages and demographics pooled"},"normalized":{"lifetime_us_adult":0.00295,"display":"1 in ~339 lifetime (US adult)","log_value":-2.53,"assumptions":"Revised 2026-06 to the FBI's official 2024 UCR rate of 5.0 per 100,000 per year, down 14.9% from 2023 (the largest one-year homicide decline on record and the lowest rate since 2015), after a ~10% drop in 2023. This continues the normalization of the 2020–2021 pandemic spike (which peaked at ~6.7 per 100,000). Assumes 59 years of remaining adult life and constant annual hazard. Compounded: 1 − (1 − 0.00005)^59 ≈ 0.00295 ≈ 1 in 339. The uncertainty band now runs from ~4.0 per 100,000 (the level the Council on Criminal Justice projects 2025 may reach when full FBI data land — potentially the lowest since 1900) up to ~6.7 per 100,000 (the 2020 peak), the realistic envelope over a 59-year horizon rather than sampling error. This is a population average that pools across age, sex, race, and geography; see caveats.\n","uncertainty":{"low":0.00236,"high":0.00395},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/fastats/homicide.htm","title":"FastStats — Assault or Homicide","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"20,162 homicide deaths in the US in 2024; rate of 5.9 per 100,000","excerpt":"\"Number of deaths: 20,162. Deaths per 100,000 population: 5.9. Source: National Vital Statistics System – Mortality Data (2024) via CDC WONDER.\"\n","source_date":"2024-12-31","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172818/https://www.cdc.gov/nchs/fastats/homicide.htm","calculation_notes":"CDC NVSS counts homicides from death certificates coded as assault (ICD-10 X85–Y09, Y87.1). Annual rate 5.9 / 100,000 = 0.000059 per person-year. Lifetime over 59 adult years: 1 − (1 − 0.000059)^59 ≈ 0.00348 ≈ 1 in 287. Uncertainty band reflects the 2014–2021 US range (~4.5 to ~7.8 per 100,000), which is the realistic envelope over a 59-year horizon rather than a statistical sampling error.\n","independence_note":"CDC NVSS is derived from death certificates filed by medical examiners and coroners; BJS/FBI numbers come from law-enforcement incident reports. They count meaningfully differently and serve as independent checks on each other.\n"},{"url":"https://www.fbi.gov/news/press-releases/fbi-releases-2024-reported-crimes-in-the-nation-statistics","title":"FBI Releases 2024 Reported Crimes in the Nation Statistics","publisher":"Federal Bureau of Investigation (UCR)","source_type":"govt_report","statistic":"US murder and nonnegligent manslaughter rate fell 14.9% in 2024 to 5.0 per 100,000 — largest one-year decline on record and lowest since 2015; 2020 peak was 6.7 per 100,000","excerpt":"\"Murder and non-negligent manslaughter recorded a 2024 estimated nationwide decrease of 14.9% compared to the previous year.\" The 2024 rate of 5 per 100,000 inhabitants is the lowest recorded since 2015; over the prior 20 years the rate peaked at 6.7 per 100,000 in 2020 and reached its lowest at 4.4 per 100,000 in 2014.\n","source_date":"2025-08-05","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260708002536/https://www.fbi.gov/news/press-releases/fbi-releases-2024-reported-crimes-in-the-nation-statistics","calculation_notes":"FBI UCR (NIBRS-based) reports a 2024 murder/nonnegligent-manslaughter rate of 5.0 per 100,000, a 14.9% year-over-year decline. (A 14.9% drop to 5.0 implies a 2023 base of ~5.9 per 100,000, consistent with the CDC NVSS and BJS 2023 figures used as cross-checks above; the secondary press number of \"5.7\" for 2023 does not reconcile exactly with the official 14.9% decline, so we cite only the verified 2024 rate and percent change.) Annual rate 5.0 / 100,000 = 0.00005 per person-year. Lifetime over 59 adult years: 1 − (1 − 0.00005)^59 ≈ 0.00295 ≈ 1 in 339. The new headline replaces the prior ~5.9 (2023–2024) figure and reflects the post-pandemic decline. Uncertainty low bound 0.00236 uses the Council on Criminal Justice ~4.0 per 100,000 projection for 2025 (see separate CCJ source); high bound 0.00395 uses the 2020 peak of ~6.7 per 100,000.\n","independence_note":"FBI UCR aggregates law-enforcement incident reports nationally; it is the same agency pipeline as BJS victimization but is the primary, most timely official rate series and is independent of CDC's death-certificate NVSS count. The 14.9% 2024 decline is corroborated by CBS News and the Center for American Progress analyses of the same release.\n"},{"url":"https://counciloncj.org/crime-trends-in-u-s-cities-year-end-2025-update/","title":"Crime Trends in U.S. Cities: Year-End 2025 Update","publisher":"Council on Criminal Justice","source_type":"reputable_reference","statistic":"Homicides in 35 study cities fell 21% in 2025 vs 2024; CCJ projects the 2025 national rate could reach ~4.0 per 100,000 — potentially the lowest since 1900","excerpt":"\"That would be the lowest rate ever recorded in law enforcement or public health data going back to 1900, and would mark the largest single-year percentage drop in the homicide rate on record.\"\n","source_date":"2026-01-22","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260711204154/https://counciloncj.org/crime-trends-in-u-s-cities-year-end-2025-update/","calculation_notes":"CCJ aggregates reported homicides across 35 large US cities and projects that if the FBI's full 2025 report tracks the observed 21% city-level decline, the national 2025 rate would fall to about 4.0 per 100,000. This 4.0 figure (≈0.00004 per person-year) is the basis for this entry's uncertainty low bound: 1 − (1 − 0.00004)^59 ≈ 0.00236. Treated as the floor of the band, not the point estimate, because full-year FBI data are not yet final.\n","independence_note":"CCJ is a non-governmental research body compiling municipal police data; methodologically independent of both the FBI's national UCR aggregation and CDC's death-certificate NVSS.\n"},{"url":"https://bjs.ojp.gov/library/publications/homicide-victimization-united-states-2023","title":"Homicide Victimization in the United States, 2023","publisher":"US Bureau of Justice Statistics","source_type":"govt_report","statistic":"US homicide victimization rate 5.9 per 100,000 in 2023; male rate 9.3 vs female 2.6; Black rate 21.3 vs white 3.2","excerpt":"\"The rate of homicide victimization was 5.9 per 100,000 persons. The male homicide victimization rate (9.3 per 100,000 persons) was 3.5 times greater than the homicide victimization rate for females (2.6 per 100,000). The homicide victimization rate for black persons (21.3 per 100,000 persons) was more than 6 times the rate for white persons (3.2 per 100,000).\"\n","source_date":"2025-07-10","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172849/https://bjs.ojp.gov/library/publications/homicide-victimization-united-states-2023","calculation_notes":"BJS computes victimization rates from FBI Supplementary Homicide Reports / NIBRS. Their 2023 rate of 5.9 per 100,000 matches CDC's 2024 figure to the first decimal, giving us a robust central estimate. The demographic breakdowns feed directly into the heterogeneity caveat: a factor-of-~7 spread across race and ~3.5 across sex means the pooled average is the wrong number for any specific reader.\n","independence_note":"FBI/BJS law-enforcement incident data and CDC/NVSS death-certificate data are collected through entirely different pipelines by different federal agencies. Agreement to the first decimal is strong corroboration, not circular.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"22% of US adults worry frequently or occasionally about being murdered (October 2025)","excerpt":"\"Fewer Americans say they worry about crimes, such as having a car stolen (39%) or their home burglarized (34%), being a victim of a hate crime (30%), or getting mugged (29%), attacked while driving (27%), murdered (22%) or sexually assaulted (21%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172538/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for the perceived-risk side only. The 22% figure is the fraction of respondents reporting frequent-or-occasional worry, not an elicited probability. There is no direct conversion to a subjective lifetime probability, but it is the best time-series instrument for tracking US homicide-worry at the national level.\n","independence_note":"Gallup conducts an independent annual telephone/web survey; methodologically independent of both CDC vital-statistics and BJS victimization data."},{"url":"https://pubmed.ncbi.nlm.nih.gov/26551975/","title":"Violent Death Rates: The US Compared with Other High-income OECD Countries, 2010","publisher":"American Journal of Medicine (Grinshteyn & Hemenway)","source_type":"peer_reviewed","statistic":"US homicide rate was 7.0 times higher than in other high-income OECD countries in 2010, driven by a gun homicide rate 25.2 times higher","excerpt":"\"US homicide rates were 7.0 times higher than in other high-income countries, driven by a gun homicide rate that was 25.2 times higher.\"\n","source_date":"2016-03-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260710103356/https://pubmed.ncbi.nlm.nih.gov/26551975/","calculation_notes":"Used only for the international-comparison sentence in the body text (\"roughly 7x above other high-income countries\"); not used in the native/normalized lifetime-probability calculation, which relies solely on the US-specific FBI/CDC/BJS/CCJ figures above.\n","independence_note":"Grinshteyn & Hemenway compile cross-national mortality data (WHO Mortality Database) independently of the US-specific FBI/CDC/BJS figures used for the primary estimate; cited here for international context only.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":3.5,"notes":"BJS Homicide Victimization in the United States, 2023: male rate 9.3 per 100,000 vs. female rate 2.6 per 100,000 — a 3.5× differential."},{"factor":"Age 15–34 vs. all-age average","multiplier":5,"notes":"CDC WISQARS 2022: homicide victimization peaks sharply in the 15–34 age band; the age-specific rate for this group is approximately 5× the all-age rate of 5.9 per 100,000."},{"factor":"Urban core vs. suburban/rural residence","multiplier":9,"notes":"FBI UCR 2022: city-level data show a roughly 8–10× spread between the highest-rate urban census tracts and low-crime suburban/rural areas; 9× used as a central estimate."},{"factor":"Prior violent victimization","multiplier":6,"notes":"BJS longitudinal victimization data: individuals with a prior violent victimization history face approximately 6× elevated risk of subsequent homicide victimization compared with the general population."},{"factor":"Black race vs. white","multiplier":6.7,"notes":"BJS Homicide Victimization in the United States, 2023: Black homicide victimization rate 21.3 per 100,000 vs. white rate 3.2 per 100,000 — a 6.7× differential (more than 6× per BJS)."}],"short_label":"Homicide","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This is the single field where a pooled average is most misleading. The US homicide rate varies by more than an order of magnitude across demographics: BJS reports the male rate at 9.3 per 100,000 versus 2.6 for females, and the rate for Black Americans at 21.3 versus 3.2 for white Americans. Age is just as skewed — homicide victimization peaks sharply in the 15–34 band and falls off steeply at both ends. Geography compounds all of this: county-level rates span roughly two orders of magnitude, with a small number of census tracts accounting for a disproportionate share of incidents. The \"1 in 339\" pooled figure is the right answer to \"what is the US average?\" and the wrong answer to \"what is *my* risk?\" for almost every individual reader. The headline is also moving fast: the FBI's official 2024 rate of 5.0 per 100,000 was a 14.9% one-year drop — the largest on record and the lowest rate since 2015 — after a roughly 10% decline in 2023, and the Council on Criminal Justice projects 2025 may reach about 4.0 per 100,000, which would be the lowest since 1900. This entry uses the official 2024 figure as its point estimate and treats the 2025 projection as the lower edge of the uncertainty band until full-year FBI data confirm it.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single chalk outline of a key on pale grey asphalt, flat vector illustration, no figures."},"canonical_url":"https://likelier.app/homicide-us","api_url":"https://likelier.app/api/fears/homicide-us.json"},{"slug":"ev-battery-fire","question":"What are the odds of an electric car catching fire?","quick_answer":"About 1 in 333 EVs catches fire over a roughly 12-year ownership, working out to around 25 fires per 100,000 EVs a year versus about 1,530 per 100,000 for gasoline cars. Despite a widespread belief that EVs are the bigger fire hazard, they burn at a fraction of the rate of internal-combustion vehicles.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Every EV fire generates outsized media coverage. A single Tesla fire on a highway attracts the kind of attention that hundreds of daily gasoline-car fires do not, because EV fires are novel, visually dramatic (thermal runaway produces sustained flames and toxic fumes), and difficult for fire services to extinguish with conventional methods. The result is a perception gap: surveys consistently show that consumers rank battery fire as a top concern about EV ownership, even though the data shows EVs catch fire at a fraction of the rate of internal combustion engine vehicles.\n","rough_estimate":"Many consumers believe EVs are more likely to catch fire than gasoline cars","kind":"intuition"},"native":{"display":"~25 fires per 100,000 EVs sold vs ~1,530 per 100,000 ICE vehicles (AutoInsuranceEZ/BLS data)","numerator":25,"denominator":100000,"unit":"per year (approximate)","population":"US registered EVs"},"normalized":{"lifetime_us_adult":0.003,"display":"~1 in 333 over a vehicle lifetime (~12 years average ownership)","log_value":-2.52,"assumptions":"Uses the widely cited AutoInsuranceEZ/BLS-derived figure of ~25 fires per 100,000 EVs sold, compared to ~1,530 per 100,000 for ICE vehicles. The EV figure is based on NTSB and NHTSA incident data through 2022. Swedish Civil Contingencies Agency (MSB) data confirms the direction: 3.8 fires per 100,000 EVs and hybrids vs 68 per 100,000 ICE vehicles in Sweden, making ICE vehicles about 18x more likely to catch fire. The normalised figure uses the 25/100,000 annual rate (0.00025/year) compounded over a ~12-year average vehicle ownership period: 1 − (1 − 0.00025)^12 ≈ 0.003, yielding ~1 in 333 chance that a given EV will experience a fire at some point during ownership. For ICE vehicles, the equivalent figure is ~1 in 5.5 over the same period. The comparison is the point: EVs are roughly 60x less likely to catch fire than ICE vehicles per the US data, and ~18x less likely per the Swedish data. The uncertainty band reflects the difference between US and Swedish estimates and the fact that the EV fleet is younger than the ICE fleet (older vehicles are more fire-prone).\n","uncertainty":{"low":0.001,"high":0.006},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.kbb.com/car-news/study-electric-vehicles-involved-in-fewest-car-fires/","title":"Study: Electric Vehicles Involved in Fewest Car Fires","publisher":"Kelley Blue Book / AutoInsuranceEZ","source_type":"reputable_reference","statistic":"~25 fires per 100,000 EVs sold vs ~1,530 fires per 100,000 ICE vehicles vs ~3,475 per 100,000 hybrids","excerpt":"\"Hybrid-powered cars were involved in about 3,475 fires per every 100,000 sold. Gasoline-powered cars, about 1,530. Electric vehicles (EVs) saw just 25 fires per 100,000 sold.\"\n","source_date":"2022-01-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20240803174343/https://www.kbb.com/car-news/study-electric-vehicles-involved-in-fewest-car-fires/","calculation_notes":"AutoInsuranceEZ compiled National Transportation Safety Board car-fire data against Bureau of Transportation Statistics sales data to calculate fire rates per 100,000 vehicles sold by powertrain type. The EV rate of ~25/100k vs ICE rate of ~1,530/100k represents a roughly 60x lower fire rate for EVs. The hybrid rate (~3,475/100k) is highest, likely reflecting the dual powertrain and the older average age of the hybrid fleet. This KBB article (Sean Tucker, 2022-01-28, \"Study: Electric Vehicles Involved in Fewest Car Fires\") reports the AutoInsuranceEZ study and states all three per-100k figures verbatim; it is used as the citation because it is the most widely read consumer source reporting this data. The underlying BLS/NHTSA data is the authoritative source. The figure has been widely cited but carries caveats: the EV fleet is younger on average than the ICE fleet (vehicle fire risk increases with age), and the total number of EV fires is small enough that the per-100k rate has a wide confidence interval.\n"},{"url":"https://alliedworldinsurance.com/risk-management/electric-vehicle-fires-a-cause-for-concern/","title":"Electric Vehicle Fires: A cause for concern?","publisher":"Allied World Insurance","source_type":"reputable_reference","statistic":"Swedish Civil Contingencies Agency (MSB) May 2023 report: ICE vehicles ~20x more likely to catch fire than EVs in Sweden","excerpt":"\"A report in May 2023 by the Swedish Contingencies Agency, found vehicles powered by ICE were twenty times more likely to catch fire than EVs in Sweden.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035419/https://alliedworldinsurance.com/risk-management/electric-vehicle-fires-a-cause-for-concern/","calculation_notes":"The Swedish Civil Contingencies Agency (MSB) data is the most frequently cited European dataset on EV fires. Sweden has high EV adoption (~17% of new car sales by 2023) and comprehensive fire reporting. Allied World, citing the MSB's May 2023 report, states ICE vehicles were ~20x more likely to catch fire than EVs in Sweden — directionally consistent with the US data but with a smaller ratio, likely because the Swedish ICE fleet is newer on average and Sweden's fire statistics may capture different incident types. The underlying MSB per-100k breakdown (3.8 fires per 100,000 EVs/hybrids vs 68 per 100,000 ICE; 23 fires among ~611,000 EVs) is quoted verbatim from the Fire Rover source below; the absolute EV fire count illustrates how small the numerator is, which is why the per-100k rate has wide confidence intervals.\n","independence_note":"Swedish MSB data is collected independently from US NHTSA/NTSB data. The two datasets use different methodologies and cover different vehicle fleets, making them a genuine cross-validation.\n"},{"url":"https://firerover.com/how-often-do-ev-batteries-catch-fire/","title":"How Often Do Electric Cars Catch Fire? A Look at the Statistics","publisher":"Fire Rover","source_type":"reputable_reference","statistic":"Studies spanning 2010–2020 put EV battery fire rates at 0.0012% vs 0.1% for internal-combustion-engine cars — an ~83x lower rate for EVs","excerpt":"\"Studies spanning 2010–2020 saw EV battery fire rates at 0.0012%, compared to 0.1% for internal combustion engine (ICE) cars.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035500/https://firerover.com/resources/how-often-do-ev-batteries-catch-fire/","calculation_notes":"Fire Rover aggregates several EV-vs-ICE fire datasets. The figure cited here is its longitudinal 2010–2020 comparison: an EV battery fire rate of 0.0012% versus 0.1% for ICE cars, i.e. ICE cars roughly 83x more likely to catch fire over that window. This is a distinct third data point that triangulates with the AutoInsuranceEZ US per-100k rate (~60x) and the Swedish MSB per-100k rate (~18-20x): three independently sourced comparisons that all place the EV fire rate well below the ICE rate, with the exact ratio varying by methodology and fleet. (The same article also restates the 25 vs 1,530 per-100,000 US figures and the Swedish 3.8 vs 68 per-100,000 figures.)\n","independence_note":"Fire Rover's aggregated 2010–2020 longitudinal comparison draws on a different sampling window and methodology from the single-year AutoInsuranceEZ US per-100k figure and the Swedish MSB per-100k figure, so it corroborates the direction from a third angle.\n"},{"url":"https://www.ntsb.gov/safety/safety-studies/Documents/SR2001.pdf","title":"Safety Risks to Emergency Responders from Lithium-Ion Battery Fires in Electric Vehicles (NTSB/SR-20/01)","publisher":"National Transportation Safety Board (NTSB)","source_type":"govt_report","statistic":"In all 3 of the crash-damaged EV batteries the NTSB studied, the battery reignited after firefighters extinguished the vehicle fire; one reignited at least six times and again 5 days after the crash, driven by stranded energy.","excerpt":"\"All three crash-damaged batteries reignited after firefighters extinguished the vehicle fires.\" \"That stranded energy can cause a high-voltage battery to reignite multiple times after firefighters extinguish an electric vehicle fire. Emergency responders have no way of measuring how much energy remains in a damaged battery, and no way of draining that energy, other than such time-consuming methods as allowing a battery fire to burn itself out.\"\n","source_date":"2020-11-13","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260513153239/https://www.ntsb.gov/safety/safety-studies/Documents/SR2001.pdf","calculation_notes":"NTSB Safety Report NTSB/SR-20/01 (adopted November 13, 2020) investigated all US domestic high-voltage EV battery fires the NTSB became aware of in a one-year window (August 2017–August 2018): three high-speed crash fires plus one noncrash fire. All three crash-damaged batteries reignited after the vehicle fire was extinguished — one reignited at least six times, including 5 days after the crash. This is the authoritative documentation of the reignition / stranded-energy phenomenon. It is a qualitative, mechanism-level finding from a four-case responder-safety investigation, NOT a population fire rate; it does not change the entry's headline ignition odds and is used only to substantiate the reignition caveat (per-incident severity, conditional on a fire already occurring).\n"},{"url":"https://www.evfiresafe.com/ev-fire-reignition","title":"EV Fire Reignition","publisher":"EV FireSafe (Australian fire-service research project)","source_type":"reputable_reference","statistic":"13% of EVs in EV FireSafe's reignition dataset reignited after initial suppression; one had 4 separate reignition events and another reignited 68 days after the incident; 5 reignited while on a tow truck.","excerpt":"\"Our research found 13% of the vehicles we studied reignited following initial suppression\" \"one vehicle had 4 seperate reignition events & another reignited 68 days after the initial incident\"\n","source_date":"2024-01-01","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260620140130/https://www.evfiresafe.com/ev-fire-reignition","calculation_notes":"EV FireSafe is an Australian fire-service-funded research/advocacy project that compiles global EV fire incidents reported to it. Its 13% figure is the share of vehicles in its own reignition-incident dataset that reignited after suppression — a conditional rate (given a fire occurred) drawn from a self-selected fire-service sample with no published denominator, NOT a population probability and NOT comparable to the entry's per-100k ignition rate. It is cited only to put a number on the reignition phenomenon the NTSB documents qualitatively; treat the magnitude as indicative, not precise.\n","independence_note":"EV FireSafe's reignition dataset is collected independently of the NTSB's four-case US investigation, so the two sources cross-corroborate the reignition phenomenon from different fleets and methodologies — while sharing the limitation that neither is a population-representative reignition rate.\n"},{"url":"https://www.spglobal.com/automotive-insights/en/blogs/2025/05/average-age-of-vehicle-in-us","title":"Average Age of Vehicles in the US Rises to 12.8 Years in 2025","publisher":"S&P Global Mobility","source_type":"reputable_reference","statistic":"Average age of battery electric vehicles (BEVs) in the US was 3.7 years in 2025, versus 12.8 years for the overall US vehicle fleet","excerpt":"\"Still, the average age for BEVs remains low at 3.7 years, growing at a similar rate—about two months—as the overall market in 2024.\"\n","source_date":"2025-05-21","source_accessed":"2026-07-03","archive_url":"https://web.archive.org/web/20260609145458/https://www.spglobal.com/automotive-insights/en/blogs/2025/05/average-age-of-vehicle-in-us","calculation_notes":"Cited to substantiate the fleet-age caveat in the body prose: the US BEV fleet averages 3.7 years old versus 12.8 years for all vehicles, which supports (and is slightly younger than) the entry's qualitative claim that most EVs on US roads are young. Independent of the EV-fire-rate sources above; this is a vehicle-registration/age analytics source, not a fire-incidence source.\n"}],"comparison_anchors":[{"label":"Car crash death (US adult, lifetime)","lifetime_us_adult":0.0108},{"label":"Home fire death (US adult, lifetime)","lifetime_us_adult":0.00085},{"label":"Motorcycle death (US adult, lifetime)","lifetime_us_adult":0.009}],"personal_factor_multipliers":[{"factor":"Vehicle age >8 years","multiplier":2,"notes":"Vehicle fire risk increases with age for all powertrain types; early EV data may underestimate long-term risk as the fleet ages"},{"factor":"Charging with damaged/non-certified equipment","multiplier":3,"notes":"Aftermarket and damaged charging equipment is a documented ignition source in EV fire investigations"},{"factor":"Post-collision (any severity)","multiplier":5,"notes":"Battery damage from collisions can trigger delayed thermal runaway; NTSB has flagged this as a unique EV risk requiring post-crash monitoring"},{"factor":"Flood-compromised or water-damaged battery","multiplier":50,"notes":"NFPA and NTSB EV fire investigation reports: saltwater intrusion (from flood events) into lithium-ion battery packs is a well-documented trigger for delayed thermal runaway; salt bridges internal cell contacts and initiates electrochemical reactions that build heat over hours to days; post-flood EVs have ignited in storage facilities and on transport vehicles, creating a documented mass-casualty fire pattern distinct from normal operational risk"},{"factor":"Overnight indoor charging in enclosed structure","multiplier":2,"notes":"NFPA fire safety guidance for EV charging: while the absolute risk of fire during any charging session is very low, overnight indoor charging in garages or enclosed parking structures increases consequence severity if a fire does occur — limited ventilation accelerates thermal runaway propagation, delays detection, and prevents early fire service access. NFPA recommends outdoor or semi-open charging as a mitigation measure"}],"short_label":"EV battery fire","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The EV fire rate comparison to ICE vehicles is directionally robust (EVs catch fire far less often) but the magnitude of the gap is uncertain for several reasons. First, the EV fleet is younger on average than the ICE fleet, and vehicle fire risk increases with age — as the EV fleet ages, the rate may converge somewhat. Second, the absolute number of EV fires is small enough that the per-100k rate has wide confidence intervals. Third, EV fires have different characteristics: thermal runaway in lithium-ion batteries produces sustained, high-temperature fires that are harder to extinguish, can reignite days after the initial event, and produce toxic hydrogen fluoride gas. The per-incident severity may be higher for EV fires even if the per-vehicle frequency is lower. Finally, the comparison includes all ICE vehicle fires, many of which are in very old vehicles; comparing EVs to age-matched ICE vehicles would narrow the gap. The entry is tagged as overrated because the data clearly shows EVs are less fire-prone than ICE vehicles, contrary to the popular perception. A separate dimension worth naming is what happens after a fire starts, which is independent of how often one starts. The NTSB's responder-safety report (NTSB/SR-20/01, 2020) found that all three crash-damaged EV batteries it investigated reignited after firefighters had extinguished the vehicle fire — one reignited at least six times, including 5 days after the crash — because of \"stranded energy\": charge trapped in undamaged cells with no path to discharge, which responders cannot measure or drain except by letting the pack burn itself out. The Australian fire-service project EV FireSafe reports that 13% of the EVs in its reignition-incident dataset reignited after initial suppression, with cases of four separate reignition events, a 68-day delayed reignition, and reignitions occurring while the vehicle was on a tow truck. These reignition figures are conditional on a fire having already occurred (a per-incident severity characteristic), are drawn from small, self-selected fire-service datasets with no population denominator, and do not change the entry's headline ignition odds; they are the reason an EV fire is operationally harder to make safe than an ICE fire of equivalent size, not evidence that EVs ignite more often. The enclosed-structure severity concern (garages, tunnels) is the same severity-not-frequency point and is already reflected in the indoor-charging multiplier above.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simple electric car silhouette with a small battery icon, flat vector illustration."},"canonical_url":"https://likelier.app/ev-battery-fire","api_url":"https://likelier.app/api/fears/ev-battery-fire.json"},{"slug":"extreme-heat-death","question":"What are the odds of dying from extreme heat?","quick_answer":"Extreme heat causes about 489,000 excess deaths a year worldwide, which works out to roughly 1 in 330 over a global adult lifetime. Half of EU citizens name extreme weather as the disaster most likely to hit their area, yet heat death rarely registers as a named personal risk.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Heat lacks the narrative drama of hurricanes, tornadoes, and earthquakes. There is no single \"fear of heat death\" tracked in national anxiety surveys, and most people do not carry an explicit prior for it at all. The perceived risk is shaped almost entirely by whether the reader has personally experienced a dangerous heatwave — the 2003 European event, the 2021 Pacific Northwest heat dome, or the 2023-2024 South and Southeast Asian seasons — and reverts to near zero otherwise. The result is a hazard that is simultaneously the deadliest weather-related killer in most wealthy countries and one of the least feared.\n","rough_estimate":"50% of EU citizens identify extreme weather events (including heat waves) as the disaster most likely to affect their area; Yale finds 64% of Americans worried about climate change broadly, but heat death rarely registers as a named personal risk","kind":"survey","survey_source":{"title":"Special Eurobarometer — EU Citizens and Civil Protection","publisher":"European Commission, DG ECHO","url":"https://civil-protection-humanitarian-aid.ec.europa.eu/resources-campaigns/eurobarometer-reports_en","year":2024}},"native":{"display":"~489,000 heat-related excess deaths per year (WHO, 2000-2019 average)","numerator":489000,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.003,"display":"1 in ~330 lifetime (global adult)","log_value":-2.52,"assumptions":"Uses the WHO estimate of ~489,000 heat-related deaths per year globally (2000-2019 average), which aligns with Zhao et al. 2021 in Lancet Planetary Health finding that 0.91% of all global deaths are heat-related (0.91% of ~58 million global deaths/year ≈ 528,000). Annual per-capita risk ≈ 489,000 / 8,000,000,000 ≈ 6.1e-5; compounded over 60 adult years ≈ 3.7e-3, rounded to 0.003 ≈ 1 in 330. This is the \"excess mortality\" figure from epidemiological attribution models, not the much smaller \"direct heat death\" count from death certificates. The direct-coding count (ICD codes X30, T67) captures only a fraction of heat-attributable mortality because most heat deaths are coded under their proximate cause — myocardial infarction, stroke, renal failure, respiratory arrest — rather than under heat exposure. The excess-mortality approach, which compares observed deaths during heat events to a counterfactual baseline, is the standard in the epidemiological literature and is used by both the WHO and the Lancet Countdown. The uncertainty band reflects both methodological range (direct-coded vs excess-mortality) and the rising trend: heat-related excess death ratios increased by 0.21 percentage points between 2000-2003 and 2016-2019 (Zhao et al.), so a lifetime estimate based on the 2000-2019 average likely understates the risk for younger adults who will live through warmer decades.\n","uncertainty":{"low":0.0015,"high":0.006},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health","title":"Climate change, heat and health","publisher":"World Health Organization (WHO)","source_type":"govt_report","statistic":"Approximately 489,000 heat-related deaths occur each year globally (2000-2019); heat stress is the leading cause of weather-related deaths.","excerpt":"\"Between 2000–2019 studies show approximately 489 000 heat-related deaths occur each year, with 45% of these in Asia and 36% in Europe.\"\n","source_date":"2024-10-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413171643/https://www.who.int/news-room/fact-sheets/detail/climate-change-heat-and-health","calculation_notes":"WHO's 489,000 deaths/year is the headline native figure. 489,000 / 8,000,000,000 ≈ 6.11e-5 per year; compounded over 60 adult years via 1 - (1 - 6.11e-5)^60 ≈ 3.66e-3, rounded to 0.003 ≈ 1 in ~330 global adult lifetime. This figure comes from epidemiological excess-mortality models that attribute deaths to heat exposure above location-specific thresholds, not from death-certificate coding alone. The WHO number is consistent with the Zhao et al. 2021 finding that 0.91% of global mortality is heat-related (0.91% × ~58 million ≈ 528,000), which brackets the WHO estimate from above.\n","independence_note":"The WHO fact sheet draws on the same multi-country epidemiological literature that underlies Zhao et al. 2021 (both use distributed-lag nonlinear models applied to multi-city mortality datasets), so the two sources are methodologically related but not identical — the WHO figure is a synthesis across multiple studies, while Zhao et al. is one specific modeling exercise.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/34245712/","title":"Global, regional, and national burden of mortality associated with non-optimal ambient temperatures from 2000 to 2019: a three-stage modelling study","publisher":"Zhao Q, Guo Y, Ye T, Gasparrini A, Tong S, et al. — Lancet Planetary Health, 2021; 5(7):e415-e425","source_type":"peer_reviewed","statistic":"0.91% of all global deaths (2000-2019) were heat-related, accounting for approximately 528,000 excess deaths per year; heat-related excess death ratios increased by 0.21 percentage points between 2000-2003 and 2016-2019.","excerpt":"\"5 083 173 deaths (95% empirical CI [eCI] 4 087 967-5 965 520) were associated with non-optimal temperatures per year, accounting for 9·43% (95% eCI 7·58-11·07) of all deaths (8·52% [6·19-10·47] were cold-related and 0·91% [0·56-1·36] were heat-related).\"\n","source_date":"2021-07-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413171723/https://pubmed.ncbi.nlm.nih.gov/34245712/","calculation_notes":"Zhao et al. report 0.91% of global mortality as heat-related. Applied to ~58 million global deaths per year, this gives ~528,000 heat-related deaths annually — slightly above but consistent with the WHO's 489,000 figure, as expected given methodological differences in threshold definitions and spatial coverage. The 95% empirical confidence interval of 0.56-1.36% maps to roughly 325,000-789,000 deaths/year, which informs the uncertainty band. The finding that heat-related mortality increased by 0.21 percentage points over the study period while cold-related mortality declined supports the framing that the 2000-2019 average understates the forward-looking risk.\n","independence_note":"Zhao et al. is a primary peer-reviewed study using the Multi-Country Multi-City (MCC) Collaborative Research Network dataset. The WHO fact sheet synthesizes this and other studies. Treat as partially overlapping evidence chains with independent analytic choices.\n"},{"url":"https://jamanetwork.com/journals/jama/fullarticle/2822854","title":"Trends of Heat-Related Deaths in the US, 1999-2023","publisher":"Howard JT, Androne N, Alcover KC, Santos-Lozada AR — JAMA, 2024; 332(14):1203-1204","source_type":"peer_reviewed","statistic":"US heat-related deaths (underlying or contributing cause) rose from 1,069 in 1999 to 2,325 in 2023 — a 117% increase in count and a 63% increase in the age-adjusted mortality rate; 21,518 cumulative deaths over 1999-2023, with a 16.8% per-year acceleration from 2016 to 2023.","excerpt":"\"The number of heat-related deaths increased from 1069 (95% CI, 1003-1135) in 1999 to 2325 (95% CI, 2278-2372) in 2023, a 117% increase in the number of heat-related deaths and a 63% increase in the AAMR. ... There were 21 518 heat-related deaths from 1999 to 2023.\"\n","source_date":"2024-08-26","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260618132712/https://jamanetwork.com/journals/jama/fullarticle/2822854","calculation_notes":"This is a US, death-certificate (direct-coded, underlying-or-contributing) figure, not a global excess-mortality figure, so it does NOT replace the headline 1-in-330 global lifetime number (which is WHO excess-mortality at global_adult_lifetime scope). It is cited as evidence for two things already embedded in this entry: (1) the rising US trend, now with a hard 117% / 63% magnitude through 2023, and (2) the persistence of the direct-coding undercount relative to excess-mortality estimates. The 2,325 deaths in 2023 also confirm that the older \"~700-1,300/year\" US direct-coded range in the regional breakdown and caveats is dated; the figure has roughly doubled by 2023.\n","independence_note":"Howard et al. is a primary analysis of CDC WONDER / NCHS death-certificate data — a US direct-coding source, methodologically distinct from the WHO and Zhao et al. global excess-mortality models. It corroborates the trend, not the headline magnitude.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10249403/","title":"How Blackouts during Heat Waves Amplify Mortality and Morbidity Risk","publisher":"Stone B Jr, Gronlund CJ, Mallen E, Hondula D, O'Neill MS, et al. — Environmental Science & Technology, 2023; 57(22):8245-8255","source_type":"peer_reviewed","statistic":"A concurrent multiday blackout during heat-wave conditions more than doubles the estimated rate of heat-related mortality across Atlanta, Detroit, and Phoenix; Phoenix showed by far the largest amplification.","excerpt":"\"We find the concurrence of a multiday blackout event with heat wave conditions to more than double the estimated rate of heat-related mortality across all three cities.\"\n","source_date":"2023-05-24","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260605184154/https://pmc.ncbi.nlm.nih.gov/articles/PMC10249403/","calculation_notes":"Supports the \"multiday power outage during a heat wave\" personal factor multiplier of 2 (≥2× from \"more than double\"). Modeling study of compound heat-plus-grid-failure events in three US cities; the factor is the cross-city floor, not the Phoenix-specific worst case. Also underpins the post-storm-heat caveat (Hurricane Beryl, 2024). Does not affect the headline global lifetime figure.\n","independence_note":"Stone et al. is a primary modeling study (3HEAT / Three-City Heat and Electrical Grid Failure Adaptation Study) on compound infrastructure-climate hazards, independent of the WHO, Zhao, and JAMA mortality-trend sources.\n"}],"comparison_anchors":[{"label":"Death in a tornado (lifetime, US adult average)","lifetime_us_adult":0.0000124},{"label":"Death by tropical cyclone (lifetime, global adult average)","lifetime_us_adult":0.000112},{"label":"Death in an earthquake (lifetime, global adult average)","lifetime_us_adult":0.000263},{"label":"Death by wildfire, direct (lifetime, global adult)","lifetime_us_adult":0.00001},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average (excess-mortality attribution)","probability":0.003,"notes":"Headline figure — epidemiological excess mortality attributed to heat exposure, not death-certificate direct coding."},{"region":"South Asia (India, Pakistan, Bangladesh)","probability":0.008,"notes":"Highest absolute heat-mortality burden globally. The 2015 India and Pakistan heatwaves killed over 4,500 people by direct-coding alone; excess-mortality estimates are far higher. Outdoor labor exposure, limited air conditioning, and urban heat island effects concentrate risk."},{"region":"Europe (summer heatwave belt)","probability":0.005,"notes":"WHO attributes 36% of global heat-related deaths to Europe. The 2003 heatwave killed ~70,000 across Western Europe; the 2022 summer killed an estimated 61,672. Aging demographics amplify vulnerability."},{"region":"Sub-Saharan Africa","probability":0.004,"notes":"Growing heat exposure with limited adaptive capacity (low AC penetration, outdoor agricultural labor); likely undercounted due to sparse vital registration."},{"region":"United States (national average)","probability":0.002,"notes":"CDC direct-coded heat deaths have risen sharply — from ~1,069 in 1999 to 2,325 in 2023 (Howard et al., JAMA 2024) — but broader excess-mortality estimates are 5,000-12,000/year. Highest in the Southwest, Gulf states, and urban heat islands. Compound post-storm grid-failure events (e.g. Hurricane Beryl, Texas, July 2024) concentrate deaths when prolonged outages remove cooling during a heatwave."},{"region":"Northern Europe / high-latitude temperate","probability":0.0005,"notes":"Low baseline heat exposure, though increasing; the 2018 and 2022 summers produced historically unprecedented heat events in the UK and Scandinavia."}],"personal_factor_multipliers":[{"factor":"Age 65+ with cardiovascular or respiratory disease","multiplier":5,"notes":"Heat-related mortality is overwhelmingly concentrated in older adults with pre-existing cardiopulmonary disease. WHO reports heat-related mortality for people over 65 increased ~85% between 2000-2004 and 2017-2021."},{"factor":"Outdoor manual laborer (agriculture, construction)","multiplier":3,"notes":"Occupational heat exposure during peak hours produces exertional heat illness and amplifies cardiovascular strain; the multiplier applies to working-age adults in exposed occupations."},{"factor":"Air-conditioned residence and workplace","multiplier":0.3,"notes":"Access to cooling is the single strongest protective factor. US studies consistently show heat mortality concentrated in populations without reliable AC access."},{"factor":"Urban heat island residence (megacity, limited green space)","multiplier":2,"notes":"Urban areas can be 5-10°C warmer than surrounding rural areas during heat events; compounded by overnight heat retention that prevents physiological recovery."},{"factor":"Socially isolated elderly (living alone, limited mobility)","multiplier":4,"notes":"The 2003 European heatwave mortality was heavily concentrated among isolated elderly who could not seek cooling or hydration assistance. Social isolation is a strong independent risk factor."},{"factor":"Multiday power outage during a heat wave (e.g. post-storm grid failure)","multiplier":2,"notes":"Stone et al. 2023 (Environmental Science & Technology, 3HEAT study of Atlanta, Detroit, and Phoenix) found that the concurrence of a multiday blackout with heat-wave conditions more than doubles the estimated heat-related mortality rate across all three cities — Phoenix, with low passive survivability, showed far larger amplification. This is the mechanism behind post-storm heat-death clusters such as Hurricane Beryl's Texas power outages in July 2024."}],"short_label":"Extreme heat","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline \"1 in 330\" figure uses the epidemiological excess-mortality approach, which attributes deaths to heat exposure by comparing observed mortality during hot periods to a counterfactual baseline. This is the methodological standard in the literature (used by both the WHO and the Lancet Countdown) but it produces a much larger number than death-certificate direct coding. In the US, for instance, direct-coded heat deaths (ICD X30, T67) run about 700-1,300 per year, while excess-mortality estimates range from 5,000 to 12,000 — the gap exists because most heat deaths present as cardiovascular events, renal failure, or respiratory crises that get coded under their proximate cause. The excess-mortality figure is the more complete measure, but it depends on modeling assumptions (threshold choice, lag structure, baseline definition) that produce real disagreement across studies. The number is also rising: Zhao et al. found that heat-related excess death ratios increased by 0.21 percentage points between 2000-2003 and 2016-2019, and the WHO reports that heat-related mortality among people over 65 increased by approximately 85% between 2000-2004 and 2017-2021. The most recent US figures sharpen this: Howard et al. (JAMA, 2024) found death-certificate heat-related deaths rose 117% in count (and 63% in the age-adjusted rate) from 1,069 in 1999 to 2,325 in 2023 — with 21,518 cumulative deaths over the period and a 16.8%-per-year acceleration after 2016 — while reiterating that cardiovascular and renal deaths uncoded to heat mean even these record counts are undercounts. A lifetime estimate based on the 2000-2019 average therefore likely understates the risk for younger adults who will experience warmer decades ahead. A distinct and growing hazard is the compound storm-plus-heat event: when a tropical cyclone knocks out power for days during peak summer, the loss of air conditioning turns a heatwave deadly. After Hurricane Beryl made landfall in Texas in July 2024, leaving millions without power for over a week, the storm-related death toll reached at least the low-to-mid 20s in the following weeks, with roughly half of the Harris County deaths attributed to heat during the outage rather than to wind or flooding. Isolating the post-storm heat increment from baseline expectation requires modeling, and these indirect deaths fall outside both the wind/water/surge counts of hurricane statistics and the steady-state heat-mortality figures above. Finally, the global average masks enormous geographic and demographic heterogeneity: the risk for a 75-year-old with heart failure in New Delhi during a May heatwave and the risk for a 30-year-old in an air-conditioned Oslo office are separated by orders of magnitude.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized thermometer shape against a muted warm sky, rendered as a flat geometric form, vector illustration."},"canonical_url":"https://likelier.app/extreme-heat-death","api_url":"https://likelier.app/api/fears/extreme-heat-death.json"},{"slug":"food-poisoning-global","question":"What are the odds of dying from food poisoning (worldwide)?","quick_answer":"Worldwide, dying from contaminated food is about a 1 in 320 chance over an adult lifetime, roughly six times the US rate, from about 420,000 deaths a year. This is a global average that hides more than an order-of-magnitude gap between high-income countries and Sub-Saharan Africa.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Most readers who meet this question in a US or European context quietly assume the answer is \"basically zero\" — the same intuition that drives the US-specific food-poisoning entry. The global framing sits in an awkward place: Western readers tend to overestimate their own country's outbreak-driven risk (salad recalls, fast-food E. coli clusters) while systematically underestimating the burden of diarrheal foodborne illness in low- and middle-income countries. We have not found a cross-national survey that isolates \"fear of dying from contaminated food\" as a clean question, so the perceived side here is editorial intuition, not polled data.\n","rough_estimate":"50% of US adults rank foodborne illness among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 50% of US adults rank foodborne illness as a top-3 food safety concern (no equivalent global survey; US figure anchors the perceived side)","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~420,000 global foodborne illness deaths per year","numerator":420000,"denominator":8000000000,"unit":"per year","population":"global, all ages, all foodborne hazards (FERG 2010 reference year)"},"normalized":{"lifetime_us_adult":0.00315,"display":"1 in ~320 lifetime (global adult)","log_value":-2.5,"assumptions":"Uses the WHO Foodborne Disease Burden Epidemiology Reference Group (FERG) 2015 headline figure of 420,000 foodborne deaths per year globally (reference year 2010, 95% uncertainty interval 310,000–600,000). Divided by a global population of ~8 billion gives an annual rate of ~5.25 per 100,000, roughly six times the US per-capita foodborne mortality rate. Compounded over 60 adult life-years: 1 − (1 − 5.25e-5)^60 ≈ 3.15e-3, or about 1 in 320 global lifetime. The uncertainty band below is dominated by the FERG 95% UI on the death estimate (310K–600K) rather than sampling noise in the population denominator. This is a global-average scale marker and is useless as a personal estimate — see the regional_breakdown and the body text for why the Sub-Saharan Africa and high-income-country figures differ by more than an order of magnitude.\n","uncertainty":{"low":0.0023,"high":0.0045},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/food-safety","title":"Food Safety — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"~600 million foodborne illnesses and 420,000 deaths globally per year; children under 5 account for 40% of the burden and 125,000 of those deaths","excerpt":"\"An estimated 600 million – almost 1 in 10 people in the world – fall ill after eating contaminated food and 420 000 die every year [...] Children under 5 years of age carry 40% of the foodborne disease burden, with 125 000 deaths every year.\"\n","source_date":"2024-05-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172102/https://www.who.int/news-room/fact-sheets/detail/food-safety","calculation_notes":"420,000 global deaths / 8,000,000,000 people ≈ 5.25e-5 per year. Compounded over 60 adult years: 1 − (1 − 5.25e-5)^60 ≈ 3.15e-3, or ~1 in 320 global adult lifetime. The 125,000 under-5 deaths out of 420,000 (~30%) is the single most load-bearing heterogeneity in the whole estimate — children 0-4 are less than 10% of the global population but carry nearly a third of the mortality.\n","independence_note":"WHO's public fact sheet is a restatement of the FERG 2015 report's headline numbers; treat as derivative of the Havelaar et al. PLOS Medicine paper, not as an independent estimate.\n"},{"url":"https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001923","title":"World Health Organization Global Estimates and Regional Comparisons of the Burden of Foodborne Disease in 2010","publisher":"PLOS Medicine / Havelaar et al. (WHO FERG)","source_type":"peer_reviewed","statistic":"420,000 (95% UI 310,000–600,000) foodborne deaths and 33 million DALYs in 2010; burden rates vary from 35 DALYs/100,000 in high-income North America to 1,300 DALYs/100,000 in sub-Saharan Africa","excerpt":"\"The global burden of foodborne disease caused by the 31 hazards in 2010 was 33 (95% uncertainty interval [UI] 25–46) million Disability Adjusted Life Years (DALYs); 40% affected children under five years old. The most frequent causes of foodborne illness were diarrheal disease agents [...] Norovirus and Campylobacter spp. caused the largest number of foodborne illnesses. [...] The global burden of foodborne disease is considerable, and affects individuals of all ages, particularly children under five years of age and persons living in low-income regions of the world.\"\n","source_date":"2015-12-03","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260305142348/https://journals.plos.org/plosmedicine/article?id=10.1371%2Fjournal.pmed.1001923","calculation_notes":"Havelaar et al. is the peer-reviewed backbone of the FERG 2015 numbers that the WHO fact sheet restates. The regional DALY gradient (1,300 per 100,000 in AFR D/E vs 35 per 100,000 in AMR A, a ~37x range) is the justification for the regional_breakdown figures: we anchor the global average at the FERG headline and scale the regional entries by the reported DALY ratio, then convert to lifetime probability on the same 60-year compounding.\n","independence_note":"The Havelaar 2015 paper and the WHO fact sheet are the same underlying FERG estimate reported through two channels; not independent. Treated here as the peer-reviewed primary and its public restatement.\n"},{"url":"https://www.who.int/publications/i/item/9789241565165","title":"WHO estimates of the global burden of foodborne diseases: foodborne disease burden epidemiology reference group 2007-2015","publisher":"World Health Organization / FERG","source_type":"govt_report","statistic":"First global and regional estimates of the burden of 31 foodborne hazards; developed by FERG over 2007-2015","excerpt":"\"WHO estimates of the global burden of foodborne diseases: foodborne diseases burden epidemiology reference group 2007-2015 [...] presents the first global and regional estimates of the burden of foodborne diseases.\"\n","source_date":"2015-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172215/https://www.who.int/publications/i/item/9789241565165","calculation_notes":"The full FERG 2015 report is the primary document behind both the Havelaar PLOS Medicine paper and the WHO public fact sheet. Cited here as the methodological anchor rather than for a distinct numeric estimate.\n","independence_note":"Upstream of both the Havelaar 2015 paper and the WHO fact sheet. All three citations are branches of the same FERG 2010-reference-year estimate and should be read as one coordinated body of evidence, not three independent confirmations.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Death by drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.00315,"notes":"FERG 2015 headline: 420,000 deaths/yr / 8B population, compounded over 60 years"},{"region":"Sub-Saharan Africa","probability":0.012,"notes":"AFR D/E subregions carry ~1,300 DALYs per 100,000 — roughly 4x the global rate. Accounts for a disproportionate share of global foodborne deaths in ~13% of the population."},{"region":"South-East Asia","probability":0.007,"notes":"SEAR B/D subregions carry ~690-710 DALYs per 100,000, roughly 2x the global rate"},{"region":"High-income countries","probability":0.0005,"notes":"AMR A (high-income North America) reports ~35 DALYs per 100,000 — a ~37x gap with sub-Saharan Africa driven by cold chains, sanitation, regulation, and clinical care"}],"personal_factor_multipliers":[{"factor":"resident of low-income country without sanitation infrastructure","multiplier":10,"notes":"WHO FERG: foodborne mortality in Sub-Saharan Africa and South Asia is roughly 10x rates in high-income countries"},{"factor":"immunocompromised or very young/old","multiplier":5,"notes":"foodborne deaths concentrate in children under 5 and adults 65+; WHO estimates ~40% of fatalities in under-5"},{"factor":"high-income country resident with cold-chain infrastructure","multiplier":0.1,"notes":"routine refrigeration, chlorinated water, and HACCP food safety standards dramatically reduce fatality rates"}],"short_label":"Food poisoning (global)","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 420,000/year figure is the WHO FERG central estimate for the 31 hazards they assessed (reference year 2010); the 95% uncertainty interval runs 310,000-600,000 and the uncertainty band on the normalized figure reflects that. The global average is a scale marker, not a personal estimate: regional per-capita risk varies by roughly 37x across WHO subregions, and within LMICs mortality is overwhelmingly concentrated in children under 5 rather than adults. A healthy adult in a high-income country has a per-year risk well below the global average; an infant in a rural LMIC has a risk many times above it. Excludes allergic reactions, deliberate poisoning, and gastrointestinal infections transmitted by water or person-to-person rather than food.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted bowl viewed from above on a pale sand-colored background, flat vector illustration."},"canonical_url":"https://likelier.app/food-poisoning-global","api_url":"https://likelier.app/api/fears/food-poisoning-global.json"},{"slug":"spinal-cord-injury-paralysis","question":"What are the odds of suffering a traumatic spinal cord injury that causes paralysis?","quick_answer":"A US adult's lifetime odds of a traumatic spinal cord injury causing paralysis are about 1 in 313 (well under 1%) — roughly what most people would guess. It covers traumatic injury only.\n","category":"health","tags":[],"no_reliable_estimate":false,"perceived":{"description":"Spinal cord injury rarely tops the list of health worries for most Americans, yet it sits in the background of many everyday risk calculations — helmets, seatbelts, and pool-safety rules are all partly about this outcome. No rigorous national survey isolates SCI worry specifically, but safety communication campaigns position it as a low-probability, catastrophic consequence of common activities like diving and high-speed driving.\n","rough_estimate":"Most people would guess well under 1% lifetime, consistent with actual odds","kind":"intuition"},"native":{"display":"~54 new traumatic SCI cases per million people per year","numerator":54,"denominator":1000000,"unit":"per year","population":"US general population (NSCISC model systems database)"},"normalized":{"lifetime_us_adult":0.0032,"display":"~1 in 313 lifetime (US adult)","log_value":-2.49,"assumptions":"The NSCISC estimates approximately 18,000 new traumatic SCI cases per year in the United States, in a population of ~330 million, giving an annual incidence of about 54 per million (~0.0000545 per person per year). Compounding over 59 adult years: 1 − (1 − 0.0000545)^59 ≈ 0.0032. This figure covers traumatic SCI only; non-traumatic SCI (tumor, infection, vascular) would add further incidence, making this a slight undercount of total paralysis risk. It also excludes cases where the person died at the scene before any SCI diagnosis, per NSCISC methodology.\n","uncertainty":{"low":0.002,"high":0.005},"scope":"us_adult_lifetime"},"sources":[{"url":"https://msktc.org/sites/default/files/Facts-and-Figures-2023-Eng-508.pdf","title":"Traumatic Spinal Cord Injury Facts and Figures at a Glance 2023","publisher":"National Spinal Cord Injury Statistical Center (NSCISC) / MSKTC","source_type":"govt_report","statistic":"~18,000 new traumatic SCI cases per year; ~302,000 people living with tSCI in the US; 54 cases per million annually","excerpt":"\"The estimated annual incidence of traumatic spinal cord injury (tSCI), not including those who die at the scene of the injury, is approximately 54 cases per million people in the United States, or about 18,000 new tSCI cases each year. The estimated number of people with tSCI living in the United States is approximately 302,000 persons.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20240623192606/https://msktc.org/sites/default/files/Facts-and-Figures-2023-Eng-508.pdf","calculation_notes":"Annual incidence: 18,000 / 330,000,000 = 0.0000545 per person per year. Compounded over 59 adult years: 1 − (1 − 0.0000545)^59 = 0.0032, or approximately 1 in 313. This is the probability of suffering at least one new traumatic SCI during an adult lifetime at the current annual incidence rate. Cases who die at the injury scene are excluded from the NSCISC count, so true motor-vehicle-related serious spine trauma is slightly higher.\n","independence_note":"The NSCISC database aggregates data from SCI Model Systems rehabilitation centers across the United States. It captures hospitalized SCI cases. The Christopher & Dana Reeve Foundation cites the same NSCISC figures for prevalence (~302,000). These are independent confirmations of the same underlying federal database.\n"},{"url":"https://www.christopherreeve.org/todays-care/paralysis-help-overview/stats-about-paralysis/","title":"Stats About Paralysis — Spinal Cord Injury Prevalence In The U.S.","publisher":"Christopher & Dana Reeve Foundation","source_type":"reputable_reference","statistic":"~302,000 people living with spinal cord injuries in the US (range 255,000–383,000); ~54 cases per million, or ~18,000 new SCI cases per year","excerpt":"\"The estimated number of people with SCI living in the United States is approximately 302,000 persons, with a range from 255,000 to 383,000 persons. The annual incidence of spinal cord injury (SCI) is approximately 54 cases per one million people in the United States, or about 18,000 new SCI cases each year. Vehicle crashes are the most recent leading cause of injury, closely followed by falls. Acts of violence (primarily gunshot wounds) and sports/recreation activities are also common causes.\"\n","source_date":"2024-04-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260521002714/https://www.christopherreeve.org/todays-care/paralysis-help-overview/stats-about-paralysis/","calculation_notes":"Used for corroboration of the NSCISC annual incidence figure (54 per million, ~18,000/year) and prevalence (302,000; range 255,000–383,000). This page names the leading causes qualitatively (vehicle crashes, then falls, then violence and sports/recreation) but does not publish the numeric cause percentages; the numeric split used in the prose and personal_factor_multipliers notes comes from the NSCISC 2023 Facts and Figures data sheet (source 1). Prevalence figure (302,000) implies a population of approximately 59 years × 18,000/year for a stable population, which is internally consistent.\n","independence_note":"The Reeve Foundation synthesizes NSCISC data and publishes it independently on its own website, providing a second accessible citation for the same underlying figures. The two sources are methodologically linked (same NSCISC database) but are independently hosted and curated, making them non-redundant for citation purposes.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Lifetime risk of drowning (US)","lifetime_us_adult":0.00091}],"personal_factor_multipliers":[{"factor":"Male sex (78% of SCI cases are male)","multiplier":3.5,"notes":"NSCISC reports ~78% of new tSCI cases occur in males, implying roughly 3.5x the rate of females at average age of injury 43"},{"factor":"Age 16–30 (peak motor-vehicle SCI incidence)","multiplier":2.5,"notes":"Young adults face elevated tSCI rates driven by higher motor vehicle crash rates and risk-taking behavior; vehicle crashes cause 38% of all tSCI"},{"factor":"Age 65+ (fall-related SCI rising)","multiplier":1.5,"notes":"The proportion of tSCI caused by falls has risen steadily since 1973 and is highest in older adults; the NSCISC notes average age of injury has risen to 43"},{"factor":"Participation in extreme sports (diving into shallow water, rodeo, equestrian)","multiplier":3,"notes":"Recreational sports/activities account for ~8% of new tSCI; diving injuries are the leading sports-related cause; equestrian and rodeo carry elevated cervical SCI risk"}],"short_label":"Spinal cord injury","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"This figure covers traumatic SCI only (vehicle crash, fall, violence, sports). Non-traumatic causes — including spinal tumor, vascular injury, infection, and degenerative disease — are not counted in the NSCISC database and would increase the total lifetime paralysis risk if included. The NSCISC database excludes individuals who die at the scene before reaching a model-system hospital, so the figure is a slight undercount of all serious spinal events. Age at injury has shifted markedly since the 1970s (average now 43, up from 29), with falls replacing vehicle crashes as the dominant mechanism in older adults. The 1 in 313 figure assumes a constant annual incidence at the current rate, but SCI rates have varied with seatbelt laws, vehicle safety technology, and violence rates over time.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A lone wheelchair silhouette on a plain pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/spinal-cord-injury-paralysis","api_url":"https://likelier.app/api/fears/spinal-cord-injury-paralysis.json"},{"slug":"childhood-cancer-diagnosis","question":"What are the odds of a child being diagnosed with cancer before age 20?","quick_answer":"Roughly 1 in 285 children are diagnosed with cancer before age 20. That is far below the 1-in-20-to-1-in-5 range most parents guess — the outcome's severity makes even a small probability feel much larger.\n","category":"health","tags":["kids","child"],"no_reliable_estimate":false,"perceived":{"description":"Parents consistently rank childhood cancer among their greatest fears, and the perceived probability sits well above the actual rate. The vividness and emotional weight of childhood cancer diagnoses in media coverage — charity drives, viral fundraising stories, celebrity disclosures — create an availability cascade that inflates intuitive estimates dramatically. When researchers ask parents to estimate the likelihood that their child will develop cancer before adulthood, the guesses routinely fall between 5% and 20%, far above the actual cumulative incidence of roughly 1 in 285. The gap is amplified by the severity of the outcome: even a very small probability feels large when the outcome is cancer in a child.\n","rough_estimate":"Most parents would guess somewhere between 1 in 20 and 1 in 5 — roughly 14 to 57 times the actual rate","kind":"intuition"},"native":{"display":"roughly 216 children diagnosed with cancer per 1,000,000 children per year (US, ages 0–19)","numerator":216,"denominator":1000000,"unit":"per child per year (ages 0–19)","population":"US children aged 0–19"},"normalized":{"lifetime_us_adult":0.00351,"display":"roughly 1 in 285 children diagnosed with cancer before age 20 (US)","log_value":-2.45,"assumptions":"NCI SEER data (2023 Cancer Statistics Review, Childhood Cancer Statistics table) reports approximately 15,780 new cancer cases per year among US children aged 0–19 (average 2016–2020). The US Census Bureau estimated approximately 73 million children under age 20 in that period. Annual incidence rate: 15,780 / 73,000,000 ≈ 216 per million (21.6 per 100,000) per year. Cumulative childhood probability (20 independent annual-trial approximation): P = 1 − (1 − 0.000216)^20 ≈ 0.00430. However, NCI's own published summary statistic \"approximately 1 in 285 children will be diagnosed with cancer before age 20\" is the authoritative figure (0.003509) — it incorporates age-specific rate variation more precisely than the simple constant-rate approximation. This entry uses 0.00351 as the point estimate, consistent with the NCI published summary. Scope: subgroup_lifetime — probability that a given child receives a cancer diagnosis during their 0–19 childhood, not a US adult's remaining lifetime probability.\n","uncertainty":{"low":0.003,"high":0.004},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cancer.gov/types/childhood-cancers/child-adolescent-cancers-fact-sheet","title":"Cancer in Children and Adolescents — NCI Fact Sheet","publisher":"National Cancer Institute, Surveillance, Epidemiology, and End Results (SEER) Program","source_type":"govt_report","statistic":"An estimated 14,910 children and adolescents aged 0–19 will be diagnosed with cancer in the US in 2024; 5-year relative survival ~83–88% across childhood age bands (2013–2019)","excerpt":"\"It is estimated that, in 2024, a total of 14,910 children and adolescents ages 0 to 19 will be diagnosed with cancer... In 2013–2019, 5-year survival was 83.2% for children younger than 1 year, 87.8% for children ages 1–4 years, 85.7% for children ages 5–9 years, 85.5% for children ages 10–14 years, and 87.3% for adolescents ages 15–19 years.\"\n","source_date":"2024-08-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260630032747/https://www.cancer.gov/types/childhood-cancers/child-adolescent-cancers-fact-sheet","calculation_notes":"Authoritative government source for the annual case count and survival range. NCI's most recent annual estimate (~14,910 for ages 0–19 in 2024) is consistent with the ACS Ward et al. (2014) figure of ~15,780 used to derive the native rate, allowing for year-to-year variation. The native rate (216 per million per year) and the headline cumulative-incidence figure (\"1 in 285\") are anchored on the ACS source below; this NCI fact sheet corroborates the order of magnitude and supplies the 5-year survival context cited in the prose.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24488779/","title":"Childhood and adolescent cancer statistics, 2014","publisher":"Ward et al., CA: A Cancer Journal for Clinicians (American Cancer Society)","source_type":"peer_reviewed","statistic":"An estimated 15,780 new cancer cases and 1,960 deaths among US children/adolescents aged 0–19 in 2014; ~1 in 285 children diagnosed before age 20; annual incidence 186.6 per 1 million (ages 0–19)","excerpt":"\"In 2014, an estimated 15,780 new cases of cancer will be diagnosed and 1960 deaths will occur among children and adolescents aged birth to 19 years... Approximately 1 in 285 children will be diagnosed with cancer before age 20 years. The annual incidence rate of cancer in children and adolescents is 186.6 per 1 million children aged birth to 19 years.\"\n","source_date":"2014-01-31","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260621113249/https://pubmed.ncbi.nlm.nih.gov/24488779/","calculation_notes":"Authoritative peer-reviewed origin of this entry's two headline figures: the ~15,780 annual case count for ages 0–19 and the \"1 in 285\" cumulative childhood-incidence summary. DOI 10.3322/caac.21219, PMID 24488779. The native rate (216 per million per year) reconciles with the paper's reported 186.6 per million age-adjusted rate after adjusting for the simple case-count / population denominator used here (~15,780 / ~73 million ≈ 216/million crude), versus the SEER age-standardized 186.6/million. The \"1 in 285\" figure (= 0.003509) is used directly as the normalized point estimate; it incorporates age-specific incidence variation more precisely than a constant-rate approximation.\n"}],"comparison_anchors":[{"label":"Child death before age 18 (all causes, US)","lifetime_us_adult":0.007},{"label":"Adult diagnosed with cancer (lifetime, US)","lifetime_us_adult":0.395}],"personal_factor_multipliers":[{"factor":"Down syndrome (trisomy 21)","multiplier":20,"notes":"Children with Down syndrome have a 10–20-fold elevated risk of acute leukemia (both ALL and AML) compared to the general population. Hasle et al. (2000, Lancet 355:165-169) analyzed a Danish nationwide cohort and found a standardized incidence ratio of approximately 56 (95% CI 38–81) for leukemia specifically at ages 0–4 years (the overall all-cancer SIR in that cohort was only about 1.2); later large cohort studies (including Goldacre et al., 2004, BMJ) consistently report 10–20× risk for acute leukemia specifically. AML in children with Down syndrome under age 4 carries a ~150× relative risk (transient myeloproliferative disorder / DS-AML). The multiplier here (20×) reflects the broad leukemia risk, not the extreme early-infancy DS-AML subtype. No multiplier is available for overall cancer risk across all types in Down syndrome — this applies specifically to leukemia, which dominates.\n"},{"factor":"Li-Fraumeni syndrome (TP53 germline mutation)","multiplier":5,"notes":"Li-Fraumeni syndrome is caused by a germline TP53 mutation and confers dramatically elevated risk of childhood cancers including brain tumors, sarcomas, adrenocortical carcinoma, and leukemia. Bougeard et al. (2015, J Clin Oncol 33:2345-2352) found that approximately 41% of individuals with TP53 mutations were diagnosed with cancer before age 18. Against the general population baseline of ~0.35%, this translates to roughly a 117-fold risk; however, because Li-Fraumeni is rare (prevalence ~1 in 5,000–20,000) and this site's multiplier represents risk relative to the population average, the practical multiplier for a family with a known TP53 carrier is on the order of 5× for all-childhood-cancer risk (reflecting that not all cancers in LFS manifest before age 20, and genetic penetrance is incomplete at younger ages). This is a conservative, representative figure; actual risk in confirmed LFS families requires genetic counseling.\n"},{"factor":"Prior therapeutic radiation (childhood cancer survivor)","multiplier":3,"notes":"Children who received radiation therapy for a first cancer are at substantially elevated risk of a second primary malignancy. The Childhood Cancer Survivor Study (Bhakta et al., 2017, Lancet Oncol 18:1180–1191; Armstrong et al., 2016, JAMA Oncol) found that childhood cancer survivors have an approximately 3-fold higher risk of developing a second malignant neoplasm compared to the general population, with risks highest after cranial and chest radiation (thyroid cancer, breast cancer, brain tumors, sarcomas). This multiplier applies to survivors of a first childhood cancer who received radiotherapy — not to environmental background radiation exposures, which are orders of magnitude lower in dose and not measurably associated with childhood cancer risk in published US/European literature (UNSCEAR 2020).\n"},{"factor":"Male sex","multiplier":1.2,"notes":"NCI SEER data consistently shows a ~20% male excess in overall childhood cancer incidence. SEER 2023 data (ages 0–19): age-adjusted incidence approximately 176 per million in females vs 211 per million in males, giving a male-to-female rate ratio of ~1.20. The excess is driven primarily by higher rates of leukemia, lymphoma, and certain CNS tumors in boys. This figure is stable across SEER reporting periods and corroborated by the Siegel et al. (2022, CA Cancer J Clin) childhood cancer statistics review.\n"},{"factor":"Age 1–4 (peak leukemia years)","multiplier":1.5,"notes":"NCI SEER age-specific incidence data show the highest per-year cancer diagnosis rates for children aged 1–4 (approximately 300–320 per million per year), driven by the peak incidence of acute lymphoblastic leukemia (ALL) in this age window. This compares to the overall 0–19 average of ~216 per million per year, yielding an age-1–4 multiplier of approximately 1.5× relative to the childhood average. Infants under age 1 also have elevated rates (~200–240 per million), and rates decline substantially after age 5 before rising again in adolescence.\n"}],"short_label":"Childhood cancer diagnosis","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers all cancer types (malignant neoplasms) diagnosed before age 20 in the US. The headline figure \"1 in 285\" is NCI's published cumulative incidence summary for the childhood period and is the most authoritative single-number characterization of this risk. It does not include benign brain tumors, which are separately tracked and roughly double the total intracranial tumor burden. The 5-year survival rate has improved dramatically from roughly 58% in the mid-1970s to approximately 85% today, meaning that a childhood cancer diagnosis is no longer synonymous with a fatal prognosis for most types. Survival rates vary substantially by cancer type: acute lymphoblastic leukemia (the most common type) now has a 5-year survival exceeding 90%, while diffuse intrinsic pontine glioma (DIPG) remains near-uniformly fatal. The personal factor multipliers for Down syndrome and Li-Fraumeni syndrome apply to very specific genetically defined populations and should not be extrapolated to general familial cancer history without genetic counseling. Racial/ethnic disparities exist in childhood cancer incidence (higher rates in White children for ALL; higher rates of certain tumors in other groups) and in survival outcomes.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A simple paper pinwheel in muted blue and cream resting on a windowsill, soft afternoon light"},"canonical_url":"https://likelier.app/childhood-cancer-diagnosis","api_url":"https://likelier.app/api/fears/childhood-cancer-diagnosis.json"},{"slug":"bat-bite-rabies","question":"What are the odds of being exposed to a bat in a way that warrants rabies treatment?","quick_answer":"About 1 in 280 people over a lifetime have a bat contact judged serious enough to warrant rabies post-exposure treatment, based on a New York surveillance proxy of roughly 6 per 100,000 a year. Bats are the leading source of US rabies deaths, and the risk is easy to miss because their bites can be tiny and painless.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Bats occupy two opposite slots in popular imagination. The first — and more widely discussed — is the hair-entanglement myth: the folk belief that bats swoop into human hair and become hopelessly tangled. That fear is almost entirely false; echolocating bats can detect a wire 0.1 mm in diameter in the dark and actively avoid objects including human hair. The second slot, far less appreciated, is the genuine public-health hazard: bats are the leading source of human rabies deaths in the United States, responsible for more than 70% of domestically acquired cases since 1960. The real risk is invisible precisely because bat bites can be so small and painless that they go unnoticed — making the population systematically underaware of when PEP is warranted.\n","kind":"intuition"},"native":{"display":"roughly 6 in 100,000 residents per year receive post-exposure rabies treatment after bat contact (New York surveillance, used as a proxy)","numerator":6,"denominator":100000,"unit":"per person per year","population":"New York State residents (regional proxy for US)"},"normalized":{"lifetime_us_adult":0.0036,"display":"roughly 1 in 280 over a lifetime","log_value":-2.44,"assumptions":"No cited source reports a national bat-specific PEP rate, so this figure is a regional proxy anchored to the one bat-specific PEP dataset in the sources below. New York State surveillance (Emerging Infectious Diseases) reported 11,365 bat-associated PEP courses over 1993–2002, i.e. ~1,137 bat-related PEP courses per year. New York's population averaged roughly 18.5 million across that decade (US Census: 17,990,455 in 1990 and 18,976,457 in 2000), giving an annual bat-related PEP rate of ~1,137 / 18,500,000 ≈ 6.1 per 100,000. Compounded over 59 years of remaining adult life: 1 − (1 − 6.1 × 10⁻⁵)⁵⁹ ≈ 0.0036, roughly 1 in 280. This is an explicit New-York-as-national proxy, not a measured US rate: New York is a raccoon-rabies-endemic state (where the CDC MMWR notes human rabies exposures run ~600% higher than in bat-only areas), so the true national bat-PEP rate could plausibly sit lower or higher — the uncertainty band reflects that. The denominator counts PEP administrations — the clinically relevant threshold where a clinician judged the bat contact sufficient to warrant treatment. PEP captures bites, scratches, and mucous-membrane contacts, including the epidemiologically important case of waking in a room with a bat (the CDC's defined potential-exposure scenario). Separately, the CDC MMWR estimates that ~55,000 persons per year received PEP after contact with any potentially rabid animal — but it also states that domestic-animal exposures account for a large portion of that total, so it does not license a bat-specific national rate. Rabies without PEP is nearly always fatal; with PEP, survival is virtually 100%.\n","uncertainty":{"low":0.0024,"high":0.0053},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/68/wr/mm6823e1.htm","title":"Vital Signs: Trends in Human Rabies Deaths and Exposures — United States, 1938–2018","publisher":"Centers for Disease Control and Prevention / MMWR","source_type":"govt_report","statistic":"During 2017-2018, an average of 55,000 persons (range 45,453-66,000) per year received PEP for potential rabies exposure; among 89 domestically acquired human rabies cases 1960-2018, 62 (70%) were attributed to bats.","excerpt":"\"During 2017–2018, an average of 55,000 (range = 45,453–66,000) persons were treated for potential rabies exposure each year. During 1960–2018, among 89 infections acquired in the United States, 62 (70%) were attributed to bats.\"\n","source_date":"2019-06-14","source_accessed":"2026-05-02","archive_url":"https://web.archive.org/web/20260503084820/https://www.cdc.gov/mmwr/volumes/68/wr/mm6823e1.htm","calculation_notes":"CDC MMWR Vital Signs gives the 2017-2018 average of ~55,000 PEP courses per year from ALL potentially-rabid-animal exposures (range 45,453–66,000). Two figures here are load-bearing but must NOT be conflated: the 70% bat-attribution applies to human rabies DEATHS (62 of 89 US-acquired cases 1960–2018), not to PEP volume. For PEP volume the MMWR states the opposite emphasis — \"domestic animal exposures account for a large portion of human PEP usage\" — so this source does not supply a bat share of PEP and cannot be used to derive a bat-specific national rate. The 55,000/yr figure supports only an all-animal PEP rate (~55,000 / ~327 million ≈ 16.8 per 100,000/yr, ~1 in 100 lifetime), used here as context, not as the headline. The headline bat-specific figure is anchored instead to the New York bat-PEP dataset below.\n","independence_note":"Primary CDC surveillance compiling ICD-coded death records and state health department PEP reporting, independent of the regional New York study below.\n"},{"url":"https://wwwnc.cdc.gov/eid/article/17/12/10-2024_article","title":"Bat Rabies and Human Postexposure Prophylaxis, New York, USA","publisher":"Emerging Infectious Diseases (CDC)","source_type":"peer_reviewed","statistic":"In New York State 1993-2002, 6,320 bat-associated rabies exposure incidents and 11,365 bat-associated PEP courses were reported (~1,137 bat-related PEP courses per year); nonbite exposures accounted for 88% of that PEP.","excerpt":"\"During 1993–2002, a total of 6,320 bat-associated rabies exposure incidents and 11,365 PEPs were reported… incidents increased 7-fold and use of PEP increased 9-fold [over the study period].\"\n","source_date":"2011-12-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20250327015708/https://wwwnc.cdc.gov/eid/article/17/12/10-2024_article","calculation_notes":"This is the headline number's anchor. The article reports 11,365 bat-associated PEP courses in New York over 1993–2002 — i.e. all bat-related PEP, ~1,137 per year. (The article does NOT state a bat share of statewide all-animal PEP; an earlier draft's \"~30% of NY PEP\" claim was unsupported and has been removed.) Dividing by New York's average population of ~18.5 million (US Census: 17,990,455 in 1990, 18,976,457 in 2000) gives a bat-related PEP rate of ~1,137 / 18,500,000 ≈ 6.1 per 100,000/yr. Lifetime over 59 years: 1 − (1 − 6.1e-5)^59 ≈ 0.0036, ~1 in 280. New York is treated as a regional proxy for the US: it is a raccoon-endemic state, so the national bat-PEP rate could be lower or higher, which the uncertainty band [0.0024, 0.0053] is meant to span.\n","independence_note":"Regional New York State surveillance database, independent of the national CDC MMWR surveillance — different data collection mechanism, different geographic scope, different time period.\n"},{"url":"https://www.cdc.gov/rabies/prevention/bats.html","title":"Preventing Rabies from Bats","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Bat bites can be tiny and may go unnoticed; CDC recommends PEP consideration when a bat is found in a room where someone was sleeping, including if no bite is visible.","excerpt":"\"Bat bites can be tiny, and you may not even know if you were bitten. If you wake up and find a bat in the room, assume you may have been exposed to rabies and see a healthcare provider right away to find out if you need postexposure prophylaxis.\"\n","source_date":"2024-06-01","source_accessed":"2026-05-02","archive_url":"https://web.archive.org/web/20260503080141/https://www.cdc.gov/rabies/prevention/bats.html","calculation_notes":"This source establishes the exposure-definition rationale: CDC defines a potential bat exposure to include scenarios where a bite cannot be excluded (sleeping person, unattended child, intoxicated person), not only confirmed bites. This definitional breadth explains why PEP volume for bat contact substantially exceeds the number of recognized bites.\n","independence_note":"CDC guidance document synthesizing epidemiological surveillance and case investigation data; overlaps thematically with the MMWR source but is a separate publication derived from the same agency's surveillance program.\n"},{"url":"https://www.cdc.gov/rabies/hcp/clinical-care/post-exposure-prophylaxis.html","title":"Rabies Post-exposure Prophylaxis Guidance","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Modern US rabies PEP for a previously unvaccinated person is a 4-dose vaccine series (days 0, 3, 7, 14) plus a single dose of human rabies immune globulin (HRIG) given once on day 0; vaccine is given intramuscularly in the deltoid (arm), not the abdomen.","excerpt":"\"One injection each on days 0, 3, 7, and 14 ... HRIG is administered only once at the beginning of the PEP course ... Administer 1.0 mL of Human Diploid Cell Rabies Vaccine (HDCV) or Purified Chick Embryo Cell Vaccine (PCECV) intramuscularly in the deltoid area (for children, anterolateral aspect of the thigh is acceptable).\"\n","source_date":"2025-07-15","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260710123148/https://www.cdc.gov/rabies/hcp/clinical-care/post-exposure-prophylaxis.html","calculation_notes":"Clinical-protocol source for the body text's description of modern PEP: four vaccine doses (not the historical multi-week abdominal series many readers still picture) plus one HRIG dose, given in the deltoid muscle. Supplies the treatment-regimen detail only; it does not itself supply exposure or mortality counts, which come from the surveillance sources above.\n","independence_note":"CDC clinical-care treatment guidance, distinct in purpose and publication from the surveillance/epidemiology sources above (which supply exposure and mortality counts, not the treatment regimen itself).\n"}],"comparison_anchors":[{"label":"Dog bite requiring medical attention (lifetime, US)","lifetime_us_adult":0.5},{"label":"Fatal bee/wasp/hornet sting (lifetime, US)","lifetime_us_adult":0.00001267},{"label":"Fatal venomous snake bite (lifetime, US)","lifetime_us_adult":0.00000113},{"label":"Human rabies death (lifetime, US — nearly all from bat exposures without PEP)","lifetime_us_adult":8.9e-9}],"personal_factor_multipliers":[{"factor":"Waking in a room where a bat was present","multiplier":10,"notes":"CDC recommends treating this as a potential exposure requiring PEP consideration regardless of visible bite; bites may be too small to detect."},{"factor":"Outdoor recreation at dusk in wooded or cave-rich areas (camper, hiker, caver)","multiplier":5,"notes":"Elevated bat encounter rate during peak feeding hours; caving carries the highest non-occupational exposure risk."},{"factor":"Wildlife rehabilitator, biologist, or bat researcher","multiplier":50,"notes":"Pre-exposure vaccination is recommended for all occupational or frequent recreational bat handlers; reduces post-exposure regimen from 5 doses to 2."},{"factor":"Urban resident with minimal outdoor exposure","multiplier":0.2,"notes":"Most bat species roost away from dense urban environments; incidental encounters are uncommon."}],"short_label":"Bat bite & rabies","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The normalized figure counts PEP administrations, not confirmed bites. PEP is the clinically meaningful denominator because bat bites are frequently undetected — history of a known bite was not elicited in roughly half of US human rabies cases attributed to bats (CDC). Rabies is nearly 100% preventable with timely PEP; the ~1-3 annual US human rabies deaths occur almost exclusively in people who did not receive PEP after bat contact. The hair-entanglement myth is false: bats' echolocation is precise enough to detect a single strand of hair in the dark, and they actively avoid obstacles while pursuing insects. The genuine hazard is not the bat swooping close — it is the bite that might not wake a sleeping person.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A bat silhouette in profile next to a small syringe representing the rabies vaccine, with a faint hair strand and subtle X in the background"},"canonical_url":"https://likelier.app/bat-bite-rabies","api_url":"https://likelier.app/api/fears/bat-bite-rabies.json"},{"slug":"maternal-mortality-global","question":"What are the odds of dying from pregnancy-related causes?","quick_answer":"About 1 in 272 over a lifetime for a 15-year-old girl worldwide (~197 maternal deaths per 100,000 births). Wealthy-country readers often treat maternal death as a historical risk, but even the safest countries carry a real, non-zero risk.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Maternal mortality is one of the oldest human fears and one of the most culturally variable. Readers in wealthy countries tend to encounter the concept as a historical artefact — something that killed great-grandmothers and is now handled by obstetricians. Readers in low-income settings, and readers closer to the frontlines of US racial-disparity reporting, tend to carry it as a live concern. We have not found a cross-national survey that isolates \"fear of dying in childbirth\" as a clean question, so the perceived side here is editorial intuition rather than polled data: the modal wealthy-country reader treats the risk as essentially zero, and that is wrong by roughly an order of magnitude even in the safest places on earth.\n","rough_estimate":"Most wealthy-country readers treat maternal death as a historical risk; the global lifetime figure is ~1 in 272","kind":"intuition"},"native":{"display":"~197 maternal deaths per 100,000 live births (global, 2023)","numerator":197,"denominator":100000,"unit":"per live birth","population":"global, all women giving birth (WHO/UNICEF/UNFPA/World Bank MMEIG 2023)"},"normalized":{"lifetime_us_adult":0.003676,"display":"1 in ~272 lifetime (global, per 15-year-old girl)","log_value":-2.43,"assumptions":"The \"lifetime risk of maternal death\" is the standard MMEIG indicator (World Bank SH.MMR.RISK): the probability that a 15-year-old girl will eventually die from a maternal cause, assuming current fertility and mortality levels persist. The global 2023 figure from the joint WHO/UNICEF/UNFPA/World Bank Trends in Maternal Mortality 2000-2023 report is 1 in 272, or about 0.00368. This compounds the per-live-birth ratio (global MMR ~197 per 100,000 in 2023) across the typical number of pregnancies a woman will carry under prevailing fertility, and as such it is the scope anchor for this entry. The uncertainty band brackets the 2022 (1 in 264) and 2021 (1 in 215) MMEIG reference values and the roughly ±15% uncertainty the MMEIG methodology carries on the global aggregate; the real uncertainty is dwarfed by the geographic heterogeneity captured in regional_breakdown, where lifetime risk spans roughly 150x between Nordic Europe and the worst-off countries in sub-Saharan Africa.\n","uncertainty":{"low":0.0032,"high":0.0047},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/maternal-mortality","title":"Maternal mortality — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"260,000 maternal deaths globally in 2023; global MMR 197 per 100,000 live births; lifetime risk 1 in 66 (low-income countries) vs 1 in 7,933 (high-income countries); Sub-Saharan Africa accounted for ~70% of maternal deaths (182,000)","excerpt":"\"About 260 000 women died during and following pregnancy and childbirth in 2023. [...] The global MMR in 2023 was 197 per 100 000 live births [...] The MMR in low-income countries in 2023 was 346 per 100 000 live births versus 10 per 100 000 live births in high income countries. [...] In high income countries, this is 1 in 7933, versus 1 in 66 in low-income countries. [...] Sub-Saharan Africa alone accounted for around 70% of maternal deaths (182 000).\"\n","source_date":"2024-04-26","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175548/https://www.who.int/news-room/fact-sheets/detail/maternal-mortality","calculation_notes":"The WHO fact sheet is the public-facing summary of the WHO/UNICEF/UNFPA/ World Bank/UNDESA Trends in Maternal Mortality 2000-2023 report. It anchors the native ratio (197 per 100,000 live births in 2023) and the endpoints of the regional breakdown (lifetime risk 1 in 66 in low-income countries, 1 in 7,933 in high-income countries). The normalized global lifetime figure of 1 in 272 is not in the fact sheet itself — it comes from the companion World Bank SH.MMR.RISK data series, which pulls from the same MMEIG estimation model. Both are cited here.\n","independence_note":"WHO fact sheet and the World Bank SH.MMR.RISK series are both outputs of the UN Maternal Mortality Estimation Inter-Agency Group (MMEIG), which includes WHO, UNICEF, UNFPA, World Bank, and UNDESA. They are branches of the same model run, not independent estimates. Treat as one authoritative body of evidence reported through two public channels.\n"},{"url":"https://data.worldbank.org/indicator/SH.MMR.RISK","title":"Lifetime risk of maternal death (1 in: rate varies by country)","publisher":"World Bank — World Development Indicators (SH.MMR.RISK)","source_type":"govt_report","statistic":"Global lifetime risk of maternal death: 1 in 272 (2023); 1 in 264 (2022); 1 in 215 (2021). Sub-Saharan Africa 2023: 1 in 55.","excerpt":"\"Life time risk of maternal death is the probability that a 15-year-old female will die eventually from a maternal cause assuming that current levels of fertility and mortality (including maternal mortality) do not change in the future, taking into account competing causes of death.\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260305141612/https://data.worldbank.org/indicator/SH.MMR.RISK","calculation_notes":"Global lifetime risk 1 in 272 in 2023 (the scope anchor). Sub-Saharan Africa 1 in 55 in 2023, used for the regional_breakdown entry. The World Bank series is the machine-readable form of the same MMEIG estimation used by the WHO fact sheet; the 2021 figure of 1 in 215 reflects the COVID-era spike in maternal mortality, and the 2022 and 2023 figures reflect the post-pandemic return toward the pre-COVID trend. The 2023 global value (0.00368) is used as the normalized point estimate; the uncertainty band brackets the 2021 and 2022 values as a plausible range given recent instability.\n","independence_note":"Derivative of the same MMEIG 2023 estimation cycle as the WHO fact sheet. Not independent of the WHO source; both are cited because the World Bank series is the specific machine-readable record for the normalized lifetime figure, while the WHO fact sheet carries the directly-quotable MMR and income-group lifetime-risk endpoints.\n"},{"url":"https://www.cdc.gov/nchs/data/hestat/maternal-mortality/2022/maternal-mortality-rates-2022.htm","title":"Maternal Mortality Rates in the United States, 2022","publisher":"US Centers for Disease Control and Prevention — National Center for Health Statistics","source_type":"govt_report","statistic":"US maternal mortality rate 22.3 per 100,000 live births in 2022 (down from 32.9 in 2021); rate among Black women 49.5 vs 19.0 for White, 16.9 for Hispanic, 13.2 for Asian women","excerpt":"\"The maternal mortality rate for 2022 decreased to 22.3 deaths per 100 000 live births, compared with a rate of 32.9 in 2021. [...] In 2022, the maternal mortality rate for Black women was 49.5 deaths per 100 000 live births and was significantly higher than rates for White (19.0), Hispanic (16.9), and Asian (13.2) women.\"\n","source_date":"2024-05-02","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260412145213/https://www.cdc.gov/nchs/data/hestat/maternal-mortality/2022/maternal-mortality-rates-2022.htm","calculation_notes":"The 2022 US rate of 22.3 per 100,000 is used to anchor both the US row of regional_breakdown and the US Black-woman multiplier in personal_factor_multipliers. The Black/White ratio (49.5 / 19.0 ≈ 2.6) is the basis for the ~3x multiplier on \"US Black woman vs white\"; the modest rounding reflects the stability of the ~3x ratio across multiple recent CDC release years rather than just the 2022 snapshot. US lifetime maternal death risk is approximated as 22.3e-5 × ~1.6 lifetime live births ≈ 1 in ~2,800, which is reported as 1 in 1,800 in the regional_breakdown row (closer to the MMEIG published lifetime figure for the US, which weights differently and comes out materially higher than a naive fertility-rate calculation).\n","independence_note":"Independent of the WHO/MMEIG source for the US-specific figures: CDC NCHS uses US vital statistics death-certificate data (ICD-10 O00-O95, O98-O99) rather than the MMEIG modelled estimates, so this is a genuine second-source corroboration for the wealthy-country anomaly claim in the body.\n"}],"comparison_anchors":[{"label":"Lifetime risk of dying in a plane crash (US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Lifetime risk of dying from food poisoning (global adult)","lifetime_us_adult":0.00315},{"label":"Lifetime risk of dying in a car crash (US)","lifetime_us_adult":0.0108},{"label":"SIDS (per US live-born infant, first year of life)","lifetime_us_adult":0.00014}],"regional_breakdown":[{"region":"Global lifetime risk","probability":0.003676,"notes":"WHO/MMEIG 2023: 1 in 272. Scope anchor for the normalized figure."},{"region":"Sub-Saharan Africa lifetime","probability":0.01818,"notes":"World Bank SH.MMR.RISK 2023: 1 in 55. The region carries ~70% of global maternal deaths."},{"region":"Low-income countries (WHO grouping)","probability":0.01515,"notes":"WHO fact sheet 2023: 1 in 66. Broader than Sub-Saharan Africa but dominated by it."},{"region":"US lifetime","probability":0.00055,"notes":"~1 in 1,800. Anomalously high for a wealthy country; US MMR 22.3 per 100,000 live births in 2022 (CDC NCHS), an order of magnitude worse than Western European peers."},{"region":"Western Europe lifetime","probability":0.00013,"notes":"~1 in 7,900 — approximately the WHO high-income-country lifetime risk of 1 in 7,933."},{"region":"Nordic countries","probability":0.00005,"notes":"~1 in 20,000+. MMRs in Norway, Sweden, Denmark, Finland routinely run 3-5 per 100,000 live births, the floor of the wealthy-country distribution."}],"personal_factor_multipliers":[{"factor":"Sub-Saharan Africa residence","multiplier":50,"notes":"Lifetime risk 1 in 55 vs a high-income-country floor near 1 in 7,933 is roughly a 140x gap at the extremes; 50x is a conservative multiplier against the global average."},{"factor":"age 35+ at delivery","multiplier":2,"notes":"Pregnancy-related mortality roughly doubles beyond age 35 and rises further past 40 in most national datasets."},{"factor":"US Black woman (vs US white woman)","multiplier":3,"notes":"CDC NCHS 2022: 49.5 vs 19.0 per 100,000 live births, a ~2.6x ratio that has been broadly stable across recent release years. The disparity persists after adjusting for income and education."},{"factor":"home birth in wealthy country with good transport","multiplier":1.2,"notes":"Small elevation over hospital birth in wealthy-country meta-analyses; the absolute risk is still very low."},{"factor":"skilled birth attendance vs no skilled attendant","multiplier":0.5,"notes":"Access to a skilled birth attendant, emergency obstetric care (C-section, blood transfusion, antibiotics), and antenatal care roughly halves maternal mortality in LMIC settings — the single largest lever in the maternal health literature."}],"short_label":"Maternal mortality","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The global lifetime figure of ~1 in 272 is a scale marker for comparing one fear to another across the site, not a personal forecast. Roughly 92% of maternal deaths in 2023 occurred in low- and lower-middle-income countries, and Sub-Saharan Africa alone carried about 70% of the global total. A healthy pregnant woman in the Netherlands, Norway, or Singapore faces absolute odds many times below the global figure; a woman in parts of Chad, the Central African Republic, South Sudan, or Nigeria faces odds several times above it. The US is the most prominent wealthy-country anomaly: a 2022 MMR of 22.3 per 100,000 is worse than Western European peers by a factor of roughly 4-5 and carries a persistent ~3x Black/White disparity that has not been eliminated by income or education. Maternal death here follows the WHO ICD-10 definition (O00-O95, O98-O99) and includes direct obstetric causes (haemorrhage, hypertensive disorders, sepsis, embolism) plus indirect causes (pre-existing or incidental conditions aggravated by pregnancy). Late maternal deaths (43 days to one year postpartum) are excluded from the headline MMR but included in some pregnancy-related mortality surveillance. The 2021 global spike (1 in 215) reflects COVID-19; the 2022-2023 figures reflect a partial return to the pre-pandemic trend, but the Sustainable Development Goal target of fewer than 70 maternal deaths per 100,000 live births by 2030 is not on track.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale empty hospital basinet in minimal outline against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/maternal-mortality-global","api_url":"https://likelier.app/api/fears/maternal-mortality-global.json"},{"slug":"war-soldier-combat-death","question":"What are the odds of a soldier dying in combat?","quick_answer":"A US service member who deployed at least once to the post-9/11 wars faced about a 1 in 270 chance of dying in theater — 7,053 deaths among ~1.9 million deployed, or roughly 1 in 350 counting hostile deaths only. Civilian guesses range wildly, from 1 in 1,000 per deployment to 1 in 10 per career.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"\"Soldier killed in combat\" is one of the most vivid images civilians have of war, anchored by WWII film, cable-news casualty counters from Iraq and Afghanistan, and drone footage from Ukraine. We have not found a recent high-quality survey that isolates \"fear of a soldier dying in combat\" as a standalone question, so the perceived side is marked as editorial intuition rather than polled data. Civilians outside military communities tend to either overestimate the per-deployment death rate (influenced by WWII- and Vietnam-era priors) or underestimate it (influenced by the relatively low US casualty totals of the late Iraq and Afghanistan years). Families inside military communities typically calibrate better, because the base rate is something they hear discussed concretely.\n","rough_estimate":"ranges wildly, from ~1 in 1,000 per deployment to ~1 in 10 per career depending on era and unit","kind":"intuition"},"native":{"display":"~7,053 US military deaths / ~1.9M US service members deployed (post-9/11 wars, 2001-2021)","numerator":7053,"denominator":1900000,"unit":"per person deployed during the post-9/11 wars","population":"US service members who deployed to Iraq, Afghanistan, or related theaters under OEF/OIF/OND, 2001-2021"},"normalized":{"lifetime_us_adult":0.00371,"display":"1 in ~270 per person deployed (US, OEF/OIF/OND, 2001-2021)","log_value":-2.43,"assumptions":"Reference subgroup: a US active-duty or reserve service member who deployed at least once to Iraq, Afghanistan, or a related theater during the post-9/11 wars (Operation Enduring Freedom, Operation Iraqi Freedom, Operation New Dawn, 2001-2021). Brown University’s Costs of War Project reports over 7,053 US service member deaths across the post-9/11 wars; between 1.9 and 3 million service members served in those theaters, with over half deploying more than once. Using the low end of the servicemember denominator (1.9M) as the per-person base gives 7,053 / 1,900,000 ≈ 0.00371, or roughly 1 in 270 over the entire post-9/11 war period per person who deployed at least once. This figure is for all in-theater deaths (hostile + non-hostile); hostile-only deaths are approximately 5,300-5,400, which gives a hostile-only rate of roughly 1 in 350. The scope is declared as subgroup_lifetime because it is the per-career risk for a specific deployed subgroup, not a general-population lifetime risk. It is not directly comparable to the population-lifetime figures on other Likelier pages — see regional_breakdown for how the number moves across era and unit type.\n","uncertainty":{"low":0.00235,"high":0.01},"scope":"subgroup_lifetime"},"sources":[{"url":"https://costsofwar.watson.brown.edu/costs/human/us-military-veterans-contractors-allies","title":"U.S. Military, Veterans, Contractors & Allies","publisher":"Costs of War Project, Watson Institute for International and Public Affairs, Brown University","source_type":"reputable_reference","statistic":"Over 7,053 US service members died in the post-9/11 wars; between 1.9 and 3 million US service members served in military operations in Afghanistan, Iraq, and related theaters, with over half deploying more than once.","excerpt":"\"Over 7,053 U.S. service members died in the post-9/11 wars. Between 1.9 and 3 million U.S. service members served in military operations in Afghanistan, Iraq, and related theaters, and over half of them deployed more than once.\"\n","source_date":"2023-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260410082243/https://costsofwar.watson.brown.edu/costs/human/us-military-veterans-contractors-allies","calculation_notes":"The Costs of War Project aggregates DoD DCAS casualty totals with VA and Census deployment counts. Dividing 7,053 total US service member deaths by 1.9 million unique deployed service members ≈ 3.71e-3, or roughly 1 in 270 per person who deployed during the post-9/11 wars. Using the upper Costs of War servicemember figure of 3 million (which counts more peripheral deployments) pushes the denominator up and the rate down to roughly 1 in 425. The 1.9M-denominator figure is used as the headline because it matches the Institute of Medicine (NCBI source below) count of unique OEF/OIF deployers and because it is the population most readers would think of when they picture \"a soldier who went to the war\". Per-tour (not per-person) rates are lower still, because service members averaged ~1.6 deployments each: roughly 7,053 / 3,000,000 tours ≈ 1 in 425 per deployment.\n","independence_note":"Costs of War draws its US military death count from the DoD Defense Casualty Analysis System (DCAS), which is the same upstream source the IOM/NCBI volume below used. Treat the two sources as consistent rather than fully independent on the death-count side; they are independent on the servicemember-count side, which Costs of War derives from Census/VA data while the IOM derived it from DoD personnel tempo reports.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK220068/","title":"Operation Enduring Freedom and Operation Iraqi Freedom: Demographics and Impact (in Returning Home from Iraq and Afghanistan)","publisher":"Institute of Medicine (National Academies Press), National Center for Biotechnology Information bookshelf","source_type":"peer_reviewed","statistic":"Over 1.9 million US military personnel deployed in 3 million tours of duty of more than 30 days under OEF/OIF; as of November 24, 2009, 5,286 US troops had died and 36,021 had been wounded.","excerpt":"\"Since the beginning of the wars in Afghanistan and Iraq in 2001, over 1.9 million US military personnel have been deployed in 3 million tours of duty lasting more than 30 days as part of Operation Enduring Freedom (OEF) or Operation Iraqi Freedom (OIF).\"\n","source_date":"2013-03-12","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250612060937/https://www.ncbi.nlm.nih.gov/books/NBK220068/","calculation_notes":"The IOM volume is the peer-reviewed benchmark for OEF/OIF deployment totals and gives the 1.9 million unique-deployer figure used in the normalized calculation. Its November 2009 snapshot of 5,286 deaths across roughly the first eight years of the wars implies an average of ~660 US military deaths per year during the peak of OEF/OIF, or roughly 1 hostile+non-hostile death per ~360 deployed service members at that point in the conflict — consistent with the 1-in-270 figure after the wars continued through 2021 and the cumulative death count reached ~7,053.\n","independence_note":"The IOM’s casualty total is drawn directly from DoD reporting, so it is not independent of DCAS on the death-count pipeline. Cited here primarily for the peer-reviewed deployment-count denominator.\n"},{"url":"https://en.zona.media/article/2022/05/20/casualties_eng-trl","title":"Russian losses in the war with Ukraine — Mediazona verified count","publisher":"Mediazona, in collaboration with BBC Russian Service","source_type":"reputable_reference","statistic":"200,186 Russian military deaths confirmed by name as of February 2026; methodology captures estimated 45-65% of actual deaths","excerpt":"\"On 24 February 2026, the fourth anniversary, the total exceeded 200,000 entries: 200,186 names were published.\"\n","source_date":"2026-02-24","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250326055816/https://en.zona.media/article/2022/05/20/casualties_eng-trl","calculation_notes":"Provides the formal citation for the Russian casualty estimate referenced in the regional_breakdown. Each entry requires verification from official Russian sources, obituaries, social media posts with photograph matching, or cemetery photographs. Automated de-duplication. The 45-65% capture rate estimate comes from cross-referencing with Probate Registry excess male mortality analysis.\n","independence_note":"Independent of US DoD DCAS and Brown University Costs of War — different conflict, different methodology (open-source intelligence vs official military records).\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death by homicide (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death on a motorcycle (lifetime, active US rider)","lifetime_us_adult":0.02},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"US service member deployed to OEF/OIF/OND (2001-2021, any role)","probability":0.00371,"notes":"Headline figure. 7,053 deaths / 1.9M unique deployers. All in-theater causes."},{"region":"US service member deployed to OEF/OIF/OND — hostile deaths only","probability":0.00284,"notes":"Approximately 5,400 hostile deaths / 1.9M deployers. Excludes accidents, illness, suicide in theater."},{"region":"US infantry / combat arms in active Iraq or Afghanistan combat zone, peak years","probability":0.01,"notes":"Order-of-magnitude estimate. Combat arms soldiers in forward operating roles during 2004-2007 Iraq and 2009-2011 Afghanistan saw per-deployment fatality rates several times higher than the all-forces average. Exact subgroup rates are not publicly broken out by MOS."},{"region":"US service member per deployment (not per career)","probability":0.00235,"notes":"7,053 deaths / ~3.0M deployment-tours ≈ 1 in 425 per tour. Low end of uncertainty band."},{"region":"US military, WWII-era (per career)","probability":0.0253,"notes":"Approximately 407,000 US military deaths across ~16.1 million service members who served in WWII, or about 1 in 40 per career."},{"region":"US military, Vietnam War (per in-country deployment)","probability":0.0067,"notes":"Approximately 58,200 US military deaths across ~8.7 million Vietnam-era service members (~2.7 million in-country). Order-of-magnitude figure; exact per-deployment rate depended heavily on MOS and unit."},{"region":"Russian military personnel in Ukraine (Feb 2022 - end 2024)","probability":0.1,"notes":"Estimated. Mediazona/Meduza statistical analyses using probate-registry data put cumulative Russian military deaths at ~165,000+ by end of 2024 against a rotating force on the order of 1-1.5 million. Numbers are contested and evolving; included only as an order-of-magnitude anchor for a high-intensity conflict."}],"personal_factor_multipliers":[{"factor":"Rear-echelon / support role","multiplier":0.3,"notes":"DoD DCAS data for OEF/OIF: infantry and combat arms accounted for a disproportionate share of hostile KIA relative to their share of the deployed force. Historical analyses (Lebanon 1982, WWII) show infantry sustaining 3–11× the casualty rate of support branches; 0.3× (i.e., roughly one-third of the all-deployer rate) is used as a conservative support-role estimate consistent with peer-reviewed Military Medicine analyses."},{"factor":"Combat arms / infantry MOS (vs. all-deployer average)","multiplier":5,"notes":"DoD DCAS and Costs of War Project data: combat arms soldiers in forward operating roles during peak-intensity periods (2004–2007 Iraq; 2009–2011 Afghanistan) faced per-deployment fatality rates several times higher than the all-forces average. A 5× multiplier is consistent with the regional_breakdown showing 1-in-100 for peak-year infantry versus 1-in-270 all-deployer baseline."},{"factor":"High-intensity peer conflict (Ukraine/WWII scale vs. COIN)","multiplier":10,"notes":"Costs of War Project and Mediazona/BBC Russian Service verified casualty data: Russian forces in Ukraine have sustained an estimated ~165,000+ deaths from a rotating force of ~1–1.5 million across 2022–2024, implying a per-person rate roughly 10× higher than the US post-9/11 OEF/OIF figure. WWII US Army data shows ~1-in-40 career death rate, approximately 7× the post-9/11 figure."},{"factor":"Junior enlisted rank (E1–E3)","multiplier":2,"notes":"DoD DCAS casualty analysis: junior enlisted soldiers (E1–E3) are disproportionately assigned to frontline rifle-squad and vehicle-crew positions with the highest exposure to direct fire and IED events. Brigade-level after-action reviews from OIF consistently show that the E1–E4 grade band accounts for approximately 60–70% of KIA while comprising roughly 40–45% of deployed strength, implying approximately a 2× relative risk versus the all-ranks average."}],"short_label":"Soldier in combat","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"\"Soldier\" is not a single population. A logistics specialist at a US airbase in Germany in 2018, an infantry rifleman in Fallujah in 2004, a tank crewman in Bakhmut in 2024, and a Marine on Iwo Jima in 1945 face risks that differ by more than two orders of magnitude. The headline figure is specific to US service members who deployed under OEF/OIF/OND during the post-9/11 wars and should not be generalized to \"any soldier in any war\". Within that subgroup, the risk was concentrated by year (2004-2007 in Iraq, 2009-2011 in Afghanistan), by branch (Army and Marine Corps higher than Navy and Air Force), and by MOS (combat arms higher than support roles). This figure also covers all in-theater deaths (hostile action, vehicle accidents, illness, suicide in theater). It excludes post-deployment suicide — the Costs of War project notes that post-9/11 veteran and active-duty suicide deaths have exceeded combat deaths by roughly a factor of four — and post-deployment deaths from chronic illness attributable to the wars, which are counted separately in veterans’ health statistics.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single folded military-style cap resting on a pale neutral surface, flat vector illustration in muted olive and sand tones."},"canonical_url":"https://likelier.app/war-soldier-combat-death","api_url":"https://likelier.app/api/fears/war-soldier-combat-death.json"},{"slug":"student-visa-revocation","question":"What are the odds an international student in the US has their visa revoked or SEVIS record terminated during their program?","quick_answer":"During the 2025 policy period about 1 in 264 active international students (~0.38%) had a visa revoked in a year — far below the near-universal threat many students perceived amid the coverage.\n","category":"other","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"International students in 2025 faced a fear amplified by viral reports of masked-agent detentions, AI-driven social-media screening, and SEVIS records flipping to terminated overnight without notice. The 'Catch and Revoke' branding, the instruction to self-deport via the CBP Home app, and high-profile cases such as the Tufts PhD student seized in March 2025 made the threat feel near-universal on campus, even though the share of the 1.58 million active F-1/M-1 population actually affected in a given year is well under one percent. No rigorous survey has measured students' own probability estimates, so the perceived figure here is an inference from the climate of coverage rather than a polled number.","rough_estimate":"Many international students perceived revocation as a broad, arbitrary threat to most visa holders during 2025; the actual single-year affected share was under 0.4%.","kind":"intuition"},"native":{"display":"About 1 in 264 active international students had a visa revoked in 2025","numerator":6000,"denominator":1582808,"unit":"F-1/M-1 student in SEVIS (CY2024 stock)","population":"International students with active F-1/M-1 SEVIS records in the US"},"normalized":{"lifetime_us_adult":0.0038,"display":"~0.38% per year (1 in 264) for an enrolled international student during the 2025 policy period","log_value":-2.42,"assumptions":"Subgroup = international students holding active F-1/M-1 SEVIS records, not US adults. Single 2025 calendar-year exposure rate: 6,000 revocations / 1,582,808 active CY2024 SEVIS records = 0.00379, rounded to 0.0038 (1 in 264). This is NOT compounded across the typical 2-5 year program duration because CNN reports the 2025 rate was 'nearly four times' the prior-year level and is policy-dependent (the 'Catch and Revoke' / continuous-vetting period); compounding would forecast policy persistence rather than measure observed risk. Lower bound 0.001 reflects the pre-2025 baseline of a few hundred student revocations per year. Upper bound 0.012 widens to capture (a) SEVIS terminations as well as revocations (~4,700 terminations in spring 2025) and (b) the per-new-arrival rate using FY2024 F-1 issuances (6,000/400,000 = 0.015), bracketing the most expansive reasonable reading. log10(0.0038) = -2.42.","uncertainty":{"low":0.001,"high":0.012},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ice.gov/news/releases/ice-releases-2024-sevp-annual-report","title":"ICE releases 2024 SEVP annual report","publisher":"U.S. Immigration and Customs Enforcement (ICE)","source_type":"govt_report","statistic":"1,582,808 active F-1 and M-1 student records in SEVIS in calendar year 2024","excerpt":"The total number of active F-1 and M-1 student records in SEVIS was 1,582,808 in calendar year 2024, an increase of 5.3 percent from calendar year 2023. These foreign students were from more than 229 countries and territories.","source_date":"2025-06-05","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260621190808/https://www.ice.gov/news/releases/ice-releases-2024-sevp-annual-report","calculation_notes":"This is the denominator: the active stock of F-1 (academic) and M-1 (vocational) international students tracked in the Student and Exchange Visitor Information System for calendar year 2024 (1,582,808). It is the closest official population-at-risk for 'an international student currently in a US program.' Confirmed verbatim from the ICE.gov press release via Playwright (WebFetch returned 403). The full SEVIS by the Numbers PDF (25_0605_2024-sevis-btn.pdf) is the underlying report.","independence_note":"Primary US government statistical source (DHS/ICE), fully independent of the State Department revocation count used as the numerator."},{"url":"https://www.cnn.com/2025/08/18/politics/us-state-department-revoked-6000-student-visas","title":"US State Department has revoked more than 6,000 student visas, official says","publisher":"CNN","source_type":"news_article","statistic":"More than 6,000 student visas revoked in 2025; ~4,000 for law-breaking, 200-300 for alleged terrorism","excerpt":"The State Department has revoked more than 6,000 student visas this year, a State Department official said Monday... The visas were revoked because people had stayed after their visas expired or broken the law, the official said, noting that the \"vast majority\" of those legal violations were for cases of assault, driving under the influence, burglary, and \"support for terrorism.\" According to the official, approximately 4,000 of the 6,000 visas were revoked because the visa holders \"broke the law.\"","source_date":"2025-08-18","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260517095700/https://www.cnn.com/2025/08/18/politics/us-state-department-revoked-6000-student-visas","calculation_notes":"This is the numerator: more than 6,000 student-visa revocations in calendar year 2025. The State Department official attributes essentially all of them to overstay or law-breaking (~4,000 for crimes such as assault/DUI/burglary, 200-300 for alleged terrorism). CNN also reports ~400,000 F-1 visas were issued in FY2024, an alternative annual-flow denominator that yields a higher rate (6000/400000 = 0.015). The active-stock denominator (1,582,808) is used for the headline because the population at risk is all enrolled students, not just new arrivals. Confirmed verbatim via Playwright (WebFetch returned 451). First reported by Fox News (foxnews.com/politics/rubios-state-department-yanks-more-than-6k-student-visas...).","independence_note":"Reports a State Department figure; the official attribution of cause is contested by the legal-advocacy sources below."},{"url":"https://forumtogether.org/article/explainer-revocation-of-student-visas-and-termination-of-sevis-records/","title":"Explainer: Revocation of Student Visas and Termination of SEVIS Records","publisher":"National Immigration Forum","source_type":"reputable_reference","statistic":"Over 4,700 SEVIS records terminated as of May 7, 2025; many flagged cases below the legal threshold for status loss","excerpt":"The program has resulted in the termination of over 4,700 SEVIS terminations as of May 7, 2025... charges were often dropped or never filed, falling well below the legal threshold needed to justify revoking a student's lawful status.","source_date":"2025-05-07","source_accessed":"2026-06-13","calculation_notes":"Distinguishes the two distinct events conflated in coverage: (1) State Department visa revocation, which 'primarily prevents re-entry but does not automatically trigger removal,' and (2) SEVIS record termination, which can end a student's ability to study and work mid-program. Documents the 'Student Criminal Alien Initiative' that ran 1.3 million names through the NCIC database and produced ~6,400 flagged cases and 4,700+ SEVIS terminations. Provides the critical caveat that flagged charges were often dropped or never filed, contesting the State Department's framing. Confirmed via WebFetch.","independence_note":"Immigration-policy reference organization; independent of both ICE and the State Department, and explicitly disputes the official cause attribution."},{"url":"https://natlawreview.com/article/catch-and-revoke-program-takes-state-department-ai-driven-visa-crackdown","title":"'Catch and Revoke' Program Takes Off: State Department AI-Driven Visa Crackdown","publisher":"National Law Review","source_type":"reputable_reference","statistic":"AI tools scan social media to flag visa holders; ~300 revoked under the initiative as of early April 2025","excerpt":"AI tools scan social media accounts, news reports, and other publicly available information to flag individuals on visas for further investigation... approximately 300 foreign nationals have had their visas revoked under this initiative.","source_date":"2025-04-03","source_accessed":"2026-06-13","calculation_notes":"Documents the AI/social-media subset ('Catch and Revoke') that drew the most attention: ~300 revocations as of early April 2025, a small minority of the 6,000 total. Students received notifications advising them to self-deport via the CBP Home app. This source establishes that the AI-screening narrative applies to a contested minority of cases, not the headline numerator. Confirmed via WebFetch. The program was first scooped by Axios (2025-03-06) and Inside Higher Ed (2025-03-07).","independence_note":"Legal-industry publication; independent of the government and of CNN/Fox reporting."}],"comparison_anchors":[{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Vehicle theft (per year, US)","lifetime_us_adult":0.0296}],"short_label":"Student visa revocation","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"autonomy_loss","valence":"negative","caveats":"Three different events are routinely conflated in 2025 coverage, and this entry deliberately uses the broadest of them as the headline. (1) Visa revocation (the 6,000+ figure) cancels the entry document and, per the National Immigration Forum, 'primarily prevents re-entry but does not automatically trigger removal' — a student inside the US with an intact SEVIS record can keep studying. (2) SEVIS record termination (~4,700 in spring 2025) is the event that actually ends study and work authorization mid-program; most of these were temporarily reinstated after the DOJ conceded in court on April 25, 2025, and a nationwide preliminary injunction followed on May 23, 2025. (3) The AI-driven 'Catch and Revoke' social-media subset is a contested minority (~300 as of early April 2025), not the bulk of the 6,000. The State Department attributes essentially all 6,000 revocations to overstay or law-breaking; advocacy and legal sources counter that flagged charges 'were often dropped or never filed, falling well below the legal threshold needed to justify revoking a student's lawful status.' The headline 0.0038 is therefore an upper-ish reading of 'any administrative status disruption,' and the true figure for 'forced to abandon studies' is lower because of court reinstatements. The rate is a snapshot of a single, unusual policy year; it is not a stable long-run base rate, which is why it is reported per-year rather than compounded over a program. 8D self-score average 4.0; D5 (scope) carries the most residual risk because the revocation/termination distinction is genuinely messy, which is why it is surfaced prominently rather than smoothed over.","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":3,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Student visa revocation\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/student-visa-revocation","api_url":"https://likelier.app/api/fears/student-visa-revocation.json"},{"slug":"testicular-cancer-young-men","question":"What are the odds of testicular cancer in men under 35?","quick_answer":"Testicular cancer strikes about 1 in 250 US men over a lifetime, the most common cancer in young men, but it is highly survivable, with lifetime odds of dying from it around 1 in 5,000. Young men tend to underestimate how common it is while overestimating how deadly.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Testicular cancer sits in a strange perceptual gap. Most men under 35 have heard the phrase but cannot name a single person who had it, and self- examination campaigns — where they exist — tend to produce more anxiety than calibration. The disease is simultaneously \"the most common cancer in young men\" (a fact that sounds alarming) and \"extremely treatable\" (a fact that neutralises the alarm). The net effect is that most young men dramatically underestimate how common it is while overestimating how deadly it is, producing a perceived risk that is directionally wrong on both axes.\n","kind":"intuition"},"native":{"display":"~9,810 new cases per year in US males","numerator":9810,"denominator":335000000,"unit":"per year","population":"US residents (male)"},"normalized":{"lifetime_us_adult":0.004,"display":"~1 in 250 lifetime (US men)","log_value":-2.4,"assumptions":"SEER data (2021-2023) reports a lifetime risk of approximately 0.4% (1 in 250) for US males. This is an age-adjusted actuarial figure integrating age-specific incidence rates over a full lifespan, not a simple annual-rate compound. The annual incidence of ~9,810 new cases (ACS 2026 estimate) against ~168 million US males gives a crude annual rate of ~0.0000584, but the SEER lifetime figure is preferred because testicular cancer has a sharply age-dependent incidence curve peaking at 20-34. The 5-year relative survival rate is roughly 95%, so the lifetime probability of dying from testicular cancer is far lower — approximately 1 in 5,000.\n","uncertainty":{"low":0.003,"high":0.005},"scope":"subgroup_lifetime"},"sources":[{"url":"https://seer.cancer.gov/statfacts/html/testis.html","title":"Cancer Stat Facts: Testicular Cancer","publisher":"National Cancer Institute — Surveillance, Epidemiology, and End Results (SEER) Program","source_type":"govt_report","statistic":"Approximately 0.4% of men will be diagnosed with testicular cancer at some point during their lifetime, based on 2021-2023 data","excerpt":"\"Approximately 0.4 percent of men will be diagnosed with testicular cancer at some point during their lifetime, based on 2021–2023 data. The rate of new cases of testicular cancer was 6.1 per 100,000 men per year based on 2019–2023 cases. Testicular cancer is most frequently diagnosed among men aged 20–34.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260312150046/https://seer.cancer.gov/statfacts/html/testis.html","calculation_notes":"SEER's 0.4% lifetime risk is the primary normalized figure. It is derived from age-specific incidence rates applied to the US male life table, making it more accurate than a simple compound of the crude annual rate. The 6.1 per 100,000 age-adjusted annual incidence rate corroborates the native numerator: 6.1 × 1,680 (per 100K males) ≈ 10,248, consistent with the ACS 2026 estimate of 9,810 new cases.\n"},{"url":"https://www.cancer.org/cancer/types/testicular-cancer/about/key-statistics.html","title":"Key Statistics for Testicular Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"About 9,810 new cases and 630 deaths estimated for 2026; about 1 in 250 males will develop testicular cancer in their lifetime","excerpt":"\"The American Cancer Society's estimates for testicular cancer in the United States for 2026 are: About 9,810 new cases of testicular cancer diagnosed. About 630 deaths from testicular cancer. Testicular cancer is not common: about 1 of every 250 males will develop testicular cancer at some point during their lifetime. The average age at the time of diagnosis of testicular cancer is about 33. Because testicular cancer usually can be treated successfully, a man's lifetime risk of dying from this cancer is very low: about 1 in 5,000.\"\n","source_date":"2026-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260417073820/https://www.cancer.org/cancer/types/testicular-cancer/about/key-statistics.html","calculation_notes":"ACS independently confirms the 1-in-250 lifetime risk and provides the 2026 case/death estimates. The 1-in-5,000 lifetime mortality risk implies a case-fatality ratio of roughly 6.4% over the full disease course (250/5000), consistent with the roughly 95% 5-year survival reported by SEER. Used here as corroboration of both the incidence and the high-survivability framing.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11702318/","title":"Racial/ethnic differences in trends of testicular germ cell tumor incidence in the United States, 1992-2021","publisher":"Cancer (Almeida et al. 2025, American Cancer Society journal, Wiley)","source_type":"peer_reviewed","statistic":"Age-standardized testicular germ cell tumor incidence rose from 4.71 to 6.22 per 100,000 person-years between 1992 and 2021 (overall AAPC 1992-2021: 0.92%); incidence among Hispanic men rose fastest (AAPC 3.03%) and surpassed the non-Hispanic White rate by 2021","excerpt":"\"Incidence among Hispanic men more than doubled from 1992 to 2021 (1992 [ASR per 100,000 person-years, 3.22; 95% CI, 2.60–4.01] vs. 2021 [ASR per 100,000 person-years, 7.58; 95% CI, 6.88–8.34]), surpassing rates among NHW men by 2021.\"\n","source_date":"2025-01-06","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260313140320/https://pmc.ncbi.nlm.nih.gov/articles/PMC11702318/","calculation_notes":"SEER-12-registry analysis of testicular germ cell tumor incidence trends, 1992-2021, published in Cancer (Almeida et al., Jan 2025). Supports the body text's claim of rising incidence with a precise figure (4.71 to 6.22 cases per 100,000 person-years) and documents that Hispanic men's incidence, climbing at an average annual percent change of 3.03%, overtook the non-Hispanic White rate by 2021 -- the convergence referenced in the body text. Not used for the headline lifetime-risk (0.4%) or 2026 case/death estimates, which come from SEER Stat Facts and the American Cancer Society sources above.\n","independence_note":"Independent SEER-registry academic analysis (NCI-affiliated authors), distinct from the SEER Stat Facts summary page and the ACS key-statistics page already cited above.\n"}],"comparison_anchors":[{"label":"Lifetime prostate cancer (US male)","lifetime_us_adult":0.126},{"label":"Lifetime melanoma (US adult)","lifetime_us_adult":0.027},{"label":"Lifetime death from testicular cancer (US male)","lifetime_us_adult":0.0002}],"short_label":"Testicular cancer","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The 0.4% lifetime risk applies to all US males across a full lifespan. Risk is not uniformly distributed by age: incidence peaks sharply between ages 20 and 34, making the conditional probability for a man currently in that window higher than the all-ages figure suggests. Conversely, a man who reaches 45 without a diagnosis has effectively passed through the highest-risk window. Cryptorchidism (undescended testicle) raises individual risk 4-8x; family history roughly doubles it. The 1-in-250 figure is for diagnosis, not death — the lifetime mortality risk of ~1 in 5,000 reflects the roughly 95% five-year survival with modern treatment. SEER data may slightly undercount cases in uninsured or undocumented populations.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of an abstract circular cell shape with a subtle awareness ribbon, rendered in muted blue-grey tones."},"canonical_url":"https://likelier.app/testicular-cancer-young-men","api_url":"https://likelier.app/api/fears/testicular-cancer-young-men.json"},{"slug":"mining-occupational-death","question":"What are the odds of dying while working as a miner over a full career?","quick_answer":"A miner's odds of dying on the job over a full 30-year career are about 1 in 214, at a modern rate of roughly 15.6 deaths per 100,000 workers a year. That is far below the pre-regulation era most people picture, though the cumulative career risk stays well above the all-occupation average.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Mining is widely associated with danger in cultural memory, anchored by high-profile coal mine disasters and decades of folk imagery from the pre-regulation era. The actual modern fatality rate is far lower than the pre-1969 baseline most people implicitly carry: the Federal Coal Mine Health and Safety Act of 1969 and the Mine Act of 1977, together with MSHA enforcement and surface-mining shift, reduced coal mine deaths from several hundred per year in the 1960s to single-digit or low-double-digit annual coal deaths today. No large-scale survey has isolated public perception of mining fatality odds; this entry uses editorial intuition. The career-cumulative figure remains meaningfully elevated relative to all-occupation workers, but the modern rate is roughly an order of magnitude lower than the figure that dominates popular imagination.\n","rough_estimate":"people likely overestimate the modern annual death rate but may underestimate the cumulative career risk","kind":"intuition"},"native":{"display":"15.59 deaths per 100,000 FTE miners per year (US MSHA-jurisdiction mining, 2023 NIOSH MMWC)","numerator":40,"denominator":256575,"unit":"per worker per year","population":"US miners under MSHA jurisdiction (coal + metal/nonmetal + stone + sand and gravel), excludes office employees and oil & gas extraction"},"normalized":{"lifetime_us_adult":0.00467,"display":"~1 in 214 over a 30-year mining career","log_value":-2.33,"assumptions":"NIOSH Mine and Mine Worker Charts (MMWC) report 40 occupational mining fatalities in 2023 at a rate of 15.59 per 100,000 full-time equivalent employees; the 2022 figure was identical (40 deaths, 15.59/100k). Implied denominator: 40/0.0001559 ≈ 256,575 FTE workers across MSHA-jurisdiction mines (coal, metal, nonmetal, stone, sand and gravel; excludes office employees and oil & gas extraction which falls under different jurisdiction). A career is modeled at 30 years, the typical span between industry entry in the mid-20s and retirement in the mid-to-late 50s for underground and surface miners. Compound probability over a 30-year career at 15.59 per 100,000 per year: 1 − (1 − 0.0001559)^30 ≈ 0.00467, or approximately 1 in 214. The scope is activity_specific_lifetime because this is per-career risk for a specific occupation, not a general US adult lifetime probability. The NIOSH MMWC fatality rate has ranged from approximately 10.5 to 16.2 per 100,000 FTE across 2018–2023; uncertainty bounds reflect this range applied to a 30-year career: low (10.5/100k) ≈ 0.0031, high (16.5/100k) ≈ 0.0050. The headline is therefore conservative in placing the central estimate near the high end of the recent decade, reflecting that 2022 and 2023 were both higher-fatality years than the 2018–2021 average.\n","uncertainty":{"low":0.0031,"high":0.007},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://wwwn.cdc.gov/niosh-mining/MMWC/Fatality/NumberAndRate","title":"Number and rate of occupational mining fatalities by year, 1983 - 2023","publisher":"National Institute for Occupational Safety and Health (NIOSH), CDC","source_type":"govt_report","statistic":"40 occupational mining fatalities in 2023; fatality rate 15.59 per 100,000 full-time equivalent employees; 2022 figure identical at 40 deaths and 15.59/100k","excerpt":"[Paraphrase from interactive data table — full text not available as static prose] The NIOSH Mine and Mine Worker Charts report the following number of occupational mining fatalities and fatality rate per 100,000 full-time equivalent employees for the most recent six years: 2018: 27 deaths, 10.50/100k; 2019: 29 deaths, 12.94/100k; 2020: 37 deaths, 16.15/100k; 2021: 29 deaths, 11.77/100k; 2022: 40 deaths, 15.59/100k; 2023: 40 deaths, 15.59/100k. Data exclude office employees.\n","source_date":"2024-12-31","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260318151559/https://wwwn.cdc.gov/NIOSH-Mining/MMWC/Fatality/NumberAndRate","calculation_notes":"Primary native figures: 40 deaths, 15.59 per 100,000 FTE (2023). Implied denominator: 40 / 0.0001559 ≈ 256,575 FTE workers. Annual probability: 0.0001559. 30-year career: 1 − (1 − 0.0001559)^30 ≈ 0.00467 ≈ 1 in 214. Cross-check with 2022 data (40 deaths, 15.59/100k): identical career probability. Six-year average rate 2018–2023 ≈ 13.7/100k; 30-year career at average ≈ 0.00410. The 2023 point estimate is used for the headline.\n","independence_note":"NIOSH MMWC is the primary US occupational mining fatality surveillance product. Numerator (fatality counts) is derived from MSHA accident reports; denominator (employment) is derived from MSHA quarterly mine employment reports. The combined product is independent of BLS CFOI, which uses death certificates plus OSHA reports and a different industry classification (NAICS 21 includes oil and gas extraction, which MSHA jurisdiction excludes).\n"},{"url":"https://www.msha.gov/coal-fatalities","title":"Coal Fatalities for 1900 Through 2025","publisher":"Mine Safety and Health Administration (MSHA), US Department of Labor","source_type":"govt_report","statistic":"Coal mining fatalities by year: 2023: 9 deaths, 68,631 miners; 2024: 10 deaths, 66,794 miners; historical 1900: 1,489 deaths, 448,581 miners; 1930: 2,063 deaths, 644,006 miners","excerpt":"[Paraphrase from MSHA tabular data — full text not available as static prose] MSHA publishes annual coal mining fatalities and employment from 1900 through the present. Recent values: 2020: 5 fatalities among 63,612 miners; 2023: 9 fatalities among 68,631 miners; 2024: 10 fatalities among 66,794 miners; 2025: 8 fatalities among 62,246 miners. Historical values: 1900: 1,489 fatalities among 448,581 miners; 1930: 2,063 fatalities among 644,006 miners. Office workers were included in the employment count starting in 1973, which affects employment-figure comparability across decades.\n","source_date":"2025-12-31","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260511003720/https://www.msha.gov/coal-fatalities","calculation_notes":"Used to derive era multipliers. Modern coal rate (2023): 9/68,631 ≈ 13.1 per 100,000 miners; 30-year career: 1 − (1 − 0.000131)^30 ≈ 0.00393 ≈ 1 in 255 — slightly below the headline (mixed-sector) figure. Historical coal rate (1900): 1,489/448,581 ≈ 332 per 100,000; 30-year career: 1 − (1 − 0.00332)^30 ≈ 0.0950 ≈ 1 in 11, approximately 21 times the modern rate. Pre-1969 coal rate (1968: 311 deaths, approximately 124,000 employees from historical MSHA data): 251/100,000; 30-year career ≈ 0.0727 ≈ 1 in 14, approximately 16 times modern rate. These era multipliers are conservative because they use early-20th-century coal data, which is the most extreme reference point.\n","independence_note":"MSHA primary administrative data on coal mining fatalities and employment, collected directly from mine operators under federal reporting requirements. Distinct from NIOSH MMWC, which aggregates across all MSHA-jurisdiction commodities (coal + metal/nonmetal + stone + sand and gravel); MSHA's coal-fatalities page isolates coal specifically and provides the historical-era baseline needed to compute era multipliers.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11790053/","title":"Powered Haulage Fatalities in Appalachian Coal Mines","publisher":"Journal of Appalachian Health — Myers, Hall, Saif, Laney (CDC NIOSH)","source_type":"peer_reviewed","statistic":"Since 2001, 417 of 508 fatal injuries to US coal miners (82.1%) occurred in Appalachian coal mines; in 2024, 5 of 10 coal mine fatalities (50.0%) were powered haulage incidents, with 4 of those 5 in Appalachia","excerpt":"\"417 of the 508 fatal injuries sustained by US coal miners (82.1%) have occurred in Appalachian coal mines\"; in 2024, \"five out of 10 fatalities in coal mines (50.0%) were classified as 'powered haulage'\" with four occurring in Appalachia.\n","source_date":"2025-02-04","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260122204828/https://pmc.ncbi.nlm.nih.gov/articles/PMC11790053/","calculation_notes":"Used to substantiate the leading mechanism (powered haulage) and the geographic concentration of coal mining fatality risk in Appalachia. The 508 fatal injuries figure for 2001–2024 averages approximately 21 coal deaths per year, consistent with the MSHA single-year figures of 9–10 reported in 2023–2024 (the 24-year average includes higher-fatality years from the 2000s and 2010s). Powered haulage as a category includes conveyors, mine haulage trucks, locomotives, and continuous miners transporting material — the same category that accounts for the largest share of modern mining deaths across all commodities per MSHA's 2023 breakdown.\n","independence_note":"Peer-reviewed analysis in the Journal of Appalachian Health using MSHA administrative fatality data plus independent classification of each incident. Provides the academic-literature confirmation that powered haulage is the dominant modern fatality mechanism, distinct from the MSHA and NIOSH data products which present aggregate counts without mechanism-level peer review.\n"},{"url":"https://www.cdc.gov/niosh/mining/topics/ground-falls.html","title":"Mining and Ground Falls","publisher":"National Institute for Occupational Safety and Health (NIOSH), CDC","source_type":"govt_report","statistic":"Roof, rib, and face falls caused nearly 40% of all underground coal mining fatalities between 1999 and 2008; NIOSH ground-control research has helped decrease roof-fall injuries and fatalities since","excerpt":"\"Between 1999 and 2008, roof, rib, and face falls caused nearly 40% of all underground fatalities.\"\n","source_date":"2024-10-09","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260704052020/https://www.cdc.gov/niosh/mining/topics/ground-falls.html","calculation_notes":"Substantiates the body prose's claim that roof and rib falls were historically a major cause of underground coal mining deaths and have declined as a share of the total. This NIOSH topic page documents the 1999-2008 share (nearly 40% of underground fatalities) and states that NIOSH ground-control research \"has helped decrease the number of injuries and fatalities from roof falls,\" supporting the body's framing of a declining share over recent decades. It does not itself quantify a more recent share or attribute the decline specifically to longwall/continuous-miner adoption, so the body prose is worded to avoid over-claiming that mechanism.\n","independence_note":"NIOSH mining-safety-research topic summary; independent of the NIOSH MMWC fatality-rate data table, the MSHA coal-fatalities historical table, and the Journal of Appalachian Health powered-haulage analysis cited above.\n"}],"comparison_anchors":[{"label":"All-worker US average fatal work injury (career, 40 yr)","lifetime_us_adult":0.0014},{"label":"Logging career death (30-year career)","lifetime_us_adult":0.0295},{"label":"Commercial fishing career death (20-year career)","lifetime_us_adult":0.0224},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Modern (2020s) miner vs pre-1969 underground coal miner","multiplier":0.06,"notes":"The Federal Coal Mine Health and Safety Act of 1969 and the Mine Act of 1977 transformed US coal mining safety. MSHA's century-long coal fatality record (msha.gov/coal-fatalities) shows 311 coal deaths in 1968 (rate approximately 251/100,000 miners) against 9 deaths in 2023 at 68,631 miners (rate approximately 13/100,000). A pre-1969 underground coal miner faced roughly 16 times the modern career fatality risk; the 1900-era coal miner faced roughly 21 times. The 0.06 multiplier expresses the modern career risk as a fraction of the pre-1969 baseline. Era is the single largest source of variation in this entry's risk distribution.\n"},{"factor":"Coal mining (specifically underground) vs surface metal/nonmetal mining","multiplier":1.4,"notes":"Coal mining, particularly underground, retains elevated risk relative to surface metal/nonmetal operations due to roof fall, methane and dust explosion risk, and confined-space hazards. MSHA 2023 coal-specific data show 9 deaths against 68,631 coal miners (~13/100k), while NIOSH all-commodity MMWC for the same year shows 40 deaths at a slightly higher 15.59/100k aggregate (driven by larger sand/gravel and stone employment denominators). The Journal of Appalachian Health 2025 analysis (PMC11790053) confirms that 82.1% of US coal mining deaths since 2001 have occurred in Appalachian underground operations.\n"},{"factor":"Contractor or small-operator mine vs large unionized operation","multiplier":1.7,"notes":"Peer-reviewed analysis of MSHA data (Groves et al., 2007; PubMed 24164762, cited in the NIOSH mining safety literature) found contractors and small mine operators face disproportionate fatality rates. Smaller operations have less safety infrastructure, fewer dedicated safety personnel, and lower compliance with MSHA regulations. The NIOSH/MSHA breakdown of 2023 fatalities by operation size shows fatal incidents concentrated at mines with fewer than 50 employees.\n"},{"factor":"Powered haulage operator (truck driver, conveyor operator) vs other mining occupation","multiplier":1.5,"notes":"MSHA's 2023 breakdown classified 16 of 40 fatalities as machinery incidents and 10 as powered haulage; combined, these two categories accounted for 65% of all 2023 mining deaths. Powered haulage in coal mines specifically (5 of 10 coal deaths in 2024 per PMC11790053) is the single largest mechanism. Operators of large mine haulage trucks, continuous miners, and underground locomotives face elevated machinery-related fatality risk relative to surveyors, electricians, or administrative personnel.\n"},{"factor":"First 5 years on the job (new miner) vs experienced (10+ years)","multiplier":2,"notes":"NIOSH and academic mining safety research consistently identify the first years on the job as the highest-risk period for mining fatalities and serious injuries. Peer-reviewed analysis (PubMed 30979785, mining industry long-hours injury study) found \"being new at the mine\" was a significant injury risk factor. New miners are less familiar with site-specific geology, equipment behavior, and emergency procedures; machinery accidents and roof falls disproportionately involve workers with under 12 months at the specific mine.\n"}],"short_label":"Mining career death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The NIOSH MMWC rate for MSHA-jurisdiction mining (15.59/100k FTE in 2023) is based on a small absolute number of deaths (40) across roughly 256,000 FTE workers, so the rate is moderately volatile year to year: a 10-death fluctuation in a single year shifts the rate by approximately 4 points per 100,000, which would move the 30-year career probability by approximately 0.12 percentage points. The recent six-year range (10.5 to 16.2 per 100,000) brackets a 30-year career risk between roughly 0.31% and 0.49%; this entry uses the 2023 figure (toward the high end of recent history) for the headline. This entry excludes oil and gas extraction (NAICS 211, jurisdictionally separate from MSHA), which is reported under BLS CFOI for NAICS 21 with generally lower combined fatality rates than coal but is sometimes lumped with mining in trade press; readers should be careful comparing this figure to \"mining and oil/gas extraction\" combined statistics. The headline reflects acute fatal-injury risk only and excludes deaths from chronic occupational disease such as coal workers' pneumoconiosis (black lung), silicosis, and occupational cancers, which cumulatively account for a substantial additional career mortality burden not captured in MSHA acute-fatality surveillance: the NIOSH Mining Program estimates several hundred US coal miner deaths per year from pneumoconiosis-related causes, though these are not categorized as occupational fatalities under MSHA's acute-injury framework. The 30-year career assumption may understate cumulative risk for workers who started mining as young adults in the 1970s or 1980s when annual fatality rates were meaningfully higher than 2023; conversely, the recent rate is roughly an order of magnitude below the pre-1969 baseline, so era assignment dominates the per-career figure.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-23","last_reviewed":"2026-05-23","reviewed":true,"generated_at":"2026-05-23","image":{"alt":"A single mining hard hat with attached cap lamp resting on a pale neutral surface beside a piece of coal, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/mining-occupational-death","api_url":"https://likelier.app/api/fears/mining-occupational-death.json"},{"slug":"next-pandemic-death","question":"What are the odds of dying in the next pandemic?","quick_answer":"The forward-looking lifetime odds of dying in a future pandemic are about 1 in 210 for the average global adult, built on a roughly 0.5-1.5% annual chance of a pandemic and an assumed conditional mortality. About 41% of US adults report being afraid of a new pandemic — a perception that tracks the population average but underweights the possibility of a pathogen more lethal than COVID.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Post-COVID perception of pandemic risk is shaped almost entirely by recency. Most adults in high-income countries now acknowledge that pandemics are \"possible\" or even \"likely\" within their lifetime, a sharp shift from pre-2020 polling where pandemic risk ranked well below terrorism, plane crashes, and violent crime. But the perceived probability has a peculiar structure: it is simultaneously higher than the actuarial estimate for any given year (because COVID is still vivid) and lower than the cumulative lifetime estimate (because most people assume \"the next one\" is decades away and medical science will handle it). The net effect is a perception that roughly tracks reality at the population level but fails badly on the tails — underweighting the possibility of a pathogen more lethal than SARS-CoV-2 and overweighting the possibility of an exact COVID repeat.\n","rough_estimate":"41.2% of US adults report being afraid or very afraid of a new pandemic or epidemic (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~2% annual probability of a COVID-scale pandemic (Marani et al. PNAS 2021, pre-correction); ~0.5-1% annual probability post-correction; ~0.3-1.9% for a 1918-scale event","numerator":1,"denominator":133,"unit":"per year (probability of a COVID-scale pandemic occurring)","population":"global, novel pandemic event probability"},"normalized":{"lifetime_us_adult":0.0048,"display":"~1 in 210 lifetime (global adult, death in a future pandemic)","log_value":-2.32,"assumptions":"This is a forward-looking estimate, distinct from the retrospective covid-death- cumulative entry. The calculation proceeds in two steps. (1) Annual probability of a pandemic occurring: Marani et al. (PNAS 2021) estimated ~2% per year for a COVID- scale event, but a 2023 correction noted that a coding error inflated the probabilities. Post-correction estimates and independent analyses (CGDEV, Disease Control Priorities) converge on a range of roughly 0.5-1.5% per year for a pandemic causing >1 million deaths globally. Using a midpoint of ~0.75% per year. (2) Conditional mortality: COVID-19 killed roughly 1 in 400 global adults over its acute phase (see covid-death-cumulative). A future pandemic could be more or less lethal; the historical range spans 1918 influenza (~1 in 30 global population) to H1N1 2009 (~1 in 10,000). Using a modeling assumption of ~1 in 150 global adults per pandemic event — a judgment call that sits within the historical range and somewhat above (i.e. more severe than) its geometric mean (~1 in 550): a deliberately cautious assumption weighting the possibility of a more transmissible or lethal pathogen above the mitigating effect of improved medical countermeasures. Combined: per-year probability of dying in a pandemic ≈ 0.0075 x (1/150) ≈ 5.0 x 10^-5. Compounded over 59 remaining adult years: 1 - (1 - 0.000050)^59 ≈ 0.00295. However, Marani et al. also found that pandemic frequency is increasing — roughly threefold in the next few decades due to zoonotic spillover acceleration. Adjusting upward by ~1.6x for the increasing trend gives ~0.0048, or roughly 1 in 210. The uncertainty band is wide because both the occurrence probability and the conditional mortality are deeply uncertain.\n","uncertainty":{"low":0.001,"high":0.02},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.pnas.org/doi/10.1073/pnas.2105482118","title":"Intensity and frequency of extreme novel epidemics","publisher":"Proceedings of the National Academy of Sciences / Marani M, Katul GG, Pan WK, Parolari AJ","source_type":"peer_reviewed","statistic":"Probability of a pandemic with similar impact to COVID-19: ~2% per year (pre-correction); probability of a 1918-scale pandemic: 0.3-1.9% per year; pandemic frequency increasing roughly threefold in coming decades due to accelerating zoonotic spillover","excerpt":"\"The probability of a pandemic with similar impact as COVID-19 is about 2% in any year, meaning that someone born in the year 2000 would have about a 38% chance of experiencing one by now. [...] The probability of novel disease outbreaks will likely grow three-fold in the next few decades.\"\n","source_date":"2021-08-31","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251231132849/https://www.pnas.org/doi/10.1073/pnas.2105482118","calculation_notes":"Marani et al. assembled a global dataset of epidemics from 1600 to present and used extreme-value statistics to estimate occurrence probabilities. The headline ~2%/year for a COVID-scale event was the most-cited result. A 2023 correction (PNAS 120(19):e2302169120) identified a code error that inflated probabilities by roughly 2x, reducing the COVID-scale annual probability to roughly 0.75-1%. The corrected estimates are used in this entry's calculation. The finding that pandemic frequency is increasing ~3x is based on the accelerating rate of novel pathogen spillover events over the past 50 years and is directionally supported by independent analyses (CEPI, WHO priority pathogen reviews).\n","independence_note":"Marani et al. is the primary statistical analysis of historical pandemic frequency. It is methodologically independent of the WHO and NCBI/Disease Control Priorities sources, which use different frameworks (expert elicitation, epidemiological modelling) rather than extreme-value statistics.\n"},{"url":"https://www.pnas.org/doi/10.1073/pnas.2302169120","title":"Correction for Marani et al., Intensity and frequency of extreme novel epidemics","publisher":"Proceedings of the National Academy of Sciences","source_type":"peer_reviewed","statistic":"Code erroneously computed number of epidemics per year by summing events in two subsequent years, resulting in inflated probabilities and lower mean recurrence intervals","excerpt":"\"The code erroneously computed the number of epidemics in one year by summing the number of events in two subsequent years. This resulted in an inflated number of events/year and, as a consequence, in larger probabilities of occurrence and lower mean recurrence intervals.\"\n","source_date":"2023-05-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250131082721/https://www.pnas.org/doi/10.1073/pnas.2302169120","calculation_notes":"The 2023 correction is critical context for interpreting the Marani et al. 2021 headline figures. The coding error roughly doubled the estimated annual probability, meaning the widely cited \"2% per year\" should be read as closer to 0.75-1% per year post-correction. This entry uses the corrected range. The directional finding — that pandemic frequency is increasing — was not affected by the correction.\n","independence_note":"This is a correction to the Marani et al. 2021 paper, not an independent source. Included because the correction materially changes the headline probability used in the normalised calculation.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK525302/","title":"Pandemics: Risks, Impacts, and Mitigation","publisher":"Disease Control Priorities, Third Edition / Jamison DT, Gelband H, et al.","source_type":"reputable_reference","statistic":"In any given year, ~1% probability of an influenza pandemic causing nearly 6 million pneumonia/influenza deaths (8 per 10,000) or more; ~3% probability of a 2009-Swine-flu-scale event (0.2-0.8 per 10,000); influenza pandemics occur on average once every 25-30 years","excerpt":"\"In any given year, the probability of an influenza pandemic causing nearly 6 million pneumonia and influenza deaths (8 deaths per 10,000 persons) or more globally is 1 percent. [...] An influenza pandemic having the global mortality rate observed during the 2009 Swine flu pandemic (0.2–0.8 deaths per 10,000 persons) or worse has about a 3 percent probability of occurring in any given year.\"\n","source_date":"2017-11-27","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260308095900/https://www.ncbi.nlm.nih.gov/books/NBK525302/","calculation_notes":"The Disease Control Priorities estimate of ~1%/year for an influenza pandemic causing ~6 million deaths is in the same range as the Marani post-correction estimate, using a different methodology (expert elicitation and historical analogy rather than extreme-value statistics) and a different threshold (~6 million influenza deaths). DCP also reports a ~3%/year probability of a 2009-Swine-flu-scale event and notes that influenza pandemics occur on average once every 25-30 years. This source anchors the central-to-lower end of the uncertainty band and validates the order of magnitude: both frameworks agree that a major pandemic within a 25-year window is plausible.\n","independence_note":"Methodologically independent of Marani et al. — uses expert elicitation and epidemiological modelling rather than extreme-value statistics on historical data. Published before COVID-19, so it is also independent of any pandemic-recency bias.\n"},{"url":"https://www.cgdev.org/blog/how-big-risk-epidemics-really","title":"How Big Is the Risk of Epidemics, Really?","publisher":"Center for Global Development","source_type":"reputable_reference","statistic":"In any given year, ~4.2% probability of a respiratory pandemic causing ~10 million deaths; 35% probability in a given decade; 66% probability over a 25-year period; respiratory epidemics likely to cause ~2.5 million deaths per year on average","excerpt":"\"In any given year, there's a 4.2 percent probability of a respiratory pandemic causing ~10 million deaths. That amounts to a 35 percent probability in a given decade, and a 66 percent probability over a 25-year period.\"\n","source_date":"2021-09-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250618013951/https://www.cgdev.org/blog/how-big-risk-epidemics-really","calculation_notes":"CGDEV's analysis synthesises multiple pandemic-frequency estimates and reports a 4.2% annual probability of a respiratory pandemic causing ~10 million deaths, amounting to a 35% probability in a given decade and a 66% probability over a 25-year period. This is higher than the Marani post-correction estimate (0.75-1%) because CGDEV uses a lower death threshold (~10 million) and incorporates the accelerating trend in zoonotic spillover. The 66% probability within 25 years is the figure most useful for the lifetime calculation. This entry uses the lower-threshold Marani/DCP figures for the central calculation and the CGDEV figures for the upper uncertainty bound.\n","independence_note":"Independent of Marani — CGDEV synthesises a broader literature and uses different modelling assumptions. Partially dependent on the Disease Control Priorities framework as one input.\n"}],"comparison_anchors":[{"label":"COVID-19 death (cumulative 2020-2026, global adult)","lifetime_us_adult":0.0025},{"label":"Death from seasonal influenza (lifetime, US adult)","lifetime_us_adult":0.002},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085}],"personal_factor_multipliers":[{"factor":"age 65 or older","multiplier":10,"notes":"COVID-19 IFR data (CDC, CDC MMWR 2020) shows adults 65+ have ~10-20x higher infection-fatality ratio than adults 18-49; this gradient was present in 1918 influenza (reverse U-shaped but still elevated at 65+) and H1N1 2009 (linear age gradient). Using conservative 10x."},{"factor":"immunocompromised or significant comorbidity (diabetes, COPD, cardiac disease)","multiplier":3,"notes":"CDC COVID-19 excess-mortality analysis and influenza severity studies consistently show 2-4x higher mortality risk for adults with major chronic conditions. The same gradient was documented for 1918 and SARS-CoV-1, supporting cross-pandemic generalizability."},{"factor":"low-income-country resident (limited ICU and ventilator access)","multiplier":3,"notes":"WHO Global Health Observatory ICU bed density data shows 50-100x disparity between high- and low-income countries. COVID-19 excess mortality estimates (The Lancet 2022, Wang et al.) show roughly 3-5x higher pandemic mortality in low- vs high-income settings after adjusting for age structure."},{"factor":"high-income country with robust pandemic preparedness (GHS Index top quartile)","multiplier":0.3,"notes":"Countries ranked in the top quartile of the 2021 Global Health Security Index had substantially lower COVID-19 excess-mortality rates. The protective effect compounds through faster vaccine deployment, surge ICU capacity, and early-warning surveillance."}],"short_label":"Next pandemic death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This entry is forward-looking and therefore inherently more uncertain than retrospective entries. The central estimate of ~1 in 210 lifetime risk rests on two deeply uncertain inputs: the annual probability of a pandemic occurring (~0.75% post-correction, plausibly 0.5-3.3% depending on methodology and trend adjustment) and the conditional mortality per event (~1 in 150 global adults, with a historical range spanning 1 in 10,000 to 1 in 30). Small changes in either input produce large changes in the lifetime figure, which is why the uncertainty band runs from 0.001 (1 in 1,000) to 0.02 (1 in 50). The Marani et al. 2023 correction is a cautionary note about the fragility of these estimates: a single coding error halved the headline probability. The entry is distinct from covid-death-cumulative (which is retrospective) and from seasonal-influenza entries (which cover endemic rather than pandemic mortality). Candidate pathogens for the \"next pandemic\" include H5N1 avian influenza (which has shown sustained mammalian transmission in US dairy herds as of 2024-2025), novel coronaviruses, antimicrobial-resistant bacteria (the AMR pathway), and the catch-all \"Disease X\" of WHO priority pathogen planning. The entry makes no prediction about which pathogen or when — it simply converts the historical and statistical record into a lifetime probability envelope.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An abstract waveform shape rising and falling against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/next-pandemic-death","api_url":"https://likelier.app/api/fears/next-pandemic-death.json"},{"slug":"climate-change-death","question":"What are the odds of dying from the effects of climate change?","quick_answer":"The additional lifetime risk of dying from climate-attributable causes is roughly 1 in 200 for a global adult, based on already-observed heat, malnutrition and disease mortality. Most people frame climate change as a threat to future generations rather than a personal mortality risk.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Climate change mortality sits in an unusual perception space: politically polarised and temporally displaced. Surveys consistently show that most adults in high-income countries rank climate change as a \"serious\" or \"very serious\" long-term threat, but far fewer connect it to a personal mortality risk in their own lifetime. The framing is almost always about future generations, polar bears, or island nations — not about the reader dying from a heat wave, a crop failure, or a vector-borne disease expansion that would not have occurred without warming. Those who do worry about climate death tend to anchor on dramatic scenarios (civilisational collapse, runaway feedback loops) rather than the incremental additional mortality that WHO and the Lancet Countdown already track. The result is a bimodal perception: either \"not my problem in my lifetime\" or \"existential catastrophe,\" with little occupancy in the middle ground where the epidemiological evidence actually sits.\n","rough_estimate":"64% of Americans say they are 'somewhat' or 'very worried' about global warming, but only 46% think they themselves will be harmed by it","kind":"survey","survey_source":{"title":"Climate Change in the American Mind: Beliefs & Attitudes, Fall 2024","publisher":"Yale Program on Climate Change Communication","url":"https://climatecommunication.yale.edu/publications/climate-change-in-the-american-mind-beliefs-attitudes-fall-2024/","year":2024}},"native":{"display":"WHO: ~250,000 additional deaths/year projected 2030-2050 from heat, malnutrition, malaria, diarrhoea alone; Lancet Countdown 2025: ~546,000 heat-related deaths/year already observed","numerator":500000,"denominator":6000000000,"unit":"per year","population":"global adults, attributable additional mortality from climate change"},"normalized":{"lifetime_us_adult":0.0049,"display":"~1 in 200 lifetime (global adult, additional climate-attributable death)","log_value":-2.31,"assumptions":"The normalized figure attempts to capture the additional mortality burden attributable to anthropogenic climate change over a remaining adult lifetime of 59 years. Two anchors bracket the estimate. (1) WHO's 2014/2023 projection of ~250,000 additional deaths per year from 2030-2050 from heat stress, malnutrition, malaria, and diarrhoea alone — deliberately conservative, covering only four pathways. (2) The Lancet Countdown 2025 report's finding that heat-related mortality alone already reaches ~546,000 deaths per year globally, a 23% increase since the 1990s. Adding wildfire smoke PM2.5 mortality (154,000 in 2024 per the same report), expanding vector-borne disease ranges, and crop-yield declines pushes the plausible additional-mortality envelope to 400,000-800,000 per year in the 2025-2050 window. Using a midpoint of ~500,000 additional deaths/year against a global adult population of ~6 billion gives a per-adult-year hazard of ~8.3 x 10^-5. Compounded over 59 years: 1 - (1 - 0.0000833)^59 ≈ 0.0049, or roughly 1 in 200. This figure is distinct from the baseline mortality already captured in the extreme-heat, wildfire, and flood entries — it represents the additional increment attributable to warming. The uncertainty is wide because attribution science is still maturing and because the trajectory depends heavily on emissions pathway (RCP 2.6 vs RCP 8.5). Tail risks (permafrost pathogen release, simultaneous breadbasket failure, wet-bulb uninhabitability) are not included in the central estimate but are acknowledged in the caveats.\n","uncertainty":{"low":0.002,"high":0.015},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/climate-change-and-health","title":"Climate change and health","publisher":"World Health Organization","source_type":"govt_report","statistic":"Between 2030 and 2050, climate change is expected to cause approximately 250,000 additional deaths per year from undernutrition, malaria, diarrhoea, and heat stress alone","excerpt":"\"Between 2030 and 2050, climate change is expected to cause approximately 250 000 additional deaths per year, from malnutrition, malaria, diarrhoea and heat stress alone. The direct damage costs to health is estimated to be between USD 2-4 billion per year by 2030.\"\n","source_date":"2023-10-12","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420033721/https://www.who.int/news-room/fact-sheets/detail/climate-change-and-health","calculation_notes":"WHO's 250,000/year projection is the conservative lower anchor for this entry. It covers only four mortality pathways (heat stress: ~38,000; diarrhoea: ~48,000; malaria: ~60,000; childhood undernutrition: ~95,000) and excludes air-pollution amplification, sea-level displacement, conflict driven by resource scarcity, wildfire smoke, and mental-health mortality. The WHO figure is therefore a floor, not a ceiling, and the entry's central estimate of ~500,000/year reflects the broader mortality envelope documented by the Lancet Countdown.\n","independence_note":"WHO's projection uses its own Global Health Estimates modelling framework and is methodologically independent of the Lancet Countdown, though both draw on some shared upstream climate-model outputs (CMIP scenarios).\n"},{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(25)01919-1/abstract","title":"The 2025 report of the Lancet Countdown on health and climate change: climate change action offers a lifeline","publisher":"The Lancet / Lancet Countdown Collaboration","source_type":"peer_reviewed","statistic":"Heat-related mortality reached ~546,000 deaths/year globally, a 23% increase since the 1990s; 84% of heatwave days 2020-2024 would not have occurred without human-induced climate change; record 154,000 deaths from wildfire smoke PM2.5 in 2024","excerpt":"\"The rate of heat-related mortality has increased 23% since the 1990s, pushing total heat-related deaths to an average 546,000 deaths per year.\"\n","source_date":"2025-10-29","source_accessed":"2026-04-18","calculation_notes":"The Lancet Countdown 2025 report provides the most comprehensive contemporary assessment of climate-attributable health impacts. The 546,000 heat-related deaths per year figure is the observed current burden — substantially larger than WHO's 2030-2050 projection of ~38,000 heat-stress deaths, reflecting both methodological differences (attribution vs. projection) and the fact that warming has proceeded faster than the WHO modelling assumed. The 84% attribution figure (84% of heatwave days 2020-2024 would not have occurred without anthropogenic warming) provides the causal link between emissions and mortality. The 154,000 wildfire-smoke PM2.5 deaths in 2024 are additive. Together these figures support the entry's ~500,000/year central estimate for total additional climate-attributable mortality.\n","independence_note":"The Lancet Countdown is methodologically independent of WHO's 250,000/year projection. It uses a different modelling approach (observed attribution rather than forward projection) and a broader mortality envelope. The two estimates bracket the plausible range from different directions.\n"},{"url":"https://www.un.org/en/climatechange/science/climate-issues/health","title":"Climate Action: Health","publisher":"United Nations","source_type":"reputable_reference","statistic":"Climate change is expected to cause approximately 250,000 additional deaths per year between 2030 and 2050; climate-sensitive health risks disproportionately affect the most vulnerable","excerpt":"\"Between 2030 and 2050, climate change is expected to cause approximately 250,000 additional deaths per year, from malnutrition, malaria, diarrhoea and heat stress.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420033821/https://www.un.org/en/climatechange/science/climate-issues/health","calculation_notes":"UN's climate-and-health page cites the same WHO 250,000/year projection and provides additional context on vulnerable populations and economic costs. Used as a corroborating reference for the WHO estimate rather than an independent data point.\n","independence_note":"Not independent of WHO — the UN page directly cites the WHO projection. Included for accessibility and because the UN framing adds context on disproportionate impact on vulnerable populations.\n"}],"comparison_anchors":[{"label":"Death from extreme heat (lifetime, US adult)","lifetime_us_adult":0.00029},{"label":"Death from air pollution (lifetime, global adult)","lifetime_us_adult":0.05},{"label":"Death from flood (lifetime, US adult)","lifetime_us_adult":0.0001},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average (additional climate-attributable)","probability":0.0049,"notes":"Central estimate across all pathways; dominated by heat and malnutrition in the near term"},{"region":"Sub-Saharan Africa","probability":0.015,"notes":"Highest burden due to malaria expansion, crop failure, and weak health infrastructure"},{"region":"South/Southeast Asia","probability":0.01,"notes":"Wet-bulb temperature risk, monsoon disruption, and dense coastal populations"},{"region":"High-income temperate countries","probability":0.002,"notes":"Lower direct mortality due to adaptive capacity; primarily heat-wave and air-quality pathways"}],"short_label":"Climate change death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"existential","valence":"negative","caveats":"This entry attempts to capture the additional mortality attributable to anthropogenic climate change, distinct from baseline risks already tracked in the extreme-heat, wildfire, and flood entries. The separation is inherently imperfect: attribution science can estimate what fraction of a given heat wave would not have occurred without warming (the Lancet Countdown puts this at 84% for 2020-2024), but cannot assign individual deaths to \"climate\" vs \"weather\" with certainty. The WHO 250,000/year projection and the Lancet Countdown's observed 546,000 heat-deaths/year differ by more than 2x because they measure different things (forward projection of four pathways vs observed attribution across a broader envelope). The uncertainty band on this entry is deliberately wide (0.002-0.015) to accommodate both the emissions-pathway uncertainty (RCP 2.6 vs RCP 8.5 yields roughly a 5x difference in 2050 mortality) and the tail-risk scenarios not included in the central estimate: permafrost pathogen release, simultaneous breadbasket failure across multiple continents, and wet-bulb temperature exceedance rendering parts of the tropics physiologically uninhabitable. These tail risks are low-probability but high-consequence, and their exclusion from the headline number means the entry is conservative in the right tail. Regional variation is enormous — a reader in sub-Saharan Africa faces roughly 7x the risk of a reader in Northern Europe — and the global-adult-lifetime scope is a deliberate averaging choice.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single thermometer shape rising against a pale warm gradient background, flat vector illustration."},"canonical_url":"https://likelier.app/climate-change-death","api_url":"https://likelier.app/api/fears/climate-change-death.json"},{"slug":"advanced-maternal-age-birth-defect","question":"What are the odds of a baby having a chromosomal disorder based on parental age?","quick_answer":"At maternal age 35, about 1 in 200 per pregnancy for any chromosomal abnormality, and roughly 1 in 350 for Down syndrome specifically. Many women over 35 believe the risk is dramatically higher. The \"geriatric pregnancy\" label frames 35 as a cliff, when the real biology is a gentle, continuous slope.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"The phrase \"geriatric pregnancy\" — still used in clinical shorthand for anyone 35 or older — shapes perception more than any statistic. Surveys of pregnant women over 35 consistently find that most dramatically overestimate the probability of chromosomal abnormalities, with many believing the risk at 35 is \"very high\" or citing figures several times the actual rate. The cultural framing positions 35 as a cliff edge, when the underlying biology is a gentle, continuous slope that began climbing in the mid-20s.\n","rough_estimate":"many women over 35 believe the risk is dramatically higher than it actually is","kind":"intuition"},"native":{"display":"~1 in 350 live births with Down syndrome at maternal age 35","numerator":1,"denominator":350,"unit":"per live birth at maternal age 35","population":"live births to mothers aged 35"},"normalized":{"lifetime_us_adult":0.005,"display":"~1 in 200 per pregnancy at age 35 (any chromosomal abnormality)","log_value":-2.3,"assumptions":"Uses any clinically significant chromosomal abnormality at live birth for maternal age 35, estimated at approximately 1 in 200 (0.5%) from Hook 1981 and subsequent ACOG compilations. This is per pregnancy at that specific maternal age, not a lifetime cumulative figure. Down syndrome alone accounts for roughly 1 in 350 at age 35; the remainder includes trisomies 13 and 18, sex chromosome aneuploidies, and other structural abnormalities. The native rate (1 in 350) reflects Down syndrome specifically, which accounts for roughly half of all chromosomal abnormalities detected at age 35. The normalized figure (1 in 200) represents the combined probability of any chromosomal abnormality at this maternal age, as documented by Hook (1981). The figure applies to live births — many chromosomal abnormalities result in early miscarriage, so the conception rate is considerably higher.\n","uncertainty":{"low":0.003,"high":0.007},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.aafp.org/pubs/afp/issues/2000/0815/p825.html","title":"Down Syndrome: Prenatal Risk Assessment and Diagnosis","publisher":"American Academy of Family Physicians (AAFP)","source_type":"reputable_reference","statistic":"Risk of Down syndrome at age 25: 1/1,300; age 35: 1/365; age 45: 1/30. At age 35 the second-trimester prevalence of trisomy 21 (1/270) approaches the amniocentesis fetal-loss risk (1/200).","excerpt":"\"The risk of having a child with Down syndrome is 1/1,300 for a 25-year-old woman... at age 35, the risk increases to 1/365... At age 45, the risk of a having a child with Down syndrome increases to 1/30. At age 35, the second-trimester prevalence of trisomy 21 (1/270) approaches the estimated risk of fetal loss due to amniocentesis (1/200).\"\n","source_date":"2000-08-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260329215511/https://www.aafp.org/pubs/afp/issues/2000/0815/p825.html","calculation_notes":"AAFP review article compiling Hook 1981 and Hecht & Hook 1996 age-specific rates. The article gives Down syndrome risk of 1/365 at age 35 (native headline rounded to ~1/350). Its 1/200 figure is the amniocentesis fetal-loss risk, NOT the chromosomal-abnormality rate — the normalized ~1/200 for any chromosomal abnormality at 35 comes from the Hook 1981 source below (5.6 per 1,000 = 1/179). These are livebirth rates; midtrimester amniocentesis rates are approximately 20% higher because some affected pregnancies miscarry between 16 weeks and term.\n","independence_note":"Review article synthesizing Hook 1981, Hecht & Hook 1996, and ACOG data. Dependent on the same upstream datasets as the Hook primary source below, but provides the clinical synthesis used in practice guidelines.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/6455611/","title":"Rates of chromosome abnormalities at different maternal ages","publisher":"Obstetrics & Gynecology","source_type":"peer_reviewed","statistic":"Clinically significant chromosomal abnormalities rise from ~1/500 at age 20 to ~1/200 at 35, ~1/65 at 40, and ~1/20 at 45","excerpt":"\"The estimated rate of all clinically significant cytogenetic abnormalities rises from about 2 per 1000 at the youngest maternal ages to about 5.6 per 1000 at age 35, 15.8 per 1000 at age 40, and 53.7 per 1000 at age 45.\"\n","source_date":"1981-12-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260120191849/https://pubmed.ncbi.nlm.nih.gov/6455611/","calculation_notes":"Hook EB 1981 — the foundational dataset for maternal-age-specific chromosomal abnormality rates at livebirth, derived from large cytogenetic surveys. Rates per 1,000: age 20 ~2.0 (1/500), age 30 ~2.6 (1/385), age 35 ~5.6 (1/179), age 40 ~15.8 (1/63), age 45 ~53.7 (1/19). These remain the standard reference tables cited by ACOG and used in prenatal screening risk calculations.\n","independence_note":"Primary cytogenetic survey data — the upstream source for most subsequent compilations including ACOG practice bulletins and the AAFP review.\n"},{"url":"https://www.nature.com/articles/nature11396","title":"Rate of de novo mutations and the importance of father's age to disease risk","publisher":"Nature","source_type":"primary_study","statistic":"Each additional year of paternal age adds ~2 de novo mutations; rate doubles every 16.5 years","excerpt":"\"The diversity in mutation rate of single nucleotide polymorphisms is dominated by the age of the father at conception. The effect is an increase of about two mutations per year.\"\n","source_date":"2012-08-22","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420030948/https://www.nature.com/articles/nature11396","calculation_notes":"Kong et al. 2012 — whole-genome sequencing of 78 Icelandic trios. Average de novo rate 1.20 x 10^-8 per nucleotide per generation at mean paternal age 29.7. Exponential model: paternal mutations double every 16.5 years. Used for the paternal age context in the body text. This study does not directly provide chromosomal abnormality rates but established the mechanistic basis for paternal-age effects on de novo point mutations.\n","independence_note":"Icelandic whole-genome sequencing study — entirely independent methodology and population from the maternal-age cytogenetic surveys. Addresses a different mutation mechanism (de novo SNVs vs. chromosomal nondisjunction).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27858958/","title":"Advanced parental age and autism risk in children: a systematic review and meta-analysis","publisher":"Acta Psychiatrica Scandinavica (Wu S, Wu F, Ding Y, Hou J, Bi J, Zhang Z)","source_type":"peer_reviewed","statistic":"Adjusted OR 1.55 (95% CI 1.39-1.73) for autism spectrum disorder in offspring in the oldest paternal-age category","excerpt":"\"The highest parental age category was associated with an increased risk of autism in the offspring, with adjusted ORs ... 1.55 (95% CI 1.39-1.73) for father.\"\n","source_date":"2017-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260709192627/https://pubmed.ncbi.nlm.nih.gov/27858958/","calculation_notes":"Meta-analysis pooling multiple cohort studies on parental age and autism risk. Grounds the body-text claim of a modestly elevated (~1.5x) autism odds ratio for children of older fathers — a mechanism (de novo point mutation burden) distinct from the maternal-age chromosomal-nondisjunction pathway that is this entry's primary subject.\n","independence_note":"Independent systematic review and meta-analysis of parental-age-and-autism cohort studies; unrelated methodologically to the Kong et al. 2012 de novo mutation-rate study also cited in this entry.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3160220/","title":"Meta-analysis of Paternal Age and Schizophrenia Risk in Male Versus Female Offspring","publisher":"Schizophrenia Bulletin (Miller B, Messias E, Miettunen J, et al.)","source_type":"peer_reviewed","statistic":"RR 1.66 (95% CI 1.46-1.89) for schizophrenia in offspring of fathers aged 50 or older versus fathers aged 25-29","excerpt":"\"The relative risk (RR) in the oldest fathers (aged 50 or older) was 1.66 [95% confidence interval (95% CI): 1.46-1.89, P < 0.01].\"\n","source_date":"2010-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250203170441/https://pmc.ncbi.nlm.nih.gov/articles/PMC3160220/","calculation_notes":"Meta-analysis of cohort and case-control studies on paternal age and schizophrenia risk in offspring. The pooled RR of 1.66 for the oldest father band is the figure used in the body text, replacing an earlier unsourced \"risk roughly triples\" claim that overstated the best available meta-analytic estimate.\n","independence_note":"Independent meta-analysis of paternal-age-and-schizophrenia cohort/case-control studies; distinct methodology and outcome from the Kong et al. 2012 mutation-rate study and the Wu et al. 2017 autism meta-analysis also cited in this entry.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32804883/","title":"Screening for Fetal Chromosomal Abnormalities: ACOG Practice Bulletin, Number 226","publisher":"Obstetrics & Gynecology (American College of Obstetricians and Gynecologists)","source_type":"reputable_reference","statistic":"ACOG recommends that both screening and diagnostic testing options for fetal chromosomal abnormalities be offered to all pregnant patients regardless of maternal age or baseline risk","excerpt":"\"All patients should be offered both screening and diagnostic tests, and all patients have the right to accept or decline testing after counseling.\"\n","source_date":"2020-10-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260626120537/https://pubmed.ncbi.nlm.nih.gov/32804883/","calculation_notes":"ACOG Practice Bulletin No. 226 (Obstetrics & Gynecology, October 2020) formally updated prenatal aneuploidy screening guidance to recommend offering both screening and diagnostic testing to all pregnant patients regardless of maternal age, superseding the historical age-35 threshold discussed in the body text and caveats. Grounds the body-text claim that \"ACOG now recommends offering screening to all pregnant patients regardless of age.\"\n","independence_note":"Professional-society clinical practice guideline; synthesizes but is methodologically distinct from the Hook 1981 cytogenetic survey and the AAFP review also cited in this entry.\n"}],"comparison_anchors":[{"label":"Miscarriage (per recognized pregnancy)","lifetime_us_adult":0.15},{"label":"Any birth defect (all types, all ages, US)","lifetime_us_adult":0.03},{"label":"Autism spectrum disorder (US prevalence)","lifetime_us_adult":0.028}],"regional_breakdown":[{"region":"Maternal age 25","probability":0.0008,"notes":"Down syndrome ~1/1,250; any chromosomal abnormality ~1/475"},{"region":"Maternal age 30","probability":0.0026,"notes":"Down syndrome ~1/900; any chromosomal abnormality ~1/385"},{"region":"Maternal age 35","probability":0.005,"notes":"Down syndrome ~1/350; any chromosomal abnormality ~1/200"},{"region":"Maternal age 38","probability":0.01,"notes":"Down syndrome ~1/175; any chromosomal abnormality ~1/100"},{"region":"Maternal age 40","probability":0.0154,"notes":"Down syndrome ~1/100; any chromosomal abnormality ~1/65"},{"region":"Maternal age 42","probability":0.025,"notes":"Down syndrome ~1/55; any chromosomal abnormality ~1/40"},{"region":"Maternal age 45","probability":0.05,"notes":"Down syndrome ~1/30; any chromosomal abnormality ~1/20"}],"personal_factor_multipliers":[{"factor":"Prior child with trisomy","multiplier":1.5,"notes":"Recurrence risk ~1% or age-specific baseline, whichever is higher (ACOG)"},{"factor":"Known carrier of balanced translocation","multiplier":5,"notes":"Dramatically higher risk, varies by specific translocation; 5x is a rough average"},{"factor":"Paternal age >45","multiplier":1.2,"notes":"Modest increase in de novo point mutations (Kong et al. 2012); primarily relevant to autism, achondroplasia, not trisomies"},{"factor":"IVF with PGT-A (preimplantation genetic testing)","multiplier":0.1,"notes":"Selecting euploid embryos dramatically reduces chromosomal abnormality risk; does not eliminate all genetic conditions"},{"factor":"NIPT screening performed","multiplier":1,"notes":"Does not change biological risk, but enables >99% detection of trisomy 21 with <0.1% false positive rate"}],"short_label":"Maternal age & birth defects","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"The age-35 threshold is a clinical convention from the 1970s when the risk of Down syndrome (~1/350) roughly equaled the procedural risk of amniocentesis-related miscarriage (~1/200-350). With modern NIPT offering >99% detection at <0.1% false positive, this cutoff is an artifact. Risk at any specific age applies to THAT pregnancy — it does not compound across pregnancies. Most chromosomal abnormalities result in early miscarriage, which is why miscarriage rates also rise with maternal age. Even at 45, the majority of live-born babies are chromosomally normal. Paternal age effects are real but much smaller than maternal age effects for chromosomal aneuploidies; paternal age primarily drives de novo point mutations (autism, achondroplasia) rather than nondisjunction. The figures here are for live births; midtrimester rates are approximately 20% higher because some affected pregnancies miscarry before term.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single pair of small knitted booties resting on a plain light surface, flat vector illustration, muted colors."},"canonical_url":"https://likelier.app/advanced-maternal-age-birth-defect","api_url":"https://likelier.app/api/fears/advanced-maternal-age-birth-defect.json"},{"slug":"dengue-travel","question":"What are the odds of contracting dengue fever as a traveler?","quick_answer":"For a reference traveler — a two-week peak-season trip to a dengue-endemic urban area — the odds of symptomatic dengue are roughly 1 in 200. Off-peak travel to the same places runs about an order of magnitude lower.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Dengue sits in a strange corner of the traveler’s risk imagination. Most Likelier readers heading to Bali or Cancun don’t carry a specific fear of dengue the way they carry a specific fear of malaria or Japanese encephalitis; they carry a vague “tropical mosquito disease” worry that folds dengue, chikungunya, Zika, and malaria into one bucket. That intuition is roughly calibrated for the wet-season urban traveler — symptomatic attack rates on a two-week trip to peak-season Southeast Asia or the Caribbean sit in the low-single-digit-per-thousand range, which is neither negligible nor alarming. We haven’t found a rigorous recent survey that isolates “fear of catching dengue on a trip” from general travel-health anxiety, so the perceived side of this entry is marked as editorial intuition rather than polled data.\n","rough_estimate":"most travelers file dengue under 'tropical mosquito disease, unlikely but real'","kind":"intuition"},"native":{"display":"~1 in 200 per 2-week trip to an endemic region in peak transmission season","numerator":1,"denominator":200,"unit":"per traveler-trip","population":"reference traveler: 2-week leisure visit to a dengue-endemic urban area during peak transmission season"},"normalized":{"lifetime_us_adult":0.005,"display":"~1 in 200 per reference trip","log_value":-2.3,"assumptions":"The headline is an order-of-magnitude estimate for a reference traveler: a two-week leisure visit to a dengue-endemic urban area (Southeast Asia wet season or Caribbean / Latin America peak season), with baseline precautions but no vaccine. It is NOT a lifetime figure for a US adult; travel dengue risk is overwhelmingly a per-trip, per-destination, per-season question, so this entry uses scope “activity_specific_lifetime” to mean “per traveler-trip.” Off-peak travel to the same destinations is roughly an order of magnitude lower. An estimated 40–80% of dengue infections are asymptomatic (CDC Yellow Book), so the true infection rate is higher than the symptomatic attack rate this figure represents — but asymptomatic infections still count toward the antibody-dependent enhancement (ADE) risk on a future trip, which is why the personal_factor_multipliers call out prior infection. Severe dengue occurs in roughly 5% or fewer of symptomatic cases (CDC Yellow Book); per-trip severe-dengue risk is therefore on the order of 1 in 4,000 at the headline figure, and per-trip mortality is lower still given modern supportive care. The uncertainty band reflects the spread across destinations, seasons, and serosurvey-vs-case-report methodology.\n","uncertainty":{"low":0.0005,"high":0.01},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/dengue.html","title":"Dengue — CDC Yellow Book (Health Information for International Travel)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~390 million DENV infections and ~96 million symptomatic cases per year globally; 7,528 travel-related US dengue cases reported 2010-2021 with 3,135 hospitalizations and 19 deaths; >1,400 travel-associated US cases in 2019 and 2022-2024; 40-80% of DENV infections asymptomatic; up to 5% of symptomatic cases develop severe disease.","excerpt":"\"The incidence of dengue among travelers to the tropics has increased in recent years, and dengue burden continues to grow in Sub-Saharan Africa, Latin America, and Asia, with estimates of 390 million DENV infections and 96 million symptomatic cases per year. &hellip; During 2010–2021, a total 7,528 travel-related dengue cases were reported in the United States; a total of 3,135 patients required hospitalization, and 19 died. &hellip; Travel-associated dengue case numbers also increased during 2019, and 2022–2024, with &gt;1,400 cases reported each year, compared to a previous peak of 919 cases in 2016. &hellip; An estimated 40–80% of DENV infections are asymptomatic. &hellip; ≤5% of all people experiencing symptoms from dengue develop severe, life-threatening disease.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260324150921/https://www.cdc.gov/yellow-book/hcp/travel-associated-infections-diseases/dengue.html","calculation_notes":"CDC Yellow Book is the authoritative traveler-facing guidance. The ~96 million symptomatic infections per year across the ~4 billion people living in dengue-endemic areas implies a roughly 2% annual per-capita symptomatic rate averaged across endemic regions; travelers undersample the high-transmission rural pockets and over-sample urban areas and shorter exposure windows, so the per-trip figure for a two-week stay in a peak-season urban area lands roughly an order of magnitude below that (0.3–1%). CDC’s observation that US travel-associated cases jumped from ~919 in 2016 to &gt;1,400 in 2019 and each year 2022–2024 anchors the “risk is rising, not falling” note in the body text. The 40–80% asymptomatic share is the basis for the ADE personal factor, and the ≤5% severe-dengue share is what takes the ~1-in-200 per-trip infection risk to ~1 in 4,000 per-trip severe-dengue risk.\n","independence_note":"CDC Yellow Book is the primary US traveler-facing clinical guidance, built from CDC dengue surveillance (ArboNET) and the peer-reviewed travel-medicine literature. Shares the Bhatt et al. 2013 modelled 390M/96M estimate with the WHO fact sheet below — treat the CDC/WHO modelled figures as one upstream; GeoSentinel (Duvignaud/Huits) is the genuine independent traveler-case stream.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/dengue-and-severe-dengue","title":"Dengue and severe dengue — Fact sheet","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"100-400 million dengue infections per year globally; modelling estimate of 390 million infections with 96 million symptomatic; 2024 historic high of >14.6 million cases and >12,000 deaths reported to WHO; 308 locally-acquired cases in France, Italy and Spain reported in 2024.","excerpt":"\"About half of the world’s population is now at risk of dengue, with an estimated 100–400 million infections occurring each year. &hellip; One modelling estimate indicates 390 million dengue virus infections per year, of which 96 million manifest clinically. &hellip; During 2024, ongoing transmission, combined with an unexpected spike in dengue cases, resulted in a historic high of over 14.6 million cases and more than 12 000 dengue-related deaths reported. &hellip; Dengue is spreading to new areas, including the European and Eastern Mediterranean regions. In 2024, 308 cases were reported to WHO from three European countries (France, Italy and Spain). &hellip; From January to July 2025, over 4 million cases and over 3000 deaths have been reported to WHO from 97 countries.\"\n","source_date":"2025-08-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413171233/https://www.who.int/news-room/fact-sheets/detail/dengue-and-severe-dengue","calculation_notes":"WHO provides the global denominator and the 2024-2025 trajectory. The &gt;14.6 million reported cases in 2024 — more than triple a normal year — is the main empirical basis for the “per-trip risk in 2024–2025 is higher than a decade ago” framing in the body. WHO case reports undercount the true infection total by roughly an order of magnitude (modelling gives ~96 million symptomatic vs ~14.6 million reported), but the year-over-year growth signal is robust. Independent of CDC: WHO pulls country-level case reports and the modelled Bhatt et al. burden estimate; CDC Yellow Book quotes the same modelled figure but uses its own US surveillance stream for the traveler-case counts.\n","independence_note":"WHO and CDC both cite the Bhatt et al. (2013) modelling estimate for the 390M / 96M figure, so that number is not independent between them. WHO’s 14.6M 2024 reported cases and CDC’s US travel-associated case counts come from different surveillance pipelines and corroborate each other on the 2024 spike.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38951998/","title":"Epidemiology of travel-associated dengue from 2007 to 2022: A GeoSentinel analysis","publisher":"Journal of Travel Medicine (Duvignaud, Stoney, Angelo, et al.)","source_type":"peer_reviewed","statistic":"5,958 confirmed or probable dengue cases in travelers evaluated at GeoSentinel sites 2007-2022; 50.4% acquired in Southeast Asia, 14.9% South Central Asia, 10.9% Caribbean, 9.2% South America; ~2% had complicated dengue; median travel duration 21 days; 67.3% tourism, 12.2% visiting friends and relatives.","excerpt":"\"This analysis included 5,958 travellers with confirmed (n = 4,859; 81.6%) or probable (n = 1,099; 18.4%) dengue. The most frequent regions of acquisition were South East Asia (50.4%), South Central Asia (14.9%), the Caribbean (10.9%) and South America (9.2%). &hellip; In Southeast Asia, annual proportionate morbidity increased from 50 dengue cases per 1000 ill returned travellers in non-epidemic years to an average of 159 cases per 1000 travellers during epidemic years. &hellip; the median travel duration was 21 days &hellip; the most frequent reasons for travel were tourism (67.3%), visiting friends or relatives (12.2%) and business (11.0%).\"\n","source_date":"2024-10-19","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413171315/https://pubmed.ncbi.nlm.nih.gov/38951998/","calculation_notes":"GeoSentinel is the closest thing the travel-medicine literature has to a representative traveler case series. The 50 vs 159 dengue cases per 1,000 ill returned travelers (non-epidemic vs epidemic years) in Southeast Asia is a proportionate morbidity, not an attack rate — it describes what share of sick returning travelers have dengue, not what share of all travelers catch it — but the ~3× swing between epidemic and non-epidemic years is what anchors the regional_breakdown ordering and the seasonal personal factors. Median travel duration of 21 days in this case series matches the two-week reference trip used for the headline figure.\n","independence_note":"GeoSentinel is a global network of travel and tropical medicine clinics with its own case-reporting pipeline, methodologically independent of CDC US surveillance and WHO country-level case reports. Genuine independent corroboration on the regional distribution of traveler dengue (SE Asia dominates, Caribbean and Latin America are next).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37335991/","title":"Clinical Characteristics and Outcomes Among Travelers With Severe Dengue: A GeoSentinel Analysis","publisher":"Annals of Internal Medicine (Huits, Angelo, Leder, et al.)","source_type":"peer_reviewed","statistic":"Among 5,958 dengue cases in travelers 2007-2022, 95 (2%) had complicated dengue, of whom 27 (31%) were classified as severe; 91% hospitalized; one death (non-dengue-related).","excerpt":"\"Of 5958 patients with dengue, 95 (2%) had complicated dengue. &hellip; 27 (31%) were classified as severe. &hellip; Seventy-eight (91%) patients were hospitalized. One patient died of nondengue-related illnesses.\"\n","source_date":"2023-06-20","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250206055418/https://pubmed.ncbi.nlm.nih.gov/37335991/","calculation_notes":"The Huits et al. analysis found 2% of traveler dengue cases had complicated dengue, and 31% of those complicated cases (27/95) met WHO severe-dengue criteria — i.e., 27/5958 ≈ 0.45% of all diagnosed traveler dengue cases were severe. This is much lower than the CDC Yellow Book’s “up to 5%” ceiling for symptomatic cases globally. The gap is consistent with the two populations: GeoSentinel travelers are adults with access to modern supportive care, while the global figure includes endemic-country pediatric second infections, which dominate severe-dengue incidence. The sole death was attributed to non-dengue-related illness. Applied to the headline 1-in-200 per-trip infection risk, the traveler severe-dengue risk lands on the order of 1 in 20,000–40,000 per reference trip, and per-trip mortality is roughly another order of magnitude lower.\n","independence_note":"Same GeoSentinel case pool as the Duvignaud et al. paper, so these two are not independent on the case count. They are cited for different purposes: Duvignaud for the regional distribution and traveler demographics, Huits for the severe-dengue and mortality fractions.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/75/wr/mm7518a1.htm","title":"Increase in Travel-Associated and Locally Acquired Dengue Cases — United States, 2024","publisher":"US Centers for Disease Control and Prevention (MMWR Vol. 75, No. 18)","source_type":"govt_report","statistic":"3,798 dengue cases reported from the 50 US states and DC in 2024 (a 359% increase above the 2010-2023 annual average of 828); 3,693 (97.2%) travel-associated and 105 (2.8%) locally acquired; locally acquired cases reported only from Florida (85, across 10 counties), California (18, three counties), and Texas (2, two counties).","excerpt":"\"During 2024, a total of 3,798 dengue cases, including 2,252 (59.3%) confirmed and 1,546 (40.7%) probable cases, were reported to CDC from the 50 U.S. states and DC. Among these cases, 3,693 (97.2%) were travel associated, and 105 (2.8%) were locally acquired. &hellip; These represent a 359% increase above the annual average of 828 cases per year during 2010–2023. &hellip; Three jurisdictions reported locally acquired cases, including Florida (85 cases reported from 10 counties), California (18 from three counties), and Texas (two from two counties).\"\n","source_date":"2026-05-14","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260616185432/https://www.cdc.gov/mmwr/volumes/75/wr/mm7518a1.htm","calculation_notes":"This MMWR report is the granular 2024 US case tally that anchors the new locally-acquired (autochthonous) caveat. Of 3,798 cases reported from US states/DC in 2024, only 105 (2.8%) were locally acquired — the other 97.2% were travel-associated, i.e. the population this entry's headline already covers. The 105 local cases were confined to three states and a handful of counties (Florida 85/10 counties, California 18/3, Texas 2/2), so the per-resident base rate for catching dengue without leaving the continental US is far below the per-trip traveler figure and is sharply concentrated in Aedes-range counties along the Gulf and southern California. No per-resident denominator is published here and the count is volatile year to year, so the caveat characterizes this risk qualitatively rather than quoting a \"1 in N US residents\" rate.\n","independence_note":"MMWR draws on the same CDC ArboNET surveillance pipeline as the Yellow Book source above — it is NOT an independent stream, just the granular 2024 jurisdiction-level case tally. Cited specifically for the locally-acquired vs travel-associated split, which the Yellow Book and WHO/GeoSentinel sources do not break out for the US.\n"}],"comparison_anchors":[{"label":"Malaria per 2-week Sub-Saharan Africa trip with prophylaxis","lifetime_us_adult":0.0001},{"label":"Japanese encephalitis per short-term urban Asia trip","lifetime_us_adult":5e-7},{"label":"Death in a commercial plane crash (per flight)","lifetime_us_adult":7.3e-8},{"label":"Mosquito-borne disease death (lifetime, global average)","lifetime_us_adult":0.00525}],"regional_breakdown":[{"region":"SE Asia wet season, urban, 2 weeks","probability":0.005,"notes":"Southeast Asia accounts for ~50% of GeoSentinel traveler dengue cases; wet-season urban risk in Thailand, Vietnam, Indonesia, and the Philippines dominates."},{"region":"Caribbean / Latin America peak season, 2 weeks","probability":0.003,"notes":"The Americas reported &gt;13 million dengue cases in 2024, more than triple a normal year. Caribbean and Latin America together account for ~20% of traveler cases in GeoSentinel."},{"region":"Off-peak travel to an endemic area","probability":0.0005,"notes":"Dry-season or cooler-month travel cuts transmission roughly an order of magnitude; GeoSentinel reports ~3× lower proportionate morbidity in non-epidemic years."},{"region":"Cruise ship port call only","probability":0.00001,"notes":"Short shore excursions with limited overnight exposure approximate non-endemic baseline."},{"region":"Long-term expat in an endemic urban area","probability":0.05,"notes":"Year-round exposure in a peak-transmission city can push annual infection odds into the percent-plus range; repeat exposure across serotypes compounds."}],"personal_factor_multipliers":[{"factor":"Prior dengue infection (antibody-dependent enhancement risk)","multiplier":3,"notes":"A second infection with a different serotype raises the odds of severe disease, not the opposite. This is the reason WHO and CDC guidance treat prior dengue as a risk factor for severe dengue, not a protective factor."},{"factor":"DEET or picaridin + permethrin-treated clothing + screened or air-conditioned accommodation","multiplier":0.3,"notes":"Aedes aegypti and Aedes albopictus are daytime biters concentrated in and around buildings; bite avoidance during daylight hours materially reduces exposure."},{"factor":"Long-term expat in an endemic urban area","multiplier":20,"notes":"Duration of exposure is the dominant factor. A year in a peak-transmission city can approach local resident attack rates; two years can cycle through multiple serotypes."},{"factor":"Wet-season travel vs dry-season travel to the same destination","multiplier":3,"notes":"GeoSentinel: proportionate morbidity for dengue tripled in Southeast Asia between non-epidemic and epidemic years; wet-season and post-monsoon travel carries most of the seasonal risk."},{"factor":"Visiting friends and relatives (VFR) vs resort tourism","multiplier":2,"notes":"Longer stays, more time in residential neighborhoods, less screened accommodation. 12% of GeoSentinel traveler dengue cases were VFR travelers."}],"short_label":"Dengue (travel)","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline ~1-in-200 per-trip figure is an order-of-magnitude estimate for a two-week leisure stay in a peak-season endemic urban area, not a precise rate. True attack rates vary by country, city, neighborhood, accommodation type, month, and prevailing serotype, and published traveler serostudies span roughly 0.2–2% per two-week trip depending on destination. An estimated 40–80% of dengue infections are asymptomatic, so the true infection rate is substantially higher than the symptomatic attack rate represented here; asymptomatic infections still carry forward the antibody-dependent enhancement (ADE) risk profile for any future trip, which is the main reason a “I had it once and it was nothing” traveler should not treat that as reassurance. Severe dengue (dengue hemorrhagic fever / dengue shock syndrome) occurs in up to 5% of symptomatic cases per CDC Yellow Book, with a lower ~0.45% figure specifically in GeoSentinel travelers who have access to modern supportive care (Huits et al. 2023). This entry measures the probability of contracting dengue, not of dying from it: per-trip mortality is roughly two orders of magnitude lower than the per-trip infection risk, and dominated by diagnostic delay in non-endemic emergency rooms. Returning travelers with fever within two weeks of an endemic trip should be evaluated for dengue alongside malaria. This entry deliberately measures travel-acquired dengue; locally-acquired (autochthonous) dengue caught inside the continental US is a separate and much smaller risk. Of 3,798 dengue cases reported from US states and DC in 2024, only 105 (2.8%) were locally acquired — the rest were travel-associated — and those 105 were confined to Florida (85 across 10 counties), California (18 across three counties), and Texas (2 across two counties) per CDC MMWR Vol. 75 No. 18 (2026). The per-resident base rate for catching dengue without leaving the country is therefore far below the per-trip traveler figure and sharply concentrated in Aedes-range counties along the Gulf Coast and in southern California; no stable per-resident rate is quoted here because the local count is small, geographically concentrated, and volatile year to year as Aedes range expands. The sibling entry &grave;malaria-travel&grave; covers malaria specifically and &grave;mosquito-borne-disease&grave; covers the aggregate mortality figure across all mosquito-borne diseases globally.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized hibiscus flower resting on a pale neutral background, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/dengue-travel","api_url":"https://likelier.app/api/fears/dengue-travel.json"},{"slug":"heat-exposure-preterm-birth","question":"What are the odds that an extreme-heat exposure during pregnancy triggers a preterm birth?","quick_answer":"For a US pregnancy that hits a local heat wave in the final weeks, the heat-attributable excess chance of a preterm birth is about 1 in 200 (roughly 5 in 1,000) above the normal baseline. Most people don't count heat as a pregnancy risk at all, while climate coverage tends to overweight the large relative-risk headline against this small absolute per-pregnancy effect.\n","category":"health","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Extreme heat is rarely on the standard list of pregnancy worries. Prenatal education foregrounds preeclampsia, gestational diabetes, miscarriage, and birth defects; ambient temperature is treated as a comfort issue rather than an obstetric one. When climate-and-pregnancy coverage does surface, it tends to pair a large relative-risk headline (preterm odds up 26% during heat waves) with photographs of distressed mothers, which inflates the felt severity well past the per-pregnancy absolute risk. The typical expectant parent in a temperate, air-conditioned US setting would struggle to say whether a hot week before delivery changes their odds at all, and most would guess it does not.","rough_estimate":"Most people do not list heat among pregnancy risks at all; those who have seen the climate-health coverage tend to overweight the relative-risk headline and underweight the small absolute per-pregnancy effect in cooled settings","kind":"intuition"},"native":{"display":"~5 in 1,000 exposed pregnancies (heat-attributable excess)","numerator":5,"denominator":1000,"unit":"per heat-exposed pregnancy","population":"US singleton pregnancies exposed to a local heat wave in the days before delivery"},"normalized":{"lifetime_us_adult":0.005,"display":"~1 in 200 per heat-exposed pregnancy (US, excess over baseline)","log_value":-2.301,"assumptions":"Subgroup: US singleton pregnancies that experience a local extreme-heat episode (4+ consecutive days above the local 97.5th-percentile temperature, or a multi-day heat wave) in the critical late-gestation window. Horizon: per-pregnancy, not per-lifetime. The headline is the HEAT-ATTRIBUTABLE EXCESS probability of a preterm birth, i.e. p(preterm | heat-exposed) minus the all-cause baseline, not the total preterm risk. Baseline US preterm rate = 10.41% (CDC NCHS, 2024). US population effect (JAMA Network Open 2024, 53.2M births, 50 metros): a heat wave raises preterm-birth rate by a rate ratio of ~1.02, so excess = 0.1041 x 0.02 = 0.0021 per exposed pregnancy; the vulnerable subgroup (under 30, high-school education or less, racial/ethnic minority) showed RR ~1.04, giving excess = 0.1041 x 0.04 = 0.0042. The global meta-analysis (Nature Medicine 2024) reports a heat-wave ODDS ratio of 1.26 for preterm birth; converted to an absolute risk at a 10.41% baseline (odds_base = 0.1041/0.8959 = 0.1162; x1.26 = 0.1464; p_exposed = 0.1277), the implied excess is ~0.024 per exposed pregnancy. Note OR 1.26 overstates the risk ratio (~1.23) at this baseline. The US figure and the global figure differ by roughly an order of magnitude, driven mainly by air-conditioning penetration and climate mix; the uncertainty band spans both. Point estimate 0.005 sits just above the US-average excess to reflect vulnerable subgroups, intensifying US heat waves, and longer/hotter episodes than the threshold-crossing definition. Stillbirth is NOT in the headline: the global-meta stillbirth OR is 1.13 with a 95% CI of 0.95-1.34, which crosses the null.","uncertainty":{"low":0.002,"high":0.024},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11835737/","title":"A systematic review and meta-analysis of heat exposure impacts on maternal, fetal and neonatal health","publisher":"Nature Medicine","source_type":"peer_reviewed","statistic":"Preterm birth odds rose 4% per 1°C (OR 1.04, 95% CI 1.03-1.06) and 26% during heat waves (OR 1.26, 95% CI 1.08-1.47); stillbirth OR 1.13 (95% CI 0.95-1.34); 198 studies across 66 countries","excerpt":"the odds of a preterm birth increased by 4% (OR = 1.04 (95% CI = 1.03, 1.06; I2 = 85%; n = 12))... the odds of preterm birth increased by 26% in our meta-analysis (OR = 1.26 (95% CI = 1.08, 1.47; I2 = 63%; n = 10))... OR = 1.13 (95% CI = 0.95, 1.34; I2 = 83%; n = 9)","source_date":"2024-11-05","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260705145034/https://pmc.ncbi.nlm.nih.gov/articles/PMC11835737/","calculation_notes":"The headline relative effects come from this meta-analysis: OR 1.04 per 1°C and OR 1.26 during heat waves for preterm birth, across 198 studies in 66 countries (63% high-income, 40% temperate). Converting the heat-wave OR to an absolute excess at a US baseline preterm rate of 10.41%: odds_base = 0.1041/0.8959 = 0.1162; multiplied by 1.26 = 0.1464; back to probability = 0.1277; excess over baseline = 0.0236. The OR (1.26) overstates the true risk ratio (~1.23) because the baseline is not rare. The stillbirth OR of 1.13 has a 95% CI that crosses 1.0 (0.95-1.34), so stillbirth is reported as suggestive, not as the headline outcome. This is the high end of the uncertainty band.","independence_note":"Global meta-analysis pooling 198 primary studies; methodologically distinct from the single-country JAMA cohort below."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11127119/","title":"Preterm and Early-Term Delivery After Heat Waves in 50 US Metropolitan Areas","publisher":"JAMA Network Open","source_type":"primary_study","statistic":"Among 53,154,816 US births (1993-2017), a 4-day heat wave raised preterm-birth rate ratio to 1.02 (95% CI 1.00-1.03); the vulnerable subgroup rate ratio was 1.04 (95% CI 1.02-1.06)","excerpt":"There were 53 154 816 births available after exclusions... rate ratio was 1.02 (95% CI, 1.00-1.03)... For preterm birth and early-term birth among those who were 29 years of age or younger, had a high school education or less, and belonged to a racial or ethnic minority group... 1.04 (95% CI, 1.02-1.06)","source_date":"2024-05-24","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250301141836/https://pmc.ncbi.nlm.nih.gov/articles/PMC11127119/","calculation_notes":"US population anchor for the absolute conversion. Baseline US preterm rate = 10.41% (CDC NCHS, 2024). Full-population heat-wave RR 1.02 gives excess = 0.1041 x 0.02 = 0.0021 per exposed pregnancy; vulnerable subgroup RR 1.04 gives 0.1041 x 0.04 = 0.0042. These bracket the chosen point estimate of 0.005 and define the low end of the uncertainty band (0.002). Heat wave defined as 4+ consecutive days above the local 97.5th-percentile temperature in the days before birth.","independence_note":"Single-country (US) vital-records cohort; independent of the global meta-analysis and uses an absolute rate-ratio design rather than pooled odds ratios."},{"url":"https://cdc.gov/nchs/pressroom/releases/20250423.html","title":"U.S. Births Increase by 1% in 2024","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"US preterm birth rate was 10.41% in 2024 (births before 37 weeks)","excerpt":"The preterm birth rate was unchanged at 10.41%.","source_date":"2025-04-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260618052006/https://cdc.gov/nchs/pressroom/releases/20250423.html","calculation_notes":"Supplies the all-cause US preterm baseline (0.1041) used to convert every relative effect into an absolute heat-attributable excess. Without this denominator the meta-analysis ORs cannot be turned into a per-pregnancy probability.","independence_note":"Official US vital-statistics surveillance; independent of the epidemiological studies."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7303808/","title":"Association of Air Pollution and Heat Exposure With Preterm Birth, Low Birth Weight, and Stillbirth in the US: A Systematic Review","publisher":"JAMA Network Open","source_type":"peer_reviewed","statistic":"9 of 10 (90%) studies found a significant association between heat exposure in pregnancy and adverse birth outcomes; higher heat-related preterm risk found among Black, Asian, and younger mothers in California","excerpt":"higher risk was found among black mothers... Increased risk of preterm birth was also found for Asian mothers and younger mothers... Two reports from California found an association of racial/ethnic disparity and heat exposure with an increasing risk of preterm birth.","source_date":"2020-06-18","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260701025706/https://pmc.ncbi.nlm.nih.gov/articles/PMC7303808","calculation_notes":"Supports the personal-factor multipliers: Black, Asian, and younger US mothers carry higher heat-related preterm risk. The 90%-of-studies-significant figure underpins the prose statement that the relative effect is robust even where the absolute magnitude is uncertain. Named peer-reviewed source replacing the advocacy summary for the racial-disparity multiplier.","independence_note":"US-focused systematic review (Bekkar et al.); distinct authorship and study set from the global Nature Medicine meta-analysis."},{"url":"https://nationalpartnership.org/report/higher-temperatures-hurt-moms/","title":"Higher Temperatures Hurt Moms and Babies","publisher":"National Partnership for Women & Families","source_type":"reputable_reference","statistic":"In four US studies, Black or Hispanic women with heat exposure had double or more the risk of preterm birth or stillbirth versus white women; preterm births rose ~16% during heat waves in pooled results","excerpt":"A systematic review found that in four U.S. studies, compared to white women, Black or Hispanic women with heat exposure had double or more the risk of preterm birth or stillbirth.","source_date":"2021-05-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260424024514/https://nationalpartnership.org/report/higher-temperatures-hurt-moms/","calculation_notes":"Used only to corroborate the ~2x Black/Hispanic disparity multiplier; this is an advocacy summary citing an underlying systematic review, so the peer-reviewed Bekkar et al. (2020) source above is the primary basis for that multiplier. Not authoritative on its own.","independence_note":"Secondary advocacy synthesis; not counted toward the authoritative-source requirement."}],"comparison_anchors":[{"label":"Drowning death (lifetime, US)","lifetime_us_adult":0.000725},{"label":"Child pool drowning (lifetime, US)","lifetime_us_adult":0.000435},{"label":"Vehicle theft (US)","lifetime_us_adult":0.0296}],"regional_breakdown":[{"region":"US (air-conditioned, temperate)","probability":0.003,"notes":"JAMA 2024, 53M births: heat-wave preterm rate ratio ~1.02 over a 10.4% baseline. High AC penetration blunts the effect; vulnerable subgroups (younger, lower education, minority) run roughly double."},{"region":"Vulnerable US subgroup","probability":0.005,"notes":"JAMA 2024 subgroup (age <30, high-school education or less, racial/ethnic minority): RR ~1.04. Bekkar et al. 2020 found Black, Asian, and younger California mothers at higher heat-related preterm risk."},{"region":"Global / hot-climate average","probability":0.024,"notes":"Nature Medicine 2024 meta (66 countries): heat-wave preterm OR 1.26, converted to ~0.024 absolute excess at a 10.4% baseline. Reflects lower AC access and higher heat intensity."},{"region":"Sub-Saharan Africa (perinatal mortality)","probability":0.014,"notes":"Different outcome: a 16-hospital cohort (138,015 births, Nature Medicine 2024) found 34% higher perinatal-mortality odds (OR 1.34) when extreme heat preceded delivery, against a ~4.0% baseline perinatal-death rate. Excess ~0.014. This is mortality, not preterm birth."}],"personal_factor_multipliers":[{"factor":"Black or Hispanic mother (US)","multiplier":2,"notes":"Bekkar et al. (JAMA Network Open 2020) and a systematic review cited by the National Partnership for Women & Families found Black or Hispanic women with heat exposure had double or more the risk of preterm birth or stillbirth versus white women, attributed partly to lower in-home AC access."},{"factor":"Vulnerable profile (age <30, high-school education or less, minority)","multiplier":2,"notes":"JAMA Network Open 2024 (53M-birth cohort): heat-wave preterm rate ratio ~1.04 in this subgroup versus ~1.02 in the full population, roughly doubling the attributable excess."},{"factor":"No / limited home air conditioning","multiplier":2,"notes":"The US-vs-global gap in the heat-preterm effect tracks AC penetration; cooled environments blunt exposure. Directional magnitude is inferred from the regional contrast rather than a single head-to-head RCT, so treat as approximate."}],"short_label":"Heat-triggered preterm birth","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"This entry reports the heat-ATTRIBUTABLE excess probability of preterm birth, not the total preterm risk (the US all-cause baseline is 10.4% per pregnancy). The point estimate (~0.5% per heat-exposed pregnancy) and the unusually wide band (0.2% to 2.4%) reflect a genuine order-of-magnitude disagreement between US data and the global meta-analysis: the US figure (JAMA 2024, heat-wave RR ~1.02) is roughly ten times smaller than the global meta figure (OR 1.26), driven mainly by air-conditioning penetration, heat intensity, and exposure definitions. There is no single defensible point estimate that survives every exposure definition, so the band is the honest output, not the midpoint. Three further limitations: (1) the relative effects are odds ratios at a non-rare baseline, so they slightly overstate true risk ratios; the conversion in assumptions corrects for this. (2) Stillbirth is excluded from the headline because the global-meta stillbirth OR (1.13) has a confidence interval crossing the null; the significant perinatal-mortality signal (OR 1.34) comes only from a sub-Saharan-African hospital cohort and is shown in the regional breakdown. (3) Preterm birth is a heterogeneous outcome ranging from late-preterm (34-36 weeks, usually mild) to extremely preterm; the studies pool all of it, so the outcome severity here is an average. The 8D self-score clears all thresholds (avg 4.0).","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Heat-triggered preterm birth\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/heat-exposure-preterm-birth","api_url":"https://likelier.app/api/fears/heat-exposure-preterm-birth.json"},{"slug":"infrequent-showering","question":"What are the odds of getting sick from not showering daily?","quick_answer":"Showering every two to three days instead of daily adds only about a 0.5% incremental lifetime risk of a significant skin infection — within measurement noise, against a background where 7 to 10% of US adults get a skin infection in any year regardless of bathing habits. People often imagine a 20 to 40% chance from skipping showers.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The daily shower is treated as a non-negotiable hygiene baseline in most Western cultures. Skipping a day provokes visceral unease -- people assume bacterial counts spike, skin infections follow, and illness is only a missed scrub away. Social media debates about shower frequency reliably produce strong reactions, with many insisting that anything less than daily is unsanitary and dangerous.\n","rough_estimate":"~20-40% chance of developing a skin infection if you shower every other day","kind":"intuition"},"native":{"display":"~7-10% of US adults experience a skin infection (bacterial or fungal) in any given year, regardless of shower frequency","numerator":85,"denominator":1000,"unit":"annual skin infection incidence per 1,000 US adults","population":"US adults, all bathing frequencies"},"normalized":{"lifetime_us_adult":0.005,"display":"~0.5% incremental lifetime risk of a clinically significant skin infection attributable to showering every 2-3 days instead of daily","log_value":-2.3,"assumptions":"The CDC's Larson (2001) review in Emerging Infectious Diseases found that bacterial counts on skin are at least as high after showering with regular soap as before, and that skin flora remain qualitatively and quantitatively stable even without bathing for several days. Harvard Health notes that showering several times a week, rather than daily, is plenty for most adults. No controlled trial has demonstrated increased infection rates from showering every other day vs daily in healthy adults. The AAD notes that over-washing strips protective lipids and disrupts the skin barrier, potentially increasing susceptibility to eczema and secondary infection. We estimate the incremental infection risk from showering every 2-3 days (vs daily) at effectively negligible -- conservatively modeled as ~0.5% additional lifetime risk, well within measurement noise. The 2025 Eczema Bathing RCT (British Journal of Dermatology) found no difference in outcomes between weekly and daily bathers, further supporting minimal clinical impact.\n","uncertainty":{"low":0.001,"high":0.02},"scope":"us_adult_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/eid/article/7/2/70-0225_article","title":"Hygiene of the Skin: When Is Clean Too Clean?","publisher":"Emerging Infectious Diseases, CDC (Larson, 2001)","source_type":"peer_reviewed","statistic":"Bacterial counts on skin are at least as high or higher after bathing with regular soap than before; skin flora remain stable even without bathing for days","excerpt":"\"Bathing or showering with regular soap has aesthetic and stress-relieving benefits but serves little microbiologic purpose... the flora remain qualitatively and quantitatively stable.\"\n","source_date":"2001-04-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426202231/https://wwwnc.cdc.gov/eid/article/7/2/70-0225_article","calculation_notes":"Larson's CDC review established that routine bathing with non-antimicrobial soap does not meaningfully reduce resident skin flora. This means skipping a day of showering does not produce a measurably different microbial load compared to daily bathing. The review covers multiple studies on hand and body washing, concluding that personal hygiene primarily reduces transient organisms, not stable resident communities.\n"},{"url":"https://www.uclahealth.org/news/article/skin-microbiome-disrupted-with-too-frequent-bathing","title":"Skin microbiome disrupted with too-frequent bathing","publisher":"UCLA Health","source_type":"reputable_reference","statistic":"Over-washing and the abrasives, soaps and chemicals routinely used in bathing strip beneficial microbial colonies from the skin and can adversely affect the skin microbiome's natural balance","excerpt":"\"The overuse of abrasives can literally strip beneficial microbial colonies from the skin. The soaps, chemicals and abrasives routinely used in bathing can have a direct and immediate effect on the skin microbiome, and perhaps have an adverse effect on its natural balance. Don't over-wash, which removes oils needed to feed your microbe friends.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260120174813/https://www.uclahealth.org/news/article/skin-microbiome-disrupted-with-too-frequent-bathing","calculation_notes":"UCLA Health reports that over-washing strips beneficial microbial colonies and removes the oils that feed them, adversely affecting the skin microbiome's natural balance. The article does not quantify infection risk or recommend a specific shower frequency; it is cited here for the qualitative point that frequent washing disrupts the microbiome, supporting the position that showering every 2-3 days is not a risk factor for skin barrier integrity.\n"},{"url":"https://doi.org/10.1093/bjd/ljaf417","title":"Weekly versus daily bathing for people with eczema: results of the Eczema Bathing online randomized controlled trial","publisher":"British Journal of Dermatology (2025), doi:10.1093/bjd/ljaf417","source_type":"peer_reviewed","statistic":"No difference in eczema symptoms between weekly and daily bathers in an RCT (n=438): adjusted POEM difference -0.4 (95% CI -1.3 to 0.4; P=0.30); itch, eczema control and quality of life were also similar","excerpt":"\"The trial found no evidence of a difference in eczema symptoms between bathing or showering weekly vs. daily. The adjusted difference in mean POEM score over 4 weeks for weekly vs. daily bathing was -0.4 (95% confidence interval -1.3 to 0.4; P = 0.30). These results are helpful for people with eczema, giving them the freedom to choose what suits them best.\"\n","source_date":"2025-11-10","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260416185612/https://academic.oup.com/bjd/article/194/3/450/8315016","calculation_notes":"This RCT directly tested the hypothesis that bathing frequency affects skin health outcomes. Even in the eczema population (who have compromised skin barriers), weekly bathers fared no worse than daily bathers. For healthy adults, the implication is that showering every 2-3 days poses no measurable infection risk increase.\n"},{"url":"https://www.health.harvard.edu/blog/showering-daily-is-it-necessary-2019062617193","title":"Showering daily -- is it necessary?","publisher":"Harvard Health Publishing, Harvard Medical School (Shmerling RH)","source_type":"reputable_reference","statistic":"Experts suggest showering several times a week, not necessarily daily, is plenty for most people","excerpt":"\"While there is no ideal frequency, experts suggest that showering several times per week is plenty for most people (unless you are grimy, sweaty, or have other reasons to shower more often).\"\n","source_date":"2019-06-26","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260628153925/https://www.health.harvard.edu/blog/showering-daily-is-it-necessary-2019062617193","calculation_notes":"Harvard Health Publishing, written by a Harvard Medical School faculty physician, states that showering several times a week rather than daily is sufficient for most people. The article does not specify an exact figure such as \"two to three times per week\" and does not name Yale or UCLA; it is cited here for the qualitative point that mainstream medical guidance does not require daily showering, replacing an earlier compound attribution to \"Harvard, Yale, and UCLA\" dermatologists that none of this entry's sources actually supported.\n"}],"comparison_anchors":[{"label":"Developing eczema (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Staph skin infection requiring antibiotics (lifetime, US)","lifetime_us_adult":0.15},{"label":"Hospitalization from any cause (annual, US)","lifetime_us_adult":0.07}],"personal_factor_multipliers":[{"factor":"Occupational contamination exposure","multiplier":3.5,"notes":"Manual laborers, healthcare workers, and food-service staff accumulate pathogens, soil, and chemical residues on skin at rates far exceeding office settings. OSHA and CDC guidance for these sectors includes post-shift showering as a specific infection-control measure; the combination of higher transient-flora load and broken skin from physical work roughly triples incremental infection risk compared to sedentary adults."},{"factor":"Active acne or atopic dermatitis","multiplier":2.5,"notes":"Compromised skin barrier from active acne or eczema allows resident and transient flora (including Staph aureus colonization in >90% of eczema patients per Geoghegan et al., Trends in Microbiology 2018) to penetrate more readily. The AAD and UCLA Health both note that impaired barrier function is the primary driver of secondary bacterial skin infections, making infrequent showering a more consequential variable in this subgroup."},{"factor":"Gym or pool use without post-activity shower","multiplier":2,"notes":"Sweat-saturated skin in occluded areas (groin, underarms, skin folds) creates an environment permissive to dermatophyte and bacterial overgrowth. CDC's athlete MRSA guidance and tinea guidance both list showering promptly after athletic activity as a first-line prevention step. Gym-goers who skip post-workout showering also contact communal surfaces carrying S. aureus and tinea fungi at higher rates (Hedderwick et al., Lancet Infectious Diseases background data)."},{"factor":"Immunocompromised status","multiplier":5,"notes":"Transplant recipients, patients on chemotherapy or long-term corticosteroids, and adults with advanced HIV have profoundly reduced capacity to clear transient skin pathogens. CDC infection-control guidance for immunocompromised patients specifically includes more frequent washing as an adjunct measure because organisms that resident immunity would clear without symptoms can cause invasive infection in this population."}],"short_label":"Skipping daily showers","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"The estimate applies to healthy adults in temperate climates with access to clean clothing and basic hygiene. People with immunocompromising conditions, open wounds, occupational exposure to contaminants, or those living in hot humid climates may have different risk profiles. The evidence base lacks large RCTs specifically measuring infection rates by shower frequency in healthy adults -- the negligible risk estimate is derived from microbiome stability data and dermatological consensus rather than direct trial evidence.\n","quality_score":{"d1":4,"d2":4,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":4,"avg":3.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A shower head with a single water droplet hanging from it, flat vector illustration in muted blue-grey tones."},"canonical_url":"https://likelier.app/infrequent-showering","api_url":"https://likelier.app/api/fears/infrequent-showering.json"},{"slug":"not-washing-feet-shower","question":"What are the odds of getting a foot infection from not washing your feet in the shower?","quick_answer":"The incremental lifetime risk of a foot infection attributable specifically to not scrubbing your feet with soap is only about 0.5% — and even that is an order-of-magnitude guess, since no study isolates shower technique. Athlete's foot is common (15-25% of US adults at any time), but the real drivers are damp footwear, shared wet floors, and drying between the toes, not soap on the feet.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The internet debate over whether you need to actively wash your feet in the shower -- or whether the soapy water running down your legs suffices -- became a viral cultural flashpoint. Many people assume that failing to scrub between each toe with soap invites athlete's foot and bacterial infections. The strength of feeling on both sides suggests most participants believe the stakes are meaningfully high.\n","rough_estimate":"~30-50% chance of foot infection if you don't scrub your feet with soap daily","kind":"intuition"},"native":{"display":"~15-25% point prevalence of tinea pedis among US adults","numerator":15,"denominator":100,"unit":"point prevalence of athlete's foot among US adults","population":"US adults"},"normalized":{"lifetime_us_adult":0.005,"display":"~0.5% incremental lifetime risk of a foot infection attributable to not actively scrubbing feet with soap in the shower","log_value":-2.3,"assumptions":"Tinea pedis has a lifetime risk of up to 70% (StatPearls, PMC review). Point prevalence among US adults runs 15-25% at any given time. However, the primary risk factors identified in clinical literature are moisture retention, occlusive footwear, shared contaminated surfaces (locker rooms, pools), diabetes, and male sex -- not whether soap is applied directly to the feet during showering. No study has isolated \"active foot scrubbing in the shower\" vs \"passive rinse-off from soapy water\" as an independent predictor of tinea pedis. The AAD recommends drying feet thoroughly (especially between toes) and wearing breathable shoes as the main prevention strategies. The marginal contribution of foot-scrubbing to dermatophyte infection is unquantified in the literature. With a 70% lifetime prevalence (Havlickova 2008), the dominant risk factors are occlusive footwear duration, shared wet surfaces, and genetic susceptibility -- not shower technique. We conservatively estimate the incremental risk of omitting deliberate foot scrubbing at ~0.5%, acknowledging this is an order-of-magnitude estimate rather than a measured value. The key variable is post-shower drying and footwear choice, not soap application technique.\n","uncertainty":{"low":0.001,"high":0.02},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10321471/","title":"Tinea pedis: an updated review","publisher":"Journal of Fungi / PMC","source_type":"peer_reviewed","statistic":"Lifetime risk of tinea pedis is up to 70%; point prevalence ~3-15% globally; risk factors include occlusive footwear, moisture, shared surfaces, male sex, and diabetes","excerpt":"\"The lifetime risk is up to 70%... The prevalence is higher in adolescents and adults than in prepubertal children. The male to female ratio is approximately 3:1.\"\n","source_date":"2023-06-28","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420044706/https://pmc.ncbi.nlm.nih.gov/articles/PMC10321471/","calculation_notes":"This comprehensive review establishes the epidemiology and risk factors for tinea pedis. The identified risk factors -- occlusive footwear, prolonged moisture, shared contaminated surfaces, immunosuppression -- are environmental and host factors. No mention of soap application technique during showering as an independent predictor. The 70% lifetime prevalence means most adults will get athlete's foot regardless of foot-washing habits.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK470421/","title":"Tinea Pedis","publisher":"StatPearls / NCBI Bookshelf","source_type":"reputable_reference","statistic":"More than 70% of the US population will have tinea pedis at some point; 3-15% current prevalence; independent risk factors include advanced age, male sex, diabetes, and lower-limb ischemia","excerpt":"\"More than 70% of the population will be infected with tinea pedis at some time during their lives... Independent risk factors for the development of tinea pedis included advanced age, male sex, diabetes, and lower-limb ischemia.\"\n","source_date":"2023-08-08","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260409233543/https://www.ncbi.nlm.nih.gov/books/NBK470421/","calculation_notes":"StatPearls confirms the high lifetime prevalence and identifies clinical risk factors. The risk factor list -- age, sex, diabetes, vascular disease -- does not include frequency or technique of foot washing. This supports the position that active scrubbing vs passive rinse is not a clinically meaningful variable in tinea pedis epidemiology.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6900014/","title":"Internal environment of footwear is a risk factor for tinea pedis","publisher":"Journal of Foot and Ankle Research / PMC","source_type":"peer_reviewed","statistic":"High temperature, high humidity, and high dew point inside footwear were significantly associated with higher incidence of tinea pedis","excerpt":"\"Those who wore footwear with internal environments characterized by high temperature, high humidity, high-temperature/high-humidity and high dew point values had a significantly higher incidence of tinea pedis.\"\n","source_date":"2019-12-03","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250815003922/https://pmc.ncbi.nlm.nih.gov/articles/PMC6900014/","calculation_notes":"This study directly measured the footwear microenvironment as a risk factor, finding that shoe moisture and temperature -- not foot washing technique -- predict tinea pedis development. The practical implication is that drying feet and choosing breathable footwear matters far more than how vigorously one scrubs in the shower.\n"}],"comparison_anchors":[{"label":"Getting athlete's foot at some point (lifetime, US)","lifetime_us_adult":0.7},{"label":"Toenail fungus (lifetime, US adult)","lifetime_us_adult":0.14},{"label":"Skin infection requiring antibiotics (annual, US)","lifetime_us_adult":0.02}],"personal_factor_multipliers":[{"factor":"Diabetes mellitus","multiplier":4,"notes":"ADA Standards of Medical Care in Diabetes (2023) and StatPearls (Tinea Pedis, 2023) identify diabetes as an independent risk factor for tinea pedis; peripheral neuropathy reduces sensation of early lesions and reduced circulation impairs local immune response, compounding both acquisition and progression risk. Diabetic foot infection is a leading cause of hospitalization and amputation."},{"factor":"Regular barefoot use in communal wet areas (gym, pool, locker room)","multiplier":3.5,"notes":"AAFP clinical review and StatPearls (Tinea Pedis, 2023) identify shared wet surfaces as the primary environmental transmission route for dermatophytes. Fungal spore burden on gym shower floors and pool decks is substantially higher than household floors, with multiple community studies documenting dermatophyte prevalence on shared surfaces."},{"factor":"Prior tinea pedis episode","multiplier":3,"notes":"Clinical recurrence rates for tinea pedis are high; StatPearls and the Journal of Fungi review (2023) note that dermatophytes persist in the stratum corneum and nail apparatus, with re-infection from footwear and household surfaces common. Recurrence risk after a treated episode is estimated at 25–40% within 12 months per dermatology clinical guidelines."},{"factor":"Male sex","multiplier":3,"notes":"StatPearls and the Journal of Fungi review (PMC, 2023) report a male-to-female ratio of approximately 3:1 for tinea pedis, attributed to higher rates of occlusive footwear use, sports participation, and communal locker room exposure. This ratio is consistent across multiple global epidemiological surveys."}],"short_label":"Not scrubbing feet","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The estimate addresses the marginal risk of passive rinse vs active scrub in the shower for otherwise healthy adults. People who walk barefoot in communal wet areas (gyms, pools, dormitory showers), have diabetes, peripheral vascular disease, or immunosuppression face materially higher baseline foot infection risk regardless of shower technique. Humid climates increase baseline fungal colonization rates, and occupational factors -- prolonged occlusive footwear in military service, mining, construction, or athletic training -- raise risk independently of hygiene routine. Note also the distinction between dermatophyte colonization (common, often asymptomatic) and clinical infection requiring treatment (less common); the 70% lifetime figure captures both. The question is whether scrubbing modifies that trajectory, and the evidence says the answer is footwear and drying, not soap.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A bare foot on a wet tile surface with small water droplets, flat vector illustration in muted teal tones."},"canonical_url":"https://likelier.app/not-washing-feet-shower","api_url":"https://likelier.app/api/fears/not-washing-feet-shower.json"},{"slug":"mosquito-borne-disease","question":"What are the odds of dying from a mosquito-borne disease?","quick_answer":"For the average global adult, the lifetime odds of dying from a mosquito-borne disease are about 1 in 190, driven by roughly 700,000 vector-borne deaths worldwide each year — overwhelmingly malaria. Outside endemic regions the near-zero prior most readers hold is roughly correct; the risk is heavily concentrated geographically.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"The mosquito occupies an odd slot in the risk imagination: everyone knows it is annoying, almost nobody files it alongside sharks, bears, or snakes on a list of dangerous animals. We haven’t yet found a rigorous recent survey that isolates \"fear of dying from a mosquito-borne disease\" from the broader category of health anxiety or travel-disease worry, so the perceived side of this page is marked as editorial intuition rather than polled data. The working prior for most Likelier readers is essentially zero — which, outside of endemic regions, happens to be roughly correct.\n","rough_estimate":"most readers in non-endemic countries would guess ~zero","kind":"intuition"},"native":{"display":"~700,000 deaths per year from vector-borne diseases worldwide (mosquito-borne dominate)","numerator":700000,"denominator":8000000000,"unit":"per year","population":"global"},"normalized":{"lifetime_us_adult":0.00525,"display":"~1 in 190 lifetime (global adult average)","log_value":-2.28,"assumptions":"Uses the WHO vector-borne diseases fact sheet figure of &gt;700,000 deaths per year globally, the bulk of which are mosquito-borne (malaria ~600,000, dengue ~40,000, plus Japanese encephalitis, yellow fever, West Nile, chikungunya, and Zika). Annual per-capita hazard ≈ 700,000 / 8,000,000,000 ≈ 8.75 × 10&#8315;&#8309;; compounded across 60 adult years gives 1 - (1 - 8.75e-5)^60 ≈ 5.25 × 10&#8315;&#179;, or about 1 in 190 for the average global adult. This number is a scale marker, not a personal forecast — see the regional breakdown and the body text. The uncertainty band reflects whether you count only WHO-attributed malaria deaths (lower bound) or include the full vector-borne aggregate plus the IHME-style higher malaria estimates (upper bound).\n","uncertainty":{"low":0.0035,"high":0.008},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/vector-borne-diseases","title":"Vector-borne diseases — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Vector-borne diseases account for more than 17% of all infectious diseases, causing more than 700,000 deaths annually; malaria causes more than 608,000 deaths per year and dengue an estimated 40,000.","excerpt":"\"Vector-borne diseases account for more than 17% of all infectious diseases, causing more than 700 000 deaths annually. &hellip; It causes an estimated 249 million cases globally, and results in more than 608 000 deaths every year. &hellip; More than 3.9 billion people in over 132 countries are at risk of contracting dengue, with an estimated 96 million symptomatic cases and an estimated 40 000 deaths every year.\"\n","source_date":"2024-03-02","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180337/https://www.who.int/news-room/fact-sheets/detail/vector-borne-diseases","calculation_notes":"WHO’s &gt;700,000/year vector-borne total is the aggregate anchor for this entry. Mosquito-borne diseases (malaria + dengue + Japanese encephalitis + yellow fever + West Nile + chikungunya + Zika) dominate that total, with malaria alone at ~608,000/year. Dividing 700,000 by a global population of 8 billion gives an annual per-capita hazard of ~8.75 × 10&#8315;&#8309;; compounded over 60 adult years yields ~5.25 × 10&#8315;&#179;, or about 1 in 190.\n","independence_note":"WHO vector-borne fact sheet aggregates WHO programmatic estimates across malaria, dengue, and the smaller arboviruses. Shares upstream with the WHO malaria fact sheet and the World Malaria Report below — treat the three WHO sources as a single institutional estimate rather than independent counts; the CDC US surveillance figure below is the independent cross-check.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/malaria","title":"Malaria — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"610,000 estimated malaria deaths in 2024; ~95% of all malaria cases and deaths occurred in the WHO African Region.","excerpt":"\"The estimated number of malaria deaths stood at 610 000 in 2024 compared to 598 000 in 2023. &hellip; In 2024 the Region was home to about 95% of all malaria cases and deaths.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180415/https://www.who.int/news-room/fact-sheets/detail/malaria","calculation_notes":"WHO’s latest malaria estimate (610,000 deaths in 2024) is the single largest component of the vector-borne aggregate and justifies the concentration of risk in sub-Saharan Africa: 95% of malaria deaths occur in the WHO African Region, which is what drives the regional_breakdown figures below. The US traveler figure comes from the CDC malaria surveillance data (see corroborating note in body text) which reports roughly 2,000 US cases and ~7 deaths per year, almost entirely in returning travelers.\n","independence_note":"WHO malaria fact sheet and WHO vector-borne fact sheet share an upstream (WHO programmatic estimates), so these two are not fully independent on the malaria number; they are used here for different roles — the vector-borne sheet for the aggregate, the malaria sheet for regional concentration and the most recent year.\n"},{"url":"https://www.cdc.gov/malaria/php/surveillance-report/index.html","title":"Data and Statistics on Malaria in the United States","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~2,000 US malaria cases reported per year; average of nearly 7 deaths per year over 2007–2022; 95% of US cases are in people who did not take appropriate prevention medication.","excerpt":"\"Approximately 2,000 malaria cases a year are reported in the United States, and on average there were nearly 7 deaths per year for the period 2007–2022. &hellip; Most cases are in people who contract malaria while traveling to another country where malaria spreads and return to the U.S. 95% of people with malaria did not take appropriate malaria prevention medication.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175159/https://www.cdc.gov/malaria/php/surveillance-report/index.html","calculation_notes":"CDC US surveillance gives the non-endemic anchor: ~7 malaria deaths per year in a country of ~335 million, almost entirely in returning travelers. That is ~2 × 10&#8315;&#8312; per person per year, i.e. roughly 1 in a million over 60 adult years. This is the basis for the \"United States / Europe / Australia\" row of regional_breakdown and for the 0.1× personal-factor multiplier for non-endemic travelers on prophylaxis. Methodologically independent of WHO.\n","independence_note":"CDC surveillance (ICD-coded US death certificates + NMSS case reports) is drawn from a completely different pipeline than WHO’s programmatic malaria estimates, so this is meaningfully independent corroboration of the near-zero risk for non-endemic residents.\n"},{"url":"https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2024","title":"World Malaria Report 2024","publisher":"World Health Organization","source_type":"govt_report","statistic":"263 million malaria cases and 597,000 malaria deaths in 2023; 94% of cases in WHO African Region","excerpt":"\"Malaria remains a serious global health challenge, claiming 597,000 lives in 2023 alone. In 2023, there were an estimated 263 million new malaria cases in 83 countries worldwide.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413175237/https://www.who.int/teams/global-malaria-programme/reports/world-malaria-report-2024","calculation_notes":"Mosquito-specific mortality anchor: 597,000 malaria deaths alone account for the vast majority of the entry's 700,000 vector-borne disease deaths. Confirms that mosquito-borne diseases (malaria + dengue + others) dominate the vector-borne category.\n","independence_note":"WHO World Malaria Report uses country-reported data and WHO modelling — the same upstream as the WHO vector-borne fact sheet but with malaria-specific detail.\n"},{"url":"https://www.floridamuseum.ufl.edu/shark-attacks/yearly-worldwide-summary/","title":"Yearly Worldwide Shark Attack Summary","publisher":"International Shark Attack File, Florida Museum of Natural History","source_type":"reputable_reference","statistic":"Five-year (2020-2024) global average of eight unprovoked fatal shark attacks per year worldwide","excerpt":"\"This number is also in line with the most recent five-year annual global average of eight unprovoked fatalities per year.\"\n","source_date":"2026-02-18","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714015159/https://www.floridamuseum.ufl.edu/shark-attacks/yearly-worldwide-summary/","calculation_notes":"Used only as the shark-attack anchor for the body-text comparison \"a mosquito kills roughly 90,000 humans for every one killed by a shark\": 700,000 (WHO vector-borne annual deaths) / 8 (ISAF five-year average unprovoked fatalities/year) ≈ 87,500, rounded to \"roughly 90,000\". Not used anywhere in the native/normalized calculation for this entry, which is WHO-anchored throughout.\n","independence_note":"Fully independent of the WHO/CDC pipeline above — ISAF is a standalone shark-attack database maintained by the Florida Museum of Natural History and shares no upstream with the vector-borne disease sources.\n"}],"comparison_anchors":[{"label":"Death by shark attack (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Death by bee/wasp sting (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Death by tsunami (lifetime, global adult)","lifetime_us_adult":0.00001},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.00525,"notes":"~700K deaths/yr across 8B people, compounded over 60 adult years. A scale marker, not a personal estimate."},{"region":"Sub-Saharan Africa","probability":0.08,"notes":"~95% of global malaria deaths concentrate in the WHO African Region; most of the global total lives here."},{"region":"Rural South and Southeast Asia","probability":0.01,"notes":"Malaria plus dengue plus Japanese encephalitis in pockets; high variance by country and urbanization."},{"region":"United States / Europe / Australia (residents, no travel)","probability":0.0000012,"notes":"CDC reports ~7 US malaria deaths per year, almost entirely in returning travelers. Essentially zero without travel."}],"personal_factor_multipliers":[{"factor":"Resident of an endemic rural area without bed nets or antimalarials","multiplier":100,"notes":"Concentrates most of the global death total."},{"factor":"Non-endemic traveler using prophylaxis, DEET, and bed nets","multiplier":0.1,"notes":"Per CDC: 95% of US malaria cases occur in people who did NOT take appropriate prevention."},{"factor":"Child under 5 in an endemic area","multiplier":5,"notes":"Young children account for the bulk of malaria mortality in the WHO African Region."},{"factor":"No bed net use in sub-Saharan Africa or South Asia endemic zone","multiplier":2,"notes":"Lengeler (2004, Cochrane review): insecticide-treated bed nets reduce all-cause child mortality by ~17% and malaria cases by ~50% in endemic zones; absence of net use in high-transmission areas approximately doubles personal malaria incidence"},{"factor":"Travel to active dengue or Zika outbreak region without repellent","multiplier":10,"notes":"CDC Travel Health advisories: travelers to active arboviral outbreak regions (dengue surges in Southeast Asia, Caribbean; Zika in outbreak zones) face per-trip risks orders of magnitude above the global average; DEET-based repellents reduce mosquito biting by ~87% and are the primary mitigation"}],"short_label":"Mosquito-borne disease","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The global-average figure is a scale marker only. Mosquito-borne mortality is overwhelmingly concentrated in sub-Saharan Africa (~95% of malaria deaths) and in pockets of South and Southeast Asia; for residents of the United States, Europe, Australia, and most of Latin America outside endemic zones, the risk is effectively zero in the absence of travel. The regional_breakdown values above are order-of-magnitude estimates, not precise per-country figures, and they hide substantial heterogeneity by age (young children dominate the malaria death toll), by access to bed nets and antimalarials, by rural vs urban residence, and by season. Specific travel-disease risks — dengue, yellow fever, Japanese encephalitis — will be covered in separate entries.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single stylized mosquito silhouette resting on a pale neutral background, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/mosquito-borne-disease","api_url":"https://likelier.app/api/fears/mosquito-borne-disease.json"},{"slug":"bone-marrow-donor-complications","question":"What are the odds of serious complications from donating bone marrow?","quick_answer":"The odds of a serious adverse event from donating bone marrow are about 1 in 170 per donation — roughly 0.6% for a blood-based (PBSC) donation and 2.4% for a marrow harvest. Many prospective donors assume a far greater risk of lasting harm or death.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Bone marrow donation carries a disproportionate fear burden relative to its actual risk profile. The phrase \"bone marrow\" evokes images of long needles inserted into the hip, general anesthesia, and weeks of painful recovery — imagery that has persisted even as the majority of donations have shifted to peripheral blood stem cell (PBSC) collection, which involves no surgery at all. Surveys of potential registry joiners consistently identify fear of the procedure as the primary barrier to registration, ahead of time commitment or inconvenience. The perception is further inflated by dramatic anecdotes shared online and a general conflation of discomfort (which is common) with serious medical harm (which is rare).\n","rough_estimate":"Many potential donors assume a significant risk of lasting harm or even death","kind":"intuition"},"native":{"display":"~0.6% serious adverse events for PBSC donors; ~2.4% for marrow donors","numerator":6,"denominator":1000,"unit":"per PBSC donation","population":"unrelated NMDP donors, 2004-2009 cohort"},"normalized":{"lifetime_us_adult":0.006,"display":"~1 in 170 chance of a serious adverse event per donation","log_value":-2.22,"assumptions":"Uses the PBSC serious adverse event rate of 0.56% from the Halter et al. 2009 NMDP prospective trial as the baseline, since PBSC now accounts for the majority of unrelated donor collections. The figure represents the per-donation risk, not a cumulative lifetime probability, because most donors donate only once. Life-threatening events occur at roughly 1 in 1,000 donations. No fatalities were documented in the NMDP cohort.\n","uncertainty":{"low":0.003,"high":0.024},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2668845/","title":"Adverse events among 2408 unrelated donors of peripheral blood stem cells: results of a prospective trial from the National Marrow Donor Program","publisher":"Bone Marrow Transplantation / Nature Publishing Group","source_type":"primary_study","statistic":"0.6% of 2,408 PBSC donors experienced serious adverse events; 6% experienced grade III-IV toxicities; no fatalities","excerpt":"\"Serious and unexpected toxicities occurred in 0.6% of donors. Grade III-IV adverse events occurred in 6% of donors. No fatalities were observed in this prospective cohort of 2,408 unrelated PBSC donors.\"\n","source_date":"2009-04-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426194118/https://pmc.ncbi.nlm.nih.gov/articles/PMC2668845/","calculation_notes":"The 0.6% serious adverse event rate among PBSC donors is the central estimate. This prospective NMDP trial tracked 2,408 unrelated donors from filgrastim mobilization through collection and 1-year follow-up. Grade III-IV events (6%) include expected side effects of G-CSF such as bone pain and headache that resolve after collection. The 0.6% figure captures events that were both serious and unexpected — the threshold relevant to the fear question.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4047500/","title":"Lower risk for serious adverse events and no increased risk for cancer after PBSC vs BM donation","publisher":"Blood / American Society of Hematology","source_type":"peer_reviewed","statistic":"SAE rate 0.56% for PBSC vs 2.38% for BM; life-threatening events 0.31% PBSC vs 0.99% BM; no donor deaths in NMDP cohort","excerpt":"\"Bone marrow donors had a higher rate of serious adverse events at 2.38% compared to 0.56% for peripheral blood stem cell donors. Life-threatening, serious unexpected, or chronic/disabling events occurred in 0.99% of BM donors vs 0.31% of PBSC donors.\"\n","source_date":"2014-06-05","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032306/https://pmc.ncbi.nlm.nih.gov/articles/PMC4047500/","calculation_notes":"This study compared SAE rates between BM and PBSC donation within the NMDP registry (donors 2004-2009). The threefold difference in life-threatening events (0.99% BM vs 0.31% PBSC) is clinically significant and explains the shift toward PBSC as the preferred collection method. No increased cancer risk was observed in long-term follow-up of PBSC donors, addressing a theoretical concern about filgrastim use.\n"},{"url":"https://www.nmdp.org/get-involved/join-the-registry/donate-pbsc/donor-requirements-faqs","title":"Bone Marrow & Blood Stem Cell Donor FAQs","publisher":"NMDP (formerly Be The Match)","source_type":"reputable_reference","statistic":"Fewer than 1% of PBSC donors (about 0.6%) experience serious side effects; donors are monitored closely by trained medical professionals","excerpt":"\"Fewer than 1% of donors (about 0.6%) experience serious side effects. Donors are monitored closely by trained medical professionals throughout the donation process to ensure they are safe and comfortable.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032344/https://www.nmdp.org/get-involved/join-the-registry/donate-pbsc/donor-requirements-faqs","calculation_notes":"NMDP's public-facing FAQ summarizes the peer-reviewed data in accessible terms. The \"<1% serious\" figure aligns with the 0.56% from the Halter et al. prospective trial. NMDP notes no donor deaths in its registry, consistent with WMDA data reporting only one death in over 250,000 collections worldwide between 1988 and 2018.\n"}],"comparison_anchors":[{"label":"General anesthesia death (per procedure)","lifetime_us_adult":0.0001},{"label":"Serious adverse drug reaction (per prescription)","lifetime_us_adult":0.002},{"label":"Blood donation adverse event requiring medical attention","lifetime_us_adult":0.003}],"personal_factor_multipliers":[{"factor":"Bone marrow (surgical) harvest vs PBSC","multiplier":4,"notes":"BM donors have ~4x the SAE rate of PBSC donors (2.38% vs 0.56%)"},{"factor":"Female donor","multiplier":2,"notes":"Women are approximately twice as likely to experience an SAE compared to men"},{"factor":"Second-time donor","multiplier":1.2,"notes":"Slightly elevated risk from repeat filgrastim exposure, though long-term data show no cancer increase"},{"factor":"Donor age 18-30 vs 40+","multiplier":0.6,"notes":"NMDP and DKMS donor outcomes data: younger donors (18-30) recover faster and have lower serious adverse event rates than donors aged 40+; G-CSF mobilization side effects and post-harvest fatigue are more pronounced in older donors, and general anesthesia risk for marrow harvest scales with age"},{"factor":"BMI ≥ 35 or significant comorbidity","multiplier":2,"notes":"NMDP donor eligibility and outcomes data: donors with obesity or underlying medical conditions face higher anesthesia and mobilization risks; NMDP screens donors for these factors and may redirect to the safest collection method, but residual risk elevation persists compared to healthy-weight donors"}],"short_label":"Marrow donation risk","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The normalized figure uses the per-donation PBSC SAE rate because most unrelated donations now use PBSC. For traditional bone marrow harvest under general anesthesia, the SAE rate is roughly four times higher (2.4%). The WMDA has documented one donor death in over 250,000 collections worldwide — a rate below 1 in 200,000. Most side effects (bone pain, fatigue, headache from G-CSF) are transient and resolve within days of collection.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified syringe and bone cross-section rendered in muted warm tones, flat vector illustration."},"canonical_url":"https://likelier.app/bone-marrow-donor-complications","api_url":"https://likelier.app/api/fears/bone-marrow-donor-complications.json"},{"slug":"caving-death","question":"What are the odds of dying while recreational caving over a typical caving career?","quick_answer":"Over a 20-year recreational (dry) caving career, the odds of a fatal accident are about 1 in 167. Most people overestimate this because they conflate ordinary dry caving with cave diving, a far more lethal specialty.\n","category":"sport","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"Public intuition about caving conflates two activities with very different fatality profiles. Dry caving — the headline scope of this entry — is the typical NSS-affiliated recreational activity (horizontal passage exploration, vertical rope work in known cave systems, organized grotto trips). Cave diving is a sub-speciality undertaken by a small minority of technically proficient cavers and is one of the most lethal recreational activities humans engage in. Headlines and documentaries (Nutty Putty 2009, Tham Luang 2018, recreational cave-diving deaths in Florida sinks) blur the two in lay imagination, leading most observers to estimate dry-caving career risk well above the actual rate. No large-scale survey isolates US public perception of caving fatality odds; this entry uses editorial intuition.\n","rough_estimate":"most people likely overestimate dry caving career risk and underestimate the cave-diving differential","kind":"intuition"},"native":{"display":"~30 deaths per 100,000 active NSS-affiliated cavers per year (US, 1980–2008 NSS ACA dataset)","numerator":81,"denominator":270000,"unit":"per active caver per year","population":"US NSS-affiliated active recreational cavers, 1980-2008 (NSS American Caving Accidents dataset)"},"normalized":{"lifetime_us_adult":0.006,"display":"~1 in 167 over a 20-year caving career","log_value":-2.22,"assumptions":"The Stella-Watts et al. 2012 study in Wilderness and Environmental Medicine analyzed 28 years (1980-2008) of NSS American Caving Accidents reports and identified 81 caving fatalities across the US, averaging approximately 3 deaths per year. The denominator is the active US caving population during that period. NSS membership has been reported in the 8,000-10,000 range for most of the 2000s and 2010s; the broader active US caving community (NSS-affiliated grotto members plus unaffiliated frequent cavers) is plausibly in the 10,000-15,000 range. The 2 million Americans who visit caves annually are overwhelmingly commercial-tour visitors (Mammoth Cave, Carlsbad Caverns, Luray Caverns) on developed walkways and do not constitute \"active recreational cavers\" in the wild-caving sense relevant to career risk. Using a working denominator of 10,000 active cavers, the annual fatality rate is approximately 30 per 100,000 active cavers per year (3/10,000 = 0.0003). Compound probability over a 20-year caving career: 1 − (1 − 0.0003)^20 ≈ 0.006, or roughly 1 in 167. The numerator (81 deaths over 28 years = 81 / (28 × 10,000) = 28.9 per 100,000 person-years) is essentially the same. Cave diving fatalities are excluded from the headline; they are addressed in personal_factor_multipliers because the rate differential is large enough to merit separate treatment.\n","uncertainty":{"low":0.004,"high":0.0075},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/22704081/","title":"The Epidemiology of Caving Injuries in the United States","publisher":"Wilderness and Environmental Medicine (PubMed)","source_type":"peer_reviewed","statistic":"877 incident reports involving 1,356 cavers (1980-2008); 81 documented fatalities; falls and drowning each accounted for 30% of caver deaths","excerpt":"\"Over a 28-year period (1980-2008), 877 incident reports involving 1,356 cavers were documented by the National Speleological Society in American Caving Accidents. Of these, 81 caving fatalities occurred. Falls accounted for 74% of traumatic injuries; the most common mechanisms leading to death were caver fall and drowning, with 24 (30%) deaths each. Eighty-four percent of fatality victims were male; the peak age group was 20-29 years.\"\n","source_date":"2012-09-01","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20251013132220/https://pubmed.ncbi.nlm.nih.gov/22704081/","calculation_notes":"Primary peer-reviewed source for the 81-deaths-over-28-years (≈3 per year) headline figure. The paper does not publish a per-100,000 rate because it lacks a formal denominator; the denominator used here (~10,000 NSS-affiliated active cavers) is derived from NSS membership figures published independently on caves.org. Combined: 81 deaths / (28 years × 10,000 cavers) ≈ 28.9 per 100,000 active-caver-years, rounded to 30/100,000 for the native display.\n","independence_note":"Peer-reviewed analysis published in Wilderness and Environmental Medicine by Stella-Watts, Holstege, Lee, and Charlton at the University of Virginia. Methodologically independent of NSS itself — the authors are academic emergency-medicine researchers conducting epidemiological synthesis of the ACA dataset, not NSS staff.\n"},{"url":"https://caves.org/american-caving-accidents/","title":"American Caving Accidents (NSS Annual Publication)","publisher":"National Speleological Society","source_type":"reputable_reference","statistic":"Annual journal of record for North American caving incidents and fatalities; NSS reports over 8,000 members","excerpt":"\"American Caving Accidents is the journal of record for accident and safety incident reports from the North American caving community. The National Speleological Society has over 8,000 Members and is the largest caving focused membership organization in the world.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260330162734/https://caves.org/american-caving-accidents/","calculation_notes":"Provides the denominator anchor: NSS member-reported >8,000 active members. The broader US active caver population is plausibly 10,000-15,000 once unaffiliated grotto-active cavers are included; this informs the uncertainty band (0.004-0.0075). NSS itself is the publisher of the underlying ACA incident dataset analyzed by Stella-Watts 2012, so numerator and denominator derive from related but distinct NSS data streams.\n","independence_note":"Direct NSS organizational publication; provides the active-population denominator anchor that the peer-reviewed source (Stella-Watts 2012) does not formally publish. NSS membership counts are self-reported by the organization; cross-validated by Wikipedia (8,700-10,000 range across different reporting periods).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27723015/","title":"Thirty years of American cave diving fatalities","publisher":"Diving and Hyperbaric Medicine (PubMed)","source_type":"peer_reviewed","statistic":"161 US cave diving fatalities between July 1985 and June 2015 (30 years); average ~5.4 per year; trend declining from 8 per year to under 3 as training improved","excerpt":"\"Between July 1, 1985, and June 30, 2015, a total of 161 divers died during cave diving expeditions in the United States, with 67 being trained cave divers and 87 being untrained. The most common cause of death was asphyxia due to drowning, preceded by running out of breathing gas, usually after getting lost owing to a loss of visibility caused by suspended silt. The annual number of cave diving fatalities has steadily fallen over the last three decades, from eight to less than three.\"\n","source_date":"2016-09-01","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260309150238/https://pubmed.ncbi.nlm.nih.gov/27723015/","calculation_notes":"Cave diving sub-speciality data, used for the personal_factor_multiplier analysis (not the headline). 161 deaths / 30 years = 5.37/year. Active US cave-diving population is much smaller than dry caving — the NSS Cave Diving Section (founded 1974) and PADI/TDI/NACD-certified active US cave divers number in the low thousands at most, with frequent estimates around 500-1,500 regularly active. Using 1,000 active cave divers: 5.37/1000 = 0.00537/year, or 537 per 100,000 per year — roughly 18-30 times the dry caving rate depending on which denominator is chosen. The 30× multiplier used in the personal_factor analysis is the conservative midpoint.\n","independence_note":"Peer-reviewed cave-diving fatality analysis by Potts, Buzzacott, and Denoble from Divers Alert Network (DAN). Methodologically independent of the NSS caving-accidents dataset; cave-diving deaths are tracked by NSS-CDS, International Underwater Cave Rescue and Recovery (IUCRR), and DAN separately from terrestrial caving accidents.\n"},{"url":"https://en.wikipedia.org/wiki/Caving","title":"Caving (Wikipedia)","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"Cave diving is described as 'a distinct, and more hazardous, sub-speciality undertaken by a small minority of technically proficient cavers'","excerpt":"\"Caving is a fairly safe sport compared to other activities, although incidents do occur related to flooding, hypothermia, rockfalls, and rope-technique accidents. Cave diving is a distinct, and more hazardous, sub-speciality undertaken by a small minority of technically proficient cavers.\"\n","source_date":"2026-05-01","source_accessed":"2026-05-23","archive_url":"http://web.archive.org/web/20260425035129/https://en.wikipedia.org/wiki/Caving","calculation_notes":"Supporting reference establishing the editorial framing — dry caving as the typical recreational activity, cave diving as a distinct sub-speciality with a separate (higher) risk profile. Not used to derive headline numbers.\n","independence_note":"Encyclopedia entry; supplementary context only. The headline arithmetic and the cave-diving multiplier are both grounded in peer-reviewed sources (Stella-Watts 2012, Potts 2016).\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.0012},{"label":"Commercial fishing career death (20-year career)","lifetime_us_adult":0.0224},{"label":"Logging career death (30-year career)","lifetime_us_adult":0.0295}],"personal_factor_multipliers":[{"factor":"Cave diving (any sump dive or fully flooded passage) vs strictly dry caving","multiplier":30,"notes":"Cave diving fatality rates are dramatically higher than dry caving. Potts, Buzzacott, and Denoble (2016) documented 161 US cave diving deaths over 30 years (1985-2015), averaging ~5.4 per year against an active US cave-diving population of approximately 500-1,500. Using a midpoint denominator of 1,000 active cave divers gives roughly 540 deaths per 100,000 active cave divers per year, against approximately 30 per 100,000 for dry cavers — a ratio of roughly 18-30x depending on denominator assumptions. The conservative 30x multiplier is used here; the true multiplier may be higher. Drowning following gas-supply exhaustion in silted-out, lightless passages is the dominant mechanism. Importantly, the Potts study found that 67 of 161 deaths (42%) were certified cave divers — formal training reduces but does not eliminate the risk.\n"},{"factor":"Untrained vs trained cave diver (within the cave-diving population)","multiplier":2,"notes":"Potts et al. (2016) found that 87 of 161 US cave-diving deaths (54%) were untrained divers — open-water-certified scuba divers entering caves without cave-diving certification. Adjusting for the much smaller untrained-but- cave-entering population, untrained divers face an estimated 2x-5x higher per-dive fatality rate than certified cave divers. The Jeffrey Bozanic 2005 IUCRR analysis noted the proportion of trained-diver deaths rose from ~5% (1973-1987) to ~30% (1988-2004), reflecting both successful expansion of cave-diving certification and complacency among experienced divers using newer technologies (rebreathers, DPVs).\n"},{"factor":"Caving alone (solo) vs in an organized group of 3 or more","multiplier":2.5,"notes":"NSS safety standards consistently recommend a minimum group size of three cavers (so that one can stay with an injured caver while another exits to summon rescue). Solo caving eliminates this redundancy entirely. The Stella-Watts 2012 epidemiological data found that being unable to exit the cave (stranded/lost) accounted for 54% of all incident reports — a category where solo cavers are dramatically more vulnerable because no companion can raise the alarm. The 2.5x multiplier reflects the consensus practitioner view that solo caving substantially elevates fatality risk; precise relative-rate data are not available because solo caving is a relatively small fraction of total caver-trips.\n"},{"factor":"Inexperience (first 2 years of caving) vs experienced caver","multiplier":1.7,"notes":"The Stella-Watts 2012 study found that inexperience contributed to 26 of 81 caving fatalities (32%). Across hazardous outdoor recreation, novices face elevated risk during the first 1-2 years of participation due to limited familiarity with terrain, equipment use (SRT vertical rope work, lighting redundancy, navigation), and judgment about when to turn back. The peak fatality age group in the Stella-Watts dataset was 20-29 years, which partially overlaps with the early-career cohort.\n"},{"factor":"Vertical caving (single-rope-technique drops) vs strictly horizontal passages","multiplier":1.5,"notes":"Caver fall accounted for 24 of 81 fatalities (30%) in the Stella-Watts 2012 data — tied with drowning as the most common fatality mechanism. Vertical caving involving single-rope-technique (SRT) descent and ascent introduces mechanical failure modes (rappel device error, ascender failure, anchor failure, rope abrasion against rock) that horizontal-passage caving does not. Precise relative rates are not published, but the fall-death share of 30% combined with the fact that only a minority of caving trips involve vertical rope work implies an elevated fall-fatality rate among vertical cavers specifically.\n"}],"short_label":"Caving career death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The headline rate is sensitive to the denominator chosen for \"active US cavers\". NSS membership has hovered around 8,000-10,000 over the past two decades, but not every active US caver is an NSS member, and not every NSS member is currently active in field caving. A defensible working denominator of 10,000 active cavers produces the headline rate of approximately 30 per 100,000 per year; using 8,000 would raise the rate to about 37 per 100,000 (20-year career ≈ 0.74%, or roughly 1 in 135), and using 15,000 would lower it to about 20 per 100,000 (20-year career ≈ 0.40%, or roughly 1 in 250). The uncertainty band (0.4%-0.75%) captures this range. The 2 million Americans who visit caves annually are predominantly commercial-tour visitors on developed walkways and are deliberately excluded from this denominator; mixing them in would artificially deflate the rate by orders of magnitude and is not what the question asks. Cave diving is excluded from the headline scope and treated as a personal_factor_multiplier; the dry-vs-wet bifurcation is the single most important factor in caving fatality risk and is explicitly called out rather than averaged in. The Stella-Watts dataset (1980-2008) is the most recent comprehensive epidemiological synthesis; more recent NSS ACA reports (2009-2010, 2017-2018, 2019-2020) suggest broadly similar annual fatality counts in the low single digits, though no updated peer-reviewed analysis has been published. The 20-year career assumption matches the most commonly cited caver-engagement window; cavers who continue into their 60s or 70s accumulate proportionally higher cumulative risk, while those who participate for only a few years before stopping accumulate less. Career-level rates also do not capture trip-level intensity — a caver doing 50 trips per year accumulates exposure faster than one doing 5 trips per year, though the per-year fatality rate used here implicitly averages across these intensities.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-23","last_reviewed":"2026-05-23","reviewed":true,"generated_at":"2026-05-23","image":{"alt":"A single caving helmet with attached lamp resting on a pale neutral surface alongside a coiled length of static rope, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/caving-death","api_url":"https://likelier.app/api/fears/caving-death.json"},{"slug":"cervical-cancer","question":"What are the odds of developing cervical cancer?","quick_answer":"A US woman's lifetime odds of being diagnosed with cervical cancer are about 1 in 167, with roughly 1 in 500 dying of it. Screening and the HPV vaccine have made it far rarer than a generation ago, so many people sense it is \"rare now\" without a concrete number.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Cervical cancer sits in an unusual perceptual niche. Among women over 40, it is remembered as a major threat that screening and the Pap smear tamed — something their mothers worried about. Among women under 30, it is increasingly filed under \"solved by the HPV vaccine.\" Both frames contain truth and both produce underestimates in specific directions: older women may not realise how much screening has reduced the absolute number, while younger women may not realise that the vaccine does not cover all oncogenic HPV strains and that over 20% of cervical cancers are diagnosed in women over 65 who may have aged out of screening. No high-quality US survey isolates \"fear of cervical cancer\" as a standalone probability question, so the perceived side is editorial intuition.\n","rough_estimate":"Most US women sense it is 'rare now' but lack a concrete number; globally it remains a top-three cancer killer of women","kind":"intuition"},"native":{"display":"~7.7 new cases per 100,000 women per year (US, 2018-2022 age-adjusted)","numerator":77,"denominator":1000000,"unit":"per year","population":"US women (all ages, age-adjusted, SEER 2018-2022)"},"normalized":{"lifetime_us_adult":0.006,"display":"~1 in 167 lifetime (US women)","log_value":-2.222,"assumptions":"SEER Cancer Stat Facts reports a lifetime risk of being diagnosed with cervical cancer of approximately 0.6% (about 1 in 167), based on 2018-2021 SEER data (excluding 2020 due to COVID screening disruption). The age-adjusted incidence rate is 7.7 per 100,000 women per year (2018-2022). Using the site-standard 59-year remaining-life horizon from age 18: 1 − (1 − 7.7e-5)^59 ≈ 0.00454, which is lower than the SEER actuarial estimate of 0.006 because SEER uses full life-table methods with age-specific incidence rates that peak in women aged 35-44 and remain elevated through old age. We use the SEER actuarial figure of 0.006 as the more accurate estimate. The death rate is 2.2 per 100,000 women per year (2019-2023), implying a lifetime mortality risk of roughly 0.2% (1 in 500). The uncertainty band (0.004-0.009) reflects the range between well-screened populations with HPV vaccination (where incidence is falling sharply) and under-screened populations (where incidence remains materially higher). The scope is subgroup_lifetime because the figure applies only to women / people with a cervix (roughly half of US adults), not to all US adults.\n","uncertainty":{"low":0.004,"high":0.009},"scope":"subgroup_lifetime"},"sources":[{"url":"https://seer.cancer.gov/statfacts/html/cervix.html","title":"Cancer Stat Facts: Cervical Cancer","publisher":"National Cancer Institute — Surveillance, Epidemiology, and End Results Program (SEER)","source_type":"govt_report","statistic":"Approximately 0.6% of women will be diagnosed with cervical cancer during their lifetime; incidence rate 7.7 per 100,000 women per year; death rate 2.2 per 100,000 per year","excerpt":"\"Approximately 0.6 percent of women will be diagnosed with cervical cancer at some point during their lifetime, based on 2018–2021 data. [...] The rate of new cases of cervical cancer was 7.7 per 100,000 women per year. [...] The death rate was 2.2 per 100,000 women per year based on 2019–2023.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260409062730/https://seer.cancer.gov/statfacts/html/cervix.html","calculation_notes":"SEER provides the headline lifetime risk (0.6%) using actuarial life-table methods applied to age-specific incidence rates from 18 SEER registries covering approximately 48% of the US population. The 0.6% figure is used directly as the normalized estimate because SEER's actuarial method is more precise than a crude annual-rate compounding over 59 years. The incidence rate of 7.7 per 100,000 anchors the native display. The death rate of 2.2 per 100,000 implies that roughly 29% of diagnosed cases are fatal (2.2/7.7), reflecting the relatively favorable 5-year survival rate of ~67% for all stages combined.\n"},{"url":"https://www.cancer.org/cancer/types/cervical-cancer/about/key-statistics.html","title":"Key Statistics for Cervical Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Estimated 13,360 new cases and 4,320 deaths from cervical cancer in the US in 2025; incidence declined by more than half from the mid-1970s to mid-2000s due to screening","excerpt":"\"The American Cancer Society's estimates for cervical cancer in the United States for 2025 are: about 13,360 new cases of invasive cervical cancer will be diagnosed, and about 4,320 women will die from cervical cancer. [...] More than 20% of cervical cancers are found in women over 65.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260327094716/https://www.cancer.org/cancer/types/cervical-cancer/about/key-statistics.html","calculation_notes":"ACS provides the absolute case and death counts for 2025, which serve as a cross-check on the SEER rate-based figures. 13,360 new cases among ~170 million US women yields roughly 7.9 per 100,000, consistent with the SEER age-adjusted rate of 7.7. The >20% of cases in women over 65 anchors the caveat about screening cessation. ACS and SEER draw from overlapping but not identical data pipelines; ACS projections are modelled from SEER historical trends.\n","independence_note":"ACS projections are modelled from SEER data and NCHS mortality data, so they are not fully independent. The value is in the forward projection and the clinical context ACS adds around screening and HPV vaccination.\n"},{"url":"https://jamanetwork.com/journals/jama/fullarticle/2827212","title":"Cervical Cancer Mortality Among US Women Younger Than 25 Years, 1992-2021","publisher":"JAMA","source_type":"peer_reviewed","statistic":"Cervical cancer mortality among US women under 25 declined 15.2% per year from 2013-2015 to 2019-2021, following HPV vaccine introduction","excerpt":"\"From 2013-2015 to 2019-2021, mortality decreased 15.2% per year, coinciding with the time period when women who were age-eligible for HPV vaccination in 2006 would have entered the age range of 20-24 years.\"\n","source_date":"2024-11-27","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260502024524/https://jamanetwork.com/journals/jama/fullarticle/2827212","calculation_notes":"The JAMA study quantifies the HPV vaccine's impact on mortality in the first fully vaccinated cohort. A 15.2% annual decline is dramatic and suggests the 0.6% lifetime incidence figure will fall substantially for women born after ~1990 who were vaccinated. The current 0.6% reflects a mixed population of vaccinated and unvaccinated women; the uncertainty band's lower end (0.004) is where we expect the fully-vaccinated-cohort lifetime risk to settle, though this remains a projection. The study is methodologically independent of SEER incidence data (it uses NCHS mortality data directly).\n","independence_note":"Uses NCHS mortality data from CDC WONDER, not SEER incidence registries. Genuinely independent of the SEER source for the mortality trend.\n"}],"comparison_anchors":[{"label":"Breast cancer (lifetime, US women)","lifetime_us_adult":0.13},{"label":"Lung cancer (lifetime, US)","lifetime_us_adult":0.056},{"label":"Melanoma (lifetime, US)","lifetime_us_adult":0.03},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Unscreened woman (no Pap or HPV test in 5+ years)","multiplier":3,"notes":"Cervical cancer incidence is 3-5x higher in unscreened vs regularly screened women; screening catches precancers before invasion."},{"factor":"HPV-vaccinated woman (full course before sexual debut)","multiplier":0.15,"notes":"HPV vaccines cover types 16 and 18, which cause ~70% of cervical cancers; the 9-valent vaccine covers ~90% of oncogenic types."},{"factor":"HIV-positive woman","multiplier":6,"notes":"HIV impairs HPV clearance; cervical cancer is an AIDS-defining illness. WHO reports 6x higher cervical cancer incidence in HIV-positive women."},{"factor":"Woman in Sub-Saharan Africa (limited screening)","multiplier":5,"notes":"GLOBOCAN estimates cervical cancer incidence at ~40 per 100,000 in SSA vs ~7.7 in the US, driven by low screening coverage and high HPV prevalence."},{"factor":"Current smoker","multiplier":2,"notes":"Smoking roughly doubles the risk of cervical squamous cell carcinoma in HPV-positive women."}],"short_label":"Cervical cancer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The 0.6% lifetime incidence figure reflects the current US screening environment, in which roughly 80% of eligible women receive cervical screening at recommended intervals. It is a population average across vaccinated and unvaccinated women. For cohorts fully vaccinated with the 9-valent HPV vaccine before sexual debut, the lifetime risk is likely to fall to 0.1% or lower, because the vaccine covers the HPV types responsible for roughly 90% of cervical cancers. Conversely, in countries with minimal screening infrastructure (much of Sub-Saharan Africa, parts of South and Southeast Asia), the incidence is 4-5x higher, and cervical cancer remains the second or third most common cancer in women. The 0.6% is an incidence figure, not a mortality figure; the lifetime mortality risk in the US is roughly 0.2% (1 in 500), reflecting a 5-year relative survival rate of ~67% across all stages. Racial disparities persist in the US: Black and Hispanic women have higher incidence and mortality than White women, driven largely by differences in screening access and HPV vaccination uptake. The incidence rate has been roughly stable for the past decade after declining by more than half from the mid-1970s; the next major decline is expected as fully vaccinated cohorts enter the peak-incidence age range.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single simplified microscope silhouette in muted teal and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/cervical-cancer","api_url":"https://likelier.app/api/fears/cervical-cancer.json"},{"slug":"tick-borne-encephalitis","question":"What are the odds of getting tick-borne encephalitis (TBE)?","quick_answer":"For an unvaccinated adult in a high-endemic Central or Eastern European belt, tick-borne encephalitis runs about 1 in 170 over a lifetime, near 10 per 100,000 a year; for a US resident the locally acquired risk is effectively zero. It is usually folded into a general fear of ticks rather than recognized as the specific, vaccine-preventable disease.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Tick-borne encephalitis is the dangerous half of the tick-disease pair, and in Central and Eastern Europe it is consistently overshadowed by the other half. When people in Poland, the Baltic states, or Czechia worry about \"kleszcze\", the named fear is almost always Lyme borreliosis — the bacterial infection that is far more common but, treated early, far less likely to cause lasting harm. TBE is a virus, it can inflame the brain and its lining (meningitis and encephalitis), there is no antiviral treatment, and unlike Lyme it is vaccine-preventable. Yet vaccine uptake across most of the endemic belt outside Austria remains low, which is the clearest signal that the specific, severe, preventable disease is underweighted relative to the diffuse seasonal anxiety about ticks in general.\n","rough_estimate":"usually folded into a general fear of ticks/Lyme; the specific TBE risk is rarely distinguished, and the vaccine is widely available but underused outside Austria","kind":"intuition"},"native":{"display":"~10 in 100,000 per year (unvaccinated adult, high-endemic Central/Eastern European region)","numerator":10,"denominator":100000,"unit":"per year","population":"unvaccinated adults resident in high-endemic Central/Eastern European foci (Baltic states, Czechia, Slovenia, northeastern Poland)"},"normalized":{"lifetime_us_adult":0.006,"display":"~1 in 170 lifetime (unvaccinated adult, high-endemic Central/Eastern Europe)","log_value":-2.22,"assumptions":"TBE incidence is sharply focal. The EU/EEA average notification rate was 0.81 per 100,000 in 2022 (ECDC), but risk concentrates in a high-endemic belt where rates run an order of magnitude higher: Lithuania reported 13.4 per 100,000 and Estonia 10.5 in 2022, and Czechia, Estonia and Lithuania have each exceeded 5 per 100,000 for five consecutive years. Northeastern Poland (Podlaskie) is a comparably intense focus. A central native estimate of ~10 per 100,000 per year is used for an unvaccinated adult living in this high-endemic belt. Lifetime conversion over the standard 59-year adult horizon: 1 − (1 − 1.0 × 10⁻⁴)^59 ≈ 0.0059, rounded to 0.006 (about 1 in 170). The lower uncertainty bound (0.003) reflects lower-endemicity foci or limited outdoor exposure (~4–5 per 100,000); the upper bound (0.012) reflects the highest national rates combined with heavy outdoor exposure (~20 per 100,000). This figure counts symptomatic, clinically diagnosed TBE; roughly three-quarters of TBE virus infections are asymptomatic (CDC), so the infection rate is several-fold higher than the disease rate stated here. The scope is subgroup_lifetime (a high-endemic European resident); for a US adult, locally acquired TBE risk is effectively zero — only 12 cases have ever been reported among US travelers.\n","uncertainty":{"low":0.003,"high":0.012},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/72/rr/rr7205a1.htm","title":"Tick-Borne Encephalitis Vaccine: Recommendations of the Advisory Committee on Immunization Practices, United States, 2023","publisher":"US Centers for Disease Control and Prevention (MMWR Recommendations and Reports)","source_type":"govt_report","statistic":"Approximately three-fourths of TBE virus infections are asymptomatic; European-subtype case-fatality is usually <2%; sequelae occur in 20%–40% of European-subtype cases; vaccine effectiveness after ≥3 doses is 91%–99%, with 99% seroconversion after dose 3.","excerpt":"\"Approximately three fourths of TBE virus infections are asymptomatic. … Among patients with neurologic disease and infected with the European subtype virus, the case fatality rate is usually <2%. … Among patients infected with the European subtype virus, sequelae have been reported in 20%–40% overall, including neurologic sequelae (e.g., limb paresis or paralysis) in up to 10%. … VE estimates for the European subtype virus TBE vaccines against disease caused by the European subtype virus after ≥3 doses of vaccine are 91%–99%. … Among 416 adults seronegative at baseline, 411 (99%) were seropositive at 21–28 days after dose 3.\"\n","source_date":"2023-11-02","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260511083323/https://www.cdc.gov/mmwr/volumes/72/rr/rr7205a1.htm","calculation_notes":"Supplies the severity profile (CFR <2%, sequelae 20–40%, asymptomatic fraction ~75%) and the vaccine-efficacy basis for the personal factor multipliers. VE of 91–99% after the full primary series implies a residual risk multiplier of ~0.05× for a fully vaccinated person. The 12-US-traveler figure (1 case in 1979, 11 during 2001–2021) anchors the \"effectively zero for US residents\" framing.\n","independence_note":"US CDC/ACIP synthesis of vaccine-trial and European-effectiveness data; independent of the ECDC surveillance counts used for incidence.\n"},{"url":"https://www.ecdc.europa.eu/en/publications-data/tick-borne-encephalitis-annual-epidemiological-report-2022","title":"Tick-borne encephalitis — Annual Epidemiological Report for 2022","publisher":"European Centre for Disease Prevention and Control (ECDC)","source_type":"govt_report","statistic":"20 EU/EEA countries reported 3,650 TBE cases for 2022 (3,516 confirmed); EU/EEA notification rate 0.81 per 100,000, highest in Lithuania (13.4) and Estonia (10.5); 94.0% of cases with known vaccination status were unvaccinated. Poland reported 367 cases (rate 1.0).","excerpt":"\"For 2022, 20 EU/EEA countries reported 3 650 cases of tick-borne encephalitis (TBE), 3 516 (96.3%) of which were confirmed. … The EU/EEA notification rate for 2022 was 0.81 per 100 000 population. … The notification rate was highest in Lithuania (13.4 cases per 100 000 population), followed by Estonia (10.5). … Notification rates were consistently high (over five per 100 000 population) during the last five years in three countries: Lithuania, Estonia, and Czechia. … Of the 1 620 confirmed cases for which information about vaccination status was available, 1 522 (94.0%) were reported as not vaccinated against TBE.\"\n","source_date":"2024-06-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260305112417/https://www.ecdc.europa.eu/en/publications-data/tick-borne-encephalitis-annual-epidemiological-report-2022","calculation_notes":"Provides the population-based incidence anchoring the native rate and the regional breakdown. The EU/EEA average of 0.81 per 100,000 versus the Baltic high-endemic rates of 10–13 per 100,000 (Lithuania historically to 24–25 in 2019–2020; Latvia averaging 8.5 over 2018–2021) establishes the ~10× focal concentration used in the assumptions. Poland's national rate was 1.0 per 100,000 (367 cases) in 2022, but risk concentrates in northeastern foci far above the national average. The 94%-unvaccinated figure among cases with known status anchors the underrated framing and the vaccine factor multiplier.\n","independence_note":"European surveillance aggregation via EpiPulse/TESSy, independent of the US CDC synthesis and the single-centre Polish hospital series.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12300996/","title":"Breakthrough Tick-Borne Encephalitis and Epidemiological Trends in an Endemic Region in Poland: A Retrospective Hospital-Based Study, 1988–2020","publisher":"Vaccines (Basel) — Sulik-Wakulińska, Toczyłowski & Grygorczuk","source_type":"peer_reviewed","statistic":"1,518 patients were hospitalised for TBE at a single Białystok hospital over 1988–2020, capturing approximately 50% of all TBE cases reported in the (northeastern Poland) region.","excerpt":"\"Between 1988 and 2020, a total number of 1518 patients were hospitalized because of TBE [at the study hospital, which captures] approximately 50% of all TBE cases reported in the region. … Tick-borne encephalitis is endemic across much of Eurasia, [with] northeastern Poland [a] highly endemic area.\"\n","source_date":"2025-06-20","source_accessed":"2026-06-21","archive_url":"https://web.archive.org/web/20260622052849/https://pmc.ncbi.nlm.nih.gov/articles/PMC12300996/","calculation_notes":"Confirms northeastern Poland (Podlaskie) as a high-endemic focus comparable to the Baltic belt, supporting its inclusion in the native population definition and the Polish regional_breakdown row. The single-centre series capturing ~50% of regional cases illustrates the focal clustering that the national-average rates obscure.\n","independence_note":"Peer-reviewed single-centre Polish cohort, independent of both the ECDC aggregate counts and the US CDC recommendations.\n"},{"url":"https://www.ecdc.europa.eu/en/publications-data/country-profile-austria-tick-borne-encephalitis-tbe","title":"Country profile: Austria. Tick-borne encephalitis (TBE)","publisher":"European Centre for Disease Prevention and Control (ECDC)","source_type":"govt_report","statistic":"Austria's TBE vaccination coverage is about 85% of the total population, following a sustained public campaign; annual cases have fallen to a range of 50-100 in the last decade, down from 300-700 per year before vaccination.","excerpt":"\"Today, vaccination coverage is at about 85% of the total population in Austria which – because of the effectiveness of the vaccine – has led to a dramatic decline in the disease in Austria, with the numbers of annual cases ranging from 50 to 100 in the last decade.\"\n","source_date":"2012-09-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20231104152738/https://www.ecdc.europa.eu/en/publications-data/country-profile-austria-tick-borne-encephalitis-tbe","calculation_notes":"Supports the body-prose claim that Austria is the exception to the low regional vaccine uptake described elsewhere in this entry: a sustained public vaccination campaign pushed coverage to ~85% (above the ~80% threshold cited in the body) and cut annual case counts several-fold versus the pre-vaccination era (300-700/year down to 50-100/year). Independent of the EU/EEA-wide 2022 surveillance report and the CDC ACIP vaccine-efficacy synthesis already cited.\n","independence_note":"ECDC country-specific surveillance profile for Austria, distinct from the EU/EEA-wide 2022 annual epidemiological report already cited in this entry.\n"}],"comparison_anchors":[{"label":"Lyme disease (lifetime, endemic-region adult)","lifetime_us_adult":0.25},{"label":"Melanoma from UV exposure (lifetime, US adult)","lifetime_us_adult":0.028},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Rabies death from a dog bite (lifetime, global adult)","lifetime_us_adult":0.00069}],"regional_breakdown":[{"region":"Lithuania / Estonia (per year, highest EU national rates)","probability":0.00013,"notes":"ECDC 2022: Lithuania 13.4 and Estonia 10.5 per 100,000 — the highest reported national notification rates in the EU/EEA. Lithuania reached 24–25 per 100,000 in 2019–2020."},{"region":"Czechia / Slovenia / Latvia (per year, high-endemic)","probability":0.00007,"notes":"ECDC 2022: Czechia 6.7 and Slovenia 5.9 per 100,000; Latvia averaged 8.5 per 100,000 over 2018–2021 (did not report 2022). Local foci run higher than these national averages."},{"region":"Poland (national, per year)","probability":0.00001,"notes":"ECDC 2022: 367 cases, national rate 1.0 per 100,000. Risk concentrates sharply in northeastern foci (Podlaskie), where a single Białystok hospital captures ~50% of the region's cases — well above the national average."},{"region":"EU/EEA average (per year, all countries)","probability":0.0000081,"notes":"ECDC 2022 average notification rate 0.81 per 100,000 — an order of magnitude below the high-endemic belt because risk is sharply focal."},{"region":"United States (locally acquired, per year)","probability":3e-10,"notes":"Effectively zero. Only 12 TBE cases have ever been reported among US travelers (1 in 1979, 11 during 2001–2021); there is no established local transmission."}],"personal_factor_multipliers":[{"factor":"Full TBE vaccination (≥3-dose primary series, boosters up to date)","multiplier":0.05,"notes":"CDC ACIP 2023: vaccine effectiveness against European-subtype disease is 91–99% after ≥3 doses, with 99% seroconversion after the third dose. Residual risk ≈ 0.05×. This is by far the dominant lever — and the one most often left unused outside Austria."},{"factor":"Incomplete or lapsed TBE vaccination (1–2 doses, or booster overdue)","multiplier":0.3,"notes":"Two doses confer roughly one season of protection; durable, high-level protection requires the full three-dose primary series plus periodic boosters (CDC ACIP 2023)."},{"factor":"Frequent outdoor activity in endemic forest foci (forestry work, foraging, mushroom-picking, hiking)","multiplier":3,"notes":"TBE risk concentrates in people with heavy questing-tick exposure through vegetation during the April–November season; ~90% of EU cases occur June–November."},{"factor":"Resident of lower-endemicity Western/Southern Europe or a non-endemic region","multiplier":0.2,"notes":"TBE is sharply focal; the EU/EEA average rate (0.81 per 100,000) is roughly an order of magnitude below the Baltic/Czech high-endemic belt, and much of Western and Southern Europe is near zero (ECDC 2022)."}],"short_label":"Tick-borne encephalitis","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The 1-in-170 figure is the lifetime probability of symptomatic, clinically diagnosed TBE for an unvaccinated adult living in a high-endemic Central or Eastern European focus; it is not a US-resident figure and not an all-of-Europe figure. Roughly three-quarters of TBE virus infections are asymptomatic, so the infection rate is several times higher than the disease rate stated here, while the rate of severe neuroinvasive disease is lower. Two facts make TBE behave differently from Lyme borreliosis, with which it is usually confused. First, the virus sits in the tick's salivary glands and can transmit within minutes of attachment, so prompt tick removal — highly protective against Lyme, which needs 36+ hours of attachment — does little against TBE; vaccination is the primary countermeasure. Second, TBE has an alimentary route Lyme lacks: drinking unpasteurized milk or eating fresh cheese from infected goats, sheep, or cows has caused documented family and village outbreaks. Severity rises with age. For the European subtype the case-fatality rate is usually under 2%, but lasting sequelae are reported in 20–40% of symptomatic cases and persistent neurologic deficits in up to 10%; most patients nonetheless recover fully. The Far Eastern subtype circulating in Russia and parts of Asia is substantially more severe (case-fatality up to 20–40%) and is outside the European framing used here. This entry does not constitute medical advice.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"A flat vector illustration of a small tick on a leaf beside a single vaccine vial, against a muted forest-toned background."},"canonical_url":"https://likelier.app/tick-borne-encephalitis","api_url":"https://likelier.app/api/fears/tick-borne-encephalitis.json"},{"slug":"cyclist-car-collision-death","question":"What are the odds of being killed by a motor vehicle while cycling?","quick_answer":"For a regular road or urban cyclist (about 2,000 miles a year over 40 years), the lifetime risk of being killed by a motor vehicle is roughly 1 in 158. That is serious, but far below the near-certainty many riders feel, and it scales with how many miles you actually ride.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Cyclists who ride regularly report that being hit by a car is the dominant fear shaping their route choices, riding posture, and willingness to cycle at all. The mental model is roughly: drivers are distracted, roads are narrow, one mistake ends everything. What the intuition rarely tracks is the difference between per-trip risk, per-mile risk, and lifetime risk for a person who cycles a specific number of miles per year. The fear operates at the level of vividness — one fatal dooring or overtaking collision is easy to imagine — rather than at the level of frequency.\n","rough_estimate":"feels very high — many regular cyclists consider it a near-certainty over a cycling lifetime","kind":"intuition"},"native":{"display":"~7.9 deaths per 100 million miles cycled (US, NTSB 2019 study)","numerator":8,"denominator":100000000,"unit":"per mile cycled","population":"US pedalcyclists in motor vehicle traffic crashes (NHTSA FARS, NTSB 2019 study)"},"normalized":{"lifetime_us_adult":0.0063,"display":"~1 in 158 lifetime (regular road/urban cyclist, ~2,000 mi/yr × 40 years)","log_value":-2.2,"assumptions":"The NTSB 2019 safety study reported a yearly US bicyclist fatality rate of 79 bicyclist fatalities per billion miles bicycled (7.9 per 100 million miles), citing Santacreu (2018). NHTSA FARS recorded 1,166 pedalcyclist fatalities in motor vehicle traffic crashes in 2023 (up about 4% from 1,117 in 2022, the highest count in the series), consistent with the ~1,000–1,200 range across 2019–2023. Normalizing to a \"regular road or urban cyclist\" who averages 2,000 miles per year over 40 active cycling years (ages 18–58) gives a lifetime exposure of 80,000 miles. At 7.9 deaths per 100 million miles, that yields a lifetime probability of 80,000 × 7.9e-8 ≈ 0.0063, or about 1 in 158. The uncertainty band reflects variation in annual mileage and trip type: a casual cyclist doing 1,000 miles/year for 30 years (30,000 miles) sits near 0.0024; a high-mileage road cyclist doing 4,000 miles/year for 50 years (200,000 miles) sits near 0.016. The per-mile fatality rate for cyclists is several times that of motor vehicle occupants, reflecting the lack of occupant protection, not a fundamentally more dangerous activity per trip — most Americans cycle very few miles annually.\n","uncertainty":{"low":0.0024,"high":0.016},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.ntsb.gov/safety/safety-studies/Documents/SS1901.pdf","title":"Bicyclist Safety on US Roadways: Crash Risks and Countermeasures (SS-19/01)","publisher":"National Transportation Safety Board (NTSB)","source_type":"govt_report","statistic":"79 bicyclist fatalities per billion miles bicycled in the US (7.9 per 100 million miles); 806 bicyclists died in crashes with motor vehicles in 2017; bicyclist fatalities reached their lowest level of 623 in 2010","excerpt":"\"Santacreu (2018) provided bicyclist fatalities data per distance bicycled for 13 European countries and the United States from 2011 through 2015. Figure 4 shows that the yearly US bicyclist fatality rate of 79 bicyclist fatalities per billion miles bicycled ranked second only to Italy.\"\n","source_date":"2019-11-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260212011120/https://www.ntsb.gov/safety/safety-studies/Documents/SS1901.pdf","calculation_notes":"The NTSB figure of 79 deaths per billion miles (= 7.9 per 100 million miles) is the primary exposure-normalized rate used for the native statistic. To convert to a lifetime probability for a regular cyclist, multiply the per-mile rate by total lifetime miles: 80,000 miles (2,000/year × 40 years) × 7.9e-8 per mile ≈ 0.0063. This report was also the NTSB's first comprehensive bicyclist-safety analysis since 1972 — a 47-year gap the report itself notes — which is why the body text describes it as \"nearly five decades\" since the prior review.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813739","title":"Traffic Safety Facts 2023 Data: Bicyclists and Other Cyclists (DOT HS 813 739)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"1,166 pedalcyclist fatalities in motor vehicle traffic crashes in 2023, representing 2.9% of all traffic fatalities; 81% occurred in urban areas; 62% at non-intersection locations; 87% of those killed were male","excerpt":"\"In 2023 there were 1,166 pedalcyclist fatalities, accounting for 2.9 percent of all traffic fatalities.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20251008191200/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813739","calculation_notes":"NHTSA FARS 2023 provides the most recent annual fatality count. The 1,166 figure is used to confirm that the NTSB per-mile rate remains plausible: if US cyclists collectively travel ~18 billion miles/year (rough estimate from American Community Survey trip data and NHTS), then 1,166 / 18,000,000,000 ≈ 6.5 deaths per 100 million miles, consistent with the NTSB figure. NHTSA FARS is the upstream source for all US traffic fatality counts and is treated as the primary authority.\n"},{"url":"https://www.iihs.org/topics/fatality-statistics/detail/bicyclists","title":"Fatality Facts 2023: Bicyclists","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"1,163 bicyclists killed in crashes with motor vehicles in 2023; about 2%-3% of motor vehicle crash deaths are bicyclists each year; deaths among bicyclists age 20 and older have increased almost fivefold since 1975","excerpt":"\"A total of 1,163 bicyclists were killed in crashes with motor vehicles in 2023. ... Each year about 2%-3% of motor vehicle crash deaths are bicyclists.\"\n","source_date":"2024-12-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250531072045/https://www.iihs.org/topics/fatality-statistics/detail/bicyclists","calculation_notes":"IIHS uses a slightly different coding methodology than NHTSA FARS, producing a 2023 figure of 1,163 vs NHTSA's 1,166. Both are used as corroboration that the annual US cyclist fatality count from motor vehicle crashes is reliably in the 1,100–1,200 range in recent years. The IIHS trend data also establishes that fatality counts have risen substantially since 2010 (from a low near 620), which means the per-mile rate used may slightly understate current risk if cycling miles have not grown proportionally.\n"}],"comparison_anchors":[{"label":"Death as a pedestrian hit by a motor vehicle (lifetime, US adult)","lifetime_us_adult":0.0076},{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Death on a motorcycle (lifetime, US adult, regular rider)","lifetime_us_adult":0.00144}],"personal_factor_multipliers":[{"factor":"casual cyclist, <500 miles/year, mostly off-road paths","multiplier":0.2,"notes":"Low total mileage on routes with reduced motor vehicle exposure."},{"factor":"regular urban commuter cyclist, ~2,000 miles/year","multiplier":1,"notes":"Baseline for this entry's headline number."},{"factor":"high-mileage road cyclist, 4,000+ miles/year on open roads","multiplier":2.5,"notes":"Higher exposure on arterial roads with faster-moving traffic."},{"factor":"cycling in the dark without lights on arterial roads","multiplier":3,"notes":"53% of pedalcyclist fatalities occur in dark conditions (NHTSA 2023); lighting and hi-vis clothing directly reduce this multiplier."}],"short_label":"Cyclist killed by car","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The per-mile fatality rate from the NTSB 2019 study (79 deaths per billion miles) is the best available US figure but relies on bicycle VMT estimates that have significant uncertainty — the US does not systematically measure bicycle miles traveled the way it measures motor vehicle miles. NHTSA's annual fatality count is accurate (FARS captures all motor-vehicle-involved deaths), but the denominator (total bike miles) is estimated from travel surveys with wide confidence intervals. The lifetime calculation is also sensitive to assumed annual mileage: doubling the assumed mileage doubles the lifetime probability. Cyclists who primarily ride on protected infrastructure (separated bike lanes, off-road trails) face lower exposure than those riding on open roads. Alcohol involvement (rider or driver) is reported in 34% of fatal crashes, indicating the underlying distribution is not uniform. The 1 in 158 headline applies to a specific profile; it is not a population-average figure for all US adults, which would be much lower because most Americans cycle very infrequently.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A lone bicycle wheel casting a long shadow on an empty road surface, flat vector illustration with muted tones."},"canonical_url":"https://likelier.app/cyclist-car-collision-death","api_url":"https://likelier.app/api/fears/cyclist-car-collision-death.json"},{"slug":"gambling-addiction-financial-ruin","question":"What are the odds of financial ruin from a gambling addiction?","quick_answer":"The lifetime odds of severe financial ruin from a gambling disorder work out to roughly 1 in 160. About 1.4% of adults have a gambling disorder, close to the 1-2% most people vaguely assume, but the financial-ruin tail gets far less attention than the addiction itself.\n","category":"other","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Gambling addiction is generally perceived as a niche problem affecting a small, identifiable population — stereotypically older men in casinos. The rapid expansion of legal sports betting since the 2018 Supreme Court decision in Murphy v. NCAA has increased awareness but not proportionally increased perceived personal risk. Most adults who gamble occasionally consider themselves immune to disorder, and the marketing of sports betting as entertainment and skill-based activity further normalizes the behavior. Public health campaigns around problem gambling remain underfunded relative to those for alcohol and drug addiction, and the financial consequences — as opposed to the psychological ones — receive comparatively little media attention.\n","rough_estimate":"~1-2% of adults affected","kind":"intuition"},"native":{"display":"~1.4% of adults globally experience gambling disorder or problematic gambling (Lancet Public Health 2024 meta-analysis)","numerator":1.4,"denominator":100,"unit":"share of adults with gambling disorder (current prevalence, DSM-5)","population":"global adults (Lancet Public Health 2024 systematic review and meta-analysis, 2010-2024 studies)"},"normalized":{"lifetime_us_adult":0.0063,"display":"~1 in 160 lifetime risk of severe financial harm from gambling disorder","log_value":-2.2,"assumptions":"The 2024 Lancet Public Health systematic review and meta-analysis (studies from 2010-2024) found that approximately 1.41% of adults experience gambling disorder or problematic gambling (current prevalence). The National Council on Problem Gambling estimates that roughly 1% of US adults meet criteria for a severe gambling problem in a given year and another 2-3% have mild-to-moderate problems. Because gambling disorder is chronic and recurrent, we take the ~2-3% \"any gambling problem\" tier as an approximate share of adults affected over a working lifetime; projecting an annual/current prevalence onto a lifetime basis is a modeling assumption, not a directly cited lifetime figure. Among those who develop gambling disorder, the PMC bankruptcy study (Journal of Gambling Studies) finds 19-23% file for bankruptcy (19.2% and 22.8% in two samples), with average gambling-related debts of $42,000-$53,000. Using bankruptcy filing as the most rigorously measured severe-harm indicator (rather than summing overlapping debt categories): ~2.5% of adults affected x ~25% severe financial harm rate (bankruptcy or equivalent financial ruin) = ~0.63% (0.0063), or approximately 1 in 160 US adults. The uncertainty range spans 0.3% (lower-bound ~1.6% affected x 19% bankruptcy-only) to 1.2% (upper-bound ~3% affected x ~35% including bankruptcy plus non-overlapping extreme debt).\n","uncertainty":{"low":0.003,"high":0.012},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ncpgambling.org/training/ngage-survey/ngage-3/","title":"NGAGE 3.0: National Survey on Gambling Attitudes and Gambling Experiences (2024)","publisher":"National Council on Problem Gambling","source_type":"primary_study","statistic":"8% of American adults reported at least one indicator of problematic gambling behavior; approximately 2-3 million adults likely meet criteria for gambling disorder","excerpt":"\"About 8% of American adults — nearly 20 million people — reported experiencing at least one indicator of problematic gambling behavior. Approximately 2-3 million adults likely meet criteria for gambling disorder.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260324143829/https://www.ncpgambling.org/training/ngage-survey/ngage-3/","calculation_notes":"IMPORTANT: NGAGE 3.0 explicitly states it is \"NOT a study of the prevalence of gambling addiction\" and was not designed to assess DSM-5 gambling disorder prevalence. It measures gambling participation, attitudes, and behavioral indicators. The ~1% and 2-3 million figures are the survey's own estimates of likely disorder but are not derived from clinical diagnostic instruments. This source is used here for context on gambling participation and behavioral indicators, not as the primary prevalence figure. The primary prevalence estimate comes from the 2024 Lancet Public Health systematic review and meta-analysis.\n","independence_note":"NGAGE is a nationally representative survey conducted by the NCPG using an independent research panel (Ipsos), methodologically distinct from clinical prevalence studies and from administrative gambling industry data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2824911/","title":"Pathological Gambling and Bankruptcy","publisher":"Journal of Gambling Studies / PMC","source_type":"peer_reviewed","statistic":"19.2% of pathological gamblers filed for bankruptcy (NORC Gambling Impact and Behavior Study); 22.8% in a later clinical sample","excerpt":"\"The Gambling Impact and Behavior Study found that 19.2% of pathological gamblers had filed bankruptcy, compared to 5.5% of low-risk gamblers. A later study found that 22.8% of pathological gamblers had declared bankruptcy with an average debt of $53,103.\"\n","source_date":"2010-03-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260426201428/https://pmc.ncbi.nlm.nih.gov/articles/PMC2824911/","calculation_notes":"The bankruptcy rate among pathological gamblers (19-23%) is roughly 4x the general population rate (~5%). This study provides the key bridge between gambling disorder prevalence and financial ruin: if ~2.5% of adults develop gambling disorder over a lifetime and ~20-25% of those experience severe financial harm (bankruptcy or equivalent), that yields ~0.5-0.63% lifetime risk of gambling-related financial ruin. The debt categories (33% with $10k-$50k, 21% with $50k-$100k) overlap substantially with the bankruptcy group and with each other, so they cannot be summed to produce an independent severe-harm rate.\n","independence_note":"This peer-reviewed analysis uses data from the NORC Gambling Impact and Behavior Study (commissioned by the National Gambling Impact Study Commission), independent from the NCPG's NGAGE survey methodology.\n"},{"url":"https://www.ncpgambling.org/help-treatment/faqs-what-is-problem-gambling/","title":"FAQs: What is Problem Gambling?","publisher":"National Council on Problem Gambling","source_type":"reputable_reference","statistic":"An estimated 2.5 million US adults meet criteria for severe gambling problems; another 5-8 million have mild or moderate problems","excerpt":"\"An estimated 2.5 million U.S. adults (1% of the population) are estimated to meet criteria for severe gambling problems. Another 5-8 million (2-3%) would be considered to have mild or moderate gambling problems.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426201312/https://www.ncpgambling.org/help-treatment/faqs-what-is-problem-gambling/","calculation_notes":"The FAQ page provides prevalence data consistent with the NGAGE 3.0 survey: ~1% severe, 2-3% mild-to-moderate. The frequently cited \"only 8% seek help\" figure does not appear on this page and is contradicted by the Bijker et al. (2022) systematic review and meta-analysis in Addiction, which found that approximately 20.6% of people with problem gambling seek help (1 in 5). The low but non-trivial treatment-seeking rate implies that the majority of gambling-related financial harm goes unaddressed until it reaches crisis levels (bankruptcy, divorce, criminal charges).\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Student loan default (US borrowers)","lifetime_us_adult":0.26},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"States with legal sports betting (38 states, 2025)","probability":0.008,"notes":"Legal sports betting expansion since 2018 has increased accessibility; early data suggests higher problem gambling rates in states with mobile betting"},{"region":"States without legal sports betting","probability":0.005,"notes":"Casino and lottery gambling still present but lower overall accessibility"}],"personal_factor_multipliers":[{"factor":"male, age 18-34","multiplier":3,"notes":"Young men are the demographic most heavily targeted by sports betting marketing and have the highest rates of gambling disorder onset"},{"factor":"history of substance use disorder","multiplier":4,"notes":"Gambling disorder co-occurs with substance use disorders at high rates; shared neurobiological pathways"},{"factor":"household income above $75,000","multiplier":0.6,"notes":"Higher-income gamblers can absorb larger losses before reaching financial ruin, though they are not immune to disorder"},{"factor":"daily sports bettor","multiplier":5,"notes":"Frequency of gambling is the strongest behavioral predictor of disorder development"}],"short_label":"Gambling financial ruin","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 0.63% lifetime estimate combines epidemiological prevalence data with financial-harm research that spans different eras of gambling availability. The post-2018 sports betting expansion is too recent to have produced 20-year longitudinal data, so the lifetime impact of ubiquitous mobile betting is necessarily speculative. The NCPG's NGAGE 3.0 actually showed a slight decline in problematic gambling indicators from 2021 to 2024, which the NCPG attributes to novelty effects wearing off and improved responsible-gambling tools. Bijker et al. (2022) found that approximately 20% of people with problem gambling seek help — low but substantially higher than the frequently cited 8% figure. Financial ruin — as opposed to clinical diagnosis — is not systematically tracked. \"Financial ruin\" is not a clinical term and is defined here as bankruptcy, loss of home, or debt exceeding annual income — a threshold that captures the severe tail but not the broader population experiencing moderate gambling-related financial stress. Note that NGAGE 3.0 explicitly states it is \"NOT a study of the prevalence of gambling addiction\" — it measures gambling participation and behavioral indicators, not DSM-5 disorder prevalence.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A pair of dice casting a long shadow over an empty wallet, muted grey and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/gambling-addiction-financial-ruin","api_url":"https://likelier.app/api/fears/gambling-addiction-financial-ruin.json"},{"slug":"schizophrenia-first-episode","question":"What is the lifetime risk of experiencing a first psychotic episode or schizophrenia?","quick_answer":"About 1 in 143 US adults (roughly 0.5-0.7%) develop schizophrenia in their lifetime. That matches the folk sense that it is rare, though people underestimate the broader category of psychotic experiences, which surveys put at 3-5%.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Most people carry a vague sense that schizophrenia is rare, severe, and happens to other people — the kind of illness they associate with news stories about violence or homelessness, not with anyone in their social circle. The folk estimate tends to land somewhere around \"very unlikely,\" which is directionally correct for narrow-definition schizophrenia (~0.5- 0.7%) but underestimates the broader category of psychotic experiences, which epidemiological surveys place at 3-5% lifetime. The stigma attached to psychosis suppresses disclosure, reinforcing the perception that it is rarer than it actually is.\n","kind":"intuition"},"native":{"display":"~0.5-0.7% lifetime prevalence of schizophrenia in the general population","numerator":7,"denominator":1000,"unit":"lifetime","population":"general adult population"},"normalized":{"lifetime_us_adult":0.007,"display":"~1 in 143 lifetime (US adult)","log_value":-2.15,"assumptions":"The 2026 systematic review and meta-analysis in Molecular Psychiatry (Jongsma et al.) reports a pooled lifetime prevalence of 0.62% (95% CI: 0.51%-0.76%) across 60 general-population studies in 24 countries. NIMH cites an international prevalence range of 0.33%-0.75% for schizophrenia among non-institutionalised persons, noting that institutionalised and homeless populations are under-counted in household surveys. The 0.7% (1 in 143) point estimate used here sits at the upper end of the NIMH range and within the confidence interval of the meta-analysis, reflecting the likelihood that true prevalence exceeds household-survey estimates because of under-enumeration of affected individuals in institutional settings, shelters, and on the streets. This is a lifetime prevalence figure, not an annual rate compounded — schizophrenia onset is heavily concentrated in adolescence through the early thirties, making annual compounding inappropriate. Uncertainty band: low 0.004 anchors to NIMH's 0.33% floor for non-institutionalised populations; high 0.012 reflects a plausible ceiling once the systematically excluded institutionalised, incarcerated, and homeless populations are accounted for.\n","uncertainty":{"low":0.004,"high":0.012},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/schizophrenia","title":"Schizophrenia — Statistics","publisher":"National Institute of Mental Health (NIMH)","source_type":"govt_report","statistic":"The international prevalence of schizophrenia among non-institutionalized persons is 0.33% to 0.75%","excerpt":"\"Precise prevalence estimates of schizophrenia are difficult to obtain due to clinical and methodological factors such as the complexity of schizophrenia diagnosis, its overlap with other disorders, and varying methods for determining diagnoses. The international prevalence of schizophrenia among non-institutionalized persons is 0.33% to 0.75%. Individuals with schizophrenia and other psychotic disorders may be under-counted in prevalence estimation studies.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260411094957/https://www.nimh.nih.gov/health/statistics/schizophrenia","calculation_notes":"NIMH provides the 0.33%-0.75% range for non-institutionalised persons, explicitly noting under-counting of institutionalised and homeless individuals. This establishes the floor for the true US prevalence. The estimated average potential life lost of 28.5 years for individuals with schizophrenia in the US contextualises the severity. Used here as the authoritative US-government source anchoring the lower bound.\n"},{"url":"https://www.nature.com/articles/s41380-026-03533-3","title":"Global lifetime prevalence of schizophrenia: A systematic review and meta-analysis","publisher":"Molecular Psychiatry / Nature","source_type":"peer_reviewed","statistic":"Pooled lifetime prevalence of schizophrenia in the general population: 0.62% (95% CI: 0.51%-0.76%)","excerpt":"\"The lifetime prevalence of schizophrenia is 0.62% (95% CI [0.51%–0.76%]) in the general population. Asia has the lowest lifetime prevalence of schizophrenia in the general population, at 0.47% (95% CI [0.35%–0.64%]). Schizophrenia is more prevalent in special populations than in the general population.\"\n","source_date":"2026-03-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260505062513/https://www.nature.com/articles/s41380-026-03533-3","calculation_notes":"This 2026 meta-analysis is the most comprehensive pooled estimate available, drawing on 60 general-population studies across 24 countries with a combined sample of over 20 million individuals. The 0.62% central estimate and 0.51%-0.76% confidence interval provide the statistical backbone for the normalized figure. The point estimate of 0.7% used for lifetime_us_adult sits at the upper edge of this CI, justified by NIMH's observation that US household surveys under-count institutionalised and homeless individuals with schizophrenia.\n","independence_note":"Methodologically independent of NIMH's range: the meta-analysis pools primary epidemiological studies worldwide using a systematic search protocol, whereas NIMH's figure synthesises US-focused surveillance and review literature.\n"}],"comparison_anchors":[{"label":"Lifetime major depression (US adult)","lifetime_us_adult":0.208},{"label":"Lifetime bipolar disorder (US adult)","lifetime_us_adult":0.028},{"label":"Lifetime any psychotic experience (general population)","lifetime_us_adult":0.05}],"short_label":"Schizophrenia","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The 0.7% figure covers schizophrenia specifically, not all psychotic disorders. Broader definitions that include schizoaffective disorder, brief psychotic episodes, and substance-induced psychosis push the lifetime rate of any psychotic experience to 3-5%. The normalised figure uses a lifetime prevalence, not an annual rate compounded over 59 years, because schizophrenia onset is concentrated in a narrow age window (late teens to early thirties) and recurrence is not well-modelled by independent annual draws. Household surveys systematically under-count affected individuals who are institutionalised, incarcerated, or homeless, so the true prevalence almost certainly exceeds the survey-derived estimates. Risk is not uniformly distributed: urban birth, migration, cannabis use during adolescence, and family history each elevate individual risk substantially. The 28.5-year average life-years-lost figure cited by NIMH reflects both the severity of the condition and the downstream effects on physical health, social integration, and access to care.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of an abstract silhouette with fragmented geometric shapes suggesting disrupted perception, in muted tones."},"canonical_url":"https://likelier.app/schizophrenia-first-episode","api_url":"https://likelier.app/api/fears/schizophrenia-first-episode.json"},{"slug":"accidental-fall-death","question":"What are the odds of dying from an accidental fall?","quick_answer":"About 1 in 135 over a full adult lifetime — but nearly all of that risk lands after age 75. Most adults under 65 assume it's essentially negligible, which is roughly right for them personally and badly wrong for the population as a whole.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Fear of falling is a well-studied construct in geriatric medicine — \"ptophobia,\" measured by instruments like the Falls Efficacy Scale — but it centers on fear of falling and injuring oneself, not specifically on fear of falling to death. No major public-opinion survey isolates \"fear of dying from an accidental fall\" as its own item. The practical intuition most non-elderly adults carry is that falls are embarrassing rather than lethal: a cause of broken wrists, bruised egos, and hospital bills, but not of the sort of death that shows up on the actuarial leaderboard. That intuition is badly wrong in aggregate and roughly right for any particular under-65 reader, which is the whole story of this page.\n","rough_estimate":"Most adults under 65 assume lifetime fall-death risk is essentially negligible","kind":"intuition"},"native":{"display":"~11.4 unintentional fall deaths per 100,000 per year (US, 2023)","numerator":1,"denominator":8772,"unit":"per year","population":"US residents, all ages, age-adjusted, unintentional falls (ICD-10 W00-W19)"},"normalized":{"lifetime_us_adult":0.0074,"display":"1 in ~135 lifetime (US adult)","log_value":-2.13,"assumptions":"Uses the NCHS age-adjusted unintentional fall death rate of 11.4 per 100,000 per year for 2023, applied across 59 years of remaining adult life. Flat-rate compounding gives 1 - (1 - 1.14e-4)^59 ≈ 6.7e-3, or about 1 in 149. Adjusted slightly upward to 7.4e-3 (1 in ~135) because the age-adjusted rate understates cumulative lifetime risk for a cohort that will actually pass through the 75+ and 85+ age bands, where rates are 75 and 340 per 100,000 respectively. The uncertainty band is wide to reflect that the true lifetime figure is very sensitive to how long the cohort lives and how risk scales with age in future decades — both of which are trending worse, not better.\n","uncertainty":{"low":0.0055,"high":0.011},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db532.htm","title":"Unintentional Fall Deaths in Adults Age 65 and Older: United States, 2003-2023","publisher":"CDC National Center for Health Statistics, Data Brief No. 532","source_type":"govt_report","statistic":"US unintentional fall death rate for adults 65+ was 69.9 per 100,000 in 2023; 339.5 per 100,000 for age 85+","excerpt":"\"In 2023, the unintentional fall death rate for adults age 65 and older was 69.9 per 100,000 population. [...] Rates of unintentional fall deaths increased between 2003 and 2023 for men and women ages 65-74, 75-84, and 85 and older.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413162844/https://www.cdc.gov/nchs/products/databriefs/db532.htm","calculation_notes":"NCHS reports an age-65+ fall death rate of 69.9 per 100,000 in 2023, with extreme age-stratification: 19.2 (65-74), 74.7 (75-84), and 339.5 (85+). These rates confirm that the all-ages age-adjusted figure of ~11.4/100,000 is a flat average across a profoundly non-flat distribution — the true cohort-lifetime risk concentrates in the last decade of life. Used here to justify the upward adjustment from the naive flat-compounded estimate of 1 in 149 to roughly 1 in 135, and to frame the heterogeneity caveat for under-65 readers.\n","independence_note":"NCHS Data Brief 532 draws from the NVSS death-certificate pipeline (ICD-10 W00-W19 unintentional fall codes). Shares the same NVSS upstream as the NCHS injury-trends brief below — treat the two as alternative analytical slices of one underlying dataset rather than independent counts.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK613832/","title":"Trends in Death Rates for Leading Methods of Injury: United States, 2003-2023","publisher":"CDC NCHS Data Brief (NCBI Bookshelf)","source_type":"govt_report","statistic":"US fall death rate rose from 6.0 per 100,000 (2003) to 11.4 per 100,000 (2021-2023)","excerpt":"\"Death rates from falls increased 90% from 2003 (6.0) to 2021 (11.4) and then remained stable through 2023.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413162918/https://www.ncbi.nlm.nih.gov/books/NBK613832/","calculation_notes":"This is the age-adjusted all-ages death rate used as the direct input to the normalized lifetime estimate. 1.14e-4 per year compounded over 59 years gives 1 - (1 - 1.14e-4)^59 ≈ 6.72e-3, rounded up to 7.4e-3 to reflect the age concentration above. Corroborates the independent 2022 NCHS figure that 47,984 Americans died from falls that year, of which 97.2% were unintentional.\n","independence_note":"Shares the NCHS National Vital Statistics System mortality files with Data Brief 532, so the two sources are not statistically independent — they are two different analytical slices of the same underlying death-certificate data. Treated here as a single authoritative pipeline with two different cuts (all-ages trend vs. 65+ detail) rather than two independent estimates.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Age 65+ vs under 65","multiplier":7,"notes":"CDC WISQARS 2022 / NCHS Data Brief 532 (2025): the unintentional fall death rate for adults 65+ was 69.9 per 100,000 vs the all-ages age-adjusted rate of ~11.4 per 100,000. At age 85+ the rate reaches 339.5 per 100,000 — roughly 30× the under-65 baseline. The 65+ multiplier of ~6-7× is a conservative midpoint."},{"factor":"Prior fall in the previous 12 months","multiplier":2.5,"notes":"Rubenstein LZ (2006, Age and Ageing): a prior fall is one of the strongest predictors of future falls; odds ratio for recurrence is approximately 2-3× in prospective studies. This is the top modifiable risk factor cited in American Geriatrics Society fall-prevention guidelines."},{"factor":"Cognitive impairment or dementia","multiplier":2,"notes":"American Geriatrics Society / British Geriatrics Society Clinical Practice Guideline for Prevention of Falls (2011, updated 2019): people with dementia fall 2× more frequently than cognitively intact peers of the same age, and sustain higher injury severity per fall due to reduced protective reflexes."},{"factor":"Balance-affecting medications (benzodiazepines, alpha-blockers, sedatives)","multiplier":1.7,"notes":"Woolcott et al. (2009, Archives of Internal Medicine): meta-analysis of 22 studies found psychotropic drugs (OR 1.73), sedatives/hypnotics (OR 1.54), and antihypertensives (OR 1.24) significantly increased fall risk. Overall medication-class OR range 1.6-2× for the highest-risk drug categories."}],"short_label":"Accidental fall","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The normalized figure is an average across the US adult population and should not be read as a personal probability for a healthy 35-year-old. Unintentional fall deaths concentrate overwhelmingly in adults 75 and older: the rate at age 85+ (339.5 per 100,000) is roughly eighteen times the rate at 65-74 (19.2 per 100,000) and more than a hundred times the rate in middle-aged adults. A reader currently under 65 who does not reach old age will accrue only a small fraction of the headline 1-in-135 risk; a reader who lives into their late eighties will accrue several times it. This cause is also one of the only major injury categories whose rate is still rising, not falling, driven primarily by population aging and possibly by medication burden in older adults. Excludes intentional falls (suicide by jumping, coded separately) and falls from transport vehicles.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale step or shallow stair edge against a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/accidental-fall-death","api_url":"https://likelier.app/api/fears/accidental-fall-death.json"},{"slug":"pancreatic-cancer","question":"What are the odds of dying from pancreatic cancer?","quick_answer":"The odds of dying from pancreatic cancer are roughly 1 in 125 over a global adult lifetime (about 1 in 71 for US adults). Unusually, the dread attached to it is fairly calibrated — it kills a larger share of those it strikes (~85–90% case-fatality) than any other common solid tumor.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Pancreatic cancer occupies an unusual position in the public mental model of cancer: most readers do not rank it among the most common killers by raw numbers (correctly — it is not), but most readers do file it as \"the bad one\", the diagnosis you do not want to hear. That intuition is approximately right. Pancreatic cancer kills fewer people each year than lung, colorectal, liver, stomach, or breast cancer, but it kills a much larger fraction of the people it diagnoses: five-year relative survival sits around 13% in the US and under 12% globally, the worst of any common solid tumor. The fear attached to the word \"pancreatic\" is a reasonably calibrated response to a genuinely grim case-fatality rate.\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); pancreatic is widely perceived as the most lethal subtype despite lower incidence","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~467,000 pancreatic cancer deaths per year globally (~511,000 new cases)","numerator":1,"denominator":17100,"unit":"per year","population":"global, all ages, pancreatic cancer only"},"normalized":{"lifetime_us_adult":0.008,"display":"~1 in 125 lifetime (global adult)","log_value":-2.097,"assumptions":"Starts from the IARC GLOBOCAN 2022 global pancreatic cancer headline: more than 500,000 new cases and almost 470,000 deaths in 2022 worldwide, making pancreatic cancer the 12th most common cancer but the 6th most common cause of cancer death — a ranking mismatch that directly reflects the aggressive case-fatality rate. Spread across a global adult population of ~5.5 billion (age 18+), ~467,000 pancreatic cancer deaths per year is ~0.85 per 10,000 adults per year. Naive 60-year compounding gives ~0.5%; age-weighting (pancreatic cancer mortality is heavily concentrated in the 60-80 band, with hazard several times higher in the last third of adult life than at the population average) pulls the realistic global adult lifetime figure up to roughly 0.8%. The direct US number from ACS is higher: lifetime risk of developing pancreatic cancer is ~1 in 56 for men and ~1 in 60 for women, and with 5-year relative survival of ~13% the implied long-run case-fatality is ~85-90%, giving a US lifetime pancreatic-cancer-death probability of ~1.4% (~1 in 71). Headline figure 0.008 (~1 in 125) with an uncertainty band of 0.005-0.015 to span the global-adult to US-adult range. Scope is global-adult-lifetime to match the cancer-lifetime parent entry; the US row in regional_breakdown anchors the top of the band.\n","uncertainty":{"low":0.005,"high":0.015},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/cancer-type/pancreatic-cancer/","title":"Pancreatic Cancer","publisher":"International Agency for Research on Cancer (IARC) / World Health Organization","source_type":"govt_report","statistic":"More than 500,000 pancreatic cancer cases diagnosed in 2022 and almost 470,000 deaths globally; 12th most common cancer but 6th most common cause of cancer death; one of the cancer types with the least favourable prognosis","excerpt":"\"more than 500 000 people estimated to have been diagnosed with pancreatic cancer in 2022 [...] almost 470 000 deaths in 2022 [...] only the 12th most common cancer type globally [...] the sixth most common cause of cancer death [...] It is one of the cancer types with the least favourable prognosis.\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413181017/https://www.iarc.who.int/cancer-type/pancreatic-cancer/","calculation_notes":"GLOBOCAN 2022 headline used directly as the native number. ~467,000 annual global pancreatic cancer deaths across ~5.5 billion adults is ~0.85 per 10,000 adult-years. Age-weighted over a 60-year adult window — pancreatic cancer hazard in the 60s and 70s is several times the adult average — gives a lifetime adult-lifetime mortality near 0.008 (~1 in 125). The ranking mismatch (12th in incidence, 6th in mortality) is the core quantitative statement of the \"diagnosis ≈ death\" story and is used directly in the long-form body text.\n","independence_note":"IARC GLOBOCAN is the upstream dataset used by WHO, ACS international comparisons, and the IHME Global Burden of Disease pancreatic cancer module. Treat this as the canonical global source; the SEER and ACS US numbers below are methodologically independent cross-checks.\n"},{"url":"https://seer.cancer.gov/statfacts/html/pancreas.html","title":"Cancer Stat Facts: Pancreatic Cancer","publisher":"Surveillance, Epidemiology, and End Results (SEER) Program, National Cancer Institute","source_type":"govt_report","statistic":"5-year relative survival 13.3% (2015-2021); ~1.6% of men and women will be diagnosed with pancreatic cancer at some point during their lifetime; estimated 67,440 new cases and 51,980 deaths in 2025; 3.3% of all new cancer cases; stage distribution at diagnosis is 15% localized, 28% regional, 51% distant, 5% unknown","excerpt":"\"5-Year Relative Survival: 13.3% [...] Approximately 1.6 percent of men and women will be diagnosed with pancreatic cancer at some point during their lifetime [...] Estimated New Cases in 2025: 67,440 [...] Estimated Deaths in 2025: 51,980 [...] 3.3% of all new cancer cases. [...] For pancreatic cancer, 15.4% are diagnosed at the local stage. [...] Stage Distribution (%): Localized 15 [...] Regional 28 [...] Distant 51 [...] Unknown 5.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260326014754/https://seer.cancer.gov/statfacts/html/pancreas.html","calculation_notes":"SEER gives direct lifetime incidence of ~1.6% for the US population. With 5-year relative survival at 13.3% — the lowest of any common solid tumor — the implied long-run case-fatality is roughly 85-90%. 1.6% lifetime incidence × ~88% long-run case-fatality yields a US lifetime pancreatic-cancer-death probability of ~1.4% (~1 in 71), consistent with the ACS lifetime-risk page. The ratio of annual deaths (~52,000) to annual new cases (~67,000) is ~0.78 on a single-year basis, which understates the long-run case-fatality because most deaths occur outside the diagnosis year. This anchors the US row in the regional breakdown and the top of the Likelier uncertainty band. The same SEER page's stage-distribution table shows 51% of cases are diagnosed at the distant (metastatic) stage versus only 15% localized — the direct source for the body text's \"roughly half of cases present\" at the distant stage, and the diagnosis-time complement to the stage-conditional survival figures in the ACS survival-rates source below.\n","independence_note":"SEER (NCI) and IARC GLOBOCAN (WHO) are methodologically independent compilation pipelines. SEER uses US vital registration and population-based cancer registries; IARC aggregates national registry data worldwide.\n"},{"url":"https://www.cancer.org/cancer/types/pancreatic-cancer/about/key-statistics.html","title":"Key Statistics for Pancreatic Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Lifetime risk of pancreatic cancer ~1 in 56 in men and ~1 in 60 in women; ~67,530 new cases and ~52,740 deaths projected for 2026 in the US; ~3% of US cancers but ~8% of US cancer deaths","excerpt":"\"The average lifetime risk of pancreatic cancer is about 1 in 56 in men and about 1 in 60 in women. [...] About 67,530 people (35,190 men and 32,340 women) will be diagnosed with pancreatic cancer. [...] About 52,740 people (27,230 men and 25,510 women) will die of pancreatic cancer.\"\n","source_date":"2026-01-16","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420045225/https://www.cancer.org/cancer/types/pancreatic-cancer/about/key-statistics.html","calculation_notes":"ACS projects ~67,530 US new cases and ~52,740 US deaths in 2026 — a ratio of ~0.78 annual deaths to annual diagnoses, which is the single most compressed deaths-to-cases ratio of any common US cancer. The \"about 3% of all cancers but about 8% of all cancer deaths\" framing is used in the body text as the plain-English version of the GLOBOCAN \"12th in incidence, 6th in mortality\" ranking. The lifetime-risk-of- developing figures (~1 in 56 men, ~1 in 60 women) combined with the ~85-90% long-run case-fatality from SEER give the ~1.4% US lifetime pancreatic-cancer-death probability used as the US anchor.\n","independence_note":"ACS Key Statistics and SEER Stat Facts share the same US vital registration and cancer registry upstream. Treat as a single institutional pipeline for the US-specific figures; the IARC source is the methodologically independent global cross-check.\n"},{"url":"https://www.cancer.org/cancer/types/pancreatic-cancer/detection-diagnosis-staging/survival-rates.html","title":"Survival Rates for Pancreatic Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"5-year relative survival by stage (SEER 2015-2021): localized 44%, regional 17%, distant 3%, all stages combined 13%","excerpt":"\"Localized 44% [...] Regional 17% [...] Distant 3% [...] All SEER stages combined 13%.\"\n","source_date":"2025-03-14","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420045300/https://www.cancer.org/cancer/types/pancreatic-cancer/detection-diagnosis-staging/survival-rates.html","calculation_notes":"Stage-conditional 5-year survival figures used in the long-form body text. The gap between localized (44%) and distant (3%) is roughly 15x, the largest stage-conditional survival gap of any common cancer in the SEER system. Combined with the fact that only about 12-15% of pancreatic cancers are diagnosed at the localized stage (per SEER staging distribution), the \"no effective screening\" story falls out arithmetically: the tumor is almost always metastatic or locally advanced at diagnosis.\n","independence_note":"Draws on the same SEER staging distribution and relative survival pipeline as the SEER Stat Facts source above. Used as the stage- conditional cross-reference, not as an independent verification of the aggregate mortality figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24840647/","title":"Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States","publisher":"Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM / Cancer Research","source_type":"peer_reviewed","statistic":"Pancreas and liver cancers are projected to surpass breast, prostate, and colorectal cancers to become the second and third leading causes of cancer-related death in the US by 2030","excerpt":"\"pancreas and liver cancers are projected to surpass breast, prostate, and colorectal cancers to become the second and third leading causes of cancer-related death by 2030.\"\n","source_date":"2014-06-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260420045336/https://pubmed.ncbi.nlm.nih.gov/24840647/","calculation_notes":"Rahib et al. 2014 is the widely cited source for the \"second leading cause of US cancer death by 2030\" projection used in the body text. The projected rise in ranking is driven almost entirely by improvements in treating the currently-larger cancers (breast, prostate, colorectal) and by population ageing, not by deteriorating pancreatic outcomes — age-standardized pancreatic cancer mortality has been roughly flat for decades. Used as the \"trend\" paragraph anchor, not as part of the headline probability calculation.\n","independence_note":"Rahib et al. projects from SEER incidence and NCHS mortality data, so upstream is the same as the SEER/ACS sources. Independent analytically in the sense that it is a projection model rather than a retrospective count, but not an independent data pipeline.\n"}],"comparison_anchors":[{"label":"All-cancer death (global adult lifetime)","lifetime_us_adult":0.14},{"label":"Lung cancer death (global adult lifetime)","lifetime_us_adult":0.018},{"label":"Colorectal cancer death (global adult lifetime)","lifetime_us_adult":0.013},{"label":"Breast cancer death (global adult women lifetime)","lifetime_us_adult":0.016},{"label":"Stroke death (global adult lifetime)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.008,"notes":"~467,000 pancreatic cancer deaths/yr across ~8B people (IARC GLOBOCAN 2022); age-weighted adult lifetime figure"},{"region":"US adult","probability":0.014,"notes":"Derived from ~1.6% SEER lifetime incidence combined with ~85-90% long-run case-fatality; ~1 in 71"},{"region":"Europe","probability":0.015,"notes":"Incidence and mortality rates among the highest worldwide; Central and Eastern European countries sit at the top of the IARC pancreatic cancer ranking"},{"region":"Sub-Saharan Africa","probability":0.003,"notes":"Lower incidence; confounded by competing mortality (infectious disease, maternal, injury) removing adults from the denominator before peak pancreatic-cancer-risk age"}],"personal_factor_multipliers":[{"factor":"current smoker","multiplier":2,"notes":"Smoking roughly doubles pancreatic cancer risk; the association is weaker than for lung or bladder cancer but is the most important modifiable risk factor and accounts for a meaningful share of cases in high-income countries"},{"factor":"family history + genetic syndrome (BRCA2, Lynch, PRSS1, CDKN2A, Peutz-Jeghers)","multiplier":5,"notes":"Hereditary pancreatitis (PRSS1) and familial syndromes raise lifetime risk by ~5x on aggregate; individual variant risks range from ~2x (BRCA2) to >50x (PRSS1 with chronic pancreatitis). These are the primary groups where surveillance imaging is recommended."},{"factor":"type 2 diabetes (long-standing)","multiplier":1.8,"notes":"Long-standing diabetes is an established risk factor; new-onset diabetes in older adults can also be a presenting sign of pancreatic cancer rather than a cause, which complicates the epidemiology"},{"factor":"chronic pancreatitis","multiplier":13,"notes":"One of the strongest non-hereditary risk factors; the multiplier is highest in hereditary pancreatitis (PRSS1 carriers) where lifetime risk approaches 40%"},{"factor":"obesity BMI 35+","multiplier":1.5,"notes":"Severe obesity is associated with roughly 50% higher pancreatic cancer mortality; the mechanism is thought to involve insulin resistance and chronic low-grade inflammation"}],"short_label":"Pancreatic cancer","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Pancreatic cancer is the entry where the case-fatality rate, not the incidence, is the load-bearing number. Lifetime *incidence* in the US is ~1.6% per SEER — lower than breast, prostate, lung, or colorectal cancer by a wide margin. What makes pancreatic cancer distinctive is the compression between diagnosis and death: ~52,000 US deaths against ~67,000 US diagnoses in the same year gives a deaths-to-cases ratio approaching 0.8, the highest of any common cancer, and 5-year relative survival of 13% is the lowest in the SEER system for a solid tumor. The biology behind that compression is structural: the pancreas sits deep in the retroperitoneum behind the stomach, symptoms are vague and late, and only about 12-15% of tumors are caught at the localized stage where 5-year survival reaches 44%. There is currently no effective population-level screening program for average-risk adults — the USPSTF continues to recommend against routine screening — because the prevalence is too low and the available tests are not specific enough to avoid net harm. Surveillance imaging is only recommended for defined high-risk groups: hereditary pancreatitis, familial pancreatic cancer kindreds, and carriers of certain germline variants. The projected rise to the second leading cause of US cancer death by 2030 reflects improvements in treating the currently-larger cancers more than deteriorating pancreatic outcomes; age-standardized pancreatic cancer mortality has been roughly flat for decades.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small pale circle offset low against a large muted sand field, flat vector illustration."},"canonical_url":"https://likelier.app/pancreatic-cancer","api_url":"https://likelier.app/api/fears/pancreatic-cancer.json"},{"slug":"petrol-car-fire","question":"What are the odds of a petrol car catching fire spontaneously?","quick_answer":"A petrol car catches fire in roughly 0.8% of cases over a 12-year ownership span (about 69 fires per 100,000 registered vehicles a year). Most drivers guess lower (~0.1–0.5%) and wrongly assume EVs are the fire-prone ones — 95% of these fires involve no collision.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Vehicle fires barely register in most drivers' risk calculus. The cultural image of a car fire is a post-crash Hollywood fireball, not a parked sedan smouldering from an electrical short. Meanwhile, EV battery fires command disproportionate media coverage relative to their incidence. The result is an inverted perception: many drivers assume electric vehicles are uniquely fire-prone while treating the baseline ICE vehicle fire rate as negligible. Few car owners realize that mechanical and electrical failures in gasoline vehicles cause roughly 190,000 highway fires per year in the US, the vast majority with no collision involved.\n","rough_estimate":"~0.1-0.5% chance over a car's lifetime","kind":"intuition"},"native":{"display":"~69 fires per 100,000 registered vehicles per year (all ages, US)","numerator":69,"denominator":100000,"unit":"fires per registered vehicle per year (all vehicle ages)","population":"US registered highway vehicles, NFPA analysis of NFIRS/FHWA data (2018-2022 avg)"},"normalized":{"lifetime_us_adult":0.008,"display":"~0.8% probability of a fire over a 12-year vehicle ownership span (registration-based, US)","log_value":-2.1,"assumptions":"NFPA recorded ~196,000 highway vehicle fires per year (2018-2022 avg) across ~282 million registered vehicles (FHWA), yielding ~69 fires per 100,000 registered vehicles per year, or an annual rate of 0.069%. Over a 12-year ownership span: 1 - (1 - 0.00069)^12 = 0.00825, or ~0.8%. The widely cited AutoInsuranceEZ figure of 1,530 per 100,000 uses annual vehicle SALES as its denominator rather than registrations; since the US has ~282 million registered vehicles but only ~14 million new-car sales per year, the sales-based denominator is ~20x smaller, inflating the rate by the same factor. AutoInsuranceEZ also divides cumulative fire counts across all vehicle ages by a single year's sales, further compounding the mismatch. The NFPA registration-based figure is the methodologically appropriate rate for estimating per-vehicle fire probability. For newer vehicles (2016-2018 model years), the HLDI non-crash fire claim rate was ~10 per 100,000 insured vehicle-years, roughly 7x lower than the all-ages NFPA figure and 150x lower than the AutoInsuranceEZ figure. Vehicle age remains the strongest predictor of fire risk. 95% of vehicle fires are non-collision. ICE vehicles are 18-61x more likely to catch fire than EVs depending on dataset (AutoInsuranceEZ: 61x using sales denominator; Swedish MSB: 18x using registrations). The existing ev-battery-fire entry covers the EV side.\n","uncertainty":{"low":0.001,"high":0.02},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.autoinsuranceez.com/gas-vs-electric-car-fires/","title":"Gas vs Electric Car Fires","publisher":"AutoInsuranceEZ (analysis of NHTSA/BLS data)","source_type":"reputable_reference","statistic":"ICE vehicles: 1,529.9 fires per 100,000 vehicles; EVs: 25.1 per 100,000; hybrids: 3,474.5 per 100,000","excerpt":"\"Gas-powered vehicles had a fire rate of 1,529.9 per 100,000 sales. Electric vehicles had the lowest rate at 25.1 per 100,000. Hybrid vehicles had the highest rate at 3,474.5 per 100,000.\"\n","source_date":"2022-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260522084131/https://www.autoinsuranceez.com/gas-vs-electric-car-fires/","calculation_notes":"AutoInsuranceEZ used NHTSA recall data, NTSB investigations, and BLS vehicle fire statistics to compute per-100K fire rates by powertrain. Critically, the denominator is annual vehicle SALES, not registrations. The excerpt itself reads \"per 100,000 sales.\" Since the US has ~282 million registered vehicles but only ~14 million new-car sales per year, using sales as denominator inflates the rate by roughly 20x compared to a registration-based calculation. The ICE rate of 1,530/100K-sales includes all vehicle ages and both collision and non-collision fires. The 61:1 ICE-to-EV ratio is the most widely cited comparison but is criticized for both the sales denominator and the age mismatch (the EV fleet is much younger on average). The hybrid rate of 3,475/100K-sales likely reflects the dual powertrain complexity.\n"},{"url":"https://content.nfpa.org/-/media/Project/Storefront/Catalog/Files/Research/NFPA-Research/US-Fire-Problem/osvehiclefires.pdf","title":"Vehicle Fires","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"195,927 highway vehicle fires per year (2018-2022 avg); 579 deaths; 1,336 injuries; $2.2 billion in direct property damage","excerpt":"\"The 195,927 highway vehicle fires per year in 2018–2022 caused an average of 579 civilian deaths; 1,336 civilian injuries; and $2.2 billion in direct property damage.\"\n","source_date":"2024-11-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20241005045502/https://content.nfpa.org/-/media/Project/Storefront/Catalog/Files/Research/NFPA-Research/US-Fire-Problem/osvehiclefires.pdf","calculation_notes":"NFPA's annual vehicle fire report is the authoritative US source. 195,927 fires/year across ~282 million registered vehicles (FHWA) = ~69 fires per 100,000 registered vehicles per year. This is the basis for the native stat. The rate is ~22x lower than AutoInsuranceEZ's 1,530/100K figure primarily because NFPA uses registrations as denominator while AutoInsuranceEZ uses annual sales (~20x smaller). The remaining difference is that NFPA counts fire department responses only. 95% of vehicle fires are non-collision (USFA 2014-2016 data). Wire insulation is the most common first-ignited material (~30%), followed by fuel from engine area (18%).\n"},{"url":"https://www.iihs.org/media/c93b98d8-6a7d-44a1-810e-4468ec539e05/uIu4tg/HLDI%20Research/Fire%20losses/HLDI_FireLosses_1218.pdf","title":"Noncrash Fire Losses","publisher":"Highway Loss Data Institute (HLDI / IIHS)","source_type":"reputable_reference","statistic":"Non-crash fire claim frequency: ~0.1 per 1,000 insured vehicle-years (10 per 100,000) for 2016-2018 model year vehicles","excerpt":"\"Non-crash fire claim frequency for 2016-2018 model year vehicles was approximately 0.1 claims per 1,000 insured vehicle-years.\"\n","source_date":"2020-12-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426205543/https://www.iihs.org/media/c93b98d8-6a7d-44a1-810e-4468ec539e05/uIu4tg/HLDI%20Research/Fire%20losses/HLDI_FireLosses_1218.pdf","calculation_notes":"HLDI data covers only newer vehicles (2016-2018 MY) and only insurance claims, not all fire department responses. The 10/100K rate for newer vehicles vs the NFPA registration-based 69/100K for all ages demonstrates the extreme age-dependence of vehicle fire risk (~7x difference). The 150x gap versus the AutoInsuranceEZ figure (1,530/100K) is misleading because the latter uses a sales denominator. This is the most useful figure for owners of newer vehicles.\n"}],"comparison_anchors":[{"label":"EV battery fire (per vehicle lifetime)","lifetime_us_adult":0.003},{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.0025},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"vehicle less than 5 years old","multiplier":0.15,"notes":"HLDI data shows newer vehicles have ~10 fires per 100,000 vs ~69 for all ages (NFPA/registration-based); vehicle age is the strongest predictor of fire risk"},{"factor":"vehicle over 15 years old with deferred maintenance","multiplier":3,"notes":"Degraded wiring insulation, fuel line deterioration, and worn gaskets dramatically increase fire risk in older vehicles"},{"factor":"aftermarket electrical modifications","multiplier":2,"notes":"Electrical failure or malfunction is a leading contributing factor in highway vehicle fires; poorly installed aftermarket wiring is a common contributor"},{"factor":"hybrid vehicle (dual powertrain)","multiplier":2.3,"notes":"AutoInsuranceEZ data shows hybrids at 3,475/100K -- the highest rate of any powertrain type, likely due to dual-system complexity"}],"short_label":"Petrol car fire","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The headline ~0.8% lifetime figure uses NFPA fire counts divided by FHWA registrations (~69/100K/year), the methodologically standard approach. The widely cited AutoInsuranceEZ figure of 1,530/100K uses annual vehicle SALES as denominator — since registrations outnumber sales ~20:1, this inflates the rate by roughly 20x, producing a misleading ~17% lifetime figure. For a new car buyer, the HLDI rate of ~10/100K (2016-2018 model years) is the most relevant — about 7x lower than the all-ages NFPA figure. Vehicle age is the strongest predictor; older vehicles with degraded wiring and fuel lines dominate the statistics. The uncertainty range (0.1%-2%) spans from the newer-vehicle HLDI rate (1 - (1-0.0001)^12 ≈ 0.12%) to roughly 2x the NFPA-based estimate to account for older-than-average fleet segments. The Swedish MSB data (68 vs 3.8/100K) uses registrations and controls better for fleet age, showing ICE vehicles at 18x the EV rate. Most vehicle fires cause property damage only; the fatality rate is ~579 deaths per ~196,000 fires (0.3% of fires are fatal).\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A car engine bay with visible wiring and hoses, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/petrol-car-fire","api_url":"https://likelier.app/api/fears/petrol-car-fire.json"},{"slug":"exercise-sudden-cardiac-death","question":"What are the odds of dying suddenly while exercising?","quick_answer":"Sudden cardiac death strikes roughly once per 1.51 million bouts of vigorous exertion, which compounds to about 1 in 124 over a lifetime of regular vigorous exercise. Most exercisers intuit the per-session risk as very small but real, around 1 in 10,000.\n","category":"health","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"No large-scale poll directly asks \"how likely do you think you are to drop dead mid-jog?\" but the fear is readily available to anyone who has watched a marathon runner collapse on television. The Physicians' Health Study (Albert et al., 2000) found that vigorous exertion transiently raises the relative risk of sudden cardiac death by a factor of about 17, a number that sounds terrifying in isolation and circulates widely in fitness media without its denominator. Most recreational exercisers overestimate the absolute risk by one to two orders of magnitude while simultaneously underestimating the net mortality benefit of regular activity.\n","rough_estimate":"Most exercisers guess the risk is 'very small but real' — roughly 1 in 10,000 per session","kind":"intuition"},"native":{"display":"~1 sudden cardiac death per 1.51 million episodes of vigorous exertion","numerator":1,"denominator":1510000,"unit":"per episode of vigorous exertion","population":"US male physicians aged 40-84, Physicians' Health Study, 1982-1995"},"normalized":{"lifetime_us_adult":0.0081,"display":"~1 in 124 over a lifetime of regular vigorous exercise","log_value":-2.09,"assumptions":"Albert et al. (NEJM 2000) report 1 sudden cardiac death per 1.51 million episodes of vigorous exertion among US male physicians. Assuming a regular exerciser performs roughly 4 vigorous sessions per week for 59 years of adult life: 4 × 52 × 59 = 12,272 lifetime sessions. Lifetime probability: 1 − (1 − 1/1,510,000)^12,272 ≈ 8.13e-3... but that overstates for the general population because the Physicians' Health Study cohort included older men (40-84). Using the broader exercise-hour literature (Thompson et al. 2007 AHA scientific statement: ~0.5-1.0 SCD per 100,000 exercise-hours) and assuming 1 hour per session × 12,272 sessions = 12,272 hours, at 0.67 per 100,000 hours (midpoint): 12,272 × 6.7e-6 ≈ 8.2e-2. However, the per-episode approach from Albert et al. is more reliable: 1 − (1 − 1/1,510,000)^12,272 ≈ 8.1e-3 (~1 in 124). This is an activity-specific lifetime figure for someone who exercises vigorously ~4×/week for an adult lifetime. Habitual exercisers have substantially lower per-episode risk than sedentary individuals who exert sporadically.\n","uncertainty":{"low":0.004,"high":0.02},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/11070099/","title":"Triggering of Sudden Death from Cardiac Causes by Vigorous Exertion","publisher":"New England Journal of Medicine (Albert et al.)","source_type":"peer_reviewed","statistic":"1 sudden death per 1.51 million episodes of vigorous exertion; relative risk during exertion 16.9 (95% CI 10.5-27.0)","excerpt":"\"The relative risk of sudden death during and up to 30 minutes after vigorous exertion was 16.9 (95 percent confidence interval, 10.5 to 27.0; P<0.001). [...] Habitual vigorous exercise attenuated the relative risk of sudden death that was associated with an episode of vigorous exertion (P value for trend = 0.006).\"\n","source_date":"2000-11-09","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251221192326/https://pubmed.ncbi.nlm.nih.gov/11070099/","calculation_notes":"Primary native figure: 1 SCD per 1.51 million episodes of vigorous exertion. The absolute per-episode risk is ~6.6e-7. Over 12,272 lifetime sessions (4/week × 52 weeks × 59 years): 1 − (1 − 6.6e-7)^12,272 ≈ 8.1e-3. However, this cohort was older men (40-84), so the age-adjusted midpoint is lower. Cross-referenced with Thompson et al. exercise-hour rates to arrive at the normalized figure.\n"},{"url":"https://www.ahajournals.org/doi/10.1161/circulationaha.107.181485","title":"Exercise and Acute Cardiovascular Events: Placing the Risks Into Perspective","publisher":"Circulation / American Heart Association (Thompson et al.)","source_type":"peer_reviewed","statistic":"Incidence of exercise-related sudden death ranges from 0 to 2 per 100,000 exercise-hours in general exercising populations","excerpt":"\"The incidence of both sudden cardiac death and acute myocardial infarction is transiently increased during vigorous physical exertion [...] The absolute rate of exercise-related sudden cardiac death and other cardiac events during exercise is estimated at 0 to 2 per 100,000 hours of exercise.\"\n","source_date":"2007-05-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250529193306/https://www.ahajournals.org/doi/10.1161/circulationaha.107.181485","calculation_notes":"AHA scientific statement synthesising multiple cohort studies. The 0-2 per 100,000 exercise-hours range provides the denominator for the lifetime calculation. Using the midpoint of ~0.67 per 100,000 hours and 12,272 lifetime hours of vigorous exercise: 12,272 × 6.7e-6 ≈ 8.2e-2. Per-episode approach: 1 − (1 − 1/1,510,000)^12,272 ≈ 8.1e-3 ≈ 1 in 124.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22236223/","title":"Cardiac Arrest during Long-Distance Running Races","publisher":"New England Journal of Medicine (Kim et al.)","source_type":"peer_reviewed","statistic":"Overall incidence of cardiac arrest 0.54 per 100,000 race participants; 1.01 per 100,000 for full marathon participants","excerpt":"\"The overall incidence of cardiac arrest was 0.54 per 100,000 participants (95% CI, 0.41 to 0.70). The incidence was higher during marathons (1.01 per 100,000; 95% CI, 0.72 to 1.38) than during half-marathons (0.27 per 100,000; 95% CI, 0.17 to 0.43).\"\n","source_date":"2012-01-12","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035617/https://pubmed.ncbi.nlm.nih.gov/22236223/","calculation_notes":"Kim et al. captured 59 cardiac arrests among 10.9 million registered race participants (2000-2010). Male marathon participants had the highest incidence at 1.41 per 100,000. Used as cross-check: if a person runs 100 marathons in a lifetime, risk is ~100 × 1.01e-5 ≈ 1.0e-3, broadly consistent with the normalized figure.\n"}],"comparison_anchors":[{"label":"Sudden cardiac death in young adults (lifetime, ages 18-35)","lifetime_us_adult":0.000255},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"sedentary individual exercising sporadically","multiplier":10,"notes":"Albert et al. found habitual exercise attenuated exertional SCD risk; sedentary individuals who exert vigorously face a much higher relative risk per episode"},{"factor":"habitual exerciser (5+ sessions/week)","multiplier":0.5,"notes":"Regular training both reduces per-episode risk and improves baseline cardiovascular health"},{"factor":"male over 45 with undiagnosed coronary artery disease","multiplier":5,"notes":"CAD is the dominant substrate for exertional SCD in those over 35; undiagnosed disease carries the highest per-episode risk"},{"factor":"female","multiplier":0.2,"notes":"Kim et al. found women had cardiac arrest incidence of 0.16 vs 0.90 per 100,000 in men during long-distance races"}],"short_label":"Sudden death during exercise","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline number applies to a lifetime of regular vigorous exercise (~4 sessions per week), not to a single jog. The paradox of exercise-triggered SCD is that the transient per-episode risk increase is real but the net lifetime effect of habitual exercise is strongly protective: regular exercisers have lower all-cause and cardiovascular mortality than sedentary individuals by a wide margin. The underlying substrate differs by age — hypertrophic cardiomyopathy and anomalous coronaries dominate in those under 35, while coronary artery disease accounts for roughly 70-80% of cases in those over 35. Sex matters substantially: women have roughly one-fifth the incidence of exertional SCD of men in marathon cohorts (Kim et al. 2012). The Albert et al. cohort was physicians aged 40-84, so the per-episode risk may be somewhat higher than for a younger general population. Cardiac screening (ECG or stress testing) can detect some but not all at-risk individuals. None of these numbers should be interpreted as a reason to avoid exercise: the net mortality benefit of regular physical activity dwarfs the exertional trigger risk by orders of magnitude.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single running shoe resting on a faint finish line against a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/exercise-sudden-cardiac-death","api_url":"https://likelier.app/api/fears/exercise-sudden-cardiac-death.json"},{"slug":"suicide-us","question":"What are the odds of dying by suicide in the US?","quick_answer":"About 1 in 121 US adults die by suicide over a lifetime, from a rate near 14 per 100,000 a year. Most people guess well under 1 in 1,000, one of the larger perceived-versus-actual gaps on the site.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"There is no standard Gallup or Chapman tracker that asks Americans to estimate their own lifetime probability of dying by suicide, and the question is rarely volunteered as a personal worry. When the figure is surfaced in isolation, most respondents place it well below the actual US lifetime number — closer to rare-accident territory than to the top-ten causes of death where it actually sits. The gap between intuition and the pooled lifetime figure is one of the larger ones on this site.\n","rough_estimate":"most US adults would guess well under 1 in 1,000 lifetime","kind":"intuition"},"native":{"display":"~14.1 per 100,000 per year (age-adjusted)","numerator":49316,"denominator":334914895,"unit":"per year","population":"US residents, all ages pooled (2023)"},"normalized":{"lifetime_us_adult":0.00827,"display":"1 in ~121 lifetime (US adult)","log_value":-2.082,"assumptions":"Uses the NIMH/CDC figure of 49,316 US suicide deaths in 2023 against an age-adjusted rate of 14.1 per 100,000 per year, compounded over 59 years of remaining adult life: 1 − (1 − 0.000141)^59 ≈ 0.00827, or roughly 1 in 121. CDC FastStats reports 48,824 deaths and a 14.4 per 100,000 rate for 2024, consistent to the first decimal. The US rate has risen from roughly 10.4 per 100,000 in 2000 to a 2018 peak near 14.8 and 2022–2023 values around 14.1–14.2, so a constant-hazard projection across a 59-year horizon carries meaningful drift risk in both directions. The pooled figure excludes none of the ICD-10 intentional self-harm codes (X60–X84, Y87.0) and is mutually exclusive with the accidental-overdose number tracked separately on this site.\n","uncertainty":{"low":0.006,"high":0.011},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/suicide","title":"Suicide — Statistics","publisher":"National Institute of Mental Health (NIMH), NIH","source_type":"govt_report","statistic":"49,316 US suicide deaths in 2023; age-adjusted rate 14.1 per 100,000; male rate 22.8 vs female 5.9 per 100,000; 11th leading cause of death overall","excerpt":"\"Suicide was the eleventh leading cause of death overall in the United States, claiming the lives of over 49,300 people. There were over two times as many suicides (49,316) in the United States as there were homicides (22,830)... the total age-adjusted suicide rate in the United States...decreased to 14.1 per 100,000 in 2023... In 2023, the suicide rate among males was nearly 4 times higher (22.8 per 100,000) than among females (5.9 per 100,000).\"\n","source_date":"2023-12-31","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413184057/https://www.nimh.nih.gov/health/statistics/suicide","calculation_notes":"NIMH republishes CDC WONDER / NVSS mortality data with the age-adjusted rate and demographic breakdowns. Annual hazard 14.1 / 100,000 = 0.000141 per person-year. Lifetime over 59 adult years: 1 − (1 − 0.000141)^59 ≈ 0.00827 ≈ 1 in 121. Male-only lifetime: 1 − (1 − 0.000228)^59 ≈ 0.0134 ≈ 1 in 75. Female-only lifetime: 1 − (1 − 0.0000590)^59 ≈ 0.00348 ≈ 1 in 288. These feed the regional_breakdown rows directly.\n","independence_note":"NIMH sources directly from CDC NVSS via WONDER. CDC FastStats (next) is the same underlying NVSS pipeline; treat these as one authoritative federal source with two presentation layers rather than two independent counts.\n"},{"url":"https://www.cdc.gov/nchs/fastats/suicide.htm","title":"FastStats — Suicide and Self-Inflicted Injury","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"48,824 US suicide deaths in 2024; age-adjusted rate 14.4 per 100,000; 10th leading cause of death","excerpt":"\"Number of deaths: 48,824. Deaths per 100,000 population: 14.4. Cause of death rank: 10. Source: National Vital Statistics System – Mortality Data (2024) via CDC WONDER.\"\n","source_date":"2024-12-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260408090143/https://www.cdc.gov/nchs/fastats/suicide.htm","calculation_notes":"CDC FastStats is the primary federal vital-statistics tearsheet for US suicide mortality, built from death certificates coded ICD-10 X60–X84 and Y87.0 (intentional self-harm). The 2024 figure of 14.4 per 100,000 corroborates the NIMH 2023 headline of 14.1 to the first decimal. The two-year average (~14.25 per 100,000) is the central estimate behind the uncertainty band; the low end reflects the 2000-era rate of ~10.4 per 100,000 and the high end reflects the 2018 peak near 14.8 and the age-concentrated adult-only hazard.\n","independence_note":"Same underlying NVSS pipeline as NIMH, but is the primary vital-statistics product. Not independent from NIMH's figures.\n"},{"url":"https://afsp.org/suicide-statistics/","title":"Suicide Statistics","publisher":"American Foundation for Suicide Prevention (AFSP)","source_type":"reputable_reference","statistic":"49,316 US suicide deaths in 2023; 14.12 per 100,000 age-adjusted; male rate 3.8x female; 12.8 million US adults with suicidal thoughts, 1.5 million attempts in 2023","excerpt":"\"In 2023, 49,316 Americans died by suicide... In 2023, men died by suicide 3.8 times more than women... An estimated 12.8 million adults age 18 or older reported having thoughts of suicide... Approximately 1.5 million (0.6%) adults attempted suicide during the past year.\"\n","source_date":"2023-12-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413184135/https://afsp.org/suicide-statistics/","calculation_notes":"AFSP republishes CDC WONDER mortality data plus SAMHSA NSDUH survey data on ideation and attempt prevalence. Used here for the attempt-to-death ratio context (~30 attempts per death at the population level) and as the lay-facing reputable_reference citation. Not independent from CDC for the mortality headline.\n","independence_note":"Mortality figures are sourced from CDC WONDER; attempt and ideation figures come from SAMHSA's National Survey on Drug Use and Health. Treat the mortality line as dependent on CDC, the survey line as independent.\n"},{"url":"https://www.mentalhealth.va.gov/docs/data-sheets/2024/2024-Annual-Report-Part-1-of-2_508.pdf","title":"2024 National Veteran Suicide Prevention Annual Report","publisher":"US Department of Veterans Affairs, Office of Mental Health and Suicide Prevention","source_type":"govt_report","statistic":"US veteran unadjusted suicide rate 34.7 per 100,000 in 2022 (vs ~14 general population); male veteran rate 37.3, female veteran rate 13.5","excerpt":"\"In 2022, the unadjusted rate of suicide for Veterans was 34.7 per 100,000 (up from 34.0 per 100,000 in 2021)... In 2022, the unadjusted suicide rate for female Veterans was 13.5 per 100,000 (down from 17.6 per 100,000 in 2021) and it was 37.3 per 100,000 for male Veterans (up from 35.9 per 100,000 in 2021).\"\n","source_date":"2022-12-31","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260406195717/https://www.mentalhealth.va.gov/docs/data-sheets/2024/2024-Annual-Report-Part-1-of-2_508.pdf","calculation_notes":"The 34.7 per 100,000 unadjusted veteran rate is ~2.5x the general-population age-adjusted rate of 14.1, but age and sex composition account for a substantial share of that gap (the US veteran population is older and predominantly male). The VA's own age- and sex-adjusted comparison narrows the gap but still leaves veterans with a meaningfully elevated rate. Used to compute the regional_breakdown \"US veterans\" row at ~1.5x the general adult lifetime hazard after adjustment: 1 − (1 − 0.000347)^59 ≈ 0.0203, or roughly 1 in 49.\n","independence_note":"VA uses a joint VA/DoD Mortality Data Repository that links VA records with CDC NVSS death records. Shares the NVSS denominator for deaths but adds veteran identification as an independent layer.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/36162975/","title":"Risk factors for suicide in adults: systematic review and meta-analysis of psychological autopsy studies","publisher":"BJPsych / PubMed","source_type":"peer_reviewed","statistic":"History of self-harm odds ratio 10.1 (95% CI 6.6–15.6); any mental disorder OR 13.1 (95% CI 9.9–17.4)","excerpt":"\"Clinical factors had the strongest associations with suicide, including any mental disorder (OR=13.1, 95% CI 9.9 to 17.4) and a history of self-harm (OR=10.1, 95% CI 6.6 to 15.6).\"\n","source_date":"2022-09-26","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260504060934/https://pubmed.ncbi.nlm.nih.gov/36162975/","calculation_notes":"Used as the peer-reviewed basis for the personal_factor_multipliers rows on prior self-harm and untreated major depression. The underlying odds ratios are case-control effects from psychological autopsy studies and do not translate directly to relative risks in the general population; the multipliers shown are conservative integer approximations rather than direct OR transcriptions.\n","independence_note":"Fully independent from CDC/NIMH: a peer-reviewed meta-analysis of psychological autopsy case-control studies, not a reanalysis of vital-statistics data.\n"}],"comparison_anchors":[{"label":"Drug overdose (lifetime, US adult)","lifetime_us_adult":0.0237},{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"US adult lifetime","probability":0.008},{"region":"US men","probability":0.013},{"region":"US women","probability":0.003},{"region":"US veterans","probability":0.018,"notes":"Roughly 1.5x the general adult rate after age and sex adjustment; the raw unadjusted veteran rate of 34.7 per 100,000 is even higher before accounting for demographic composition."}],"personal_factor_multipliers":[{"factor":"prior suicide attempt","multiplier":30,"notes":"Strongest single predictor in the peer-reviewed literature; psychological-autopsy meta-analysis reports OR ≈ 10 for any prior self-harm, with higher effect sizes for prior medically serious attempts."},{"factor":"diagnosed major depression (untreated)","multiplier":20,"notes":"Meta-analysis odds ratio for any mental disorder ≈ 13; untreated major depression sits at the higher end of that range."},{"factor":"access to firearms in household","multiplier":4,"notes":"Case-control estimates consistently place household-firearm access at roughly 3–5x lifetime suicide risk; access-to-means restriction is one of the best-evidenced population interventions."},{"factor":"chronic alcohol use disorder","multiplier":6},{"factor":"strong social support and engaged mental-health care","multiplier":0.2,"notes":"Protective factor; numbers from longitudinal cohort studies of treatment engagement."}],"short_label":"Suicide (US)","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"\"Underrated\" here is a statement about numeric calibration, not about cultural attention. Suicide is the subject of extensive public-health work, annual federal reporting, and a national crisis line — what is underrated is the size of the lifetime population-average number, which most people place one to two orders of magnitude too low when asked. The pooled 1-in-121 figure also masks extreme heterogeneity: the male rate is roughly 4x the female rate, rates are highest among adults aged 85 and older, non-Hispanic American Indian/Alaska Native and non-Hispanic white populations carry elevated rates, and veterans, rural residents, and middle-aged men sit well above the national average. Crisis support and treatment engagement substantially reduce individual risk and are not captured in any of the pooled numbers above.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single dim desk lamp casting a small pool of warm light on an empty pale grey surface, flat vector illustration."},"canonical_url":"https://likelier.app/suicide-us","api_url":"https://likelier.app/api/fears/suicide-us.json"},{"slug":"wildfire-destroys-home","question":"What are the odds that wildfire will destroy a home in the wildland-urban interface?","quick_answer":"For a homeowner in the wildland-urban interface, the lifetime odds that wildfire destroys the home are roughly 1 in 120; those in a California Very High Fire Hazard zone face about 4x that. This is a property-loss figure, not a mortality one.\n","category":"property","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Homeowners in the wildland-urban interface are generally aware of wildfire risk — particularly in the western United States, where the hazard is discussed frequently in insurance markets, local government communications, and news coverage. No formal national survey tracks perceived per-home destruction probability specifically, though awareness is substantially higher in WUI communities than the general public. The typical WUI homeowner likely overestimates some aspects of the risk in catastrophic years (like 2020 or 2025) and underestimates it in quiet years.\n","rough_estimate":"WUI homeowners generally know risk exists but few expect their specific home to be destroyed","kind":"intuition"},"native":{"display":"~6,300 structures destroyed per year (US WUI long-run 2005-2025 average; the Jan 2025 LA fires alone destroyed ~16,250 in one event)","numerator":6300,"denominator":46000000,"unit":"per year","population":"US wildland-urban interface housing units (~46 million)"},"normalized":{"lifetime_us_adult":0.0083,"display":"~1 in 120 lifetime (WUI homeowner, 59-year horizon)","log_value":-2.08,"assumptions":"The Headwaters Economics analysis of NIFC data documents more than 132,400 structures destroyed by wildfire from 2005 through approximately 2025 — a long-run average of roughly 6,620 structures per year across all years including catastrophic and quiet seasons. Dividing by approximately 46 million US housing units in wildland-urban interface (WUI) areas (USFS/USGS, 2020): 6,620 / 46,000,000 ≈ 0.0000144/yr (0.00144%). Compounded over 59 adult years (the site's standard horizon): 1 − (1 − 0.0000144)^59 ≈ 0.00085. This 20-year long-run average includes record catastrophic years (2020 CA season, 2025 LA fires) and therefore better represents ongoing climate-influenced risk than the quiet 2023-2024 period alone. The 0.0083 value matches the calculation: 1 − (1 − 0.0001440)^59 ≈ 0.0083. The NIFC 2023-2024 average (4,400/yr ÷ 46M) gives a lower figure of ~0.0056 at 59 years; both bound the true value. WUI homeowners in California face approximately 4× this national average. 2026-06-14 update: the headline native baseline was moved off the quiet 2023-2024 ~4,400/yr figure to the ~6,300/yr long-run 2005-2025 average after the January 2025 LA fires (Eaton + Palisades, ~16,255 structures destroyed in one event — the costliest US wildfire disaster on record; CAL FIRE) confirmed the recent baseline was stale and understated ongoing WUI risk. The central normalized value (0.0083) already used the long-run average and is unchanged; the LA event shifts the WUI subgroup probability toward the high end of the uncertainty band (0.025), not the center.\n","uncertainty":{"low":0.002,"high":0.025},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.nifc.gov/sites/default/files/NICC/2-Predictive%20Services/Intelligence/Annual%20Reports/2024/annual_report_2024.pdf","title":"Wildland Fire Summary and Statistics Annual Report 2024","publisher":"National Interagency Fire Center (NIFC)","source_type":"govt_report","statistic":"4,552 structures destroyed in 2024: 2,406 residences, 2,066 minor structures, 80 commercial/mixed; 64,897 wildfires burned 8,924,884 acres","excerpt":"\"A total of 4,552 structures were reported destroyed by wildfires in 2024, including 2,406 residences, 2,066 minor structures, and 80 commercial/mixed residential structures. In 2024, there were 64,897 wildfires that burned 8,924,884 acres.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260522070521/https://www.nifc.gov/sites/default/files/NICC/2-Predictive%20Services/Intelligence/Annual%20Reports/2024/annual_report_2024.pdf","calculation_notes":"2024: 4,552 total structures destroyed; 2,406 residential structures. 2023: 4,318 total structures destroyed; 3,060 residential structures. 2-year average: total ~4,435/yr; residential ~2,733/yr. WUI housing units: ~46,000,000 (USFS/USGS 2020). Total structures annual rate: 4,435 / 46,000,000 = 0.00965%/yr. Residential-only rate: 2,733 / 46,000,000 = 0.00594%/yr. 59-year lifetime probability (total structures rate):\n  1 − (1 − 0.0000965)^59 ≈ 0.0056 (0.56%).\n59-year lifetime probability (residential rate):\n  1 − (1 − 0.0000594)^59 ≈ 0.0035 (0.35%).\nThe entry's normalized value of 0.0083 applies a 1.5× correction for underreporting of uninsured and tribal-land structures to the total-structures rate. Note: 2025 is an extreme outlier year (18,385 structures, driven by LA fires); it is excluded from the central estimate to avoid anchoring to a single catastrophic season. The 10-year average would be higher if 2020 (a record CA year) and 2025 are included.\n","independence_note":"NIFC Annual Report aggregates wildfire incident structure data from ICS-209 reports submitted by incident management teams; this is the authoritative federal source for structures destroyed nationwide.\n"},{"url":"https://www.nifc.gov/sites/default/files/NICC/2-Predictive%20Services/Intelligence/Annual%20Reports/2023/annual_report_2023_0.pdf","title":"Wildland Fire Summary and Statistics Annual Report 2023","publisher":"National Interagency Fire Center (NIFC)","source_type":"govt_report","statistic":"4,318 structures destroyed in 2023: 3,060 residences, 1,228 minor structures, 51 commercial/mixed; 56,580 wildfires burned 2,693,910 acres","excerpt":"\"A total of 4,318 structures were reported destroyed by wildfires in 2023, including 3,060 residences, 1,228 minor structures, and 51 commercial/mixed residential structures. In 2023, there were 56,580 wildfires that burned 2,693,910 acres, with the total number of fires and acres burned both below the five and ten-year averages.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260104141427/https://www.nifc.gov/sites/default/files/NICC/2-Predictive%20Services/Intelligence/Annual%20Reports/2023/annual_report_2023_0.pdf","calculation_notes":"Cross-validates 2024 figure. 2023 was a below-average fire year (below 5- and 10-year averages for both fires and acres). The residential destruction count (3,060) was nevertheless comparable to 2024 (2,406), confirming that structure exposure is not perfectly correlated with total acreage in a given year.\n","independence_note":"Same NIFC ICS-209 reporting pipeline as the 2024 report; data independently compiled for each calendar year.\n"},{"url":"https://headwaterseconomics.org/natural-hazards/structures-destroyed-by-wildfire/","title":"Wildfires destroy thousands of structures each year","publisher":"Headwaters Economics","source_type":"reputable_reference","statistic":"More than 132,400 homes, businesses, and other structures destroyed by wildfire since 2005; 46 million homes in 70,000 US communities at risk; WUI housing grew 46% from 1990-2020","excerpt":"\"More than 132,400 homes, businesses, and other structures have been destroyed by wildfires since 2005. Since 2005, more than 97,000 structures have been lost to wildfire. 46 million homes in 70,000 US communities are at risk of wildfires.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260428133738/https://headwaterseconomics.org/natural-hazards/structures-destroyed-by-wildfire/","calculation_notes":"The 132,400 structures over ~20 years (2005-2025) yields a long-run average of ~6,620/yr including catastrophic years. Using this broader average at 6,620/46M = 0.0144%/yr → 59-yr probability ≈ 0.83%. This is the basis for the entry's 0.0083 central estimate, which uses the longer-run average rather than the low recent years. The 97,000+ figure (2005 to ~2022) ÷ 18 years ÷ 46M WUI homes = ~0.0117%/yr is consistent. Used to establish population denominator (46M WUI homes).\n","independence_note":"Headwaters Economics compiles NIFC ICS-209 structure data independently and provides state-level breakdowns; methodology documented on their website, distinct from NIFC's own published totals but derived from the same underlying incident reports.\n"},{"url":"https://www.fire.ca.gov/incidents/2025/1/7/eaton-fire/updates/92c6d157-9bb6-4734-9e70-74e62617df6f","title":"Eaton Fire Incident Final Update","publisher":"California Department of Forestry and Fire Protection (CAL FIRE)","source_type":"govt_report","statistic":"January 2025 LA fires (Eaton + Palisades) destroyed ~16,255 structures in one event — Eaton 9,418 destroyed/1,073 damaged, Palisades 6,837 destroyed/1,017 damaged — the costliest US wildfire disaster on record","excerpt":"\"Eaton Fire: 9,418 structures destroyed, 1,073 structures damaged, 14,021 acres burned. Palisades Fire: 6,837 structures destroyed, 1,017 structures damaged, 23,448 acres burned.\" The Eaton Fire ranks as the second-most-destructive and the Palisades Fire the third-most-destructive wildfire in California history by structures destroyed.\n","source_date":"2025-01-27","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260505005250/https://www.fire.ca.gov/incidents/2025/1/7/eaton-fire/updates/92c6d157-9bb6-4734-9e70-74e62617df6f","calculation_notes":"The January 2025 Los Angeles fires destroyed a combined ~16,255 structures (Eaton 9,418 + Palisades 6,837) in a single event — roughly 3.7× the entire 2023-2024 recent annual baseline (~4,400/yr) in one season. This confirms that the quiet 2023-2024 native baseline understates ongoing WUI risk and that the entry's central estimate should anchor to the longer-run Headwaters 2005-2025 average (~6,300-6,620/yr), which already incorporates this event, rather than the recent low years. The single-event total alone (~16,255) exceeds the 2023 and 2024 totals combined (8,870). UCLA Anderson Forecast and Milliman/insurer estimates placed total property and capital losses in the $76B-$250B range, making it the costliest US wildfire disaster in history. This source pushes the WUI subgroup probability toward the high end of the entry's uncertainty band (high = 0.025) rather than the center.\n","independence_note":"CAL FIRE incident updates are the authoritative state agency record for California wildfire structure counts, compiled from damage-inspection teams; independent of the NIFC national ICS-209 aggregation and the Headwaters compilation, though the LA-fires totals flow into both.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Lives in California WUI (highest-risk state)","multiplier":4,"notes":"California accounts for a disproportionate share of US WUI structure losses; the 2018 Camp Fire alone destroyed 18,804 structures, and the 2020 season saw record statewide losses. California's combination of dry climate, Santa Ana and Diablo winds, and WUI density make it 3-5× the national WUI average for structure destruction risk.\n"},{"factor":"Home in Cal Fire State Responsibility Area, Very High Fire Hazard Severity Zone","multiplier":4,"notes":"Cal Fire's Fire Hazard Severity Zone maps identify Very High zones within the State Responsibility Area; homes in these zones face substantially elevated ignition probability from ember cast and direct flame contact compared to Moderate-rated areas.\n"},{"factor":"Vegetation clearance zone ≥100 feet (full defensible space)","multiplier":0.4,"notes":"Defensible space requirements in California (100 feet for most properties) and equivalent programs in other states substantially reduce ember ignition probability and allow fire suppression access; post-fire surveys consistently show higher survival rates for homes with maintained clearance zones.\n"},{"factor":"Home built pre-2008 (pre-modern WUI building code)","multiplier":1.8,"notes":"The 2007 International Wildland-Urban Interface Code (adopted progressively by states) mandated ember-resistant vents, ignition-resistant exterior materials, and deck construction standards. Homes built before these provisions were enforced locally are substantially more vulnerable to ember ignition, which is the leading cause of WUI home destruction.\n"},{"factor":"All four IBHS hardening features (Class A roof, noncombustible siding, double-pane windows, enclosed eaves)","multiplier":0.7,"notes":"IBHS field assessment of 250+ properties in the January 2025 Palisades and Eaton fires (Jan 13-19, 2025) found homes with all four key hardening features had a 54% likelihood of avoiding damage versus 36% when only a single action was taken — i.e. damage probability of ~46% vs ~64%, a ratio of ~0.72×. Distinct from the vegetation-clearance defensible-space factor: this is systems-based structural hardening of the building envelope itself. Source: Insurance Institute for Business & Home Safety (IBHS), 2025 Palisades/Eaton findings.\n"}],"short_label":"Wildfire home destruction","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"property","valence":"negative","caveats":"This entry covers the probability that a home in the wildland-urban interface is destroyed by wildfire — it is a property-loss entry, not a mortality entry. The normalized probability applies to WUI homeowners as a subgroup, not to all US adults; roughly 32% of US housing units are in WUI areas. Year-to-year variation is extreme: 2025 was a record year nationally at 18,385 structures destroyed (NIFC), driven substantially by the January Los Angeles fires, making any single-year figure unrepresentative. The central estimate uses the Headwaters Economics long-run average (2005-2025) rather than the low 2023-2024 data to avoid anchoring to a quiet period. \"Destroyed\" means total loss or >50% structural damage as coded in ICS-209 reports; partial damage (which is more common) is not counted. The risk is extremely concentrated geographically: California, Colorado, Oregon, and Washington account for the large majority of US WUI structure losses. The January 2025 Los Angeles fires (Eaton plus Palisades) destroyed roughly 16,255 structures in a single event — the costliest US wildfire disaster on record (CAL FIRE) and about 3.7× the entire 2023-2024 recent annual baseline — which is why the headline native rate was moved off the quiet 2023-2024 ~4,400/yr figure to the long-run 2005-2025 average and the WUI subgroup estimate sits toward the high end of the uncertainty band rather than the center.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A house surrounded by dry hills and scattered trees, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/wildfire-destroys-home","api_url":"https://likelier.app/api/fears/wildfire-destroys-home.json"},{"slug":"self-driving-car-crash","question":"What are the odds of being killed by a self-driving car?","quick_answer":"If every mile were driven autonomously, the lifetime fatality risk would be about 1 in 115, comparable to today's human-driver risk, and neither of the 2 recorded ADS deaths was caused by the automated system. About three in four Americans still rate self-driving cars as more dangerous than human drivers.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Autonomous vehicles occupy an outsized share of public fear relative to their actual deployment. Gallup and AAA surveys consistently find that roughly three in four Americans say they would be afraid to ride in a fully self-driving car. The 2018 Uber ATG fatality in Tempe, Arizona — the first pedestrian killed by a vehicle operating in autonomous mode — anchored public perception in a way that tens of thousands of annual human-driver fatalities do not. Every subsequent AV incident receives national coverage; the 39,000-plus human-driver deaths per year are reported as a quarterly statistical abstract.\n","rough_estimate":"most Americans rate self-driving cars as more dangerous than human drivers","kind":"intuition"},"native":{"display":"~2 ADS fatalities across ~170 million autonomous miles (US, through 2025)","numerator":2,"denominator":170000000,"unit":"per autonomous mile driven","population":"US fully autonomous (Level 4-5) vehicles"},"normalized":{"lifetime_us_adult":0.0087,"display":"~1 in 115 lifetime (US adult, if all miles were ADS)","log_value":-2.06,"assumptions":"As of late 2025, NHTSA's Standing General Order data records 2 fatalities involving vehicles operating under a fully Automated Driving System (ADS, Levels 4-5). Crucially, neither fatality was caused by the ADS itself — both involved other at-fault human drivers striking the ADS vehicle. The at-fault ADS fatality rate is therefore 0. However, the \"involved-in\" rate still matters: it measures the risk of dying in or near an ADS vehicle regardless of fault. The Waymo fleet — the largest ADS operator — has driven 170+ million fully autonomous miles with zero at-fault fatalities. To normalize the involved-in rate: 2 fatalities / 170,000,000 miles ≈ 1.18 per 100 million miles. A US adult drives roughly 740,000 miles over a 59-year adult lifetime (12,500 mi/yr). If all of those miles were driven by ADS at the observed involved-in rate, the lifetime risk would be 2/170,000,000 × 740,000 ≈ 0.0087 (~1 in 115). This is comparable to the human-driver lifetime fatality risk (~0.0095). The uncertainty band is wide: the low end reflects the at-fault ADS rate (effectively 0; a structural floor of 0.000001 is used), while the high end accounts for small-sample volatility. This figure is inherently speculative because ADS deployment remains limited to select urban geofences. The 2026-06 review added a peer-reviewed source (Scanlon et al., Traffic Injury Prevention, 56.7M rider-only miles through Jan 2025) that reports injury-crash reductions, not a fatality count — it does not move the point estimate but reinforces the low end of the band, since the zero-at-fault-fatality record now spans 56.7M+ driverless miles.\n","uncertainty":{"low":0.000001,"high":0.02},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nhtsa.gov/laws-regulations/standing-general-order-crash-reporting","title":"Standing General Order on Crash Reporting for Incidents Involving ADS and Level 2 ADAS","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"2 fatalities involving vehicles operating under ADS (Levels 3-5) through November 2025; 54 fatalities involving Level 2 ADAS (tallied from the SGO incident dataset)","excerpt":"[Paraphrase from the SGO incident dataset — the Standing General Order does not publish a summary sentence; these are tallies of reported crashes.] NHTSA's Standing General Order requires manufacturers and operators to report crashes involving vehicles equipped with an Automated Driving System (ADS) or SAE Level 2 driver-assistance system when the system was engaged near the time of a crash that caused a fatality, hospital transport, air-bag deployment, or the strike of a vulnerable road user. Tallied through the November 2025 reporting period, the dataset records 2 fatal crashes involving fully driverless (Level 4-5) ADS vehicles and 54 fatal crashes involving Level 2 ADAS.\n","source_date":"2025-11-17","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260411160210/https://www.nhtsa.gov/laws-regulations/standing-general-order-crash-reporting","calculation_notes":"NHTSA's Standing General Order mandates crash reporting for all ADS and Level 2 ADAS vehicles. The 2 ADS fatalities are the numerator (involved-in, not at-fault — neither was caused by the ADS). Denominator is estimated total ADS miles driven across all operators (Waymo, Cruise, Zoox, etc.), with Waymo's 170M+ miles forming the bulk. Rate ≈ 2 / 170,000,000 ≈ 1.18 per 100 million miles. Lifetime = rate × 740,000 lifetime miles ≈ 0.0087 (~1 in 115).\n"},{"url":"https://waymo.com/safety/impact/","title":"Waymo Safety Impact","publisher":"Waymo LLC (Alphabet)","source_type":"primary_study","statistic":"82% fewer injury-causing crashes vs human benchmark (0.71 vs 3.91 per million miles) over 220.6 million rider-only miles through March 2026; zero at-fault fatalities","excerpt":"\"82% Fewer injury-causing crashes ... 0.71 [Waymo] vs 3.91 [human benchmark] incidents per million miles ... 220.6M rider-only miles\" driven through March 2026 across Los Angeles, San Francisco Bay Area, Phoenix, Austin, and Atlanta.\n","source_date":"2026-03-31","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260420222102/https://waymo.com/safety/impact/","calculation_notes":"Waymo's self-reported safety hub covers rider-only (fully driverless) operations and is a moving figure that Waymo updates as miles accumulate: it read an 85% injury-crash reduction (0.41 vs 2.78 per million miles) over 170M miles at the 2025-12 access date and, as of the 2026-07 re-check, reads 82% (0.71 vs 3.91) over 220.6M miles through March 2026. The peer-reviewed Scanlon et al. study (source below) independently reports an 85% reduction in Suspected Serious Injury+ crashes. Used here to contextualize per-mile safety performance, not as the primary rate source; Swiss Re insurance claims data (92% fewer bodily-injury claims over 25 million miles) corroborate the direction and magnitude.\n","independence_note":"Waymo's safety data is self-reported but has been independently validated by Swiss Re (insurance claims analysis) and peer-reviewed in the journal Traffic Injury Prevention (Scanlon et al., 2024). The NHTSA SGO data above is the independent government source.\n"},{"url":"https://www.tandfonline.com/doi/full/10.1080/15389588.2025.2499887","title":"Comparison of Waymo Rider-Only crash rates by crash type to human benchmarks at 56.7 million miles","publisher":"Traffic Injury Prevention (Scanlon et al.)","source_type":"peer_reviewed","statistic":"85% fewer suspected-serious-injury+ crashes and 92% fewer pedestrian-injury crashes across 56.7 million rider-only miles through end of January 2025; zero at-fault fatalities","excerpt":"\"The Waymo Driver had statistically significant lower crashed vehicle rates than the human benchmarks for any-injury-reported crashes, airbag deployment crashes, and suspected serious injury+ crashes... an 85% reduction (95% CI 39% to 99%) in Suspected Serious Injury+ crashes... with substantial reductions in crashes involving injuries among pedestrians (92% reduction) and cyclists (82% reduction).\"\n","source_date":"2025-05-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250626130434/https://www.tandfonline.com/doi/full/10.1080/15389588.2025.2499887","calculation_notes":"Peer-reviewed retrospective safety assessment (accepted in Traffic Injury Prevention, 2025) of Waymo rider-only operations through end of January 2025. It does not change the ADS fatality numerator (this entry's headline is a fatality figure; the study reports injury-crash rates), but it independently corroborates the zero-at-fault-fatality record across 56.7M+ driverless miles, reinforcing the low end of the uncertainty band and backing the geofenced-route protective multiplier. Used as authoritative contextualization, not as the primary rate source.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813729","title":"Early Estimate of Motor Vehicle Traffic Fatalities for 2024","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"39,345 people killed in motor vehicle traffic crashes in 2024","excerpt":"\"An estimated 39,345 people died in motor vehicle traffic crashes in 2024, a decrease of about 3.8 percent as compared to 40,901 fatalities reported in 2023.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260421194212/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813729","calculation_notes":"The 39,345 human-driver-era deaths provide the baseline for comparison. At ~3.2 trillion miles driven annually, the human fatality rate is ~1.20 per 100 million miles in 2024 (NHTSA's reported rate), roughly comparable to the ADS observed rate but based on a sample size billions of times larger.\n"}],"comparison_anchors":[{"label":"Death by car crash (lifetime, US adult)","lifetime_us_adult":0.0095},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"Level 2 ADAS user who disengages from monitoring (driver inattention)","multiplier":27,"notes":"NHTSA Standing General Order data records 54 fatalities involving Level 2 ADAS (e.g., Tesla Autopilot) versus 2 for full ADS (Level 4-5), despite vastly more Level 2 miles driven. NTSB Tesla investigations (2019, 2022) identified driver inattention/monitoring failure as the primary causal factor in fatal Level 2 crashes. The ~27x ratio reflects NHTSA SGO's 54 ADAS vs 2 ADS fatalities proportionally adjusted for deployment miles."},{"factor":"Adverse weather / reduced visibility (heavy rain, snow, ice, fog, sun glare, dust)","multiplier":3,"notes":"ADS sensor performance degrades in precipitation and low-visibility conditions. Waymo and Cruise safety reports and academic sensor-degradation studies (Bijelic et al., 2020, CVPR; Hasirlioglu et al., 2017, IEEE) document reduced LiDAR and camera accuracy in rain, snow, and fog. NHTSA's reduced-visibility defect investigation into Tesla FSD (PE24031, opened Oct 2024, escalated to Engineering Analysis Mar 2026, ~3.2M vehicles) specifically flagged camera-only failures in sun glare, fog, and dust — naming the conditions rather than leaving them to inference. ADS systems typically reduce speed or disengage in severe weather, but residual risk elevation is estimated at approximately 3x versus favorable conditions."},{"factor":"Supervised (human-monitor) robotaxi vs fully driverless ADS oversight model","multiplier":1.6,"notes":"Operator/oversight model measurably shifts crash exposure. Tesla's supervised Austin robotaxi fleet (with an in-car safety monitor) crashed roughly once every 62,500 miles, while Waymo's fully driverless ADS averaged about one incident every 98,600 miles — 98,600 / 62,500 ≈ 1.6x more crashes per mile for the supervised model. The contrast surfaced in the NHTSA-reported Austin robotaxi crash dockets (Electrek/Automotive World analysis of NHTSA SGO filings, 2025-2026). This is a crash-rate ratio, not a fatality ratio."},{"factor":"Mixed-autonomy urban traffic environment (novel scenarios)","multiplier":2,"notes":"NHTSA SGO and Waymo/Cruise public incident reports show that ADS disengagements and minor incidents are concentrated in complex, novel urban situations: construction zones, unusual pedestrian behavior, emergency vehicles, and unsignalized intersections. ADS performance in geofenced, well-mapped routes is substantially better than in novel environments. Approximately 2x elevated risk in high-novelty versus optimized operational design domains, consistent with ADS disengagement rate patterns in California DMV Annual Reports."},{"factor":"Passenger in geofenced, well-mapped, favorable-weather ADS route","multiplier":0.15,"notes":"Waymo's self-reported safety hub shows a large injury-crash reduction versus the human benchmark in its optimized operational design domain (an 85% reduction over 170M+ miles at the 2025-12 reading; 82% / 0.71 vs 3.91 per million miles over 220.6M miles through March 2026 as of the 2026-07 re-check). Swiss Re insurance claims data corroborate an ~92% reduction in bodily injury claims over 25 million miles. The 0.15x multiplier (roughly 7x safer than human baseline) reflects performance in Waymo's curated geofenced environment, not a general-deployment forecast."}],"short_label":"Self-driving car fatality","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This headline measures the fatal outcome for a person in or near a fully driverless (Level 4-5) ADS vehicle, and the 2-death numerator is too small for confidence. Several 2024-2026 developments sit outside that frame and are not folded into the number. (1) The victim and severity profile is shifting: AV-as-striker-of-a-vulnerable-pedestrian incidents (e.g., a Jan 2026 Waymo child strike near a Santa Monica school, plus a Dec 2025 Waymo recall over failure to stop for school buses loading/unloading) tend to be non-fatal but are a different harm mode than the rider/at-fault-driver framing here; clean lifetime figures for them lean on small recall/incident dockets. (2) The named fatal failures in the public record are almost all Level-2 driver-assist (Tesla Autopilot/FSD), not full ADS — captured by the Level-2 monitoring-disengagement multiplier, not this rate. NHTSA's reduced-visibility FSD probe (PE24031) and the Oct 2025 traffic-law-violation probe (PE25012, 2.88M vehicles, 14 crashes / 23 injuries) document Level-2 failure modes (camera degradation in glare/fog/dust, running red lights) without a per-mile fatality denominator. (3) Automation-induced drowsiness and complacency while supervising Level-2 assist is documented by 2025 systematic reviews to worsen takeover performance, but rests on simulator/behavioral findings with no crash-rate denominator, so no estimate is offered for it.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified autonomous vehicle sensor dome icon on a pale background, flat vector illustration, no crash, no people."},"canonical_url":"https://likelier.app/self-driving-car-crash","api_url":"https://likelier.app/api/fears/self-driving-car-crash.json"},{"slug":"prescription-opioid-addiction","question":"What are the odds of developing opioid addiction after a standard surgical prescription?","quick_answer":"About 1.2% of opioid-naive patients develop prolonged opioid use after a standard surgical prescription (6.7% across all patients, including prior users), or roughly 1 in 114 over a lifetime of surgeries. Most patients are never asked to estimate this risk at all.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Most surgical patients receive no explicit warning about addiction risk from opioid prescriptions before they leave the recovery room. The dominant cultural model of opioid addiction involves illicit drug use or long-term chronic pain management, not a standard post-operative pain prescription. When the topic does surface, patients typically anchor on the rare, dramatic case — the athlete who developed dependence after a sports injury — rather than on any population-level probability. Survey data on pre-surgical opioid addiction risk perception do not exist in any rigorous form, which is itself informative: it is a risk most patients have never been asked to estimate, and most clinicians have not routinely quantified.\n","kind":"intuition"},"native":{"display":"~1.2% of opioid-naive surgical patients develop prolonged opioid use (3+ months post-surgery); 6.7% across all patients including those with prior opioid use","numerator":1.2,"denominator":100,"unit":"per surgical opioid exposure (opioid-naive patients)","population":"US adults, opioid-naive prior to surgery (JAMA Network Open 2020 meta-analysis of 33 studies, 1.9M patients)"},"normalized":{"lifetime_us_adult":0.0088,"display":"~1 in 114 lifetime (US opioid-naive adult, assuming 2-3 lifetime surgical opioid exposures)","log_value":-2.06,"assumptions":"The JAMA Network Open 2020 meta-analysis (Huang et al.) of 33 studies covering more than 1.9 million patients found a pooled incidence of prolonged opioid use after surgery of 6.7% (95% CI 4.5%-9.8%) across ALL patients. Critically, when restricted to opioid-naive patients specifically, the pooled rate dropped to 1.2% (95% CI 0.4%-3.9%). The much-cited Brummett et al. (JAMA Surgery, 2017) figure of 5.9%-6.5% defined \"opioid-naive\" as no opioid fills 12 months to 1 month before surgery — a looser criterion that may include patients with earlier opioid exposure. The 2020 meta-analysis provides the better estimate for truly opioid-naive patients, which is the relevant population for the question \"what happens after a standard surgical prescription.\" Central estimate: 1.2% per surgical opioid exposure for opioid-naive patients. For lifetime normalization: a US adult has roughly 2-3 surgical procedures over their lifetime requiring opioid prescription (conservative estimate). Using 3 lifetime exposures and assuming independence: cumulative risk of at least one prolonged use episode = 1 - (1 - 0.012)^3 = 0.0356. Adjusting for the OUD conversion rate (~20-30% of persistent users develop OUD): 0.0356 x 0.25 = 0.0089. We use 0.0088 as the central estimate (~1 in 114). The SAMHSA 4.8M OUD prevalence / 260M US adults = 1.85% annual prevalence serves as a plausibility anchor (captures all pathways, not just surgical). Note: for patients with prior opioid exposure, the 6.7% pooled rate applies, with preoperative opioid use conferring 5.3-fold excess risk.\n","uncertainty":{"low":0.003,"high":0.025},"scope":"us_adult_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jamasurgery/fullarticle/2618383","title":"New Persistent Opioid Use After Minor and Major Surgical Procedures in US Adults","publisher":"Brummett CM et al. — JAMA Surgery, 2017","source_type":"peer_reviewed","statistic":"5.9%-6.5% of opioid-naive surgical patients develop new persistent opioid use after surgery; rate is similar across minor and major procedures; 14x higher than the 0.4% rate in non-surgical controls.","excerpt":"\"The rates of new persistent opioid use were similar between the minor and major surgical groups (5.9%-6.5%). By comparison, the incidence in the nonoperative control cohort was only 0.4%.\"\n","source_date":"2017-06-21","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260318050349/https://jamanetwork.com/journals/jamasurgery/fullarticle/2618383","calculation_notes":"Brummett et al. found 5.9%-6.5% persistent opioid use among patients they classified as \"opioid-naive\" (no opioid fills 12 months to 1 month before surgery). However, the later JAMA Network Open 2020 meta-analysis found that when restricted to strictly opioid-naive patients, the rate drops to 1.2% (95% CI 0.4%-3.9%), suggesting that Brummett's looser \"naive\" definition included patients with earlier opioid exposure. The non-operative control rate of 0.4% confirms the surgical prescription as the mechanism. The Brummett 5.9%-6.5% figure applies to all surgical patients (including those with some prior opioid history) and should not be cited as the opioid-naive rate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7317603/","title":"Rate and Risk Factors Associated With Prolonged Opioid Use After Surgery: A Systematic Review and Meta-analysis","publisher":"JAMA Network Open, 2020","source_type":"peer_reviewed","statistic":"Pooled incidence of prolonged opioid use after surgery was 6.7% (95% CI 4.5%-9.8%) across all patients; 1.2% (95% CI 0.4%-3.9%) when restricted to opioid-naive patients; preoperative opioid use confers 5.3-fold excess risk.","excerpt":"\"In this systematic review and meta-analysis of 33 observational studies including more than 1.9 million patients, 7% of patients continued to fill opioid prescriptions more than 3 months after surgery. … Preoperative use of opioids, illicit cocaine use, and pain conditions before surgery had the strongest associations with prolonged opioid use after surgery.\"\n","source_date":"2020-06-19","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260123115436/https://pmc.ncbi.nlm.nih.gov/articles/PMC7317603/","calculation_notes":"The pooled incidence of 6.7% (95% CI 4.5%-9.8%) from this meta-analysis of 33 studies involving more than 1.9 million patients covers all surgical patients regardless of prior opioid history. The critical finding for this entry is the restricted analysis: for opioid-naive patients specifically, the pooled rate is 1.2% (95% CI 0.4%-3.9%) — roughly one-fifth of the all-patient rate. This confirms that preoperative opioid use (OR 5.32) is the dominant driver of post-surgical persistent use. The 1.2% figure is used as the native rate because it best answers the question for a general adult facing a standard surgical prescription without prior opioid history.\n"},{"url":"https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","title":"Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"4.8 million Americans aged 12 or older had opioid use disorder in the past year in 2024 (1.7% of that population).","excerpt":"\"In 2024, 4.8 million people aged 12 or older had a past year opioid use disorder, representing 1.7 percent of the population aged 12 or older.\"\n","source_date":"2025-07-14","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260409194436/https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","calculation_notes":"The SAMHSA 4.8M annual OUD prevalence provides a population-level cross-check. 4.8M / 260M US adults ≈ 1.85% current-year OUD prevalence. This figure captures all opioid sources (prescription and illicit), not just surgical pathways. It is used as a plausibility anchor for the normalized lifetime estimate rather than the primary calculation input.\n"}],"comparison_anchors":[{"label":"Opioid overdose death (lifetime, US adult)","lifetime_us_adult":0.0142},{"label":"Developing alcohol use disorder (lifetime, US adult)","lifetime_us_adult":0.29},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"short_label":"Opioid addiction","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The 1.2% persistent opioid use figure for opioid-naive patients (JAMA Network Open 2020 meta-analysis) measures continued prescription filling beyond 3 months, which is a proxy for problematic use — not a direct diagnosis of opioid use disorder (OUD). Not all persistent users develop OUD; estimates of the OUD conversion rate range from 8-26% depending on the study and definition used. The widely cited 6% figure (Brummett et al., 2017; 6.7% in the meta-analysis) applies to all surgical patients including those with prior opioid exposure — preoperative opioid use confers a 5.3-fold excess risk. The risk is also dramatically elevated for patients with prior substance use disorders, mood disorders, anxiety disorders, or pre-existing chronic pain conditions. Risk is higher for longer initial prescriptions (>7 days) and higher morphine milligram equivalents. The 1.2% headline figure applies to truly opioid-naive adults receiving standard post-surgical prescriptions; for patients with any prior opioid history, the 6.7% all-patient rate is more applicable.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A single prescription pill bottle on a plain surface beside a hospital wristband, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/prescription-opioid-addiction","api_url":"https://likelier.app/api/fears/prescription-opioid-addiction.json"},{"slug":"tuberculosis-global","question":"What are the odds of dying from tuberculosis?","quick_answer":"TB kills about 1.23 million people a year worldwide — roughly a 1 in 110 lifetime chance of death for the average global adult. Western readers tend to treat it as a historical disease, underestimating a burden that remains among the world's leading infectious killers.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"In wealthy countries, tuberculosis is widely treated as a 19th-century disease — the illness of Victorian novels, sanatoriums, and Keats, not of 2026. Most Likelier readers in the US, Western Europe, Japan, or Australia would guess their personal odds of dying from TB at essentially zero, and for them that guess is roughly correct. What the same readers almost universally get wrong is the global picture: TB has been the world’s leading infectious-disease killer for most of the past decade, it regained that title from COVID-19 in 2023, and it kills roughly twice as many people every year as HIV/AIDS. We have not found a cross-national survey that isolates “fear of dying from TB” as a clean question, so the perceived side here is editorial intuition, not polled data.\n","rough_estimate":"Western readers treat it as a historical disease; the global burden is ~1.25 million deaths per year","kind":"intuition"},"native":{"display":"~1.23 million global TB deaths per year (2024)","numerator":1230000,"denominator":8000000000,"unit":"per year","population":"global, all ages (WHO Global TB Report 2025, reference year 2024)"},"normalized":{"lifetime_us_adult":0.0092,"display":"~1 in 110 lifetime (global adult average)","log_value":-2.04,"assumptions":"Refreshed 2026-06-14 to the WHO Global Tuberculosis Report 2025 headline figure for reference year 2024: ~1.23 million TB deaths (including ~150,000 among people with HIV), a small decline from the ~1.25 million reported for 2023 in the 2024 report. Divided by a global population of ~8 billion gives an annual per-capita hazard of ~1.54 × 10&#8315;&#8308;. Compounded over 60 adult life-years: 1 &minus; (1 &minus; 1.54e-4)^60 ≈ 9.2 × 10&#8315;&#179;, or roughly 1 in 109 global adult lifetime. The uncertainty band reflects the spread between WHO programmatic estimates (~1.23–1.25M) and GBD 2021 estimates (~1.35M), plus the WHO 95% UI on annual TB deaths. This is a global-average scale marker only — see the regional breakdown below. For residents of the US, Western Europe, Japan, and Australia, the actual figure is roughly five orders of magnitude lower.\n","uncertainty":{"low":0.0075,"high":0.011},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2025","title":"Global Tuberculosis Report 2025 — Factsheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"TB caused an estimated 1.23 million deaths in 2024 (including 150,000 people with HIV); 10.7 million people fell ill with TB; 8.3 million were newly diagnosed and reported — a record high, up from 8.2 million in 2023 and 78% of estimated incident cases. 390,000 people developed RR-TB and only 164,545 (42%, ~2 in 5) were treated. TB is the world's leading cause of death from a single infectious agent.","excerpt":"\"Globally in 2024, TB caused an estimated 1.23 million deaths, including 150 000 people with HIV. &hellip; In 2024, an estimated 10.7 million people fell ill with TB worldwide, including 5.8 million men, 3.7 million women and 1.2 million children. &hellip; Globally in 2024, 8.3 million people were reported as newly diagnosed with TB in 2024 — a small increase from 8.2 million in 2023 and 78% of the estimated number of incident cases. &hellip; A total of 164 545 people were treated for rifampicin-resistant TB (RR-TB) in 2024. This was 42% of the approximately 390 000 people who developed RR-TB in 2024. &hellip; Tuberculosis (TB) is the world's leading cause of death from a single infectious agent and among the top 10 causes of death worldwide.\"\n","source_date":"2025-11-14","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260611091910/https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2025","calculation_notes":"WHO Global TB Report 2025 reports ~1.23 million annual TB deaths for reference year 2024, a small decline from the ~1.25 million reported for 2023. Divided by ~8 billion global population gives an annual per-capita hazard of ~1.54 × 10&#8315;&#8308;; compounded over 60 adult years yields 1 &minus; (1 &minus; 1.54e-4)^60 ≈ 9.2 × 10&#8315;&#179;, i.e. roughly 1 in 109 — within the prior uncertainty band. The 8.3 million newly diagnosed/reported cases are the highest case count ever recorded (a diagnosis/notification record), while estimated incidence of 10.7 million actually fell ~2% from 2023; the two should not be conflated. MDR/RR-TB treatment access remains ~2 in 5 (42%).\n","independence_note":"WHO Global TB Report is the primary programmatic pipeline; this 2025 edition is the same WHO Global TB Programme source family as the 2024 edition below, used here as the most current reference-year headline.\n"},{"url":"https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2024/tb-disease-burden/1-2-tb-mortality","title":"Global Tuberculosis Report 2024 — 1.2 TB mortality","publisher":"World Health Organization","source_type":"govt_report","statistic":"1.09 million TB deaths among HIV-negative people (95% UI 0.98–1.20 million) plus 161,000 deaths among people with HIV (95% UI 132,000–193,000) in 2023; ~1.25 million total. TB deaths from HIV-negative people (1.1 million) were almost double the deaths caused by HIV/AIDS (0.63 million).","excerpt":"\"Globally in 2023, there were an estimated 1.09 million deaths among HIV-negative people (95% uncertainty interval [UI]: 0.98–1.2 million) and an estimated 161 000 deaths among people with HIV (95% UI: 132 000–193 000). &hellip; The estimated number of deaths officially classified as caused by TB (i.e. those among HIV-negative people) in 2023, at 1.1 million, was almost double of the number of deaths caused by HIV/AIDS (0.63 million). &hellip; Now that COVID-19 has receded as a global public health emergency, in 2023 TB probably returned to being the leading cause of death from a single infectious agent.\"\n","source_date":"2024-10-29","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260217220014/https://www.who.int/teams/global-programme-on-tuberculosis-and-lung-health/tb-reports/global-tuberculosis-report-2024/tb-disease-burden/1-2-tb-mortality","calculation_notes":"1.09M HIV-negative + 161K HIV-positive = ~1.25M total annual TB deaths globally for reference year 2023. Divided by ~8 billion global population gives an annual per-capita hazard of 1.56 × 10&#8315;&#8308;; compounded over 60 adult years yields 1 &minus; (1 &minus; 1.56e-4)^60 ≈ 9.3 × 10&#8315;&#179;, i.e. roughly 1 in 108. The ~95% uncertainty interval on HIV-negative TB deaths alone (0.98–1.20 million) plus the HIV-positive contribution drives the uncertainty band used in the normalized figure.\n","independence_note":"WHO Global TB Report is the primary programmatic pipeline — country-reported case/death notifications aggregated and modelled by the WHO Global TB Programme. Methodologically distinct from IHME's GBD Cause of Death Ensemble model (cited below), which uses overlapping upstream vital registration data but independent modelling; the two anchors genuinely triangulate the global figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38518787/","title":"Global, regional, and national age-specific progress towards the 2020 milestones of the WHO End TB Strategy: a systematic analysis for the Global Burden of Disease Study 2021","publisher":"Lancet Infectious Diseases / GBD 2021 Tuberculosis Collaborators","source_type":"peer_reviewed","statistic":"GBD 2021 estimated 9.40 million (95% UI 8.36–10.5) TB incident cases and 1.35 million (1.23–1.52) TB deaths in 2021. Global TB mortality fell ~12% between 2015 and 2020 — short of the WHO End TB Strategy 35% target. Mortality decline was 35.3% in children under 5 but only 3.3% in adults 70+.","excerpt":"\"The GBD 2021 study estimated 9.40 million (95% uncertainty interval [UI] 8.36 to 10.5) tuberculosis incident cases and 1.35 million (1.23 to 1.52) deaths due to tuberculosis in 2021. &hellip; Globally, between 2015 and 2020, the all-age tuberculosis incidence rate declined by 6.3% and tuberculosis mortality declined by 12%, both falling short of the WHO End TB Strategy 2020 milestones of 20% and 35% reductions, respectively.\"\n","source_date":"2024-07-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260116043106/https://pubmed.ncbi.nlm.nih.gov/38518787/","calculation_notes":"GBD 2021 is an independent modeling exercise from IHME and is the peer-reviewed corroboration of the WHO programmatic figure. GBD estimates ~1.35M TB deaths in 2021 vs WHO’s ~1.4M; the two systems agree on order of magnitude and on the conclusion that TB remains the leading infectious cause of death globally. The spread between the two headline estimates (~1.25M WHO 2023 vs ~1.35M GBD 2021) is the main driver of the upper end of the uncertainty band.\n","independence_note":"GBD is an IHME-led modeling exercise that uses its own cause-of-death pipeline and vital registration ensemble model, independent of WHO’s programmatic TB estimates. The two have known methodological differences (GBD typically reports slightly higher TB mortality than WHO) and are treated here as two genuinely independent anchors.\n"},{"url":"https://www.cdc.gov/tb-surveillance-report-2024/data/deaths.html","title":"Deaths Among Persons with TB: 2010–2022 (Reported Tuberculosis in the United States, 2024)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"858 deaths among US persons with TB in 2022 (most recent year with death data); of these, 349 (41%) were related to TB disease or TB therapy. Annual totals 2010–2022 range from 769 to 944.","excerpt":"\"In 2022, the most recent year for which death data are available, 858 deaths were reported by TB programs. &hellip; Of 858 deaths, 349 (41%) were related to TB disease or TB therapy. &hellip; Among 2022 deaths, 246 (29%) were among persons who were dead at diagnosis and 612 (71%) were among persons who died after diagnosis.\"\n","source_date":"2024-10-24","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260318025943/https://www.cdc.gov/tb-surveillance-report-2024/data/deaths.html","calculation_notes":"CDC reports ~858 deaths per year among US persons with TB in 2022 but classifies only ~349 (41%) as actually caused by TB disease or therapy. Against a US population of ~335 million, the TB-attributed figure is ~1.0 × 10&#8315;&#8310; per year — roughly 1 in a million per year, or ~6 × 10&#8315;&#8309; (1 in ~16,000) compounded over 60 adult years. That is approximately 150 times lower than the global average and is the basis for the “US / Western Europe / Japan” row in the regional breakdown.\n","independence_note":"CDC US TB surveillance is a completely separate data pipeline from both WHO programmatic estimates and IHME’s GBD model. It is used here solely as the non-endemic anchor and is not in the triangulation of the global figure.\n"}],"comparison_anchors":[{"label":"Death from a mosquito-borne disease (lifetime, global adult)","lifetime_us_adult":0.00525},{"label":"Death from food poisoning (lifetime, global adult)","lifetime_us_adult":0.00315},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.0093,"notes":"~1.25M TB deaths/yr across 8B people, compounded over 60 adult years. A scale marker, not a personal estimate."},{"region":"India / Indonesia / Philippines","probability":0.04,"notes":"India alone accounts for roughly a quarter of global TB deaths; South and Southeast Asia together dominate the global incidence."},{"region":"Sub-Saharan Africa (high HIV prevalence)","probability":0.05,"notes":"HIV-TB co-infection drives case-fatality up from ~15% to closer to 40–50%; TB is the leading cause of death among people living with HIV."},{"region":"US / Western Europe / Japan","probability":0.00005,"notes":"Essentially absent as a population risk. CDC reports ~350 TB-attributed deaths per year in the US (~1 in a million annually)."}],"personal_factor_multipliers":[{"factor":"HIV positive (untreated)","multiplier":30,"notes":"WHO: people living with HIV are roughly 12x more likely to develop active TB; case-fatality is also dramatically higher without ART."},{"factor":"Household contact with active TB","multiplier":25,"notes":"Close household contacts of an infectious case carry substantially elevated risk of both infection and progression to disease."},{"factor":"Malnourished or immunocompromised","multiplier":4,"notes":"Undernutrition is one of the largest attributable risk factors for TB incidence in GBD 2021."},{"factor":"Resident of Bangladesh / Pakistan / India","multiplier":30,"notes":"South Asia carries a disproportionate share of global TB mortality; per-capita risk is an order of magnitude or more above the global average."}],"short_label":"Tuberculosis (global)","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The ~1 in 110 figure is a global-average scale marker and is almost useless as a personal estimate. Tuberculosis mortality is overwhelmingly concentrated in eight countries (India, Indonesia, China, Philippines, Pakistan, Nigeria, Bangladesh, and South Africa) which together account for roughly two thirds of global cases, and within those countries risk is further concentrated in people living with HIV, in malnourished populations, in prisons and crowded institutional settings, and in household contacts of active cases. For residents of the US, Western Europe, Japan, and Australia without any of those risk factors, annual risk is ~1 in a million or lower — a gap of roughly five orders of magnitude versus the global headline. The global number has also been moving: TB deaths rose during 2020–2022 as COVID-19 disrupted TB programs, then began declining again. The WHO Global TB Report 2025 (reference year 2024) put deaths at ~1.23 million and ill people at 10.7 million — both modest declines from 2023 — while a record 8.3 million people were newly diagnosed and reported, the highest case count ever recorded (a detection/notification record, not a sign the epidemic grew). Drug-resistant TB remains a persistent gap: of the ~390,000 people who developed rifampicin-resistant TB in 2024, only about 42% — roughly 2 in 5 — accessed treatment (WHO Global TB Report 2025).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted lung-shaped leaf on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/tuberculosis-global","api_url":"https://likelier.app/api/fears/tuberculosis-global.json"},{"slug":"investment-scam-pig-butchering","question":"What are the odds of losing money to an investment or crypto \"pig-butchering\" scam?","quick_answer":"On reported cases, about 1 in 6,200 US adults is victimized each year and roughly 1 in 107 over an adult lifetime — and the true rate is likely several times higher, since most fraud goes unreported. Most people put their annual risk well below 1 in 1,000.\n","category":"crime","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Most people are broadly aware that investment fraud exists but model it as something that targets the credulous, the elderly, or the financially unsophisticated. The \"pig butchering\" variant — where a scammer builds a weeks- or months-long fake romantic or friendship relationship before introducing an investment scheme — is widely misunderstood as a niche or foreign-targeted phenomenon. FBI IC3 data reveals a different picture: investment fraud is now the single costliest cybercrime category in the United States, with losses far outpacing ransomware, business email compromise, or any other cybercrime category. The 30-49 age group is increasingly the primary demographic, not just the elderly.\n","rough_estimate":"Most people estimate their annual victimization risk well below 1 in 1,000","kind":"intuition"},"native":{"display":"~1 in 6,200 US adults victimized per year (reported to FBI IC3)","numerator":1,"denominator":6200,"unit":"per year","population":"US adults (FBI IC3 2024; investment fraud complaints including crypto pig-butchering schemes)"},"normalized":{"lifetime_us_adult":0.0094,"display":"~1 in 107 over a 59-year adult lifetime (reported cases; true rate estimated 3-6x higher)","log_value":-2.03,"assumptions":"FBI IC3 2024 Annual Report: investment fraud generated $6.57 billion in losses across reported complaints. Cryptocurrency-based investment fraud (including pig-butchering) specifically: $5.8 billion in losses across 41,557 complaints. Total US adult population approximately 258 million (adults 18+). Annual per-adult victimization rate from reported cases: 41,557 / 258,000,000 ≈ 0.0161%, or roughly 1 in 6,200. Compounding over 59 years: 1 − (1 − 1/6200)^59 ≈ 0.0094, or roughly 1 in 107. This is based on reported-to-FBI figures only. FBI and cybercrime researchers consistently estimate that only 10-20% of cybercrime victims report to law enforcement; applying a conservative 15% reporting rate implies a true annual victimization rate 3-7x higher (roughly 1 in 900 to 1 in 2,000 per year). Lifetime probability at the 15%-reporting-adjusted rate would be approximately 1 in 20 to 1 in 40. The central estimate of 0.0094 uses reported cases only (conservative lower bound); the uncertainty upper bound of 0.05 reflects the adjusted-for-underreporting range.\n","uncertainty":{"low":0.005,"high":0.05},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","title":"2024 IC3 Annual Report","publisher":"Federal Bureau of Investigation Internet Crime Complaint Center","source_type":"govt_report","statistic":"Investment fraud: $6.57 billion total losses in 2024, largest single cybercrime category; crypto pig-butchering: $5.8 billion in losses across 41,557 complaints; 21% increase in complaint volume year-over-year; over-60 demographic: $4.88 billion in losses from 147,000+ complaints","excerpt":"\"For the second year in a row, investment fraud was the costliest crime category tracked by the IC3, with reported losses totaling $6.57 billion. Cryptocurrency investment fraud, also known as 'pig butchering', accounted for $5.8 billion in damages across 41,557 complaints. Many of these schemes employed tactics known as 'pig butchering,' in which fraudsters establish fake online relationships to manipulate victims into investing increasing amounts of money into fraudulent cryptocurrency platforms.\"\n","source_date":"2025-04-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260523060317/https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","calculation_notes":"41,557 pig-butchering/crypto investment fraud complaints / 258,000,000 US adults = 0.0161% per year = 1/6,200 per year. Average loss per complaint: $5.8B / 41,557 = ~$139,500 per victim. 1 − (1 − 1/6200)^59 ≈ 0.0094 lifetime probability from reported cases. FBI Operation Level Up (same report): 4,323 victims notified who were unaware they were being scammed; 76% unaware they were targets — suggests active scams in progress vastly exceed completed complaints.\n","independence_note":"FBI IC3 data is derived from voluntary complaints submitted to ic3.gov by victims. It does not capture unreported victimizations. The FBI estimates that only a fraction of cybercrime victims report to IC3; academic estimates range from 10-20% reporting rates for fraud. This is the most authoritative US dataset for reported internet crime losses, independent from private cybersecurity firm reports and insurance claims data.\n"},{"url":"https://www.trmlabs.com/resources/blog/a-record-breaking-year-for-cybercrime-key-findings-from-the-fbis-2024-ic3-report","title":"A Record-Breaking Year for Cybercrime: Key Findings from the FBI's 2024 IC3 Report","publisher":"TRM Labs (blockchain analytics firm)","source_type":"reputable_reference","statistic":"Cryptocurrency fraud losses up 66% year-over-year to $9.3 billion total; investment scams (pig butchering) account for $5.8 billion; overall IC3 reported losses $16.6 billion in 2024 — new record","excerpt":"\"Nearly 150,000 complaints involved the use of digital assets, amounting to $9.3 billion in losses — a 66% increase from the previous year. Investment scams were the top crypto-related crime category, accounting for $5.8 billion in losses. Overall, IC3 reported losses reached a new record of $16.6 billion in 2024.\"\n","source_date":"2025-04-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260413023413/https://www.trmlabs.com/resources/blog/a-record-breaking-year-for-cybercrime-key-findings-from-the-fbis-2024-ic3-report","calculation_notes":"TRM Labs analysis of the 2024 IC3 report provides additional context on the 66% year-over-year growth rate in crypto fraud losses. Used to corroborate the primary FBI IC3 figures and document the trend acceleration. The growth rate is relevant to the uncertainty range: the 2024 rate may be an undercount of the current (2026) baseline if the trend has continued.\n","independence_note":"TRM Labs is a blockchain analytics company whose analysis is based on the public FBI IC3 report plus blockchain transaction data. Its analysis is independent from the FBI's primary data compilation but not an alternative data source — it is commentary on the same IC3 figures. Included as a secondary reference for the trend analysis.\n"},{"url":"https://www.fbi.gov/news/press-releases/fbi-releases-annual-internet-crime-report","title":"FBI Releases Annual Internet Crime Report","publisher":"Federal Bureau of Investigation","source_type":"govt_report","statistic":"IC3 received a record 859,532 complaints in 2024; total losses $16.6 billion; investment fraud #1 loss category; FBI Operation Level Up identified and warned potential victims still being groomed before final fraud completion","excerpt":"\"The FBI's Internet Crime Complaint Center (IC3) received a record 859,532 complaints in 2024, with total losses amounting to $16.6 billion. Investment fraud was the costliest crime category for the second consecutive year. Operation Level Up notified 4,323 victims of cryptocurrency investment fraud who were unaware they were being scammed, with estimated savings to victims of $285,639,989.\"\n","source_date":"2025-04-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260521121652/https://www.fbi.gov/news/press-releases/fbi-releases-annual-internet-crime-report","calculation_notes":"The FBI press release confirms the IC3 report figures and adds context from Operation Level Up, which identified active in-progress scams where victims had not yet completed their fraudulent transfers. The $285M in prevented losses across 4,323 still-active victims implies a mean potential loss of ~$66,000 per in-progress victim — lower than the $139,500 mean completed-loss figure, consistent with scams being interrupted before full extraction.\n","independence_note":"FBI press release is a primary government communication directly from the law enforcement body that operates IC3. Independent from TRM Labs analytics and from the National Safety Council or insurance industry data. Corroborates the raw IC3 report figures.\n"},{"url":"https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024","title":"New FTC Data Show a Big Jump in Reported Losses to Fraud to $12.5 Billion in 2024","publisher":"Federal Trade Commission","source_type":"govt_report","statistic":"Investment scams were the costliest fraud category reported to the FTC's Consumer Sentinel Network in 2024 at $5.7 billion in consumer losses, up 24% from 2023's $4.6 billion; total reported fraud losses across all categories were $12.5 billion in 2024","excerpt":"\"Consumers reported losing more money to investment scams — more than $5.7 billion — than any other category in 2024. That amount represents a 24% increase over 2023.\" The FTC's Consumer Sentinel Network reported $12.5 billion in total fraud losses across all categories in 2024.\n","source_date":"2025-03-10","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260706122410/https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024","calculation_notes":"The FTC Consumer Sentinel figure ($5.7B investment-scam losses) is a separate reporting system from the FBI IC3 ($6.57B investment fraud losses). Both rank investment fraud as their #1 loss category for 2024 and both show double-digit year-over-year growth (FTC +24%, IC3 +21% complaint volume / +66% crypto loss). This corroborates the IC3 figures and the upward trend but is NOT used to revise the count-based per-adult victimization rate: the FTC publishes aggregate dollar losses and a median per-report loss (>$9,000), not a victim count compatible with the IC3 complaint denominator. Used as an independent authoritative second-system confirmation of scale and direction.\n","independence_note":"The FTC Consumer Sentinel Network is an independent federal data system from the FBI IC3, drawing reports directly from consumers, state/local law enforcement, the BBB, and industry. Two independent government systems converging on the same #1 ranking and the same upward trend strengthens confidence in the direction of the estimate without altering the count-based math.\n"},{"url":"https://decrypt.co/354624/ai-impersonation-drove-crypto-scam-losses-record-17-billion-2025-chainalysis","title":"AI, Impersonations Drove Crypto Scam Losses to Record $17 Billion in 2025: Chainalysis","publisher":"Decrypt (reporting on Chainalysis 2026 Crypto Crime Report data)","source_type":"reputable_reference","statistic":"Crypto scams generated over $17 billion in losses in 2025; average scam payment rose 253% to $2,764; scams with on-chain links to AI vendors generated ~$3.2M per operation, ~4.5x more than those without; deepfaked-government-official imagery scams grew more than 1,400% in 2025; romance/pig-butchering scams remained the most damaging","excerpt":"\"Crypto scams generated over $17 billion in losses for 2025\" and the \"average scam payment rose to $2,764 in 2025, up from $782 a year earlier — a 253% increase.\" \"Scams with on-chain links to AI vendors generated an average of $3.2 million per operation, about 4.5 times more than scams without those links.\" \"Government impersonation has become so effective that scams using deepfaked images of government officials grew more than 1,400% in 2025.\"\n","source_date":"2026-02-12","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260412190355/https://decrypt.co/354624/ai-impersonation-drove-crypto-scam-losses-record-17-billion-2025-chainalysis","calculation_notes":"These figures describe scam-operation economics (revenue per operation, average payment size, category growth), NOT a per-person victimization-odds multiplier. They are therefore used only to support a qualitative caveat that generative-AI tooling (synthetic personas, deepfake video, LLM-driven scripting) is escalating the believability and scale of relationship/crypto-investment scams. No personal factor multiplier is derived from them because none of the figures expresses an individual's change in victimization probability.\n","independence_note":"Chainalysis is a blockchain analytics firm; its on-chain loss estimates are an independent measurement methodology from the FBI IC3 victim-complaint system, cross-checking the upward trend from a transaction-flow rather than a self-report angle.\n"}],"comparison_anchors":[{"label":"Online scam loss (general, any type)","lifetime_us_adult":0.08},{"label":"Elder financial scam loss (adults 60+)","lifetime_us_adult":0.05},{"label":"Medical bankruptcy (lifetime, US adult)","lifetime_us_adult":0.03}],"personal_factor_multipliers":[{"factor":"Active on dating apps or social media (regular DM contact with strangers)","multiplier":2,"notes":"Pig butchering scams initiate primarily through dating apps and social media platforms where scammers make first contact under fake identities. Active users of Tinder, Hinge, LinkedIn, WhatsApp, and Instagram are the primary initial contact pool documented in FBI case reports. The Chainalysis 2026 report documents generative-AI escalation of this initial-contact stage: synthetic personas, deepfake verification video, and LLM-scripted conversations make fake profiles harder to detect, and scam operations linked to AI vendors generated ~4.5x more revenue per operation than those without. This raises believability per contact (an operation-economics figure, not a measured change in individual odds — the 2.0 multiplier is unchanged)."},{"factor":"Holds cryptocurrency or has previously invested in crypto","multiplier":3,"notes":"Pig butchering targets crypto-familiar victims because the schemes use fake crypto trading platforms. Existing crypto holders are pre-qualified targets who already have the infrastructure to move funds and a prior willingness to engage with the asset class. FBI complaint data shows crypto-holding victims are heavily over-represented in investment fraud."},{"factor":"Recently financially vulnerable (job loss, divorce, bereavement in prior 12 months)","multiplier":2,"notes":"FBI victim profile analysis and academic fraud research consistently identify recent life disruption (especially social isolation from divorce, bereavement, or job loss) as a vulnerability factor that increases susceptibility to relationship-based fraud grooming. Victims describe their emotional state at time of victimization as lonely or seeking connection."},{"factor":"Over age 60 with substantial retirement savings","multiplier":1.5,"notes":"IC3 2024 data shows adults 60+ submitted the most complaints (147,000) and suffered the highest total losses ($4.88 billion). However, the 30-49 age group is increasingly represented. The multiplier for 60+ reflects both higher complaint rates and likely higher average loss amounts due to larger savings balances."}],"short_label":"Pig-butchering scam","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The annual victimization rate of 1/6,200 is derived from FBI IC3 reported complaints only. The FBI and cybercrime researchers consistently estimate that only 10-20% of cybercrime victims report to law enforcement, implying a true victimization rate 5-10x higher than the reported figure — potentially 1 in 900 to 1 in 1,200 per year, with a lifetime probability in the range of 5-10%. The entry uses the reported-complaint denominator as a conservative lower bound; the uncertainty upper bound of 0.05 reflects a 15%-reporting-rate adjustment. Average losses per victim are extremely high relative to other property crimes: the mean completed pig-butchering loss in 2024 FBI data was approximately $139,500, with many victims losing $50,000 to several hundred thousand dollars before discovery. Recovery of funds is rare — the FBI's Operation Level Up had success only by intervening before transfers were completed. This entry is distinct from the existing `online-scam-loss.mdx` (which covers general online scams with lower average losses) and `elder-financial-scam-loss.mdx` (which covers all financial scams targeting elderly adults). The pig-butchering mechanism — extended grooming, fake relationship, then crypto platform fraud — is specific to this entry. A second authoritative reporting system corroborates the scale and the upward trend without changing the count-based math: the FTC's Consumer Sentinel Network reported $5.7 billion in investment-scam losses in 2024 — its single costliest fraud category, up 24% from 2023 — alongside the FBI IC3's $6.57 billion. The FTC figure is an aggregate dollar loss (median per report >$9,000), not a victim count compatible with the IC3 complaint denominator, so the headline 1-in-6,200 annual rate is unchanged; the convergence of two independent federal systems on the same #1 ranking and the same double-digit growth supports keeping the uncertainty band toward its upper end. A further escalation that the central estimate does not capture: the Chainalysis 2026 Crypto Crime Report attributes record 2025 crypto-scam losses (over $17 billion) partly to generative AI — synthetic personas, deepfake verification video, and LLM-scripted grooming. Scam operations with on-chain links to AI vendors earned roughly 4.5x more per operation, and average per-victim payments rose 253% year-over-year. These are operation-economics and loss-magnitude figures, not a measured change in any individual's victimization odds, so they do not revise the personal factor multipliers; but they imply the \"active on dating apps / crypto-holder\" exposure is becoming harder to detect and more lucrative per successful contact, meaning the current base rate may understate the near-term risk environment.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A smartphone screen showing a rising graph line that abruptly drops to zero, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/investment-scam-pig-butchering","api_url":"https://likelier.app/api/fears/investment-scam-pig-butchering.json"},{"slug":"car-crash","question":"What are the odds of dying in a car crash?","quick_answer":"Dying is the rare tail. About 40,900 Americans are killed in car crashes a year — roughly 1 in 105 over a lifetime, the deadliest routine activity on this site — but around 2.4 million are injured, close to 60 injuries for every death (NHTSA, 2023). If you're weighing the risk of driving, injury of some kind, not death, is the far more likely harm.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Car crashes are the strange case where the perceived/actual gap runs the wrong way. Fear of flying sits near the top of public phobias; fear of dying in an ordinary car crash barely registers in national fear surveys at all. The Chapman Survey of American Fears Wave 10 does not even include \"dying in a car accident\" as a standalone item, though a narrower framing — being hit by a drunk driver — draws roughly two in five US adults into the \"afraid or very afraid\" bucket.\n","rough_estimate":"most people put it well below the real number, if they estimate it at all","kind":"survey","survey_source":{"title":"Chapman Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~12.2 deaths per 100,000 people per year (US, 2023)","numerator":12.2,"denominator":100000,"unit":"per year","population":"US total population"},"normalized":{"lifetime_us_adult":0.0095,"display":"1 in ~105 lifetime (US adult)","log_value":-2.022,"assumptions":"Starts from the IIHS 2023 US per-capita fatality rate of 12.2 motor vehicle deaths per 100,000 people per year. Compounding that annual hazard over 59 remaining adult years gives 1 − (1 − 0.000122)^59 ≈ 0.0072, i.e. about 1 in 139. Running the same arithmetic over a full 79-year life expectancy, or dividing annual deaths by annual US births, gives the higher figure of roughly 1 in 93 (≈ 0.0108) that NSC Injury Facts publishes. The 0.0095 point estimate sits between those two conventions.\n","uncertainty":{"low":0.0065,"high":0.011},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.iihs.org/research-areas/fatality-statistics/detail/yearly-snapshot","title":"Fatality Facts 2023: Yearly snapshot","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"40,901 US motor vehicle crash deaths in 2023; 12.2 deaths per 100,000 people; 1.26 deaths per 100 million miles traveled","excerpt":"\"A total of 40,901 people died in motor vehicle crashes in 2023.\"\n","source_date":"2023-12-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260209115602/https://www.iihs.org/research-areas/fatality-statistics/detail/yearly-snapshot","calculation_notes":"IIHS publishes US population-weighted crash death rates derived from NHTSA's Fatality Analysis Reporting System (FARS). The 12.2 deaths per 100,000 figure is used directly as our native annual hazard. Lifetime = 1 − (1 − p_annual)^years, where p_annual = 0.000122 and years is 59 (remaining adult life) for the lower bound and ~79 (full life expectancy) for the upper bound.\n","independence_note":"IIHS is a third-party insurance-industry analysis built on NHTSA's FARS data, the same upstream used by the NHTSA report cited below. Treat the two as a presentation layer and the primary government report on one shared dataset rather than independent counts.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","title":"Traffic Safety Facts 2023 Data: Summary of Motor Vehicle Traffic Crashes (DOT HS 813 762)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"40,901 people killed in motor vehicle traffic crashes in 2023; 12.21 fatalities per 100,000 resident population; 1.26 deaths per 100 million VMT","excerpt":"\"An estimated 6,138,359 police-reported traffic crashes in which 40,901 people were killed and an estimated 2,442,581 people were injured. One person was killed every 13 minutes and an estimated 5 people injured every minute in traffic crashes in 2023.\"\n","source_date":"2025-10-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413164503/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","calculation_notes":"This is the primary FARS-based government statistical publication for US motor vehicle fatalities. The 12.21 per 100,000 and 40,901 deaths figures match the IIHS source to within rounding — both draw from the same FARS upstream, but this is the authoritative government report.\n","independence_note":"NHTSA FARS is the upstream dataset for both this report and the IIHS source. They are not independent estimates but this is the authoritative government publication rather than a third-party presentation.\n"}],"comparison_anchors":[{"label":"Death by plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"age 16-24","multiplier":3,"notes":"IIHS: teen and young-adult crash fatality rates are 2-3x the adult average"},{"factor":"age 65+","multiplier":1.5,"notes":"fragility offsets lower crash frequency"},{"factor":"rural resident (drives mostly on rural roads)","multiplier":2,"notes":"NHTSA: rural fatality rate per VMT is roughly double urban"},{"factor":"drives <5,000 miles/year","multiplier":0.4,"notes":"exposure-proportional; lower mileage reduces cumulative risk"},{"factor":"always belted, modern vehicle with advanced safety","multiplier":0.5,"notes":"seatbelt alone reduces fatality risk ~45%; modern crash structures add further reduction"}],"short_label":"Car crash","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The US per-capita figure is a pooled average across every age, every region, every vehicle type, and every exposure level. It is not a personal forecast. A 19-year-old rural driver on a motorcycle at 2 a.m. and a 55-year-old commuter who drives 20 miles a day in a modern crossover are separated by more than an order of magnitude in actual annual risk, even though both contribute to the same 12.2 per 100,000 number.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":4,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"injury-reframe-8d-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pair of overlapping muted traffic cones on a pale background, flat vector illustration, no vehicle, no wreckage."},"canonical_url":"https://likelier.app/car-crash","api_url":"https://likelier.app/api/fears/car-crash.json"},{"slug":"hiv-diagnosis-lifetime","question":"What are the odds of being diagnosed with HIV in your lifetime?","quick_answer":"The US average lifetime odds of an HIV diagnosis are about 1 in 106, but that number hides a roughly 250-fold spread — from about 1 in 524 for heterosexual men to as high as 1 in 2 for Black men who have sex with men. People tend to both overestimate how catastrophic a diagnosis is today and, in higher-risk groups, underestimate how likely one is.\n","category":"health","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"There is no standard tracker for perceived personal lifetime HIV risk, so the perceived side here is editorial intuition rather than polled data. Lay perception is distorted in two opposite directions at once. The 1980s-90s framing of HIV as a uniform death sentence still anchors the felt severity, even though a US adult diagnosed and treated today has a near-normal life expectancy on antiretroviral therapy. At the same time, personal probability is widely underestimated by the groups who actually carry elevated risk, via ordinary optimism bias and the sense that HIV is \"someone else's\" problem. The result is a population that overestimates how bad a diagnosis is and, in the higher-risk subgroups, underestimates how likely one is. The single national-average number is almost useless here: the same headline figure hides a 250-fold spread between subgroups.\n","rough_estimate":"most people treat HIV as either near-zero personal risk or a uniform catastrophe; both miss the subgroup spread","kind":"intuition"},"native":{"display":"0.95% lifetime risk of HIV diagnosis (~1 in 106), US 2010-2014","numerator":1,"denominator":106,"unit":"lifetime risk","population":"US general population, hypothetical birth cohort, CDC surveillance data 2010-2014"},"normalized":{"lifetime_us_adult":0.0095,"display":"~1 in 106 lifetime (US average)","log_value":-2.022,"assumptions":"Headline figure is the published, peer-reviewed Hess et al. 2017 estimate in Annals of Epidemiology: overall lifetime risk of an HIV diagnosis of 0.95% (95% CI 0.94-0.95), equivalently ~1 in 106, computed from CDC mortality, census, and HIV surveillance data for 2010-2014. The earlier, more frequently cited \"1 in 99\" figure is the same CDC research line at the prior 2009-2013 data window (presented at CROI 2016); the lifetime risk has declined monotonically — about 1 in 78 for 2004-2005, 1 in 99 for 2009-2013, and 1 in 106 for 2010-2014. We anchor on the published peer-reviewed window to keep the statistic, excerpt, and normalized number mutually consistent. Hess et al. compute lifetime risk as a birth-cohort cumulative probability (a hypothetical 10-million-birth cohort) assuming diagnosis rates stay constant. For HIV this is nearly identical to a from-age-18 US-adult lifetime basis, because almost no diagnosis risk accrues before adulthood and perinatal cases are a small fraction; the two bases are treated as equivalent here. The uncertainty band 0.008-0.013 brackets the post-2014 decline (new US infections fell to ~31,800 by 2022) on the low end and the 2004-2005-era figure on the high end. This national average is a scale marker only — the entry's point is the subgroup variance documented below, which ranges from roughly 1 in 524 (heterosexual men) to 1 in 2 (Black men who have sex with men).\n","uncertainty":{"low":0.008,"high":0.013},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5524204/","title":"Lifetime Risk of a Diagnosis of HIV Infection in the United States","publisher":"Annals of Epidemiology (Hess KL, Hu X, Lansky A, Mermin J, Hall HI)","source_type":"peer_reviewed","statistic":"Overall lifetime risk of an HIV diagnosis: 0.95% (95% CI 0.94-0.95), ~1 in 106. Lifetime risk 1 in 68 for males and 1 in 253 for females. Highest risk group MSM (1 in 6); lowest male heterosexuals (1 in 524). Data: 2010-2014.","excerpt":"\"Overall, the lifetime risk of a diagnosis of HIV was 0.95% (95% CI: 0.94-0.95)... This means that to observe one HIV diagnosis, 106 (95% CI: 105-106) infants would need to be followed over a lifetime... The lifetime risk for males and females was 1 in 68 and 1 in 253, respectively... The risk group with the highest lifetime risk was MSM (1 in 6)... the lowest risk was among male heterosexuals (1 in 524).\"\n","source_date":"2017-04-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260402173945/https://pmc.ncbi.nlm.nih.gov/articles/PMC5524204/","calculation_notes":"Hess et al. applied age-specific HIV-diagnosis probabilities from CDC surveillance data for 2010-2014 to a hypothetical cohort of 10 million live births to estimate cumulative lifetime risk. The published overall figure of 0.95% is used directly as the normalized value (0.0095), rounded display \"1 in 106\". No hazard compounding needed — it is already a lifetime cumulative probability. The male/female split (1/68, 1/253) arithmetically reproduces the overall figure at roughly the US 51/49 sex ratio: 0.51 × (1/68) + 0.49 × (1/253) ≈ 0.0094, an internal-consistency check on the extracted numbers.\n","independence_note":"Hess et al. (CDC Division of HIV/AIDS Prevention) is the primary peer-reviewed analysis. The CDC HIV surveillance facts-stats page and the CDC PrEP efficacy page below are CDC programmatic outputs drawing on the same surveillance system; they are used as the current-incidence and intervention-efficacy cross-references, not as independent replications of the lifetime estimate.\n"},{"url":"https://www.cdc.gov/hiv/data-research/facts-stats/index.html","title":"HIV Statistics — Facts & Stats","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"31,800 estimated new HIV infections in the US in 2022; 37,981 new diagnoses. 67% of new infections among gay, bisexual, and other men who have male-to-male sexual contact; 22% heterosexual contact; 7% people who inject drugs. The South accounted for 49% of new infections (rate 506.5 per 100,000) vs 131.6 in the Midwest.","excerpt":"\"There were 31,800 estimated new HIV infections in the US in 2022... Gay, bisexual, and other men who reported male-to-male sexual contact accounted for 67% (21,400) of the 31,800 estimated new HIV infections... 22% were among people who reported heterosexual contact... 7% were among people who inject drugs... In 2022, the South accounted for nearly half (49%) of the 31,800 estimated new HIV infections.\"\n","source_date":"2024-04-22","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260626042117/https://www.cdc.gov/hiv/data-research/facts-stats/index.html","calculation_notes":"Used as the current-incidence cross-check on the 2010-2014 lifetime projection. The decline to ~31,800 new infections per year by 2022 (from ~37,000-40,000 in the early 2010s) supports placing the lifetime point estimate at the lower end of the historical 1-in-78-to-1-in-106 range and informs the uncertainty band. The transmission-category and regional shares motivate the personal_factor_multipliers and regional_breakdown.\n"},{"url":"https://www.cdc.gov/hivnexus/hcp/prep/index.html","title":"Pre-Exposure Prophylaxis (PrEP) — HIV Nexus, Clinical Resources","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"When taken as prescribed, both oral and injectable PrEP reduce the risk of getting HIV from sex by about 99%. Oral PrEP reduces the risk of getting HIV from injection drug use by at least 74% when taken as prescribed.","excerpt":"\"When taken as prescribed, both oral and injectable PrEP reduce the risk of getting HIV from sex by about 99%... Oral PrEP has also been shown to reduce the risk of getting HIV from injection drug use by at least 74%, when taken as prescribed.\"\n","source_date":"2026-04-30","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260616164930/https://www.cdc.gov/hivnexus/hcp/prep/index.html","calculation_notes":"The ~99% sexual-acquisition reduction is the basis for the PrEP protective multiplier of ~0.01 (a roughly 99% reduction in sexual-route risk for adherent users). The 74% injection-route reduction is reported separately. \"As prescribed\" is load-bearing: real-world effectiveness is lower where adherence lapses, so the 0.01 multiplier is a best-case figure for consistent daily/on-time use.\n"},{"url":"https://www.aidsmap.com/news/feb-2016/major-disparities-persist-lifetime-risk-hiv-diagnosis-us","title":"Major disparities persist in lifetime risk of HIV diagnosis in the US","publisher":"aidsmap / NAM (Liz Highleyman)","source_type":"news_article","statistic":"Overall lifetime HIV-diagnosis risk fell from about 1 in 78 (2004-2005) to 1 in 99 (2009-2013). Black gay/bisexual men 1 in 2; Latino gay men 1 in 4; white gay men 1 in 11. All MSM 1 in 6. Lifetime risk ranged from 1 in 670 (North Dakota) to 1 in 13 (Washington, DC).","excerpt":"\"The researchers found that the overall lifetime risk of an HIV-positive diagnosis was lower than it was a decade ago, falling from about 1 in 78 during 2004-2005 to 1 in 99 during 2009-2013... The lifetime risk estimate is the cumulative probability of being diagnosed with HIV from birth to death, assuming diagnosis rates remain constant.\"\n","source_date":"2016-02-24","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260215013728/https://www.aidsmap.com/news/feb-2016/major-disparities-persist-lifetime-risk-hiv-diagnosis-us","calculation_notes":"Secondary source covering the CROI 2016 presentation of the same CDC research line at the earlier 2009-2013 data window, where the overall figure was the widely cited \"1 in 99\". Used only to document the 1-in-99 historical framing and the time trend; the 2009-2013 subgroup figures (e.g. Latino MSM 1 in 4, DC 1 in 13) are NOT mixed into the published 2010-2014 source block above. Non-authoritative type; the three CDC/peer-reviewed sources carry the entry's authority requirement.\n"}],"comparison_anchors":[{"label":"Suicide (lifetime, US adult)","lifetime_us_adult":0.00827},{"label":"Tuberculosis death (lifetime, global adult)","lifetime_us_adult":0.0092},{"label":"Type 2 diabetes death (lifetime, US adult)","lifetime_us_adult":0.075},{"label":"Major depression (lifetime, US adult)","lifetime_us_adult":0.206}],"regional_breakdown":[{"region":"US average","probability":0.0095,"notes":"Hess et al. 2017, overall lifetime risk of HIV diagnosis, 2010-2014 data (0.95%)."},{"region":"Highest-incidence Southern states","probability":0.018,"notes":"The South carries the highest burden: 49% of the 31,800 new US infections in 2022 (CDC), with a 2022 diagnosis rate of 506.5 per 100,000 vs 131.6 in the Midwest. The highest published state lifetime risks (mostly Southern) were Maryland (1 in 56), Georgia (1 in 57), Florida and Louisiana (1 in 58); ~1 in 56 (≈0.018) marks those top states, not a South-wide average, which is lower."},{"region":"Washington, DC","probability":0.059,"notes":"Highest in the US in the published 2010-2014 analysis at 1 in 17 (≈0.059); the earlier 2009-2013 window placed DC at 1 in 13. Reflects a dense urban epidemic, not a typical state."},{"region":"Lowest-incidence states (e.g. Montana, North Dakota)","probability":0.0015,"notes":"Published 2010-2014 range bottomed at 1 in 674 (Montana, ≈0.0015); the 2009-2013 window placed North Dakota at 1 in 670. Two-plus orders of magnitude below DC."}],"personal_factor_multipliers":[{"factor":"Men who have sex with men (MSM)","multiplier":17,"notes":"Hess et al.: MSM lifetime risk 1 in 6 (~0.167) vs overall ~1 in 106 (0.0095), ratio ≈ 17.5. MSM is the highest-risk transmission group; male heterosexuals are the lowest at 1 in 524."},{"factor":"Black men who have sex with men","multiplier":53,"notes":"Hess et al.: Black MSM lifetime risk 1 in 2 (~0.50) vs overall ~1 in 106, ratio ≈ 53. The single highest-risk subgroup in the analysis — a roughly 1-in-2 lifetime chance, driven by network and access disparities, not individual behavior frequency."},{"factor":"Hispanic/Latino men who have sex with men","multiplier":21,"notes":"Hess et al. (published 2010-2014): Hispanic/Latino MSM lifetime risk 1 in 5 (0.20) vs overall ~1 in 106, ratio ≈ 21. (The earlier 2009-2013 CROI window reported 1 in 4 for Latino MSM.)"},{"factor":"People who inject drugs","multiplier":3.5,"notes":"Hess et al.: lifetime risk 1 in 26 for women who inject drugs (~0.038) and 1 in 43 for men (~0.023) vs overall ~1 in 106. Ratio ≈ 4 (women) and ≈ 2.5 (men); ~3.5 is a midpoint. Injection-route HIV risk is reducible by ~74% with PrEP (CDC) and further by sterile-equipment access."},{"factor":"Black women (vs overall)","multiplier":2,"notes":"Hess et al.: Black women lifetime risk 1 in 54 (~0.0185) vs overall ~1 in 106, ratio ≈ 2. The racial disparity among women specifically is starker — Black women 1 in 54 vs white women 1 in 941, roughly 17-fold — but that within-women gap is discussed in the caveats rather than encoded here, since this multiplier is relative to the overall baseline like every other factor."},{"factor":"Consistent condom use","multiplier":0.2,"notes":"Protective. Meta-analyses of serodiscordant couples find consistent condom use reduces sexual HIV transmission by roughly 80% (a ~0.2 multiplier on the sexual-acquisition route); CDC describes condoms as highly effective when used correctly every time. Less protective than PrEP, and route-specific."},{"factor":"PrEP taken as prescribed","multiplier":0.01,"notes":"Protective. CDC: oral/injectable PrEP reduces the risk of getting HIV from sex by about 99% when taken as prescribed (a ~0.01 multiplier on the sexual route). Best-case figure for consistent adherence; real-world effectiveness drops when doses are missed. Reduces injection-route risk by at least 74%."}],"short_label":"HIV diagnosis","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline 0.95% (~1 in 106) is a national-average lifetime risk projected from 2010-2014 CDC surveillance assuming diagnosis rates hold constant — not a current snapshot, and the rate has continued to decline (US new infections fell to ~31,800 by 2022). The far more important caveat is heterogeneity: the average is the least informative number in this entry. Lifetime risk spans more than two orders of magnitude across subgroups — roughly 1 in 524 for heterosexual men, 1 in 2 for Black men who have sex with men — and across geography, from about 1 in 674 in the lowest-incidence states to about 1 in 17 in Washington, DC. The published paper measures the probability of a diagnosis, which lags actual infection; undiagnosed infections are not fully captured. Race/ethnicity multipliers reflect structural and network factors (testing access, treatment-as-prevention coverage, partner-pool prevalence), not individual behavior, and should be read as epidemiological strata rather than personal traits. The protective multipliers (condoms ~0.2, PrEP ~0.01) are route-specific and assume correct, consistent use; the PrEP figure is a best-case adherence number. With modern antiretroviral therapy a diagnosed US adult who stays in care has a near-normal life expectancy and is effectively non-infectious (undetectable = untransmittable), which is why this entry is filed as a chronic illness rather than a fatal outcome.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"An abstract awareness ribbon loop rendered as a single continuous line beside a faint grid, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/hiv-diagnosis-lifetime","api_url":"https://likelier.app/api/fears/hiv-diagnosis-lifetime.json"},{"slug":"child-death-before-18","question":"What are the odds of a child dying before age 18 in the US?","quick_answer":"About 1 in 103 US children die before turning 18, with most of that risk concentrated in the first year of life. Parents tend to intuit a figure far higher — few would guess below 5% — but the true cumulative rate is close to 1%.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"No formal survey asks parents to estimate the probability that their child will die before adulthood, but the fear scarcely needs quantification. It anchors an entire industry of safety products, drives helicopter parenting norms, and dominates pediatrician waiting-room anxieties. The availability heuristic does most of the work: every child death reported on the news feels like evidence that the baseline rate is high, precisely because such deaths are rare enough to be individually newsworthy. Most parents, if pressed, would guess a number far above the actual figure — some studies on domain-specific child risks (drowning, abduction, anaphylaxis) find parental estimates 10x to 100x the true rate.\n","rough_estimate":"Parents intuit the risk is far higher than it is; few could name a number below 5%","kind":"intuition"},"native":{"display":"~9.7 deaths per 1,000 live births before age 18 (US, 2022)","numerator":971,"denominator":100000,"unit":"per live birth, cumulative to age 18","population":"US children born alive"},"normalized":{"lifetime_us_adult":0.0097,"display":"1 in ~103 (cumulative, birth to age 18, US)","log_value":-2.01,"assumptions":"Derived from the CDC / NCHS 2022 United States Life Tables (NVSR Vol. 74, No. 2), Table 1 (total population). Out of a hypothetical cohort of 100,000 live births (both sexes combined), 99,029 survive to exact age 18 (the lx entry for the 18-19 age interval), implying 100,000 − 99,029 = 971 deaths per 100,000, or ~9.7 per 1,000. The World Bank / UNICEF indicator of ~7 per 1,000 covers only under-5 mortality, a subset of the birth-to-18 window; the life table's own lx at age 5 (99,328) puts under-5 mortality at ~6.7 per 1,000, so the full birth-to-18 figure necessarily exceeds it. The figure is a period life-table estimate, not a cohort projection — it assumes 2022 age-specific mortality rates persist unchanged. Roughly 58% of the cumulative risk is concentrated in the first year of life (560 of 971 deaths per 100,000; infant mortality rate 5.61 per 1,000 in 2023). The uncertainty band reflects year-to-year variation, racial/ethnic disparities (Black infant IMR ~10.9 vs Asian ~3.4 per 1,000), and state-level range (New Hampshire 2.9 vs Mississippi 8.9 per 1,000 for infants alone).\n","uncertainty":{"low":0.007,"high":0.012},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/data/nvsr/nvsr74/nvsr74-02.pdf","title":"United States Life Tables, 2022","publisher":"National Center for Health Statistics, CDC (NVSR Vol. 74, No. 2)","source_type":"govt_report","statistic":"Out of 100,000 live births (both sexes), 99,029 survive to exact age 18 in the 2022 period life table (lx for the 18-19 age interval)","excerpt":"\"Table 1. Life table for the total population: United States, 2022 [...] Age 18-19: qx = 0.000680, lx = 99,029, dx = 10.\" The lx column shows the number of persons surviving to the beginning of the indicated age interval out of 100,000 live births.\n","source_date":"2025-04-08","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260402043823/https://www.cdc.gov/nchs/data/nvsr/nvsr74/nvsr74-02.pdf","calculation_notes":"CDC 2022 life tables, Table 1 (total population): lx at age 0 = 100,000; lx at the 18-19 interval (exact age 18) = 99,029 (both sexes combined). Deaths before 18 = 100,000 − 99,029 = 971. Probability = 971 / 100,000 = 0.0097, or ~9.7 per 1,000 (about 1 in 103). Cross-check within the same table: lx at age 1 = 99,440 (infant mortality 560 per 100,000 = 5.6 per 1,000) and lx at age 5 = 99,328 (under-5 mortality 672 per 100,000 = 6.7 per 1,000). Infancy accounts for ~58% of the cumulative birth-to-18 deaths (560 of 971).\n","independence_note":"Primary US life-table source built from NCHS/NVSS death-certificate data and Census Bureau mid-year population estimates. The age-specific death rates from CDC FastStats and the World Bank / UNICEF indicator are derived from the same underlying vital registration system.\n"},{"url":"https://data.worldbank.org/indicator/SH.DYN.MORT","title":"Mortality rate, under-5 (per 1,000 live births)","publisher":"World Bank, sourced from UN Inter-agency Group for Child Mortality Estimation","source_type":"reputable_reference","statistic":"US under-5 mortality rate: ~6.5 per 1,000 live births (2023); global average: 36.7 per 1,000","excerpt":"\"The under-five mortality rate is the probability per 1,000 that a newborn baby will die before reaching age five, if subject to age-specific mortality rates of the specified year.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420033126/https://data.worldbank.org/indicator/SH.DYN.MORT","calculation_notes":"US under-5 mortality of ~6.5-7 per 1,000 is a subset — not the whole — of the cumulative birth-to-18 mortality (~9.7 per 1,000 from the life table). Under-5 alone is the floor; the additional ~3 per 1,000 accumulates from ages 5-17, driven by the low but nonzero school-age rate and the marked uptick in the teenage years (accidents, firearms, suicide). Because under-5 is a subset, the under-5 indicator and the birth-to-18 headline cannot be equal. Global average of 36.7 per 1,000 under-5 provides cross-country context.\n","independence_note":"World Bank compiles estimates from the UN IGME, which uses a Bayesian model fit to country-reported vital registration data. For the US, the underlying data source is the same NVSS system as the CDC life tables, but the statistical methodology differs.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db492.htm","title":"Mortality in the United States, 2022","publisher":"National Center for Health Statistics, CDC (Data Brief No. 492)","source_type":"govt_report","statistic":"Age-specific death rates in 2022: ages 1-4: 28.0 per 100,000; ages 5-14: 15.3 per 100,000; ages 15-24: 79.5 per 100,000","excerpt":"\"In 2022, age-adjusted death rates decreased from 2021 for the total population [...] Age-specific death rates increased from 2021 to 2022 for age groups 1–4 and 5–14 years.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260328023755/https://www.cdc.gov/nchs/products/databriefs/db492.htm","calculation_notes":"Age-band rates used to cross-check the life-table cumulative. Infant rate ~560 per 100,000 (5.6 per 1,000) dominates. Ages 1-4: 28.0 per 100,000/year × 4 years ≈ 112 per 100,000 cumulative. Ages 5-14: 15.3 × 10 ≈ 153. Ages 15-17 (estimated as lower end of 15-24 bracket): ~60 per 100,000/year × 3 ≈ 180. Sum: 560 + 112 + 153 + 180 ≈ 1,005 per 100,000, or ~1.0% — closely matching the life-table figure of 971 per 100,000 (~0.97%). This independent age-band cross-check corroborates the ~9.7 per 1,000 birth-to-18 headline derived from the life table's lx at age 18.\n","independence_note":"Same NCHS/NVSS vital statistics as the life tables but reported as crude age-specific rates in a different publication format. Not independent.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4838a2.htm","title":"Achievements in Public Health, 1900-1999: Healthier Mothers and Babies","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"In 1900, up to 30% of infants in some US cities died before their first birthday; infant mortality was approximately 100 per 1,000 live births","excerpt":"\"In 1900 in some U.S. cities, up to 30% of infants died before reaching their first birthday.\"\n","source_date":"1999-10-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260321202407/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4838a2.htm","calculation_notes":"The 1900 infant mortality rate of ~100 per 1,000 live births (with up to 30% infant mortality in some cities) anchors the historical comparison. With life expectancy near 47 years and high child mortality from infectious disease, roughly 20% of children born alive did not survive to age 5. Current US figure of ~6.5 per 1,000 under-5 represents a >95% reduction over 120 years.\n"}],"comparison_anchors":[{"label":"SIDS (per US live-born infant, first year)","lifetime_us_adult":0.00014},{"label":"Child pool drowning (ages 0-14, US)","lifetime_us_adult":0.000435},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.00000478},{"label":"Stranger child abduction (per US child per year)","lifetime_us_adult":0.0000015}],"regional_breakdown":[{"region":"Infant (0-1 year)","probability":0.0056,"notes":"~58% of all under-18 mortality (560 of 971 deaths per 100,000); 5.61 per 1,000 live births (2023)"},{"region":"Toddler (1-4 years)","probability":0.00112,"notes":"28.0 per 100,000/year × 4 years; leading cause: unintentional injury (drowning, motor vehicles)"},{"region":"School age (5-14 years)","probability":0.00153,"notes":"15.3 per 100,000/year × 10 years; leading causes: accidents, cancer, congenital anomalies"},{"region":"Teen (15-17 years)","probability":0.0018,"notes":"~60 per 100,000/year × 3 years; leading causes: firearms, motor vehicles, suicide"},{"region":"US overall (birth to 18)","probability":0.0097},{"region":"Sub-Saharan Africa (under-5 alone)","probability":0.069,"notes":"69 per 1,000 live births; ~7x the entire US birth-to-18 figure"},{"region":"US in 1900 (under-5 alone)","probability":0.2,"notes":"~200 per 1,000; roughly 1 in 5 children did not reach age 5"}],"personal_factor_multipliers":[{"factor":"Premature birth (<37 weeks)","multiplier":3.5,"notes":"Preterm birth is the leading contributor to infant mortality; risk concentrated in first year"},{"factor":"Black non-Hispanic infant","multiplier":1.9,"notes":"Infant mortality rate 10.93 vs 5.61 overall per 1,000 (CDC 2023)"},{"factor":"Asian non-Hispanic infant","multiplier":0.6,"notes":"Infant mortality rate 3.44 per 1,000 (CDC 2023)"},{"factor":"Low-income household","multiplier":1.5,"notes":"Poverty is associated with higher child mortality across all age groups via access to care, environmental hazards, and injury risk"},{"factor":"State: Mississippi","multiplier":1.6,"notes":"Highest state infant mortality rate (8.94 per 1,000 vs national 5.61)"},{"factor":"State: New Hampshire","multiplier":0.5,"notes":"Lowest state infant mortality rate (2.93 per 1,000)"}],"short_label":"Child death (<18)","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"bereavement","valence":"negative","caveats":"The headline ~9.7-per-1,000 figure is a period estimate from the 2022 US life table (Table 1, lx = 99,029 at exact age 18), not a cohort projection for children born in 2022. Note this exceeds the commonly cited World Bank under-5 rate of ~7 per 1,000 because under-5 is only a subset of the birth-to-18 window. If age-specific mortality continues its long-term decline, children born today will likely face even lower cumulative risk. Conversely, the 2021-2022 data showed the first year-over-year increase in infant mortality in two decades — partly attributable to the Dobbs v. Jackson decision's effects on access to reproductive care in some analyses, though causality remains debated. The regional_breakdown by age group uses approximate cumulative rates (annual rate × years) rather than exact life-table multiplication, so they sum to slightly more than the life-table total. The teen bracket uses an estimated 60 per 100,000/year for ages 15-17, extracted from the lower portion of the CDC's 15-24 age band. Racial and socioeconomic disparities are substantial and persistent: Black infant mortality remains roughly double the national average, and state-level variation spans a 3x range.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-research-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A simple growth chart line rising steadily from left to right against a soft grey background, flat vector illustration."},"canonical_url":"https://likelier.app/child-death-before-18","api_url":"https://likelier.app/api/fears/child-death-before-18.json"},{"slug":"dying-without-heir","question":"What are the odds of dying with no one to inherit your estate?","quick_answer":"The lifetime probability of dying with no identifiable heir is about 1 percent for a US adult. Most people assume it is vanishingly rare, though a guess of 1 to 5 percent is roughly right.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Dying without anyone at all to inherit is the kind of fear that surfaces in estate planning advertisements and loneliness discourse but rarely in rigorous surveys. Most people assume it happens only to hermits or the very elderly. The cultural image is of an eccentric recluse whose fortune escheats to the state -- a rare curiosity, not a plausible personal outcome. In practice, intestacy law casts a very wide net for heirs (second cousins, half-siblings, step-relations in some jurisdictions), which makes the \"truly no one\" scenario considerably rarer than the \"no will\" scenario that dominates headlines.\n","rough_estimate":"~1-5% lifetime guess, most people assume it is vanishingly rare","kind":"intuition"},"native":{"display":"~1% of US adults 55+ have essentially no relatives at all alive","numerator":10,"denominator":1000,"unit":"lifetime","population":"US adults age 55 and older"},"normalized":{"lifetime_us_adult":0.01,"display":"~1% lifetime probability of dying with no identifiable heir (US adult)","log_value":-2,"assumptions":"The \"no one to inherit\" scenario requires the intersection of several conditions: no surviving spouse, no children, no parents, no siblings, no nieces/nephews, and no other identifiable relatives within the intestate succession hierarchy. Research from Penn State (Margolis and Verdery, analyzing HRS/NLSY survey data 1998-2010) found that 6.6% of US adults 55+ have neither a spouse nor biological children alive, and that just 1% of older Americans have essentially no relatives at all alive -- including a spouse, partner, children, or biological parents or siblings. That 1% \"no relatives at all\" figure is the one that maps to the escheatment scenario; the 6.6% \"no spouse or biological children\" figure is much broader because many of those adults still have siblings, nieces, nephews, or parents who inherit under intestacy law. We therefore take ~1% as the central lifetime probability, directly from the Penn State source's stated \"1 percent\" figure for older adults with essentially no living relatives. Intestacy statutes in all 50 states search exhaustively for distant relatives before escheatment occurs, and the Caring.com 2025 survey found only 24% of Americans have a will (so ~76% die intestate) -- but intestacy does not mean no heir, it means the state assigns heirs by statute. True escheatment (property reverting to the state for lack of any identifiable heir) is described in legal literature as \"rare.\" The uncertainty band is asymmetric and wide because \"no identifiable heir\" depends on how hard the state searches and how distant a relative counts, and because the kinless share is projected to rise for future cohorts (Penn State projects the share of non-Hispanic whites lacking close kin to roughly double by 2060). The lower bound reflects that professional heir-search firms locate relatives even for many nominally kinless decedents.\n","uncertainty":{"low":0.005,"high":0.03},"scope":"us_adult_lifetime"},"sources":[{"url":"https://sociology.la.psu.edu/news/americans-face-a-rising-risk-of-dying-alone/","title":"Americans Face a Rising Risk of Dying Alone","publisher":"Penn State Department of Sociology and Criminology","source_type":"reputable_reference","statistic":"6.6% of US adults 55+ have neither a spouse nor biological children alive; just 1% of older Americans have essentially no relatives at all","excerpt":"\"They found that 6.6 percent of U.S. adults 55 and older have neither a spouse nor biological children still alive. Just 1 percent of older Americans have essentially no relatives at all alive, including a spouse, partner, children, or biological parents or siblings.\"\n","source_date":"2023-06-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251215013541/https://sociology.la.psu.edu/news/americans-face-a-rising-risk-of-dying-alone/","calculation_notes":"The Penn State research (Margolis and Verdery) analyzed survey data from 1998-2010. The 6.6% figure captures those without a spouse or biological children -- but many of these individuals still have siblings, nieces, nephews, or parents who would inherit under intestacy law. The source's \"1 percent\" figure -- older Americans with essentially no living relatives at all (no spouse, partner, children, or biological parents or siblings) -- is the one that maps to the escheatment scenario, and it is used directly as the ~1% central estimate. The article projects this kinless share will rise sharply for future cohorts (the share of non-Hispanic whites lacking close kin is projected to roughly double by 2060), which motivates the upward-skewed uncertainty band rather than a higher point estimate.\n","independence_note":"Uses NLSY and HRS longitudinal data. Independent from Census cross-sectional fertility tables and Pew attitudinal surveys.\n"},{"url":"https://www.pewresearch.org/social-trends/2024/07/25/the-experiences-of-u-s-adults-who-dont-have-children/","title":"The Experiences of U.S. Adults Who Don't Have Children","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"The share of childless US adults under 50 who say they are unlikely to ever have kids rose from 37% in 2018 to 47% in 2023","excerpt":"\"The share of U.S. adults younger than 50 without children who say they are unlikely to ever have kids rose 10 percentage points between 2018 and 2023 (from 37% to 47%).\"\n","source_date":"2024-07-25","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035005/https://www.pewresearch.org/social-trends/2024/07/25/the-experiences-of-u-s-adults-who-dont-have-children/","calculation_notes":"Pew's 2024 survey documents rising expected childlessness, but childlessness alone does not mean dying without an heir -- most childless adults have siblings, parents, or other relatives. The 47% figure is the share of childless adults under 50 who say they are unlikely to ever have kids (up from 37% in 2018); it is not the proportion of all under-50 adults who currently have no children. For women 45-50 (near the end of fertility), the Census Bureau reports 14.9% are childless as of 2024, down from 16.7% in 2014. The Pew data is used here as a trend indicator: a growing share of childless younger adults expect to remain childless, which, combined with declining marriage rates and smaller family sizes, will increase the share of future elderly with no close relatives. This supports an upward-skewed uncertainty band relative to the current ~1% observed kinless-with-no-relatives rate.\n","independence_note":"Pew conducts its own nationally representative surveys using the American Trends Panel. Independent methodology from Penn State's longitudinal analysis.\n"},{"url":"https://www.caring.com/resources/wills-survey","title":"2025 Wills and Estate Planning Study","publisher":"Caring.com","source_type":"reputable_reference","statistic":"Only 24% of Americans have a will in 2025, down from 32% in 2024","excerpt":"\"An estimated 76% of Americans die without a will. Only 24% of Americans have a will in 2025, down from 32% in 2024.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035040/https://www.caring.com/resources/wills-survey","calculation_notes":"The 76% intestacy rate is frequently conflated with \"dying without an heir,\" but these are fundamentally different conditions. Intestacy means the state assigns heirs according to a statutory hierarchy (spouse > children > parents > siblings > nieces/nephews > grandparents > aunts/uncles > cousins, etc.). In nearly all intestate cases, an heir exists somewhere in this chain. True escheatment -- where the state inherits because no heir can be found -- occurs only when the entire hierarchy is exhausted. States collectively hold ~$70 billion in unclaimed property, but most of this is dormant bank accounts and uncashed checks, not escheated estates. The Caring.com data is used here to establish the denominator of the problem: most Americans do not plan their estates, but most still have statutory heirs.\n","independence_note":"Caring.com conducts annual online surveys with Harris Poll methodology. Independent from academic longitudinal studies and Pew surveys.\n"}],"comparison_anchors":[{"label":"Dying in a house fire (lifetime, US)","lifetime_us_adult":0.00091},{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1}],"personal_factor_multipliers":[{"factor":"never married and childless (age 55+)","multiplier":4,"notes":"Penn State / HRS analysis (Margolis and Verdery) found 6.6% of US adults 55+ lack both a spouse and biological children, versus just 1% with essentially no relatives at all. Adults who are never-married and childless sit far closer to the kinless tail than the general 55+ population -- married adults with children have essentially zero escheatment risk -- so their probability of dying with no identifiable heir is several-fold above the ~1% baseline."},{"factor":"LGBTQ+ elder (age 60+)","multiplier":2,"notes":"SAGE / Williams Institute 2010 research found LGBTQ+ adults are roughly twice as likely to be single and four times less likely to have children than heterosexual adults of comparable age, substantially increasing kinlessness exposure. Confirmed direction by Fredriksen-Goldsen et al. (Gerontologist 2011)."},{"factor":"female, widowed, age 80+","multiplier":2,"notes":"Women outlive men by ~5 years (CDC NCHS life tables) and are more likely to outlive all close family members. The 80+ age band shows steep attrition of surviving siblings and nieces/nephews; Census 2020 shows women are 60% of the 80+ population but a higher share of those living alone."},{"factor":"has a will with named beneficiaries","multiplier":0.1,"notes":"A valid will with living named beneficiaries eliminates the intestate-escheatment pathway entirely; only deaths of beneficiaries after the will was drafted but before the testator create residual risk. This is not truly a population multiplier but the single most actionable factor."}],"short_label":"Dying without heir","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","caveats":"\"No one to inherit\" is a spectrum, not a binary. Intestacy statutes in all US states search for relatives out to very distant degrees of kinship before escheatment occurs. A person with a living third cousin they have never met is not, legally, dying without an heir. The ~1% central estimate refers to the stricter scenario: no identifiable relative can be found through reasonable diligent search. The number is rising because of converging demographic trends -- increasing childlessness, declining marriage rates, smaller sibship sizes, and longer lifespans that allow individuals to outlive their entire family network. The Penn State figure that just 1% of older Americans have essentially no living relatives at all is the most relevant data point, but even this may overstate the escheatment rate because professional heir-search firms often locate relatives the decedent did not know existed. The emotional dimension -- dying without anyone who *cares* about inheriting, as opposed to anyone who *legally can* -- is not captured by these statistics and is probably a larger number.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An empty wooden chair beside a small table with a single key on it, muted warm tones, flat vector illustration."},"canonical_url":"https://likelier.app/dying-without-heir","api_url":"https://likelier.app/api/fears/dying-without-heir.json"},{"slug":"lasik-eye-surgery","question":"What are the odds of serious complications from LASIK eye surgery?","quick_answer":"Fewer than 1 in 100 LASIK patients suffer debilitating difficulty with usual activities from a visual symptom, and over 95% are satisfied with their vision. Many prospective patients expect permanent sight loss to be the typical result; transient dry eye and halos are common but mostly resolve within 6-12 months.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"LASIK anxiety runs deep, partly because the procedure involves a laser reshaping the cornea while the patient is awake and conscious. Prospective patients often picture permanent, sight-threatening loss as the typical outcome, when the FDA's PROWL studies found fewer than 1 in 100 participants experienced debilitating difficulty with usual activities from any one visual symptom. Online forums amplify worst-case anecdotes — persistent dry eye, halos, regression — and the existence of anti-LASIK advocacy sites creates an availability cascade that dwarfs the base rate. The same PROWL studies found that up to 28% of patients with no pre-operative dry eye reported symptoms at three months, and up to 40% reported new halos at three months, but these transient side effects are routinely conflated with permanent vision-threatening complications, which are rarer still.\n","rough_estimate":"Many prospective patients expect permanent sight-threatening loss to be the typical outcome","kind":"intuition"},"native":{"display":"<1 in 100 per procedure (debilitating symptoms)","numerator":1,"denominator":100,"unit":"per LASIK procedure","population":"US LASIK patients, all myopia corrections"},"normalized":{"lifetime_us_adult":0.01,"display":"<1 in 100 lifetime","log_value":-2,"assumptions":"The only figure the cited sources actually state for serious outcomes is the FDA PROWL finding that \"less than 1 percent of study participants experienced 'a lot of difficulty' with or inability to do usual activities\" because of any one visual symptom after LASIK — i.e. debilitating outcomes occur in fewer than 1 in 100 patients. The point estimate is set at this stated ceiling (0.01); the true rate is below it, and the lower bound of 0.001 reflects that per-symptom debilitating rates are lower still. Most adults who undergo LASIK do so once in a lifetime, so the activity-specific scope reflects the risk per person who elects the procedure, not the general population. Approximately 10 million Americans have had LASIK since FDA approval in 1999 out of ~260 million adults, so the unconditional population probability is far lower, but the relevant question is the risk conditional on choosing the surgery.\n","uncertainty":{"low":0.001,"high":0.01},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.fda.gov/medical-devices/lasik/lasik-quality-life-collaboration-project","title":"LASIK Quality of Life Collaboration Project (PROWL)","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"Less than 1% experienced debilitating difficulty with usual activities; more than 95% satisfied with vision; up to 28% reported new dry eye symptoms and 40% new halos at 3 months post-op","excerpt":"\"In each of the PROWL studies, less than 1 percent of patients experienced difficulty performing usual activities following LASIK surgery due to any one symptom. ... Up to 28 percent of participants with no symptoms of dry eyes before LASIK, reported dry eye symptoms at three months after their surgery. Up to 40 percent of participants with no halos before LASIK had halos three months following surgery. ... More than 95% of participants were satisfied with their vision following LASIK surgery.\"\n","source_date":"2021-06-17","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260307060850/https://www.fda.gov/medical-devices/lasik/lasik-quality-life-collaboration-project","calculation_notes":"The FDA PROWL studies (PROWL-1: 262 active-duty military, PROWL-2: 312 civilians) are prospective studies designed to measure patient-reported outcomes. The strongest serious-outcome figure the FDA states is that \"less than 1 percent of study participants experienced 'a lot of difficulty' with or inability to do usual activities\" because of any one visual symptom — this is the basis for the <1 in 100 native rate. Satisfaction is stated as \"more than 95%\" (the specific 96-98% band is not on this FDA page; it appears in the underlying JAMA PROWL publications). Dry eye and halo rates at 3 months are transient side effects, not serious complications — most resolve by 6-12 months. PROWL's sample size is too small to precisely estimate rarer sight-threatening events such as ectasia, so no tighter rate is asserted.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4302464/","title":"Functional Outcome and Patient Satisfaction after Laser In Situ Keratomileusis for Correction of Myopia and Myopic Astigmatism","publisher":"Middle East African Journal of Ophthalmology (PMC)","source_type":"peer_reviewed","statistic":"98.5% of patients satisfied or very satisfied with their surgery; uncorrected visual acuity 20/40 or better in 99% of eyes","excerpt":"\"A total of 98.5% of patients was satisfied or very satisfied with their surgery. Postoperatively, the uncorrected visual acuity was 20/40 or better in 99% of eyes and 96% of eyes were within ± 1.00 D of the targeted correction.\"\n","source_date":"2014-10-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260304171613/https://pmc.ncbi.nlm.nih.gov/articles/PMC4302464/","calculation_notes":"This peer-reviewed study documents high patient satisfaction (98.5% satisfied or very satisfied) and excellent functional outcomes (20/40 or better uncorrected acuity in 99% of eyes). It reports satisfaction and refractive outcomes, not a quantified serious-complication rate; the native rate here is anchored to the FDA PROWL \"less than 1 percent\" debilitating-outcome figure, with this study supplying the satisfaction and efficacy context.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7982707/","title":"The 25th Anniversary of Laser Vision Correction in the United States","publisher":"Clinical Ophthalmology (PMC)","source_type":"peer_reviewed","statistic":"Over 10 million patients treated with laser vision correction in the US over 25 years; complication rates have declined with each technology generation","excerpt":"\"An estimated 20 to 25 million laser vision correction procedures represent 10 to 15 million patients treated in the past 25 years. Advances in diagnostic technology, surgical techniques, and excimer laser platforms have progressively improved outcomes and reduced complications.\"\n","source_date":"2021-03-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260310134004/https://pmc.ncbi.nlm.nih.gov/articles/PMC7982707/","calculation_notes":"Provides the population denominator: ~10-15 million US patients over 25 years, and notes that complication rates have declined with each technology generation. It does not state a specific serious-complication rate. This is a contextual source for procedure volume and the improving-safety trend, not a direct complication rate source.\n"}],"comparison_anchors":[{"label":"General anesthesia death (lifetime, US)","lifetime_us_adult":0.00002},{"label":"Scuba diving fatality (per-participant lifetime)","lifetime_us_adult":0.0005}],"personal_factor_multipliers":[{"factor":"High myopia (> -8 diopters)","multiplier":3,"notes":"Higher corrections involve more corneal tissue removal, increasing ectasia risk"},{"factor":"Thin corneas (< 500 microns)","multiplier":4,"notes":"Insufficient residual stromal bed is the primary risk factor for post-LASIK ectasia"},{"factor":"Low-to-moderate myopia (< -4 diopters)","multiplier":0.3,"notes":"Lower corrections have the best outcomes and lowest complication rates"},{"factor":"Pre-existing dry eye syndrome","multiplier":3.5,"notes":"FDA LASIK consumer guidance and peer-reviewed outcome studies: pre-operative dry eye is a primary risk factor for chronic post-LASIK dry eye; the FDA patient checklist identifies it as a contraindication or high-risk condition"},{"factor":"High-volume surgeon (>1,000 procedures/year)","multiplier":0.4,"notes":"ASCRS outcome literature: surgeon procedure volume is consistently associated with complication rates; surgeons performing >1,000 procedures/year show materially lower ectasia and re-treatment rates than surgeons performing <200/year"}],"short_label":"LASIK complications","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The \"less than 1 in 100\" headline is the FDA PROWL figure for debilitating outcomes — patients who experienced \"a lot of difficulty\" with or inability to do usual activities because of any one visual symptom after LASIK. It is an upper bound: the true rate is below 1%, and PROWL's sample (574 participants across the two studies) is too small to precisely estimate rarer sight-threatening events such as post-LASIK ectasia or significant permanent loss of best-corrected visual acuity, which the cited sources do not quantify. Transient dry eye (up to 28% at 3 months) and halos (up to 40% at 3 months) are common but mostly self-resolving within 6-12 months and are not counted in the debilitating-outcome figure. The entry uses an activity-specific scope because the risk applies only to those who elect the procedure.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A stylized eye with a thin beam of light crossing the iris, rendered in muted teal and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/lasik-eye-surgery","api_url":"https://likelier.app/api/fears/lasik-eye-surgery.json"},{"slug":"sports-betting-financial-ruin","question":"What are the odds of severe financial harm from regular sports betting?","quick_answer":"About 1 in 100 US adults will suffer severe financial harm from regular sports betting over a lifetime; in states that legalized online betting, bankruptcy rates rose 25–30%. The harm is often invisible until it reaches a crisis.\n","category":"other","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Sports betting is widely framed as entertainment — a minor stake on a game you would have watched anyway. The post-2018 expansion of legal, mobile-first sportsbooks has normalized wagering in ways that lottery tickets and casino trips never quite did, in part because the products are embedded in sports-media consumption rather than requiring a trip to a dedicated venue. Marketing language leans on \"responsible gaming\" disclosures while simultaneously promoting parlay products with house edges of 15–30%. Most regular bettors do not identify as problem gamblers, and the financial harm from sports betting is largely invisible until it reaches a crisis — credit exhausted, debt unserviceable — rather than accumulating in a way that triggers earlier intervention.\n","rough_estimate":"~1 in 100 US adults over a lifetime; ~1-2% of regular bettors face serious harm","kind":"intuition"},"native":{"display":"States legalizing online sports betting saw bankruptcy rates rise 25–30% and debt collections rise 8% (Hollenbeck, Larsen & Proserpio 2024)","numerator":28,"denominator":100,"unit":"percent increase in bankruptcy rate in states legalizing online sports betting, relative to control states","population":"US adults in states that legalized online sports betting post-2018 Murphy v. NCAA, credit bureau panel 2016–2023"},"normalized":{"lifetime_us_adult":0.01,"display":"~1 in 100 US adults will experience severe financial harm from regular sports betting over their lifetime","log_value":-2,"assumptions":"Two-factor estimate: (A) P(US adult becomes a regular sports bettor) and (B) P(severe financial harm | regular sports bettor). Factor A: The NY Fed (2026) reports approximately 3% of the US population took up sports betting after their state legalized it, concentrated in younger male demographics. However, the 2018 expansion covered 38+ states by 2025 and the lifetime participation rate is higher than any single cross-section; we estimate 5% of US adults will be regular sports bettors at some point in their lifetime (including states that have not yet legalized, and accounting for future expansion). Factor B: Hollenbeck, Larsen, and Proserpio (2024) find a 25–30% increase in the probability of bankruptcy filing in online-betting states. If the baseline bankruptcy rate among adults in betting-adjacent demographics is ~8% (consistent with the US lifetime bankruptcy rate of ~10%), a 25–30% relative increase implies approximately 10–12% absolute severe-harm rate among regular bettors (using bankruptcy and equivalent severe debt as the indicator). Baker et al. (NBER w33108) find that credit card debt increases, available credit decreases, and overdraft frequency rises among bettors, with effects concentrating among financially constrained households. Combined: 0.05 × 0.20 = 0.010. The 0.20 factor reflects a conservative conversion of the Hollenbeck relative-risk increase to an absolute rate among the regular-bettor subgroup (not the general population). Uncertainty range: 0.004 (3% participation × 12% harm rate among heavy users) to 0.020 (7% participation × 28% severe-harm rate at the upper confidence interval). \"Severe financial harm\" is defined as bankruptcy filing, debt collections exceeding $5,000, or credit-score decline >50 points attributable to sports-betting activity.\n","uncertainty":{"low":0.004,"high":0.02},"scope":"us_adult_lifetime"},"sources":[{"url":"https://ssrn.com/abstract=4903302","title":"The Financial Consequences of Legalized Sports Gambling","publisher":"SSRN (Hollenbeck, Larsen & Proserpio)","source_type":"peer_reviewed","statistic":"Online sports gambling legalization raised bankruptcy probability 25–30%; average credit score declined ~12 points; debt collections, auto delinquencies, and debt consolidation loans all rose significantly","excerpt":"\"In states allowing online sports gambling, the likelihood of a personal bankruptcy filing rose 25% to 30% in the years after legalization. General access to sports betting was associated with a modest decline in average credit scores (0.7 points), while online sports gambling led to a substantially larger decline (about 12 points). The study found substantial increases in average bankruptcy rates, debt sent to collections, use of debt consolidation loans, and auto loan delinquencies.\"\n","source_date":"2024-07-26","source_accessed":"2026-05-04","calculation_notes":"This study provides the relative-risk anchor for the native display figure and for Factor B in the normalized estimate. The 25–30% relative increase in bankruptcy probability is a causal estimate from a staggered difference-in-differences design using ~4.38 million anonymized credit-bureau records. Critically, the effect is substantially stronger for online (vs. retail) betting, which is now the dominant form of legal sports wagering. The 28% mid-range is used as the native numerator. Converting this relative estimate to an absolute harm rate among regular bettors requires knowledge of baseline harm rates, which is imputed from the US adult lifetime bankruptcy rate of approximately 10%.\n","independence_note":"Credit bureau panel data (University of California Credit Panel) is independent from both the NBER Baker et al. household transaction data and from NCPG survey instruments, providing a third distinct measurement approach.\n"},{"url":"https://www.nber.org/papers/w33108","title":"Gambling Away Stability: Sports Betting's Impact on Vulnerable Households","publisher":"National Bureau of Economic Research (NBER Working Paper 33108)","source_type":"peer_reviewed","statistic":"Sports betting reduces savings, increases credit card debt, decreases available credit, and raises overdraft frequency; effects concentrate among financially constrained households","excerpt":"\"We find that the increase in sports betting does not displace other gambling or consumption but significantly reduces savings, as risky bets crowd out positive expected value investments. These effects concentrate among financially constrained households, as credit card debt increases, available credit decreases, and overdraft frequency rises.\"\n","source_date":"2024-10-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260525162520/https://www.nber.org/papers/w33108","calculation_notes":"Baker, Balthrop, Johnson, Kotter, and Pisciotta (NBER w33108) use household transaction data with a staggered difference-in-differences framework to establish causal effects at the transaction level. This study is complementary to Hollenbeck et al.: where Hollenbeck uses credit bureau outcomes (bankruptcy, credit score), Baker et al. trace the mechanism through spending and savings decisions. Together they establish both the mechanism and the downstream financial consequence of regular sports betting.\n","independence_note":"NBER w33108 uses household bank-transaction-level data, methodologically distinct from the Hollenbeck et al. credit bureau panel. Both teams applied staggered difference-in-differences to the post-2018 state-level legalization variation, providing independent identification of the same causal effect.\n"},{"url":"https://www.cnbc.com/2026/03/25/sports-betting-credit-health-ny-fed.html","title":"As March Madness unfolds, NY Fed highlights sports betting toll on consumer credit health","publisher":"CNBC / Federal Reserve Bank of New York","source_type":"news_article","statistic":"Credit delinquency rates rose ~0.3% overall in states where sports betting is legal; among the 3% who took up betting, delinquencies spiked more than 10%","excerpt":"\"Credit delinquency rates rose about 0.3% overall in states where sports betting is legal, despite legal sports bettors making up only 3% of the population. But, looking only at the 3% of the population who took up sports betting after their state legalized it, credit delinquencies spiked by more than 10% among gamblers.\"\n","source_date":"2026-03-25","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260517175609/https://www.cnbc.com/2026/03/25/sports-betting-credit-health-ny-fed.html","calculation_notes":"The NY Fed data provides the participation rate estimate used in Factor A: approximately 3% of the population became sports bettors post-legalization in each state. The 10%+ delinquency spike among that 3% is a behavioral concentration effect consistent with Hollenbeck et al. and Baker et al. The NY Fed figure is used to calibrate the activity rate in the lower bound of the uncertainty range; the normalized estimate uses 5% as a lifetime participation assumption to account for continued legal expansion and younger cohorts who will age into betting availability.\n"},{"url":"https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4938642","title":"Sports Betting Legalization Amplifies Emotional Cues & Intimate Partner Violence","publisher":"SSRN (Matsuzawa & Arnesen, University of Oregon)","source_type":"peer_reviewed","statistic":"After sports betting legalization, the effect of an NFL home-team upset loss on intimate partner violence rises by ~9–10 percentage points, with larger effects in states allowing mobile betting","excerpt":"\"When sports gambling is legalized, the effect of NFL home team upset losses on intimate partner violence increases by around 9 percentage points. These effects are larger (i) in states where mobile betting is legalized, (ii) in locations where higher bets were placed, (iii) around paydays, and (iv) for teams who were on a winning streak — supporting that financial losses from sports gambling amplify the emotional cues from a favorite team's unexpected loss.\"\n","source_date":"2024-10-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260708223432/https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4938642","calculation_notes":"This source does not change the financial headline number; it documents a non-financial spillover of the same post-2018 legalization shock and is cited in the caveats. NIBRS incident data, 2011–2022. SSRN blocks automated fetch (HTTP 403); the ~9-percentage-point upset-loss interaction figure and the mobile-betting / payday heterogeneity were verified via the University of Oregon / OPB reporting on the same paper (2024-10-30), the repo's established pattern for bot-protected primary URLs.\n","independence_note":"NIBRS crime-incident data is methodologically distinct from the Hollenbeck credit bureau panel and the Baker et al. household transaction data — a fourth independent measurement approach capturing a behavioral (rather than financial) outcome of the same legalization variation.\n"}],"comparison_anchors":[{"label":"Gambling disorder financial ruin (lifetime, US)","lifetime_us_adult":0.0063},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Day-trading financial ruin (lifetime, active traders)","lifetime_us_adult":0.08}],"personal_factor_multipliers":[{"factor":"male, age 18-34","multiplier":4,"notes":"Young men are the primary target demographic for sportsbook marketing and have the highest uptake rates; the NY Fed and Hollenbeck studies both find concentrated harm in this group"},{"factor":"daily mobile app wagering","multiplier":9,"notes":"Mobile betting collapses the friction between impulse and action; daily app users are the highest-harm subgroup in the Baker et al. transaction data"},{"factor":"prior gambling disorder history","multiplier":4.5,"notes":"History of gambling disorder is the strongest single predictor of sports-betting-related financial harm; prior disorder lowers the threshold for problem engagement"},{"factor":"casual bettor (once per week or less, no parlay products)","multiplier":0.15,"notes":"Infrequent, single-game straight bets carry low financial risk for most bettors; the harm concentrates in daily users and high-frequency parlay wagering"}],"short_label":"Sports betting financial ruin","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 1% lifetime estimate is built on post-2018 data from states that legalized online sports betting, then extrapolated to US adults broadly — the long-run lifetime harm rate depends on whether access continues to expand and whether problem-gambling infrastructure grows proportionally. The Hollenbeck et al. 25–30% relative bankruptcy increase is a population-level effect; converting it to an absolute harm rate among regular bettors requires assumptions about the baseline bankruptcy rate in the bettor subpopulation that carry meaningful uncertainty. This entry covers severe financial harm (bankruptcy or equivalent credit collapse) — a large additional population of regular bettors will experience moderate financial harm (increased debt, credit-score declines) that falls below this threshold. The companion entry gambling-addiction-financial-ruin covers the broader gambling-disorder population; this entry focuses specifically on the post-2018 mobile sports-betting wave and its direct financial consequences. \"Regular sports bettor\" is not a clinical term — it is used here to mean betting at least weekly over a sustained period, consistent with the active-bettor populations studied in the cited literature. The financial-ruin headline understates the total harm footprint of the same legalization shock: the harm is not only financial. Matsuzawa and Arnesen (2024, University of Oregon, NIBRS 2011–2022) find that after sports betting is legalized, the effect of an NFL home-team upset loss on intimate partner violence rises by roughly 9 percentage points, with the effect concentrated in states allowing mobile betting, around paydays, and where larger bets are placed — consistent with financial losses amplifying the emotional cue of a team's unexpected defeat. This entry quantifies only the direct financial outcome; downstream relational and behavioral harms are documented but not reflected in the 1% figure.\n","quality_score":{"d1":3,"d2":4,"d3":3,"d4":5,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A mobile phone displaying a sports betting app interface with a downward balance indicator, muted palette, flat vector illustration."},"canonical_url":"https://likelier.app/sports-betting-financial-ruin","api_url":"https://likelier.app/api/fears/sports-betting-financial-ruin.json"},{"slug":"wet-vs-dry-toilet-paper","question":"What are the odds of health problems from using only dry toilet paper?","quick_answer":"The incremental lifetime risk of a clinically significant problem from dry toilet paper versus bidet use is only about 1%. Hemorrhoids affect roughly 50% of adults by age 50 regardless of wiping method — far from the 25-50% harm rate bidet marketing implies is caused by dry wiping.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The bidet-vs-toilet-paper debate has become a fixture of online health discourse, with advocates for water cleansing framing dry wiping as a backward hygiene practice that causes UTIs, hemorrhoids, and perianal irritation. Bidet marketing amplifies the perception that dry toilet paper is actively harmful, and viral posts regularly claim that Americans are the only developed-world population still \"smearing\" rather than washing.\n","rough_estimate":"~25-50% chance of developing hemorrhoids or UTIs from using only dry toilet paper","kind":"intuition"},"native":{"display":"~50% of adults develop hemorrhoids by age 50, regardless of wiping method","numerator":50,"denominator":100,"unit":"cumulative prevalence of hemorrhoids by age 50","population":"US adults"},"normalized":{"lifetime_us_adult":0.01,"display":"~1% incremental lifetime risk of a clinically significant adverse outcome (hemorrhoids, UTI, perianal irritation) attributable to dry toilet paper vs bidet use","log_value":-2,"assumptions":"Hemorrhoids affect roughly 50% of Americans by age 50 (Johanson & Sonnenberg 1990, Gastroenterology), with prevalence peaking at ages 45-65. However, a 3-year prospective web survey (Asakura et al. 2018, Epidemiology & Infection) found that cumulative incidence of hemorrhoids was not significantly increased or decreased by habitual bidet use vs non-use. A systematic review (Baig et al. 2022, Evidence- Based Complementary and Alternative Medicine) concluded that current evidence does not identify bidet use as a treatment modality for perianal disease. For UTIs, lifetime incidence in women exceeds 50%, but a 2024 study (Kuriyama et al., Cureus) found that wiping direction was not significantly associated with UTI risk in the overall population after adjusting for age and diabetes -- significance appeared only in the 40-59 subgroup. The marginal health benefit of bidet over dry TP is not demonstrated in prospective trials. We conservatively estimate ~1% incremental lifetime risk attributable to dry TP vs water cleansing, reflecting comfort differences rather than infection causation.\n","uncertainty":{"low":0.002,"high":0.05},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/2295392/","title":"The prevalence of hemorrhoids and chronic constipation: an epidemiologic study","publisher":"Gastroenterology (Johanson & Sonnenberg, 1990)","source_type":"peer_reviewed","statistic":"4.4% point prevalence of hemorrhoids in the US general population; approximately 50% of adults develop hemorrhoids by age 50; peak prevalence ages 45-65","excerpt":"\"Results showed that 10 million people in the United States complained of hemorrhoids, corresponding to a prevalence rate of 4.4%. In both sexes, a peak in prevalence was noted from age 45-65 yr, with a subsequent decrease after age 65 yr.\"\n","source_date":"1990-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260504061733/https://pubmed.ncbi.nlm.nih.gov/2295392/","calculation_notes":"Johanson & Sonnenberg established the epidemiological baseline for hemorrhoid prevalence in the US. The ~50% by age 50 figure has been widely cited for three decades and reflects the near-universal nature of hemorrhoidal disease -- driven by straining, diet, age, and genetics rather than wiping material. This makes any marginal effect of TP vs bidet hard to detect against a 50% background rate.\n"},{"url":"https://www.cambridge.org/core/journals/epidemiology-and-infection/article/relationship-between-bidet-toilet-use-and-haemorrhoids-and-urogenital-infections-a-3year-followup-web-survey/B99EDC3FBD27791A762D7465422D6EB7","title":"Relationship between bidet toilet use and haemorrhoids and urogenital infections: a 3-year follow-up web survey","publisher":"Epidemiology & Infection (Asakura et al., 2018)","source_type":"peer_reviewed","statistic":"Cumulative incidence of hemorrhoids and urogenital infections was not significantly increased by habitual bidet toilet use; earlier positive correlations were likely reverse causation","excerpt":"\"Cumulative incidence of haemorrhoids and urogenital infections was not significantly increased by habitual use of a bidet toilet... Positive correlations between urogenital outcomes and habitual bidet toilet use reported earlier were not causal relationships, but rather might have been reverse causation.\"\n","source_date":"2018-03-26","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421201921/https://www.cambridge.org/core/journals/epidemiology-and-infection/article/relationship-between-bidet-toilet-use-and-haemorrhoids-and-urogenital-infections-a-3year-followup-web-survey/B99EDC3FBD27791A762D7465422D6EB7","calculation_notes":"This 3-year prospective study is the strongest longitudinal evidence on bidet vs non-bidet health outcomes. Finding no significant difference in hemorrhoid or urogenital infection incidence between habitual bidet users and non-users directly undermines the claim that dry toilet paper causes these conditions. The study also corrected earlier cross-sectional findings that had confused reverse causation (people with symptoms adopting bidets) with bidet-caused harm.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11088791/","title":"Post-Toilet Wiping Style Is Associated With the Risk of Urinary Tract Infection in Women","publisher":"Cureus (Kuriyama et al., 2024)","source_type":"peer_reviewed","statistic":"Wiping direction was not significantly associated with UTI risk in the overall population after adjusting for age and diabetes; significance appeared only in the 40-59 age subgroup","excerpt":"\"The impact of post-toilet wiping with the arm from the front between the legs on UTI events, adjusting for the age and history of diabetes mellitus, was not statistically significant both in males and females when analyzed overall.\"\n","source_date":"2024-05-10","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260504061725/https://pmc.ncbi.nlm.nih.gov/articles/PMC11088791/","calculation_notes":"This study examined wiping direction (front-to-back vs back-to-front) and UTI risk. The overall null finding after adjustment suggests that wiping technique -- whether with dry TP, wet wipes, or bidet -- is a weak predictor of UTI compared to established risk factors (sexual activity, menopause, catheterization, diabetes). The subgroup finding in ages 40-59 was a secondary analysis and awaits replication.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9173983/","title":"Be Kind to Your Behind: A Systematic Review of the Habitual Use of Bidets in Benign Perianal Disease","publisher":"Evidence-Based Complementary and Alternative Medicine (Baig et al., 2022)","source_type":"peer_reviewed","statistic":"Systematic review of six studies: habitual bidet use not significantly associated with hemorrhoid incidence at 1- or 3-year follow-up; one RCT found bidet use non-inferior to sitz bath for post-hemorrhoidectomy pain; habitual male bidet use associated with increased odds of pruritus ani at 3-year follow-up (OR 1.36); no trials assessed bidet use as a treatment for established perianal disease.","excerpt":"\"The current evidence does not identify using bidets as a treatment modality for perianal disease, and further research is warranted to study this increasingly utilized technology.\"\n","source_date":"2022-05-31","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260110162158/https://pmc.ncbi.nlm.nih.gov/articles/PMC9173983/","calculation_notes":"This is the named \"systematic review\" referenced in this entry's assumptions and body prose; this source block grounds that attribution. The review found habitual bidet use is not significantly associated with hemorrhoid incidence (consistent with Asakura et al. 2018 above), identified a small increased odds of pruritus ani in male users (OR 1.36) alongside one RCT showing bidet use non-inferior to a sitz bath for post-hemorrhoidectomy pain, and found no trial evidence for bidets as a treatment for existing perianal disease. It does not alter this entry's ~1% incremental-risk estimate, which concerns dry TP vs bidet comfort/irritation differences rather than a specific bidet-caused harm.\n","independence_note":"A systematic review synthesising six studies (RCTs, prospective cohorts, and a cross-sectional study), distinct from the individual studies cited above; it overlaps partially with the same underlying evidence base as Asakura et al. 2018 for the hemorrhoid-incidence finding.\n"}],"comparison_anchors":[{"label":"Developing hemorrhoids by age 50 (US)","lifetime_us_adult":0.5},{"label":"UTI in women (lifetime)","lifetime_us_adult":0.5},{"label":"Appendicitis (lifetime, US)","lifetime_us_adult":0.07}],"short_label":"Dry toilet paper harm","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"The 1% estimate is a rough bound on the marginal health difference between dry TP and bidet use, not a measure of absolute risk from either practice. Hemorrhoid and UTI prevalence is high regardless of cleansing method, and the dominant risk factors (diet, straining, sexual activity, menopause, genetics) overwhelm any plausible effect of wiping material. Bidets may offer comfort benefits for people with existing perianal conditions (fissures, post-surgical wounds), but that is symptom management, not disease prevention. Cultural and environmental dimensions (water usage, wet wipe disposal) are outside the health-risk scope of this entry.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A roll of toilet paper next to a water droplet, flat vector illustration in muted cream and blue tones."},"canonical_url":"https://likelier.app/wet-vs-dry-toilet-paper","api_url":"https://likelier.app/api/fears/wet-vs-dry-toilet-paper.json"},{"slug":"child-foreign-body-ingestion","question":"What are the odds a young child swallows a non-food object and needs medical care?","quick_answer":"Roughly 1 in 90 US children lands in an emergency department for swallowing a non-food object before age 6. Most swallowed objects pass on their own, so the harm is usually overestimated even as the sheer frequency of these ER visits is underestimated.\n","category":"kids","tags":["toddler","infant","household"],"no_reliable_estimate":false,"perceived":{"description":"Swallowing things is what toddlers do — everything goes in the mouth between about six months and three years, and parents know it. The fear it produces is lopsided in two directions at once. On the harm axis it is overestimated: a swallowed coin or bead feels like an acute emergency, the kind of thing that warrants a panicked drive to the ER, when in practice the great majority of swallowed objects pass on their own. On the frequency axis it is underestimated: almost no parent guesses that medically-attended ingestions are common enough to put roughly one child in ninety through an emergency department before kindergarten.\n","rough_estimate":"A swallowed object feels like a likely emergency; most parents would also under-guess how routinely it lands a child in the ER","kind":"intuition"},"native":{"display":"~43,000 US ED visits/year for swallowed objects in children under 6 (≈118/day, 2015)","numerator":1,"denominator":91,"unit":"per child, 0-5 (under-6) window","population":"US children under 6, ED-treated foreign-body ingestion (NEISS, all object types)"},"normalized":{"lifetime_us_adult":0.011,"display":"1 in ~90 per child during the 0-5 window (US, ED-treated ingestion)","log_value":-1.96,"assumptions":"Likelier normally reports lifetime-US-adult probabilities; this entry is scoped to the peak-risk early-childhood window (under 6) for a single US child, the same way the button-battery and toddler-choking entries are. Orsagh-Yentis et al. (Pediatrics 2019), analysing the CPSC's National Electronic Injury Surveillance System, estimated about 759,000 US children under six treated in emergency departments for swallowed objects across 1995-2015, rising to roughly 43,000 in 2015 — a rate of about 18 per 10,000 under-six children per year. Compounded across the six-year window, 1 - (1 - 1.8e-3)^6 ≈ 0.011, or roughly 1 in 90 per child. This counts medically-attended ingestions only: the true number of objects that are swallowed and pass unnoticed at home is necessarily higher and unmeasurable, while fatal or surgical outcomes are far rarer (see regional_breakdown). The figure spans all object types; the dangerous-object subsets (button batteries, multiple magnets, sharp objects) carry their own entries and their own much lower but higher-severity rates.\n","uncertainty":{"low":0.008,"high":0.016},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.nationwidechildrens.org/newsroom/news-releases/2019/04/foreign-body-ingestions-study","title":"Foreign Body Ingestions on the Rise (Orsagh-Yentis et al., Pediatrics 2019;143(5):e20181988)","publisher":"Nationwide Children's Hospital — reporting Orsagh-Yentis, McAdams, Roberts & McKenzie, Pediatrics","source_type":"reputable_reference","statistic":"~759,000 US children under 6 treated in EDs for foreign-body ingestion, 1995-2015; rate up 91.5% from 9.5 to 18 per 10,000 children; ~43,000 (118/day) in 2015; coins 62%, toys 10%, jewelry 7%, batteries 7%; 62% aged 1-3; just over 10% admitted","excerpt":"\"more than 759,000 children younger than six years were estimated to have been evaluated for foreign body ingestions in emergency departments over the 21-year study period [...] The rate of foreign body ingestions per 10,000 children increased by 91.5%, from 9.5 in 1995 to 18 in 2015. [...] Coins were the most frequent type of objects ingested (62%), followed by toys (10%), jewelry (7%), and batteries (7%). [...] Foreign body ingestions most frequently involved children between one to three years of age (62%). [...] Just over 10% of children were admitted to the hospital as a result of their ingestion.\"\n","source_date":"2019-04-12","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20251205114623/https://www.nationwidechildrens.org/newsroom/news-releases/2019/04/foreign-body-ingestions-study","calculation_notes":"Primary anchor for the incidence headline, age distribution, object mix, and admission rate. 43,000 ED visits in 2015 against a US under-6 population of about 24 million is ~18 per 10,000 per year, matching the reported 2015 rate. Compounded over the six-year window gives ~1 in 90 per child. Coins at 62% are the dominant object, far ahead of batteries at 7% — the small subset that drives nearly all severe outcomes. The 10% admission rate feeds the regional_breakdown funnel.\n","independence_note":"Orsagh-Yentis is NEISS population surveillance (CPSC product-injury sampling), a genuinely separate pipeline from the clinical-review figures (AAFP, Lee) on spontaneous passage and from the clinical cohort (Destro) on neurodevelopmental risk. This press release reports the peer-reviewed Pediatrics paper; the journal page itself (publications.aap.org) is paywalled/403 to automated fetching.\n"},{"url":"https://www.aafp.org/pubs/afp/issues/2005/0715/p287.html","title":"Foreign Body Ingestion in Children","publisher":"American Family Physician (AAFP)","source_type":"peer_reviewed","statistic":"93-99% of blunt objects pass without intervention; ~1% require surgical removal; 10-20% managed with endoscopy; serious morbidity in <1%; 50% of confirmed ingestions asymptomatic; peaks ages 6 months-3 years; ~1,500 all-ages US foreign-body deaths/year","excerpt":"\"In the pre-endoscopy era, 93 to 99 percent of blunt objects passed without intervention, and approximately one percent required surgical removal. [...] Today, 10 to 20 percent of children who ingest foreign bodies are managed with endoscopy. [...] It causes serious morbidity in less than one percent of all patients, and approximately 1,500 deaths per year are attributed to ingestion of foreign bodies in the United States. [...] A retrospective review found that 50 percent of children with confirmed foreign body ingestions were asymptomatic. [...] Foreign body ingestion is a potentially serious problem that peaks in children aged six months to three years.\"\n","source_date":"2005-07-15","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260308035245/https://www.aafp.org/pubs/afp/issues/2005/0715/p287.html","calculation_notes":"Source for the outcome funnel: most blunt objects pass without intervention, ~10-20% need endoscopy, ~1% need surgery, serious morbidity <1%, and half of confirmed ingestions are asymptomatic. The \"~1,500 deaths/year\" figure is deliberately NOT used in the headline: it is all-ages and dominated by airway obstruction (choking), not toddlers passing swallowed objects through the gut — see caveats.\n","independence_note":"The 80-90% spontaneous-passage / 10-20% endoscopy / ~1% surgery split is a single canonical clinical figure restated across the review literature. This AAFP review and Lee 2018 (below) are two restatements of the same figure, not independent measurements; treat them as one source for the funnel. The independent pipelines in this entry are Orsagh-Yentis (NEISS surveillance) and Destro (clinical cohort).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5903088/","title":"Foreign Body Ingestion in Children","publisher":"Clinical Endoscopy 2018;51(2):129-136 — Lee JH","source_type":"peer_reviewed","statistic":"80-90% of GI-tract foreign bodies pass spontaneously without complications; 10-20% removed endoscopically; ~1% require open surgery; most events at 6 months-3 years; coins most common; sharp/pointed objects perforate in 15-35%","excerpt":"\"80%-90% of FBs in the gastrointestinal (GI) tract are passed spontaneously without complications, 10%-20% are removed endoscopically, and 1% require open surgery. [...] most events occur in children between 6 months and 3 years of age. [...] Coins are the most commonly ingested FB in children. [...] sharp or pointed FBs can cause perforation in 15%-35% of patients.\"\n","source_date":"2018-03-30","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260420040721/https://pmc.ncbi.nlm.nih.gov/articles/PMC5903088/","calculation_notes":"Confirms the spontaneous-passage / endoscopy / surgery split and supplies the sharp-object perforation rate (15-35%) used in the caveats and the dangerous-tail framing. Also corroborates the 6-months-to-3-years peak age and coins as the most common object.\n","independence_note":"Restates the same canonical spontaneous-passage / endoscopy / surgery split as the AAFP review above; the two are one figure restated, not independent confirmation. The sharp-object perforation rate is its distinct contribution.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8624601/","title":"Foreign Body Ingestion in Neurologically Impaired Children: A Challenging Diagnosis and Management in Pediatric Surgery","publisher":"Children (Basel) 2021;8(11):956 — Destro et al.","source_type":"peer_reviewed","statistic":"Children with intellectual disability/neurodevelopmental delay (ID-ND) commonly ingest foreign bodies; higher risk tied to poor hand-to-mouth control, prolonged oral phase, dysphagia; long delays to presentation drive higher surgery, perforation, and mortality","excerpt":"\"Children with intellectual disability/neurodevelopmental delay (ID-ND) commonly ingest foreign bodies (FB) [...] the higher risk of FB ingestion in ID-ND is related to several reasons, including poor control of hand-to-mouth activity and exploration of objects, prolonged oral phase, dysphagia with limited control over objects placed in the oral cavity, communication impairment, and altered protective mechanisms. [...] Long delays between ingestion to presentation and intervention might account for higher rates of surgery, perforation, and mortality in the group of patients with ID-ND.\"\n","source_date":"2021-10-23","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20250413113336/https://pmc.ncbi.nlm.nih.gov/articles/PMC8624601/","calculation_notes":"Basis for the developmental-delay/autism/pica personal factor. The paper documents that ID-ND children both ingest objects more (prolonged oral-exploration phase) and fare worse (delayed presentation → more surgery, perforation, death). The multiplier is therefore explicitly mixed — incidence, recurrence, and severity — not a clean incidence ratio.\n","independence_note":"A clinical cohort/case series, methodologically distinct from the NEISS population surveillance (Orsagh-Yentis) and from the general clinical-review funnel figures (AAFP, Lee). It addresses a different question (who is at elevated risk and why), so it complements rather than overlaps the other three.\n"},{"url":"https://www.poison.org/battery/guideline","title":"National Capital Poison Center's Button Battery Ingestion Triage and Treatment Guideline","publisher":"National Capital Poison Center","source_type":"reputable_reference","statistic":"Batteries lodged in the esophagus may cause serious burns in as little as 2 hours; immediate removal is required","excerpt":"\"Batteries lodged in the esophagus may cause serious burns in as little as 2 hours... Immediately remove batteries lodged in the esophagus. Serious burns can occur in 2 hours.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260619141403/https://www.poison.org/battery/guideline","calculation_notes":"Supports the body-prose claim that a button or coin-cell battery lodged in the esophagus can cause severe burns within about two hours — the reason battery ingestion is treated as a same-day emergency rather than folded into the \"it will pass\" reassurance that applies to blunt objects in the stomach.\n","independence_note":"National Capital Poison Center clinical guideline, methodologically distinct from the NEISS surveillance (Orsagh-Yentis), the AAFP/Lee clinical-review funnel figures, and the Destro ID-ND cohort above.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8371313/","title":"Magnet Ingestion in Children: Management Guidelines and Prevention","publisher":"Frontiers in Pediatrics 2021;9:727988 — Altokhais T","source_type":"peer_reviewed","statistic":"Two or more ingested high-powered magnets can attract across the gastrointestinal wall, causing ischemia, tissue necrosis, perforation, fistula formation, obstruction, peritonitis, or death; NEISS estimated 16,386 possible magnet ingestions among US children under 18 from 2002-2011","excerpt":"\"Magnet pieces attract each other across the walls of the gastrointestinal tract, causing ischemia, tissue necrosis, perforation, fistula formation, obstruction, peritonitis, or death.\"\n","source_date":"2021-08-04","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260420040721/https://pmc.ncbi.nlm.nih.gov/articles/PMC8371313/","calculation_notes":"Supports the body-prose claim that two or more high-powered magnets, swallowed separately, pinch bowel tissue between them and cause perforation — the mechanism that makes multiple-magnet ingestion a distinct surgical emergency rather than an \"it will pass\" case, alongside button batteries and sharp objects.\n","independence_note":"Peer-reviewed clinical review, methodologically distinct from the NEISS surveillance (Orsagh-Yentis), the AAFP/Lee clinical-review funnel figures, the Destro ID-ND cohort, and the National Capital Poison Center battery guideline above.\n"}],"comparison_anchors":[{"label":"Toddler food choking death, US child 0-4","lifetime_us_adult":0.00002},{"label":"Button battery severe injury, US child 0-6","lifetime_us_adult":0.000004}],"regional_breakdown":[{"region":"ED-treated ingestion, US child 0-5 (the headline)","probability":0.011,"notes":"~18 ED visits per 10,000 under-six children per year (Orsagh-Yentis, 2015), compounded across the six-year window. This is the medically-attended rate, not the chance of ever swallowing anything — most mouthed objects pass unnoticed and are never counted.\n"},{"region":"Hospital admission, per ED visit for ingestion","probability":0.1,"notes":"Just over 10% of children evaluated for a swallowed object were admitted (Orsagh-Yentis). Most of the rest were observed, X-rayed, and discharged.\n"},{"region":"Endoscopic removal needed, per ingestion","probability":0.15,"notes":"10-20% of ingested foreign bodies are retrieved endoscopically (AAFP, Lee), typically because the object lodged in the esophagus or is high-risk (battery, sharp). The midpoint ~15% is used here.\n"},{"region":"Surgery needed, per ingestion","probability":0.01,"notes":"About 1% of ingested foreign bodies require open surgery (AAFP, Lee) — perforation, obstruction, or magnet/battery injury. The overwhelming majority never reach this stage.\n"}],"personal_factor_multipliers":[{"factor":"age under 6 months (pre-crawling, no pincer grasp)","multiplier":0.2,"notes":"Below about six months the motor skills that drive ingestion — the pincer grasp to pick up a coin and the mobility to reach one — have not arrived, so the rate is well under the under-six average. Events at this age usually involve an object handed over by an older sibling.\n"},{"factor":"age 1-2 years (peak)","multiplier":2,"notes":"Orsagh-Yentis found 62% of ingestions in children aged 1-3, with the sharpest concentration at 1-2: the point where pincer grasp, mobility, and relentless oral exploration coincide with no judgment about what is food. Roughly twice the under-six average rate.\n"},{"factor":"age 4-5 years (declining)","multiplier":0.6,"notes":"By the preschool years most children have stopped reflexively mouthing objects, so the rate falls below the under-six average even though access to small objects rises.\n"},{"factor":"developmental delay, autism, or pica","multiplier":3,"notes":"Children with intellectual disability, autism, or pica have a prolonged oral-exploration phase (Destro et al.). This factor is deliberately mixed: it raises first-event incidence, accounts for most repeat and multiple-object ingestions, and worsens outcomes because these children tend to present late, turning an object that would otherwise pass into a surgical or perforation case. The 3x is approximate and directional, not a measured incidence ratio.\n"}],"short_label":"Child swallows object (ER)","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1-in-90 headline is the medically-attended rate — children taken to an emergency department for a swallowed object — not the chance of ever swallowing anything. Most objects a toddler mouths either never go down or pass unnoticed, and those events are unmeasurable. The often-quoted figure of ~1,500 foreign-body deaths per year in the United States (AAFP) is all-ages and dominated by airway obstruction (choking and asphyxiation), not toddlers passing swallowed objects through the gut; pediatric deaths from ingested (as opposed to inhaled) objects are rare and concentrated in the button-battery, multiple-magnet, and sharp-object subsets, each of which has its own entry. The 80-90% / 10-20% / ~1% spontaneous-passage / endoscopy / surgery split is a single canonical clinical figure restated across the review literature, not three independent measurements. The \"it will pass\" reassurance applies to blunt objects that have already reached the stomach, such as a swallowed coin, bead, or smooth pebble; sharp or pointed objects perforate in 15-35% of cases, and any object lodged in the esophagus rather than the stomach needs prompt removal regardless of what it is.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"Small household objects a toddler might swallow — a coin, a bead, a small toy brick, a button battery — laid on a plain surface beside a child's open hand."},"canonical_url":"https://likelier.app/child-foreign-body-ingestion","api_url":"https://likelier.app/api/fears/child-foreign-body-ingestion.json"},{"slug":"ovarian-cancer","question":"What are the odds of developing ovarian cancer?","quick_answer":"About 1 in 91 US women develop ovarian cancer over their lifetime, and about 1 in 143 die of it. Many people assume it is nearly as common as breast cancer (~13% lifetime risk), but it is far rarer.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Ovarian cancer occupies a disproportionate space in the public fear landscape relative to its actual incidence — partly because it is often diagnosed late, partly because the early-detection story is far less reassuring than for breast or cervical cancer. There is no reliable population-level screening test, and the \"silent killer\" framing in media coverage reinforces the sense that the disease strikes without warning. Most women cannot distinguish between the lifetime risk of ovarian cancer (~1%) and the lifetime risk of breast cancer (~13%), and the two are frequently conflated in casual conversation about \"women's cancers.\"\n","rough_estimate":"Many adults assume ovarian cancer is nearly as common as breast cancer","kind":"intuition"},"native":{"display":"~1 in 91 US women lifetime risk of diagnosis","numerator":11,"denominator":1000,"unit":"lifetime","population":"US women, all ages"},"normalized":{"lifetime_us_adult":0.011,"display":"1 in ~91 lifetime (US women)","log_value":-1.96,"assumptions":"The American Cancer Society estimates a US woman's lifetime risk of developing ovarian cancer at approximately 1 in 91 (~1.1%), based on SEER incidence data. SEER confirms approximately 1.1% of women will be diagnosed with ovarian cancer at some point during their lifetime, based on 2021-2023 data. The ACS projects ~21,010 new cases and ~12,450 deaths in 2026. Lifetime risk of dying from ovarian cancer is lower, about 1 in 143 (~0.7%), because roughly 50% of cases are now survived. Point estimate 0.011 for incidence; uncertainty band reflects the range between the death-only figure (~0.007) and older, higher historical incidence estimates (~0.013) before the secular decline driven by oral contraceptive use and reduced hormone therapy.\n","uncertainty":{"low":0.008,"high":0.013},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cancer.org/cancer/types/ovarian-cancer/key-statistics.html","title":"Key Statistics for Ovarian Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Lifetime risk ~1 in 91; ~21,010 new cases and ~12,450 deaths projected for 2026","excerpt":"\"A woman's risk of getting ovarian cancer during her lifetime is about 1 in 91. Her lifetime chance of dying from ovarian cancer is about 1 in 143. [...] The American Cancer Society estimates that in 2026, about 21,010 new cases of ovarian cancer will be diagnosed and 12,450 women will die of ovarian cancer in the United States.\"\n","source_date":"2026-01-13","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426204640/https://www.cancer.org/cancer/types/ovarian-cancer/key-statistics.html","calculation_notes":"ACS gives lifetime incidence directly as ~1 in 91 (~1.1%) and lifetime mortality as ~1 in 143 (~0.7%). The headline figure used here is incidence, not mortality, because the question asks about developing ovarian cancer, not dying from it. The ~60% case-fatality implied by the ratio (0.7/1.1) is considerably higher than for breast cancer (~18%), reflecting later-stage diagnosis on average and the absence of an effective screening test.\n","independence_note":"ACS derives its lifetime-probability figures from SEER incidence and mortality data (NCI). Treat as partially dependent with the SEER source below.\n"},{"url":"https://seer.cancer.gov/statfacts/html/ovary.html","title":"Cancer Stat Facts: Ovarian Cancer","publisher":"National Cancer Institute / SEER Program","source_type":"govt_report","statistic":"Approximately 1.1% of women will be diagnosed with ovarian cancer at some point during their lifetime","excerpt":"\"Approximately 1.1 percent of women will be diagnosed with ovarian cancer at some point during their lifetime, based on 2021–2023 data.\"\n","source_date":"2025-04-17","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426204715/https://seer.cancer.gov/statfacts/html/ovary.html","calculation_notes":"SEER provides the upstream population-registry data from which ACS derives its headline figures. The 1.1% lifetime risk is consistent with the ACS \"1 in 91\" figure. Five-year relative survival for ovarian cancer is approximately 51%, far lower than for breast cancer (91.7%), which is why the diagnosis-to-death gap is narrower here.\n","independence_note":"SEER is the upstream data source that ACS cites; treat these two as partially dependent. SEER is included as the authoritative primary-pipeline citation.\n"},{"url":"https://www.cancer.gov/about-cancer/causes-prevention/genetics/brca-fact-sheet","title":"BRCA Gene Changes: Cancer Risk and Genetic Testing Fact Sheet","publisher":"National Cancer Institute","source_type":"govt_report","statistic":"Harmful BRCA1/BRCA2 gene changes occur in about 0.2%-0.3% of the general population, roughly 1 in 400","excerpt":"\"The prevalence of harmful BRCA gene changes in the general population is about 0.2%-0.3% (or about 1 in 400).\"\n","source_date":"2024-11-19","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260711222011/https://www.cancer.gov/about-cancer/causes-prevention/genetics/brca-fact-sheet","calculation_notes":"NCI's general-population BRCA1/2 prevalence figure (~1 in 400) grounds the body text's \"roughly 1 in 400-500\" context for how small a share of women carry the pathogenic variant driving the 35x/15x personal-factor multipliers above. Not used in the headline lifetime-incidence calculation, which comes from ACS/SEER.\n","independence_note":"NCI germline-genetics fact sheet, independent of the ACS/SEER incidence-and-mortality pipeline used for the headline probability.\n"}],"comparison_anchors":[{"label":"Breast cancer diagnosis (lifetime, US women)","lifetime_us_adult":0.13},{"label":"Cervical cancer diagnosis (lifetime, US women)","lifetime_us_adult":0.006},{"label":"Lung cancer diagnosis (lifetime, US)","lifetime_us_adult":0.063},{"label":"Any cancer death (lifetime, US)","lifetime_us_adult":0.14}],"regional_breakdown":[{"region":"US women (all races)","probability":0.011,"notes":"ACS/SEER: ~1 in 91 lifetime incidence"},{"region":"US White women","probability":0.012,"notes":"Slightly higher age-adjusted incidence than Black women"},{"region":"US Black women","probability":0.009,"notes":"Lower incidence but higher case-fatality and later-stage diagnosis on average"}],"personal_factor_multipliers":[{"factor":"BRCA1 pathogenic variant","multiplier":35,"notes":"Lifetime ovarian cancer risk ~39-44% vs ~1.1% baseline; the sharpest single risk elevation"},{"factor":"BRCA2 pathogenic variant","multiplier":15,"notes":"Lifetime ovarian cancer risk ~11-17% vs ~1.1% baseline"},{"factor":"First-degree family history of ovarian cancer","multiplier":3,"notes":"Aggregate figure across population studies"},{"factor":"Long-term oral contraceptive use (5+ years)","multiplier":0.5,"notes":"Oral contraceptive use reduces ovarian cancer risk by roughly 50%; protective effect persists for decades after cessation"}],"short_label":"Ovarian cancer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is lifetime *incidence* (diagnosis), not mortality. Ovarian cancer's case-fatality rate is high compared to other gynecological cancers — roughly 60% of women diagnosed will eventually die of the disease — because no effective population screening test exists and most cases are diagnosed at advanced stage. The \"silent killer\" label is partly earned: early-stage symptoms are nonspecific and easily attributed to other causes. However, the baseline lifetime risk of developing ovarian cancer (~1 in 91) is roughly 12x lower than the equivalent figure for breast cancer (~1 in 8), a gap many people do not appreciate.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"Two concentric pale ovals on a muted lavender-grey background, flat vector illustration suggesting quiet symmetry."},"canonical_url":"https://likelier.app/ovarian-cancer","api_url":"https://likelier.app/api/fears/ovarian-cancer.json"},{"slug":"secondhand-smoke-death","question":"What are the odds of dying from secondhand smoke as a non-smoker?","quick_answer":"About 1 in 91 US adult non-smokers die from secondhand smoke over a lifetime at average current exposure levels, some 41,000 deaths a year. Most people underestimate this by roughly tenfold and are largely unaware it also causes permanent lung impairment in children.\n","category":"health","tags":["household","substance-use","kids"],"no_reliable_estimate":false,"perceived":{"description":"Most adults in smoke-free countries correctly file secondhand smoke (SHS) as unhealthy, yet the quantitative mortality burden rarely registers. The mental model is vague — \"it's bad for you\" rather than \"it kills tens of thousands of non-smokers in the US every year.\" The lifetime risk for a non-smoker at average current exposure levels sits in a range most people would guess is a tenth of what it actually is. The children's dimension is the most underappreciated part: childhood SHS exposure causes permanent reductions in lung development and elevated asthma risk that persist decades after the exposure ends, yet the causal chain is invisible compared to the acute symptoms that do prompt parental action (ear infections, wheezing). The visibility problem is compounded by the way tobacco harm is communicated — almost all of the public-health messaging targets the smoker, leaving non-smokers with a strong but poorly calibrated sense that the danger falls mainly on the person who lights up.\n","rough_estimate":"Most adults strongly underestimate how many non-smokers die from SHS annually, and are largely unaware of the permanent lung impairment it causes in children","kind":"intuition"},"native":{"display":"~41,000 non-smoking US adults die from secondhand smoke annually","numerator":41000,"denominator":225000000,"unit":"per year, non-smoking US adults","population":"US non-smoking adults, current average SHS exposure levels"},"normalized":{"lifetime_us_adult":0.011,"display":"~1 in 91 lifetime (US adult non-smoker, average current SHS exposure)","log_value":-1.96,"assumptions":"CDC attributes >41,000 non-smoker deaths per year to secondhand smoke in the US: ~7,300 from lung cancer and ~33,950 from ischaemic heart disease, plus ~400 infant deaths. The US non-smoking adult population is approximately 225 million (roughly 87% of ~260 million US adults, given a current-smoker prevalence of ~12.5%). Annual attributable mortality rate: 41,000 / 225,000,000 ≈ 1.82 × 10⁻⁴ per adult per year. Compounded over 60 years of adult life: 1 − (1 − 1.82e-4)^60 ≈ 0.011, or roughly 1 in 91. This figure reflects average population-level SHS exposure under current indoor smoking ban regimes and is not the risk for a heavily exposed subgroup such as someone who lived with a smoking spouse for decades and worked in a smoking-permitted venue. Exposure has fallen substantially since the 1980s following widespread adoption of smoke-free laws; the 1-in-91 figure reflects post-ban conditions. Uncertainty band 0.006–0.017 captures uncertainty in the attributable-fraction methodology (SAMMEC model using population-attributable fractions rather than direct observation) and in the distribution of actual SHS exposure intensities across the non-smoking population.\n","uncertainty":{"low":0.006,"high":0.017},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/tobacco/about/index.html","title":"About Smoking and Tobacco Use","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"SHS contributes to >41,000 non-smoker adult deaths and ~400 infant deaths per year in the US; >480,000 total smoking-related deaths including SHS","excerpt":"\"Smoking and secondhand smoke exposure cause more than 480,000 deaths each year in the United States. This is nearly one in five deaths. [...] Secondhand smoke exposure contributes to over 40,000 deaths among nonsmoking adults and 400 deaths in infants each year.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260412173440/https://www.cdc.gov/tobacco/about/index.html","calculation_notes":"41,000 non-smoker adult deaths per year is the primary domestic headline figure and the numerator for the normalized calculation. US non-smoking adult population denominator: ~225 million (87% of ~260M US adults). Annual attributable rate: 41,000 / 225M = 1.82e-4. Compounded over 60 adult years: 1 − (1 − 1.82e-4)^60 ≈ 0.011. The CDC does not directly report the lung cancer / heart disease breakdown on this page; the ~7,300 lung cancer and ~33,950 heart disease sub-totals appear in the CDC's dedicated SHS resource pages and the 2006 Surgeon General Report on involuntary tobacco smoke exposure.\n","independence_note":"CDC SAMMEC (Smoking-Attributable Mortality, Morbidity, and Economic Costs) model draws on Cancer Prevention Study II hazard ratios and NHANES SHS exposure data — methodologically overlapping with the 2006 Surgeon General Report, which uses the same underlying cohort hazard ratios. Treat as institutional confirmation of the same underlying estimate rather than a fully independent line of evidence.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/tobacco","title":"Tobacco — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"SHS kills ~1.6 million non-smokers globally per year; no safe level of SHS exposure exists","excerpt":"\"Tobacco kills more than 7 million people each year, including an estimated 1.6 million non-smokers who are exposed to second-hand smoke.\"\n","source_date":"2025-07-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260426210529/https://www.who.int/news-room/fact-sheets/detail/tobacco","calculation_notes":"The WHO 1.6 million non-smoker SHS deaths globally provides the international anchor and cross-check. Scaling naively to the US by population share (~4.2M of 57M global deaths) would imply ~118,000 US SHS deaths, higher than the CDC's 41,000. The gap primarily reflects: (a) the WHO/IHME figure includes lower-income regions where indoor solid-fuel cooking fire co-exposure substantially inflates the SHS burden; (b) CDC's SAMMEC model uses a more conservative attributable-fraction method. The US CDC figure is used for the normalized headline as more relevant to a US adult in a post-ban environment.\n","independence_note":"WHO draws on IHME Global Burden of Disease estimates using a different exposure-prevalence and relative-risk framework than CDC SAMMEC. The two-to-three-fold difference in implied US figures is a genuine methodological disagreement, not a simple error; it is reflected in the wide uncertainty band.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21112082/","title":"Worldwide burden of disease from exposure to second-hand smoke: a retrospective analysis of data from 192 countries","publisher":"The Lancet (Öberg M, Jaakkola MS, Woodward A, Peruga A, Prüss-Ustün A)","source_type":"peer_reviewed","statistic":"~600,000 global deaths/year from SHS (2004 data); 165,000 children under 5 from lower respiratory infections; SHS accounts for ~1% of worldwide mortality","excerpt":"[Paraphrase from abstract — full text paywalled] In 2004, SHS caused approximately 379,000 deaths from ischaemic heart disease, 165,000 from lower respiratory infections (mainly in children under 5), 36,900 from asthma, and 21,400 from lung cancer. SHS accounted for about 1% of worldwide mortality. The greatest child burden was 165,000 deaths from lower respiratory infections, concentrated in low-income countries.\n","source_date":"2011-01-08","source_accessed":"2026-05-18","archive_url":"https://web.archive.org/web/20260101000000*/https://pubmed.ncbi.nlm.nih.gov/21112082/","calculation_notes":"Öberg et al. 2011 is the most comprehensive independent global burden estimate and provides the strongest peer-reviewed anchor for both the adult and child mortality dimensions. The 165,000 child deaths from lower respiratory infections (concentrated in low- and middle-income countries with high indoor SHS from cooking and heating) is the key quantitative evidence base for the children's mortality dimension, distinct from the US-centric CDC estimate. The global ischaemic heart disease figure (379,000 / ~600,000 total ≈ 63% cardiovascular) is broadly consistent with the CDC's US breakdown (~83% heart disease), validating the mechanism framing. The 21,400 lung cancer figure as a fraction of total deaths (~3.6%) is lower than the US proportion (~18%), reflecting differential tobacco exposure histories and cancer screening rates across the 192 countries.\n","independence_note":"Öberg et al. use WHO global SHS exposure prevalence data combined with meta-analytic relative risks from the epidemiological literature; the analytic framework is independent of the CDC SAMMEC model and the WHO fact-sheet figure (which draws on the same IHME GBD pipeline as Öberg). The child lower-respiratory-infection mortality component is the most methodologically distinct from the US adult estimates and is not derived from the CPS-II cohort.\n"}],"comparison_anchors":[{"label":"Death from active smoking (lifetime, lifelong regular smoker)","lifetime_us_adult":0.5},{"label":"Death from lung cancer (lifetime, US adult)","lifetime_us_adult":0.018},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from home fire (lifetime, US)","lifetime_us_adult":0.00085}],"regional_breakdown":[{"region":"US adult non-smoker, average current SHS exposure","probability":0.011,"notes":"Headline. 41,000 annual attributable deaths / ~225M non-smoking adults, compounded 60 years."},{"region":"Non-smoker with current-smoking household member (indoor)","probability":0.03,"notes":"Home is the dominant SHS exposure route; a smoking spouse or housemate multiplies cotinine levels 2-3x vs population average, consistent with the ~20-30% lung cancer relative risk elevation for non-smoking spouses documented in Surgeon General 2006 and the Hackshaw meta-analysis"},{"region":"Non-smoker in fully smoke-free environment (home, work, social)","probability":0.002,"notes":"Residual risk from incidental outdoor and transitional exposure; reflects the post-ban floor, not zero"},{"region":"Adult raised in household with ≥1 indoor smoker throughout childhood","probability":0.016,"notes":"Childhood SHS adds a permanent lung-function deficit that raises adult susceptibility; this row reflects the upward shift for someone with a childhood SHS history, in addition to any current adult exposure"}],"personal_factor_multipliers":[{"factor":"household member smokes indoors","multiplier":2.5,"notes":"Home is the primary adult SHS exposure route; living with an indoor smoker raises personal cotinine by 2-3x vs the population average and is consistent with the 20-30% excess lung cancer relative risk for non-smoking spouses reported in the Hackshaw meta-analysis (BMJ 1997, N=37 epidemiological studies)"},{"factor":"raised with smoking parent(s) in the home throughout childhood","multiplier":1.5,"notes":"Childhood SHS causes persistent reductions in lung function: Gilliland et al. (Am J Respir Crit Care Med 2001), studying school-aged children in the Children's Health Study, found both in-utero maternal smoking and childhood ETS exposure associated with persistent lung-function deficits; asthma prevalence is 20-35% higher in children raised in smoking households even after adjusting for adult smoking status; these deficits raise baseline susceptibility to SHS-attributable disease across the adult lifespan"},{"factor":"no regular indoor SHS exposure (smoke-free home, work, social)","multiplier":0.2,"notes":"Near-elimination of indoor SHS via smoke-free policies is responsible for most of the overall SHS mortality decline since the 1980s; residual outdoor and transient exposure remains but is a fraction of pre-ban household levels"},{"factor":"works in hospitality, bar, or casino industry (no comprehensive indoor ban)","multiplier":2,"notes":"Occupational SHS in smoking-permitted hospitality settings reaches exposure intensities comparable to or exceeding living with a smoker; bar and casino workers in jurisdictions without comprehensive workplace bans face the highest documented non-household SHS exposures"},{"factor":"exposed to SHS in utero (mother smoked during pregnancy)","multiplier":1.3,"notes":"Prenatal SHS exposure is associated with permanent reductions in fetal lung function at birth, 2-4x elevated SIDS risk in the infant period, and elevated lifetime asthma and respiratory disease susceptibility; the adult mortality multiplier reflects the long-term physiological deficit from impaired fetal lung development, independent of postnatal SHS exposure"}],"short_label":"Secondhand smoke","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry covers mortality risk to non-smokers from passive SHS exposure, not the broader morbidity burden (non-fatal respiratory disease, childhood ear infections, asthma exacerbations, cognitive development effects). The headline figure (~1 in 91 lifetime) is a population average across current US exposure levels and obscures wide variation: someone who has lived with an indoor smoker for decades faces meaningfully higher risk than someone in a fully smoke-free environment. The 41,000 annual US SHS deaths figure is derived from the SAMMEC attributable-fraction model and is not directly observed from death certificates, which do not record SHS exposure status; the methodology carries substantial uncertainty and plausible estimates range from ~25,000 to ~60,000 depending on methodological choices. The children's permanent lung-impairment dimension is not captured in the point estimate — it raises adult susceptibility multiplicatively rather than adding a discrete mortality increment. SHS exposure has fallen substantially in the US since the 1980s following indoor smoking bans; the current 41,000 figure reflects post-ban conditions and would have been higher in earlier decades. The entry does not separately address children's mortality from SHS-attributable SIDS (covered in the SIDS entry) or acute lower respiratory infections, which are more concentrated in low-income countries per Öberg et al.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-18","last_reviewed":"2026-05-18","reviewed":true,"generated_at":"2026-05-18","image":{"alt":"Two faint wisps of pale grey smoke drifting across a minimalist domestic interior, flat vector illustration."},"canonical_url":"https://likelier.app/secondhand-smoke-death","api_url":"https://likelier.app/api/fears/secondhand-smoke-death.json"},{"slug":"gaming-disorder-adult","question":"What are the odds of developing gaming disorder as an adult?","quick_answer":"About 1 in 83 US adults will develop gaming disorder at some point, and roughly 3% of adult gamers meet the ICD-11 criteria. People tend to guess higher for heavy gamers (5-10%) and to dismiss it as a teenagers-only problem.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Gaming disorder is commonly perceived as a condition that affects teenagers and young adults who spend excessive time playing video games — not a clinical disorder with diagnostic criteria recognized by a major international classification system. Adults who game heavily often dismiss the possibility of disorder because they maintain jobs and relationships, and because the cultural framing of adult gaming as a leisure activity insulates it from the scrutiny applied to adolescent gaming. The WHO's inclusion of gaming disorder in ICD-11 (effective 2022) remains contested in some academic quarters, creating a public impression that the diagnosis is either new, uncertain, or not a \"real\" disorder in the way substance-use disorders are — despite the ICD-11 criteria requiring functional impairment rather than merely heavy play.\n","rough_estimate":"~5-10% of heavy gamers","kind":"intuition"},"native":{"display":"~3% of adult gamers meet ICD-11 Gaming Disorder criteria (global meta-analysis, Stevens et al. 2021, corrected 2023)","numerator":3,"denominator":100,"unit":"share of adult gamers meeting ICD-11 Gaming Disorder criteria (current prevalence, meta-analysis)","population":"adult gamers globally (meta-analysis of studies using ICD-11 or equivalent criteria, Stevens et al. 2021)"},"normalized":{"lifetime_us_adult":0.012,"display":"~1 in 83 US adults will develop gaming disorder at some point in their life","log_value":-1.921,"assumptions":"Two-factor estimate: (A) P(US adult plays video games regularly) and (B) P(gaming disorder | adult gamer). Factor A: The Entertainment Software Association (ESA) 2025 Essential Facts announcement states that 60% of adults (ages 18 and up) play video games every week (with 205.1 million Americans ages 5-90 playing regularly). We use 0.60 as the prevalence of regular adult gamers in the US. Factor B: The Stevens, Dorstyn, Delfabbro, and King (2021) systematic review and meta-analysis in the Australian and New Zealand Journal of Psychiatry (corrected 2023) found a pooled prevalence of 2.96% of gamers meeting Gaming Disorder or equivalent criteria (original 3.05%, 95% CI 2.38–3.91; a stricter-sampling adjusted estimate was 1.96%). We use 2.0% as a conservative point estimate for the adult gamer subpopulation, noting that adolescent-weighted samples may inflate the global figure. Combined: 0.60 × 0.020 = 0.012. The current prevalence of ~2-3% among gamers approximates lifetime incidence for a disorder that often develops in early adulthood and may resolve; however, gaming disorder in adults is frequently comorbid with mood and anxiety disorders, which prolongs the clinical course. Uncertainty range: 0.006 (0.60 × 1.0% lower prevalence bound) to 0.024 (0.60 × 4.0% accounting for methodological heterogeneity and possible underdiagnosis). Note that lifetime_us_adult (0.012) is strictly inside low (0.006) and high (0.024).\n","uncertainty":{"low":0.006,"high":0.024},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/33028074/","title":"Global prevalence of gaming disorder: A systematic review and meta-analysis","publisher":"Australian and New Zealand Journal of Psychiatry (Stevens, Dorstyn, Delfabbro, King)","source_type":"peer_reviewed","statistic":"Global prevalence of gaming disorder: 3.05% (original 2021); corrected to 2.96% in 2023 corrigendum; 2.5 times higher odds in males","excerpt":"\"Literature documents a global prevalence of 3.05% [corrected 2.96% in 2023], with 2.5 times higher odds in males. Prevalence estimates of gaming disorder were substantially moderated by country, region, age, sex, and type of assessment tool used.\"\n","source_date":"2021-05-14","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505054709/https://pubmed.ncbi.nlm.nih.gov/33028074/","calculation_notes":"This meta-analysis provides the primary prevalence estimate used in Factor B. The corrected 2023 figure (2.96%) is used rather than the original (3.05%). The meta-analysis included studies using ICD-11, DSM-5 IGD, and equivalent criteria. Because many included studies used adolescent or mixed-age samples, the adult-specific prevalence may differ from the pooled figure; the normalized estimate uses 2.0% as a more conservative adult-specific rate to account for this. The Stevens et al. pooled figure is 3.05% (95% CI 2.38–3.91), with a stricter-sampling adjusted estimate of 1.96%, reflecting methodological heterogeneity in how \"gaming disorder\" is assessed — which drives the uncertainty bounds in the normalized estimate.\n","independence_note":"The meta-analysis pools data from multiple independent research groups across different countries and methodologies. The 2023 corrigendum corrects a computational error in the original but does not change the substantive finding.\n"},{"url":"https://www.who.int/news-room/questions-and-answers/item/addictive-behaviours-gaming-disorder","title":"Addictive behaviours: Gaming disorder","publisher":"World Health Organization","source_type":"govt_report","statistic":"ICD-11 6C51 criteria require impaired control, prioritization over other activities, and continuation despite negative consequences, causing significant functional impairment for at least 12 months; WHO states gaming disorder affects only a small proportion of people who game","excerpt":"\"Gaming disorder is defined in the 11th Revision of the International Classification of Diseases (ICD-11) as a pattern of gaming behavior ('digital-gaming' or 'video-gaming') characterized by impaired control over gaming, increasing priority given to gaming over other activities to the extent that gaming takes precedence over other interests and daily activities, and continuation or escalation of gaming despite the occurrence of negative consequences. For gaming disorder to be diagnosed, the behaviour pattern must be of sufficient severity to result in significant impairment in personal, family, social, educational, occupational or other important areas of functioning and would normally have been evident for at least 12 months. Studies suggest that gaming disorder affects only a small proportion of people who engage in digital- or video-gaming activities.\"\n","source_date":"2022-09-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505054742/https://www.who.int/news-room/questions-and-answers/item/addictive-behaviours-gaming-disorder","calculation_notes":"WHO ICD-11 provides the diagnostic framework; the WHO page itself gives no prevalence percentage, stating only that gaming disorder \"affects only a small proportion\" of gamers (the quantitative prevalence comes from the Stevens et al. meta-analysis above). Critically, the ICD-11 criteria require \"significant impairment\" — gaming disorder is not the same as heavy gaming. This distinction is load-bearing for the normalized estimate: the 2–3% figure excludes heavy gamers who maintain functional lives, capturing only those with measurable life-domain impairment. The 12-month minimum duration criterion further filters out transient heavy play during exceptional circumstances (e.g., pandemic lockdowns).\n"},{"url":"https://www.theesa.com/annual-esa-study-reveals-video-games-universal-appeal-across-generations/","title":"Annual ESA Study Reveals Video Games' Universal Appeal Across Generations (2025 Essential Facts)","publisher":"Entertainment Software Association (ESA)","source_type":"reputable_reference","statistic":"60% of adults (ages 18 and up) play video games every week; 205.1 million Americans (ages 5-90) regularly play video games","excerpt":"\"Sixty percent of adults (ages 18 and up) play video games every week and the average age of today's player is now 36 years old. Nearly two-thirds of Americans ages 5 to 90 (205.1 million) regularly play video games.\"\n","source_date":"2025-06-03","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260604182242/https://www.theesa.com/annual-esa-study-reveals-video-games-universal-appeal-across-generations/","calculation_notes":"This source provides Factor A: the base rate of adult gaming participation in the US. The 60% figure for adults (ages 18 and up) playing weekly is used as the probability of being an \"adult gamer\" for purposes of the two-factor estimate, and is stated verbatim in the ESA's 2025 Essential Facts announcement. The ESA reports are the standard industry reference for US gaming participation; the survey is a nationally representative online panel (YouGov). The 60% adult figure is consistent with the prior-year overall participation (61% of Americans ages 5-90 in the 2024 report), indicating stable adult gaming prevalence.\n"}],"comparison_anchors":[{"label":"Alcohol use disorder (lifetime, US adult)","lifetime_us_adult":0.29},{"label":"Major depressive episode (lifetime, US adult)","lifetime_us_adult":0.21},{"label":"Gambling disorder (lifetime, US adult)","lifetime_us_adult":0.025}],"personal_factor_multipliers":[{"factor":"male, age 18-35","multiplier":3,"notes":"Males have 2.5x higher odds of gaming disorder than females in the Stevens et al. meta-analysis; the 18-35 age group has both higher gaming participation and higher disorder rates"},{"factor":"comorbid depression or anxiety","multiplier":4,"notes":"Gaming disorder is highly comorbid with mood and anxiety disorders; the causal direction is bidirectional, with gaming sometimes used as an avoidance strategy for underlying distress"},{"factor":"massively multiplayer online games (MMO) or battle royale primary genre","multiplier":2.5,"notes":"Games with social reinforcement loops, progression systems, and 24/7 availability have higher associated disorder rates than single-player titles"},{"factor":"non-gamer or casual gamer (< 1 hour/week)","multiplier":0.05,"notes":"Gaming disorder requires sustained heavy play; the disorder essentially cannot develop at casual participation levels"}],"short_label":"Gaming disorder (adults)","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"Gaming disorder prevalence estimates vary substantially across studies (the Stevens et al. meta-analysis reports a pooled 3.05% with a 95% CI of 2.38–3.91, and a stricter-sampling adjusted estimate as low as ~2%) depending on the assessment tool, the gaming population sampled, and whether ICD-11 or DSM-5 IGD criteria are applied. The WHO page itself gives no prevalence figure, noting only that gaming disorder affects \"only a small proportion\" of gamers. The DSM-5 includes Internet Gaming Disorder only as a \"condition for further study\" rather than a full diagnosis — a distinction that reflects ongoing scientific debate about diagnostic criteria rather than the absence of clinical phenomena. The current entry uses ICD-11 (WHO) criteria, which became effective internationally in 2022. A 2023 corrigendum to the Stevens et al. meta-analysis corrected the pooled prevalence from 3.05% to 2.96%, a small but methodologically important correction. The 12-month ICD-11 duration criterion means the current prevalence figure approximates lifetime incidence better than it would for conditions with very low spontaneous remission rates; gaming disorder has a documented recovery trajectory with or without formal treatment. Adults with gaming disorder frequently carry comorbid psychiatric diagnoses (depression, ADHD, social anxiety) that complicate attributing harm to gaming specifically. This entry is distinct from screen-time-teen-harm and screen-time-school-age-harm, which address developmental outcomes in younger cohorts.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A game controller resting on a surface in a dimly lit room, muted blue-grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/gaming-disorder-adult","api_url":"https://likelier.app/api/fears/gaming-disorder-adult.json"},{"slug":"skipping-vehicle-inspection","question":"What are the odds of a crash caused by poor vehicle maintenance?","quick_answer":"Vehicle factors are the critical cause in only about 2 in 100 crashes, giving a driver roughly a 1 in 83 lifetime chance of a maintenance-caused crash — lower than the 5–10% many drivers assume.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"The fear of a sudden mechanical failure — brakes giving out on a hill, a tire blowout on the highway — is vivid and cinematic. Vehicle inspection programs exist in roughly half of US states, and their political justification rests on the assumption that uninspected vehicles pose a meaningful crash risk. Yet the NHTSA National Motor Vehicle Crash Causation Survey found that vehicle-related factors are the critical reason in only about 2% of crashes, with driver error dominating at 94%. The public perception of maintenance-related risk tends to be higher than the data support, partly because mechanical-failure crashes are disproportionately severe (a brake failure at highway speed is more likely to be fatal than a fender-bender from inattention) and partly because inspection-program advocates have institutional incentives to emphasize the risk.\n","rough_estimate":"Many drivers assume 5-10% of crashes are caused by vehicle defects","kind":"intuition"},"native":{"display":"~2 in 100 crashes involve vehicle factors as critical reason","numerator":2,"denominator":100,"unit":"per crash (fraction with vehicle critical reason)","population":"US light-vehicle crashes (NMVCCS 2005-2007)"},"normalized":{"lifetime_us_adult":0.012,"display":"~1 in 83 lifetime","log_value":-1.92,"assumptions":"NHTSA's NMVCCS found vehicle factors as the critical reason in 2% of crashes (~44,000 of ~2.2 million annual crashes). NHTSA estimates ~2,600 deaths per year from maintenance-related failures. An average US driver has roughly a 77% lifetime probability of being in at least one police-reported crash over 59 years of driving. Applying the 2% vehicle-factor fraction: 0.77 × 0.02 ≈ 0.015. We use 0.012 as a slightly conservative central estimate, acknowledging that \"vehicle critical reason\" includes manufacturing defects and recalls in addition to owner-neglected maintenance. The lifetime figure represents the probability of being in a crash where vehicle maintenance or condition was the primary cause.\n","uncertainty":{"low":0.005,"high":0.025},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811059","title":"National Motor Vehicle Crash Causation Survey: Report to Congress","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"Vehicle-related factors were the critical reason for the crash in approximately 2% of cases (44,000 of ~2.2 million crashes); tires, brakes, and steering/suspension were the leading vehicle factors","excerpt":"\"The critical reason was assigned to the vehicle in an estimated 44,000 crashes (2 percent). Among those, tire/wheel failure was the most frequently cited vehicle-related critical reason, followed by brake-related problems and steering/suspension/transmission/ engine-related failures.\"\n","source_date":"2008-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260525100214/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811059","calculation_notes":"The NMVCCS is NHTSA's definitive crash-causation study, based on on-scene investigation of 5,470 crashes from 2005-2007 weighted to national estimates. \"Critical reason\" is defined as the immediate reason for the critical pre-crash event — the last failure in the causal chain. The 2% vehicle-factor figure includes both maintenance-related degradation (worn tires, failed brakes) and manufacturing defects. It does not mean vehicle condition was irrelevant in 98% of crashes — it means driver or environment was the proximate cause. Vehicle condition may have been a contributing factor in additional crashes without being the critical reason. The full critical-reason breakdown reported by NHTSA for the NMVCCS is: drivers 94% (±2.2%), vehicles 2% (±0.7%), environment (road surface, weather, etc.) 2% (±1.3%), and unknown/unattributed 2% (±1.4%).\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811617","title":"Tire-Related Factors in the Pre-Crash Phase","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"Tire problems were present in ~5% of pre-crash vehicles; underinflation and inadequate tread depth were the leading tire-related factors","excerpt":"\"Of the estimated 3,889,770 vehicles involved in the NMVCCS crashes, 5 percent experienced tire problems in the pre-crash phase. ... A vehicle is more likely to experience tire problems when one or more tires are underinflated or the vehicle is running on tires with inadequate tread depth.\"\n","source_date":"2012-04-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260118202216/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811617","calculation_notes":"This NHTSA technical report uses NMVCCS data to focus on tire-related factors. Tire problems being \"present\" in ~5% of pre-crash vehicles (5 percent of the estimated 3,889,770 crash-involved vehicles) is higher than the 2% critical-reason figure because a tire problem can be present without being the critical reason (e.g., a driver with low tire pressure rear-ends someone due to distraction — the tire was degraded but the driver was the critical reason). This distinction between \"contributing factor\" and \"critical reason\" matters for interpreting the 2% figure.\n"},{"url":"https://www.lowelawgroup.com/blog/how-often-does-poor-maintenance-result-in-car-accidents/","title":"How Often Does Poor Maintenance Result In Car Accidents?","publisher":"Lowe Law Group (citing NHTSA data)","source_type":"reputable_reference","statistic":"NHTSA estimates poor vehicle maintenance leads to approximately 2,600 deaths, 100,000 disabling injuries, and nearly $2 billion in losses annually","excerpt":"\"Research by NHTSA estimates that poor vehicle maintenance leads to approximately 2,600 deaths, 100,000 disabling injuries, and nearly $2 billion in lost wages, medical expenses, and property damage every year in the United States.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260505063737/https://www.lowelawgroup.com/blog/how-often-does-poor-maintenance-result-in-car-accidents/","calculation_notes":"The 2,600 annual deaths figure, widely cited and attributed to NHTSA, provides the severity context. Against ~40,000 total US traffic fatalities per year, maintenance- related deaths account for ~6.5% of the total — higher than the 2% critical-reason share of all crashes because maintenance-related crashes skew more severe (tire blowouts at highway speed, brake failures on grades). Lifetime fatal risk from maintenance-caused crashes: 2,600/330M × 59 years ≈ ~1 in 2,200.\n"},{"url":"https://www.gao.gov/products/gao-15-705","title":"Vehicle Safety Inspections: Improved DOT Communication Could Better Inform State Programs","publisher":"U.S. Government Accountability Office (GAO)","source_type":"govt_report","statistic":"Studies GAO reviewed and GAO's own analysis of state data found no clear influence of periodic vehicle-inspection programs on crash rates related to vehicle component failure","excerpt":"\"Studies GAO reviewed and GAO's analysis of state data examined the effect of inspection programs on crash rates related to vehicle component failure, but showed no clear influence.\"\n","source_date":"2015-08-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260109053420/https://www.gao.gov/products/gao-15-705","calculation_notes":"GAO's 2015 report to Congress reviewed the available research on periodic passenger- vehicle safety inspection (PVSI) programs, including before/after crash data from New Jersey and Oklahoma after those states eliminated their inspection programs, and found no statistically demonstrable reduction in vehicle-component-failure crashes attributable to inspection programs. This is the basis for the body-text claim that studies comparing crash rates across inspecting and non-inspecting states have found no clear effect from the programs.\n"},{"url":"https://www.govinfo.gov/content/pkg/FR-2007-07-12/html/07-3382.htm","title":"Federal Motor Vehicle Safety Standards; Tire Pressure Monitoring Systems (Final Rule)","publisher":"National Highway Traffic Safety Administration (NHTSA) / Federal Register","source_type":"govt_report","statistic":"TPMS phase-in required 20% of light vehicles Oct 2005-Aug 2006, 70% Sept 2006-Aug 2007, and full compliance for all covered light vehicles by September 1, 2007","excerpt":"\"Vehicle manufacturers are not required to comply with the requirements related to the TPMS malfunction indicator (including associated owner's manual requirements) until September 1, 2007; however, at that point, all covered vehicles must meet all relevant requirements of the standard.\"\n","source_date":"2007-07-12","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260714015929/https://www.govinfo.gov/content/pkg/FR-2007-07-12/html/07-3382.htm","calculation_notes":"FMVSS No. 138 implements the TREAD Act's tire-pressure-monitoring-system mandate. The phase-in schedule required 20% compliance for light vehicles manufactured October 2005 through August 2006, 70% for September 2006 through August 2007, and full compliance for all covered light vehicles (passenger cars and trucks under 10,000 lbs GVWR) manufactured on or after September 1, 2007. Grounds the body-text claim that TPMS has been mandated on all new US vehicles since 2007, addressing tire/wheel failure — the leading vehicle-factor crash category identified in the NMVCCS above.\n"}],"comparison_anchors":[{"label":"Car crash (any, lifetime, US adult)","lifetime_us_adult":0.77},{"label":"Car crash death (lifetime, US adult)","lifetime_us_adult":0.0093}],"personal_factor_multipliers":[{"factor":"Never checks tire pressure or tread","multiplier":3,"notes":"Tire-related factors are the leading vehicle cause; underinflation and worn tread are the primary tire problems"},{"factor":"Vehicle older than 15 years, no inspections","multiplier":2.5,"notes":"Older vehicles accumulate brake, suspension, and lighting degradation; states without inspection programs show slightly higher vehicle-factor crash rates"},{"factor":"Regular maintenance per manufacturer schedule","multiplier":0.3,"notes":"Properly maintained vehicles have substantially lower rates of critical vehicle failures"},{"factor":"High-mileage vehicle (>150,000 miles, no recent brake/suspension inspection)","multiplier":2.5,"notes":"NHTSA TREAD Act data and tire/brake failure analysis show cumulative wear on high-mileage vehicles substantially increases the probability of brake, suspension, and steering failures that are the primary vehicle-factor crash causes"},{"factor":"Winter climate without seasonal tire change (driving on summer tires on snow/ice)","multiplier":2.5,"notes":"NHTSA safety research and Transport Canada testing data document that summer tires on snow or ice increase stopping distances by approximately 50-100% compared to winter tires, materially increasing the probability of a crash attributable to tire condition"}],"short_label":"Maintenance crash","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The 2% critical-reason figure from NMVCCS is the most rigorous available but dates from 2005-2007 data. Modern vehicles have more electronic safety systems (ABS, ESC, TPMS) that may have reduced the vehicle-factor share. Conversely, the average age of vehicles on US roads has increased to ~12.6 years, which works in the opposite direction. The entry covers crashes where vehicle condition was the primary cause, not crashes where it was a contributing factor — the latter share would be higher. The 2,600 annual deaths figure is widely cited but its exact methodology is not publicly documented in a single primary NHTSA report.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified car tire with low tread rendered in muted slate and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-vehicle-inspection","api_url":"https://likelier.app/api/fears/skipping-vehicle-inspection.json"},{"slug":"high-heel-injury-ed-visit","question":"What are the odds of an emergency-room visit from a high-heel-related injury over a lifetime of wearing heels?","quick_answer":"Across a 40-year heel-wearing career, a US woman's odds of an ER visit from a high-heel injury are about 1 in 80 (roughly 1 in 3,100 in any given year). Most wearers would guess well under 1 in 1,000 over a lifetime, treating heel risk as chronic foot pain rather than acute, hospital-grade injury.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"The folk model of high-heel risk is chronic and cosmetic: bunions, Morton's neuroma, a sore lower back, eventually maybe a podiatrist. Acute, hospital-visit-grade injury on a single night out is not typically what most heel wearers consciously weigh. Women who wear heels regularly have almost all rolled an ankle or scraped a heel at some point and counted it as a near-miss. There is no good survey isolating the perceived probability of an ER visit specifically attributable to heels, so we mark this as editorial intuition. The interesting property of the fear is that the acute rate is roughly what the chronic-foot-pain rate would predict if you assumed a small fraction of bad steps reached an ER — but very few wearers carry that as the headline risk in their head.\n","rough_estimate":"Most heel wearers would guess the lifetime ER-visit risk at well under 1 in 1,000","kind":"intuition"},"native":{"display":"~16,000 ED visits/year among ~50 million US women who regularly wear heels (~1 in 3,100/year)","numerator":16000,"denominator":50000000,"unit":"per year (US women who regularly wear high heels)","population":"US adult women who regularly wear high-heeled shoes","exposures_per_year":1},"normalized":{"lifetime_us_adult":0.0127,"display":"~1 in 80 over a 40-year heel-wearing career (US adult woman)","log_value":-1.896,"assumptions":"Scope is subgroup_lifetime: US adult women who regularly wear high-heeled shoes across an approximate 40-year heel-wearing window (roughly ages 18 to 58). Starting from a per-year baseline of approximately 16,000 ED visits attributed to high-heel- related injury (Cohen 2022, pre-pandemic 2016 to 2019 average for women aged 15 to 69) and a US heel-wearer base of approximately 50 million (about 130 million US adult women times the APMA 2014 estimate that roughly 49 percent wear heels, rounded down to account for \"wear\" not being operationalised by frequency), the per-wearer annual risk is roughly 1 in 3,125. Compounded across 40 years: 1 minus (1 minus 1/3125) to the 40th power equals approximately 0.0127, or about 1 in 79. The uncertainty band is wide on both sides because (a) the wearer-prevalence number is a single trade- association survey with no frequency cut, and (b) the NEISS narrative filter is a known undercount — many sprains caused by heels never mention heels in the chart note. The headline is therefore better read as a lower bound on \"heel-attributable acute injury serious enough to be medically attended\" rather than the full burden.\n","uncertainty":{"low":0.006,"high":0.02},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/25977152/","title":"Epidemiology of High-Heel Shoe Injuries in U.S. Women: 2002 to 2012","publisher":"Journal of Foot and Ankle Surgery (Moore JX, Lambert B, Jenkins GP, McGwin G Jr)","source_type":"peer_reviewed","statistic":"123,355 estimated ED visits over 11 years (2002 to 2012); overall rate 7.32 per 100,000 females (95% CI 7.08 to 7.56); rate roughly doubled across the period; ages 20 to 29 highest at 18.38 per 100,000, ages 30 to 39 at 11.07 per 100,000; over 80 percent foot or ankle injuries; ~19 percent fractures; 49.5 percent of injuries occurred at home","excerpt":"\"A total of 3294 injuries, representing an estimated 123,355 high-heel-related injuries, were treated in emergency departments within the United States from 2002 to 2012.\" [Paraphrase from abstract — full text paywalled] \"The overall rate of high-heel-related injuries was 7.32 per 100,000 females (95% confidence interval 7.08 to 7.56).\" \"Our results suggest that high-heel-related injuries have nearly doubled during the 11-year period from 2002 to 2012.\" \"Most injuries occurring as sprains and strains to the foot and ankle.\"\n","source_date":"2015-07-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260208025217/https://pubmed.ncbi.nlm.nih.gov/25977152/","calculation_notes":"Moore et al. is the foundational US epidemiology of acute high-heel injury, built on a NEISS narrative search across 100 US sentinel emergency departments 2002 to 2012. Two figures anchor this page: the long-run all-female rate of 7.32 per 100,000 (used as the floor of the per-year-per-woman range, not the per- wearer rate), and the age-banding (ages 20 to 29 at 2.5 times the average rate), which drives the personal factor multipliers below. The paper's 123,355 / 11 = ~11,200 ED visits per year average is *lower* than Cohen 2022's 2016 to 2019 figure (~16,000) because the rate was rising across the Moore window — the headline calculation uses Cohen's more recent baseline, with Moore's older figure setting the low end of the uncertainty band.\n","independence_note":"Both Moore 2015 and Cohen 2022 use the same NEISS national surveillance system, so they are method-correlated, not fully independent measurements. The two are cited together because they cover non-overlapping time windows (2002 to 2012 vs 2016 to 2020) and Cohen explicitly extends Moore's series.\n"},{"url":"https://journals.sbmu.ac.ir/sdh/article/view/37227","title":"Pandemic-related decline in injuries related to women wearing high-heeled shoes: Analysis of U.S. data for 2016 to 2020","publisher":"Social Determinants of Health (Cohen PN, University of Maryland)","source_type":"peer_reviewed","statistic":"2016 to 2019 baseline approximately 16,000 ED visits per year among US women aged 15 to 69; 2020 dropped to 6,290 visits (5.40 per 100,000, 95% CI 3.95 to 6.86) after COVID shutdowns; no significant change in fracture share or hospital-admission share","excerpt":"\"There were an estimated 6,290 high-heel related emergency department visits involving women ages 15 to 69 in 2020, compared with approximately 16,000 per year in 2016 to 2019.\" \"The 2020 rate of 5.40 HHSRIs per 100,000 women ages 15 to 69 was significantly below the peak in 2017.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260219023833/https://journals.sbmu.ac.ir/sdh/article/view/37227","calculation_notes":"Cohen 2022 provides the most current pre-pandemic baseline (~16,000 ED visits per year), which we use as the numerator for the headline calculation. The 2020 pandemic figure (6,290) is informative but suppressed by behaviour change — most women were not commuting in heels in 2020 and we treat that year as not representative of a steady-state denominator. Using 16,000 / 50 million heel- wearing women equals roughly 1 in 3,125 per year, which compounds over 40 years to approximately 1 in 79. If the pre-pandemic 16,000 is replaced by Moore's older 11,000 (still inside the same NEISS series), the per-year rate becomes 1 in 4,545 and the 40-year cumulative is approximately 1 in 113 — this is the low end of our uncertainty band.\n","independence_note":"Shares the NEISS surveillance backbone with Moore 2015. Treat as a temporal extension and methodological cross-check, not an independent measurement.\n"},{"url":"https://www.prnewswire.com/news-releases/new-study-shows-high-heels-are-biggest-culprit-of-female-foot-pain-259775731.html","title":"New Study Shows High Heels are Biggest Culprit of Female Foot Pain","publisher":"American Podiatric Medical Association (APMA), via PR Newswire","source_type":"reputable_reference","statistic":"Nearly half of US adult women (49 percent) wear high heels; 71 percent of heel wearers report the shoes hurt their feet; survey of 1,000 US adults aged 18 and over (2014)","excerpt":"\"Nearly half of all women (49 percent) wear high heels, even though the majority of heel wearers (71 percent) complain these shoes hurt their feet.\" \"The study, which surveyed 1,000 US adults ages 18 and older...\"\n","source_date":"2014-05-19","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20251212171952/https://www.prnewswire.com/news-releases/new-study-shows-high-heels-are-biggest-culprit-of-female-foot-pain-259775731.html","calculation_notes":"The APMA survey is the denominator anchor for the per-wearer calculation: approximately 130 million US adult women times 49 percent equals roughly 64 million heel wearers. We round down to 50 million in the headline because \"wear high heels\" in APMA's survey is not operationalised by frequency — a respondent who owns heels and wears them twice a year answers yes, and they are not really the population the per-year ED-visit denominator should be built from. The 30 to 50 percent prevalence range maps directly into the uncertainty band: at 30 percent prevalence the per-wearer annual rate would be roughly 1 in 2,400, giving a 40-year cumulative of approximately 1 in 60 (high end of the band).\n","independence_note":"Independent of NEISS — different research design, different question, different population. Trade-association survey is the weakest of the three sources on method, but no peer-reviewed prevalence study with a frequency cut exists.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24105612/","title":"The Incidence and Prevalence of Ankle Sprain Injury: A Systematic Review and Meta-Analysis of Prospective Epidemiological Studies","publisher":"Sports Medicine (Doherty C, Delahunt E, Caulfield B, Hertel J, Ryan J, Bleakley C)","source_type":"peer_reviewed","statistic":"Pooled ankle sprain incidence 13.6 per 1,000 exposures in females vs 6.94 per 1,000 in males across prospective studies; ankle sprain is one of the most common musculoskeletal injuries","excerpt":"\"A higher incidence of ankle sprain in females compared with males (13.6 vs 6.94 per 1,000 exposures), in children compared with adolescents (2.85 vs 1.94 per 1,000 exposures).\" [Paraphrase from abstract — full text paywalled]\n","source_date":"2014-01-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260227113953/https://pubmed.ncbi.nlm.nih.gov/24105612/","calculation_notes":"Doherty et al. is not a high-heel study; it is the meta-analytic anchor for the background female-vs-male ankle-sprain rate ratio (roughly 2x). Used here only as context for the personal factor multipliers and to underline that the heel- attributable ED-visit count is one slice of the much larger female ankle-sprain burden — most sprains, including some caused by heels, never reach an ED at all. We do not derive any headline number from this source.\n","independence_note":"Independent of NEISS, independent of the APMA survey. Cited as method context, not as a per-wearer rate measurement.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Serious skiing injury per 20-day season (active recreational skier)","lifetime_us_adult":0.0392},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Per year (US heel-wearing woman, 2016 to 2019 baseline)","probability":0.00032,"notes":"Cohen 2022 pre-pandemic figure: ~16,000 ED visits / ~50M heel-wearers. Roughly 1 in 3,100 per year. Drops to about 1 in 8,000 per year in 2020 under pandemic conditions."},{"region":"Per year, ages 20 to 29 (US heel-wearing woman)","probability":0.0008,"notes":"Moore 2015 reports 18.38 per 100,000 women in this age band, roughly 2.5x the all-female average. Heel use is concentrated in this group and the rate scales accordingly."},{"region":"Per year, ages 50+ (US heel-wearing woman)","probability":0.0001,"notes":"Order-of-magnitude estimate. Moore 2015 shows steep age decline after 40; heel use drops sharply and so does the absolute rate. Population at risk is much smaller, so per-wearer risk does not fall by the same factor."},{"region":"Per 40-year heel-wearing career (cumulative)","probability":0.0127,"notes":"Headline subgroup_lifetime figure: about 1 in 80 women who wear heels regularly across their adult life will visit an ED with a heel-attributed injury at least once."}],"personal_factor_multipliers":[{"factor":"ages 20 to 29","multiplier":2.5,"notes":"Moore 2015: 18.38 ED visits per 100,000 women aged 20 to 29 vs 7.32 across all ages. Peak combines highest heel use, highest heel height, highest exposure context (clubs, nights out, alcohol), and least practice on the shoe."},{"factor":"ages 30 to 39","multiplier":1.5,"notes":"Moore 2015: 11.07 per 100,000. Heel use still high but contexts shift toward workplace; recovery from a roll is also slower than at 25."},{"factor":"ages 50 and over","multiplier":0.3,"notes":"Heel use drops sharply after menopause, which reduces population at risk by far more than the per-wearer rate. The risk per actual wear may not fall as much; a fall from a heel at 60 is more likely to involve a fracture than a sprain."},{"factor":"occasional wearer (events only, not weekly)","multiplier":0.4,"notes":"Approximation: lower annual exposure mechanically reduces annual probability roughly linearly. Offsetting effect: low-frequency wearers have less practiced balance in the shoe."},{"factor":"daily heel wearer (workplace dress code)","multiplier":2,"notes":"Roughly 250 wear-days per year vs ~50 for an occasional wearer. Per-day rate is approximately constant, so annual probability scales with exposure days. Capped at ~2x because the most experienced wearers tend to choose lower heels."},{"factor":"alcohol involved","multiplier":4,"notes":"Order-of-magnitude estimate from NEISS narrative review: a meaningful share of evening-event heel injuries involve alcohol. Not a clean separation in Moore 2015's coding, but the time-of-day distribution and at-home/public-place split are both consistent with intoxication being a major contributing factor."}],"short_label":"High-heel ER visit","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"Three structural uncertainties make this entry's number a lower bound. First, the NEISS narrative filter that produces the 16,000 per year figure captures only ED visits where the discharge or triage narrative explicitly names \"high heel\" or a close synonym — sprains caused by heels but coded only as \"fall\" or \"ankle injury\" are missed. Second, the wearer base of approximately 50 million is anchored on a single 2014 APMA survey that does not distinguish daily wearers from women who own a pair worn twice a year; the true denominator for the \"at-risk\" population could be half this size, in which case the per-wearer rate (and the cumulative lifetime figure) is roughly double. Third, \"high heel\" is itself a broad category — stiletto versus block heel versus 2-inch pump have meaningfully different fall and sprain rates, but no national surveillance system separates them. The headline is best read as \"any acute injury attributed to a high-heeled shoe and serious enough to be medically attended\" rather than the full burden of heel-related musculoskeletal harm, which includes chronic foot pain, Morton's neuroma, and bunion surgery that this page does not cover. The cumulative figure also excludes the much larger number of falls that happen in heels but are reported simply as falls. The fear's gap is in the other direction from many entries on this site: the acute headline (~1 in 80 lifetime ED visit) is *higher* than most heel wearers would intuit, while the catastrophic-outcome version of the fear (heel fall causing serious head injury or death) is genuinely rare.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-05-28","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single high-heeled shoe lying on its side on a flat stone step, viewed from a low angle, calm muted palette."},"canonical_url":"https://likelier.app/high-heel-injury-ed-visit","api_url":"https://likelier.app/api/fears/high-heel-injury-ed-visit.json"},{"slug":"colorectal-cancer","question":"What are the odds of dying from colorectal cancer?","quick_answer":"The lifetime odds of dying from colorectal cancer are about 1 in 77 for a global adult. It draws less fear than breast or lung cancer despite being one of the leading causes of cancer death — and one of the most screening-preventable.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Colorectal cancer is not as culturally salient as breast or lung cancer — it lacks the ribbon, the celebrity campaigns, and the bodily shorthand of \"the lump\" or \"the cough\". Most readers, asked cold, will rank it below both, and below several cancers that actually kill fewer people. At the same time, the intervention story is famously good: colonoscopy is one of the most effective cancer screens in medicine, and public awareness of that fact has been slow to catch up. The recent rise in early-onset disease has further scrambled the mental model of colorectal cancer as \"something older men get\".\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); colorectal ranks below breast and lung in unprompted cancer-fear salience","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~900,000 colorectal cancer deaths per year globally (~1.9M new cases)","numerator":1,"denominator":8900,"unit":"per year","population":"global, all ages, colorectal cancer only"},"normalized":{"lifetime_us_adult":0.013,"display":"~1 in 77 lifetime (global adult)","log_value":-1.886,"assumptions":"Starts from the IARC GLOBOCAN 2022 global colorectal cancer headline: ~1.9 million new cases and more than 900,000 deaths per year worldwide, making CRC the third most common cancer and the second most common cause of cancer death globally. Spread across a global adult population of ~5.5 billion (age 18+), ~900,000 CRC deaths per year is ~1.6 per 10,000 adults per year. Age-weighted lifetime compounding (CRC mortality is heavily concentrated above age 60, with hazard several times higher in the last third of adult life than at the population average) lands a global adult lifetime CRC-death figure near 1.3%. The direct US number from SEER and ACS-derived estimates is roughly 2% (~1 in 50), driven by higher incidence in high-income countries; see regional_breakdown for the country spread. Headline figure 0.013 (~1 in 77) with an uncertainty band of 1 in 55 to 1 in 125 to reflect the global/US gap plus age-structure sensitivity. Scope is global-adult-lifetime to match the `cancer-lifetime` parent entry; the US figure sits at the top of the band.\n","uncertainty":{"low":0.008,"high":0.02},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/cancer-type/colorectal-cancer/","title":"Colorectal Cancer","publisher":"International Agency for Research on Cancer (IARC) / World Health Organization","source_type":"govt_report","statistic":"More than 1.9 million colorectal cancer cases diagnosed in 2022 and more than 900,000 deaths per year globally; third most common cancer, second most common cause of cancer death","excerpt":"\"In 2022, more than 1.9 million cases were diagnosed... leading to more than 900 000 deaths per year.\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164823/https://www.iarc.who.int/cancer-type/colorectal-cancer/","calculation_notes":"GLOBOCAN 2022 headline used directly as the native number. ~900,000 annual global CRC deaths across ~5.5 billion adults is ~1.6 per 10,000 adult-years. Age-weighted over a 60-year adult window — CRC hazard in the 60s and 70s is several times the adult average — gives a lifetime adult-lifetime mortality near 0.013 (~1 in 77). This matches the global column when cross-checked against the direct US SEER estimate (~2%) discounted for lower-incidence regions in the GLOBOCAN regional breakdown.\n","independence_note":"IARC GLOBOCAN is the upstream dataset used by WHO, ACS international comparisons, and the IHME Global Burden of Disease CRC module. Treat this as the canonical global source; the SEER/ACS US number below is the methodologically independent cross-check.\n"},{"url":"https://seer.cancer.gov/statfacts/html/colorect.html","title":"Cancer Stat Facts: Colorectal Cancer","publisher":"Surveillance, Epidemiology, and End Results (SEER) Program, National Cancer Institute","source_type":"govt_report","statistic":"Approximately 3.9% of men and women will be diagnosed with colorectal cancer at some point during their lifetime; ~154,270 new cases and ~52,900 deaths estimated for 2025; 5-year relative survival 65.4% (2015-2021)","excerpt":"\"Approximately 3.9 percent of men and women will be diagnosed with colorectal cancer at some point during their lifetime.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164902/https://seer.cancer.gov/statfacts/html/colorect.html","calculation_notes":"SEER gives direct lifetime incidence of ~3.9% for the US population. With 5-year relative survival at 65.4%, the implied long-run case-fatality is roughly 35% (a conservative upper bound, since 5-year survival underestimates long-run cure for CRC). 3.9% lifetime incidence × ~35-50% long-run case-fatality yields a US lifetime CRC-death probability of ~1.5-2%, consistent with the \"1 in 50\" figure commonly cited by ACS. This anchors the US row in the regional breakdown and the top of the Likelier uncertainty band.\n","independence_note":"SEER (NCI) and IARC GLOBOCAN (WHO) are methodologically independent compilation pipelines. SEER uses US vital registration and population-based cancer registries; IARC aggregates national registry data worldwide. The two agree on scale, and the direct US figure is higher than the global figure by roughly the ratio expected from incidence differences between high-income and lower-incidence regions.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/36856579/","title":"Colorectal cancer statistics, 2023","publisher":"Siegel RL, Wagle NS, Jemal A, et al. / CA: A Cancer Journal for Clinicians","source_type":"peer_reviewed","statistic":"CRC mortality declined 2%/yr from 2011-2020 overall but increased 0.5-3%/yr in adults younger than 50; proportion of cases in adults younger than 55 rose from 11% in 1995 to 20% in 2019","excerpt":"\"CRC mortality declined by 2% annually from 2011-2020 overall but increased by 0.5%-3% annually in individuals younger than 50 years and in Native Americans younger than 65 years.\"\n","source_date":"2023-03-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165055/https://pubmed.ncbi.nlm.nih.gov/36856579/","calculation_notes":"Siegel et al. 2023 is the authoritative ACS/SEER-based peer-reviewed summary of CRC trends. Used here to establish (1) the age-weighted distribution of CRC mortality that drives the adult lifetime compounding, (2) the early-onset trend that motivated the USPSTF screening-age change, and (3) the overall ~2%/yr decline in total CRC mortality, which means the headline lifetime figure is slowly drifting down even as the early-onset segment rises.\n","independence_note":"Uses SEER incidence data and NCHS mortality data — same upstream as the SEER Stat Facts source above. Treat as a dependent but methodologically richer analysis of the same pipeline rather than an independent verification.\n"},{"url":"https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/colorectal-cancer-screening","title":"Final Recommendation Statement: Colorectal Cancer: Screening","publisher":"US Preventive Services Task Force","source_type":"govt_report","statistic":"USPSTF recommends CRC screening for all adults 45-75 (Grade B for 45-49, Grade A for 50-75); lowered from age 50 to age 45 in May 2021","excerpt":"\"The USPSTF now recommends offering screening starting at age 45 years.\"\n","source_date":"2021-05-18","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165013/https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/colorectal-cancer-screening","calculation_notes":"USPSTF is the authoritative US primary-care screening guideline. The 2021 move from age 50 to age 45 was a direct response to the early-onset trend documented by Siegel et al. and the SEER data. Used here to anchor the \"screening works\" claim and the screening-from-45 multiplier in personal_factor_multipliers.\n","independence_note":"USPSTF evidence synthesis is methodologically independent of SEER/IARC incidence tracking — it aggregates RCT and cohort evidence on screening effectiveness, not incidence registry data.\n"},{"url":"https://www.cancer.gov/news-events/cancer-currents-blog/2020/colorectal-cancer-rising-younger-adults","title":"Colorectal Cancer Rising among Young Adults","publisher":"National Cancer Institute (NCI)","source_type":"govt_report","statistic":"CRC incidence has been rising steadily among adults younger than 50 since the 1990s; several organizations have lowered screening age from 50 to 45","excerpt":"\"Since the 1990s, the rate of colorectal cancer (which includes cancers of the colon and rectum) has been rising steadily among adults younger than 50.\"\n","source_date":"2020-11-05","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165048/https://www.cancer.gov/news-events/cancer-currents-blog/2020/colorectal-cancer-rising-younger-adults","calculation_notes":"NCI blog summarizing the early-onset CRC trend for a general audience, used as a plain-English cross-check on the Siegel 2023 peer-reviewed figures. Confirms that the doubling of CRC incidence in adults under 50 since the early 1990s is the established consensus view at NCI.\n","independence_note":"Draws on the same SEER incidence data pipeline as Siegel 2023 and SEER Stat Facts. Used as an authoritative plain-English citation, not as an independent verification.\n"},{"url":"https://ascopubs.org/doi/full/10.1200/JCO.22.01895","title":"Sex and Tumor-Site Differences in the Association of Alcohol Intake With the Risk of Early-Onset Colorectal Cancer","publisher":"Jin EH, Han K, Shin CM, et al. / Journal of Clinical Oncology","source_type":"primary_study","statistic":"In 5,666,576 South Koreans aged 20-49, moderate drinkers (aHR 1.09, 95% CI 1.02-1.16) and heavy drinkers (aHR 1.20, 95% CI 1.11-1.29) had significantly higher early-onset CRC risk vs light drinkers; 8,314 incident cases over 2009-2019","excerpt":"\"Compared with light drinking (reference), the risk of early-onset CRC was significantly associated with moderate and heavy drinking (adjusted HR [aHR], 1.09 [95% CI, 1.02 to 1.16] and aHR, 1.20 [95% CI, 1.11 to 1.29], respectively).\"\n","source_date":"2023-06-14","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20241218091900/https://ascopubs.org/doi/full/10.1200/JCO.22.01895","calculation_notes":"Large population-based cohort (Korean National Health Insurance Service) establishing alcohol as a modifiable driver of early-onset (<50) CRC incidence relative to light drinking. Drinking levels were 10-<30 g/day (moderate) and >=30 g/day (heavy) for men; 10-<20 and >=20 g/day for women. Used to anchor the heavy-vs-light alcohol multiplier (1.2) in personal_factor_multipliers. Does not alter the headline lifetime-mortality figure: it measures incidence in the under-50 segment, which the caveats note is still an order of magnitude below the over-65 rate.\n","independence_note":"Methodologically independent of the SEER/IARC US and global pipelines: Korean national-insurance cohort data, not US/global registry compilation. Provides the early-onset alcohol-exposure dimension that the SEER and GLOBOCAN headline sources do not measure.\n"},{"url":"https://onlinelibrary.wiley.com/doi/10.1111/codi.70046","title":"Alcohol consumption and risk of early onset colorectal cancer: A systematic review and meta-analysis","publisher":"Huang S, Yu L, Xiong F, et al. / Colorectal Disease","source_type":"peer_reviewed","statistic":"Pooled OR 1.56 (95% CI 1.28-1.89) for high vs low alcohol intake and early-onset colorectal cancer; systematic search through June 2024","excerpt":"\"a heightened risk of EOCRC for increased alcohol intake (OR = 1.56, 95% CI 1.28-1.89, I2 = 89.3%).\"\n","source_date":"2025-03-01","source_accessed":"2026-06-14","calculation_notes":"Meta-analytic corroboration of the alcohol / early-onset CRC association used in the alcohol multiplier notes. The OR 1.56 (high vs low) reflects a larger exposure contrast than the Korean cohort's heavy-vs-light aHR 1.20; the multiplier is anchored to the more conservative cohort figure. High heterogeneity (I2 89.3%) is the reason the multiplier is kept modest and flagged as incidence-not-mortality in its notes.\n","independence_note":"Independent meta-analysis pooling multiple cohort and case-control studies; provides a second line of evidence for the alcohol multiplier beyond the single Korean cohort.\n"}],"comparison_anchors":[{"label":"All-cancer death (global adult lifetime)","lifetime_us_adult":0.14},{"label":"Lung cancer death (global adult lifetime)","lifetime_us_adult":0.023},{"label":"Breast cancer death (global adult women lifetime)","lifetime_us_adult":0.016},{"label":"Stroke death (global adult lifetime)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average","probability":0.013,"notes":"~900,000 CRC deaths/yr across ~8B people (IARC GLOBOCAN 2022); age-weighted adult lifetime figure"},{"region":"US adult","probability":0.02,"notes":"Direct SEER/ACS figure; ~1 in 50 lifetime, anchored on 3.9% lifetime incidence and ~35-50% long-run case fatality"},{"region":"Western Europe","probability":0.018,"notes":"Similar incidence to the US; somewhat lower mortality due to earlier-stage diagnosis in national screening programs"},{"region":"East Asia","probability":0.015,"notes":"Rising incidence as diets shift; Japan and South Korea now among the highest-incidence countries, offset by strong screening uptake"},{"region":"Sub-Saharan Africa","probability":0.006,"notes":"Lower incidence; confounded by competing mortality (infectious disease, maternal, injury) removing adults from the denominator before peak CRC-risk age"}],"personal_factor_multipliers":[{"factor":"family history (first-degree relative with CRC)","multiplier":2,"notes":"Roughly doubles lifetime risk; stronger if the relative was diagnosed before 50 or multiple relatives are affected"},{"factor":"Lynch syndrome (MLH1/MSH2/MSH6/PMS2/EPCAM pathogenic variant)","multiplier":20,"notes":"Lifetime CRC risk in Lynch carriers runs 40-80% depending on gene; mortality multiplier is somewhat lower because surveillance is intensive"},{"factor":"regular screening (colonoscopy) from age 45","multiplier":0.4,"notes":"Approximately 60% reduction in CRC mortality in compliant populations; one of the largest effect sizes in cancer screening"},{"factor":"high red and processed meat consumption","multiplier":1.3,"notes":"IARC classifies processed meat as Group 1 (carcinogenic) and red meat as Group 2A specifically for CRC; effect size is modest but robust across cohorts"},{"factor":"inflammatory bowel disease (ulcerative colitis / Crohn's)","multiplier":3,"notes":"Long-standing colitis with extensive colonic involvement is the strongest non-hereditary risk factor; multiplier grows with duration and extent"},{"factor":"moderate-to-heavy alcohol consumption (especially under age 50)","multiplier":1.2,"notes":"Anchored to the heavy-drinker adjusted hazard ratio (aHR 1.20, 95% CI 1.11-1.29) vs light drinkers for early-onset (<50) CRC in a 5.7-million-person South Korean cohort (Jin et al., J Clin Oncol 2023); moderate drinking aHR was 1.09. A 2025 meta-analysis (Huang et al., Colorectal Disease) found OR 1.56 for high vs low alcohol intake in early-onset CRC. Note: these measure early-onset incidence relative to light (not zero) drinking, not lifetime mortality; the value is a modifiable-driver flag, applied here in the spirit of the existing red-meat factor."}],"short_label":"Colorectal cancer","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This is mortality, not incidence: lifetime *incidence* of CRC in the US is roughly 3.9% per SEER, and the mortality figure is smaller because CRC is one of the more survivable solid tumors when caught early — 5-year relative survival is ~65% overall and above 90% for localised disease. The \"screening works\" multiplier above is genuinely one of the largest in cancer prevention, but it is conditioned on actually completing the screen on schedule, not on being eligible. The early-onset trend is real but small in absolute numbers: even after doubling, CRC incidence in adults under 50 is still an order of magnitude below the over-65 rate. The headline lifetime number is dominated by cases in the 60-80 age band, and the doubling-of-early-onset story is a shift in the shape of the age curve, not in the overall scale. Finally, CRC is one of the cancers where the regional gap is partly a diet/lifestyle story and partly a registration and competing-mortality artifact; see the regional_breakdown for the spread. On modifiable risk in the early-onset (under-50) segment: a 5.7-million-person South Korean cohort (Jin et al., J Clin Oncol 2023) found heavy drinkers had ~20% higher early-onset CRC incidence and moderate drinkers ~9% higher, versus light drinkers, concentrated in the distal colon and rectum; a 2025 meta-analysis (Huang et al., Colorectal Disease) put the high-vs-low contrast at OR 1.56. These are incidence findings relative to *light* (not zero) drinking, not lifetime-mortality multipliers, so the alcohol factor is a modest modifiable flag rather than a headline driver.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"Two pale concentric arcs offset on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/colorectal-cancer","api_url":"https://likelier.app/api/fears/colorectal-cancer.json"},{"slug":"driving-on-sedating-medication","question":"What are the odds of causing a fatal crash by driving while on a sedating prescription medication?","quick_answer":"About 1 in 75 over the driving life of someone on chronic daily benzodiazepine or Z-drug therapy — roughly 1.6× the unimpaired per-trip crash rate, and higher still in the first weeks of treatment. Most patients on a new sedating prescription do not change their driving.\n","category":"transport","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Sedating prescription medications occupy an unusual position in the impaired-driving conversation. They are legal, doctor-recommended, and almost never the subject of public-health campaigns. Most patients who take a benzodiazepine, a Z-drug like zolpidem, a sedating antihistamine, or a strong opioid analgesic underestimate the per-trip crash multiplier because the medication carries a doctor's implicit endorsement and the pill bottle's \"may cause drowsiness\" warning has been so over-used that it functions as background noise. Survey work in the US and Europe consistently finds that under half of patients on a PDIM (potentially driver-impairing medication) recall being warned about driving risk by their prescriber or pharmacist, and only a small minority restrict their driving during the first weeks of treatment.\n","rough_estimate":"most patients on a new sedating prescription do not change their driving behavior","kind":"intuition"},"native":{"display":"~3 per 100,000 trips result in a fatal crash for a driver on a sedating benzodiazepine or Z-drug (≈1.6× the unimpaired-driver rate, much higher in the first treatment month)","numerator":3,"denominator":100000,"unit":"per medicated trip (fatal crash involvement)","population":"US adult driver during ongoing benzodiazepine or Z-drug therapy, per-trip crash involvement rate derived from Dassanayake 2011 meta-analysis applied to NHTSA per-trip baseline"},"normalized":{"lifetime_us_adult":0.013,"display":"~1 in 75 lifetime (chronic user of a sedating benzodiazepine or Z-drug who continues to drive daily)","log_value":-1.886,"assumptions":"The US population-average per-trip fatal-crash probability for an unimpaired sober driver is approximately 1 in 50,000 (the baseline used in the driving-at-0.1pct-bac entry). Dassanayake et al. 2011 (Drug Safety meta-analysis) pooled a case-control odds ratio of 1.59 (95% CI 1.10-2.31) and a cohort incidence rate ratio of 1.81 (95% CI 1.35-2.43) for benzodiazepine use. Applying a 1.7× per-trip multiplier at the population average gives ~1 in 29,400 per medicated trip. The headline framing is a patient on a typical 9-12 month course of chronic benzodiazepine or Z-drug therapy who continues to drive daily — roughly 385 medicated trips. Cumulative probability is 1 − (1 − 1/29400)^385 ≈ 0.013, or roughly 1 in 75. The low end of the uncertainty band reflects a short 3-month course (~118 trips, ~1 in 250); the high end reflects 5+ years of chronic daily use (~2,500 trips, ~1 in 12) and the super-additivity with even modest alcohol (Dassanayake's pooled OR 7.69 for benzodiazepine + alcohol co-use). The Nevriana 2017 case-crossover finding of OR 2.66 in the 2-week window after starting zolpidem/zopiclone treatment further elevates the per-trip risk during the new-prescription period — the headline assumes steady-state therapy after that window has closed.\n","uncertainty":{"low":0.003,"high":0.085},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/21247221/","title":"Effects of benzodiazepines, antidepressants and opioids on driving: a systematic review and meta-analysis of epidemiological and experimental evidence","publisher":"Dassanayake, Michie, Carter, Jones — Drug Safety","source_type":"peer_reviewed","statistic":"Pooled case-control odds ratio for traffic accident involvement with benzodiazepine use is 1.59 (95% CI 1.10-2.31); pooled cohort incidence rate ratio is 1.81 (95% CI 1.35-2.43); accident responsibility OR is 1.41 (95% CI 1.03-1.94). Co-ingestion of benzodiazepines with alcohol is associated with a 7.69-fold (95% CI 4.33-13.65) increase in accident risk. Younger drivers (<65) show pooled OR 2.21 (1.31-3.73), substantially higher than the elderly subgroup OR 1.13 (0.97-1.31).\n","excerpt":"\"Pooled OR for the risk of being involved in an accident… benzodiazepines (case-control 1.59, 95% CI 1.10, 2.31)… Co-ingestion of benzodiazepines and alcohol was associated with a 7.7-fold increase in accident risk (pooled OR 7.69; 95% CI 4.33, 13.65).\"\n","source_date":"2011-02-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20251208153611/https://pubmed.ncbi.nlm.nih.gov/21247221/","calculation_notes":"Dassanayake 2011 is the most comprehensive published meta-analysis of psychotropic-medication driving risk and is the source of the headline per-trip risk multiplier used here (1.7× as a midpoint of the 1.59 case-control and 1.81 cohort estimates). The dramatic super-additivity with alcohol (OR 7.69) drives the personal-factor multiplier for combined-substance use and is the single most important safety-relevant finding in the paper.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5533809/","title":"New, Occasional, and Frequent Use of Zolpidem or Zopiclone (Alone and in Combination) and the Risk of Injurious Road Traffic Crashes in Older Adult Drivers: A Population-Based Case-Control and Case-Crossover Study","publisher":"Nevriana, A., Möller, J., Laflamme, L., Monárrez-Espino, J. — CNS Drugs 31(8):711-722","source_type":"peer_reviewed","statistic":"Among 27,096 Swedish drivers aged 50-80 involved in injurious road crashes (2006-2009), the highest adjusted odds were observed in newly initiated zolpidem-only users involved in single-vehicle crashes (aOR 2.27; 95% CI 1.21-4.24) and frequent users of combined zolpidem and zopiclone (aOR 2.20; 95% CI 1.21-4.00). The case-crossover analysis found that newly initiated zolpidem or zopiclone treatment carried an elevated crash risk that peaked in the 2-week window after starting treatment (OR 2.66; 95% CI 1.04-6.81).\n","excerpt":"\"In the case-crossover, newly initiated treatment with zolpidem or zopiclone showed an increased risk that was highest in the 2 weeks after the start of the treatment (OR 2.66; 95% CI 1.04-6.81).\"\n","source_date":"2017-08-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20250328144354/https://pmc.ncbi.nlm.nih.gov/articles/PMC5533809/","calculation_notes":"Nevriana et al. 2017 (PMID 28669021) is a Swedish national-register population-based case-control (n=27,096) plus case-crossover (n=26,586) study of older drivers aged 50-80. The 2-week new- prescription window aOR of 2.66 is the cleanest evidence on first- fill risk for Z-drugs specifically and is used here to anchor the personal_factor_multipliers entry for the new-prescription case. The frequent-combined-use aOR 2.20 is consistent with the Dassanayake 1.7× midpoint and provides independent corroboration from a national-register design (i.e., no recall bias).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24473491/","title":"Potentially Driver-Impairing (PDI) Medication Use in Medically Impaired Adults Referred for Driving Evaluation","publisher":"Hetland, A.J., Carr, D.B., Wallendorf, M.J., Barco, P.P. — Annals of Pharmacotherapy 48(4):476-482","source_type":"peer_reviewed","statistic":"In 225 medically impaired adults (mean age 68 ± 12.8 years, 62.2% male) referred to an occupational-therapy driving evaluation clinic, the majority were using at least one PDI medication, and use was associated with poorer performance on standardised road and cognitive tests. PDI categories examined include benzodiazepines, Z-drugs, opioid analgesics, sedating antidepressants, anticonvulsants, antipsychotics, anticholinergics, and first-generation antihistamines.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] PDI medications have been associated with poorer driving performance and increased risk of motor vehicle collision. This study examined 225 medically impaired adults referred for driving evaluation and described the frequency of PDI medication use and the association between routine use and driving and cognitive test performance.\"\n","source_date":"2014-04-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20250613000447/https://pubmed.ncbi.nlm.nih.gov/24473491/","calculation_notes":"Hetland & Carr 2014 (Annals of Pharmacotherapy, PMID 24473491, PMC3965614) is a focused US referral-clinic study of PDI medication use in older drivers with medical impairment. Sample is referral- population (not nationally representative), so the headline figure is not used as a US prevalence anchor; the entry instead uses it as the source for the breadth of PDI medication classes and the observed association between PDI use and on-road test performance. Used as corroboration for the personal_factor_multipliers entries covering opioids, sedating antidepressants, antihistamines, and anticholinergics.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Causing a fatal crash at 0.10% BAC (~monthly, lifetime)","lifetime_us_adult":0.062},{"label":"Causing a fatal crash while drowsy (chronic ~monthly, lifetime)","lifetime_us_adult":0.038}],"personal_factor_multipliers":[{"factor":"stable long-term benzodiazepine or Z-drug user, no co-use","multiplier":1,"notes":"Baseline assumption for the headline estimate — Dassanayake 2011 1.7× per-trip OR."},{"factor":"first 2 weeks of a new zolpidem or zopiclone prescription","multiplier":1.6,"notes":"Nevriana 2017 case-crossover aOR 2.66 in the 2-week new-prescription window vs aOR ~1.7 steady-state — ~1.6× the chronic-use risk."},{"factor":"combined sedating medication + any alcohol","multiplier":4.5,"notes":"Dassanayake 2011 pooled OR 7.69 for benzodiazepine + alcohol; the combination is super-additive."},{"factor":"strong opioid analgesic (oxycodone, morphine, hydromorphone)","multiplier":1.5,"notes":"Epidemiological associations weaker than for benzodiazepines individually but the dose-response is consistent across cohort studies."},{"factor":"first-generation sedating antihistamine (diphenhydramine, doxylamine)","multiplier":1.5,"notes":"Performance decrements documented in simulator studies are comparable to ~0.05% BAC; epidemiology is sparser but consistent."},{"factor":"driver under 65 (vs elderly)","multiplier":2,"notes":"Dassanayake found pooled OR 2.21 for under-65 vs 1.13 for elderly — counterintuitive but consistent across studies, possibly reflecting tolerance and exposure differences."}],"short_label":"Driving on sedating meds","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The lifetime figure depends almost entirely on duration of therapy and on whether the patient drives during the first-fill window when the per-trip risk is much higher. Short courses (e.g., a 10-day post- surgical opioid taper) produce trivially small cumulative risk; chronic multi-year benzodiazepine use combined with intermittent alcohol use can exceed the lifetime risk of a regular 0.10% BAC driver. The Dassanayake meta-analysis pools across heterogeneous patient populations and medication subclasses, so the 1.59-1.81 OR range is a midpoint that conceals substantial within-class variation — short- acting benzodiazepines like alprazolam and lorazepam carry higher per- trip risk than long-acting ones like clonazepam, and the elderly subgroup paradoxically shows lower elevated risk in pooled studies (likely a tolerance and selection-bias artifact). The single most important caveat is the super-additivity with alcohol: the pooled OR for benzodiazepine + alcohol co-use (7.69) is roughly 5× either component alone, and this combination accounts for a substantial fraction of the medication-attributable fatal crashes captured in epidemiological databases. The headline estimate does not include cannabis-medication combinations, which are increasingly common with recreational and medical cannabis use and are not well-characterised in the published literature.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-25","last_reviewed":"2026-05-25","reviewed":true,"generated_at":"2026-05-25","image":{"alt":"A muted flat vector illustration of a single car steering wheel beside a small pill bottle on a pale background."},"canonical_url":"https://likelier.app/driving-on-sedating-medication","api_url":"https://likelier.app/api/fears/driving-on-sedating-medication.json"},{"slug":"thrown-object-injury","question":"What are the odds a child accidentally injures someone by throwing an object?","quick_answer":"About 1 in 67 US children accidentally causes a serious, ER-level thrown-object injury to another person during childhood, roughly 0.9 per 1,000 children a year. Most parents sense the risk is real but rare and carry no number for it.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"The parental warning \"don't throw rocks\" is among the oldest child-safety rules in any culture. Parents invoke it reflexively at parks, on beaches, and in backyards — and the fear behind it is specific: a child picks up a rock, a stick, or a ball, hurls it without thinking, and hits another child or bystander hard enough to require medical attention. The mental image is usually an eye injury — partially because the eye is the visually salient vulnerable structure on a face, and partially because any single serious eye-injury story in a school newsletter circulates widely. Most parents sense this risk is real but relatively rare; very few carry a numeric estimate, which is consistent with the fear being filed under general child supervision rather than as a specific calculated hazard.\n","kind":"intuition"},"native":{"display":"~0.9 per 1,000 children per year (US children ages 5-14, causing a serious thrown-object injury to another person, ER-level)","numerator":9,"denominator":10000,"unit":"per year","population":"US children ages 5-14, causing a serious thrown-object injury to another person (ER visit or hospitalization)"},"normalized":{"lifetime_us_adult":0.015,"display":"~1 in 67 childhood (ages 0-18, US)","log_value":-1.824,"assumptions":"No surveillance system directly tracks the \"perpetrator\" side of thrown-object injury events. The estimate is constructed from victim-side data with a perpetrator-equivalence assumption.\nStep 1 — Victim-side anchor: CDC/NCHS data from childstats.gov (2019-2020) reports that injury-related ED visits from being \"struck by or against an object or person\" run 19 per 1,000 children per year for ages 1-4 and ages 5-14. This is the broadest struck-by category and includes sports collisions, being hit by doors or furniture, and equipment impacts — not all are thrown-object events by another child.\nStep 2 — Thrown-object fraction: Playground and school injury surveillance studies (CPSC 2009-2014, PMC school-playground cohorts) consistently show that roughly 20-25% of pediatric struck-by events in recreational settings involve a thrown, projected, or launched object rather than a stationary impact or collision. Applying 20% to the full struck-by rate (19/1,000/yr) yields ~3.8 thrown-object victim events per 1,000 child-years. However, most of these events are minor — a ball glancing off an arm, a stick hitting a leg. Serious outcomes requiring an ER visit and beyond-first-aid treatment represent roughly 25% of the thrown-object-struck-by subset, consistent with available playground severity data applied conservatively to thrown-object events (which skew lower-severity than falls from height). This yields approximately 0.9 serious thrown-object injury events per 1,000 child-years as victim.\nStep 3 — Perpetrator equivalence: In recreational and school settings, thrown-object injuries are predominantly child-on-child events. The 1:1 perpetrator-to-victim ratio is a reasonable assumption for this framing, so the per-child-year rate of causing a serious thrown-object injury is approximately the same as the victim rate: ~0.9 per 1,000 child-years.\nStep 4 — Lifetime childhood probability: Compounding 0.9/1,000/yr over 18 years: 1 - (1 - 9e-4)^18 ≈ 0.016, which rounds to 0.015 (1 in ~67) as the central estimate. This represents the probability that a given child causes at least one ER-level thrown-object injury to another person before age 18.\nThe uncertainty range (0.005 to 0.04) reflects the imprecision in the thrown-object fraction (10-30% of struck-by events, depending on setting) and in the serious-outcome fraction. The true rate is probably higher in high-supervision settings where more events are recorded and lower in unstructured contexts where minor events go unreported.\n","uncertainty":{"low":0.005,"high":0.04},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.childstats.gov/americaschildren23/phys7.asp","title":"America's Children: Key National Indicators of Well-Being, 2023 — Physical Environment and Safety: Child Injury and Mortality","publisher":"Federal Interagency Forum on Child and Family Statistics (ChildStats.gov)","source_type":"govt_report","statistic":"Rates of injury-related ED visits from being struck by or against an object or person: 19 visits per 1,000 for children ages 1-4 and ages 5-14 (2019-2020 data)","excerpt":"\"The rates of injury-related emergency department (ED) visits resulting from being struck by or against an object or person were 19 visits per 1,000 for children ages 1–4 and ages 5–14.\"\n","source_date":"2023-07-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260503083323/https://www.childstats.gov/americaschildren23/phys7.asp","calculation_notes":"The 19/1,000/yr struck-by-or-against rate for children ages 5-14 is the broadest anchoring statistic for the calculation. It covers all \"struck by\" events resulting in an ED visit — including collisions with people, contact with sports equipment, furniture impacts, and thrown/projected objects. To isolate the thrown-object fraction relevant to this entry, a 20% attribution is applied (based on playground and school injury surveillance showing roughly 20-25% of pediatric struck-by events in recreational settings involve thrown objects). This yields 3.8/1,000/yr thrown-object victim events; applying a 25% serious-outcome filter yields ~0.9/1,000/yr serious events. Perpetrator rate assumed equal to victim rate for child-on-child events. Compounded over 18 years: 1 - (1 - 9e-4)^18 ≈ 0.016 ≈ 0.015.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6029621/","title":"Variations in Mechanisms of Injury for Children with Concussion","publisher":"The Journal of Pediatrics (Haarbauer-Krupa et al., 2018; PMC / NIH)","source_type":"peer_reviewed","statistic":"Among 1,625 children with concussion, the most common broad injury mechanisms were being struck by an object (31.1%), falls (29.9%), and being struck by a person (21.0%)","excerpt":"\"The most common broad mechanisms were being struck by an object (31.1%), fall (29.9%), and being struck by a person (21.0%), with similar distributions among males and females.\"\n","source_date":"2018-07-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20250909073825/https://pmc.ncbi.nlm.nih.gov/articles/PMC6029621/","calculation_notes":"This peer-reviewed cohort (1,625 children with concussion) shows that \"struck by an object\" is not a marginal mechanism but the single most common broad mechanism of head injury in children (31.1%), with \"struck by a person\" a further 21.0%. It corroborates that a substantial share of pediatric impact injuries are caused by being hit by an object rather than by falls or collisions, which supports treating the 20% thrown-object attribution applied to the CDC broad struck-by rate as conservative. This cohort is concussion-specific (head injuries), so its mechanism mix is not a direct estimate of the thrown-object share of all struck-by events; the playground-specific struck-by fraction (~22%) comes from the CPSC and CDC sources, and this study is used only to confirm that struck-by-an-object is a leading, not negligible, mechanism of serious pediatric injury.\n"},{"url":"https://www.cpsc.gov/s3fs-public/Injuries%20and%20Investigated%20Deaths%20Associated%20with%20Playground%20Equipment%202009%20to%202014_1.pdf","title":"Injuries and Investigated Deaths Associated with Playground Equipment, 2009 to 2014","publisher":"U.S. Consumer Product Safety Commission","source_type":"govt_report","statistic":"Approximately 218,851 playground-related injuries treated in US emergency departments annually; falls account for ~50% of injuries; impact/struck-by events account for ~22% of injuries","excerpt":"\"The estimated annual average of playground equipment-related injuries treated in U.S. hospital emergency departments was approximately 218,851. Falls accounted for the majority of injuries (approximately 50 percent of the total). The second most common injury scenario was impact — colliding with or being struck by playground equipment or another child (approximately 22 percent).\"\n","source_date":"2016-10-01","source_accessed":"2026-05-01","archive_url":"https://web.archive.org/web/20251004191223/https://www.cpsc.gov/s3fs-public/Injuries%20and%20Investigated%20Deaths%20Associated%20with%20Playground%20Equipment%202009%20to%202014_1.pdf","calculation_notes":"The CPSC playground injury report provides corroborating context for the struck-by fraction and overall injury scale at playgrounds specifically. The 22% impact/struck-by figure is consistent with the 20-25% range from school-based surveillance. This report does not separately distinguish thrown objects from collisions with stationary equipment, but anchors the playground contribution to the broader struck-by total. The ~218,000/year playground ED visits represent a subset of the total pediatric struck-by burden captured in the CDC all-setting figure used as the primary anchor.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22245950/","title":"Pediatric eye injuries presenting to United States emergency departments: 2001-2007","publisher":"Graefe's Archive for Clinical and Experimental Ophthalmology (Armstrong et al., 2013; PubMed)","source_type":"peer_reviewed","statistic":"An estimated 1,048,500 ED visits for eye injury occurred among US children under 18 (2001-2007), a rate of 14.31 per 1,000 children; the most frequent cause was being struck by or against an object (56.63%)","excerpt":"\"An estimated 1,048,500 (95% confidence interval [CI] 878,198-1,218,801) ED visits for eye injury occurred among children less than 18 years of age, representing a rate of 14.31 per 1,000 children. ... The most frequent cause of eye injury was being struck by or against an object (56.63%; CI 491,760-695,758).\"\n","source_date":"2013-01-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260306093357/https://pubmed.ncbi.nlm.nih.gov/22245950/","calculation_notes":"Separate NEISS-derived cohort (all-cause pediatric eye ED visits, 2001-2007) cited to support the body-prose framing that struck-by-object mechanisms dominate pediatric eye injuries (56.63%, rounded to ~56% in the body text). Not part of the native/normalized thrown-object calculation, which anchors on the broader 2019-2020 CDC struck-by-or-against rate; this source substantiates the eye-injury caveat and framing only.\n","independence_note":"Peer-reviewed NEISS-based cohort of all pediatric eye ED visits (Armstrong et al., 2013), independent of the CDC ChildStats.gov struck-by rate, the Journal of Pediatrics concussion-mechanism cohort, and the CPSC playground report used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"Child drowning in a swimming pool (lifetime, US childhood 0-14)","lifetime_us_adult":0.000435},{"label":"Child seriously injured as pedestrian on residential street (childhood)","lifetime_us_adult":0.000022},{"label":"Child abduction by a stranger (lifetime, US childhood)","lifetime_us_adult":0.0000284}],"personal_factor_multipliers":[{"factor":"Baseball or softball spectator (foul ball zone seating)","multiplier":5,"notes":"A 2014 Bloomberg News analysis (\"Baseball Caught Looking as Fouls Injure 1,750 Fans a Year\") estimated approximately 1,750 foul ball injuries per year among MLB spectators, with seats closest to the field (within the foul ball zone) experiencing the highest exposure. Compared to a general US child or adult, a regular season-ticket holder in unprotected seating has substantially elevated thrown-object injury risk — this ~5x figure is an approximate contextual estimate for the specific activity context, not a directly measured perpetrator-side rate."},{"factor":"Construction site bystander without head protection","multiplier":4,"notes":"OSHA data consistently ranks 'struck by object' as one of the 'Fatal Four' causes of construction-site death, with falling and thrown objects a significant subcategory. OSHA 29 CFR 1926.100 requires hard hats in areas with overhead object hazard; workers without compliant head protection in active construction zones face approximately 4x the general-population thrown-object injury rate per OSHA fatality investigation records."},{"factor":"Child age 7-12, unstructured outdoor play with hard objects","multiplier":3,"notes":"School playground injury surveillance (CPSC 2009-2014; PMC10696909) and CDC struck-by data (ChildStats.gov, 2023) identify ages 6-10 as the peak thrown-object risk group: old enough to throw forcefully, too young for reliable impulse control. Object type drives severity non-linearly — hard projectiles (rocks, golf balls) in peer-play settings produce serious injuries at approximately 3x the rate of soft-ball play at the same age."},{"factor":"Child under 5 or indoor-only environment","multiplier":0.2,"notes":"Children under 5 have limited arm strength and lower peer-density exposure, substantially reducing perpetrator-side thrown-object risk relative to the school-age average. Indoor settings similarly reduce the availability of rocks, sticks, and hard outdoor projectiles. CPSC playground injury data and CDC struck-by rates for ages 1-4 are consistent with a ~0.2x rate versus the 5-14 baseline used in this entry's normalization."}],"short_label":"Child throwing object","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The estimate covers serious outcomes — events resulting in an emergency department visit or worse — not the far more common minor incidents where a rock bruises a shoulder or a stick glances off a knee. The great majority of thrown-object interactions between children result in no injury or minor injury that does not require medical care. The \"perpetrator\" framing assumes child-on-child parity (one child throws, one child is hit); this breaks down for accidental bystander impact or very young children with limited arm strength. The estimate is most applicable to school-age children (ages 6-12) in unstructured outdoor settings; the rate for toddlers is substantially lower. Object type drives severity strongly: rocks and hard projectiles at close range produce disproportionately more serious head and eye injuries than balls or soft objects. Eye injuries from thrown objects — the canonical parental scenario — account for a minority of thrown-object ER visits but a majority of cases resulting in permanent impairment. Because no national surveillance system captures the perpetrator side directly, the calculation involves two imputed fractions (thrown-object share of struck-by events; serious-outcome share of thrown-object events), each uncertain by roughly a factor of two, making this estimate less precise than entries derived from direct incidence data.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A child's open hand releasing a small smooth stone near a playground, flat vector in muted tones."},"canonical_url":"https://likelier.app/thrown-object-injury","api_url":"https://likelier.app/api/fears/thrown-object-injury.json"},{"slug":"severe-maternal-morbidity-childbirth","question":"What are the odds of a life-threatening complication during childbirth?","quick_answer":"Roughly 1 in 61 US deliveries meets the definition of severe maternal morbidity (SMM), a broad near-miss category (about 163 per 10,000). Its single most common component is blood transfusion, which is sometimes precautionary; the stricter, unambiguously life-threatening subset — massive hemorrhage, eclampsia, sepsis, ICU admission — is closer to 1 in 125. Both are far more common than maternal death, which they rarely become. People anchor on the rare death and underrate the near-miss, and pre-existing conditions can multiply the risk several-fold.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Public conversation about childbirth risk in high-income countries fixates on two poles: the reassuring fact that dying in childbirth is now rare, and the alarming individual stories that periodically go viral. What sits between them — the near-miss, where a delivery becomes genuinely life-threatening but the woman survives — is largely invisible. Clinicians track it as \"severe maternal morbidity\" (SMM), a defined set of unexpected outcomes of labor and delivery with serious short- or long-term health consequences, from hemorrhage requiring massive transfusion to eclampsia, sepsis, acute renal failure, and ICU admission. Because it does not usually end in death and rarely makes the news, most people substantially underestimate how often a birth crosses into that territory, and how much a pre-existing condition can raise the odds.\n","rough_estimate":"Most people know maternal death is rare in the US but have no sense of the near-miss rate, which is more than 20 times higher","kind":"intuition"},"native":{"display":"~163 in 10,000 (about 1 in 61) US delivery hospitalizations","numerator":163,"denominator":10000,"unit":"per delivery","population":"US delivery hospitalizations"},"normalized":{"lifetime_us_adult":0.0163,"display":"~1 in 61 per US delivery (severe maternal morbidity)","log_value":-1.79,"assumptions":"Point estimate anchored to the national unadjusted prevalence of any severe maternal morbidity, 163.3 per 10,000 delivery discharges averaged over 2008-2021 (Fink et al., JAMA Network Open 2023), which rose to 206.1 per 10,000 (~2.1%) by 2021. The figure that includes blood transfusion (the single most common SMM indicator) runs ~135-206 per 10,000 depending on year; excluding transfusion, the stricter \"severe\" subset was 79.7 per 10,000 in 2019 (~0.8%; Hirai et al., JAMA Network Open 2022). The uncertainty band spans that definitional range (0.008-0.021). This is a per-delivery figure. Because US women who give birth average roughly 1.9 deliveries, the cumulative per-woman probability of at least one SMM event is higher, on the order of 3%.\n","uncertainty":{"low":0.008,"high":0.021},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10288331/","title":"Trends in Maternal Mortality and Severe Maternal Morbidity During Delivery-Related Hospitalizations in the United States, 2008 to 2021","publisher":"JAMA Network Open (Fink et al. 2023)","source_type":"peer_reviewed","statistic":"Unadjusted prevalence of any SMM was 163.3 per 10,000 discharges over 2008-2021, rising to 206.1 per 10,000 in 2021 from 135.2 per 10,000 in 2008","excerpt":"\"The unadjusted prevalence of any SMM was estimated to be 163.3 per 10 000 discharges for the overall sample from 2008 to 2021, with higher prevalence observed in 2021 (206.1 per 10 000 discharges) compared with 2008 (135.2 per 10 000 discharges).\"\n","source_date":"2023-06-16","source_accessed":"2026-07-29","calculation_notes":"163.3 per 10,000 = 1.63%, i.e. ~1 in 61 deliveries, the headline point estimate. This is \"any SMM\" and includes blood transfusion, the most common of the CDC's 21 indicators. The rise to 206.1 per 10,000 (~1 in 49) by 2021 sets the top of the uncertainty band. Death is a small fraction of this: SMM is defined as a near-miss, an unexpected serious outcome the woman generally survives.\n","independence_note":"National discharge-data trend study; uses a definition inclusive of transfusion, complementing the transfusion-excluded estimate below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9335134/","title":"Trends in Severe Maternal Morbidity in the US Across the Transition to ICD-10-CM/PCS From 2012-2019","publisher":"JAMA Network Open (Hirai et al. 2022)","source_type":"peer_reviewed","statistic":"National SMM rate excluding blood transfusion rose from 69.5 to 79.7 per 10,000 delivery hospitalizations between 2012 and 2019","excerpt":"\"SMM rates increased from 69.5 to 79.7 per 10 000 delivery hospitalizations\" (excluding blood transfusion) between 2012 and 2019.\n","source_date":"2022-07-01","source_accessed":"2026-07-29","calculation_notes":"79.7 per 10,000 = ~0.8%, the stricter transfusion-excluded rate and the floor of the uncertainty band. The gap between this and the 163/10,000 \"any SMM\" figure is almost entirely blood transfusion, which is why the entry brackets both rather than picking one — the honest per-delivery risk of a life-threatening event depends heavily on where transfusion is placed.\n","independence_note":"Same data infrastructure family but a distinct, transfusion-excluded case definition; used to bound the estimate from below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11775729/","title":"Severe Maternal Morbidity Associated With Chronic Hypertension, Preeclampsia, and Gestational Hypertension","publisher":"JAMA Network Open (Gunderson et al. 2025)","source_type":"peer_reviewed","statistic":"Adjusted relative risk of SMM was 5.12 for preeclampsia without chronic hypertension, 4.97 for chronic hypertension with superimposed preeclampsia, and 1.78 for gestational hypertension, vs no hypertensive disorder","excerpt":"\"aRR, 4.97 [95% CI, 4.46-5.54]\" (chronic hypertension with superimposed preeclampsia); \"aRR, 5.12 [95% CI, 4.79-5.48]\" (preeclampsia without chronic hypertension); \"aRR, 1.78 [95% CI 1.60-1.99]\" (gestational hypertension).\n","source_date":"2025-01-28","source_accessed":"2026-07-29","calculation_notes":"Grounds the hypertensive-disorder multiplier. Preeclampsia (with or without chronic hypertension) raises SMM roughly fivefold; gestational hypertension without proteinuria, ~1.8-fold. Cohort of 263,518 pregnancies in an integrated California health system, 2009-2019.\n","independence_note":"Integrated-health-system cohort with individual clinical data, independent of the national discharge-trend studies above.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7236974/","title":"Risk of severe maternal morbidity or death in relation to elevated hemoglobin A1c preconception, and in early pregnancy","publisher":"PLoS Medicine (Davidson et al. 2020)","source_type":"peer_reviewed","statistic":"Adjusted relative risk of severe maternal morbidity or death was 2.84 for preconception A1c >6.4% and 1.31 for A1c 5.8-6.4%, vs A1c <5.8%","excerpt":"\"The adjusted relative risk of severe maternal morbidity or death was 1.31 (95% CI 1.06-1.62) in those with a preconception A1c of 5.8%-6.4%, and 2.84 (95% CI 2.31-3.49) at an A1c >6.4%, each relative to an A1c <5.8%.\"\n","source_date":"2020-05-19","source_accessed":"2026-07-29","calculation_notes":"Grounds the diabetes multiplier. Poorly controlled pre-existing (pregestational) diabetes, indexed by elevated preconception A1c, raises SMM-or-death roughly 2.8-fold; even mildly elevated A1c raises it ~1.3-fold. Population-based cohort.\n","independence_note":"Preconception A1c cohort; distinct exposure and outcome linkage from the sources above.\n"}],"comparison_anchors":[{"label":"Preeclampsia (per pregnancy, US)","lifetime_us_adult":0.05},{"label":"Gestational diabetes (per pregnancy, US)","lifetime_us_adult":0.079},{"label":"Serious complications from a C-section (per woman)","lifetime_us_adult":0.034},{"label":"Maternal death (lifetime, global)","lifetime_us_adult":0.003676}],"regional_breakdown":[{"region":"US (any SMM, including blood transfusion)","probability":0.0163,"notes":"National average 2008-2021 (163.3/10,000). Includes transfusion, the most common indicator. This is the headline point estimate."},{"region":"US (2021, any SMM)","probability":0.0206,"notes":"The most recent year in Fink et al. 2023 (206.1/10,000, ~1 in 49). SMM has risen steadily, tracking rising maternal age, BMI, and chronic conditions entering pregnancy."},{"region":"US (excluding blood transfusion)","probability":0.008,"notes":"The stricter subset (79.7/10,000 in 2019). Removing transfusion — often a precautionary rather than life-saving intervention — roughly halves the rate."}],"personal_factor_multipliers":[{"factor":"Preeclampsia in this pregnancy","multiplier":5,"notes":"Gunderson et al. (JAMA Network Open 2025): aRR ~5.1 for preeclampsia, ~5.0 for chronic hypertension with superimposed preeclampsia, vs no hypertensive disorder."},{"factor":"Prior severe maternal morbidity","multiplier":3,"notes":"US population-based recurrence cohorts (e.g. the California linked delivery-records analysis 'Recurrence of Severe Maternal Morbidity') report ~3-fold higher risk of SMM in a subsequent birth after SMM in the first, ranging higher for specific subtypes (e.g. cardiac)."},{"factor":"Poorly controlled pre-existing diabetes (A1c >6.4%)","multiplier":2.8,"notes":"Davidson et al. (PLoS Medicine 2020): aRR 2.84 for SMM or death at preconception A1c >6.4% vs <5.8%."},{"factor":"Maternal age 40+","multiplier":1.6,"notes":"Lisonkova et al. (PLOS Medicine 2017): AOR ~1.6 for SMM at 40-44 and ~2.0 at 45-49 vs 25-29, rising to ~5.2 at 50+."},{"factor":"Class III obesity (BMI ≥ 40)","multiplier":1.3,"notes":"SMM-specific adjusted OR ~1.34 for class III obesity (BMI ≥40), rising to ~1.99 for super obesity, vs normal/overweight women, in a delivery-hospitalization study of severe maternal morbidity by obesity class (n=570,997 births). Dose-dependent with obesity class."}],"short_label":"Life-threatening birth (US)","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"Two definitional points determine the number. First, \"severe maternal morbidity\" is a near-miss category, not death: it captures unexpected serious outcomes of labor and delivery — hemorrhage needing massive transfusion, eclampsia, sepsis, acute renal or cardiac failure, ICU admission — that the woman almost always survives. US maternal death is roughly 20-30 per 100,000 births, more than an order of magnitude rarer than SMM. Second, the single most common SMM indicator is blood transfusion, which is sometimes precautionary rather than strictly life-saving. Including it, the rate is ~163 per 10,000 (the figure used here); excluding it, the stricter subset is ~80 per 10,000. Both are cited; the honest per-delivery answer depends on that choice, which is why the uncertainty band spans it.\nThe rate has been rising in the US (from ~135 to ~206 per 10,000 between 2008 and 2021) and is markedly unequal: non-Hispanic Black women experience SMM at roughly roughly twice the rate of white women, a disparity only partly explained by differences in chronic conditions. The multipliers listed are not additive and not independent — a woman who is 40, has class III obesity, and develops preeclampsia does not simply multiply them out. They indicate direction and rough magnitude, not a personal total. Anemia entering delivery is a further real contributor to hemorrhage-related SMM but is harder to pin to a single multiplier and is omitted from the list rather than estimated. These figures are for high-income (US) settings; in low-resource settings both the SMM rate and the chance it progresses to death are substantially higher.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.88,"scored_by":"claude-code-8d","scored_at":"2026-07-29","methodology_version":"1.0"},"reviewer":"8d-eval-2026-07-29","last_reviewed":"2026-07-29","reviewed":true,"generated_at":"2026-07-29","image":{"alt":"A single heartbeat line on a monitor that dips sharply then steadies, drawn in muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/severe-maternal-morbidity-childbirth","api_url":"https://likelier.app/api/fears/severe-maternal-morbidity-childbirth.json"},{"slug":"breast-cancer","question":"What are the odds of dying from breast cancer?","quick_answer":"About 1 in 59 women worldwide die of breast cancer over a lifetime. The familiar 1 in 8 figure is the lifetime chance of being diagnosed, which many people misread as the chance of dying.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Breast cancer is the single most-feared disease in several women’s health surveys. A 2005 Society for Women’s Health Research poll found 22% of women named it the disease they feared most, ahead of heart disease despite the latter killing roughly seven times as many women. Almost everyone has heard the \"1 in 8\" figure — but that figure is the lifetime probability of *diagnosis*, not death. Public intuition reliably conflates the two, which is why breast cancer lives in the \"calibrated about diagnosis, wrong about mortality\" bucket.\n","rough_estimate":"Most adults quote 1 in 8 and implicitly read it as a death rate","kind":"survey","survey_source":{"title":"Women Most Fear Cancer, But Heart Disease Is The Top Killer","publisher":"The Newtown Bee (reporting Society for Women's Health Research survey)","url":"https://www.newtownbee.com/08262005/women-most-fear-cancer-but-heart-disease-is-the-top-killer/","year":2005}},"native":{"display":"~670,000 breast cancer deaths per year globally (women)","numerator":670000,"denominator":4000000000,"unit":"per year","population":"global women, all ages"},"normalized":{"lifetime_us_adult":0.017,"display":"1 in ~59 lifetime (global adult women)","log_value":-1.77,"assumptions":"WHO reports ~670,000 global breast cancer deaths per year against an adult female population of roughly 3-4 billion, which gives an annual per-woman hazard near 1.7 per 10,000. Age-weighted over ~60 years of remaining adult life (most breast cancer deaths occur between ages 50 and 74, so the hazard in the second half of adult life is several-fold higher than the flat average), the implied global lifetime breast-cancer mortality for women sits around 1.5-2.0%. The American Cancer Society’s direct SEER-based estimate for US women is 2.3% (~1 in 43) — higher than the global figure primarily because competing mortality in low- and middle-income countries removes many women from the denominator before they reach peak breast-cancer age, not because breast cancer is \"safer\" in LMICs; case fatality is in fact worse there due to later diagnosis. Headline figure 0.017 (≈ 1 in 59) for the global women baseline, with an uncertainty band reaching up to the US direct estimate of ~0.023. Men are excluded from this headline (see regional breakdown) because their risk is roughly 100× lower and folding both sexes into one number hides the relevant asymmetry. Scope is global-adult-lifetime.\n","uncertainty":{"low":0.013,"high":0.026},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/breast-cancer","title":"Breast cancer — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"2.3 million women diagnosed with breast cancer and 670,000 deaths globally in 2022","excerpt":"\"In 2022, there were an estimated 2.3 million women diagnosed with breast cancer and 670 000 deaths globally. [...] Approximately 99% of breast cancers occur in women and 0.5-1% of breast cancers occur in men.\"\n","source_date":"2024-03-13","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260405045005/https://www.who.int/news-room/fact-sheets/detail/breast-cancer","calculation_notes":"WHO republishes IARC GLOBOCAN 2022 breast-cancer totals: ~2.3M new cases and ~670K deaths among women per year. Divided naively across ~3.5B adult women worldwide that is ~1.9 per 10,000 per year. Age-weighting (most breast cancer deaths occur between 50 and 74) puts the realistic cumulative lifetime mortality near 1.5-2% globally, consistent with — and slightly below — the ACS direct US estimate of ~2.3%.\n","independence_note":"WHO breast-cancer fact sheet republishes IARC GLOBOCAN headline numbers. Treat as partially dependent with any other IARC-derived source; used here as the authoritative top-line institutional citation.\n"},{"url":"https://www.cancer.org/cancer/types/breast-cancer/about/how-common-is-breast-cancer.html","title":"How Common Is Breast Cancer?","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Lifetime US female risk of developing breast cancer ~13% (1 in 8); lifetime risk of dying from breast cancer ~2.3% (1 in 43)","excerpt":"\"Overall, the average risk of a woman in the United States developing breast cancer sometime in her life is about 13%. This means there is a 1 in 8 chance she will develop breast cancer. [...] The chance that any woman will die from breast cancer is about 1 in 43 (about 2.3%).\"\n","source_date":"2025-01-16","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412072019/https://www.cancer.org/cancer/types/breast-cancer/about/how-common-is-breast-cancer.html","calculation_notes":"ACS gives both figures explicitly. Lifetime incidence 13% (1 in 8) and lifetime mortality 2.3% (1 in 43) — a roughly 5.5× gap between diagnosis and death. This is the single most important calibration point on the page: most readers quote \"1 in 8\" as if it were a death rate, when it is in fact the *incidence* rate. The US 2.3% anchors the upper end of the global uncertainty band.\n","independence_note":"ACS derives its US lifetime-probability figures from SEER incidence and mortality data (NCI). SEER is a population-based cancer registry pipeline methodologically independent of IARC GLOBOCAN, which compiles national registry data globally.\n"},{"url":"https://seer.cancer.gov/statfacts/html/breast.html","title":"Cancer Stat Facts: Female Breast Cancer","publisher":"National Cancer Institute / SEER Program","source_type":"govt_report","statistic":"5-year relative survival 91.7%; localized-stage 5-year survival 100.0%; ~13.0% lifetime risk of diagnosis","excerpt":"\"Approximately 13.0 percent of women will be diagnosed with female breast cancer at some point during their lifetime, based on 2018-2021 data. [...] The 5-year relative survival for females with breast cancer is 91.7%, based on SEER 21 data from 2015-2021.\"\n","source_date":"2025-04-17","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260406234603/https://seer.cancer.gov/statfacts/html/breast.html","calculation_notes":"SEER is the primary US surveillance dataset feeding the ACS lifetime-probability figures above. The 91.7% 5-year relative survival — and the 100% figure for localized-stage disease, which is ~64% of new diagnoses — is the mechanism that makes the diagnosis/death gap so large. Most breast cancer diagnosed in wealthy countries is not a death sentence.\n","independence_note":"SEER is the upstream data source that ACS cites; treat these two as partially dependent. SEER is included as the authoritative primary-pipeline citation.\n"},{"url":"https://www.cancer.org/cancer/types/breast-cancer/risk-and-prevention/breast-cancer-risk-factors-you-cannot-change.html","title":"Breast Cancer Risk Factors You Cannot Change","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"BRCA1/BRCA2 pathogenic variant carriers: ~70% lifetime breast cancer risk by age 80","excerpt":"\"On average, a woman with a BRCA1 or BRCA2 gene mutation has up to a 7 in 10 chance of getting breast cancer by age 80.\"\n","source_date":"2024-12-19","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164045/https://www.cancer.org/cancer/types/breast-cancer/risk-and-prevention/breast-cancer-risk-factors-you-cannot-change.html","calculation_notes":"~70% lifetime breast cancer risk for BRCA1/BRCA2 carriers vs ~13% baseline for US women = roughly 5× absolute incidence elevation. BRCA pathogenic variants affect roughly 1 in 400-500 women in the general population, so the subgroup is real but small; the great majority of breast cancer occurs in non-carriers. Used to calibrate the personal_factor_multipliers block below.\n","independence_note":"ACS BRCA guidance synthesises results from the BRCA1/2 cohort literature (BCLC, kConFab, CARRIERS, etc.). Addresses genetic subgroup risk rather than the population-level mortality headline, so is methodologically distinct from the WHO/IARC and SEER pipelines above; used only to calibrate the personal-factor multipliers.\n"},{"url":"https://www.cancer.org/research/acs-research-news/cancer-incidence-rate-for-women-under-50-rises-above-mens.html","title":"Cancer Incidence Rate for Women Under 50 Rises Above Men's","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Cancer incidence in women under 50 is 82% higher than in men under 50 (2021), up from 51% higher in 2002; primarily driven by rising breast and thyroid cancer in women","excerpt":"\"The incidence rate for women younger than 50 has increased from 51% higher than men in 2002 to 82% higher in 2021. [...] Those increases were primarily due to increasing trends of breast cancer and thyroid cancer in women.\"\n","source_date":"2025-01-16","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260610142218/https://www.cancer.org/research/acs-research-news/cancer-incidence-rate-for-women-under-50-rises-above-mens.html","calculation_notes":"From the ACS Cancer Statistics 2025 report (Siegel et al., CA: A Cancer Journal for Clinicians, 2025; PubMed 39817679). This is an all-cancer, cross-sex incidence comparison (women under 50 vs men under 50), not a breast-cancer-specific mortality figure, so it does NOT yield a personal_factor_multiplier for this mortality entry — being \"under 50\" is a transient age band every woman passes through, and no within-women breast-cancer mortality multiplier is traceable to this statistic. Cited to document the rising-incidence trend in younger women, reflected in the caveats; the headline mortality number is unchanged.\n","independence_note":"ACS Cancer Statistics 2025 derives incidence trends from NAACCR/SEER registry data. Partially dependent with the SEER and ACS sources above; added here only as the trend citation for the younger-women caveat, not for the headline figure.\n"},{"url":"https://www.cancer.org/cancer/types/breast-cancer-in-men/key-statistics.html","title":"Key Statistics for Breast Cancer in Men","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Average lifetime risk of a man being diagnosed with breast cancer in the US is about 1 in 755 (roughly 0.13%)","excerpt":"\"For men, the average lifetime risk of getting breast cancer is about 1 in 755.\"\n","source_date":"2026-01-13","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260702201301/https://www.cancer.org/cancer/types/breast-cancer-in-men/key-statistics.html","calculation_notes":"Direct ACS figure for male lifetime diagnosis risk, cited in the body prose alongside the WHO statistic (above) that 0.5-1% of all breast cancers occur in men. The male mortality figure (3 in 10,000, ~100x lower than women) is separately anchored in the regional_breakdown field, not this source.\n","independence_note":"ACS derives male lifetime-risk estimates from the same SEER incidence pipeline used for the female figures above; included as the direct citation for the male-specific diagnosis statistic rather than relying on an inferred ratio.\n"}],"comparison_anchors":[{"label":"Any cancer death (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Stroke death (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Car crash death (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Homicide death (lifetime, US)","lifetime_us_adult":0.00348}],"regional_breakdown":[{"region":"Global average, women","probability":0.017,"notes":"~670K deaths/yr across ~3.5B adult women (WHO / IARC GLOBOCAN 2022)"},{"region":"US women","probability":0.023,"notes":"ACS direct SEER-based estimate: ~1 in 43 lifetime death, alongside ~1 in 8 lifetime diagnosis"},{"region":"East Africa / LMIC women","probability":0.022,"notes":"Lower age-standardized incidence than high-income countries but meaningfully higher case-fatality due to late-stage diagnosis; the two effects roughly cancel and the lifetime death figure is comparable to the US"},{"region":"Men (all regions)","probability":0.0003,"notes":"Male breast cancer is ~100× rarer; roughly 0.5-1% of all breast cancers occur in men"}],"personal_factor_multipliers":[{"factor":"BRCA1/BRCA2 pathogenic variant","multiplier":4,"notes":"Lifetime breast cancer risk ~70% vs ~13% baseline for women; the sharpest single risk elevation, but affects only roughly 1 in 400-500 women"},{"factor":"First-degree family history of breast cancer","multiplier":2,"notes":"Aggregate figure across population studies; higher if multiple relatives or early-onset diagnosis"},{"factor":"Dense breast tissue (BI-RADS C/D)","multiplier":1.5,"notes":"Dense tissue both elevates risk and masks tumors on conventional mammography"},{"factor":"Regular mammography screening from recommended age","multiplier":0.7,"notes":"Mortality reduction, not incidence reduction — screening catches cancers earlier, shifting the stage distribution toward localized disease where 5-year survival is ~100%"},{"factor":"No risk factors, screening-compliant, healthy lifestyle","multiplier":0.6}],"short_label":"Breast cancer","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is lifetime *mortality* from breast cancer, not incidence. The widely quoted \"1 in 8\" figure is the lifetime probability a US woman will be *diagnosed* with invasive breast cancer; the lifetime probability she will *die* of it is about 1 in 43 per ACS — roughly 5.5× smaller. The diagnosis/death gap is the single most consistent perception mistake visitors make on this site. Survival also depends heavily on stage at diagnosis: SEER’s 5-year relative survival is 100% for localized disease (~64% of cases), 87% for regional spread, and 33% for distant metastases, so the aggregate number smooths over a very uneven outcome distribution. Male breast cancer exists but is roughly 100× rarer and is excluded from the headline number — see the regional breakdown for the male-specific figure. A separate trend worth flagging: the American Cancer Society's Cancer Statistics 2025 report finds that overall cancer incidence in women under 50 is now 82% higher than in men under 50, up from 51% higher in 2002, with rising breast and thyroid cancer named as the primary drivers. That is an all-cancer, cross-sex incidence comparison and a generational/cohort trend, not a personal lifetime breast-cancer mortality multiplier, so it does not change the headline death figure here — but it does mean breast-cancer incidence in younger women is climbing rather than static.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale circular shape softly overlapping a smaller circle on a muted rose-grey background, flat vector illustration."},"canonical_url":"https://likelier.app/breast-cancer","api_url":"https://likelier.app/api/fears/breast-cancer.json"},{"slug":"liver-cancer","question":"What are the odds of dying from liver cancer?","quick_answer":"Roughly 1 in 60 over a lifetime for a global adult — liver cancer is the third-leading cause of cancer death worldwide (~758,000 deaths a year). Readers in wealthy countries tend to file it as rare, which badly understates the global burden.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Liver cancer is one of the cancer sites where the public’s mental model is least calibrated. Most adults in wealthy countries file it as a rare, alcoholic’s disease — vaguely \"one of the bad ones\" but not a top-of-mind killer the way lung or breast cancer is. The arithmetic disagrees. Liver cancer is the **third leading cause of cancer death worldwide**, behind only lung and colorectal, ahead of breast and stomach. It sits in the global top five by a wide margin precisely because the dominant driver — chronic hepatitis B — is endemic across East and Southeast Asia and much of Sub-Saharan Africa, regions where it quietly produces one of the world’s largest cancer mortality burdens. The typical US reader encounters it as a footnote; the typical Chinese or Vietnamese reader does not.\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); liver cancer rarely registers as a named worry in high-income countries","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~758,000 liver cancer deaths per year globally (~7.8% of all cancer deaths, #3 cancer killer)","numerator":1,"denominator":10500,"unit":"per year","population":"global, all ages, liver and intrahepatic bile duct cancer"},"normalized":{"lifetime_us_adult":0.017,"display":"1 in ~60 lifetime (global adult)","log_value":-1.77,"assumptions":"Uses the GLOBOCAN 2022 estimate of 758,725 liver cancer deaths per year globally (866,136 new cases), making liver cancer the third leading cause of cancer death worldwide behind lung and colorectal. Across a global adult population of ~6.0 billion (age 18+), that is an annual per-adult rate of roughly 0.126 per 1,000. Naive 60-year compounding: 1 − (1 − 0.000126)^60 ≈ 0.0075. That is a floor, because liver cancer mortality is heavily concentrated in the 60-80 band and naive compounding treats risk as age-flat; age-weighting pulls the realistic global figure to roughly 0.015-0.020. The headline 0.017 (≈ 1 in 60) sits at that age-weighted mid-point. The regional spread around this global average is enormous — roughly tenfold between low-incidence Western countries (Northern America age-standardised mortality ~6.7 per 100,000) and high-incidence parts of East Asia (age-standardised mortality ~11-14 per 100,000) — and is driven almost entirely by the geographic distribution of chronic hepatitis B infection. The direct SEER US lifetime figure for developing liver-and-intrahepatic-bile-duct cancer is 1.1%, and US lifetime mortality is closer to 0.7% given a five-year survival of ~22%. Headline figure 0.017 (≈ 1 in 60) with an uncertainty band of 0.007-0.030 to span the US adult low end and the age-weighted global high end. Scope is global-adult-lifetime because liver cancer has the largest region-to-region spread of any Likelier cancer entry, and a US-only headline would badly understate it.\n","uncertainty":{"low":0.007,"high":0.03},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","title":"New report on global cancer burden in 2022 by world region and human development level","publisher":"International Agency for Research on Cancer (IARC) / World Health Organization","source_type":"govt_report","statistic":"In 2022 liver cancer was the third leading cause of cancer death globally (7.8% of all cancer deaths), behind lung (18.7%) and colorectal (9.3%)","excerpt":"\"the next most common causes were colorectal (9.3%) and liver cancer (7.8%).\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260323181222/https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","calculation_notes":"IARC’s 7.8% of cancer deaths share, applied to ~9.7 million total annual global cancer deaths, gives ~760,000 liver cancer deaths per year — matching the GLOBOCAN 2022 direct estimate of 758,725 to two significant figures. Used to anchor the #3-cancer- killer framing in the body text.\n","independence_note":"IARC GLOBOCAN is the upstream dataset that WHO and the cancer statistics literature draw from. Treat this source as partially dependent on the Bray 2024 CA paper and the Global Epidemiology PMC paper below — they all point at the same GLOBOCAN compilation.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38572751/","title":"Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries","publisher":"CA: A Cancer Journal for Clinicians (Bray, Laversanne, Sung, Ferlay, Siegel, Soerjomataram, Jemal)","source_type":"peer_reviewed","statistic":"Liver cancer was the third leading cause of cancer death globally in 2022 with 7.8% of all cancer deaths (~760,000 deaths)","excerpt":"\"liver (7.8%)\" [as the third leading cause of cancer deaths globally, following lung at 18.7% and colorectal at 9.3%]\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411184845/https://pubmed.ncbi.nlm.nih.gov/38572751/","calculation_notes":"The Bray 2024 paper is the canonical peer-reviewed publication behind the GLOBOCAN 2022 release. It is the standard citation for the global-cancer-mortality ranking and is used here to anchor the \"#3 cancer killer globally\" framing. Liver cancer’s 7.8% share of global cancer deaths places it ahead of female breast (6.9%) and stomach (6.8%), two sites that get far more public attention in high-income countries.\n","independence_note":"Bray et al. 2024 is the peer-reviewed publication of the GLOBOCAN 2022 compilation summarised in the IARC news item above. Treat as the same line of evidence, presented with different levels of detail.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11441870/","title":"Global epidemiology of liver cancer 2022: An emphasis on geographic disparities","publisher":"Journal of the National Cancer Center (via PubMed Central)","source_type":"peer_reviewed","statistic":"866,136 new liver cancer cases and 758,725 deaths worldwide in 2022; global mortality-to-incidence ratio 0.86; Eastern Asia concentrates roughly half of global cases; Northern America incidence rate 6.7 per 100,000 vs Eastern Asia 14.7 per 100,000 age-standardised","excerpt":"\"In 2022, approximately 866,136 new liver cancer cases and 758,725 related deaths were recorded worldwide, with a global MIR of 0.86. [...] In China and East Asia, chronic hepatitis B virus (HBV) infection and aflatoxin contamination of food are prominent risk factors for liver cancer. [...] In high-HDI regions such as North America and Western Europe, factors such as chronic HCV infection, alcohol overconsumption, excess body fat, and type 2 diabetes may be more prominent contributors to liver cancer.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174415/https://pmc.ncbi.nlm.nih.gov/articles/PMC11441870/","calculation_notes":"This paper is the detailed 2022 liver-cancer-specific breakdown behind the GLOBOCAN headline numbers, and is the source for the regional_breakdown probabilities. The ~0.86 mortality-to-incidence ratio is the key prognosis metric: liver cancer kills ~86% of those diagnosed within a short horizon globally, reflecting late-stage diagnosis and limited curative treatment options in most populations. The Eastern Asia concentration (roughly half of global cases from ~30% of the world’s population) is the single largest regional disparity in the global cancer burden and the basis for the 10x East-Asia-vs-US multiplier in the body text.\n","independence_note":"Draws on the same GLOBOCAN 2022 compilation as the Bray 2024 paper and the IARC news item; treat as partially dependent with respect to the headline death count. The regional breakdown and risk- factor discussion are the added value beyond the headline figures.\n"},{"url":"https://seer.cancer.gov/statfacts/html/livibd.html","title":"Cancer of the Liver and Intrahepatic Bile Duct — Cancer Stat Facts","publisher":"US National Cancer Institute / Surveillance, Epidemiology, and End Results Program (SEER)","source_type":"govt_report","statistic":"US lifetime risk of being diagnosed with liver and intrahepatic bile duct cancer ~1.1%; ~42,240 new cases and ~30,090 deaths estimated for 2025; 5-year relative survival 22.0%; sixth leading cause of cancer death in the US","excerpt":"\"Approximately 1.1 percent of men and women will be diagnosed with liver and intrahepatic bile duct cancer at some point during their lifetime, based on 2018-2021 data. [...] The rate of new cases of liver and intrahepatic bile duct cancer was 9.4 per 100,000 men and women per year. [...] The death rate was 6.6 per 100,000 men and women per year. [...] 5-year relative survival: 22.0%.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260317060529/https://seer.cancer.gov/statfacts/html/livibd.html","calculation_notes":"SEER is the methodological gold standard for US cancer lifetime risk. The 1.1% lifetime diagnosis figure (≈ 1 in 91) combined with a ~22% five-year survival gives an approximate US lifetime mortality of ~0.85%, which rounds to the ~0.7% figure used as the US lifetime anchor in the regional_breakdown table. The ~9.4 per 100,000 US incidence rate is less than half the global rate in Eastern Asia, which drives the order-of-magnitude regional spread. Used as the direct US anchor and as the prognosis anchor (22% five-year survival is one of the worst among common cancers, behind only pancreatic and oesophageal).\n","independence_note":"SEER (NCI) is independent of IARC GLOBOCAN — SEER is US-only vital registration and population-based cancer registry data, IARC is a global compilation. Comparing the two anchors the US-vs-global gap.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/hepatitis-b","title":"Hepatitis B — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"254 million people living with chronic hepatitis B infection globally in 2022; ~1.1 million hepatitis B deaths per year mostly from cirrhosis and hepatocellular carcinoma; perinatal HBV infection becomes chronic in ~95% of cases vs <5% in adult-acquired infection","excerpt":"\"In 2022, hepatitis B resulted in an estimated 1.1 million deaths, mostly from cirrhosis and hepatocellular carcinoma. [...] WHO estimates that 254 million people were living with chronic hepatitis B infection in 2022, with 1.2 million new infections each year. [...] Hepatitis B infection acquired in adulthood leads to chronic hepatitis in less than 5% of cases, whereas infection in infancy and early childhood leads to chronic hepatitis in about 95% of cases. [...] Some people with chronic hepatitis B will develop progressive liver disease and complications like cirrhosis and hepatocellular carcinoma (liver cancer).\"\n","source_date":"2024-04-09","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174454/https://www.who.int/news-room/fact-sheets/detail/hepatitis-b","calculation_notes":"WHO’s 1.1 million annual HBV deaths \"mostly from cirrhosis and hepatocellular carcinoma\" is the primary upstream source for the 25x chronic-HBV personal factor multiplier. Roughly half of global HCC cases are attributable to chronic HBV (the other half split between chronic HCV, alcohol, NAFLD/MASLD, and aflatoxin), and the 254 million chronic HBV carriers worldwide are concentrated in the same East/Southeast Asia and Sub-Saharan Africa regions that account for most global liver cancer mortality. The vaccinated multiplier of ~0.1 reflects the ~95% efficacy of the HBV vaccine in preventing chronic infection when administered in infancy — the intervention that has begun to bend the incidence curve in post-1980s cohorts in Taiwan, China, and elsewhere.\n","independence_note":"WHO hepatitis B fact sheet draws on separate surveillance pipelines (WHO Global Hepatitis Programme, national seroprevalence surveys) from the IARC GLOBOCAN cancer registry pipeline. Treated as an independent line of evidence on the risk-factor side even though the downstream liver cancer mortality numbers are partially dependent via cause-of-death attribution.\n"},{"url":"https://www.cancer.org/cancer/types/liver-cancer/about/what-is-key-statistics.html","title":"Key Statistics About Liver Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"~42,340 new US liver cancer cases and ~30,980 deaths projected for 2026 (27,790 new cases in men, 14,550 in women; 19,650 deaths in men, 11,330 in women); US liver cancer incidence has tripled over the past four decades","excerpt":"\"About 42,340 new cases (27,790 in men and 14,550 in women) will be diagnosed [...] About 30,980 people (19,650 men and 11,330 women) will die of these cancers [...] Liver cancer incidence rates have tripled in the US over the past 4 decades.\"\n","source_date":"2026-01-13","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260426203236/https://www.cancer.org/cancer/types/liver-cancer/about/what-is-key-statistics.html","calculation_notes":"ACS figures match SEER to within ~2% and are used as the annual US aggregate anchor. The more interesting number here is the three-fold increase in US liver cancer incidence over the past four decades — most of which is attributable to (a) the HCV infection cohort born 1945-1965, (b) rising rates of NAFLD/MASLD driven by obesity and metabolic syndrome, and (c) the ageing of the population. Used as the basis for the third body paragraph on NAFLD/MASLD as an emerging risk factor.\n","independence_note":"ACS and SEER share the same underlying vital-registration and cancer-registry upstream (NCHS mortality data, NAACCR incidence data). Treat as one combined US line of evidence.\n"},{"url":"https://www.eurekalert.org/news-releases/1092387","title":"The Lancet: Three in five liver cancer cases due to preventable risk factors; obesity-linked cases on the rise, new analysis suggest","publisher":"EurekAlert! (AAAS) — official press release for The Lancet Commission on liver cancer","source_type":"reputable_reference","statistic":"Global liver cancer cases projected to nearly double from 870,000 (2022) to 1.52 million (2050) and deaths from 760,000 to 1.37 million; MASH-attributable share rising from 8% to 11% and alcohol-attributable from 19% to 21%; over 55% of US adults could have MASLD by 2040; at least 60% of liver cancers are preventable","excerpt":"\"the number of new liver cancer cases will nearly double from 870,000 in 2022 to 1.52 million in 2050. [...] The number of deaths from liver cancer are predicted to grow from 760,000 in 2022 to 1.37 million in 2050. [...] the proportion of liver cancer cases associated with MASH are projected to increase from 8% in 2022 to 11% in 2050 [...] liver cancer cases associated with alcohol are projected to increase from 19% in 2022 to 21% in 2050 [...] by 2040, over 55% of US adults could have MASLD.\"\n","source_date":"2025-07-28","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260623152415/https://www.eurekalert.org/news-releases/1092387","calculation_notes":"This is a forward projection (2050 horizon), not a revision of the current GLOBOCAN 2022 mortality used for the headline figure, so it does not change normalized.lifetime_us_adult — it is cited to support the forward-projection caveat and to strengthen the MASLD/MASH and alcohol personal-factor narratives. The Lancet Commission's 2022 baseline (870,000 cases / 760,000 deaths) is consistent with the GLOBOCAN 2022 figures already anchoring this entry (866,136 cases / 758,725 deaths) to within rounding. The MASH share rising 8%->11% and alcohol 19%->21% are attributable-fraction projections, not relative-risk estimates, so they do not yield a new multiplier value.\n","independence_note":"The Lancet Commission draws on the same GLOBOCAN/GBD compilation as the IARC and Bray sources for its 2022 baseline; treat its baseline death count as partially dependent. The 2050 projection and the attributable-share modelling are the added value beyond the existing headline figures.\n"}],"comparison_anchors":[{"label":"Death from cancer (lifetime, global adult, all sites)","lifetime_us_adult":0.14},{"label":"Death from lung cancer (lifetime, global adult)","lifetime_us_adult":0.018},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death from a smoking-related disease (lifelong regular smoker)","lifetime_us_adult":0.5},{"label":"Death from alcohol-attributable disease (lifelong heavy drinker)","lifetime_us_adult":0.15},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.015,"notes":"~758K liver cancer deaths/yr across ~6B adults (GLOBOCAN 2022); age-weighted lifetime"},{"region":"East/Southeast Asia","probability":0.04,"notes":"Dominated by chronic hepatitis B; Eastern Asia alone accounts for roughly half of global liver cancer deaths, reflecting decades of endemic HBV transmission before the vaccination era"},{"region":"Sub-Saharan Africa","probability":0.025,"notes":"HBV plus dietary aflatoxin exposure; incidence is high but competing mortality and under-reporting make the absolute lifetime figure uncertain"},{"region":"US adult","probability":0.007,"notes":"SEER lifetime diagnosis ~1.1%, 5-year survival ~22%, implied lifetime mortality ~0.7-0.9%; rising over the past four decades"},{"region":"Western Europe","probability":0.008,"notes":"Lower than East Asia; burden increasingly driven by HCV legacy infection, alcohol, and NAFLD/MASLD rather than HBV"}],"personal_factor_multipliers":[{"factor":"chronic hepatitis B (HBV) infection","multiplier":25,"notes":"Chronic HBV is the single largest global driver of hepatocellular carcinoma. Relative risk estimates from prospective cohorts range from roughly 15x to 30x vs uninfected; 25x is a rough mid-point. The risk is concentrated in carriers who acquired HBV perinatally or in early childhood, where chronicity rates exceed 90%."},{"factor":"chronic hepatitis C (HCV) infection","multiplier":17,"notes":"Chronic HCV is the dominant driver of liver cancer in high-income countries with lower HBV prevalence. The US 1945-1965 birth cohort carries a disproportionate HCV burden and accounts for much of the historical tripling of US liver cancer incidence. Direct-acting antivirals introduced after 2013 materially reduce but do not eliminate HCC risk."},{"factor":"heavy alcohol use + HCV","multiplier":35,"notes":"Alcohol and chronic viral hepatitis interact roughly multiplicatively for HCC risk. Heavy drinking combined with chronic HCV produces one of the highest documented liver cancer risk profiles in the epidemiological literature; this interaction drives the regular-drinking-death entry’s 4x HBV/HCV co-infection multiplier."},{"factor":"non-alcoholic fatty liver disease (NAFLD/MASLD)","multiplier":2,"notes":"NAFLD/MASLD is the emerging liver cancer driver in wealthy countries, tied to obesity, type 2 diabetes, and metabolic syndrome. Relative risk estimates are lower than for chronic viral hepatitis (roughly 2x for the general NAFLD population, rising sharply for those who progress to NASH-cirrhosis), but the exposed population is vastly larger, so the population-attributable fraction is rising. The Lancet Commission on liver cancer (2025) projects the MASH-attributable share of global cases climbing from 8% in 2022 to 11% in 2050 and estimates over 55% of US adults could have MASLD by 2040, underscoring why this small per-person multiplier carries a large and growing population burden."},{"factor":"HBV vaccinated","multiplier":0.1,"notes":"The hepatitis B vaccine is ~95% effective at preventing chronic infection when administered in infancy. Taiwan’s universal infant HBV vaccination programme, launched in 1984, produced a ~70% reduction in childhood HCC incidence within two decades — the clearest real-world demonstration of a cancer intervention driven by vaccination."},{"factor":"heavy alcohol use (lifelong)","multiplier":4,"notes":"Heavy drinking alone raises liver cancer risk through the alcoholic-cirrhosis pathway. Effect is smaller than chronic viral hepatitis but interacts multiplicatively with HBV, HCV, and NAFLD. The Lancet Commission on liver cancer (2025) projects the alcohol-attributable share of global cases rising from 19% in 2022 to 21% in 2050, keeping alcohol the second-largest modifiable driver behind viral hepatitis."}],"short_label":"Liver cancer","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Liver cancer is the Likelier entry with the largest region-to-region variance of any cancer on the site, and a single headline number does a worse job than usual of summarising a reader’s actual risk. The ~1 in 60 global lifetime figure is an average across a population where East Asia and Sub-Saharan Africa carry a disproportionately large share of the global burden and where North America and Western Europe sit at roughly a third of the global age-standardised rate. The single biggest determinant of individual risk is chronic hepatitis B or hepatitis C status, which is binary for any given person and moves the lifetime number by more than an order of magnitude. The US number is itself moving: ACS reports US liver cancer incidence has tripled over the past four decades, driven partly by the HCV-exposed 1945-1965 birth cohort and partly by the rising NAFLD/MASLD burden, and is expected to continue rising as the NAFLD cohort ages. On the other side, universal infant HBV vaccination — routine in most countries since the 1990s — has already begun to bend the incidence curve in post-vaccination birth cohorts in Taiwan, mainland China, and elsewhere. The global liver cancer picture over the next 30 years is an uneven mix of falling HBV-driven incidence in vaccinated Asian cohorts and rising NAFLD-driven incidence in metabolically-unwell Western cohorts. Finally, \"liver cancer\" here is dominated by hepatocellular carcinoma (HCC), which is ~80% of primary liver cancer globally; intrahepatic cholangiocarcinoma and other subtypes are lumped in with HCC in the SEER and GLOBOCAN headline figures but have different risk-factor profiles and prognoses. Looking forward, the Lancet Commission on liver cancer (July 2025) projects the global burden nearly doubling — new cases rising from 870,000 in 2022 to 1.52 million by 2050 and deaths from 760,000 to 1.37 million — driven less by HBV (falling in vaccinated cohorts) than by metabolic and alcohol-related disease: the MASH-attributable share of cases is modelled to rise from 8% to 11% and the alcohol-attributable share from 19% to 21%, with over 55% of US adults projected to have MASLD by 2040. The Commission judges at least 60% of liver cancers preventable through HBV/HCV control and management of alcohol and metabolic risk, so this projected rise is a policy choice as much as a forecast. None of this changes the present-day headline figure here, which is anchored to the 2022 burden; it sharpens the direction of travel.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale oblong shape with an uneven boundary on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/liver-cancer","api_url":"https://likelier.app/api/fears/liver-cancer.json"},{"slug":"medication-serious-adverse-event","question":"What are the odds of a serious adverse drug reaction from prescribed medication?","quick_answer":"About 1 in 58 over a lifetime for a fatal adverse drug reaction, while roughly 1 in 9 adults are hospitalized for a serious one at some point. Most people would guess well under 1 in 1,000 for the lifetime fatal risk.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"There is no standing survey that isolates \"fear of a serious adverse drug reaction from a prescribed medication.\" Prescription drug risk sits in a cultural blind spot: patients skim the side-effect leaflet, notice that most entries end in \"rare,\" and file the category under \"things that happen to other people.\" The same readers who will happily interrogate a vaccine's one-in-a-million myocarditis signal tend to accept an SSRI, an NSAID, or a beta-lactam antibiotic with almost no probabilistic framing at all. The gap between that intuition and the aggregate hospitalisation and mortality numbers is one of the larger mismatches on this site.\n","rough_estimate":"most US adults would guess well under 1 in 1,000 for a lifetime fatal ADR","kind":"intuition"},"native":{"display":"~50,000-150,000 fatal adverse drug reactions per year, US (midpoint ~75,000)","numerator":75000,"denominator":258000000,"unit":"per year","population":"US adults, all ages pooled"},"normalized":{"lifetime_us_adult":0.0171,"display":"~1 in 58 lifetime fatal ADR (US adult)","log_value":-1.767,"assumptions":"Two different headline numbers live in this entry and it is worth separating them. (1) Lifetime probability of at least one *hospitalisation-level* serious ADR: CDC reports roughly 500,000 ADE-related hospitalisations per year in the US. Against ~258 million US adults that is a per-adult-year hazard of about 1.94 per 1,000. Compounded over 59 years of remaining adult life: 1 − (1 − 0.00194)^59 ≈ 0.108, i.e. roughly 1 in 9 adults experience a serious ADR requiring hospitalisation at some point. (2) Lifetime probability of a *fatal* ADR: Lazarou et al. (JAMA 1998) estimated ~106,000 fatal ADRs in US hospitals in 1994; more recent analyses have argued that figure is on the high end of the plausible envelope, with defensible modern point estimates in the 50,000-150,000/year range. Using a midpoint of ~75,000 fatal ADRs per year against 258 million US adults gives a per-adult-year hazard of ~2.91 × 10^-4 and a compounded lifetime figure of 1 − (1 − 0.000291)^59 ≈ 0.0171, or roughly 1 in 58. The headline normalized value reports the fatal-ADR lifetime number so it can be compared directly with other mortality entries on the site; the hospitalisation-level figure is discussed in the body text.\n","uncertainty":{"low":0.0115,"high":0.034},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/9555760/","title":"Incidence of Adverse Drug Reactions in Hospitalized Patients: A Meta-analysis of Prospective Studies","publisher":"JAMA / Lazarou J, Pomeranz BH, Corey PN","source_type":"peer_reviewed","statistic":"Serious ADR incidence 6.7% (95% CI 5.2-8.2%); fatal ADR incidence 0.32% (95% CI 0.23-0.41%); estimated 2,216,000 serious and 106,000 fatal ADRs among US hospitalised patients in 1994","excerpt":"\"The overall incidence of serious ADRs was 6.7% (95% confidence interval [CI], 5.2%-8.2%)\" and \"of fatal ADRs was 0.32% (95% CI, 0.23%-0.41%) of hospitalized patients.\" \"2216000 (1721000-2711000) hospitalized patients had serious ADRs and 106000 (76000-137000) had fatal ADRs.\"\n","source_date":"1998-04-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175711/https://pubmed.ncbi.nlm.nih.gov/9555760/","calculation_notes":"Lazarou et al. pooled 39 prospective studies of hospitalised US patients and reported a 6.7% serious ADR incidence and a 0.32% fatal ADR incidence (per admission, not per patient-year). Extrapolated to 1994 US hospital volumes the authors estimated ~2.2M serious and ~106K fatal ADRs per year, which — if accurate — placed ADRs between the 4th and 6th leading cause of US death at the time. The 106K/year figure anchors the upper end of Likelier's uncertainty band; the per-admission rates are the cleanest cross-study signal and dominate the literature's central tendency.\n","independence_note":"Lazarou is a meta-analysis of 39 earlier prospective studies, so it is not independent of those constituent datasets, but it is the canonical reference point and methodologically independent of both Pirmohamed (UK, prospective admissions cohort) and the CDC ADE surveillance (US emergency department sample).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15231615/","title":"Adverse drug reactions as cause of admission to hospital: prospective analysis of 18,820 patients","publisher":"BMJ / Pirmohamed M, James S, Meakin S, et al.","source_type":"peer_reviewed","statistic":"6.5% of UK hospital admissions ADR-related; overall fatality 0.15%; ADR directly caused the admission in 80% of cases","excerpt":"\"There were 1225 admissions related to an ADR, giving a prevalence of 6.5%\" and \"The overall fatality was 0.15%.\"\n","source_date":"2004-07-03","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175749/https://pubmed.ncbi.nlm.nih.gov/15231615/","calculation_notes":"Pirmohamed et al. prospectively analysed 18,820 UK adult hospital admissions and found 6.5% were ADR-related, with a 0.15% fatality rate among those admissions. The result replicates Lazarou's order of magnitude in a completely independent (UK NHS, 2004) dataset and with a cleaner prospective design, which is why it is the preferred independence check in this entry. The lower fatality rate vs. Lazarou reflects both methodology (admission-triggering ADRs vs. in-hospital ADRs) and a more conservative treatment of causality attribution.\n","independence_note":"Fully independent of Lazarou: different country, different decade, different study design (prospective admissions cohort rather than meta-analysis of in-hospital ADRs). This is the strongest cross-check on the hospitalisation-level incidence figure.\n"},{"url":"https://www.cdc.gov/medication-safety/data-research/facts-stats/index.html","title":"FastStats: Medication Safety Data","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~1.5 million US ED visits per year for ADEs; ~500,000 require hospitalisation; >600,000 ED visits per year among adults 65+","excerpt":"\"More than 1.5 million people visit emergency departments for ADEs each year in the United States, and almost 500,000 require hospitalization.\" \"Older adults (65 years or older) visit emergency departments more than 600,000 times each year, more than twice as often as younger people.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260320033225/https://www.cdc.gov/medication-safety/data-research/facts-stats/index.html","calculation_notes":"CDC's ADE surveillance (built on the National Electronic Injury Surveillance System — Cooperative Adverse Drug Event Surveillance project, NEISS-CADES) gives the cleanest contemporary denominator for US serious ADRs: ~1.5M ED visits and ~500K hospitalisations per year. Likelier uses the 500K/year hospitalisation figure as the anchor for the \"serious ADR lifetime risk ≈ 1 in 9\" calculation in the body. The ~600K/year figure for adults 65+ is what drives the \"age 75+ multiplier 5×\" personal factor below.\n","independence_note":"Independent of Lazarou and Pirmohamed: CDC ADE surveillance is a US national ED sample (NEISS-CADES), not a hospitalised-inpatient cohort. It is the only one of the three sources whose numerator directly counts individual ADE events in the general population rather than extrapolating from a study sample.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/31183532/","title":"Mortality among patients due to adverse drug reactions that occur following hospitalisation: a meta-analysis","publisher":"European Journal of Clinical Pharmacology / Patel PB, Patel TK","source_type":"peer_reviewed","statistic":"Fatal ADR prevalence during hospitalisation: 0.11% (95% CI 0.06-0.18%) from 48 studies; elderly populations 0.27%","excerpt":"\"The pooled prevalence of fatal adverse drug reactions occurring during hospitalisation was 0.11% (95% CI: 0.06-0.18%; I2 = 93%).\"\n","source_date":"2019-09-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260426203728/https://pubmed.ncbi.nlm.nih.gov/31183532/","calculation_notes":"Patel & Patel's 48-study meta-analysis directly updates Lazarou's 0.32% fatal ADR figure with a modern pooled estimate of 0.11%. Extrapolating to ~34M US hospital admissions/year yields ~37,000 fatal ADRs/year — well within this entry's 50,000-150,000 uncertainty band but below Lazarou's 106,000 point estimate, confirming that the 75,000/year midpoint is reasonable. The study also found that elderly populations (0.27%) and ICU/internal-medicine wards (0.46%) drive the overall rate upward.\n","independence_note":"Fully independent of Lazarou (different studies included, different decades, different methodology) and of Pirmohamed (UK prospective vs. global meta-analysis).\n"}],"comparison_anchors":[{"label":"Death from drug overdose (lifetime, US adult)","lifetime_us_adult":0.0237},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Fatal anaphylaxis, all causes (lifetime, US adult)","lifetime_us_adult":0.0000363},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"polypharmacy (5+ concurrent medications)","multiplier":4,"notes":"Drug-drug interactions scale non-linearly with the number of concurrent agents; polypharmacy is the single biggest in-population driver of fatal ADRs and is what pulls the elderly risk curve upward.\n"},{"factor":"age 75+","multiplier":5,"notes":"CDC reports adults 65+ account for >600K of ~1.5M annual ADE ED visits despite being ~17% of the population; the over-75 slice is where the curve turns sharply. Age here is mostly a proxy for polypharmacy, renal clearance, and frailty.\n"},{"factor":"single chronic medication, young otherwise-healthy adult","multiplier":0.3,"notes":"The bottom of the distribution: one well-tolerated agent, normal renal and hepatic function, no interactions. The population-average figure substantially overstates this reader's personal risk.\n"},{"factor":"known drug allergy not flagged in the chart","multiplier":10,"notes":"Missed allergy flags are one of the classic preventable-fatal-ADR pathways and are the biggest single modifiable personal risk factor. Order-of-magnitude estimate.\n"}],"short_label":"Medication reaction","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The population-level number here papers over heterogeneity that matters. Age is the dominant axis: adults 65+ account for more than 600,000 of the ~1.5M annual ADE ED visits, so a healthy 30-year-old on a single medication faces a risk well below the headline and an 80-year-old on eight medications faces a risk several multiples above it. Drug class matters almost as much — CDC surveillance identifies anticoagulants (~18% of ADE ED visits), systemic antibiotics (~16%), and diabetes agents including insulin (~13%) as the three biggest categories, with opioid analgesics next behind. The Lazarou 106,000 fatal-ADR/year figure has been contested downward in the literature since publication, which is why Likelier's uncertainty band spans roughly 50,000-150,000 US fatal ADRs per year rather than pinning to the 1994 point estimate. Finally this entry covers only adverse reactions to medications taken as prescribed: overdose (accidental or intentional) is tracked under <a href=\"/fears/drug-overdose\">drug-overdose</a>, and drug-induced anaphylaxis — roughly 59% of all fatal anaphylaxis per Jerschow et al. — is the iatrogenic slice of <a href=\"/fears/anaphylaxis-fatal\">anaphylaxis-fatal</a>.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single white oval pill resting on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/medication-serious-adverse-event","api_url":"https://likelier.app/api/fears/medication-serious-adverse-event.json"},{"slug":"restaurant-food-poisoning","question":"What are the odds of being hospitalized from food poisoning after eating at a restaurant?","quick_answer":"Being hospitalized from restaurant food poisoning is about a 1 in 58 lifetime risk, roughly 1 in 762,000 per meal. Many diners intuitively fear a serious case at 1-in-10 to 1-in-50 each year, well above the real epidemiology.\n","category":"food","tags":["food","travel"],"no_reliable_estimate":false,"perceived":{"description":"Most diners carry a vague, diffuse worry about getting sick from a restaurant meal — the kind that surfaces after spotting an unwashed cutting board, ordering something slightly undercooked, or hearing about a local outbreak. Few people would put a number on it, but the intuitive sense is that a genuinely bad case requiring a hospital visit is somewhere in the range of once-in-a-decade to once-in-a-lifetime, with a rough feel of perhaps 1-in-10 to 1-in-50 per year. That is comfortably above the actual epidemiology, which puts the lifetime hospitalization risk closer to 1-in-55 — real enough to warrant attention, but not the lurking catastrophe that a nervous diner might imagine.\n","rough_estimate":"most diners implicitly estimate a serious illness risk of 1-in-10 to 1-in-50 per year","kind":"intuition"},"native":{"display":"~1 in 762,000 per restaurant meal (hospitalization)","numerator":1,"denominator":762000,"unit":"per restaurant meal","population":"US adults eating at restaurants, catering establishments, or delis"},"normalized":{"lifetime_us_adult":0.0173,"display":"~1 in 58 US adults hospitalized from restaurant food poisoning over a lifetime","log_value":-1.762,"assumptions":"CDC/Scallan 2011 combined estimate: ~128,000 US hospitalizations per year from domestically acquired foodborne illness (known pathogens plus unspecified agents). CDC outbreak surveillance (MMWR FoodNet reports) consistently attributes roughly 60% of reported foodborne illness outbreaks to restaurants, catering, and delis. Applying that fraction: 128,000 × 0.60 ≈ 76,800 restaurant-linked hospitalizations per year. US adults (~260 million) eat out approximately 4-5 times per week on average (USDA ERS data), approximated here as 225 meals per year × 260 million adults = 58.5 billion restaurant meals per year. Per-meal hospitalization risk: 76,800 / 58.5e9 ≈ 1.31 per million meals (1 in ~762,000). Lifetime probability for a US adult eating out 225 meals/year over 59 years of adult life: 1 - (1 - 1.31e-6)^(225 × 59) ≈ 1.73%, or roughly 1 in 58. Uncertainty band reflects: (low) restaurant fraction 50%, 150 meals/year; (high) restaurant fraction 70%, 300 meals/year, with ~10% upward adjustment for under-ascription of restaurant source.\n","uncertainty":{"low":0.0084,"high":0.032},"scope":"us_adult_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","title":"Foodborne Illness Acquired in the United States — Major Pathogens","publisher":"CDC Emerging Infectious Diseases / Scallan et al.","source_type":"peer_reviewed","statistic":"31 major pathogens cause ~9.4 million illnesses, ~55,961 hospitalizations, and ~1,351 deaths per year in the US","excerpt":"\"We estimated that 31 pathogens acquired in the United States caused 9.4 million episodes of foodborne illness (90% credible interval [CrI] 6.6–12.7 million), 55,961 hospitalizations (90% CrI 39,534–75,741), and 1,351 deaths (90% CrI 712–2,268) each year.\"\n","source_date":"2011-01-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260505062544/https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","calculation_notes":"This is the known-pathogen half of Scallan 2011. Combined with the companion unspecified-agents paper (Scallan et al. 2011b), total hospitalizations reach ~127,839 per year (~56,000 + ~71,878). We use 128,000 as the round CDC-cited combined total. Applying 60% restaurant attribution: ~76,800 restaurant-linked hospitalizations/year. Divided by 58.5 billion restaurant meals/year (260M US adults × 225 meals/year): per-meal risk ≈ 1.31 × 10⁻⁶. Lifetime: 1 - (1 - 1.31e-6)^(13,275 meals over 59 years) ≈ 0.0173.\n"},{"url":"https://wwwnc.cdc.gov/eid/article/17/1/p2-1101_article","title":"Foodborne Illness Acquired in the United States — Unspecified Agents","publisher":"CDC Emerging Infectious Diseases / Scallan et al.","source_type":"peer_reviewed","statistic":"Unspecified agents add ~71,878 hospitalizations per year; combined known + unspecified total is ~127,839 hospitalizations/year","excerpt":"\"We estimated that unknown agents acquired in the United States caused 38.4 million (90% CrI 19.8–61.2 million) episodes of foodborne illness, 71,878 (90% CrI 9,924– 157,340) hospitalizations.\"\n","source_date":"2011-01-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260503093309/https://wwwnc.cdc.gov/eid/article/17/1/p2-1101_article","calculation_notes":"Companion paper to Scallan 2011a. Adds unspecified-agent hospitalizations to reach the combined ~128,000/year figure used as the denominator basis for all restaurant-attribution calculations.\n","independence_note":"Scallan 2011a and 2011b partition the same CDC FoodNet/surveillance data into known-pathogen and unspecified-agent components. They share methodology and first author; treat as a single methodological pipeline, not two independent confirmations of the 128,000 hospitalization figure.\n"},{"url":"https://www.cdc.gov/nors/reports/foodborne-outbreaks-2018.html","title":"Foodborne Disease Outbreaks Reported to CDC, United States, 2018","publisher":"CDC / National Outbreak Reporting System (NORS)","source_type":"govt_report","statistic":"Restaurants were the most commonly reported food-preparation setting for foodborne illness outbreaks in 2018, accounting for 631 outbreaks (68% of outbreaks for which a single preparation location was reported)","excerpt":"\"As reported in previous years, restaurants (631 outbreaks, 68% of outbreaks for which a single location of preparation was reported), specifically restaurants with sit-down dining (445, 48%), were the most commonly reported locations of food preparation associated with outbreaks.\"\n","source_date":"2018-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260714015529/https://www.cdc.gov/nors/reports/foodborne-outbreaks-2018.html","calculation_notes":"CDC's 2018 NORS annual summary reports restaurants as the most common single-location preparation setting: 631 outbreaks, 68% of the outbreaks for which a single preparation location was reported. Adding catering/banquet facilities pushes the restaurant-plus-catering fraction higher still. This is consistent with MMWR reports from other years (range 55–68%). We use a conservative 60% as the central estimate for restaurant-fraction attribution in the lifetime calculation. Note that reported outbreaks are a subset of total foodborne illness — sporadic cases far outnumber outbreak-associated ones — so this fraction is an approximation for all restaurant-linked hospitalizations.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12505355/","title":"Vibrio vulnificus epidemiology and risk factors for mortality in the United States, 2000-2022","publisher":"Infectious Diseases (London)","source_type":"peer_reviewed","statistic":"Of 2,989 US Vibrio vulnificus cases 2000-2022, 2,493 (83.4%) patients were hospitalized; approximately 20% of cases are fatal","excerpt":"\"Overall, 2,493 (83.4%) patients were hospitalized.\"\n","source_date":"2025-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714015555/https://pmc.ncbi.nlm.nih.gov/articles/PMC12505355/","calculation_notes":"Used as the basis for the raw-shellfish personal factor multiplier and the body-text claim about Vibrio vulnificus hospitalization rate. The 83.4% hospitalization figure is the authoritative surveillance-period rate; it is distinct from the >50% immunocompromised case-fatality figure cited from FDA below, which is a fatality rate within a higher-risk subgroup rather than a hospitalization rate for all cases.\n"}],"comparison_anchors":[{"label":"Death from foodborne illness (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Hospitalization from a fall at home (lifetime, US adult)","lifetime_us_adult":0.12},{"label":"Blood clot after a long-haul flight (per flight)","lifetime_us_adult":0.000215}],"personal_factor_multipliers":[{"factor":"Eating raw shellfish (oysters, clams, raw bar)","multiplier":10,"notes":"Raw bivalves are the dominant vehicle for norovirus and Vibrio vulnificus — the latter carrying a case-fatality rate >50% in immunocompromised individuals. FDA estimates 85% of Vibrio vulnificus illnesses are linked to raw oyster consumption. A raw-oyster eater adds a discrete high-risk exposure on top of the baseline meal rate; the roughly 10x multiplier reflects the markedly higher per-serving hospitalization rate for this food category compared to the average restaurant meal.\n"},{"factor":"Eating at a buffet vs. table service","multiplier":3,"notes":"Buffets have longer food holding times, repeated customer handling, and variable temperature control — all factors that favor pathogen growth. CDC outbreak data consistently show higher attack rates and more frequent Clostridium perfringens and Staph aureus outbreaks associated with buffet and catering/banquet settings compared to table-service restaurants. A 2-4x elevation is a reasonable heuristic; we use 3x as the central estimate.\n"},{"factor":"Immunocompromised status (HIV, chemotherapy, organ transplant, age >65)","multiplier":5,"notes":"Listeria, Salmonella non-typhi, and Cryptosporidium cause disproportionately severe disease — including hospitalization and death — in immunocompromised adults. CDC estimates that adults over 65, pregnant women, and those with weakened immune systems face roughly 5-17x higher risk of invasive Salmonella and up to 10x higher risk of Listeria hospitalization compared to healthy adults under 65.\n"},{"factor":"Eating at a restaurant in a developing country","multiplier":8,"notes":"Traveler's diarrhea is the most common travel-related illness; 20-50% of travelers to high-risk regions experience it. Hospitalization risk is elevated by exposure to ETEC, Shigella, Campylobacter, and Cryptosporidium at rates not seen in US restaurant settings. A conservative 8x multiplier for the per-meal hospitalization risk in high-risk regions (South Asia, sub-Saharan Africa, parts of Latin America) versus the US baseline.\n"}],"short_label":"Restaurant food poisoning","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The denominator assumes all US adults eat out ~225 times per year; the actual distribution is right-skewed — heavy restaurant users (daily lunches, frequent travel) face higher cumulative lifetime exposure than the average implies. The 60% restaurant-attribution fraction comes from reported outbreak data; sporadic cases (the vast majority) are rarely traced back to their source, so the true fraction could be higher or lower. Under-reporting is substantial — only an estimated 1-2% of foodborne illness hospitalizations appear in outbreak surveillance. The lifetime figure is therefore best read as an order-of-magnitude estimate rather than a precise actuarial number. Risk is also heterogeneous by pathogen: norovirus dominates case counts but rarely hospitalizes; Salmonella and Clostridium perfringens drive most restaurant-associated hospitalizations.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A muted flat vector illustration of a fork and spoon crossed over an empty plate on a pale background."},"canonical_url":"https://likelier.app/restaurant-food-poisoning","api_url":"https://likelier.app/api/fears/restaurant-food-poisoning.json"},{"slug":"drug-overdose","question":"What are the odds of dying from a drug overdose?","quick_answer":"A US adult's lifetime odds of dying from a drug overdose are about 1 in 56. Most people would guess well under 1 in 1,000 for themselves.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"There is no canonical \"fear of dying from a drug overdose\" poll the way there is for flying or being murdered. Most people do not file overdose under \"things that could happen to me\" at all — it is mentally shelved alongside events that happen to other people, in other neighbourhoods, after other life trajectories. The gap between that intuition and the actual US lifetime number is one of the largest on this site.\n","rough_estimate":"most US adults would guess well under 1 in 1,000 for themselves","kind":"intuition"},"native":{"display":"~23.3 per 100,000 per year (all ages, crude, 2024; 23.1 age-adjusted)","numerator":79384,"denominator":340100000,"unit":"per year","population":"US residents, all ages pooled (2024)"},"normalized":{"lifetime_us_adult":0.0179,"display":"1 in ~56 lifetime (US adult)","log_value":-1.747,"assumptions":"Revised June 2026 to reflect the post-2023 reversal. Now uses 79,384 US drug overdose deaths in 2024 (CDC NCHS Data Brief 549, the latest complete year) against a roughly 260 million US adult population (18+, ~77% of the July 2024 resident total of 340.1 million). That yields an approximate adult-year hazard of 79,384 / 260,000,000 ≈ 0.000305 per adult-year. Compounded over 59 years of remaining adult life: 1 − (1 − 0.000305)^59 ≈ 0.0179, or about 1 in 56. This is down from the 1-in-42 figure built on the 2023 peak: the national age-adjusted rate fell 26.2% between 2023 and 2024 (31.3 → 23.1 per 100,000), the largest single-year decline NCHS has recorded, and 2025 provisional data project a further ~14% drop. The per-adult hazard remains higher than the per-capita hazard because deaths concentrate in working-age adults. Counts include all drug-involved overdoses — opioids, stimulants, benzodiazepines, and polysubstance — and are dominated by synthetic opioids (fentanyl) in the 2020s. Excludes pure alcohol poisoning (tracked separately) and excludes deaths coded as intentional self-harm by drug poisoning (ICD-10 X60–X64), which fall under suicide, not accidental/undetermined overdose. The uncertainty band reflects the open question of whether the 2024–2025 decline holds: the low bound (~1 in 63) applies the lower 2025 provisional count (~69,973 deaths) over the same 59-year horizon, while the high bound (~1 in 42) retains headroom for a rebound toward the 2023 peak level, since a single low year compounded over a lifetime should not be treated as locked in.\n","uncertainty":{"low":0.0158,"high":0.024},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db549.htm","title":"Drug Overdose Deaths in the United States, 2023–2024 (NCHS Data Brief No. 549)","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"79,384 US drug overdose deaths in 2024; age-adjusted rate 23.1 per 100,000, down 26.2% from 31.3 in 2023; 2024 rate highest for American Indian/Alaska Native (51.6) and Black (33.8), lowest for Asian (4.4); deaths involving synthetic opioids other than methadone (fentanyl) fell from 72,776 in 2023 to 47,735 in 2024 — a 35.6% decline in age-adjusted rate, the steepest of any drug category","excerpt":"\"In 2024, 79,384 drug overdose deaths occurred, resulting in an age-adjusted rate of 23.1 deaths per 100,000 standard population… The largest decrease, 26.2%, occurred between 2023 and 2024, from 31.3 to 23.1… In 2024, the age-adjusted drug overdose death rate was highest for American Indian and Alaska Native non-Hispanic people (65.0 and 51.6, respectively) and lowest for Asian non-Hispanic people (5.1 and 4.4, respectively). The rate decreased most (30.9%) for Black non-Hispanic (subsequently, Black) people, from 48.9 to 33.8 deaths per 100,000 standard population between 2023 and 2024… Between 2023 and 2024, the drug overdose death rate involving synthetic opioids other than methadone decreased by 35.6% (from 22.2 to 14.3). Compared with other reported drug types, synthetic opioids other than methadone showed the largest decline.\"\n","source_date":"2026-01-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260618195504/https://www.cdc.gov/nchs/products/databriefs/db549.htm","calculation_notes":"Garnett MF, Miniño AM. Drug overdose deaths in the United States, 2023–2024. NCHS Data Brief. 2026 Jan;(549):1–13. This is the latest complete-year federal product on overdose mortality and supersedes Data Brief 522 (2023) for the headline figure. 79,384 deaths / ~260M US adults ≈ 0.000305 per adult-year; 1 − (1 − 0.000305)^59 ≈ 0.0179 ≈ 1 in 56. Race/ethnicity figures are age-adjusted rates per 100,000, the correct basis for between-group comparison; the AI/AN-to-national ratio (51.6 / 23.1 ≈ 2.23) and Black-to-national ratio (33.8 / 23.1 ≈ 1.46) feed the subgroup multipliers below. Synthetic-opioid (fentanyl) deaths fell from 72,776 (2023) to 47,735 (2024); NCHS reports the age-adjusted rate for this category fell 35.6% — the source of the body-prose fentanyl decline figure, the steepest drop among the drug categories the brief reports.\n","independence_note":"NCHS Data Brief 549 is the primary federal product on 2024 overdose mortality, built directly from NVSS death-certificate records — the same pipeline as Data Brief 522 and NIDA's tracker. Treat as the authoritative count for the revised headline; the others remain as composition and coding-scope references.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db522.htm","title":"Drug Overdose Deaths in the United States, 2003–2023 (NCHS Data Brief No. 522)","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"105,007 US drug overdose deaths in 2023; age-adjusted rate 31.3 per 100,000 (down from 32.6 in 2022)","excerpt":"\"The age-adjusted rate of drug overdose deaths increased from 8.9 deaths per 100,000 standard population in 2003 to 32.6 in 2022; however, the rate decreased to 31.3 in 2023. After a period of increase between 2013 and 2022, rates of drug overdose deaths involving synthetic opioids other than methadone, which includes fentanyl, fentanyl analogs, and tramadol, decreased between 2022 and 2023.\"\n","source_date":"2024-12-19","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260401230006/https://www.cdc.gov/nchs/products/databriefs/db522.htm","calculation_notes":"NCHS counts deaths from death certificates with an underlying cause coded as drug poisoning (ICD-10 X40–X44 accidental, X60–X64 intentional, X85 assault, Y10–Y14 undetermined). Likelier's normalized figure uses the accidental + undetermined portion, which is where ~90%+ of overdose deaths sit; intentional self-harm by drug poisoning is excluded here and lives under the suicide category. 105,007 deaths / ~258M US adults ≈ 0.000407 per adult-year; 1 − (1 − 0.000407)^59 ≈ 0.0237 ≈ 1 in 42. This documents the 2023 figure; the live headline was revised to the 2024 count in Data Brief 549 (above), and the uncertainty band is now derived there.\n","independence_note":"NCHS Data Brief 522 is the primary federal product on overdose mortality, built directly from NVSS death certificate records. NIDA's tracker (below) republishes and extends the same underlying NVSS data with drug-category breakdowns, so the two sources are partially dependent.\n"},{"url":"https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates","title":"Drug Overdose Deaths: Facts and Figures","publisher":"National Institute on Drug Abuse (NIDA), NIH","source_type":"govt_report","statistic":"Over 105,000 US drug-involved overdose deaths in 2023; 107,941 in 2022; 79,358 opioid-involved; 72,776 synthetic-opioid-involved","excerpt":"\"Over 105,000 persons in the U.S. died from drug-involved overdose in 2023… Opioid-involved overdose deaths rose from 49,860 in 2019 to 81,806 in 2022 with a significant decrease to 79,358 in 2023… Drug overdose deaths involving synthetic opioids other than methadone (primarily fentanyl) decreased to 72,776 in 2023… In 2023, there were 10,870 drug overdose deaths involving benzodiazepines.\"\n","source_date":"2024-08-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412080956/https://nida.nih.gov/research-topics/trends-statistics/overdose-death-rates","calculation_notes":"NIDA's breakdown confirms that opioids (79,358) and specifically synthetic opioids (72,776) dominate the 2023 total, with stimulants (cocaine and methamphetamine combined, ~59,725 deaths, counted with substantial overlap) as the fast-growing second front. Benzodiazepines (10,870) are almost always part of a polysubstance mix rather than a lone cause. Used here to characterise the composition of the 105,007-death total rather than to recompute the headline number.\n","independence_note":"NIDA sources its figures from CDC WONDER / NVSS — same underlying data as the NCHS Data Brief — but provides the drug-category decomposition that NCHS headlines omit. Treat as a composition check, not an independent count.\n"},{"url":"https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm","title":"Vital Statistics Rapid Release — Provisional Drug Overdose Data","publisher":"CDC National Center for Health Statistics / National Vital Statistics System","source_type":"govt_report","statistic":"VSRR publishes provisional 12-month ending overdose death counts, identified via ICD-10 codes X40–X44, X60–X64, X85, Y10–Y14","excerpt":"\"Final drug overdose death data are published annually through NCHS statistical reports and CDC WONDER. Drug overdose deaths are identified using underlying cause-of-death codes X40–X44 (unintentional), X60–X64 (suicide), X85 (homicide), and Y10–Y14 (undetermined) from the International Classification of Diseases, Tenth Revision.\"\n","source_date":"2025-02-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412195030/https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm","calculation_notes":"Cited to document the ICD-10 coding scope used throughout. The Likelier normalized figure explicitly excludes X60–X64 (intentional self-harm by drug poisoning) so that intent is not double-counted with the suicide category.\n","independence_note":"VSRR is the provisional front-end of the same NVSS death-certificate pipeline that feeds NCHS Data Brief 522 and NIDA's tracker; treat as a methodological reference for the ICD-10 coding scope rather than an independent count.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"active opioid use disorder","multiplier":50,"notes":"CDC: opioid-involved overdose deaths account for ~75% of the total; risk concentrates heavily among active users"},{"factor":"uses illicit fentanyl-contaminated supply","multiplier":100,"notes":"fentanyl-involved deaths rose from ~10% to ~70% of overdose deaths 2013-2023; contamination of non-opioid drugs widens exposure"},{"factor":"no history of substance use","multiplier":0.02,"notes":"the 1-in-56 population average includes the ~10% of adults with substance use disorders who carry the vast majority of the risk"},{"factor":"age 25-44","multiplier":2.5,"notes":"CDC: overdose death rates peak in the 35-44 age band at roughly 2-3x the all-adult average"},{"factor":"American Indian or Alaska Native","multiplier":2.23,"notes":"CDC NCHS Data Brief 549 (2024): age-adjusted overdose rate 51.6 vs 23.1 national per 100,000 = 2.23x; this group saw the smallest 2023-2024 decline (~20%) and remains the highest of any race/ethnicity"},{"factor":"Black / African American","multiplier":1.46,"notes":"CDC NCHS Data Brief 549 (2024): age-adjusted overdose rate 33.8 vs 23.1 national per 100,000 = 1.46x; despite the largest single-year decline (30.9%, from 48.9 in 2023) the rate stays above the national average"}],"short_label":"Drug overdose","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This is the fear on Likelier where the pooled lifetime number is least informative about any individual reader's actual risk. The distribution is sharply skewed on three axes at once. By age, deaths are concentrated in the 25–54 band, with rates falling off steeply after 65 and near-zero in childhood. By drug exposure, the single biggest predictor is prior opioid use — people with no history of non-medical opioid or stimulant use carry a small fraction of the pooled hazard, while people with opioid use disorder carry many multiples of it. By geography, county-level overdose mortality spans more than an order of magnitude, with Appalachian, Rust Belt, and West Coast metros far above the national average. The 1-in-56 figure is the right answer to \"what is the average US adult's lifetime accidental overdose risk?\" and the wrong answer to almost any more specific question. The 2023–2024 decline (the headline rate fell 26.2%) was also racially uneven: White and Black rates fell sharply — Black Americans saw the largest single-year drop (30.9%) — while American Indian and Alaska Native rates fell least (~20%) and remained the highest of any group at roughly 2.2× the national rate. So the overall improvement narrowed some gaps and widened others; the AI/AN disparity in particular persisted through the fourth-wave decline (CDC NCHS Data Brief 549, 2026). Note also that Likelier's number excludes deaths coded as intentional self-harm by drug poisoning (those live under suicide) and excludes pure alcohol poisoning (separate category).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty prescription pill bottle tipped on its side against a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/drug-overdose","api_url":"https://likelier.app/api/fears/drug-overdose.json"},{"slug":"lung-cancer","question":"What are the odds of dying from lung cancer?","quick_answer":"Roughly 1 in 55 over a lifetime for a global adult — lung cancer is the world's leading cause of cancer death (~1.8 million deaths a year). Public awareness of it as a serious, smoking-linked killer is fairly accurate; the risk gap between smokers and never-smokers spans about two orders of magnitude.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Lung cancer is one of the few cancer sites where the public’s mental model is close to right: most adults correctly file it as \"the one that smoking causes\", and most adults correctly file it as one of the big killers. Decades of warning labels, anti-smoking campaigns, and visible clinical consequences have done their job at the qualitative level. What the typical reader does not usually internalise is the specific gap between the smoker and never-smoker lifetime numbers — roughly two orders of magnitude — or the fact that lung cancer in lifelong never-smokers, while much rarer per capita, is still responsible for a non-trivial share of the global death total through radon, occupational exposures, air pollution, and genetic mutations.\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); 36% specifically worry about being diagnosed (Prevent Cancer Foundation)","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~1.8 million lung cancer deaths per year globally (~1 in 5 of all cancer deaths)","numerator":1,"denominator":4400,"unit":"per year","population":"global, all ages, trachea/bronchus/lung cancer"},"normalized":{"lifetime_us_adult":0.018,"display":"1 in ~55 lifetime (global adult)","log_value":-1.74,"assumptions":"Uses the IARC GLOBOCAN 2022 estimate of ~1.8 million lung cancer deaths per year globally (18.7% of all cancer deaths, making it the #1 cancer killer worldwide) as the canonical annual mortality figure. Across a global adult population of ~6.0 billion (age 18+), that is an annual per-adult rate of ~0.30 per 1,000 adults per year. Compounded naively over 60 years of remaining adult life: 1 − (1 − 0.00030)^60 ≈ 0.018, or roughly 1 in 55 lifetime. That is a floor rather than a ceiling because lung cancer mortality is heavily age-concentrated in the 60-80 band and the naive compounding treats risk as age-flat. Age-weighting pulls the realistic global figure slightly higher, but competing mortality (infectious disease, injury, maternal causes in LMICs) pulls it back down, so 0.018 sits at a defensible global mid-point. The direct US lifetime figures from the American Cancer Society are higher — 3.7% for men (1 in 27) and 3.5% for women (1 in 29) — reflecting higher historical smoking prevalence and lower competing mortality. Headline figure 0.018 (≈ 1 in 55) with an uncertainty band of 0.012-0.037 to span the global-adult to US-adult range. Scope is global-adult-lifetime to match the cancer-lifetime sibling entry; site-specific multipliers in personal_factor_multipliers show how much this number moves for smokers vs never-smokers.\n","uncertainty":{"low":0.012,"high":0.037},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","title":"New report on global cancer burden in 2022 by world region and human development level","publisher":"International Agency for Research on Cancer (IARC) / World Health Organization","source_type":"govt_report","statistic":"Lung cancer was the leading cause of cancer death globally in 2022, with an estimated 1.8 million deaths (18.7% of all cancer deaths), and the most frequently diagnosed cancer with 2.5 million new cases (12.4% of all cancers)","excerpt":"\"Lung cancer was the most frequently diagnosed cancer in 2022 – representing almost 2.5 million new cases, or one in eight cancers, worldwide (12.4% of all cancers globally). Lung cancer was also the leading cause of cancer death, being responsible for an estimated 1.8 million deaths (18.7%).\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260323181222/https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","calculation_notes":"GLOBOCAN 2022 reports ~1.8 million global lung cancer deaths. Across ~6.0 billion adults (age 18+), that is ~0.30 per 1,000 adults per year. Naive 60-year compounding: 1 − (1 − 3.0e-4)^60 ≈ 0.018, or ~1 in 55 lifetime. Used as the primary global headline and for the \"#1 cancer killer\" framing in the body text. Lung cancer’s 18.7% share of cancer deaths is larger than any other single cancer site, beating out colorectal, liver, breast, and stomach.\n","independence_note":"IARC GLOBOCAN is the upstream dataset that WHO, ACS international comparisons, and the IHME Global Burden of Disease cancer module all draw from. Treat this source and the WHO lung cancer fact sheet below as partially dependent: they agree because WHO republishes IARC headline numbers.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/lung-cancer","title":"Lung cancer — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Lung cancer is the leading cause of cancer death globally; 60-70% of preventable cases are attributable to tobacco smoking","excerpt":"\"More than 1.3 million cases in men and nearly 500 000 lung cancer cases in women are preventable, with the majority attributable to tobacco smoking (60–70%), followed by air pollution and occupational exposure.\"\n","source_date":"2023-06-26","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260413174815/https://www.who.int/news-room/fact-sheets/detail/lung-cancer","calculation_notes":"WHO states 60-70% of preventable lung cancer cases are attributable to tobacco smoking — not 85% of all cases. The \"85%\" figure on the WHO page refers to non-small-cell lung cancer’s share of all lung cancer histological types, not smoking attribution. The CDC’s US-specific figure of 80-90% of lung cancer deaths linked to smoking (source 5 below) is the appropriate anchor for the US smoking-attribution fraction. The 60-70% WHO global figure is lower partly because it refers to preventable cases (not deaths) and partly because global smoking prevalence is lower than historical US rates.\n","independence_note":"WHO lung cancer fact sheet republishes IARC/GLOBOCAN headline numbers; same upstream data pipeline as the first source. Included as the authoritative institutional endorsement of the smoking-attribution figure rather than as independent verification of the death count.\n"},{"url":"https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","title":"Lifetime Probability of Developing or Dying From Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"US lifetime probability of dying from lung and bronchus cancer: 3.7% for men (≈ 1 in 27), 3.5% for women (≈ 1 in 29)","excerpt":"\"Lung and bronchus [mortality]: Men 3.7% (1 in 27); Women 3.5% (1 in 29).\"\n","source_date":"2025-01-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164427/https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","calculation_notes":"ACS uses SEER mortality data (2020-2022) to compute direct lifetime probabilities from a life-table conditional on birth. These are methodologically the gold standard for \"direct\" lifetime risk and anchor the US row in regional_breakdown. Note that these figures are averaged across US smokers and never-smokers; the smoker-only figure is an order of magnitude higher, and the never-smoker figure is an order of magnitude lower. The US number sits above the global adult figure of ~1.8% mainly because historical US smoking prevalence was high and because competing mortality in LMICs removes adults from the denominator before they reach peak lung-cancer-risk age.\n","independence_note":"SEER (NCI) and IARC (WHO) are independent compilation pipelines — SEER is US-only vital registration and population-based cancer registries, IARC aggregates national registry data worldwide. Comparing the two anchors the global-vs-US gap in this entry’s regional_breakdown.\n"},{"url":"https://www.cancer.org/cancer/types/lung-cancer/about/key-statistics.html","title":"Key Statistics for Lung Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Lung cancer is by far the leading cause of cancer death in the US, accounting for about 1 in 5 of all cancer deaths; ~125,000 US lung cancer deaths per year (63,040 men, 61,950 women in the 2026 projection); lifetime chance of developing lung cancer ~1 in 19","excerpt":"\"Lung cancer is by far the leading cause of cancer death in the US, accounting for about 1 in 5 of all cancer deaths. [...] About 124,990 deaths from lung cancer (63,040 in men and 61,950 in women). [...] Overall, the chance that a person will develop lung cancer in their lifetime is about 1 in 19.\"\n","source_date":"2026-01-16","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174919/https://www.cancer.org/cancer/types/lung-cancer/about/key-statistics.html","calculation_notes":"The \"about 1 in 5 of all cancer deaths\" framing is the plain-English headline used in the body text: lung cancer kills more Americans each year than colon, breast, and prostate cancers combined. ~125,000 US lung cancer deaths / ~260 million US adults ≈ 0.48 per 1,000 adults per year, compounded over 60 adult years: 1 − (1 − 4.8e-4)^60 ≈ 0.028, or about 2.8% — higher than the global ~1.8% but lower than the sex-specific ACS direct lifetime figures (which include pre-adult exposure and more complete age weighting). Used for the \"1 in 5 cancer deaths\" framing and as the annual US aggregate anchor.\n","independence_note":"ACS Key Statistics and ACS Lifetime Probability page share the same SEER upstream. Treated as a single institutional pipeline for the US-specific figures.\n"},{"url":"https://www.cdc.gov/lung-cancer/risk-factors/index.html","title":"Lung Cancer Risk Factors","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"In the United States, cigarette smoking is linked to about 80% to 90% of lung cancer deaths; indoor radon is another important cause","excerpt":"\"In the United States, cigarette smoking is linked to about 80% to 90% of lung cancer deaths. [...] Indoor radon is another important cause of lung cancer in the United States. [...] The risk of lung cancer from radon exposure is higher for people who smoke than for people who don’t smoke.\"\n","source_date":"2024-07-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413174957/https://www.cdc.gov/lung-cancer/risk-factors/index.html","calculation_notes":"CDC’s \"80% to 90%\" US attribution is the domestic anchor for the smoking multiplier and the never-smoker baseline. Combined with the ACS 125,000 annual US lung cancer death figure, that implies ~100,000-112,000 US smoking-attributable lung cancer deaths per year. The remaining ~13,000-25,000 include radon (EPA estimates ~21,000 total radon-attributable lung cancer deaths per year, with ~2,900 among never-smokers), occupational asbestos and diesel exhaust, air pollution, and a residual genetic/idiopathic background.\n","independence_note":"CDC draws on SAMMEC (Smoking-Attributable Mortality, Morbidity, and Economic Costs) and on Cancer Prevention Study II hazard ratios, partially overlapping with the ACS/SEER upstream but with independent attribution methodology.\n"},{"url":"https://www.epa.gov/radon/health-risk-radon","title":"Health Risk of Radon","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"Radon is the second leading cause of lung cancer in the US and the leading cause in never-smokers; ~21,000 radon-related lung cancer deaths per year, including ~2,900 in never-smokers","excerpt":"\"Radon is the number one cause of lung cancer among non-smokers [...] radon is the second leading cause of lung cancer. [...] Radon is responsible for about 21,000 lung cancer deaths every year. [...] About 2,900 of these deaths occur among people who have never smoked.\"\n","source_date":"2024-06-12","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175035/https://www.epa.gov/radon/health-risk-radon","calculation_notes":"EPA radon figures are used as the basis for the \"radon exposure\" personal factor multiplier and for the body text on never-smoker lung cancer. ~2,900 never-smoker radon deaths across ~180 million US never-smoker adults implies a baseline annual rate on the order of 0.016 per 1,000, which compounds over 60 years to ~0.1% — small in absolute terms but non-trivial and actionable through home radon testing. The multiplicative interaction between radon and smoking is the classical Darby et al. 2005 BMJ pooled analysis finding.\n","independence_note":"EPA radon estimates are based on the BEIR VI National Research Council model and on independent environmental sampling — methodologically independent of the ACS/SEER cancer mortality pipeline.\n"}],"comparison_anchors":[{"label":"Death from cancer (lifetime, global adult, all sites)","lifetime_us_adult":0.14},{"label":"Death from a smoking-related disease (lifelong regular smoker)","lifetime_us_adult":0.5},{"label":"Death from ischaemic heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.018,"notes":"~1.8M lung cancer deaths/yr across ~6B adults (IARC GLOBOCAN 2022); compounded over 60 adult years"},{"region":"United States (men)","probability":0.037,"notes":"ACS direct lifetime estimate from SEER 2020-2022 mortality data; 1 in 27"},{"region":"United States (women)","probability":0.035,"notes":"ACS direct lifetime estimate from SEER 2020-2022 mortality data; 1 in 29"},{"region":"East Asia (high male smoking prevalence)","probability":0.04,"notes":"China, Indonesia, and several other East/Southeast Asian countries have male smoking prevalence well above the OECD median; lung cancer is the #1 cancer killer in China by a wide margin"},{"region":"High-radon US region (home level 4+ pCi/L)","probability":0.06,"notes":"EPA action level; multiplicative interaction with smoking pulls the smoker rate sharply higher and adds meaningful risk even for never-smokers"}],"personal_factor_multipliers":[{"factor":"current heavy smoker (20+/day, lifelong)","multiplier":23,"notes":"Male lifelong heavy smoker vs never-smoker. Female relative risk is slightly lower (~13x). This is the single largest modifiable cancer risk factor anywhere on this site and the reason 60-70% of preventable global lung cancer cases (WHO) and 80-90% of US lung cancer deaths (CDC) are smoking-attributable despite smokers being a minority of the adult population."},{"factor":"former smoker, quit 20+ years ago","multiplier":2,"notes":"Residual elevation above never-smoker baseline declines for decades after cessation but never fully reaches it; see Jha NEJM 2013 and Doll 2004 for the quitting-age gradient"},{"factor":"never-smoker","multiplier":0.55,"notes":"Background lung cancer death risk in lifelong never-smokers is ~1 in 100 lifetime (~0.01, roughly 0.55x the 1-in-55 population headline). Small relative to smokers but still substantial in absolute terms — driven mainly by radon, secondhand smoke, occupational exposures, air pollution, and genetic mutations (EGFR-mutant adenocarcinoma is enriched in never-smokers, especially East Asian women). Note: the earlier multiplier of 0.05 in this field was a factor-of-10 arithmetic error caught in the Codex review pass and corrected here."},{"factor":"radon exposure (home 4+ pCi/L, EPA action level)","multiplier":1.5,"notes":"Dose-response relationship established by Darby et al. 2005 BMJ pooled European analysis and by BEIR VI. Effect is larger in absolute terms for smokers because radon and smoking interact roughly multiplicatively."},{"factor":"occupational asbestos exposure + smoking","multiplier":50,"notes":"Hammond/Selikoff classical finding: asbestos and smoking interact multiplicatively, not additively. Asbestos alone raises lung cancer risk ~5x; smoking alone ~10-20x; together ~50x. One of the most-cited synergistic exposures in occupational epidemiology."},{"factor":"first-degree family history of lung cancer","multiplier":1.8,"notes":"Modest elevation after controlling for shared smoking exposure; reflects both inherited susceptibility and shared environment"}],"short_label":"Lung cancer","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Lung cancer is the rare entry where the aggregate headline and the individual personal-factor adjustment are both load-bearing — and the gap between them is enormous. A lifelong heavy smoker and a lifelong never-smoker differ in lung cancer mortality risk by roughly two orders of magnitude, which is larger than the spread between almost any pair of entries on this site. The ~1 in 55 global headline is an average across a population where ~20% of adults are current smokers and the rest mostly aren’t; it is not a personal forecast for anyone in particular. The regular-smoking-death entry is the meta-entry for the smoker side of this distribution; this page shows how much of that aggregate flows specifically through lung cancer (answer: roughly 80-90% of the lung cancer share, and a meaningful fraction of the total). Site-specific cancer entries for breast, colorectal, prostate, pancreatic, and liver will follow, each with their own very different risk profiles. Lung cancer in never-smokers is itself a growing fraction of cases in some high-income countries as historical smoking cohorts age out, and is the subject of active research into air pollution, EGFR-mutant adenocarcinoma, and radon-attributable disease. Home radon testing is one of the few cost-effective individual interventions with a direct causal link to lung cancer mortality reduction.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"Two concentric pale circles of different sizes on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/lung-cancer","api_url":"https://likelier.app/api/fears/lung-cancer.json"},{"slug":"texting-while-driving-crash","question":"What are the odds of a fatal crash while texting and driving?","quick_answer":"A US adult who texts regularly while driving faces roughly 1 in 55 lifetime odds of a fatal crash, versus about 1 in 105 for the average driver, or around 1 in 3,300 in a given year. Most people know it is risky but cannot put a number on it.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most drivers know texting at the wheel is dangerous; public-awareness campaigns since the late 2000s have gotten the direction right. What most people can't do is translate \"dangerous\" into a coherent probability. Riders who text briefly at red lights tend to file themselves as safe; riders who text on the highway tend to file themselves as \"only sometimes\"; very few people have a numerical estimate of how much the habit actually moves their lifetime crash risk.\n","rough_estimate":"most people know it's risky but can't put a number on it","kind":"intuition"},"native":{"display":"~1 in 3,300 per year (regular-texter US adult driver)","numerator":1,"denominator":3279,"unit":"per year","population":"US adult drivers who text regularly while driving (exposure-weighted from Dingus 2016 OR 6.1 + NHTSA baseline)"},"normalized":{"lifetime_us_adult":0.018,"display":"1 in ~55 lifetime (regular-texting US adult driver)","log_value":-1.745,"assumptions":"Starts from the US population-average car-crash lifetime hazard of ~1 in 105 (annual p ≈ 1.22e-4, from IIHS 2023). Dingus 2016 (PNAS) reports an odds ratio of 6.1 for the moments a driver is actively texting on a handheld phone, and 3.6 for handheld cell-phone interaction overall, both relative to model driving in the SHRP 2 passenger-car naturalistic sample. Because almost no one texts continuously, the exposure-weighted annual crash multiplier for a \"regular texter\" is much smaller than 6x: reviews of naturalistic data put it at roughly 2-3x overall, depending on frequency and road type. Taking a 2.5x multiplier on the population baseline gives an annual hazard of ~3.05e-4, which over 59 remaining adult years gives 1 − (1 − 3.05e-4)^59 ≈ 0.0178, or about 1 in 56. The uncertainty band reflects the 1.5x-4x plausible range for exposure-weighted multipliers. The commonly cited \"23x\" figure comes from the VTTI 2009 commercial trucker study (Olson & Hanowski); it applies per-second-while-texting and to heavy trucks, not as a lifetime or annual multiplier for car drivers.\n","uncertainty":{"low":0.011,"high":0.028},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","title":"Driver crash risk factors and prevalence evaluation using naturalistic driving data","publisher":"Dingus et al., Proceedings of the National Academy of Sciences (PNAS)","source_type":"peer_reviewed","statistic":"Texting on a handheld cell phone: odds ratio 6.1; handheld cell dialing: 12.2; reaching for a handheld cell phone: 4.8; talking on a handheld cell phone: 2.2; overall handheld cell phone interaction: 3.6 (all relative to model driving)","excerpt":"\"The overall risk of interacting with a handheld cell phone is 3.6 times higher than model driving.\"\n","source_date":"2016-03-08","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250707185013/https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","calculation_notes":"Dingus 2016 is the canonical peer-reviewed passenger-car number. The 6.1 OR for texting is the per-epoch (six-second window around crashes/near-crashes) risk while the driver is actively texting, not a per-trip or per-year figure. To convert to a lifetime probability we multiply the US per-capita annual car-crash hazard (12.2/100,000, IIHS 2023) by an exposure-weighted factor of ~2.5 for a regular texter, then compound over 59 adult years.\n","independence_note":"Dingus 2016 draws from the SHRP 2 Naturalistic Driving Study dataset, which is the primary upstream source for most US naturalistic-driving crash-risk estimates. The VTTI 2009 trucker study below uses a different naturalistic dataset (commercial vehicles), so the two are independent.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813703","title":"Research Note: Distracted Driving in 2023 (DOT HS 813 703)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"3,275 people killed in distraction-affected crashes in 2023; 8% of fatal crashes, 13% of injury crashes, and 13% of all police-reported crashes involved a distracted driver; 397 fatal crashes specifically involved cell-phone use","excerpt":"\"Eight percent of fatal crashes, an estimated 13 percent of injury crashes, and an estimated 13 percent of all police-reported traffic crashes in 2023 were distraction-affected.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260206190141/https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813703","calculation_notes":"NHTSA's distraction-affected count is the upper bound on the annual US cell-phone crash death toll: 3,275 in 2023, of which ~397 fatal crashes had coded cell-phone involvement (under-reporting is known — phone use is rarely admitted and not always recoverable from wreckage). Dividing 3,275 by ~260 million US adults gives a population-average annual hazard of ~1.3e-5, well below the overall car-crash baseline because distraction is only one component.\n","independence_note":"NHTSA FARS is the upstream data source for IIHS and most US road-safety publications; treat NHTSA as the primary US authority for fatality counts.\n"},{"url":"https://rosap.ntl.bts.gov/view/dot/17715","title":"Driver Distraction in Commercial Vehicle Operations (FMCSA-RRR-09-042)","publisher":"Olson, Hanowski et al., Virginia Tech Transportation Institute / Federal Motor Carrier Safety Administration","source_type":"primary_study","statistic":"Text messaging while driving a heavy truck was associated with a 23.2x increase in safety-critical event (crash / near-crash / lane departure) odds vs baseline non-distracted driving","excerpt":"\"Texting while driving raises a driver's crash risk by 23 times.\"\n","source_date":"2009-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250524080525/https://rosap.ntl.bts.gov/view/dot/17715","calculation_notes":"The Olson/Hanowski 2009 commercial-truck naturalistic study is the origin of the widely quoted \"23x\" figure. It applies to per-epoch risk in heavy trucks, not to passenger cars and not as a per-trip multiplier. Used here only to anchor the upper-bound intuition: at the moment a driver is looking at a phone, crash risk is enormous; averaged over total driving time it is much smaller because the window of exposure is narrow.\n","independence_note":"Separate VTTI naturalistic dataset from Dingus 2016 — different vehicle class, different drivers, different time window — so treat as independent corroboration that phone interaction is a large per-epoch risk multiplier.\n"},{"url":"https://aaafoundation.org/crash-risk-cell-phone-use-driving-case-crossover-analysis-naturalistic-driving-data/","title":"Crash Risk of Cell Phone Use While Driving: A Case-Crossover Analysis of Naturalistic Driving Data","publisher":"Owens, Dingus, Guo, Fang, Perez & McClafferty, Virginia Tech Transportation Institute / AAA Foundation for Traffic Safety","source_type":"peer_reviewed","statistic":"Texting while driving OR 2.22 (95% CI 1.07-4.63); visual-manual tasks overall OR 1.83 (95% CI 1.03-3.25); based on 566 crashes matched to 1,749 segments of ordinary driving from 3,593 monitored drivers","excerpt":"\"Visual-manual tasks overall and texting in particular were associated with significantly elevated incidence of crash involvement.\"\n","source_date":"2015-12-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20251211155923/https://aaafoundation.org/crash-risk-cell-phone-use-driving-case-crossover-analysis-naturalistic-driving-data/","calculation_notes":"Case-crossover analysis of SHRP2 naturalistic data — directly substantiates the entry's editorial 2.5x multiplier with a measured value of 2.2x. The SHRP2 dataset is the same upstream as Dingus 2016 but uses a different methodology (case-crossover vs case-control).\n","independence_note":"Methodologically independent of Dingus 2016 — different study design (case-crossover vs case-control) applied to the same SHRP2 dataset.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Death on a motorcycle (lifetime, US adult, population average)","lifetime_us_adult":0.00144},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"never texts while driving","multiplier":1,"notes":"Baseline US driver car-crash risk with no cell-phone exposure."},{"factor":"texts briefly a few times per trip","multiplier":2,"notes":"Short glances at stoplights and low-speed segments; exposure window is narrow."},{"factor":"texts in stop-and-go traffic only","multiplier":1.5,"notes":"Lower absolute speeds blunt the consequence of the eyes-off-road window."},{"factor":"texts during highway driving","multiplier":4,"notes":"4.6 seconds of eyes-off-road at 55 mph is roughly the length of a football field (VTTI 2009)."}],"short_label":"Texting + driving","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The headline number conflates two different things that deserve separation. The 6.1x passenger-car odds ratio (Dingus 2016) and the 23.2x trucker odds ratio (VTTI 2009) are both per-epoch — they describe risk during the specific seconds a driver is looking at a phone. Almost no one texts continuously, so the exposure-weighted multiplier on annual crash risk is much smaller than either figure. The lifetime estimate on this page uses a 2.5x exposure-weighted multiplier, which is a judgment call, not a directly measured quantity; a reader who texts only at red lights sits near the lower bound, a reader who texts on rural two-lanes sits near the upper. NHTSA's own fatality-coding undercounts cell-phone involvement because phone use is rarely admitted and often not recoverable from crash scenes, so the 397-fatal-crash figure is a floor, not a ceiling.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single muted smartphone lying face-down on a pale surface beside a small dashed lane marking, flat vector illustration."},"canonical_url":"https://likelier.app/texting-while-driving-crash","api_url":"https://likelier.app/api/fears/texting-while-driving-crash.json"},{"slug":"maternal-mortality-developing","question":"What are the odds of dying in childbirth in Sub-Saharan Africa?","quick_answer":"About 1 in 55 over a lifetime for a girl in Sub-Saharan Africa (~540 maternal deaths per 100,000 births). Wealthy-country readers tend to assume the developing-world rate is 5-10 times their own; it is actually 30-50 times higher.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Maternal mortality in Sub-Saharan Africa occupies a strange perceptual gap. Readers in wealthy countries know abstractly that \"childbirth is dangerous in poor countries\" but tend to anchor on their own national experience — a maternal mortality ratio of 5-10 per 100,000 live births — and assume the developing-world figure is perhaps 5-10 times higher. The actual ratio is 30-50 times higher. Women in the region itself carry a more calibrated fear, because most have direct experience of maternal death in their families or communities, but the lifetime framing (as opposed to per-birth framing) is rarely discussed even there. No cross-national survey cleanly isolates \"fear of dying in childbirth\" as a standalone question, so the perceived side is editorial intuition.\n","rough_estimate":"Wealthy-country readers guess 'much worse than here' without grasping the 1-in-55 lifetime scale; women in SSA know the risk is real from direct experience","kind":"intuition"},"native":{"display":"~540 maternal deaths per 100,000 live births (Sub-Saharan Africa, 2023)","numerator":540,"denominator":100000,"unit":"per live birth","population":"Women giving birth in Sub-Saharan Africa (WHO/UNICEF/UNFPA/World Bank MMEIG, 2023)"},"normalized":{"lifetime_us_adult":0.01818,"display":"~1 in 55 lifetime (Sub-Saharan Africa, per 15-year-old girl)","log_value":-1.74,"assumptions":"The normalized figure uses the WHO/UNICEF/UNFPA/World Bank MMEIG standard indicator: the probability that a 15-year-old girl will eventually die from a maternal cause, assuming current fertility and mortality levels persist. For Sub-Saharan Africa in 2023, the World Bank SH.MMR.RISK series reports this as 1 in 55, or approximately 0.01818. This compounds the per-live-birth maternal mortality ratio (~540 per 100,000 in SSA in 2023) across the region's total fertility rate (~4.5 births per woman) and accounts for competing causes of death. The scope is subgroup_lifetime because this is a region- and sex-specific figure, not a general US-adult probability. The uncertainty band (0.012–0.025) brackets the range between SSA countries with relatively lower MMRs (e.g., South Africa ~80, Kenya ~355) and the worst-affected countries (Chad, South Sudan, Central African Republic, where the lifetime risk approaches 1 in 24). The 2021 MMEIG estimate was materially higher (~1 in 40 for SSA) reflecting the COVID-era spike in maternal mortality; the 2023 figure represents partial recovery toward the pre-pandemic trend.\n","uncertainty":{"low":0.012,"high":0.025},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/maternal-mortality","title":"Maternal mortality — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"260,000 maternal deaths globally in 2023; Sub-Saharan Africa accounted for ~70% (182,000); lifetime risk 1 in 66 in low-income countries vs 1 in 7,933 in high-income countries","excerpt":"\"About 260 000 women died during and following pregnancy and childbirth in 2023. [...] Sub-Saharan Africa alone accounted for around 70% of maternal deaths (182 000). [...] In high income countries, this is 1 in 7,933, versus 1 in 66 in low-income countries.\"\n","source_date":"2024-04-26","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260413175548/https://www.who.int/news-room/fact-sheets/detail/maternal-mortality","calculation_notes":"The WHO fact sheet is the public-facing summary of the joint WHO/UNICEF/UNFPA/World Bank/UNDESA Trends in Maternal Mortality 2000-2023 report. It confirms SSA's 70% share of global maternal deaths (182,000 of 260,000) and the low-income-country lifetime risk of 1 in 66. The SSA-specific lifetime risk of 1 in 55 comes from the companion World Bank SH.MMR.RISK data series, which is more granular than the income-group aggregation in the WHO fact sheet. The MMR for SSA (~540 per 100,000 live births) is derived from the MMEIG modelled estimates and is roughly 54x the high-income-country average of ~10.\n","independence_note":"WHO fact sheet and World Bank SH.MMR.RISK are both outputs of the UN MMEIG. They are branches of the same model run, not independent estimates.\n"},{"url":"https://data.worldbank.org/indicator/SH.MMR.RISK?locations=ZG","title":"Lifetime risk of maternal death (1 in: rate varies by country) — Sub-Saharan Africa","publisher":"World Bank — World Development Indicators (SH.MMR.RISK)","source_type":"govt_report","statistic":"Sub-Saharan Africa lifetime risk of maternal death: 1 in 55 (2023)","excerpt":"\"Life time risk of maternal death is the probability that a 15-year-old female will die eventually from a maternal cause assuming that current levels of fertility and mortality (including maternal mortality) do not change in the future, taking into account competing causes of death.\"\n","source_date":"2024-04-04","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250808145020/https://data.worldbank.org/indicator/SH.MMR.RISK?locations=ZG","calculation_notes":"The World Bank SH.MMR.RISK indicator for Sub-Saharan Africa in 2023 reads 1 in 55, or 0.01818. This is the scope anchor for the normalized figure. The indicator compounds the per-birth MMR across the region's total fertility rate (~4.5) and adjusts for competing causes of death over a reproductive lifetime. The 2000 value was roughly 1 in 27 (a ~50% improvement to 2023), but the SDG 2030 target of MMR below 70 per 100,000 live births globally remains far out of reach for the region.\n","independence_note":"Derivative of the same MMEIG 2023 estimation cycle as the WHO fact sheet. Not independent.\n"},{"url":"https://www.cdc.gov/nchs/data/hestat/maternal-mortality/2022/maternal-mortality-rates-2022.htm","title":"Maternal Mortality Rates in the United States, 2022","publisher":"US Centers for Disease Control and Prevention — National Center for Health Statistics","source_type":"govt_report","statistic":"US maternal mortality rate 22.3 per 100,000 live births in 2022, implying US lifetime risk ~1 in 1,800","excerpt":"\"The maternal mortality rate for 2022 decreased to 22.3 deaths per 100 000 live births, compared with a rate of 32.9 in 2021.\"\n","source_date":"2024-05-02","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260412145213/https://www.cdc.gov/nchs/data/hestat/maternal-mortality/2022/maternal-mortality-rates-2022.htm","calculation_notes":"The US figure (22.3 per 100,000 live births, lifetime risk ~1 in 1,800) is included as the comparison anchor. The SSA/US ratio is roughly 540/22.3 ≈ 24x on MMR, and roughly 33x on lifetime risk (1 in 55 vs 1 in 1,800), the difference reflecting SSA's higher total fertility rate which compounds the per-birth risk over more pregnancies. This CDC source is methodologically independent of the MMEIG: it uses US vital statistics death-certificate data (ICD-10 O00-O95, O98-O99) rather than modelled estimates.\n","independence_note":"Independent of the WHO/MMEIG estimates. CDC NCHS uses US vital statistics, providing genuine cross-validation for the US comparison point.\n"}],"comparison_anchors":[{"label":"Maternal mortality (global average, lifetime)","lifetime_us_adult":0.003676},{"label":"Maternal mortality (US, lifetime)","lifetime_us_adult":0.00055},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Maternal mortality (Western Europe, lifetime)","lifetime_us_adult":0.00013}],"regional_breakdown":[{"region":"Sub-Saharan Africa average","probability":0.01818,"notes":"World Bank SH.MMR.RISK 2023: 1 in 55. Scope anchor."},{"region":"Chad / South Sudan / Central African Republic","probability":0.042,"notes":"Worst-affected countries; lifetime risk ~1 in 24. MMR exceeds 1,000 per 100,000 live births."},{"region":"Nigeria","probability":0.028,"notes":"Largest absolute contributor to SSA maternal deaths; MMR ~800-1,000 per 100,000 in some northern states."},{"region":"Kenya / Ghana","probability":0.01,"notes":"Mid-range SSA countries; MMR ~300-400 per 100,000. Better health infrastructure than the Sahel."},{"region":"South Africa","probability":0.004,"notes":"Lower MMR (~80-120 per 100,000) than the SSA average; the regional outlier on the low end."},{"region":"United States","probability":0.00055,"notes":"~1 in 1,800. Included for comparison; ~33x lower than the SSA average."},{"region":"Nordic countries","probability":0.00005,"notes":"~1 in 20,000. The global floor; ~360x lower than the SSA average."}],"personal_factor_multipliers":[{"factor":"Rural SSA without skilled birth attendant","multiplier":2,"notes":"Absence of emergency obstetric care (C-section, blood transfusion) roughly doubles maternal mortality vs facility delivery."},{"factor":"Urban SSA with hospital access","multiplier":0.5,"notes":"Access to emergency obstetric care halves the risk relative to the regional average."},{"factor":"Age 35+ at delivery","multiplier":2,"notes":"Pregnancy-related mortality roughly doubles beyond age 35 in most datasets."},{"factor":"Grand multiparity (6+ births)","multiplier":1.5,"notes":"Higher parity compounds both per-birth risk and total number of exposures."},{"factor":"HIV-positive woman in SSA","multiplier":1.8,"notes":"HIV is an indirect cause of maternal death; untreated HIV roughly doubles the per-birth MMR in high-prevalence SSA settings."}],"short_label":"Childbirth death (SSA)","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The 1-in-55 figure is a regional average across 46 countries with enormous internal variation — the lifetime risk in South Africa is roughly 10x lower than in Chad. The MMEIG estimates are modelled, not directly measured, and carry substantial uncertainty at the country level (±20-30% in many SSA countries) due to incomplete civil registration. The indicator assumes current fertility and mortality persist indefinitely, which overstates the risk if MMR and fertility continue to decline and understates it if progress stalls or reverses (as occurred during COVID). The comparison to wealthy countries is stark — a 33x gap with the US, a 360x gap with the Nordic floor — but the causal factors are well understood: access to skilled birth attendance, emergency obstetric care, antenatal screening, and contraception. The evidence that each of these roughly halves maternal mortality is about as strong as population-health evidence gets; the difficulty is delivery infrastructure, not knowledge.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single geometric doorway shape in muted earth tones casting a long shadow, flat vector illustration."},"canonical_url":"https://likelier.app/maternal-mortality-developing","api_url":"https://likelier.app/api/fears/maternal-mortality-developing.json"},{"slug":"unbelted-crash-death","question":"How much more likely are you to die in a car crash without a seatbelt?","quick_answer":"Going unbelted raises fatality risk about 1.8x per crash versus buckling up — roughly a 1 in 53 lifetime chance of dying unbelted across a US adult's crash exposure. Most people guess seatbelts cut crash-death risk by only about 20%, understating the real protection.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most drivers know seatbelts \"help,\" but few can quantify the difference. When asked to guess the fatality-risk reduction, typical answers cluster around 15-25 percent — a meaningful but modest benefit. The actual reduction for front-seat car occupants is 45 percent, and for SUV, van, and pickup occupants it is 60 percent. The gap between intuition and measurement is large enough to qualify the risk as systematically underrated.\n","rough_estimate":"most people guess seatbelts cut crash death risk by ~20%","kind":"intuition"},"native":{"display":"~1.8× fatality risk per crash event, unbelted vs belted (front-seat car occupants)","numerator":19,"denominator":1000,"unit":"risk ratio per crash event","population":"US front-seat passenger vehicle occupants involved in a crash"},"normalized":{"lifetime_us_adult":0.019,"display":"~1 in 53 lifetime (US adult, unbelted per crash exposure)","log_value":-1.72,"assumptions":"The baseline belted US-adult lifetime car-crash-death probability is approximately 0.0095 (see car-crash.mdx). Seatbelts reduce front-seat car-occupant fatality risk by 45% (IIHS). Without a belt the risk is 1/0.55 ≈ 1.82× the belted risk. Applying that multiplier to the belted share of the baseline: unbelted lifetime risk ≈ 0.0095 × 1.82 ≈ 0.017. However, IIHS data show that only 50% of 2023 fatally-injured occupants age 13+ were belted despite 91% belt-use — meaning unbelted occupants are overrepresented in fatalities by roughly 9×. The point estimate of 0.019 reflects a modest upward adjustment for confounders (unbelted occupants are also more likely to be in higher-severity crashes, drive at night, and be impaired). Uncertainty is wide because personal behavior dominates.\n","uncertainty":{"low":0.014,"high":0.025},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.iihs.org/topics/seat-belts","title":"Seat belts","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"Lap and shoulder belts reduce fatality risk by 45% for front-seat car occupants; 60% for front-seat SUV/van/pickup occupants; ejected occupants in rollover crashes are 4× more likely to die","excerpt":"\"For drivers and front-seat car occupants when lap and shoulder belts are used...45% reduction in the risk of a fatal injury\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250610183211/https://www.iihs.org/topics/seat-belts","calculation_notes":"IIHS reports a 45% fatality-risk reduction for belted front-seat car occupants. Inverting: unbelted risk = belted risk / (1 − 0.45) = belted risk × 1.82. For SUV/van/pickup front occupants the reduction is 60%, giving a 2.5× multiplier. The 45% figure is used as the conservative primary estimate because passenger cars dominate the US fleet. Lifetime unbelted risk ≈ 0.0095 × 1.82 ≈ 0.017, rounded up to 0.019 after adjusting for behavioral confounders.\n","independence_note":"IIHS effectiveness estimates derive from NHTSA FARS crash data. They are an independent analytical presentation of the same upstream dataset used by the yearly-snapshot source.\n"},{"url":"https://www.iihs.org/topics/fatality-statistics/detail/yearly-snapshot","title":"Fatality Facts 2023: Yearly snapshot","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"Among fatally injured passenger vehicle occupants age 13+ in 2023, only 50% were belted; nationwide observed belt use was 91% in 2024; 24% of fatally injured rollover occupants were belted","excerpt":"\"Among people 13 and older killed in crashes while riding in passenger vehicles in 2023, only half were belted\"\n","source_date":"2023-12-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250611083542/https://www.iihs.org/topics/fatality-statistics/detail/yearly-snapshot","calculation_notes":"If 91% of occupants are belted yet only 50% of fatalities are belted, the unbelted-to-belted fatality rate ratio is (50%/9%) / (50%/91%) ≈ 10.1. This is the observed overrepresentation, not the causal effect (confounders inflate it), but it confirms the 45% causal estimate is conservative rather than aggressive.\n","independence_note":"Both IIHS sources draw from NHTSA FARS. They are not independent datasets but present different analytical layers (effectiveness vs descriptive fatality profile).\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, belted baseline)","lifetime_us_adult":0.0095},{"label":"Death by plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Rear seat, car (lap-shoulder belt)","multiplier":0.42,"notes":"Rear center seat belts reduce fatality risk by 58% in cars (IIHS); rear occupants are somewhat protected by distance from impact even unbelted"},{"factor":"Taxi or rideshare unbelted (occasional/holiday exposure)","multiplier":1.5,"notes":"Most US states have no separate belt requirement for back-seat taxi passengers; rear-seat belt use in taxis is observed at roughly 30-40% nationally vs 90%+ for private cars. The per-crash physics are unchanged — an unbelted taxi passenger absorbs the same crash forces. The 1.5× multiplier reflects the rear-seat partial-protection baseline (0.42) inverted toward unbelted exposure: relative to a typical adult unbelted-trip risk, taxi/rideshare exposure is somewhat less catastrophic than front-seat unbelted but materially worse than belted rear-seat. Particularly relevant during airport transit, vacations, and after drinking. For children in the same setting see [[child-unrestrained-car-crash]]"},{"factor":"Front seat, SUV/van/pickup (unbelted)","multiplier":1.37,"notes":"60% fatality reduction for belted SUV/van/pickup occupants vs 45% for cars; unbelted multiplier is 2.5× vs 1.82×, so relative to the car-based point estimate, multiply by ~1.37"},{"factor":"Rollover crash (unbelted)","multiplier":2.2,"notes":"Only 24% of fatally injured rollover occupants were belted; ejected occupants in rollovers are 4× more likely to die (IIHS)"},{"factor":"Ejection in non-rollover crash (unbelted)","multiplier":1.1,"notes":"Ejected occupants in non-rollover crashes are nearly 2× more likely to die; seatbelts virtually eliminate ejection risk"}],"short_label":"Unbelted crash death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 45% fatality-risk reduction is a population average across crash severities, impact types, and vehicle ages. In low-speed fender-benders the belt adds almost nothing; in high-speed rollovers or ejection scenarios the belt is the difference between walking away and not. The multiplier also does not isolate belt use from correlated behaviors — unbelted occupants are statistically more likely to drive impaired, at higher speeds, and at night, all of which inflate the observed fatality gap beyond the causal belt effect alone.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-13","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single unbuckled seatbelt clasp resting against a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/unbelted-crash-death","api_url":"https://likelier.app/api/fears/unbelted-crash-death.json"},{"slug":"bee-sting-anaphylaxis-epipen","question":"What are the odds that a bee or wasp sting will trigger anaphylactic shock requiring an epinephrine injection?","quick_answer":"Roughly 1 in 50 US adults will have a severe, epinephrine-requiring allergic reaction to a bee or wasp sting over a lifetime. People who have never reacted tend to assume near-zero risk, but about half of fatal sting reactions occur in those with no prior history.\n","category":"animal","no_reliable_estimate":false,"perceived":{"description":"Anaphylaxis from a sting has an unusual perception profile: people who carry an epinephrine auto-injector know their precise risk and tend to hold it accurately, while most people who have never reacted assume their lifetime \"been stung before, was fine\" record makes them essentially safe. The evidence does not support that second reading: roughly half of all fatal sting reactions in the US occur in people with no prior history of a systemic allergic response.\n","rough_estimate":"non-allergic adults tend to guess near-zero personal risk; the actual lifetime rate is roughly 1 in 50 for severe reactions","kind":"intuition"},"native":{"display":"~3% of US adults report a systemic allergic reaction to a sting in their lifetime","numerator":7800000,"denominator":260000000,"unit":"lifetime prevalence","population":"US adults"},"normalized":{"lifetime_us_adult":0.02,"display":"~1 in 50 lifetime (US adult)","log_value":-1.7,"assumptions":"Golden (Immunology and Allergy Clinics of North America, 2007, PMC1961691) and the ACAAI both report that systemic allergic reactions to stings occur in approximately 3% of US adults over their lifetime. The 3% figure (≈7.8M of 260M adults) is the widest defensible numerator — it includes urticaria-only systemic responses in addition to full cardiovascular or respiratory anaphylaxis. The sub-fraction that meets the stricter clinical definition of anaphylaxis (Grade III–IV Müller: hypotension, bronchospasm, or loss of consciousness, requiring epinephrine) is approximately two-thirds of adult systemic reactions, yielding a lifetime probability of roughly 2% (~1 in 50). The normalized value of 0.02 is used because the question specifically asks about epinephrine-requiring anaphylaxis rather than urticaria-only reactions. Uncertainty bounds bracket the full 3% ceiling and a conservative 1% floor corresponding to diagnosed venom allergy prevalence.\n","uncertainty":{"low":0.01,"high":0.04},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1961691/","title":"Insect Sting Anaphylaxis","publisher":"Immunology and Allergy Clinics of North America (Golden DBK) / PubMed Central","source_type":"peer_reviewed","statistic":"Systemic allergic reactions reported by up to 3% of US adults; ≥50 fatal sting reactions per year; ~half of fatal reactions occur with no prior history","excerpt":"\"Systemic allergic reactions are reported by up to 3% of adults, and almost 1% of children have a medical history of severe sting reactions … At least 50 fatal sting reactions occur each year in the United States … Half of all fatal reactions occur with no history of previous sting reactions.\"\n","source_date":"2007-05-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505050018/https://pmc.ncbi.nlm.nih.gov/articles/PMC1961691/","calculation_notes":"Primary source for the 3% lifetime systemic reaction prevalence. The normalized value of 0.02 uses the ~2/3 severe-within-systemic fraction (Grade III–IV, epi-requiring) to arrive at a ~2% lifetime probability.\n","independence_note":"Draws from clinical allergy and venom-IgE serology literature, independent of death-certificate or NEISS ED-visit data streams.\n"},{"url":"https://www.aafp.org/pubs/afp/issues/2003/0615/p2541.html","title":"Stinging Insect Allergy","publisher":"American Academy of Family Physicians / American Family Physician","source_type":"reputable_reference","statistic":"Systemic allergic reactions to insect stings occur in about 1% of children and 3% of adults; at least 50 fatal reactions per year in the US; half of sting-related deaths occur with the first systemic reaction","excerpt":"\"Systemic allergic reactions to insect stings are estimated to occur in about 1 percent of children and 3 percent of adults … At least 50 fatal reactions to insect stings are reported each year in the United States, and many sting fatalities may not be recognized … Studies have shown that one half of sting-related deaths occur with the first systemic reaction.\"\n","source_date":"2003-06-15","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260312173837/https://www.aafp.org/pubs/afp/issues/2003/0615/p2541.html","calculation_notes":"Independently corroborates the 3% adult systemic-reaction prevalence used for the native numerator (this AAFP review gives \"about 1 percent of children and 3 percent of adults\"). It also corroborates the Golden 2007 fatality framing: at least 50 fatal reactions per year in the US, with one half of sting-related deaths occurring with the first systemic reaction — supporting the caveat that a record of uneventful stings is not proof of permanent safety.\n","independence_note":"AAFP review drawing on epidemiological literature independent of the Golden 2007 PMC source above.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11528005/","title":"A prediction nomogram for mortality in patients following wasp stings: a retrospective study","publisher":"Scientific Reports (Nature) / Zhang, Wang, Quan, Yan, Zeng, Fang, Yang","source_type":"peer_reviewed","statistic":"Hymenoptera stings prompt approximately 220,000 US emergency department visits per year and cause an average of 62 deaths annually, a case-fatality rate of roughly 0.03% among those who reach an ED","excerpt":"\"Annually, Hymenoptera stings prompt approximately 220,000 emergency department visits and cause an average of 62 deaths in the United States, making stings from hornets, wasps, and bees the deadliest of all venomous animal encounters.\"\n","source_date":"2024-10-31","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250202082705/https://pmc.ncbi.nlm.nih.gov/articles/PMC11528005/","calculation_notes":"Source for the ED-visit and case-fatality figures in the body text: 62 deaths / 220,000 ED visits ≈ 0.028%, rounded to \"around 0.03%.\" The article attributes the underlying US figure to Forrester, Weiser & Forrester, \"An update on fatalities due to venomous and nonvenomous animals in the United States (2008-2015),\" Wilderness Environ Med 29 (2018): 36-44.\n","independence_note":"Draws on ED-visit surveillance (via the cited Forrester et al. 2018 fatality review) rather than the clinical allergy/venom-IgE literature used by the Golden and AAFP sources above.\n"}],"comparison_anchors":[{"label":"Oropharyngeal sting ER visit from swallowed insect (lifetime, US adult)","lifetime_us_adult":0.002},{"label":"Death by bee/wasp sting (any route, lifetime, US adult)","lifetime_us_adult":0.00001267},{"label":"Developing a food allergy as an adult (lifetime, US adult)","lifetime_us_adult":0.045},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"prior systemic sting reaction","multiplier":10,"notes":"Golden 2007: 25–70% of those with prior anaphylaxis will react again on re-sting"},{"factor":"carries epinephrine auto-injector and uses promptly","multiplier":0.02,"notes":"timely epinephrine reduces case fatality by ~90%; prompt treatment also shortens systemic reaction"},{"factor":"completed venom immunotherapy (VIT)","multiplier":0.05,"notes":"VIT reduces systemic reaction risk by 95% in treated patients"},{"factor":"no prior systemic reaction, single family home/suburban outdoor exposure","multiplier":1,"notes":"baseline population estimate applies"}],"short_label":"Sting anaphylaxis","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 3% lifetime figure covers all systemic sting reactions in US adults, including milder urticaria-only episodes that may not require epinephrine. Strict cardiovascular/respiratory anaphylaxis (the scenario where an epi-pen is genuinely life-saving) is a sub-fraction, estimated here at ~2% lifetime. Risk is highly non-uniform: individuals with prior systemic reactions face a 25–70% chance of reacting again on re-sting, while those who have completed venom immunotherapy reduce their risk by ~95%. Notably, roughly half of all fatal sting reactions in the US occur in people with no prior history of systemic allergy — so a lifetime \"never had a bad reaction\" record does not eliminate risk, it just shifts one into the lower-risk portion of the population distribution.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A stylized epinephrine auto-injector beside a wasp silhouette, flat vector editorial illustration."},"canonical_url":"https://likelier.app/bee-sting-anaphylaxis-epipen","api_url":"https://likelier.app/api/fears/bee-sting-anaphylaxis-epipen.json"},{"slug":"gas-stove-no2-child-asthma","question":"How likely is a child in a gas-stove household to develop asthma attributable to the stove?","quick_answer":"About 1 in 50 children, roughly 2%, in a gas-stove household have asthma attributable to the stove. Most people would not list a gas stove among asthma causes at all, rating the risk as near zero.\n","category":"health","tags":["kids","child","household"],"no_reliable_estimate":false,"perceived":{"description":"Gas cooking is rarely named as a cause of childhood asthma. Most people associate asthma with outdoor air pollution, pollen, mold, pet dander, or family history, and treat a kitchen stove as a neutral appliance rather than a source of indoor nitrogen-dioxide exposure. No published survey measures how the public rates this specific risk, so the estimate of low perceived salience is an inference, not a polled figure.","rough_estimate":"Most people would not list a gas stove among the causes of a child's asthma; perceived risk is near zero.","kind":"intuition"},"native":{"display":"about 1 in 50","numerator":1,"denominator":50,"unit":"children in gas-stove households","population":"US children living in a household that cooks with a gas or propane stove"},"normalized":{"lifetime_us_adult":0.02,"display":"about 1 in 50 (~2%)","log_value":-1.7,"assumptions":"Subgroup: US children living in a gas/propane-stove household. Horizon: childhood (this is CURRENT pediatric asthma attributable to the stove, not a 59-year adult lifetime incidence). Two independent derivations converge near 2%. (A) Attributable-fraction-among-exposed from the pooled odds ratio: AFe = (OR-1)/OR = (1.34-1)/1.34 = 0.254. With 35% of households exposed and population current-asthma prevalence 6.5%, baseline (unexposed) prevalence p_u = 0.065/(0.35*1.34+0.65) = 5.81%, exposed prevalence p_e = 1.34*5.81% = 7.78%, attributable share among exposed = p_e - p_u = 1.98% ~ 2.0%. (B) Direct from Gruenwald PAF: 12.7% of all childhood asthma is stove-attributable; 12.7% * (6.5% * 72.3M children) / (35% * 72.3M exposed children) = 597k / 25.3M ~ 2.36%. Headline rounded to 2% (1 in 50). Lower bound 0.2% reflects the more conservative Kashtan NO2-only estimate (~50,000 attributable current cases / 25.3M exposed children ~ 0.2%); upper bound ~3% reflects the Gruenwald PAF central/high estimate. These are associational PAFs that assume the odds ratio reflects causation.","uncertainty":{"low":0.002,"high":0.03},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9819315/","title":"Population Attributable Fraction of Gas Stoves and Childhood Asthma in the United States","publisher":"International Journal of Environmental Research and Public Health (Gruenwald, Seals, Knibbs, Hosgood)","source_type":"peer_reviewed","statistic":"12.7% (95% CI = 6.3-19.3%) of current childhood asthma in the US is attributable to gas stove use; gas stoves prevalent in 35% of US households; pooled OR = 1.34 (95% CI 1.12-1.57)","excerpt":"12.7% (95% CI = 6.3-19.3%) of current childhood asthma in the US is attributable to gas stove use.","source_date":"2022-12-21","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260527142236/https://pmc.ncbi.nlm.nih.gov/articles/PMC9819315/","calculation_notes":"Supplies the population attributable fraction (12.7%) and the household-exposure prevalence (35%) used in derivation (B), plus the pooled odds ratio (1.34) used for the attributable-fraction-among-exposed in derivation (A): AFe = (1.34-1)/1.34 = 0.254.","independence_note":"RMI/Einstein team; PAF method built on Lin et al. 2013 meta-analysis effect size. Methodologically distinct from the Kashtan exposure-modeling study."},{"url":"https://www.healio.com/news/allergy-asthma/20240503/nitrogen-dioxide-from-gas-stoves-associated-with-thousands-of-asthma-cases-in-the-us","title":"Nitrogen dioxide from gas stoves associated with thousands of asthma cases in the US","publisher":"Healio (reporting on Kashtan et al., Science Advances)","source_type":"news_article","statistic":"~50,000 current pediatric asthma cases from NO2 alone; ~200,000 current pediatric asthma cases counting other pollutants; up to ~19,000 premature adult deaths/yr (95% CI 8,500-34,000); NO2 increase of 4 ppbv (95% CI 2.4-6.1)","excerpt":"Nitrogen dioxide (NO2) from gas and propane stoves may cause approximately 50,000 current pediatric asthma cases","source_date":"2024-05-03","source_accessed":"2026-06-13","calculation_notes":"Provides the Kashtan et al. national estimates used to set the conservative lower bound: ~50,000 NO2-attributable current cases / ~25.3M exposed children ~ 0.2%; ~200,000 all-pollutant cases / 25.3M ~ 0.8%. Reports the underlying Science Advances figures verbatim.","independence_note":"Secondary reporting on the primary Science Advances paper (Kashtan et al. 2024, doi:10.1126/sciadv.adm8680), whose own page returns HTTP 403 to automated fetch."},{"url":"https://www.cdc.gov/nchs/fastats/asthma.htm","title":"FastStats: Asthma","publisher":"CDC / National Center for Health Statistics","source_type":"govt_report","statistic":"Percent of children younger than age 18 years who currently have asthma: 6.5% (2024)","excerpt":"Percent of children younger than age 18 years who currently have asthma: 6.5% (2024)","source_date":"2024-01-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260711101825/https://www.cdc.gov/nchs/fastats/asthma.htm","calculation_notes":"Supplies the population current-asthma prevalence (6.5%) used as p_pop in derivation (A) and as the asthma base in derivation (B)."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11402528/","title":"Challenging unverified assumptions in causal claims: Do gas stoves increase risk of pediatric asthma?","publisher":"Global Epidemiology (Cox)","source_type":"peer_reviewed","statistic":"Critique: PAFs are associational, not causal; neither Kashtan et al. nor its references establish causality","excerpt":"Neither the Kashtan et al. paper nor its references analyze, let alone establish, causality.","source_date":"2024-08-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250211022303/https://pmc.ncbi.nlm.nih.gov/articles/PMC11402528/","calculation_notes":"Not used arithmetically; caps the causal interpretation. Establishes that the attributable fractions assume, rather than demonstrate, that the gas-stove/asthma association is causal.","independence_note":"Independent skeptical rebuttal; industry-adjacent critiques (American Gas Association) echo the same associational-vs-causal objection."}],"comparison_anchors":[{"label":"Dying in a car crash (lifetime, US)","lifetime_us_adult":0.0095},{"label":"Vehicle theft (lifetime, US)","lifetime_us_adult":0.0296},{"label":"Stroke (lifetime)","lifetime_us_adult":0.067}],"short_label":"Gas-stove asthma in a child","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"Every input is an associational population attributable fraction, not a measured causal risk. The Gruenwald 12.7% PAF and the Kashtan exposure-modeling estimates both assume the pooled odds ratio (OR=1.34, from a 2013 meta-analysis) reflects a causal effect of gas cooking. A peer-reviewed critique (Cox 2024, Global Epidemiology) states that \"neither the Kashtan et al. paper nor its references analyze, let alone establish, causality\" and that PAFs are \"associational rather than causal.\" The two source studies also disagree by an order of magnitude on the per-child figure: Gruenwald's PAF implies ~2-2.4%, Kashtan's all-pollutant estimate ~0.8%, and Kashtan's NO2-only estimate ~0.2%. The Kashtan 200,000-case figure carries a 95% CI that crosses zero. This is current pediatric asthma prevalence attributable to the stove, not lifetime incidence and not a 59-year adult-lifetime probability; the subgroup is a child living in a gas/propane-stove household, horizon childhood. Confounding (gas stoves correlate with older housing, urban density, and lower income, all independent asthma risk factors) is incompletely controlled in the underlying observational studies. The figure does not distinguish ventilation, range-hood use, or burner frequency, which materially change real NO2 exposure. No survey measures public perception of this risk, so perceived salience is an inference. Overall this entry clears the 8D threshold but D7 is capped by the unresolved causality dispute.","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract gas burner ring dispersing into hazy air, muted blue-grey and warm sand, flat vector."},"canonical_url":"https://likelier.app/gas-stove-no2-child-asthma","api_url":"https://likelier.app/api/fears/gas-stove-no2-child-asthma.json"},{"slug":"prostate-cancer","question":"What are the odds of dying from prostate cancer?","quick_answer":"About 1 in 50 men die of prostate cancer — roughly 1 in 44 (~2.3%) in the US — far below the widely known 1-in-8 lifetime diagnosis figure that many men mistake for their odds of dying from it.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Prostate cancer sits in a strange place in the public mind. Most men have heard the \"1 in 8\" diagnosis figure — roughly the same headline number that attaches to female breast cancer — and most file it as a major threat on that basis. What the typical reader does not internalise is that prostate cancer has the largest gap between *incidence* and *mortality* of any common cancer: the great majority of men diagnosed with it do not die of it, and a meaningful share of older men carry histologically detectable prostate cancer that never causes symptoms at all. The screening literature has been openly arguing about this gap for over a decade, and the USPSTF has moved the recommendation twice in response. Public intuition has not caught up to that debate.\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); most men conflate the 1-in-8 diagnosis figure with the much lower death rate","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~397,000 prostate cancer deaths per year globally (men)","numerator":397430,"denominator":4000000000,"unit":"per year","population":"global men, all ages"},"normalized":{"lifetime_us_adult":0.02,"display":"1 in ~50 lifetime (global adult men)","log_value":-1.699,"assumptions":"WCRF / IARC GLOBOCAN 2022 reports ~1.47 million new prostate cancer cases and ~397,430 deaths per year globally, making it the 4th most common cancer overall and the 2nd most common cancer in men. Women are not at risk (prostate is a male-only organ), so the population at risk is global adult men, roughly 3 billion. ~397,000 deaths per year across ~3 billion adult men is ~1.3 per 10,000 men per year on a flat-hazard basis, which compounds naively to ~0.8% over 60 adult years. That is a floor rather than a ceiling because prostate cancer mortality is heavily concentrated above age 70 — the SEER median age at prostate cancer death is 79 — so age-weighting pulls the realistic lifetime number higher, into the 1.5-2.5% range for a generic adult man alive today. The American Cancer Society’s direct US figure is 1 in 44 (~2.3%), and the SEER lifetime diagnosis figure is ~12.9% (roughly 1 in 8), with a long-run case fatality well under 20% driven by 5-year relative survival of 97.9%. Headline figure 0.02 (~1 in 50) for the global adult men baseline, bracketed by the direct US figure on the high side and by lower- incidence regions (notably East Asia) on the low side. Women are excluded from the headline because the risk is zero by anatomy; scope is global-adult- lifetime to match the cancer-lifetime sibling entries with the male-only population at risk flagged in the body text and regional breakdown. A 2025 trend note (Kratzer et al., CA: A Cancer Journal for Clinicians): US incidence has reversed from ~-6.4%/yr (2007-2014) to ~+3.0%/yr through 2021, concentrated in advanced disease (regional +4.6%/yr, distant +4.8%/yr, 2017-2021), while the mortality decline slowed from 3.5%/yr (1993-2012) to 0.6%/yr (2012-2023). This is a stage-shift / late-detection signal, not a change to the cumulative lifetime-mortality headline, which is unchanged.\n","uncertainty":{"low":0.012,"high":0.028},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.wcrf.org/cancer-trends/prostate-cancer-statistics/","title":"Prostate cancer statistics","publisher":"World Cancer Research Fund International","source_type":"reputable_reference","statistic":"1,467,854 new prostate cancer cases and ~397,430 deaths globally in 2022; 4th most common cancer worldwide and 2nd most common cancer in men","excerpt":"\"There were 1,467,854 new cases of prostate cancer in 2022. [...] Prostate cancer is the 4th most common cancer worldwide. It is the 2nd most common cancer in men.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20241117074955/https://www.wcrf.org/cancer-trends/prostate-cancer-statistics/","calculation_notes":"WCRF republishes the IARC GLOBOCAN 2022 prostate cancer totals: ~1.47M new cases and ~397K deaths per year. Divided across ~3B adult men worldwide, that is ~1.3 per 10,000 per year on a flat-hazard basis. Age-weighting (prostate cancer mortality is concentrated above age 70, with a median age at death of 79 per SEER) pulls the realistic cumulative lifetime mortality near 1.5-2.5% globally. Used as the primary global headline and for the \"4th most common / 2nd in men\" framing in the body text. The ~3.7x ratio between global cases and deaths is the largest for any common cancer and is the central story of this entry.\n","independence_note":"WCRF’s cancer statistics pages are a downstream republication of IARC GLOBOCAN 2022. Treated as partially dependent with any other IARC-derived source; used here because the direct IARC news release for GLOBOCAN 2022 does not break out prostate cancer totals in its text.\n"},{"url":"https://seer.cancer.gov/statfacts/html/prost.html","title":"Cancer Stat Facts: Prostate Cancer","publisher":"Surveillance, Epidemiology, and End Results (SEER) Program, National Cancer Institute","source_type":"govt_report","statistic":"~12.9% US lifetime risk of prostate cancer diagnosis; 5-year relative survival 97.9% (2015-2021); median age at death 79; ~313,780 new cases and ~35,770 deaths estimated for 2025; age-adjusted death rate declining ~0.6% per year","excerpt":"\"Approximately 12.9 percent of men will be diagnosed with prostate cancer at some point during their lifetime, based on 2018–2021 data, excluding 2020 due to COVID.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413181914/https://seer.cancer.gov/statfacts/html/prost.html","calculation_notes":"SEER gives direct US lifetime incidence of ~12.9% (the \"roughly 1 in 8\" that most men have heard). With 5-year relative survival of 97.9% and a median age at death of 79, the implied long-run case fatality is small: ~12.9% incidence multiplied by a roughly 15-20% long-run case fatality gives a US lifetime prostate-cancer-death probability near 2.0-2.5%, consistent with ACS’s direct \"1 in 44\" figure. Anchors the US row in the regional breakdown and the top of the Likelier uncertainty band. The 97.9% 5-year survival is the mechanism that creates the diagnosis/ death gap flagged in the body text — it is higher than the figure for any other common cancer.\n","independence_note":"SEER (NCI) and IARC GLOBOCAN (WHO/WCRF) are methodologically independent compilation pipelines. SEER uses US vital registration and population- based cancer registries; IARC aggregates national registry data worldwide. The two are used here as independent anchors on the US and global ends of the regional breakdown.\n"},{"url":"https://www.cancer.org/cancer/types/prostate-cancer/about/key-statistics.html","title":"Key Statistics for Prostate Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"About 1 in 8 US men will be diagnosed with prostate cancer during their lifetime; about 1 in 44 will die of it; ~333,830 new cases and ~36,320 deaths projected for 2026","excerpt":"\"About 1 in 8 men will be diagnosed with prostate cancer during their lifetime. [...] About 1 in 44 men will die of prostate cancer. [...] Prostate cancer risk is also higher in Black men in the US and the Caribbean. [...] About 6 in 10 prostate cancers are diagnosed in men who are 65 or older, and it is rare in men under 40.\"\n","source_date":"2026-01-16","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260405152053/https://www.cancer.org/cancer/types/prostate-cancer/about/key-statistics.html","calculation_notes":"ACS gives both the 1-in-8 diagnosis and the 1-in-44 death figure explicitly. The 5.5x gap between the two is the load-bearing fact for this entry and the main calibration story: readers reliably quote \"1 in 8\" as if it were a death rate, when it is in fact the incidence rate — the death rate is ~2.3%, an order of magnitude below what the headline implies. The age-skew (\"6 in 10 diagnoses at 65 or older\") and the race-disparity framing are used to support the body text and the regional_breakdown / personal_factor_multipliers blocks.\n","independence_note":"ACS derives its US lifetime-probability figures from SEER incidence and mortality data. Treat these two as a single pipeline for US-specific numbers rather than as independent verification of each other.\n"},{"url":"https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/prostate-cancer-screening","title":"Final Recommendation Statement: Prostate Cancer: Screening","publisher":"US Preventive Services Task Force","source_type":"govt_report","statistic":"USPSTF recommends individual decision-making on PSA screening for men 55-69 (Grade C, upgraded from D in 2012); recommends against PSA screening for men 70+ (Grade D); notes 20-50% of screen-detected cases may be overdiagnosed","excerpt":"\"The decision to undergo periodic prostate-specific antigen (PSA)-based screening for prostate cancer should be an individual one. [...] The USPSTF recommends against PSA-based screening for prostate cancer in men 70 years and older.\"\n","source_date":"2018-05-08","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260322095529/https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/prostate-cancer-screening","calculation_notes":"USPSTF is the authoritative US primary-care screening guideline. The 2018 move from a blanket Grade D (recommend against) to Grade C (individual decision) for men 55-69 was a direct response to the longer-term ERSPC trial follow-up, which showed screening prevents about 1.3 prostate cancer deaths per 1,000 men screened over 13 years and reduces metastatic disease by about 3 per 1,000. That is a real but modest mortality benefit, set against a 20-50% overdiagnosis rate. Used here as the authoritative basis for the overdiagnosis framing in the body text and the myth_framing: overrated tag — \"overrated\" in the sense that the headline incidence figure massively overstates the death risk, not that the disease itself is not a major cancer.\n","independence_note":"USPSTF evidence synthesis is methodologically independent of the SEER/IARC incidence-registry pipelines; it aggregates RCT and cohort evidence on screening effectiveness. Treated as an independent source here for the screening and overdiagnosis claims.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38230766/","title":"Cancer statistics, 2024","publisher":"Siegel RL, Giaquinto AN, Jemal A / CA: A Cancer Journal for Clinicians","source_type":"peer_reviewed","statistic":"Prostate cancer mortality rates are approximately two-fold higher in Black men than in White men in the US, alongside stomach and uterine corpus cancers","excerpt":"\"Compared to White people, mortality rates are two-fold higher for prostate, stomach and uterine corpus cancers in Black people.\"\n","source_date":"2024-01-17","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411215337/https://pubmed.ncbi.nlm.nih.gov/38230766/","calculation_notes":"Siegel et al. 2024 is the authoritative annual ACS peer-reviewed cancer statistics summary. The ~2x Black-vs-White prostate cancer mortality ratio is the canonical figure used for the African American row in the regional_breakdown block and the \"African ancestry\" row in personal_factor_multipliers. The gap reflects both biology (higher incidence and more aggressive disease at diagnosis in men of African descent) and differential access to timely treatment; the paper does not attempt to decompose the two contributions precisely. The overall prostate cancer death rate has been declining ~0.6% per year (per SEER), so the absolute gap is narrowing even as the ratio persists.\n","independence_note":"Uses SEER incidence data and NCHS mortality data — same upstream as the SEER Stat Facts source above. Treated as a dependent but methodologically richer peer-reviewed analysis of the same pipeline.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12593258/","title":"Prostate cancer statistics, 2025","publisher":"Kratzer TB, et al. / CA: A Cancer Journal for Clinicians","source_type":"peer_reviewed","statistic":"US prostate cancer incidence reversed from −6.4% per year (2007-2014) to +3.0% per year through 2021, confined to advanced stage (regional +4.6%/yr, distant +4.8%/yr, 2017-2021); the mortality decline decelerated from 3.5% per year (1993-2012) to 0.6% per year (2012-2023); ~313,780 new cases and ~35,770 deaths estimated for 2025","excerpt":"\"Prostate cancer incidence trends have reversed from a decline of 6.4% per year during 2007 through 2014 to an increase of 3.0% annually during 2014 through 2021 [...] Over the past 5 data years (2017–2021), the rate has increased by 2.4% annually for localized-stage, 4.6% annually for regional-stage, and 4.8% annually for distant-stage disease [...] rapid declines of 3.5% per year from 1993 to 2012 have decelerated to 0.6% per year during 2012–2023.\"\n","source_date":"2025-09-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260305190704/https://pmc.ncbi.nlm.nih.gov/articles/PMC12593258/","calculation_notes":"Kratzer et al. 2025 is the dedicated ACS peer-reviewed prostate cancer statistics report. It documents a recent reversal that the older diagnosis/death framing does not capture: after years of decline, US incidence began rising ~3.0% per year (2014-2021), and the rise is concentrated in advanced disease — regional-stage +4.6%/yr and distant-stage +4.8%/yr over 2017-2021. Meanwhile the long mortality decline has slowed sharply, from 3.5%/yr (1993-2012) to 0.6%/yr (2012-2023). This does NOT change the headline lifetime-mortality figure (still ~1 in 50 global / ~1 in 44 US per ACS/SEER) — that figure is a cumulative lifetime probability, not a year-over-year trend — but it supports the \"stage-shift / late detection\" caveat added to this entry and is consistent with the screening-pullback context already discussed via USPSTF. Used for the trend note in assumptions and the caveat.\n","independence_note":"Uses SEER (NCI) incidence data and NCHS mortality data — the same upstream registry pipeline as the SEER Stat Facts and Siegel et al. sources. Treated as dependent on those for the underlying counts, but contributes the stage-specific and temporal-trend analysis those sources do not break out.\n"}],"comparison_anchors":[{"label":"All-cancer death (global adult lifetime)","lifetime_us_adult":0.14},{"label":"Lung cancer death (global adult lifetime)","lifetime_us_adult":0.018},{"label":"Colorectal cancer death (global adult lifetime)","lifetime_us_adult":0.013},{"label":"Breast cancer death (global adult women lifetime)","lifetime_us_adult":0.017},{"label":"Stroke death (global adult lifetime)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Global average (men)","probability":0.02,"notes":"~397K deaths/yr across ~3B adult men (WCRF / IARC GLOBOCAN 2022); age-weighted global adult lifetime figure"},{"region":"US men","probability":0.023,"notes":"ACS direct SEER-based estimate: ~1 in 44 lifetime death alongside ~1 in 8 lifetime diagnosis"},{"region":"African American men","probability":0.04,"notes":"~2x higher mortality than white Americans per Siegel et al. 2024 in CA: A Cancer Journal for Clinicians; partly biology (higher incidence and more aggressive disease), partly differential access to timely screening and treatment"},{"region":"East Asia (men)","probability":0.005,"notes":"Age-standardized prostate cancer mortality is an order of magnitude lower in East Asian populations than in men of European or African descent; the gap partly survives migration, suggesting a real genetic component alongside diet and screening differences"},{"region":"Women (all regions)","probability":0,"notes":"Not anatomically possible; women do not have a prostate gland"}],"personal_factor_multipliers":[{"factor":"African ancestry","multiplier":2,"notes":"~2x higher prostate cancer mortality in Black men vs White men in the US per Siegel et al. 2024; similar elevation in Caribbean populations of African descent per ACS"},{"factor":"first-degree family history (father or brother with prostate cancer)","multiplier":2,"notes":"Roughly doubles lifetime risk; stronger if multiple relatives or early-onset diagnosis"},{"factor":"BRCA2 pathogenic variant carrier","multiplier":3,"notes":"BRCA2 is the strongest common genetic driver of aggressive prostate cancer; BRCA1 carriers have a smaller but measurable elevation"},{"factor":"age 70+ vs age 50 baseline","multiplier":5,"notes":"Prostate cancer mortality is overwhelmingly age-driven — median age at death is 79 per SEER, and roughly 6 in 10 diagnoses occur at 65 or older. The annual hazard in the 70s is several times the hazard in the 50s, and negligible before 40."}],"short_label":"Prostate cancer","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is lifetime *mortality* from prostate cancer, not incidence. The widely quoted \"1 in 8\" figure is the lifetime probability a US man will be *diagnosed* with prostate cancer; the lifetime probability he will *die* of it is about 1 in 44 per ACS — roughly 5.5x smaller. The diagnosis/death gap is larger for prostate cancer than for any other common cancer on this site. Five-year relative survival is 97.9% per SEER, higher than for any other common cancer, and a meaningful share of older men carry histologically detectable prostate cancer that never causes symptoms: autopsy studies have found microscopic prostate cancer in a majority of men over 70 who died of unrelated causes. This is the main reason USPSTF recommendations on PSA screening have moved twice in the past fifteen years and still describe the decision as individual rather than routine — overdiagnosis and overtreatment are the dominant harms, and the mortality benefit, while real, is modest and context-dependent. Prostate cancer is also a male-only disease by anatomy, so the \"global adult\" framing of the headline number is population-weighted across the roughly half of adults at risk. For women the probability is zero; for men the population-average figure near 1 in 50 lifetime hides an enormous gap between indolent low-grade disease (where many men die *with* prostate cancer, not *of* it) and aggressive high-grade disease (where the 5-year survival drops sharply). The Black-White mortality gap is the largest demographic disparity for any common cancer in the US and is partly biological and partly structural — current evidence does not let the two contributions be cleanly separated. One recent caveat cuts against the reassuring diagnosis/death gap: per Kratzer et al. 2025 in CA: A Cancer Journal for Clinicians, US prostate cancer incidence has reversed from a decline of ~6.4% per year (2007-2014) to an increase of ~3.0% per year through 2021, and that rise is concentrated in advanced disease — regional- stage incidence rose ~4.6% per year and distant-stage ~4.8% per year over 2017-2021. Over the same period the long decline in prostate cancer mortality slowed sharply, from ~3.5% per year (1993-2012) to ~0.6% per year (2012-2023). This stage-shift toward later-stage diagnosis — widely attributed in part to reduced PSA screening after the 2012 USPSTF Grade D recommendation — does not change the cumulative lifetime-mortality figure used here, but it is a reminder that the favourable survival statistics apply overwhelmingly to localized disease, and that distant-stage prostate cancer remains far more lethal than the all-stages 5-year survival figure implies.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A large pale circle enclosing a much smaller dark circle on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/prostate-cancer","api_url":"https://likelier.app/api/fears/prostate-cancer.json"},{"slug":"speeding-crash-severity","question":"How much does driving 20% over the speed limit raise your odds of a fatal crash?","quick_answer":"Consistently driving 20% over the limit roughly doubles fatal-crash risk, lifting a regular speeder's lifetime odds of dying in a crash to about 1 in 50 — a far bigger jump than the 10–20% increase most drivers assume.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most drivers treat a 10-15 mph overage as a minor infraction, not a meaningful safety decision. Public understanding of crash physics is weak: surveys and focus groups consistently find that drivers underestimate how steeply crash severity scales with speed. Kinetic energy grows with the square of velocity, and the probability of a fatal outcome at impact grows even faster. Drivers who routinely exceed the limit by 20% tend to report that they feel in control and that their driving is only marginally riskier than keeping to the posted speed — the actual multiplier, around 2x on fatality risk, is far larger than most would guess.\n","rough_estimate":"most drivers underestimate the risk multiplier; gut feel is often 10–20% elevated risk, not 2x","kind":"intuition"},"native":{"display":"~1 fatal crash per ~3,850 driver-years at 20% over the speed limit","numerator":1,"denominator":3850,"unit":"per driver-year at 1.2x posted limit","population":"US driver travelling at 20% over posted speed limit. Baseline one-year motor-vehicle-crash death odds ≈ 1 in 7,979 (NSC lifetime odds 1 in 101 × 79.0-year life expectancy, per NSC's stated method) × 2.07 Power-Model factor → 1 in ~3,855, shown as ~3,850."},"normalized":{"lifetime_us_adult":0.02,"display":"~1 in 50 lifetime (regular speeder, 20%+ over limit)","log_value":-1.699,"assumptions":"Step 1 — Baseline lifetime motor-vehicle-crash fatality: the National Safety Council's \"Odds of Dying\" (2024 data) publishes US lifetime odds of dying in a motor-vehicle crash of 1 in 101 ≈ 0.0099. (NSC derives this by dividing the one-year odds by the 79.0-year life expectancy of a person born in 2024.) Step 2 — Power Model multiplier: Nilsson (1981/2004) gives the fatality multiplier as (v₂/v₁)⁴. At v₂/v₁ = 1.2 the multiplier is 1.2⁴ = 2.0736 ≈ 2.07. Elvik's 2013 meta-analysis of 98 studies confirms this exponent; at highway speeds the empirical exponent may exceed 4, so 2.07 is a conservative lower bound. Step 3 — Lifetime estimate: A driver who consistently travels at 1.2× the limit for a significant fraction of highway miles applies roughly the full 2.07× to their baseline risk: 0.0099 × 2.07 ≈ 0.0205, rounded to 0.020 as the central estimate. The display \"1 in 50\" = 1 / 0.020. Uncertainty band lower bound (0.012) corresponds to a ~1.2× exposure-weighted multiplier (half of miles at limit, half at 1.2×): 0.0099 × 1.2 ≈ 0.012. Upper bound (0.023) corresponds to 2.3× (highway driving mostly at 25% over limit, exponent ~4.5): 0.0099 × 2.3 ≈ 0.023.\n","uncertainty":{"low":0.012,"high":0.023},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.toi.no/getfile.php?mmfileid=13206","title":"The Power Model of the relationship between speed and road safety","publisher":"Nilsson, G. — Swedish National Road and Transport Research Institute (VTI) / Transport Research Institute (TØI)","source_type":"primary_study","statistic":"Fatal crash frequency scales as (v₂/v₁)⁴ relative to a baseline speed; a 20% speed increase → 1.2⁴ ≈ 2.07× fatality risk; a 10% increase → 1.1⁴ ≈ 1.46×","excerpt":"\"The number of fatal accidents is proportional to the fourth power of the ratio between the new and old mean speed of traffic.\"\n","source_date":"2004-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250805230617/https://www.toi.no/getfile.php?mmfileid=13206","calculation_notes":"Nilsson's original 1981 study established the power relationships; this 2004 VTI/TØI report is the canonical summary and reference document. The exponent of 4 for fatal crashes is the central estimate; Elvik's 2004 meta-analysis of 98 studies (460 estimates) broadly confirms this value across road types. Used here as the primary mechanism linking speed to fatality risk.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813721","title":"Traffic Safety Facts 2023 Data: Speeding (DOT HS 813 721)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"11,775 fatalities in speeding-related crashes in 2023, representing 29% of all 40,901 traffic fatalities; decrease of 3% from 12,157 in 2022; estimated 332,598 people injured in speeding-related crashes","excerpt":"\"In 2023, there were 11,775 fatalities in speeding-related crashes, which represents 29 percent of the total number of traffic fatalities for the year.\"\n","source_date":"2024-11-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260101201711/http://crashstats.nhtsa.dot.gov/Api/Public/Publication/813721","calculation_notes":"NHTSA FARS is the authoritative US count for speeding-related fatalities. The 29% share (11,775 of 40,901) establishes that speeding is the second largest behavioral factor in US traffic deaths. This source supplies the annual US fatality count and speeding share; the lifetime population baseline that the Power Model multiplier is applied to comes from NSC's stated lifetime odds (1 in 101), not from a hazard recomputed here.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22840212/","title":"A re-parameterisation of the Power Model of the relationship between the speed of traffic and the number of accidents and accident victims","publisher":"Elvik, R. — Accident Analysis and Prevention (Elsevier)","source_type":"peer_reviewed","statistic":"Meta-analysis of 98 studies with 460 estimates confirms the Power Model; exponent for fatal accidents empirically estimated at approximately 4 (range ~3–6 depending on road type and initial speed)","excerpt":"\"The Power Model of the relationship between speed and road safety proposes that changes in the mean speed of traffic are associated with changes in the number of accidents and accident victims according to power functions.\"\n","source_date":"2013-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250206132608/https://pubmed.ncbi.nlm.nih.gov/22840212/","calculation_notes":"Elvik's 2013 re-parameterisation (published in Accident Analysis & Prevention) is the peer-reviewed confirmation of Nilsson's empirically derived exponent. The update shows the exponent varies with initial speed — at highway speeds the fatality exponent is closer to 4–5, strengthening the case that 20% overspeed at 60 mph is more dangerous than the simple 2.07× central estimate implies. Used here to corroborate the Power Model as the consensus scientific framework.\n"},{"url":"https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","title":"Odds of Dying (2024 Data) — Motor-vehicle crash","publisher":"National Safety Council (NSC), Injury Facts","source_type":"reputable_reference","statistic":"Lifetime odds of a US resident dying in a motor-vehicle crash: 1 in 101 (2024 data). Lifetime odds are derived by dividing one-year odds by the 79.0-year life expectancy of a person born in 2024.","excerpt":"\"Motor-vehicle crash … 1 in 101.\" (Lifetime odds of death for selected causes, United States, 2024.) \"The lifetime odds are approximated by dividing the one-year odds by the life expectancy of a person born in 2024 (79.0 years).\"\n","source_date":"2026-01-01","source_accessed":"2026-07-10","archive_url":"http://web.archive.org/web/20260706033319/https://injuryfacts.nsc.org/all-injuries/preventable-death-overview/odds-of-dying/","calculation_notes":"NSC Injury Facts is the standard US reference for population-average lifetime odds of death by external cause; it computes odds from National Center for Health Statistics mortality data and Census population/life-expectancy figures. The 1-in-101 (≈0.0099) lifetime motor-vehicle-crash figure is the population baseline this entry multiplies by the Power Model factor. (An earlier NSC release, based on 2022 data, gave 1 in 93 ≈ 0.0108; the 2024-data figure is 1 in 101 and is the value used here.) Neither this baseline nor the \"1 in 93\" predecessor is an IIHS statistic — the prior IIHS attribution was an error.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0099},{"label":"Death while texting and driving (lifetime, regular-texter US adult)","lifetime_us_adult":0.018},{"label":"Death on a motorcycle (lifetime, US adult, population average)","lifetime_us_adult":0.00144}],"personal_factor_multipliers":[{"factor":"drives at or below the posted limit","multiplier":1,"notes":"Baseline US driver car-crash fatality risk; no speed-related multiplier applied."},{"factor":"drives 10% over the limit (5–6 mph on a 55-mph road)","multiplier":1.46,"notes":"1.1⁴ ≈ 1.46× fatality multiplier per the Power Model."},{"factor":"drives 20% over the limit (12 mph on a 60-mph road)","multiplier":2.07,"notes":"1.2⁴ ≈ 2.07× — the headline scenario for this entry."},{"factor":"drives 30% over the limit (20 mph on a 65-mph road)","multiplier":2.86,"notes":"1.3⁴ ≈ 2.86× — common on US interstates during off-peak hours."}],"short_label":"Speeding 20% over limit","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The Power Model (v₂/v₁)⁴ gives a point estimate for the per-trip fatality multiplier at a given speed ratio. It does not account for road type (the exponent is lower on urban arterials, higher on rural two-lanes), traffic density, vehicle safety ratings, or driver skill. The 2.07× figure is a central estimate for rural/highway conditions; it understates risk on undivided rural roads and overstates it in dense urban traffic where absolute speeds are low. NHTSA's speeding classification uses \"any speeding-related factor\" in the crash report — it includes drivers cited for going 1 mph over as well as those going 50 mph over — so the 29% share of fatalities does not mean the average speeding driver faced a 2× elevated risk; it means speeding was a coded contributing factor across a very wide range of magnitudes. The lifetime estimate on this page applies only to drivers who regularly travel 20%+ over the limit for a meaningful fraction of their miles; it is not a population-average figure.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A single speed limit sign on a pale road surface, flat vector illustration with muted colors."},"canonical_url":"https://likelier.app/speeding-crash-severity","api_url":"https://likelier.app/api/fears/speeding-crash-severity.json"},{"slug":"cat-litter-toxoplasmosis","question":"What are the odds of acquiring a toxoplasma infection from cleaning a cat's litter box?","quick_answer":"A US adult who cleans a litter box regularly for decades faces roughly a 1 in 48 lifetime chance of a litter-attributable toxoplasma infection. That is much lower than the 10-20% many imagine — most US infections trace to undercooked meat, not the litter box.\n","category":"health","tags":["pets","household"],"no_reliable_estimate":false,"perceived":{"description":"The cat-litter-as-toxoplasmosis-vector story is one of the most durable pieces of medical folk knowledge in the United States. Pregnant women are routinely told to hand over litter duty entirely, and many cat owners believe that scooping the box without gloves is a meaningful health hazard. The mental model is that cats are a primary transmission reservoir: share a home with a cat, clean its box, get infected. Online forums treat the risk as roughly equivalent to eating undercooked meat. Most people assume that a significant fraction of toxoplasmosis cases in the US come from litter-box exposure, and that immunocompetent adults who clean litter regularly face a meaningful lifetime probability of acquiring the infection from that source.\n","rough_estimate":"~10-20% lifetime chance for a regular cat-litter cleaner","kind":"intuition"},"native":{"display":"~1 litter-attributable infection per 2,700 litter-cleaning adults per year","numerator":1,"denominator":2700,"unit":"per year","population":"US adults who regularly clean a cat litter box","exposures_per_year":365},"normalized":{"lifetime_us_adult":0.021,"display":"~1 in 48 lifetime risk for a US adult who cleans litter regularly for 59 years","log_value":-1.68,"assumptions":"The CDC-referenced estimate of approximately 1.1 million new T. gondii infections per year in the US (Jones et al. 2014) implies an annual incidence of roughly 0.42% among the ~260 million US adults. However, the proportion attributable specifically to cat litter is contested and not well-measured; the best available evidence (USDA estimate, Jones et al. 1999-2000 analysis) attributes roughly half of US infections to undercooked or raw meat. Of the remaining half, oocyst-source infections (litter box, soil, contaminated water, unwashed produce) are split across multiple routes. CAPC and CDC guidance suggest direct contact with domestic cats is not a primary risk factor, especially for indoor cats on commercial diets, which rarely re-acquire and shed oocysts. Conservatively attributing ~15% of total infections to litter-box handling (range 5-30%) gives an annual litter-attributable incidence of ~0.063% across all adults, but only ~45% of US households own cats. Among active litter-cleaners the effective annual rate is approximately 1/2,700 (0.037%). Over a 59-year adult horizon, lifetime probability ≈ 1-(1-0.00037)^59 ≈ 2.1%. Daily litter cleaning (as recommended) substantially further reduces this because oocysts require 1-5 days to sporulate into the infectious form; an adult who cleans within 24 hours of deposition faces near-zero sporulation-based risk per event. The 2.1% figure therefore represents a worst-case estimate for less frequent cleaning; daily cleaners' actual risk is materially lower. Seroprevalence of 13.2% overall (NHANES 2009-2010) captures all routes over a lifetime; litter is a minority contributor.\n","uncertainty":{"low":0.005,"high":0.06},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4047742/","title":"Toxoplasma gondii Seroprevalence in the United States 2009-2010 and Comparison with the Past Two Decades","publisher":"American Journal of Tropical Medicine and Hygiene (Jones & Dubey)","source_type":"peer_reviewed","statistic":"Overall T. gondii seroprevalence among persons ≥6 years was 13.2% (age-adjusted 12.4%); among women 15-44 it was 9.1%; among US-born persons 12-49 the age-adjusted prevalence declined from 14.1% in NHANES III (1988-1994) to 6.7% in 2009-2010, and among US-born women 15-44 from 13.0% to 7.8%","excerpt":"\"The overall T. gondii antibody seroprevalence among persons ≥ 6 years of age in 2009–2010 was 13.2% (95% confidence limit [CL] 11.8%, 14.5%) and age-adjusted seroprevalence was 12.4% (95% CL 11.1%, 13.7%). Among women 15–44 years of age, the age-adjusted T. gondii antibody seroprevalence was 9.1%.\"\n","source_date":"2014-06-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505050923/https://pmc.ncbi.nlm.nih.gov/articles/PMC4047742/","calculation_notes":"A lifetime seroprevalence of 13.2% represents cumulative infection across all transmission routes over a lifetime. This is the upper ceiling for litter-box risk: litter is one of several routes (undercooked meat, soil, unwashed produce, water), so the litter-attributable fraction must be well below 13.2%. The declining trend (age-adjusted seroprevalence among US-born persons 12-49 fell from 14.1% in NHANES III to 6.7% in 2009-2010) coincides with food-safety improvements, consistent with meat being the dominant modifiable route. Used here as denominator context for the litter-specific calculation.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4015566/","title":"Neglected Parasitic Infections in the United States: Toxoplasmosis","publisher":"American Journal of Tropical Medicine and Hygiene (Jones et al., CDC)","source_type":"govt_report","statistic":"T. gondii infects an estimated 1.1 million persons each year in the US; ~327 deaths annually; proportion from cat litter vs meat is not known for a representative sample","excerpt":"\"Toxoplasma gondii infects an estimated 1.1 million persons each year in the United States. The proportion of human T. gondii infections acquired by eating meat containing infective cysts versus ingesting oocysts from cat feces contamination is not known for a representative sample of the general population.\"\n","source_date":"2014-05-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505050935/https://pmc.ncbi.nlm.nih.gov/articles/PMC4015566/","calculation_notes":"1.1 million infections per year among ~260 million US adults = annual incidence of approximately 0.42%. The authors explicitly state the litter vs. meat split is unknown. The USDA estimate (cited in the companion 1999-2000 seroprevalence paper) attributes roughly half to undercooked meat. Conservatively assigning 15% of the 1.1M annual infections to litter-box contact gives ~165,000 litter-attributable infections per year. Adjusted to the ~45% of US adults in cat-owning households (~117M), the annual rate among litter-cleaners is approximately 165,000/117,000,000 ≈ 0.14%. Even this conservative calculation is likely an overestimate because many oocyst-source infections come from soil and contaminated produce, not directly from litter boxes. We use 0.037% (half of 0.14%, reflecting that not all cat-owners clean litter and not all oocyst infections come from household litter) as the central per-cleaner annual estimate.\n"},{"url":"https://capcvet.org/guidelines/toxoplasma-gondii/","title":"Toxoplasma gondii — Companion Animal Parasite Council Guidelines","publisher":"Companion Animal Parasite Council (CAPC)","source_type":"reputable_reference","statistic":"~1% of cats shed oocysts at any time; oocysts sporulate (become infective) in 1-5 days after excretion; direct contact with cats is not considered a risk factor when cats are kept indoors on commercial diets","excerpt":"\"At any point in time, approximately 1% of cats have intestinal infection and will be shedding oocysts. Oocysts shed by cats become infective (sporulated) in 1 to 5 days and survive for months to years in the environment. Direct contact with cats is not considered to be a risk factor for T. gondii infection in people, particularly when cats are kept indoors and fed a commercial diet.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505050902/https://capcvet.org/guidelines/toxoplasma-gondii/","calculation_notes":"The 1% shedding prevalence at any point in time and the 1-5 day sporulation window are the two biological constraints that make daily litter cleaning so effective. If oocysts are removed before day 1, they are not yet infectious. Even without daily cleaning, only 1 in 100 cats sheds oocysts at any given time, meaning the large majority of litter-box interactions carry zero parasite exposure regardless of hygiene. This supports the conservative low end of the uncertainty range (0.5% lifetime) for daily-cleaning, indoor-cat households.\n"}],"comparison_anchors":[{"label":"Lifetime T. gondii seroprevalence (all routes, US adults)","lifetime_us_adult":0.132},{"label":"Foodborne illness requiring hospitalization (lifetime, US)","lifetime_us_adult":0.16},{"label":"Congenital toxoplasmosis (lifetime risk, US pregnancies)","lifetime_us_adult":0.001}],"personal_factor_multipliers":[{"factor":"cleans litter daily with gloves, washes hands after","multiplier":0.1,"notes":"Daily removal before the 1-5 day sporulation window, combined with barrier and hand hygiene, reduces risk to near-negligible; CAPC notes indoor cats on commercial diets rarely re-shed"},{"factor":"cleans litter every 3-7 days without gloves","multiplier":3,"notes":"Allowing oocysts to sit 3+ days permits full sporulation; handling without gloves adds direct fecal-oral exposure"},{"factor":"immunocompromised (HIV/AIDS, transplant, chemotherapy)","multiplier":20,"notes":"Reactivation of latent infection poses the greatest risk; any new infection can disseminate; immunocompromised individuals should avoid litter duty entirely"},{"factor":"pregnant, previously seronegative","multiplier":15,"notes":"Primary infection during pregnancy can cause congenital toxoplasmosis; CDC and CAPC recommend pregnant women delegate litter cleaning for the duration of pregnancy"},{"factor":"cat is strictly indoor and eats only commercial food","multiplier":0.2,"notes":"Indoor cats rarely hunt or ingest infected prey; re-infection and subsequent oocyst shedding is uncommon; CAPC explicitly notes lower risk in this scenario"}],"short_label":"Cat litter toxoplasmosis","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The litter-attributable fraction of annual US toxoplasmosis infections is not measured directly in any population-based study; the 15% figure used here is a conservative midpoint estimate. The biological constraints (1% feline shedding prevalence, 1-5 day sporulation window) are well-established, but they make clean quantification of per-cleaning-event risk difficult. Most seroprevalence surveys do not ask participants about litter-cleaning frequency or cat-ownership history in sufficient detail to isolate litter-box exposure as a risk factor. The 2.1% central estimate should be interpreted as \"upper-bound for an infrequent cleaner with an outdoor cat\"; the true figure for a daily-cleaning, indoor-cat household is likely under 0.5% lifetime.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A litter box with a scoop next to it on a tiled floor, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/cat-litter-toxoplasmosis","api_url":"https://likelier.app/api/fears/cat-litter-toxoplasmosis.json"},{"slug":"fighter-pilot-death","question":"What are the odds of a US military fighter pilot dying in an aviation mishap or combat over a career?","quick_answer":"A modern US fighter pilot faces roughly 0.60 fatal aviation mishaps per 100,000 flight hours, about 1 in 48 over a 20-year flying career. Most people substantially overestimate this, imagining tens of percent rather than the low single-digit percent the data shows.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Popular culture — from Top Gun to news coverage of every fighter crash — frames the profession as extraordinarily hazardous. The archetype of the \"danger zone\" pilot dying in a screaming fireball is one of the most durable pieces of occupational mythology in American life. No rigorous survey of public estimates of fighter pilot career mortality was identified, but the cultural baseline strongly implies most people place the career death risk somewhere between 10 and 50 percent — especially if they conflate Cold War training attrition rates (which were genuinely severe) with modern peacetime operations. The actual non-combat mishap fatality risk for a current US Air Force fighter pilot is far lower than that cultural prior, though it remains meaningfully above the average American job.\n","rough_estimate":"most people substantially overestimate; the actual career fatal-mishap risk for a modern US fighter pilot is in the low single-digit percent range, not tens of percent","kind":"intuition"},"native":{"display":"~0.60 fatal aviation mishaps per 100,000 fighter flight hours (estimated, FY2020–FY2021 basis)","numerator":0.6,"denominator":100000,"unit":"estimated fatal aviation mishaps per 100,000 fighter flight hours","population":"US Air Force fighter/attack aircraft pilots, FY2020–FY2021"},"normalized":{"lifetime_us_adult":0.021,"display":"~1 in 48 over a 20-year flying career","log_value":-1.68,"assumptions":"Reference subgroup: a US Air Force fighter pilot flying a 20-year career with approximately 3,500 total fighter flight hours (midpoint of the documented 3,000–4,000 hour range for careers spanning entry at ~age 24 through separation or retirement by ~age 44). Annual flight hours: fighter pilots averaged approximately 16.4 hours/month in 2018 USAF readiness data (approximately 197 hours/year), somewhat below the NATO-minimum 180 hours/year standard, yielding roughly 3,500 hours over a typical 20-year active flying career. The fatal mishap rate for USAF fighter/attack aircraft is derived in two steps. Step 1: USAF fleet-wide manned aircraft fatal mishap rate from Air Force Times reporting of official USAF Safety Center data: 0.45 fatal mishaps per 100,000 flight hours (FY2020) and 0.19 per 100,000 hours (FY2021, the lowest since at least 2014). Midpoint: approximately 0.30 per 100,000 hours. Step 2: Fighter aircraft apply a ~2× multiplier over the fleet-wide average, based on the documented finding that fighter/attack aircraft account for approximately 49% of all manned Class A and Class B USAF mishaps in FY2019, while constituting a smaller share of total manned flight hours; this is consistent with the ScienceDirect 2020 risk-classification study showing fighter-type platforms in higher-probability fatality categories than transport or trainer aircraft. Estimated fighter fatal mishap rate: 0.30 × 2 = ~0.60 per 100,000 hours. Career probability: 1 − (1 − 0.0000060)^3500 ≈ 1 − e^(−0.021) ≈ 0.021 (2.1%, roughly 1 in 48). This is a mishap-only figure; it excludes combat losses. Post-2003 US fighter combat losses to hostile fire have been effectively zero (no manned USAF fixed-wing aircraft was shot down during Operation Iraqi Freedom, Operation Enduring Freedom, or Operation Inherent Resolve), so the non-combat mishap rate dominates total career mortality risk for any pilot whose career falls entirely in the post-2003 era. The scope is activity_specific_lifetime because this is a career probability for a defined occupational subgroup, not a general US-adult figure.\n","uncertainty":{"low":0.008,"high":0.055},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.airforcetimes.com/news/your-air-force/2022/01/31/deadly-aircraft-accidents-declined-in-2021-air-force-says/","title":"Deadly aircraft accidents declined in 2021, Air Force says","publisher":"Air Force Times","source_type":"reputable_reference","statistic":"USAF manned aircraft fatal mishap rate: 0.19 per 100,000 flight hours (FY2021, lowest since 2014); 0.45 per 100,000 hours (FY2020); 21 Class A mishaps in FY2021 vs 30 in FY2020; four personnel killed in FY2021 vs seven in FY2020; approximately 1.24 million manned flying hours funded in FY2021","excerpt":"\"0.94 accidents per 100,000 flying hours for manned aircraft in 2021, the lowest since 2014. Four personnel were killed in Air Force accidents during the fiscal year, representing a decrease from seven fatalities in 2020. The death rate dropped to 0.19 per 100,000 flying hours compared to 0.45 the previous year.\"\n","source_date":"2022-01-31","source_accessed":"2026-05-10","archive_url":"https://web.archive.org/web/20260525161426/https://www.airforcetimes.com/news/your-air-force/2022/01/31/deadly-aircraft-accidents-declined-in-2021-air-force-says/","calculation_notes":"The article provides two key fleet-wide manned aircraft fatal mishap rates sourced from official USAF Safety Center data: 0.45/100k hours (FY2020) and 0.19/100k hours (FY2021). These are the primary basis for the fleet-wide fatal mishap rate midpoint of ~0.30/100k hours used in the headline calculation. The fighter-specific estimate applies a 2× multiplier derived from the FY2019 fighter mishap share data (see second source). Combined estimate: 0.60 fatal mishaps per 100,000 fighter flight hours. Career probability over 3,500 hours: 1 − (1 − 0.000006)^3500 ≈ 0.021.\n","independence_note":"Air Force Times is a DoD-independent trade publication that routinely obtains and reports official Air Force Safety Center statistical releases; this article directly cites USAF Safety Center figures. It is used here because the USAF Safety Center PDFs (safety.af.mil) are not accessible to automated fetch tools but are the underlying source for the reported numbers.\n"},{"url":"https://www.airforcetimes.com/news/your-air-force/2020/03/19/fighter-accidents-rose-in-2019-despite-overall-decline-in-mishap-rates/","title":"Fighter accidents rose in 2019, despite overall decline in mishap rates","publisher":"Air Force Times","source_type":"reputable_reference","statistic":"Fighter/attack aircraft accounted for 49% of all manned Class A and B USAF mishaps in FY2019; F-16 Class A–C mishaps rose from 67 (FY2018) to 90 (FY2019); F-15 from 61 to 76; F-22 from 45 to 57; F-35 from 17 to 20","excerpt":"\"Fighter aircraft accounted for 49 percent of all manned Class A and B mishaps in 2019, partly attributable to advanced technology in fifth-generation platforms elevating repair costs into higher mishap categories.\"\n","source_date":"2020-03-19","source_accessed":"2026-05-10","archive_url":"https://web.archive.org/web/20260525161458/https://www.airforcetimes.com/news/your-air-force/2020/03/19/fighter-accidents-rose-in-2019-despite-overall-decline-in-mishap-rates/","calculation_notes":"This article, sourcing USAF Safety Center FY2019 data, establishes that fighter/ attack aircraft are disproportionately represented in the most severe mishap categories. The 49% share of all manned Class A and B mishaps provides the empirical basis for the 2× multiplier applied to the fleet-wide fatal mishap rate to estimate the fighter-specific fatal mishap rate. Fighter platforms constitute a smaller share of total manned flight hours than their share of mishaps, making the direction of the multiplier well-supported; the specific magnitude (2×) is a conservative lower bound consistent with the mishap-share data and the ScienceDirect classification study (PMID/DOI: 10.1016/j.jsr.2020.01.003).\n","independence_note":"Same publication as the first source, but different reporters, different data year (FY2019 vs FY2021), and used for a distinct quantity (fighter mishap share ratio vs fleet-wide fatal rates). The two sources together are used to construct the fighter-specific fatal rate rather than as duplicative corroboration of the same statistic.\n"},{"url":"https://www.rand.org/pubs/research_reports/RRA257-1.html","title":"Trends in U.S. Air Force Aircraft Mishap Rates (1950–2018)","publisher":"RAND Corporation","source_type":"primary_study","statistic":"USAF pilot fatality rates from aviation mishaps showed persistent improvement from 1950s to 2010s; rates of improvement in Class A mishaps have been less dramatic since the 1970s; newer aircraft designs tend to experience lower mishap rates; greatest safety improvements occurred in the 1950s and 1960s","excerpt":"\"Trends in average mishap rates suggest major improvements in flight safety have been achieved, with the greatest rate of improvement occurring in the 1950s and 1960s. The rates of improvement in Class A mishaps and destroyed aircraft, although still meaningful, have been less dramatic since the 1970s. Mishaps involving pilot fatalities, however, have shown a more persistent rate of improvement. Aircraft introduced more recently have tended to experience lower mishap rates.\"\n","source_date":"2020-12-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260202215142/https://www.rand.org/pubs/research_reports/RRA257-1.html","calculation_notes":"The RAND study (Light, Hamilton, Pfeifer 2020) covers 55 USAF aircraft types from 1950 through 2018 using Air Force Safety Center data. It provides the historical context that Cold War-era fatal mishap rates were dramatically higher than modern rates — the era multiplier in personal_factor_multipliers. The study confirms the sustained improvement trend that contextualizes why modern estimates (~0.60/100k fighter flight hours) are far below historical Cold War figures (~5–10/100k hours in the 1950s–1960s). This source is used for the historical trend claim, not for a specific modern point estimate.\n","independence_note":"Independent RAND study using primary Air Force Safety Center data; methodologically distinct from the Air Force Times articles which report contemporary Safety Center statistical releases. RAND's 1950–2018 longitudinal analysis provides the era- comparison basis unavailable in contemporary reporting.\n"}],"comparison_anchors":[{"label":"US combat soldier (Iraq/Afghanistan peak era, career)","lifetime_us_adult":0.037},{"label":"Commercial airline pilot (career, US, fatal accident)","lifetime_us_adult":0.0003},{"label":"US police officer (traumatic line-of-duty death, 25-year career)","lifetime_us_adult":0.0032},{"label":"NASA astronaut (career, spaceflight fatality)","lifetime_us_adult":0.038}],"personal_factor_multipliers":[{"factor":"Career era: Cold War (1950s–1980s) vs post-2000 modern era","multiplier":10,"notes":"USAF pilot fatality rates from aviation mishaps were dramatically higher during the Cold War. Fatal mishap rates for fighter aircraft in the 1950s and early 1960s were roughly 5–10 per 100,000 flight hours; RAND (2020) documents sustained improvement from that era to the ~0.5–1.0 range by the 2010s, representing an approximately 10-fold reduction in fatal mishap rate over the period. A pilot who flew fighters in 1960 rather than 2020 faced roughly 10× the fatal mishap probability per flight hour."},{"factor":"Single-engine vs twin-engine fighter (F-16/F-35 vs F-15E/F-22)","multiplier":1.5,"notes":"Single-engine aircraft (F-16, F-35A) have no backup powerplant; engine failure in a single-engine jet is immediately life-threatening. The F-16's Class A mishap history at the USAF Safety Center shows elevated rates relative to the twin-engine F-15E. The USAF Safety Center FY2019 data showed F-16 with 90 Class A–C mishaps vs the F-15's 76, despite the F-15 fleet flying comparable hours. The multiplier is estimated at ~1.5× for single-engine vs twin-engine fighter platforms, consistent with the engine-failure contribution to total fatal mishaps."},{"factor":"Night or instrument meteorological conditions (IMC) operations","multiplier":2.5,"notes":"USAF mishap investigations consistently identify controlled flight into terrain (CFIT), spatial disorientation, and loss of situational awareness as leading causes of fatal fighter crashes. These risks concentrate in night low-level and IMC operations. USAF Safety Center data show that spatial disorientation alone accounts for approximately 15–20% of fatal fighter mishaps and occurs almost exclusively during night or instrument flight. Pilots flying predominantly night low-level training profiles face roughly 2–3× the fatal mishap risk of those in daytime VFR-dominant roles."},{"factor":"Combat deployment to actively contested airspace (e.g., Vietnam-era threat environment)","multiplier":5,"notes":"During the Vietnam War, the US Air Force flew over 5 million sorties and lost more than 2,250 aircraft — roughly 1 loss per 2,200 sorties, or a loss rate orders of magnitude above modern peacetime training. In the highest-threat periods of Rolling Thunder and Linebacker, combat loss rates for fighter aircraft reached 1–2 per 100 sorties in the most dangerous missions. Post-2003, US fighters flying over Iraq, Syria, and Afghanistan in low-threat environments experienced zero combat losses to hostile fire. A pilot whose career includes combat operations in a high-threat environment (peer or near-peer adversary with modern SAMs and fighters) faces roughly 3–8× the total career death risk of a peacetime-only pilot; 5× is used as a central estimate for a Vietnam-analog threat environment."}],"short_label":"Fighter pilot death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The headline figure of ~2.1% is an estimate built from two separately verified data points rather than a direct USAF published fighter-pilot mortality figure. The USAF Safety Center publishes per-aircraft mishap data (F-16FY23.pdf, etc.) but these PDFs were not machine-readable for this entry; the calculation relies on fleet-wide fatal rates from Air Force Times reporting of Safety Center releases, plus a fighter multiplier from mishap-share data. Users should treat the point estimate as order-of-magnitude accurate (range: 1–6%) rather than precise. The estimate covers only non-combat fatal mishaps — combat deaths post-2003 have been effectively zero for USAF fixed-wing aircraft and are discussed qualitatively. The figure does not include occupational disease, long-term health consequences of ejection (spinal compression injuries affect ~50% of ejecting pilots), or deaths in non-flying military roles. Career length and flight-hour assumptions vary significantly across pilots: those who leave the cockpit early for staff tours accumulate fewer hours and thus face lower absolute risk; those who fly in Reserve or Guard units throughout a longer career may accumulate more hours. The Defense One (2025) analysis found the overall USAF Class A mishap rate rose from 1.72 (2020) to 1.9 (2024) per 100,000 hours, suggesting the safety trajectory since 2021 has not continued to improve, which may push the true current rate toward the higher end of the uncertainty interval.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A single fighter jet helmet resting on a plain surface, flat vector illustration."},"canonical_url":"https://likelier.app/fighter-pilot-death","api_url":"https://likelier.app/api/fears/fighter-pilot-death.json"},{"slug":"mental-health-disability-claim","question":"What are the odds of filing a long-term disability claim for a mental health condition?","quick_answer":"About 1 in 45 US adults files a long-term disability claim for a mental health condition over their working life, and mental and nervous conditions make up roughly 9 in 100 LTD claims — the fourth-largest category. The figure likely understates the true burden, since most policies cap mental-health benefits at 24 months.\n","category":"health","tags":["workplace","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Given that roughly 1 in 5 US adults experiences a mental illness in any given year, most people might expect mental health to rank among the top causes of long-term disability. Instead, the Council for Disability Income Awareness finds mental and nervous conditions account for approximately 9% of LTD claims — the fourth-largest category, behind musculoskeletal disorders (~30%), cancer (~15%), and injuries (~11%). This underrepresentation relative to prevalence is partly explained by benefit duration limits (most group LTD policies cap mental health claims at 24 months), stigma-driven underreporting, and the reality that most mental health episodes, even serious ones, do not produce 12+ months of continuous work absence.\n","rough_estimate":"under 1 in 20 lifetime, most people guess","kind":"intuition"},"native":{"display":"~9 in 100 long-term disability claims","numerator":9,"denominator":100,"unit":"share of LTD claims","population":"US workers with long-term disability insurance claims"},"normalized":{"lifetime_us_adult":0.022,"display":"~1 in 45 US adults over their working lifetime","log_value":-1.658,"assumptions":"Two-step calculation. Step 1: SSA Fact Sheet states that just over 1 in 4 of today's 20-year-olds will become disabled before reaching normal retirement age. This implies a ~25% lifetime working disability probability. Step 2: CDIA's LTD claims data places mental and nervous conditions at approximately 9% (9.1% in the most precise CDIA figure) of all LTD claims. Combining: 0.25 × 0.091 ≈ 0.023, rounded to 0.022. This is almost certainly an underestimate of the true mental-health-related work disability rate because: (a) most group LTD policies limit mental health benefits to 24 months, reducing the count of ongoing claimants; (b) many mental health disabilities are coded under a comorbid physical diagnosis; (c) the SSA SSDI program shows higher mental disorder prevalence among approved claims than private LTD data suggests. Uncertainty range: 0.01-0.04, reflecting both the definitional sensitivity and the substantial and growing trend in mental health LTD claims.\n","uncertainty":{"low":0.01,"high":0.04},"scope":"us_adult_lifetime"},"sources":[{"url":"https://thecdia.org/the-top-5-reasons-why-people-go-out-of-work-and-stay-out-of-work/","title":"The Top 5 Reasons Why People Go Out of Work and Stay Out of Work","publisher":"The Council for Disability Income Awareness (CDIA)","source_type":"reputable_reference","statistic":"Mental health challenges account for 9.1% of long-term disability claims; musculoskeletal disorders account for nearly one-third","excerpt":"\"Mental health challenges — including depression and anxiety disorders — account for 9.1 percent of long-term disability claims.\"\n","source_date":"2018-04-30","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525162123/https://thecdia.org/the-top-5-reasons-why-people-go-out-of-work-and-stay-out-of-work/","calculation_notes":"CDIA's analysis of group LTD insurer claims data. The 9.1% figure covers mental and nervous conditions including depression, anxiety disorders, bipolar disorder, PTSD, and schizophrenia. CDIA notes that the true incidence may be higher because depression often goes untreated or is attributed to a comorbid physical cause in claims coding. Step 1: SSA 25% lifetime disability × 9.1% mental health share = 0.023 ≈ 0.022. The 24-month benefit duration limit in most group LTD policies (noted by the DOL ERISA Advisory Council 2023) means that long-duration mental health claimants are likely undercounted relative to physical disability claimants.\n","independence_note":"CDIA compiles data from multiple large group LTD insurer claim databases. This is distinct from SSA SSDI administrative data (which shows mental disorders accounting for a larger share of SSDI approvals) and from general population mental health prevalence surveys.\n"},{"url":"https://www.ssa.gov/news/press/factsheets/basicfact-alt.pdf","title":"Social Security Basic Facts — Disability Statistics","publisher":"Social Security Administration (SSA)","source_type":"govt_report","statistic":"Just over 1 in 4 of today's 20-year-olds will become disabled before reaching normal retirement age (age 67)","excerpt":"\"Just over 1 in 4 of today's 20 year-olds can expect to be out of work for at least a year because of a disabling condition before they reach the normal retirement age.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260517081611/https://www.ssa.gov/news/press/factsheets/basicfact-alt.pdf","calculation_notes":"SSA actuarial fact sheet provides the baseline 25% lifetime working disability probability used in Step 1. This figure covers all disabling conditions meeting SSA/SSDI criteria. Combined with the CDIA 9.1% mental health share: 0.25 × 0.091 ≈ 0.023, rounded down slightly to 0.022 as a conservative estimate given the undercount dynamics described above.\n","independence_note":"SSA administrative data is independent of CDIA's insurer claims database. The two sources use different populations and eligibility definitions; both converge on the ~25% overall lifetime disability incidence from independent methodologies.\n"}],"comparison_anchors":[{"label":"Musculoskeletal LTD claim (lifetime)","lifetime_us_adult":0.073},{"label":"Experiencing major depressive episode (lifetime, US adult)","lifetime_us_adult":0.21},{"label":"Workplace injury requiring days away from work (annual)","lifetime_us_adult":0.14}],"personal_factor_multipliers":[{"factor":"Prior major depressive episode requiring hospitalization or intensive treatment","multiplier":3,"notes":"A prior severe episode is the strongest predictor of future depression-related work disability; recurrence rates for major depression are high (50-80% have recurrent episodes)"},{"factor":"High-stress occupation (healthcare worker, first responder, social worker)","multiplier":1.5,"notes":"Burnout, secondary trauma, and PTSD are elevated in these occupations; mental health LTD claim rates are above average among healthcare workers and emergency services personnel"},{"factor":"Adequate access to mental health treatment (therapy plus medication when indicated)","multiplier":0.5,"notes":"Early and sustained treatment substantially reduces the probability of any mental health episode progressing to long-term work disability; access is the critical modifier"},{"factor":"Employer group LTD policy with 24-month mental health benefit cap","multiplier":1,"notes":"Most group LTD policies limit mental health benefits to 24 months regardless of clinical need; this affects claim duration and benefit receipt but not the probability of a disability episode occurring"}],"short_label":"Mental health LTD claim","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The 2.2% lifetime figure is a chained estimate with substantial uncertainty on both inputs. The SSA's 25% lifetime working disability figure is an actuarial projection and includes all causes; realized cohort rates have varied. The CDIA 9.1% mental health share is from insured employer group LTD claims, which systematically undercounts mental health for two reasons: (1) most policies limit mental health benefits to 24 months, reducing the ongoing claimant count at any point in time; (2) mental health conditions are frequently coded under a comorbid physical diagnosis in claims data. The DOL ERISA Advisory Council's 2023 report found that only 1% of group LTD policies lack a 24-month mental health benefit duration limit, and concluded that these limits are discriminatory and not supported by current clinical standards. The true lifetime probability of experiencing a mental-health-related work disability of any duration is substantially higher than 2.2% — the figure here captures only the subset reaching a formal LTD claim. Mental health conditions are the fastest-growing category in both short-term and long-term disability claims, with the pandemic period accelerating a multi-year trend. Growing-trend data suggest the claim rate will be higher for younger cohorts than historical averages.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simple office desk with an empty chair and a closed laptop, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/mental-health-disability-claim","api_url":"https://likelier.app/api/fears/mental-health-disability-claim.json"},{"slug":"motorcycle-helmetless-death","question":"What are the odds of dying on a motorcycle without a helmet?","quick_answer":"A regular unhelmeted rider faces roughly a 2.2% chance of dying in a crash over a 30-year riding career, at about 75 fatalities per 100,000 unhelmeted riders a year. Riders often picture the helmet effect as marginal — 10-15% — but it is closer to the difference between a survivable and a fatal crash.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Motorcycle helmets occupy an unusual position in risk perception. Riders in states without universal helmet laws often frame helmet choice as personal freedom, with the implicit assumption that skill and attention substitute for protective equipment. The \"loud pipes save lives\" school of thought emphasizes conspicuity over crash protection. Many unhelmeted riders acknowledge motorcycling is dangerous but underestimate the specific magnitude of the helmet effect, treating it as marginal rather than the difference between a survivable crash and a fatal one.\n","rough_estimate":"~10-15% higher death risk without a helmet","kind":"intuition"},"native":{"display":"~75 fatalities per 100,000 unhelmeted motorcycle riders per year","numerator":75,"denominator":100000,"unit":"annual fatality rate per unhelmeted motorcycle rider","population":"US unhelmeted motorcycle riders, derived from FARS/NHTSA 2023 data"},"normalized":{"lifetime_us_adult":0.022,"display":"~2.2% lifetime probability of dying in a motorcycle crash for a regular unhelmeted rider over a 30-year riding career","log_value":-1.66,"assumptions":"The overall 2023 motorcycle fatality rate was 47.6 per 100,000 registered motorcycles (6,335 deaths / 13.3M registered). NOPUS reports 73.8% helmet use. NHTSA's 37% effectiveness figure means unhelmeted riders are 1/(1-0.37) = 1.59x more likely to die per crash than helmeted riders. The overall rate is a weighted average: 47.6 = 0.738 × R_helmeted + 0.262 × R_unhelmeted, where R_unhelmeted = 1.59 × R_helmeted. Solving: R_helmeted = 47.6 / (0.738 + 0.262 × 1.59) = 47.6 / 1.154 = 41.2 per 100K. R_unhelmeted = 41.2 × 1.59 = 65.5 per 100K. Rounding up to ~75 per 100K to account for the fact that unhelmeted riders may also take other risks (night riding, alcohol) that correlate with helmet non-use. Over a 30-year riding career: 1 - (1 - 0.00075)^30 = 0.0223, or ~2.2%. CDC estimates 42% effectiveness (vs NHTSA's 37%), which would push the unhelmeted rate slightly higher. The IIHS analysis of 22,058 excess deaths from absent universal helmet laws (1976-2022) provides independent confirmation of the magnitude.\n","uncertainty":{"low":0.015,"high":0.04},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/809715","title":"Motorcycle Helmet Effectiveness Revisited","publisher":"National Highway Traffic Safety Administration (Deutermann, 2004)","source_type":"govt_report","statistic":"Helmets are 37% effective in preventing fatalities for motorcycle operators and 41% effective for passengers","excerpt":"\"Using FARS data and the double-pair comparison method, helmets are estimated to be 37 percent effective in preventing fatal injuries to motorcycle operators and 41 percent effective for motorcycle passengers.\"\n","source_date":"2004-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251226073044/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/809715","calculation_notes":"The NHTSA 37% figure is the canonical helmet effectiveness estimate used in all subsequent NHTSA \"lives saved\" calculations. It means that for every 100 unhelmeted riders who die, only 63 would have died had they been helmeted. The double-pair comparison method controls for crash severity by comparing helmeted and unhelmeted occupants within the same crash. This figure has been stable across multiple replications and is the basis for NHTSA's estimate that helmets saved 1,872 lives in 2017 and could have saved an additional 749.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813732","title":"Motorcycles: 2023 Data","publisher":"National Highway Traffic Safety Administration","source_type":"govt_report","statistic":"6,335 motorcycle fatalities in 2023; 35% of killed riders were unhelmeted; fatality rate 31.39 per 100M VMT","excerpt":"\"In 2023, 6,335 motorcyclists were killed in traffic crashes, the highest number since FARS began recording in 1975. The motorcyclist fatality rate was 31.39 per 100 million vehicle miles traveled. Among fatally injured motorcycle operators with known helmet use, 64 percent were helmeted.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251212202139/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813732","calculation_notes":"6,335 fatalities with 64% helmeted among known-use cases means ~2,281 unhelmeted deaths. With NOPUS showing 73.8% general helmet use, unhelmeted riders (26.2% of riders) account for 36% of deaths -- a 1.4x overrepresentation confirming the protective effect. The 31.39 per 100M VMT rate makes motorcycles ~28x more dangerous per mile than passenger cars (1.13 per 100M VMT).\n"},{"url":"https://www.iihs.org/news/detail/lax-helmet-laws-have-killed-more-than-20-000-motorcyclists-study-shows","title":"Lax helmet laws have killed more than 20,000 motorcyclists since 1976","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"An estimated 22,058 additional motorcyclists would have survived from 1976-2022 had universal helmet laws been in effect in all states","excerpt":"\"If every state had required all riders to wear helmets from 1976 to 2022, a total of 22,058 motorcyclists' lives could have been saved. The number represents 11% of all rider fatalities over those years.\"\n","source_date":"2024-07-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426204135/https://www.iihs.org/news/detail/lax-helmet-laws-have-killed-more-than-20-000-motorcyclists-study-shows","calculation_notes":"IIHS analysis using NHTSA's lives-saved methodology applied across all 50 states over 46 years. The 22,058 figure represents the cumulative cost of partial or absent helmet laws. Natural experiments from law repeals confirm the effect: Kentucky (+50% deaths after repeal), Louisiana (+100%), Texas (+31% operator fatalities). CDC systematic review found law repeals decreased helmet use by a median of 39 percentage points and increased deaths by a median of 42%.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/28526357/","title":"Universal Motorcycle Helmet Laws to Reduce Injuries: A Community Guide Systematic Review","publisher":"Peng, Vaidya, Finnie, Reynolds, Dumitru, Njie, Elder, Ivers, Sakashita, Shults, Sleet, Compton / American Journal of Preventive Medicine (CDC Community Preventive Services Task Force)","source_type":"peer_reviewed","statistic":"Helmet use associated with an estimated 42% reduction in fatal-injury risk and 69% reduction in head-injury risk; universal helmet law repeals associated with a median 39-percentage-point drop in helmet use and a median 42% increase in total deaths","excerpt":"\"Motorcycle helmet use can lead to an estimated 42% reduction in risk for fatal injuries and a 69% reduction in risk for head injuries.\"\n","source_date":"2017-06-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714015251/https://pubmed.ncbi.nlm.nih.gov/28526357/","calculation_notes":"This CDC Community Preventive Services Task Force systematic review of 60 US studies is the direct source for the \"CDC's slightly higher effectiveness estimate of 42%\" figure used in the body text and in normalized.assumptions and caveats above, and independently confirms the \"CDC systematic review\" repeal-effect figures (median 39-percentage-point drop in helmet use, median 42% increase in deaths) cited in the IIHS source's calculation_notes.\n","independence_note":"A meta-analytic synthesis of 60 US studies conducted for the CDC Community Preventive Services Task Force — methodologically distinct from NHTSA's single double-pair comparison analysis (Deutermann 2004) above — providing independent corroboration of the general magnitude of the helmet effect.\n"}],"comparison_anchors":[{"label":"Motorcycle death (general, lifetime rider)","lifetime_us_adult":0.02},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Cycling without helmet head injury (per crash)","lifetime_us_adult":0.17}],"personal_factor_multipliers":[{"factor":"wearing a DOT-compliant helmet","multiplier":0.6,"notes":"NHTSA: 37% reduction in fatality risk for operators, 41% for passengers; CDC cites 42% reduction"},{"factor":"riding in a state without a universal helmet law","multiplier":1.8,"notes":"States without universal laws have ~35% helmet use among fatalities vs 89% in universal-law states; fatality rates are correspondingly higher"},{"factor":"riding at night","multiplier":2,"notes":"NHTSA data shows disproportionate motorcycle fatalities at night; 33% of fatal crashes occur between 6pm and midnight"},{"factor":"alcohol involvement (BAC ≥ 0.08)","multiplier":3,"notes":"28% of fatally injured motorcycle operators in 2023 had BAC ≥ 0.08; alcohol impairs both riding ability and the decision to wear a helmet"},{"factor":"rider aged 40+ on a large-displacement bike","multiplier":1.5,"notes":"Older riders on high-powered motorcycles are the fastest-growing fatality demographic; average age of killed motorcyclists has risen steadily"}],"short_label":"Motorcycle no helmet","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 37% effectiveness figure from NHTSA is based on 2004 FARS analysis using the double-pair comparison method. The CDC estimates 42% effectiveness using different methodology. Both figures measure relative risk reduction in fatality, not absolute risk. The lifetime estimate of ~2.2% for unhelmeted riders depends heavily on assumed annual mileage and riding years; a weekend-only rider faces far less cumulative exposure than a daily commuter. The 31.39 per 100M VMT rate includes all riders regardless of helmet status; extracting a clean unhelmeted-only rate requires assumptions about VMT distribution by helmet use. FARS only records helmet use for fatalities and seriously injured riders, creating survivorship bias in some analyses. Helmet quality matters -- novelty helmets and non-DOT-compliant helmets provide significantly less protection.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A motorcycle parked next to a helmet resting on the seat, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/motorcycle-helmetless-death","api_url":"https://likelier.app/api/fears/motorcycle-helmetless-death.json"},{"slug":"commercial-fishing-death","question":"What are the odds of dying while working as a commercial fisher over a full career?","quick_answer":"Over a full 20-year career, a US commercial fisher's odds of dying on the job are about 1 in 45 — a fatality rate of 114 per 100,000 workers a year. Most people know the work is dangerous but underestimate how much.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Most people understand that commercial fishing is dangerous, partly because the industry has been dramatized in television programs covering Alaskan crab and halibut fleets. However, the precise magnitude of the risk tends to be underestimated. Viewers associate the danger with extreme weather events and distant Alaskan seas, underweighting the everyday vessel-disaster and man-overboard risks that occur along all US coastlines. No large-scale survey directly asks the general public to estimate career fatality odds for commercial fishers; perceived risk is classified as editorial intuition based on available media and occupational discourse.\n","rough_estimate":"most people guess a small but real risk, perhaps 1–2% per year, without a clear lifetime figure","kind":"intuition"},"native":{"display":"114 deaths per 100,000 FTE workers per year (US commercial fishing, 2000–2017 average)","numerator":114,"denominator":100000,"unit":"per worker per year","population":"US commercial fishing workers, NIOSH Commercial Fishing Incident Database 2000–2017"},"normalized":{"lifetime_us_adult":0.0224,"display":"~1 in 45 over a 20-year commercial fishing career","log_value":-1.65,"assumptions":"NIOSH's Commercial Fishing Incident Database recorded 791 work-related fatalities among US commercial fishers from 2000 through 2017 (18 years), yielding an average of approximately 44 deaths per year. With approximately 39,000 FTE commercial fishing workers in the US (BLS OES), the annualized fatality rate is 114 per 100,000 FTE workers (NIOSH, cdc.gov/niosh/maritime). A career is modeled at 20 years, consistent with the physically demanding nature of the occupation and median industry tenure. Compound probability over a 20-year career at 114 per 100,000 per year: 1 − (1 − 0.00114)^20 ≈ 0.0224. The scope is activity_specific_lifetime because this is per-career risk for a specific occupation, not a general US adult lifetime probability. The rate has varied substantially across years and regions (from ~86/100k in 2016 to ~204/100k in 2009), so the NIOSH 18-year average is used as the primary anchor; year-to-year variance is captured in the uncertainty band. The BLS notes the 2019 fatality rate for commercial fishers was more than 40 times the national average (all-worker rate ~3.5/100k → implied ~140/100k), consistent with the NIOSH long-run average.\n","uncertainty":{"low":0.015,"high":0.035},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/niosh/maritime/about/commercial-fishing.html","title":"About Commercial Fishing Safety — NIOSH Maritime Safety and Health","publisher":"National Institute for Occupational Safety and Health (NIOSH), CDC","source_type":"govt_report","statistic":"Fatality rate of 114 deaths per 100,000 FTE workers during 2000–2017; 791 total fatalities; 39,000 FTE workers","excerpt":"\"Commercial fishermen experienced a fatality rate of 114 deaths per 100,000 full-time equivalent (FTE) workers\" during 2000–2017, \"compared with an average of 4 deaths per 100,000 FTE workers among all U.S. workers.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260511103424/https://www.cdc.gov/niosh/maritime/about/commercial-fishing.html","calculation_notes":"Primary native rate: 114 deaths per 100,000 FTE workers per year (2000–2017 average). Total deaths: 791 over 18 years ≈ 43.9 deaths per year. Worker population: ~39,000 FTE. Annual probability: 114/100,000 = 0.00114. 20-year career compound: 1 − (1 − 0.00114)^20 ≈ 0.0224 ≈ 1 in 45.\n","independence_note":"NIOSH Commercial Fishing Incident Database (CFID) is an independent surveillance system distinct from BLS CFOI; CFID captures coast-guard-reported fatalities that may not reach BLS CFOI records, and uses a different denominator methodology.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5919605/","title":"Fatal Falls Overboard in Commercial Fishing — United States, 2000–2016","publisher":"MMWR Supplements / CDC NIOSH","source_type":"peer_reviewed","statistic":"2016 work-related fatality rate 86.0 deaths per 100,000 FTE workers, 23 times higher than all US workers (3.6/100k); 204 falls-overboard fatalities studied 2000–2016","excerpt":"\"2016 work-related fatality rate (86.0 deaths per 100,000 full-time equivalent workers) 23 times higher than that for all U.S. workers (3.6)\"\n","source_date":"2018-01-19","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260525094149/https://pmc.ncbi.nlm.nih.gov/articles/PMC5919605/","calculation_notes":"The 2016 rate of 86/100k provides the lower-bound anchor for the uncertainty range (86/100k over 20 years → 1−(1−0.00086)^20 ≈ 1.7%). The NIOSH 2000–2017 average of 114/100k sits above this, consistent with 2016 being a relatively safer year. Falls overboard (27% of all industry deaths) are methodologically documented separately from vessel disasters (48% of deaths), confirming that the dominant causes are vessel loss and man-overboard incidents, not onshore or gear-handling accidents.\n","independence_note":"Peer-reviewed MMWR study using NIOSH CFID data; independently published analysis distinct from the NIOSH overview page, with separate methodology covering falls overboard specifically.\n"},{"url":"https://www.bls.gov/opub/btn/volume-1/facts-of-the-catch-occupational-injuries-illnesses-and-fatalities-to-fishing-workers-2003-2009.htm","title":"Facts of the catch: occupational injuries, illnesses, and fatalities to fishing workers, 2003–2009","publisher":"Bureau of Labor Statistics (Beyond the Numbers)","source_type":"govt_report","statistic":"203.6 per 100,000 FTE workers fatality rate for fishers and related fishing workers, 2009","excerpt":"\"In 2009, the rate of fatal injury for fishers and related fishing workers was 203.6 per 100,000 full-time equivalent workers, which is more than 50 times the all-worker rate of 3.5 per 100,000.\"\n","source_date":"2012-10-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260515235052/https://www.bls.gov/opub/btn/volume-1/facts-of-the-catch-occupational-injuries-illnesses-and-fatalities-to-fishing-workers-2003-2009.htm","calculation_notes":"The 2009 rate of 203.6/100k provides the upper-bound anchor for the uncertainty range: 1−(1−0.002036)^20 ≈ 3.97%, rounded to 4%. Used to set the upper bound of the uncertainty band at 0.035 (hedging that peak years are extreme outliers). The BLS OES denominator for 2009 implied approximately 35,000–40,000 fishing workers, consistent with the NIOSH estimate of ~39,000 FTE.\n"}],"comparison_anchors":[{"label":"All-worker US average fatal work injury (career, 40 yr)","lifetime_us_adult":0.0014},{"label":"Police officer line-of-duty death (career)","lifetime_us_adult":0.0018},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Logging worker career death (30-year career)","lifetime_us_adult":0.0295}],"personal_factor_multipliers":[{"factor":"Alaska Bering Sea or Gulf of Alaska fisheries (crab, pollock, halibut)","multiplier":2.5,"notes":"NIOSH CFID data show Alaska accounted for 27% of all US commercial fishing fatalities 2000–2017 despite representing a smaller share of total US fishing employment. Alaskan crab fisheries in the Bering Sea have historically recorded the highest fatality rates of any US fishery, driven by severe weather, cold water immersion, and the short-season intensity of operations.\n"},{"factor":"Vessel less than 40 feet in length","multiplier":2,"notes":"NIOSH CFID analyses consistently show smaller vessels account for a disproportionate share of vessel-disaster fatalities. Smaller vessels have less stability margin, less redundant safety equipment, and are more likely to be operated by sole proprietors without onboard safety protocols. The MMWR overboard fatality study (2018) found working alone was a contributing factor in 48.5% of falls-overboard deaths.\n"},{"factor":"First year of commercial fishing (novice crew)","multiplier":1.8,"notes":"Occupational fatality analyses across hazardous industries consistently show elevated risk in the first 1–2 years of employment, when workers are learning vessel-specific procedures, gear handling, and emergency protocols. NIOSH safety training programs specifically target new-entry fishers as the highest-risk subgroup; formal deck safety and immersion-suit training significantly reduce first-year incident rates.\n"},{"factor":"East Coast inshore fisheries (lobster, scallop, groundfish)","multiplier":0.6,"notes":"NIOSH CFID regional data show the East Coast had the highest absolute count of fatalities (288, 33%) but across a much larger fishing workforce than Alaska, yielding a lower per-worker rate. Inshore lobster and scallop operations in New England show lower fatality rates than offshore or Alaskan deep-water fisheries, though still substantially above the all-industry average.\n"}],"short_label":"Commercial fishing career death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 114/100k headline rate is a 2000–2017 long-run average and conceals substantial year-to-year variance: the BLS CFOI records rates as low as 86/100k (2016) and as high as 204/100k (2009) for this occupation. The NIOSH CFID and BLS CFOI use different methodologies and denominators, so the two datasets are not directly interchangeable. The NIOSH figure uses coast-guard-incident reports as the numerator, which may capture fatalities missed by CFOI, while CFOI uses employer OSHA reports. The denominator (~39,000 FTE) is an estimate from BLS OES, not a head count, and may undercount part-time or seasonal fishing workers who nonetheless face full occupational exposure during their time on the water. The 20-year career assumption reflects the physically demanding nature of the work; workers with longer careers face a higher cumulative probability, and those who exit early face lower cumulative risk. The figure does not include non-fatal injuries, which are substantially undercounted due to the self-employed status of many commercial fishers.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A single empty rain slicker hanging on a hook against a pale grey wall, flat vector illustration in muted blues and grey."},"canonical_url":"https://likelier.app/commercial-fishing-death","api_url":"https://likelier.app/api/fears/commercial-fishing-death.json"},{"slug":"astronaut-spaceflight-death","question":"What are the odds of dying as an astronaut on a spaceflight mission?","quick_answer":"Spaceflight has killed about 19 of the 791 people flown to space, roughly 1 in 42 per astronaut per mission. This is one of the few risks where public intuition tracks the number — informed consent, not risk blindness, explains why people still fly.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Most people sense that spaceflight is genuinely dangerous — one of the few popular fears where public intuition tracks the statistics reasonably well. Media coverage is anchored by the two Space Shuttle losses, Apollo 1, and Soyuz 1/11, and the general reaction to \"would you take a 1 in 50 risk for a seat?\" is a cautious no. We have not found a standalone survey isolating \"fear of dying as an astronaut\", so perceived risk is marked as editorial intuition. The interesting property of this fear is that it is among the few Likelier entries where the perceived risk is roughly calibrated against the actual number, and where informed consent — not risk blindness — is what explains the behavior.\n","rough_estimate":"most people correctly guess spaceflight is in the low single-digit percent range per mission","kind":"intuition"},"native":{"display":"~19 deaths across ~791 people flown into space (1961-2026)","numerator":19,"denominator":791,"unit":"per astronaut per mission","population":"all humans who have flown to space, 1961-2026"},"normalized":{"lifetime_us_adult":0.024,"display":"1 in ~42 per astronaut per mission (all human spaceflight, 1961-2026)","log_value":-1.62,"assumptions":"Reference subgroup: a crew member on a crewed orbital or suborbital spaceflight mission of any program (Vostok, Mercury, Voskhod, Gemini, Soyuz, Apollo, Skylab, Space Shuttle, Shenzhou, Crew Dragon, etc.) between the first human spaceflight in 1961 and April 2026. The Wikipedia list of spaceflight-related accidents and incidents, which aggregates the standard program-level fatality records, reports that 791 people had flown into space as of April 2026 and that 19 of them died in spaceflight-related incidents, for a headline rate of 19/791 ≈ 2.4 percent. The scope is declared as activity_specific_lifetime because this is per-person-per-mission risk for a specific activity, not a general-population lifetime risk, and it is not directly comparable to the population-lifetime figures on other Likelier pages. The 19-death figure includes the Apollo 1 ground-test fire (3 deaths), Soyuz 1 reentry (1), Soyuz 11 decompression (3), X-15 Flight 3-65-97 (1), Challenger STS-51-L (7), and Columbia STS-107 (7). Excluding Apollo 1, which happened during a ground plugs-out test rather than in flight, yields 16 deaths / 791 ≈ 2.0 percent, which matches Rick Hauck's Shuttle-era calculation of \"closer to two percent\" across the Shuttle program. Because the two canonical figures (2.0 percent in-flight-only and 2.4 percent including Apollo 1) bracket the same order of magnitude, we use 0.024 as the headline point estimate and a wider uncertainty band to reflect that this is a small-sample statistic dominated by a handful of catastrophic events, each of which killed its entire crew.\n","uncertainty":{"low":0.02,"high":0.042},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.mdpi.com/2226-4310/9/11/675","title":"Sixty Years of Manned Spaceflight — Incidents and Accidents Involving Astronauts between Launch and Landing","publisher":"Aerospace (MDPI) / Schmitz J, Komorowski M, Russomano T, Ullrich O, Hinkelbein J","source_type":"peer_reviewed","statistic":"327 manned spaceflights from 1961-2020, 1,294 astronaut-missions, 19 astronaut deaths; fatality rate 5.8% per spaceflight","excerpt":"\"The number of astronauts who have died during spaceflight is represented by n = 19.\" \"The current statistical fatality rate is 5.8% (deaths per spaceflight) with the highest fatality rate in the 1960s (0.013 deaths/day spent in space), and the lowest rates in the 1990s and the period from 2010 until the present (no deaths).\"\n","source_date":"2022-11-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260331180534/https://www.mdpi.com/2226-4310/9/11/675","calculation_notes":"Schmitz et al. count 1,294 person-launches across 327 missions, confirming 19 deaths. The per-mission rate (19/327 = 5.8%) is higher than the per-unique-person rate (19/791 ≈ 2.4%) because many astronauts flew multiple missions. Likelier uses the per-unique-person rate as the headline since the entry frames risk per astronaut. The 19-death roster matches the Wikipedia aggregate exactly.\n","independence_note":"Peer-reviewed open-access paper drawing on NASA, Roscosmos, and ESA mission records. Independently compiled and methodologically distinct from the Space Review secondary reporting of Hauck's AIAA lecture.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32592663/","title":"Celestial Versus Terrestrial Travel — An Analysis of Spaceflight Fatalities and Comparison to Other Modes of Transportation","publisher":"American Journal of Medicine / Freiberg AS, Zhou S","source_type":"peer_reviewed","statistic":"18 fatalities in 4 fatal missions across 326 launches (1961-2019); per-trip fatality rate 1.2%, per-person rate 1.4%","excerpt":"\"There has yet to be a fatality in orbit, and there have been none on any space flight since 2003.\"\n","source_date":"2020-11-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413163424/https://pubmed.ncbi.nlm.nih.gov/32592663/","calculation_notes":"Freiberg and Zhou count 18 deaths (likely excluding X-15 Flight 191 depending on definition) across 326 launches through ~2019, giving a per-trip rate of 1.2% and per-person rate of 1.4%. The lower count vs Schmitz et al. (18 vs 19) reflects the definitional boundary around X-15 suborbital flights. Both papers independently confirm zero in-orbit fatalities and no deaths since Columbia (2003).\n","independence_note":"Fully independent of Schmitz et al. — different research group, different journal, different publication year, same underlying agency records but independently compiled.\n"},{"url":"https://thespacereview.com/article/36/2","title":"Weighing the risks of human spaceflight (page 2)","publisher":"The Space Review (reporting on Rick Hauck's AIAA lecture)","source_type":"reputable_reference","statistic":"18 of 430 humans who have flown in space had died, for a fatality rate of just over four percent; Shuttle-specific rate closer to two percent across ~600 seats on 113 flights","excerpt":"\"18 of the 430 humans who have flown in space have died\" ... \"a fatality rate of just over four percent.\" ... \"Would I have flown if I had known there was a four percent chance of death? No, I don't think I would have flown.\"\n","source_date":"2003-11-10","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163538/https://thespacereview.com/article/36/2","calculation_notes":"Hauck's 2003 AIAA analysis used a person-basis denominator (unique individuals flown) rather than a seat-basis denominator, producing a ~4 percent per-person figure immediately after the Columbia loss. Over the next two decades the denominator grew much faster than the numerator (no in-flight fatalities since 2003), pulling the figure down to ~2.4 percent by 2026. Hauck's separate seat-basis Shuttle calculation across \"over 600 seats filled on 113 flights\" produced \"closer to two percent\", which matches the modern 19/791 figure within rounding and is used as the cross-check that sets the low end of the uncertainty band. The Space Review's article is a secondary reputable-reference summary of a former Shuttle commander's technical talk, which is why it is classified reputable_reference rather than primary_study.\n","independence_note":"Hauck's person-basis denominator (430 people flown as of 2003) and Wikipedia's current 791-person denominator both derive ultimately from space-agency astronaut rosters, so the two sources are not independent on the denominator. They are independent in methodology — Hauck's lecture was computed ad hoc from program data in 2003, while the Wikipedia roster was aggregated from subsequent mishap-report publications. Treat them as method cross-check rather than two independent measurements.\n"},{"url":"https://www.nasa.gov/history/rogersrep/v2appf.htm","title":"Report of the Presidential Commission on the Space Shuttle Challenger Accident — Appendix F: Personal Observations on Reliability of Shuttle (R.P. Feynman)","publisher":"NASA / Rogers Commission","source_type":"govt_report","statistic":"Feynman's Rogers Commission appendix reported pre-Challenger shuttle failure-probability estimates ranging from roughly 1 in 100 (working engineers) to 1 in 100,000 (NASA management)","excerpt":"\"The estimates range from roughly 1 in 100 to 1 in 100,000. The higher figures come from the working engineers, and the very low figures from management.\"\n","source_date":"1986-06-06","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260706004005/https://www.nasa.gov/history/rogersrep/v2appf.htm","calculation_notes":"Supports the body's claim about the three-order-of-magnitude gap between NASA management's pre-Challenger reliability estimate (~1 in 100,000, already referenced in this entry's caveats) and working engineers' estimate (~1 in 100). Feynman's personal-observations appendix to the Rogers Commission report is the primary citation for the engineer-side figure. It does not change this entry's headline 2.4 percent historical fatality rate, which is a retrospective count rather than a pre-flight risk estimate.\n","independence_note":"A first-person NASA-hosted primary document, independent of the Schmitz et al., Freiberg & Zhou, and Space Review sources already cited, which compile post-hoc fatality counts rather than pre-flight risk-perception estimates.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death on a motorcycle (lifetime, active US rider)","lifetime_us_adult":0.02},{"label":"US combat death (per post-9/11 deployed service member)","lifetime_us_adult":0.00371}],"regional_breakdown":[{"region":"All crewed missions (1961-2026, per-person-per-mission)","probability":0.024,"notes":"Headline figure. 19 deaths / 791 unique people flown. Dominated by five catastrophic events."},{"region":"Space Shuttle era (per seat, 135 missions)","probability":0.016,"notes":"14 deaths across ~850 seats filled over 135 Shuttle missions, or roughly 1.6 percent per seat. Two of 135 missions (Challenger STS-51-L, Columbia STS-107) were loss-of-crew events."},{"region":"Soyuz program (per seat, 1967-2024)","probability":0.003,"notes":"4 deaths (Soyuz 1, Soyuz 11) across more than a thousand cosmonaut/astronaut seat-flights on Soyuz crewed missions. The Russian crewed program has had no in-flight fatalities since 1971."},{"region":"SpaceX Crew Dragon era (2020-2026)","probability":0.0001,"notes":"No fatalities across ~50 crew seats flown on Dragon as of April 2026. Small sample; the upper bound of the 95 percent confidence interval on a zero-event Poisson denominator of 50 is still above 5 percent per seat. Reported here as 'effectively zero so far'."},{"region":"NASA internal PRA estimate, late Shuttle program","probability":0.011,"notes":"NASA's post-Challenger probabilistic risk assessment settled on roughly 1 in 90 per flight for mature Shuttle operations; an earlier internal estimate put the first nine flights at roughly 1 in 9. Per-seat risk is similar because loss-of-crew events killed the whole crew."}],"personal_factor_multipliers":[{"factor":"Space Shuttle era (STS, 1981–2011)","multiplier":0.67,"notes":"14 of 19 total spaceflight deaths occurred on two Shuttle missions (Challenger, Columbia). 14 deaths across ~850 seat-flights ≈ 1.6%, vs the all-program headline of 2.4% — roughly 0.67× the average. Schmitz et al. (2022) document the Shuttle era as contributing the highest absolute death count. The per-seat rate is slightly lower than the unique-person headline because many Shuttle crew members flew only once."},{"factor":"Soyuz / Russian crewed program (1971–2024)","multiplier":0.12,"notes":"4 deaths (Soyuz 1, Soyuz 11) across more than 1,000 cosmonaut/astronaut seat-flights on Soyuz crewed missions since 1967; no in-flight fatalities since 1971. Regional breakdown data in this entry shows ~0.3% per seat for the Soyuz program vs the 2.4% all-program headline — roughly 0.12× the average."},{"factor":"Pioneer / early program era (1961–1975)","multiplier":2.5,"notes":"The 1960s had the highest fatality rate per day in space (0.013 deaths/day, per Schmitz et al. 2022). The early Soviet and NASA programs included unproven vehicles, first-generation abort systems, and limited safety culture compared to post-Challenger standards. Soyuz 1, Soyuz 11, and Apollo 1 all occurred in this window."},{"factor":"NASA post-Challenger PRA era (mature Shuttle, 1988–2003)","multiplier":0.45,"notes":"After Challenger, NASA's probabilistic risk assessment settled on ~1 in 90 per flight (~1.1%) for Shuttle operations — roughly half the all-time historical average. The Rogers Commission and subsequent Safety Advisory Panel reforms reduced per-flight risk relative to the 1970s-80s baseline. Columbia (2003) ended this window before further improvement was measurable."}],"short_label":"Spaceflight (astronaut)","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Spaceflight fatality statistics are a small-sample problem. Nineteen deaths across 65 years of human spaceflight is not enough data to support a stable per-mission rate in the frequentist sense; the headline 2.4 percent figure is dominated by five catastrophic events, each of which killed its entire crew, and would move substantially with a single future event in either direction. The figure also collapses risk profiles that differ by orders of magnitude: a Crew Dragon ferry flight to the ISS in 2025, a first-generation Vostok orbital flight in 1961, a Space Shuttle mission in 1986, and a future Mars transit mission are not drawing from the same risk distribution. The \"1 in 42 per mission\" headline should be read as a long-run historical average across the whole 65-year crewed-spaceflight program, not as a forecast for any specific future mission or vehicle. NASA's own internal probabilistic risk assessments for the late-Shuttle era settled on approximately 1 in 90 per flight, a figure that was itself a roughly tenfold revision upward from pre-Challenger management estimates of 1 in 100,000. The gap between engineering PRA and program management risk perception is one of the recurring themes of the Rogers Commission, CAIB, and Aerospace Safety Advisory Panel reports.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty space helmet resting on a pale neutral surface, flat vector illustration in muted greys and soft blue."},"canonical_url":"https://likelier.app/astronaut-spaceflight-death","api_url":"https://likelier.app/api/fears/astronaut-spaceflight-death.json"},{"slug":"benzodiazepine-dependence-after-prescription","question":"What are the odds of developing benzodiazepine dependence after a standard prescription?","quick_answer":"About 1 in 40 US adults develops benzodiazepine dependence over a lifetime; among people who receive at least one prescription, roughly 10% do. Patients often assume the risk is lower — closer to 5%, and only with long-term use.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Benzodiazepines — marketed under names like Xanax, Valium, Ativan, and Klonopin — occupy an unusual position in perceived risk. They are FDA-approved, prescribed by physicians, and widely regarded as medically sanctioned in a way that street drugs are not. The most common patient model of risk is that dependence applies to people who misuse them, take higher doses than prescribed, or use them recreationally — not to patients following a standard prescription for anxiety or insomnia. Many prescribers share this intuition, despite clinical guidelines recommending that prescriptions be limited to two to four weeks to minimize dependence risk. The FDA updated the Boxed Warning for all benzodiazepines in September 2020 to explicitly address physical dependence occurring even with prescribed therapeutic use, a change the agency described as necessary because the risk was systematically underestimated across the prescriber and patient population.\n","rough_estimate":"~5% of patients on long-term prescriptions","kind":"intuition"},"native":{"display":"~10% of adults who receive at least one benzodiazepine prescription develop physiological dependence (estimated from long-term use rates and per-exposure dependence rates)","numerator":10,"denominator":100,"unit":"estimated share of benzo prescription recipients who develop physiological dependence","population":"US adults who receive a benzodiazepine prescription (estimated from Soyka 2017, Bachhuber 2016, FDA 2020)"},"normalized":{"lifetime_us_adult":0.025,"display":"~1 in 40 US adults develops benzodiazepine dependence over a lifetime","log_value":-1.6,"assumptions":"This estimate requires two sequential inputs: (1) the lifetime probability of receiving at least one benzodiazepine prescription; (2) the per-prescription dependence rate given the observed mix of short and long courses.\nStep 1 — Lifetime prescription rate: Bachhuber et al. (Am J Public Health, 2016, PMC4816010) found that 5.6% of US adults filled at least one benzodiazepine prescription annually as of 2013. Over a 59-year adult life span (ages 18–77), cumulative exposure assuming independence and constant annual rate: 1 - (1 - 0.056)^59 ≈ 0.964, but this substantially overstates true lifetime exposure because the same individuals often have prescriptions across years. Lader (2011, Addiction) noted that approximately one-third of long-term benzo users continue for more than one year, suggesting substantial chronic use concentrated in a minority of the prescription population. A conservative estimate: approximately 25–35% of US adults receive at least one benzodiazepine prescription over their lifetime (accounting for the repeated-use concentration effect). We use 0.30 (30%) as the central estimate.\nStep 2 — Per-prescription dependence rate: Soyka (NEJM, 2017) summarizes the clinical literature: physiological dependence (defined by withdrawal symptoms on cessation) occurs in 20–44% of patients after long-term use (>4 weeks). However, the majority of benzodiazepine prescriptions are for shorter courses. Applying a rough distribution: approximately 30–40% of recipients receive courses >4 weeks (where dependence risk is 20–44%), and 60–70% receive shorter courses with lower but non-zero dependence risk. Weighted average per-prescription dependence rate: approximately 8–12%. Central estimate: 10%.\nStep 3 — Lifetime normalization: 0.30 (lifetime Rx probability) × 0.10 (per-prescription dependence rate) = 0.030. Adjusted slightly downward to 0.025 to account for potential overlap (repeat prescriptions to the same patient counted once) and the distinction between physiological dependence and clinically significant benzodiazepine use disorder (BzdUD), which requires functional impairment beyond physical tolerance. Central estimate: 0.025 (1 in 40 US adults).\n","uncertainty":{"low":0.01,"high":0.06},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJMra1611832","title":"Treatment of Benzodiazepine Dependence","publisher":"Soyka M — New England Journal of Medicine, 2017","source_type":"peer_reviewed","statistic":"Physiological dependence (withdrawal symptoms on cessation) occurs in 20-44% of patients after long-term benzodiazepine use (>4 weeks); physical dependence can occur in days to weeks even with therapeutic doses","excerpt":"\"Long-term use of benzodiazepines (longer than 4 weeks) can lead to physical dependence, with 20 to 44% of patients experiencing withdrawal symptoms on discontinuation. Physical dependence can occur even with therapeutic doses and can develop within days to weeks of continuous treatment. Symptoms of withdrawal include anxiety, irritability, confusion, seizures, and sleep disorders.\"\n","source_date":"2017-03-23","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260118193145/https://www.nejm.org/doi/full/10.1056/NEJMra1611832","calculation_notes":"The 20–44% dependence rate applies to patients on long-term (>4 weeks) benzodiazepine courses. This is used as the per-prescription dependence rate for the long-course subset. The weighted average across all prescription durations (short and long) is estimated at ~10%, which is the primary per-prescription input for the normalized calculation. This figure specifically measures physiological dependence, not DSM-5 benzodiazepine use disorder, which requires additional functional impairment criteria.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4816010/","title":"Increasing Benzodiazepine Prescriptions and Overdose Mortality in the United States, 1996–2013","publisher":"Bachhuber MA et al. — American Journal of Public Health, 2016","source_type":"peer_reviewed","statistic":"The percentage of US adults filling a benzodiazepine prescription increased from 4.1% in 1996 to 5.6% in 2013, with the quantity of benzodiazepines filled increasing from 1.1 to 3.6 kg lorazepam equivalents per 100,000 adults","excerpt":"\"Between 1996 and 2013, the percentage of adults filling a benzodiazepine prescription increased from 4.1% to 5.6%, with an annual percent change of 2.5%. The quantity of benzodiazepines filled increased from 1.1 to 3.6 kilogram lorazepam equivalents per 100,000 adults (annual percent change = 9.0%).\"\n","source_date":"2016-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260215021808/https://pmc.ncbi.nlm.nih.gov/articles/PMC4816010/","calculation_notes":"The 5.6% annual benzodiazepine prescription rate (as of 2013) is the primary input for estimating the lifetime probability of receiving at least one benzodiazepine prescription. This is used to establish the exposure base: approximately 25-35% of US adults are estimated to receive at least one benzodiazepine prescription over a 59-year adult lifespan (accounting for concentration of use in chronic users and avoiding naive independence assumptions across years). The notable increase in quantity per prescription (3.3x) relative to prevalence (1.4x) reflects growing long-term use in the patient population already receiving benzos — directly increasing dependence risk in that subset.\n"},{"url":"https://www.fda.gov/drugs/drug-safety-and-availability/fda-requiring-boxed-warning-updated-improve-safe-use-benzodiazepine-drug-class","title":"FDA Requiring Boxed Warning Updated to Improve Safe Use of Benzodiazepine Drug Class","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"FDA required updated Boxed Warning for all benzodiazepines in September 2020, explicitly addressing physical dependence occurring even with prescribed therapeutic use and normal dosing","excerpt":"\"Physical dependence can occur when benzodiazepines are taken steadily for several days to weeks, even as prescribed, and stopping them abruptly or reducing the dosage too quickly can result in withdrawal reactions, including seizures, which can be life-threatening. The FDA required the Boxed Warning to be updated with information describing the risks of abuse, misuse, addiction, physical dependence, and withdrawal reactions consistently across all medicines in the benzodiazepine drug class.\"\n","source_date":"2020-09-23","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260318051007/https://www.fda.gov/drugs/drug-safety-and-availability/fda-requiring-boxed-warning-updated-improve-safe-use-benzodiazepine-drug-class","calculation_notes":"The FDA Boxed Warning update is used as a regulatory anchor confirming that physiological dependence from prescribed therapeutic doses is a recognized, systematically documented risk — not a misuse phenomenon. The FDA's characterization of dependence occurring \"within several days to weeks even as prescribed\" is consistent with the Soyka (2017) clinical literature review. This source establishes regulatory standing for the risk, not a quantified prevalence figure; it is combined with the Soyka and Bachhuber sources for the arithmetic.\n"}],"comparison_anchors":[{"label":"Opioid addiction after surgical prescription (lifetime, US adult)","lifetime_us_adult":0.0088},{"label":"Cannabis use disorder (lifetime, US adult)","lifetime_us_adult":0.063},{"label":"Alcohol use disorder (lifetime, US adult)","lifetime_us_adult":0.291}],"personal_factor_multipliers":[{"factor":"long-term prescription (>4 weeks of daily use)","multiplier":4,"notes":"Soyka 2017: 20-44% dependence rate for long-term use; the per-prescription rate of ~10% is an average across short and long courses"},{"factor":"age 65 or older","multiplier":2,"notes":"Older adults receive more long-term prescriptions and have slower drug clearance, increasing dependence risk; also more likely to have had multiple prior prescriptions"},{"factor":"prescribed for insomnia (vs acute anxiety)","multiplier":1.5,"notes":"Chronic insomnia often leads to longer prescriptions than acute anxiety episodes; longer duration increases dependence risk"},{"factor":"history of alcohol or other substance use disorder","multiplier":3,"notes":"Cross-addiction vulnerability; benzodiazepines act on the same GABA-A receptor system affected by alcohol"},{"factor":"never received a benzodiazepine prescription","multiplier":0.05,"notes":"Near-zero prescription-pathway risk; the small residual reflects rare non-prescription exposure pathways not covered by this entry"}],"short_label":"Benzo dependence","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry specifically measures physiological dependence occurring via a standard prescription pathway — it does not include benzodiazepine use disorder (BzdUD) arising from illicit or non-prescription use. Physiological dependence (tolerance and withdrawal on cessation) is pharmacologically expected after several weeks of daily dosing and is distinct from DSM-5 benzodiazepine use disorder, which additionally requires functional impairment, loss of control, or continued use despite harm. Some degree of physiological adaptation is anticipated and manageable with gradual dose tapering; the disorder threshold captures the population experiencing clinically significant distress or dysfunction. The Bachhuber 2016 prescription rate data is from 2013 and does not capture more recent prescribing trends, including the co-prescription with opioids (which carries additional overdose risk) or the increased scrutiny following the 2020 Boxed Warning update, which may have modestly reduced long-term prescribing. The lifetime normalization arithmetic involves several estimates (lifetime Rx rate, per-prescription dependence rate for mixed-duration courses) that are not directly measured in a single study; the 0.025 central estimate should be understood as an order-of-magnitude figure. The actual risk for any individual depends heavily on prescription duration: a single 7-day course for acute anxiety carries meaningfully lower dependence risk than a 6-month prescription for chronic anxiety disorder.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A prescription pill bottle beside a calendar showing weeks elapsing, flat vector illustration in muted blue and grey tones."},"canonical_url":"https://likelier.app/benzodiazepine-dependence-after-prescription","api_url":"https://likelier.app/api/fears/benzodiazepine-dependence-after-prescription.json"},{"slug":"tap-water-lead","question":"What are the odds of lead poisoning from your home tap water?","quick_answer":"About 1 in 40 US children ages 1-5 have an elevated blood lead level, but that counts all sources combined (paint, dust, soil, and water), with water estimated at about 20% of exposure in homes with lead pipes. Clinical lead poisoning from modern municipal water alone is extremely rare.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Lead in drinking water became a national fixation after the Flint, Michigan crisis in 2014-2015, when corrosion-control failures sent lead levels to over 13,000 ppb in some homes — hundreds of times above EPA's 15 ppb action level. Coverage was justified: children were harmed, officials were indicted, and the phrase \"lead service line\" entered the public vocabulary. Post-Flint polling (Gallup 2016-2024) consistently finds that roughly 60% of US adults express concern about lead or heavy metals in their tap water, and environmental-advocacy messaging reinforces the impression that 9.2 million lead service lines make this a universal danger. The intuitive risk estimate for many consumers — especially parents — is that lead in tap water is both common and seriously harmful at typical municipal supply levels.\n","rough_estimate":"41% of US adults rank heavy metals in food among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 41% rank heavy metals in food as a top-3 concern; lead in tap water is the most prominent heavy-metal pathway post-Flint","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"~2.5% of US children aged 1-5 had BLL ≥3.5 µg/dL (2021 CDC reference value)","numerator":1,"denominator":40,"unit":"prevalence among US children 1-5","population":"US children aged 1-5 years, blood lead level ≥3.5 µg/dL (CDC 2021 lowered reference value)"},"normalized":{"lifetime_us_adult":0.025,"display":"~1 in 40 (childhood elevated BLL from all sources)","log_value":-1.6,"assumptions":"The 2.5% is definitional: CDC sets the blood lead reference value of 3.5 µg/dL at the 97.5th percentile of the blood lead distribution among US children aged 1-5 (NHANES 2015-2016 + 2017-2018), so by construction ~2.5% of that age cohort are at or above it. This is prevalence of elevated blood lead from ALL sources — paint, dust, soil, water, and consumer products — not water alone. Water contributes an estimated 20% of total lead exposure in homes with lead service lines or pre-1986 solder (EPA 2006 cost-benefit analysis for LCR). For homes with modern plumbing, water's contribution is negligible. Clinical lead poisoning (BLL ≥45 µg/dL) from water alone is extremely rare in modern municipal systems. The normalized figure represents the probability that a US child will have an elevated BLL at some point during ages 1-5, which is the peak exposure window. For adults, the BLL reference value of 3.5 µg/dL is less meaningful — adult reference ranges are higher and clinical toxicity starts at higher thresholds. The lifetime figure of 0.025 reflects childhood prevalence, which is the epidemiologically load-bearing number for tap-water lead concerns.\n","uncertainty":{"low":0.015,"high":0.04},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/lead-prevention/php/data/index.html","title":"Blood Lead Levels in Children — CDC Lead Prevention","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC's blood lead reference value of 3.5 µg/dL is the 97.5th percentile of the blood lead distribution in US children ages 1-5 (NHANES 2015-2016 + 2017-2018), so ~2.5% of that cohort are at or above it by definition","excerpt":"\"CDC analyzed NHANES data of the blood lead values among U.S. children ages 1-5 years from 2015-2016 and 2017-2018 NHANES cycles to determine the 97.5th percentile of blood lead distribution in children. This helped to determine the reference value at 3.5 µg/dL.... The childhood BLL data on CDC's website are not a population-based estimate.\"\n","source_date":"2024-08-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260412174443/https://www.cdc.gov/lead-prevention/php/data/index.html","calculation_notes":"The 2.5% headline is definitional, not a division: CDC sets the blood lead reference value at the 97.5th percentile of the blood lead distribution among US children 1-5 (NHANES 2015-2016 + 2017-2018 cycles), which places 3.5 µg/dL at that percentile — so ~2.5% of children in that age cohort are at or above it by construction. (An older CDC estimate put roughly 500,000 children above the reference value out of ~20 million aged 1-5, ~2.5%, but this page does not carry that count and explicitly notes its surveillance data \"are not a population-based estimate\"; the definitional 97.5th-percentile framing is used instead.) This prevalence captures all lead sources (paint, dust, soil, water, consumer products), not water alone. Tap water's contribution varies enormously by housing stock and service-line material — EPA's 2006 economic analysis for the Lead and Copper Rule estimated water contributes ~20% of total lead intake for children in homes with lead plumbing, and near zero for homes with modern copper or PEX plumbing.\n"},{"url":"https://www.epa.gov/newsreleases/biden-harris-administration-issues-final-rule-requiring-replacement-lead-pipes-within","title":"EPA Issues Final Rule Requiring Replacement of Lead Pipes Within 10 Years (Lead and Copper Rule Improvements)","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"EPA estimates up to 9 million homes are served by legacy lead pipes; the 2024 LCRI requires water systems to identify and replace lead pipes within 10 years and lowers the action level to 10 ppb","excerpt":"\"EPA estimates that up to 9 million homes are served through legacy lead pipes across the country... issued a final rule requiring drinking water systems across the country to identify and replace lead pipes within 10 years. The Lead and Copper Rule Improvements (LCRI) also require more rigorous testing of drinking water and a lower threshold requiring communities to take action to protect people from lead exposure in water.\"\n","source_date":"2024-10-08","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260605012650/https://www.epa.gov/newsreleases/biden-harris-administration-issues-final-rule-requiring-replacement-lead-pipes-within","calculation_notes":"The LCRI (89 FR 86018) finalizes mandatory lead-service-line replacement on a 10-year timeline, replacing the 1991 Lead and Copper Rule's partial-replacement regime. The cited EPA press release states EPA estimates \"up to 9 million homes are served through legacy lead pipes\"; EPA's parallel national lead-service-line inventory (systems were required to submit inventories by October 2024) puts the count at roughly 9.2 million service lines, serving roughly 22 million Americans. The prior LCR action level of 15 ppb required action only when >10% of tap samples exceeded the threshold; the LCRI lowers the action level to 10 ppb and shifts to proactive full replacement regardless of sampling results, reflecting the consensus that no lead level in water is safe for children.\n"},{"url":"https://www.atsdr.cdc.gov/toxprofiles/tp13.pdf","title":"Toxicological Profile for Lead","publisher":"Agency for Toxic Substances and Disease Registry (ATSDR), CDC","source_type":"govt_report","statistic":"No safe blood lead level has been identified; neurodevelopmental effects (IQ loss, attention deficits) begin below 5 µg/dL; clinical poisoning symptoms at BLL ≥45 µg/dL","excerpt":"\"There is no identified threshold for the adverse effects of lead in children. Effects on IQ and academic achievement have been demonstrated at blood lead levels below 5 µg/dL. Clinical signs of lead poisoning — abdominal pain, constipation, encephalopathy — generally occur at blood lead levels above 45 µg/dL in children and 70 µg/dL in adults.\"\n","source_date":"2020-08-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260318152824/https://www.atsdr.cdc.gov/ToxProfiles/tp13.pdf","calculation_notes":"The ATSDR profile synthesizes dose-response data from prospective cohort studies (Lanphear et al. 2005 pooled analysis, Rochester longitudinal study, Cincinnati Lead Study) establishing that each 1 µg/dL increase in BLL below 10 µg/dL is associated with a 1-2 point IQ decrement — a steeper dose-response slope at low levels than at high levels. The NTP 2012 monograph reached the same conclusion. Clinical lead poisoning (BLL ≥45 µg/dL) from drinking water alone is effectively limited to catastrophic infrastructure failures like Flint. In typical homes with lead service lines, first-draw water after overnight stagnation may reach 10-50 ppb, contributing perhaps 1-3 µg/dL to a child's BLL — a subclinical increment, not acute poisoning.\n"},{"url":"https://ntp.niehs.nih.gov/whatwestudy/assessments/noncancer/completed/lead","title":"NTP Monograph on Health Effects of Low-Level Lead","publisher":"National Toxicology Program, NIEHS","source_type":"peer_reviewed","statistic":"Sufficient evidence that BLL <5 µg/dL is associated with reduced IQ, reduced academic achievement in children, and increased incidence of attention-related behavioral problems","excerpt":"\"NTP concludes that there is sufficient evidence that blood lead levels less than 5 µg/dL are associated with adverse health effects in children, including decreased academic achievement, decreased IQ, and increased incidence of attention-related behavioral problems, and increased incidence of delayed puberty.\"\n","source_date":"2012-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250407012318/https://ntp.niehs.nih.gov/whatwestudy/assessments/noncancer/completed/lead","calculation_notes":"The NTP monograph was the pivotal document that led CDC to abandon the \"level of concern\" framework (previously 10 µg/dL) in favor of a reference value approach acknowledging no safe threshold. The monograph reviewed 17 prospective studies and 43 cross-sectional studies on IQ and lead. The sufficient-evidence classification for effects below 5 µg/dL established the scientific consensus that even modest lead exposure from any source — including tap water in homes with lead plumbing — carries measurable neurodevelopmental cost, even if the individual increment from water is small relative to paint and dust.\n"}],"comparison_anchors":[{"label":"PFAS illness from tap water (no reliable estimate)","lifetime_us_adult":0.001},{"label":"Heart disease death (lifetime, US adult)","lifetime_us_adult":0.2058}],"regional_breakdown":[{"region":"Modern plumbing (post-1986, no lead service line)","probability":0.001,"notes":"Homes built after 1986 (when Congress banned lead solder in plumbing) with copper or PEX service lines contribute negligible lead to tap water. BLL elevation from water in these homes is effectively zero. Residual probability reflects brass fixtures containing trace lead, which leaches at very low levels.\n"},{"region":"Pre-1986 solder, no lead service line","probability":0.01,"notes":"Lead solder in joints of copper pipes can leach 5-15 ppb in first-draw water after overnight stagnation. Contribution to childhood BLL is modest (~0.5-1 µg/dL) but measurable. Flushing the tap for 30 seconds before use reduces exposure by 50-90%.\n"},{"region":"Confirmed lead service line","probability":0.05,"notes":"Lead service lines can produce first-draw concentrations of 10-50 ppb and contribute 1-3 µg/dL to a child's BLL. Roughly 22 million Americans are served by the ~9.2 million lead service lines in the EPA inventory. The LCRI mandates full replacement by 2034.\n"},{"region":"Corrosion-control failure (Flint-type event)","probability":0.15,"notes":"When corrosion control fails catastrophically — as in Flint (2014-2015), where water chemistry changes stripped protective pipe scale — lead levels can exceed 1,000 ppb and BLL in children can spike above 10 µg/dL. These events are rare (a handful documented in US history) but produce genuine clinical harm.\n"}],"personal_factor_multipliers":[{"factor":"Child under 6","multiplier":5,"notes":"Children absorb 40-50% of ingested lead vs 3-10% for adults. Higher intake per body weight, hand-to-mouth behavior, and developing nervous system make children the primary concern for tap-water lead.\n"},{"factor":"Pregnant woman","multiplier":3,"notes":"Lead crosses the placenta. Maternal BLL ≥5 µg/dL is associated with reduced birth weight and neurodevelopmental effects. Bone-stored lead from prior exposure can remobilize during pregnancy.\n"},{"factor":"Home built pre-1986","multiplier":2,"notes":"Pre-1986 homes may have lead solder in copper pipe joints. First-draw water after overnight stagnation shows highest concentrations. Post-1986 plumbing uses lead-free solder (≤0.2% lead).\n"},{"factor":"Confirmed lead service line","multiplier":4,"notes":"Lead service lines are the largest single contributor to lead in tap water. EPA's LCRI mandates full replacement within 10 years. Until replacement, flushing and filtration (NSF-certified pitcher or faucet filter) reduce exposure by 90%+.\n"},{"factor":"Modern plumbing, municipal water with corrosion control","multiplier":0.05,"notes":"Homes with post-1986 plumbing on a municipal system maintaining proper corrosion control (orthophosphate or pH adjustment) have near-zero lead contribution from tap water.\n"}],"short_label":"Tap water lead","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The normalized figure (2.5% prevalence of BLL ≥3.5 µg/dL among US children 1-5) captures ALL lead sources, not water alone. Lead paint and paint dust remain the dominant exposure pathway, responsible for an estimated 70% of elevated BLLs in pre-1978 housing. Water's contribution ranges from negligible (modern plumbing) to significant (lead service line) to dominant (corrosion-control failure). Clinical lead poisoning (BLL ≥45 µg/dL) attributable to tap water alone is vanishingly rare outside catastrophic infrastructure events. The subclinical effects (IQ loss of 1-2 points per µg/dL BLL) are real and have no safe threshold, but the individual increment from tap water in most US homes is small. The entry's myth_framing of \"overrated\" refers to the perception that typical municipal tap water poses a clinical poisoning risk — not to the genuine concern for homes with confirmed lead service lines or the subclinical neurodevelopmental effects that accumulate at population scale. See also the [PFAS in tap water](/pfas-tap-water) entry for a distinct contaminant with a different evidence profile.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":2,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A kitchen faucet dripping into a glass, with faint pipe cross-sections visible in the background, flat vector illustration."},"canonical_url":"https://likelier.app/tap-water-lead","api_url":"https://likelier.app/api/fears/tap-water-lead.json"},{"slug":"pickpocket-tourist","question":"What are the odds of being pickpocketed while traveling?","quick_answer":"Roughly 1 in 1,000 to 1 in 2,000 per trip to a high-risk European tourist city, and about 1 in 38 over a lifetime of frequent international travel (~40 such trips). Most travelers dramatically overestimate this, often treating it as near-certain.\n","category":"crime","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Pickpocketing occupies a disproportionate share of the travel-anxiety landscape. Guidebooks, travel forums, and airport-bookshop thrillers have elevated it into a near-certainty in the minds of first-time visitors to Barcelona, Rome, or Paris. Travel-insurance comparison sites reinforce the framing by ranking \"pickpocket capitals,\" and Rick Steves has built a minor media empire partly on anti-pickpocket advice. No rigorous probability survey asks travelers to estimate their per-trip theft risk, so the perceived side here is editorial intuition. The directional finding is clear: most travelers to Western European tourist cities dramatically overestimate their personal risk of being pickpocketed on any given trip.\n","rough_estimate":"most travelers to European tourist cities overestimate; many treat it as near-certain","kind":"intuition"},"native":{"display":"~1 in 1,000 to 1 in 2,000 visitors pickpocketed per trip to a high-risk European city","numerator":1,"denominator":1500,"unit":"per trip to a high-risk European tourist city (Barcelona, Rome, Paris)","population":"international tourists visiting major European tourist cities"},"normalized":{"lifetime_us_adult":0.026,"display":"~1 in 38 lifetime (frequent international traveler, ~40 trips over a career)","log_value":-1.585,"assumptions":"Scope is activity_specific_lifetime — this is the probability for a regular international traveler who takes roughly 40 trips to high-risk tourist cities over a lifetime of travel (roughly ages 25-65, ~1 trip per year to a destination where pickpocketing is common).\nThe per-trip rate is estimated from multiple converging data points: (1) Rome police recorded 33,455 pickpocketing cases in 2024 against roughly 35 million annual visitors, implying a reported-crime rate of ~1 per 1,050 visitors. (2) The Quotezone European Pickpocketing Index 2024 records 478 pickpocketing mentions per million British visitors to Italy's top tourist attractions — about 1 per 2,100 visitors. (3) Barcelona's 1-in-70 figure circulated in travel media is almost certainly inflated by counting all reported thefts (including bag snatches, hotel-room thefts, and car break-ins) rather than pickpocketing alone. Restricting to pocket-picking and person-theft from pedestrian tourists, a defensible per-trip rate for a high-risk city is roughly 1 in 1,500. Most pickpocketing goes unreported, so the true rate may be 2-3x higher, but the relevant question for this page is \"did the tourist notice a loss,\" which is the reported-crime numerator.\nOver 40 lifetime trips at 1/1,500 per trip: 1 − (1 − 1/1500)^40 ≈ 0.026 ≈ 1 in 38.\nThis is an activity-specific figure. The vast majority of US adults take far fewer than 40 trips to high-risk European cities over a lifetime; for the median US adult (who may take 0-5 such trips), the lifetime probability is correspondingly lower.\n","uncertainty":{"low":0.01,"high":0.065},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.quotezone.co.uk/travel-insurance/guides/pickpocketing-statistics-in-europe","title":"European Pickpocketing Index 2024: Pickpocketing Statistics in Europe","publisher":"Quotezone (travel insurance comparison)","source_type":"reputable_reference","statistic":"Italy: 478 pickpocketing mentions per million British visitors; France: 251; Spain: 111; Germany: 111; Netherlands: 100","excerpt":"\"Italy has 478 pickpocketing mentions for every million British visitors to Italy's top tourist attractions — the highest proportion of any European country.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260403093653/https://www.quotezone.co.uk/travel-insurance/guides/pickpocketing-statistics-in-europe","calculation_notes":"Quotezone's European Pickpocketing Index 2024 uses TripAdvisor review mentions of pickpocketing normalized by visitor volume. Italy leads at 478 per million visitors, or approximately 1 mention per 2,092 visitors. This is a lower bound on the true rate because (a) not all victims leave TripAdvisor reviews and (b) not all review-leaving victims mention the theft. It is an upper bound in the sense that some mentions may be warnings rather than personal victimization reports. Treating it as an order-of-magnitude anchor: ~1 in 2,000 per trip to Italy's top attractions. This anchors the lower end of the native range.\n"},{"url":"https://www.chinatravelnews.com/article/187640/","title":"From romance to risk: How pickpockets are plaguing Europe's tourist hubs","publisher":"ChinaTravelNews","source_type":"news_article","statistic":"Rome: 33,455 pickpocketing cases in 2024, a 68% increase from 2019; over 2,000 reported robberies, up 51.3% from 2019","excerpt":"\"Pickpocketing incidents surged to 33,455 cases in 2024, marking a 68.0% increase compared to 2019.\"\n","source_date":"2024-05-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260117105826/https://www.chinatravelnews.com/article/187640","calculation_notes":"Rome's 33,455 reported pickpocketing cases in 2024 against approximately 35 million annual visitors to Rome implies a reported-victimization rate of roughly 1 per 1,050 visitors. This is the highest per-visitor reported rate among major European tourist cities and anchors the upper end of the native range. The figure includes all reported pocket-picking offenses, not just those against international tourists, so the tourist-specific rate may differ. However, pickpocketing in Rome disproportionately targets tourists at attractions like the Trevi Fountain and Colosseum.\n"},{"url":"https://popcenter.asu.edu/content/crimes-against-tourists-0","title":"Crimes Against Tourists","publisher":"ASU Center for Problem-Oriented Policing","source_type":"reputable_reference","statistic":"Theft is the most common crime against tourists; tourists are disproportionately targeted due to carrying cash, being distracted, and being unfamiliar with local crime patterns","excerpt":"\"Tourists are lucrative targets, since they typically carry large sums of money and other valuables, and they are vulnerable because they are more likely to be relaxed and off guard — and sometimes careless — while on vacation.\"\n","source_date":"2010-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421192809/https://popcenter.asu.edu/content/crimes-against-tourists-0","calculation_notes":"The ASU Center for Problem-Oriented Policing guide on crimes against tourists is a widely cited criminological reference establishing that tourists face elevated property-crime risk relative to residents. It does not provide a per-trip probability but establishes the qualitative framework: tourists are disproportionately victimized because they carry more cash, are less situationally aware, and concentrate at predictable locations. This supports the finding that per-trip pickpocketing risk in high-tourist-density areas is meaningfully higher than the background property-crime rate for residents.\n"}],"comparison_anchors":[{"label":"Losing any item (lost or stolen) on an international trip","lifetime_us_adult":0.46},{"label":"Violent crime victimization (lifetime, US adult, Koppel 1987)","lifetime_us_adult":0.83},{"label":"Car broken into (lifetime, US urban adult)","lifetime_us_adult":0.3},{"label":"Identity theft (annual, US adult)","lifetime_us_adult":0.07}],"personal_factor_multipliers":[{"factor":"visiting Rome, Barcelona, or Paris versus lower-risk destinations","multiplier":3,"notes":"The top three European pickpocketing cities have roughly 3x the per-visitor reported theft rate of mid-tier cities like Berlin or Vienna."},{"factor":"using crowded public transit at peak tourist season","multiplier":2,"notes":"Metro systems in Barcelona, Rome, and Paris are the single highest-concentration pickpocketing environments. Avoiding rush-hour metro use substantially reduces exposure."},{"factor":"carrying a backpack or open-top bag in a crowd","multiplier":2,"notes":"Visible, accessible bags are the modal target. Front-carry or zippered cross-body bags reduce risk."},{"factor":"traveling to low-theft destinations (Japan, Scandinavia, Iceland)","multiplier":0.1,"notes":"Per-visitor pickpocketing rates in Japan and Nordic countries are an order of magnitude lower than in Southern European tourist cities."}],"short_label":"Pickpocketed while traveling","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry covers pickpocketing — theft from the person without force or threat — at tourist destinations. It does not cover mugging (robbery with force), hotel-room theft, or rental-car break-ins, which are separate risk categories with different denominators. The native rate is calibrated to high-risk European tourist cities; travelers to Japan, Scandinavia, or rural destinations face a per-trip rate that is one to two orders of magnitude lower. The \"1 in 1,500\" per-trip figure is a reported-crime rate; the true rate including unnoticed and unreported thefts may be 2-3x higher, but unreported theft that the victim never notices is arguably not the event most travelers are afraid of. The lifetime figure assumes 40 trips to high-risk destinations, which is characteristic of a frequent international traveler, not the median US adult.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single wallet resting on a pale surface next to a folded city map, flat vector illustration in muted tones with a quiet amber accent."},"canonical_url":"https://likelier.app/pickpocket-tourist","api_url":"https://likelier.app/api/fears/pickpocket-tourist.json"},{"slug":"catalytic-converter-theft","question":"What are the odds of having your catalytic converter stolen?","quick_answer":"For a vehicle-owning US adult, catalytic-converter theft works out to roughly 1 in 37 over a lifetime even at the 2022 peak-year rate. That is far below the 5-10% per year some owners feared during the surge, and thefts fell sharply after 2022.\n","category":"crime","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Catalytic converter theft entered the public consciousness sharply around 2021--2022, when surge coverage made it feel ubiquitous in many metro areas. By 2024 the perceived risk had tempered somewhat as laws tightened and insurance claims data showed a steep decline from peak years. No rigorous national survey measures worry about this specific crime separately from general vehicle theft.\n","rough_estimate":"Vehicle owners in affected metros might put it at 5-10% per year during the 2022 peak; the actual average is far lower","kind":"intuition"},"native":{"display":"~64,700 insurance-claim thefts in 2022 (NICB peak year)","numerator":64701,"denominator":140000000,"unit":"per year","population":"US vehicle-owning households (approximate)"},"normalized":{"lifetime_us_adult":0.027,"display":"~1 in 37 lifetime (vehicle-owning US adult, at the 2022 NICB peak-year rate)","log_value":-1.57,"assumptions":"NICB reports 64,701 catalytic converter theft insurance claims in 2022 -- the peak of the surge cycle -- up from 16,660 claims in 2020. There are roughly 140 million vehicle-owning households in the US. Annual probability per vehicle-owning household at the 2022 peak rate: 64,701 / 140,000,000 = 0.000462 per year. Over 59 adult years at this rate: 1 − (1 − 0.000462)^59 ≈ 0.027 (1 in 37). The uncertainty band is anchored on the two cited NICB annual data points: the 2020 pre-surge rate (16,660 claims → ~0.007 lifetime, the low bound) and the 2022 peak (the point value). The high bound (0.04) reflects the NICB caveat that claims undercount total thefts -- NICB calls this \"an underreported crime\" -- so actual incidents run above the claims count. The headline deliberately anchors on the 2022 peak because it is the strongest figure the cited sources state verbatim; thefts fell sharply after 2022. State Farm, a second large insurer, independently reports a 74% year-over-year drop in H1 2024 claims (over 3,800 vs over 14,800), corroborating the decline, but as a single-insurer book its counts cannot be scaled to a national rate, so it is used qualitatively only. No cited source states a national annual theft total for 2023 or 2024, so no post-peak national rate is computed here.\n","uncertainty":{"low":0.007,"high":0.04},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nicb.org/news/news-releases/catalytic-converter-thefts-surge-nationwide-according-new-report","title":"Catalytic Converter Thefts Surge Nationwide, According To New Report","publisher":"National Insurance Crime Bureau (NICB)","source_type":"reputable_reference","statistic":"Insurance claims for catalytic converter theft rose from 16,660 in 2020 to 64,701 in 2022; California and Texas together saw more than 32,000 thefts in 2022","excerpt":"\"In total, the nation experienced more than 64,000 catalytic converter thefts in 2022. Leading the country are California and Texas, which experienced more than 32,000 catalytic converter thefts last year. ... Insurance claims for these thefts increased from 16,660 claims in 2020 to 64,701 in 2022.\"\n","source_date":"2023-05-10","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508021244/https://www.nicb.org/news/news-releases/catalytic-converter-thefts-surge-nationwide-according-new-report","calculation_notes":"Peak figure of 64,701 insurance claims in 2022 / ~140M vehicle-owning households = 0.000462/year. Over 59 years at this rate: 1 − (0.999538)^59 ≈ 0.027 (1 in 37). This is the headline point value. The 2020 figure of 16,660 claims gives ~0.007 lifetime, used as the low bound of the uncertainty range.\n","independence_note":"NICB is the insurance-industry crime data clearinghouse, drawing on claims submitted by member insurance companies. It does not capture thefts that go uninsured or unreported to insurers. The figures represent a lower bound on actual thefts.\n"},{"url":"https://www.nbcnews.com/data-graphics/catalytic-converter-thefts-decline-cars-steal-from-most-rcna91262","title":"Catalytic converter thefts decline after pandemic-era surge","publisher":"NBC News (citing NICB data)","source_type":"reputable_reference","statistic":"NICB monthly-average claims show catalytic converter thefts declining through 2023 from the 2022 peak, as rhodium prices fell to one-sixth of their 2021 peak and 21 states enacted anti-theft laws","excerpt":"\"Claims data from the National Insurance Crime Bureau shows that the average number of catalytic converters stolen each month is down significantly from its peak in 2022 through the first nine months of this year. ... The prices of the precious metals have tumbled, with rhodium's price down to one-sixth of its 2021 peak. ... This year alone, 21 states have enacted laws aimed at combating thefts.\"\n","source_date":"2023-11-05","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260115142657/https://www.nbcnews.com/data-graphics/catalytic-converter-thefts-decline-cars-steal-from-most-rcna91262","calculation_notes":"This NBC report (Nov. 5, 2023) documents the post-2022 decline qualitatively using NICB monthly-average claims data, but does not state a national annual theft total, so it contributes no numeric point to the rate calculation. It corroborates the direction of the decline (falling rhodium prices, legislative crackdowns) that anchors the caveat that the current rate sits below the 2022 peak used as the headline.\n","independence_note":"NBC News independently reports on and editorially processes the NICB claims data. It confirms the timing and drivers of the decline but, like the NICB release itself, ultimately traces to NICB insurance claims figures.\n"},{"url":"https://newsroom.statefarm.com/state-farm-sees-74-percent-drop-in-catalytic-converter-thefts/","title":"State Farm data reveals 74 percent drop in catalytic converter thefts","publisher":"State Farm","source_type":"reputable_reference","statistic":"State Farm auto claims show a 74% drop in catalytic converter theft claims in H1 2024 vs H1 2023 (3,800 vs 14,800 claims); the 2022 peak was about 45,000 claims","excerpt":"\"State Farm auto claims data shows a 74% drop in catalytic converter theft claims for the first half of 2024 compared to the same period in 2023. There were over 3,800 claims the first half of this year, down from the same time last year when State Farm received over 14,800 catalytic converter theft claims. Over 3,800 claims, totaling approximately $11.2 million paid to customers to repair car and replace the stolen part.\"\n","source_date":"2024-09-12","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260508020655/https://newsroom.statefarm.com/state-farm-sees-74-percent-drop-in-catalytic-converter-thefts/","calculation_notes":"State Farm's own book of business is not normalized to a national rate here; it serves as independent corroboration of the steep, sustained decline already captured by the NICB figures. State Farm reports a second consecutive year of decline (H1 2024), with claims down 74% YoY (over 3,800 vs over 14,800), and its full-year comparison table shows the book falling from ~45,000 claims in 2022 to 21,200 in 2023. This confirms that the current rate sits well below the 2022 NICB peak used as the headline point value, and does not alter the rate basis.\n","independence_note":"State Farm is a single large private insurer, distinct from the NICB clearinghouse, so its claims data is a genuinely independent corroboration of the decline rather than a re-publication of the same NICB figures. As a single-insurer book it cannot be scaled to a national rate without an assumed market share, so it is used qualitatively only.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Motor vehicle theft (lifetime, US)","lifetime_us_adult":0.05}],"personal_factor_multipliers":[{"factor":"Owns Toyota Prius or other hybrid vehicle","multiplier":6,"notes":"Hybrid vehicles contain higher concentrations of platinum-group metals in their converters; Toyota Prius is consistently the most-targeted vehicle in NICB data"},{"factor":"Lives or parks in California","multiplier":3,"notes":"California is consistently the highest-volume state -- it alone made up over half of State Farm's first-half-2024 converter-theft claims, and California and Texas together accounted for more than 32,000 of the 64,000-plus NICB-reported 2022 thefts; other high-volume states include Illinois, Texas, and New York"},{"factor":"Vehicle parked outdoors on street overnight vs in garage","multiplier":2,"notes":"Converters are typically stolen quickly with a battery-powered saw; outdoor and overnight parking substantially increases exposure"},{"factor":"Owns any hybrid vs conventional gasoline vehicle","multiplier":3,"notes":"Hybrid converters operate at lower temperatures and accumulate more platinum-group metals, making them significantly more valuable to thieves than conventional converters"}],"short_label":"Catalytic converter theft","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"NICB data counts insurance claims, not total thefts. Many converter thefts are not covered by insurance (comprehensive coverage required) or not reported to insurers, so the claims counts are a lower bound on actual thefts. The sharp rise and subsequent fall in thefts (NICB claims peaked in 2022) reflects a crime wave driven by precious-metal prices and suppressed by state legislation requiring vehicle title documentation for scrap metal sales. Future rates are uncertain; precious-metal prices and legislative enforcement effectiveness will shape the trend. State Farm, a second large insurer, independently reported a 74% year-over-year drop in H1 2024 claims (over 3,800 vs over 14,800) and a second consecutive year of decline, its full-year book falling from about 45,000 claims in 2022 to 21,200 in 2023 -- so the headline peak-year figure now sits well above the current observed rate. The decline is not guaranteed to hold: it tracks palladium and rhodium prices, which collapsed after 2022 but remain volatile, so a rebound in precious-metal prices could pull theft rates back up. The lifetime figure deliberately anchors on the 2022 NICB peak because it is the strongest annual total the cited sources state verbatim; the current rate is lower. Only vehicle-owning households face this risk, excluding roughly 10% of US households without a registered vehicle.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simplified car undercarriage with a gap where the catalytic converter was, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/catalytic-converter-theft","api_url":"https://likelier.app/api/fears/catalytic-converter-theft.json"},{"slug":"knife-assault-injury","question":"What are the odds of being stabbed in an assault?","quick_answer":"The lifetime odds of being the victim of a knife-involved aggravated assault are about 1 in 37 — higher than most expect for a \"rare\" violent crime, but lower than the anxiety it draws. It counts incidents where a knife was present, not only those where someone was actually stabbed and wounded, which is rarer still.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Knife attacks occupy a disproportionate share of crime anxiety, partly because they are viscerally frightening and heavily covered in local news. No major US survey directly asks about worry over being stabbed, but proxy data suggests elevated concern: Gallup's 2025 poll finds 27% of US adults worry about being \"attacked while driving\" and 29% about being mugged, placing interpersonal violence fears in the 20-30% range. Knife-specific fear likely sits within that band, amplified by media salience.\n","rough_estimate":"~1 in 5 adults express general worry about violent attack","kind":"intuition"},"native":{"display":"~155,000 aggravated assaults with knife/cutting instrument per year","numerator":155,"denominator":100000,"unit":"per year (estimated)","population":"US population (~335 million; FBI 2023 aggravated-assault count × 2019 UCR 17.5% knife weapon-share)"},"normalized":{"lifetime_us_adult":0.027,"display":"~1 in 37 lifetime","log_value":-1.57,"assumptions":"FBI UCR data for 2019 (last year with full legacy reporting) shows knives or cutting instruments accounted for 17.5% of aggravated assaults. The 2023 aggravated assault rate is 264.1 per 100,000 (Statista/FBI), implying roughly 885,000 aggravated assaults nationally (264.1 × 3.35 million hundreds of population). Applying the historical 17.5% knife share yields ~155,000 knife-involved aggravated assaults per year, or roughly 46.3 per 100,000 population (~0.000463 annual probability). Compounding over 59 adult years: 1 − (1 − 0.000463)^59 ≈ 0.027. This is the probability of being a victim of at least one aggravated assault involving a knife/cutting instrument, NOT the probability of being physically stabbed with a wound. The FBI weapon category counts incidents where a knife was brandished, displayed, or threatened as well as those where it made contact, and the national data does not break out the wounded fraction. The true \"stabbed with a wound\" probability is therefore some unquantified amount lower than 0.027; no cited national source states that fraction, so this entry does not assign it a precise value.\n","uncertainty":{"low":0.013,"high":0.045},"scope":"us_adult_lifetime"},"sources":[{"url":"https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-2019/topic-pages/aggravated-assault","title":"Aggravated Assault — Crime in the United States, 2019","publisher":"Federal Bureau of Investigation, Uniform Crime Reporting Program","source_type":"govt_report","statistic":"Knives or cutting instruments were used in 17.5% of aggravated assaults in 2019","excerpt":"\"In 2019, of the aggravated assault offenses for which law enforcement provided expanded data, 27.6 percent were committed with firearms. Personal weapons, such as hands, fists, or feet, were used in 25.2 percent of aggravated assaults, and knives or cutting instruments were used in 17.5 percent. Other weapons were used in 29.8 percent of aggravated assaults.\"\n","source_date":"2020-09-28","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426202844/https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-2019/topic-pages/aggravated-assault","calculation_notes":"FBI UCR 2019 is the last year of full legacy Summary Reporting System data with a clear weapon-type percentage table. The 17.5% knife share has been stable within 1-2 percentage points since 2010 (ranging 17.2%-18.0%). Applied to the 2023 estimated aggravated assault count of ~885,000 (derived from the 2023 rate of 264.1/100K × 335M population): 0.175 × 885,000 ≈ 155,000 knife-involved aggravated assaults per year. Annual risk per person: 155,000 / 335,000,000 ≈ 0.000463. Lifetime: 1 − (1 − 0.000463)^59 ≈ 0.027.\n"},{"url":"https://www.fbi.gov/news/press-releases/fbi-releases-2023-crime-in-the-nation-statistics","title":"FBI Releases 2023 Crime in the Nation Statistics","publisher":"Federal Bureau of Investigation","source_type":"govt_report","statistic":"Aggravated assault decreased an estimated 2.8% in 2023; rate of 264.1 per 100,000","excerpt":"\"Aggravated assault figures decreased an estimated 2.8 percent. The FBI released detailed data on over 14 million criminal offenses for 2023 reported to the Uniform Crime Reporting (UCR) Program by participating law enforcement agencies. More than 16,000 agencies, covering 94.3% of inhabitants, submitted data through NIBRS.\"\n","source_date":"2024-09-23","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260422065043/https://www.fbi.gov/news/press-releases/fbi-releases-2023-crime-in-the-nation-statistics","calculation_notes":"The 2023 aggravated assault rate of 264.1 per 100,000 comes from FBI NIBRS estimates. With a US population of approximately 335 million, this implies roughly 885,000 reported aggravated assaults. The 2.8% decrease from 2022 is consistent with a multi-year declining trend in violent crime. These are law-enforcement-reported incidents; the NCVS estimates substantially more aggravated assaults when unreported incidents are included.\n","independence_note":"FBI NIBRS collects data from law enforcement agencies. The weapon-type breakdown (17.5% knife) is from the 2019 UCR legacy system, applied to the 2023 aggregate count. The two data points are from the same FBI program but different reporting years and systems.\n"},{"url":"https://counciloncj.org/wp-content/uploads/2024/07/assault-fact-sheet.pdf","title":"Trends in Assault: What You Need to Know","publisher":"Council on Criminal Justice","source_type":"reputable_reference","statistic":"Aggravated assault trends and weapon-type context for US violent crime","excerpt":"\"Aggravated assault is by far the most common form of serious violent crime in the United States. Aggravated assaults accounted for 68% of all violent crime reported to police in 2022.\"\n","source_date":"2024-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260310053410/https://counciloncj.org/wp-content/uploads/2024/07/assault-fact-sheet.pdf","calculation_notes":"Contextual source confirming that aggravated assault is the dominant category of violent crime. The 68% share of all reported violent crime underscores that aggravated assault is not a rare event class. Used to validate the FBI NIBRS aggregate count and provide trend context.\n","independence_note":"Council on Criminal Justice is an independent policy organization that synthesizes FBI and BJS data. Not a primary data collector.\n"}],"comparison_anchors":[{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Male aged 18-24","multiplier":3.5,"notes":"Young men are disproportionately both perpetrators and victims of aggravated assault per FBI and NCVS data"},{"factor":"Urban resident in high-crime metro","multiplier":2.5,"notes":"NCVS urban violent victimization rate is roughly 2-3x the suburban rate"},{"factor":"Female","multiplier":0.5,"notes":"Women are substantially less likely to be victims of stranger aggravated assault with a weapon"},{"factor":"Rural resident","multiplier":0.6,"notes":"Rural aggravated assault rates are lower than urban rates in NCVS data"}],"short_label":"Knife-involved assault","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline \"1 in 37\" measures being the victim of a knife-INVOLVED aggravated assault, not being physically stabbed with a wound. The FBI UCR category \"aggravated assault with knife or cutting instrument\" includes incidents where a knife was brandished, displayed, or threatened but no stabbing occurred. The fraction of these incidents that result in an actual stab wound is not separately reported in national data, so the true \"stabbed with a wound\" lifetime probability is meaningfully lower than 1 in 37 by an amount this entry does not attempt to quantify — no cited authoritative source states that fraction, and manufacturing one would be false precision. FBI data captures only incidents reported to law enforcement; the NCVS, which includes unreported assaults, would yield a higher figure. The 17.5% weapon share is from 2019 (last full legacy UCR year) applied to the 2023 aggregate — the NIBRS transition makes direct weapon-type percentages for 2023 harder to extract, though historical stability in the 17-18% range suggests limited error. Young urban men face dramatically higher risk than the population average.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A closed folding knife resting on a grey surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/knife-assault-injury","api_url":"https://likelier.app/api/fears/knife-assault-injury.json"},{"slug":"toddler-stair-fall-injury","question":"What are the odds of a toddler suffering serious injury from falling down stairs?","quick_answer":"Of toddler stair falls serious enough to reach an ER, about 2.7% require hospitalization, and most tumbles never reach a hospital at all. Parents often imagine a 20-30% chance of serious injury per fall; stair-fall deaths number in the single digits nationally each year.\n","category":"kids","tags":["toddler","household"],"no_reliable_estimate":false,"perceived":{"description":"A toddler tumbling down a flight of stairs is one of the most viscerally terrifying moments in early parenthood. The sound alone is enough to trigger a 911 call. Baby gates are a multi-hundred-million-dollar market built on the assumption that an unguarded staircase is a near-certain path to skull fractures or worse. Many parents imagine the fall as a single long drop from top to bottom, mentally equating it with falling off a balcony. Pediatric ER waiting rooms are full of parents convinced the tumble they just witnessed caused brain damage.\n","rough_estimate":"~20-30% chance of serious injury per stair fall","kind":"intuition"},"native":{"display":"~2.7% of stair-fall ER visits in children under 5 require hospitalization","numerator":27,"denominator":1000,"unit":"hospitalization rate per ER-presenting stair fall in children aged 0-4","population":"US children under 5 treated in emergency departments for stair-related injuries, 1999-2008 (NEISS)"},"normalized":{"lifetime_us_adult":0.027,"display":"~2.7% probability of hospitalization per stair fall that reaches the ER (per fall event, not per US adult)","log_value":-1.57,"assumptions":"Zielinski, Rochette & Smith (2012) analyzed NEISS data (1999-2008) and found 931,886 stair-related injuries in children under 5 over 10 years, or ~93,000 ER visits per year. Of these, 2.7% required hospitalization (~2,500/year). With ~20 million US children under 5, the annual ER visit rate is ~46.5 per 10,000 children, and the annual hospitalization rate is ~1.25 per 10,000. Over the 5-year period from birth to age 5, the cumulative probability of any ER-treated stair fall is approximately 2.3%, and of hospitalization ~0.06%. The 2.7% figure used as the headline represents the per-fall-event probability of serious injury (hospitalization) given that the fall was serious enough to reach an ER. Many stair tumbles never reach a hospital at all -- the denominator of all stair falls is far larger than 93,000/year, making the true per-fall serious-injury rate even lower. Soffer et al. (2024) found only 2.0% of stair-tumble presentations were at high risk for clinically important TBI (ciTBI). Deaths are in the single digits nationally per year; no study in the literature (Joffe n=363, Chiaviello n=69, Soffer n=344) recorded a stair-fall death in their cohort.\n","uncertainty":{"low":0.01,"high":0.07},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/22412031/","title":"Stair-Related Injuries to Young Children Treated in US Emergency Departments, 1999-2008","publisher":"Pediatrics (Zielinski, Rochette, Smith)","source_type":"peer_reviewed","statistic":"93,000 ER visits per year for stair-related injuries in children under 5; 2.7% hospitalized; rate of 46.5 per 10,000 children per year","excerpt":"\"An estimated 931,886 children younger than 5 years of age were treated in US emergency departments for stair-related injuries during the 10-year study period, averaging 93,189 per year. Approximately 2.7% required hospitalization. The injury rate was 46.5 per 10,000 children.\"\n","source_date":"2012-03-12","source_accessed":"2026-04-23","archive_url":"https://web.archive.org/web/20260503094552/https://pubmed.ncbi.nlm.nih.gov/22412031/","calculation_notes":"Zielinski et al. used NEISS data (CPSC) from 1999-2008. 931,886 total injuries over 10 years = 93,189/year. 76.3% of injuries were to the head/neck. 2.7% hospitalized = ~2,500/year. Rate declined from 53.0 to 42.4 per 10,000 over the study period. One child treated in a US ED every 6 minutes for a stair injury. With ~20 million US children under 5, annual ER rate = 93,189 / 20M = 0.00466, or ~0.47% per year. Over 5 years: 1-(1-0.00466)^5 = ~2.3%. Hospitalization: 2,500/20M = 0.000125 per year; over 5 years ~0.06%.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/39332970/","title":"Low Risk of Clinically Important Traumatic Brain Injury in Children Who Tumble Down Stairs","publisher":"Journal of Pediatric Surgery (Soffer et al.)","source_type":"peer_reviewed","statistic":"Only 2.0% of children presenting after stair tumbles were at high risk for clinically important TBI; number of steps fallen does not predict injury severity","excerpt":"\"Only 2.0% of children were at high risk for clinically important traumatic brain injury. Tumbling down stairs should not be treated as equivalent to a free fall in risk assessment. The number of steps fallen does not independently predict the need for head CT.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-23","archive_url":"http://web.archive.org/web/20250204065605/https://pubmed.ncbi.nlm.nih.gov/39332970/","calculation_notes":"Soffer et al. (2024) studied 344 children who presented to a pediatric ED after stair tumbles. The 2.0% ciTBI high-risk rate is the most directly relevant clinical outcome metric -- it represents the fraction of stair-fall patients where emergency physicians should be genuinely concerned about brain injury. The finding that step count does not predict severity supports the biomechanical model that stair tumbles are a series of low-energy impacts, not a single fall from height.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/7936895/","title":"Stairway-Related Injuries in Children","publisher":"Pediatrics (Chiaviello, Christoph, Bond)","source_type":"peer_reviewed","statistic":"Of 69 children, 22% had significant injuries; 16% concussions; 7% skull fractures; 3% cerebral contusions; one subdural hematoma","excerpt":"\"Fifteen patients (22%) suffered significant injuries, including concussion in 11 (16%), skull fracture in 5 (7%), cerebral contusion in 2 (3%), subdural hematoma in 1 (1%), and a C-2 fracture in 1 (1%).\"\n","source_date":"1994-11-01","source_accessed":"2026-04-23","archive_url":"http://web.archive.org/web/20260504061122/https://pubmed.ncbi.nlm.nih.gov/7936895/","calculation_notes":"Chiaviello et al. studied 69 children at a single ED. The 22% significant injury rate is higher than Zielinski's 2.7% hospitalization rate because Chiaviello used a broader definition of \"significant\" (including concussions diagnosed clinically) and because single-ED studies skew toward more severe presentations. Chiaviello's own cohort reported a 7% overall skull fracture rate (5 of 69) with no carried-infant stratification. The separate carried-infant finding (40% skull fracture rate, 4 of 10) comes from Joffe & Ludwig (1988): the mechanism changes entirely when a caregiver drops a child from adult height onto stairs, which is biomechanically different from a toddler tumbling down steps under their own locomotion.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/3405681/","title":"Stairway Injuries in Children","publisher":"Pediatrics (Joffe M, Ludwig S)","source_type":"peer_reviewed","statistic":"363 children treated for stairway injuries; children who fell down more than four steps had no greater number or severity of injury than those who fell down four steps or fewer (P=.67); a carried-infant subgroup had a substantially higher skull-fracture rate","excerpt":"\"A total of 363 consecutive patients seen in a pediatric emergency department were studied.\" ... \"Children who fell down more than four steps had no greater number or severity of injury than those who fell down less than four steps (P = .67).\"\n","source_date":"1988-09-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20250204193659/https://pubmed.ncbi.nlm.nih.gov/3405681/","calculation_notes":"Joffe & Ludwig (1988) is the original source for the \"steps fallen doesn't predict severity\" finding (P=.67) that the body text and this entry's caveats attribute to it, and that Soffer et al. independently replicated 36 years later. It is also the origin of the carried-infant subgroup finding (4 of 10 infants sustained skull fractures when a caregiver fell while carrying them) referenced in Chiaviello's calculation_notes above and in the personal_factor_multipliers \"infant being carried\" row. n=363 is the same cohort referenced as \"Joffe n=363\" in this entry's normalized.assumptions. Previously cited only indirectly through another source's calculation_notes; added here as its own source since the body text names Joffe & Ludwig directly and quotes specific findings from it.\n"}],"comparison_anchors":[{"label":"Infant fall from furniture serious injury (per fall event)","lifetime_us_adult":0.01},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"infant being carried by caregiver on stairs","multiplier":10,"notes":"Joffe & Ludwig (1988) found a 40% skull fracture rate (4 of 10) when infants fell while being carried -- the fall height is adult-level, not a toddler-level tumble"},{"factor":"child in a baby walker on stairs","multiplier":5,"notes":"Walker-aided stair falls peak at 8 months; CPSC documented 25,700 walker-related injuries in 1992, mostly stair falls; walkers add momentum and prevent the child from arresting the fall"},{"factor":"child under 12 months (crawler/cruiser)","multiplier":2,"notes":"Children under 1 had the highest proportion of skull fractures (63.1%) and intracranial hemorrhage (65.5%) among all ages 0-4; thinner skulls and proportionally larger heads increase vulnerability"},{"factor":"child aged 2-4 years walking independently","multiplier":0.5,"notes":"Older toddlers have better protective reflexes, more skull ossification, and tend to tumble rather than fall head-first"},{"factor":"baby gate installed at top and bottom of stairs","multiplier":0.1,"notes":"Gates prevent the fall event itself; CPSC has set safety standards (ASTM F1004) specifically for stair gates after documenting injuries from accordion-style gates"}],"short_label":"Toddler stair fall","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 2.7% hospitalization rate is conditioned on ER presentation -- it does not capture the vast majority of stair tumbles that never reach a hospital. The true per-tumble serious-injury rate is substantially lower but cannot be estimated because the denominator (all stair falls, including those managed at home) is unknown. The Chiaviello and Joffe studies are single-ED case series from the 1980s and 1990s with small sample sizes; their injury proportions may not reflect current national patterns. The critical distinction is between a toddler tumbling under their own locomotion (overwhelmingly benign) and an infant being dropped from caregiver height onto stairs (genuinely dangerous). These are biomechanically different events that the aggregate statistics conflate. Stair tumbles are not equivalent to free falls of the same total height -- Joffe (1988) and Soffer (2024) both found no correlation between number of steps and injury severity, because a tumble is a series of low-energy sequential impacts, not one high-energy impact.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-23","reviewed":true,"generated_at":"2026-04-23","image":{"alt":"A residential staircase with a baby gate at the top, viewed from below, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/toddler-stair-fall-injury","api_url":"https://likelier.app/api/fears/toddler-stair-fall-injury.json"},{"slug":"cosleeping-infant-death","question":"What are the odds of an infant dying from co-sleeping or bed-sharing?","quick_answer":"For a breastfed infant of non-smoking, non-drinking parents, the modeled first-year bed-sharing SIDS risk is about 1 in 4,348. For a higher-risk configuration (bottle-fed, both parents smoking, maternal alcohol use) it rises to roughly 1 in 36. The risk is real but depends heavily on these factors, which most parents sense but cannot quantify.\n","category":"kids","tags":["infant"],"no_reliable_estimate":false,"perceived":{"description":"Bed-sharing occupies a peculiar spot in new-parent risk perception: it is simultaneously widespread practice and official taboo. The AAP recommends against it; attachment-parenting communities promote it; exhausted parents do it whether or not they planned to. Most parents who bed-share sense that the risk is real but small, while most who avoid it believe they are dodging something meaningfully dangerous. Neither group could produce a number. The emotional weight comes partly from conflation with SIDS (see the sibling entry), even though the mechanisms differ.\n","rough_estimate":"Most parents sense elevated risk but cannot quantify it; estimates vary wildly by parenting philosophy","kind":"intuition"},"native":{"display":"~23 SIDS deaths per 100,000 live births in the first year of life, for a breastfed infant of non-smoking, non-drinking parents who bed-share","numerator":1,"denominator":4348,"unit":"per live-born infant during the first year of life","population":"breastfed infants of non-smoking parents with no maternal alcohol or drug use, bed-sharing"},"normalized":{"lifetime_us_adult":0.0275,"display":"~1 in 36 (per infant, first year, higher-risk configuration: bottle-fed + both parents smoking + maternal alcohol use)","log_value":-1.56,"assumptions":"The native rate (1 in ~4,348; 0.23 per 1,000 live births) is Carpenter et al. 2013's own modeled first-year absolute risk for the lowest-risk bed-sharing group: breastfed infant, neither parent smoking, no maternal alcohol or drug use (Table 4, \"minimum risk\" row). This is a cumulative first-year-of-life rate, not a per-night figure — the paper reports no per-night estimate anywhere. The normalized headline uses the highest row modeled in that same table: bottle-fed infant, both parents smoking, maternal alcohol use, still for an otherwise \"normal\" mother (26-30 years old, second child, normal birth weight) — predicted first-year bed-sharing rate 27.5 per 1,000 (~1 in 36). Carpenter et al.'s discussion section separately models a more extreme combination of demographic and behavioral risk factors (18-year-old mother, 2.25 kg bottle-fed infant, both parents smoking, maternal alcohol use) at 125 per 1,000 (12.5%, ~1 in 8) — the ceiling reported anywhere in the paper. That extreme figure is used as the top of the uncertainty band and in regional_breakdown rather than as the headline, since it stacks demographic risk factors (young maternal age, low birth weight) on top of the behavioral ones. All three figures are Carpenter et al.'s own modeled outputs; none are compounded or re-derived by this entry.\n","uncertainty":{"low":0.00023,"high":0.125},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/23793691/","title":"Bed sharing when parents do not smoke: is there a risk of SIDS? An individual level analysis of five major case-control studies","publisher":"BMJ Open — Carpenter, McGarvey, Mitchell, Tappin, Vennemann, Smuk, Carpenter","source_type":"primary_study","statistic":"Modeled first-year cumulative SIDS risk during bed-sharing: 0.23 per 1,000 live births for the lowest-risk group (breastfed, neither parent smoking, no maternal alcohol/drug use), rising to 27.5 per 1,000 for the highest-risk group modeled in the paper's main table (bottle-fed, both parents smoking, maternal alcohol use); a separate modeled scenario stacking young maternal age and low birth weight on top of those behavioral factors reaches 125 per 1,000. The paper reports no per-night estimate.","excerpt":"\"Bed sharing for sleep when the parents do not smoke or take alcohol or drugs increases the risk of SIDS. [...] The absolute risk estimates were very low at 0.08 per 1,000 live births for room sharing infants, rising to 0.23 per 1,000 when bed sharing occurred. [...] Table 4 shows that for room sharing breast-fed babies placed supine, and whose parents do not smoke and have no other risk factors, the SIDS rate is predicted to be 0.08 (0.05 to 0.14)/1000 live-births. This rate is predicted to increase by 2.7 times (1.4 to 5.3) to 0.23 (0.11 to 0.49)/1000 when bed sharing. [...] a 2.25 kg bottle-fed baby bed sharing with an 18-year-old mother, who smokes and regularly takes 2+ units of alcohol and whose partner also smokes, has a predicted SIDS rate of 125/1000, that is, 12.5% (see online supplementary table b).\"\n","source_date":"2013-01-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426195226/https://pubmed.ncbi.nlm.nih.gov/23793691/","calculation_notes":"Carpenter et al. pooled individual-level data from five major SIDS case-control studies (1,472 SIDS cases, 4,679 controls) and modeled predicted SIDS rates (Table 4) for a \"normal\" cohabiting mother aged 26-30 with a second normal-weight baby. All rates are CUMULATIVE FIRST-YEAR (ages 0 up to 1 year) absolute risks per 1,000 live births — the paper reports no per-night estimate anywhere. Minimum-risk row (breastfed, neither parent smoking, no alcohol): bed sharing 0.23 (0.11 to 0.49)/1000 live-births versus 0.08 (0.05 to 0.14)/1000 for room sharing (ratio 2.7). This 0.23/1000 (1 in ~4,348) is the native rate. The highest row modeled in the same table (bottle-fed, both parents smoking, maternal alcohol use, same \"normal\" mother profile): bed sharing 27.5 (10.4 to 68.4)/1000, used as the normalized headline (1 in ~36). A separate, more extreme combination modeled in the Discussion section (2.25 kg bottle-fed infant, 18-year-old mother who smokes and drinks 2+ units of alcohol, smoking partner) reaches 125/1000 (12.5%, ~1 in 8) — the ceiling reported anywhere in the paper, used as the top of the uncertainty band. None of these figures involve per-night compounding; they are the paper's own first-year model outputs.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25238618/","title":"Bed-Sharing in the Absence of Hazardous Circumstances: Is There a Risk of Sudden Infant Death Syndrome? An Analysis from Two Case-Control Studies Conducted in the UK","publisher":"PLOS One — Blair, Sidebotham, Pease, Fleming","source_type":"peer_reviewed","statistic":"Adjusted OR for SIDS/SUID with bed-sharing in the absence of hazards was not significant (OR=1.1, 95% CI 0.6-2.0); sofa-sharing OR=18.3 and alcohol >2 units OR=18.3; smoking for infants under 3 months OR=8.9","excerpt":"\"The multivariable risk associated with bed-sharing in the absence of these hazards was not significant overall (OR = 1.1 [95% CI: 0.6-2.0]), for infants less than 3 months old (OR = 1.6 [95% CI: 0.96-2.7]), and was in the direction of protection for older infants (OR = 0.1 [95% CI: 0.01-0.5]). [...] Sofa-sharing carried extremely high risk (OR=18.3). Co-sleeping with alcohol-intoxicated parents showed very elevated risk (OR=18.3). Parental smoking posed significant danger for infants under 3 months (OR=8.9).\"\n","source_date":"2014-06-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503085758/https://pubmed.ncbi.nlm.nih.gov/25238618/","calculation_notes":"Blair et al. reanalyzed two large UK case-control datasets. The OR of ~2.9 for bed-sharing in the absence of hazards (from the broader meta-analytic literature including Vennemann 2012) is used as the baseline multiplier. With parental smoking the OR rises to ~6.3; with alcohol and smoking combined the effective OR reaches ~18 — a similar order of magnitude to Carpenter et al.'s own bed-sharing \"both smoke\" multiplier (Table 3: ~15.6x at 3 months and over). These ORs are shown for context on how risk factors stack; they are not compounded against a per-night baseline to derive this entry's native/normalized headline figures, which come directly from Carpenter et al.'s own modeled first-year rates (see that source's calculation_notes). Age-stratified, the hazard-free bed-sharing OR was not significant overall (1.1, 95% CI 0.6-2.0) and not quite significant for infants under 3 months (1.6, CI 0.96-2.7), but was statistically significant and protective for infants 3 months and older (0.1, CI 0.01-0.5), though the source paper flags that older-infant subgroup as numerically small.\n"},{"url":"https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","title":"Data and Statistics for SUID and SIDS","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"In 2022, ~3,700 SUID deaths in the US; 1,040 classified as accidental suffocation and strangulation in bed (ASSB)","excerpt":"\"In 2022, there were about 3,700 sudden unexpected infant deaths (SUID) in the United States. These deaths occur among infants less than 1 year old and have no immediately obvious cause. [...] 1,040 deaths from accidental suffocation and strangulation in bed.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420034001/https://www.cdc.gov/sudden-infant-death/data-research/data/index.html","calculation_notes":"1,040 ASSB deaths out of ~3.67 million US live births in 2022 ≈ 28.3 per 100,000 live births, or roughly 1 in 3,530 infants in the first year of life. Not all ASSB deaths involve bed-sharing (some occur in cribs with soft bedding, car seats, or other surfaces), and not all bed-sharing deaths are classified as ASSB (some fall under SIDS or unknown cause). The ASSB figure provides an order-of-magnitude anchor for suffocation-mechanism deaths rather than a direct measure of bed-sharing risk.\n","independence_note":"CDC vital registration data. The same underlying NCHS mortality files inform the sibling SIDS entry and the AAP 2022 policy statement; treat as methodologically linked sources.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35726558/","title":"Sleep-Related Infant Deaths: Updated 2022 Recommendations for Reducing Infant Deaths in the Sleep Environment","publisher":"Pediatrics (American Academy of Pediatrics) — Moon, Carlin, Hand et al.","source_type":"peer_reviewed","statistic":"AAP recommends against bed-sharing; ~3,500 US sleep-related infant deaths per year; bed-sharing identified as a modifiable risk factor","excerpt":"\"Each year in the United States, ∼3500 infants die of sleep-related infant deaths, including sudden infant death syndrome (SIDS) [...] It is recommended that infants sleep on a separate, flat, noninclined sleep surface [...] It is recommended that infants not sleep on adult beds, couches, or armchairs.\"\n","source_date":"2022-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260413183441/https://pubmed.ncbi.nlm.nih.gov/35726558/","calculation_notes":"AAP 2022 policy statement synthesizing the epidemiological literature. Provides the institutional recommendation context: bed-sharing is listed as a modifiable risk factor. The ~3,500 figure covers all sleep-related infant deaths (SIDS + ASSB + unknown), not bed-sharing specifically.\n","independence_note":"Policy synthesis drawing on the same Carpenter, Blair, and CDC datasets cited above; not an independent estimate.\n"}],"comparison_anchors":[{"label":"SIDS (per US infant, first year of life)","lifetime_us_adult":0.00014},{"label":"Choking death (lifetime, US adult)","lifetime_us_adult":0.00091},{"label":"Drowning (lifetime, US adult)","lifetime_us_adult":0.00086},{"label":"Car crash death (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Minimum risk (breastfed, neither parent smokes, no alcohol)","probability":0.00023,"notes":"Carpenter et al. 2013, Table 4: 0.23 per 1,000 live births, first year of life, bed-sharing"},{"region":"Moderate risk (breastfed, both parents smoke, no alcohol)","probability":0.00188,"notes":"Carpenter et al. 2013, Table 4: 1.88 per 1,000 live births, first year of life, bed-sharing"},{"region":"Higher risk (bottle-fed, both parents smoke, maternal alcohol use)","probability":0.0275,"notes":"Carpenter et al. 2013, Table 4: 27.5 per 1,000 live births, first year of life, bed-sharing — the highest row modeled in the paper's main table"},{"region":"Extreme stacked risk (18-year-old mother, 2.25kg bottle-fed infant, both parents smoke, maternal alcohol use)","probability":0.125,"notes":"Carpenter et al. 2013, Discussion section: 125 per 1,000 live births (12.5%), first year of life, bed-sharing — the most extreme combination modeled anywhere in the paper"}],"personal_factor_multipliers":[{"factor":"parental smoking","multiplier":8.9,"notes":"Blair et al. OR 8.9 for bed-sharing with smoking parent, infants under 3 months"},{"factor":"maternal alcohol consumption","multiplier":3,"notes":"Blair et al. 2014; compounds with smoking"},{"factor":"infant under 12 weeks","multiplier":10.4,"notes":"Vennemann 2012 pooled OR 10.37 for bed-sharing with infant <12 weeks"},{"factor":"soft bedding or couch/armchair","multiplier":5,"notes":"AAP 2022; couches and armchairs carry the highest suffocation risk"},{"factor":"breastfeeding (protective)","multiplier":0.5,"notes":"Carpenter et al. 2013; breastfeeding roughly halves the risk vs formula"},{"factor":"premature / low birth weight","multiplier":4,"notes":"Consistent across Carpenter and Vennemann analyses"}],"short_label":"Co-sleeping death","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry is distinct from the sibling SIDS entry. SIDS is a diagnosis of exclusion that occurs regardless of sleep surface; bed-sharing deaths are primarily attributable to accidental suffocation and strangulation in bed (ASSB), a mechanistically different category even though both fall under the SUID umbrella. The baseline risk (~1 in 4,348; 0.23 per 1,000 live births in the first year of life) comes from Carpenter et al.'s modeled estimates, not direct measurement, and assumes the lowest-risk configuration; the source paper reports no per-night figure. Real-world bed-sharing is heterogeneous: a breastfeeding mother without alcohol or smoking faces a fundamentally different risk profile than a parent who smokes and drinks. The higher-risk figures in the normalized field and regional_breakdown are Carpenter et al.'s own modeled first-year rates for infants in each named risk category (bottle-fed plus parental smoking plus alcohol, and beyond that a stacked scenario adding young maternal age and low birth weight) — they are not extrapolated from a single occasional exposure and should not be read as the risk for a family that only occasionally bed-shares under low-risk conditions. CDC's ASSB category (~1,040 deaths in 2022) is the closest surveillance proxy for suffocation-mechanism deaths, but not all ASSB involves bed-sharing and not all bed-sharing deaths are coded as ASSB. Racial and socioeconomic disparities mirror those seen in SIDS: American Indian / Alaska Native and non-Hispanic Black infants are disproportionately represented in SUID data. This entry publishes risk estimates, not sleep recommendations; the AAP's clinical guidance is the appropriate reference for practice decisions.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A minimal flat illustration of a bed with a small pillow beside it, rendered in muted tones against a soft background."},"canonical_url":"https://likelier.app/cosleeping-infant-death","api_url":"https://likelier.app/api/fears/cosleeping-infant-death.json"},{"slug":"medication-misuse-interaction","question":"What are the odds of being harmed by not reading drug labels or mixing medications?","quick_answer":"About 1 in 35 over a lifetime that a US adult's death is attributable to medication non-adherence, dosing error, or a drug interaction. Most people would place their risk well below 1 in 100.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"No standing survey isolates public perception of harm from medication non-adherence or drug-drug interactions as a discrete fear. Most adults file drug-label compliance under \"common sense I already follow,\" which is precisely the problem: CDC data show that roughly half of patients with chronic conditions do not take medications as directed, yet almost none of them would describe themselves as at elevated risk. Polypharmacy interactions occupy an even deeper blind spot. The average reader knows vaguely that \"grapefruit and statins\" is a thing, but cannot name the mechanism (CYP3A4 inhibition) or estimate the magnitude (up to 260% elevation in simvastatin blood levels). The gap between that hazy awareness and the 125,000 deaths attributed annually to non-adherence alone is one of the wider perception mismatches in the medication-risk space.\n","rough_estimate":"Most adults would place their lifetime risk of serious harm from medication misuse well below 1 in 100","kind":"intuition"},"native":{"display":"~125,000 US deaths/year from non-adherence; ~1.5 million ER visits/year for adverse drug events; ~99,628 emergency hospitalisations/year in adults 65+","numerator":125000,"denominator":258000000,"unit":"per year","population":"US adults, all ages"},"normalized":{"lifetime_us_adult":0.0284,"display":"~1 in 35 lifetime (US adult, fatal medication misuse or interaction)","log_value":-1.547,"assumptions":"The 125,000 deaths/year figure is the widely cited estimate for US deaths attributable to medication non-adherence, drawn from Osterberg & Blaschke (NEJM 2005) and corroborated by the Annals of Internal Medicine and CDC medication-safety literature. Against a US adult population of ~258 million, this gives a per-adult-year hazard of ~4.84 x 10^-4. Compounded over 59 years of remaining adult life (from age 18): 1 - (1 - 0.000484)^59 ≈ 0.0282, or roughly 1 in 35. This figure encompasses deaths from non-adherence (skipped doses, premature discontinuation), incorrect dosing (not reading labels), and drug-drug/drug-food interactions (polypharmacy, grapefruit- statin, warfarin-vitamin K). It deliberately excludes intentional overdose, which is tracked under drug-overdose. The number is conservative in one direction (many non-adherence deaths are attributed to the underlying disease rather than the treatment failure) and aggressive in another (the 125,000 figure has been contested as poorly sourced). The uncertainty band reflects this methodological spread.\n","uncertainty":{"low":0.017,"high":0.045},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/22111719/","title":"Emergency Hospitalizations for Adverse Drug Events in Older Americans","publisher":"New England Journal of Medicine / Budnitz DS, Lovegrove MC, Shehab N, Richards CL","source_type":"peer_reviewed","statistic":"An estimated 99,628 emergency hospitalisations per year for adverse drug events in US adults aged 65+, 2007-2009; nearly two thirds due to unintentional overdoses; warfarin, insulin, oral antiplatelet agents, and oral hypoglycaemic agents most commonly implicated","excerpt":"\"There were an estimated 99,628 emergency hospitalizations (95% confidence interval [CI], 55,531 to 143,724) for adverse drug events in U.S. adults 65 years of age or older each year from 2007 through 2009. Nearly half of these hospitalizations were among adults 80 years of age or older (48.1%).\"\n","source_date":"2011-11-24","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420043531/https://pubmed.ncbi.nlm.nih.gov/22111719/","calculation_notes":"Budnitz et al. used nationally representative NEISS-CADES data to estimate ~99,628 annual emergency hospitalisations for ADEs in adults 65+. Nearly two thirds (65.7%) were due to unintentional overdoses — i.e., dosing errors or failure to read labels correctly — rather than idiosyncratic reactions. The four most commonly implicated drug classes (warfarin, insulin, oral antiplatelets, oral hypoglycaemics) are precisely the medications where label-reading and interaction awareness matter most. This study anchors the \"older adult\" slice of the entry and validates the claim that label misuse and interaction ignorance, not exotic side-effects, drive the bulk of ADE hospitalisations.\n","independence_note":"Independent of the CDC FastStats aggregate and the Osterberg non-adherence mortality estimate. Uses the same NEISS-CADES surveillance platform as CDC FastStats but reports a distinct age-stratified analysis from a different time window.\n"},{"url":"https://www.cdc.gov/medication-safety/data-research/facts-stats/index.html","title":"FastStats: Medication Safety Data","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"More than 1.5 million US ED visits per year for adverse drug events; almost 500,000 require hospitalisation; adults 65+ account for more than 600,000 ED visits","excerpt":"\"More than 1.5 million people visit emergency departments for ADEs each year in the United States, and almost 500,000 require hospitalization. Older adults (65 years or older) visit emergency departments more than 600,000 times each year, more than twice as often as younger people.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260423050603/https://www.cdc.gov/medication-safety/data-research/facts-stats/index.html","calculation_notes":"CDC's aggregate ADE surveillance provides the contemporary US denominator. Of the ~1.5 million annual ED visits, a substantial fraction involve medication misuse (wrong dose, missed dose, drug interaction) rather than idiosyncratic reactions to correctly taken medications. The 500,000 hospitalisations/year figure is the basis for the non-fatal serious-harm estimate. Combined with the 125,000 deaths/year non-adherence figure, it implies roughly 375,000 hospitalisations/year result in recovery — consistent with the overall case-fatality rate in the ADE literature.\n","independence_note":"CDC FastStats is the authoritative US national surveillance estimate. It shares the NEISS-CADES platform with Budnitz et al. but reports a broader age range and more recent time window.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6045499/","title":"The Unmet Challenge of Medication Nonadherence","publisher":"The Permanente Journal / Kleinsinger F","source_type":"peer_reviewed","statistic":"Medication non-adherence causes approximately 125,000 preventable deaths and $100 billion in preventable medical costs annually in the US; 40-50% of patients with chronic diseases do not adhere to prescribed regimens","excerpt":"\"Nonadherence is thought to cause at least 100,000 preventable deaths and $100 billion in preventable medical costs per year.\"\n","source_date":"2018-06-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420043640/https://pmc.ncbi.nlm.nih.gov/articles/PMC6045499/","calculation_notes":"Kleinsinger's review synthesises the medication non-adherence literature and reports the 100,000-125,000 deaths/year figure that anchors this entry's native numerator. The 40-50% chronic-disease non-adherence rate is the behavioural denominator: roughly half of all adults on long-term medications are not taking them as directed, and a fraction of those non-adherent patients die from the resulting treatment failure. The $100 billion cost estimate is not used in the probability calculation but contextualises the scale of the problem.\n","independence_note":"Kleinsinger is a narrative review synthesising prior primary studies (including Osterberg & Blaschke NEJM 2005 and others). It is not independent of those upstream estimates but provides the most accessible modern summary of the 125,000 figure.\n"},{"url":"https://www.fda.gov/consumers/consumer-updates/grapefruit-juice-and-some-drugs-dont-mix","title":"Grapefruit Juice and Some Drugs Don't Mix","publisher":"U.S. Food and Drug Administration (FDA)","source_type":"govt_report","statistic":"Grapefruit can increase blood levels of certain statins (simvastatin, atorvastatin) by inhibiting CYP3A4; one study found a 260% increase in simvastatin blood levels with grapefruit juice","excerpt":"\"Grapefruit juice can cause the body to metabolize drugs abnormally, resulting in higher or lower levels of the drug in the blood. When there is too much drug in the blood, you may have more side effects.\"\n","source_date":"2021-07-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260322055802/https://www.fda.gov/consumers/consumer-updates/grapefruit-juice-and-some-drugs-dont-mix","calculation_notes":"The FDA consumer update documents the CYP3A4-mediated grapefruit-statin interaction mechanism that is the canonical example of a food-drug interaction in this entry. The 260% simvastatin blood-level increase is the headline figure used to illustrate how a common food can turn a safe medication into a dangerous one. This source is used for the drug-interaction narrative, not for the headline mortality calculation.\n","independence_note":"Independent of the CDC and Budnitz sources. FDA consumer guidance based on pharmacokinetic studies, not epidemiological surveillance.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10337167/","title":"Prevalence and trends of polypharmacy in U.S. adults, 1999-2018","publisher":"Global Health Research and Policy / Wang X, Liu K, Shirai K, Tang C, Hu Y, Wang Y, Hao Y, Dong JY","source_type":"peer_reviewed","statistic":"Polypharmacy (5+ concurrent medications) prevalence among US adults aged 65+ rose from 23.5% (1999-2000) to 44.1% (2017-2018)","excerpt":"\"The polypharmacy prevalence was consistently higher in participants aged 65 years and older ... the prevalence increased from 23.5% (95% CI 20.6 to 26.4) to 44.1% (95% CI 40.4 to 47.8).\"\n","source_date":"2023-07-12","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260404180610/https://pmc.ncbi.nlm.nih.gov/articles/PMC10337167/","calculation_notes":"Grounds the body claim that roughly 40% of US adults over 65 meet the polypharmacy threshold: NHANES-based estimate of 44.1% by 2017-2018, up from 23.5% in 1999-2000. \"Roughly 40%\" is a conservative rounding of the most recent (2017-2018) figure. Used as descriptive context for why polypharmacy is a structural accelerant of medication-misuse/ interaction risk; not part of the headline native/normalized calculation.\n","independence_note":"Independent of the CDC, Budnitz, and Kleinsinger sources above; uses NHANES survey data rather than ED/hospitalisation surveillance or narrative-review synthesis.\n"}],"comparison_anchors":[{"label":"Serious adverse drug reaction (lifetime, US adult)","lifetime_us_adult":0.0171},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from drug overdose (lifetime, US adult)","lifetime_us_adult":0.0237},{"label":"Fatal anaphylaxis (lifetime, US adult)","lifetime_us_adult":0.0000363}],"personal_factor_multipliers":[{"factor":"polypharmacy (5+ concurrent medications)","multiplier":6,"notes":"Drug-drug interactions scale non-linearly with the number of concurrent agents. A patient on 5+ medications faces roughly 6x the population-average risk of a harmful interaction or dosing error.\n"},{"factor":"age 75+","multiplier":5,"notes":"Budnitz et al. found nearly half of ADE hospitalisations were in adults 80+. Age is a proxy for polypharmacy, declining renal clearance, and cognitive difficulty with complex regimens.\n"},{"factor":"single medication, healthy adult under 50","multiplier":0.15,"notes":"Young adults on one well-tolerated medication with normal organ function sit far below the population average. The headline number substantially overstates this reader's personal risk.\n"},{"factor":"no pharmacist medication review in past year","multiplier":2,"notes":"Maher et al. (2014, BJGP) and AGS Beers Criteria: pharmacist-led medication reviews detect and resolve clinically significant interactions and dosing errors; patients who have not had a recent review have roughly 2× the rate of undetected harmful interactions, especially under polypharmacy"},{"factor":"CYP2D6 or CYP3A4 poor metabolizer genotype","multiplier":2.5,"notes":"Pharmacogenomic literature: approximately 7–10% of the US population are CYP2D6 poor metabolizers and 5–7% are CYP3A4 low expressers; these genotypes roughly double drug accumulation for affected medications, increasing toxicity risk without dose changes"}],"short_label":"Medication misuse","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 125,000 deaths/year figure for medication non-adherence is widely cited but poorly sourced at its origin — it traces to a chain of secondary citations rather than a single definitive primary study, and the true number may be anywhere from 75,000 to 200,000 depending on how strictly \"non-adherence death\" is defined. The entry deliberately uses the 125,000 midpoint and places a wide uncertainty band around the lifetime figure. This entry covers harm from misuse (wrong dose, skipped dose, unrecognised interaction) of medications taken with therapeutic intent; it does not cover intentional overdose or recreational drug misuse, which are tracked under drug-overdose. The grapefruit-statin and warfarin-vitamin K examples are illustrative of the interaction mechanism, not the primary drivers of mortality — anticoagulant dosing errors and insulin errors are the bigger killers in the Budnitz data. Readers on a single medication with no comorbidities face a risk well below the headline; readers on complex multi-drug regimens face a risk several multiples above it.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Several overlapping pill shapes in muted tones against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/medication-misuse-interaction","api_url":"https://likelier.app/api/fears/medication-misuse-interaction.json"},{"slug":"vehicle-theft","question":"What are the odds of having your car stolen?","quick_answer":"Having your car stolen runs about 1 in 34 over a typical 12-year ownership period — roughly 250 thefts per 100,000 registered vehicles a year. Many people put it nearer 1 in 10 lifetime, well above the actual rate.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Gallup's annual crime-worry poll consistently places vehicle theft among the most-worried-about property crimes. In the October 2025 wave, 39% of US adults said they worry frequently or occasionally about having their car stolen or broken into. That figure has held in the mid-to- high 30s for over a decade, even as the actual theft rate has swung by 40% in both directions during that period. Vehicle theft sits just below home burglary in the worry ranking and well above violent crimes like homicide and sexual assault.\n","rough_estimate":"~1 in 10 lifetime feels about right to many respondents","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"native":{"display":"~250 per 100,000 registered vehicles per year (2024)","numerator":250,"denominator":100000,"unit":"per year","population":"Registered motor vehicles in the US"},"normalized":{"lifetime_us_adult":0.0296,"display":"~1 in 34 over 12 years of typical vehicle ownership","log_value":-1.53,"assumptions":"In 2024, NICB reported 850,708 motor vehicles stolen out of approximately 283 million registered vehicles (FHWA), yielding an annual per-vehicle theft rate of about 0.0030 (850,708 / 283,000,000). The average vehicle ownership period is roughly 12 years (based on average vehicle age of 12.6 years per S&P Global Mobility 2024). Compounding over 12 years: 1 − (1 − 0.00301)^12 ≈ 0.0355. However, the 2024 rate is already well below the 2023 peak (1.02M thefts), and the long-run average (2015–2024) is closer to 0.0025 per vehicle-year. Using the 10-year average gives 1 − (1 − 0.0025)^12 ≈ 0.0296. The central estimate uses this smoothed figure. Scope is activity_specific_lifetime because the denominator is per vehicle over an ownership period, not per person over a biological lifetime.\n","uncertainty":{"low":0.018,"high":0.042},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.nicb.org/news/news-releases/vehicle-thefts-united-states-fell-17-2024","title":"Vehicle Thefts in United States Fell 17% in 2024","publisher":"National Insurance Crime Bureau","source_type":"reputable_reference","statistic":"850,708 vehicles stolen in the US in 2024; national rate 250.2 per 100,000 residents; 17% decline from 2023","excerpt":"\"In 2024, 850,708 vehicles were stolen nationwide, a 17% decline from 2023 when 1,020,729 vehicles were stolen, marking the largest annual decrease in stolen vehicles in the last 40 years.\"\n","source_date":"2025-03-18","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260321025934/https://www.nicb.org/news/news-releases/vehicle-thefts-united-states-fell-17-2024","calculation_notes":"NICB compiles theft data from law enforcement agencies nationwide. The 850,708 figure for 2024 is divided by approximately 283 million registered vehicles (FHWA Highway Statistics) to yield a per-vehicle annual rate of ~0.00301. The NICB-reported rate of 250.2 per 100,000 residents uses population as the denominator; the per-vehicle rate is slightly lower because there are more registered vehicles than people. Compounded over 12 years of ownership: 1 − (1 − 0.00301)^12 ≈ 0.0355. Using the 2015–2024 average rate of ~0.0025 gives 1 − (1 − 0.0025)^12 ≈ 0.0296.\n","independence_note":"NICB aggregates data from law enforcement agencies; the FBI Crime Data Explorer independently reports motor vehicle theft through the UCR/NIBRS pipeline. Both converge on similar totals.\n"},{"url":"https://cde.ucr.cjis.gov/LATEST/resources/reports/UCR%20Summary%20of%20Reported%20Crimes%20in%20the%20Nation%202024.pdf","title":"UCR Summary of Reported Crimes in the Nation, 2024","publisher":"FBI Crime Data Explorer","source_type":"govt_report","statistic":"Motor vehicle theft rate of 258.8 per 100,000 inhabitants in 2024; estimated 19.4% decrease from 2023","excerpt":"\"In 2024, the estimated motor vehicle theft rate dropped 19.4% compared with the previous year. The motor vehicle theft rate was 258.8 per 100,000 inhabitants.\"\n","source_date":"2025-07-02","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260318045311/https://cde.ucr.cjis.gov/LATEST/resources/reports/UCR%20Summary%20of%20Reported%20Crimes%20in%20the%20Nation%202024.pdf","calculation_notes":"FBI UCR data uses a population denominator (inhabitants, not vehicles). The 258.8 per 100,000 inhabitants figure is consistent with NICB's 250.2 per 100,000 residents — the small difference reflects slightly different population estimates and reporting coverage. Both confirm the order of magnitude: roughly 850K–880K thefts in 2024. The 19.4% year-over-year drop is the largest single-year decline in four decades, following the post-pandemic theft spike driven partly by Hyundai/Kia ignition vulnerabilities.\n","independence_note":"FBI UCR/NIBRS is an independent law-enforcement reporting pipeline. NICB data comes from insurance-industry and law-enforcement sources. Agreement on the theft total and direction of change provides strong cross-validation.\n"},{"url":"https://www.nicb.org/news/news-releases/vehicles-stolen-keys-left-inside-rise","title":"Vehicles Stolen With Keys Left Inside On The Rise","publisher":"National Insurance Crime Bureau","source_type":"reputable_reference","statistic":"By the end of 2021, just over 100,000 thefts facilitated by keys or key fobs were reported nationally, accounting for 11% of all US vehicle thefts that year; up more than 20% from 2019.","excerpt":"\"A recent NICB report analyzed vehicle theft data from 2019 through 2021 and found an increase of more than 20% in thefts facilitated by keys. This also includes vehicle thefts where key fobs were left inside the vehicle. By the end of 2021, just over 100,000 thefts facilitated by keys or key fobs were reported nationally. This total accounts for 11% of vehicle thefts of all types reported in the U.S. in 2021.\"\n","source_date":"2023-01-26","source_accessed":"2026-05-25","archive_url":"https://web.archive.org/web/20241206014848/https://www.nicb.org/news/news-releases/vehicles-stolen-keys-left-inside-rise","calculation_notes":"Supports the keys-in / unlocked dimension of the caveats. The 11%-of-thefts figure is a share-of-incidents statistic, not a per-vehicle-night rate ratio, so it cannot be converted to a clean relative-risk multiplier without an estimate of what fraction of parked-vehicle nights have keys left inside (NICB notes the true share is likely higher than reported because drivers under-disclose). The number is cited as a behavioural-risk anchor in the caveats rather than added to personal_factor_multipliers, where every other entry is grounded in a directly measured rate ratio (HLDI claim rates, FBI UCR metro vs national, etc.).\n","independence_note":"NICB compiles this figure from member-insurer claims and law-enforcement reports. It is the only US-wide tally of keys-facilitated thefts; FBI UCR does not break thefts out by circumstance.\n"},{"url":"https://www.iihs.org/news/detail/anti-theft-software-tamps-down-viral-theft-trend-targeting-hyundai-kia-vehicles","title":"Anti-theft software tamps down viral theft trend targeting Hyundai, Kia vehicles","publisher":"Insurance Institute for Highway Safety / Highway Loss Data Institute","source_type":"reputable_reference","statistic":"Affected Hyundai/Kia models that received the free anti-theft software update by December 2023 had theft claim frequencies 53% lower than non-upgraded vehicles; whole-vehicle theft fell 64% for upgraded vehicles.","excerpt":"\"Those vehicles that received it as of December 2023 — a total of 30% of the eligible Hyundais and 28% of the eligible Kias in HLDI's database — had theft claim frequencies that were 53% lower than vehicles that didn't get the upgrade. The frequency of whole vehicle theft, which excludes thefts of parts and contents, fell by a larger 64% for vehicles with the upgrade.\"\n","source_date":"2024-08-07","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260515173819/https://www.iihs.org/news/detail/anti-theft-software-tamps-down-viral-theft-trend-targeting-hyundai-kia-vehicles","calculation_notes":"A 53% lower theft claim frequency for upgraded vehicles implies non-upgraded affected vehicles run at roughly 1 / (1 − 0.53) ≈ 2.1× the upgraded rate; the 64% whole-vehicle reduction implies ≈ 2.8× for theft of the entire car. The personal_factor_multiplier uses a conservative 2.0× to capture \"affected make/model still missing the software patch\" relative to a patched equivalent. This is a within-fleet update-status ratio, distinct from the separate \"no factory engine immobilizer at all\" factor (HLDI 2023, 6–8×) which compares affected Hyundai/Kia to the broader vehicle population.\n","independence_note":"HLDI/IIHS derives this from member-insurer collision/theft claim data, independent of the NICB and FBI UCR counts that anchor the headline rate. It cross-validates the Hyundai/Kia driver of the 2022–2023 surge and its 2024 reversal.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US adult)","lifetime_us_adult":0.072},{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"personal_factor_multipliers":[{"factor":"Urban core vs. suburban/rural parking","multiplier":4,"notes":"FBI UCR 2022 and NICB 2024: metro areas like Washington D.C. (842 per 100,000) and San Francisco-Oakland (477 per 100,000) show 3–5× the national average rate of 250 per 100,000; 4× used as a central estimate."},{"factor":"2019–2023 Kia/Hyundai without engine immobilizer","multiplier":7,"notes":"Highway Loss Data Institute (HLDI) 2023: Kia and Hyundai models lacking factory engine immobilizers experienced theft claim rates 6–8× higher than comparable vehicles; HLDI report on Hyundai/Kia theft rates, 2023."},{"factor":"Overnight street parking vs. locked private garage","multiplier":3,"notes":"NICB and insurance industry data consistently show vehicles stored overnight in locked garages have approximately one-third the theft rate of vehicles left on public streets; NICB Vehicle Theft Report 2022."},{"factor":"Pre-2000 vehicle without immobilizer","multiplier":2,"notes":"HLDI research shows vehicles manufactured before factory immobilizer mandates (pre-2000 in the US) are approximately 2× more likely to be stolen than post-immobilizer equivalents in the same geographic area; HLDI report on theft deterrents, 2022."},{"factor":"Affected Hyundai/Kia without the free anti-theft software update","multiplier":2,"notes":"IIHS/HLDI Bulletin 42-07 (Aug 2024): affected 2015–2019 Hyundai and Kia models that did NOT receive the free anti-theft software update had theft claim frequencies 53% higher than upgraded ones (whole-vehicle theft 64% lower for upgraded vehicles), i.e. roughly 2× the rate — these makes drove the 2022–2023 'Kia Challenge' surge."}],"short_label":"Vehicle theft","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The per-vehicle rate masks enormous geographic variation. The District of Columbia's 2024 rate was 842 per 100,000 residents — more than three times the national average. Metro areas like San Francisco-Oakland and Bakersfield topped 477 per 100,000. Meanwhile, many rural counties recorded rates well below 100 per 100,000. Vehicle make and model matter too: the Hyundai/Kia ignition vulnerability (models lacking engine immobilizers) drove a disproportionate share of the 2022–2023 theft spike and the subsequent 2024 decline as software patches and steering- wheel locks were distributed. Older vehicles without immobilizers remain significantly easier to steal. The 12-year ownership horizon used for normalization is an average; someone who keeps a car for 5 years faces roughly half the cumulative risk, while a 20-year owner faces roughly double. Behaviour matters too. NICB reports that roughly 100,000 US vehicles a year — about 11% of all 2021 thefts, and up more than 20% from 2019 — are stolen after the keys or fob were left inside, often during winter \"puffing\" (leaving a car running unattended to warm up). The unlocked-vehicle share is much higher than the keys-in share: the Houston Police Department's Auto Theft Division states that approximately 50% of vehicles stolen in Houston were left unlocked, and Lincoln (NE) Police Department data reported by Nebraska TV in July 2025 puts the unlocked share among stolen vehicles in Lincoln at 41% in 2024 and 63% in the first half of 2025; multi-year LPD figures range from 57% to 69% across 2020–2023. The true unlocked share is likely higher than reported because drivers under-disclose. These are share-of-incidents figures, not per-night rate ratios: converting them to a clean \"leave it unlocked overnight in the city\" multiplier would require an exposure denominator (what fraction of parked-vehicle nights are unlocked) that no published study reports, which is why this behavioural dimension sits in caveats rather than personal_factor_multipliers.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"An empty parking spot marked by faded paint lines on grey asphalt, flat vector illustration, muted tones."},"canonical_url":"https://likelier.app/vehicle-theft","api_url":"https://likelier.app/api/fears/vehicle-theft.json"},{"slug":"logging-occupational-death","question":"What are the odds of dying while working as a logger over a full career?","quick_answer":"About 1 in 34 over a 30-year logging career, at roughly 100 deaths per 100,000 workers a year — one of the highest fatal-injury rates of any US job. Most people sense logging is dangerous but underestimate the cumulative career risk.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Logging's reputation for danger is generally known among people who live in timber-producing regions, but the specific magnitude of the risk tends to be poorly calibrated in the general population. The occupation does not receive sustained national media attention the way commercial fishing does, and most people have little direct exposure to the work. No large-scale survey has isolated public perception of logging fatality odds; this entry uses editorial intuition. The BLS has consistently listed logging as either the highest or second-highest fatal work injury rate occupation in the US for more than two decades, a fact that is not widely known outside the industry.\n","rough_estimate":"most people likely guess logging is dangerous but underestimate the career-level cumulative risk","kind":"intuition"},"native":{"display":"100.7 deaths per 100,000 FTE workers per year (US logging workers, 2022 BLS CFOI via NIOSH)","numerator":54,"denominator":53625,"unit":"per worker per year","population":"US logging workers, BLS Census of Fatal Occupational Injuries 2022 (as reported by NIOSH)"},"normalized":{"lifetime_us_adult":0.0298,"display":"~1 in 34 over a 30-year logging career","log_value":-1.53,"assumptions":"Headline basis is the verbatim-confirmed NIOSH figure for BLS CFOI 2022: 54 logging worker fatalities at a rate of 100.7 per 100,000 FTE workers (implied denominator: 54/0.001007 ≈ 53,625 FTE workers). A career is modeled at 30 years, reflecting the longer working tenure typical of logging compared to commercial fishing; the median retirement age in the industry is commonly cited as the late 50s, with entry typically in the early-to-mid 20s. Compound probability over a 30-year career at 100.7 per 100,000 per year: 1 − (1 − 0.001007)^30 ≈ 0.0298. The scope is activity_specific_lifetime because this is per-career risk for a specific occupation, not a general US adult lifetime probability. The BLS CFOI rate has been broadly stable at 80–130 per 100,000 for logging workers over the past decade; the 2023 CFOI figure of roughly 98.9 per 100,000 is essentially identical and produces the same rounded 1-in-34 career probability, but the 2022 NIOSH-reported rate is used as the headline because it is the figure this entry's sources confirm verbatim (the aggregate Farming/fishing/forestry SOC group rate in the BLS 2023 news release was 24.4 per 100,000; the detailed logging-worker rate sits in CFOI's supplementary occupation data). The NIOSH 2024 blog post states the 2022 figure (100.7/100k) is more than 27 times higher than the all-occupation rate (3.7/100k).\n","uncertainty":{"low":0.02,"high":0.04},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/niosh/bulletin/2024/forest-operations.html","title":"Perspectives on Forest Operations Safety (NIOSH Science Blog)","publisher":"National Institute for Occupational Safety and Health (NIOSH), CDC","source_type":"govt_report","statistic":"54 fatalities to logging workers in 2022; rate 100.7 per 100,000 FTE, more than 27 times higher than all occupations (3.7/100k)","excerpt":"\"In 2022, there were 54 fatalities to logging workers. The work-related fatality rate for logging workers is 100.7 per 100,000 FTE, which is more than 27 times higher than the rate for all occupations at 3.7 per 100,000 FTE.\"\n","source_date":"2024-10-29","source_accessed":"2026-05-10","archive_url":"https://web.archive.org/web/20260525161926/https://www.cdc.gov/niosh/bulletin/2024/forest-operations.html","calculation_notes":"2022 native rate: 100.7 per 100,000 FTE. Annual probability: 0.001007. 30-year career compound: 1−(1−0.001007)^30 ≈ 0.0295. Used as the primary NIOSH confirmation of the BLS CFOI 2022 data. The 2023 BLS CFOI reports 52 deaths at 98.9/100k, slightly lower than 2022 but within the same range; the 2023 figure is used as the native headline.\n","independence_note":"NIOSH blog citing BLS CFOI 2022 data; provides independent narrative synthesis of the fatality data and confirms the 100.7/100k rate from a government occupational health perspective distinct from the primary BLS release.\n"},{"url":"https://www.bls.gov/news.release/cfoi.nr0.htm","title":"Census of Fatal Occupational Injuries Summary, 2023","publisher":"Bureau of Labor Statistics, US Department of Labor","source_type":"govt_report","statistic":"5,283 US fatal work injuries in 2023; all-worker rate 3.5 per 100,000 FTE (down from 3.7 in 2022); Farming, fishing, and forestry occupation group rate 24.4 per 100,000 FTE, the highest of any major occupation group","excerpt":"\"There were 5,283 fatal work injuries recorded in the United States in 2023... The fatal work injury rate was 3.5 fatalities per 100,000 full-time equivalent (FTE) workers, down from 3.7 in 2022.\"\n","source_date":"2024-12-19","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20250308125441/https://www.bls.gov/news.release/cfoi.nr0.htm","calculation_notes":"Primary BLS CFOI release establishing the national baseline the logging rate is measured against: all-worker rate 3.5 per 100,000 FTE (2023), and the Farming/fishing/forestry major occupation group at 24.4 per 100,000 FTE (Table 4), the highest of any major group. The detailed per-occupation \"logging workers\" rate (~98.9/100k in 2023, ~100.7/100k in 2022) appears in CFOI's supplementary occupation data, not in the news-release tables; this entry therefore takes its headline from the NIOSH source, which reports the 2022 figure (54 deaths, 100.7/100k) verbatim. Career math: 1−(1−0.001007)^30 ≈ 0.0298 ≈ 1 in 34.\n","independence_note":"Primary BLS CFOI release; methodologically independent of NIOSH CFID. BLS CFOI uses OSHA incident reports and death certificates; numerator and denominator both derive from BLS survey infrastructure.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10398574/","title":"Job Factors Associated with Occupational Injuries and Deaths in the United States Forestry Industry","publisher":"International Journal of Environmental Research and Public Health (MDPI)","source_type":"peer_reviewed","statistic":"Forestry workers fatality rate 92 per 100,000 FTE in 2014, approximately 28 times higher than all US industries (3.3/100k)","excerpt":"\"forestry workers faced a fatality rate of 92 per 100,000 FTE (full-time equivalents) in 2014, compared to 3.3 per 100,000 across all U.S. industries\"\n","source_date":"2023-07-26","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20250207142109/https://pmc.ncbi.nlm.nih.gov/articles/PMC10398574/","calculation_notes":"2014 rate of 92/100k used as historical cross-check. 1−(1−0.00092)^30 ≈ 2.7%, consistent with the 2022–2023 CFOI data. Confirms the rate has been persistently above 90/100k for at least a decade. Study also notes OSHA recordable non-fatal injury rate of 5.1 per 100 FTE for forestry, versus 3.2 for all industries, indicating the physical risk concentration is not limited to fatalities.\n"}],"comparison_anchors":[{"label":"All-worker US average fatal work injury (career, 40 yr)","lifetime_us_adult":0.0014},{"label":"Commercial fishing career death (20-year career)","lifetime_us_adult":0.0224},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Police officer line-of-duty death (career)","lifetime_us_adult":0.0018}],"personal_factor_multipliers":[{"factor":"Manual felling (chainsaw operator) vs mechanized harvesting","multiplier":2,"notes":"Struck-by incidents involving falling trees, logs, and limbs account for approximately 49.8% of all logging fatalities (BLS CFOI 2022 data per NIOSH blog). Manual chainsaw fallers working in standing timber face the highest struck-by exposure; mechanical harvester operators in enclosed cabs have lower struck-by risk but elevated machinery and rollover risk. NIOSH Forest Operations blog (2024) notes the transition toward mechanization as a priority safety intervention precisely because it reduces the struck-by hazard.\n"},{"factor":"Pacific Northwest (Oregon, Washington, California) old-growth or steep terrain","multiplier":1.5,"notes":"Logging on steep slopes (greater than 35%) in Pacific Northwest old-growth and second-growth forests involves larger-diameter trees and more complex terrain than flat southern pine operations. The Pacific Northwest Agricultural Safety and Health Center (PNASH) documents elevated risk in steep-slope cable logging compared to ground-based logging. Larger trees produce wider, less predictable fall zones and heavier limb debris.\n"},{"factor":"Fewer than 5 years in the occupation (new-entry worker)","multiplier":1.7,"notes":"Across hazardous manual occupations, new-entry workers face consistently elevated fatality rates relative to experienced workers. In logging, struck-by deaths from unexpected tree fall directions and kicked-back or barber-chaired stems are substantially more common among workers learning to read tree lean, root stability, and cutting geometry. NIOSH safety training programs for loggers specifically target inexperienced workers as the highest-risk subgroup.\n"},{"factor":"Contract or independent crew vs employer-crew with full OSHA 1910.266 compliance","multiplier":1.4,"notes":"OSHA's logging standard (29 CFR 1910.266) requires training, personal protective equipment, first-aid capability, and safe work procedures. Contract and independent logging crews have lower rates of verified compliance than larger employer-operated crews. The BLS CFOI data show disproportionate fatalities at smaller operations, consistent with OSHA compliance gaps driving additional risk beyond the industry-average rate.\n"}],"short_label":"Logging career death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The CFOI rate for logging workers (100.7/100k in 2022, ~98.9/100k in 2023) is based on a small absolute number of deaths (54 in 2022) in a small workforce (~53,600 FTE), so the rate is statistically volatile: a difference of 10 deaths in a single year shifts the rate by roughly 19 points per 100,000, which would move the 30-year career probability by approximately 0.5 percentage points. Year-to-year rates for logging have ranged from below 80 to above 130 per 100,000 over the past decade, all consistently ranking logging as one of the top two or three most dangerous US occupations. The 30-year career assumption may overstate career length for workers in physically intensive manual felling roles, who often transition to mechanized equipment or leave the industry in their 40s; longer actual careers produce higher cumulative risk, shorter careers lower. The rate does not distinguish between fatalities on public and private timberlands, or between federal contract crews and state and private operations, which have different regulatory oversight. Non-fatal serious injuries (chainsaw lacerations, musculoskeletal injuries, traumatic brain injury from limb strikes) are substantially undercounted due to self-employment and small-crew reporting gaps in OSHA injury surveillance.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A single hard hat resting against a sawn log cross-section on a pale neutral surface, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/logging-occupational-death","api_url":"https://likelier.app/api/fears/logging-occupational-death.json"},{"slug":"child-play-arm-swing-injury","question":"What are the odds of a child being injured by common play activities like arm-swinging or going down a slide together?","quick_answer":"Roughly 3 in 100 children get an injury like nursemaid's elbow, a pulled elbow from arm-swinging, before age 6 — more common than the negligible harm parents tend to assume, though the injuries are usually minor and easily treated.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Swinging children by the arms, riding down a slide together, and shoulder carries feel playful and safe — injury seems like an unlikely outcome of normal bonding play. The activities are common enough that most parents have done all of them multiple times without incident, which reinforces the intuition that harm is essentially hypothetical. The specific mechanisms by which each activity can cause injury are not widely known.\n","kind":"intuition"},"native":{"display":"~6 per 10,000 children under 6 per year reach an emergency department with nursemaid's elbow (radial head subluxation)","numerator":6,"denominator":10000,"unit":"per child per year (under age 6, ED-presenting cases)","population":"US children aged 0–5","exposures_per_year":1},"normalized":{"lifetime_us_adult":0.03,"display":"roughly 3 in 100 children before age 6 (cumulative across peak risk years)","log_value":-1.52,"assumptions":"The NEISS-based ED rate of 6 per 10,000 children ≤6/year (Rangan et al. 2021) is a lower bound — it counts only ED-treated cases. Community-level estimates from clinical literature (Aberdeen, 1990, cited widely) place the true event rate at approximately 1% per year (10 per 1,000) for children in the peak-risk window. Peak risk is age 1–4 (approximately 3 years of meaningful exposure); annular ligament strengthens markedly after age 5. Central estimate: 1% × 3 years = 3%, with lower bound using NEISS-only rate (0.6/1,000 × 3 = 0.18%, rounded to 0.012 for conservative single-tail lower) and upper bound including mild underdiagnosis (2%/year × 3 = 6%). Scope: subgroup_lifetime (early childhood, not adult lifetime).\n","uncertainty":{"low":0.012,"high":0.06},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/34363491/","title":"Trends and epidemiology of radial head subluxation in the United States from 2004 to 2018","publisher":"European Journal of Orthopaedic Surgery & Traumatology","source_type":"peer_reviewed","statistic":"6.03 per 10,000 children ≤6 years per year (ED-treated); ~253,578 total cases 2004–2018","excerpt":"\"The overall annual rate of RHS per 10,000 children ≤ 6 years was 6.03 (95% CI = 4.85–7.58). The rate significantly increased from 5.18 (95% CI 2.96–7.39) in 2004 to 7.69 (95% CI = 4.63–10.75) in 2018. An estimated total 253,578 children 6 years or younger were treated for RHS.\"\n","source_date":"2021-08-01","source_accessed":"2026-05-02","archive_url":"https://web.archive.org/web/20260503080009/https://pubmed.ncbi.nlm.nih.gov/34363491/","calculation_notes":"NEISS national sample, 2004–2018. The rate of 6.03 per 10,000/year is used as the native numerator (6) per 10,000 denominator. This is an ED-treated rate only; community-level incidence (including primary care and self-resolved reductions) is higher. Pulling mechanisms accounted for 36.2% of cases, swinging and similar arm-traction mechanisms for the remainder alongside falls. The native figure intentionally uses the conservative ED rate; the normalized central estimate uses the higher 1%/year community-level figure (Aberdeen-origin, cited in clinical literature) over 3 peak-risk years.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29637487/","title":"Playground slide-related injuries in preschool children: increased risk of lower extremity injuries when riding on laps","publisher":"Injury Epidemiology (Springer)","source_type":"peer_reviewed","statistic":"OR 49.5 (95% CI 31.7–77.4) for lower leg/ankle fracture when child rides on adult's lap vs. riding solo","excerpt":"\"When a young child is going down a slide on the lap of another person, their foot may catch on the slide's surfaces including the inner side or bottom of the slide. The lower leg can then twist and be pulled backward as both proceed down the slide. Children identified as riding on another person's lap had 43 times higher odds (OR 43.0, 95% CI 32.0–58.0) of lower extremity injury versus other body parts, and 49.5 times higher odds (95% CI 31.7–77.4) of lower leg/ankle fracture compared with other fracture locations.\"\n","source_date":"2018-04-09","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260505051428/https://pubmed.ncbi.nlm.nih.gov/29637487/","calculation_notes":"NEISS-based study of 12,686 playground slide injuries in children ≤5 years (2002–2015), estimating >350,000 total US slide injuries in the period. Lap-riding was documented in 644 cases (5% of total), with infants most affected (34% of under-1s, 15% of 1-year-olds). The OR of 49.5 for lower leg/ankle fracture is used as the effect-size anchor for the lap-slide paragraph in prose. Not used in the native→normalized arithmetic (which uses nursemaid's elbow as the primary axis) but establishes the lap-slide fracture risk as the secondary quantified finding.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19700991/","title":"Tibia fractures in children sustained on a playground slide","publisher":"Gaffney JT — Journal of Pediatric Orthopedics, 2009","source_type":"peer_reviewed","statistic":"8 of 58 tibia fractures (13.8%) reviewed over an 11-month period were sustained on a playground slide; all 8 occurred while the child rode down on an adult's lap, none while sliding independently","excerpt":"\"During the period of study, 58 fractures of the tibia were found. Eight (13.8%) of the tibia fractures were sustained while playing on a playground slide... All tibia fractures associated with playing on a slide were sustained while going down the slide on the lap of an adult. None of the 8 children studied went down the slide alone.\"\n","source_date":"2009-09-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20230921123851/https://pubmed.ncbi.nlm.nih.gov/19700991/","calculation_notes":"Single-institution case series (Gaffney, 2009) reviewing all tibia fractures presenting over an 11-month period; 8 of 58 (13.8%) were slide-related, and all 8 occurred while the child rode down the slide on an adult's lap rather than independently. Used in body prose as a second, independent confirmation of the lap-riding mechanism established in the larger Jennissen et al. (2018) NEISS-based study; not used in the native/normalized arithmetic, which is anchored on the nursemaid's-elbow NEISS rate.\n","independence_note":"Single-institution case-series chart review, methodologically independent of the national NEISS-based surveillance study (Jennissen et al. 2018) and of the StatPearls nursemaid's-elbow reference.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK430777/","title":"Nursemaid Elbow — StatPearls","publisher":"NCBI Bookshelf / StatPearls","source_type":"reputable_reference","statistic":"Represents >20% of upper extremity injuries in children; recurrence rate ~20%; resolves with closed reduction in seconds","excerpt":"\"Radial head subluxation (RHS) is common in children 1 to 4 years of age and represents more than 20% of upper extremity injuries in children. The injury occurs when a child is swung around by the arms, or lifted by one arm. Recurrence rate is approximately 20%. Treatment involves closed reduction, and this can be performed in a few seconds without sedation. It occurs less commonly in children older than the age of 5 because the annular ligament strengthens with age.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260504190711/https://www.ncbi.nlm.nih.gov/books/NBK430777/","calculation_notes":"Used to establish: (1) swinging by arms as a named mechanism of nursemaid's elbow, (2) recurrence rate of ~20% quoted in the personal_factor_multipliers section, (3) that the condition accounts for >20% of upper extremity injuries in children, (4) age boundary (resolves after 5) underlying the 3-year peak window used in normalized assumptions. Not used in the primary arithmetic; supports narrative claims.\n"}],"comparison_anchors":[{"label":"Child requiring stitches from any injury (lifetime)","lifetime_us_adult":0.25},{"label":"Child sustaining any arm fracture (lifetime under 18)","lifetime_us_adult":0.05}],"personal_factor_multipliers":[{"factor":"Age 1–3 (peak nursemaid's elbow window)","multiplier":3,"notes":"Annular ligament is immature; radial head subluxation can occur with minimal traction. Risk falls sharply after age 5."},{"factor":"Repeated arm-swinging play","multiplier":2,"notes":"Recurrence rate after first episode is approximately 20%; the annular ligament remains vulnerable until it strengthens."},{"factor":"Prior nursemaid's elbow episode","multiplier":4,"notes":"After the first subluxation, recurrence risk is approximately 20–26% per Macias et al. (2000, Pediatric Emergency Care) and StatPearls (NCBI, 2024). The annular ligament remains lax at the site of previous injury, raising the per-episode risk roughly 4× above the baseline population rate until the ligament fully matures (typically after age 5).\n"},{"factor":"Uninstructed caregiver (unaware of traction risk)","multiplier":2,"notes":"Caregivers who are unaware that pulling or swinging by the wrist/hand is the primary mechanism are more likely to repeat the activity after a first injury. AAP Pediatric Orthopedics education guidelines note that caregiver counseling on avoiding forearm traction significantly reduces recurrence in clinical follow-up; uninstructed caregivers face higher repeat-injury rates.\n"}],"short_label":"Play swing & slide injury","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"Nursemaid's elbow is usually treated with a simple manual reduction manoeuvre that takes seconds and requires no sedation, casting, or anaesthesia; most children resume normal use of the arm within minutes of reduction. The condition does not involve a true fracture. Lap-slide fractures are less common in absolute terms but more clinically consequential — the tibial fracture typically requires casting for 3–6 weeks. These activities are not inherently dangerous on any given occasion: the risk per episode is low, and most children who are swung by the arms or taken down a slide on a parent's lap will never be injured. The primary risk factor for nursemaid's elbow is age under 3, when the radial head is not yet fully developed. Shoulder-carry and piggyback play carry fall risk if the adult loses balance, but no specific population-based injury rate is well established for these activities.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"Two large adult hands gently gripping a small child's wrist, the child's feet just off the ground, with a playground slide visible softly in the background"},"canonical_url":"https://likelier.app/child-play-arm-swing-injury","api_url":"https://likelier.app/api/fears/child-play-arm-swing-injury.json"},{"slug":"traumatic-brain-injury-serious","question":"What are the odds of a serious traumatic brain injury in a lifetime?","quick_answer":"A hospitalization-level traumatic brain injury carries about 1 in 33 lifetime odds for a US adult, with the fatal subset near 1 in 125. Most people picture TBI as a sports or military problem and underweight their own exposure, which actually concentrates sharply in old age.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Public framing of traumatic brain injury in the US is dominated by two images: the young athlete concussed on a football field, and the young soldier concussed by an IED. Both are real, both are important, and neither is where the bulk of serious TBI in the US actually happens. No major fear survey isolates \"getting a serious TBI\" as its own item — the closest proxies are fear of concussion in youth sports (high among parents of athletes) and fear of dementia (high among older adults, and often conflated with post-TBI cognitive decline in the popular press). Non-athlete, non-military adults under 65 generally treat serious TBI as a freak event that happens to someone else. That intuition is roughly right for the current year of their life and badly wrong for their remaining lifetime, which is the standard shape of a risk that concentrates sharply in old age.\n","rough_estimate":"Most adults treat serious TBI as a sports-or-military problem and underweight their own lifetime exposure","kind":"intuition"},"native":{"display":"~85 TBI-related hospitalizations per 100,000 per year (US, 2013)","numerator":85,"denominator":100000,"unit":"per year","population":"US residents, all ages, TBI-related hospitalizations (Taylor et al. MMWR 2017)"},"normalized":{"lifetime_us_adult":0.03,"display":"1 in ~33 lifetime hospitalization-level TBI (US adult)","log_value":-1.523,"assumptions":"Anchored on Taylor, Bell, Breiding & Xu (MMWR Surveillance Summaries 2017), which reports approximately 2.8 million TBI-related ED visits, hospitalizations, and deaths (TBI-EDHDs) in the US in 2013, of which ~2.5 million were ED visits, ~282,000 were hospitalizations, and ~56,000 were deaths. Against a 2013 US population of ~316 million, that is ~89 hospitalizations per 100,000 per year and ~18 deaths per 100,000 per year. The CDC's more recent (2020-2021) TBI data page reports ~214,110 TBI-related hospitalizations in 2020 and 69,473 TBI-related deaths in 2021 — a similar order of magnitude against a ~332 million US population (~64/100k hospitalizations, ~21/100k deaths), with some of the drop in hospitalizations attributable to pandemic-era reporting and coding changes.\nFor a US adult over a 59-year remaining lifetime, flat-compounding the ~85/100k annual hospitalization rate would give 1 − (1 − 0.00085)^59 ≈ 0.049 (1 in 20), but that rate is a cross-sectional age-adjusted average that bakes in the very high pediatric and 75+ tails. Removing the pediatric years (the compounding starts at adulthood) and down-weighting the middle-age decades, the realistic US-adult lifetime probability of a hospitalization-level TBI lands near 0.03 (1 in ~33). The fatal subset — compounding ~20/100k/year over 59 years — lands near 0.008 (1 in ~125), and the \"persistent disability from TBI\" subset, using the ~5.3 million Americans living with TBI-related disability (CDC surveillance program legacy estimate) against the ~260 million US adult population, sits near 0.01 (1 in 100) as a cross-sectional prevalence that roughly approximates lifetime incidence.\n","uncertainty":{"low":0.02,"high":0.05},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/66/ss/ss6609a1.htm","title":"Traumatic Brain Injury-Related Emergency Department Visits, Hospitalizations, and Deaths — United States, 2007 and 2013","publisher":"CDC Morbidity and Mortality Weekly Report, Surveillance Summaries (Taylor, Bell, Breiding & Xu 2017)","source_type":"govt_report","statistic":"2013: approximately 2.8 million TBI-EDHDs in the US, comprising ~2.5 million ED visits, ~282,000 hospitalizations, and ~56,000 deaths. Falls were the leading mechanism at 413.2 per 100,000 (age-adjusted). Rate for age 75+ was 2,232.2 per 100,000.","excerpt":"\"In 2013, a total of approximately 2.8 million TBI-EDHDs occurred in the United States. This consisted of approximately 2.5 million TBI-related ED visits, approximately 282,000 TBI-related hospitalizations, and approximately 56,000 TBI-related deaths.\"\n","source_date":"2017-03-17","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260318233501/https://www.cdc.gov/mmwr/volumes/66/ss/ss6609a1.htm","calculation_notes":"Taylor et al. is the canonical US TBI surveillance paper and is the direct source of the native numerator (~282,000 US TBI hospitalizations/year ≈ 89/100,000/year in 2013) that we use for normalization. The same paper is also the source of the age-stratified rates used in the personal_factor_multipliers section: TBI rates at age 75+ are roughly 20x the rate among middle-aged adults, and falls are the leading mechanism for those older-adult hospitalizations. The shift noted in the paper — that intentional self-harm (primarily firearm suicide) surpassed motor vehicle crashes as the leading cause of TBI-related death between 2007 and 2013 — also feeds the fatal-TBI subset estimate and explains why the mortality number has risen while the hospitalization rate has flattened.\n","independence_note":"Upstream of most US TBI surveillance numbers. The CDC TBI data page we cite as the second authoritative source draws from the same National Vital Statistics System (for deaths) and state hospital discharge databases (for hospitalizations) as Taylor et al., so treat the two as one authoritative pipeline with two different vintages rather than as independent estimates.\n"},{"url":"https://www.cdc.gov/traumatic-brain-injury/data-research/index.html","title":"TBI Data","publisher":"CDC National Center for Injury Prevention and Control","source_type":"govt_report","statistic":"~214,110 TBI-related hospitalizations in 2020; 69,473 TBI-related deaths in 2021. Age-adjusted hospitalization rate: 79.9 per 100,000 (males), 43.7 per 100,000 (females). Age-adjusted death rate: 28.3 per 100,000 (males), 8.4 per 100,000 (females).","excerpt":"\"There were over 69,000 TBI-related deaths in the United States in 2021. That's about 190 TBI-related deaths every day.\" ... \"Older adults are more likely to be hospitalized and die from a TBI compared to all other age groups.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421200145/https://www.cdc.gov/traumatic-brain-injury/data-research/index.html","calculation_notes":"This is the CDC's current TBI data landing page, providing the most recent headline figures: ~214,110 TBI-related hospitalizations in 2020 and 69,473 TBI-related deaths in 2021. These figures are slightly lower than the 2013 Taylor et al. headline (282,000 hospitalizations, 56,000 deaths) for hospitalizations and modestly higher for deaths, reflecting the ongoing increase in firearm suicide TBI and some decline in hospitalization coding during the pandemic. Combined with the ~332 million US population in 2021, the current rates are approximately 64/100,000/year (hospitalizations) and 21/100,000/year (deaths), which is the check figure against the 2013 numerator used in the normalized estimate. The roughly 3:1 male:female ratio on both hospitalizations and deaths is documented directly on this page.\n","independence_note":"Shares upstream datasets (NVSS, HCUP, NEISS-AIP) with Taylor et al. Used here for currency rather than as an independent measurement.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33947273/","title":"Epidemiology of Chronic Effects of Traumatic Brain Injury","publisher":"Journal of Neurotrauma (Haarbauer-Krupa et al. 2021)","source_type":"peer_reviewed","statistic":"Review of long-term outcomes following TBI; most patients recover within weeks but a meaningful subset develop persistent symptoms lasting months to years that significantly affect quality of life. Risk factors for persistent disability include age, pre-injury status, comorbid conditions, and injury mechanism.","excerpt":"\"Although many patients diagnosed with traumatic brain injury (TBI), particularly mild TBI, recover from their symptoms within a few weeks, a small but meaningful subset experience symptoms that persist for months or years after injury and significantly impact quality of life for the person and their family.\"\n","source_date":"2021-05-05","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20251207005841/https://pubmed.ncbi.nlm.nih.gov/33947273/","calculation_notes":"Haarbauer-Krupa et al. is the modern peer-reviewed review of chronic TBI effects and is the source for the \"persistent disability\" framing of the headline number. It establishes that the persistent-symptoms subset is a minority of TBI cases overall but concentrates disproportionately in moderate-to-severe TBI, which is approximately the hospitalization-level subset we use as the main numerator. The review does not produce its own prevalence figure; for that we lean on the older CDC TBI Surveillance Program legacy estimate of ~5.3 million Americans living with a long-term TBI-related disability, which against a ~260 million US adult population gives the ~1 in 100 persistent-disability lifetime anchor.\n","independence_note":"Independent of the CDC surveillance pipeline: Haarbauer-Krupa et al. is a peer-reviewed synthesis that draws on the same underlying studies but applies independent editorial framing and risk-factor analysis.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK580076/","title":"The Scope and Burden of Traumatic Brain Injury (in: Traumatic Brain Injury: A Roadmap for Accelerating Progress)","publisher":"National Academies of Sciences, Engineering, and Medicine (NASEM 2022)","source_type":"reputable_reference","statistic":"TBI is the leading cause of death and disability in young adults in high-income countries; falls are the leading mechanism of TBI in the US overall, and motor vehicle crashes remain a major cause of severe TBI in younger adults. An estimated 5.3 million Americans live with a long-term TBI-related disability per the CDC TBI Surveillance Program.","excerpt":"\"Traumatic brain injury (TBI) is a serious public health problem in the United States. In 2014, approximately 2.87 million TBI-related emergency department visits, hospitalizations, and deaths occurred in the United States, including over 837,000 of these health events among children.\"\n","source_date":"2022-02-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250617034955/https://www.ncbi.nlm.nih.gov/books/NBK580076/","calculation_notes":"The NASEM Roadmap is the most recent authoritative synthesis of the US TBI burden and is used here as a secondary independent cross-check on the 2.8 million total TBI-EDHD figure. It also consolidates the long-term disability framing and the age-distribution caveats that are load-bearing for this page. Not used as the primary numerator because it draws on the same Taylor et al. surveillance vintage.\n","independence_note":"Synthesizes Taylor et al. and CDC TBI surveillance data, so it is not an independent measurement. Included for editorial framing and because it is the most recent authoritative expert review of the field.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death from accidental fall (lifetime, US adult)","lifetime_us_adult":0.0074},{"label":"Serious head injury, frequent unhelmeted urban cyclist (30-yr career)","lifetime_us_adult":0.125},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Any TBI ED visit (mild + severe), lifetime","probability":0.4,"notes":"Concussion-level events are common. The ~2.5 million annual TBI ED visits compounded across an adult lifetime produce a lifetime any-TBI probability approaching one in two, though most of those are mild TBI (concussion) that resolve within weeks."},{"region":"Hospitalization-level TBI, lifetime","probability":0.03,"notes":"Headline number on this page. Roughly 1 in 33 US adults will be hospitalized for a TBI at some point in their remaining adult lifetime — derived from the ~282,000 annual TBI hospitalizations in the Taylor et al. 2013 surveillance and the age distribution of those hospitalizations."},{"region":"Persistent disability from TBI, lifetime","probability":0.01,"notes":"Roughly 1 in 100. Approximated from the CDC TBI Surveillance Program legacy estimate of ~5.3 million Americans living with long-term TBI-related disability, against the ~260 million US adult population. Concentrates in moderate-to-severe TBI rather than concussion."},{"region":"Fatal TBI, lifetime","probability":0.008,"notes":"Roughly 1 in 125. Compounds the ~20/100,000/year age-adjusted TBI death rate across an adult lifetime. Firearm suicide is now the single largest contributor to this number, followed by elderly falls and motor vehicle crashes."}],"personal_factor_multipliers":[{"factor":"age 75+","multiplier":3,"notes":"Fall-related TBI dominates elderly incidence. Taylor et al. 2017 reports the 75+ TBI rate at 2,232.2 per 100,000 per year — roughly 20x the middle-aged rate — but not all of that elevation carries forward into a straight lifetime multiplier, because much of the baseline lifetime risk is already allocated to the older-adult years. A 3x multiplier is the practical effect for a reader currently 75+ who is re-reading this page as a cohort-specific forecast."},{"factor":"contact sport history","multiplier":2,"notes":"Order-of-magnitude estimate. Football, boxing, MMA, rugby, ice hockey and soccer heading all elevate per-hour TBI risk meaningfully, but the single-impact hospitalization-level TBI is the outcome in scope here — not chronic traumatic encephalopathy, which is a distinct pathology from repeated subconcussive impacts. The CTE multiplier would be much higher but belongs on a separate page."},{"factor":"motorcycle rider","multiplier":4,"notes":"Motor vehicle crashes remain a major TBI cause, and motorcyclists are overrepresented among severe TBI hospitalizations relative to their share of vehicle miles traveled by a factor of roughly 30x. Helmet use reduces but does not eliminate the elevated baseline."},{"factor":"military combat deployment","multiplier":5,"notes":"Post-9/11 veterans have elevated lifetime TBI prevalence driven by blast exposure, falls during training and combat, and vehicle crashes. The DoD TBI Center of Excellence documents several hundred thousand service-member TBIs across the post-2000 deployment cohort; the multiplier here is the rough ratio of deployed-cohort lifetime serious TBI to the general US adult baseline."}],"short_label":"Traumatic brain injury","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The single biggest caveat on this page is that \"serious TBI\" covers an enormous severity range, from a concussion admitted overnight for observation to a severe focal injury requiring craniotomy and leaving permanent deficits. The regional_breakdown rows pull those layers apart — roughly 1 in 2 adults will see a mild TBI in their lifetime, 1 in 33 will be hospitalized, 1 in 100 will live with persistent disability, and 1 in 125 will die of TBI — and the right number depends on which outcome you are actually asking about. The headline 1-in-33 is the hospitalization-level figure because that is both the best-measured stratum in the CDC surveillance pipeline and the outcome most readers mean when they say \"serious TBI.\"\nThe second caveat is the demographic mismatch between the cultural framing of TBI and the statistics. The US public conversation centers on young athletes (concussion in football, soccer heading, MMA, high school sports) and young service members (blast TBI from the post-9/11 deployments). Both are real, both produce the multipliers in the personal_factor_multipliers section, and neither describes where the majority of severe-TBI hospitalizations actually originate: the CDC and Taylor et al. data are unambiguous that falls in adults 75 and older are the largest single mechanism of severe TBI in the US, by a wide margin, and the rate is rising with population aging. The median severe-TBI patient in the US is not a 19-year-old linebacker but an 82- year-old who fell in a bathroom.\nThird: chronic traumatic encephalopathy (CTE) from repeated subconcussive impacts in contact sports and combat is a real and distinct entity from the single-impact TBI framework used on this page. CTE risk depends on cumulative subconcussive exposure, not on whether the athlete ever had a hospitalization-level concussion, and it deserves its own entry rather than being folded into this one. The contact-sport multiplier above applies to single-impact severe TBI only.\nFinally, the numerator on this page excludes TBIs that never reach an emergency department or hospital — solo falls that the patient \"walked off\", ring injuries that never got checked, military field-treated concussions that were not entered into a medical record. The true all-severity lifetime incidence is higher than what the Taylor et al. surveillance captures, by an unknown but non-trivial amount.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single plain medical-style paper gown or hospital wristband resting on a pale surface, flat vector illustration in muted colors."},"canonical_url":"https://likelier.app/traumatic-brain-injury-serious","api_url":"https://likelier.app/api/fears/traumatic-brain-injury-serious.json"},{"slug":"watching-video-while-driving","question":"What are the odds of a crash from watching video on a phone while driving?","quick_answer":"For a driver who regularly watches video at the wheel, the crash risk is about 1 in 1,850 a year, compounding to roughly 1 in 32 over a lifetime. Most people rank it below texting on the danger scale, but video demands sustained gaze and is at least as risky.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Drivers who watch video at the wheel tend to classify the behavior as less dangerous than texting because they are not \"doing\" anything — they are just watching. This framing is backwards. Texting involves short, repeated glances; watching video requires sustained gaze, often ten seconds or more at a stretch. Most people cannot give a numerical probability for a crash from the habit, and the subset who have considered it at all typically place video-watching below texting on the danger scale.\n","rough_estimate":"most people think it's risky but less so than texting — this is wrong","kind":"intuition"},"native":{"display":"~1 in 1,850 per year (regular video-watcher US adult driver)","numerator":1,"denominator":1850,"unit":"per year","population":"US adult drivers who regularly watch video on a handheld phone while driving (exposure-weighted from Dingus 2016 OR estimates and NHTSA baseline)"},"normalized":{"lifetime_us_adult":0.031,"display":"1 in ~32 lifetime (regular video-watching US adult driver)","log_value":-1.508,"assumptions":"Starts from the US population-average car-crash lifetime hazard of ~1 in 105 (annual p ≈ 1.22e-4, from IIHS 2023). Dingus et al. 2016 (PNAS) reports an odds ratio of 9.9 for reading/writing on a handheld phone and 12.2 for handheld cell dialing — both high-visual-demand tasks with sustained eyes-off-road windows. Watching video is at minimum a reading-class task (sustained gaze, no manual input) and more plausibly sits at or above that range because video viewing is designed to hold attention for seconds at a time rather than the brief glances texting requires (OR 6.1 in Dingus 2016). NHTSA and IIHS research on glance duration confirms crash risk rises steeply beyond 2 seconds of eyes-off-road; Simons-Morton et al. 2014 found OR 6.0 for glances exceeding 3 seconds. Using a conservative per-epoch OR of ~10 for video-watching episodes and an exposure-weighted multiplier of ~4x for a regular video-watcher (higher than the 2.5x for texting because each episode is longer), the annual hazard becomes ~4.88e-4. Over 59 remaining adult years: 1 − (1 − 4.88e-4)^59 ≈ 0.028, approximately 1 in 36. Rounding to a central estimate of 0.031 (1 in 32) reflects the plausible range of per-epoch ORs from 8–12. Uncertainty band reflects the 3x–6x plausible range for exposure-weighted multipliers.\n","uncertainty":{"low":0.016,"high":0.052},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","title":"Driver crash risk factors and prevalence evaluation using naturalistic driving data","publisher":"Dingus et al., Proceedings of the National Academy of Sciences (PNAS)","source_type":"peer_reviewed","statistic":"Reading/writing on handheld cell phone: OR 9.9; handheld cell dialing: OR 12.2; texting: OR 6.1; browsing: OR 2.7; overall handheld cell interaction: OR 3.6. All relative to model driving in SHRP 2 passenger-car naturalistic sample.","excerpt":"\"Reading or writing on a handheld cell phone (e.g., e-mail, text, browsing) was 9.9 times more likely to result in a crash than model driving.\"\n","source_date":"2016-03-08","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250707185013/https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","calculation_notes":"Dingus 2016 does not report a specific \"watching video\" category as a discrete secondary task. The 9.9 OR for reading/writing is the closest analogue: both tasks demand sustained eyes-off-road gaze, typically 5–15 seconds per episode. Video-watching differs from reading/writing in that it holds attention longer and is less likely to be interrupted voluntarily, suggesting a per-epoch OR at or above 9.9. This entry uses ~10 as the working per-epoch estimate, consistent with the reading/writing figure. To convert to a lifetime probability, the US per-capita annual car-crash hazard (12.2/100,000, IIHS 2023) is multiplied by an exposure-weighted factor of ~4x for a regular video-watcher, then compounded over 59 adult years.\n","independence_note":"Dingus 2016 draws from the SHRP 2 Naturalistic Driving Study. The Simons-Morton 2014 source below also uses SHRP 2 data, so treat both as drawing from a shared upstream dataset; they are methodologically distinct but not independent samples.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3999409/","title":"Keep Your Eyes on the Road: Young Driver Crash Risk Increases According to Duration of Distraction","publisher":"Simons-Morton BG, Guo F, et al., Journal of Adolescent Health / PMC","source_type":"peer_reviewed","statistic":"Crash risk increased with eye-glance duration during secondary tasks: OR 3.8 for glances >2 s; OR 6.0 for glances >3 s. Crash risk during wireless secondary tasks: OR 5.5 for engagement >2 s.","excerpt":"\"Crash risk increased with the duration of single longest glance during all secondary tasks (odds ratio=3.8 for >2 s) and wireless secondary task engagement (odds ratio=5.5 for >2 s).\"\n","source_date":"2014-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260118054055/https://pmc.ncbi.nlm.nih.gov/articles/PMC3999409/","calculation_notes":"This study directly measures the effect of glance duration on crash risk, independent of task type. Video watching predictably produces glances well above the 2-second threshold (often 5–15 s per episode), placing it firmly in the highest-risk category identified. The OR of 6.0 for glances >3 s serves as a conservative lower bound for the per-epoch risk of video viewing. The study instrumented 42 newly licensed teen drivers; crash patterns for sustained-glance tasks are consistent with adult naturalistic studies.\n","independence_note":"This study uses a subset of SHRP 2 naturalistic driving data (teen driver cohort). Methodologically distinct from Dingus 2016 (case-crossover design vs. case-control) but draws from the same upstream database.\n"},{"url":"https://www.iihs.org/topics/bibliography/ref/2264","title":"Prevalence of distracted driving by driver characteristics in the United States","publisher":"Cox AE, Cicchino JB — Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"About 15% of surveyed US drivers reported watching movies, TV shows, or video clips regularly while driving (about 25% at some point in the past 30 days); over 21% engaged in at least one modern smartphone-based distraction (video, social media, video calls) on most or all trips.","excerpt":"\"Males, parents of children ages 18 and younger, and participants who drive in the gig economy had higher adjusted odds of engaging in 'modern' device-based distractions enabled by smartphones (e.g., making video calls, watching videos, using social media) than other drivers.\"\n","source_date":"2022-11-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250520232309/https://www.iihs.org/topics/bibliography/ref/2264","calculation_notes":"The Cox & Cicchino survey establishes that video watching while driving is not rare: about 15% of US adult drivers do it regularly (about 25% at some point in the past 30 days). This prevalence figure is used here only to validate that the habit is common enough to model as \"regular\" exposure rather than a fringe behavior. The survey does not directly measure crash risk; crash-risk estimates come from Dingus 2016 and Simons-Morton 2014.\n","independence_note":"Survey-based prevalence data; fully independent of both naturalistic driving datasets (Dingus 2016, Simons-Morton 2014). Primary source for the ~15% regular-video-watcher prevalence figure.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Crash lifetime risk — regular texter while driving","lifetime_us_adult":0.018},{"label":"Death on a motorcycle (lifetime, US adult, population average)","lifetime_us_adult":0.00144}],"personal_factor_multipliers":[{"factor":"never watches video while driving","multiplier":1,"notes":"Baseline US driver car-crash risk with no video-viewing exposure."},{"factor":"glances briefly at a paused or thumbnail screen occasionally","multiplier":1.5,"notes":"Short glance at a static image; closer to texting-class distraction."},{"factor":"watches video clips regularly at stoplights, resuming after light changes","multiplier":2.5,"notes":"Video content is engaging and continuation past the light is common."},{"factor":"watches video continuously on the highway or during active traffic","multiplier":6,"notes":"Sustained gaze away from the road at speed; near the upper bound of modeled risk."}],"short_label":"Video watching + driving","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"No large-scale naturalistic study has yet isolated \"watching video on a phone\" as a discrete coded secondary task with its own reported odds ratio. The per-epoch OR of ~10 used here is inferred from the reading/writing category in Dingus 2016 (OR 9.9) and the glance-duration findings in Simons-Morton 2014 (OR 6.0 for glances >3 s). Video content holds attention longer than reading a text message, which pushes the estimate upward relative to texting; but the rarity of dedicated video-watching data means the uncertainty band is wider than for texting. The exposure-weighted multiplier (4x) is also a judgment call: a driver who watches one 15-second clip per hour spends far more seconds with eyes off-road than a driver who sends two texts per hour, making 4x a plausible midpoint, not a measured value. The ~15% regular-watcher prevalence (Cox & Cicchino 2022) suggests that video watching while driving is common enough to take seriously, but rarer than texting, so the population-average contribution to US crash statistics is smaller even though the per-exposure risk is higher.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":3,"d5":5,"d6":4,"d7":3,"d8":4,"avg":3.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A single muted smartphone screen showing a small play-button icon, resting flat on a pale surface with a faint lane-marker stripe beside it, flat vector illustration."},"canonical_url":"https://likelier.app/watching-video-while-driving","api_url":"https://likelier.app/api/fears/watching-video-while-driving.json"},{"slug":"autism-diagnosis-child","question":"What are the odds of your child being diagnosed with autism spectrum disorder?","quick_answer":"About 1 in 31 US children are identified with autism spectrum disorder by age 8. Most parents are roughly right on the prevalence but overestimate severity — the diagnosis spans a wide spectrum, and lifelong dependency applies to only a fraction of it.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Autism sits near the top of parental anxieties in the United States. The steady upward march of CDC prevalence numbers — from 1 in 150 in 2000 to 1 in 31 in 2022 — produces a feeling that something unprecedented is happening to children. Media coverage tends toward the word \"epidemic,\" which implies a cause that must be found and stopped. Most parents can cite a rough prevalence figure (typically \"about 1 in 36\" or \"1 in 30-something\"), but the fear usually outweighs the number: the diagnosis carries connotations of lifelong dependency that apply to only a fraction of the spectrum.\n","rough_estimate":"Most parents are roughly correct on prevalence but overestimate severity","kind":"intuition"},"native":{"display":"1 in 31 (3.2%) among 8-year-olds (US, 2022 surveillance year)","numerator":32,"denominator":1000,"unit":"per child (cumulative identification by age 8)","population":"US children aged 8, CDC ADDM Network 2022"},"normalized":{"lifetime_us_adult":0.032,"display":"1 in ~31 per child (US)","log_value":-1.49,"assumptions":"The CDC ADDM Network 2025 community report (2022 surveillance year) identifies 3.2% of 8-year-olds as having ASD. Because ASD is a developmental condition diagnosed in childhood, the 8-year-old prevalence is the standard proxy for per-child lifetime diagnosis probability. Some diagnoses occur after age 8 (particularly in girls and in children without intellectual disability), so 3.2% is likely a slight undercount of eventual lifetime identification. The figure is used directly as lifetime_us_adult for schema compatibility; the scope field clarifies this is a subgroup (per-child) lifetime figure.\n","uncertainty":{"low":0.025,"high":0.042},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/74/ss/ss7402a1.htm","title":"Prevalence and Early Identification of Autism Spectrum Disorder Among Children Aged 4 and 8 Years — ADDM Network, 16 Sites, United States, 2022","publisher":"CDC MMWR Surveillance Summaries","source_type":"govt_report","statistic":"3.2% (1 in 31) of 8-year-olds identified with ASD across 16 ADDM sites in 2022; boys 4.9% (1 in 20), girls 1.4% (1 in 71); ratio 3.4:1","excerpt":"\"About 1 in 31 (3.2%) children aged 8 years has been identified with ASD according to estimates from CDC's Autism and Developmental Disabilities Monitoring (ADDM) Network.\"\n","source_date":"2025-04-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260412090444/https://www.cdc.gov/mmwr/volumes/74/ss/ss7402a1.htm","calculation_notes":"The ADDM 2022 report gives 3.2% prevalence among 8-year-olds as the headline figure. This is used directly as the native and normalized value because 8-year-old prevalence is the standard epidemiological benchmark for ASD. Numerator 32 per denominator 1,000 = 0.032. Boys: 4.9% (1 in 20). Girls: 1.4% (1 in 71). Ratio: 3.4 boys per girl.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26709141/","title":"Heritability of autism spectrum disorders: a meta-analysis of twin studies","publisher":"Journal of Child Psychology and Psychiatry","source_type":"peer_reviewed","statistic":"ASD heritability estimated at 64-91% across twin studies (point estimates 74% [95% CI 0.70-0.87] and, with thresholds set to 5% prevalence, 93% [95% CI 0.77-0.99])","excerpt":"\"The meta-analytic heritability estimates were substantial: 64-91%.\"\n","source_date":"2016-01-21","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260210205701/https://pubmed.ncbi.nlm.nih.gov/26709141/","calculation_notes":"The meta-analysis aggregates seven twin studies. The first meta-analysis yielded a heritability of 74% (95% CI 0.70-0.87); when thresholds were fixed to 5% prevalence the estimate rose to 93% (95% CI 0.77-0.99), giving the reported 64-91% range. This figure is used in the body text as context for the genetics discussion. It does not feed into the prevalence calculation directly but establishes that genetic factors dominate ASD etiology.\n","independence_note":"This is an independent academic meta-analysis of twin concordance data, entirely separate from the CDC ADDM surveillance system used for the prevalence figure.\n"},{"url":"https://publichealth.jhu.edu/2025/is-there-an-autism-epidemic","title":"Is There an Autism Epidemic?","publisher":"Johns Hopkins Bloomberg School of Public Health","source_type":"reputable_reference","statistic":"Broadened diagnostic criteria, better screening, and increased awareness account for the majority of the prevalence rise over the past two decades","excerpt":"\"A gradual rise over the past 20 years is due to broadened diagnostic definitions, better screening, and increased awareness.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260310214917/https://publichealth.jhu.edu/2025/is-there-an-autism-epidemic","calculation_notes":"This is a contextual source explaining the trajectory of ASD prevalence. It does not contribute a numeric estimate but provides the expert consensus that most of the increase from 1 in 150 (2000) to 1 in 31 (2022) reflects diagnostic expansion rather than a true increase in the underlying condition.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/72/ss/ss7202a1.htm","title":"Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2020","publisher":"CDC MMWR Surveillance Summaries","source_type":"govt_report","statistic":"27.6 per 1,000 (1 in 36) of 8-year-olds identified with ASD across 11 ADDM sites in the 2020 surveillance year; 3.8x higher among boys than girls (43.0 vs 11.4 per 1,000)","excerpt":"\"The overall ASD prevalence was 27.6 per 1,000 (one in 36) children aged 8 years... was 3.8 times as prevalent among boys as among girls (43.0 versus 11.4).\"\n","source_date":"2023-03-24","source_accessed":"2026-07-03","archive_url":"https://web.archive.org/web/20260711034933/https://www.cdc.gov/mmwr/volumes/72/ss/ss7202a1.htm","calculation_notes":"The ADDM 2020 report (11 sites) is the surveillance year immediately preceding the 2022 report (16 sites, first source above) and supplies the intermediate data point cited in the body text's prevalence trajectory: 1 in 150 (2000) -> 1 in 68 (2010) -> 1 in 36 (2020) -> 1 in 31 (2022). Site count grew from 11 to 16 between the two most recent surveillance years, which is one contributor to the apparent trend alongside genuine diagnostic-practice change.\n"}],"comparison_anchors":[{"label":"Child pool drowning (ages 0-14, US)","lifetime_us_adult":0.000435},{"label":"SIDS (per live birth, US)","lifetime_us_adult":0.00034}],"regional_breakdown":[{"region":"Boys","probability":0.049,"notes":"1 in 20; 4.9% prevalence among 8-year-old males (CDC ADDM 2022)"},{"region":"Girls","probability":0.014,"notes":"1 in 71; 1.4% prevalence among 8-year-old females (CDC ADDM 2022)"},{"region":"US overall (2000)","probability":0.0067,"notes":"1 in 150; earliest CDC ADDM estimate, 2000 surveillance year"},{"region":"US overall (2010)","probability":0.0147,"notes":"1 in 68; CDC ADDM 2010 surveillance year"},{"region":"US overall (2020)","probability":0.0276,"notes":"1 in 36; CDC ADDM 2020 surveillance year (11 sites), published March 2023"},{"region":"US overall (2022)","probability":0.032,"notes":"1 in 31; CDC ADDM 2022 surveillance year (latest as of April 2025)"}],"personal_factor_multipliers":[{"factor":"male child","multiplier":1.53,"notes":"Boys 4.9% vs overall 3.2% (CDC ADDM 2022)"},{"factor":"female child","multiplier":0.44,"notes":"Girls 1.4% vs overall 3.2%; underdiagnosis in girls is increasingly recognized"},{"factor":"sibling with ASD","multiplier":6.25,"notes":"Recurrence risk ~20% when an older sibling has ASD (Ozonoff et al. 2011)"},{"factor":"paternal age >40","multiplier":1.5,"notes":"Advanced paternal age associated with ~1.5x risk (Sandin et al. 2016)"},{"factor":"extremely preterm birth (<28 weeks)","multiplier":2.5,"notes":"Very preterm infants have roughly 2-3x the ASD prevalence of term infants"}],"short_label":"Autism diagnosis","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"The 1-in-31 figure comes from the CDC's ADDM Network, which uses health and education records rather than direct clinical assessment. It captures children identified by age 8; some individuals — especially girls and those without intellectual disability — receive a diagnosis later in adolescence or adulthood, so the true lifetime prevalence is likely somewhat higher. Prevalence varies considerably by ADDM site; California reported rates as high as 1 in 12.5 in the 2022 data. The spectrum ranges from profound disability requiring full-time support to subclinical traits compatible with full independence, so a single prevalence number conceals enormous heterogeneity in outcomes.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single puzzle piece in muted blue resting on a pale background, flat vector illustration, calm and understated."},"canonical_url":"https://likelier.app/autism-diagnosis-child","api_url":"https://likelier.app/api/fears/autism-diagnosis-child.json"},{"slug":"hospital-acquired-infection","question":"What are the odds of getting a serious infection during a hospital stay?","quick_answer":"About 1 in 31 US hospital patients picks up an infection they did not arrive with, and roughly 1 in 200 admissions involves a fatal one — far above the under-1-in-1,000 rate most patients assume.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"There is no standing survey that isolates fear of a healthcare-associated infection, but the category sits in a near-universal blind spot. Patients being wheeled into an acute-care bed tend to model the hospital as a place that neutralises infection risk, not a place that generates it. The iatrogenic framing — that roughly one admission in thirty picks up a bloodstream, surgical-site, urinary, pneumonia, or C. difficile infection that was not present on arrival — is absent from almost every informed- consent conversation and from almost every lay intuition about hospital safety.\n","rough_estimate":"most patients assume the per-admission rate is well under 1 in 1,000","kind":"intuition"},"native":{"display":"~1 in 31 US hospital patients has at least one HAI on any given day","numerator":1,"denominator":31,"unit":"per hospitalised patient (point prevalence)","population":"US acute care hospital patients"},"normalized":{"lifetime_us_adult":0.032,"display":"~1 in 31 per US hospital admission (any HAI)","log_value":-1.495,"assumptions":"The headline figure is per hospital admission, not per adult lifetime. CDC's 2015 point-prevalence survey (the follow-up to Magill et al. 2014) found that about 3% of hospitalised patients had one or more HAIs on any given day, which the agency rounds to \"1 in 31 hospital patients.\" Magill et al. 2014 reported a slightly higher 4.0% prevalence (1 in 25) from 2011 data; the 16% relative decline between the two surveys is real and documented. Likelier reports the more recent 2015 figure as the headline (0.032) with Magill's 0.040 as the upper end of the uncertainty band. Note that point prevalence slightly understates per-admission cumulative incidence — patients admitted briefly and discharged without an HAI are fully represented in the denominator, but a patient who develops an HAI on day 10 is only counted on days when the infection is present — so the \"per admission\" framing here is a lower bound. The fatal-HAI per-admission figure (~0.005) comes from 72,000 HAI-associated deaths against roughly 14.4 million annual US acute-care admissions with HAI exposure windows, which is the second regional_breakdown row below.\n","uncertainty":{"low":0.025,"high":0.05},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/24670166/","title":"Multistate Point-Prevalence Survey of Health Care-Associated Infections","publisher":"New England Journal of Medicine / Magill SS, Edwards JR, Bamberg W, et al.","source_type":"peer_reviewed","statistic":"4.0% point prevalence of HAI among hospitalised patients (95% CI 3.7-4.4); estimated 648,000 patients with 721,800 HAIs in US acute care hospitals in 2011","excerpt":"\"Of 11,282 patients, 452 had 1 or more health care-associated infections (4.0%; 95% confidence interval, 3.7 to 4.4).\" \"We estimated that there were 648,000 patients with 721,800 health care-associated infections in U.S. acute care hospitals in 2011.\"\n","source_date":"2014-03-27","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172940/https://pubmed.ncbi.nlm.nih.gov/24670166/","calculation_notes":"Magill et al. is the canonical US HAI point-prevalence study and the origin of the widely-cited \"1 in 25 hospitalised patients has an HAI\" figure. Pneumonia (21.8%) and surgical-site infections (21.8%) were the two most common categories, followed by gastrointestinal infections (17.1%, dominated by C. difficile, which was the single most common pathogen at 12.1% of all HAIs). The 4.0% point- prevalence figure anchors the upper end of Likelier's uncertainty band; the 2015 CDC follow-up survey (3%, \"1 in 31\") is used as the headline because it is more recent and reflects a genuine decline in HAI rates post-2011.\n","independence_note":"Methodologically upstream of CDC's 2015 survey: both are point-prevalence surveys run through the Emerging Infections Program using the same case definitions, so the two should be read as one dataset compared against itself across years rather than as two fully independent estimates. WHO's global report is independent of both and is the cleaner cross-check.\n"},{"url":"https://www.cdc.gov/healthcare-associated-infections/php/data/index.html","title":"Healthcare-Associated Infections (HAIs) Data","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"~1 in 31 US hospital patients has at least one HAI on any given day; ~687,000 HAIs and ~72,000 HAI-associated in-hospital deaths in 2015","excerpt":"\"On any given day, about one in 31 hospital patients has at least one healthcare- associated infection.\" \"There were an estimated 687,000 HAIs in U.S. acute care hospitals in 2015.\" \"About 72,000 hospital patients with HAIs died during their hospitalizations.\"\n","source_date":"2015-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412034755/https://www.cdc.gov/healthcare-associated-infections/php/data/index.html","calculation_notes":"CDC's 2015 HAI Hospital Prevalence Survey is the direct successor to Magill et al. 2014 and is the source of the headline \"1 in 31\" figure used throughout this entry. The 72,000 annual HAI-associated in-hospital deaths, divided across roughly ~14.4 million at-risk US acute-care admissions, gives a per-admission fatal-HAI rate of roughly 0.5% — the second regional_breakdown row. The CDC also notes a 16% relative decline in HAI prevalence between 2011 (Magill) and 2015 and further decreases between 2023 and 2024 across CAUTI and C. difficile infections, which is why the 2011 4.0% figure is the ceiling of the uncertainty band rather than the headline.\n","independence_note":"CDC's 2015 survey shares methodology with Magill et al. 2014 (same point- prevalence design, same Emerging Infections Program network). Treat as a time- series update of the same dataset, not a fully independent estimate.\n"},{"url":"https://www.who.int/news/item/06-05-2022-who-launches-first-ever-global-report-on-infection-prevention-and-control","title":"WHO launches first ever global report on infection prevention and control","publisher":"World Health Organization","source_type":"govt_report","statistic":"7 per 100 patients in high-income countries and 15 per 100 patients in low- and middle-income countries acquire at least one HAI during hospital stay; ~1 in 10 affected patients dies","excerpt":"\"out of every 100 patients in acute-care hospitals, seven patients in high-income countries and 15 patients in low- and middle-income countries will acquire at least one health care-associated infection.\" \"On average, 1 in every 10 affected patients will die from their HAI.\"\n","source_date":"2022-05-06","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173024/https://www.who.int/news/item/06-05-2022-who-launches-first-ever-global-report-on-infection-prevention-and-control","calculation_notes":"WHO's 2022 Global Report on Infection Prevention and Control gives the cleanest cross-country comparison. The 7%/15% figures are per admission (cumulative incidence over the stay, not point prevalence), which is why they are higher than the US point-prevalence number used as the headline: a 7% per-admission cumulative incidence in a high-income system is consistent with a ~3% point prevalence on any given day once you account for admission length and infection-onset timing. The 10% HAI case-fatality rate, applied to the WHO high-income figure, gives ~0.7% per-admission fatal-HAI probability, bracketing the CDC-derived 0.5% figure. The 15% LMIC figure is the source of the regional_breakdown row for LMIC admissions below.\n","independence_note":"Fully independent of Magill and CDC: WHO synthesises hundreds of national surveys across dozens of countries, most of which are not part of the US Emerging Infections Program network. This is the strongest cross-check in the entry.\n"}],"comparison_anchors":[{"label":"Serious ADR per hospital admission (Lazarou 1998)","lifetime_us_adult":0.067},{"label":"Lifetime fatal ADR (US adult)","lifetime_us_adult":0.0171},{"label":"Lifetime death in a car crash (US)","lifetime_us_adult":0.0108},{"label":"Lifetime death in a plane crash (US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Any HAI per US admission","probability":0.032,"notes":"CDC 2015 point-prevalence figure — \"1 in 31 hospital patients.\" Magill 2014 put the equivalent 2011 number at 4.0%.\n"},{"region":"Fatal HAI per US admission","probability":0.005,"notes":"~72,000 HAI-associated in-hospital deaths divided across roughly 14.4M at-risk acute-care admissions. Order-of-magnitude figure.\n"},{"region":"Any HAI per LMIC admission","probability":0.12,"notes":"WHO 2022 global report: 15 per 100 admissions in low- and middle-income countries acquire at least one HAI, roughly double the 7/100 high-income country rate. Point estimate rounded to 0.12.\n"},{"region":"ICU admission, any HAI","probability":0.15,"notes":"WHO reports overall ICU HAI incidence around 30% globally, with an order of magnitude more variation across countries; the 0.15 figure here is closer to the high-income ICU baseline (Magill 2014 reported ICU HAI point prevalence several multiples higher than ward averages). Use 0.30 as the LMIC ICU ceiling.\n"}],"personal_factor_multipliers":[{"factor":"ICU admission","multiplier":5,"notes":"ICU patients face 5-10x higher HAI rates than general ward patients — the combination of invasive devices, longer stays, broad-spectrum antibiotics, and sicker hosts. WHO's global ICU HAI average is ~30%.\n"},{"factor":"mechanical ventilation","multiplier":8,"notes":"Ventilator-associated pneumonia is one of the top HAI categories and scales roughly linearly with ventilator-days. The 8x multiplier is an order-of-magnitude estimate against the per-admission baseline.\n"},{"factor":"central line","multiplier":4,"notes":"Central line-associated bloodstream infections (CLABSI) are the textbook preventable HAI — rates have fallen sharply since the 2000s checklist era but remain 3-5x the general hospital baseline.\n"},{"factor":"length of stay >7 days","multiplier":3,"notes":"HAI hazard accumulates roughly linearly with bed-days once past the initial colonisation window. A two-week stay is a different risk profile than a 48-hour admission for observation.\n"},{"factor":"immunocompromised","multiplier":3,"notes":"Oncology, transplant, and high-dose steroid patients face elevated rates across almost every HAI category, particularly invasive fungal and opportunistic bacterial infections.\n"}],"short_label":"Hospital infection","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline point-prevalence figure (\"1 in 31\") is a snapshot, not a cumulative per-admission risk, and slightly understates the probability that a given admission includes an HAI somewhere along its timeline. The WHO per-admission figures (7/100 high-income, 15/100 LMIC) are the cleaner per-admission numbers and are closer to what a patient being admitted should mentally budget. The distribution is also radically non-uniform: a 48-hour observation admission on a general ward is a very different risk than a three-week ICU stay with a ventilator and two central lines, which is what the personal_factor_multipliers above are trying to capture. Finally, the US figure has improved meaningfully since 2011 (a 16% relative decline between Magill 2014 and CDC 2015, with further gains through 2024 in CAUTI and C. difficile), so the headline is a trailing indicator — the 2026 figure is probably somewhat lower than 1 in 31, though no current survey cleanly replaces the 2015 baseline.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty hospital bed rendered as a minimal flat vector on a muted background."},"canonical_url":"https://likelier.app/hospital-acquired-infection","api_url":"https://likelier.app/api/fears/hospital-acquired-infection.json"},{"slug":"c-section-complications","question":"What are the odds of serious complications from a C-section vs vaginal birth?","quick_answer":"Serious complications occur in about 8-12% of individual cesarean deliveries, which averages to roughly 3.4% per woman over her births. That sits within the 5-15% range most people expect.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Cesarean delivery occupies an unusual spot in risk perception: some parents view it as the safer, more controlled option, while others see it as major abdominal surgery fraught with danger. Surveys of expectant mothers in high-income countries show wide variance — many underestimate surgical risks, while a vocal minority overestimates them. The net effect is that perceived risk tracks actual risk only loosely, with strong anchoring to personal birth stories and media framing rather than population data.\n","rough_estimate":"~5-15% serious complication rate, depending on framing","kind":"intuition"},"native":{"display":"~8-12% serious complications per cesarean delivery","numerator":10,"denominator":100,"unit":"per delivery","population":"women undergoing cesarean delivery in high-income countries"},"normalized":{"lifetime_us_adult":0.034,"display":"~3.4% per woman (assuming ~1.06 cesarean deliveries among women who deliver)","log_value":-1.47,"assumptions":"US cesarean rate is 32.3% (CDC 2023). Among US women who give birth (~85% of women), average ~2 deliveries, so ~0.65 cesareans per mother. Serious complication rate of ~5% per cesarean (requiring ICU, hysterectomy, transfusion, or organ injury) gives ~3.4% cumulative. The native display of 8-12% includes all complications (minor and serious); the 5% used for normalization is restricted to Clavien-Dindo Grade III+ events only. Vaginal baseline serious complications ~2-4% per delivery. \"Serious\" defined as Grade III-IV Clavien-Dindo or equivalent. Emergency cesarean carries roughly 2x the complication rate of planned cesarean.\n","uncertainty":{"low":0.015,"high":0.06},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/29360829/","title":"Long-term risks and benefits associated with cesarean delivery for mother, baby, and subsequent pregnancies: Systematic review and meta-analysis","publisher":"PLoS Medicine / Keag, Norman & Stock","source_type":"peer_reviewed","statistic":"Placenta previa OR 1.74 (1.62-1.87), placenta accreta OR 2.95 (1.32-6.60), uterine rupture OR 25.81 (10.96-60.76) in subsequent pregnancies after cesarean","excerpt":"\"Previous cesarean delivery was associated with increased risk of placenta previa (OR 1.74, 95% CI 1.62 to 1.87), placenta accreta (OR 2.95, 1.32 to 6.60), and uterine rupture (OR 25.81, 10.96 to 60.76) in subsequent pregnancies. Cesarean delivery was associated with reduced urinary incontinence (OR 0.56, 0.47 to 0.66) and pelvic organ prolapse (OR 0.29, 0.17 to 0.51).\"\n","source_date":"2018-01-23","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413171021/https://pubmed.ncbi.nlm.nih.gov/29360829/","calculation_notes":"Keag et al. is a systematic review of 79 studies covering long-term outcomes. ORs reported are for subsequent pregnancies after an index cesarean vs vaginal delivery. Placenta accreta risk rises with each additional cesarean (from ~0.3% at first to ~6% at fifth). These long-term cumulative risks are factored into the normalized lifetime estimate. The review did not find studies on maternal death directly.\n","independence_note":"Meta-analysis pooling individual cohort studies; overlapping populations possible across included studies but each adds independent weight.\n"},{"url":"https://www.who.int/news/item/16-06-2021-caesarean-section-rates-continue-to-rise-amid-growing-inequalities-in-access","title":"Caesarean section rates continue to rise, amid growing inequalities in access","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"Global cesarean rate 21% (2021), projected 29% by 2030; unnecessary surgical procedures can be harmful for woman and baby","excerpt":"\"Caesarean sections are absolutely critical to save lives in situations where vaginal deliveries would pose risks, so all health systems must ensure timely access for all women when needed. Unnecessary surgical procedures can be harmful, both for a woman and her baby.\"\n","source_date":"2021-06-16","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260407020615/https://www.who.int/news/item/16-06-2021-caesarean-section-rates-continue-to-rise-amid-growing-inequalities-in-access","calculation_notes":"WHO provides the global context and framing. The 21% global rate and wide regional variance (5% sub-Saharan Africa to 43% Latin America) contextualizes the risk discussion. WHO does not provide per-delivery complication rates in this release but establishes the authoritative position that both under- and over-use carry risk.\n","independence_note":"WHO press release synthesising country-level cesarean rates from DHS / MICS surveys and national health statistics. Not an independent estimate of complication rates — used only for global-rate context and the institutional framing around appropriate use.\n"},{"url":"https://www.cdc.gov/nchs/fastats/delivery.htm","title":"FastStats — Delivery Methods","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"US cesarean delivery rate 32.3% in 2023 (1,161,896 cesarean deliveries out of 3,593,396 total births)","excerpt":"\"Cesarean delivery: 32.3% of all deliveries in 2023. Vaginal deliveries: 2,431,500. Cesarean deliveries: 1,161,896.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260406175745/https://www.cdc.gov/nchs/fastats/delivery.htm","calculation_notes":"CDC rate used as the baseline for normalizing US cesarean exposure. At 32.3% of ~3.6 million annual births, roughly 1.16 million cesareans occur per year in the US. Combined with average fertility rate (~1.6 children per woman), the typical US mother has a meaningful probability of at least one cesarean.\n","independence_note":"CDC natality data is an independent administrative dataset from US birth certificates, fully independent of the Keag meta-analysis which drew primarily on European cohorts.\n"}],"comparison_anchors":[{"label":"Serious complication from vaginal delivery (per delivery)","lifetime_us_adult":0.02},{"label":"Maternal death, cesarean (US, per delivery)","lifetime_us_adult":0.000085},{"label":"Maternal death, vaginal (US, per delivery)","lifetime_us_adult":0.000035}],"regional_breakdown":[{"region":"Planned cesarean (high-income)","probability":0.04,"notes":"Elective/scheduled cesarean at term with no emergency factors; lowest surgical complication profile"},{"region":"Emergency cesarean (high-income)","probability":0.12,"notes":"Unplanned cesarean during labor; roughly 2-3x complication rate of planned cesarean"},{"region":"Vaginal delivery (high-income)","probability":0.025,"notes":"Serious complications including severe perineal tear, hemorrhage, or infection requiring intervention"},{"region":"Cesarean (low-resource setting)","probability":0.18,"notes":"Higher rates of surgical site infection, anesthesia complications, and delayed care"}],"personal_factor_multipliers":[{"factor":"Emergency vs planned cesarean","multiplier":2.5,"notes":"ACOG and Keag et al. (PLoS Medicine 2018) meta-analysis; emergency cesarean carries roughly 2-3x the serious complication rate of a scheduled cesarean at term due to underlying obstetric emergency, time pressure, and physiological compromise"},{"factor":"Obesity (BMI > 35)","multiplier":2,"notes":"ACOG Practice Bulletin: BMI > 35 approximately doubles wound complication rate (surgical site infection, wound dehiscence, seroma) following cesarean; also increases risk of VTE and anesthesia complications"},{"factor":"Multiple prior cesareans (3 or more)","multiplier":4,"notes":"Keag et al. (PLoS Medicine 2018) and ACOG: placenta accreta spectrum (PAS) risk rises from ~0.3% after first cesarean to ~3-6% after third or subsequent cesarean — a roughly 10-20x increase over baseline, which translates to ~4x increase in major hemorrhage and hysterectomy risk for the overall serious-complication composite"},{"factor":"Low-resource or unplanned delivery setting","multiplier":2,"notes":"WHO (2021) and global comparative studies: cesarean in low-resource settings carries approximately 1.5-2x higher serious complication rates due to surgical site infection, limited blood banking, anesthesia limitations, and delayed response to emergencies"}],"short_label":"C-section complications","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"\"Serious complication\" definitions vary across studies (some include only life-threatening events, others include extended hospital stay or readmission). The figures here use a composite definition roughly equivalent to Clavien-Dindo Grade III+. Individual risk depends heavily on indication (planned vs emergency), number of prior cesareans, maternal age, BMI, and comorbidities. Cesarean delivery also confers benefits — notably reduced pelvic floor injury and elimination of labor-related fetal distress risk. This entry is neutral on birth choice; both modes carry trade-offs that are best discussed with an obstetrician in context.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":3.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-12","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A simplified surgical light hovering above an empty operating table, rendered in muted teal and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/c-section-complications","api_url":"https://likelier.app/api/fears/c-section-complications.json"},{"slug":"melanoma-uv-exposure","question":"What are the odds of getting melanoma from regular unprotected sun exposure?","quick_answer":"Fair-skinned US adults have about a 1 in 29 (men) to 1 in 40 (women) lifetime chance of a melanoma diagnosis; dying of it is far rarer, around 1 in 269 (men) and 1 in 496 (women). About 86% of the risk is UV-attributable, and so behavior-modifiable.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Most people know that sun exposure causes skin cancer in a vague, background-knowledge sort of way, yet UV protection remains inconsistent: fewer than half of US adults report regular sunscreen use, and tanning beds still attract millions of users per year. Melanoma is the fifth most common cancer in the United States, but its lethality relative to its incidence does not register the way lung or pancreatic cancer does. The result is a risk that most people file as \"probably fine if I'm careful\" rather than one they actively calibrate against.\n","rough_estimate":"50% of US adults are very or somewhat worried about getting cancer (Gallup, all sites); melanoma-specific worry is lower and inconsistent with actual UV behavior","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~1 in 29 lifetime (non-Hispanic white US men, diagnosis)","numerator":1,"denominator":29,"unit":"lifetime","population":"non-Hispanic white US adults"},"normalized":{"lifetime_us_adult":0.034,"display":"1 in ~29 lifetime (fair-skinned US men); 1 in ~40 (fair-skinned US women)","log_value":-1.47,"assumptions":"Uses the American Cancer Society direct lifetime probability from SEER 2020-2022 data: 3.5% (1 in 29) for non-Hispanic white men and 2.5% (1 in 40) for non-Hispanic white women. The headline figure of 0.034 is the sex-averaged midpoint weighted toward the higher male rate, reflecting the population-level lifetime diagnosis probability for fair-skinned US adults with typical (i.e., inconsistent) UV protection behavior. This is a diagnosis probability, not a mortality probability — the 5-year relative survival for melanoma is ~95% overall and ~99% for localized disease. Mortality lifetime risk is much lower: ~0.4% for white men (1 in 269), ~0.2% for white women (1 in 496). The ~86% UV-attributable fraction (per the American Academy of Dermatology and WHO) means the vast majority of this risk is behavior-modifiable.\n","uncertainty":{"low":0.025,"high":0.045},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","title":"Lifetime Probability of Developing or Dying From Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Melanoma of the skin: men 3.5% (1 in 29) risk of developing, 0.4% (1 in 269) risk of dying; women 2.5% (1 in 40) risk of developing, 0.2% (1 in 496) risk of dying — figures are for non-Hispanic White people","excerpt":"\"Melanoma of the skin: 3.5% risk of developing (1 in 29), 0.4% risk of dying from (1 in 269) [men]. Melanoma of the skin: 2.5% risk of developing (1 in 40), 0.2% risk of dying from (1 in 496) [women]. [...] The risk numbers for melanoma are for non-Hispanic White people.\"\n","source_date":"2025-01-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164427/https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","calculation_notes":"ACS uses SEER mortality and incidence data (2020-2022) to compute direct lifetime probabilities from a life-table conditional on birth. The 3.5% male and 2.5% female figures are the gold standard for US non-Hispanic white lifetime melanoma diagnosis risk. Sex-averaged midpoint: (0.035 + 0.025) / 2 = 0.030; weighted slightly toward the male rate (higher incidence, larger share of melanoma burden) gives ~0.034 as the headline. Uncertainty band 0.025-0.045 spans the female-only to the upper bound for high-UV-exposure male subgroups.\n","independence_note":"ACS lifetime probability tables are built directly on SEER 2020-2022 incidence/mortality data and life tables from the same NCHS pipeline referenced by the SEER Stat Facts source below. Treat ACS and SEER as one analytical pipeline on a shared upstream dataset; the AAD and Green (Nambour RCT) sources provide the genuine independent verification.\n"},{"url":"https://seer.cancer.gov/statfacts/html/melan.html","title":"Cancer Stat Facts: Melanoma of the Skin","publisher":"National Cancer Institute / SEER Program","source_type":"govt_report","statistic":"Approximately 2.2% of men and women will be diagnosed with melanoma at some point during their lifetime; non-Hispanic white males 39.7 per 100,000/year, non-Hispanic white females 26.8 per 100,000/year; 5-year relative survival 94.7%","excerpt":"\"Approximately 2.2 percent of men and women will be diagnosed with melanoma of the skin at some point during their lifetime.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413175938/https://seer.cancer.gov/statfacts/html/melan.html","calculation_notes":"SEER's 2.2% overall figure covers all races/ethnicities. Non-Hispanic white incidence rates (39.7 M, 26.8 F per 100,000/year) are roughly 30-40x the rates for Black Americans (1.0 M, 0.9 F), confirming the enormous racial disparity that justifies scoping this entry to fair-skinned adults. The 94.7% 5-year survival rate underscores that melanoma diagnosis risk far exceeds mortality risk.\n","independence_note":"SEER is the upstream data source for the ACS lifetime probability calculations in source 1. Treat as the same pipeline for incidence/mortality figures; included here for the race-stratified incidence rates and overall population figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21135266/","title":"Reduced melanoma after regular sunscreen use: randomized trial follow-up","publisher":"Journal of Clinical Oncology / Green AC, Williams GM, Logan V, Strutton GM","source_type":"peer_reviewed","statistic":"Daily sunscreen users had half the melanoma incidence of discretionary users (HR 0.50, 95% CI 0.24-1.02); invasive melanoma reduction was 73% (HR 0.27, 95% CI 0.08-0.97)","excerpt":"\"11 new primary melanomas had been identified in the daily sunscreen group, and 22 had been identified in the discretionary group (hazard ratio [HR], 0.50; 95% CI, 0.24 to 1.02; P = .051). The reduction in invasive melanomas was substantial (n = 3 in active v 11 in control group; HR, 0.27; 95% CI, 0.08 to 0.97).\"\n","source_date":"2011-01-20","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413180017/https://pubmed.ncbi.nlm.nih.gov/21135266/","calculation_notes":"The Nambour RCT (Queensland, Australia; 1,621 participants; 1992-2006 follow-up) is the only large-scale randomized trial of sunscreen for melanoma prevention. The overall HR of 0.50 (borderline significance at p=.051) and the invasive-melanoma HR of 0.27 (significant) anchor the \"~50% melanoma reduction with regular sunscreen\" claim used in personal_factor_multipliers. The wide confidence intervals reflect the relatively small number of melanoma events (33 total) over the 10-year follow-up.\n","independence_note":"Fully independent of SEER/ACS pipeline — Australian population-based RCT with its own endpoint ascertainment.\n"},{"url":"https://www.aad.org/media/stats-skin-cancer","title":"Skin cancer statistics","publisher":"American Academy of Dermatology","source_type":"reputable_reference","statistic":"1 in 28 men and 1 in 39 women lifetime melanoma risk; 5+ blistering sunburns age 15-20 increases melanoma risk by 80%; indoor tanning before 30 increases melanoma risk 6x for women","excerpt":"\"It is estimated that melanoma will affect 1 in 28 men and 1 in 39 women in their lifetime. [...] Experiencing five or more blistering sunburns between ages 15 and 20 increases one's melanoma risk by 80% and nonmelanoma skin cancer risk by 68%.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413180054/https://www.aad.org/media/stats-skin-cancer","calculation_notes":"AAD figures (1 in 28 M, 1 in 39 F) are marginally different from ACS (1 in 29 M, 1 in 40 F) due to different data vintage and rounding; both are within expected variation. The 80% sunburn multiplier (5+ blistering sunburns age 15-20) and the indoor tanning risk figures are used in personal_factor_multipliers.\n","independence_note":"AAD compiles statistics from multiple sources including SEER, ACS, and independent dermatology literature. The sunburn and tanning bed figures derive from separate epidemiological studies (Lew et al., Lazovich et al.) and are independently sourced from the SEER lifetime probability.\n"},{"url":"https://www.aimatmelanoma.org/wp-content/uploads/melanoma-statistics-2026.pdf","title":"Melanoma Statistics 2026","publisher":"AIM at Melanoma Foundation (citing American Cancer Society, Cancer Facts & Figures 2026)","source_type":"reputable_reference","statistic":"An estimated 234,680 people will be diagnosed with melanoma in the US in 2026 (122,680 in situ, 112,000 invasive)","excerpt":"\"In 2026, an estimated 234,680 people will be diagnosed with melanoma.\"\n","source_date":"2026-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260308180320/https://www.aimatmelanoma.org/wp-content/uploads/melanoma-statistics-2026.pdf","calculation_notes":"The 234,680 figure originates from the American Cancer Society's \"Cancer Facts & Figures 2026\" annual report (122,680 in situ + 112,000 invasive = 234,680 total new US melanoma cases estimated for 2026); AIM at Melanoma Foundation's statistics sheet reproduces the ACS estimate. Used here to source the body text's \"234,680 new cases projected for 2026\" figure, which supports the entry's fifth-most-common-cancer underrated framing.\n","independence_note":"Distinct from the ACS lifetime-probability tables (source 1) and SEER Stat Facts (source 2) used elsewhere in this entry — this source is used solely for the annual new-case count estimate, not the lifetime probability figures.\n"}],"comparison_anchors":[{"label":"Death from lung cancer (lifetime, US)","lifetime_us_adult":0.036},{"label":"Death from cancer (all sites, lifetime, US)","lifetime_us_adult":0.14},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Non-Hispanic white US men","probability":0.035,"notes":"ACS direct lifetime probability from SEER 2020-2022; 1 in 29"},{"region":"Non-Hispanic white US women","probability":0.025,"notes":"ACS direct lifetime probability from SEER 2020-2022; 1 in 40"},{"region":"US overall (all races)","probability":0.022,"notes":"SEER all-race figure; ~2.2% lifetime diagnosis probability"},{"region":"Hispanic US adults","probability":0.005,"notes":"ACS estimate; 1 in 200"},{"region":"Black US adults","probability":0.001,"notes":"ACS estimate; 1 in 1,000 — incidence ~30-40x lower than non-Hispanic white"},{"region":"Australia (fair-skinned)","probability":0.05,"notes":"Australia has the highest melanoma incidence rate globally; estimated ~1 in 20 lifetime for fair-skinned Australians"}],"personal_factor_multipliers":[{"factor":"Fitzpatrick skin type I-II (very fair, always burns)","multiplier":2,"notes":"Relative to the population average for all white adults; very fair skin types have roughly double the melanoma incidence of olive-complexioned white adults"},{"factor":"regular broad-spectrum sunscreen use (SPF 15+)","multiplier":0.5,"notes":"Green et al. 2011 Nambour RCT: daily sunscreen users had half the melanoma incidence (HR 0.50); invasive melanoma reduction was even greater (HR 0.27). This is the only RCT evidence for sunscreen and melanoma."},{"factor":"5+ blistering sunburns age 15-20","multiplier":1.8,"notes":"AAD statistic: 80% increased melanoma risk from adolescent blistering sunburns. Reflects the critical window of UV damage during skin development."},{"factor":"indoor tanning before age 35","multiplier":1.75,"notes":"Meta-analyses estimate 59-75% increased melanoma risk from indoor tanning before age 35; AAD reports women under 30 who tan indoors are 6x more likely to develop melanoma"},{"factor":"first-degree family history of melanoma","multiplier":2,"notes":"Approximately doubles lifetime melanoma risk independent of shared UV exposure; reflects inherited variants in MC1R, CDKN2A, and other melanoma susceptibility genes"},{"factor":"50+ common moles or any atypical moles","multiplier":2.5,"notes":"High mole count is an independent melanoma risk factor; atypical mole syndrome (dysplastic nevi) confers additional risk beyond mole count alone"}],"short_label":"Melanoma (UV)","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry reports diagnosis probability, not mortality. Melanoma caught early (localized stage) has a 99% five-year survival rate; the lifetime mortality risk for white US adults is roughly an order of magnitude lower than the diagnosis risk (~0.4% for men, ~0.2% for women). The \"underrated\" framing refers specifically to the gap between the high population-level incidence and the inconsistent UV protection behavior: melanoma is the 5th most common US cancer, yet regular sunscreen use remains a minority behavior. The ~86% UV-attributable fraction (widely cited by WHO, AAD, and others) makes melanoma one of the most preventable common cancers, yet prevention behavior does not match. The Green et al. RCT confidence intervals are wide (the overall HR 0.50 had p=.051), so the \"50% reduction\" figure should be read as the best available RCT point estimate rather than a precise effect size. Fair-skinned subgroup scoping means the headline number does not apply to Black, Hispanic, or Asian Americans, for whom melanoma risk is dramatically lower.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-13","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single abstract sun shape rendered in muted amber tones against a pale cream background, flat vector illustration."},"canonical_url":"https://likelier.app/melanoma-uv-exposure","api_url":"https://likelier.app/api/fears/melanoma-uv-exposure.json"},{"slug":"air-pollution-death","question":"What are the odds of dying prematurely from air pollution?","quick_answer":"About 1 in 29 over a lifetime for a US adult. Most people file air pollution as an annoyance rather than a leading killer and guess well below the real toll. The deaths are diffuse — heart disease, stroke, COPD, lung cancer — and never make a headline, which is exactly why they are underestimated.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Most people register air pollution as a quality-of-life annoyance — hazy skies, asthma triggers, maybe a cough on a bad-air day — rather than a leading cause of death. When pressed for a number, guesses tend to cluster around traffic-accident territory or lower, well below the actual attributable-mortality figures. The disconnect is partly structural: air-pollution deaths are never acute, never photographed, and never lead a news cycle. Nobody drops dead on a sidewalk with \"PM2.5\" on the death certificate. The deaths are diffused across cardiovascular disease, stroke, COPD, lung cancer, and lower respiratory infections, and the causal chain is long enough that it rarely triggers the availability heuristic the way a plane crash or shark attack does.\n","rough_estimate":"47.0% of US adults report being afraid or very afraid of air pollution (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~100,000–200,000 premature deaths per year attributable to PM2.5 in the US","numerator":100000,"denominator":331000000,"unit":"per year","population":"US population (all ages)"},"normalized":{"lifetime_us_adult":0.035,"display":"~1 in 29 lifetime (US adult)","log_value":-1.456,"assumptions":"Native rate: EPA and Health Effects Institute estimates attribute approximately 100,000–200,000 premature US deaths per year to ambient PM2.5 exposure. We use 100,000 as the conservative point estimate against a US population of ~331 million, yielding an annual attributable mortality rate of ~30 per 100,000 (~0.0003). Lifetime conversion: 1 − (1 − 0.0003)^59 ≈ 0.0176 using the site-standard 59-year remaining-life horizon from age 18. However, the annual rate underestimates the cumulative effect because PM2.5-attributable mortality is heavily concentrated in ages 55+, where baseline mortality is also higher and the attributable fraction rises. Using the GBD 2019 age-weighted attributable fractions and US life tables, the integrated lifetime attributable probability is approximately 0.03–0.05. We use 0.035 as the central estimate. This aligns with the GBD 2019 finding that ambient particulate matter pollution is the sixth-leading risk factor for death globally, and with Burnett et al. 2018 (GEMM) estimates of the PM2.5-mortality concentration-response function at US average exposure levels (~8–10 µg/m³). Uncertainty band 0.02–0.06 reflects the range between the EPA's more conservative attributable-fraction estimates and the higher figures from the GEMM integrated-exposure-response model. The scope is us_adult_lifetime because the normalization uses US-specific exposure levels and US life tables. Global-anchor update (2026-06-14): State of Global Air 2024 (HEI/IHME, GBD 2021) raised the global air-pollution death total to 8.1 million in 2021 — now the second-leading global mortality risk factor — and first quantified NO2 health effects. That report does not provide a US-specific count diverging from the ~100,000–200,000/yr range used here, so the US-adult lifetime point estimate (0.035), its log_value, and uncertainty band are unchanged; only the global framing is refreshed.\n","uncertainty":{"low":0.02,"high":0.06},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health","title":"Ambient (outdoor) air pollution — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Ambient air pollution is estimated to have caused 4.2 million premature deaths worldwide in 2019","excerpt":"\"Ambient (outdoor) air pollution is estimated to have caused 4.2 million premature deaths worldwide in 2019. Some 89% of those premature deaths occurred in low- and middle-income countries, and the greatest number in the WHO South-East Asia and Western Pacific regions.\"\n","source_date":"2024-12-19","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420031756/https://www.who.int/news-room/fact-sheets/detail/ambient-(outdoor)-air-quality-and-health","calculation_notes":"The WHO 4.2 million global figure sets the worldwide scale. To derive a US share: the GBD 2019 attributes roughly 100,000–200,000 US deaths to ambient PM2.5, consistent with the US having ~5% of world population but much lower average PM2.5 exposure (~8 µg/m³ vs global population-weighted ~40 µg/m³). The WHO figure is used as the global anchor; the US-specific estimate comes from EPA/HEI and GBD sources below.\n","independence_note":"WHO draws on IHME GBD estimates for its headline figure but applies its own methodology for risk-factor attribution. Partially dependent on GBD 2019 data cited separately below.\n"},{"url":"https://www.stateofglobalair.org/news-events/2024/new-state-global-air-report-finds-air-pollution-second-leading-risk-factor-death","title":"New State of Global Air Report Finds Air Pollution is Second Leading Risk Factor for Death Worldwide","publisher":"Health Effects Institute / IHME (State of Global Air 2024, GBD 2021)","source_type":"reputable_reference","statistic":"Air pollution ranked as the second-leading risk factor for death globally in 2021, accounting for 8.1 million deaths; the report also quantified NO2 exposure and health effects for the first time","excerpt":"\"Air pollution now ranks as the second leading risk factor for death, accounting for 8.1 million deaths globally\" and \"For the first time, this year's report includes exposure levels and related health effects of nitrogen dioxide (NO2). Traffic exhaust is a major source of NO2, which means densely populated urban areas, particularly in high-income countries, often see the highest levels of NO2 exposures and health impacts.\"\n","source_date":"2024-06-18","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260309162906/https://www.stateofglobalair.org/news-events/2024/new-state-global-air-report-finds-air-pollution-second-leading-risk-factor-death","calculation_notes":"State of Global Air 2024 (HEI/IHME, drawing on GBD 2021) updates the global anchor from the WHO 4.2 million / GBD 2019 4.14 million ambient-PM2.5 figures to a combined air-pollution total of 8.1 million deaths in 2021 — now the second-leading global mortality risk factor, behind only high blood pressure. The larger total reflects (a) a 2021 base year vs 2019, (b) inclusion of NO2 and ozone alongside ambient and household PM2.5, and (c) updated GBD 2021 exposure-response curves. The report does not publish a US-specific attributable death count that differs from the ~100,000–200,000/yr range already used for the US-adult-lifetime headline, so the normalized US lifetime estimate (0.035) is unchanged. This source is added as the current global anchor and as the first authoritative quantification of NO2 health effects, whose exposure is concentrated in dense high-income urban / heavy-traffic settings (already captured by the highway-proximity factor below).\n","independence_note":"State of Global Air 2024 is the HEI/IHME GBD 2021 reporting product and is therefore methodologically continuous with the GBD 2019 Lancet analysis cited below — treat as the updated edition of the same GBD lineage, not an independent line of evidence.\n"},{"url":"https://doi.org/10.1016/S0140-6736(20)30752-2","title":"Global burden of 87 risk factors in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019","publisher":"The Lancet (GBD 2019 Risk Factors Collaborators)","source_type":"peer_reviewed","statistic":"Ambient particulate matter pollution was the sixth-leading risk factor for death globally in 2019, responsible for 4.14 million deaths (95% UI 3.45–4.80 million)","excerpt":"\"In 2019, the leading Level 2 risk factors globally for attributable deaths were high systolic blood pressure (10.8 million deaths), tobacco (8.71 million), dietary risks (7.94 million), air pollution (6.67 million, of which ambient particulate matter 4.14 million), and high fasting plasma glucose (6.50 million).\"\n","source_date":"2020-10-17","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420031830/https://www.thelancet.com/retrieve/pii/S0140673620307522","calculation_notes":"GBD 2019 assigns 4.14 million deaths (95% uncertainty interval 3.45–4.80 million) to ambient particulate matter pollution globally. For the US specifically, GBD 2019 country-level results estimate approximately 100,000– 200,000 attributable deaths, depending on the concentration-response function used. The annual US attributable mortality rate of ~30–60 per 100,000 is then compounded over the 59-year horizon using age-weighted life-table methods to arrive at the 0.035 central lifetime estimate.\n","independence_note":"GBD 2019 is methodologically independent from the Pope et al. ACS CPS-II cohort studies and from the Burnett GEMM model, though GEMM has influenced GBD exposure-response curves in later iterations.\n"},{"url":"https://doi.org/10.1001/jama.287.9.1132","title":"Lung Cancer, Cardiopulmonary Mortality, and Long-term Exposure to Fine Particulate Air Pollution","publisher":"JAMA (Pope, Burnett, Thun, Calle, Krewski, Ito, Thurston)","source_type":"primary_study","statistic":"Each 10 µg/m³ increase in fine particulate air pollution was associated with approximately 4%, 6%, and 8% increased risk of all-cause, cardiopulmonary, and lung cancer mortality, respectively","excerpt":"\"Each 10-µg/m³ elevation in fine particulate air pollution was associated with approximately a 4%, 6%, and 8% increased risk of all-cause, cardiopulmonary, and lung cancer mortality, respectively.\"\n","source_date":"2002-03-06","source_accessed":"2026-04-18","calculation_notes":"Pope et al. 2002 is the landmark ACS Cancer Prevention Study II cohort analysis linking long-term PM2.5 exposure to mortality. The 6% increase in cardiopulmonary mortality per 10 µg/m³ is the core concentration-response coefficient used by EPA in its Integrated Science Assessment for Particulate Matter. At the US national average exposure of ~8 µg/m³ (compared to a counterfactual of ~2.4 µg/m³ per GBD), the excess relative risk is modest for any single individual but applies to the entire population, generating the large attributable death count. The cohort followed ~500,000 adults in 116 metropolitan areas over 16 years.\n","independence_note":"Pope et al. ACS CPS-II is the foundational independent cohort study. GBD and WHO estimates ultimately calibrate their concentration-response functions partly on this study plus the Harvard Six Cities study, so they are not fully independent — but the cohort data itself is primary.\n"},{"url":"https://doi.org/10.1073/pnas.1803222115","title":"Global estimates of mortality associated with long-term exposure to outdoor fine particulate matter","publisher":"PNAS (Burnett et al.)","source_type":"peer_reviewed","statistic":"The Global Exposure Mortality Model (GEMM) estimates 8.9 million deaths/year globally attributable to ambient PM2.5, roughly double previous WHO/GBD estimates","excerpt":"\"We estimated that ambient PM2.5 was associated with 8.9 million deaths in 2015 globally (95% confidence interval 7.5–10.3), which is substantially larger than previous estimates.\"\n","source_date":"2018-09-18","source_accessed":"2026-04-18","calculation_notes":"The GEMM uses a non-linear concentration-response function that produces higher attributable-mortality estimates than the GBD's integrated-exposure- response (IER) model, especially at low exposure levels relevant to the US. Under the GEMM, the US-specific attributable death count is at the higher end of the 100,000–200,000 range. The GEMM estimates are the basis for the upper bound of the uncertainty interval (0.06). The difference between GBD and GEMM estimates is the primary source of structural uncertainty in this entry.\n","independence_note":"Burnett et al. 2018 is methodologically independent of the GBD IER model and provides the strongest alternative concentration-response framework. Burnett is also a co-author on Pope et al. 2002, but the GEMM uses 41 additional cohorts beyond the ACS CPS-II study.\n"},{"url":"https://www.epa.gov/isa/integrated-science-assessment-isa-particulate-matter","title":"Integrated Science Assessment for Particulate Matter","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"EPA concludes there is a causal relationship between long-term PM2.5 exposure and total (non-accidental) mortality","excerpt":"\"The body of evidence is sufficient to conclude that a causal relationship exists between long-term PM2.5 exposure and total (non-accidental) mortality, including cardiovascular and respiratory mortality, as well as lung cancer mortality.\"\n","source_date":"2019-12-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420031927/https://www.epa.gov/isa/integrated-science-assessment-isa-particulate-matter","calculation_notes":"The EPA ISA is the regulatory basis for US National Ambient Air Quality Standards (NAAQS) for PM2.5 and represents the most comprehensive US institutional assessment of the PM2.5-mortality evidence. EPA's regulatory impact analyses have estimated 100,000+ premature deaths per year at current US ambient PM2.5 levels using Pope et al. and subsequent cohort concentration-response coefficients. This is the domestic institutional anchor for the native figure.\n","independence_note":"EPA ISA is an independent institutional review but draws heavily on Pope et al. 2002 and subsequent ACS CPS-II reanalyses for its concentration-response functions. Treat as an authoritative institutional endorsement rather than a fully independent line of evidence.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from lung cancer (lifetime, US)","lifetime_us_adult":0.056},{"label":"Death from heart disease (lifetime, US)","lifetime_us_adult":0.2005},{"label":"Death from a lightning strike (lifetime, US)","lifetime_us_adult":0.0000135}],"regional_breakdown":[{"region":"US average (~8 µg/m³ PM2.5)","probability":0.035,"notes":"Headline estimate at national average exposure"},{"region":"US urban (10–15 µg/m³)","probability":0.05,"notes":"Higher exposure in major metro areas; roughly 1.4x the national average risk"},{"region":"EU average (~12 µg/m³)","probability":0.045,"notes":"European Environment Agency estimates; higher than US but below South/East Asian levels"},{"region":"Delhi / Beijing (50–100+ µg/m³)","probability":0.15,"notes":"Exposure 6–12x the US average; GEMM concentration-response is sub-linear, so risk does not scale proportionally"},{"region":"US wildfire-affected areas (seasonal spikes)","probability":0.04,"notes":"Episodic high exposure during fire season; growing concern with increasing wildfire frequency"}],"personal_factor_multipliers":[{"factor":"COPD or asthma","multiplier":2.5,"notes":"Pre-existing respiratory disease amplifies PM2.5 mortality risk"},{"factor":"Outdoor worker","multiplier":1.5,"notes":"Higher cumulative exposure from occupational time outdoors"},{"factor":"Lives within 150m of a major highway","multiplier":1.8,"notes":"Elevated ultrafine particle and NO2 exposure from traffic emissions"},{"factor":"Living in developing-world megacity","multiplier":4,"notes":"Exposure levels 5–10x US average; see regional_breakdown"},{"factor":"Never-smoker in low-exposure rural area","multiplier":0.3,"notes":"Minimal ambient PM2.5; baseline risk well below national average"}],"short_label":"Air pollution","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"Air-pollution deaths are entirely an attribution exercise — nobody dies with \"PM2.5\" on a death certificate. The figures are population-level estimates of excess mortality derived from cohort studies comparing age-adjusted death rates across exposure gradients, not individual-level predictions. The two major concentration-response models (GBD IER and Burnett GEMM) disagree by roughly a factor of two on global attributable deaths, and the uncertainty at US exposure levels is proportionally larger because the US sits on the flat part of the exposure-response curve where small changes in the slope coefficient translate to large changes in the attributable count. The 0.035 central estimate should be read as \"plausible order of magnitude\" rather than a precise lifetime probability. Indoor air pollution (cooking fuels, household particulates) is excluded from this entry — the WHO attributes a further 3.2 million deaths/year globally to household air pollution, overwhelmingly in low-income countries using solid fuels. Wildfire smoke is included in ambient PM2.5 measurements but its health-effect profile may differ from combustion-engine or industrial PM2.5 due to differences in particle composition. The global anchor here has since risen: the State of Global Air 2024 report (HEI/IHME, GBD 2021) puts total air-pollution deaths at 8.1 million in 2021 — the second-leading global mortality risk factor — and for the first time quantified nitrogen dioxide (NO2) exposure and health effects, which are highest in dense, heavy-traffic urban areas of high-income countries. That larger total partly reflects a later base year and the addition of NO2 and ozone alongside ambient and household PM2.5; it does not change the US-specific attributable count or the US-adult lifetime estimate used above.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A faint haze hanging over a simplified city skyline in muted greys and soft blues, flat vector illustration."},"canonical_url":"https://likelier.app/air-pollution-death","api_url":"https://likelier.app/api/fears/air-pollution-death.json"},{"slug":"end-stage-renal-disease-lifetime","question":"What are the lifetime odds of developing end-stage kidney disease requiring dialysis or transplant?","quick_answer":"The lifetime odds of developing end-stage kidney disease requiring dialysis or a transplant are about 1 in 29, roughly 3 to 4 percent. Most healthy people would guess under 1 percent.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Kidney failure rarely ranks highly in surveys of health fears, despite being among the most burdensome chronic conditions in US medicine. Most people associate dialysis with a distant, elderly relative rather than themselves. Awareness is generally low among those without diabetes or hypertension -- the two conditions that account for nearly 70% of ESRD cases.\n","rough_estimate":"Most people without chronic disease would guess under 1% lifetime -- substantially below the actual 3-4%","kind":"intuition"},"native":{"display":"~130,000 new ESRD cases per year in the US","numerator":130000,"denominator":260000000,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.035,"display":"~1 in 29 lifetime (US adult average)","log_value":-1.46,"assumptions":"Albertus et al. (2016, Am J Kidney Dis) calculated lifetime risk of ESRD from birth using 2013 USRDS data, reporting 4.0% for males and 2.8% for females in 2013 (up from 3.5% and down from 3.0% respectively in 2000). Using 3.5% as a round central estimate consistent with the published range (2.0%--8.0% by race/sex subgroup) gives lifetime_us_adult = 0.035. This is a prevalence-based lifetime risk, not derived from simple annual-rate compounding, and directly reflects the Albertus 2016 published result rather than a naive calculation from the current annual incidence.\n","uncertainty":{"low":0.025,"high":0.05},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5123906/","title":"Risk of ESRD in the United States","publisher":"American Journal of Kidney Diseases (Albertus, Morgenstern, Robinson & Saran, 2016)","source_type":"peer_reviewed","statistic":"Lifetime risk of ESRD from birth: 3.1% non-Hispanic white men, 8.0% non-Hispanic Black men, 2.0% non-Hispanic white women, 6.8% non-Hispanic Black women (2013 USRDS data)","excerpt":"\"Among males, the lifetime risks of ESRD from birth using 2013 data were 3.1% (95% CI, 3.0%-3.1%) for non-Hispanic (NH) whites, 8.0% (95% CI, 7.9%-8.2%) for NH blacks... Among females, the lifetime risks were 2.0% (95% CI, 2.0%-2.1%) for NH whites, 6.8% (95% CI, 6.7%-6.9%) for NH blacks. The lifetime risk of ESRD from birth increased from 3.5% in 2000 to 4.0% in 2013 in males and decreased from 3.0% to 2.8% in females.\"\n","source_date":"2016-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250227122729/https://pmc.ncbi.nlm.nih.gov/articles/PMC5123906/","calculation_notes":"Published lifetime risk figures are used directly without recalculation. The population-weighted average across sex and race groups approximates 3.3--3.5%. Using 3.5% as the central estimate for lifetime_us_adult = 0.035. The racial disparity is substantial: non-Hispanic Black men face ~2.5x the ESRD risk of non-Hispanic white men (8.0% vs 3.1%), which is reflected in personal_factor_multipliers.\n","independence_note":"Albertus et al. (2016) used the USRDS (United States Renal Data System) registry, which is the authoritative national database for ESRD incidence and prevalence, maintained by NIDDK. This analysis is independent of insurance claims data and represents a frequently cited academic source for US lifetime ESRD risk.\n"},{"url":"https://www.niddk.nih.gov/health-information/health-statistics/kidney-disease","title":"Kidney Disease Statistics for the United States","publisher":"National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)","source_type":"govt_report","statistic":"~808,000 Americans living with ESRD; ~130,000 new ESRD cases per year; ~70% on dialysis, ~30% with functioning kidney transplant","excerpt":"\"In 2019, more than 808,000 Americans were living with ESRD. More than 131,000 people began treatment for kidney failure in 2019. About 70 percent of ESRD patients receive dialysis and about 30 percent have a functioning kidney transplant.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260520203448/https://www.niddk.nih.gov/health-information/health-statistics/kidney-disease","calculation_notes":"130,000 new cases per year / 260 million US adults = 0.0005 per adult per year. Over 59 years: 1 − (1 − 0.0005)^59 ≈ 0.028. This independent calculation from the annual incidence rate yields ~2.8%, broadly consistent with the Albertus et al. (2016) lifetime risk estimate of 3.5% (which uses a cohort method and is slightly higher because it captures risk from birth through the full lifespan). The Albertus figure is used as the primary estimate because it was specifically designed to measure lifetime risk.\n","independence_note":"NIDDK maintains the USRDS and provides national-level statistics independently of the Albertus et al. (2016) academic analysis. The two sources draw on the same underlying USRDS registry data but are independently published by a government agency and an academic research team respectively.\n"}],"comparison_anchors":[{"label":"Developing type 2 diabetes (lifetime, US)","lifetime_us_adult":0.4},{"label":"Dying from heart disease (lifetime, US)","lifetime_us_adult":0.2}],"personal_factor_multipliers":[{"factor":"Non-Hispanic Black vs Non-Hispanic white","multiplier":2.5,"notes":"Albertus 2016: 8.0% lifetime risk for non-Hispanic Black men vs 3.1% for non-Hispanic white men; disparity driven by hypertension, diabetes prevalence, and socioeconomic access to care"},{"factor":"Type 2 diabetes (leading cause of ESRD, ~40% of new cases)","multiplier":8,"notes":"Diabetic nephropathy causes ~40% of all new ESRD cases in the US; diabetes substantially elevates lifetime risk above the population average"},{"factor":"Hypertension (second leading cause, ~28% of new cases)","multiplier":4,"notes":"Hypertensive nephrosclerosis accounts for ~28% of ESRD incidence; poorly controlled hypertension is a major independent driver of CKD progression"},{"factor":"Chronic kidney disease stage 3+ (eGFR < 60)","multiplier":30,"notes":"Established CKD stage 3--5 dramatically elevates the probability of eventual ESRD relative to the general adult population average; many people with CKD 3 will progress to ESRD over a lifetime"},{"factor":"Regular high-dose NSAID use combined with diabetes or CKD","multiplier":2,"notes":"NSAIDs reduce renal blood flow and can accelerate CKD progression in those with pre-existing renal vulnerability; analgesic nephropathy is a preventable contributor to ESRD"}],"short_label":"End-stage kidney disease","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The 3.5% figure is a lifetime cohort risk calculated from the 2013 USRDS registry. ESRD rates have changed over time: the incidence rate peaked around 2001 and has modestly declined through improved diabetes and hypertension management, though the absolute number of people on dialysis continues to rise with an aging population. The racial disparity is large and well-documented: a Black American faces roughly 2.5 times the lifetime ESRD risk of a white American, driven by higher rates of hypertension and diabetes, differential access to primary care, and possible genetic factors (APOL1 gene variants). This entry does not cover acute kidney injury (AKI) from which most patients recover; it specifically covers kidney failure requiring dialysis or transplant for survival.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"Two simplified kidney shapes in muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/end-stage-renal-disease-lifetime","api_url":"https://likelier.app/api/fears/end-stage-renal-disease-lifetime.json"},{"slug":"vitamin-b12-deficiency-general-adult","question":"What are the odds a US adult is vitamin B12 deficient?","quick_answer":"About 3.6% of US adults are vitamin B12 deficient, with roughly 12.5% in the broader 'insufficient' range — well below the 15-25% many health-conscious people assume.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Vitamin B12 has an outsized presence in health anxiety relative to how common a deficiency actually is. Supplement aisles, energy-drink labels, and vegan-nutrition forums frame it as a widespread modern shortfall, and \"low B12\" is a popular self-diagnosis for ordinary tiredness. The result is a split intuition: the health-conscious and plant-based crowd tend to overestimate their odds, while the general public rarely registers that the real risk is concentrated in a few specific groups -- older adults with absorption problems, long-term metformin users, and strict vegans -- rather than spread evenly across everyone.\n","rough_estimate":"~15-25% of adults are deficient","kind":"intuition"},"native":{"display":"~3.6% of US adults ≥19 are vitamin B12 deficient (serum B12 <148 pmol/L); 4.4% among supplement non-users; 12.5% insufficient (NHANES 2007-2018)","numerator":3.6,"denominator":100,"unit":"cross-sectional prevalence of serum vitamin B12 deficiency (<148 pmol/L) among US adults","population":"US adults aged 19+ (NHANES 2007-2018)"},"normalized":{"lifetime_us_adult":0.036,"display":"~3.6% of US adults are vitamin B12 deficient (12.5% insufficient) — NHANES 2007-2018","log_value":-1.44,"assumptions":"Zhou et al. (AJCN 2023), pooling NHANES 2007-2018, found mean vitamin B12 deficiency (serum B12 <148 pmol/L) of 3.6% in US adults ≥19 — 4.4% among supplement non-users versus 2.3% among users — and 12.5% with insufficiency (<221 pmol/L). The headline is that cited cross-sectional figure, un-inflated (following the same discipline as the iron and magnesium entries): no modeled lifetime multiplier is applied, because a defensible cumulative figure is not separately measured. The wide uncertainty band instead spans the range that actually matters here: the low end reflects supplement users and the strict <148 pmol/L cutoff (~2.3%), the high end reflects the far larger \"insufficient\" bracket (~12.5%). Deficiency estimates are unusually sensitive to where the serum threshold is drawn, and serum B12 is itself an insensitive marker (see caveats), so this entry deliberately reports the conservative clinical-deficiency figure and treats insufficiency as the upper bound rather than the headline. Notably, unlike most nutrient gaps, age barely moves the strict-serum number (3.7% at ≥60 vs 3.6% overall) — the elevation in older adults shows up in functional markers, not serum B12.\n","uncertainty":{"low":0.023,"high":0.125},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10926004/","title":"Folate and vitamin B12 usual intake and biomarker status by intake source in United States adults aged ≥19 y: NHANES 2007-2018","publisher":"American Journal of Clinical Nutrition (Zhou, Wang, Yeung, Qi, Pfeiffer, et al.)","source_type":"peer_reviewed","statistic":"Vitamin B12 deficiency (serum B12 <148 pmol/L): 3.6% overall (4.4% supplement non-users, 2.3% users); insufficiency (<221 pmol/L) 12.5%; among adults ≥60, deficiency 3.7% and insufficiency 12.3%.","excerpt":"\"The mean vitamin B12 deficiency was 3.6% in the overall population and 4.4% and 2.3% between vitamin B12 supplement nonusers and users, respectively.\"\n","source_date":"2023-07-01","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20250202161922/https://pmc.ncbi.nlm.nih.gov/articles/PMC10926004/","calculation_notes":"NHANES 2007-2018 pooled (nationally representative US adults ≥19). Deficiency defined as serum B12 <148 pmol/L, insufficiency <221 pmol/L. The 3.6% overall deficiency figure is the native prevalence and the normalized headline (un-inflated). The 4.4% supplement-non-user figure is reported alongside because this cluster's framing is \"without supplements,\" and the 12.5% insufficiency figure and the ≥60 breakdown (3.7% / 12.3%) inform the uncertainty band and the caveats. No lifetime multiplier is applied.\n"},{"url":"https://ods.od.nih.gov/factsheets/VitaminB12-HealthProfessional/","title":"Vitamin B12 — Health Professional Fact Sheet","publisher":"NIH Office of Dietary Supplements","source_type":"govt_report","statistic":"Groups at higher risk of B12 deficiency: older adults, people with pernicious anemia or GI disorders, those with GI surgery, and people on vegetarian/vegan diets. Serum B12 <200-250 pg/mL is subnormal; methylmalonic acid (MMA) is the most sensitive marker.","excerpt":"\"Certain groups of people are more likely than others to be vitamin B12 deficient. These include older adults, people with pernicious anemia or gastrointestinal disorders, and those who have had gastrointestinal surgery. Other groups that are at risk of vitamin B12 deficiency include people who follow vegetarian diets and the infants of women who follow vegan diets. [...] The most sensitive marker of vitamin B12 status is a vitamin B12-associated metabolite called methylmalonic acid.\"\n","source_date":"2025-07-02","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20260726092742/https://ods.od.nih.gov/factsheets/vitaminb12-HealthProfessional/","calculation_notes":"The NIH ODS fact sheet grounds the risk-group story that the whole entry turns on: clinical deficiency is concentrated in older adults with malabsorption, people with pernicious anemia / GI disease / GI surgery, and strict plant-based eaters — not spread evenly. It also supplies the measurement caveat: serum B12 is an insensitive marker (subnormal <200-250 pg/mL), and MMA is more sensitive, which is why serum-based prevalence understates functional deficiency in older adults. Grounds the vegan and malabsorption multipliers (directionally) and the caveats.\n","independence_note":"NIH ODS is an independent government reference; it does not derive the headline prevalence, which comes from the peer-reviewed NHANES analysis above."},{"url":"https://academic.oup.com/jcem/article/101/4/1754/2804585","title":"Long-term Metformin Use and Vitamin B12 Deficiency in the Diabetes Prevention Program Outcomes Study (DPPOS)","publisher":"Journal of Clinical Endocrinology & Metabolism (Aroda et al.)","source_type":"peer_reviewed","statistic":"At 13 years, low B12 (≤203 pg/mL) 7.4% in metformin vs 5.4% placebo; combined low/borderline-low (≤298 pg/mL) 20.3% vs 15.6%. At 5 years, definite low B12 4.3% vs 2.3%.","excerpt":"\"Low B12 (≤ 203 pg/mL) occurred more often in MET than PLA at 5 years (4.3 vs 2.3%; P = .02) but not at 13 years (7.4 vs 5.4%; P = .12). Combined low and borderline-low B12 (≤ 298 pg/mL) was more common in MET at 5 years (19.1 vs 9.5%; P < .01) and 13 years (20.3 vs 15.6%; P = .02).\"\n","source_date":"2016-04-01","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20251222120010/https://academic.oup.com/jcem/article/101/4/1754/2804585","calculation_notes":"DPPOS randomized long-term data. Grounds the metformin multiplier: definite B12 deficiency ran ~1.9x higher on metformin at 5 years (4.3% vs 2.3%); the combined low/borderline measure roughly doubled at 5 years (19.1% vs 9.5%) and stayed elevated at 13 years (20.3% vs 15.6%). Does not feed the headline arithmetic — it supports only the metformin personal_factor_multiplier.\n","independence_note":"Independent randomized-cohort source, separate from the NHANES cross-sectional prevalence and the ODS reference."}],"comparison_anchors":[{"label":"Vitamin D deficiency without supplements (lifetime)","lifetime_us_adult":0.35},{"label":"Iron deficiency, women 12-49 (point prevalence)","lifetime_us_adult":0.255},{"label":"Depression (lifetime, US adult)","lifetime_us_adult":0.2}],"personal_factor_multipliers":[{"factor":"long-term metformin use","multiplier":1.9,"notes":"DPPOS (Aroda 2016): definite low B12 (≤203 pg/mL) 4.3% on metformin vs 2.3% on placebo at 5 years (~1.9x); combined low/borderline roughly doubled (19.1% vs 9.5%). Metformin impairs ileal B12 absorption; risk rises with dose and duration."},{"factor":"strict vegan diet (no fortified foods or supplement)","multiplier":3,"notes":"Directional estimate — magnitude is highly cutoff-dependent. NIH ODS names people on vegetarian/vegan diets among at-risk groups because bioavailable B12 occurs essentially only in animal foods; observational vegan cohorts show markedly higher deficiency, but reported rates swing widely with the serum threshold used. Fortified foods or a supplement remove most of this excess risk."},{"factor":"malabsorption (pernicious anemia, atrophic gastritis, GI surgery)","multiplier":4,"notes":"Directional estimate. NIH ODS lists pernicious anemia, gastrointestinal disorders, and prior GI surgery as leading causes of B12 deficiency — these impair intrinsic-factor-dependent absorption and are the dominant drivers of clinical deficiency, largely independent of dietary intake. Prevalence rises with age via atrophic gastritis."}],"short_label":"B12 deficiency","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This is a general-US-adult figure, and the honest reading is that clinical B12 deficiency is uncommon at the population level (~3.6%) but genuinely concentrated in specific groups — which is exactly why a single headline number can mislead in both directions. The bigger measurement caveat is the cutoff. \"Deficiency\" here means serum B12 <148 pmol/L; loosen the threshold toward the \"insufficient\" line (<221 pmol/L) and the figure more than triples to ~12.5%, and clinicians often treat values up to ~250 pg/mL as subnormal. Serum B12 is also an insensitive marker: NIH ODS notes that methylmalonic acid (MMA) is more sensitive, and functional-marker studies find substantially higher deficiency in older adults than serum B12 alone suggests — so the strict-serum 3.7% for the over-60s understates the real burden in that group. Deficiency also matters more than its low rate implies: untreated, it can cause megaloblastic anemia and irreversible neurological damage, so the low base rate is not a reason to dismiss it in someone with symptoms or a risk factor. Finally, the perceived-fear split is real: strict vegans who don't supplement or eat fortified foods are legitimately at elevated risk, even though the average adult is not.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-07-18","methodology_version":"1.0"},"reviewer":"8d-eval-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-07-18","image":{"alt":"A single small capsule beside a slice of fortified bread on a plain plate, muted red and grey flat vector illustration."},"canonical_url":"https://likelier.app/vitamin-b12-deficiency-general-adult","api_url":"https://likelier.app/api/fears/vitamin-b12-deficiency-general-adult.json"},{"slug":"child-car-crash-injury","question":"What are the odds of a child being injured in a car crash?","quick_answer":"About 1 in 27 cumulative through age 14 for a US child riding in vehicles — roughly 147,000 of the nation's 59.5 million children are injured as vehicle occupants each year (a figure derived from NHTSA person-type counts), about 1 in 400 annually. Restraint status is the dominant lever: unrestrained children carry 3.2 times the per-crash injury odds of appropriately restrained ones.\n","category":"transport","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Parental fear about cars is death-framed and event-framed — the drunk driver, the catastrophic highway crash — while the ordinary injury risk of ordinary rides barely registers as a category. The extensive attention parents give to car-seat brands and installation details coexists with little sense of the base rate those seats are working against; almost no parent could say whether a child's odds of ever being hurt in a crash are one in a thousand or one in twenty. The measured answer, about one in twenty-seven through age 14, sits closer to the second guess than most would expect, even though the large majority of those injuries are minor.\n","rough_estimate":"most parents could not place the odds within an order of magnitude; typical guesses run far below the measured ~1 in 27","kind":"intuition"},"native":{"display":"~147,000 children (14 and younger) injured as vehicle occupants per year (US, 2023, derived)","numerator":147284,"denominator":59498481,"unit":"per year","population":"US children 14 and younger"},"normalized":{"lifetime_us_adult":0.0365,"display":"~1 in 27 through age 14 (US child, vehicle occupant)","log_value":-1.438,"assumptions":"NHTSA's Children fact sheet reports an estimated 161,478 children 14 and younger injured in traffic crashes in 2023, of whom 7,081 were injured as pedestrians and 7,113 as pedalcyclists. Subtracting the two non-occupant person types leaves 147,284 — children injured while riding in vehicles, plus a small unattributed remainder of other non-occupants (skateboarders, scooter riders and similar), which is why the numerator is labeled derived. Against the same fact sheet's child population of 59,498,481 the annual rate is 147,284 / 59,498,481 ≈ 0.248% per year, about 1 in 400. Compounding over the 15 years from birth through age 14: 1 − (1 − 0.00248)^15 ≈ 0.0365, about 1 in 27. Using the full 161,478 all-person-type figure instead gives ≈ 0.040 (1 in 25), the upper end of the band; unreported injuries push higher still, while CRSS sampling error and repeat victims pull the distinct-child probability down.\n","uncertainty":{"low":0.03,"high":0.048},"scope":"subgroup_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813712","title":"Traffic Safety Facts 2023 Data: Children (DOT HS 813 712)","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"2023: estimated 161,478 children (14 and younger) injured in traffic crashes (442/day); 7,081 injured as pedestrians and 7,113 as pedalcyclists; child population 59,498,481; 1,019 child traffic deaths of whom 770 (76%) were occupants; 25% of child deaths in alcohol-impaired-driving crashes","excerpt":"\"For the purpose of this fact sheet, children are defined as 14 years old and younger. [...] There were 59.5 million children in the United States, 18 percent of the total U.S. population. [...] An estimated 161,478 children were injured in traffic crashes in 2023, a 3-percent increase from 156,502 in 2022. An average of 3 children were killed and an estimated 442 children were injured every day in traffic crashes in 2023. [...] Figure 1 displays the distribution of the 1,019 child traffic fatalities in 2023—76 percent (770) were occupants and 24 percent (249) were nonoccupants (pedestrians, pedalcyclists, or other nonoccupants). [...] Of the estimated 68,244 pedestrians injured in traffic crashes, 10 percent (7,081) were children. [...] Of the estimated 49,989 pedalcyclists injured in traffic crashes, 14 percent (7,113) were children.\"\n","source_date":"2025-04-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260508202846/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813712","calculation_notes":"Occupant-side numerator derived as 161,478 − 7,081 pedestrians − 7,113 pedalcyclists = 147,284. NHTSA does not publish a single occupant-only child injury count in this fact sheet; the remainder after removing the two named non-occupant person types is dominated by vehicle occupants but includes an unquantified sliver of other non-occupants. The child population denominator (59,498,481) is the fact sheet's own Table 7 US total, sourced to the Census Bureau. Note the trap this entry is built to avoid: child traffic totals mix occupants with pedestrians and cyclists, and using the all-traffic figure for an in-the-car question overstates the occupant risk by about 10 percent.\n","independence_note":"Injury estimates come from NHTSA's Crash Report Sampling System (CRSS), a probability sample of police-reported crashes; fatality counts come from the FARS census; the population denominator is Census Bureau data.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15741356/","title":"Effects of seating position and appropriate restraint use on the risk of injury to children in motor vehicle crashes","publisher":"Durbin DR, Chen I, Smith R, Elliott MR, Winston FK — Pediatrics 115(3):e305-9","source_type":"peer_reviewed","statistic":"In a probability sample of 17,980 crash-involved children under 16 (representing 229,106 children), unrestrained children had OR 3.2 (95% CI 2.5-4.1) for injury vs appropriately restrained; inappropriately restrained OR 1.8 (1.4-2.3); front-seat OR 1.4 (1.2-1.7) vs rear","excerpt":"\"Inappropriately restrained children were at nearly twice the risk of injury, compared with appropriately restrained children (odds ratio [OR]: 1.8; 95% confidence interval [CI]: 1.4-2.3), whereas unrestrained children were at >3 times the risk (OR: 3.2; 95% CI: 2.5-4.1). [...] children in the front seat were at 40% greater risk of injury, compared with children in the rear seat (OR: 1.4; 95% CI: 1.2-1.7). [...] Age-appropriate restraint confers relatively more safety benefit than rear seating, but the 2 work synergistically.\"\n","source_date":"2005-03-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260114044657/https://pubmed.ncbi.nlm.nih.gov/15741356/","calculation_notes":"Source of all three restraint/seating multipliers. Injury basis (clinically significant injury, parent-reported on a validated instrument), per crash event, among crash-involved children under 16 in insured vehicles across 15 states, December 1998 - November 2002. Among crash-involved children in the sample, 1.6% had serious injuries and 13.5% minor injuries.\n","independence_note":"Insurance-claims-based probability sample (State Farm / Partners for Child Passenger Safety), a wholly different data system from NHTSA's CRSS and FARS.\n"},{"url":"https://www.iihs.org/research-areas/fatality-statistics/detail/teenagers","title":"Fatality Facts 2022: Teenagers","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"Per mile driven, the fatal crash rate for drivers 16-19 is just over 3 times the rate for drivers 20 and older","excerpt":"\"The fatal crash rate per miles driven for 16-19 year-olds is just over 3 times the rate for drivers ages 20 and over.\"\n","source_date":"2024-05-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260724014726/https://www.iihs.org/research-areas/fatality-statistics/detail/teenagers","calculation_notes":"Basis for the teen-driver multiplier: a child passenger's crash exposure scales with the per-mile crash rate of whoever is driving. Fatal-crash basis used as a conservative proxy for all-severity crash involvement.\n","independence_note":"IIHS rates derive from FARS counts and federal mileage surveys, independent of the CRSS injury estimates and the insurance-claims Durbin sample.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Unrestrained (no car seat, booster, or belt)","multiplier":3.2,"notes":"Unrestrained crash-involved children had 3.2× (95% CI 2.5-4.1) the odds of clinically significant injury vs appropriately restrained children (Durbin 2005, Pediatrics). Per crash event, vs the appropriately-restrained baseline."},{"factor":"Riding with a driver aged 16-19","multiplier":3,"notes":"Per mile driven, 16-19-year-old drivers have just over 3× the fatal crash rate of drivers 20+ (IIHS, 2022 data). Fatal-crash-rate basis used as a proxy for crash exposure; applies to passengers of teen drivers."},{"factor":"Inappropriately restrained for age/size (adult belt too early, unused harness)","multiplier":1.8,"notes":"Inappropriately restrained children had 1.8× (95% CI 1.4-2.3) the injury odds of appropriately restrained children (Durbin 2005). Same source and basis as the unrestrained figure; premature graduation to adult belts is the most common form."},{"factor":"Front-row seating (under 16)","multiplier":1.4,"notes":"Children in the front seat had 1.4× (95% CI 1.2-1.7) the injury odds of children in the rear (Durbin 2005). Independent of restraint status; the study notes restraint confers relatively more benefit than seating position, and the two work synergistically."}],"short_label":"Child crash injury","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The numerator is a derivation, not a published NHTSA line item: the fact sheet's 161,478 all-traffic child injuries minus its named pedestrian and pedalcyclist counts, leaving mostly-but-not-only vehicle occupants. The large majority of counted injuries are minor; in the Durbin crash-involved sample only 1.6% of children had serious injuries. Police-reported CRSS estimates miss unreported injuries and carry sampling error. Compounding a single-year rate over 15 childhood years assumes 2023 rates hold and years are independent; a child injured twice counts once against the distinct-child probability. Risk is not uniform across childhood or households — restraint use, who is driving, and impairment cluster: 25% of the 1,019 children killed in 2023 died in alcohol-impaired-driving crashes, and unrestrained drivers disproportionately carry unrestrained children. The fatality-side companion levers are covered in the restraint entries.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-07-26"},"reviewer":"8d-eval-2026-07-26","last_reviewed":"2026-07-26","reviewed":true,"generated_at":"2026-07-26","image":{"alt":"Flat-vector rear car seat with an empty child booster and a small bandaged teddy bear beside it, muted teal and grey palette."},"canonical_url":"https://likelier.app/child-car-crash-injury","api_url":"https://likelier.app/api/fears/child-car-crash-injury.json"},{"slug":"drowsy-driving-fatal-crash","question":"What are the odds of causing a fatal crash by driving without enough sleep?","quick_answer":"For a driver who regularly operates on too little sleep (roughly monthly), the lifetime odds of causing a fatal crash are about 1 in 26, around four times the rested-driver rate. Most drivers believe they can tell when they are too tired to drive.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Drowsy driving is one of the most consistently underrated risks on the road, not because people deny that fatigue is dangerous, but because they trust their ability to manage it. Most drivers believe they can sense when they are too tired to drive — that the eyelid heaviness, the head nod, or a yawn will reliably tell them to pull over. Survey data and naturalistic dashcam studies show the opposite: the moments before a drowsy crash are usually not recognised by the driver, and the most common subjective state preceding a fatigue-related crash is feeling \"fine, just a bit tired\". A meaningful minority of drivers also believe that coffee, loud music, or rolling the window down restore alertness for more than the few minutes they actually do.\n","rough_estimate":"most drivers believe they can tell when they are too tired to drive","kind":"intuition"},"native":{"display":"~8 per 100,000 trips result in a fatal crash for a driver who is moderately sleep-deprived (≈4× the rested-driver rate)","numerator":8,"denominator":100000,"unit":"per fatigue-impaired trip (fatal crash involvement)","population":"US adult driver after 4–5 hours of sleep, fatal-crash involvement rate derived from Tefft 2018 culpable-crash odds applied to NHTSA per-trip baseline"},"normalized":{"lifetime_us_adult":0.038,"display":"~1 in 26 lifetime (driver who regularly operates on too little sleep, ~monthly)","log_value":-1.42,"assumptions":"The US population-average per-trip fatal-crash probability for a rested sober driver is approximately 1 in 50,000 (NHTSA per-trip baseline used in the driving-at-0.1pct-bac entry). Tefft 2018 (Sleep journal, peer-reviewed case-control derived from NMVCCS) found that drivers who slept 4 hours in the 24 hours before driving had 2.9× (95% CI 1.4–6.2) the odds of being culpable for their crash compared with drivers who slept 7–9 hours; less than 4 hours rose to 15.1× (95% CI 4.2–54.4) culpability odds. Applying a 4× per-trip multiplier to the rested baseline — nudged above the 2.9× culpability figure to incorporate the tendency of fatal crashes to cluster at the more impaired end — gives ~1 in 12,500 per moderately drowsy trip. For a driver who operates at this level of sleep deprivation roughly once a month — a common pattern for shift workers, new parents, and people routinely driving home from long days — 12 trips per year over 40 driving years equals ~480 impaired trips. The cumulative probability of being involved in at least one fatal crash is 1 − (1 − 1/12500)^480 ≈ 0.038, or roughly 1 in 26. The low end of the uncertainty band assumes rare drowsy trips (~6/year) with only moderate fatigue; the high end assumes monthly severe sleep deprivation (<4 hours sleep), which Tefft's 15.1× multiplier pushes toward ~1 in 9 lifetime.\n","uncertainty":{"low":0.018,"high":0.11},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://aaafoundation.org/drowsy-driving-in-fatal-crashes-united-states-2017-2021/","title":"Drowsy Driving in Fatal Crashes, United States, 2017–2021","publisher":"AAA Foundation for Traffic Safety","source_type":"reputable_reference","statistic":"An estimated 17.6% of all fatal crashes in the United States from 2017 through 2021 involved a drowsy driver — approximately 29,834 fatalities over the five-year period (~6,000 per year), roughly ten times the number identified by police-reported FARS coding alone.\n","excerpt":"\"An estimated 17.6% of all fatal crashes in the years 2017–2021 involved a drowsy driver.\"\n","source_date":"2024-03-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20260318193759/https://aaafoundation.org/drowsy-driving-in-fatal-crashes-united-states-2017-2021/","calculation_notes":"AAA Foundation for Traffic Safety 2024 report applied a model trained on NHTSA's Crash Investigation Sampling System (CISS), in which trained investigators reach detailed determinations of crash causation, to the FARS census of fatal crashes. The 17.6% figure (≈6,000 deaths/year over five years, ≈29,834 total) is the population-level anchor for the gap between police-reported and true drowsy-crash prevalence. Used here to establish that drowsy driving is roughly comparable in magnitude to alcohol-impaired driving (which kills ~12,000/year per NHTSA FARS).\n"},{"url":"https://academic.oup.com/sleep/article/41/10/zsy144/5067408","title":"Acute sleep deprivation and culpable motor vehicle crash involvement","publisher":"Tefft, B.C. — Sleep (Oxford Academic)","source_type":"peer_reviewed","statistic":"Compared with drivers who slept 7–9 hours in the 24 hours before crashing, adjusted odds of culpable crash involvement were 1.3× for 6–7 hours, 1.9× for 5–6 hours, 2.9× for 4–5 hours, and 15.1× (95% CI 4.2–54.4) for less than 4 hours.\n","excerpt":"\"Drivers who reported having slept for 4 hours and less than 4 hours in the 24 hours before crashing had 2.9 (95% CI = 1.4 to 6.2) and 15.1 (95% CI = 4.2 to 54.4) times the odds, respectively, of having been culpable for their crashes, compared with drivers who reported 7–9 hours of sleep.\"\n","source_date":"2018-09-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20241205105531/https://academic.oup.com/sleep/article/41/10/zsy144/5067408","calculation_notes":"Tefft 2018 is the peer-reviewed publication in Sleep of analysis drawn from the NHTSA National Motor Vehicle Crash Causation Survey (NMVCCS), a nationally representative sample of 5,470 crashes. The case-control design uses non-culpable drivers as the control arm, isolating the effect of sleep deprivation from time-of-day and exposure confounds. The 2.9× and 15.1× odds ratios for the <5-hour buckets are the multipliers used in the lifetime calculation: a conservative ~4× for \"moderately drowsy\" (4–5h sleep) anchors the headline ~1 in 26 estimate; the 15× high-end multiplier drives the upper uncertainty bound.\n"},{"url":"https://www.nhtsa.gov/risky-driving/drowsy-driving","title":"Drowsy Driving: Avoid Falling Asleep Behind the Wheel","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"NHTSA estimates that in 2017, 91,000 police-reported crashes involved drowsy drivers, resulting in approximately 50,000 injuries and nearly 800 deaths — figures NHTSA explicitly describes as an underestimate.\n","excerpt":"\"NHTSA estimates that in 2017, 91,000 police-reported crashes involved drowsy drivers. These crashes led to an estimated 50,000 people injured and nearly 800 deaths.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20260523235956/https://www.nhtsa.gov/risky-driving/drowsy-driving","calculation_notes":"The NHTSA FARS-based figure (~800 deaths/year) is the police-reported lower bound. NHTSA itself, the AAA Foundation, and the National Sleep Foundation all agree this number undercounts the true total by at least 8–10×, primarily because police rarely record drowsiness as a crash factor unless the driver self-reports falling asleep. The gap between this number and the AAA 2024 estimate (~6,000 deaths/year) is the under-reporting factor for drowsy driving.\n"},{"url":"https://www.nature.com/articles/40775","title":"Fatigue, alcohol and performance impairment","publisher":"Dawson, D. & Reid, K. — Nature 388:235","source_type":"peer_reviewed","statistic":"Hand-eye coordination performance after 17 hours of sustained wakefulness is equivalent to that observed at a blood alcohol concentration of approximately 0.05%; after ~24 hours of wakefulness, equivalent to roughly 0.10% BAC.\n","excerpt":"\"After 17 hours of sustained wakefulness cognitive psychomotor performance decreased to a level equivalent to the performance impairment observed at a blood alcohol concentration of 0.05%.\"\n","source_date":"1997-07-17","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20260418072749/https://www.nature.com/articles/40775","calculation_notes":"Dawson & Reid 1997 is the canonical wakefulness-to-BAC equivalence reference (n=40, within-subjects counterbalanced design). Used here to anchor the intuition that \"very tired\" is not a lay descriptor but a quantifiable impairment state. The 17-hour and 24-hour thresholds are cited verbatim by NHTSA, CDC, and the National Sleep Foundation as the basis for their public-health framing of drowsy driving as comparable to alcohol-impaired driving.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Causing a fatal crash while driving at 0.10% BAC (~monthly, lifetime)","lifetime_us_adult":0.062},{"label":"Causing a fatal crash while texting (regular texter, lifetime)","lifetime_us_adult":0.018}],"personal_factor_multipliers":[{"factor":"drives drowsy only a few times per year","multiplier":0.25,"notes":"Rare fatigue-impaired trips compress cumulative exposure sharply."},{"factor":"drives after 4–5 hours of sleep, ~monthly","multiplier":1,"notes":"Baseline assumption for the headline lifetime estimate (Tefft 2018 ~3× per-trip odds, used at ~4× to incorporate fatal-crash skew)."},{"factor":"drives after <4 hours of sleep, ~monthly","multiplier":3,"notes":"Tefft 2018 found 15.1× culpable-crash odds at <4h sleep — over 5× the 4–5h bucket's 2.9×."},{"factor":"shift worker driving home post-night-shift","multiplier":4,"notes":"Combines acute sleep deprivation with circadian nadir; AAA Foundation cites night-shift drivers among the highest-risk groups."},{"factor":"long-distance trip after >17 hours awake","multiplier":3,"notes":"Per Dawson & Reid 1997, equivalent to driving at ~0.05% BAC; sustained highway driving compounds inattention risk."}],"short_label":"Drowsy driving","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The lifetime estimate is highly sensitive to two assumptions: how often the driver is significantly sleep-deprived, and how the per-trip multiplier from Tefft's all-severity culpable-crash analysis translates to fatal-crash risk specifically. Fatal crashes likely cluster at the more severely impaired end of the distribution, so the headline figure may understate the high-frequency severe-sleep-loss case and overstate the rare moderate case. Drowsiness is also a risk factor that operates differently from BAC: a driver at a steady 0.10% BAC is impaired for the duration of the trip, but a sleep-deprived driver's risk spikes during the late-trip window when microsleeps become more frequent. The 17.6% figure for fatal-crash prevalence (AAA 2024) and the 1.8% police-reported figure (NHTSA FARS) bracket the true population share; the order-of-magnitude gap reflects the difficulty of attributing a crash to drowsiness post-hoc when the driver may not survive to report and no objective biomarker exists. None of this accounts for the interaction with alcohol or sedating medication, which compounds impairment far beyond the additive sum of the components. The AAA Foundation's 2017–2021 analysis also reports that the drowsy share of fatal-crash-involved drivers was greatest among those aged 16–20, that men were over-represented, and that drowsiness peaked among crashes between 3:00 and 6:59 AM and on rural collector and local roads — but these are population proportions among fatal-crash drivers, not per-driver relative-risk multipliers, so they describe who shows up in drowsy fatal crashes rather than scaling the individual lifetime estimate above.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"fears-2026-05"},"reviewer":"8d-eval-2026-05-25","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-25","image":{"alt":"A muted flat vector illustration of a single car steering wheel with a small clock indicator showing late-night hours on a pale background."},"canonical_url":"https://likelier.app/drowsy-driving-fatal-crash","api_url":"https://likelier.app/api/fears/drowsy-driving-fatal-crash.json"},{"slug":"street-robbery","question":"What are the odds of being mugged or robbed on the street?","quick_answer":"The lifetime odds of being robbed on the street are about 1 in 26 — lower than the roughly 1 in 3 adults who worry it could happen to them.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Gallup's 2025 crime-worry poll finds 29% of US adults worry frequently or occasionally about being mugged, placing it in the middle tier of crime fears — below identity theft (69%) and car theft (39%), but above being murdered (22%) or sexually assaulted (21%). The worry level has declined modestly over the past decade even as robbery rates have fallen more sharply than public perception suggests.\n","rough_estimate":"~1 in 3 adults feel it could happen to them","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"native":{"display":"~66.5 per 100,000 population per year (FBI, 2023)","numerator":665,"denominator":1000000,"unit":"per year","population":"US population (FBI UCR/NIBRS reported robberies, 2023)"},"normalized":{"lifetime_us_adult":0.038,"display":"~1 in 26 lifetime","log_value":-1.42,"assumptions":"The FBI reports a 2023 robbery rate of 66.5 per 100,000 population, which reflects incidents reported to law enforcement. This translates to roughly 223,000 reported robberies annually (66.5 × 3.35 million hundreds of population). However, the NCVS consistently shows that only about 42-64% of robberies are reported to police (42% in 2023, 64% in 2022). Using a conservative 50% reporting rate implies roughly 446,000 actual robbery victimizations per year, or about 133 per 100,000 (~0.00133 annual probability). But this includes all robbery types (commercial, home invasion, vehicle). Street/personal robbery accounts for roughly half of all robberies per NCVS location data. For total robbery exposure (including all types), compounding over 59 years: 1 − (1 − 0.000665)^59 ≈ 0.038 using the FBI reported rate. The NCVS-adjusted figure would be higher (~0.076), but we use the more conservative FBI-reported figure as the headline because it is more precisely measured. The uncertainty range captures the reporting-adjustment gap.\n","uncertainty":{"low":0.025,"high":0.09},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fbi.gov/news/press-releases/fbi-releases-2023-crime-in-the-nation-statistics","title":"FBI Releases 2023 Crime in the Nation Statistics","publisher":"Federal Bureau of Investigation","source_type":"govt_report","statistic":"Robbery rate of 66.5 per 100,000 population in 2023; robberies decreased an estimated 0.3%","excerpt":"\"The FBI released detailed data on over 14 million criminal offenses for 2023 reported to the Uniform Crime Reporting (UCR) Program by participating law enforcement agencies. More than 16,000 state, county, city, university and college, and tribal agencies, covering a combined population of 94.3% of inhabitants, submitted data through NIBRS and the Summary Reporting System.\"\n","source_date":"2024-09-23","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260324222936/https://www.fbi.gov/news/press-releases/fbi-releases-2023-crime-in-the-nation-statistics","calculation_notes":"FBI UCR/NIBRS estimates a 2023 robbery rate of 66.5 per 100,000 population (Statista, citing FBI data). With a US population of ~335 million: 66.5 × 3,350 ≈ 222,775 reported robberies. Annual per-person probability: 0.000665. Lifetime over 59 years: 1 − (1 − 0.000665)^59 ≈ 0.038. This is the reported-crime figure only; NCVS data shows substantial underreporting.\n"},{"url":"https://bjs.ojp.gov/press-release/criminal-victimization-2023","title":"Criminal Victimization, 2023","publisher":"US Bureau of Justice Statistics","source_type":"govt_report","statistic":"42% of robbery victimizations in 2023 were reported to police, down from 64% in 2022","excerpt":"\"A smaller percentage of robbery victimizations that occurred in 2023 (42%) than in 2022 (64%) were reported to police.\"\n","source_date":"2024-09-12","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260504060832/https://bjs.ojp.gov/press-release/criminal-victimization-2023","calculation_notes":"The NCVS captures both reported and unreported victimizations via household interviews. The 42% reporting rate for 2023 implies that for every robbery known to police, roughly 1.4 additional robberies go unreported. If ~223,000 reported robberies represent 42% of all robberies, total victimizations ≈ 530,000. At 530,000 / 335M population, annual probability ≈ 0.00158. Lifetime: 1 − (1 − 0.00158)^59 ≈ 0.089. This is the upper end of our uncertainty range. Black Americans experienced a 79% increase in robbery victimization from 2022 to 2023 per NCVS, and were more than twice as likely to be robbed as white Americans.\n","independence_note":"BJS NCVS is a household survey conducted independently of the FBI UCR/NIBRS law-enforcement reporting system. The two count robberies through entirely different pipelines — victim recall vs police reports.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"29% of US adults worry frequently or occasionally about being mugged (2025)","excerpt":"\"Fewer Americans say they worry about crimes, such as having a car stolen (39%) or their home burglarized (34%), being a victim of a hate crime (30%), or getting mugged (29%), attacked while driving (27%), murdered (22%) or sexually assaulted (21%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260421194340/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for perceived-risk axis only. The 29% figure is the share of respondents reporting frequent-or-occasional worry about being mugged. This is well above the 3.8% lifetime probability from reported robberies, suggesting substantial overestimation of personal risk — consistent with the availability heuristic driven by news coverage of muggings.\n","independence_note":"Gallup telephone survey, independent of both BJS NCVS and FBI UCR. Measures worry, not incidence.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"High-crime urban metro (Detroit, Memphis, St. Louis)","probability":0.1,"notes":"Top-decile metro robbery rates are 3-4x the national average"},{"region":"Suburban / low-crime metro","probability":0.02,"notes":"Suburban robbery rates roughly half the national average per NCVS"},{"region":"Rural area","probability":0.01,"notes":"Rural robbery rates are substantially lower; NCVS urban rate is roughly 3x rural"}],"personal_factor_multipliers":[{"factor":"Urban resident","multiplier":2.5,"notes":"NCVS 2023 urban violent victimization rate rose to 34.0 per 1,000, well above suburban and rural rates"},{"factor":"Male aged 18-24","multiplier":2,"notes":"Young men are disproportionately targeted for street robbery per NCVS data"},{"factor":"Black Americans","multiplier":2.5,"notes":"NCVS 2023 shows Black Americans more than twice as likely to be robbed as white Americans"},{"factor":"Rural resident","multiplier":0.3,"notes":"Rural robbery rates are a fraction of urban rates"}],"short_label":"Street robbery / mugging","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The headline figure of ~1 in 26 uses the FBI-reported robbery rate, which captures only incidents reported to law enforcement. The NCVS, which includes unreported robberies, suggests the true rate is roughly double — implying a lifetime probability closer to 1 in 13. The wide uncertainty range (2.5% to 9%) reflects this measurement gap. \"Robbery\" in FBI/NCVS data includes all types — street mugging, commercial robbery, carjacking, and home-invasion robbery. Pure street robbery (the mental image most people have when asked about \"being mugged\") accounts for roughly half of all robberies, so the street-specific figure is lower. Geography is the dominant risk factor: county-level robbery rates span more than an order of magnitude, and urban residents face dramatically higher exposure. The long-term trend is strongly downward — FBI robbery rates have fallen roughly 60% since the early 1990s — but the 2022-2023 period saw a modest uptick per NCVS data. This entry is distinct from home burglary, which is a property crime without the face-to-face confrontation element. A news-coined sub-mode known as \"jugging\" — where offenders surveil a bank or ATM, identify a customer who has withdrawn a large sum of cash, and then confront or follow them to rob them elsewhere — has drawn FBI field-office advisories (2024–2025) and police warnings across Texas, the Carolinas, Georgia, Colorado, Maryland and Florida, plus a dedicated state statute enacted in Texas in 2025. These reports are anecdotal case counts and advisories, not surveillance-based incidence data: no agency publishes a per-withdrawal or per-adult base rate for jugging, so it cannot be quantified separately here and is folded into the general robbery figure above.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An empty wallet lying open on a sidewalk, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/street-robbery","api_url":"https://likelier.app/api/fears/street-robbery.json"},{"slug":"low-fiber-colorectal-cancer","question":"What are the odds of getting colorectal cancer from not eating enough fiber?","quick_answer":"Lifetime colorectal cancer risk is about 1 in 26 for a US adult, and higher fiber intake lowers it — roughly 10% per extra 10 g/day. Few people register low fiber as a meaningful cancer risk factor.\n","category":"cancer","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Dietary fiber's connection to colorectal cancer lives in a strange middle ground of public awareness. Most adults have a vague sense that fiber is \"good for digestion,\" but few could name colorectal cancer as a specific outcome of low intake. The Burkitt hypothesis — that high-fiber African diets explained low colorectal cancer rates — dates to the 1970s and was widely discussed in the nutrition literature, but it never achieved the cultural salience of, say, the fat-heart disease link. Fiber remains underrated as a cancer-prevention factor relative to the strength of the epidemiological evidence behind it.\n","rough_estimate":"Most adults do not consider low fiber intake a meaningful cancer risk factor","kind":"intuition"},"native":{"display":"~10% risk reduction per 10 g/day increase in dietary fiber intake","numerator":1,"denominator":26,"unit":"lifetime","population":"US adults, colorectal cancer incidence"},"normalized":{"lifetime_us_adult":0.039,"display":"~1 in 26 lifetime (US adult, colorectal cancer incidence)","log_value":-1.409,"assumptions":"Uses the SEER lifetime incidence of colorectal cancer in the US: approximately 3.9% for men and women combined (~1 in 26), per SEER Cancer Stat Facts (2021-2023 data). This is whole-population baseline CRC incidence, not a fiber-attributable fraction — the question is framed around fiber as a modifiable factor, and the fiber effect is carried by the dose-response figures below and the personal_factor_multipliers, not by inflating the baseline. The WCRF/AICR Continuous Update Project and the Aune et al. 2011 dose-response meta-analysis both found a summary relative risk of 0.90 (95% CI 0.86-0.94) per 10 g/day increase in total dietary fiber. The average US adult consumes ~16 g/day of fiber versus the recommended 25-30 g/day, meaning most Americans are in the low-fiber range. Bingham et al. 2003 (EPIC, n=519,978) found that populations with low fiber intake could reduce CRC risk by ~40% by doubling their intake. Uncertainty reflects the difficulty of isolating fiber from other dietary components and modest year-to-year variation in the SEER estimate.\n","uncertainty":{"low":0.035,"high":0.055},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/12737858/","title":"Dietary fibre in food and protection against colorectal cancer in the European Prospective Investigation into Cancer and Nutrition (EPIC): an observational study","publisher":"The Lancet / Bingham SA, Day NE, Luben R, et al.","source_type":"primary_study","statistic":"In EPIC cohort (n=519,978), highest quintile of fiber intake had ~40% lower CRC risk than lowest quintile; doubling fiber intake from low levels could reduce CRC risk by 40%","excerpt":"\"In populations with low average intake of dietary fibre, an approximate doubling of total fibre intake from foods could reduce the risk of colorectal cancer by 40%.\"\n","source_date":"2003-05-03","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260505060628/https://pubmed.ncbi.nlm.nih.gov/12737858/","calculation_notes":"Bingham et al. 2003 is the landmark EPIC analysis — the largest prospective study of fiber and CRC at the time of publication, covering 519,978 participants across 10 European countries with dietary assessment by 7-day food diary (considered more accurate than FFQ). The ~40% reduction for doubling fiber intake from low levels is the headline finding. This is an observational association, not a causal estimate from an RCT, and cannot fully exclude confounding by other dietary and lifestyle factors. However, the dose-response relationship was consistent across countries with very different dietary patterns, strengthening the case for a real biological effect.\n","independence_note":"EPIC is methodologically independent of the US-based cohort studies (PLCO, NHS/HPFS) and the WCRF meta-analysis cited below, though the WCRF analysis includes EPIC data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3213242/","title":"Dietary fibre, whole grains, and risk of colorectal cancer: systematic review and dose-response meta-analysis of prospective studies","publisher":"BMJ / Aune D, Chan DSM, Lau R, et al.","source_type":"peer_reviewed","statistic":"Summary RR for CRC per 10 g/day of total dietary fiber: 0.90 (95% CI 0.86-0.94) across 16 prospective studies","excerpt":"\"The summary relative risk of developing colorectal cancer for 10 g daily of total dietary fibre was 0.90 (95% confidence interval 0.86 to 0.94).\"\n","source_date":"2011-11-10","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260328074134/https://pmc.ncbi.nlm.nih.gov/articles/PMC3213242/","calculation_notes":"Aune et al. 2011 is the dose-response meta-analysis conducted for the WCRF/AICR Continuous Update Project. Pooled 16 prospective cohort studies and found a consistent 10% reduction in CRC risk per 10 g/day increase in fiber intake. The dose-response curve was approximately linear with no threshold, suggesting benefit even at modest increases. Cereal fiber and whole grains showed the strongest associations. This is the basis for the WCRF's \"convincing\" evidence rating for fiber and CRC prevention.\n","independence_note":"Meta-analysis pooling multiple independent cohort studies; includes EPIC data but also NIH-AARP, PLCO, and several European cohorts.\n"},{"url":"https://www.wcrf.org/wp-content/uploads/2024/10/Colorectal-cancer-report.pdf","title":"Diet, Nutrition, Physical Activity and Colorectal Cancer","publisher":"World Cancer Research Fund / American Institute for Cancer Research","source_type":"reputable_reference","statistic":"WCRF Panel judges foods containing dietary fiber as having 'convincing' evidence for decreasing colorectal cancer risk","excerpt":"\"The Panel judges that consumption of foods containing dietary fibre decreases the risk of colorectal cancer.\"\n","source_date":"2024-10-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260408111037/https://www.wcrf.org/wp-content/uploads/2024/10/Colorectal-cancer-report.pdf","calculation_notes":"The WCRF/AICR Continuous Update Project is the most comprehensive ongoing review of diet and cancer evidence. The \"convincing\" grade is the highest evidence level, reserved for associations with consistent results across multiple study designs, biological plausibility, and evidence of a dose-response. For fiber and CRC, this grade has been maintained through multiple update cycles. The report estimates that roughly 12% of CRC cases in high-income countries could be prevented by adequate fiber intake alone — a population-attributable fraction that is larger than most dietary factors.\n","independence_note":"WCRF/AICR panel review synthesises global evidence independently; uses the Aune et al. meta-analysis as one input but evaluates biological plausibility and experimental evidence separately.\n"}],"comparison_anchors":[{"label":"Colorectal cancer death (global adult, lifetime)","lifetime_us_adult":0.013},{"label":"All cancer (US adult, lifetime incidence)","lifetime_us_adult":0.4},{"label":"Heart disease death (US adult, lifetime)","lifetime_us_adult":0.2}],"personal_factor_multipliers":[{"factor":"High fiber intake (>30 g/day)","multiplier":0.7,"notes":"~30% reduction versus low intake, based on dose-response meta-analysis"},{"factor":"Very low fiber intake (<10 g/day)","multiplier":1.3,"notes":"Below-average intake associated with ~30% higher risk in prospective cohorts"},{"factor":"Family history of CRC (first-degree relative)","multiplier":2,"notes":"Genetic risk is additive with dietary risk; fiber benefit likely preserved but baseline is higher"},{"factor":"Regular screening (colonoscopy from 45)","multiplier":0.4,"notes":"Screening reduces CRC mortality by ~60% independent of fiber intake"}],"short_label":"Low-fiber CRC risk","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline number is US lifetime colorectal cancer incidence (~3.9%, ~1 in 26), not mortality and not a fiber-attributable fraction — it is the whole-population baseline, with fiber's effect expressed through the dose-response figures and multipliers rather than baked into the headline. CRC 5-year survival is ~65% overall and >90% for localized disease, so the death figure is lower (~1.5-2%). The fiber-CRC association is from observational epidemiology and cannot fully exclude confounding by other dietary and lifestyle factors that correlate with high fiber intake (e.g., higher vegetable consumption, lower red meat, higher physical activity, lower BMI). No large RCT has tested whether increasing fiber intake reduces CRC incidence over a multi-decade follow-up. The WCRF's \"convincing\" grade reflects the consistency and biological plausibility of the association, not proof from experimental intervention.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A few stalks of wheat and a scattering of grain on a pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/low-fiber-colorectal-cancer","api_url":"https://likelier.app/api/fears/low-fiber-colorectal-cancer.json"},{"slug":"traveler-unsafe-water","question":"What are the odds of getting seriously ill from drinking water while traveling to a developing country?","quick_answer":"Serious, hospitalization-level illness runs about 8 in 1,000 per two-week trip to a high-risk region — roughly 1 in 26 over a lifetime of five such trips. Travelers typically guess a 10-30% chance of serious illness per trip, well above the real rate for the severe illness the fear is about.\n","category":"health","tags":["travel","food"],"no_reliable_estimate":false,"perceived":{"description":"\"Don't drink the water\" is one of the most widely repeated pieces of travel advice, issued by parents, guidebooks, and doctors alike before any trip to a developing country. The warning conjures images of days lost to severe gastroenteritis, emergency IV drips, and ruined itineraries. Most travelers heading to South Asia, Sub-Saharan Africa, or Central America assume the risk of serious illness from local water is high — perhaps 10–30% per trip. That intuition is directionally correct for any diarrheal illness but overstates the risk of the serious, hospitalization-level illness that the fear is really about.\n","rough_estimate":"travelers typically guess 10–30% chance of serious illness per trip to a developing country","kind":"intuition"},"native":{"display":"~8 in 1,000 per 2-week trip to a high-risk region (serious traveler's diarrhea)","numerator":8,"denominator":1000,"unit":"per 2-week trip to high-risk region","population":"travelers to high-risk regions (South/Southeast Asia, Sub-Saharan Africa, Central America)"},"normalized":{"lifetime_us_adult":0.039,"display":"~1 in 26 over a lifetime of travel (5 trips to high-risk regions)","log_value":-1.41,"assumptions":"CDC Yellow Book estimates traveler's diarrhea (TD) affects 30–70% of travelers to high-risk regions over a 2-week trip, with ~40% as a widely cited midpoint. Of those TD cases, approximately 1–5% require medical attention at the level of hospitalization or IV rehydration; ~2% is a reasonable central estimate from the literature, giving a serious-TD rate of ~0.8% per trip (8/1,000). The native figure uses this serious-TD estimate rather than all-cause diarrhea, because the \"any loose stool\" figure (~400/1,000) overstates the fear's actual content — the concern is being truly incapacitated, not merely inconvenienced. Normalized to activity_specific_lifetime: assuming 5 trips to high-risk regions over a lifetime, P(at least one serious TD) = 1 - (1 - 0.008)^5 ≈ 0.039 (~3.9%). Uncertainty range reflects variation in destination risk (SE Asia higher, Central America lower), trip duration, accommodation type (backpacker vs. resort), and adherence to food/water precautions.\n","uncertainty":{"low":0.01,"high":0.1},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/travel/yellowbook/2024/preparing/travelers-diarrhea","title":"Travelers' Diarrhea","publisher":"US Centers for Disease Control and Prevention — CDC Yellow Book 2024","source_type":"govt_report","statistic":"Traveler's diarrhea affects approximately 30–70% of travelers, depending on destination and season; estimated 10 million cases annually among international travelers","excerpt":"\"Traveler's diarrhea (TD) is the most predictable travel-related illness. Attack rates range from 30% to 70% of travelers during a 2-week trip to high-risk destinations, depending on the destination and season of travel. An estimated 10 million cases occur annually among international travelers.\"\n","source_date":"2023-05-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20250421062226/https://wwwnc.cdc.gov/travel/yellowbook/2024/preparing/travelers-diarrhea","calculation_notes":"CDC Yellow Book gives a 30–70% incidence range; midpoint ~40% (400/1,000) is used for any-TD incidence. Approximately 2% of TD cases are estimated to require hospitalization or IV rehydration (CDC and Steffen et al. 2015 both cite roughly 1–5%); 40% × 2% = 0.8% (8/1,000) per trip for serious TD. This is the native numerator/denominator.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22099542/","title":"Inconvenience due to travelers' diarrhea: a prospective follow-up study","publisher":"BMC Infectious Diseases (Soonawala D, Vlot JA, Visser LG)","source_type":"peer_reviewed","statistic":"In a prospective follow-up cohort, 41% of travelers contracted diarrhea (median duration 2.5 days); of those affected, 21% had to alter their program, 5% consulted a physician, and 2/160 (1%) were admitted to hospital","excerpt":"\"Of 160 travelers who contracted TD ... the majority (107/160, 67%) could conduct their activity program as planned despite having diarrhea. However, 21% (33/160) were forced to alter their program and an additional 13% (20/160) were confined to their accommodation for one or more daylight days. Eight travelers (5%) consulted a physician for the diarrheal illness.\" Two travelers (2/160; 1%) were admitted to hospital.\n","source_date":"2011-11-20","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20250323114711/https://pubmed.ncbi.nlm.nih.gov/22099542/","calculation_notes":"Soonawala et al.'s prospective follow-up cohort (160 travelers with TD, 41% incidence) confirms the CDC hospitalization estimate: 2/160 (1%) of affected travelers were admitted to hospital and 5% consulted a physician. Using the CDC 40% any-TD incidence midpoint: 40% × 1–2% = 0.4–0.8% serious TD per trip. We use 0.8% (2% hospitalization fraction) as the central native estimate, which is conservative and consistent with the range. The Soonawala figure (~1% hospitalization of affected travelers) supports the lower bound of the uncertainty range.\n","independence_note":"Soonawala et al. is a fully independent prospective follow-up cohort of Dutch travelers, providing a separate estimate of TD severity from CDC's guidance compilation. Both converge on a roughly 1–5% serious-illness fraction of all TD cases.\n"}],"comparison_anchors":[{"label":"Any traveler's diarrhea per 2-week trip to high-risk region","lifetime_us_adult":0.4},{"label":"Blood clot (VTE) from a long-haul flight (per flight)","lifetime_us_adult":0.000215},{"label":"Dengue per 2-week trip to endemic area","lifetime_us_adult":0.005}],"regional_breakdown":[{"region":"High-risk (South/Southeast Asia, Sub-Saharan Africa, Central America)","probability":0.008,"notes":"Serious-TD rate ~0.8% per trip; CDC Yellow Book places any-TD incidence at 60–70% in highest-risk rural areas."},{"region":"Intermediate-risk (Eastern Europe, Southern Africa, parts of Caribbean/South America)","probability":0.003,"notes":"Any-TD incidence 8–20% per trip; serious-TD fraction yields ~0.3% per trip."},{"region":"Low-risk (Western Europe, Japan, Australia, New Zealand, Canada, US)","probability":0.0001,"notes":"Tap water is safe; serious TD essentially zero. Resort hotels in moderate-risk areas approach this level."}],"personal_factor_multipliers":[{"factor":"High-risk destination (rural South/Southeast Asia, Sub-Saharan Africa)","multiplier":2.5,"notes":"CDC Yellow Book notes incidence can reach 60–70% per 2-week trip in the highest-risk destinations; rural or budget-accommodation travel compounds risk further vs. the 40% midpoint."},{"factor":"Low-risk destination (resort hotel, Western Europe, Japan, Australia)","multiplier":0.05,"notes":"Western Europe and Japan have tap water safety comparable to the US; incidence essentially zero. Even resort hotels in moderate-risk areas with filtered water and managed food supply reduce TD risk by ~90%."},{"factor":"Immunocompromised traveler (HIV, chemotherapy, IBD)","multiplier":4,"notes":"Immunocompromised travelers face substantially higher rates of severe and prolonged TD; CDC advises prophylactic antibiotics for some high-risk immunocompromised travelers specifically."},{"factor":"Backpacker / street food / tap water consumption","multiplier":3,"notes":"Budget travelers eating street food, drinking tap water, and consuming ice are at the upper end of the 30–70% any-TD range; adhering strictly to bottled water and cooked food compresses risk toward the lower end."}],"short_label":"Traveler's diarrhea (water)","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"Water is only one of several routes for traveler's diarrhea; food is often the larger contributor. Raw produce washed in tap water, ice made from untreated water, buffet dishes at room temperature, and undercooked meat collectively account for a substantial proportion of TD cases even when travelers are careful about drinking water. The \"don't drink the water\" framing captures the fear accurately but may lead travelers to over-focus on beverages while eating freely from street stalls — a risk-allocation error. The native figure represents serious TD (hospitalization or IV-rehydration level), not the more common self-limiting 2–3 day diarrhea that resolves without treatment. The ~40% all-cause TD rate is the figure most often cited in warnings, but the proportion causing meaningful trip disruption or medical evacuation is substantially lower. Antibiotic prophylaxis (e.g., rifaximin) reduces TD incidence by 70–90% in high-risk travelers but is generally reserved for immunocompromised or short-trip travelers where onset timing is critical; the CDC and IDSA do not recommend routine prophylaxis for healthy travelers.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A glass of tap water on a tiled surface, with a faint world map in the background, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/traveler-unsafe-water","api_url":"https://likelier.app/api/fears/traveler-unsafe-water.json"},{"slug":"skiing-serious-injury","question":"What are the odds of serious injury or death while skiing?","quick_answer":"An active recreational skier faces roughly a 1-in-25 chance of serious injury per 20-day season, about 1 per 500 ski days; death is far rarer, near 1 in 1.4 million skier visits. People overestimate the fatal end while their sense of injury risk is roughly calibrated.\n","category":"health","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"Most non-skiers rate downhill skiing as obviously dangerous and most skiers know someone who has torn a knee ligament on a lift-served mountain. The perceived fear clusters around two very different outcomes — the vivid, rare fatality (tree collision, avalanche) and the common, unglamorous ACL tear — and the public discussion rarely distinguishes them. We have not found a standalone survey isolating \"fear of skiing injury\", so perceived risk is marked as editorial intuition. The interesting property of this fear is that the fatal framing is heavily overestimated relative to the roughly one-in-a-million-visits reality, while the serious-injury framing is roughly calibrated.\n","rough_estimate":"most people expect roughly 1 in 100 chance of meaningful injury per ski day","kind":"intuition"},"native":{"display":"~1 serious injury per ~500 skier visits (per ski day)","numerator":1,"denominator":500,"unit":"per skier visit (one ski day)","population":"US alpine skiers and snowboarders at lift-served resorts"},"normalized":{"lifetime_us_adult":0.0392,"display":"~1 in 25 per 20-day ski season (active recreational skier)","log_value":-1.407,"assumptions":"Scope is activity_specific_lifetime, expressed on a per-season basis for reader relatability. Starting from a per-ski-day serious injury probability of roughly 1 in 500 (the midpoint of the 1-in-500 to 1-in-1000 range reported across the modern epidemiology — Shealy/Johnson/Ettlinger's long-running Vermont series finds about one medically significant injury per 400 skier visits, of which roughly half are serious enough to warrant medical follow-up beyond first aid), the probability of at least one serious injury across a typical 20-day season is 1 − (1 − 0.002)^20 ≈ 0.0392, or about 1 in 25. Compounded across a 30-season (~600 ski-day) active lifetime the implied probability rises to roughly 1 − (1 − 0.002)^600 ≈ 0.70 — most lifetime active skiers will eventually experience at least one hospitalization-level injury. Fatality risk is far smaller: the NSAA's long-run figure of roughly 0.7 fatalities per million skier visits implies ~1 in 1.4 million per ski day, ~1 in 70,000 per 20-day season, and ~1 in 2,300 over a 600-day active career. We report the serious-injury number as the headline because fatalities are a small subset of the outcome the fear is actually about.\n","uncertainty":{"low":0.02,"high":0.08},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6299353/","title":"Alpine Skiing Injuries","publisher":"Sports Health / National Library of Medicine (PMC)","source_type":"peer_reviewed","statistic":"~38 fatalities per US ski season, 0.67 fatalities per million skier visits; ACL injury incidence 0.23 per 1,000 skier days","excerpt":"\"approximately 38 fatalities occur each ski season in the United States, equaling 0.67 fatalities per million skier visits.\"\n","source_date":"2018-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183550/https://pmc.ncbi.nlm.nih.gov/articles/PMC6299353/","calculation_notes":"The peer-reviewed Sports Health review by Hébert-Losier and Holmberg synthesizes alpine skiing injury epidemiology including Shealy, Johnson, and Ettlinger's long-running series. Two numbers from this paper anchor the page: the population-level fatality rate of 0.67 per million skier visits (used as the ceiling on per-day fatality risk, giving ~1 in 1.5 million per skier visit), and the ACL-specific rate of 0.23 per 1,000 skier days (a lower bound on serious knee injuries alone, since ACL tears are one subset of serious injury). The paper also reports that 43-77 percent of alpine skiing injuries involve the lower extremity, with the knee accounting for 27-41 percent of all injuries — consistent with the common characterization of recreational alpine skiing as a knee-injury sport.\n","independence_note":"The review draws on Shealy-Johnson-Ettlinger primary data, which the Vermont Ski Safety source also cites. The two sources are not fully independent on the underlying injury dataset, but the review is independent in the sense that it synthesizes the Shealy-Johnson series with international studies and places the US numbers in a global epidemiology context.\n"},{"url":"https://vermontskisafety.com/ufaqs/is-skiing-becoming-more-dangerous-as-claimed-by-the-la-times/","title":"Is skiing becoming more dangerous as claimed by the LA Times?","publisher":"Vermont Ski Safety (Shealy / Johnson / Ettlinger)","source_type":"reputable_reference","statistic":"~1 medically significant injury per 400 skier visits; ~20,000 to 25,000 severe knee (ACL) injuries per year among Alpine winter sports participants; overall injury rates fell ~50 percent over 27 years","excerpt":"\"about one medically significant injury in every 400 skier visits\" ... \"Over the past twenty-seven years, skiing injury rates have declined by half.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260426211546/https://vermontskisafety.com/ufaqs/is-skiing-becoming-more-dangerous-as-claimed-by-the-la-times/","calculation_notes":"The Vermont Ski Safety FAQ is the public face of the Shealy-Johnson-Ettlinger research program, which has catalogued more than 16,000 injuries at a single northern Vermont ski area over 27 years and is the longest continuous alpine ski injury dataset in the world. The \"1 medically significant injury per 400 skier visits\" figure is the direct source for the 1 in 400 to 1 in 1,000 per-day range used to set this page's native rate. We report 1 in 500 as the point estimate because \"medically significant\" as Shealy defines it includes injuries that are treated and released without hospitalization, while the headline we want is closer to \"hospitalization-level\" — conservatively about half of the full medically-significant set. The 20,000 to 25,000 annual serious knee injuries across ~60 million US skier visits also implies a knee-specific rate of roughly 1 in 2,500 to 1 in 3,000 per visit, consistent with the Sports Health paper's 0.23 per 1,000 skier days.\n","independence_note":"Shares the Shealy-Johnson primary dataset with the peer-reviewed Sports Health review above. Treat as method cross-check, not two independent measurements.\n"},{"url":"https://coloradosun.com/2025/05/02/colorado-ski-deaths-2024-25/","title":"At least 13 people died on Colorado ski slopes during the 2024-25 season","publisher":"The Colorado Sun","source_type":"news_article","statistic":"NSAA reported 35 ski-related deaths in 2023-24 vs a 10-year average of 42; approximately 1 fatality per million skier visits nationally","excerpt":"\"In the 2023-24 ski season, the association reported 35 deaths, which was below the 10-year average of 42 deaths.\" ... \"Colorado's death rate is significantly higher than the national average, with about one fatality for every million visits.\"\n","source_date":"2025-05-02","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260503082803/https://coloradosun.com/2025/05/02/colorado-ski-deaths-2024-25/","calculation_notes":"The Colorado Sun is used as a corroboration source for the NSAA fatality figures, because the NSAA's own fatality fact sheet is a binary PDF that could not be directly verified via our standard web-fetch check. The 35 deaths across roughly 60 million skier visits in 2023-24 implies 0.58 deaths per million skier visits, which is the 10-year low and is consistent with the Sports Health review's long-run figure of 0.67 per million. The 10-year NSAA average of 42 deaths per season was used to sanity-check the paper's \"approximately 38 per season\" figure; they differ because the paper was written before the 2019-2024 data were available, not because the two sources disagree about methodology.\n","independence_note":"Quotes NSAA directly, so is not an independent measurement of the fatality rate, but is independent of the Shealy-Johnson injury dataset and serves as the NSAA cross-check for the two authoritative sources above.\n"},{"url":"https://www.sciencedirect.com/science/article/abs/pii/S0020138323004916","title":"Incidence of alpine skiing and snowboarding injuries","publisher":"Injury (Elsevier) / Wagner M, Liebensteiner M, Dammerer D, et al.","source_type":"peer_reviewed","statistic":"43,283 injury cases across 98.1 million skier days (2017-2022) in Tyrol, Austria; overall incidence 0.44 injuries per 1,000 skier days","excerpt":"\"43,283 cases across 98.1 million skier days with an overall incidence of 0.44 injuries per 1,000 skier days, representing a significant reduction compared with previous studies.\"\n","source_date":"2023-06-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20230524105215/https://www.sciencedirect.com/science/article/abs/pii/S0020138323004916","calculation_notes":"Austrian Tyrolean registry data — genuinely independent of the US Shealy-Johnson Vermont dataset. The 0.44/1,000 rate is lower than the US figure (~1-2/1,000) partly due to different injury-capture thresholds but confirms the order of magnitude.\n","independence_note":"Fully independent of Shealy-Johnson — different country, different data source (Austrian emergency dispatch vs US ski patrol), different time period.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death on a motorcycle (lifetime, active US rider)","lifetime_us_adult":0.02},{"label":"Death during spaceflight (per astronaut per mission)","lifetime_us_adult":0.024}],"regional_breakdown":[{"region":"Per ski day (serious injury, recreational alpine skier)","probability":0.002,"notes":"Point estimate from Shealy-Johnson Vermont series, midpoint of 1 in 500 to 1 in 1,000 per skier visit. Headline per-day risk for the fear."},{"region":"Per 20-day ski season (at least one serious injury)","probability":0.0392,"notes":"Compounded from the per-day rate. Roughly 1 in 25, or about 4 percent per season. Rises to ~6 percent for a heavy 30-day season."},{"region":"Per 600-day active lifetime (~30 seasons of 20 days)","probability":0.699,"notes":"Most lifetime active skiers will eventually experience at least one hospitalization-level injury. This is the activity_specific_lifetime probability; it is not directly comparable to the population-lifetime figures on other Likelier pages."},{"region":"Per ski day (fatality)","probability":7e-7,"notes":"NSAA long-run rate of ~0.7 fatalities per million skier visits. The fear's vivid headline — death on the slopes — is roughly 3,000 times less likely per ski day than serious injury."},{"region":"Per 600-day active lifetime (fatality)","probability":0.00042,"notes":"~1 in 2,400 over a 30-season active career. Comparable to the lifetime odds of dying in a bicycle crash or drowning for a US adult."},{"region":"Backcountry / out-of-bounds (per equivalent day)","probability":0.02,"notes":"Order-of-magnitude estimate. Avalanche fatality rate dominates; per-hour fatality risk is roughly 10x the in-bounds resort rate depending on conditions and slope choice. 'Skiing' aggregates wildly different activities."}],"personal_factor_multipliers":[{"factor":"helmet worn","multiplier":0.7,"notes":"Roughly 30 percent reduction in head-injury risk, not zero. Helmet usage is now ~91 percent of US skiers, up from ~25 percent in 2003. Helmets are most effective against mild-to-moderate head injuries; they do not meaningfully change tree-collision survival at high speed."},{"factor":"beginner on green runs","multiplier":0.5,"notes":"Lower closing speeds and less aggressive terrain. Partly offset by higher fall rate in absolute terms — the net effect is a reduction in serious injury rate, not fall rate."},{"factor":"backcountry / off-piste","multiplier":4,"notes":"Avalanche risk dominates. In a bad snow year or on unfamiliar terrain the multiplier is meaningfully higher; this is a working average, not a ceiling."},{"factor":"under 18 or over 65","multiplier":1.5,"notes":"Younger skiers have higher per-visit injury rates largely driven by risk exposure and terrain park use; older skiers have lower fall rates but higher case-fatality given a serious injury."},{"factor":"snowboarder rather than alpine skier","multiplier":1.5,"notes":"Snowboarders have a higher overall injury rate per visit than alpine skiers, concentrated in wrist and upper-body injuries. Serious head and lower-extremity injury rates are closer to parity."}],"short_label":"Skiing injury","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"\"Skiing injury\" aggregates outcomes that differ by three or more orders of magnitude. The per-day fatality rate is around 1 in 1.4 million; the per-day rate of any medically attended injury is around 1 in 400; and the rate of career-ending or paraplegia-level catastrophic injury is around 1 per million skier visits, similar to the fatality rate. We have chosen \"serious non-fatal injury\" — roughly the hospitalization-or-longer-term- follow-up threshold — as the headline because that is the outcome the fear is actually about for most readers. The underlying Shealy-Johnson-Ettlinger series is drawn from a single northern Vermont ski area, so there is an open question about generalizability to larger western resorts with higher average speeds and longer runs, though the NSAA national aggregates do not suggest a large regional gap in the fatality rate once traffic is accounted for. The backcountry subgroup is genuinely a different activity: avalanche fatalities at lift-served resorts are rare, but out-of-bounds and sidecountry terrain carry order-of-magnitude higher per-hour fatality risk and are not covered by the headline number on this page. Finally, helmet adoption has risen from roughly 25 percent in 2003 to roughly 91 percent in 2024, which has meaningfully reduced the head- injury fraction of the total injury mix but has not moved the overall serious-injury rate by a large factor.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pair of crossed ski poles resting in calm snow, flat vector illustration in muted greys and cool blue."},"canonical_url":"https://likelier.app/skiing-serious-injury","api_url":"https://likelier.app/api/fears/skiing-serious-injury.json"},{"slug":"ai-deepfake-intimate-image-adult","question":"What are the odds an AI-generated intimate deepfake of you will be created or shared without consent in your lifetime?","quick_answer":"About 1 in 25 over a typical adult lifetime, with roughly 2% already reporting it by 2023, and the risk falls overwhelmingly on women. Most people file this under \"happens to celebrities.\" It is more common, and more personal, than that framing suggests.\n","category":"tech","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Public discourse about non-consensual intimate deepfakes is dominated by a handful of celebrity cases — Taylor Swift in January 2024, women in the US Congress later that year, K-pop idols across multiple incidents. The mental model that emerges is \"this happens to famous women on the internet,\" which lets ordinary adults file the risk under things-that-happen-to-other- people. The cost trajectory of the underlying tools does not support that framing: open-source models that required a research-grade GPU in 2019 now run on consumer phones, and a usable face-swap on a single photograph takes minutes. Asked directly, most adults significantly underestimate the population prevalence of personal victimization, which a peer-reviewed survey of ~16,000 respondents across 10 countries put at 2.2% as of 2023 — in absolute terms, several million people, drawn overwhelmingly from a single half of the population.\n","rough_estimate":"Most adults file this under 'happens to celebrities' rather than a personal risk","kind":"intuition"},"native":{"display":"~2 in 100 (ever, across 10 countries 2023)","numerator":22,"denominator":1000,"unit":"lifetime-to-date (2023)","population":"adults aged 16+ across 10 countries (Australia, Belgium, Denmark, France, Italy, Netherlands, Mexico, South Korea, UK, US), Umbach et al. CHI 2024"},"normalized":{"lifetime_us_adult":0.04,"display":"~1 in 25 over a typical adult lifetime","log_value":-1.4,"assumptions":"The 2.2% figure from Umbach et al. (2024) is a contemporaneous \"ever experienced\" prevalence measured in 2023, when consumer-grade deepfake tools had been broadly accessible for roughly five years. Translating it into a remaining-lifetime probability for a US adult requires two adjustments in opposite directions. First, the headline understates true victimization because (a) many victims do not know intimate deepfakes of them exist — the imagery is shared in closed channels or hosted on dedicated sites the subject never visits, and (b) self-report surveys chronically under-capture sexual-violence categories due to recall and disclosure barriers, with Eaton et al.'s (2017) US-only NCII survey finding 8% lifetime prevalence on a broader definition that includes real (non-AI) intimate imagery. Second, the technology curve is steep: the Home Security Heroes census found a 550% increase in detected deepfake videos from 2019 to 2023, almost entirely pornographic and targeting women. A reasonable lifetime estimate combining the peer-reviewed contemporary base rate with a moderate growth assumption over a 59-year horizon lands near 4% — roughly double the 2023 prevalence, well below Eaton's broader-definition 8% which captured fifteen years of pre-AI image-based abuse. The uncertainty band is wide (2%-10%) and skews upward because the technology is still in rapid diffusion.\n","uncertainty":{"low":0.02,"high":0.1},"scope":"us_adult_lifetime"},"sources":[{"url":"https://arxiv.org/abs/2402.01721","title":"Non-Consensual Synthetic Intimate Imagery: Prevalence, Attitudes, and Knowledge in 10 Countries","publisher":"Umbach, Henry, Beard & Berryessa — Proceedings of CHI 2024 (arXiv preprint)","source_type":"peer_reviewed","statistic":"2.2% of all respondents indicated personal victimization with non-consensual synthetic intimate imagery; 1.8% indicated perpetration; survey of >16,000 respondents across 10 countries (Australia, Belgium, Denmark, France, Italy, Netherlands, Mexico, South Korea, UK, US)","excerpt":"\"2.2% of all respondents indicated personal victimization, and 1.8% all of respondents indicated perpetration behaviors regarding deepfake pornography.\"\n","source_date":"2024-02-02","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260301234528/https://arxiv.org/abs/2402.01721","calculation_notes":"Umbach et al. (2024) is the first peer-reviewed multi-country population survey specifically measuring synthetic (AI-generated) intimate-imagery victimization rather than the broader image-based-abuse category. The paper was accepted to CHI 2024 (the flagship ACM human-computer interaction conference). Sample is >16,000 respondents across the 10 listed countries; the 2.2% figure is the pooled personal-victimization rate across all respondents. This is the headline native rate. The fielding year is 2023, so the figure is best interpreted as \"ever experienced as of 2023\" — a contemporaneous lifetime-to-date prevalence with the technology only ~5 years into mainstream availability. The lifetime extrapolation to ~4% combines this base with a moderate forward-growth assumption (Home Security Heroes census shows 550% growth 2019-2023), capped well below the 8% lifetime rate Eaton et al. found for the broader NCII category that includes real imagery.\n"},{"url":"https://internationalaisafetyreport.org/publication/2026-report-extended-summary-policymakers","title":"International AI Safety Report 2026 — Extended Summary for Policymakers","publisher":"International AI Safety Report (chaired by Yoshua Bengio), backed by 30+ countries, the EU, OECD and UN","source_type":"govt_report","statistic":"Restates the multi-country survey finding that 2.2% of >16,000 respondents across 10 countries reported someone had generated non-consensual intimate images of them; notes 96% of deepfake videos are pornographic, 15% of UK adults report having seen deepfake pornographic images, and that the vast majority of 'nudify' apps explicitly target women","excerpt":"\"In a survey of over 16,000 respondents across 10 countries, 2.2% of respondents said that someone had generated non-consensual intimate images of them.\"\n","source_date":"2026-02-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260522182518/https://internationalaisafetyreport.org/publication/2026-report-extended-summary-policymakers","calculation_notes":"The International AI Safety Report is the intergovernmental synthesis assessment (the AI-equivalent of the IPCC) commissioned after the Bletchley Park AI Safety Summit and chaired by Yoshua Bengio, with contributions vetted across 30+ governments plus the EU, OECD and UN. It is cited here as an independent authoritative corroboration of the Umbach et al. (2024) 2.2% headline rather than as a new base rate: the 2026 report restates the same multi-country figure, confirming it remains the best available adult-population prevalence estimate as of early 2026 and that no superseding adult-lifetime rate has been published. It does NOT move the headline number — it stabilises it. The additional context it supplies (\"nudify\" apps overwhelmingly target women; 15% of UK adults have seen pornographic deepfakes) reinforces the gender asymmetry and the rapid-diffusion framing already in this entry.\n","independence_note":"The AI Safety Report is a secondary synthesis that cites Umbach et al., so it is not statistically independent of the headline figure. It is retained because its govt-report authority and 2026 vintage confirm the 2.2% number had not been revised upward in any peer-reviewed adult survey by mid-2026, which is exactly why this entry's headline was left at 4% rather than raised on the strength of youth-only (Thorn) data.\n"},{"url":"https://www.cybercivilrights.org/wp-content/uploads/2017/06/CCRI-2017-Research-Report.pdf","title":"2017 Nationwide Online Study of Nonconsensual Porn Victimization and Perpetration","publisher":"Eaton, Jacobs & Ruvalcaba — Cyber Civil Rights Initiative","source_type":"primary_study","statistic":"1 in 12 (8.0%) of US adult respondents reported lifetime victimization of nonconsensual pornography; 1 in 20 (5%) reported perpetration; women's victimization rate was higher than men's; n=3,044 US adults","excerpt":"\"Among participants (54% women), 1 in 12 reported at least one instance of nonconsensual pornography victimization in their lifetime, and 1 in 20 reported perpetration of nonconsensual pornography. Women reported higher rates of victimization and lower rates of perpetration than men.\"\n","source_date":"2017-06-12","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260529100326/https://www.cybercivilrights.org/wp-content/uploads/2017/06/CCRI-2017-Research-Report.pdf","calculation_notes":"The Eaton/CCRI 2017 study is the most-cited US adult NCII prevalence figure and was published as a peer-reviewed follow-up in Psychology of Violence (Ruvalcaba & Eaton, 2019). It covers all non-consensual intimate imagery — overwhelmingly authentic photographs taken consensually and later distributed without consent — not just AI-generated deepfakes, which were not yet a measurable category in 2016 when the survey was fielded. The 8% lifetime figure is included here as the broader-category ceiling: any plausible deepfake-specific lifetime rate should sit at or below this number, because synthetic imagery is one mechanism within a larger problem that pre-dates the AI tools. The CCRI sample (3,044 US adults via online panel) has the usual online-panel skew but is the most authoritative US-specific headline number available. Used to set the upper bound on the uncertainty interval, not the central estimate.\n","independence_note":"Eaton et al. used a US-only online probability panel through CCRI, entirely separate from the Umbach et al. 10-country fielding. The broader definition (real + synthetic NCII) and the seven-year gap make the two figures complementary, not redundant: Umbach captures the synthetic share of a recent year; Eaton captures lifetime prevalence of the broader category before consumer deepfake tools existed.\n"},{"url":"https://www.siliconrepublic.com/enterprise/deepfakes-non-consensual-porn-research-deeptrace","title":"96pc of deepfakes online are pornographic in nature (Deeptrace report coverage)","publisher":"Silicon Republic — coverage of Deeptrace (Sensity) State of Deepfakes 2019","source_type":"news_article","statistic":"Deeptrace 2019 census found 14,678 deepfake videos online; 96% were non-consensual pornography; every observed pornographic deepfake portrayed a female subject","excerpt":"\"Deepfake pornography is a phenomenon that exclusively targets and harms women.\"\n","source_date":"2019-10-08","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20250115060625/https://www.siliconrepublic.com/enterprise/deepfakes-non-consensual-porn-research-deeptrace","calculation_notes":"The Deeptrace (later renamed Sensity AI) 2019 report is the earliest systematic census of deepfake videos online and established the gender asymmetry that has held in every subsequent measurement: pornographic deepfakes overwhelmingly target women. Used here purely to source the gender multiplier — the headline native rate is from Umbach, the severity asymmetry is from Deeptrace. The original report PDF is published on Sensity's site; this Silicon Republic coverage is used because the primary host returns 403 to automated retrieval. The 96% pornographic / ~100% female-target finding has been replicated by Home Security Heroes (2023) on a much larger 95,820-video sample with 99% female-target.\n"},{"url":"https://www.securityhero.io/state-of-deepfakes/","title":"2023 State of Deepfakes: Realities, Threats, and Impact","publisher":"Home Security Heroes","source_type":"reputable_reference","statistic":"95,820 deepfake videos identified online in 2023 (a 550% increase from 2019); 98% of deepfake videos online are pornographic; 99% of deepfake pornography targets women","excerpt":"\"Deepfake porn makes up 98 percent of all deepfake videos online, with 99 percent of them targeting women.\"\n","source_date":"2023-09-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260527182119/https://www.securityhero.io/state-of-deepfakes/","calculation_notes":"The 2023 census is included for trajectory rather than base-rate: the 550% growth from 2019 to 2023 (14,678 → 95,820 detected videos) is the diffusion signal that drives the upward skew on the uncertainty band. Home Security Heroes is a commercial security publication rather than peer-reviewed academia, but its census methodology mirrors Deeptrace's 2019 work and the corroborating 99% female-target figure aligns with both the academic literature and the Sensity follow-ups. Used to justify treating the 2.2% contemporary prevalence as a lower bound on true lifetime exposure given the rapid capability and accessibility growth.\n"},{"url":"https://www.thorn.org/research/library/deepfake-nudes-and-young-people/","title":"Deepfake Nudes & Young People","publisher":"Thorn (child-safety technology nonprofit)","source_type":"reputable_reference","statistic":"1 in 17 (6%) of US young people aged 13-20 reported having been the target of someone creating deepfake nudes of them; this experience was most prevalent among younger teen boys and young-adult women, with 1 in 10 (10%) in each group reporting victimization; around 1 in 8 (13%) personally knew an under-18 victim; n=1,200, surveyed Sept 27-Oct 7 2024","excerpt":"\"1 in 17 (6%) respondents reported they had been the target of someone using technology to create deepfake nudes of them. This experience was most prevalent among younger teen boys and young adult women, with 1 in 10 (10%) in each group reporting this form of victimization.\"\n","source_date":"2025-03-03","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260707140112/https://www.thorn.org/research/library/deepfake-nudes-and-young-people/","calculation_notes":"Thorn's March 2025 youth survey (n=1,200 US respondents aged 13-20, fielded Sept 27-Oct 7 2024) is the first US prevalence measurement specific to deepfake-nude victimization among young people. Its 6% headline is NOT comparable to this entry's adult-lifetime base rate: it is a current-prevalence figure for a 13-20 cohort, a different population and a different (point-in-time, not lifetime) basis. It is cited here for two narrow purposes only: (1) to source the young-adult woman multiplier (10% in that subgroup / 6% overall ≈ 1.7x, an internal-to-Thorn ratio), and (2) to document the younger-teen-boy subgroup and peer-on-peer \"nudify\" pattern in the caveats. It does not contribute to the adult-lifetime headline number.\n"},{"url":"https://www.missingkids.org/theissues/generative-ai","title":"Generative AI — CyberTipline reporting trajectory","publisher":"National Center for Missing & Exploited Children (NCMEC)","source_type":"govt_report","statistic":"CyberTipline reports of generative-AI child sexual abuse material rose from 4,700 (2023) to 67,000 (2024) to 1.5 million (2025), though ~1.1 million of the 2025 reports came from a single ESP (Amazon AI Services) with no actionable information; NCMEC categorized 158,000+ images/videos as AI-generated Jan 2023-Dec 2025; 275+ direct victims of GAI CSAM identified in 2024-2025","excerpt":"\"In 2023, NCMEC received 4,700 CyberTipline reports of GAI... In 2025, this increased to 1.5 million CyberTipline reports of GAI from the public and ESPs.\"\n","source_date":"2026-01-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260605012611/https://www.missingkids.org/theissues/generative-ai","calculation_notes":"NCMEC is the congressionally-designated US clearinghouse for child sexual abuse material; its CyberTipline figures are supply-side report counts, not a victim-prevalence rate, and they concern MINORS rather than the adults this entry is scoped to. They are cited solely to document the trajectory referenced in the caveats — the order-of- magnitude jump in AI-generated CSAM reporting 2023-2025. The 1.5 million 2025 figure must be read with the caveat that ~1.1 million originated from a single ESP (Amazon AI Services) and were non-actionable, so the raw count overstates the underlying growth; the 4,700 -> 67,000 jump from 2023 to 2024 and the 158,000+ human-categorized AI images are the cleaner diffusion signals. Not used for any number in this adult-scoped entry.\n"},{"url":"https://www.euronews.com/next/2026/04/30/virtual-rape-ai-and-deepfakes-are-silencing-women-in-public-life-un-report","title":"'Virtual rape': AI and deepfakes are silencing women in public life, UN report","publisher":"Euronews, reporting on UN Women's 'Tipping Point' study (with City St George's, University of London and TheNerve)","source_type":"news_article","statistic":"UN Women survey of 640+ women in public-facing roles across 119 countries (late 2025) found 41%+ had self-censored on social media to avoid abuse and 19% had reduced professional public speaking; researchers describe AI-facilitated image-based abuse as enabling 'virtual rape'","excerpt":"\"AI-assisted 'virtual rape' is now at the fingertips of perpetrators. This phenomenon accelerates the harm from online violence inflicted on women in public life.\"\n","source_date":"2026-04-30","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260702144012/https://www.euronews.com/next/2026/04/30/virtual-rape-ai-and-deepfakes-are-silencing-women-in-public-life-un-report","calculation_notes":"UN Women's \"Tipping Point: Online Violence Impacts, Manifestations and Redress in the AI Age\" report, produced with City St George's, University of London and the digital-forensics nonprofit TheNerve, surveyed 640+ women in public-facing roles (journalists, activists, politicians) across 119 countries in late 2025. Cited here for the self-censorship / chilling-effect pathway and the \"virtual rape\" framing of AI-facilitated image-based abuse referenced in the body text, not for any lifetime-probability figure: the survey measures self-reported behavioural response among a self-selected, high-public- visibility sample, not general-population victimization prevalence, so it does not feed the native/normalized headline.\n","independence_note":"UN Women is a UN entity; the study is co-produced with an academic research centre and a digital-forensics nonprofit. Independent of the Umbach et al. academic prevalence survey and of the Home Security Heroes / Deeptrace supply-side censuses; it measures self-reported behavioural impact among public-facing women rather than population prevalence.\n"}],"comparison_anchors":[{"label":"Lifetime NCII victimization (any kind, US adults, Eaton 2017)","lifetime_us_adult":0.08},{"label":"Image-based abuse victimization (Australia, eSafety 2017)","lifetime_us_adult":0.23},{"label":"Identity theft (US adult, lifetime)","lifetime_us_adult":0.3}],"personal_factor_multipliers":[{"factor":"Woman (vs man, baseline)","multiplier":4,"notes":"Deeptrace 2019 found pornographic deepfakes were exclusively female-targeted; Home Security Heroes 2023 found 99% targeted women on a 95,820-video census; Umbach et al. 2024 reported higher victimization among women but did not publish the exact split in the abstract. The 4x multiplier averages the gender skew across all categories of synthetic intimate imagery (not just the most-replicated pornographic videos, where the female share approaches 100%)."},{"factor":"Public-facing online presence (creator, journalist, politician)","multiplier":3,"notes":"The Markup / DeepMedia 2024 analysis found 1 in 6 US Congresswomen had been targeted with sexually explicit AI imagery — roughly 16%, several multiples of the general-population base rate. Public-figure status materially elevates targeting because reference images are abundant and the social payoff to perpetrators is higher."},{"factor":"Young adult (18-29)","multiplier":2,"notes":"eSafety Commissioner (Australia, 2017) found 24% of women aged 18-24 had experienced image-based abuse vs 15% of all women; the age skew toward young adults in image-based abuse generally is documented across multiple national surveys (Henry, Powell & Flynn 2017; Eaton et al. 2017). Synthetic-imagery victimization is expected to follow the same age gradient because young adults' photographic footprint online is largest."},{"factor":"Young-adult woman (the sex-by-age peak)","multiplier":1.7,"notes":"Thorn's 2025 youth survey ('Deepfake Nudes & Young People', n=1,200 US ages 13-20) found young-adult women among the two highest-victimization subgroups at 1 in 10 (10%), versus the 6% all-respondent rate — a 10/6 ≈ 1.7x within-survey ratio. This isolates the sex-by-age interaction (woman AND young adult) beyond either the standalone woman factor or the standalone young-adult factor: the two demographics combine at the leading edge of the diffusion curve. Multiplier is the internal Thorn ratio, not estimated."},{"factor":"K-12 or higher-ed student in a community where peer-generated deepfakes have been reported","multiplier":5,"notes":"Internet Watch Foundation and multiple US/UK school incidents in 2023-2024 documented clusters of student-on-student synthetic intimate imagery — typically through 'nudify' apps applied to school yearbook or social-media photographs. The base rate inside such cohorts is materially elevated. Treated as a setting multiplier rather than a personal trait."}],"short_label":"Intimate deepfake","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"Three caveats matter more than the headline number. First, the gender asymmetry is severe enough that the population-average figure misleads: the 4% lifetime estimate is roughly a weighted average of a female rate near 7-8% and a male rate well under 2%. Every census of deepfake-video supply finds the pornographic share targets women essentially exclusively. Reporting only the population average would understate the risk for women and overstate it for men, which is why the personal-factor multipliers are load-bearing rather than decorative.\nSecond, the survey data is almost certainly conservative. Many victims of synthetic intimate imagery never learn the imagery exists — it can be generated, shared, and consumed entirely outside the subject's social graph. The 2.2% contemporary figure measures self-aware victimization; the true rate is necessarily higher and unknowable by self-report. This is structurally different from cyberbullying or harassment, where the victim is by definition aware of the event.\nThird, no comparable peer-reviewed time series exists. Umbach et al. (2024) is the first multi-country population survey on this specific category. Deeptrace 2019, Home Security Heroes 2023, and the Internet Watch Foundation reports measure supply (videos online) rather than demand-side victim prevalence. The uncertainty band reflects both measurement gaps and the steepness of the technology adoption curve — any number computed from 2023 data may be substantially low by 2030.\nFourth, this entry is scoped to US-adult lifetime risk, but the fastest- moving frontier is minors and the school context, which the slug does not cover. Thorn's 2025 youth survey (n=1,200, ages 13-20) found 1 in 17 (6%) young people had been targeted with deepfake nudes of themselves, rising to 1 in 10 among both younger teen boys and young-adult women — so among teens the victimization is not female-skewed the way the adult video-supply censuses are. The dominant pattern is peer-on-peer: classmates running \"nudify\" apps on yearbook or social-media photographs, producing what is legally child sexual abuse material. NCMEC's CyberTipline reports of generative-AI CSAM rose from 4,700 in 2023 to 67,000 in 2024, with 158,000+ images human-categorized as AI-generated through 2025 (the headline 1.5 million 2025 report count is inflated by ~1.1 million non-actionable reports from a single provider and should not be read as a prevalence jump). The US TAKE IT DOWN Act, signed 19 May 2025, now criminalizes publishing non-consensual intimate imagery including deepfakes and obliges platforms to remove flagged content within 48 hours of notice (a one-year compliance window for platforms ran to May 2026). A dedicated minors / peer-generated entry with its own base rate and NCMEC/IWF/Thorn source set is warranted; the figures above are recorded here as context, not folded into the adult-lifetime headline.\nFinally, on why the headline was not revised in this update: the 2025-2026 survey evidence (Thorn) is youth-only and point-in-time, and the NCMEC trajectory is a supply-side minor-CSAM report count, not an adult-lifetime prevalence. The 2026 International AI Safety Report restates Umbach's 2.2% rather than publishing a superseding adult figure. No official adult- population survey has revised the rate, so the 4% lifetime central estimate and the 2-10% band are retained; the new sources stabilise the number and sharpen the subgroup picture rather than move it.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single closed laptop on a plain desk surface with a small blue notification dot visible at the lid edge, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/ai-deepfake-intimate-image-adult","api_url":"https://likelier.app/api/fears/ai-deepfake-intimate-image-adult.json"},{"slug":"cat-collar-strangulation","question":"What are the odds of a cat being seriously injured or killed by its collar?","quick_answer":"For a cat that wears a collar, the lifetime chance of any collar-related injury is about 4%, and roughly 0.6% die from a collar incident. That is below the 5-15% many owners fear, and breakaway collars lower it further.\n","category":"animal","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Collar anxiety is one of the quieter but persistent fears in cat ownership. Online forums overflow with stories of cats found hanging from fences, limbs caught under elasticated bands, or deep axillary wounds discovered weeks after a collar shifted. Veterinary organizations universally recommend breakaway collars, reinforcing the impression that non-breakaway collars are ticking time bombs. Many cat owners skip collars entirely rather than accept the risk, even forgoing the identification benefit that would help a lost cat get home.\n","rough_estimate":"~5-15% chance over a cat's lifetime","kind":"intuition"},"native":{"display":"~4% of collar-wearing cats experience a collar-related injury over their lifetime; ~0.6% die","numerator":4,"denominator":100,"unit":"lifetime cumulative injury rate per collar-wearing cat","population":"collar-wearing pet cats in Central Europe (Arhant et al., 2022, n > 5,000)"},"normalized":{"lifetime_us_adult":0.04,"display":"~4% lifetime probability of any collar-related injury for a cat that wears a collar (0.6% for death)","log_value":-1.4,"assumptions":"Arhant et al. (2022) surveyed over 5,000 cat owners in Central Europe and found that 4% of collar-wearing cats experienced a collar-related injury at some point in their lives, and 0.6% died from a collar incident. This is cumulative lifetime risk, not annual. The annual risk is much lower: Brinkley (2007), as restated in the Paulino & Williams (2023) review, reported 26 chronic axillary (collar-entrapment) wounds out of ~15,000 cat cases over 4 years (~0.17% of presentations), suggesting an annual per-cat incidence on the order of 0.04%. Calver et al. (2013) surveyed 107 Australian veterinarians who collectively reported 686 collar incidents and only 1 death across over 1,500 practice-years. We use the 4% lifetime injury figure from the largest available survey as the primary estimate. The 0.6% death rate is included as context but the headline figure captures all serious outcomes. Only ~20-30% of pet cats in Western countries regularly wear collars, so the population-level risk (across all cats, collared or not) is roughly 4% x 0.25 = ~1%. Breakaway collars were associated with zero injuries in the Calver study, while elasticated collars accounted for 39% of incidents and fixed-buckle collars for 18%.\n","uncertainty":{"low":0.01,"high":0.08},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.sciencedirect.com/science/article/pii/S1558787822001204","title":"Risks associated with free-roaming and collar use in cats","publisher":"Journal of Veterinary Behavior (Arhant, Lesch, Heizmann et al.)","source_type":"peer_reviewed","statistic":"4% of collar-wearing cats experienced a collar-related injury; 0.6% died from a collar incident; collar death risk was considerably lower than traffic accidents (70.5% fatality rate for cats hit by cars)","excerpt":"\"Among cats that wore collars, 4% experienced a collar-related injury and 0.6% died from a collar incident. The risk of collar-related death was considerably lower than the risk from traffic accidents or animal-inflicted injuries.\"\n","source_date":"2022-09-01","source_accessed":"2026-04-23","archive_url":"http://web.archive.org/web/20230410044552/https://www.sciencedirect.com/science/article/pii/S1558787822001204","calculation_notes":"Arhant et al. surveyed over 5,000 cat owners in Central Europe (Austria, Germany, Switzerland). The 4% injury rate and 0.6% death rate are cumulative lifetime figures for collar-wearing cats. This is the largest survey to date on collar safety outcomes. The study also found that collar-related risks were dwarfed by other outdoor hazards (traffic, predation), providing important context for risk calibration.\n"},{"url":"https://www.cambridge.org/core/journals/animal-welfare/article/abs/assessing-the-safety-of-collars-used-to-attach-predation-deterrent-devices-and-id-tags-to-pet-cats/A95144AB39DD3C9075A68DC7B3515568","title":"Assessing the safety of collars used to attach predation deterrent devices and ID tags to pet cats","publisher":"Animal Welfare (Calver, Adams, Clark & Pollock)","source_type":"peer_reviewed","statistic":"686 collar incidents and 1 death reported by 107 Australian veterinarians across 1,500+ practice-years; zero injuries from breakaway collars; 39% of incidents from elasticated collars","excerpt":"\"Across 107 veterinary respondents with a combined 1,500+ practice-years of experience, 686 collar safety incidents were reported with only one fatality. No injuries were associated with snap-release (breakaway) buckle collars. Elastic collars accounted for 39% of incidents and fixed-buckle collars for 18%.\"\n","source_date":"2013-02-01","source_accessed":"2026-04-23","archive_url":"http://web.archive.org/web/20250701035507/https://www.cambridge.org/core/journals/animal-welfare/article/abs/assessing-the-safety-of-collars-used-to-attach-predation-deterrent-devices-and-id-tags-to-pet-cats/A95144AB39DD3C9075A68DC7B3515568","calculation_notes":"Calver et al. surveyed 107 vets in Western Australia. Each vet saw roughly one non-fatal collar incident every 2.3 years. The zero-injury finding for breakaway collars is the strongest evidence that collar type is the dominant risk variable. 686 incidents / 1,500 practice-years = ~0.46 incidents per vet per year. Given that a typical small-animal vet sees hundreds of cats annually, the per-cat annual incidence is very low -- consistent with Brinkley's 0.17% of all cases.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10811975/","title":"Chronic axillary wounds in cats: what do we know and how should we manage them?","publisher":"Journal of Feline Medicine and Surgery (Paulino & Williams)","source_type":"peer_reviewed","statistic":"Feline chronic non-healing axillary (collar-entrapment) wounds are uncommon: 26 cases out of 15,000 cats seen over a 4-year period (~0.17%) in one case series (Brinkley 2007)","excerpt":"\"Feline chronic non-healing axillary wounds are uncommon in small animal practice, with 26 cases out of 15,000 cats seen over a 4-year period being reported in one case series.\"\n","source_date":"2023-04-01","source_accessed":"2026-06-30","archive_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10811975/","calculation_notes":"This 2023 peer-reviewed review (Paulino & Williams, J Feline Med Surg) restates the prevalence figure originally reported by Brinkley (2007, J Small Anim Pract 48:111-115, cited as ref 3): 26 chronic axillary wounds out of ~15,000 cat cases over a 4-year period, i.e. ~0.17% of presentations, or roughly 0.04% per cat per year. Axillary wounds are forelimb-entrapment injuries from neck collars. This figure captures only cases serious enough to present to a vet and does not include minor incidents resolved at home.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4494319/","title":"Frequency of Lost Dogs and Cats in the United States and the Methods Used to Locate Them","publisher":"Animals (Basel) / PMC (Weiss, Slater & Lord)","source_type":"peer_reviewed","statistic":"In an ASPCA-conducted national random-digit-dial survey of 1,015 pet-owning households (506 cat owners), 75% (95% CI: 64-85%) of lost cats were recovered, versus 93% of lost dogs; most recovered cats returned home on their own","excerpt":"\"75% (95% CI: 64-85%) of cats were recovered.\"\n","source_date":"2012-06-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260611065222/https://pmc.ncbi.nlm.nih.gov/articles/PMC4494319","calculation_notes":"Weiss, Slater & Lord surveyed 1,015 US households by random-digit-dial telephone (Sept-Nov 2010) about pets lost in the prior five years; 506 respondents were cat owners. The 75% cat-recovery rate is markedly lower than the 93% dog-recovery rate, and most recovered cats returned home on their own rather than via active search. Used here to support the entry's closing point that forgoing a collar (and its ID tag) trades a small, largely breakaway-collar-mitigable injury risk for a real cost if the cat is ever lost.\n","independence_note":"Independent of the veterinary-incident sources (Arhant, Calver, Paulino & Williams) used for the collar-injury statistics; this source addresses only the identification/return-rate argument in the closing paragraph.\n"}],"comparison_anchors":[{"label":"Indoor cat escape serious harm (per escape event)","lifetime_us_adult":0.1},{"label":"Dog chocolate poisoning death (per dog lifetime)","lifetime_us_adult":0.0005},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013}],"personal_factor_multipliers":[{"factor":"breakaway (snap-release) collar","multiplier":0.05,"notes":"Calver et al. found zero injuries from breakaway collars across 686 total incidents; breakaway collars are designed to release under ~2 kg of force, preventing strangulation"},{"factor":"elasticated (stretch) collar","multiplier":3,"notes":"Elasticated collars accounted for 39% of incidents in the Calver study; they stretch enough for a limb to pass through but not enough to release, causing chronic axillary wounds"},{"factor":"fixed-buckle (non-safety) collar","multiplier":2,"notes":"Fixed-buckle collars accounted for 18% of incidents; they cannot release under load, creating strangulation risk on fences and branches"},{"factor":"outdoor-access cat","multiplier":2,"notes":"Outdoor cats encounter far more entanglement hazards (fences, branches, undergrowth) than indoor-only cats wearing collars"},{"factor":"indoor-only cat with collar","multiplier":0.3,"notes":"Indoor cats face fewer entanglement hazards; main risk is furniture snag or limb entrapment during grooming"}],"short_label":"Cat collar injury","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"The 4% lifetime injury figure from Arhant et al. (2022) combines all collar types. Breakaway collars, which are now universally recommended by veterinary organizations, had zero associated injuries in the only study that stratified by collar type (Calver et al. 2013). The real-world injury rate for cats wearing modern breakaway collars is likely far below 4%. Conversely, the data may undercount collar deaths because cats that die outdoors from collar strangulation may never be found or reported. The Arhant survey was Central European and may not perfectly generalize to US cat-keeping practices. Only 20-30% of pet cats wear collars, so population-level risk is diluted further. Collar technology has improved since the Brinkley (2007) and Calver (2013) studies; current breakaway designs are more reliable than earlier versions.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-23","reviewed":true,"generated_at":"2026-04-23","image":{"alt":"A simple cat collar with a breakaway buckle laid flat on a clean surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/cat-collar-strangulation","api_url":"https://likelier.app/api/fears/cat-collar-strangulation.json"},{"slug":"medical-bankruptcy","question":"What are the odds of going bankrupt or suffering severe financial hardship due to medical bills?","quick_answer":"About 1 in 25 over a lifetime for a US adult by a causal estimate. The familiar '60%+ of bankruptcies are medical' figure overstates the specific causal role of medical bills, conflating illness with the job loss that often accompanies it.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"The claim that \"62% of bankruptcies are caused by medical bills\" has circulated in US policy debates since Himmelstein et al. published their survey-based estimate in 2009. It is one of the most-cited statistics in American health policy and anchors a widespread belief that a single hospital stay can financially destroy a middle-class family. The fear is directionally correct: medical costs do cause real financial hardship, and the US is an outlier among wealthy nations in this respect. But the headline overstates the causal role of medical expenses specifically, conflating correlation (illness often accompanies income loss) with causation.\n","rough_estimate":"most people recall '60% of bankruptcies are medical' and assume the risk is very high","kind":"intuition"},"native":{"display":"~530,000 medical-related bankruptcy filings per year (estimated, US)","numerator":530000,"denominator":131000000,"unit":"per year","population":"US households"},"normalized":{"lifetime_us_adult":0.04,"display":"~1 in 25 lifetime (US adult, causal estimate)","log_value":-1.4,"assumptions":"Two fundamentally different methodologies produce very different answers. The survey approach (Himmelstein et al. 2019): 66.5% of ~500,000 annual nonbusiness bankruptcy filings cite a medical component, yielding ~330,000 \"medical bankruptcies\" per year and a naive lifetime rate of ~13%. The causal approach (Dobkin et al. 2018, AER): hospital admissions increase bankruptcy probability by about 0.5 percentage points, and hospital admissions cause fewer than 5% of observed bankruptcies in their linked credit-report data. Applied to ~500,000 annual filings, that gives ~25,000 causally attributable medical bankruptcies per year, implying a lifetime rate of roughly 1-2%. The truth plausibly lies between these extremes. The Himmelstein figure overcounts because it includes filers whose primary driver was job loss that happened to coincide with illness. The Dobkin figure undercounts because it captures only hospitalization-triggered debt, not chronic-disease costs, outpatient bills, or medication expenses that accumulate without a discrete hospital admission. A central estimate of ~4% lifetime probability (roughly 1 in 25) reflects a judgment that medical costs are a substantial but not dominant causal factor in perhaps 50,000-80,000 filings per year, compounded over a 59-year adult horizon. This is inherently imprecise, and the uncertainty band is wide.\n","uncertainty":{"low":0.01,"high":0.13},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/30726124/","title":"Medical Bankruptcy: Still Common Despite the Affordable Care Act","publisher":"American Journal of Public Health","source_type":"peer_reviewed","statistic":"66.5% of all bankruptcies filed 2013-2016 were tied to medical issues, either because of high costs or time lost from work","excerpt":"\"Using a court-based sampling strategy, we surveyed a random sample of 910 Americans who filed for personal bankruptcy between 2013 and 2016. Medical problems contributed to 66.5% of all bankruptcies. This figure is virtually unchanged since our previous 2007 study.\"\n","source_date":"2019-02-06","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420043238/https://pubmed.ncbi.nlm.nih.gov/30726124/","calculation_notes":"Himmelstein et al. surveyed 910 bankruptcy filers from the Consumer Bankruptcy Project, 2013-2016. They classified a filing as \"medical\" if the debtor reported medical debt exceeding $5,000, lost two or more weeks of work due to illness, or cited illness or medical bills as a bankruptcy reason. This broad definition captures correlation between health events and financial distress but does not isolate the causal contribution of medical expenses specifically. The 66.5% figure applied to ~500,000 annual nonbusiness filings yields ~330,000 \"medical bankruptcies\" per year, or an annual household rate of ~0.25%. Naively compounded over 59 years: 1 - (1 - 0.0025)^59 = ~13.7%. This is the upper bound of our uncertainty range, rounded down to 13%.\n","independence_note":"Himmelstein et al. use self-reported survey data from bankruptcy filers. This is methodologically independent from the Dobkin et al. causal study, which links hospital admission records to credit reports without relying on self-report.\n"},{"url":"https://www.aeaweb.org/articles?id=10.1257/aer.20161038","title":"The Economic Consequences of Hospital Admissions","publisher":"American Economic Review","source_type":"peer_reviewed","statistic":"Hospital admissions increase out-of-pocket spending by ~$6,000 over 3 years, increase unpaid bills by ~$5,000, and raise bankruptcy rates; hospital admissions cause fewer than 5% of bankruptcies","excerpt":"\"We use an event study approach to examine the economic consequences of hospital admissions for adults in two datasets: survey data from the Health and Retirement Study, and hospitalization data linked to credit reports. [...] Hospital admissions increase out-of-pocket medical spending, unpaid medical bills, and bankruptcy, and reduce earnings, income, access to credit, and consumer borrowing.\"\n","source_date":"2018-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503081812/https://www.aeaweb.org/articles?id=10.1257%2Faer.20161038","calculation_notes":"Dobkin, Finkelstein, Kluender, and Notowidigdo linked California hospital records to credit-report data to estimate the causal effect of hospitalization on financial outcomes. They found hospital admissions increase unpaid bills by ~$5,000 and raise bankruptcy probability by approximately 0.5 percentage points among the uninsured non-elderly. Crucially, they estimate hospital admissions trigger fewer than 5% of all observed bankruptcies, a far lower figure than Himmelstein's 62-67%. Applying 5% to ~500,000 annual filings gives ~25,000 causally attributable filings per year, or an annual household rate of ~0.019%. Over 59 years: 1 - (1 - 0.00019)^59 = ~1.1%. This is the lower bound of our uncertainty range, rounded down to 1%.\n","independence_note":"Uses administrative hospital and credit-report data, methodologically independent from the Himmelstein survey-based approach. The causal identification strategy (event study around hospital admission) is distinct from the correlational survey methodology.\n"},{"url":"https://www.kff.org/health-costs/the-burden-of-medical-debt-in-the-united-states/","title":"The Burden of Medical Debt in the United States","publisher":"Kaiser Family Foundation (KFF)","source_type":"reputable_reference","statistic":"41% of US adults have some form of health care debt; about 1 in 10 adults owe $5,000+; roughly 100 million Americans carry medical debt","excerpt":"\"About four in ten adults (41%) report having some type of debt due to medical or dental bills. About 1 in 4 adults with health care debt owe at least $5,000, and about 1 in 8 owe $10,000 or more.\"\n","source_date":"2022-06-16","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426203526/https://www.kff.org/health-costs/the-burden-of-medical-debt-in-the-united-states/","calculation_notes":"KFF's 2022 Health Care Debt Survey (n=2,375 US adults, Feb-Mar 2022) provides the prevalence denominator: 41% of adults carry some medical debt, and ~10% owe $5,000+. The \"100 million Americans\" headline figure comes from KFF Health News / NPR investigation applying this survey proportion to the adult population. This establishes the at-risk population for medical bankruptcy but does not directly estimate bankruptcy probability. It confirms that medical debt exposure is far more common than medical bankruptcy, indicating that most medical debt is resolved without filing.\n"},{"url":"https://www.consumerfinance.gov/about-us/newsroom/cfpb-estimates-88-billion-in-medical-bills-on-credit-reports/","title":"CFPB Estimates $88 Billion in Medical Bills on Credit Reports","publisher":"Consumer Financial Protection Bureau","source_type":"govt_report","statistic":"$88 billion in medical debt on consumer credit reports as of June 2021; medical collections appear on 43 million credit reports; 58% of bills in collections are medical","excerpt":"\"As of the second quarter of 2021, 58% of bills that are in collections and on people's credit records are medical bills. The CFPB estimates consumers owed $88 billion in medical debt on consumer credit reports.\"\n","source_date":"2022-03-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426203603/https://www.consumerfinance.gov/about-us/newsroom/cfpb-estimates-88-billion-in-medical-bills-on-credit-reports/","calculation_notes":"CFPB's analysis of credit-bureau data establishes the scale of medical debt in the financial system. The $88 billion and 43 million credit reports with medical collections provide an independent, administrative-data check on the KFF survey estimates. Medical bills represent the single largest category of collections tradelines, confirming the outsized role of medical costs in consumer financial distress, even if not all medical debt leads to bankruptcy.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy, any cause (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Cancer diagnosis (lifetime, US adult)","lifetime_us_adult":0.4}],"regional_breakdown":[{"region":"Uninsured adults","probability":0.1,"notes":"Dobkin et al. find the uninsured non-elderly experience substantially larger increases in unpaid bills and bankruptcy following hospitalization; lifetime risk roughly 2-3x the insured average"},{"region":"Underinsured (high deductible, >$5,000)","probability":0.06,"notes":"High-deductible plans shift cost exposure; KFF finds adults with deductibles >$1,500 are twice as likely to carry medical debt"},{"region":"Well-insured (employer plan, low deductible)","probability":0.015,"notes":"Comprehensive employer coverage limits out-of-pocket exposure; medical bankruptcy risk approaches the background rate"},{"region":"Medicare/Medicaid enrolled","probability":0.01,"notes":"Gross & Notowidigdo 2011: Medicaid expansion reduces personal bankruptcies by ~8% per 10pp eligibility increase; Medicare eliminates most acute cost exposure for 65+"}],"personal_factor_multipliers":[{"factor":"uninsured at time of major medical event","multiplier":3,"notes":"Dobkin et al.: uninsured non-elderly face much larger financial shocks from hospitalization"},{"factor":"household income below 200% federal poverty level","multiplier":2.5,"notes":"KFF: low-income adults are disproportionately likely to carry medical debt >$5,000 and lack savings buffer"},{"factor":"household income top quintile, comprehensive insurance","multiplier":0.2,"notes":"high income plus good coverage makes medical bankruptcy extremely rare"},{"factor":"cancer or major chronic disease diagnosis","multiplier":2,"notes":"cancer treatment costs frequently exceed $100,000; chronic conditions generate sustained outpatient costs that the Dobkin hospitalization-only estimate may undercount"}],"short_label":"Medical bankruptcy","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"The central estimate of ~4% lifetime probability is a judgment call between two methodological extremes, not a precisely measured quantity. The Himmelstein survey approach (upper bound ~13%) captures all financially distressed filers who happen to have medical debt, regardless of whether medical costs were the primary cause. The Dobkin causal approach (lower bound ~1%) captures only hospitalization-triggered bankruptcies and misses chronic-disease costs, outpatient accumulation, and medication expenses. Neither study isolates the full causal chain from medical event to bankruptcy filing. The ACA's Medicaid expansion and marketplace subsidies likely reduced medical bankruptcy rates post-2014, but Himmelstein et al. 2019 found the overall share of bankruptcies citing medical causes was essentially unchanged at 66.5%. This may reflect rising deductibles and out-of-pocket maximums offsetting coverage gains. The CFPB's 2025 rule to remove medical debt from credit reports, if implemented, would reduce the downstream credit consequences of medical debt but would not directly reduce bankruptcy filings. State-level variation is substantial: states that expanded Medicaid show lower medical debt burdens than non-expansion states.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single medical bill and a thin wallet side by side on a plain surface, muted blue-grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/medical-bankruptcy","api_url":"https://likelier.app/api/fears/medical-bankruptcy.json"},{"slug":"cruise-ship-epidemic","question":"What are the odds of being on a cruise ship voyage that has a norovirus outbreak?","quick_answer":"A CDC-reportable norovirus outbreak occurs on about 42 of every 10,000 cruise voyages. For an active cruiser taking 10 voyages, the chance of being aboard one is roughly 1 in 24 (about 4.1%), far rarer than the 'floating petri dish' reputation implies. Being aboard an outbreak voyage is not the same as personally falling ill.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"\"Floating petri dish\" is a reliable piece of travel journalism that resurfaces every January as cruise season peaks. The narrative is vivid — thousands of passengers confined in a self-contained vessel, sharing dining halls and handrails, sometimes stuck at sea for days after illness spreads. Media coverage of high-profile outbreaks (Queen Mary 2, Celebrity cruise lines, Royal Caribbean ships) reinforces the impression that norovirus is almost a standard feature of cruising rather than an occasional event. No rigorous polling specifically asks Americans about cruise-illness fear, but industry surveys suggest a meaningful share of cruise-avoiders cite disease concerns. The fear is directionally right but orders of magnitude larger than the data support.\n","rough_estimate":"Most prospective cruisers believe illness outbreaks are common — a common industry-cited figure is that 53% of Americans who avoid cruising cite illness or disease concerns","kind":"intuition"},"native":{"display":"~42 per 10,000 voyages have a CDC-reportable gastroenteritis outbreak (≥3% of passengers ill)","numerator":42,"denominator":10000,"unit":"per voyage","population":"All cruise ship voyages under CDC VSP jurisdiction, 2006-2019 (MIDRS dataset, n=37,258 AGE surveillance reports)"},"normalized":{"lifetime_us_adult":0.041,"display":"~1 in 24 for an active cruiser over 10 voyages (4.1%)","log_value":-1.39,"assumptions":"The CDC Maritime Illness Database and Reporting System (MIDRS) recorded 156 reportable outbreaks among 37,258 AGE surveillance reports from 252 cruise ships over 2006-2019: a per-voyage outbreak rate of 0.418%, rounded to ~42 per 10,000. A CDC-reportable outbreak requires ≥3% of passengers or crew to report acute gastroenteritis symptoms to the ship's medical staff — a threshold designed to capture true outbreaks rather than background GI complaints. For someone taking 10 cruises across a lifetime, the probability of being aboard at least one outbreak voyage is 1 − (1 − 0.0042)^10 ≈ 4.1%. This is the probability of being on the ship during a reportable outbreak, not the probability of personally falling ill: within an outbreak voyage, the typical passenger attack rate is ~7% (meta-analysis of 45 outbreaks, Simou et al. 2024), so the per-voyage probability of personally becoming sick in a reportable outbreak is roughly 0.42% × 7% ≈ 0.03%. Over 59 years, total US cruise exposure per adult (averaging across all US adults, most of whom cruise rarely or never) is ~1-2 voyages, giving a ~0.5-1.0% lifetime probability of being on an outbreak voyage for the general US adult population.\n","uncertainty":{"low":0.013,"high":0.082},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8480991/","title":"Acute Gastroenteritis on Cruise Ships — Maritime Illness Database and Reporting System, United States, 2006–2019","publisher":"Centers for Disease Control and Prevention (MIDRS surveillance report)","source_type":"govt_report","statistic":"156 passenger outbreaks in 37,258 AGE surveillance reports from 252 cruise ships (0.42% per voyage); ~90% of outbreaks with known causative agents involved noroviruses","excerpt":"\"During 2006–2019, a total of 37,258 AGE surveillance reports were obtained from 252 cruise ships arriving at U.S. ports... During 2006–2019, VSP investigated 156 outbreaks among passengers and 16 outbreaks among crew... approximately 90% of cruise ship outbreaks with known causative agents involved noroviruses.\"\n","source_date":"2021-09-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250428232851/https://pmc.ncbi.nlm.nih.gov/articles/PMC8480991/","calculation_notes":"156 outbreaks / 37,258 AGE surveillance reports = 0.4187% per voyage. Rounded to 0.42% (42 per 10,000) for the native numerator. MIDRS is the definitive longitudinal dataset for CDC-jurisdiction cruise ship illness surveillance; it is larger and more comprehensive than the earlier MMWR 2016 dataset (which covered 2008-2014 only).\n"},{"url":"https://www.cdc.gov/mmwr/volumes/65/wr/mm6501a1.htm","title":"Acute Gastroenteritis on Cruise Ships — United States, 2008–2014","publisher":"CDC Morbidity and Mortality Weekly Report, January 8, 2016, Vol. 65, No. 1","source_type":"govt_report","statistic":"133 outbreaks in 29,107 voyages (0.46% per voyage); 73,599,005 total passengers; 129,678 reported GI cases (0.18% of passengers)","excerpt":"\"Among 73,599,005 passengers on cruise ships during 2008–2014, a total of 129,678 (0.18%) cases of acute gastroenteritis were reported... Of 29,107 voyages with submitted AGE reports, 133 (0.46%) had outbreaks... Norovirus caused 92 (96.8%) of the 95 laboratory-confirmed GI outbreaks.\"\n","source_date":"2016-01-08","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260511083035/https://www.cdc.gov/mmwr/volumes/65/wr/mm6501a1.htm","calculation_notes":"The MMWR 2016 dataset (2008-2014) gives 0.46% per voyage vs. the MIDRS (2006-2019) 0.42%. The rates are consistent; we use the larger MIDRS dataset as primary. The 0.18% of all passengers experiencing any GI illness per voyage — not just in outbreak voyages — represents the broader background illness rate that is two to three times higher than the outbreak-only figure.\n","independence_note":"MMWR 2016 is an earlier CDC publication drawing on the same VSP surveillance system as MIDRS. The two datasets partially overlap (2008-2014) but cover different time windows and use the same outbreak definition. They are corroborating rather than independent estimates.\n"},{"url":"https://www.cdc.gov/yellow-book/hcp/travel-air-sea/cruise-ship-travel.html","title":"Cruise Ship Travel — CDC Yellow Book 2024","publisher":"Centers for Disease Control and Prevention, Travelers' Health Yellow Book","source_type":"govt_report","statistic":"Rates of GI illness fell from 32.5 to 16.9 per 100,000 travel days (2006-2019); average of 12 norovirus outbreaks annually on US-port cruise ships (2006-2019)","excerpt":"\"During 2006–2019, rates of GI illness among passengers on voyages lasting 3–21 days fell from 32.5 to 16.9 cases per 100,000 travel days... On cruise ships, >90% of GI illness outbreaks with a confirmed cause are due to norovirus... between 2006–2019, an average of 12 norovirus outbreaks occurred annually on international cruise ships using U.S. ports.\"\n","source_date":"2024-05-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260512012110/https://www.cdc.gov/yellow-book/hcp/travel-air-sea/cruise-ship-travel.html","calculation_notes":"The 16.9 per 100,000 travel-days figure (2019, most recent available) translates to roughly 0.12% of passengers per 7-day voyage experiencing any GI symptoms reported to medical staff. This is lower than the overall 0.18% figure because rates trended downward over the surveillance period. The Yellow Book is the primary clinical reference for US travel medicine practitioners advising cruise passengers.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5718393/","title":"Gastroenteritis outbreaks on cruise ships: contributing factors and thresholds for early outbreak detection","publisher":"Eurosurveillance, Vol. 22, Issue 45, 2017 — Mouchtouri, Verykouki, Zamfir, Hadjipetris, Lewis, Hadjichristodoulou","source_type":"peer_reviewed","statistic":"9 outbreaks in 760 cruises (1.18%) on 5 European ships; attack rate 23.9% in outbreak vs 1.4% in non-outbreak voyages; overall incidence 2.81 per 10,000 traveller-days","excerpt":"\"The overall incidence rate of AG was 2.81 cases per 10,000 traveller-days (95% CI: 0.00–17.60)... nine outbreaks of AG occurred... in 760 cruises; 1.18%... The attack rate was 14.05 (95% CI: 0.00–55.08) per 10,000 travellers for non-outbreak cruises and 238.80 (95% CI: 0.00–738.70) per 10,000 travellers for outbreak cruises.\"\n","source_date":"2017-11-09","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250202170813/https://pmc.ncbi.nlm.nih.gov/articles/PMC5718393/","calculation_notes":"This European fleet dataset uses a lower outbreak threshold than the CDC's 3% rule and covers only 5 ships, so its 1.18% outbreak rate is not directly comparable to the CDC's 0.42-0.46%. It is included as a peer-reviewed independent corroboration of the per-voyage outbreak order of magnitude. The 2.81 per 10,000 traveller-days incidence rate translates to roughly 0.20% of passengers experiencing GI illness per 7-day voyage, close to the CDC's 0.18% all-passengers figure.\n","independence_note":"Mouchtouri et al. use European cruise fleet data with ship-level surveillance reporting, independent of the US CDC VSP system. The methods and fleet are distinct, making this a genuine independent corroboration.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10986668/","title":"Systematic literature review and meta-analysis on preventing and controlling norovirus outbreaks on cruise ships, 1990 to 2020","publisher":"Eurosurveillance, 2024 — Simou et al.","source_type":"peer_reviewed","statistic":"45 outbreaks on 26 cruise ships (1990-2020); weighted average passenger attack rate 7% (95% CI 5-9%); crew attack rate 2%","excerpt":"\"weighted average of prevalence (attack rate) for passengers of 7% (95% CI: 5.00–9.00)... Data from 45 outbreaks on 26 cruise ships from 1990 to 2020 were identified... Attack rates increased as the length of the cruise voyage increased, ranging from 3.57% for cruises under 7 days to 7.36% for 7–10 days and 10.59% for cruises 11+ days (Table 2).\"\n","source_date":"2024-03-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250206041021/https://pmc.ncbi.nlm.nih.gov/articles/PMC10986668/","calculation_notes":"The 7% weighted average attack rate within outbreak voyages is the multiplier used to convert per-voyage outbreak probability (0.42%) into per-voyage personal illness probability (0.42% × 7% ≈ 0.03%). This is the primary source for the within-outbreak attack rate; it synthesizes 45 outbreaks spanning 30 years and is independent of the CDC VSP surveillance pipeline.\n"}],"comparison_anchors":[{"label":"Blood clot (DVT/PE) on a long-haul flight (per flight)","lifetime_us_adult":0.0005},{"label":"Turbulence serious injury on any commercial flight (lifetime)","lifetime_us_adult":0.000004},{"label":"Traveler's diarrhea on a 2-week trip to South/Southeast Asia","lifetime_us_adult":0.4}],"short_label":"Cruise ship norovirus","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 0.42% per-voyage figure counts only CDC-reportable outbreaks — those where ≥3% of passengers report GI symptoms to the ship's medical staff. This threshold is designed to detect genuine outbreaks rather than background complaint rates; it excludes the much larger number of individual GI illnesses that do not escalate into reportable events. The broader \"any passenger GI illness per voyage\" figure from MMWR 2016 is 0.18% of all passengers, which implies roughly 1 in 50 voyages has at least one passenger reporting GI symptoms — a meaningfully higher burden than the outbreak-only metric. The surveillance data covers ships using US ports that report to the CDC VSP; vessels operating exclusively under non-US flags with different reporting standards are not captured. Outbreak rates trended downward from 2006 to 2019 (rates per 100,000 travel-days fell nearly in half), suggesting improved shipboard hygiene. The lifetime figure is activity-specific: it applies to people who cruise, not the general US adult population.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A large ocean cruise ship viewed from a low angle against a grey-white sky, flat vector illustration with muted colours."},"canonical_url":"https://likelier.app/cruise-ship-epidemic","api_url":"https://likelier.app/api/fears/cruise-ship-epidemic.json"},{"slug":"wrongful-conviction","question":"What is the probability that a criminal conviction is wrongful?","quick_answer":"The best estimate puts about 4.1% of US death-sentence convictions, roughly 1 in 24, as wrongful. Most people guess well under 1%, so the figure runs higher than intuition. It is specific to capital cases, where post-conviction scrutiny is unusually intensive.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Wrongful conviction is one of those risks that generates intense concern when it surfaces in the news — an exoneration story, a DNA reversal, a podcast — and fades between headlines. There is no national poll tracking how often Americans worry about being wrongfully convicted, so the perceived risk is best characterized as an intuition shaped by media salience. Most people, when asked, guess the error rate is \"very low\" (well under 1%), which turns out to be optimistic relative to the best available estimate.\n","rough_estimate":"Most guess well under 1%","kind":"intuition"},"native":{"display":"~4.1% of death-sentenced defendants are innocent","numerator":41,"denominator":1000,"unit":"per capital conviction","population":"US defendants sentenced to death, 1973–2004"},"normalized":{"lifetime_us_adult":0.041,"display":"~1 in 24 capital convictions is wrongful","log_value":-1.39,"assumptions":"The central estimate comes from Gross et al. (2014, PNAS), which used survival analysis on 7,482 death sentences imposed between 1973 and 2004. Of those, 117 resulted in exoneration by end-2004, and the model estimates that if all defendants remained under sentence of death indefinitely, at least 4.1% (95% CI: 2.8–5.2%) would be exonerated. This is a per-conviction probability, not a per-person lifetime probability, so scope is activity_specific_lifetime. The 4.1% figure is specific to capital cases, where post-conviction scrutiny is unusually intensive; for ordinary felonies, most researchers estimate 2–6%, but no comparably rigorous study exists. The normalized figure is used as-is because the unit of analysis is a conviction, not a person-year.\n","uncertainty":{"low":0.028,"high":0.06},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.pnas.org/doi/abs/10.1073/pnas.1306417111","title":"Rate of false conviction of criminal defendants who are sentenced to death","publisher":"Proceedings of the National Academy of Sciences","source_type":"peer_reviewed","statistic":"4.1% estimated false conviction rate among US death-sentenced defendants, 1973–2004 (95% CI: 2.8–5.2%)","excerpt":"\"We use survival analysis to model this effect, and estimate that if all death-sentenced defendants remained under sentence of death indefinitely, at least 4.1% would be exonerated. We conclude that this is a conservative estimate of the proportion of false conviction among death sentences in the United States.\"\n","source_date":"2014-04-28","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20251020072953/https://www.pnas.org/doi/abs/10.1073/pnas.1306417111","calculation_notes":"Gross et al. applied Kaplan-Meier survival analysis to 7,482 death sentences (1973–2004), identifying 117 exonerations. The 4.1% figure accounts for censoring (defendants removed from death row before exoneration could occur). The 95% CI of 2.8–5.2% is used directly as the uncertainty band, widened to 6% at the high end to account for the possibility that non-capital felonies have similar or higher base rates with less post-conviction scrutiny.\n","independence_note":"This is the only peer-reviewed study using formal survival analysis to estimate wrongful conviction rates. The National Registry of Exonerations provides independent case-level corroboration but uses a different methodology (cumulative count, not modeled rate).\n"},{"url":"https://www.law.umich.edu/special/exoneration/Pages/about.aspx","title":"National Registry of Exonerations — About","publisher":"University of Michigan Law School / Michigan State University","source_type":"reputable_reference","statistic":"3,698 known exonerations in the US since 1989 (as of mid-2025); 147 exonerations recorded in 2024","excerpt":"\"The National Registry of Exonerations provides detailed information about every known exoneration in the United States since 1989. As of June 2025, the Registry lists 3,698 exonerations.\"\n","source_date":"2025-06-28","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250414151841/https://www.law.umich.edu/special/exoneration/Pages/about.aspx","calculation_notes":"The Registry is a case-level database, not a rate estimate. It records known exonerations but cannot estimate the denominator (total convictions) or the number of wrongful convictions that were never discovered. The 147 exonerations in 2024, with an average of 13.5 years of wrongful imprisonment each, illustrate the scale but do not yield a rate. Used here as corroborating evidence for the Gross et al. estimate, not as a competing figure.\n","independence_note":"The Registry is maintained by the University of Michigan and Michigan State, independently of the Gross et al. PNAS study, though Gross is a co-founder of both. The data pipelines are distinct: the PNAS study uses survival modeling on a closed 1973–2004 cohort, while the Registry is an ongoing, open-ended case collection.\n"}],"comparison_anchors":[{"label":"Medical diagnostic error rate (US hospitals)","lifetime_us_adult":0.05},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"Eyewitness misidentification present in case","multiplier":3.5,"notes":"Innocence Project, 2023: eyewitness misidentification is a contributing factor in approximately 70% of DNA exonerations, making it the single most common contributing cause; cases relying on eyewitness ID as primary evidence face substantially elevated wrongful-conviction risk."},{"factor":"False confession in case record","multiplier":2.5,"notes":"Innocence Project: false confessions or incriminating statements were present in roughly 30% of DNA exoneration cases; false confessions are disproportionately associated with youth defendants and individuals with intellectual disabilities, who confess falsely at higher rates under interrogation pressure (Drizin & Leo, 2004, North Carolina Law Review vol. 82, catalogued 125 proven false confessions)."},{"factor":"Capital charge vs. non-capital felony","multiplier":1.5,"notes":"Gross et al. (2014, PNAS): the 4.1% exoneration rate for capital cases exceeds the observed rate for non-capital felonies, partly because post-conviction scrutiny (multiple appeal rounds, innocence-project attention, mandatory DNA review) is far more intensive for death sentences."},{"factor":"Public defender with caseload exceeding ABA standards","multiplier":2,"notes":"American Bar Association Standing Committee on Legal Aid (2022) and documented in multiple Innocence Project case reviews: inadequate defense counsel — particularly public defenders carrying caseloads above ABA's 150-case-per-year guideline — appears as a contributing factor in a substantial share of documented wrongful convictions."}],"short_label":"Wrongful conviction","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"autonomy_loss","valence":"negative","caveats":"The 4.1% figure applies specifically to capital cases, where the stakes drive extraordinary post-conviction review — multiple rounds of appeals, innocence-project involvement, and DNA testing that most felony defendants never receive. For ordinary felonies, the true wrongful conviction rate is almost certainly higher than the observed exoneration rate (since most wrongful convictions are never discovered) but may be lower than 4.1% if the capital-case selection process is unusually error-prone. Estimates for all felonies range from roughly 2% to 10%, depending on methodology and crime type. The rate also varies by offense category: wrongful convictions are disproportionately concentrated in homicide and sexual-assault cases, where eyewitness misidentification and forensic-science errors are most consequential. The figure should not be read as \"4.1% of people in prison are innocent\" — that is a related but distinct question with a different denominator.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single gavel resting on a cracked marble surface, flat vector illustration, muted tones."},"canonical_url":"https://likelier.app/wrongful-conviction","api_url":"https://likelier.app/api/fears/wrongful-conviction.json"},{"slug":"bipolar-disorder-lifetime","question":"What are the odds of developing bipolar disorder at some point in your lifetime?","quick_answer":"About 1 in 23 US adults (4.4%) develops a bipolar spectrum disorder at some point in life — considerably more common than the well-under-1% figure most people would guess.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Bipolar disorder carries a reputation as rare — a severe condition afflicting a small fraction of the population. Most adults, asked to estimate the lifetime risk for a randomly chosen adult, tend to guess something well under 1%. The actual figure is closer to 1 in 23. That gap exists partly because bipolar I (classic mania-and-depression cycling) accounts for only about a quarter of the total spectrum: bipolar II and sub-threshold presentations make up the rest, and these milder variants are frequently misdiagnosed as recurrent depression or anxiety disorders, keeping them out of the public-awareness frame. The entry covers any bipolar spectrum diagnosis (I, II, or sub-threshold) ever received over a lifetime, which is the figure most epidemiological studies report.\n","rough_estimate":"Most people guess well under 1% lifetime risk","kind":"intuition"},"native":{"display":"4.4% of US adults develop bipolar spectrum disorder at some point in their lives","numerator":44,"denominator":1000,"unit":"lifetime","population":"US adults (National Comorbidity Survey Replication, N=9,282)"},"normalized":{"lifetime_us_adult":0.044,"display":"~1 in 23 lifetime (US adult)","log_value":-1.3565,"assumptions":"The 4.4% lifetime prevalence figure is drawn from the National Comorbidity Survey Replication (NCS-R), a nationally representative sample of 9,282 English-speaking US adults aged 18 and older surveyed 2001–2003, and is the institutional figure published by NIMH. It covers the full bipolar spectrum: bipolar I (1.0% lifetime), bipolar II (1.1%), and sub-threshold bipolar disorder (2.4%). The narrower clinical definition excluding sub-threshold cases yields 2.1% (1 in 48), which is the relevant figure for comparing against diagnoses of BP-I or BP-II specifically. The 4.4% figure is the one cited on the NIMH statistics page and is the entry's headline, since the question asks about developing \"bipolar disorder\" in any form. Internationally, the World Mental Health Survey Initiative across 11 countries (N=61,392) found a lower worldwide spectrum prevalence of 2.4%, with the US being the highest-prevalence country at 4.4%; this cross-national variation likely reflects diagnostic threshold differences and measurement methodology rather than true biological variation of that magnitude. Uncertainty band of 0.030–0.060 reflects the range between the narrower DSM-IV BP-I+II-only definition (~2%) and the upper boundary of spectrum definitions, and accounts for measurement variance in population surveys.\n","uncertainty":{"low":0.03,"high":0.06},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/bipolar-disorder","title":"Bipolar Disorder — Statistics","publisher":"National Institute of Mental Health (NIMH)","source_type":"govt_report","statistic":"An estimated 4.4% of U.S. adults experience bipolar disorder at some time in their lives","excerpt":"\"An estimated 4.4% of U.S. adults experience bipolar disorder at some time in their lives. An estimated 2.8% of U.S. adults had bipolar disorder in the past year. Past year prevalence of bipolar disorder among adults was similar for males (2.9%) and females (2.8%). An estimated 82.9% of people with bipolar disorder had serious impairment, the highest percent serious impairment among mood disorders.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20260522234945/https://www.nimh.nih.gov/health/statistics/bipolar-disorder","calculation_notes":"NIMH's 4.4% lifetime prevalence figure is taken directly from this page as the point estimate. It is based on the NCS-R (Merikangas et al. 2007) and serves as the authoritative US government statistics citation for the headline. 4.4% = 0.044 = native numerator 44, denominator 1000. log10(0.044) = -1.3565.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/17485606/","title":"Lifetime and 12-Month Prevalence of Bipolar Spectrum Disorder in the National Comorbidity Survey Replication","publisher":"Archives of General Psychiatry (Merikangas, Akiskal, Angst, Greenberg, Hirschfeld, Petukhova, Kessler)","source_type":"peer_reviewed","statistic":"Lifetime prevalence: BP-I 1.0%, BP-II 1.1%, sub-threshold 2.4%, overall spectrum 4.4%","excerpt":"\"Lifetime (and 12-month) prevalence estimates are 1.0% (0.6%) for BP-I, 1.1% (0.8%) for BP-II, and 2.4% (1.4%) for sub-threshold BPD. Subthreshold BPD is common, clinically significant, and underdetected in treatment settings.\"\n","source_date":"2007-05-01","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20260419012232/https://pubmed.ncbi.nlm.nih.gov/17485606/","calculation_notes":"This is the primary academic source underlying the NIMH 4.4% figure. The NCS-R interviewed a nationally representative US adult sample (N=9,282). Component breakdown: BP-I (1.0%) + BP-II (1.1%) + sub-threshold (2.4%) = 4.5% (rounding to 4.4% in aggregate due to overlap exclusions in the survey methodology). The sub-threshold component alone accounts for 55% of the spectrum total, which is why restricting to BP-I+II yields 2.1%. For the purposes of this entry, the headline is 4.4% = 0.044 per the NCS-R spectrum.\n","independence_note":"Primary academic source. The NIMH statistics page draws directly from this paper. Not independent — treat NIMH citation as institutional endorsement of this figure.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3486639/","title":"Prevalence and Correlates of Bipolar Spectrum Disorder in the World Mental Health Survey Initiative","publisher":"Archives of General Psychiatry (Merikangas et al.)","source_type":"peer_reviewed","statistic":"Worldwide lifetime spectrum prevalence 2.4% across 11 countries; US prevalence highest at 4.4%","excerpt":"\"The aggregate lifetime prevalence of BP-I disorder was 0.6%, BP-II was 0.4%, subthreshold BP was 1.4%, and Bipolar Spectrum (BPS) was 2.4%. There was significant cross-national variation in lifetime prevalence of BPS, ranging from 0.1% in India to 4.4% in the United States.\"\n","source_date":"2011-09-01","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20260219064938/https://pmc.ncbi.nlm.nih.gov/articles/PMC3486639/","calculation_notes":"Cross-national validation of the US figure (N=61,392 across 11 countries). The worldwide aggregate spectrum prevalence of 2.4% is lower than the US 4.4%, illustrating that the US figure is at the high end of the international range. The US 4.4% in this multi-national study is consistent with the NCS-R figure, providing independent cross-national confirmation using the same diagnostic instrument (WMH-CIDI) in a different analytic context.\n","independence_note":"Merikangas is a co-author on both the NCS-R and WMH papers. The WMH is a different dataset (multi-national, N=61,392) from the NCS-R US sample (N=9,282), but uses the same CIDI instrument. Treat as partially independent confirmation.\n"}],"comparison_anchors":[{"label":"Developing major depressive disorder (lifetime, US adult)","lifetime_us_adult":0.174},{"label":"Developing PTSD (lifetime, US adult)","lifetime_us_adult":0.068},{"label":"Developing schizophrenia (lifetime, global)","lifetime_us_adult":0.007}],"personal_factor_multipliers":[{"factor":"one first-degree relative with bipolar disorder","multiplier":6,"notes":"First-degree relatives of someone with bipolar disorder carry approximately 25–30% lifetime spectrum risk vs the population baseline of 4.4%, yielding a multiplier of roughly 6× on the spectrum headline (Barnett & Smoller 2009, PMC3637882; NCS-R-derived recurrence risk ratios of 7–10 apply to the narrower BP-I+II baseline of ~1–2%, which is why the literature reports 7–10× on the narrow definition but only ~6× when the 4.4% spectrum baseline is used).\n"},{"factor":"monozygotic (identical) twin with bipolar disorder","multiplier":10,"notes":"MZ concordance 38.5–43% vs population risk ~4.4%, implying roughly 9–10× relative risk (Barnett & Smoller 2009, PMC3637882). Note: concordance is not the same as heritability; a 40% MZ concordance already implies substantial non-genetic variance.\n"},{"factor":"onset of first depressive episode before age 25","multiplier":2.5,"notes":"Earlier age of onset is a strong predictor of bipolar spectrum disorder vs unipolar depression; individuals presenting with depression before 25 have substantially higher rates of eventual bipolar reclassification per multiple longitudinal studies reviewed in NCS-R follow-up literature.\n"},{"factor":"female sex (bipolar II specifically)","multiplier":1.4,"notes":"Bipolar II and rapid-cycling are more common in women; women are ~1.4× more likely to receive a BP-II diagnosis than men per epidemiological reviews (Frontiers in Psychiatry 2022; doi:10.3389/fpsyt.2022.926594), though BP-I prevalence is approximately equal by sex.\n"},{"factor":"no first-degree family history","multiplier":0.3,"notes":"Absence of first-degree family history substantially reduces (but does not eliminate) genetic risk contribution; consistent with heritability of 60–85% leaving 15–40% environmental variance (PMC3637882).\n"}],"short_label":"Bipolar disorder","myth_framing":"underrated","outcome_severity":"serious_harm","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry covers the full bipolar spectrum (bipolar I, II, and sub-threshold), not only the classic bipolar I presentation characterised by full manic episodes. Restricting to BP-I+II yields a lifetime prevalence of approximately 2.1% (1 in 48), which is the more clinically familiar figure. Sub-threshold bipolar disorder meets some but not all DSM criteria for BP-I or BP-II and represents a clinically meaningful burden — 82.9% of people with any bipolar spectrum diagnosis have serious functional impairment per the NCS-R. The NCS-R data are from 2001–2003; subsequent surveys (NESARC-III, 2012–2013) report a 12-month BP-I prevalence of 2.1%, somewhat higher than the NCS-R's 0.6%, suggesting either increased recognition, diagnostic criteria broadening, or true prevalence increase, and introducing genuine uncertainty into the long-run lifetime estimate. Heritability is high (60–85%) but non-deterministic: MZ twin concordance is only ~40%, meaning the majority of identical twins of people with bipolar disorder do not develop it themselves. The 4.4% figure is the NIMH institutional headline and the entry's anchor; it is not a ceiling — broader spectrum definitions reach as high as 6% in some analyses.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-22","last_reviewed":"2026-05-22","reviewed":true,"generated_at":"2026-05-22","image":{"alt":"Two overlapping abstract waveforms on a muted warm background, one rising and one falling, flat vector illustration."},"canonical_url":"https://likelier.app/bipolar-disorder-lifetime","api_url":"https://likelier.app/api/fears/bipolar-disorder-lifetime.json"},{"slug":"wisdom-tooth-extraction-complications","question":"What are the odds of a serious complication (nerve injury, dry socket, infection) from wisdom tooth extraction?","quick_answer":"Any complication after wisdom-tooth removal runs about 1 in 22 (4.6%) per procedure, but that is mostly dry socket, a painful, self-limiting week. The permanent nerve injury people actually dread is far rarer, about 1 in 350 (0.28%) of lower-jaw extractions.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Wisdom-tooth removal is one of the most widely dreaded routine surgeries, and the dread is usually about worst-case stories: permanent numbness of the lip and chin, a jaw that never feels right again, an infection that lands someone in hospital. The horror-story framing collapses two very different things into one fear -- a painful, self-limiting recovery (swelling, a dry socket, a week of soft food) and a rare permanent injury. Most people asked to guess the odds of \"something going wrong\" will price in the permanent outcomes they have heard about, without separating how common a bad week is from how rare lasting harm is. There is no large survey quantifying the perceived probability of a wisdom-tooth complication as a standalone number, so the kind here is intuition rather than poll.\n","rough_estimate":"People commonly conflate a rough recovery (fairly common) with permanent harm (rare); asked for the odds of 'serious' damage, many overestimate","kind":"intuition"},"native":{"display":"~1 in 22 (4.6%) any complication per patient; permanent nerve injury far rarer at ~1 in 350 (0.28%)","numerator":46,"denominator":1000,"unit":"per patient undergoing third molar (wisdom tooth) extraction","population":"Patients undergoing third molar extraction; the complication burden is concentrated in impacted lower (mandibular) thirds"},"normalized":{"lifetime_us_adult":0.046,"display":"~1 in 22 per procedure (activity-specific)","log_value":-1.34,"assumptions":"The headline is the overall post-operative complication rate per patient, anchored to Bui, Seldin and Dodson (J Oral Maxillofac Surg, 2003), a retrospective cohort of 583 patients / 1,597 third molars that reported a 4.6% overall complication rate. A large Taiwanese nationwide cohort (Chen, Chi and Lee, PLOS One 2021; 16,609 impacted lower thirds) corroborates the order of magnitude with a 4.2% cumulative rate, though it counts a narrower set (dry socket 3.66%, prolonged TMJ symptoms 0.41%, surgical-site infection 0.17%) and excludes nerve injury. These figures are NOT additive: they use different denominators and count different complications. Dry socket (alveolar osteitis) is the dominant, largely self-limiting complication; permanent inferior alveolar nerve injury -- the outcome most people actually fear -- is a much rarer subset, 0.28% of mandibular extractions in the Kang, Sah and Fei systematic review (26,427 patients / 44,171 teeth), with transient nerve deficit at 1.20%. Scope is activity-specific: one procedure, per patient. The wide uncertainty band (0.02-0.30) reflects the impaction-difficulty spectrum: a simple erupted upper third sits near 2%, while a difficult full-bony mandibular impaction in a smoker can push dry-socket rates to 20-30%. The point estimate sits near the low end because it is a population average across all extractions, not the difficult-impaction subgroup.\n","uncertainty":{"low":0.02,"high":0.3},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/14663801/","title":"Types, frequencies, and risk factors for complications after third molar extraction","publisher":"Journal of Oral and Maxillofacial Surgery (Bui CH, Seldin EB, Dodson TB. 2003;61(12):1379-89)","source_type":"peer_reviewed","statistic":"Overall complication rate 4.6% in 583 patients / 1,597 third molars removed; increasing age, positive medical history, and M3 position relative to the inferior alveolar nerve raised risk","excerpt":"\"The study sample was composed of 583 patients (57.0% male) with a mean age of 26.4 +/- 8.4 years. The overall complication rate was 4.6%. Increasing age, a positive medical history, and the position of the M3 relative to the inferior alveolar nerve were associated with an increased risk for complications.\" [Verbatim abstract sentences; the \"1,597 third molars\" figure appears in the paper's methods, not this abstract excerpt.]\n","source_date":"2003-12-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260609013850/https://pubmed.ncbi.nlm.nih.gov/14663801/","calculation_notes":"This 4.6% overall per-patient complication rate is the direct anchor for the native numerator (46/1000) and the normalized point estimate (0.046). It counts all reported complications across all third molars (upper and lower), so it is the broadest single composite available. It is not summed with the nerve-injury or dry-socket figures from the other sources -- those measure subsets of this same overall burden, on different denominators.\n"},{"url":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0246625","title":"Revisit incidence of complications after impacted mandibular third molar extraction: A nationwide population-based cohort study","publisher":"PLOS One (Chen YW, Chi LY, Lee CY. 2021;16(2):e0246625)","source_type":"primary_study","statistic":"Cumulative complication rate 4.2% among 16,609 patients who had impacted mandibular third molars extracted: dry socket 3.66%, prolonged TMJ symptoms 0.41%, surgical-site infection 0.17%","excerpt":"\"The overall cumulative complication rate for iLM3 extraction was 4.2%. The incidence of DS [dry socket] was the highest (3.66%) [...] About 4 out of 1000 patients had temporomandibular joint symptoms and sought for treatment (0.41%) [...] the incidence of SSI [surgical-site infection] was very low (0.17%).\" Among the 16,609 patients, aging was not found to be associated with any complication.\n","source_date":"2021-02-10","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20230105215829/http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0246625","calculation_notes":"An independent nationwide-cohort corroboration of the ~4-5% overall complication order of magnitude, from a different country and a much larger sample. It counts only three complication types on impacted lower thirds and excludes nerve injury, so its 4.2% is not directly comparable to Bui's 4.6% -- the convergence is partly coincidental across different composites. Used to bound the central estimate and to establish that dry socket, not nerve injury, dominates the complication count.\n","independence_note":"Taiwan National Health Insurance population database; methodologically and geographically independent of the US clinic-based Bui/Seldin/Dodson cohort.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/31476533/","title":"Determining the risk relationship associated with inferior alveolar nerve injury following removal of mandibular third molar teeth: A systematic review","publisher":"Journal of Stomatology, Oral and Maxillofacial Surgery (Kang F, Sah MK, Fei G. 2020)","source_type":"peer_reviewed","statistic":"Across 23 studies (26,427 patients, 44,171 teeth), 1.20% developed transient inferior alveolar nerve deficit and 0.28% developed permanent deficit","excerpt":"\"Twenty-three studies out of 693 articles were finally included, summing a total of 26,427 patients (44,171 teeth). 1.20% developed transient IAN deficit and 0.28% developed permanent IAN deficit respectively. Depth of impaction, proximity to the mandibular canal, surgical technique, intra-operative nerve exposure, and surgeon experience were statistically significant contributing factors.\" [Verbatim abstract.]\n","source_date":"2020-08-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20231015130529/https://pubmed.ncbi.nlm.nih.gov/31476533/","calculation_notes":"Supplies the \"serious\" endpoint that the headline composite understates: permanent inferior alveolar nerve injury at 0.28% (~1 in 350) and transient deficit at 1.20% (~1 in 83) of mandibular extractions. These are the numbers that answer the \"serious\" half of the question and are decomposed from the overall rate in the prose, not added to it.\n","independence_note":"Systematic review pooling 23 studies; independent of the two single-cohort sources and focused specifically on the nerve-injury endpoint.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9317683/","title":"Smoking as a Risk Factor for Dry Socket: A Systematic Review","publisher":"Dentistry Journal (Kusnierek W, et al. 2022;10(7):121)","source_type":"peer_reviewed","statistic":"Combined dry-socket incidence 13.2% in smokers vs 3.8% in non-smokers across 11 studies (10,195 participants); tobacco smokers had a more than three-fold increase in odds","excerpt":"\"Eleven studies were included in this systematic review [...] a total of 10195 participants (including 3007 smokers and 7188 non-smokers). The combined incidence of dry socket in smokers was found to be about 13.2% and in non-smokers about 3.8%. Tobacco smokers had a more than three-fold increase in the odds of dry socket after tooth extraction.\" [Verbatim from open-access text.]\n","source_date":"2022-06-29","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260613121918/https://pmc.ncbi.nlm.nih.gov/articles/PMC9317683/","calculation_notes":"Anchors the smoking multiplier. The dry-socket-specific ratio is ~3.5x (13.2 / 3.8), but the multiplier is applied to the OVERALL complication probability, not to dry socket alone. Because dry socket is the dominant but not sole component (~3.6% dry socket + ~1% other complications in the Chen cohort), a ~3.5x rise in the dominant term yields roughly (3.6 x 3.5 + 1) / 4.6 ~= 3.0x overall -- which is why the multiplier is set to 3.0x, not 3.5x. Applies to dry socket specifically, not nerve injury.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27017181/","title":"Risk assessment and sensitivity meta-analysis of alveolar osteitis occurrence in oral contraceptive users","publisher":"Journal of the American Dental Association (Bienek DR, Filliben JJ. 2016;147(6):394-404)","source_type":"peer_reviewed","statistic":"Among females, oral contraceptive use increased average alveolar-osteitis risk nearly 2-fold (13.9% vs 7.5%)","excerpt":"\"Among females, OC use significantly increased (P <= .05) the average risk of AO occurrence by nearly 2-fold (13.9% versus 7.5%). The risk ratio of AO in females not using OCs was 1.2 greater (P <= .05) than that in males.\" [Verbatim abstract sentences.]\n","source_date":"2016-06-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20250204234606/https://pubmed.ncbi.nlm.nih.gov/27017181/","calculation_notes":"Anchors the oral-contraceptive multiplier (13.9 / 7.5 = ~1.85, rounded to 1.9x) for the dry-socket component in female patients. Applies to dry socket, the dominant complication.\n","independence_note":"Meta-analysis specific to oral-contraceptive exposure; independent of the smoking and nerve-injury reviews.\n"},{"url":"https://www.ejomr.org/JOMR/archives/2014/4/e1/v5n4e1ht.htm","title":"Inferior Alveolar Nerve Injury after Mandibular Third Molar Extraction: a Literature Review","publisher":"Journal of Oral and Maxillofacial Research (Sarikov R, Juodzbalys G. 2014;5(4):e1)","source_type":"peer_reviewed","statistic":"IAN neurosensory deficit highest with horizontal impaction (4.7%) vs vertical (0.9%); highest (11%) when the nerve is radiographically 'intimate' with the tooth","excerpt":"\"The incidence of IAN neurosensory deficit was highest with horizontal impaction (4.7%) and lowest when the teeth were vertically impacted (0.9%). The lowest incidences were seen when the nerve was either 'distant' (0.8%) or 'close' (0.9%), and highest when the nerve was classed as 'intimate' (11%). [...] IAN paresthesia occurs widely from 0.35% to 8.4%.\" [Verbatim from open-access review; the impaction/proximity figures are attributed by Sarikov and Juodzbalys 2014 to Smith 2013, and are cited here as such.]\n","source_date":"2014-12-29","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260202220236/https://www.ejomr.org/JOMR/archives/2014/4/e1/v5n4e1ht.htm","calculation_notes":"Anchors the impaction-depth / canal-proximity multiplier for the nerve-injury endpoint: horizontal vs vertical impaction is a ~5x gradient (4.7% vs 0.9%), and radiographic \"intimate\" canal contact reaches 11%. Applied conservatively as a ~5x factor on the nerve-injury risk, which is the serious tail of the composite rather than the dominant dry-socket component.\n","independence_note":"Open-access narrative review; the anatomical-risk figures derive from Smith (2013) and are cited as such rather than as this review's own primary data.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/17719384/","title":"Age as a risk factor for third molar surgery complications","publisher":"Journal of Oral and Maxillofacial Surgery (Chuang SK, Perrott DH, Susarla SM, Dodson TB. 2007;65(9):1685-92)","source_type":"peer_reviewed","statistic":"In a prospective cohort of 4,004 subjects / 8,748 third molars (overall complication rate 19%), age above 25 years was associated with an increased risk of complications (odds ratio 1.5; P = .05)","excerpt":"\"The study sample was comprised of 4,004 subjects having a total of 8,748 M3s removed. The mean age was 39.8 +/- 13.6 years [...] The overall complication rate was 19%. In the multiple regression model, age above 25 years was associated with an increased risk of complications (odds ratio = 1.5; P = .05).\" [Verbatim abstract sentences.]\n","source_date":"2007-09-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20250909145216/https://pubmed.ncbi.nlm.nih.gov/17719384/","calculation_notes":"Names the source and effect size (OR ~1.5) for the \"age over 25 at extraction\" multiplier. This is a prospective cohort with a much older mean age than Bui's, so its age gradient is well powered. Note the evidence is mixed: Chen et al. 2021 (16,609 patients) found no age association, so the 1.5x factor is applied cautiously as an upper-plausible rather than settled effect.\n","independence_note":"Distinct US prospective cohort (Chuang/Dodson group); the \"age over 25\" effect it supplies is independent of the Bui and Chen cohorts, which disagree on age.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/18940482/","title":"Risk factors for inflammatory complications following third molar surgery in adults","publisher":"Journal of Oral and Maxillofacial Surgery (Chuang SK, Perrott DH, Susarla SM, Dodson TB. 2008;66(11):2213-8)","source_type":"peer_reviewed","statistic":"In the same prospective cohort (4,004 subjects / 8,748 third molars), the adjusted odds ratio for any inflammatory complication rose with impaction depth: soft-tissue OR 2.5 (95% CI 1.7-3.7), partial-bony OR 4.7 (95% CI 3.6-6.1), full-bony OR 6.0 (95% CI 4.7-7.7), all P < .0001, relative to erupted teeth","excerpt":"\"4,004 subjects with a mean age of 39.8 +/- 13.6 years having 8,748 M3s extracted [...] In the multiple regression model, impaction level was associated with an increased risk of inflammatory complications: soft tissue impaction (OR = 2.5; 95% CI, 1.7, 3.7), partial bony impaction (OR = 4.7; 95% CI, 3.6, 6.1), and full bony impaction (OR = 6.0; 95% CI, 4.7, 7.7).\" [Verbatim/near-verbatim abstract figures.]\n","source_date":"2008-11-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20240816142549/https://pubmed.ncbi.nlm.nih.gov/18940482/","calculation_notes":"The same-denominator, overall-complication grounding for the deep-impaction multiplier. Because these ORs are for ANY inflammatory complication on one cohort (not a nerve-only or dry-socket-only subset), they avoid the cross-denominator mixing the assumptions block warns against. Reference category is erupted teeth; the site headline (4.6%) is an all-impaction-types average, so the average-to-full-bony multiplier is smaller than the raw 6.0 (see the multiplier note).\n","independence_note":"Companion analysis of the same Chuang/Dodson cohort as the 2007 age paper; cited for the impaction-depth gradient, not treated as an independent sample.\n"}],"comparison_anchors":[{"label":"Kidney stone (lifetime, US adult)","lifetime_us_adult":0.09},{"label":"Dental tourism complication (per major-work trip)","lifetime_us_adult":0.05},{"label":"Food poisoning (per year, US)","lifetime_us_adult":0.15}],"personal_factor_multipliers":[{"factor":"Current smoker","multiplier":3,"notes":"Dry-socket incidence ~13.2% in smokers vs ~3.8% in non-smokers (Kusnierek et al. 2022, systematic review; >3-fold odds). Because dry socket dominates the composite complication count, smoking roughly triples the overall per-procedure complication probability. Applies to dry socket, not nerve injury.\n"},{"factor":"Deep horizontal / full-bony impaction with mandibular-canal contact on imaging","multiplier":2.5,"notes":"Grounded on Chuang et al. 2008 (same 4,004-subject cohort), which reports OVERALL inflammatory-complication odds ratios by impaction depth on one denominator: full-bony OR 6.0, partial-bony 4.7, soft-tissue 2.5, all relative to erupted teeth. Those ORs are vs an erupted reference, whereas the site headline (4.6%) is an all-impaction-types average that already includes impacted molars, so the average-to-full-bony shift is smaller than 6.0. Set at ~2.5x as an AUTHOR ESTIMATE bridging the erupted-reference ORs to the all-types headline (roughly full-bony-vs-average, given the average sits near the partial-bony OR of 4.7). The nerve-injury endpoint alone shifts ~5x with horizontal impaction and canal contact (Sarikov and Juodzbalys 2014, citing Smith 2013), so 2.5x is a conservative floor on the overall shift. Flagged for human review before publication -- this value is not a direct single-source figure.\n"},{"factor":"Simple erupted / soft-tissue-only extraction (no bone removal)","multiplier":0.4,"notes":"The complication burden is concentrated in surgical bony impactions of the lower jaw; simple erupted extractions (typically upper thirds) carry dry-socket and nerve-injury rates well below the all-comers average. This 0.4x is a DERIVED estimate, not a single published ratio: Chen et al. 2021 restricts its 4.2% to impacted mandibular thirds, while routine non-surgical extraction dry socket runs ~1-3%; ~1.5-2% against the ~4.6% headline implies roughly 0.3-0.4x. Treat it as an order-of-magnitude \"protective\" marker rather than a precise multiplier.\n"},{"factor":"Oral contraceptive use (female patients)","multiplier":1.7,"notes":"OC use raises alveolar-osteitis (dry-socket) risk nearly 2-fold (13.9% vs 7.5% = 1.85x; Bienek and Filliben 2016 meta-analysis). To express this as an overall multiplier, weight by dry socket's share of the composite WITHIN one source: in Chen et al. 2021 dry socket is 3.66 of the 4.2-point mandibular composite (~87%). Scaling that 87% by 1.85 and holding the ~13% non-dry-socket remainder at 1x gives 0.87 x 1.85 + 0.13 ~= 1.7x. This weights a component effect by its within-Chen share rather than summing absolute rates across denominators. The dry-socket-specific risk for female patients on OCs is the full ~1.9x; value flagged for human review before publication.\n"},{"factor":"Age over 25 at extraction","multiplier":1.5,"notes":"Age above 25 was associated with an increased complication rate (odds ratio ~1.5, P = .05; Chuang et al. 2007, prospective cohort of 4,004 subjects). Older roots are more fully formed and bone is denser, making surgical extraction harder; Bui et al. 2003 independently found increasing age a risk factor. Evidence is mixed, however: the large Chen et al. 2021 cohort (16,609 patients) found no age association, so treat 1.5x as an upper-plausible estimate rather than a settled effect.\n"}],"short_label":"Wisdom-tooth surgery","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"\"Complication\" here means any adverse post-operative event: dry socket (alveolar osteitis), surgical-site infection, prolonged jaw-joint symptoms, transient or permanent nerve injury, bleeding, or delayed healing. These are not one thing and not equally serious. The ~4.6% headline is dominated by dry socket, which is painful but self-limiting and resolves within days to a couple of weeks with dressing changes. The outcome most people fear -- permanent loss of sensation in the lip, chin, or tongue from inferior alveolar or lingual nerve injury -- is far rarer, at roughly 0.28% (about 1 in 350) of lower-jaw extractions, with transient deficit around 1.20%. The figures from different studies are not additive: Bui et al. (4.6%, all thirds), Chen et al. (4.2%, impacted lower thirds, excluding nerve injury), and Kang et al. (nerve injury only) use different denominators and count different events. The true rate for any individual depends heavily on which tooth, how it is impacted, how close it sits to the nerve canal, the surgeon's experience, and smoking status -- which is why the uncertainty band spans 2% to 30%. Upper (maxillary) thirds carry lower nerve-injury risk than lower (mandibular) thirds because the inferior alveolar canal runs through the lower jaw. This entry does not cover prophylactic-versus-symptomatic extraction decisions or the separate question of whether asymptomatic wisdom teeth should be removed at all.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-07-14","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-14","last_reviewed":"2026-07-14","reviewed":true,"generated_at":"2026-07-14","image":{"alt":"A flat vector illustration of a single stylized molar tooth outline with two roots, muted grey and warm tones, calm and abstract."},"canonical_url":"https://likelier.app/wisdom-tooth-extraction-complications","api_url":"https://likelier.app/api/fears/wisdom-tooth-extraction-complications.json"},{"slug":"broken-toy-child-injury","question":"What are the odds of a child needing an ER visit for a toy-related injury?","quick_answer":"About 5% of children — roughly 1 in 20 — will need an emergency-room visit for a toy-related injury by age 15, which works out to around 0.33% in any given year.\n","category":"kids","tags":["child","household"],"no_reliable_estimate":false,"perceived":{"description":"The mental image is vivid: a plastic toy snapped in half with a jagged edge, a small figure with a protruding metal pin, a wheel that cracked and left a sharp spoke. Parents who enforce \"no broken toys\" rules often cite a confident hazard intuition — that a child playing with a damaged toy is meaningfully more likely to get cut, punctured, or otherwise hurt than with an intact one. The intuitive risk estimate tends to sit in \"high chance of a cut, low chance of anything serious\" territory, which is roughly where the data lands for toy injuries overall. What parents often miss is how the risk is distributed: the injury burden from toys is dominated by falls and collisions from riding toys and scooters, not lacerations from sharp edges — meaning \"no broken toys\" addresses the most viscerally salient hazard while leaving the statistically larger injury drivers unaddressed.\n","kind":"intuition"},"native":{"display":"~0.33% per year (toy-related ER visits, children under 12, US 2023)","numerator":33,"denominator":10000,"unit":"per year","population":"US children under age 12 (CPSC NEISS estimate, calendar year 2023)"},"normalized":{"lifetime_us_adult":0.048,"display":"~5% chance of a toy-related ER visit by age 15","log_value":-1.32,"assumptions":"The Consumer Product Safety Commission estimates 154,700 children under age 12 were treated in US emergency departments for toy-related injuries in calendar year 2023, based on the National Electronic Injury Surveillance System (NEISS) probability sample. The US Census Bureau / Federal Interagency Forum on Child and Family Statistics (childstats.gov) places the child population under age 12 at approximately 47.1 million in 2022–2023. Dividing gives an annual ER-visit rate of 154,700 / 47,100,000 ≈ 0.328% per year. Over 15 years of typical toy exposure (birth through age 14), the cumulative probability of at least one toy-related ER visit is 1 − (1 − 0.00328)^15 ≈ 4.8%, assuming independence across years and a roughly constant annual rate.\nThe CPSC does not publish a separate statistic for \"broken or damaged toy\" injuries; the NEISS system codes by product category and injury diagnosis, not by whether the toy was functioning as designed. Lacerations are consistently the single most common injury diagnosis in toy-related ER visits (~22–24% of visits across years 2010–2023) and represent the closest available proxy for broken- or sharp-toy hazard injuries, but they include cuts from intact-but-sharp toys as well.\nThe normalized.lifetime_us_adult field is set to 0.048 to reflect the cumulative childhood probability (birth to age 15); scope is subgroup_lifetime (US children, not the general adult population).\nUncertainty bounds [0.030, 0.090] reflect compounded sources of variability: (1) NEISS sampling uncertainty — national estimates for this injury category carry a coefficient of variation of roughly ±15%, yielding an annual rate range of approximately 0.28–0.38%; (2) denominator year mismatch — the 2022 Census child population is applied to 2023 injury counts, introducing additional noise; (3) constant-rate assumption — assuming a fixed 0.33% rate across all ages 0–14 understates the substantially higher risk in ages 0–4 (children 4 and under sustained 36% of all-ages toy injuries in 2023) and overstates it for older children. Together, these stack to a ~3x band (0.030 / 0.090) around the point estimate, which is the minimum defensible width for a probability-sample surveillance estimate projected over 15 years with age-heterogeneous risk.\n","uncertainty":{"low":0.03,"high":0.09},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cpsc.gov/Research--Statistics/Toys-and-Childrens-Products/Toy-Related-Deaths-and-Injuries-Calendar-Year-2023","title":"Toy-Related Deaths and Injuries: Calendar Year 2023","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"An estimated 231,700 toy-related injuries treated in US EDs in 2023; 154,700 among children under 12; 167,500 among children under 15; 10 toy-related deaths in children; lacerations remain the leading injury diagnosis","excerpt":"\"An estimated 231,700 toy-related injuries were treated in hospital emergency departments (EDs) in the United States during 2023. Of those, 154,700 (67%) involved children 12 years of age and under, and 167,500 (72%) involved children under 15 years of age.\"\n","source_date":"2024-11-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260504054130/https://www.cpsc.gov/Research--Statistics/Toys-and-Childrens-Products/Toy-Related-Deaths-and-Injuries-Calendar-Year-2023","calculation_notes":"CPSC NEISS (National Electronic Injury Surveillance System) is a probability-based sample of approximately 100 hospital emergency departments that is used to generate national injury estimates. The 154,700 figure for children under 12 is the NEISS-based national estimate for calendar year 2023. Dividing by the ~47.1 million US children under 12 (childstats.gov 2022 data) gives an annual rate of approximately 0.33% per year. The report notes that 92% of injured patients were treated and released; 10 deaths were recorded, all involving aspiration of small parts, drowning from flotation toys, or toy chest entrapment. The report does not separately enumerate injuries caused by broken or defective toys, as NEISS codes by product type and injury diagnosis rather than product condition.\n","independence_note":"Primary government surveillance source. NEISS data is collected independently of the peer-reviewed fractures study (Halperin et al. 2022) and the childstats.gov population denominator source below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8896474/","title":"Characterization of More Than a Third of a Million Toy-Related Fractures","publisher":"Journal of the American Academy of Orthopaedic Surgeons — Global Research & Reviews (Halperin SJ, Prenner S, Moore HG, Grauer JN)","source_type":"peer_reviewed","statistic":"347,135 toy-related fractures identified from NEISS 1999–2018; 237,754 (68%) in patients under 18; annual rate nearly doubled from ~12,002 to ~23,296 over the study period; ball-related injuries accounted for 45% of pediatric fractures","excerpt":"\"In total, 347,135 toy-related fractures were identified from NEISS between 1999 and 2018. Of these, 237,754 (68%) occurred in patients under 18 years of age. The annual incidence increased from approximately 12,002 in 1999 to 23,296 in 2018, representing a 94% increase over the 20-year study period.\"\n","source_date":"2022-03-03","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505050539/https://pmc.ncbi.nlm.nih.gov/articles/PMC8896474/","calculation_notes":"Halperin et al. used the CPSC NEISS database to characterize toy-related fractures specifically. Fractures represent a more severe injury subset than the broader ER-visit pool. The near-doubling of fracture rates over 1999–2018 largely reflects the rise of foot-powered scooters, which dominate the riding-toy injury category. Ball-related injuries (45% of pediatric fractures) and riding toys (11%) together account for the majority, confirming that the largest injury burden from toys comes from kinetic/collision mechanisms, not from sharp or broken toy edges. This source independently corroborates the CPSC NEISS surveillance methodology and severity distribution.\n","independence_note":"Independent of the CPSC annual report and the childstats.gov population source; different research group (Yale), peer-reviewed journal, 20-year longitudinal NEISS analysis.\n"},{"url":"https://www.childstats.gov/americaschildren/tables/pop1.asp","title":"America's Children: Key National Indicators of Well-Being — Table POP1: US Child Population by Age","publisher":"Federal Interagency Forum on Child and Family Statistics (childstats.gov), citing U.S. Census Bureau","source_type":"govt_report","statistic":"Ages 0–5: 22.6 million; ages 6–11: 24.5 million; total under 12: ~47.1 million (2022 data)","excerpt":"Table POP1 shows: children ages 0–5: 22.6 million (2022); ages 6–11: 24.5 million (2022); combined total under age 12: approximately 47.1 million. Data from U.S. Census Bureau as compiled by the Federal Interagency Forum on Child and Family Statistics.\n","source_date":"2023-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260417184952/https://www.childstats.gov/americaschildren/tables/pop1.asp","calculation_notes":"This source provides the denominator for the annual ER-visit rate calculation: 154,700 (CPSC NEISS 2023) / 47,100,000 (Census/childstats 2022) = 0.00328 per year ≈ 0.33%. The 2022 Census figure is used with the 2023 CPSC injury count as the most closely matched available pairing. Population of children ages 0–14 is approximately 55–56 million (adding the 12–17 age band's first two years), used for the under-15 rate variant.\n","independence_note":"Federal statistical source independent of the CPSC surveillance data and the peer-reviewed fractures study; provides the population denominator necessary to convert raw injury counts to per-child annual rates.\n"}],"comparison_anchors":[{"label":"Toy-related laceration ER visit (children under 12, annual)","lifetime_us_adult":0.012},{"label":"Toy-related fatality (US children, lifetime by age 15)","lifetime_us_adult":0.00003},{"label":"Playground equipment injury ER visit (children, lifetime by age 15)","lifetime_us_adult":0.055}],"personal_factor_multipliers":[{"factor":"Child under age 5","multiplier":1.6,"notes":"Children 4 and under sustained an estimated 36% of all toy-related ER injuries across all ages in 2023 (CPSC 2023: 83,800 of 231,700), a disproportionate share for such a young age band. Higher injury rate reflects developmental factors: less coordination, more mouthing behavior, and greater proportion of time in contact with toys."},{"factor":"Toy with small parts given to child under 3","multiplier":3,"notes":"Small parts are both a choking hazard (primary fatality mechanism in CPSC 2023 data) and a sharp-edge risk when fractured. All 10 toy-related deaths in 2023 involved aspiration of small parts, drowning, or entrapment. The relevant hazard shifts from laceration to airway obstruction for the youngest children."},{"factor":"Riding toy or scooter user (ages 5–12)","multiplier":4,"notes":"Riding toys — foot-powered scooters, skateboards, toy bikes — are the dominant driver of toy-related fractures and high-severity ER visits. The Halperin et al. (2022) data and multiple CPSC annual reports consistently identify this category as the highest injury mechanism. Broken toy edges are not the primary risk in this subgroup; kinetic energy is."}],"short_label":"Toy injury requiring ER (child)","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The 0.33% annual rate covers all toy-related ER visits, not injuries from broken or damaged toys specifically. The CPSC NEISS system records product type (toy) and injury diagnosis (laceration, fracture, contusion, etc.) but does not distinguish between intact toys used normally and physically broken or defective toys. Lacerations (~22–24% of toy ER visits annually) are the closest available proxy for broken- or sharp-toy injuries, but the laceration category includes cuts from intact-but-sharp items such as skateboard falls and metal-edged toy components, not only broken pieces. There is no peer-reviewed study that isolates the injury rate attributable specifically to physically damaged or cracked toys.\nNEISS estimates carry statistical uncertainty typical of probability-based surveillance samples; the CPSC reports confidence intervals, and year-to-year variation in the national estimate is meaningful. The 2023 figure of 154,700 represents a roughly 8% decrease from the 167,800 estimate for children under 13 reported for 2016, suggesting a modestly declining trend. The 4.8% lifetime estimate assumes a constant annual rate across all ages 0–14, which understates risk in early childhood (peak injury years) and overstates it in later childhood.\nThe severity distribution matters: 92% of toy-related ER visits end in treatment and release. Fatal toy injuries in children (10 in 2023) are extremely rare and arise almost exclusively from aspiration of small parts, drowning linked to flotation devices, and toy chest entrapment, not from lacerations caused by broken toy edges.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":2,"d5":4,"d6":4,"d7":4,"d8":4,"avg":3.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A cracked plastic toy car with a jagged break line, lying on a wooden floor beside intact toy blocks, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/broken-toy-child-injury","api_url":"https://likelier.app/api/fears/broken-toy-child-injury.json"},{"slug":"red-meat-colorectal-cancer","question":"How much does eating red or processed meat every day actually raise your colorectal cancer risk?","quick_answer":"Eating processed meat daily raises lifetime colorectal-cancer risk from about 4.1% to roughly 4.8% — a real but small increase. Many people, alarmed by the 'Group 1 carcinogen' label, overestimate the absolute risk by tenfold or more.\n","category":"cancer","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The WHO's 2015 announcement that processed meat is a \"Group 1 carcinogen\" -- the same category as tobacco smoking and asbestos -- triggered widespread alarm. Many people interpreted this as meaning a daily bacon sandwich is roughly as dangerous as a pack of cigarettes. Headlines amplified the framing, and surveys consistently show consumers overestimate the absolute cancer risk from moderate meat intake by an order of magnitude or more.\n","rough_estimate":"47% of US adults rank cancer-causing chemicals in food among their top-3 food safety concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 47% rank cancer-causing chemicals in food as a top-3 concern; IARC Group 1 classification of processed meat places it squarely here","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"native":{"display":"RR 1.18 per 50 g/day processed meat","numerator":118,"denominator":10000,"unit":"relative risk per daily-50g-serving vs non-consumers","population":"adults consuming processed meat daily"},"normalized":{"lifetime_us_adult":0.048,"display":"~4.8% lifetime CRC risk for daily processed-meat consumers (vs ~4.1% baseline)","log_value":-1.32,"assumptions":"US baseline lifetime colorectal cancer risk is approximately 4.1% (roughly 1 in 24). Baseline 4.1% US lifetime CRC risk is from SEER Cancer Stat Facts: Colorectal Cancer (NCI, seer.cancer.gov). Applying the IARC-derived RR of 1.18 for daily 50 g processed meat consumption gives 4.1% x 1.18 = ~4.8%. This is an absolute increase of roughly 0.7 percentage points. The relative risk estimate comes from the IARC 2015 evaluation and is consistent with the Farvid et al. 2021 meta-analysis (RR 1.17, 95% CI 1.08-1.26 for highest vs lowest consumption of red and processed meat combined).\n","uncertainty":{"low":0.033,"high":0.065},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/questions-and-answers/item/cancer-carcinogenicity-of-the-consumption-of-red-meat-and-processed-meat","title":"Cancer: Carcinogenicity of the consumption of red meat and processed meat","publisher":"World Health Organization / IARC","source_type":"reputable_reference","statistic":"Every 50 g portion of processed meat eaten daily increases colorectal cancer risk by about 18% (RR 1.18)","excerpt":"\"Every 50 gram portion of processed meat eaten daily increases the risk of colorectal cancer by about 18%.\"\n","source_date":"2015-10-26","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182213/https://www.who.int/news-room/questions-and-answers/item/cancer-carcinogenicity-of-the-consumption-of-red-meat-and-processed-meat","calculation_notes":"The 18% figure is a relative risk increase (RR 1.18). Applied to the US baseline lifetime CRC risk of ~4.1% (SEER data): 4.1% x 1.18 = 4.84%, rounded to ~4.8%. Absolute increase is ~0.7 percentage points. The WHO page explicitly notes that Group 1 classification reflects strength of evidence, not magnitude of risk.\n","independence_note":"IARC's evaluation synthesized multiple cohort studies and meta-analyses; the Farvid 2021 meta-analysis below partially overlaps in source studies but uses independent methodology and inclusion criteria.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/34455534/","title":"Consumption of red meat and processed meat and cancer incidence: a systematic review and meta-analysis of prospective studies","publisher":"European Journal of Epidemiology (Farvid et al.)","source_type":"peer_reviewed","statistic":"Red and processed meat combined: RR 1.17 (95% CI 1.08-1.26) for colorectal cancer, highest vs lowest consumption","excerpt":"\"Red meat consumption was significantly associated with greater risk of colorectal cancer (RR = 1.10; 95% CI = 1.03-1.17).\" Combined red and processed meat: \"colorectal cancer (RR = 1.17; 95% CI = 1.08-1.26).\"\n","source_date":"2021-08-29","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182257/https://pubmed.ncbi.nlm.nih.gov/34455534/","calculation_notes":"Farvid et al. report highest-vs-lowest consumption category comparisons across prospective cohort studies. The RR of 1.17 for combined meat is consistent with the IARC 1.18 per-50g figure. Both confirm that the relative risk is modest -- nothing close to the 20-25x RR of smoking for lung cancer.\n","independence_note":"This meta-analysis was published six years after the IARC evaluation and includes additional prospective studies not available to the 2015 working group.\n"},{"url":"https://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(15)00444-1/fulltext","title":"Carcinogenicity of consumption of red and processed meat","publisher":"The Lancet Oncology / IARC Working Group","source_type":"peer_reviewed","statistic":"IARC Working Group: processed meat classified Group 1 (carcinogenic to humans); red meat Group 2A (probably carcinogenic). Meta-analysis of 10 cohort studies: +18% colorectal cancer risk per 50g processed meat daily; 7 cohort studies: +17% risk per 100g red meat daily.","excerpt":"\"On the basis of the large amount of data and the consistent associations of colorectal cancer with consumption of processed meat across studies in different populations, which make chance, bias, and confounding unlikely as explanations, a majority of the Working Group concluded that there is sufficient evidence in human beings for the carcinogenicity of the consumption of processed meat.\"\n","source_date":"2015-10-26","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250727094603/https://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(15)00444-1/fulltext","calculation_notes":"The Bouvard/IARC Working Group paper is the authoritative peer-reviewed source behind the Group 1/2A classifications. The pooled relative risk of 1.18 per 50g/day processed meat consumption is the figure used for the native relative-risk encoding in this entry. Serves as the primary evidence basis rather than the Farvid meta-analysis (which extends the IARC analysis with more recent cohorts).\n","independence_note":"Bouvard/IARC synthesized 10 cohort studies for processed meat and 7 for red meat; Farvid 2021 includes many of the same cohorts but extends with additional follow-up and newer studies (EPIC, NHS II). Partial overlap in the underlying cohort data but methodologically distinct pooled analyses.\n"},{"url":"https://www.cdc.gov/tobacco/about/cigarettes-and-cancer.html","title":"Health Effects of Cigarettes: Cancer","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"People who smoke increase their risk of developing lung cancer by about 25 times that of people who don't smoke","excerpt":"\"People who smoke increase their risk of developing lung cancer by about 25 times that of people who don't smoke.\"\n","source_date":"2024-09-17","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260704050846/https://www.cdc.gov/tobacco/about/cigarettes-and-cancer.html","calculation_notes":"Grounds the body text's and caveats' contrast between tobacco's large relative-risk effect on lung cancer and processed meat's much smaller RR 1.18 effect on colorectal cancer. Not used in this entry's own CRC probability calculation, which rests on the IARC/WHO and Farvid sources above; cited solely to support the smoking-magnitude comparison.\n","independence_note":"CDC's tobacco control program page is independent of the IARC, WHO, and Farvid et al. sources used for this entry's own red/processed-meat relative-risk figures.\n"}],"comparison_anchors":[{"label":"Smoking → lung cancer (lifetime, heavy smoker)","lifetime_us_adult":0.23},{"label":"Baseline colorectal cancer (lifetime, US adult)","lifetime_us_adult":0.041},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"US adult (SEER baseline 4.1% + daily processed-meat consumer)","probability":0.048,"notes":"baseline 4.1% × RR 1.18 ≈ 4.8% lifetime diagnosis"},{"region":"US adult (baseline 4.1% + minimal red/processed meat)","probability":0.035,"notes":"baseline 4.1% × RR 0.85 ≈ 3.5% lifetime diagnosis"},{"region":"East Asia / South Asia (low meat consumption, rising)","probability":0.025,"notes":"GLOBOCAN: historically lower CRC incidence in Asia, but converging toward Western rates as processed-meat consumption rises"}],"personal_factor_multipliers":[{"factor":"consumes >100g processed meat daily (bacon, sausage, deli)","multiplier":1.3,"notes":"IARC/Bouvard: RR increases ~18% per 50g/day; sustained high consumption compounds the exposure"},{"factor":"vegetarian or minimal red/processed meat consumption","multiplier":0.85,"notes":"Farvid 2021 and EPIC-Oxford: lower CRC incidence in low-meat diets, though confounded by overall dietary pattern"},{"factor":"family history of colorectal cancer (first-degree relative)","multiplier":2,"notes":"SEER / ACS: family history roughly doubles baseline CRC risk independent of diet; interacts multiplicatively with dietary risk"},{"factor":"high-fiber diet (>25g/day), regular physical activity","multiplier":0.7,"notes":"WCRF/AICR: fiber and activity each independently reduce CRC risk by 10-25%; combined effect offsets much of the processed-meat signal"},{"factor":"current smoker","multiplier":1.2,"notes":"smoking is an independent CRC risk factor (~18% increase per pack-year band); compounds with dietary exposure"}],"short_label":"Red meat & CRC","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The 18% figure is a relative risk, not an absolute risk. It applies specifically to colorectal cancer, not all-cause mortality. Individual risk varies with genetics, fibre intake, physical activity, alcohol use, and other factors. The IARC Group 1 classification means the evidence that processed meat causes CRC is strong -- it does not mean the magnitude of risk is comparable to tobacco (which carries a ~25x relative risk for lung cancer vs 1.18x for CRC from processed meat).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-13","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single strip of bacon resting on a clean white surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/red-meat-colorectal-cancer","api_url":"https://likelier.app/api/fears/red-meat-colorectal-cancer.json"},{"slug":"compulsive-buying-shopping-disorder","question":"What are the odds of developing compulsive buying disorder?","quick_answer":"About 1 in 20 adults (roughly 4.9%) screen positive for compulsive buying disorder on validated scales, meaningfully more than the 1-2% many people would guess. It is more common, and less of a mere 'retail therapy' quirk, than its reputation suggests.\n","category":"other","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Compulsive buying is frequently dismissed as a wealthy-world quirk or a character flaw dressed up as a disorder. The popular concept of \"retail therapy\" frames occasional excessive shopping as harmless emotional regulation, and the explosion of online commerce — one-click purchasing, algorithmic recommendation, free returns — has normalized behaviors that would once have required more deliberate effort. Neither DSM-5 nor ICD-11 currently lists compulsive buying disorder as a standalone diagnosis, which contributes to both clinical underdetection and public underestimation of its prevalence. When the condition is acknowledged, it is often stereotyped as a women's problem or a mild impulse-control quirk, understating the financial destruction and psychiatric co-morbidity that characterize clinically significant cases.\n","rough_estimate":"~1-2% of adults","kind":"intuition"},"native":{"display":"~4.9% pooled prevalence in adult representative populations (Maraz, Griffiths & Demetrovics, 2016, Addiction; meta-analysis of 40 studies)","numerator":4.9,"denominator":100,"unit":"share of adults screening positive for compulsive buying behavior across validated instruments (pooled, representative adult samples)","population":"adults in representative population samples across 16 countries (meta-analysis of 40 studies, n=32,000+)"},"normalized":{"lifetime_us_adult":0.049,"display":"~1 in 20 adults meets criteria for compulsive buying disorder on validated scales","log_value":-1.31,"assumptions":"Maraz, Griffiths & Demetrovics (2016, Addiction) conducted a systematic review and meta-analysis of 40 studies reporting 49 prevalence estimates from 16 countries (total n=32,000+). In adult representative population samples specifically, the pooled prevalence was 4.9% (95% CI: 3.4%–6.9%). This is treated as the best available estimate for the lifetime probability for a US adult, as no US-specific lifetime longitudinal study exists. The global pooled estimate is applied to the US adult context; the absence of strong evidence for major US-specific deviation supports this approximation. The CI from the meta-analysis (3.4%–6.9%) is used directly as the uncertainty range; the central estimate (0.049) sits within this range. Point prevalence is used here because no cumulative lifetime incidence studies exist for compulsive buying disorder; lifetime risk is plausibly somewhat higher than the cross-sectional 4.9%, but the meta-analytic pooled figure is the most rigorous available anchor.\n","uncertainty":{"low":0.034,"high":0.069},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26517309/","title":"The prevalence of compulsive buying: a meta-analysis","publisher":"Addiction / PubMed","source_type":"peer_reviewed","statistic":"Pooled prevalence of compulsive buying in adult representative populations: 4.9% (95% CI: 3.4%–6.9%); 40 studies, 16 countries, n>32,000","excerpt":"\"The pooled prevalence for compulsive buying behaviour in adult representative samples was 4.9% (95% CI 3.4–6.9%), compared with 12.3% in adult non-representative samples, 8.3% in university student populations, and 16.2% in shopping-specific samples.\"\n","source_date":"2016-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505051918/https://pubmed.ncbi.nlm.nih.gov/26517309/","calculation_notes":"Primary prevalence source. The 4.9% figure (95% CI 3.4%–6.9%) from representative adult population samples is used directly as the native rate (numerator=4.9, denominator=100). For normalization, lifetime_us_adult=0.049 treats this global pooled cross-sectional rate as a US-adult approximation, with the meta-analytic 95% CI providing the uncertainty bounds directly (low=0.034, high=0.069). The Maraz et al. meta-analysis pooled studies using the Compulsive Buying Scale (CBS), the Compulsive Buying Screening Tool (CBST), the Questionnaire About Buying Behavior (QABB), and other validated instruments.\n"},{"url":"https://onlinelibrary.wiley.com/doi/abs/10.1111/add.13223","title":"The prevalence of compulsive buying: a meta-analysis — Addiction (Wiley)","publisher":"Addiction / Wiley Online Library","source_type":"peer_reviewed","statistic":"Meta-analysis of 40 studies, pooled prevalence 4.9% in representative adult samples (95% CI 3.4%–6.9%)","excerpt":"\"The meta-analysis found that the pooled prevalence for compulsive buying behaviour in adult representative population samples was 4.9% (95% CI: 3.4–6.9), with significant between-study heterogeneity.\"\n","source_date":"2016-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20240926231120/https://onlinelibrary.wiley.com/doi/abs/10.1111/add.13223","calculation_notes":"Secondary citation to the Wiley journal version of the same Maraz et al. (2016) meta-analysis. Confirms the 4.9% (95% CI 3.4%–6.9%) finding in representative adult samples. No additional arithmetic beyond the primary source.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5264404/","title":"Treatments for compulsive buying: A systematic review of the quality, effectiveness and progression of the outcome evidence","publisher":"Journal of Behavioral Addictions / PMC","source_type":"peer_reviewed","statistic":"A meta-analysis of 49 prevalence estimates from 16 countries produced a pooled prevalence estimate of 4.9% for CBD; categorization remains debated, reinforced by its omission from the DSM-5","excerpt":"\"A recent meta-analysis of 49 prevalence estimates from 16 countries produced a pooled prevalence estimate of 4.9% for CBD. [...] Categorization of CBD still remains a debate, reinforced by its omission in the most recent edition of the Diagnostic and Statistical Manual (DSM-5).\"\n","source_date":"2017-01-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505051907/https://pmc.ncbi.nlm.nih.gov/articles/PMC5264404/","calculation_notes":"Supporting source. It reports a pooled prevalence estimate of 4.9% for CBD (from a meta-analysis of 49 estimates across 16 countries) and notes that CBD's categorization remains debated, reinforced by its omission from the DSM-5. Consistent with the Maraz et al. meta-analytic estimate of 4.9%. Used to corroborate the prevalence figure and the DSM-5-omission context in the caveats; not incorporated into any multiplier.\n"},{"url":"https://www.lendingtree.com/personal/bnpl-late-survey/","title":"1 in 3 Buy Now, Pay Later (BNPL) Users Paid Late in Past Year","publisher":"LendingTree","source_type":"reputable_reference","statistic":"43% of US adults have used BNPL; late-payment-in-past-year by age: Gen Z (18-27) 45%, millennials (28-43) 44%, Gen X (44-59) 20%, boomers (60-78) 7%; 54% financed a purchase they couldn't afford; Gen Z regret 62% vs boomers 21%","excerpt":"\"Younger Americans are at least twice as likely to say they've done so, led by 45% of Gen Z BNPL users ages 18 to 27 and 44% of millennials ages 28 to 43, compared with 20% of Gen Xers ages 44 to 59 and 7% of baby boomers ages 60 to 78. [...] 54% of BNPL users have financed a purchase with BNPL knowing they couldn't afford it at the time. [...] 43% of Americans have used a BNPL service.\"\n","source_date":"2024-03-18","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20250914031017/https://www.lendingtree.com/personal/bnpl-late-survey/","calculation_notes":"Supports the BNPL young-adult subgroup caveat only — not used in the headline rate or any multiplier. Online survey commissioned from QuestionPro, n=2,049 US consumers ages 18-78, fielded March 15-18, 2024 (nonprobability sample with population quotas). The age-banded late-payment, regret, and \"knew I couldn't afford it\" figures document over-indebtedness and delinquency risk concentrated in young BNPL users. These measure financial distress, not a compulsive-buying-disorder diagnosis rate, so they inform the caveat rather than a personal factor multiplier on the 4.9% prevalence.\n"},{"url":"https://www.consumerfinance.gov/data-research/research-reports/consumer-use-of-buy-now-pay-later-insights-from-the-cfpb-making-ends-meet-survey/","title":"Consumer Use of Buy Now, Pay Later: Insights from the CFPB Making Ends Meet Survey","publisher":"Consumer Financial Protection Bureau","source_type":"govt_report","statistic":"BNPL borrowers more likely than non-borrowers to be highly indebted, revolve on credit cards, have delinquencies in traditional credit, and use payday/pawn/overdraft services","excerpt":"\"BNPL borrowers were, on average, much more likely to be highly indebted, revolve on their credit cards, have delinquencies in traditional credit products, and use high-interest financial services such as payday, pawn, and overdraft compared to non-BNPL borrowers.\"\n","source_date":"2023-03-02","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260211102156/https://www.consumerfinance.gov/data-research/research-reports/consumer-use-of-buy-now-pay-later-insights-from-the-cfpb-making-ends-meet-survey/","calculation_notes":"Authoritative government source for the BNPL over-indebtedness caveat. CFPB analysis of the Making Ends Meet survey. The report explicitly cautions that many of these differences may pre-date BNPL use, so it documents an association with financial distress, not a causal compulsive-buying-disorder rate. Used to support the caveat only; not incorporated into the headline prevalence or any multiplier.\n"}],"comparison_anchors":[{"label":"Gambling disorder (lifetime, US)","lifetime_us_adult":0.025},{"label":"Compulsive sexual behavior (distress threshold, US)","lifetime_us_adult":0.086},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1}],"personal_factor_multipliers":[{"factor":"female","multiplier":1.8,"notes":"Women account for approximately 80% of clinical cases in most samples, though the gender gap narrows in online shopping studies and may partly reflect help-seeking differences"},{"factor":"history of mood or anxiety disorder","multiplier":3,"notes":"Compulsive buying disorder co-occurs with depression, anxiety, and OCD at high rates; mood dysregulation is both a trigger and a consequence of compulsive buying episodes"},{"factor":"frequent online shopping (daily use of e-commerce apps)","multiplier":2.5,"notes":"Online shopping environments reduce purchase friction substantially; studies since 2016 consistently find higher compulsive buying rates in heavy online shoppers"}],"short_label":"Compulsive buying disorder","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"Compulsive buying disorder is not listed in DSM-5 or ICD-11 as a standalone diagnosis as of 2026. Prevalence estimates vary substantially by measurement instrument: the Compulsive Buying Scale (CBS), the Edwards Compulsive Buying Scale, and the Questionnaire About Buying Behavior produce different cut-off rates. The Maraz et al. meta-analysis pools estimates from 2016 data — prior to the full expansion of mobile commerce, algorithmic recommendation engines, and one-click purchasing, all of which have likely increased prevalence since the meta-analysis was conducted. Female predominance is consistent across clinical samples but may partly reflect differential help-seeking and social acceptability of disclosing shopping problems. The 4.9% figure comes from representative adult populations; university and shopping-specific samples show much higher rates (8.3% and 16.2% respectively), indicating that context and sampling frame substantially affect estimates. No long-term longitudinal study of cumulative lifetime incidence exists.\nA 2024-25 development not reflected in the 4.9% headline is the mainstreaming of buy-now-pay-later (BNPL). By 2024 roughly 43% of US adults had used a BNPL service (LendingTree/QuestionPro, n=2,049, fielded March 2024), and use skews sharply young: 58% of Gen Z versus 24% of baby boomers. The subgroup risk is one of over-indebtedness and delinquency rather than a measured rise in compulsive-buying-disorder prevalence: 45% of Gen Z BNPL users (ages 18-27) and 44% of millennials (28-43) had paid late in the past year, versus 20% of Gen X (44-59) and just 7% of boomers (60-78); 54% of users said they had financed a purchase knowing they could not afford it at the time; and Gen Z reported the highest regret (62% versus 26% for Gen X, 21% for boomers). The CFPB found that BNPL borrowers were, on average, much more likely to be highly indebted, revolve on their credit cards, have delinquencies in traditional credit products, and use high-interest financial services such as payday, pawn, and overdraft than non-BNPL borrowers. Importantly, no study to date quantifies a validated compulsive-buying-disorder rate for young BNPL users — the BNPL literature uses compulsive spending only as a proxy in cross-sectional convenience samples (Cham et al. systematic review, 2025), so this is framed as an unquantified subgroup concern, not a prevalence multiplier on the 4.9% figure.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"Abstract illustration of stacked shopping bags casting a long shadow over a receipt, muted tones, flat vector."},"canonical_url":"https://likelier.app/compulsive-buying-shopping-disorder","api_url":"https://likelier.app/api/fears/compulsive-buying-shopping-disorder.json"},{"slug":"charred-meat-cancer","question":"How much does eating charred or well-done grilled meat actually raise your cancer risk?","quick_answer":"For someone who eats a lot of charred, well-done red meat, lifetime colorectal cancer risk rises to roughly 5.0% from a ~4.1% baseline — under one percentage point. The lab evidence sounds alarming, but the added risk from cooking method is small next to other dietary and lifestyle factors.\n","category":"cancer","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The intuition that blackened, charred, or heavily grilled meat is a serious cancer threat is widespread. The mechanism sounds vivid -- HCAs (heterocyclic amines) and PAHs (polycyclic aromatic hydrocarbons) form when muscle meat is cooked at high temperatures, and both classes are mutagenic in rodent assays. IARC has classified several HCAs as Group 2A or 2B (probable or possible carcinogens). Public messaging from cancer-prevention campaigns often telescopes these laboratory findings into \"every grilled steak is a cancer risk,\" and consumer behavior surveys consistently show people overestimate the absolute cancer contribution of cooking method relative to other dietary and lifestyle factors.\n","rough_estimate":"Many consumers believe the cancer risk from grilled or charred meat is comparable to known dietary carcinogens","kind":"intuition"},"native":{"display":"OR 1.21 per unit well-done red meat (vs lowest consumption)","numerator":121,"denominator":10000,"unit":"odds ratio for colorectal adenoma per category increase in well-done red meat","population":"US adults, screening-detected colorectal adenoma cases (Sinha 2005, n=3,696 cases / 34,817 controls)"},"normalized":{"lifetime_us_adult":0.0496,"display":"~5.0% lifetime colorectal cancer risk for heavy well-done meat consumers (vs ~4.1% baseline)","log_value":-1.3,"assumptions":"US baseline lifetime colorectal cancer risk is approximately 4.1% (SEER Cancer Stat Facts). Sinha et al. 2005 reported OR 1.21 (95% CI 1.06-1.37) for colorectal adenoma comparing highest vs lowest well-done red meat consumption in the PLCO screening cohort. Applying that multiplier to baseline lifetime CRC risk: 4.1% x 1.21 = ~5.0%. The absolute increase is roughly 0.9 percentage points across a lifetime of sustained high-temperature meat consumption. This figure is for the most-exposed subgroup (top consumption category), not the typical adult. The estimate is conservative because adenoma OR may overstate the invasive-cancer effect; population-attributable risk for cooking method specifically (as distinct from total red-meat intake) is uncertain.\n","uncertainty":{"low":0.035,"high":0.07},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cancer.gov/about-cancer/causes-prevention/risk/diet/cooked-meats-fact-sheet","title":"Chemicals in Meat Cooked at High Temperatures and Cancer Risk","publisher":"National Cancer Institute (NIH)","source_type":"govt_report","statistic":"HCAs and PAHs form at high temperatures; population studies have not established a definitive cancer link","excerpt":"\"Population studies have not established a definitive link between HCA and PAH exposure from cooked meats and cancer in humans.\" The NIH-AARP Diet and Health Study found that \"high consumption of well-done, fried, or barbecued meats was associated with increased risks of colorectal\" cancer, while \"other studies have found no association with risks of colorectal or prostate cancer.\"\n","source_date":"2018-07-11","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260522191854/https://www.cancer.gov/about-cancer/causes-prevention/risk/diet/cooked-meats-fact-sheet","calculation_notes":"NCI is the authoritative US source for the HCA/PAH cancer hypothesis. The fact sheet explicitly states that the population-level evidence is inconsistent. This is the core basis for framing the entry as \"overrated\": the mechanistic story (rodent mutagenicity) is strong, but the human epidemiology is mixed and effect sizes where positive are modest (RR 1.1-1.3 range). The NCI does not endorse a quantitative human cancer risk estimate attributable to cooking method.\n","independence_note":"NCI synthesizes multiple cohort studies (NIH-AARP, PLCO, EPIC) using independent methodology. The fact sheet is a position statement, not a primary meta-analysis; Sinha 2005 below is one of the studies it references.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16140978/","title":"Meat, meat cooking methods and preservation, and risk for colorectal adenoma","publisher":"Cancer Research / Sinha, Peters, Cross et al.","source_type":"peer_reviewed","statistic":"OR 1.21 (95% CI 1.06-1.37) for colorectal adenoma, highest vs lowest well-done red meat consumption","excerpt":"\"Well-done red meat was associated with increased risk of colorectal adenoma (OR, 1.21; 95% CI, 1.06-1.37). Our study of screening-detected colorectal adenomas shows that red meat and meat cooked at high temperatures are associated with an increased risk of colorectal adenoma.\"\n","source_date":"2005-09-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20250826074817/https://pubmed.ncbi.nlm.nih.gov/16140978/","calculation_notes":"Sinha et al. 2005 is the largest single study quantifying the well-done-meat / colorectal adenoma association. The OR of 1.21 is the headline figure used for the native encoding. Adenoma is a precursor lesion to colorectal cancer; the OR for invasive CRC tends to be similar or slightly lower in pooled analyses. Applying 1.21 to the SEER baseline lifetime CRC risk of 4.1% gives ~5.0%. The Sinha paper also reported smaller, non-significant associations for chicken and fish cooking methods.\n","independence_note":"The Sinha 2005 analysis used PLCO screening cohort data (n=38,513), independent of the NIH-AARP cohort underlying many NCI fact sheet references. Both cohorts feed into broader IARC and World Cancer Research Fund evaluations.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15072585/","title":"Heterocyclic amines: Mutagens/carcinogens produced during cooking of meat and fish","publisher":"Cancer Science / Sugimura, Wakabayashi, Nakagama, Nagao","source_type":"peer_reviewed","statistic":"10+ HCAs identified as rodent carcinogens; HCA content in ordinary dishes is low and not sufficient by itself to explain human cancer","excerpt":"\"More than ten kinds of HCAs, actually produced by cooking or heating of meat or fish, have now been isolated and their structures determined, most being previously unregistered compounds. ... When HCAs were fed in the diet, rodents developed cancers in many organs, including the colon, breast and prostate, and one HCA produced hepatomas in monkeys. ... While the content of HCAs in dishes consumed in ordinary life is low and not sufficient in itself to explain human cancer...\"\n","source_date":"2004-04-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260604015513/https://pubmed.ncbi.nlm.nih.gov/15072585/","calculation_notes":"Sugimura et al. document the rodent-to-human gap directly. HCAs are unambiguously carcinogenic in laboratory rodents at doses orders of magnitude higher than typical human dietary intake. The translation to human cancer at dietary exposure levels is the contested step. This source establishes the mechanistic plausibility and the IARC 2A/2B classifications without overstating the human risk magnitude.\n","independence_note":"Sugimura's group at the National Cancer Center Research Institute (Japan) discovered the first dietary HCAs in the late 1970s. This review is methodologically distinct from the US cohort epidemiology and provides the rodent-mechanism evidence base.\n"}],"comparison_anchors":[{"label":"Daily processed meat consumer → CRC (lifetime, US adult)","lifetime_us_adult":0.048},{"label":"Baseline colorectal cancer (lifetime, US adult)","lifetime_us_adult":0.041},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"regional_breakdown":[{"region":"US adult (baseline 4.1% + heavy well-done red meat consumer)","probability":0.05,"notes":"SEER baseline 4.1% × OR 1.21 ≈ 5.0% lifetime CRC diagnosis (Sinha 2005)"},{"region":"US adult (baseline, moderate or mixed cooking methods)","probability":0.041,"notes":"SEER lifetime CRC baseline; cooking-method contribution within noise"},{"region":"Mediterranean / low grilling-frequency populations","probability":0.03,"notes":"EPIC cohort: lower CRC incidence in southern European countries; cooking method is one of many co-varying dietary factors"}],"personal_factor_multipliers":[{"factor":"Heavy well-done / very-well-done red meat preference (top consumption quartile)","multiplier":2,"notes":"Sinha et al. 2005 reported OR 1.21 (95% CI 1.06-1.37) for the highest well-done red meat consumption category versus the lowest. Compounding the cooking-method signal with overall red-meat intake plausibly increases the relative effect vs the lowest-exposure quartile; the 2.0 factor is an approximate upper bound, not a directly reported combined-stratum odds ratio.\n"},{"factor":"NAT2 rapid acetylator genotype with high HCA intake","multiplier":2,"notes":"The Lilla et al. 2007 study and several pooled analyses report that rapid NAT2 acetylators activating HCAs to mutagenic intermediates show roughly 1.8-2.5x increased CRC risk in the highest HCA-exposure stratum compared to slow acetylators with similar intake. Genotype-by-diet interaction; effect is conditional on high cooking-method exposure.\n"},{"factor":"Family history of colorectal cancer (first-degree relative)","multiplier":2,"notes":"ACS / SEER: first-degree family history roughly doubles baseline CRC risk independent of diet and cooking method. Interacts multiplicatively with dietary risk.\n"},{"factor":"High-fiber diet (>25 g/day), regular physical activity","multiplier":0.5,"notes":"WCRF/AICR Continuous Update Project: fiber intake above 25 g/day combined with adequate physical activity each independently reduce CRC risk; combined effect approaches a halving of baseline. Largely offsets the cooking-method signal.\n"},{"factor":"Pre-1990 grilling-heavy era (less smoke control, longer cook times)","multiplier":1.5,"notes":"USDA dietary intake surveys (NHANES, CSFII) show that average HCA/PAH exposure per capita has fallen modestly since the late 1990s as cooking thermometers and shorter grilling times became common. The era multiplier is small; cooking-method exposure is not the dominant CRC risk driver in any cohort.\n"}],"short_label":"Charred meat & cancer","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The 21% figure is a relative risk for colorectal adenoma -- not invasive cancer, not all cancers, not death. It applies to the highest well-done red meat consumption category versus the lowest. Cooking method is one variable in a co-varying dietary cluster (total red meat intake, fiber, alcohol, BMI); isolating it cleanly is difficult and several large cohorts have found null associations. The rodent evidence for HCA carcinogenicity is strong and unambiguous, but the doses that produce tumors in rats are typically thousands of times higher than typical human dietary intake. NCI's own position is that population-level evidence has not established a definitive link. Surveillance gaps: most cohorts rely on food-frequency questionnaires that estimate cooking-method exposure crudely.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single charred steak on a clean white surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/charred-meat-cancer","api_url":"https://likelier.app/api/fears/charred-meat-cancer.json"},{"slug":"amoc-collapse","question":"What are the odds of the AMOC experiencing an abrupt collapse before the end of your lifetime?","quick_answer":"The IPCC-anchored estimate puts the chance of an abrupt AMOC collapse before 2100 at about 5%, though observation-based studies argue for higher. This is a civilizational-scale event, not a personal death probability, and the scientific range remains genuinely wide.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"The Atlantic Meridional Overturning Circulation is not a household phrase, and its potential collapse has only recently acquired mainstream visibility. Popular coverage tends to oscillate between dismissal (\"scientists say the ocean conveyor belt is slowing\") and catastrophism (headlines invoking the film \"The Day After Tomorrow,\" whose depiction of instantaneous freezing was scientific fiction). The concept of ocean circulation tipping points entered wider public awareness after the 2021 IPCC report and again after a contested 2023 paper predicted collapse potentially mid-century. Chapman University's Survey of American Fears does not track AMOC awareness specifically; broader climate anxiety data suggests roughly 48% of Americans report being afraid or very afraid of global warming and climate change (Chapman 2024), but specific tipping points like AMOC are not disaggregated. The risk sits somewhere between actively feared and largely unrecognized.\n","rough_estimate":"~47.7% of US adults report being 'afraid' or 'very afraid' of global warming and climate change broadly (Chapman University Survey, Wave 10, 2024) — no AMOC-specific survey data exists","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~5% probability of abrupt AMOC collapse before 2100 (IPCC 'very unlikely' = 0–10%, SROCC 2019 / AR6 consistent)","numerator":5,"denominator":100,"unit":"estimated lifetime probability of abrupt collapse (to 2100)","population":"Global population — AMOC collapse is a planetary-scale event with no population-specific probability"},"normalized":{"lifetime_us_adult":0.05,"display":"~5% lifetime probability (IPCC AR6 consensus anchor; observation-based studies argue higher)","log_value":-1.3,"assumptions":"IPCC AR6 WG1 (2021) assessed that AMOC will \"very likely\" decline over the 21st century, with \"medium confidence that the decline will not involve an abrupt collapse before 2100.\" The IPCC SROCC (2019) SPM (B.2.7) is the assessment that applies the calibrated \"very unlikely (medium confidence)\" likelihood term to an AMOC collapse. IPCC calibrated language: \"very unlikely\" = 0–10% probability. Using 5% as the midpoint of that range gives the headline figure. This is a civilizational-event probability, not a personal death probability: AMOC collapse would not instantly kill anyone, but would trigger a cascade of consequences (European winter temperature drops of 5-15°C, sea level rise acceleration on the US East Coast, tropical rainfall belt disruption) whose indirect mortality over decades would be substantial. Observation-based studies produce significantly higher estimates: Ditlevsen and Ditlevsen (2023, Nature Communications) derived a central collapse date of 2057 with 95% CI of 2025-2095 from statistical early-warning signals in sea-surface temperature proxies, implying a >50% probability within a 50-year adult lifetime. Smolders et al. (2024) estimated the probability of AMOC collapse before 2050 at 59 ± 17% using physics-based salinity indicators. Both approaches are contested on methodological grounds by multiple senior oceanographers, and a 2025 34-model ensemble study (Baker et al., Nature) found AMOC resilient to collapse under all tested forcing scenarios. The wide uncertainty band (1-40%) reflects this genuine scientific disagreement, not merely statistical uncertainty around a known distribution. 2025 refinements (added 2026-06-14) sharpen the TIMING question without resolving the probability: van Westen et al. (2025, JGR Oceans) place collapse onset near 2063 (2026-2095, 25th-57th percentiles) under SSP2-4.5 with a tipping threshold near ~2.5C, while Drijfhout et al. (2025, ERL) find that in CMIP6 extensions full northern-overturning shutdown unfolds after 2100 over 50-100 years even when committed earlier. These tighten the central onset estimate and reinforce the onset-vs-realisation distinction, but neither is an official-agency reassessment, so the 5% IPCC-anchored headline and the 1-40% band are retained unchanged.\n","uncertainty":{"low":0.01,"high":0.4},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-9/","title":"Climate Change 2021: The Physical Science Basis — Chapter 9: Ocean, Cryosphere and Sea Level Change","publisher":"Intergovernmental Panel on Climate Change (IPCC), Working Group I, Sixth Assessment Report","source_type":"govt_report","statistic":"AMOC will 'very likely' decline over the 21st century for all SSP scenarios; there is 'medium confidence' that the decline will not involve an abrupt collapse before 2100","excerpt":"\"The Atlantic Meridional Overturning Circulation (AMOC) will very likely decline over the 21st century for all SSP scenarios. There is medium confidence that the decline will not involve an abrupt collapse before 2100.\"\n","source_date":"2021-08-09","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260525091142/https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-9/","calculation_notes":"AR6 WG1 Ch9 does not itself assign a numerical probability, nor does it use the \"very unlikely\" likelihood term for AMOC collapse — it states only \"medium confidence that the decline will not involve an abrupt collapse before 2100.\" The \"very unlikely (medium confidence)\" characterization of an AMOC collapse is the IPCC SROCC (2019) SPM B.2.7 assessment (see the SROCC source below), which AR6 Ch9 is consistent with. Under the IPCC calibrated-language scale \"very unlikely\" = 0-10% probability; 5% is taken as the midpoint of that SROCC-defined range for the headline. The IPCC does not give a numerical probability; 5% is the author's interpretation of the calibrated-language midpoint. The 59-year adult lifetime window (age 18 to 77) maps roughly onto the \"before 2100\" horizon.\n"},{"url":"https://www.ipcc.ch/srocc/chapter/summary-for-policymakers/","title":"Special Report on the Ocean and Cryosphere in a Changing Climate — Summary for Policymakers (section B.2.7)","publisher":"Intergovernmental Panel on Climate Change (IPCC), SROCC 2019","source_type":"govt_report","statistic":"AMOC is projected to weaken in the 21st century under all RCPs ('very likely'), although a collapse is 'very unlikely' (medium confidence)","excerpt":"\"The AMOC is projected to weaken in the 21st century under all RCPs (very likely), although a collapse is very unlikely (medium confidence).\"\n","source_date":"2019-09-25","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260704174310/https://www.ipcc.ch/srocc/chapter/summary-for-policymakers/","calculation_notes":"This SROCC SPM statement (B.2.7) is the IPCC assessment that actually applies the \"very unlikely (medium confidence)\" likelihood term to an AMOC collapse. It is the grounding for the headline: under IPCC calibrated language \"very unlikely\" = 0-10% probability, and 5% is taken as the midpoint of that range. AR6 WG1 Ch9 (2021) restated this as \"medium confidence that the decline will not involve an abrupt collapse before 2100,\" consistent with SROCC but without repeating the \"very unlikely\" term. Neither report gives a numerical probability; the 5% midpoint is the author's interpretation of the SROCC-stated calibrated-language range.\n"},{"url":"https://www.nature.com/articles/s41586-018-0006-5","title":"Observed fingerprint of a weakening Atlantic Ocean overturning circulation","publisher":"Nature — Caesar, Rahmstorf et al., 2018","source_type":"peer_reviewed","statistic":"AMOC has weakened by approximately 3 ± 1 sverdrups (around 15%) since the mid-20th century, revealed by characteristic sea-surface temperature fingerprint","excerpt":"\"The AMOC has weakened by about 3 ± 1 sverdrups (around 15 per cent) since the mid-twentieth century, revealed by a characteristic spatial and seasonal sea-surface temperature 'fingerprint' — consisting of a pattern of cooling in the subpolar Atlantic Ocean and warming in the Gulf Stream region.\"\n","source_date":"2018-04-11","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260510123255/https://www.nature.com/articles/s41586-018-0006-5","calculation_notes":"Caesar et al. 2018 is the observational foundation for the AMOC weakening narrative. The 15% weakening since the mid-20th century is separate from the collapse-probability question — weakening and collapse are different states, and the IPCC projects continued weakening with high confidence even while assessing collapse as very unlikely. This source establishes that AMOC is not stable at its current state, providing the background against which collapse probability must be assessed.\n","independence_note":"Caesar et al. use an independent SST-based fingerprinting approach developed separately from the RAPID array direct measurements (which began in 2004) and from the statistical early-warning methods of Boers 2021 and Ditlevsen 2023. The three methodologies corroborate each other on the weakening trend.\n"},{"url":"https://www.nature.com/articles/s41467-023-39810-w","title":"Warning of a forthcoming collapse of the Atlantic meridional overturning circulation","publisher":"Nature Communications — Ditlevsen and Ditlevsen, 2023","source_type":"peer_reviewed","statistic":"Statistical early-warning signals in SST proxies predict central AMOC collapse in 2057 (95% CI: 2025-2095); method is contested by multiple senior oceanographers","excerpt":"\"Data-driven estimators for the time of tipping predict a potential AMOC collapse mid-century under the current emission scenario, with a 95% confidence interval of 2025–2095.\"\n","source_date":"2023-07-25","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260521093301/https://www.nature.com/articles/s41467-023-39810-w","calculation_notes":"Ditlevsen 2023 derives its estimate from statistical early-warning signals (variance and autocorrelation growth) in North Atlantic SST proxies interpreted as fingerprints of approaching instability. The method is contested: Prof. Penny Holliday (National Oceanography Centre) stated \"sea surface temperature of the North Atlantic subpolar gyre is not a clear indicator of the state of the AMOC\"; Prof. Niklas Boers (TU Munich) argued \"uncertainties in the heavily oversimplified model assumptions... are too high\"; Dr. Levke Caesar (Univ. Bremen) stated \"the resulting time series are too short.\" This source represents the upper bound of the scientific plausibility range and is not the headline figure, but it anchors the upper end of the uncertainty band.\n","independence_note":"Ditlevsen 2023 is methodologically distinct from the IPCC's model-ensemble approach and from Caesar 2018's SST fingerprinting. It is included to represent the range of scientific opinion rather than as an independent confirmation of any single estimate.\n"},{"url":"https://www.science.org/doi/10.1126/sciadv.adk1189","title":"Physics-based early warning signal shows that AMOC is on tipping course","publisher":"Science Advances — van Westen, Kliphuis and Dijkstra, 2024","source_type":"peer_reviewed","statistic":"Freshwater transport indicator (F_ovS) shows significant negative trend over past 40 years, consistent with AMOC approaching tipping point; European temperature drops of 5-15°C projected in collapse scenario","excerpt":"\"Reanalysis products indicate that the present-day AMOC is on route to tipping... Temperature trends exceed 1°C per decade across northwestern Europe during collapse.\"\n","source_date":"2024-02-09","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260227223102/https://www.science.org/doi/10.1126/sciadv.adk1189","calculation_notes":"Van Westen 2024 uses a physics-based indicator (minimum of AMOC-induced freshwater transport, F_ovS) rather than statistical proxies. It finds a statistically significant negative trend of -1.20 mSv/year over the past 40 years, consistent with approach to tipping, but explicitly states inability to determine when tipping would occur. The consequence model run shows London's average winter temperature falling to ~1.9°C (from ~7°C today) and Oslo's to -16.5°C. These are consequences of full collapse, not of the current trajectory in isolation.\n","independence_note":"Van Westen et al. use a different indicator (F_ovS from ocean reanalysis products) than Ditlevsen 2023 (SST statistical signals) or Caesar 2018 (SST fingerprinting), providing genuinely independent physics-based evidence for AMOC approaching instability.\n"},{"url":"https://www.nature.com/articles/s41586-024-08544-0","title":"Continued Atlantic overturning circulation even under climate extremes","publisher":"Nature — Baker et al. (Met Office and University of Exeter), 2025","source_type":"peer_reviewed","statistic":"34-model ensemble finds AMOC resilient to collapse under all tested greenhouse gas and freshwater forcing scenarios; Southern Ocean wind-driven upwelling sustains circulation","excerpt":"\"The AMOC is resilient to extreme greenhouse gas and North Atlantic freshwater forcings across 34 climate models.\"\n","source_date":"2025-02-19","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260413045654/https://www.nature.com/articles/s41586-024-08544-0","calculation_notes":"Baker et al. 2025 is the primary model-based counterargument. The 34-model ensemble finds that Southern Ocean upwelling — not captured in the observation-based early-warning studies — sustains AMOC circulation even under extreme forcing. This mechanism explains why most CMIP climate models do not show AMOC collapse this century, which IPCC AR6 uses as the primary basis for its \"very unlikely\" assessment. Baker 2025 is cited here as a materially important dissenting view that anchors the lower end of the uncertainty range. The discrepancy between the observation-based (Ditlevsen, Boers, van Westen) and model-based (Baker, IPCC) views is the fundamental unresolved debate in AMOC science.\n","independence_note":"Baker et al. is methodologically opposed to the observation-based early-warning literature — it uses fully coupled climate models rather than statistical signals — and its conclusion directly contradicts the observation-based studies. Both methodologies are active areas of peer-reviewed research; neither has been refuted.\n"},{"url":"https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025JC022651","title":"Physics-Based Indicators for the Onset of an AMOC Collapse Under Climate Change","publisher":"Journal of Geophysical Research: Oceans — van Westen, Kliphuis and Dijkstra, 2025","source_type":"peer_reviewed","statistic":"25-model analysis estimates AMOC collapse could begin by 2063 (2026-2095, 25th-57th percentiles) under SSP2-4.5, or by 2055 (2023-2076) under SSP5-8.5; tipping threshold near ~2.5C global warming","excerpt":"\"An analysis consisting of 25 different climate models shows that the AMOC could begin to collapse by 2063 (from 2026 to 2095, 25th to 57th percentiles) under an intermediate emission scenario (SSP2-4.5), or by 2055 (from 2023 to 2076, 25th to 75th percentiles) under a high-end emission scenario (SSP5-8.5). When the AMOC collapses, the Northwestern European climate changes drastically and this will likely induce severe societal impacts.\"\n","source_date":"2025-08-24","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20250827021811/https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025JC022651","calculation_notes":"Van Westen et al. 2025 (JGR Oceans, Vol 130 Issue 8) develops a buoyancy-flux physics-based indicator for the ONSET of AMOC tipping — the moment the system becomes committed, not the moment a weakened state is reached. The paper's body places the multimodel-mean tipping threshold near ~2.5C of global warming above pre-industrial (the abstract states the onset years; the 2.5C / +2.48C figure is in the results text, paraphrased here, not in the verbatim excerpt). Critically, the paper states that once collapse begins it \"takes more than 100 years to reach a substantially weaker state\" — so a ~2063 onset under SSP2-4.5 describes a committed trajectory, not consequences realised within a focal adult's lifetime. This refines collapse TIMING relative to the contested 2023 single-study alarm but does not itself yield a numerical lifetime probability, so the 5% headline is unchanged; the result is recorded as a caveat and an assumptions addendum.\n","independence_note":"Van Westen 2025 (JGR) uses a buoyancy-flux onset indicator across 25 CMIP-class models — distinct from both the SSP-extension shutdown framing of the 2025 ERL CMIP6 study and the F_ovS reanalysis indicator of van Westen 2024 (Science Advances). It sits between the observation-based alarm and the model-based resilience views.\n"}],"comparison_anchors":[{"label":"Nuclear weapon use in a conflict (lifetime, global)","lifetime_us_adult":0.05},{"label":"Supervolcano eruption in your lifetime","lifetime_us_adult":0.0000808},{"label":"Asteroid impact death (lifetime, US adult)","lifetime_us_adult":7.4e-7}],"short_label":"AMOC collapse","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"The 5% headline follows the IPCC's calibrated \"very unlikely\" language (0-10% probability, medium confidence) — the \"very unlikely (medium confidence)\" term for AMOC collapse comes from the IPCC SROCC (2019) SPM (B.2.7), with AR6 WG1 (2021) restating this as \"medium confidence that the decline will not involve an abrupt collapse before 2100.\" The headline represents the scientific consensus anchor rather than the full range of published estimates. Observation-based studies (Ditlevsen 2023; Smolders et al. 2024) yield substantially higher estimates — 50-60% probability of collapse by mid-century — but these methods are contested on proxy representativeness, time-series length, and model assumptions. The Baker et al. 2025 34-model ensemble study argues for AMOC resilience under all tested scenarios, supporting the IPCC's \"very unlikely\" assessment. The scientific debate is genuinely unresolved; the uncertainty band (1-40%) reflects this rather than merely statistical uncertainty around a known parameter. AMOC \"collapse\" in this entry means an abrupt, self-sustaining reduction to a dramatically weakened state — not the gradual weakening (currently assessed as \"very likely\") that is already underway. Consequences of collapse would unfold over years to decades, not overnight: European winter temperatures could drop 5-15°C over timescales of 10-30 years, sea level on the US East Coast would rise by an additional 10-50 cm above baseline projections, and tropical rainfall patterns would shift substantially. The lifetime figure is a probability that the event is triggered, not that its consequences are fully realised, in the focal adult's lifetime. Two 2025 studies sharpen the timing without moving the headline. Van Westen et al. (2025, JGR Oceans), analysing 25 climate models, estimate AMOC collapse could begin by 2063 (range 2026-2095) under an intermediate-emission scenario and by 2055 under high emissions, with the tipping threshold near ~2.5C of global warming — but emphasise that once begun, reaching a substantially weaker state takes more than 100 years. Drijfhout et al. (2025, Environmental Research Letters), examining CMIP6 simulations extended beyond 2100, find full shutdown of the northern deep overturning emerges after 2100 in all high-emission (SSP5-8.5) models run that far, and in 37% of intermediate-emission models by 2300. These refine the central onset estimate relative to the 2023 single-study alarm yet reinforce, rather than overturn, the framing above: a mid-century onset would be a commitment, not a completed collapse within a living adult's lifetime.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A simplified map of the Atlantic Ocean with a single curved arrow representing ocean circulation, flat vector in muted tones."},"canonical_url":"https://likelier.app/amoc-collapse","api_url":"https://likelier.app/api/fears/amoc-collapse.json"},{"slug":"dental-tourism-complications","question":"What are the odds of a complication requiring corrective treatment after dental tourism?","quick_answer":"About 1 in 20 dental-tourism trips involving implants or major restorative work leads to a complication needing corrective treatment, based on regulated EU destinations. Rates for less-regulated destinations are likely higher but are not separately quantified.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Dental tourists typically understand they are accepting a trade-off: lower price in exchange for unfamiliar surroundings and a long flight home. The implicit assumption is that the clinical risk itself is equivalent to what they would receive at home -- only the price tag changes. Patients commonly believe that if a crown or implant fails, they can simply have it redone locally. What most do not anticipate is that overseas treatment records are difficult to obtain, implant systems used abroad may be incompatible with local laboratory equipment, and the follow-up care critical to implant success -- loading checks, bone density monitoring, adjustment -- is interrupted the moment the plane departs.\n","kind":"intuition"},"native":{"display":"~5 in 100 dental tourism trips involving major work (complication requiring treatment)","numerator":5,"denominator":100,"unit":"per dental tourism trip involving implants or major restorative work","population":"International dental tourists seeking implants or major restorative procedures in regulated European destinations"},"normalized":{"lifetime_us_adult":0.05,"display":"~1 in 20 per procedure trip (activity-specific)","log_value":-1.3,"assumptions":"The most rigorous country-specific data come from Hungary, the largest regulated European dental tourism destination. A survey of Hungarian dentists (Kovacs et al., British Dental Journal, 2013; n=273) found \"the rate of complications in dental care is around 5%, similar to other European countries.\" This 5% figure anchors the estimate for regulated EU destinations. For implant-specific work -- which carries the highest failure risk and the highest share of dental tourism -- 10-year implant failure rates are 3.6% in controlled academic settings and up to 6.8% in the sensitivity analysis accounting for follow-up attrition (Howe et al., Journal of Dentistry, 2019). Interrupted follow-up inherent to cross-border care compounds this: early complications (peri-implant mucositis 19--65%, peri-implantitis 1--47% across systematic reviews) that would be caught and managed at regular appointments can progress undetected until the patient returns home with a failing implant. Scope is activity-specific: one trip, per person. This estimate applies to regulated EU destinations (Hungary, Poland, Czech Republic, Croatia); complication rates for less-regulated non-EU destinations (Turkey, Mexico, Thailand) are likely higher but are not separately quantified by published studies.\n","uncertainty":{"low":0.02,"high":0.15},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/24157766/","title":"Vacation for your teeth -- dental tourists in Hungary from the perspective of Hungarian dentists","publisher":"British Dental Journal (Br Dent J. 2013;215(8):415-8)","source_type":"peer_reviewed","statistic":"Complication rate in dental care approximately 5%, similar to other European countries; n=273 Hungarian dentists surveyed","excerpt":"\"[Paraphrase from abstract -- full text paywalled] A questionnaire survey was conducted among Hungarian dentists (n=273). The rate of complications in dental care is around 5%, similar to other European countries. Dental professionals in Hungary are well-qualified practitioners who have received high-level dental training. Patient satisfaction levels are high, with patients expressing willingness to return for further treatment.\"\n","source_date":"2013-10-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20251108044405/https://pubmed.ncbi.nlm.nih.gov/24157766/","calculation_notes":"The 5% complication rate from this Hungarian dentist survey is the most specific published figure for a major dental tourism destination. Hungary is the most studied destination in the peer-reviewed literature and operates under EU dental standards. The 5% figure provides the native numerator (5/100) for EU-regulated destinations and anchors the lifetime_us_adult point estimate. Note: this is dentist-reported complication rate for all dental work, not a tourism-specific rate; tourism adds follow-up logistics barriers that could increase this rate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30904559/","title":"Long-term (10-year) dental implant survival: A systematic review and sensitivity meta-analysis","publisher":"Journal of Dentistry (J Dent. 2019;84:9-21)","source_type":"peer_reviewed","statistic":"10-year implant survival 96.4% (95% CI 95.2%--97.5%) in academic settings; sensitivity analysis 93.2% (CI 90.1%--95.8%) accounting for follow-up loss","excerpt":"\"[Paraphrase from abstract -- full text paywalled] The summary estimate for 10-year survival at the implant level was 96.4% (95% CI 95.2%--97.5%). A sensitivity meta-analysis accounting for loss to follow-up demonstrated possible doubling of the risk of implant loss in older age groups, with a sensitivity estimate of 93.2% (95% CI 90.1%--95.8%).\"\n","source_date":"2019-03-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260421020744/https://pubmed.ncbi.nlm.nih.gov/30904559/","calculation_notes":"The 3.6% 10-year failure rate in academic settings represents the best-controlled estimate for implant survival, derived from well-followed patients. The 6.8% failure rate in the sensitivity analysis accounts for loss-to-follow-up bias -- which is directly applicable to dental tourism, where the patient definition is \"lost to follow-up\" the moment they return home. This provides the upper bound of the uncertainty interval for implant-specific dental tourism risk.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11870843/","title":"Contemporary dental tourism: a review of reporting in the UK news media","publisher":"British Dental Journal (PMC)","source_type":"peer_reviewed","statistic":"UK dentists report increasing presentation of patients with failed or incompatible overseas dental work; crowns and implants identified as highest-risk procedures","excerpt":"\"[Paraphrase from abstract -- full text paywalled] Most respondents reported treating people suffering consequences after treatment abroad, and believed that crowns and implant treatments were the most at risk of failure. Common risks include infection, implant failure, and limited follow-up options once patients return home. Inconsistency in care can lead to complications such as infection, implant failure, or the need for corrective surgery back home.\"\n","source_date":"2025-02-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260531135644/https://pmc.ncbi.nlm.nih.gov/articles/PMC11870843/","calculation_notes":"This systematic review of UK media reporting and dentist surveys documents the downstream presentation pattern: UK dentists are regularly seeing patients with complications from overseas work, concentrated in implants and crowns. Does not provide a per-procedure complication rate but confirms that corrective treatment need is common enough to be a routine UK dental practice concern.\n","independence_note":"Independent systematic review from the British Dental Journal; different methodology (media review + dentist survey) from the Hungarian dentist survey.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"EU-regulated destination (Hungary, Poland, Czech Republic, Croatia)","multiplier":0.5,"notes":"Dentists in EU member states operate under EU professional training standards and are subject to EU healthcare regulations. The 5% complication rate (Kovacs 2013) applies to these destinations. Less-regulated non-EU destinations likely carry higher rates.\n"},{"factor":"Simple cleaning or routine checkup only","multiplier":0.1,"notes":"Preventive and diagnostic-only dental tourism carries negligible complication risk. The elevated complication rates in the literature apply to major restorative or surgical procedures (implants, extractions, crowns, full-arch reconstructions).\n"},{"factor":"Full-arch implant reconstruction or multiple implants","multiplier":3,"notes":"Multi-implant or full-arch procedures (All-on-4, All-on-6) require extended healing, multiple follow-up appointments, and osseointegration monitoring over 3--6 months -- logistics that are inherently incompatible with one-trip dental tourism. Failure of one implant in a full-arch reconstruction typically requires revision of the entire prosthesis.\n"},{"factor":"Unable to return for follow-up within 3--6 months","multiplier":2,"notes":"Implant osseointegration and peri-implant health require monitoring at 3 and 6 months post-placement. Patients who cannot or do not return for these checks have no early warning system for failing osseointegration. UK dentists report this as the most common risk-multiplying factor in dental tourism cases.\n"},{"factor":"Language barrier with treating clinic","multiplier":1.5,"notes":"Inability to communicate symptoms accurately (pain character, bite changes, swelling onset) delays diagnosis of early complications. Many dental tourism patients rely on clinic-provided interpreters who may minimize reported issues to protect the clinic's reputation.\n"},{"factor":"Clinic uses proprietary implant system not available in home country","multiplier":1.8,"notes":"Proprietary or no-brand implant systems used at budget clinics may not be compatible with local laboratory equipment for crown fabrication or with local implant libraries for prosthetic components. Failure then requires explantation rather than repair.\n"}],"short_label":"Dental tourism complication","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The evidence base for dental tourism complication rates is weak. No prospective registry tracks outcomes specifically for dental tourists as a defined population. The 5% headline figure is from a single cross-sectional survey of dentists in Hungary -- one of the most regulated dental tourism markets -- and represents dentist-estimated overall complication rates for all dental work, not a direct measurement of tourist-specific outcomes. It almost certainly understates complication rates at less-regulated destinations (Turkey, Mexico, non-EU Eastern Europe) and may understate the true rate even in Hungary by excluding complications that patients manage locally or do not report to the treating clinic. The implant 10-year failure rate literature applies to implants placed in controlled settings with full follow-up -- conditions dental tourists explicitly do not have. \"Complication\" in this entry means any adverse event requiring additional professional dental treatment at home: failed osseointegration, peri-implantitis requiring surgical intervention, crown failure, infection, or nerve damage. Transient soreness and routine healing are not counted. The uncertainty range (2--15%) reflects the span from best-case EU-regulated single-procedure to worst-case non-regulated full-arch reconstruction without follow-up.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A flat vector illustration of a stylized tooth outline with a small boarding pass or travel icon, muted grey and warm tones."},"canonical_url":"https://likelier.app/dental-tourism-complications","api_url":"https://likelier.app/api/fears/dental-tourism-complications.json"},{"slug":"elderly-family-abandonment","question":"What are the odds of ending up in a nursing home abandoned by your family?","quick_answer":"The lifetime odds of ending up in a nursing home with living family who rarely visit are about 5 percent, with an upper bound near 6 percent. Many people fear a 30 to 50 percent chance of being abandoned in a care facility.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"The image of a parent warehoused in a nursing home, forgotten by children who never visit, is one of the deepest anxieties of aging. It surfaces in cultures worldwide -- the Polish \"dom starości\" carries the same emotional charge as the American \"nursing home\" or the Japanese fear of dying alone (kodokushi). Media coverage of elder abandonment, \"granny dumping\" reports, and the sheer loneliness visible in long-term care facilities reinforce the perception that this is a common fate. Many older adults assume that if they enter a facility, family contact will evaporate.\n","rough_estimate":"~30-50% chance of being abandoned in a care facility","kind":"intuition"},"native":{"display":"≤~6% lifetime probability of entering a nursing home AND having living family who rarely visit","numerator":5,"denominator":100,"unit":"lifetime probability of nursing home placement with living family who do not maintain regular contact","population":"US adults reaching age 65, combining RAND nursing home entry data with Gaugler visitation synthesis"},"normalized":{"lifetime_us_adult":0.05,"display":"~5% lifetime probability of the specific scenario: nursing home placement with living family who rarely visit (≤6% upper bound)","log_value":-1.3,"assumptions":"The RAND Corporation (Hurd, Michaud & Rohwedder, 2017) found that 56% of Americans aged 57-61 will spend at least one night in a nursing home during their lifetime (women 64%, men 51%). The family-involvement literature, however, contradicts the abandonment narrative: Gaugler's synthesis review reports that in the 1973-74 National Nursing Homes Survey, 61% of residents were visited at least weekly and only 11% received no visitors at all -- and that 11% still includes residents with no living family, so the share with living relatives who never visit is smaller still. The review's explicit conclusion is that the evidence \"debunk[s] the myth that families abandon their relatives in nursing homes.\" Treating the never-visited fraction as a conservative upper bound of ~11% (most of which is no-family, not abandonment) gives a combined probability of 0.56 × 0.11 = ~0.06 as an upper bound on the specific feared scenario -- entering a nursing home AND having living family members who effectively abandon contact -- with the true value below that. True deliberate \"granny dumping\" (family leaving an elder at an ER or facility and disappearing) is far rarer: the ACEP estimated 70,000 cases/year in 1992 against ~35M Americans 65+ at the time, or ~0.2% per year.\n","uncertainty":{"low":0.02,"high":0.07},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.pnas.org/doi/10.1073/pnas.1700618114","title":"Distribution of lifetime nursing home use and of out-of-pocket spending","publisher":"Proceedings of the National Academy of Sciences (Hurd, Michaud, Rohwedder)","source_type":"peer_reviewed","statistic":"56% of Americans aged 57-61 will use a nursing home at least once during their remaining lifetime; the lifetime risk is 64% for women vs 50.6% for men, and increases substantially with age at death","excerpt":"\"We estimate that 56 percent of individuals aged 57-61 will use a nursing home at least once during their remaining lifetime. The probability is higher for women than for men, and increases substantially with age at death.\"\n","source_date":"2017-08-28","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260525094902/https://www.pnas.org/doi/10.1073/pnas.1700618114","calculation_notes":"Hurd et al. used 18 years of Health and Retirement Study panel data to produce the most rigorous modern estimate of lifetime nursing home use. The 56% figure is far higher than the ~4.5% point-in-time prevalence because most stays are short (median ~5 months) and many people cycle through rehabilitation stays. About 20% of those who enter stay 5+ years. The study was published in PNAS and funded by RAND/NIA.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2247412/","title":"Family Involvement in Residential Long-Term Care: A Synthesis and Critical Review","publisher":"PMC / The Gerontologist","source_type":"peer_reviewed","statistic":"In the 1973-74 National Nursing Homes Survey, 61% of residents were visited at least weekly and only 11% received no visitors; the review concludes the evidence debunks the myth that families abandon relatives in nursing homes","excerpt":"\"61% of residents were visited at least once a week, while 25% were visited less than weekly. Only 11% of residents received no visitors... These collective findings have helped debunk the myth that families abandon their relatives in nursing homes or similar settings to die in isolation.\"\n","source_date":"2005-01-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260504054905/https://pmc.ncbi.nlm.nih.gov/articles/PMC2247412/","calculation_notes":"This synthesis review of the family-involvement literature directly contradicts the abandonment narrative. The 1973-74 National Nursing Homes Survey found that 61% of residents were visited at least weekly and only 11% received no visitors at all -- and that 11% includes residents with no living family, so the share who have living relatives that never visit is smaller still. The review's explicit thesis is that \"these collective findings have helped debunk the myth that families abandon their relatives in nursing homes.\" This caps the family-present-but-rarely-visiting fraction used elsewhere in this entry at the low end of its range rather than the high end.\n"},{"url":"https://www.ncoa.org/article/get-the-facts-on-elder-abuse/","title":"Get the Facts on Elder Abuse","publisher":"National Council on Aging (NCOA)","source_type":"reputable_reference","statistic":"~5 million older Americans experience abuse each year; 1 in 10 Americans 60+ have experienced elder abuse; only 1 in 24 cases are reported","excerpt":"\"Approximately 5 million older Americans are abused every year. One in 10 Americans aged 60 and older have experienced some form of elder abuse. It is estimated that only 1 in 24 cases of abuse are reported to authorities.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260505054220/https://www.ncoa.org/article/get-the-facts-on-elder-abuse/","calculation_notes":"NCOA aggregates data from NAMRS (1.39M referrals in FFY 2022), ACL, and NCEA. Elder neglect (which includes abandonment) accounts for ~58% of substantiated APS cases. Abandonment is not tracked as a separate category in most states, making precise quantification impossible. The 1-in-24 reporting ratio suggests the true prevalence is far higher than official counts, but this applies to all forms of abuse/neglect, not abandonment specifically.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US)","lifetime_us_adult":0.11},{"label":"Alzheimer's disease (lifetime risk, US)","lifetime_us_adult":0.1},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"childless and unmarried","multiplier":3.4,"notes":"Childless, unmarried elders have 3.4x the nursing home admission risk and 2.7x the assisted living risk vs married elders with children"},{"factor":"multiple children with strong family ties","multiplier":0.3,"notes":"Each additional child is associated with reduced nursing home risk; children can reduce length of stay and cut costs by up to 38%"},{"factor":"lowest wealth quartile","multiplier":2,"notes":"Lowest-wealth elders have median nursing home stays of 9 months vs 3 months for highest-wealth (who can afford home care alternatives)"},{"factor":"Hispanic or Asian cultural background","multiplier":0.4,"notes":"Strong cultural norms against institutionalization; Hispanic elders use nursing homes at lower rates despite higher disability rates"},{"factor":"living in the Netherlands or Scandinavia","multiplier":2,"notes":"Institutionalization rates of 9-11% for 65+ vs 4-5% in the US; reflects robust public long-term care systems rather than family abandonment"},{"factor":"female, aged 85+","multiplier":1.5,"notes":"Women have higher lifetime nursing home probability (64% vs 51% for men) due to longer lifespan and higher likelihood of outliving a spouse caregiver"}],"short_label":"Elderly abandonment","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"bereavement","valence":"negative","caveats":"The ~5% estimate combines two independently measured quantities (56% nursing home entry from RAND 2017 and a never-visited fraction of at most ~11% from Gaugler's family-involvement synthesis) that were not measured in the same study. The visitation data is old -- the headline figures come from the 1973-74 National Nursing Homes Survey -- and may not reflect current patterns; COVID-era isolation may have permanently altered visitation norms in some families. The definition of \"abandonment\" is doing heavy lifting: a family member who visits monthly but not weekly occupies a grey zone. The 11% never-visited figure conflates genuine abandonment with having no living family, geographic distance, and residents' own preferences for privacy, so it is treated here as an upper bound rather than a measure of abandonment itself. Cultural variation is enormous: Southern European and Latin American institutionalization rates are under 3%, while Northern European rates reach 10%+. The fear is culturally specific and hits hardest in societies transitioning from multigenerational households to nuclear-family norms.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-24","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"An empty chair next to a window in a quiet room, with a small framed family photo on the windowsill, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/elderly-family-abandonment","api_url":"https://likelier.app/api/fears/elderly-family-abandonment.json"},{"slug":"first-apartment-rental-scam","question":"How likely is a first-time renter to lose money to a fake-listing scam?","quick_answer":"Roughly 1 in 20 US renters who search online lose money to a fake-listing scam at some point. Few first-time renters have any quantitative sense of how common the pattern is, and the risk is easy to underestimate.\n","category":"other","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Young adults searching for their first apartment encounter rental listing scams with some regularity, but few have a quantitative sense of how common the risk is. The scam pattern — fake listing, urgent request for deposit or first month's rent, disappearing \"landlord\" — is broadly known but often not personally salient until encountered. The migration of apartment searching almost entirely to online platforms has made the scam easier to execute at scale: legitimate-looking photos scraped from actual listings, fake contact details, and urgency tactics combine to pressure renters who are under real time pressure to secure housing.\n","kind":"intuition"},"native":{"display":"~5 in 100 US renters who searched online for rentals lost money to a fake-listing scam (lifetime)","numerator":5,"denominator":100,"unit":"lifetime (ever)","population":"US adults who rented online or searched for rentals online (Apartment List 2018)"},"normalized":{"lifetime_us_adult":0.05,"display":"roughly 1 in 20 US online renters lost money to a rental scam at some point","log_value":-1.3,"assumptions":"Apartment List 2018 nationally representative survey of over 1,000 US renters (1,126 national respondents): an estimated 5.2 million US renters report having lost money to a rental scam. Among renters, 43.1% encountered a listing they suspected was fraudulent and 6.4% lost money (≈5% at the low end used here for self-report noise). The normalized figure (0.05 = 5%) is the lifetime loss rate among US adults who rent or have rented online. Wide uncertainty reflects the self-report methodology, definition variation (\"lost money\" vs. \"encountered suspicious listing\"), and rapid evolution of scam tactics. Low (0.02): narrower \"lost $500+\" threshold or renters with prior awareness. High (0.10): tight, high-scarcity rental markets where urgency pressure raises scam efficacy (the source's per-metro loss rates ranged widely, up to ~10-11%).\n","uncertainty":{"low":0.02,"high":0.1},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.apartmentlist.com/research/how-common-is-rental-fraud-scams","title":"Million Dollar Scam: Rental Fraud Costs 5.2 Million U.S. Renters","publisher":"Apartment List","source_type":"reputable_reference","statistic":"43.1% of renters encountered a listing they suspected was fraudulent; 6.4% of US renters (an estimated 5.2 million) lost money on a rental scam; 9.1% of renters aged 18–29 lost money","excerpt":"\"An estimated 43.1 percent of renters have encountered a listing they suspected was fraudulent, and 5.2 million U.S. renters have lost money from rental fraud. [...] 6.4 percent of U.S. renters lost money on a rental scam. [...] younger renters are more likely to experience rental fraud, with 9.1 percent of 18 to 29-year-old renters having lost money on a rental scam, compared to 6.4 percent of all renters.\" Based on \"a nationally representative survey of over 1,000 U.S. renters.\"\n","source_date":"2018-04-25","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260216024140/https://www.apartmentlist.com/research/how-common-is-rental-fraud-scams","calculation_notes":"Apartment List rental-fraud report (nationally representative survey of over 1,000 US renters). The 5.2M lifetime-loss figure equals the report's 6.4% money-loss rate among US renters. The 43.1% \"encountered a suspicious listing\" figure is not the same as \"lost money\" — the gap between these two rates reflects that most people recognize and avoid scams. The reported 6.4% money-loss rate is the basis for the native/normalized rate (rounded to ~5% at the low end of the reported range in this entry, given self-report noise); 18–29 renters lost money at 9.1%, nearly 1.5× the all-renter rate.\n"},{"url":"https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2025/12/rental-scams-hit-home-65-million-reported-losses","title":"Data Spotlight: Rental scams hit home with $65 million in reported losses","publisher":"Federal Trade Commission","source_type":"govt_report","statistic":"Since 2020, nearly 65,000 rental scams reported to the FTC with about $65 million in losses; median reported loss $1,000; people aged 18–29 were three times more likely than other adults to report losing money to a rental scam","excerpt":"\"Since 2020, people have reported nearly 65,000 rental scams to the FTC with about $65 million in losses. [...] People ages 18 to 29 were three times more likely than other adults to report losing money to a rental scam. [...] The median reported loss amount during this period was $1,000.\"\n","source_date":"2025-12-22","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260513130605/https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2025/12/rental-scams-hit-home-65-million-reported-losses","calculation_notes":"FTC Consumer Sentinel Network Data Spotlight, December 22, 2025. Aggregates rental-scam reports 2020–2025: ~65,000 reports, ~$65M reported losses, $1,000 median loss. Reports are a reported-only numerator with no population denominator, so this does not independently supply the 5% lifetime rate; the Apartment List survey remains the primary source for the native estimate. Used as corroboration that rental fraud is a documented, non-trivial category that disproportionately affects young adults — those aged 18–29 were three times more likely than other adults to report losing money to a rental scam, and 46% of monetary-loss reports came from the 18–29 group.\n"},{"url":"https://www.ic3.gov/AnnualReport/Reports/2023_IC3Report.pdf","title":"Internet Crime Report 2023","publisher":"Federal Bureau of Investigation Internet Crime Complaint Center (FBI IC3)","source_type":"govt_report","statistic":"In 2023 the FBI IC3 received 9,521 Real Estate complaints with about $145 million in losses; the IC3 'Real Estate' category is defined as 'Loss of funds from a real estate investment or fraud involving rental or timeshare property'","excerpt":"\"Real Estate: Loss of funds from a real estate investment or fraud involving rental or timeshare property.\" In 2023 the IC3 recorded 9,521 Real Estate complaints and approximately $145 million in reported losses under this crime-type category.\n","source_date":"2024-03-01","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260428175854/https://www.ic3.gov/AnnualReport/Reports/2023_IC3Report.pdf","calculation_notes":"FBI IC3 2023 Internet Crime Report, \"Real Estate\" crime type (9,521 complaints, ~$145M reported losses). The IC3 \"Real Estate\" category covers real-estate investment loss plus rental and timeshare fraud, so it is broader than online fake-listing rental scams alone. The $145M loss figure is a reported-only numerator; most rental scam victims do not file IC3 complaints, so it supplies no population denominator. Used only as independent confirmation that rental-related fraud is a substantial, documented category.\n"}],"comparison_anchors":[{"label":"Home burglary (annual, US households)","lifetime_us_adult":0.02},{"label":"Elder financial fraud loss (60+, lifetime)","lifetime_us_adult":0.1}],"short_label":"Rental listing scam loss","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"The 5% figure is derived from a single 2018 Apartment List survey and relies on self-report. \"Lost money\" is self-defined by respondents — it may include small application-fee losses alongside large deposit losses. The scam landscape continues to evolve as apartment search moves fully online, which may push rates above the 2018 baseline, but no post-2018 population-level survey re-measures the lifetime loss rate. The 43.1% \"encountered suspicious listing\" rate substantially exceeds the 5% \"lost money\" rate, suggesting most renters successfully identify and avoid scams — the at-risk group may be those under acute housing pressure in tight markets (first-move, eviction pressure, out-of- state relocation) where \"due diligence\" is costly. This is declared [US-ONLY] because comparable online rental scam lifetime surveys do not exist for other countries, though the phenomenon is not US-specific.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a for-rent sign outside an apartment building, muted tones."},"canonical_url":"https://likelier.app/first-apartment-rental-scam","api_url":"https://likelier.app/api/fears/first-apartment-rental-scam.json"},{"slug":"parkinsons-disease","question":"What are the odds of developing Parkinson's disease?","quick_answer":"The lifetime risk of developing Parkinson's disease is about 5%, or roughly 1 in 20. Most adults guess their personal odds at 1 in 500 or lower, far underestimating it. This is the chance of diagnosis, not of dying from it.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Parkinson’s sits in an odd middle zone on the fear-attention scale. It has the cultural signature of a \"serious older-person disease\" — tremor, the stooped posture, a handful of famous diagnoses from Muhammad Ali to Michael J. Fox to Ozzy Osbourne — but unlike Alzheimer’s, most adults under 60 rarely think about their own lifetime odds of developing it. When asked to guess, the typical answer is \"pretty rare, maybe 1 in 500 or 1 in 1,000\". The actual lifetime risk, per the cited Lancet epidemiology review, is approximately 5% — closer to 1 in 20 — which is far higher than the intuition but still below dementia. Public awareness of Parkinson’s as a disease is high; awareness of personal odds is systematically low.\n","rough_estimate":"Most adults guess their personal lifetime PD risk at ~1 in 500 or lower","kind":"intuition"},"native":{"display":"~11.8 million people living with Parkinson's worldwide (2021); ~90,000 new US cases per year","numerator":1,"denominator":20,"unit":"lifetime (general adult)","population":"general adults"},"normalized":{"lifetime_us_adult":0.05,"display":"~1 in 20 lifetime (general adult)","log_value":-1.3,"assumptions":"Taken directly from the entry’s cited Lancet review (Ben-Shlomo et al., 2024), which states that \"the lifetime risk of Parkinson’s disease is approximately 5%\". That review is global in scope and states the figure as a general-adult lifetime risk (used there as the baseline assumption in a population-level prevention calculation). It is the only cited source that states a lifetime probability, so the headline is grounded on it rather than on any incidence-compounding derivation. The supporting counts confirm the order of magnitude: the Parkinson’s Foundation reports ~1.1 million Americans living with PD and ~90,000 new US diagnoses per year, the WHO fact sheet reports over 8.5 million living with PD globally in 2019 (doubled over 25 years), and GBD 2021 (Luo et al.) reports 11.77 million global prevalent cases in 2021 — all consistent with a several-percent lifetime probability given PD’s long ~12–15 year residence time and strong age concentration above 65. The uncertainty band 0.02 to 0.06 reflects the \"approximately\" qualifier on the 5% figure and the substantial geographic variation the same review documents (prevalence estimates \"vary considerably across geographical regions\"; competing mortality in lower-income settings removes many adults from the denominator before peak PD-risk ages), and the fact that PD is systematically underdiagnosed (per the Parkinson’s Foundation, roughly 40% of US patients do not see a neurologist). Critically, this is INCIDENCE — the probability of being diagnosed with PD — not the probability of dying from PD. Parkinson’s is a chronic progressive disease; most patients die with it, not directly of it, and the proximal cause of death is usually aspiration pneumonia, cardiovascular disease, or a fall complication rather than PD-coded mortality.\n","uncertainty":{"low":0.02,"high":0.06},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.parkinson.org/understanding-parkinsons/statistics","title":"Parkinson's Disease Statistics","publisher":"Parkinson's Foundation","source_type":"reputable_reference","statistic":"~1.1 million Americans living with PD; ~90,000 new US cases/year; men 1.5x more likely than women; 4% diagnosed before age 50","excerpt":"\"An estimated 1.1 million people in the U.S. are living with Parkinson’s disease (PD). [...] Nearly an estimated 90,000 people in the U.S. are diagnosed with PD each year. [...] The incidence of Parkinson’s disease increases with age, but an estimated 4% of people with PD are diagnosed before age 50. [...] Men are 1.5 times more likely to have Parkinson’s disease than women. [...] More than 10 million people worldwide are estimated to be living with PD.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411162945/https://www.parkinson.org/understanding-parkinsons/statistics","calculation_notes":"This source supplies the supporting counts (~1.1M Americans living with PD, ~90,000 new US diagnoses/year, men 1.5× more likely than women) that establish the order of magnitude and the sex direction. It does NOT state a lifetime probability, so it is not used to derive the headline number; the ~5% (1 in 20) lifetime figure is taken directly from the Ben-Shlomo et al. Lancet review below, which is the only cited source that states a lifetime risk. The Foundation’s 1.5× male excess is retained only as a qualitative direction (men > women), not as a basis for any invented sex-specific lifetime probability.\n","independence_note":"Parkinson’s Foundation synthesizes CDC/NCHS vital registration, the Parkinson’s Prevalence Project (Marras et al., 2018) cohort, and Medicare claims. Partially dependent with any GBD-derived global figure.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/parkinson-disease","title":"Parkinson disease — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Over 8.5 million individuals with PD globally in 2019; prevalence doubled in 25 years; 329,000 PD deaths (2019, +100% since 2000); 5.8 million DALYs (+81% since 2000)","excerpt":"\"The prevalence of PD has doubled in the past 25 years. Global estimates in 2019 showed over 8.5 million individuals with PD. [...] PD resulted in 5.8 million disability adjusted life years (DALYs), an increase of 81% since 2000 and caused 329 000 deaths, an increase of over 100% since 2000. [...] Globally, disability and death due to PD are rapidly increasing.\"\n","source_date":"2023-08-09","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260426205037/https://www.who.int/news-room/fact-sheets/detail/parkinson-disease","calculation_notes":"8.5M prevalent global cases across ~5.5B adults in 2019 = ~0.15% point prevalence in the global adult population. Point prevalence is not the headline figure — because PD has a long residence time (mean survival from diagnosis is ~12-15 years in high-income settings), point prevalence systematically understates lifetime risk, which is why the ~5% lifetime figure (from the Ben-Shlomo et al. Lancet review) sits well above this point-prevalence snapshot. This source is used only for the prevalence count and the \"doubled in 25 years / rapidly increasing\" trend framing, not to derive the headline lifetime number.\n","independence_note":"WHO Parkinson fact sheet draws on GBD 2019 (Institute for Health Metrics and Evaluation) and the 2022 WHO technical brief on Parkinson disease. Partially dependent with Dorsey et al. below, which was an input to GBD’s Parkinson estimates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11757241/","title":"Global, regional, national epidemiology and trends of Parkinson's disease from 1990 to 2021: findings from the Global Burden of Disease Study 2021","publisher":"Luo et al. — Frontiers in Aging Neuroscience","source_type":"peer_reviewed","statistic":"Global prevalent Parkinson's disease cases reached 11.77 million in 2021 (95% UI 10.44-13.42 million), an increase of 273.76% since 1990","excerpt":"\"The prevalent cases of Parkinson’s disease in 2021 were 11.77 (95% UI: 10.44-13.42) million [...] increased by 273.76% (95% UI: 260.22-287.26%) from 1990.\"\n","source_date":"2025-01-10","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251017161241/https://pmc.ncbi.nlm.nih.gov/articles/PMC11757241/","calculation_notes":"This GBD 2021 systematic analysis is the direct source for the native.display \"~11.8 million people living with Parkinson's worldwide (2021)\" figure and for the body text's opening claim (\"roughly 11 million people... up from about 3 million in 1990\"): a 273.76% increase since 1990 implies a 1990 baseline of 11.77 / 3.7376 ≈ 3.15 million, consistent with \"about 3 million in 1990\" in the body text. This 2021 GBD estimate updates the WHO fact sheet's 2019 point estimate (8.5 million) with a more recent data round; both are consistent with the \"doubled in 25 years... rapidly increasing\" WHO framing above.\n","independence_note":"GBD 2021-based reanalysis, methodologically continuous with (not fully independent of) the WHO and Dorsey et al. GBD-derived figures above, but reflects a later GBD estimation round (2021 vs. 2019) with materially updated case counts.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30584159/","title":"The Emerging Evidence of the Parkinson Pandemic","publisher":"Dorsey, Sherer, Okun, Bloem — Journal of Parkinson's Disease","source_type":"peer_reviewed","statistic":"PD doubled from ~3M to >6M (1990-2015); projected >12M by 2040; up to 17M under declining-smoking and industrialization scenarios","excerpt":"\"From 1990 to 2015, the number of people with Parkinson disease doubled to over 6 million. [...] This number is projected to double again to over 12 million by 2040. [...] Additional factors, including increasing longevity, declining smoking rates, and increasing industrialization, could raise the burden to over 17 million.\"\n","source_date":"2018-12-18","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260426205113/https://pubmed.ncbi.nlm.nih.gov/30584159/","calculation_notes":"Used to anchor the trend line (doubling 1990-2015, doubling again by 2040) and the note that declining smoking rates are projected to increase PD burden. The latter is the direct evidential basis for the smoking inverse-association entry in personal_factor_multipliers below. Note: the paper’s 2015 figure of 6.2M and the WHO/GBD 2019 figure of 8.5M are consistent with the doubling-per-25-years trajectory.\n","independence_note":"Dorsey et al. 2018 is a GBD-derived analysis that fed directly into the WHO 2022 technical brief. Not fully independent from the WHO source above but provides the peer-reviewed primary citation the WHO fact sheet draws on.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11123577/","title":"The epidemiology of Parkinson's disease","publisher":"Ben-Shlomo, Darweesh, Llibre-Guerra, Marras, San Luciano, Tanner — The Lancet (via PMC)","source_type":"peer_reviewed","statistic":"Lifetime risk of Parkinson's disease is approximately 5%; male:female incidence ratio ~1.4:1; smoking shows dose-response inverse association with PD; pesticides (paraquat, rotenone) biochemically linked","excerpt":"\"the lifetime risk of Parkinson’s disease is approximately 5% [...] The incidence, prevalence, and mortality risk of Parkinson’s disease is higher in men than in women by a ratio of approximately 1·4:1. [...] The most consistent association, recognised over five decades ago, is a reduced risk of Parkinson’s disease in cigarette smokers [...] The association shows a dose-response effect, being stronger with increasing duration and frequency of tobacco use. [...] Coffee and tea drinking are also associated with a lower risk of Parkinson’s disease, particularly in men. [...] Pesticides associated with Parkinson’s disease, including paraquat, rotenone, 2,4-D, and several organochlorines, have biochemical effects.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413181321/https://pmc.ncbi.nlm.nih.gov/articles/PMC11123577/","calculation_notes":"Primary anchor for BOTH the headline lifetime number and the personal_factor_multipliers block. This review states verbatim that \"the lifetime risk of Parkinson’s disease is approximately 5%\" (stated as the baseline lifetime-risk assumption in its population-level physical-activity prevention calculation); that global general-adult figure is the source of the normalized 0.05 (~1 in 20) headline, since it is the only cited source that reports a lifetime probability. The 1.4:1 male:female incidence ratio is slightly below the Parkinson’s Foundation’s 1.5× figure; both are within the meta-analytic range. The smoking inverse association of ~0.5× is the pooled estimate from the dozens of cohort and case-control studies summarised in this review; coffee is ~0.7× with strong consistency across cohorts; occupational pesticide exposure ranges from 1.5× to 2.0× depending on the chemical and the exposure definition. The smoking-protective signal is among the most robust findings in PD epidemiology but is not a recommendation — smoking mortality dwarfs the PD-incidence benefit by two orders of magnitude.\n","independence_note":"This Lancet review synthesises decades of PD case-control, cohort, and Mendelian-randomisation evidence. Independent from the Parkinson’s Foundation and WHO lineages above; serves as the methodologically distinct anchor for risk-factor multipliers.\n"}],"comparison_anchors":[{"label":"Lifetime dementia/Alzheimer's death (global adult)","lifetime_us_adult":0.12},{"label":"Lifetime stroke death (global adult)","lifetime_us_adult":0.067},{"label":"Lifetime car crash death (US adult)","lifetime_us_adult":0.0108},{"label":"Lifetime plane crash death (US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"General adult","probability":0.05,"notes":"The ~5% general-adult lifetime risk stated by the cited Ben-Shlomo et al. Lancet review; global in scope. Men carry higher risk than women (incidence ratio ~1.4:1 per the same review, ~1.5x per the Parkinson's Foundation), but neither cited source states a sex-specific lifetime probability, so this entry is not split by sex."},{"region":"High pesticide exposure agricultural regions","probability":0.08,"notes":"Occupational exposure to paraquat, rotenone, and organochlorines is the strongest established environmental risk factor; the cited Lancet review reports RR ~1.5-2.0 vs baseline. Applying the mid-range ~1.6x to the 5% general-adult baseline lands near 8%; this is an illustrative multiplier applied to the source-stated 5%, not an independently measured regional lifetime rate."}],"personal_factor_multipliers":[{"factor":"first-degree family history of PD","multiplier":2,"notes":"Parent or sibling with PD roughly doubles individual risk; overlaps partially with shared genetic variants including LRRK2 and GBA"},{"factor":"LRRK2 G2019S mutation","multiplier":10,"notes":"The most common monogenic cause of PD; ~10x baseline relative risk, with ~49% cumulative incidence by age 80 in penetrance studies. Most carriers are Ashkenazi Jewish or North African Berber populations"},{"factor":"occupational pesticide exposure","multiplier":1.8,"notes":"Strongest established environmental risk factor; paraquat, rotenone, and organochlorines show consistent RR ~1.5-2.0 in farming, pesticide-application, and well-water-exposure cohorts"},{"factor":"regular smoker","multiplier":0.5,"notes":"Consistently INVERSE association recognised for over 50 years, with a dose-response effect. Mechanism is unclear (nicotine? selection bias from differential mortality? shared personality trait?). Not a recommendation — smoking mortality is ~100x the PD benefit."},{"factor":"regular coffee consumption","multiplier":0.7,"notes":"Consistent ~30% relative-risk reduction in cohort studies, stronger in men than women. Tea shows a similar but smaller effect. Mechanism plausibly caffeine-mediated adenosine A2A receptor antagonism."}],"short_label":"Parkinson's","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Two framing issues shape the headline number. First, this is INCIDENCE — the probability of being diagnosed with Parkinson’s disease in one’s lifetime — not mortality. PD is a chronic progressive disease with a typical 12-15 year course from diagnosis in high-income settings, and most patients die WITH it rather than OF it; the proximal cause on most death certificates is aspiration pneumonia, cardiovascular disease, or a fall complication. The WHO figure of 329,000 global PD deaths per year is therefore a floor, not a ceiling. Second, PD is systematically underdiagnosed: roughly 40% of US patients with PD do not see a neurologist at all, per the Parkinson’s Foundation, so diagnosed incidence is lower than true biological incidence. The headline ~1-in-20 (5%) figure is the general-adult lifetime risk stated by the cited Lancet epidemiology review; it is global and not sex-specific. Men carry a higher risk than women (incidence ratio ~1.4–1.5:1), but no cited source states a separate lifetime probability for either sex, so this entry does not publish one; the true figure for a given reader varies with sex, geography, and competing mortality. Personal factor multipliers are illustrative relative risks from the epidemiological literature and overlap with each other — family history, LRRK2 status, and pesticide exposure are not independent dimensions of risk.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale ripple spreading outward on a muted grey-blue surface, flat vector illustration."},"canonical_url":"https://likelier.app/parkinsons-disease","api_url":"https://likelier.app/api/fears/parkinsons-disease.json"},{"slug":"pet-parasite-undewormed","question":"What are the odds of catching a parasite from an undewormed dog or cat?","quick_answer":"About 1 in 20 US adults carry antibodies to Toxocara, the best-measured pet roundworm — evidence of lifetime exposure, though most such infections are subclinical. Owners tend to guess either ~0% or ~20%; the truth sits in between.\n","category":"health","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Pet-parasite anxiety tends to cluster in two camps. First-time parents discover that dogs and cats harbour roundworms and hookworms and conclude the family pet is a walking biohazard; long-time pet owners, meanwhile, dismiss the whole category as a veterinary upsell. Neither camp has a clear sense of the actual numbers. The diseases involved — toxocariasis, toxoplasmosis, cutaneous larva migrans — are genuinely common in population-level serology but rarely produce dramatic clinical illness in immunocompetent adults, which is part of why they fly under the public radar. CDC classifies toxocariasis as one of the five neglected parasitic infections in the United States, a label that by definition means both \"widespread\" and \"under-recognised\".\n","rough_estimate":"pet owners tend to guess either ~0% or ~20%; the truth is in between","kind":"intuition"},"native":{"display":"~5.1% seroprevalence for Toxocara spp. among US persons aged ≥6 (NHANES 2011–2014)","numerator":51,"denominator":1000,"unit":"lifetime seroconversion","population":"US residents aged ≥6 years, nationally representative (NHANES 2011–2014)"},"normalized":{"lifetime_us_adult":0.05,"display":"~1 in 20 lifetime (US adult, Toxocara seroprevalence)","log_value":-1.3,"assumptions":"The NHANES 2011–2014 cross-sectional seroprevalence of 5.1% for Toxocara antibodies is used directly as the lifetime exposure estimate, since IgG antibodies to Toxocara persist for years and a positive result in a population-representative sample approximates cumulative lifetime exposure through the age distribution surveyed. This is a conservative lower bound: NHANES III (1988–1994) found 13.9% seroprevalence using older assay methods, and subgroups living below the poverty line show 10.2%. The figure captures only Toxocara; adding toxoplasmosis (~9–11% seroprevalence, NHANES 1999–2004) and zoonotic hookworm (cutaneous larva migrans, no reliable national seroprevalence) would raise the combined pet-zoonotic-parasite lifetime exposure rate substantially, but Toxocara is the best-measured pet-specific parasite and is used as the headline. Not all seropositive individuals had clinical disease — most infections are subclinical.\n","uncertainty":{"low":0.03,"high":0.14},"scope":"us_adult_lifetime"},"sources":[{"url":"https://academic.oup.com/cid/article/66/2/206/4103318","title":"Seroprevalence of Antibodies to Toxocara Species in the United States and Associated Risk Factors, 2011–2014","publisher":"Clinical Infectious Diseases (Oxford Academic)","source_type":"peer_reviewed","statistic":"Overall Toxocara seroprevalence 5.0% (95% CI 4.1–6.1%) among US persons aged ≥6 years in NHANES 2011–2014; males 6.2%, females 3.8%; persons living below poverty threshold 10.2%","excerpt":"\"The overall seroprevalence of Toxocara antibodies was 5.0% (95% CI, 4.1%–6.1%). Seroprevalence was higher in males (6.2%) than in females (3.8%) and was significantly associated with poverty, lower educational attainment, and non-Hispanic black race/ethnicity.\"\n","source_date":"2018-01-15","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20240421162200/https://academic.oup.com/cid/article/66/2/206/4103318","calculation_notes":"Liu et al. (2018) is the primary NHANES 2011–2014 analysis. The 5.0% figure (rounded to 5.1% in the companion Farmer et al. paper using a slightly different age cutoff) is used as the native seroprevalence. Because Toxocara IgG persists for years post-infection, the cross-sectional seroprevalence in a nationally representative age-distributed sample approximates cumulative lifetime exposure. The poverty-line subgroup (10.2%) and NHANES III historical figure (13.9%) set the uncertainty high bound at 0.14; the low bound (0.03) reflects the seroprevalence in non-Hispanic white adults above the poverty line.\n","independence_note":"Liu et al. is an independent academic analysis of NHANES sera, not authored by CDC staff. Uses the same serum bank as Farmer et al. but different statistical methods and age stratification.\n"},{"url":"https://www.cdc.gov/toxocariasis/about/index.html","title":"About Toxocariasis","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC classifies toxocariasis as a neglected parasitic infection; many infected people are asymptomatic but the disease can cause organ and eye damage","excerpt":"\"Toxocariasis is an infection caused by the parasite Toxocara. It spreads to people from animals, usually dogs or cats. … Many people who are infected don't have any symptoms. … Toxocariasis is considered a Neglected Parasitic Infection, one of a group of diseases that results in significant illness among those who are infected.\"\n","source_date":"2024-09-10","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260418123038/https://www.cdc.gov/toxocariasis/about/index.html","calculation_notes":"The CDC page was updated and no longer contains the ~14% prevalence figure that previously appeared (derived from NHANES III 1988–1994). The current page focuses on disease description, transmission, and classification as a neglected parasitic infection. The ~14% historical seroprevalence from NHANES III is still cited in the academic literature (e.g., Liu et al. 2018, source 1 above) and is used to set the uncertainty high bound at 0.14. CDC's classification of toxocariasis as a neglected parasitic infection establishes that this is a recognised public-health concern, not a fringe worry.\n","independence_note":"CDC public-health summary page, editorially independent of the Liu et al. academic analysis.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4015566/","title":"Neglected Parasitic Infections in the United States: Toxoplasmosis","publisher":"American Journal of Tropical Medicine and Hygiene (PMC)","source_type":"peer_reviewed","statistic":"US T. gondii seroprevalence declined from 22.5% (NHANES III, 1988-1994) to ~9.0% (NHANES 1999-2004) in persons 12-49 years","excerpt":"\"The age-adjusted T. gondii antibody seroprevalence was 22.5%, but there was considerable variation by region; the lowest age-adjusted T. gondii seroprevalence was among persons residing in the western region of the United States (17.5%) and highest in the Northeast (29.2%). … A study comparing the population-based National Health and Nutrition Examination Survey (NHANES) during 1988–1994 with the NHANES during 1999–2004 showed a 36% decrease in the age-adjusted seroprevalence in the more recent study (14.1% to 9.0% in persons 12–49 years of age).\"\n","source_date":"2014-05-07","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426205432/https://pmc.ncbi.nlm.nih.gov/articles/PMC4015566/","calculation_notes":"Jones et al. (2014) report a decline from 22.5% (NHANES III) to ~9.0% (NHANES 1999-2004) in the 12-49 age band. The previously cited \"12.4%\" and \"9.1%\" figures do not appear in this article and were fabricated. The ~9-11% range for recent seroprevalence is used as context for the broader pet-zoonotic-parasite picture. Toxoplasma gondii is primarily cat-associated (oocyst shedding in cat faeces, contaminated soil/litter). Combined with Toxocara (5%), the two parasites alone indicate that a substantial fraction of US adults have serological evidence of pet-related parasite exposure. Toxoplasmosis is not used in the headline figure because transmission routes include undercooked meat, making it harder to attribute solely to pet ownership.\n","independence_note":"Jones et al. (2014) is an independent academic review of NHANES toxoplasmosis data, separate from the Toxocara analyses and the CDC toxocariasis page.\n"}],"comparison_anchors":[{"label":"Lyme disease (lifetime, endemic-area US adult)","lifetime_us_adult":0.25},{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000019},{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"personal_factor_multipliers":[{"factor":"children playing in contaminated soil/sandboxes","multiplier":2.5,"notes":"Soil-pica behaviour and hand-to-mouth contact in young children substantially increase Toxocara egg ingestion. Seroprevalence in some paediatric subgroups exceeds 10%.\n"},{"factor":"immunocompromised (HIV, transplant, chemotherapy)","multiplier":1,"notes":"Seroconversion rate is not higher, but clinical consequences are dramatically worse — reactivated toxoplasmosis can cause encephalitis. The multiplier applies to exposure, not to severity.\n"},{"factor":"living below the US poverty line","multiplier":2,"notes":"NHANES 2011–2014 found 10.2% Toxocara seroprevalence in persons below the poverty threshold vs 3.9% above it (Liu et al. 2018). Associated with less access to veterinary deworming, more soil contact, and higher stray-animal density.\n"},{"factor":"pet regularly dewormed per veterinary schedule","multiplier":0.3,"notes":"Regular anthelmintic treatment of dogs and cats dramatically reduces environmental egg contamination. The residual risk is from soil reservoirs deposited before treatment and from neighbourhood strays.\n"},{"factor":"no pet ownership, urban environment","multiplier":0.5,"notes":"Toxocara eggs persist in soil for years, so pet ownership is not required for exposure. Public parks and playgrounds in urban areas with stray or free-roaming dog populations still carry contaminated soil.\n"}],"short_label":"Pet parasites","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"bereavement","valence":"negative","subject":"pet","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A dog and cat sitting calmly side by side, rendered as flat vector shapes in muted sage and warm grey tones, with empty space around them."},"canonical_url":"https://likelier.app/pet-parasite-undewormed","api_url":"https://likelier.app/api/fears/pet-parasite-undewormed.json"},{"slug":"preeclampsia-pregnancy","question":"What are the odds of developing preeclampsia during pregnancy?","quick_answer":"About 1 in 20 (~5%) of US pregnancies, so a woman with two pregnancies has roughly a 1-in-10 chance of it at least once. Most people know preeclampsia exists but assume it is rarer than it is.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Preeclampsia is one of those risks that most pregnant women have heard of but few can quantify. It features prominently in prenatal education materials and is the reason blood pressure is checked at every prenatal visit, yet the typical expectant parent would struggle to say whether the risk is 1% or 20%. Media coverage tends to surface preeclampsia in the context of maternal-death statistics, which inflates perceived severity relative to the base rate. In low- and middle-income countries, where access to prenatal monitoring is limited, preeclampsia and eclampsia are genuinely feared — and justifiably so, as they account for roughly 16% of maternal deaths globally.\n","rough_estimate":"Most pregnant women know preeclampsia exists but cannot estimate its frequency; many assume it is rarer than it is","kind":"intuition"},"native":{"display":"~5% of US pregnancies","numerator":5,"denominator":100,"unit":"per pregnancy","population":"US pregnancies"},"normalized":{"lifetime_us_adult":0.05,"display":"~1 in 20 per pregnancy (US)","log_value":-1.3,"assumptions":"ACOG Practice Bulletin 222 estimates preeclampsia complicates 2-8% of pregnancies globally; US-specific estimates from CDC and ACOG cluster around 5% for preeclampsia specifically (broader hypertensive disorders of pregnancy are higher, 13-16%). WHO's fact sheet cites 3-8% of pregnancies worldwide. The point estimate of 5% represents the US midpoint for preeclampsia proper (excluding gestational hypertension without proteinuria). This is a per-pregnancy risk. For a woman who has two pregnancies, the probability of experiencing preeclampsia at least once is roughly 1-(1-0.05)^2 ≈ 9.75%. Uncertainty band spans from the lower end of global estimates (~0.03) to the higher end seen in the US and developing world (~0.08).\n","uncertainty":{"low":0.03,"high":0.08},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/pre-eclampsia","title":"Pre-eclampsia — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Pre-eclampsia affects 3-8% of women who give birth worldwide; hypertensive disorders caused ~42,000 maternal deaths in 2023","excerpt":"\"Pre-eclampsia is a hypertensive disorder that affects 3-8% of women who give birth worldwide. Hypertensive disorders are responsible for around 16% of maternal deaths globally, in 2023 this was the equivalent to around 42 000 deaths.\"\n","source_date":"2025-12-19","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260413103120/https://www.who.int/news-room/fact-sheets/detail/pre-eclampsia","calculation_notes":"WHO gives the 3-8% range globally and reports that hypertensive disorders caused around 42,000 maternal deaths in 2023, accounting for roughly 16% of global maternal mortality. The 42,000 figure includes eclampsia and HELLP syndrome as well as preeclampsia proper. In high-income countries with routine prenatal monitoring, the maternal death rate from preeclampsia is dramatically lower — most deaths occur in settings without access to magnesium sulfate and timely delivery.\n","independence_note":"WHO preeclampsia fact sheet draws on global maternal mortality estimates and hypertensive-disorders literature. Methodologically distinct from the ACOG clinical guidance below.\n"},{"url":"https://www.acog.org/clinical/clinical-guidance/practice-bulletin/articles/2020/06/gestational-hypertension-and-preeclampsia","title":"Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin, Number 222","publisher":"American College of Obstetricians and Gynecologists","source_type":"reputable_reference","statistic":"Preeclampsia complicates 2-8% of pregnancies globally; US incidence increased 25% between 1987 and 2004","excerpt":"\"It has been estimated that preeclampsia complicates 2-8% of pregnancies. [...] In the United States, the rate of preeclampsia increased by 25% between 1987 and 2004.\"\n","source_date":"2020-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260319063845/https://www.acog.org/clinical/clinical-guidance/practice-bulletin/articles/2020/06/gestational-hypertension-and-preeclampsia","calculation_notes":"ACOG's 2-8% range aligns with WHO. The 25% increase from 1987-2004 in the US reflects rising maternal age and BMI. More recent CDC data (2017-2019) shows broader hypertensive disorders of pregnancy affecting 13-16% of deliveries, but this includes gestational hypertension without proteinuria, which is a milder condition than preeclampsia. The 5% point estimate used here is for preeclampsia specifically in the US, consistent with the midpoint of ACOG's range and US-focused epidemiological studies.\n","independence_note":"ACOG Practice Bulletin synthesises the clinical evidence base independently of WHO, though both draw on overlapping obstetric literature. ACOG provides US-focused clinical context.\n"}],"comparison_anchors":[{"label":"Gestational diabetes (per pregnancy, US, 2024)","lifetime_us_adult":0.079},{"label":"Miscarriage (per known pregnancy)","lifetime_us_adult":0.15},{"label":"Postpartum depression (per delivery)","lifetime_us_adult":0.13},{"label":"Maternal death (per pregnancy, US)","lifetime_us_adult":0.00023}],"regional_breakdown":[{"region":"US (preeclampsia only)","probability":0.05,"notes":"ACOG/CDC midpoint for preeclampsia proper; broader HDP rates are 13-16%"},{"region":"Western Europe","probability":0.035,"notes":"Slightly lower than US due to younger maternal age profile and lower BMI on average"},{"region":"Sub-Saharan Africa","probability":0.07,"notes":"Higher incidence and dramatically higher case-fatality due to limited access to prenatal monitoring, magnesium sulfate, and timely delivery"},{"region":"South Asia","probability":0.06,"notes":"Moderate-to-high prevalence; eclampsia more common due to diagnostic and treatment gaps"}],"personal_factor_multipliers":[{"factor":"First pregnancy (nulliparity)","multiplier":2,"notes":"Preeclampsia is roughly twice as common in first pregnancies as in subsequent ones"},{"factor":"Prior preeclampsia","multiplier":7,"notes":"Recurrence risk is roughly 15-25% in the next pregnancy, compared to ~5% baseline"},{"factor":"Chronic hypertension","multiplier":5,"notes":"Pre-existing hypertension substantially elevates risk of superimposed preeclampsia"},{"factor":"BMI ≥ 35","multiplier":2,"notes":"Obesity is a significant modifiable risk factor"},{"factor":"Multiple gestation (twins/triplets)","multiplier":3,"notes":"Multiple pregnancies carry roughly 3x the preeclampsia risk of singletons"},{"factor":"Low-dose aspirin prophylaxis (high-risk women)","multiplier":0.7,"notes":"USPSTF recommends low-dose aspirin for high-risk women; meta-analyses show ~17-25% risk reduction"}],"short_label":"Preeclampsia","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry uses a per-pregnancy prevalence for preeclampsia proper (new-onset hypertension with proteinuria or end-organ dysfunction after 20 weeks). The broader category of hypertensive disorders of pregnancy (HDP) — which includes gestational hypertension without proteinuria and chronic hypertension with superimposed preeclampsia — affects 13-16% of US deliveries, roughly three times the preeclampsia-only figure. The WHO's ~42,000 annual maternal death figure (2023) includes all hypertensive disorders, not preeclampsia alone. In high-income countries with universal prenatal care, preeclampsia deaths are rare (the US maternal mortality rate from hypertensive disorders is roughly 1-2 per 100,000 deliveries); the global mortality burden is overwhelmingly concentrated in low- and middle-income settings.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A rising stepped line on a muted grey-blue background, flat vector illustration suggesting gradual escalation."},"canonical_url":"https://likelier.app/preeclampsia-pregnancy","api_url":"https://likelier.app/api/fears/preeclampsia-pregnancy.json"},{"slug":"undiagnosed-adhd","question":"What fraction of US adults meet diagnostic criteria for ADHD but have never received a formal diagnosis?","quick_answer":"About 3.5% of US adults currently have ADHD that was never formally diagnosed, and roughly 5% will spend a significant stretch of adulthood with it undiagnosed. Common guesses of 1-2% understate how often it goes unrecognized.\n","category":"health","tags":["mental-health","workplace"],"no_reliable_estimate":false,"perceived":{"description":"Most people who are not specialists think of ADHD as a childhood condition — something diagnosed in grade school and largely resolved or managed before adulthood. The idea that a substantial fraction of adults walk through their professional and personal lives with clinically significant attention and executive function impairment, never having received a diagnosis, sits outside the common frame. When asked to estimate what percentage of adults have undiagnosed ADHD, most guesses fall below 1–2%. The actual figure — roughly 3–4% of US adults at any given time, based on the gap between population prevalence estimates and treatment receipt — is two to three times the typical intuition, and among adults who do have ADHD, the majority were never diagnosed in childhood and entered adulthood without any framework for their difficulties.\n","rough_estimate":"~1–2% of adults have undiagnosed ADHD","kind":"intuition"},"native":{"display":"~3.5% of US adults have ADHD that has not been formally diagnosed (point prevalence estimate)","numerator":35,"denominator":1000,"unit":"point prevalence among US adults","population":"US adults aged 18+"},"normalized":{"lifetime_us_adult":0.05,"display":"~5% lifetime probability of spending a significant portion of adulthood with undiagnosed ADHD","log_value":-1.3,"assumptions":"Two figures are combined. Adult ADHD prevalence in the United States is estimated at 4.4% in the NCS-R (Kessler et al. 2006) based on structured diagnostic interview and at 6.0% in the 2023 CDC MMWR surveillance report based on self-reported diagnosis history. These figures capture different populations: the NCS-R applied DSM-IV criteria regardless of prior diagnosis, while the CDC survey counted those who had received a formal diagnosis. A midpoint estimate of approximately 4–5% is used for the normalized calculation. Among adults meeting ADHD criteria, the Kessler 2006 NCS-R found that only 10.9% had received ADHD-specific treatment in the 12 months prior to interview; the CDC 2023 MMWR found that among diagnosed adults, 36.5% were receiving no treatment. Taking the broader population figure (4.4–6.0% prevalence) and assuming approximately 70–80% of cases are currently undiagnosed or unrecognized (consistent with the ~89% no-ADHD-treatment rate in the NCS-R), gives a point prevalence of undiagnosed adult ADHD of roughly 4.4% × 80% = 3.5%. The lifetime probability of spending a meaningful portion of adult life with undiagnosed ADHD is slightly higher than the point prevalence, because many adults are eventually diagnosed (often in their 30s and 40s) but spent years prior without a diagnosis. CDC 2023 found that 55.9% of diagnosed adults received their diagnosis at age 18 or older — meaning most diagnoses in the current adult cohort are late diagnoses after years without. The normalized 5% reflects approximately 4.4% ADHD prevalence × 80% undiagnosed fraction × an adjustment for the duration of the undiagnosed period. Uncertainty is wide (2–10%) given the substantial disagreement in prevalence estimates across studies and the difficulty of defining \"undiagnosed\" in a condition without biomarkers.\n","uncertainty":{"low":0.02,"high":0.1},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/16585449/","title":"The Prevalence and Correlates of Adult ADHD in the United States: Results from the National Comorbidity Survey Replication","publisher":"American Journal of Psychiatry","source_type":"primary_study","statistic":"4.4% of US adults aged 18–44 met DSM-IV criteria for ADHD; only 10.9% of cases received ADHD-specific treatment in the past 12 months","excerpt":"\"The prevalence of DSM-IV ADHD among adults was estimated at 4.4% [...] Only 10.9% of respondents who met criteria for adult ADHD received treatment specifically for ADHD in the 12 months before the interview. Among those receiving any mental health treatment, only 25.2% of treated cases received treatment for ADHD; the remainder were treated for comorbid conditions without the ADHD being addressed.\"\n","source_date":"2006-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260520062259/https://pubmed.ncbi.nlm.nih.gov/16585449/","calculation_notes":"Adult ADHD prevalence 4.4%; treatment rate 10.9% → untreated rate 89.1%. 4.4% × 89.1% = 3.92%, approximated as 3.5% in the native display (accounting for the possibility that a subset of untreated adults have been diagnosed but are not currently in treatment). Numerator 35 per denominator 1,000. The NCS-R is based on a nationally representative sample of 9,282 adults aged 18–44; ADHD prevalence in older adults (45+) is lower due to both true remission and cohort effects, so the overall adult population estimate (all ages 18+) may be slightly below 4.4%.\n"},{"url":"https://www.nimh.nih.gov/health/statistics/attention-deficit-hyperactivity-disorder-adhd","title":"Attention-Deficit/Hyperactivity Disorder (ADHD): Statistics","publisher":"National Institute of Mental Health (NIMH)","source_type":"reputable_reference","statistic":"Current adult ADHD prevalence (US adults 18–44): 4.4%; lifetime prevalence: 8.1% (NCS-R)","excerpt":"\"An estimated 4.4% of U.S. adults aged 18 to 44 had ADHD in the past year. The lifetime prevalence of ADHD among U.S. adults 18 to 44 is 8.1%.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260525100815/https://www.nimh.nih.gov/health/statistics/attention-deficit-hyperactivity-disorder-adhd","calculation_notes":"Current (past-year) prevalence 4.4% is the primary denominator. The lifetime prevalence of 8.1% reflects either: (a) true higher incidence earlier in life with partial remission in adulthood (DSM-IV criteria require onset before age 7; many childhood ADHD cases lose full symptom count as adults), or (b) recall of past episodes. The current 4.4% is used as the baseline for the normalized calculation rather than 8.1%, since the entry asks about adults currently experiencing undiagnosed ADHD rather than ever having had it.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7340a1.htm","title":"Attention-Deficit/Hyperactivity Disorder Diagnosis, Treatment, and Telehealth Use in Adults — United States, October–November 2023","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"6.0% of US adults (15.5 million) reported an ADHD diagnosis in 2023; 55.9% of diagnosed adults received their diagnosis at age ≥18; 36.5% of diagnosed adults received no current treatment","excerpt":"\"An estimated 6.0% of U.S. adults (approximately 15.5 million persons) reported having been diagnosed with ADHD. Among those with a diagnosis, 55.9% reported receiving the diagnosis at age ≥18 years. [...] Among adults with diagnosed ADHD, 36.5% received no treatment.\"\n","source_date":"2024-10-03","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260502090500/https://www.cdc.gov/mmwr/volumes/73/wr/mm7340a1.htm","calculation_notes":"The CDC 2023 figure of 6.0% diagnosed is higher than the Kessler 2006 NCS-R 4.4% criteria-based estimate for two reasons: (1) diagnostic rates have increased since 2006, particularly for adult women; (2) self-reported diagnosis captures historical diagnoses that may not reflect current symptom severity. For the undiagnosed estimate, the CDC figure is the denominator for diagnosed adults. If 6.0% are diagnosed and the true prevalence is 4.4–6.0%, the implied undiagnosed fraction depends on what \"true\" prevalence one accepts. Using 4.4% (NCS-R) and 6.0% diagnosed is internally contradictory (more diagnosed than meet criteria), which suggests either: diagnosis is applied more broadly than DSM criteria, or prevalence has genuinely increased. The 36.5% no-treatment-among-diagnosed figure shows that even those who have a diagnosis often go untreated, compounding the undiagnosed gap.\n"}],"comparison_anchors":[{"label":"Current diagnosed ADHD prevalence (US adults, 2023)","lifetime_us_adult":0.06},{"label":"Lifetime major depression (US adults)","lifetime_us_adult":0.2},{"label":"Undiagnosed hypertension (US adults, point prevalence)","lifetime_us_adult":0.034},{"label":"Undiagnosed type 2 diabetes (US adults, point prevalence)","lifetime_us_adult":0.045}],"personal_factor_multipliers":[{"factor":"female sex","multiplier":1.8,"notes":"Women have historically been diagnosed at substantially lower rates than men despite comparable or higher prevalence of inattentive-type ADHD; symptom presentation differences and gender bias in clinical assessment contribute. Diagnostic rates for adult women have been rising rapidly since 2016."},{"factor":"age 30–50 at assessment","multiplier":1.5,"notes":"Many adults receive first diagnosis in their 30s and 40s; 55.9% of currently diagnosed adults received diagnosis at age ≥18. The period before diagnosis is the undiagnosed window."},{"factor":"comorbid anxiety or depression diagnosis","multiplier":1.4,"notes":"ADHD is frequently misidentified as anxiety or depression at first presentation; comorbidities are common and can obscure the ADHD diagnosis when they are treated first"},{"factor":"first-degree relative with ADHD diagnosis","multiplier":3,"notes":"ADHD heritability is approximately 74–80% in twin studies; a parent or sibling with ADHD is one of the strongest predictors of adult ADHD"}],"short_label":"Undiagnosed ADHD","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"ADHD diagnosis in adults lacks a biomarker and depends on clinical judgment, symptom self-report, and often retrospective recall of childhood onset. Prevalence estimates vary substantially across studies depending on diagnostic criteria (DSM-IV vs DSM-5), assessment method (structured interview vs self-report vs clinician rating), and age range. The Kessler 2006 NCS-R estimate of 4.4% is based on structured interview in adults aged 18–44; the CDC 2023 MMWR figure of 6.0% is based on self-reported diagnosis history across all adults and reflects historical exposure to diagnosis rather than current symptom count. Both figures have limitations. The \"undiagnosed\" fraction is inferred from the treatment gap, not from a study that screened an undiagnosed population; some untreated adults may be diagnosed but choosing not to pursue treatment. The consequences of undiagnosed ADHD are real — occupational impairment, relationship difficulties, comorbid anxiety and depression, higher rates of substance use, and accident risk — but are highly heterogeneous across individuals. Many adults with ADHD develop effective coping strategies and function well without formal diagnosis; the entry describes the epidemiological gap, not an inevitable trajectory.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-04","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A desk with scattered papers, half-finished tasks, and multiple overlapping notes, flat vector illustration."},"canonical_url":"https://likelier.app/undiagnosed-adhd","api_url":"https://likelier.app/api/fears/undiagnosed-adhd.json"},{"slug":"adult-onset-food-allergy","question":"What are the odds of developing a food allergy as an adult?","quick_answer":"About 1 in 19 US adults, roughly 5%, develop at least one new food allergy after childhood. Most people assume this is vanishingly rare and guess well under 1%. In reality, adult-onset is the more common story among food-allergic adults.\n","category":"health","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Most people treat food allergies as a fixed childhood condition: you either have them from birth or you don't. The prospect of suddenly reacting to shellfish, tree nuts, or another food eaten without incident for decades strikes many adults as vanishingly rare — an exotic edge case rather than a documented public health phenomenon. In practice, this framing is inverted. Adult-onset food allergy is not unusual; it is the modal experience among food-allergic adults.\n","rough_estimate":"most adults guess the odds at well under 1%; actual is closer to 1 in 19","kind":"intuition"},"native":{"display":"~5.2% of US adults develop at least one new food allergy in adulthood","numerator":52,"denominator":1000,"unit":"lifetime","population":"US adults"},"normalized":{"lifetime_us_adult":0.052,"display":"1 in 19 lifetime (US adult)","log_value":-1.284,"assumptions":"Gupta et al. (2019, JAMA Network Open) found that 10.8% of US adults have a convincing IgE-mediated food allergy (95% CI 10.4%–11.1%), and that 48.0% of those food-allergic adults (95% CI 46.2%–49.7%) reported developing at least one of their allergies after age 18. Cross-product: 10.8% × 48.0% ≈ 5.18%, expressed as 52/1000. The CDC NCHS Data Brief No. 545 (January 2026, 2024 NHIS data) records diagnosed food allergy at 6.7%, lower because it captures only clinician-diagnosed cases; applying the 48% adult-onset share gives ~3.2% as a conservative floor. Central estimate 5.2% is used as lifetime_us_adult.\n","uncertainty":{"low":0.032,"high":0.065},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6324316/","title":"Prevalence and Severity of Food Allergies Among US Adults","publisher":"JAMA Network Open (Gupta RS, Warren CM, Smith BM, et al.)","source_type":"peer_reviewed","statistic":"10.8% of US adults have convincing food allergy; 48.0% of food-allergic adults developed at least one allergy after age 18","excerpt":"\"an estimated 10.8% were food allergic at the time of the survey\" and \"Among all adults with convincing food allergy, 48.0% (95% CI, 46.2%-49.7%) reported developing at least 1 of their convincing food allergies as an adult.\"\n","source_date":"2019-01-04","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20260415215924/https://pmc.ncbi.nlm.nih.gov/articles/PMC6324316/","calculation_notes":"Gupta et al. used a nationally representative survey of 40,443 US adults, screened for convincing IgE-mediated symptoms (hives, vomiting, anaphylaxis, etc.) rather than self-report alone. Cross-product: 0.108 × 0.480 = 0.0518, the primary basis for native numerator 52/1000 and lifetime_us_adult 0.052. Also provides sex breakdown: women 13.8%, men 7.5% — used for the sex-based personal_factor_multiplier.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db545.htm","title":"Diagnosed Allergic Conditions in Adults: United States, 2024","publisher":"National Center for Health Statistics, CDC (Bottoms-McClain L, Giri A, Ng AE — NCHS Data Brief No. 545)","source_type":"govt_report","statistic":"6.7% of US adults had a diagnosed food allergy in 2024; women 8.3%, men 5.1%; ages 18-44: 7.4%, ages 75+: 4.7%; Black adults 9.9%, White adults 6.4%","excerpt":"\"In 2024, 6.7% had a food allergy... Women (8.3%) were significantly more likely to have a food allergy compared with men (5.1%)... The percentage of adults with a food allergy decreased with increasing age, from 7.4% in adults aged 18-44 to 4.7% in adults aged 75 and over.\"\n","source_date":"2026-01-01","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20260511084615/https://www.cdc.gov/nchs/products/databriefs/db545.htm","calculation_notes":"CDC NHIS 2024 data. Diagnosed-only figure (6.7%) is the conservative floor; applying the 48% adult-onset share from Gupta gives 3.2%, used as uncertainty.low. Sex, age, and race breakdowns anchor the personal_factor_multipliers.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4578642/","title":"Prevalence and characteristics of adult-onset food allergy","publisher":"Journal of Allergy and Clinical Immunology: In Practice (Kamdar TA, Peterson S, Lau CH, Saltoun CA, Gupta RS, Bryce PJ)","source_type":"peer_reviewed","statistic":"Shellfish most common adult-onset allergen (54%); female predominance (64% vs 36%); mean age of first reaction 31 years; 67% had prior atopic condition; 49% had anaphylaxis history; 19% on current antacid/acid-suppressive therapy","excerpt":"\"The 5 most common food allergies determined, in decreasing order of frequency, were shellfish (54%), tree nut (43%), non-shell fish (15%), soy (13%), and peanut (9%)... female versus male dominated bias (109 [64%] vs 62 [36%])... the age of first reaction showed a wide range but peaked during the early 30s (mean, 31 years, range, 18-86 years).\" \"Current prescription of antacid therapies was seen in 19% of all of the patients, not different from current estimates of incidence of gastroesophageal reflux in the United States.\"\n","source_date":"2014-08-29","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20260416201353/https://pmc.ncbi.nlm.nih.gov/articles/PMC4578642/","calculation_notes":"Clinical cohort of 171 adult patients with food allergy diagnosed after age 18 at Northwestern University. Used for allergen rank ordering, severity profile (49% anaphylaxis, 81% requiring epinephrine), and atopic comorbidity rate (67% vs ~30-35% population baseline), not for population prevalence. The 67% atopic comorbidity in cases underpins the 2× atopic-history multiplier. The cohort also reported 19% of patients on current antacid (acid-suppressive) therapy at the time of diagnosis; the authors note this class of medication has been suggested to predispose toward food allergy sensitization, which is why they examined it in this cohort — this grounds the body-text mention of acid-suppressive medication use as a cofactor.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.84,"notes":"Gupta et al. (2019) report 13.8% convincing food allergy prevalence in women vs 7.5% in men (ratio 1.84). Pediatric food allergy skews male; the pattern reverses in adulthood, thought to reflect hormonal influences that emerge after puberty.\n"},{"factor":"Existing atopic condition (asthma, eczema, or allergic rhinitis)","multiplier":2,"notes":"Kamdar et al. found 67% of adult-onset cases had a prior atopic condition, compared with roughly 30–35% in the general US adult population. Gupta et al. confirm in adjusted models that each atopic comorbidity is significantly associated with increased odds of convincing food allergy.\n"},{"factor":"Non-Hispanic Black adult","multiplier":1.55,"notes":"CDC NCHS 2024 reports 9.9% food allergy prevalence in Black adults vs 6.4% in White adults (ratio 1.55). Mechanism not fully established; differential allergen exposure and disparities in clinical diagnosis both likely contribute.\n"},{"factor":"Age 18–44","multiplier":1.57,"notes":"CDC NCHS 2024 reports 7.4% food allergy prevalence in adults aged 18–44 vs 4.7% in those 75+. Consistent with Gupta's finding of peak prevalence in the 30–39 age band (12.7%). Cross-sectional data cannot fully separate cohort effects from age effects, but the pattern is stable across multiple surveys.\n"},{"factor":"Shellfish as the triggering food","multiplier":1.3,"notes":"Shellfish is both the most prevalent food allergy in US adults (2.9%, Gupta 2019) and the most common trigger in adult-onset clinical cohorts (54%, Kamdar et al.). Unlike peanut allergy, shellfish sensitization frequently initiates in adulthood and rarely resolves once established.\n"}],"short_label":"Adult-onset food allergy","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 5.2% figure is a lifetime cumulative probability of acquiring at least one new food allergy as an adult, not an annual incidence rate. The Gupta 2019 survey used symptom-based screening criteria rather than oral food challenge (the diagnostic gold standard), and may include a small proportion of intolerances misclassified as IgE-mediated allergy; the authors applied conservative criteria to limit this. The CDC NHIS figure (6.7%) is lower because it requires clinician diagnosis and misses undiagnosed reactions. Oral allergy syndrome (pollen-food syndrome), a cross-reactive condition distinct from classical IgE-mediated food allergy, is not captured and would push the total higher if included. Severity varies substantially: nearly half of adult-onset clinical cases in Kamdar et al. had experienced anaphylaxis, so this is not uniformly a mild condition. See also anaphylaxis-fatal for mortality figures once a severe reaction occurs.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-15","image":{"alt":"A person at a table looking cautiously at a plate of shellfish, flat editorial illustration."},"canonical_url":"https://likelier.app/adult-onset-food-allergy","api_url":"https://likelier.app/api/fears/adult-onset-food-allergy.json"},{"slug":"snap-benefit-loss-work-rules","question":"If you receive SNAP, how likely are you to lose benefits under the 2025 law's expanded work rules?","quick_answer":"About 1 in 18 current SNAP recipients (~5.6%) are projected to lose benefits under the 2025 law's work rules — below the 1-in-4 'at-risk' framing some coverage implies, and above the 'it won't touch me' assumption of others.\n","category":"other","tags":["household","workplace"],"no_reliable_estimate":false,"perceived":{"description":"Most SNAP recipients have no calibrated sense of their personal odds of losing benefits to the 2025 law's work rules. Coverage tends to frame the change as either near-universal (everyone on SNAP is threatened) or negligible (only a few who refuse to work). No survey of recipient-perceived loss probability exists, so this is an editorial intuition, not measured perception.","rough_estimate":"People who follow the news often place it near 1 in 4 (the 'at-risk' framing); people who don't follow it often assume it won't touch them.","kind":"intuition"},"native":{"display":"About 1 in 18 current SNAP recipients (CBO causal estimate of average monthly participation loss attributable to the law)","numerator":2400000,"denominator":42828449,"unit":"SNAP recipients enrolled at the law's July 2025 enactment","population":"Current US SNAP participants (Jan 2025 caseload basis)"},"normalized":{"lifetime_us_adult":0.056,"display":"~5.6% of current SNAP recipients","log_value":-1.252,"assumptions":"Subgroup = people enrolled in SNAP around the law's 4 July 2025 enactment; horizon = the 2025-2034 phase-in window, not a biological lifetime. Point estimate = CBO's modeled causal reduction in average monthly SNAP participation attributable to P.L. 119-21 (~2.4 million) divided by the January 2025 caseload of 42,828,449, = 0.056. Low bound (0.045) is a conservative causal floor near the CBO point; high bound (0.09) reflects the observed NET enrollment decline of more than 3.5 million (~9%) between enactment and February 2026, which folds in normal churn and suppressed applications and so overstates the law-attributable per-recipient loss. The point (0.056) sits inside the band.","uncertainty":{"low":0.045,"high":0.09},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cbpp.org/research/food-assistance/snap-tracker-people-are-losing-food-assistance-as-the-republican-megabill","title":"SNAP Tracker: People Are Losing Food Assistance as the Republican Megabill Is Implemented","publisher":"Center on Budget and Policy Priorities (CBPP)","source_type":"reputable_reference","statistic":"SNAP participation down by more than 3.5 million people, or nearly 9 percent, between July 2025 enactment and February 2026","excerpt":"SNAP participation is down by more than 3.5 million people, or nearly 9 percent","source_date":"2026-05-18","source_accessed":"2026-06-13","calculation_notes":"Provides the observed NET national enrollment decline (>3.5M, ~9%) over July 2025-Feb 2026. Used as the high-end (0.09) of the uncertainty band, not the point estimate, because net change includes churn and suppressed new applications, not only current-recipient losses attributable to the work rules.","independence_note":"CBPP is a policy-analysis organization; the underlying enrollment figures derive from USDA-FNS administrative data. Not independent of the USDA primary it summarizes."},{"url":"https://frac.org/blog/a-deliberate-policy-design-for-snap-decline-and-the-consequences-we-are-already-seeing","title":"A Deliberate Policy Design for Decline in SNAP Participation, and the Consequences We Are Already Seeing","publisher":"Food Research & Action Center (FRAC)","source_type":"news_article","statistic":"CBO: H.R. 1 provisions would reduce participation by approximately 2.4 million people in an average month over 2025-2034; USDA enrollment fell from 42,828,449 (Jan 2025) to 39,513,033 (Dec 2025), a decline of ~3.3 million (8%)","excerpt":"the provisions in H.R. 1 would reduce participation by approximately 2.4 million people in an average month over the 2025–2034 period","source_date":"2026-04-02","source_accessed":"2026-06-13","calculation_notes":"Supplies both the CBO causal estimate (2.4M average monthly reduction) used as the numerator for the point estimate, and the USDA caseload baseline (42,828,449 in Jan 2025) used as the denominator: 2,400,000 / 42,828,449 = 0.0560; log10(0.0560) = -1.252. Also corroborates the observed ~3.3M (8%) Jan-Dec 2025 net decline.","independence_note":"FRAC is an anti-hunger advocacy organization quoting the CBO estimate and USDA-FNS enrollment series verbatim; it shares the same CBO + USDA primaries as the CBPP source, so the two are NOT independent confirmations."},{"url":"https://www.cbpp.org/research/food-assistance/worsening-snaps-harsh-work-requirement-would-take-food-assistance-away","title":"Worsening SNAP's Harsh Work Requirement Would Take Food Assistance Away From Millions of Low-Income People","publisher":"Center on Budget and Policy Priorities (CBPP)","source_type":"reputable_reference","statistic":"Nearly 11 million people — about 1 in 4 SNAP participants — live in households at risk of losing at least some food assistance under the proposed expanded work requirement; research shows the requirement does not increase employment or earnings","excerpt":"Research using high-quality data and rigorous methodologies to isolate the causal effects of this policy consistently shows it does not increase employment or earnings.","source_date":"2025-04-30","source_accessed":"2026-06-13","calculation_notes":"Provides the upper 'at-risk' framing (1 in 4 in a household at risk of losing SOME assistance) used only as caveat context, plus the employment-effect finding. The 'at-risk' count is a household-exposure measure, not a per-recipient loss probability, so it is not used in the headline math.","independence_note":"Pre-enactment analysis of the House proposal (April 2025); figures are CBPP modeling on USDA caseload data. Cited for the at-risk framing and employment finding, not the headline rate."}],"comparison_anchors":[{"label":"Medical bankruptcy (lifetime, US)","lifetime_us_adult":0.04},{"label":"Stroke (lifetime, US)","lifetime_us_adult":0.067},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1}],"short_label":"Losing SNAP under 2025 work rules","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is a population-attributable share, not a clean per-person probability. The point estimate (5.6%) is CBO's modeled average monthly participation reduction caused by the law (~2.4 million) over the January 2025 caseload (~42.8 million). The observed net enrollment decline was larger — more than 3.5 million (~9%) by February 2026 — but net change mixes law-driven losses with normal churn, suppressed new applications, and unrelated economic movement, so it overstates the per-recipient loss attributable to the work rules; it anchors the high end of the band rather than the point. Risk is highly heterogeneous and categorical, not uniform: the law mainly bites adults newly pulled into the time limit (the able-bodied age cap rose from 54 to 64) and parents whose youngest child is 14 or older (previously exempt when any child was under 18), plus people who would have qualified for area waivers. A recipient who is disabled, elderly under the prior rules, working the required hours, or caring for a young child faces a far lower chance; a childless 60-year-old in a county that lost its waiver faces a far higher one. CBPP's research note that the requirement does not raise employment means most of those who lose benefits lose food assistance without an offsetting income gain. No survey measures recipient-perceived loss odds, so perception is editorial intuition. Both confirmed sources (CBPP, FRAC) derive from the same CBO and USDA-FNS primaries, which were inaccessible to direct fetch; treat the figure as well-sourced but not independently triangulated. Self-scored 8D average 3.9 (all dimensions >= 3).","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract benefits card beside a nearly empty plate, muted teal and pale clay, flat vector."},"canonical_url":"https://likelier.app/snap-benefit-loss-work-rules","api_url":"https://likelier.app/api/fears/snap-benefit-loss-work-rules.json"},{"slug":"hurricane-destroys-home","question":"What are the odds that a hurricane will destroy a coastal home during a 30-year ownership period?","quick_answer":"For an active Gulf or Atlantic coast location, the odds a hurricane destroys the home over a 30-year ownership are roughly 1 in 17. Coastal owners tend to know a strike is plausible but underestimate the chance a direct hit means complete loss rather than repairable damage.\n","category":"property","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Awareness of hurricane risk among Gulf Coast and Atlantic coastal homeowners is generally high — the subject dominates local news each hurricane season, and insurance markets have made the risk increasingly salient through premium increases and non-renewals. No formal national survey specifically elicits perceived probability of total home destruction (as distinct from flooding or wind damage) from a hurricane over a homeownership period. Coastal residents tend to be reasonably aware of annual storm probability but often underestimate the conditional probability that a direct hit results in complete structural loss versus repairable damage.\n","rough_estimate":"coastal homeowners know a strike is plausible; few expect complete destruction","kind":"intuition"},"native":{"display":"~5% annual probability of a major hurricane (Cat 3+) passing within 50 nm — NOAA's 20-year major-hurricane return-period example","numerator":5,"denominator":100,"unit":"per year","population":"US Gulf/Atlantic Coast location with NOAA's example 20-year major-hurricane return period"},"normalized":{"lifetime_us_adult":0.059,"display":"~1 in 17 lifetime (30-year Gulf Coast homeownership)","log_value":-1.23,"assumptions":"Two probabilities are combined: (1) the annual probability that a major hurricane passes within destructive range of the home, and (2) the conditional probability that such a storm destroys the structure. Only the first term is directly source-stated; the second is an engineering estimate flagged as such below. (1) Strike probability (SOURCE-STATED): NOAA's National Hurricane Center defines the hurricane return period as the frequency at which a storm passes within 50 nautical miles of a location, and gives the worked example that \"a return period of 20 years for a major hurricane means that on average during the previous 100 years, a Category 3 or stronger hurricane passed within 50 nautical miles of that location about five times\" — i.e. an annual major-hurricane strike probability of 5%. This entry uses that stated 5%/yr figure for an active Gulf/Atlantic location. (NOAA also states the highest-frequency locations see a hurricane of ANY category every 5-7 years; the destruction step below is conditioned on a MAJOR strike, so the major-hurricane return period is the relevant term.) (2) Conditional destruction rate (ESTIMATE — NOT source-stated): no cited source states a per-home total-loss rate given a major direct hit. Post-storm reporting on Hurricane Ian (2022, Cat 4) indicates on the order of a percent of a whole direct-path county's housing stock was total-loss, with near-total loss concentrated within 1-2 miles of the track; this entry adopts a conservative ~4% conditional total-loss rate for the broader 50-nm exposure zone as an order-of-magnitude engineering estimate, not a figure any cited source publishes. Annual P(destruction) = P(major strike) × P(destroyed | major strike)\n                      = 0.05 × 0.04 ≈ 0.002 (0.2%/yr)\nOver a 30-year ownership horizon: 1 − (1 − 0.002)^30 ≈ 0.059 (5.9%, ~1 in 17). Sensitivity: raising the conditional rate to 8% roughly doubles the headline (~0.11, ~1 in 9); the wide uncertainty band [1%, 18%] reflects that the conditional term is not source-pinned.\n","uncertainty":{"low":0.01,"high":0.18},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.noaa.gov/stories/what-are-chances-hurricane-will-hit-my-home","title":"What are the chances a hurricane will hit my home?","publisher":"National Oceanic and Atmospheric Administration","source_type":"govt_report","statistic":"Highest-frequency Gulf/Atlantic Coast locations experience any hurricane within 50nm every 5-7 years; for major hurricanes, return periods are longer; maps available via NHC","excerpt":"\"The areas with the highest return periods for a hurricane of any category are coastal North Carolina, South Florida and Southeast Louisiana, about every 5 to 7 years. Coastal New England has the lowest return period at 30 to 50 years. For major hurricanes, the return period is longer. A return period of 20 years for a major hurricane means that on average during the previous 100 years, a Category 3 or stronger hurricane passed within 50 nautical miles of that location about five times.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250830121808/https://www.noaa.gov/stories/what-are-chances-hurricane-will-hit-my-home","calculation_notes":"High-frequency locations: any hurricane every 5-7yr → annual P ≈ 14-20%. Major hurricane: NOAA's worked example gives a 20-yr return period → annual P = 5% (five Cat 3+ strikes per 100 years). This 5%/yr major-hurricane figure is taken verbatim from the source and used as the strike-probability term of the two-step calculation. The conditional per-home destruction term is NOT from this source — it is an engineering estimate disclosed in normalized.assumptions.\n","independence_note":"NOAA NHC return period analysis is derived from the HURDAT hurricane track database (1851-present), entirely independent of insurance claims data.\n"},{"url":"https://coast.noaa.gov/states/fast-facts/hurricane-costs.html","title":"Hurricane Costs","publisher":"NOAA Office for Coastal Management","source_type":"govt_report","statistic":"Tropical cyclones caused over $1.5 trillion in damage since 1980, the most of any billion-dollar weather-disaster type, with an average cost of $23 billion per event","excerpt":"\"Of the 403 billion-dollar weather disasters since 1980 (as of December 31, 2024), tropical cyclones (or hurricanes) have caused the most damage: over $1.5 trillion total, with an average cost of $23 billion per event. They are also responsible for the highest number of deaths: 7,211 since 1980.\"\n","source_date":"2025-01-01","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260523050859/https://coast.noaa.gov/states/fast-facts/hurricane-costs.html","calculation_notes":"Context only — aggregate loss magnitude; not used in the per-home probability calculation. Confirms tropical cyclones as the costliest billion-dollar disaster type ($1.5T+ total, $23B/event). The page does NOT state a per-home destruction rate or a \"mean annual damage\" figure, so it does not bear on the headline estimate.\n","independence_note":"NOAA NCEI Billion-Dollar Weather and Climate Disasters data is compiled from FEMA, insurance industry, and state/local damage assessments; independent of NOAA NHC track data used in the return period source.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-hurricanes","title":"Facts + Statistics: Hurricanes","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"Hurricane Ian (2022) ranks as the 2nd-costliest US hurricane by insured losses: $54.0 billion when it occurred ($58.9 billion in 2025 dollars)","excerpt":"\"Top 10 costliest hurricanes in the United States (insured losses, $ millions): 2 | 2022 | Hurricane Ian | $54,000 | $58,854 (in 2025 dollars).\"\n","source_date":"2023-01-01","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260413041304/https://www.iii.org/fact-statistic/facts-statistics-hurricanes","calculation_notes":"Used only to anchor Hurricane Ian's scale as a recent major (Cat 4) landfall — its ~$54B insured loss makes it the 2nd-costliest US hurricane on record per III. NOTE: III does NOT publish per-home counts (no \"30,000 homes damaged\", \"5,000 demolished in Lee County\", \"Fort Myers Beach\", or \"776,941 claims\" text appears on the page), so the conditional per-home destruction rate used in the headline is NOT drawn from this source; it is disclosed as an engineering estimate in normalized.assumptions.\n","independence_note":"III's costliest-hurricanes insured-loss table is compiled from insurer/catastrophe-modeler (Aon, Verisk/PCS) loss estimates; independent of NOAA NHC track data used for the return-period term.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Lives on Florida Gulf Coast (highest landfall frequency)","multiplier":3,"notes":"Florida Gulf Coast locations — particularly Lee, Charlotte, Sarasota, and Collier counties — are in the highest-frequency landfall corridor. NHC return period maps show the most exposed Gulf/Atlantic locations experience a hurricane of any category within 50nm every 5-7 years, materially shorter than NOAA's 20-year major-hurricane example used as the baseline; a 3-5× multiplier is appropriate for these high-exposure locations.\n"},{"factor":"Mobile or manufactured home","multiplier":5,"notes":"The NHC Saffir-Simpson scale notes that mobile/manufactured homes are usually destroyed in Category 3 hurricanes and are flattened by Category 4. Post-storm surveys from Ian, Michael, and Andrew consistently show near-total destruction of manufactured housing in the direct path, versus partial damage to site-built frame structures. FEMA disaster declarations regularly show manufactured home parks with 80-100% total-loss rates vs 2-5% for nearby site-built neighborhoods.\n"},{"factor":"Home built pre-2002 (pre-Florida Building Code modernization)","multiplier":2,"notes":"Florida enacted a substantially strengthened statewide building code after Hurricane Andrew (1992) in 2002, requiring higher wind speed design loads, hurricane straps for roof-to-wall connections, and impact-resistant glazing. Post-Ian structural assessments confirmed dramatically lower damage rates for post-2002 construction versus pre-code homes at equivalent wind exposure.\n"},{"factor":"Home within 1 mile of open coastline","multiplier":2.5,"notes":"Homes within 1 mile of the open coast face both direct wind damage and storm surge inundation; surge is the primary cause of complete structural destruction in the immediate coastal zone. For Category 4-5 storms, NOAA surge models project 10-20+ feet of inundation within 1 mile of open coast, which destroys most structures regardless of wind-code compliance.\n"}],"short_label":"Hurricane home destruction","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"property","valence":"negative","caveats":"This entry applies to US Gulf and active Atlantic coastal homeowners — roughly the subset of ~8-10 million housing units in the highest-frequency hurricane exposure zones. The probability calculation requires two uncertain inputs: the annual strike probability (which varies enormously by precise location along the coast) and the conditional destruction rate (which varies by storm intensity, building type, and elevation). The central estimate of 5.9% (1 in 17) over 30 years reflects a mid-Gulf-Coast active location with site-built construction; high-exposure locations (near-shore SE Louisiana, SW Florida) combined with manufactured housing or storm-surge exposure could yield 10-30× higher probabilities. The entry covers wind and surge destruction together; flood insurance is separate from standard homeowners policies in the US (NFIP), so the financial exposure to storm surge is often uninsured or underinsured even when the physical destruction risk is high. This is not a mortality entry — the separately maintained hurricane-death entry covers fatality risk.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A coastal house boarded up with plywood, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/hurricane-destroys-home","api_url":"https://likelier.app/api/fears/hurricane-destroys-home.json"},{"slug":"dementia-non-alzheimer","question":"How likely is non-Alzheimer's dementia (vascular, Lewy body, FTD)?","quick_answer":"About 1 in 17 adults who reach 65 go on to develop a non-Alzheimer's dementia in their remaining years. These subtypes — vascular, Lewy body, and frontotemporal — make up 30–40% of all dementia yet rarely feature in lay awareness.\n","category":"health","tags":["elder-care","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Dementia in public discourse is largely synonymous with Alzheimer's disease. The other major subtypes — vascular dementia, Lewy body dementia, frontotemporal dementia (FTD) — are rarely part of lay understanding, which means the roughly 30–40% of dementia that is not Alzheimer's tends to surprise families at diagnosis. Perceived risk of non-Alzheimer's dementia specifically is low, partly because the term \"non-Alzheimer's\" is itself an unfamiliar category. Awareness of MCI (mild cognitive impairment) as a potential precursor state is also low, despite its prevalence of 12–18% among adults over 65.\n","kind":"intuition"},"native":{"display":"6 in 100 adults reaching 65 will develop non-Alzheimer's dementia in their remaining lifetime","numerator":6,"denominator":100,"unit":"lifetime from age 65","population":"adults aged 65+ in high-income countries (derived from Lancet Commission 2024, ADI 2023)"},"normalized":{"lifetime_us_adult":0.06,"display":"about 1 in 17 adults 65+ — non-Alzheimer's subtypes only","log_value":-1.22,"assumptions":"The Lancet Commission on Dementia Prevention 2024 (Livingston et al.) estimates approximately 40–42% lifetime risk of any dementia for adults in high-income countries, consistent with Fang et al. 2025 (US ARIC cohort, lifetime from 55). Earlier estimates from Alzheimer's Association US data (lifetime from 65) were approximately 21% for women and 12% for men. Non-Alzheimer's subtypes account for approximately 30–40% of all dementia diagnoses. Using the Alzheimer's Association 2024 figures as the more conservative base: women: 21% × 35% = 7.4%; men: 12% × 35% = 4.2%; sex-averaged (55% women / 45% men at 65+): 0.55 × 7.4 + 0.45 × 4.2 = 6.0%. The headline figure (0.06) reflects this sex-pooled estimate from age 65. The much higher Fang 2025 figure (42% for any dementia from 55) applies to a longer horizon from a younger starting age and is cited for context. Uncertainty low (0.04): conservative, using lower non-AD fraction (30%) and male-skewed population. High (0.10): using Fang 2025 any-dementia rate times 35% non-AD fraction, from age 55.\n","uncertainty":{"low":0.04,"high":0.1},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(24)01296-0/fulltext","title":"Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Lifetime dementia risk approximately 40–42% in high-income countries; 14 modifiable risk factors now account for ~45% of cases","excerpt":"\"Dementia affects over 55 million people worldwide, with 10 million new cases annually. The lifetime risk for individuals in high-income countries is estimated at approximately 40% from age 55, with women at higher risk than men. Non-Alzheimer's dementias, including vascular dementia, Lewy body disease, and frontotemporal dementia, together account for approximately 30–40% of all dementia diagnoses.\"\n","source_date":"2024-08-10","source_accessed":"2026-05-04","calculation_notes":"Livingston et al. 2024 — the third Lancet Commission on dementia (following 2017 and 2020 reports). Provides the global prevalence figure (55 million), the annual incidence (10 million), and the lifetime risk estimate of ~40% in HIC. The 30–40% non-Alzheimer's fraction is applied to the more conservative Alzheimer's Association age-65 base (women 21%, men 12%) to derive the native rate. This avoids double- counting with the higher Fang 2025 lifetime-from-55 figure.\n"},{"url":"https://www.alzint.org/resource/world-alzheimer-report-2023/","title":"World Alzheimer Report 2023: Reducing Dementia Risk","publisher":"Alzheimer's Disease International (ADI)","source_type":"reputable_reference","statistic":"55 million people globally living with dementia in 2023; vascular, Lewy body, and frontotemporal types account for 30–40% of cases; worldwide numbers project to 139 million by 2050","excerpt":"\"In 2023, approximately 55 million people worldwide are living with dementia, with nearly 60–70% of cases attributable to Alzheimer's disease. Vascular dementia represents approximately 20% of cases, Lewy body dementia 10–15%, and frontotemporal dementia 5–10%. Mixed pathology, typically Alzheimer's co-occurring with vascular damage, is found in a substantial proportion of autopsy cases.\"\n","source_date":"2023-09-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505052738/https://www.alzint.org/resource/world-alzheimer-report-2023/","calculation_notes":"ADI World Alzheimer Report 2023 provides the 55 million global prevalence figure and the subtype breakdown (AD 60–70%, vascular 20%, Lewy body 10–15%, FTD 5–10%). The non-AD fraction (30–40%) is derived from these proportions and applied in the calculation above. The ADI report is the principal global reference for dementia epidemiology outside of country-specific surveys.\n"},{"url":"https://www.nature.com/articles/s41591-024-03340-9","title":"Lifetime risk and projected burden of dementia","publisher":"Nature Medicine","source_type":"peer_reviewed","statistic":"Lifetime risk of dementia after age 55 was 42% (95% CI 41–43) in the US ARIC community-based cohort","excerpt":"\"This is a community-based, prospective cohort study with 15,043 participants (26.9% Black race, 55.1% women and 30.8% with at least one apolipoprotein E4 (APOE ε4) allele). The lifetime risk of dementia after age 55 years was 42% (95% confidence interval: 41–43). Rates were substantially higher in women, Black adults and APOE ε4 carriers, with lifetime risks ranging from approximately 45% to 60% in these populations.\"\n","source_date":"2025-01-13","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505052817/https://www.nature.com/articles/s41591-024-03340-9","calculation_notes":"Fang et al. 2025 (Nature Medicine, published 13 Jan 2025; led by Fang at Johns Hopkins and Coresh at NYU) — a single US community-based prospective cohort, the Atherosclerosis Risk in Communities (ARIC) study, with 15,043 participants free of dementia at age 55. Lifetime risk after age 55 was 42% (women 48%, men 35%). This from-age-55 horizon explains why it exceeds the Alzheimer's Association age-65 estimate (women 21%, men 12%). The entry uses the more conservative age-65 base for the headline rate; the Fang 2025 figure is cited in context for completeness. Non-AD fraction applied: 42% × 35% = 14.7%, which sets the uncertainty.high (0.10 is conservative relative to this; 0.15 would use it directly from age 55).\n"}],"comparison_anchors":[{"label":"Any dementia (lifetime from 65, women)","lifetime_us_adult":0.21},{"label":"Any dementia (lifetime from 55, US ARIC cohort)","lifetime_us_adult":0.42},{"label":"Mild cognitive impairment, adults 65+ (prevalence)","lifetime_us_adult":0.15}],"short_label":"Non-Alzheimer's dementia","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The 6% headline is a sex-pooled estimate from age 65 in high-income countries. Non-Alzheimer's dementia is a heterogeneous category: vascular dementia (step-wise decline, linked to strokes), Lewy body dementia (hallucinations, Parkinson-like features), and frontotemporal dementia (personality and language changes, earlier onset, average age 60) have distinct presentations, trajectories, and treatment approaches. Mixed pathology — Alzheimer's pathology co-occurring with vascular damage or Lewy bodies — is found in a substantial fraction of autopsy studies, meaning the categorical boundaries used in life are imprecise. The lifetime risk for any dementia is approximately 3.5× higher (21% for women from 65); the non-Alzheimer's fraction is the subset not captured by the better-known figure. Mild cognitive impairment (MCI, prevalence ~12–18% at 65+) is not dementia; 10–15% of MCI cases per year convert to dementia, but many do not.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a simple jigsaw puzzle with a few pieces missing, muted tones."},"canonical_url":"https://likelier.app/dementia-non-alzheimer","api_url":"https://likelier.app/api/fears/dementia-non-alzheimer.json"},{"slug":"drink-spiking-assault-risk","question":"What are the odds of being drugged without consent in a social setting?","quick_answer":"About 1 in 17 US adults may experience a suspected drink-spiking incident in a lifetime, well below the 10–25% social perception. Forensically confirmed covert drugging is rarer still — alcohol alone accounts for the vast majority of drug-facilitated assault cases.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Drink spiking occupies a disproportionately large space in risk perception relative to its confirmed prevalence. Media coverage, campus safety campaigns, and social media have elevated the image of a stranger slipping a drug into an unattended drink to near-iconic status as a sexual assault mechanism. Surveys of college students find that awareness of drink spiking is near-universal and fear of being drugged is common, even among students who have never personally experienced or witnessed it. The perception is not baseless — drink spiking does occur — but the specific mechanism (covert administration of GHB, Rohypnol, or ketamine) is substantially rarer than the broader category of drug-facilitated sexual assault, where voluntary alcohol consumption by the victim is the dominant intoxicant in the vast majority of forensically confirmed cases.\n","rough_estimate":"~10-25% lifetime chance (social perception)","kind":"intuition"},"native":{"display":"~7.8% of college students self-report suspected drugging; ~2-4% of sexual assault tox screens detect classic spiking drugs","numerator":78,"denominator":1000,"unit":"self-reported suspected drugging among college students","population":"US college students aged 18-24, multi-campus survey studies"},"normalized":{"lifetime_us_adult":0.06,"display":"~1 in 17 US adults may experience suspected drink spiking in a lifetime","log_value":-1.22,"assumptions":"The best available prevalence data come from college student surveys, which are not nationally representative of all US adults. Swan et al. (2016) found 7.8% of 6,064 students at three universities reported suspected drugging. Other studies report 6-9% among college-aged women. However, college students have substantially higher social drinking exposure than the general population, and many self-reported drugging events may reflect unexpectedly strong alcohol effects rather than actual spiking — the researchers explicitly note they cannot verify actual drugging. Forensic toxicology studies of confirmed drug-facilitated sexual assault cases find classic spiking agents (GHB, Rohypnol, ketamine) in only 2-4% of samples; alcohol alone accounts for the vast majority. Estimating a lifetime figure: if the college-period risk is ~7.8% over 4 years of elevated exposure, and non-college-period risk is much lower, a rough lifetime estimate of ~6% accounts for both the peak-exposure college years and lower- exposure adult years. This is highly uncertain and likely an overestimate of actual covert drugging (vs. self-attribution of excessive intoxication). Uncertainty band: low end uses forensic confirmation rates extrapolated to general population (~2%); high end uses self-reported suspected rates (~10%).\n","uncertainty":{"low":0.02,"high":0.1},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.apa.org/news/press/releases/2016/05/drink-spiking","title":"More than a myth: Drink spiking happens","publisher":"American Psychological Association (APA)","source_type":"reputable_reference","statistic":"7.8% of 6,064 college students reported suspected drugging incidents","excerpt":"\"A survey of 6,064 students at three U.S. universities found that 462 students (7.8 percent) reported 539 incidents in which they said they had been drugged.\"\n","source_date":"2016-05-26","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260210113733/https://www.apa.org/news/press/releases/2016/05/drink-spiking","calculation_notes":"Primary self-reported prevalence from Swan et al. (2016), published in Psychology of Violence. 7.8% of students reported suspected drugging. Used as the anchor for the college-period risk estimate. The researchers note: \"We have no way of knowing if the drugging victims were actually drugged or not, and many of the victims were not certain either.\"\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S2589871X24000925","title":"The prevalence of selected licit and illicit drugs in drug facilitated sexual assaults","publisher":"Forensic Science International: Synergy (Elsevier)","source_type":"peer_reviewed","statistic":"GHB detected in ~1-4% of DFSA toxicology screens; alcohol is the dominant substance","excerpt":"\"Unexpected drugs found on toxicological screening included cannabinoids (40.2%), cocaine (32.2%), amphetamines (13.8%), MDMA (9.2%), ketamine (2.3%), and GHB (1.1%). A 26-month study of 1,179 urine samples from suspected drug-facilitated sexual assaults found 4% positive for GHB.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-24","calculation_notes":"Forensic toxicology data showing that classic \"date rape drugs\" (GHB, Rohypnol, ketamine) are detected in a small minority of drug-facilitated sexual assault cases. GHB: 1.1-4% depending on study. Rohypnol: <2%. The dominant substances are alcohol and recreational drugs the victim may have consumed voluntarily. This anchors the low end of the uncertainty band — actual covert spiking with specific agents is considerably rarer than self-reported suspicion of being drugged.\n"},{"url":"https://journals.sagepub.com/doi/10.1177/00220426231197826","title":"Spiking Versus Speculation? Perceived Prevalence, Probability, and Fear of Drink and Needle Spiking","publisher":"Journal of Drug Issues (SAGE)","source_type":"peer_reviewed","statistic":"Self-reported spiking prevalence substantially exceeds forensically confirmed rates","excerpt":"\"Considering toxicological and police findings, the reported prevalence and perceived likelihood of spiking seem to be highly overestimated.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250527233207/https://journals.sagepub.com/doi/10.1177/00220426231197826","calculation_notes":"Peer-reviewed analysis directly addressing the perception-reality gap. Confirms that self-reported suspected spiking rates (6-9%) are much higher than forensic confirmation rates (1-4% for specific spiking agents). Supports the \"overrated\" myth_framing for the specific mechanism of covert drug administration, while noting that drug-facilitated sexual assault via alcohol remains a serious and prevalent crime.\n"}],"comparison_anchors":[{"label":"Sexual assault (lifetime, US adult, contact)","lifetime_us_adult":0.34},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.11},{"label":"Food poisoning (lifetime, US adult)","lifetime_us_adult":0.9965}],"short_label":"Drink spiking","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry addresses a specific mechanism — covert administration of drugs into someone's drink — rather than the broader category of drug-facilitated sexual assault, where voluntary alcohol consumption by the victim is the dominant intoxicant. The \"overrated\" framing applies to the specific spiking mechanism, not to drug-facilitated sexual assault as a whole, which is both common and serious. Self-reported suspicion of being drugged (6-9% of college students) likely overstates actual covert spiking because the symptoms attributed to spiking — unexpected intoxication, memory gaps, loss of motor control — are also produced by drinking more alcohol than intended, combining alcohol with medications, or drinking on an empty stomach. GHB is detectable in urine for only 6-12 hours after ingestion, so some genuine spiking cases may be missed by toxicology screens conducted after that window. The college-student data are not generalizable to all US adults; social-drinking patterns differ substantially by age, and the majority of reported spiking occurs in the 18-24 age bracket. This entry should not be read as minimizing the reality of drink spiking, which is a serious crime when it occurs, but as calibrating the frequency of the specific mechanism relative to public perception.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A half-full glass on a bar counter with a single drop falling in, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/drink-spiking-assault-risk","api_url":"https://likelier.app/api/fears/drink-spiking-assault-risk.json"},{"slug":"perineal-tearing-oasis","question":"What are the odds of severe perineal tearing during vaginal delivery?","quick_answer":"Severe (3rd/4th-degree) perineal tearing affects about 3 in 100 vaginal deliveries, and about 5.8 in 100 for a first birth. That is below the 10–30% many expectant mothers fear, partly because minor tears get conflated with the rare sphincter-damaging kind.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Fear of tearing during vaginal delivery is widespread and often conflated: many women picture catastrophic injury when reading about perineal tears, not distinguishing the minor 1st- or 2nd-degree tears that affect most vaginal deliveries from the rare 3rd- or 4th-degree tears that damage the anal sphincter. Some surveys suggest most expectant mothers believe their risk of serious tearing is substantially higher than population data support, while others underestimate because they assume episiotomy prevents it.\n","rough_estimate":"perceived risk of serious tearing often cited anecdotally at 10–30%","kind":"intuition"},"native":{"display":"~3 in 100 vaginal deliveries (5.8 in 100 for a first vaginal birth)","numerator":3,"denominator":100,"unit":"per vaginal delivery","population":"women delivering vaginally in US and UK hospital settings"},"normalized":{"lifetime_us_adult":0.06,"display":"~6% for women who deliver vaginally (one to two births)","log_value":-1.22,"assumptions":"Scope is for US women who deliver vaginally at least once (approximately 54% of all US adult women, given ~80% give birth and ~68% of deliveries are vaginal at the current US C-section rate of 32.3%). First vaginal delivery carries a ~5.8% OASIS risk; a second vaginal delivery adds ~1.5% (cumulative ~7.2% for two vaginal births). The 6% central estimate represents women with a typical reproductive trajectory of 1–2 vaginal deliveries. The figure is not normalized to all US women (which would be ~3.2%) because the risk is only relevant for women who deliver vaginally. Instrumental delivery (forceps, vacuum) sharply elevates risk and is addressed in caveats.\n","uncertainty":{"low":0.038,"high":0.11},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3132187/","title":"Characteristics Associated With Severe Perineal and Cervical Lacerations During Vaginal Delivery","publisher":"Obstetrics & Gynecology / Landy et al. (Consortium on Safe Labor)","source_type":"peer_reviewed","statistic":"2.9% overall OASIS rate per vaginal delivery; 5.8% nulliparous, 0.6% multiparous, in 87,267 US vaginal deliveries 2002–2008","excerpt":"\"A total of 2,516 third- or fourth-degree perineal lacerations (2.9%) were observed. Nulliparity conferred a 7.2-fold increased risk.\"\n","source_date":"2011-03-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260525162314/https://pmc.ncbi.nlm.nih.gov/articles/PMC3132187/","calculation_notes":"Landy et al. (2011) analyzed 87,267 successful vaginal deliveries at 19 US hospitals from 2002–2008 as part of the Consortium on Safe Labor. Overall OASIS rate: 2,516 / 87,267 = 2.88% ≈ 2.9%. Parity breakdown: nulliparous 2,223 of approximately 38,300 = 5.80%; multiparous 293 of approximately 48,967 = 0.60%. Nulliparity OR for OASIS: 7.2 (reported directly). These parity-stratified rates form the basis for the native stat and the normalized lifetime assumptions.\n","independence_note":"Multi-site US cohort using electronic health record data from 19 hospitals; fully independent of the Spinelli systematic review and RCOG guideline below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8347477/","title":"Obstetric Anal Sphincter Injuries (OASIS): Risk Factors and Repair Techniques","publisher":"Journal of Clinical Medicine / Spinelli et al.","source_type":"peer_reviewed","statistic":"OASIS incidence ranges 0.5–17% across published studies; primiparous 5.7%, multiparous without prior OASIS 1.5%","excerpt":"\"The reported incidence of OASIS ranges between 0.5% and 17% depending on the study population. Stratified by parity: primiparous women 5.7%, multiparous women without prior OASIS 1.5%, multiparous women with prior OASIS 6.8–10%.\"\n","source_date":"2021-07-26","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505061400/https://pmc.ncbi.nlm.nih.gov/articles/PMC8347477/","calculation_notes":"Spinelli et al. (2021) systematic review of OASIS epidemiology and surgical management. The 0.5–17% global range reflects differences in episiotomy policy (episiotomy protects against OASIS in primiparae when performed mediolaterally), maternal age distribution, and instrumental delivery rates across studies. The parity-stratified rates (5.7% primip, 1.5% subsequent) closely corroborate the Landy consortium data (5.8% / 0.6%). These two independent datasets anchor the native and normalized estimates.\n","independence_note":"Systematic review of predominantly European and Australian cohort studies; independent of the US Consortium on Safe Labor (Landy) and the UK RCOG guideline.\n"},{"url":"https://www.rcog.org.uk/guidance/browse-all-guidance/green-top-guidelines/third-and-fourth-degree-perineal-tears-management-green-top-guideline-no-29/","title":"Third- and Fourth-Degree Perineal Tears, Management (Green-Top Guideline No. 29)","publisher":"Royal College of Obstetricians and Gynaecologists","source_type":"govt_report","statistic":"Overall UK OASIS incidence ~2.9% (range 0–8%); primiparous 6.1%, multiparous 1.7%","excerpt":"\"The incidence of OASIS in the UK is approximately 2.9% (range 0–8%). For primiparous women the incidence is approximately 6.1% and for multiparous women approximately 1.7%.\"\n","source_date":"2023-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260218084329/https://www.rcog.org.uk/guidance/browse-all-guidance/green-top-guidelines/third-and-fourth-degree-perineal-tears-management-green-top-guideline-no-29/","calculation_notes":"RCOG Green-Top Guideline No. 29 (most recent edition) cites national UK hospital audit data. The overall UK rate (2.9%) is almost identical to the US Consortium figure (2.9%), reflecting similar obstetric practice. The primiparous rate (6.1%) is slightly higher than the Landy cohort (5.8%), consistent with the range from Spinelli (5.7%). These three independent national-level sources (US, UK, systematic review) converge on 5.7–6.1% for first vaginal birth and 0.6–1.7% for subsequent, providing high confidence in the native estimate.\n","independence_note":"RCOG guideline draws on UK national birth register audit data; fully independent of the Landy US consortium and Spinelli international systematic review.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6937116/","title":"Sexual problems and vaginal changes 12 months after vaginal delivery: a prospective cohort study","publisher":"BMJ Open / Gommesen et al.","source_type":"peer_reviewed","statistic":"Dyspareunia at 12 months postpartum: 53% of women with 3rd/4th-degree tears vs 25% with no or minor tears (aRR 2.09, 95% CI 1.55–2.81)","excerpt":"\"The proportion of women with dyspareunia was 25%, 38% and 53% of women with no/labia/first-degree, second-degree or third-degree/fourth-degree tears, respectively. Adjusted relative risk for third/fourth-degree vs no-tear group: aRR 2.09; 95% CI 1.55 to 2.81.\"\n","source_date":"2019-11-12","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505061506/https://pmc.ncbi.nlm.nih.gov/articles/PMC6937116/","calculation_notes":"Gommesen et al. (2019) prospective Danish cohort of 554 primiparous women, assessed at 12 months postpartum. Dyspareunia rates stratified by tear severity. The 53% figure represents 12-month dyspareunia among the 3rd/4th-degree group; the aRR of 2.09 adjusts for age, BMI, and other confounders. Used here to quantify the key long-term sexual-function consequence rather than to derive the native probability — arithmetic derivation is from Landy/RCOG/Spinelli.\n","independence_note":"Prospective cohort at Odense University Hospital, Denmark; entirely independent of the US, UK, and Italian sources above in population, institution, and methodology.\n"}],"comparison_anchors":[{"label":"Any perineal trauma per vaginal delivery","lifetime_us_adult":0.85},{"label":"OASIS with instrumental delivery (forceps)","lifetime_us_adult":0.14},{"label":"OASIS recurrence after prior sphincter injury","lifetime_us_adult":0.085}],"personal_factor_multipliers":[{"factor":"First (nulliparous) vaginal birth","multiplier":2,"notes":"~5.8% vs ~3% average; approximately 7x compared to multiparous"},{"factor":"Forceps-assisted delivery","multiplier":4.7,"notes":"~14% OASIS rate for forceps in recent US national data (Sarker 2025)"},{"factor":"Vacuum-assisted delivery","multiplier":1.2,"notes":"~3.7% OASIS rate; moderate elevation above spontaneous vaginal"},{"factor":"Prior OASIS (recurrence in next vaginal delivery)","multiplier":2.8,"notes":"6.8–10% recurrence vs 1.5% multiparous baseline"}],"short_label":"Severe birth tearing","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"\"OASIS\" (obstetric anal sphincter injury) means 3rd- or 4th-degree tears that extend into or through the anal sphincter; this is distinct from the much more common 1st- and 2nd-degree tears, which affect roughly 60–85% of vaginal deliveries and typically heal without lasting complication. The 3% overall / 6% first-birth figures here refer exclusively to sphincter injuries. Instrumental delivery risk varies sharply by instrument: forceps carries approximately 14% OASIS in recent US national data; vacuum approximately 3.7%. Mediolateral episiotomy (at ≥45–60° angle) reduces OASIS risk by 50–80% in primiparous women per RCOG guidance; routine episiotomy in all vaginal deliveries is not recommended. Long-term outcomes after OASIS: roughly 60–80% of women are asymptomatic at 12 months, but 38% report significant bowel symptoms at 4 years (systematic review) and anal incontinence rates rise to 23.8% at 20 years versus 11.7% in women with no sphincter injury. Sexual dysfunction (dyspareunia) is present in 53% at 12 months postpartum vs 25% for women without OASIS (Gommesen 2019). Scope is subgroup_lifetime; the 6% estimate applies to women who deliver vaginally, not to all US adult women (~3.2% when denominator is all US women).\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A simple anatomical diagram outline showing a birth canal cross-section, flat vector in muted clinical tones."},"canonical_url":"https://likelier.app/perineal-tearing-oasis","api_url":"https://likelier.app/api/fears/perineal-tearing-oasis.json"},{"slug":"cannabis-use-disorder","question":"What are the odds of developing cannabis use disorder?","quick_answer":"About 1 in 16 US adults (6.3%) meets criteria for cannabis use disorder at some point in life. That sits well above the common impression that only 1-2% of adults are affected, making dependence more likely than cannabis's \"least addictive\" reputation suggests.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Cannabis occupies an unusual position in public risk perception: it is widely regarded as the drug least likely to cause dependence, often in explicit contrast to alcohol, opioids, or stimulants. The political framing of legalization debates has reinforced this view — advocates have emphasized relative safety compared to alcohol, and the term \"marijuana use disorder\" does not have the cultural salience of \"alcoholism\" or \"opioid addiction.\" Many users and non-users alike believe cannabis is simply not addictive in any meaningful sense. That belief is incorrect for a substantial minority of users. The post-2018 legalization wave has also made high-potency products — concentrates, edibles, and vapes with THC concentrations far above what was available in prior decades — the norm in legal markets, changing the pharmacological exposure profile in ways that older survey data may not fully capture.\n","rough_estimate":"~1-2% of adults","kind":"intuition"},"native":{"display":"6.3% of US adults meet DSM-5 criteria for cannabis use disorder at some point in their lifetime (NESARC-III, 2012–2013)","numerator":6.3,"denominator":100,"unit":"share of US adults with lifetime DSM-5 cannabis use disorder diagnosis","population":"US adults aged 18 and older (NESARC-III, N=36,309, face-to-face interviews 2012–2013)"},"normalized":{"lifetime_us_adult":0.063,"display":"~1 in 16 US adults develops cannabis use disorder at some point in their lifetime","log_value":-1.2,"assumptions":"Hasin et al. (American Journal of Psychiatry, 2016) used the NESARC-III data (N=36,309 US adults, 2012–2013) with DSM-5 diagnostic criteria to estimate lifetime cannabis use disorder prevalence at 6.3% and 12-month prevalence at 2.5%. The lifetime figure is used directly as the normalized estimate: it already represents the US adult population and encompasses the full adult lifespan captured by retrospective structured interviews. Among adults who have ever used cannabis, the conditional probability of developing CUD is substantially higher. The Lopez-Quintero et al. (2011, Drug and Alcohol Dependence) analysis of NESARC-I data found that approximately 8.9% of ever-users transition to cannabis dependence — the per-user conditional rate. The NESARC-III all-adult 6.3% figure is used here because the question asks about population-level lifetime risk for a US adult, not conditional risk given use. SAMHSA 2024 NSDUH found 20.6 million past-year CUD among US residents 12+, consistent with the high end of prevalence estimates once cannabis use rates are applied.\n","uncertainty":{"low":0.04,"high":0.09},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5026387/","title":"Prevalence and Correlates of DSM-5 Cannabis Use Disorder, 2012–2013: Findings from the National Epidemiologic Survey on Alcohol and Related Conditions–III","publisher":"Hasin DS et al. — American Journal of Psychiatry, 2016","source_type":"peer_reviewed","statistic":"Lifetime and 12-month prevalences of DSM-5 cannabis use disorder among US adults were 6.3% and 2.5%, respectively (NESARC-III, N=36,309)","excerpt":"\"The prevalences of 12-month and lifetime cannabis use disorder were 2.5% and 6.3%. Odds of 12-month and lifetime cannabis use disorder were higher for men, Native Americans, unmarried individuals, those with low incomes, and young adults; cannabis use disorder was associated with other substance use disorders, affective disorders, anxiety, and personality disorders.\"\n","source_date":"2016-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505050556/https://pmc.ncbi.nlm.nih.gov/articles/PMC5026387/","calculation_notes":"The 6.3% lifetime prevalence is used directly as the native numerator (6.3 per 100 US adults). This is the primary calculation input, representing the first nationally representative DSM-5 CUD prevalence estimate using structured diagnostic interviews. The 2.5% 12-month figure confirms the disorder is active for many adults at any given time, not merely historical.\n","independence_note":"NESARC-III was conducted by NIAAA using probability sampling and structured clinical interviews (AUDADIS-5), methodologically distinct from SAMHSA NSDUH self-report instruments.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5037576/","title":"Prevalence of Marijuana Use Disorders in the United States Between 2001-2002 and 2012-2013","publisher":"Hasin DS et al. — JAMA Psychiatry, 2015","source_type":"peer_reviewed","statistic":"Past-year prevalence of DSM-IV marijuana use disorder increased from 1.5% in 2001-2002 to 2.9% in 2012-2013; marijuana use more than doubled over the same period","excerpt":"\"The past-year prevalence of marijuana use was 4.1% in 2001-2002 and 9.5% in 2012-2013. The past-year prevalence of DSM-IV marijuana use disorder was 1.5% in 2001-2002 and 2.9% in 2012-2013 (P < .05). Significant increases were found across all demographic subgroups.\"\n","source_date":"2015-12-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260223205053/https://pmc.ncbi.nlm.nih.gov/articles/PMC5037576/","calculation_notes":"This study uses DSM-IV criteria and past-year prevalence, so the absolute figures are lower than the DSM-5 lifetime estimates in Hasin et al. 2016. It is used here to establish the trend: both cannabis use and cannabis use disorder have increased substantially since 2001-2002, with disorder prevalence roughly doubling. The post-legalization era (2018 onward) is not captured in either NESARC study and likely represents a further increase in exposure and conditional disorder rates.\n","independence_note":"Both Hasin 2015 (JAMA Psychiatry) and Hasin 2016 (Am J Psychiatry) draw on NESARC wave data. They are not fully independent sources but use different diagnostic criteria (DSM-IV vs DSM-5) and different comparison years, providing genuine methodological triangulation on the trend.\n"},{"url":"https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","title":"Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"20.6 million people aged 12 or older had past-year cannabis use disorder in 2024, making it the most common drug use disorder in the US","excerpt":"\"Marijuana use disorder was the most common drug use disorder (20.6 million), followed by opioid use disorder (4.8 million) and central nervous system stimulant use disorder (4.3 million). In 2024, 44.3 million individuals reported marijuana use in the past month.\"\n","source_date":"2025-07-14","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260512142703/https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","calculation_notes":"SAMHSA 2024 NSDUH: 20.6M past-year CUD / ~260M US adults ≈ 7.9% past-year CUD prevalence. This figure is notably higher than the NESARC-III 2.5% 12-month estimate, reflecting both the increase in cannabis use since 2012-2013 and instrument differences between NSDUH and NESARC. It is used here as a cross-validation anchor showing that post-legalization cannabis use disorder rates are higher than the NESARC-III figures, supporting the upper end of the uncertainty range. The 6.3% lifetime estimate from 2012-2013 is almost certainly an undercount for today's adult cohort.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/41099091/","title":"High potency cannabis flower use is associated with heavier consumption and risk for cannabis use disorder among young adults in California, United States","publisher":"Dunbar MS et al. — Addiction, 2025 (RAND Corporation)","source_type":"peer_reviewed","statistic":"A one-point increase on the perceived THC potency scale was associated with 3.33 more use days per month, 0.13 more grams of flower per day, and 1.21 greater CUDIT-R scores; results are inconsistent with the notion that individuals titrate use of higher-potency cannabis (n=512 young adults, California, 2022–2023)","excerpt":"\"Results from separate adjusted linear regression models showed that a one-point increase in the perceived THC potency scale was associated with 3.33 more use days per month [B = 3.33, (standard error, SE = 0.46), P < 0.0001], 0.13 more grams of flower used per day [B = 0.13 (SE = 0.04), P < 0.01] and 1.21 greater CUDIT-R scores [B = 1.21 (SE = 0.29), P < 0.0001].... Self-reported use of higher potency cannabis flower appears to be associated with more frequent and higher-quantity cannabis consumption and greater Cannabis Use Disorder Identification Test - Revised (CUDIT-R) scores, which is inconsistent with the notion that individuals titrate use of higher-potency cannabis.\"\n","source_date":"2025-08-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20251028230603/https://pubmed.ncbi.nlm.nih.gov/41099091/","calculation_notes":"This study reports associations between perceived THC potency and (a) frequency and quantity of use and (b) CUDIT-R screening scores — not a per-person risk ratio for developing cannabis use disorder. No clean CUD multiplier is derivable from the published coefficients without an unstated assumption, so it is NOT added as a personal_factor_multiplier; it is cited as supporting evidence in the caveats that high-potency products are associated with heavier consumption and higher CUD-screening risk, and that users do not titrate down to compensate for higher potency.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21145178/","title":"Probability and predictors of transition from first use to dependence on nicotine, alcohol, cannabis, and cocaine: results of the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC)","publisher":"Lopez-Quintero C et al. — Drug and Alcohol Dependence, 2011","source_type":"peer_reviewed","statistic":"Cumulative probability of transition to dependence: 67.5% for nicotine users, 22.7% for alcohol users, 20.9% for cocaine users, and 8.9% for cannabis users","excerpt":"\"The cumulative probability estimate of transition to dependence was 67.5% for nicotine users, 22.7% for alcohol users, 20.9% for cocaine users, and 8.9% for cannabis users.\"\n","source_date":"2011-05-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260427112522/https://pubmed.ncbi.nlm.nih.gov/21145178/","calculation_notes":"Supplies the per-substance conditional conversion rates the body text uses to compare cannabis's ever-user transition rate against nicotine, alcohol, and cocaine. This is the same NESARC-I analysis (Lopez-Quintero et al. 2011) referenced elsewhere in this entry's assumptions and caveats for the cannabis-specific 8.9% figure; added here as a full citation because the body also states the nicotine (67.5%), alcohol (22.7%), and cocaine (20.9%) rates from the same analysis.\n","independence_note":"NESARC-I (2001-2002 and 2004-2005 waves) is a distinct survey wave from NESARC-III (2012-2013) used for the primary 6.3% lifetime CUD estimate above, though both are NIAAA-conducted nationally representative surveys using structured diagnostic interviews.\n"}],"comparison_anchors":[{"label":"Alcohol use disorder (lifetime, US adult)","lifetime_us_adult":0.291},{"label":"Opioid addiction after surgical prescription (lifetime, US adult)","lifetime_us_adult":0.0088},{"label":"Gambling disorder causing financial ruin (lifetime, US adult)","lifetime_us_adult":0.0063}],"personal_factor_multipliers":[{"factor":"daily or near-daily cannabis user","multiplier":5,"notes":"Frequency of use is the dominant behavioral predictor; daily users have substantially higher conditional CUD rates"},{"factor":"use began in adolescence (before age 18)","multiplier":3,"notes":"Early onset of use is a robust predictor of CUD; the developing adolescent brain is more vulnerable to dependence"},{"factor":"male","multiplier":1.5,"notes":"Men have higher CUD rates than women in NESARC-III, consistent with higher rates of heavy use"},{"factor":"co-occurring anxiety or mood disorder","multiplier":2.5,"notes":"NESARC-III found strong comorbidity; cannabis is commonly used for self-medication of anxiety and insomnia"},{"factor":"uses only occasionally (monthly or less)","multiplier":0.15,"notes":"CUD is strongly use-pattern-dependent; occasional users face much lower conditional risk"}],"short_label":"Cannabis use disorder","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The 6.3% lifetime prevalence is drawn from 2012-2013 NESARC-III data, before legalization in most US states. Since 2018, both cannabis use prevalence and use disorder rates have increased substantially; SAMHSA 2024 found 20.6 million past-year CUD cases, suggesting the lifetime prevalence for today's younger cohort will be higher than 6.3% by the time they reach the age of the NESARC-III respondents. The conditional risk among ever-users is substantially higher than the all-adult 6.3% — approximately 8.9% of ever-users develop dependence, per Lopez-Quintero et al. (2011, Drug and Alcohol Dependence, NESARC-I data). Post-legalization products — concentrates, vapes, and edibles with THC concentrations of 40-90% versus the 5-10% of typical cannabis sold in 2000 — represent a meaningfully different pharmacological exposure than the products used by most NESARC-III respondents. The 6.3% figure should be treated as a lower bound for current and future adult cohorts. DSM-5 CUD requires at least 2 of 11 criteria; tolerance and withdrawal are included but not required, meaning the disorder captures a wider range of problematic use patterns than colloquial \"addiction.\" Product potency is an emerging but not-yet-quantifiable risk axis. A 2025 study of 512 young adults in California (Dunbar et al., Addiction) found that each one-point step up the perceived THC potency scale was associated with 3.33 more use days per month, 0.13 more grams of flower per day, and a 1.21-point higher CUDIT-R screening score — and that higher-potency users do not \"titrate\" down to compensate, contrary to a common assumption. This is an association with consumption intensity and a CUD screening measure, not a per-person risk multiplier for developing the disorder, so it is reported here as a caveat rather than as a personal factor. It suggests the post-2018 shift toward high-potency flower and concentrates (average THC roughly 3–5x higher than two decades ago) compounds the heavy-use pathway that drives CUD, and is a further reason to treat the 2012-2013 6.3% figure as a floor for current cohorts.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A small calendar showing months accumulating beside a single leaf silhouette, flat vector illustration in muted green and grey tones."},"canonical_url":"https://likelier.app/cannabis-use-disorder","api_url":"https://likelier.app/api/fears/cannabis-use-disorder.json"},{"slug":"stroke","question":"What are the odds of dying from a stroke?","quick_answer":"The lifetime odds of dying from a stroke are about 1 in 15 for a typical adult — much higher than the common intuition that strokes are rare and mostly happen to other people. Lifetime stroke incidence (fatal plus non-fatal) is higher still, but this figure counts deaths only.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Stroke almost never shows up in lists of things people are afraid of. It doesn't feature in Chapman's top fears, it isn't a staple of disaster movies, and most adults under 50 file it somewhere between \"old people's problem\" and \"unlikely to happen to me\". The intuitive mental model is that strokes are rare, sudden, and mostly fatal for other people — which gets the sudden-and-fatal part roughly right and the rare part badly wrong.\n","rough_estimate":"35% of US adults say they are very or somewhat worried about personally experiencing a stroke","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~1.2 per 1,000 adults per year (~6.5 million deaths globally)","numerator":1,"denominator":806,"unit":"per year","population":"global adults (age 18+), ischaemic and haemorrhagic stroke combined"},"normalized":{"lifetime_us_adult":0.067,"display":"1 in ~15 lifetime (global adult)","log_value":-1.17,"assumptions":"Uses WHO's 2024 Top 10 Causes of Death update, which places stroke at approximately 10% of the world's 68 million annual deaths in 2021 — roughly 6.5 to 6.8 million stroke deaths per year globally. Divided across a global adult population of ~5.5 billion (age 18+), that is an annual rate of ~1.2 per 1,000 adults per year. Compounded over 60 years of remaining adult life: 1 - (1 - 0.00124)^60 ≈ 0.072. Rounded slightly down to 0.067 (≈ 1 in 15) to account for competing mortality (an adult who dies of heart disease at 72 never gets a chance to die of stroke at 84) and for the fact that a significant share of stroke deaths occur above age 80 where many readers will have already been removed from the denominator by other causes. Note that *incidence* of stroke over a lifetime (fatal + non-fatal) is much higher — the GBD / WSO estimates put it near 1 in 4 — but this entry reports the mortality number only, consistent with other Likelier health entries. Scope is global-adult-lifetime, not US-adult-lifetime, because stroke mortality is heavily concentrated in low- and middle-income countries (86% of deaths) and the US-only figure would understate the global baseline by roughly a factor of two.\n","uncertainty":{"low":0.04,"high":0.1},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","title":"The top 10 causes of death","publisher":"World Health Organization","source_type":"govt_report","statistic":"Stroke responsible for approximately 10% of global deaths in 2021 (≈6.5-6.8 million), third leading cause of death worldwide","excerpt":"\"The world's biggest killer is ischaemic heart disease, responsible for 13% of the world's total deaths. [...] Instead of being the second and third leading causes of death as in 2019, stroke and chronic obstructive pulmonary disease became the third and fourth in 2021, responsible for approximately 10% and 5% of total deaths, respectively. [...] The top global causes of death, in order of total number of lives lost, are associated with two broad topics: cardiovascular (ischaemic heart disease, stroke) and respiratory (COVID-19, chronic obstructive pulmonary disease, lower respiratory infections).\"\n","source_date":"2024-08-07","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413165125/https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","calculation_notes":"WHO reports 68 million total deaths globally in 2021; stroke at ~10% implies roughly 6.8 million annual stroke deaths. Across ~5.5 billion adults (age 18+), that is ~1.24 per 1,000 adults per year. Compounded over 60 adult years: 1 - (1 - 1.24e-3)^60 ≈ 0.072, adjusted to 0.067 for competing mortality. COVID-19 displaced stroke from #2 to #3 in 2020-2021; pre-pandemic baselines in GBD and WSO data place stroke solidly at #2, and most forecasts expect stroke to return to #2 as COVID recedes.\n","independence_note":"WHO Global Health Estimates draw on national vital-registration and IHME / GBD modelling pipelines, so this figure is not fully independent of Feigin et al. / GBD stroke estimates. Used here as the canonical institutional source.\n"},{"url":"https://www.world-stroke.org/news-and-blog/news/wso-global-stroke-fact-sheet-2022","title":"WSO Global Stroke Fact Sheet 2022","publisher":"World Stroke Organization (Feigin et al., Int J Stroke)","source_type":"peer_reviewed","statistic":"Stroke is the second-leading cause of death worldwide; 43% increase in stroke deaths between 1990 and 2019; 86% of stroke deaths occur in lower- and lower-middle-income countries","excerpt":"\"Stroke remains the second-leading cause of death and the third-leading cause of death and disability combined [...] The estimated global cost of stroke is over US$721 billion (0.66% of the global GDP) [...] From 1990 to 2019, there was a substantial increase of [...] 70.0% increase in incident strokes, 43.0% deaths from stroke, 102.0% prevalent strokes, and 143.0% DALYs [...] 86.0% of deaths and 89.0% of DALYs [occur in] lower-income and lower-middle-income countries.\"\n","source_date":"2022-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260217081915/https://www.world-stroke.org/news-and-blog/news/wso-global-stroke-fact-sheet-2022","calculation_notes":"WSO's pre-pandemic baseline (stroke as #2 cause of death) is the number Likelier uses as the underlying rate. The 43% rise in absolute stroke deaths from 1990 to 2019 reflects population ageing more than rising per-capita risk; age-standardised stroke mortality has been falling globally. Used as the authoritative peer-reviewed cross-check on the WHO figure, and as the source of the \"second leading cause\" framing in the long-form text.\n","independence_note":"WSO Fact Sheet uses Global Burden of Disease 2019 data, which WHO Global Health Estimates also draws from; the two sources should be treated as partially dependent.\n"},{"url":"https://www.cdc.gov/stroke/data-research/facts-stats/index.html","title":"Stroke Facts","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"In the US, someone dies of stroke every 3 minutes 14 seconds; stroke accounts for about 1 in 6 cardiovascular deaths (17.5%)","excerpt":"\"Every 3 minutes and 14 seconds, someone dies of stroke in this country. [...] 1 in 6 deaths (17.5%) from cardiovascular disease was due to stroke. [...] Stroke is a leading cause of death for Americans. [...] Stroke risk increases with age, but strokes can — and do — occur at any age.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260407133417/https://www.cdc.gov/stroke/data-research/facts-stats/index.html","calculation_notes":"One stroke death every 194 seconds in the US works out to ~163,000 US stroke deaths per year, or ~49 per 100,000 — less than half the crude global rate implied by the WHO figures, consistent with the WSO finding that 86% of stroke deaths are in low- and middle-income countries. Used here to anchor the gap between the US-only number (which would give a lifetime figure closer to 1 in 25) and the global-adult figure Likelier reports for this entry.\n","independence_note":"CDC Stroke Facts draws from the NCHS/NVSS death-certificate pipeline, a different upstream from WHO GHE and WSO/GBD modelled estimates. Provides an independent US-only anchor for the crude-rate sanity check, even though WHO figures ultimately incorporate US vital-registration data.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Death by drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Hypertension (untreated)","multiplier":4,"notes":"AHA/ACC guidelines and Framingham Heart Study cohort data: hypertension is the single largest modifiable stroke risk factor, increasing stroke risk approximately 4-fold; well-controlled BP reduces risk back toward baseline"},{"factor":"Atrial fibrillation","multiplier":5,"notes":"AHA/ACC: atrial fibrillation confers a 4-5x independent increase in ischemic stroke risk; anticoagulation therapy (warfarin, DOACs) reduces this by approximately 60-70%"},{"factor":"Current smoker","multiplier":2.5,"notes":"Framingham Heart Study and multiple meta-analyses: smoking approximately doubles to triples stroke risk; the Framingham data specifically shows ~2.5x relative risk for current smokers vs never-smokers; risk falls toward baseline within 2-5 years of quitting"},{"factor":"Prior TIA (transient ischemic attack)","multiplier":10,"notes":"AHA/ASA guidelines: 90-day stroke risk after TIA is approximately 10-15% without treatment; short-term relative risk elevation vs general population is approximately 10-fold; risk is highest in the first 48 hours"}],"short_label":"Stroke","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The normalized figure is a global-adult-lifetime number, not a US-adult number. A typical US or Western European adult with controlled blood pressure and no diabetes has a materially lower lifetime stroke mortality risk — roughly 1 in 25 to 1 in 30, reflecting the lower US crude rate and the concentration of global stroke deaths in low- and middle-income countries. Conversely, adults in low-income regions, adults with untreated hypertension, and adults with atrial fibrillation all run meaningfully higher than the headline number. This entry also reports stroke *mortality*, not stroke *incidence*: lifetime risk of having any stroke (fatal or non-fatal) is roughly 1 in 4 globally per GBD / WSO, which is the figure usually cited in clinical contexts.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":5,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale branching line on a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/stroke","api_url":"https://likelier.app/api/fears/stroke.json"},{"slug":"ptsd-lifetime","question":"What are the odds of developing PTSD in your lifetime?","quick_answer":"About 1 in 15 US adults (6.8%) develop PTSD in their lifetime. Most people link PTSD almost exclusively to combat veterans and underestimate how much of the lifetime caseload is civilian and female.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"There is no standard tracker for perceived lifetime PTSD risk, and the lay mental model is unusually distorted by a single cultural anchor: combat. Most people associate PTSD almost exclusively with soldiers, so when asked how common it is in the general population they either picture the condition as a near-exclusively military problem or estimate a small fraction of the population. The civilian reality is the opposite — the large majority of lifetime PTSD in the United States follows non-combat trauma (sexual assault, serious accidents, sudden bereavement, disasters, interpersonal violence), and the condition skews strongly female rather than male. The other systematic error runs the other way: because most adults will experience at least one potentially traumatic event, some assume that trauma exposure and PTSD are nearly the same thing. They are not. Exposure is common; the disorder is a minority outcome conditional on exposure.\n","rough_estimate":"most adults associate PTSD with combat veterans and underestimate how civilian and how female the lifetime caseload actually is","kind":"intuition"},"native":{"display":"6.8% lifetime prevalence (NCS-R, DSM-IV PTSD)","numerator":68,"denominator":1000,"unit":"lifetime prevalence","population":"US adults 18+ (NCS-R, 2001-2003)"},"normalized":{"lifetime_us_adult":0.068,"display":"~1 in 15 lifetime","log_value":-1.167,"assumptions":"Uses the National Comorbidity Survey Replication (NCS-R) lifetime prevalence of 6.8% for DSM-IV PTSD among US adults as the headline figure, as reported by NIMH. This is a direct survey-based lifetime estimate — no compounding or hazard conversion required. The earlier National Comorbidity Survey (Kessler et al. 1995) found 7.8% lifetime prevalence under DSM-III-R criteria, which anchors the upper bound of the uncertainty band. The later NESARC-III survey (2012-2013), reported by the VA's National Center for PTSD, found about 6% lifetime under DSM-5 criteria, anchoring the lower-central range. The point estimate of 0.068 sits between these independent surveys. The uncertainty band 0.06-0.085 reflects this cross-survey methodological range (different diagnostic editions, sampling frames, and interview instruments). PTSD is a conditional outcome: most US adults experience at least one potentially traumatic event in their lifetime, but only a minority develop the disorder, and the conditional rate varies enormously by trauma type — highest for rape and sexual assault, lower for accidents and disasters.\n","uncertainty":{"low":0.06,"high":0.085},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/post-traumatic-stress-disorder-ptsd","title":"Post-Traumatic Stress Disorder (PTSD) — Statistics","publisher":"National Institute of Mental Health (NIMH), NIH","source_type":"govt_report","statistic":"Past-year prevalence of PTSD among US adults 3.6% (females 5.2%, males 1.8%); lifetime prevalence of PTSD 6.8%","excerpt":"\"An estimated 3.6% of U.S. adults had PTSD in the past year... Past year prevalence of PTSD among adults was higher for females (5.2%) than for males (1.8%)... The lifetime prevalence of PTSD was 6.8%.\"\n","source_date":"2003-12-31","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260618145237/https://www.nimh.nih.gov/health/statistics/post-traumatic-stress-disorder-ptsd","calculation_notes":"NIMH reports National Comorbidity Survey Replication (NCS-R) data, a face-to-face household survey of US adults aged 18+ conducted 2001-2003. The 6.8% lifetime figure is used directly as the normalized estimate — no hazard compounding needed because it is already a lifetime estimate. The past-year female-to-male ratio of 5.2%/1.8% ≈ 2.9 and the lifetime ratio (see VA source, 8%/4% ≈ 2.0) provide the female multiplier of ~2.0, applied conservatively to the more-replicated lifetime split.\n","independence_note":"This NIMH page republishes NCS-R (Harvard/Kessler/NIMH) data. It is therefore the same underlying survey as the broader NCS-R disorder analyses, but is independent from the NESARC-III (NIAAA) figures cited via the VA source and from the original 1995 NCS.\n"},{"url":"https://www.ptsd.va.gov/professional/treat/essentials/epidemiology.asp","title":"Epidemiology of PTSD","publisher":"US Department of Veterans Affairs, National Center for PTSD","source_type":"govt_report","statistic":"NESARC-III lifetime PTSD prevalence ~6% overall (women 8%, men 4%); veteran lifetime PTSD by service era: Vietnam 10%, OEF/OIF 29%; female veteran lifetime 13% vs male 6%","excerpt":"\"lifetime PTSD prevalence according to DSM-5 criteria of 6% overall in a sample of over 36,000 U.S. adults... lifetime prevalence of PTSD among Veterans is 7%. Lifetime prevalence was higher among female Veterans (13%) than male Veterans (6%).\"\n","source_date":"2023-01-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260622052813/https://www.ptsd.va.gov/professional/treat/essentials/epidemiology.asp","calculation_notes":"The VA National Center for PTSD synthesizes NESARC-III (2012-2013, 36,000+ US adults) for the general-population DSM-5 figures and the National Health and Resilience in Veterans Study (NHRVS) for service-era breakdowns. The general-population 6% lifetime anchors the lower-central range of the uncertainty band. Veteran service-era lifetime rates (WWII/Korea 3%, Vietnam 10%, Persian Gulf 21%, OEF/OIF 29%) provide the veteran multipliers: OEF/OIF 29% / 6.8% general ≈ 4.3x; Vietnam 10% / 6.8% ≈ 1.5x. The female 8% / male 4% lifetime split gives the ~2.0 female multiplier and ~0.6 male multiplier.\n","independence_note":"NESARC-III (NIAAA, Hasin/Goldstein) and the NHRVS are independent of the NCS-R (Harvard) survey pipeline, with different sampling frames, field operations, and diagnostic instruments. Agreement to within roughly one percentage point of lifetime prevalence (NESARC-III ~6% DSM-5 vs NCS-R 6.8% DSM-IV) across two independent surveys a decade apart and under different DSM editions is the strongest cross-validation available for US PTSD prevalence.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/7492257/","title":"Posttraumatic stress disorder in the National Comorbidity Survey","publisher":"Archives of General Psychiatry (Kessler RC, Sonnega A, Bromet E, Hughes M, Nelson CB)","source_type":"peer_reviewed","statistic":"Lifetime prevalence of PTSD 7.8% (DSM-III-R); elevated among women; traumas most associated with PTSD are combat and witnessing among men, rape and sexual molestation among women","excerpt":"\"The estimated lifetime prevalence of PTSD is 7.8%. Prevalence is elevated among women and the previously married. The traumas most commonly associated with PTSD are combat exposure and witnessing among men and rape and sexual molestation among women... more than one third of people with an index episode of PTSD fail to recover even after many years.\"\n","source_date":"1995-12-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260608090927/https://pubmed.ncbi.nlm.nih.gov/7492257/","calculation_notes":"The original National Comorbidity Survey (NCS) was a face-to-face survey of 5,877 persons aged 15-54 conducted under DSM-III-R criteria. The 7.8% lifetime figure is the highest of the three independent estimates and anchors the upper bound (0.085) of the uncertainty band. The abstract identifies trauma type as the key conditional modifier — rape and sexual molestation are the highest-association traumas among women — but quantified conditional rates per trauma type live in the full-text tables, not the abstract, so the trauma-type gradient is described qualitatively in prose and caveats rather than as a numeric multiplier.\n","independence_note":"The 1995 NCS is the predecessor survey to the NCS-R and used a different DSM edition (DSM-III-R), a younger age range (15-54), and a separate field wave. It is independent of both the NESARC-III (NIAAA) and the NCS-R replication figures.\n"}],"comparison_anchors":[{"label":"Suicide (lifetime, US adult)","lifetime_us_adult":0.00827},{"label":"Eating disorder diagnosis (lifetime, US adult)","lifetime_us_adult":0.09},{"label":"Alzheimer's disease (lifetime, US adult)","lifetime_us_adult":0.107},{"label":"Major depression (lifetime, US adult)","lifetime_us_adult":0.206}],"personal_factor_multipliers":[{"factor":"Female","multiplier":2,"notes":"VA/NESARC-III lifetime prevalence: ~8% women vs ~4% men, ratio ≈ 2.0. NCS-R past-year split (5.2% vs 1.8%) is even wider. The female skew is one of the most replicated findings in PTSD epidemiology, driven partly by higher exposure to high-conditional-risk traumas such as sexual assault."},{"factor":"Male","multiplier":0.6,"notes":"VA/NESARC-III lifetime ~4% men vs ~6.8% general; ratio ≈ 0.6. Men experience more trauma overall but more often the lower-conditional-risk types (accidents, witnessing) rather than sexual assault."},{"factor":"Vietnam-era veteran","multiplier":1.5,"notes":"VA National Center for PTSD (NHRVS): Vietnam-era veteran lifetime PTSD ~10% vs ~6.8% general population; ratio ≈ 1.5. Service era matters: combat intensity and post-deployment support differ markedly across cohorts."},{"factor":"OEF/OIF (Iraq/Afghanistan) veteran","multiplier":4.3,"notes":"VA National Center for PTSD (NHRVS): OEF/OIF-era veteran lifetime PTSD ~29% vs ~6.8% general; ratio ≈ 4.3. The highest service-era rate measured, reflecting repeated deployments and high combat-exposure rates."}],"short_label":"PTSD","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"PTSD is defined by DSM criteria following exposure to a qualifying traumatic event: intrusive re-experiencing, avoidance, negative changes in cognition and mood, and hyperarousal, persisting more than one month and causing functional impairment. The headline figure is population lifetime prevalence — the share of all US adults who have ever met criteria — not the conditional risk given trauma. Those are very different quantities and conflating them is the central error to avoid. The large majority of US adults experience at least one potentially traumatic event in their lifetime, yet only roughly 6-8% ever develop PTSD, because the conditional probability of PTSD given a trauma is itself low and varies enormously by trauma type. The NCS (Kessler et al. 1995) found that rape and sexual molestation are the traumas most associated with PTSD among women, and combat and witnessing among men; accidents and natural disasters carry far lower conditional rates. This trauma-type gradient, rather than any single population number, is what most changes a careful reader's interpretation. Estimates also differ by DSM edition: the 1995 NCS (7.8%, DSM-III-R) and NCS-R (6.8%, DSM-IV) used broader criteria than DSM-5; the NESARC-III DSM-5 figure (~6%) is slightly lower, partly reflecting the tightened diagnostic definition. Veteran subgroups, especially recent combat cohorts, run several times the general-population rate, but veterans are a small fraction of the adult population, so most lifetime PTSD cases are civilian.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"A still pond surface with a single fading ripple spreading outward, flat vector illustration in muted grey-blue tones."},"canonical_url":"https://likelier.app/ptsd-lifetime","api_url":"https://likelier.app/api/fears/ptsd-lifetime.json"},{"slug":"mortgage-foreclosure","question":"What are the odds of losing your home to mortgage foreclosure?","quick_answer":"A US homeowner with a mortgage has roughly a 1 in 14 lifetime chance of foreclosure, against an annual filing rate of about 0.26% of housing units in 2025. Anxious homeowners often guess 10-20%, anchoring on 2008-crisis imagery rather than the tighter post-Dodd-Frank lending era.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"The 2008 foreclosure crisis left a deep imprint on the American psyche. News footage of boarded-up subdivisions, underwater borrowers, and bank-owned signs became the dominant image of homeownership risk for an entire generation. Most adults who lived through it overestimate the current probability of foreclosure, anchoring on crisis-era imagery rather than the post-Dodd-Frank lending environment. Surveys of financial anxiety consistently rank \"losing my home\" near the top, often above risks that are statistically far more likely, such as job loss or medical bankruptcy.\n","rough_estimate":"~10-20% lifetime guess for anxious homeowners","kind":"intuition"},"native":{"display":"~367,460 foreclosure filings in 2025 (0.26% of US housing units)","numerator":367460,"denominator":142000000,"unit":"per year","population":"US housing units"},"normalized":{"lifetime_us_adult":0.07,"display":"~1 in 14 lifetime (US homeowner with a mortgage)","log_value":-1.15,"assumptions":"ATTOM Data reported 367,460 properties with foreclosure filings in 2025, representing 0.26% of approximately 142 million US housing units. However, completed foreclosures (bank repossessions / REO) numbered only 46,439, or about 0.09% of mortgaged properties. The annual filing rate of 0.26% is a post-pandemic normalization figure, well below the pre-pandemic 2019 rate (~0.36%) and far below the 2010 peak (~2.23%). A BLS longitudinal cohort study (NLSY79, tracking baby boomers born 1957-1964 from 1988 to 2008) found that about 2% received a foreclosure notice and roughly 1.4% lost a home to foreclosure over a 20-year span — but that cohort happened to include the worst housing crisis in modern history. For a lifetime estimate spanning 30-35 years of potential mortgage exposure (age 25-60, with most homeowners holding 1-3 mortgages across the period), compounding the long-run average annual filing rate of roughly 0.2-0.3% across normal years but including one severe cycle gives approximately 5-10%. The central estimate of 7% reflects the reality that most homeowners will live through at least one serious recession but that post-2008 regulatory reforms (Qualified Mortgage rules, Dodd-Frank, tighter underwriting) have structurally reduced the tail risk relative to the pre-crisis era. The St. Louis Fed estimated that nearly 10 million homeowners lost homes between 2006 and 2014, out of roughly 50-55 million mortgaged homes at the time — a crisis-era cumulative rate of roughly 18-20%, which should not be treated as a baseline.\n","uncertainty":{"low":0.03,"high":0.15},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.attomdata.com/news/market-trends/foreclosures/2025-year-end-foreclosure-market-report/","title":"U.S. Foreclosure Activity Increases in 2025","publisher":"ATTOM Data Solutions","source_type":"reputable_reference","statistic":"367,460 US properties with foreclosure filings in 2025 (0.26% of housing units), up 14% from 2024; 46,439 bank repossessions (REO), up 27% from 2024 but down 96% from the 2010 peak of 1,050,500","excerpt":"\"In 2025, foreclosure filings — default notices, scheduled auctions and bank repossessions — were reported on 367,460 U.S. properties, up 14 percent from 2024 and up 3 percent from 2023 but down 25 percent from 2019. Lenders repossessed 46,439 properties through foreclosures (REO) in 2025, up 27 percent from 2024 but down 68 percent from 143,955 in 2019, and down 96 percent from a peak of 1,050,500 in 2010.\"\n","source_date":"2026-01-15","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260525162150/https://www.attomdata.com/news/market-trends/foreclosures/2025-year-end-foreclosure-market-report/","calculation_notes":"Primary source for the 2025 annual filing count and bank repossession count. 367,460 filings on ~142 million housing units = 0.26% annual filing rate. Completed foreclosures (REO) of 46,439 on ~52 million mortgaged homes = ~0.09% annual completion rate. The gap between filings and completions reflects the multi-stage foreclosure process: many filings are cured through loan modification, short sale, or reinstatement before reaching REO.\n","independence_note":"ATTOM aggregates county recorder data from over 3,000 counties nationwide. It is the primary private-sector foreclosure tracking source used by industry, media, and researchers. Independent of MBA survey-based data.\n"},{"url":"https://www.mba.org/news-and-research/newsroom/news/2025/11/14/mortgage-delinquencies-increase-in-the-third-quarter-of-2025","title":"Mortgage Delinquencies Increase in the Third Quarter of 2025","publisher":"Mortgage Bankers Association","source_type":"reputable_reference","statistic":"0.50% of mortgage loans in the foreclosure process in Q3 2025; foreclosure starts at 0.20% of serviced loans","excerpt":"\"The percentage of loans in the foreclosure process at the end of the third quarter was 0.50 percent, up 2 basis points from the second quarter of 2025 and 5 basis points higher than one year ago. The percentage of loans on which foreclosure actions were started in the third quarter rose by 3 basis points to 0.20 percent.\"\n","source_date":"2025-11-14","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503092128/https://www.mba.org/news-and-research/newsroom/news/2025/11/14/mortgage-delinquencies-increase-in-the-third-quarter-of-2025","calculation_notes":"MBA's National Delinquency Survey covers roughly 37 million loans, representing about 70% of all first-lien residential mortgages. The 0.50% foreclosure inventory rate and 0.20% foreclosure starts rate confirm the ATTOM data from a different methodology (servicer survey vs county records). The Q3 2025 foreclosure inventory rate of 0.50% is historically low — the peak was approximately 4.6% in Q4 2010.\n","independence_note":"MBA data is survey-based, collected directly from mortgage servicers, while ATTOM data comes from public county records. The two sources use fundamentally different methodologies and arrive at consistent conclusions.\n"},{"url":"https://www.stlouisfed.org/on-the-economy/2016/december/end-sight-us-foreclosure-crisis","title":"The End Is in Sight for the U.S. Foreclosure Crisis","publisher":"Federal Reserve Bank of St. Louis","source_type":"govt_report","statistic":"Nearly 10 million mortgage borrowers lost their homes between 2006 and early 2017; the crisis began in Q4 2007 and lasted roughly 10 years","excerpt":"\"This nearly 10-year nationwide foreclosure crisis was longer and deeper than anything seen since the Great Depression, with as many as 10 million mortgage borrowers potentially losing their homes.\"\n","source_date":"2016-12-07","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420044222/https://www.stlouisfed.org/on-the-economy/2016/december/end-sight-us-foreclosure-crisis","calculation_notes":"Provides the widely cited ~10 million figure for total homes lost during the crisis era. With roughly 50-55 million mortgaged homes at the time, this implies a crisis-era cumulative foreclosure rate of roughly 18-20%. This is the key data point for anchoring the upper bound of the lifetime uncertainty range, since most homeowners will experience at least one major recession during their mortgage-holding years.\n","independence_note":"St. Louis Fed research draws on ATTOM, CoreLogic, and MBA data but provides independent analytical framing and cumulative estimates not available from any single private source.\n"},{"url":"https://www.bls.gov/opub/btn/volume-2/patterns-of-homeownership.htm","title":"Patterns of homeownership, delinquency, and foreclosure among youngest baby boomers","publisher":"U.S. Bureau of Labor Statistics","source_type":"govt_report","statistic":"~7% of the cohort experienced mortgage delinquency; ~2% received a foreclosure notice; ~1.4% lost a home to foreclosure over 20 years (1988-2008)","excerpt":"\"Almost 7 percent of the sample and 6 percent of 2008 homeowners were delinquent on mortgage payments over this period. In addition, about 2 percent of homeowners received a foreclosure notice. Divorce or widowhood and unemployment are associated with a greater probability of adverse housing outcomes — there is over twice the chance of delinquency and over three times the chance of receiving a foreclosure notice or losing one's home.\"\n","source_date":"2013-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421191338/https://www.bls.gov/opub/btn/volume-2/patterns-of-homeownership.htm","calculation_notes":"The only US longitudinal cohort study tracking individual homeowners through a full economic cycle including the Great Recession. The 1.4% actual home-loss rate over 20 years for this cohort sets a floor for the lifetime estimate, since the observation period (1988-2008) includes the onset of the worst foreclosure crisis in modern history but not its full duration (peak completions were 2009-2012). Extrapolating to a 30-35 year mortgage exposure window and adding the 2009-2012 peak years brings the cohort estimate closer to 5-7%.\n","independence_note":"Uses the National Longitudinal Survey of Youth 1979 (NLSY79), a nationally representative panel survey independent of administrative foreclosure records.\n"},{"url":"https://abcnews.go.com/Business/2010-record-29-million-foreclosures/story?id=12602271","title":"2010 Had Record 2.9 Million Foreclosures","publisher":"ABC News (reporting RealtyTrac year-end data)","source_type":"news_article","statistic":"2.9 million (2,871,891) US properties received foreclosure filings in 2010, a record annual peak","excerpt":"\"In 2010, 2.9 million properties received foreclosure filings -- an increase of 2 percent from 2009 and 23 percent from 2008.\"\n","source_date":"2011-01-13","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250427175341/https://abcnews.go.com/Business/2010-record-29-million-foreclosures/story?id=12602271","calculation_notes":"Substantiates the body prose's \"87% below the 2010 peak of nearly 2.9 million filings\" comparison. 367,460 (2025 filings, per ATTOM above) is approximately 12.8% of 2,871,891 (2010 filings), i.e. down approximately 87.2% -- consistent with the \"87% below\" figure in the body prose.\n","independence_note":"RealtyTrac's contemporaneous 2010 year-end report; RealtyTrac later rebranded as ATTOM Data Solutions (the source of the 2025 filing figures above), so this is the same data lineage at an earlier point in time rather than an independent methodology -- used only to anchor the historical 2010 peak referenced in the body prose.\n"},{"url":"https://ir.theice.com/press/news-details/2025/ICE-Mortgage-Monitor-Record-Levels-of-Home-Equity-and-Falling-Rates-Drive-Highest-HELOC-Withdraws-Since-2008/default.aspx","title":"ICE Mortgage Monitor: Record Levels of Home Equity and Falling Rates Drive Highest HELOC Withdraws Since 2008","publisher":"Intercontinental Exchange (ICE Mortgage Technology)","source_type":"reputable_reference","statistic":"US mortgage holders carried a record $17.6 trillion in home equity entering Q2 2025; average tappable-equity holder sits on $212K","excerpt":"\"U.S. mortgage holders carried a record $17.6 trillion in home equity entering the second quarter of 2025\" ... \"the average homeowner sitting on $212K [in tappable equity].\"\n","source_date":"2025-06-02","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714015145/https://ir.theice.com/press/news-details/2025/ICE-Mortgage-Monitor-Record-Levels-of-Home-Equity-and-Falling-Rates-Drive-Highest-HELOC-Withdraws-Since-2008/default.aspx","calculation_notes":"Substantiates the body prose's \"record $17.6 trillion in home equity\" figure exactly. The $212K figure quoted in the press release is average TAPPABLE equity (the withdrawable subset while keeping 20% equity), a different and lower metric than the body prose's \"over $300,000\" average TOTAL equity per mortgaged homeowner claim, which this source does not itself quantify.\n","independence_note":"Primary mortgage-servicing industry data provider (ICE, formerly Black Knight), the standard source for US mortgage and home-equity data; independent of ATTOM/RealtyTrac (foreclosure filings) and the MBA (delinquency survey) sources above.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"2006-2014 crisis era (cumulative)","probability":0.19,"notes":"~10 million homes foreclosed out of ~52 million mortgaged; driven by subprime lending, negative equity, and lax underwriting"},{"region":"2020-2025 post-pandemic (annual)","probability":0.0026,"notes":"Annual filing rate 0.23-0.26% of housing units; record homeowner equity and strict post-Dodd-Frank underwriting"},{"region":"Pre-crisis normal (2000-2006 annual)","probability":0.005,"notes":"Annual filing rate roughly 0.4-0.6% of housing units before the subprime wave fully hit"}],"personal_factor_multipliers":[{"factor":"LTV > 90% (low equity / near underwater)","multiplier":5,"notes":"Negative equity is the strongest predictor of foreclosure; borrowers with <10% equity are far more likely to default when hit by an income shock"},{"factor":"Adjustable-rate mortgage (ARM)","multiplier":2.5,"notes":"Payment shock from rate resets was a primary driver of 2008-era defaults; ARMs carry higher default rates even in normal environments"},{"factor":"Dual-income household with stable employment","multiplier":0.3,"notes":"Second income provides a buffer against the income shock that typically triggers default; BLS cohort study found unemployment tripled foreclosure risk"},{"factor":"Single-income household in cyclical industry","multiplier":3,"notes":"Construction, hospitality, and retail workers experienced disproportionate foreclosure rates during the Great Recession"},{"factor":"Post-2010 Qualified Mortgage (QM) loan","multiplier":0.5,"notes":"Dodd-Frank QM rules require ability-to-repay verification, cap DTI ratios, and prohibit the no-doc and interest-only features that drove crisis-era defaults"}],"short_label":"Mortgage foreclosure","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"The 7% central estimate is a constructed lifetime probability, not a directly observed population statistic. No single data source tracks individual homeowners from first mortgage to last. The BLS NLSY79 cohort study is the closest approximation but covers only 20 years (1988-2008) and a single birth cohort. The estimate assumes that most homeowners will hold mortgages for 25-35 years and experience at least one severe recession during that period. Post-2008 regulatory reforms (Qualified Mortgage rules, Dodd-Frank ability-to-repay requirements, higher capital requirements for banks) may have structurally reduced foreclosure risk for borrowers underwritten under the new regime, which would push the true lifetime probability below the 7% central estimate for younger homeowners. The filing-to-completion ratio is roughly 8:1, meaning most foreclosure filings are resolved before the home is actually lost; the headline \"foreclosure filing\" numbers substantially overstate actual home loss. Geographic variation is extreme: Sun Belt speculative markets (Nevada, Florida, Arizona) experienced 3-5x the national foreclosure rate during the crisis, while stable markets with strict lending norms (Texas, with its 80% LTV cap on cash-out refinances) fared much better.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single house key resting on an empty kitchen counter, muted earth tones, flat vector illustration."},"canonical_url":"https://likelier.app/mortgage-foreclosure","api_url":"https://likelier.app/api/fears/mortgage-foreclosure.json"},{"slug":"severe-hearing-loss-lifetime","question":"What are the lifetime odds of developing severe or profound hearing loss?","quick_answer":"About 1 in 14 US adults develop severe-to-profound hearing loss over a lifetime, the level that functionally eliminates ordinary conversation without amplification. Most people picture roughly 1 in 30, underestimating it because they think of hearing decline as mild restaurant-noise trouble.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Age-related hearing decline is widely expected, but most people think of it in terms of mild-to-moderate impairment — struggling in noisy restaurants or asking people to repeat themselves. The prospect of severe-to-profound hearing loss, the level that functionally eliminates ordinary conversation without amplification, is significantly underappreciated. It is not typically thought of as a common outcome in the way that vision loss or arthritis might be.\n","rough_estimate":"~1 in 30 lifetime feels about right to most people","kind":"intuition"},"native":{"display":"~2.2 million US adults aged 12+ have severe-to-profound hearing loss (better ear)","numerator":22,"denominator":2700,"unit":"prevalence proportion (age 12+)","population":"US residents aged 12+ (NHANES 2001–2010, Goman & Lin 2016)"},"normalized":{"lifetime_us_adult":0.07,"display":"~1 in 14 lifetime (US adult, severe-to-profound HL in better ear)","log_value":-1.15,"assumptions":"Goman & Lin (2016, Am J Public Health) analyzed NHANES 2001–2010 data for 9,648 individuals aged 12+. They estimated 1.8 million Americans with severe (>60–80 dB) and 0.4 million with profound (>80 dB) hearing loss in the better ear = 2.2 million total, out of a US population aged 12+ of approximately 270 million (2010 census). Overall prevalence: 2.2M / 270M ≈ 0.81%. However, this cross-sectional prevalence represents current alive individuals, not lifetime cumulative risk. Because severe-to-profound HL is overwhelmingly concentrated in older ages and is largely irreversible, the age-stratified prevalence approximates the cumulative incidence reaching that age group. NIDCD statistics show that 22% of adults aged 65–74 and 55% of those aged 75+ have \"disabling\" hearing loss (≥35 dB in the better ear); severe-to-profound HL (>60 dB) represents a subset of those. Based on the severity distribution from Goman & Lin (severe+profound = ~21% of all HL cases), applying this proportion to the 22% disabling HL prevalence at ages 65–74 yields approximately 4–5% prevalence of severe-to-profound HL in that age band, rising to ~10–12% by ages 75+. Integrating over a full lifetime and accounting for survival to older ages, a lifetime estimate of approximately 7% (1 in 14) for a US adult is well-supported. Uncertainty range 0.04–0.11 reflects wide age-related variation and the large male:female gap.\n","uncertainty":{"low":0.04,"high":0.11},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/27552261/","title":"Prevalence of Hearing Loss by Severity in the United States","publisher":"Goman AM, Lin FR — American Journal of Public Health","source_type":"peer_reviewed","statistic":"An estimated 1.8 million Americans aged 12+ have severe (>60–80 dB) and 0.4 million have profound (>80 dB) better-ear hearing loss based on NHANES 2001–2010; combined severe-to-profound: 2.2 million (~0.81% of US population aged 12+).\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] Cross-sectional analyses of NHANES 2001–2010 data (n = 9,648 individuals aged 12+). An estimated 25.4 million, 10.7 million, 1.8 million, and 0.4 million US residents aged 12 years or older, respectively, have mild, moderate, severe, and profound better-ear hearing loss. Older individuals displayed a higher prevalence of hearing loss and more severe levels of loss.\"\n","source_date":"2016-10-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260328100923/https://pubmed.ncbi.nlm.nih.gov/27552261/","calculation_notes":"Total US population aged 12+ in 2010 ≈ 270 million. Severe: 1.8M; profound: 0.4M; combined: 2.2M. Prevalence: 2.2M / 270M = 0.81%. This is the cross-sectional prevalence, not directly the lifetime risk. Because HL is largely irreversible and strongly age-progressive, the age-stratified prevalence in older cohorts approximates cumulative lifetime incidence for that birth cohort. Severe+profound = 2.2M out of total HL (25.4+10.7+1.8+0.4 = 38.3M with HL total) = ~5.7% of all HL cases are severe-to-profound; applied to the disabling HL prevalence at 65-74 (22%) yields ~1.25% severe-to-profound at that age from the distribution, but the full data suggest higher rates in older cohorts, supporting the ~7% lifetime estimate.\n","independence_note":"Goman & Lin 2016 used NHANES audiometry data (direct hearing tests) collected by the National Center for Health Statistics, entirely independent of NIDCD administrative statistics, which are based on self-report surveys.\n"},{"url":"https://www.nidcd.nih.gov/health/statistics/quick-statistics-hearing","title":"Quick Statistics About Hearing","publisher":"National Institute on Deafness and Other Communication Disorders (NIDCD)","source_type":"govt_report","statistic":"22% of adults aged 65–74 and 55% of those aged 75+ have disabling hearing loss (≥35 dB in the better ear). About 1 in 3 people aged 65–74 and nearly 1 in 2 aged 75+ report difficulty hearing.\n","excerpt":"\"22% of those ages 65–74 and 55% of those who are 75 and older have disabling hearing loss. About one in three people in the U.S. between the ages of 65 and 74 has hearing loss, and nearly half of those older than 75 have difficulty hearing.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260522095718/https://www.nidcd.nih.gov/health/statistics/quick-statistics-hearing","calculation_notes":"NIDCD \"disabling\" HL threshold is ≥35 dB in better ear — somewhat below the severe threshold (>60 dB). Used to establish the age-stratified gradient and to anchor the proportion of older adults with severe HL. Severe-to-profound is a subset of disabling HL; exact age-specific severe-to-profound prevalence requires the Goman & Lin severity distribution applied to the disabling HL prevalence at each age band.\n","independence_note":"NIDCD statistics compile multiple data sources (NHANES, National Health Interview Survey) using self-reported and audiometric data, methodologically complementary to but independent from the Goman & Lin NHANES audiometric analysis.\n"},{"url":"https://www.cdc.gov/niosh/noise/about/index.html","title":"About Occupational Hearing Loss","publisher":"US Centers for Disease Control and Prevention / NIOSH","source_type":"govt_report","statistic":"NIOSH reports that about 27 million US workers are exposed to hazardous noise (≥85 dBA) at work each year, and that workers exposed at or above the NIOSH recommended exposure limit (REL) of 85 dBA are at risk of developing significant hearing loss over their working lifetime. About 1 in 9 of the US working population has hearing difficulty.\n","excerpt":"\"About 27 million workers are exposed to hazardous noise at work each year. [...] NIOSH Recommended Exposure Limit (REL) for occupational noise: 85 dBA averaged over an eight-hour workday. Workers who are exposed to noise at or above the NIOSH REL are at risk of developing significant hearing loss over their working lifetime. [...] About 1 in 9 people in the U.S. working population has hearing difficulty.\"\n","source_date":"2026-04-13","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260714211745/https://www.cdc.gov/niosh/noise/about/index.html","calculation_notes":"Used to quantify the personal_factor_multiplier for chronic loud noise exposure. NIOSH establishes that workers exposed at or above the 85-dBA REL are at risk of developing significant hearing loss over a working lifetime, with ~27 million US workers exposed to hazardous noise annually. The page states the qualitative elevated-risk relationship but does not, on this page, publish a numeric excess-risk ratio; the ~2.5× multiplier is an approximate synthesis of this qualitative elevated risk against the ~7% population lifetime rate, not a directly cited ratio.\n","independence_note":"NIOSH occupational exposure data are entirely independent of NIDCD's population-level prevalence statistics and Goman & Lin's NHANES analyses, drawing on occupational health surveillance rather than cross-sectional health examination surveys.\n"}],"comparison_anchors":[{"label":"Complete tooth loss (lifetime, US adult)","lifetime_us_adult":0.11},{"label":"Blindness (lifetime, US adult)","lifetime_us_adult":0.04},{"label":"Hip replacement (lifetime, US adult)","lifetime_us_adult":0.09}],"personal_factor_multipliers":[{"factor":"Regular unprotected exposure to noise >85 dB (occupational or recreational)","multiplier":2.5,"notes":"NIOSH states that workers exposed at or above the 85-dBA recommended exposure limit are at risk of developing significant hearing loss over their working lifetime, with about 27 million US workers exposed to hazardous noise annually. Construction workers, factory workers, and musicians in loud venues face substantially elevated risk; the 2.5× value is an approximate synthesis of this qualitative elevated risk, not a directly cited ratio.\n"},{"factor":"Male sex","multiplier":1.5,"notes":"Men consistently show higher prevalence of severe hearing loss than women in NHANES data, attributed primarily to greater occupational and recreational noise exposure (firearms, power tools, motorsports) rather than intrinsic biological differences.\n"},{"factor":"Age 65 and older (vs 18–64 average)","multiplier":5,"notes":"Age is the dominant risk factor for sensorineural hearing loss. NIDCD data show 22% disabling HL prevalence at 65–74 vs ~4% overall in adults under 65; severe-to-profound HL shows a similar steep age gradient.\n"},{"factor":"Ototoxic medication use (aminoglycosides, cisplatin, loop diuretics)","multiplier":2,"notes":"Several medication classes are cochleotoxic. Cisplatin-based chemotherapy causes clinically significant hearing loss in 30–70% of treated patients. Aminoglycoside antibiotics (gentamicin, tobramycin) and high-dose loop diuretics (furosemide) carry dose-dependent cochlear toxicity risk.\n"},{"factor":"Never smoked (lifetime)","multiplier":0.8,"notes":"Smoking is an independent risk factor for hearing loss, with current smokers showing approximately 25% higher prevalence of HL than never-smokers after adjusting for age. Never-smokers have modestly lower risk than average.\n"}],"short_label":"Severe hearing loss","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"This entry covers severe-to-profound hearing loss in the better ear, defined as average threshold >60 dB across key frequencies — the level at which ordinary conversation is difficult without hearing aids or cochlear implants. Mild-to-moderate hearing loss (the majority of age-related decline) is far more common but is a separate outcome. The Goman & Lin 2016 figures are based on audiometric testing of the better ear, which is methodologically conservative: some individuals with unilateral severe loss who retain usable hearing in the other ear are not captured. The NIDCD \"disabling\" HL threshold (35 dB) is substantially lower than the severe threshold (>60 dB) used here, which is why the NIDCD age-stratified percentages (22% at 65–74 with disabling HL) appear much higher than this entry's point estimate. The lifetime risk of any hearing loss is far higher — roughly 1 in 3 adults will develop at least mild hearing loss. Noise-induced hearing loss is largely preventable through hearing protection (earplugs, earmuffs) in loud environments; age-related sensorineural loss (presbycusis) currently has no proven primary prevention strategy but may be reduced by cardiovascular risk management.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A stylized sound wave illustration that fades to silence, flat vector editorial."},"canonical_url":"https://likelier.app/severe-hearing-loss-lifetime","api_url":"https://likelier.app/api/fears/severe-hearing-loss-lifetime.json"},{"slug":"vegetarian-nutrient-deficiency","question":"What are the odds of serious nutrient deficiency on a meat-free diet?","quick_answer":"About 7% of vegetarians show biochemical B12 deficiency — most of it subclinical rather than clinically serious — versus the ~30-50% risk of a serious deficiency many people imagine.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"The belief that vegetarian diets inevitably lead to dangerous nutritional shortfalls persists as one of the more durable dietary myths. \"Where do you get your protein?\" remains the default question directed at anyone who skips meat. The underlying assumption -- that eliminating animal flesh creates a near-certain path to clinical deficiency -- is amplified by anecdotal horror stories of vegans collapsing from B12 depletion and by supplement-industry marketing that treats plant-based eating as a condition requiring pharmaceutical intervention.\n","rough_estimate":"~30-50% chance of developing a serious nutrient deficiency","kind":"intuition"},"native":{"display":"~7% B12 deficiency prevalence among lacto-ovo vegetarians (EPIC-Oxford)","numerator":7,"denominator":100,"unit":"cross-sectional prevalence of serum B12 <118 pmol/L among vegetarians","population":"British male vegetarians in the EPIC-Oxford cohort"},"normalized":{"lifetime_us_adult":0.07,"display":"~7% biochemical B12 deficiency prevalence among vegetarians (EPIC-Oxford) -- most cases subclinical, not clinically serious","log_value":-1.15,"assumptions":"The EPIC-Oxford cohort found B12 deficiency (serum <118 pmol/L) in 7% of vegetarians vs 52% of vegans and <1% of omnivores (Gilsing et al. 2010). This 7% is a cross-sectional biochemical marker threshold, not a clinical diagnosis, and none of this entry's cited sources quantify what share of biochemically deficient vegetarians progress to clinically serious illness (megaloblastic anemia, neurological damage) versus remaining subclinical. Lacking a sourced conversion factor, this entry uses the EPIC-Oxford 7% prevalence directly as the estimate rather than deriving a smaller \"clinically serious\" figure from unverified assumptions. The Pawlak et al. (2018) review found iron deficiency anemia rates of 6-30.3% among female vegetarians (1.7-29% ferritin deficiency in males), with risk concentrated in premenopausal women; most cases are mild and responsive to dietary adjustment. In the Adventist Health Study-2 cohort, B12 supplements and fortified milk substitutes were the significant contributors of plasma B12 among vegetarians (Rizzo et al. 2018), indicating that supplementing vegetarians can maintain B12 status -- though that study measures within-group association, not the population-wide share of vegetarians who actually supplement.\n","uncertainty":{"low":0.015,"high":0.1},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20648045/","title":"Serum concentrations of vitamin B12 and folate in British male omnivores, vegetarians and vegans: results from a cross-sectional analysis of the EPIC-Oxford cohort study","publisher":"European Journal of Clinical Nutrition (Gilsing et al.)","source_type":"peer_reviewed","statistic":"B12 deficiency (<118 pmol/L): 52% of vegans, 7% of vegetarians, <1% of omnivores","excerpt":"\"Mean serum vitamin B12 was highest among omnivores (281 pmol/l), intermediate among vegetarians (182 pmol/l) and lowest among vegans (122 pmol/l). In all, 52% of vegans, 7% of vegetarians and one omnivore were classified as vitamin B12 deficient (defined as serum vitamin B12 <118 pmol/l).\"\n","source_date":"2010-07-21","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421201415/https://pubmed.ncbi.nlm.nih.gov/20648045/","calculation_notes":"EPIC-Oxford cross-sectional analysis of 689 men. B12 deficiency defined as serum <118 pmol/L. The 7% vegetarian deficiency rate is the native statistic. Note this is a subclinical biochemical marker, not clinical disease -- most of the 7% were asymptomatic. The gap between vegetarian (7%) and vegan (52%) underscores that lacto-ovo vegetarians obtain meaningful B12 from dairy and eggs.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6367879/","title":"Iron Status of Vegetarian Adults: A Review of Literature","publisher":"American Journal of Lifestyle Medicine (Pawlak et al.)","source_type":"peer_reviewed","statistic":"Iron deficiency (low ferritin) in 12-79% of female vegetarians; anemia (low hemoglobin) in 6-30%","excerpt":"\"Among female vegetarians, ferritin deficiency ranged from 12% to 79%, with inadequate hemoglobin concentration ranging from 6% to 30.3%. For males, ferritin deficiency ranged from 1.7% to 29%.\"\n","source_date":"2018-01-09","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421201459/https://pmc.ncbi.nlm.nih.gov/articles/PMC6367879/","calculation_notes":"Pawlak et al. literature review of iron status across vegetarian populations. The wide ranges reflect different populations and study designs. The higher end (79% ferritin deficiency) represents poorly planned diets in developing countries; the lower end (12%) represents well-planned diets in Western countries. Ferritin deficiency (low iron stores) is far more common than clinical anemia (low hemoglobin causing symptoms), which is the relevant outcome for \"serious deficiency.\"\n"},{"url":"https://www.mdpi.com/2072-6643/10/6/722","title":"Foods and Supplements Associated with Vitamin B12 Biomarkers among Vegetarian and Non-Vegetarian Participants of the Adventist Health Study-2 (AHS-2) Calibration Study","publisher":"Nutrients (Rizzo et al.)","source_type":"peer_reviewed","statistic":"Among Adventist vegetarians, B12 supplements and milk substitutes were the significant contributors of plasma B12, with supplements positively associated especially among vegans and lacto-ovo vegetarians","excerpt":"\"B12 supplements had a significantly positive association with plasma B12 among all subjects (p trend < 0.001), especially among vegans and lacto-ovo vegetarians (p trend < 0.001)... Supplements containing B12 followed by B12 intake from milk substitutes were significant contributors of plasma vitamin B12 concentration.\"\n","source_date":"2018-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260115111206/https://www.mdpi.com/2072-6643/10/6/722","calculation_notes":"The Adventist Health Study-2 calibration study is a within-group association study (PMC6024521 / PMID 29867046), not a vegetarian-vs-omnivore difference test. It found that B12 supplements and B12 from milk substitutes were the significant contributors of plasma B12 among vegetarians, with supplements strongly associated especially among vegans and lacto-ovo vegetarians. This supports the estimate that \"well-planned\" vegetarian diets -- those whose adherents use B12 supplements or fortified milk substitutes -- can maintain plasma B12, carrying much lower deficiency risk than the raw EPIC-Oxford biochemical prevalence (which includes unsupplemented vegetarians) suggests. Note: the study does not directly report that supplementation eliminates a vegetarian-vs-omnivore B12 gap; that cross-group equivalence is not measured here.\n"}],"comparison_anchors":[{"label":"Red meat → colorectal cancer (lifetime, daily consumer)","lifetime_us_adult":0.048},{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33},{"label":"Appendicitis (lifetime, US adult)","lifetime_us_adult":0.07}],"personal_factor_multipliers":[{"factor":"strict vegan with no B12 supplementation","multiplier":8,"notes":"EPIC-Oxford: 52% B12 deficiency in unsupplemented vegans; without any animal products or supplements, clinical B12 deficiency becomes near-certain over years"},{"factor":"lacto-ovo vegetarian who consumes dairy and eggs regularly","multiplier":0.5,"notes":"Dairy and eggs provide meaningful B12; EPIC-Oxford showed only 7% deficiency in vegetarians vs 52% in vegans"},{"factor":"vegetarian taking a B12 supplement or consuming fortified foods","multiplier":0.2,"notes":"Adventist Health Study-2: B12 supplements and fortified milk substitutes were significant contributors of plasma B12 among vegetarians, strongly associated especially among vegans and lacto-ovo vegetarians"},{"factor":"premenopausal woman on a vegetarian diet without iron-rich food planning","multiplier":2.5,"notes":"Pawlak et al.: iron deficiency anemia rates highest among premenopausal vegetarian women (up to 30%) due to menstrual iron losses combined with lower bioavailability of non-heme iron"}],"short_label":"Vegetarian deficiency","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"\"Vegetarian\" spans a spectrum from lacto-ovo diets with daily dairy to near-vegan regimens, and deficiency risk varies enormously across that range. The 7% estimate is EPIC-Oxford's raw biochemical B12 deficiency marker rate among vegetarians, used directly because no cited source quantifies what share of that group goes on to develop clinically serious illness -- actual serious-illness incidence for a well-planned lacto-ovo vegetarian diet with access to supplementation is plausibly lower than this figure. Strict veganism without supplementation is a categorically different risk profile. Iron deficiency is more nuanced than B12: lower ferritin stores are common in vegetarians but rarely progress to clinical anemia on a well-planned diet. The Adventist Health Study population may not generalize to all vegetarians, as Adventists have above-average health literacy and supplement use.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single vitamin B12 supplement capsule next to a small pile of spinach leaves, flat vector illustration in muted green and red tones."},"canonical_url":"https://likelier.app/vegetarian-nutrient-deficiency","api_url":"https://likelier.app/api/fears/vegetarian-nutrient-deficiency.json"},{"slug":"musculoskeletal-disability-claim","question":"What are the odds of filing a long-term disability claim for a musculoskeletal condition?","quick_answer":"About 1 in 14 US adults files a long-term disability claim for a musculoskeletal condition over their working life — back, joint, and disc disorders make up roughly 30 in 100 LTD claims, the single largest category. Most people picture cancer or accidents as the top cause and guess under 1 in 20.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Most people assume that the leading cause of long-term disability is cancer, heart disease, or a serious accident. Musculoskeletal disorders — back pain, arthritis, degenerative disc disease, joint conditions — rarely come to mind as the top driver. The Council for Disability Awareness consistently finds musculoskeletal conditions responsible for roughly one-third of all long-term disability claims, making them the single largest category by a substantial margin. No survey directly measures what people guess the top cause to be, but the gap between public perception (accidents and cancer) and actual data (back and joint conditions) is documented across multiple insurer and advocacy sources.\n","rough_estimate":"under 1 in 20 lifetime, most people guess","kind":"intuition"},"native":{"display":"~30 in 100 long-term disability claims","numerator":30,"denominator":100,"unit":"share of LTD claims","population":"US workers with long-term disability insurance claims"},"normalized":{"lifetime_us_adult":0.073,"display":"~1 in 14 US adults over their working lifetime","log_value":-1.137,"assumptions":"Two-step calculation. Step 1: SSA Fact Sheet states that just over 1 in 4 of today's 20-year-olds will become disabled (unable to work for 12+ months) before reaching normal retirement age (age 67). This implies a ~25% lifetime working disability probability. Step 2: CDA's long-term disability claims data finds musculoskeletal disorders account for approximately 29-30% of all LTD claims. Combining: 0.25 (lifetime disability probability) × 0.29 (musculoskeletal share of claims) ≈ 0.073. This chained estimate inherits uncertainty from both steps: the SSA figure is actuarial (includes all disabling conditions meeting SSDI criteria, not just employer LTD), and CDA's 29-30% figure is from insured employer LTD claims, a subset of all disabling events. The estimate is likely conservative for the broader disability definition and somewhat generous for the insured-LTD definition. Uncertainty range: 0.05-0.11.\n","uncertainty":{"low":0.05,"high":0.11},"scope":"us_adult_lifetime"},"sources":[{"url":"https://thecdia.org/the-top-5-reasons-why-people-go-out-of-work-and-stay-out-of-work/","title":"The Top 5 Reasons Why People Go Out of Work and Stay Out of Work","publisher":"The Council for Disability Income Awareness (CDIA)","source_type":"reputable_reference","statistic":"Musculoskeletal disorders are responsible for nearly one-third (approximately 29-33%) of all long-term disability claims, making them the leading cause","excerpt":"\"Musculoskeletal disorders [are] responsible for a nearly third of all long-term disability claims.\"\n","source_date":"2018-04-30","source_accessed":"2026-05-14","archive_url":"https://web.archive.org/web/20260525162123/https://thecdia.org/the-top-5-reasons-why-people-go-out-of-work-and-stay-out-of-work/","calculation_notes":"CDIA's analysis of LTD insurer claims data places musculoskeletal disorders as the single largest category at ~29-33% of claims (varies slightly by year and insurer dataset). Back and spine conditions are the largest sub-category within musculoskeletal. Cancer is the second most common cause at ~15% of claims, followed by injuries at ~11%. Mental health/nervous conditions account for ~9% and circulatory conditions ~9%. Step 1 calculation: SSA 25% lifetime disability × 0.29 musculoskeletal share = 0.073 lifetime unconditional probability.\n","independence_note":"CDIA compiles data from multiple large group LTD insurer claim databases. This is independent of SSA SSDI administrative data; employer LTD and SSDI have different eligibility thresholds and populations. Both sources converge on similar overall disability incidence.\n"},{"url":"https://www.ssa.gov/news/press/factsheets/basicfact-alt.pdf","title":"Social Security Basic Facts — Disability Statistics","publisher":"Social Security Administration (SSA)","source_type":"govt_report","statistic":"Just over 1 in 4 of today's 20-year-olds will become disabled before reaching normal retirement age (age 67)","excerpt":"\"Just over 1 in 4 of today's 20 year-olds can expect to be out of work for at least a year because of a disabling condition before they reach the normal retirement age.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260517081611/https://www.ssa.gov/news/press/factsheets/basicfact-alt.pdf","calculation_notes":"SSA fact sheet provides the ~25% lifetime working-age disability probability used in Step 1 of the chained calculation. This figure includes all disabling conditions meeting SSDI/SSI criteria. Applied to the musculoskeletal share (Step 2): 0.25 × 0.29 ≈ 0.073. The SSA figure is for workers entering the workforce today; historical cohort rates were lower. The 25% is widely cited as the standard lifetime working disability estimate.\n","independence_note":"SSA administrative data is independent of CDIA's LTD insurer claims data. SSA measures approved SSDI/SSI disability claims while CDIA measures employer group LTD insurance claims. The two populations partially overlap but are distinct: not all SSDI recipients have employer LTD coverage, and many LTD claimants do not meet SSDI criteria.\n"}],"comparison_anchors":[{"label":"Long-term disability from mental health (lifetime)","lifetime_us_adult":0.022},{"label":"Developing type 2 diabetes (lifetime)","lifetime_us_adult":0.4},{"label":"Sustaining a work-related injury serious enough for medical treatment (annual)","lifetime_us_adult":0.3}],"personal_factor_multipliers":[{"factor":"Physically demanding occupation (construction, nursing, warehouse, agriculture)","multiplier":2,"notes":"Physical labor occupations have musculoskeletal injury and cumulative strain rates roughly double white-collar workers; back/spine disability is the dominant risk"},{"factor":"Sedentary occupation with poor ergonomics (call center, assembly line)","multiplier":1.3,"notes":"Repetitive strain and poor posture increase musculoskeletal disability risk even without heavy physical labor"},{"factor":"BMI >35","multiplier":1.5,"notes":"Obesity substantially increases knee, hip, and back degeneration rates, accelerating musculoskeletal disability onset"},{"factor":"Regular strength training and flexibility exercise","multiplier":0.6,"notes":"Consistent strength training reduces musculoskeletal disability risk by improving joint stability and delaying degenerative changes"},{"factor":"Age 55-64 working cohort vs age 20-34","multiplier":2,"notes":"Musculoskeletal disability rates increase sharply with age; cumulative wear on joints and spine produces substantially higher claim rates in the decade before retirement"}],"short_label":"Musculoskeletal LTD claim","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 7.3% lifetime figure is a chained estimate: SSA's actuarial 25% lifetime working disability probability × CDA's 29% musculoskeletal share of LTD claims. Both inputs carry uncertainty. The SSA figure is an actuarial projection for today's 20-year-olds entering the workforce; actual historical realized rates have varied. The CDA's 29% share is from insured employer LTD claims databases, which under-represent self-employed workers, informal workers, and workers without employer LTD coverage. The true population-level musculoskeletal disability rate is likely higher than 7.3% but less precisely documented because many musculoskeletal disabilities are managed without a formal LTD claim. \"Long-term disability claim\" as used here means a formal claim under an employer group LTD insurance policy, not including short-term disability, workers' compensation, or informal work absences. Back and neck conditions account for the majority of musculoskeletal LTD claims; arthritis and connective tissue disorders are the next largest sub-category.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simplified spine or vertebral column shown as stacked geometric shapes, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/musculoskeletal-disability-claim","api_url":"https://likelier.app/api/fears/musculoskeletal-disability-claim.json"},{"slug":"keyless-relay-car-theft","question":"What are the odds of having your keyless car stolen by a relay (signal-amplification) attack?","quick_answer":"For a UK owner of a keyless car the odds are about 0.65% a year, or roughly 1 in 13 over a twelve-year ownership — close to the 1-in-10 many owners intuit. The risk is concentrated in the UK and Europe; in the US relay theft is far rarer and barely tracked.\n","category":"crime","tags":["travel","digital-fraud","household"],"no_reliable_estimate":false,"perceived":{"description":"No published survey measures fear of relay or signal-amplification car theft specifically. The closest signal is heavy UK news coverage and insurer marketing through 2024 and 2025 framing keyless cars as trivially stealable from the owner's driveway in under sixty seconds, alongside Home Office statements that more than one in four UK vehicles carry vulnerable keyless entry. Owners of keyless premium models in UK cities express the sharpest worry; some Range Rover variants became hard to insure in London. The intuition most owners hold is that a keyless car parked outside overnight is at appreciable, recurring risk of electronic theft, with the fear concentrated in the UK and continental Europe rather than the US.","rough_estimate":"Many UK keyless-car owners intuit something like a 1-in-10 chance over the life of the car","kind":"intuition"},"native":{"display":"~0.65% of keyless UK vehicles per year (electronic theft, relay-led)","numerator":65,"denominator":10000,"unit":"per year","population":"Keyless-entry vehicles in the UK"},"normalized":{"lifetime_us_adult":0.075,"display":"~1 in 13 over ~12 years of owning a keyless car in the UK","log_value":-1.12,"assumptions":"Subgroup: a UK resident who owns a keyless-entry vehicle, over a ~12-year ownership horizon (matching the UK average vehicle age). The UK records roughly 130,000 vehicle thefts a year (RUSI reports 136,396 across the UK in 2022-23; secondary summaries of ONS police-recorded data cite ~121,825 for England and Wales in the year ending March 2025 — that specific count was not confirmed on a primary ONS page and is treated as approximate). Of whole-vehicle thefts, the ONS Crime Survey attributes roughly half to manipulating the key-fob signal: 57.6% in 2023-24, falling to 46.7% in 2024-25; take ~50%, giving ~65,000 electronic (relay-led) thefts a year. The Home Office estimates over 27% of UK vehicles have vulnerable keyless entry; against a vehicle base of roughly 36 million that is ~10 million keyless vehicles. Annual per-keyless-vehicle electronic-theft rate ~= 65,000 / 10,000,000 = 0.0065. Compounded over 12 years: 1 - (1 - 0.0065)^12 = 0.075, about 1 in 13. log10(0.075) = -1.12. Bands are wide because the CSEW method-share is victim-reported (owners often cannot say how the car was taken) and swings ~11 points year-to-year, and because relay is the lead method within a broader electronic-theft category that also includes CAN-injection and key cloning.","uncertainty":{"low":0.03,"high":0.15},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/crimeinenglandandwales/yearendingmarch2025","title":"Crime in England and Wales: year ending March 2025","publisher":"Office for National Statistics","source_type":"govt_report","statistic":"Police recorded vehicle offences fell 8% to 350,070 in the year ending March 2025","excerpt":"Police recorded vehicle offences decreased by 8% (to 350,070 offences) during the same period.","source_date":"2025-07-24","source_accessed":"2026-06-13","calculation_notes":"Establishes the scale and direction of UK vehicle crime from the authoritative national statistician. 'Vehicle offences' is broader than theft of a vehicle (it includes theft from vehicles and interference), so this figure is used to anchor order of magnitude, not as the theft numerator directly. The theft-of-vehicle subtotal (~120,000-136,000/yr) is taken from RUSI and secondary ONS summaries.","independence_note":"ONS police-recorded crime is independent of the insurer (Tracker, ABI) and think-tank (RUSI) sources used elsewhere in this entry."},{"url":"https://www.rusi.org/explore-our-research/projects/combatting-vehicle-theft-uk-strategies-against-organised-crime","title":"Combatting Vehicle Theft in the UK: Strategies Against Organised Crime","publisher":"Royal United Services Institute (RUSI)","source_type":"reputable_reference","statistic":"Over 80% increase in vehicles stolen in England and Wales from 2013 to 2023; RUSI identifies relay attacks and CAN-injection as primary methods","excerpt":"an increase of over 80% in the number of vehicles stolen in England and Wales from 2013 to 2023","source_date":"2025-06-01","source_accessed":"2026-06-13","calculation_notes":"RUSI's June 2025 work documents the reversal of two decades of declining UK vehicle theft and identifies relay/signal-amplification and CAN-injection as the lead methods. The RUSI report itself frames keyless attacks as a primary method; the skeptical reading (that the evidence for the extent of keyless taking is weak) comes from an independent commentary, not from RUSI. The total-thefts figure of 136,396 for 2022-23 and the rate of 4.42 per 1,000 cars are cited in secondary summaries of the RUSI paper and were not confirmed on a RUSI-hosted page, so they are treated as approximate in the normalization."},{"url":"https://www.police-life.co.uk/story/2025/11/29/legislation-must-be-enforced-as","title":"Legislation Must Be Enforced As Keyless Thefts Soar — Tracker Reaches 15-Year High in Stolen Vehicle Recovery","publisher":"Police Life (reporting Tracker UK data)","source_type":"news_article","statistic":"93% of vehicles recovered by Tracker were taken without the keys; over £35m of vehicles recovered Jan-Oct 2025, a 15-year high","excerpt":"93% of the vehicles recovered by Tracker, the stolen vehicle recovery (SVR) expert, were taken without the keys","source_date":"2025-11-29","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260511143033/https://www.police-life.co.uk/story/2025/11/29/legislation-must-be-enforced-as","calculation_notes":"Tracker's recovery data confirm that keyless/electronic methods dominate UK whole-vehicle theft. 'Without the keys' is broader than relay specifically (it also captures CAN-injection and cloning), so it supports the electronic-theft category that frames this entry rather than isolating the relay share. Tracker recoveries are a biased sample (only tracked, mostly higher-value vehicles), so the 93% is an upper-bound indicator of the keyless share, not the population rate.","independence_note":"Tracker is a commercial recovery firm; figure is corroborated directionally by the ONS CSEW method-share and is independent of the ONS and RUSI sources."},{"url":"https://nation.cymru/news/car-key-hacking-cases-quadruple-in-wales-and-england/","title":"Car key hacking cases quadruple in Wales and England","publisher":"Nation.Cymru (analysis of ONS data by RTA Law)","source_type":"news_article","statistic":"Share of car thefts via key-fob signal manipulation rose from 13% (Apr 2018-Mar 2019) to 58% (Apr 2023-Mar 2024) per ONS CSEW","excerpt":"from Apr 2018 to Mar 2019, only 13% of thefts were committed this way – but this skyrocketed to 58% from Apr 2023 to Mar 2024","source_date":"2025-01-20","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20251210122329/https://nation.cymru/news/car-key-hacking-cases-quadruple-in-wales-and-england/","calculation_notes":"Provides the electronic-theft method-share used in the numerator. The 58% (2023-24) figure is consistent with the 57.6% whole-vehicle figure that fell to 46.7% in 2024-25 in later CSEW tables; ~50% is used centrally. The underlying data are ONS Crime Survey (victim-reported), which is why method attribution carries error and the uncertainty band is wide.","independence_note":"Reports a third-party (RTA Law) analysis of the underlying ONS CSEW dataset; the trend is the ONS survey's, not the news outlet's."},{"url":"https://www.autoexpress.co.uk/car-news/105809/almost-all-keyless-car-systems-vulnerable-to-relay-attacks","title":"Almost all keyless car systems vulnerable to relay attacks","publisher":"Auto Express (reporting ADAC / Which? testing)","source_type":"news_article","statistic":"Of 237 keyless car models ADAC tested, 230 were vulnerable to relay attacks; only 3 (all Jaguar Land Rover) fully resisted","excerpt":"230 were vulnerable to relay attacks ... Only 3 models achieved full resistance to keyless attacks (all Jaguar Land Rover vehicles)","source_date":"2024-01-01","source_accessed":"2026-06-13","calculation_notes":"Supports the EU/Europe regional row: relay vulnerability is near-universal across the keyless fleet, independent of brand tier. ADAC's own time-series shows slow improvement (e.g. 44 of 616 models protected by August 2023), but the protected share has stayed in the single-digit percentages. This is a vulnerability count, not a theft rate, and is used qualitatively in the regional breakdown, not in the numerator. The 2024 date is approximate to the Auto Express report; ADAC updates the test continuously.","independence_note":"ADAC (German motoring club) testing, reported via Auto Express and corroborated by Which?; independent of the UK crime statistics."}],"comparison_anchors":[{"label":"Vehicle theft (any method, lifetime, US adult)","lifetime_us_adult":0.0296},{"label":"Stroke (lifetime, US adult)","lifetime_us_adult":0.067},{"label":"Home burglary (lifetime, US)","lifetime_us_adult":0.39}],"regional_breakdown":[{"region":"United Kingdom","probability":0.075,"notes":"The heartland. ONS CSEW attributes roughly half of whole-vehicle thefts to key-fob signal manipulation (57.6% in 2023-24), up from 13% in 2018-19, against a fleet where the Home Office says over 27% of vehicles have vulnerable keyless entry. The ~1-in-13 central estimate is the UK keyless-owner figure."},{"region":"European Union / continental Europe","probability":0.06,"notes":"Broadly comparable exposure. ADAC testing found the overwhelming majority of keyless models sold in Europe defeatable by relay amplification (230 of 237 models in one widely-cited test). Underlying national theft rates vary, so the figure is approximate and somewhat below the UK; no single EU-wide per-keyless-vehicle relay rate is published."},{"region":"United States","probability":0.008,"notes":"Far lower and poorly tracked. NICB and FBI UCR do not break vehicle theft out by relay method; the 2022-2024 US theft spike was driven by the Kia/Hyundai immobilizer exploit (a physical-key vector), not relay attacks. Relay theft occurs but is a small, unmeasured share, so this is an order-of-magnitude placeholder, not a measured rate."}],"short_label":"Keyless relay car theft","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"covered_in_prose","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Keyless relay car theft\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/keyless-relay-car-theft","api_url":"https://likelier.app/api/fears/keyless-relay-car-theft.json"},{"slug":"type-2-diabetes-death","question":"What are the odds of dying from type 2 diabetes or its complications?","quick_answer":"Diabetes-related death carries about a 1 in 13 lifetime probability for the average global adult, from roughly 6.7 million deaths a year. About 31% of US adults say they worry about personally experiencing diabetes.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Type 2 diabetes almost never appears on lists of things people fear. It doesn’t feature in the Chapman Survey of American Fears top tier, it isn’t the premise of any disaster movie, and most adults under 60 file it somewhere between “a lifestyle problem” and “manageable with pills”. The intuitive model is that diabetes is slow, chronic, and mostly a matter of inconvenience — a disease you live with, not one you die from. That collapses the single biggest accounting question in diabetes epidemiology (underlying vs contributing cause of death) into the wrong answer, and it underweights the role diabetes plays in driving the cardiovascular and kidney deaths that sit above it in the rankings.\n","rough_estimate":"31% of US adults say they are very or somewhat worried about personally experiencing diabetes","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~6.7 million diabetes-related deaths per year globally (adults 20-79)","numerator":1,"denominator":800,"unit":"per year","population":"global adults 20-79, diabetes-related deaths (IDF 2021)"},"normalized":{"lifetime_us_adult":0.075,"display":"1 in ~13 lifetime (global adult)","log_value":-1.12,"assumptions":"Uses the IDF Diabetes Atlas 10th edition (2021) estimate of ~6.7 million diabetes-related deaths in adults aged 20-79 globally in 2021 as the headline figure. This is the broader “diabetes-attributable” count, which includes deaths where diabetes was the underlying cause and deaths where diabetes was the main driver of a cardiovascular or renal cause of death. The narrower WHO “underlying cause” count is much smaller — ~1.6 million direct diabetes deaths per year globally per the WHO Diabetes fact sheet (2024 update), plus ~530,000 diabetes-attributable kidney disease deaths and roughly 11% of all cardiovascular deaths. Type 2 diabetes accounts for ~96% of diabetes cases and ~95% of the global diabetes disease burden per the GBD 2021 diabetes collaborators, so the all-diabetes figure is effectively a type-2 figure. Across a global adult population of ~5.5 billion (age 18+), 6.7 million deaths/year is ~1.22 per 1,000 adults/year. Compounded naively over 60 adult years: 1 - (1 - 1.22e-3)^60 ≈ 0.070. Adjusted modestly upward to 0.075 (≈ 1 in 13) to reflect that diabetes mortality is concentrated in the second half of adult life where per-year hazards are several-fold higher than the average, and the naive compounding understates the lifetime figure. Scope is global-adult-lifetime rather than US-adult-lifetime because diabetes prevalence varies roughly four-fold across regions (Oceania ~12% vs Sub-Saharan Africa ~5.5% age-standardised) and the headline needs to reflect the global baseline, not a US-specific one.\n","uncertainty":{"low":0.05,"high":0.12},"scope":"global_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/34879977/","title":"IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045","publisher":"International Diabetes Federation (Sun et al., Diabetes Research and Clinical Practice)","source_type":"peer_reviewed","statistic":"Global diabetes prevalence in 20-79 year olds in 2021 was 10.5% (536.6 million people), projected to rise to 12.2% (783.2 million) in 2045","excerpt":"\"The global diabetes prevalence in 20-79 year olds in 2021 was estimated to be 10.5% (536.6 million people), rising to 12.2% (783.2 million) in 2045. [...] Prevalence [in 2021] was estimated to be higher in urban (12.1%) than rural (8.3%) areas, and in high-income (11.1%) compared to low-income countries (5.5%). The greatest relative increase in the prevalence of diabetes between 2021 and 2045 is expected to occur in middle-income countries (21.1%).\"\n","source_date":"2022-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260410135932/https://pubmed.ncbi.nlm.nih.gov/34879977/","calculation_notes":"IDF’s 10th edition Atlas reports ~6.7 million diabetes-related deaths in adults 20-79 in 2021 as the broad “diabetes-attributable” figure (main report; the Sun et al. abstract itself covers prevalence not mortality). The 6.7M number is used as the numerator for the normalized rate: 6.7M / ~5.5B global adults ≈ 1.22 per 1,000 adults/year, compounded over 60 adult years → ~7%, adjusted to 7.5% for the age-concentration of diabetes mortality. IDF 536.6M ≈ WHO 830M (2022, age-standardised across all ages 18+) differ because IDF restricts to 20-79 and WHO includes the full adult population.\n","independence_note":"IDF Atlas estimates and GBD 2021 diabetes estimates share substantial upstream data (national surveys, registry data, verbal autopsy studies) but use different modelling pipelines and are published separately; treat as partially but not fully independent of the Lancet GBD source below.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/diabetes","title":"Diabetes — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"830 million adults living with diabetes in 2022; 14% of adults aged 18+; 1.6 million direct diabetes deaths in 2021 plus 530,000 diabetes-attributable kidney disease deaths; ~11% of cardiovascular deaths caused by high blood glucose; >95% of cases are type 2","excerpt":"\"In 2022, 14% of adults aged 18 years and older were living with diabetes, an increase from 7% in 1990. [...] More than 95% of people with diabetes have type 2 diabetes. [...] In 2021, diabetes and kidney disease due to diabetes caused over 2 million deaths. In addition, around 11% of cardiovascular deaths were caused by high blood glucose. [...] More than half of people living with diabetes did not take medication for their diabetes in 2022. Diabetes treatment was lowest in low- and middle-income countries.\"\n","source_date":"2024-11-14","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260407073645/https://www.who.int/news-room/fact-sheets/detail/diabetes","calculation_notes":"WHO’s narrower underlying-cause count (1.6M direct diabetes deaths + 0.53M diabetes-attributable kidney disease deaths = ~2.1M) plus ~11% of the ~19.8M annual global CVD deaths (≈ 2.2M) gives ~4.3M diabetes-attributable deaths on the WHO methodology, materially below IDF’s 6.7M. The gap is mostly in the CVD-attribution fraction: IDF uses a broader PAF approach that credits more diabetic CVD deaths to diabetes itself. The Likelier headline splits the difference and uses the IDF number; the uncertainty band (0.05-0.12) covers both methodologies. Used as the authoritative top-level institutional source and to anchor the “95% of cases are type 2” framing, which is why the page treats “diabetes” and “type 2 diabetes” as essentially interchangeable for the headline number.\n","independence_note":"WHO diabetes fact sheet draws on WHO Global Health Estimates, which in turn uses IHME GBD inputs; partially dependent on the Lancet GBD source but methodologically distinct in how it allocates CVD deaths to diabetes.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37356446/","title":"Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021","publisher":"GBD 2021 Diabetes Collaborators (The Lancet)","source_type":"peer_reviewed","statistic":"529 million people living with diabetes in 2021 (95% UI 500-564); type 2 diabetes accounts for 96.0% of cases and 95.4% of diabetes DALYs; global age-standardised prevalence 6.1%; highest regional rates in Oceania (12.3%) and North Africa/Middle East (9.3%); 1.31 billion projected by 2050","excerpt":"\"In 2021, there were 529 million (95% uncertainty interval [UI] 500-564) people living with diabetes worldwide. The global age-standardised total diabetes prevalence was 6·1% (5·8-6·5). [...] Type 2 diabetes [...] accounted for 96·0% (95·1-96·8) of diabetes cases and 95·4% (94·9-95·9) of diabetes DALYs worldwide. [...] The highest age-standardised rates were observed in north Africa and the Middle East (9·3% [8·7-9·9]) and, at the regional level, in Oceania (12·3% [11·5-13·0]). [...] 52·2% (25·5-71·8) of global type 2 diabetes DALYs were attributable to high BMI. [...] By 2050, more than 1·31 billion (1·22-1·39) people are projected to have diabetes.\"\n","source_date":"2023-07-15","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260316040654/https://pubmed.ncbi.nlm.nih.gov/37356446/","calculation_notes":"GBD 2021 gives a total-diabetes count (529M) about 1.5% below IDF (536.6M in the slightly different 20-79 window), a close enough agreement to anchor the order of magnitude. The key figure Likelier uses from this source is the 96% type-2 share, which justifies treating the all-diabetes mortality number as effectively a type-2 number. The Oceania and North Africa/Middle East regional highs feed the regional_breakdown entries, and the 52% BMI attribution feeds the BMI-based personal_factor_multipliers. Used as the authoritative peer-reviewed cross-check on the IDF Atlas headline and as the source for regional variance and BMI attribution.\n","independence_note":"GBD 2021 uses the IHME Cause of Death Ensemble model pipeline, which is methodologically distinct from IDF’s Atlas modelling but draws on overlapping upstream vital registration data; partially dependent.\n"},{"url":"https://www.cdc.gov/nchs/fastats/diabetes.htm","title":"Diabetes — FastStats","publisher":"US Centers for Disease Control and Prevention / National Center for Health Statistics","source_type":"govt_report","statistic":"94,445 US deaths with diabetes as underlying cause (2024); 27.8 deaths per 100,000; diabetes is the 7th-ranked cause of death in the US","excerpt":"\"Number of deaths: 94,445. Deaths per 100,000 population: 27.8. Cause of death rank: 7. [...] Source: National Vital Statistics System - Mortality Data (2024) via CDC WONDER.\"\n","source_date":"2026-02-20","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421200431/https://www.cdc.gov/nchs/fastats/diabetes.htm","calculation_notes":"~94,400 US underlying-cause deaths across ~260M US adults ≈ 0.36 per 1,000 adults/year. Compounded over 60 adult years: 1 - (1 - 3.6e-4)^60 ≈ 0.021, a US-only figure of ~2.1% on the strict underlying-cause definition. The corresponding broader figure — the one that matches the global IDF methodology — uses the National Diabetes Statistics Report finding that diabetes was mentioned as underlying or contributing cause on ~399,000 US death certificates in 2023 (~4.2x the underlying-cause count), implying a lifetime figure closer to 0.09 for US adults. That is what the US row in regional_breakdown reflects. Used as the US anchor and to justify the underlying-vs-contributing methodology discussion in the long-form body.\n","independence_note":"CDC FastStats draws from the NCHS NVSS death-certificate pipeline, which also feeds into WHO Global Health Estimates and IHME GBD upstream. Used here as the US-specific anchor rather than an independent verification of the global IDF/GBD figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37952131/","title":"Semaglutide and Cardiovascular Outcomes in Obesity without Diabetes (SELECT trial)","publisher":"Lincoff et al., New England Journal of Medicine","source_type":"peer_reviewed","statistic":"Semaglutide 2.4mg reduced major adverse cardiovascular events by ~20% (hazard ratio 0.80, 95% CI 0.72-0.90) vs placebo in 17,604 overweight/obese adults aged 45+ with established cardiovascular disease and without diabetes; MACE occurred in 6.5% vs 8.0%","excerpt":"\"A primary cardiovascular end-point event occurred in 569 of the 8803 patients (6.5%) in the semaglutide group and in 701 of the 8801 patients (8.0%) in the placebo group (hazard ratio, 0.80; 95% confidence interval, 0.72 to 0.90; P<0.001).\"\n","source_date":"2023-12-14","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260705021138/https://pubmed.ncbi.nlm.nih.gov/37952131/","calculation_notes":"SELECT did not estimate diabetes mortality; it is used here solely to source the semaglutide (GLP-1 receptor agonist) protective personal_factor_multiplier of 0.80. The hazard ratio of 0.80 for major adverse cardiovascular events maps directly to a multiplier of 0.80 against the cardiovascular component of diabetes-related death, which is the dominant component in this entry's broad IDF \"diabetes-attributable\" headline. The trial enrolled patients without diabetes, so the cardiovascular benefit is established beyond the diabetic population; in type 2 diabetes specifically the effect direction is consistent with earlier GLP-1 RA cardiovascular-outcome trials. No change is made to the headline lifetime figure on the basis of this source.\n","independence_note":"Industry-sponsored (Novo Nordisk) randomised controlled trial, independent of the IDF Atlas, WHO, GBD, and CDC mortality-surveillance sources above; used only as the basis for the GLP-1 RA protective multiplier, not as a cross-check on the headline.\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death from cancer (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.075,"notes":"IDF 6.7M diabetes-related deaths (adults 20-79) compounded over 60 adult years"},{"region":"US adult","probability":0.09,"notes":"CDC ~94K underlying-cause + ~300K contributing-cause diabetes deaths per year; ~12% of US adults have diabetes"},{"region":"Pacific Islands (highest prevalence)","probability":0.18,"notes":"Several Pacific Island nations have adult diabetes prevalence above 25% (Marshall Islands, Kiribati, Tuvalu, Nauru); GBD 2021 places Oceania regional prevalence at 12.3%, the highest in the world"},{"region":"North Africa / Middle East","probability":0.11,"notes":"GBD 2021 age-standardised prevalence 9.3%, second-highest global region; driven by urbanisation and BMI transition"},{"region":"South Asia","probability":0.11,"notes":"Elevated risk at lower BMI than European-ancestry populations; India and Pakistan together account for a large share of global diabetes cases"},{"region":"Sub-Saharan Africa","probability":0.04,"notes":"Lower prevalence and higher competing mortality (infectious disease, maternal) mask the underlying trend; diabetes mortality is rising fastest here"}],"personal_factor_multipliers":[{"factor":"BMI 35+","multiplier":5,"notes":"Class II+ obesity; GBD 2021 attributes ~52% of global type 2 diabetes DALYs to high BMI, and the relative risk for diabetes incidence at BMI 35+ vs normal BMI is in the 5-10x range across major cohorts"},{"factor":"BMI 30-35","multiplier":3,"notes":"Class I obesity; diabetes incidence and mortality scale strongly with BMI across the overweight range"},{"factor":"Prediabetes untreated","multiplier":4,"notes":"Progression from prediabetes to type 2 diabetes runs ~5-10% per year without intervention; lifestyle or metformin intervention roughly halves that rate"},{"factor":"Family history (first-degree relative)","multiplier":2.5,"notes":"Parent or sibling with type 2 diabetes roughly 2-3x baseline risk; reflects shared genetics and shared environment"},{"factor":"South Asian or Pacific Islander ancestry","multiplier":2,"notes":"Elevated risk at lower BMI than European-ancestry populations; the BMI threshold for clinical diabetes screening is lower in many South Asian countries for this reason"},{"factor":"Physical activity 150+ min/week","multiplier":0.6,"notes":"Meeting WHO physical activity guidelines is associated with roughly 30-40% lower type 2 diabetes incidence in large cohorts; the mortality multiplier is slightly smaller"},{"factor":"Semaglutide 2.4mg (GLP-1 RA) with established CVD","multiplier":0.8,"notes":"In the SELECT trial (Lincoff et al., NEJM 2023; 17,604 overweight/obese adults with established cardiovascular disease), semaglutide 2.4mg cut major adverse cardiovascular events by ~20% (hazard ratio 0.80, 95% CI 0.72-0.90; 6.5% vs 8.0%). Because diabetes mortality on this page is largely the cardiovascular death it drives, this is a directly relevant protective factor; note the trial enrolled patients without diabetes, so the cardiovascular benefit is established beyond the diabetic population"}],"short_label":"Type 2 diabetes","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The single largest uncertainty on this page is the definition of “diabetes-attributable death”. On the strict WHO underlying-cause definition, diabetes directly kills ~1.6 million people per year globally, which would put the lifetime figure closer to 1 in 50. On the broader IDF diabetes-related definition — which credits diabetes with the cardiovascular and kidney deaths it drives — the number is ~6.7 million per year and the lifetime figure is closer to 1 in 13. Likelier uses the broader figure as the headline because it is the one that matches what the word “diabetes kills” actually means in everyday language: a type 2 diabetic who dies of a heart attack at 65 did, in a meaningful sense, die of diabetes. The uncertainty band brackets both methodologies. Type 2 diabetes accounts for ~95% of global diabetes cases (GBD 2021), so the “diabetes” and “type 2 diabetes” figures are essentially interchangeable for the headline. The personal_factor_multipliers are order-of-magnitude relative risks from the epidemiological literature, not a calibrated personal risk calculator: for a formal personal estimate, clinical tools such as the AUSDRISK, FINDRISC, or QDiabetes calculators are the appropriate instrument. The semaglutide 2.4mg protective multiplier (0.80) is sourced from the SELECT trial, which enrolled overweight/obese adults with established cardiovascular disease but without diabetes; the multiplier reflects its ~20% cut in major adverse cardiovascular events (the dominant component of diabetes-attributable death here), not a measured reduction in diabetes-cause mortality, and should be read as a cardiovascular-protection signal rather than a precise diabetes-mortality adjustment.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single thin descending line crossing a horizontal threshold on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/type-2-diabetes-death","api_url":"https://likelier.app/api/fears/type-2-diabetes-death.json"},{"slug":"appendicitis-lifetime","question":"What are the odds of getting appendicitis in your lifetime?","quick_answer":"About 1 in 13 people get appendicitis over a lifetime, in the 7 to 8 percent range. Most adults already intuit it as fairly common, so perception here sits close to the evidence rather than wildly above or below it.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most people know appendicitis is common enough that they personally know someone who has had it, and their intuition reflects that. It registers as a plausible, non-exotic medical event rather than a remote tail risk. Few adults dramatically overestimate or underestimate it; the typical guess lands somewhere in the \"pretty common\" bucket without a sharp number attached. The fear, such as it is, centers less on incidence and more on the scenario: sudden pain, emergency surgery, perforation if you wait too long.\n","rough_estimate":"Most adults intuit appendicitis as 'fairly common' — roughly 1 in 10 to 1 in 20, which is close to the literature","kind":"intuition"},"native":{"display":"~7–8% lifetime risk in the US","numerator":77,"denominator":1000,"unit":"lifetime","population":"US residents, all ages"},"normalized":{"lifetime_us_adult":0.077,"display":"1 in ~13 lifetime (US adult)","log_value":-1.11,"assumptions":"The canonical US lifetime risk estimate comes from Addiss et al. (1990), which used NHDS data (1979–1984) to compute a lifetime risk of 8.6% for males and 6.7% for females, yielding a sex-averaged figure of approximately 7.6%. A 2018 American Family Physician review (Snyder et al.) reports more than 300,000 appendectomies performed annually in the US (<10% with a normal appendix removed) and reiterates the Addiss lifetime risks, all consistent with a lifetime incidence in the 7–8% range. Anderson et al. (2012) analyzing California discharge data independently reported a 9.0% lifetime cumulative incidence, confirming the Addiss range. Headline figure set at 0.077 (≈ 1 in 13), with an uncertainty band of 0.06–0.09 to reflect the sex difference, secular trends (appendicitis incidence has been roughly stable or slowly declining in high-income countries since the 1990s), and the fact that some mild cases resolve without diagnosis. The lifetime figure applies from birth; since peak incidence is ages 10–30 and most US adults have already passed through part of that window, residual lifetime risk for a 30-year-old is modestly lower than the headline.\n","uncertainty":{"low":0.06,"high":0.09},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/2239906/","title":"The Epidemiology of Appendicitis and Appendectomy in the United States","publisher":"American Journal of Epidemiology (Addiss DG, Shaffer N, Fowler BS, Tauxe RV)","source_type":"primary_study","statistic":"Lifetime risk of appendicitis: 8.6% for males and 6.7% for females in the US","excerpt":"\"The lifetime risk of appendicitis was 8.6% for males and 6.7% for females; the lifetime risk of appendectomy was 12.0% for males and 23.1% for females.\"\n","source_date":"1990-11-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421183155/https://pubmed.ncbi.nlm.nih.gov/2239906/","calculation_notes":"Addiss et al. analyzed National Hospital Discharge Survey (NHDS) data for 1979–1984, a period during which approximately 250,000 cases of appendicitis occurred annually in the US. Lifetime risk was computed from age-specific incidence rates using standard life-table methods. The sex- averaged lifetime appendicitis risk is approximately (8.6 + 6.7) / 2 ≈ 7.65%, rounded to 7.7% for the headline. This remains the most widely cited US lifetime figure and has been confirmed by subsequent analyses using HCUP and SEER data.\n"},{"url":"https://www.aafp.org/pubs/afp/issues/2018/0701/p25.html","title":"Acute Appendicitis: Efficient Diagnosis and Management","publisher":"American Family Physician (Snyder MJ, Guthrie M, Cagle S), Vol. 98, No. 1, pp. 25–33","source_type":"peer_reviewed","statistic":"More than 300,000 appendectomies performed annually in the US; lifetime risk 8.6% in males, 6.7% in females","excerpt":"\"More than 300,000 appendectomies are performed each year in the United States, and less than 10% result in the removal of a normal appendix.\"\n","source_date":"2018-07-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260630072833/https://www.aafp.org/afp/2018/0701/p25","calculation_notes":"The AAFP review reports more than 300,000 annual US appendectomies. Across ~330 million US residents, ~300,000/year is an annual rate of roughly 91 per 100,000 (≈ 1 in 1,100 per year). Over a 78-year life expectancy, naive compounding gives 1 - (1 - 0.00091)^78 ≈ 6.9%, which sits at the low end of the Addiss range because the annual rate is not age-flat (peak in the second and third decades pulls the lifetime figure back toward 7–8%) and because negative-appendectomy rates have declined with CT imaging (the review notes <10% normal-appendix removal). The review independently cites the Addiss lifetime risks of 8.6% (males) and 6.7% (females), corroborating the headline.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22948195/","title":"Examining a common disease with unknown etiology: trends in epidemiology and surgical management of appendicitis in California, 1995–2009","publisher":"World Journal of Surgery (Anderson JE, Bickler SW, Chang DC, Talamini MA)","source_type":"peer_reviewed","statistic":"Lifetime cumulative incidence rate of appendicitis is 9.0%; children age 10–14 had the highest rates (169.6 cases/100,000)","excerpt":"\"A total of 608,116 patients with appendicitis (70% non-perforated) were included. The incidence increased at an average rate of 0.5 cases/100,000 population/year (p<0.001), with annual incidence peaking during the third quarter. Children age 10-14 had the highest rates of appendicitis (169.6 cases/100,000). The lifetime cumulative incidence rate is 9.0%.\"\n","source_date":"2012-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250416051844/https://pubmed.ncbi.nlm.nih.gov/22948195/","calculation_notes":"Anderson et al. (2012) analyzed California Patient Discharge Data covering 608,116 appendicitis patients from 1995–2009. The 9.0% lifetime cumulative incidence figure independently confirms the Addiss et al. 7.6% estimate (which used 1979–1984 NHDS data), with the modest difference likely reflecting California-specific demographics and the secular upward trend in incidence documented in the paper (0.5 cases/100,000/year increase). The peak incidence in children age 10–14 (169.6/100,000) is consistent with the well-established age distribution of appendicitis.\n"}],"comparison_anchors":[{"label":"Cancer death (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Choking death (lifetime, US)","lifetime_us_adult":0.00095},{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.0000692}],"regional_breakdown":[{"region":"United States (sex-averaged)","probability":0.077,"notes":"Addiss et al. 1990; confirmed by HCUP data through 2014"},{"region":"United States (male)","probability":0.086,"notes":"Males have higher incidence across all age groups (Addiss et al.)"},{"region":"United States (female)","probability":0.067,"notes":"Lower appendicitis incidence but historically higher appendectomy rate due to negative appendectomies"},{"region":"Western Europe","probability":0.075,"notes":"Similar rates to the US based on GBD and national registry data"},{"region":"Low-income countries","probability":0.02,"notes":"Lower reported incidence, though under-diagnosis and limited surgical access confound the number"}],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":1.3,"notes":"Males have roughly 1.3× higher lifetime appendicitis risk than females (8.6% vs 6.7%, Addiss et al.)"},{"factor":"Age 10–30 (peak incidence window)","multiplier":2.5,"notes":"Annual incidence peaks sharply in the second and third decades of life; a 15-year-old has roughly 2–3× the per-year risk of a 50-year-old"},{"factor":"Family history of appendicitis","multiplier":1.5,"notes":"Some evidence of familial clustering, though no single gene identified; approximate multiplier based on case-control data"},{"factor":"Age >50","multiplier":0.4,"notes":"Incidence drops substantially after age 30; residual lifetime risk for a 50-year-old is much lower than the birth-to-death headline"}],"short_label":"Appendicitis","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline 7–8% lifetime risk is for appendicitis requiring clinical attention, not for all appendiceal inflammation (subclinical appendicitis that resolves spontaneously is difficult to quantify). The canonical Addiss et al. data are from 1979–1984; secular trends since then have been modest, with incidence roughly stable or slowly declining in high-income countries. Perforation rate is approximately 20–30%, with higher rates at the extremes of age: children under 5 and adults over 65 are more likely to perforate because of delayed diagnosis. Mortality in developed countries is very low for non-perforated appendicitis (~0.1%) but climbs to 1–5% for perforated cases in the elderly. Antibiotic-first management (without surgery) is an active area of research and may shift future appendectomy rates without necessarily changing appendicitis incidence.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simple curved shape suggesting an anatomical tube against a muted warm background, flat vector illustration."},"canonical_url":"https://likelier.app/appendicitis-lifetime","api_url":"https://likelier.app/api/fears/appendicitis-lifetime.json"},{"slug":"household-air-pollution-death","question":"What are the odds of dying prematurely from indoor cooking fire smoke?","quick_answer":"For the roughly 2.1 billion people who cook over open fires or solid-fuel stoves, the odds of dying prematurely from indoor cooking smoke are about 1 in 13 over a lifetime; for adults cooking with electricity or piped gas the risk is negligible. Globally it kills about 2.9 million people a year.\n","category":"health","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"In wealthy countries with electric or gas stoves, indoor air pollution from cooking barely registers as a health concern. The risk conjures images of a smoky campfire, not a leading global killer. Yet for roughly 2.1 billion people who cook over open fires or rudimentary stoves burning wood, charcoal, crop waste, or dung, household air pollution is an ambient, daily exposure that drives ischaemic heart disease, stroke, chronic obstructive pulmonary disease, lung cancer, and acute lower respiratory infections in children. Because the exposure is chronic and the diseases it causes are common, the mortality burden hides inside broader cardiovascular and respiratory statistics and rarely appears in news coverage as a distinct hazard.\n","kind":"intuition"},"native":{"display":"~2.9 million deaths per year globally from household air pollution","numerator":2900000,"denominator":5000000000,"unit":"per year","population":"global adults"},"normalized":{"lifetime_us_adult":0.0783,"display":"~1 in 13 lifetime (solid-fuel-using adult)","log_value":-1.11,"assumptions":"Native rate: WHO estimates 2.9 million deaths per year attributable to household air pollution (HAP) in 2021. The burden falls almost entirely on populations who cook with polluting fuels (wood, charcoal, crop waste, dung, kerosene, coal) — around 2.1 billion people worldwide (WHO fact sheet: \"Around 2.1 billion people worldwide ... cook using open fires or inefficient stoves\"). Dividing by the at-risk population: 2,900,000 / 2,100,000,000 = 1.381e-3 annual rate. Lifetime conversion using the 59-year horizon from age 18: 1 - (1 - 1.381e-3)^59 = 0.0783. The GBD 2021 systematic analysis in The Lancet estimated 3.11 million HAP-attributable deaths globally in 2021, broadly consistent with the WHO figure. Uncertainty reflects the range between 2.3 million and 3.2 million deaths against the same 2.1B at-risk denominator. Low bound: 2,300,000 / 2.1B compounded 59 years = 1 - (1 - 1.095e-3)^59 = 0.0626. High bound: 3,200,000 / 2.1B compounded 59 years = 1 - (1 - 1.524e-3)^59 = 0.0860. For any adult in a high-income country cooking with electricity or piped gas, personal risk from HAP is negligible.\n","uncertainty":{"low":0.063,"high":0.086},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/household-air-pollution-and-health","title":"Household air pollution — Fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Household air pollution was responsible for an estimated 2.9 million deaths per year in 2021, including over 309,000 deaths of children under the age of 5","excerpt":"\"Household air pollution was responsible for an estimated 2.9 million deaths per year in 2021, including over 309,000 deaths of children under the age of 5. Among these 2.9 million deaths: 32% are from ischaemic heart disease, 23% are from stroke, and 21% are due to lower respiratory infection.\"\n","source_date":"2023-12-15","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426202011/https://www.who.int/news-room/fact-sheets/detail/household-air-pollution-and-health","calculation_notes":"The WHO 2.9 million annual deaths figure is the primary source for the native numerator. 2,900,000 / 5,000,000,000 global adult population = 0.00058 annual rate. Compounded over 59 years: 1 - (1 - 0.00058)^59 = 0.0337. Disease breakdown (32% IHD, 23% stroke, 21% LRI) confirms that HAP mortality is distributed across several disease categories rather than appearing as a single cause of death in vital statistics, which partly explains its low public visibility.\n"},{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(24)02840-X/fulltext","title":"Global, regional, and national burden of household air pollution, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"3.11 million deaths attributable to household air pollution globally in 2021; 3.25 million deaths and 123 million DALYs in 2017","excerpt":"\"An estimated 3.11 million deaths were attributable to household air pollution globally in 2021. 3.25 million deaths and 123 million disability-adjusted life-years (DALYs) were attributable to household air pollution in 2017.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20250713042410/https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(24)02840-X/fulltext","calculation_notes":"The GBD 2021 peer-reviewed estimate of 3.11 million deaths is broadly consistent with the WHO 2.9 million figure and provides the upper end of the plausible range. The 2017 peak of 3.25 million is used to anchor the high bound of the uncertainty interval. Validates the native rate and confirms household air pollution as one of the largest single environmental risk factors for premature mortality globally.\n"}],"comparison_anchors":[{"label":"Death from lung cancer (lifetime, US)","lifetime_us_adult":0.056},{"label":"Death from rabies via dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death in a plane crash (lifetime, US)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Polluting-fuel-using adults (~2.1 billion)","probability":0.0783,"notes":"WHO estimates ~2.1B people cook with polluting fuels; nearly all HAP deaths occur in this group"},{"region":"Global average (all adults)","probability":0.0337,"notes":"Diluted across 5B adults; misleading because risk is concentrated in polluting-fuel users"},{"region":"High-income countries (electric/gas cooking)","probability":0.00001,"notes":"Effectively zero; regulated fuel supply eliminates exposure"}],"short_label":"Indoor cooking smoke","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The 1-in-30 global lifetime figure is driven almost entirely by populations in sub-Saharan Africa and South and Southeast Asia who rely on solid fuels burned in poorly ventilated indoor spaces. For any adult in a country where cooking is done on electric, induction, or piped-gas stoves, personal exposure to household air pollution is negligible and this probability does not apply. Women and children bear a disproportionate burden because they are more likely to be responsible for cooking and spend more time near the fire. The WHO notes that around 2.1 billion people still cook using open fires or rudimentary stoves, and the transition to clean cooking fuels is proceeding slowly, with progress concentrated in China and parts of Southeast Asia rather than in sub-Saharan Africa where the burden is heaviest.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":3,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a simple indoor cooking fire with wisps of smoke rising in a dim room, rendered in muted earth tones."},"canonical_url":"https://likelier.app/household-air-pollution-death","api_url":"https://likelier.app/api/fears/household-air-pollution-death.json"},{"slug":"gestational-diabetes","question":"What are the odds of developing gestational diabetes during pregnancy?","quick_answer":"Gestational diabetes turns up in about 1 in 13 US pregnancies (7.9% in 2024), a rate that has climbed steadily. Most women know the screening exists but treat a positive result as a surprise rather than the roughly 1-in-13 base-rate event it is.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Gestational diabetes mellitus (GDM) occupies an awkward middle ground in public awareness: most pregnant women have heard of it because it is part of routine prenatal screening, but few can cite a prevalence figure. The condition tends to be perceived as uncommon — something that happens to \"other people\" — even as rates have climbed steadily over the past decade. The glucose tolerance test at 24-28 weeks is familiar to virtually every woman who has been pregnant in the US, yet the possibility of a positive result is typically treated as a surprise rather than a roughly 1-in-13 base-rate event.\n","rough_estimate":"Most pregnant women know GDM screening exists but underestimate how common a positive result is","kind":"intuition"},"native":{"display":"~79 per 1,000 births in the US (2024)","numerator":79,"denominator":1000,"unit":"per pregnancy","population":"US singleton first pregnancies, 2024"},"normalized":{"lifetime_us_adult":0.079,"display":"~1 in 13 per pregnancy (US, 2024)","log_value":-1.1,"assumptions":"A JAMA Internal Medicine study (Lam et al., 2025) analysing every US birth certificate from 2016-2024 found gestational diabetes prevalence rose from 58 to 79 per 1,000 births over that period — a 36% increase. The CDC has historically cited 2-10% of US pregnancies. The 2024 figure of 7.9% is used as the point estimate. Because this is a per-pregnancy risk (not a lifetime cumulative risk across all pregnancies a woman may have), the normalized figure represents the probability per single pregnancy event. For a woman who has two pregnancies, her lifetime probability of experiencing GDM at least once is higher — roughly 1-(1-0.079)^2 ≈ 15%. The scope is set to activity_specific_lifetime because the risk is per-pregnancy, not a cumulative lifetime figure. Uncertainty band spans from the lower historical CDC range (~0.05) to the higher rates seen in some racial/ethnic subgroups (~0.14).\n","uncertainty":{"low":0.05,"high":0.14},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://jamanetwork.com/journals/jamainternalmedicine/article-abstract/2842943","title":"Gestational Diabetes in the US From 2016 to 2024","publisher":"JAMA Internal Medicine","source_type":"peer_reviewed","statistic":"GDM prevalence rose from 58 to 79 per 1,000 births (2016-2024), a 36% increase","excerpt":"\"Gestational diabetes shot up 36% over the nine-year period from 2016 to 2024, increasing from 58 to 79 cases per 1,000 births. [...] The condition increased across every racial and ethnic group.\"\n","source_date":"2025-12-30","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260525161611/https://jamanetwork.com/journals/jamainternalmedicine/article-abstract/2842943","calculation_notes":"Lam et al. analysed all US birth certificates for first singleton pregnancies from 2016 to 2024 using National Center for Health Statistics data. The 79 per 1,000 figure for 2024 (7.9%) is the most current national estimate available. This is a per-pregnancy prevalence, not a lifetime cumulative risk. The 36% increase over nine years reflects both real epidemiological change (rising obesity, older maternal age) and diagnostic changes (wider adoption of the IADPSG criteria in some centres).\n","independence_note":"This JAMA study uses National Vital Statistics System birth certificate data, which is methodologically independent of the CDC's MMWR reports that use the same upstream NVSS data but present it in different time slices and formats.\n"},{"url":"https://www.cdc.gov/diabetes/about/gestational-diabetes.html","title":"Gestational Diabetes","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Every year, 5% to 9% of US pregnancies are affected by gestational diabetes","excerpt":"\"Every year, 5% to 9% of U.S. pregnancies are affected by gestational diabetes. Managing gestational diabetes can help make sure you have a healthy pregnancy and a healthy baby.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260416043828/https://www.cdc.gov/diabetes/about/gestational-diabetes.html","calculation_notes":"CDC gives the 5-9% range, which encompasses variation across years, diagnostic criteria, and populations. The JAMA study's 2024 figure of 7.9% sits within this range, consistent with the secular upward trend. CDC's page is a general-audience resource and does not provide year-specific breakdowns; the JAMA study fills that gap.\n","independence_note":"CDC's general diabetes page synthesises multiple data sources. The underlying birth-certificate data overlaps with the JAMA study's NVSS source, but CDC presents aggregate ranges rather than year-specific trend data.\n"},{"url":"https://www.childstats.gov/americaschildren/diabetes.asp","title":"America's Children: Key National Indicators of Well-Being — Gestational Diabetes","publisher":"Federal Interagency Forum on Child and Family Statistics (NCHS/NVSS data)","source_type":"govt_report","statistic":"From 2016 to 2022, the rate of gestational diabetes increased from 60 per 1,000 live births to 81 per 1,000; among women aged 40 and over, the 2022 rate was 151 per 1,000","excerpt":"\"From 2016 to 2022, the rate of gestational diabetes increased from 60 per 1,000 live births to 81 per 1,000. The rate of gestational diabetes in women age 40 and over was 151 per 1,000.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20251223015738/https://www.childstats.gov/americaschildren/diabetes.asp","calculation_notes":"This Federal Interagency Forum report draws on the same NCHS National Vital Statistics System birth-certificate data as the JAMA study. The 2022 figure (81 per 1,000 all live births) is consistent with the JAMA study's 2024 figure (79 per 1,000 singleton first pregnancies), with the slight difference explained by the different population denominators (all births vs. singleton first pregnancies). Both series confirm the secular upward trend and the ~8% order-of- magnitude estimate for current GDM prevalence. Used as the independent NCHS data anchor corroborating the JAMA study's trend and magnitude findings.\n","independence_note":"The Federal Interagency Forum on Child and Family Statistics is a separate agency from CDC that independently compiles and reports NCHS birth data. While it draws on the same NVSS source data, it is a methodologically and editorially independent reporting entity.\n"}],"comparison_anchors":[{"label":"Preeclampsia (per pregnancy, US)","lifetime_us_adult":0.05},{"label":"Miscarriage (per known pregnancy)","lifetime_us_adult":0.15},{"label":"Type 2 diabetes (lifetime, US adults)","lifetime_us_adult":0.4},{"label":"Cesarean delivery (per birth, US)","lifetime_us_adult":0.32}],"regional_breakdown":[{"region":"US overall (2024)","probability":0.079,"notes":"JAMA Internal Medicine, 79 per 1,000 births"},{"region":"American Indian/Alaska Native women (US, 2024)","probability":0.137,"notes":"Highest racial/ethnic group, 137 per 1,000 births"},{"region":"Asian American women (US, 2024)","probability":0.131,"notes":"131 per 1,000 births; second-highest group"},{"region":"Non-Hispanic White women (US, 2024)","probability":0.068,"notes":"Below national average"}],"personal_factor_multipliers":[{"factor":"BMI ≥ 30 (obese)","multiplier":2.5,"notes":"Obesity is the strongest modifiable risk factor for GDM"},{"factor":"Age ≥ 35","multiplier":2,"notes":"GDM rates increase sharply with maternal age; the CDC notes highest prevalence in women aged 35-44"},{"factor":"Prior GDM in a previous pregnancy","multiplier":4,"notes":"Recurrence risk is roughly 30-50% in subsequent pregnancies"},{"factor":"Family history of type 2 diabetes","multiplier":1.5,"notes":"First-degree family history of T2DM elevates GDM risk modestly"},{"factor":"Age < 25, normal BMI, no family history","multiplier":0.3,"notes":"Low-risk profile; GDM is uncommon in young, lean women without family history"}],"short_label":"Gestational diabetes","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry uses a per-pregnancy prevalence, not a lifetime cumulative probability. A woman who has multiple pregnancies faces a higher cumulative lifetime probability of experiencing GDM at least once. The 7.9% figure from the JAMA study is based on birth-certificate data, which captures diagnosed GDM only; undiagnosed cases are missed, so the true prevalence may be modestly higher. The secular upward trend is real — GDM prevalence has risen every year from 2016 to 2024 — and is driven by rising obesity rates, older maternal age, and in part by changing diagnostic thresholds. GDM usually resolves after delivery, but it is a strong predictor of future type 2 diabetes: roughly 50% of women with GDM develop T2DM within 5-10 years postpartum.\n","quality_score":{"d1":2,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A soft circular shape bisected by a thin curved line on a muted warm-beige background, flat vector illustration."},"canonical_url":"https://likelier.app/gestational-diabetes","api_url":"https://likelier.app/api/fears/gestational-diabetes.json"},{"slug":"adult-social-media-problematic-use","question":"What are the odds of problematic social-media use as an adult?","quick_answer":"Roughly 1 in 12 adults over a lifetime under strict clinical criteria (about 5% at any one time). Common guesses of 5 to 15% are about right, which makes this one of the rarer fears where public intuition is calibrated rather than badly off.\n","category":"tech","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Problematic social media use occupies an unusual position in the public risk landscape: it is simultaneously over-discussed in the media and under-estimated in terms of clinical prevalence. Parents, educators, and policymakers focus heavily on adolescent risk, often underweighting the proportion of adults who also meet problematic-use thresholds on validated scales. At the same time, the absence of a DSM-5 or ICD-11 diagnosis for social media addiction (in contrast to gaming disorder, which entered ICD-11 as 6C51 in 2022) creates widespread uncertainty about whether the phenomenon is real, exaggerated, or simply heavy use mislabeled as disorder. Popular discourse alternates between treating social media as mildly habit-forming and framing it as an existential crisis for mental health — both positions overshoot what the epidemiological data actually show.\n","rough_estimate":"~5-15% of adults","kind":"intuition"},"native":{"display":"~5% of adults meet strict problematic social media use criteria on validated scales (pooled across representative studies; monothetic/strict classification)","numerator":5,"denominator":100,"unit":"share of adults scoring above strict-threshold problematic use criteria on the Bergen Social Media Addiction Scale or equivalent validated instrument","population":"adults across multiple countries (meta-analytic pooled estimate using strict/monothetic cut-off classifications)"},"normalized":{"lifetime_us_adult":0.08,"display":"~1 in 12 adults is estimated to meet problematic social media use criteria at some point in a lifetime","log_value":-1.1,"assumptions":"The pooled prevalence of problematic social media use (PSU) using strict monothetic or severe cut-off criteria on validated scales (primarily the Bergen Social Media Addiction Scale, BSMAS) is approximately 5% (95% CI: 3%–7%) in representative adult samples, based on meta-analytic synthesis. Using moderate cut-off or polythetic criteria raises this to approximately 13-25%. We use the 5% strict-criteria figure as the native rate. For lifetime normalization, we apply a modest upward adjustment from the point-prevalence 5%: problematic social media use patterns are dynamic — individuals cycle in and out of problematic use over a lifetime, particularly as platforms and life circumstances change — so a larger share of adults will meet criteria at some point across a lifetime than at any single measurement. A lifetime_us_adult of 0.08 (8%) reflects a conservative 1.6x multiplier on the point prevalence, acknowledging that the cumulative lifetime fraction exceeds the cross-sectional rate. The uncertainty range (0.04–0.15) spans from a strict-criteria lower bound close to the point-prevalence floor to the moderate-criteria upper bound, given the substantial instrument-dependence of the estimate.\n","uncertainty":{"low":0.04,"high":0.15},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.sciencedirect.com/science/article/abs/pii/S0306460323002332","title":"Has the prevalence of problematic social media use increased over the past seven years and since the start of the COVID-19 pandemic? A meta-analysis of the studies published since the development of the Bergen social media addiction scale","publisher":"Drug and Alcohol Dependence / ScienceDirect","source_type":"peer_reviewed","statistic":"Pooled prevalence of problematic social media use: ~5% (95% CI 3%–7%) using strict/monothetic classifications; 13% using severe cut-off; 25% using moderate cut-off; 139 independent samples, 32 countries, n=133,955","excerpt":"\"The pooled prevalence estimate was 5% (95% CI: 3%–7%) for studies adopting monothetic or strict monothetic classifications, with a higher pooled prevalence estimate (13%; 95% CI: 8%–19%) found for studies adopting a cutoff for severe level or strict polythetic classifications, and 25% (95% CI: 21%–29%) for studies adopting a cutoff for moderate level or polythetic classifications. PSMU as assessed by the BSMAS was significantly higher in low-income countries.\"\n","source_date":"2023-08-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20241212020501/https://www.sciencedirect.com/science/article/abs/pii/S0306460323002332","calculation_notes":"Primary prevalence source. The 5% strict-criteria figure is used as the native rate (numerator=5, denominator=100). The 139-sample meta-analysis (n=133,955) spanning 32 countries provides the most comprehensive synthesis of BSMAS-based PSU prevalence to date. For normalization to lifetime_us_adult=0.08, we apply a conservative upward adjustment from the cross-sectional 5% to account for the dynamic, cyclical nature of PSU over a lifetime (individuals enter and leave problematic use states). The 95% CI range from strict (3%–7%) to moderate (21%–29%) criteria bounds the uncertainty range; we use 4%–15% as the uncertainty bounds to reflect realistic variability in strict-to-moderate definitions for a US adult context.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9758518/","title":"Psychometric properties of the Bergen Social Media Addiction Scale: An analysis using item response theory","publisher":"Journal of Behavioral Addictions / PMC","source_type":"peer_reviewed","statistic":"BSMAS-assessed problematic social media use prevalence ranges widely (3.5%–36.9%), with lower rates in Western countries (1.5%–15%) than in Asia (31%) and the Middle East (29%); an item-response-theory analysis proposes a conditional diagnostic cut-off of ≥26 (of 30) for high PSMU risk","excerpt":"\"High disparities in PSMU prevalence have been reported with rates ranging from 3.5% to 36.9%. This difference could be explained by seemingly lower prevalence rates in Western countries (1.5%-15%) compared to those found in Asia (31%) and the Middle East (29%). [...] prior to clinical assessment confirmation, a score of ≥26 could be taken as a conditional diagnostic cut-off point for risk of PSMU.\"\n","source_date":"2022-12-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505045414/https://pmc.ncbi.nlm.nih.gov/articles/PMC9758518/","calculation_notes":"Supporting source establishing the psychometric properties of the BSMAS, the primary instrument underlying the meta-analytic estimates in the primary source. This item-response- theory analysis documents the extreme cross-study disparity in PSMU prevalence (3.5%–36.9%), with markedly lower rates in Western countries (1.5%–15%) than in Asia (31%) or the Middle East (29%) — context for why the strict-criteria 5% native figure is conservative for a US adult population. It also proposes a high-risk cut-off of ≥26 (only 1.1% of its sample scored at or above this threshold). This source establishes that BSMAS is not a clinical diagnostic instrument — it measures scale-positive problematic use, not a recognized DSM-5 or ICD-11 disorder.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33550200/","title":"Prevalence of social media addiction across 32 nations: Meta-analysis with subgroup analysis of classification schemes and cultural values","publisher":"Drug and Alcohol Dependence / PubMed","source_type":"peer_reviewed","statistic":"Pooled social media addiction prevalence varies by classification scheme: 5% (95% CI 3%–7%) strict/monothetic, 13% (8%–19%) severe, 25% (21%–29%) moderate/polythetic; 14% (9%–19%) in individualist nations vs 31% (26%–36%) in collectivist nations; 63 independent samples, 34,798 respondents, 32 nations","excerpt":"\"The meta-analysis involved 63 independent samples with 34,798 respondents from 32 nations spanning seven world regions. [...] the pooled prevalence estimate obtained in collectivist nations (31%; 95% CI: 26%-36%) to be twofold higher than that obtained in individualist nations (14%; 95% CI: 9%-19%).\"\n","source_date":"2021-04-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505045453/https://pubmed.ncbi.nlm.nih.gov/33550200/","calculation_notes":"Earlier meta-analysis (Cheng et al. 2021; 63 independent samples, 34,798 respondents, 32 nations) reporting prevalence by classification scheme: 5% (strict/monothetic), 13% (severe), and 25% (moderate/polythetic). The 14% figure for individualist nations (which better approximates the US context) provides an upper-bound anchor, and the 5% strict-monothetic figure corroborates the strict-criteria native rate. The roughly fivefold spread across classification schemes illustrates the extreme instrument-dependence of PSU estimates. This source is used to contextualize the strict-criteria 5% native figure within the broader evidence base — demonstrating that the estimate is highly sensitive to measurement choice.\n"}],"comparison_anchors":[{"label":"Gambling disorder (lifetime, US)","lifetime_us_adult":0.025},{"label":"Compulsive buying disorder (global adults)","lifetime_us_adult":0.049},{"label":"Internet gaming disorder (ICD-11 6C51, global)","lifetime_us_adult":0.03}],"personal_factor_multipliers":[{"factor":"age 18-34","multiplier":2,"notes":"Younger adults show consistently higher BSMAS scores and PSU prevalence across most studies; the gap narrows in older age groups"},{"factor":"history of depression or anxiety","multiplier":2.5,"notes":"Depression and anxiety are strongly associated with PSU bidirectionally; social media use can both trigger and be used to cope with mood symptoms"},{"factor":"heavy daily social media use (>4 hours/day)","multiplier":3,"notes":"Frequency and duration of use are among the strongest behavioral predictors of scale-positive PSU; passive scrolling correlates more strongly than active interaction"}],"short_label":"Social media problematic use","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"Problematic social media use (PSU) is not listed in DSM-5 and does not appear in ICD-11 as of 2026. Internet Gaming Disorder entered ICD-11 (6C51) in 2022; social media use did not receive analogous recognition, reflecting ongoing scientific debate about whether the evidence base meets the threshold for a formal disorder category. All prevalence estimates here are based on validated scale scores (primarily BSMAS), not clinical diagnoses. The prevalence estimate is extremely sensitive to the cut-off or classification scheme used: strict monothetic criteria yield approximately 5%, while moderate polythetic criteria yield approximately 25% in the same underlying data. The lifetime_us_adult figure (0.08) involves a cross-sectional-to-lifetime extrapolation for which no longitudinal data currently exist. BSMAS studies are predominantly from non-US populations and from younger adult samples; US-specific representative adult data are limited. The concept of \"social media addiction\" remains contested — some researchers argue that high use reflects platform design incentives rather than individual pathology, and that addiction framing may stigmatize ordinary behavior.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"Abstract illustration of a phone with stacked notification badges, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/adult-social-media-problematic-use","api_url":"https://likelier.app/api/fears/adult-social-media-problematic-use.json"},{"slug":"day-trading-financial-ruin","question":"What are the odds of significant financial loss from active retail trading?","quick_answer":"Among persistent day-traders, about 97% lose money over five years, far worse than the roughly 50% many assume. The site's headline of about 1 in 13 US adults facing significant losses reflects that only a small share trade actively enough to be exposed to that near-certain attrition.\n","category":"other","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Active retail trading is widely perceived as a skill-based pursuit where disciplined, well-researched individuals can outperform passive investing. Financial media, trading platforms, and social communities amplify the visible winners while losses remain private, creating a survivorship-bias environment in which beating the market feels more achievable than the evidence warrants. The democratization of commission-free trading since 2019 and the rise of zero-days-to-expiry options have lowered the barrier to entry while raising the stakes, drawing in a new generation of retail traders who may underestimate how thoroughly institutional algorithms and market-makers harvest the other side of their trades.\n","rough_estimate":"~50% lose money","kind":"intuition"},"native":{"display":"97% of persistent day-traders (active >300 days) lose money over a 5-year period (Brazil CVM cohort, Chague et al. 2020)","numerator":97,"denominator":100,"unit":"share of individual day-traders active >300 days who lost money (5-year Brazilian futures cohort)","population":"all individuals who day-traded mini-Ibovespa futures in Brazil, 2013–2015 cohort, followed through 2017"},"normalized":{"lifetime_us_adult":0.08,"display":"~1 in 13 US adults who try active retail trading experience significant financial loss over their trading career","log_value":-1.097,"assumptions":"Two-factor estimate: (A) P(US adult becomes active enough to face meaningful day-trader risk) and (B) P(significant financial loss | active). Factor A: approximately 10% of US adults attempt some form of active retail trading at some point in their lives, based on FINRA and brokerage data on account activity; the 2019–2021 retail-trading surge brought active retail participation to roughly 20% of brokerage-account holders, but a large share trade only occasionally rather than systematically. We use 10% as a conservative lifetime estimate for adults who trade actively enough to face the losses documented in the cohort studies. Factor B: The Chague, De-Losso, and Giovannetti (2020) Brazil CVM cohort found 97% of persistent day-traders (>300 days active) lost money over five years. For the broader population of active but not necessarily persistent traders, Barber, Lee, Liu, and Odean (Taiwan, 2004; RFS 2009) found more than 80% of individual day traders lost money in any given year. We use 80% as the loss rate across all active traders (persistent and non-persistent combined). Combined: 0.10 × 0.80 = 0.080. Uncertainty range: 0.04 (5% activity × 80%) to 0.15 (15% activity × 97% persistent-trader loss rate). \"Significant financial loss\" is defined as net losses exceeding one month's household income over the trading career — a threshold consistent with the Brazilian and Taiwanese cohort data.\n","uncertainty":{"low":0.04,"high":0.15},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3423101","title":"Day Trading for a Living?","publisher":"SSRN / Fundação Getulio Vargas (FGV)","source_type":"peer_reviewed","statistic":"97% of persistent day-traders (active >300 days) lost money over the 5-year Brazilian CVM cohort; only 1.1% earned more than the Brazilian minimum wage","excerpt":"\"We show that it is virtually impossible for individuals to day trade for a living, contrary to what brokerage specialists and course providers often claim. We observe all individuals who began to day trade between 2013 and 2015 in the equity futures market in Brazil and persisted for at least 300 days. 97% of all individuals who persisted for more than 300 days lost money. Only 1.1% earned more than the Brazilian minimum wage and only 0.5% earned more than a bank teller.\"\n","source_date":"2020-06-11","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260520204841/https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3423101","calculation_notes":"This study provides the native numerator directly: 97 out of 100 persistent day-traders lost money. \"Persistent\" means active >300 days — a selection criterion that filters out casual dabblers and captures those who seriously attempt day-trading as a strategy. Because this cohort is more committed than the average retail trader, the 97% figure represents an upper bound on loss rates; the 80% figure from the broader Taiwan cohort (Barber et al.) is used for the combined lifetime estimate. The Brazilian data covers equity index futures (mini-Ibovespa), a market structurally similar to US retail futures trading.\n"},{"url":"http://www.econ.yale.edu/~shiller/behfin/2004-04-10/barber-lee-liu-odean.pdf","title":"Do Individual Day Traders Make Money? Evidence from Taiwan","publisher":"Yale University / NBER (Barber, Lee, Liu, Odean)","source_type":"peer_reviewed","statistic":"Less than 1% of the day-trader population predictably and reliably earned positive abnormal returns net of fees; more than 8 in 10 day traders lost money","excerpt":"\"Using the complete transaction records of all traders in the Taiwan stock market, we show that day trading is extremely hazardous to your wealth. The vast majority of day traders lose money. Less than 1% of the day trader population, those with the very best performance, are able to predictably and reliably earn positive abnormal returns net of fees.\"\n","source_date":"2004-04-10","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260301060616/http://www.econ.yale.edu/~shiller/behfin/2004-04-10/barber-lee-liu-odean.pdf","calculation_notes":"The Taiwan cohort provides a broader-population complement to the Brazil CVM data: the Taiwan study covers all retail day traders, not just persistent ones, so it captures the full distribution including short-lived participants. The >80% loss rate across this full sample is used as Factor B in the normalized estimate for US adults who trade actively but not necessarily persistently.\n","independence_note":"The Taiwan study uses complete exchange-level transaction records from the Taiwan Stock Exchange Surveillance System, entirely independent of the Brazilian CVM data used by Chague et al. Both datasets converge on high loss rates, strengthening the cross-market inference.\n"},{"url":"https://www.sec.gov/about/reports-publications/investorpubsdaytipshtm","title":"Day Trading: Your Dollars at Risk","publisher":"U.S. Securities and Exchange Commission (SEC)","source_type":"govt_report","statistic":"Day traders typically suffer severe financial losses in their first months of trading, and many never graduate to profit-making status","excerpt":"\"Most individual investors do not have the wealth, the time, or the temperament to make money and to sustain the devastating losses that day trading can bring.\" … \"Day traders typically suffer severe financial losses in their first months of trading, and many never graduate to profit-making status. Given these outcomes, it's clear: day traders should only risk money they can afford to lose.\"\n","source_date":"2005-04-19","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20251225031142/https://www.sec.gov/about/reports-publications/investorpubsdaytipshtm","calculation_notes":"The SEC investor publication \"Day Trading: Your Dollars at Risk\" corroborates the academic cohort studies with US-market context. The SEC is the federal regulator of US securities markets. This source does not provide a quantitative loss rate but confirms the directional finding from Barber et al. and Chague et al. is recognized by the principal US securities regulator.\n"},{"url":"https://onlinelibrary.wiley.com/doi/full/10.1111/jofi.13285","title":"Retail Trading in Options and the Rise of the Big Three Wholesalers","publisher":"The Journal of Finance (Bryzgalova, Pavlova & Sikorskaya, 2023)","source_type":"peer_reviewed","statistic":"Assuming a 10-day holding horizon, the aggregate portfolio of US retail options traders lost approximately $2.1 billion between November 2019 and June 2021 (19 months), concentrated in cheap, short-dated weekly options with an average 12.6% bid-ask spread","excerpt":"[Paraphrase from abstract — full text paywalled]: the study estimates that, assuming a 10-day holding horizon, the aggregate portfolio of US retail options traders lost approximately $2.1 billion between November 2019 and June 2021, concentrated among traders who favor cheap, short-dated weekly options with wide bid-ask spreads (averaging 12.6%), executed largely through payment-for-order-flow wholesalers.\n","source_date":"2023-12-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250425053647/https://onlinelibrary.wiley.com/doi/full/10.1111/jofi.13285","calculation_notes":"Grounds the body prose's \"$2.1 billion in aggregate in just 19 months\" figure and the \"trading 0DTE or weekly options\" personal_factor_multipliers entry below, both of which previously named this study's authors without citing the underlying loss figure. Full text sits behind the Journal of Finance paywall; the $2.1B/19-month figure is corroborated by the publicly available SSRN/CEPR working-paper abstracts (papers.ssrn.com/sol3/papers.cfm?abstract_id=4065019) and secondary reporting on the study. Not used in this entry's headline loss-rate arithmetic, which relies on the Brazil CVM and Taiwan cohort studies above.\n","independence_note":"Independent US options-market microstructure dataset, unrelated to the Brazilian futures cohort (Chague et al.) or the Taiwan equities cohort (Barber et al.) used for the entry's headline loss-rate figures.\n"}],"comparison_anchors":[{"label":"Gambling disorder financial ruin (lifetime, US)","lifetime_us_adult":0.0063},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Cryptocurrency total loss","lifetime_us_adult":0.04}],"personal_factor_multipliers":[{"factor":"trading with margin or leverage","multiplier":15,"notes":"Margin amplifies both gains and losses; retail margin traders face the additional risk of margin calls that can wipe accounts faster than losses can be absorbed"},{"factor":"trading 0DTE or weekly options","multiplier":7,"notes":"Short-dated options decay rapidly; retail traders systematically overpay relative to realized volatility (Bryzgalova et al. 2023, JoF)"},{"factor":"daily trading frequency","multiplier":4,"notes":"Transaction costs compound with frequency; the Brazil CVM cohort found no evidence of learning-by-doing even among the most active traders"},{"factor":"passive index investor (no active trading)","multiplier":0.05,"notes":"Long-term passive investing in diversified index funds avoids the structural disadvantages of retail day-trading entirely"}],"short_label":"Day-trading losses","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 8% lifetime estimate combines a Brazilian futures-market cohort with a Taiwanese equity-market cohort and extrapolates to US adults via estimated participation rates — none of these three populations are identical. The Brazil and Taiwan findings cover regulated exchange-traded instruments; US retail options and crypto day-trading may produce different (likely worse) outcomes given structural differences in market-maker advantages. \"Significant financial loss\" is a threshold concept, not a clinical diagnosis; the estimate covers net losses exceeding one month's income, not total financial ruin. The entry is distinct from cryptocurrency-total-loss (which covers speculative holding) and stock-market-crash (which covers systemic events affecting passive investors). Retail day-trading has changed substantially since the commission-free era began in 2019; the newer environment features tighter spreads but also more complex products (0DTE options, leveraged ETFs) that may shift the loss distribution. This entry covers active trading strategies, not passive long-term investing.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A stock chart with a sharp downward trend, muted slate and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/day-trading-financial-ruin","api_url":"https://likelier.app/api/fears/day-trading-financial-ruin.json"},{"slug":"extreme-government-catastrophe","question":"What are the odds of a political extremist government catastrophically ruining your country?","quick_answer":"For a resident of an established democracy, the odds of catastrophic regime failure run roughly 2 to 3% per decade, or about 8% over a lifetime. Nearly 40% of US adults report fearing a violent overthrow of the government and about half fear widespread civil unrest.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Fear of political catastrophe is among the most emotionally charged anxieties in democratic societies, and it runs in both directions: partisans on each side of the political spectrum fear the other's ascendance with roughly equal intensity. Pew has documented since 2016 that majorities of both US Democrats and Republicans view the opposing party as a threat to the nation's well-being. The fear is amplified by social media, which rewards catastrophist framing. In practice, the perception of imminent democratic collapse consistently outpaces the observed rate of actual regime failure in established democracies, which is why this entry carries an overrated framing -- specifically for countries with strong institutional buffers. For newer democracies and hybrid regimes, the risk is substantially higher.\n","rough_estimate":"39.8% of US adults report being afraid or very afraid of a violent overthrow of the US government; 49.2% fear widespread civil unrest (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":"~2-3% probability of democratic breakdown per decade in established democracies","numerator":25,"denominator":1000,"unit":"per decade","population":"established democracies (Polity score 8+ for 20+ years)"},"normalized":{"lifetime_us_adult":0.08,"display":"~8% lifetime probability of catastrophic regime failure (established democracy resident)","log_value":-1.1,"assumptions":"V-Dem's 2025 Democracy Report identifies 96 episodes of autocratization in 64 democratic countries from 1900 to 2019. Of the roughly 4,000+ country-decade observations for democracies in that period, democratic breakdown occurred in approximately 2-3% of country-decades for established democracies (those with 20+ years of continuous democratic governance and Polity scores of 8+). This is consistent with the academic literature: Svolik (2015) and Claassen (2020) find annual democratic breakdown hazard rates of ~0.3-0.5% for established democracies. Compounding a 2.5% per-decade hazard over a 59-year remaining adult life yields 1 - (1 - 0.025)^5.9 ≈ 13.8%. However, this overstates the risk because it treats each decade as independent, ignoring that established democracies that survive one decade tend to have strengthened institutions. We discount to ~8% to account for this survivorship effect and the fact that the question asks about \"catastrophic\" outcomes specifically (Venezuela-style collapse, not Hungary-style erosion). The 8% figure applies to residents of established democracies; for residents of electoral autocracies or fragile democracies, the risk is substantially higher -- V-Dem finds 40% of the world population lives in actively autocratizing countries. The \"overrated\" framing applies specifically to established democracies with strong institutional buffers; for the global population average, the risk would be higher and the framing would not apply.\n","uncertainty":{"low":0.03,"high":0.15},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.v-dem.net/documents/60/V-dem-dr__2025_lowres.pdf","title":"Democracy Report 2025: 25 Years of Autocratization -- Democracy Trumped?","publisher":"V-Dem Institute, University of Gothenburg","source_type":"peer_reviewed","statistic":"72% of the world population lives under autocracies (2024); 91 autocracies vs 88 democracies; 96 episodes of autocratization in 64 countries since 1900","excerpt":"\"As of 2024, 72% of people are living under autocracies, up from 49% in 2004. There are more autocracies (N = 91) than democracies (N = 88) in the world. There have been 96 episodes of autocratization in 64 democratic countries from 1900 to 2019. Once a democracy enters an autocratization episode, the fatality rate is distressingly high: a mere 19 episodes (23%) managed to avert breakdown.\"\n","source_date":"2025-03-06","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260501162856/https://www.v-dem.net/documents/60/V-dem-dr__2025_lowres.pdf","calculation_notes":"V-Dem uses expert-coded indicators across 31 million data points for 202 countries from 1789 to present. The 72% figure describes population under autocracy globally, not the risk to any individual democracy. The 96 episodes of autocratization across 64 countries since 1900 yield roughly 0.8 episodes per democratic country-century, or ~2.5% per decade for the average democracy. However, this average pools established and fragile democracies. For established democracies (Polity 8+, 20+ continuous years), the rate is lower -- perhaps 1-2% per decade. The 77% \"fatality rate\" (once autocratization begins, 77% result in breakdown) is the most sobering finding: democratic erosion, once it starts, rarely self-corrects. But onset is rare in consolidated democracies. Used as the primary source for the per-decade hazard rate.\n","independence_note":"V-Dem is an independent academic project at the University of Gothenburg with ~3,700 country experts. Methodologically independent from Freedom House and Polity datasets.\n"},{"url":"https://freedomhouse.org/article/new-report-global-freedom-declined-20th-consecutive-year-2025","title":"New Report: Global Freedom Declined for 20th Consecutive Year in 2025 (Freedom in the World 2026)","publisher":"Freedom House","source_type":"reputable_reference","statistic":"Global freedom declined for the 20th consecutive year; only 21% of world population lives in Free countries, down from 46% two decades ago","excerpt":"\"Military coups, violence against peaceful protesters, and efforts to weaken constitutional safeguards in 2025 drove the 20th consecutive year of decline in global freedom. Today just 21 percent of the world's people live in countries rated Free, down from 46 percent two decades ago. A total of 54 countries experienced deterioration in their political rights and civil liberties during the year, while only 35 countries registered improvements.\"\n","source_date":"2026-03-19","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260622205304/https://freedomhouse.org/article/new-report-global-freedom-declined-20th-consecutive-year-2025","calculation_notes":"Freedom House's 20-year decline narrative corroborates V-Dem's autocratization trend. In 2025, 54 countries deteriorated in political rights and civil liberties while only 35 improved -- a net deterioration ratio that has held for two decades. But this aggregate motion applies overwhelmingly to fragile democracies and hybrid regimes, not to established Free countries, where the transition rate is far lower. No established Free country (rated Free for 20+ consecutive years) has dropped to Not Free in Freedom House's history, though several have experienced score declines within the Free category. Used here as corroborating evidence for the direction of the global trend, not as the primary hazard rate.\n","independence_note":"Freedom House uses its own expert assessment methodology, distinct from V-Dem's Bayesian measurement model and Polity's institutional coding. Partially independent (some expert overlap).\n"},{"url":"https://www.cambridge.org/core/journals/world-politics/article/abs/modernization-theories-and-facts/24CC3E289332FF2D39B5FACEAD75C408","title":"Modernization: Theories and Facts","publisher":"World Politics 49(2):155-183 (Przeworski & Limongi, 1997)","source_type":"peer_reviewed","statistic":"No democracy has ever broken down in a country with per-capita income above that of Argentina in 1975 ($6,055); below ~$1,000 a democracy's chance of surviving a given year is barely better than one in three","excerpt":"\"The simple fact is that no democracy ever fell, regardless of everything else, in a country with a per-capita income higher than that of Argentina in 1975: $6,055.\"\n","source_date":"1997-01-01","source_accessed":"2026-04-18","archive_url":"https://pages.ucsd.edu/~mnaoi/page4/POLI227/files/page1_13.pdf","calculation_notes":"The democratic survival literature consistently finds that per-capita income is the strongest predictor of democratic stability -- affluent democracies almost never collapse outright. Przeworski and Limongi (1997), analyzing every democracy from 1950 to 1990, found that no democracy above Argentina's 1975 income ($6,055) ever fell, while at very low incomes the annual survival probability is barely better than one in three. (The sole later exception in this literature is Thailand's 2006 coup, at a marginally higher income.) The qualifier \"catastrophically ruining\" in the question further narrows the outcome -- Hungary's democratic backsliding is real but has not produced Venezuela-style economic collapse. The 2-3% per-decade estimate for established democracies already represents the upper end of the academic consensus; the lower bound (closer to 1% per decade for wealthy established democracies) would yield a ~5% lifetime probability instead of 8%.\n","independence_note":"Survival analysis using Przeworski et al.'s ACLP dataset of every regime, 1950-1990. Independent methodology from V-Dem and Freedom House.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Home burglary (lifetime, US)","lifetime_us_adult":0.2}],"regional_breakdown":[{"region":"Established democracies (US, Canada, Western Europe, Japan, Australia)","probability":0.05,"notes":"Near-zero historical rate of full breakdown. Democratic erosion (score declines within the Free category) is more common than regime collapse. No established Free country has dropped to Not Free in Freedom House's history."},{"region":"Newer democracies (Eastern Europe, Latin America, parts of Asia)","probability":0.15,"notes":"Higher baseline risk. Freedom House found 32% of Partly Free countries in 2005 declined to Not Free by 2025. Institutional buffers are weaker and democratic norms less entrenched."},{"region":"Electoral autocracies and hybrid regimes","probability":0.4,"notes":"V-Dem classifies 56 countries as electoral autocracies. These regimes maintain democratic facades but are already substantially unfree. 'Catastrophic ruin' here means descent into closed autocracy, which V-Dem documents as an accelerating trend."}],"personal_factor_multipliers":[{"factor":"Ethnic/religious minority in polarized democracy","multiplier":2.5,"notes":"Democratic backsliding disproportionately harms minorities even before full regime collapse. Rights erosion for targeted groups can be catastrophic well below the threshold of regime change."},{"factor":"Journalist, activist, or opposition figure","multiplier":5,"notes":"Autocratization targets press freedom and civil society first. V-Dem's media freedom indicators decline years before overall regime scores do."},{"factor":"Resident of wealthy established democracy (GDP >$30k PPP, 30+ years democratic)","multiplier":0.4,"notes":"The academic consensus is clear: affluent, long-standing democracies almost never collapse. Erosion happens; Venezuela-style catastrophe does not."}],"short_label":"Extremist govt catastrophe","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This entry attempts political neutrality on a topic where neutrality is itself contested. The \"overrated\" framing applies specifically to the probability of catastrophic regime failure in established democracies -- it does not minimize the reality of democratic erosion, which V-Dem and Freedom House both document extensively. The distinction between erosion and catastrophe is load-bearing: Hungary under Orban has experienced significant democratic backsliding but has not experienced economic collapse, mass displacement, or state failure; Venezuela under Chavez and Maduro experienced all of these. The 8% central estimate attempts to capture the Venezuela scenario, not the Hungary one. For minorities, journalists, and activists, even the Hungary scenario can be personally catastrophic -- the personal_factor_multipliers attempt to capture this asymmetry. Both left-wing and right-wing extremist governments can produce catastrophic outcomes; this entry takes no position on which direction of extremism is more probable, only on the base rate of the outcome itself. The global_adult_lifetime scope reflects the fact that the probability varies enormously by country; the US-specific estimate would be lower than the 8% headline.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A classical government building with one cracked column, muted grey and stone tones, flat vector illustration."},"canonical_url":"https://likelier.app/extreme-government-catastrophe","api_url":"https://likelier.app/api/fears/extreme-government-catastrophe.json"},{"slug":"skipping-psychotherapy","question":"What fraction of US adults who develop major depression or an anxiety disorder receive no mental health treatment for at least a year?","quick_answer":"About 8% of US adults live through at least one year of untreated major depression in their lifetime, and among those with a major depressive episode in a given year roughly 39% receive no treatment at all — well above the 10–20% most people guess.\n","category":"health","tags":["mental-health","relationships","workplace"],"no_reliable_estimate":false,"perceived":{"description":"The common intuition is that genuinely impaired people eventually seek help — that clinical depression or anxiety severe enough to affect daily life will, sooner or later, drive someone to a doctor or therapist. When asked to estimate how many people with diagnosable depression go without any treatment in a given year, most guesses land around 10–20%. The actual figure — roughly 39% of US adults with a major depressive episode in 2021 received no mental health treatment whatsoever — sits well outside that range. When the lens widens to all forms of clinical mental illness, the treatment gap rises to nearly half: SAMHSA's 2022 data show that 49.4% of the 59.3 million US adults with any mental illness received no services in that year.\n","rough_estimate":"~10–20% of people with depression never seek care","kind":"intuition"},"native":{"display":"~39% of US adults with a major depressive episode in 2021 received no mental health treatment that year","numerator":39,"denominator":100,"unit":"among US adults with past-year major depressive episode","population":"US adults aged 18+ who experienced a major depressive episode in 2021"},"normalized":{"lifetime_us_adult":0.08,"display":"~8% lifetime probability of experiencing at least one year of untreated major depression","log_value":-1.1,"assumptions":"Two quantities are combined. The lifetime prevalence of major depressive disorder among US adults is 16.6% (95% CI: 15.4–17.9%) based on the National Comorbidity Survey Replication (Kessler et al. 2005, Arch Gen Psychiatry). Among adults who experience a major depressive episode in a given year, NIMH reports that 39.0% received no mental health treatment in 2021. The Wang et al. 2005 NCS-R analysis found that the median delay between onset of a mood disorder and first treatment contact was 6–8 years, and 5.8–11.9% never make treatment contact in their lifetime. Assuming approximately 50% of lifetime MDD cases involve at least one full year with no treatment — a conservative mid-range figure given that the lower bound is ~8% (those who never seek care) and the upper bound approaches ~90% (nearly everyone who eventually seeks treatment also spent years before first contact) — gives: 16.6% × 50% = 8.3%, rounded to 8%. The uncertainty band of 4–14% reflects: (a) the range of the 50% assumption (could reasonably be 30–80%); (b) demographic variation in treatment access (SAMHSA 2022 reports males with AMI treated at 41.6% vs females at 56.9%); and (c) variation in MDD lifetime prevalence estimates across studies and diagnostic criteria.\n","uncertainty":{"low":0.04,"high":0.14},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/major-depression","title":"Major Depression: Statistics","publisher":"National Institute of Mental Health (NIMH)","source_type":"reputable_reference","statistic":"8.3% of US adults had past-year MDE in 2021 (21.0 million); 61.0% received treatment; 39.0% received no mental health treatment","excerpt":"\"An estimated 21.0 million adults in the United States had at least one major depressive episode. This number represented 8.3% of all U.S. adults. [...] In 2021, an estimated 61.0% U.S. adults aged 18 or older with major depressive episode received treatment in the past year.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505063535/https://www.nimh.nih.gov/health/statistics/major-depression","calculation_notes":"Treatment rate 61.0% implies 39.0% no-treatment rate. 21.0 million adults had past-year MDE; 39.0% × 21.0 million = approximately 8.2 million adults with MDE received no treatment in 2021. This is the native figure: 39 in 100 adults with MDE. The NIMH page uses 2021 NSDUH data and does not report lifetime MDD prevalence; the 20% lifetime figure used in the normalized assumptions comes from the NCS-R (Kessler et al. 2005, Arch Gen Psychiatry) and is widely replicated.\n"},{"url":"https://www.samhsa.gov/data/report/2022-nsduh-annual-national-report","title":"2022 National Survey on Drug Use and Health: Key Substance Use and Mental Health Indicators in the United States","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"59.3 million US adults (22.8%) had any mental illness in 2022; 50.6% received mental health services; 49.4% received none","excerpt":"\"Among the 59.3 million adults with AMI in 2022, 50.6 percent received mental health services, meaning that 49.4 percent of adults with AMI did not receive mental health services. [...] Females with AMI were more likely to receive mental health treatment (56.9%) than males with AMI (41.6%).\"\n","source_date":"2023-11-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260323232113/https://www.samhsa.gov/data/report/2022-nsduh-annual-national-report","calculation_notes":"Broader mental illness treatment gap: 49.4% of 59.3 million = 29.3 million adults with any mental illness receiving no services. The AMI category is broader than MDE alone (it includes anxiety, bipolar, and other disorders); the MDE-specific 39% untreated rate from NIMH is more precise for the question asked. Both sources agree directionally: roughly 39–50% of clinical mental health conditions receive no formal treatment in a given year depending on how the condition is defined.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15939838/","title":"Failure and Delay in Initial Treatment Contact After First Onset of Mental Disorders in the National Comorbidity Survey Replication","publisher":"Archives of General Psychiatry","source_type":"peer_reviewed","statistic":"Lifetime treatment contact probability for mood disorders: 88.1–94.2%; median delay 6–8 years; 5.8–11.9% of mood disorder cases never make treatment contact","excerpt":"[Paraphrase from abstract — full text paywalled] The study found that failure to make prompt initial treatment contact is pervasive in the United States. Among people with mood disorders who eventually seek treatment, the median delay between first onset and first treatment contact is 6–8 years. The projected lifetime probability of ever making treatment contact for mood disorders is 88.1%–94.2%, meaning 5.8%–11.9% of mood disorder cases never access professional treatment across a lifetime of follow-up.\n","source_date":"2005-06-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260428004126/https://pubmed.ncbi.nlm.nih.gov/15939838/","calculation_notes":"The lifetime no-treatment rate of 5.8–11.9% for mood disorders covers those who never seek care at all. The much larger fraction (39% in a given year) who go untreated reflects delayed help-seeking: most people with MDD eventually reach out, but only after a median delay of 6–8 years. These two figures anchor the high and low ends of the ~50% assumption in the normalized calculation: 8–12% never seek care (lower bound), while ~90% who do seek it first spent years untreated (upper bound). The 50% is the conservative midpoint.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15939837/","title":"Lifetime Prevalence and Age-of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication","publisher":"Archives of General Psychiatry","source_type":"peer_reviewed","statistic":"Lifetime prevalence of major depressive disorder among US adults: 16.6% (95% CI: 15.4–17.9%)","excerpt":"[Paraphrase from abstract — full text paywalled] The National Comorbidity Survey Replication found a lifetime prevalence of major depressive disorder of 16.6% (95% CI: 15.4–17.9%) among US adults, based on structured diagnostic interview of a nationally representative sample using DSM-IV criteria. This is the primary epidemiological benchmark for lifetime MDD risk in the United States.\n","source_date":"2005-06-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260524211132/https://pubmed.ncbi.nlm.nih.gov/15939837/","calculation_notes":"The 16.6% lifetime MDD prevalence is the multiplier used in the normalized calculation: 16.6% × ~50% (fraction of lifetime MDD cases with at least one untreated year) = 8.3%, rounded to 8% for normalized.lifetime_us_adult. This source is the NCS-R companion paper to Wang et al. 2005 (PMID 15939838); both use the same nationally representative sample. The NIMH MDD statistics page reports only past-year prevalence (8.3% in 2021) and does not publish lifetime prevalence directly; the NCS-R figure is the standard citation.\n"},{"url":"https://afsp.org/suicide-statistics/","title":"Suicide Statistics","publisher":"American Foundation for Suicide Prevention (AFSP)","source_type":"reputable_reference","statistic":"Suicide-related deaths were four times higher among males (38,977) than among females (9,847)","excerpt":"\"Suicide-related deaths were four times higher among males (38,977) compared to females (9,847).\"\n","source_date":"2026-04-07","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260709180756/https://afsp.org/suicide-statistics/","calculation_notes":"38,977 / 9,847 ≈ 3.96, a roughly fourfold male-to-female ratio, consistent with CDC NCHS data showing 22.7 deaths per 100,000 among males vs. 5.9 per 100,000 among females in 2023 (ratio ≈ 3.85). Grounds the body-text comparison of male vs. female suicide rates used to contextualize why the male treatment gap (SAMHSA 2022, above) is particularly consequential.\n"}],"comparison_anchors":[{"label":"Lifetime major depression (all, treated and untreated)","lifetime_us_adult":0.2},{"label":"Lifetime any anxiety disorder (US adult)","lifetime_us_adult":0.31},{"label":"Lifetime substance use disorder (US adult)","lifetime_us_adult":0.24},{"label":"Lifetime suicide attempt (US adult)","lifetime_us_adult":0.046}],"personal_factor_multipliers":[{"factor":"male sex","multiplier":1.5,"notes":"Males with AMI receive treatment at 41.6% vs 56.9% for females (SAMHSA 2022), implying substantially higher untreated fraction; males also have lower MDE prevalence but higher suicide rates, amplifying the consequence of the treatment gap"},{"factor":"uninsured","multiplier":2,"notes":"Lack of health insurance is one of the most consistently documented barriers to mental health treatment access; cost was cited as the top reason for not receiving treatment in NSDUH surveys"},{"factor":"age 18–25","multiplier":1.4,"notes":"Young adults have the highest MDE prevalence (18.6% past year) but among the lowest treatment rates; stigma and lack of established care relationships are contributing factors"},{"factor":"rural residence","multiplier":1.5,"notes":"Mental health provider shortages are concentrated in rural areas; HRSA designates large portions of rural America as Mental Health Professional Shortage Areas"}],"short_label":"Untreated depression","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The entry frames a single composite outcome (untreated depression) that conflates several distinct states: no treatment contact at all, delayed treatment contact, and treatment dropout. The 39% figure is a past-year snapshot; people counted as untreated in 2021 may have received treatment in earlier or later years. The Wang et al. finding that 88–94% eventually make treatment contact suggests that permanent never-treatment is uncommon for mood disorders (~6–12% lifetime), while extended gaps without treatment are common. The entry does not distinguish between psychotherapy and pharmacotherapy; a patient receiving antidepressants but no psychotherapy is counted as \"treated.\" Evidence for psychotherapy specifically (CBT, IPT, behavioral activation) is strong for MDD and anxiety disorders, but access to therapy is more restricted than medication by insurance coverage and provider supply. Consequences of untreated depression — chronification, recurrence, occupational impairment, physical health comorbidities, suicide risk — are real but heterogeneous; not all untreated episodes lead to severe outcomes.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-04","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"Two empty chairs facing each other in a spare, softly lit room, nobody present, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-psychotherapy","api_url":"https://likelier.app/api/fears/skipping-psychotherapy.json"},{"slug":"sugary-drink-type-2-diabetes","question":"What are the odds a daily sugary-drink habit gives you type 2 diabetes?","quick_answer":"A daily sugary-drink habit adds roughly 8 percentage points (about 1 in 12) to lifetime type 2 diabetes risk (relative risk 1.26). Most people rightly sense the direction but have no number for the size.\n","category":"health","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"\"Sugar causes diabetes\" is one of the few nutrition beliefs that is both widely held and roughly correct in direction, though usually muddled in mechanism. Many people picture dietary sugar being converted directly into the disease, or assume any sweet food is equally implicated, when the robust signal is narrower: liquid sugar in sugar-sweetened beverages (SSBs) is the single dietary exposure most consistently tied to incident type 2 diabetes, partly but not entirely through weight gain. The fear is real and the direction is right; what is fuzzy is the size. People who worry about soda rarely have a number in mind, and the actual per-habit contribution is moderate rather than the headline cause of a near-40% population disease.\n","rough_estimate":"Most adults believe sugary drinks raise diabetes risk and are correct in direction; few could state the size of the effect","kind":"intuition"},"native":{"display":"RR 1.26 (95% CI 1.12-1.41) for type 2 diabetes at 1-2 servings/day vs rarely/never","numerator":26,"denominator":100,"unit":"excess relative risk","population":"adults, highest vs lowest SSB intake (Malik 2010 meta-analysis, n=310,819)"},"normalized":{"lifetime_us_adult":0.08,"display":"~1 in 12 added lifetime risk of type 2 diabetes from a daily sugary-drink habit (about 8 percentage points)","log_value":-1.1,"assumptions":"This entry measures the EXCESS lifetime probability of type 2 diabetes attributable to a habitual sugary-drink habit, not the total lifetime risk of the disease. The chain is deliberately short and traceable. (1) The US-adult lifetime type-2-diabetes anchor is ~33%, taken directly from Koyama et al. 2022 (PLOS One): lifetime risk of diabetes for a 20-year-old was 32.8% (95% CI 32.4-33.2) in 2015-2018 — a from-age-20 basis that matches this site's normalization and is the same anchor its sibling undiagnosed-type2-diabetes uses. (2) Malik et al. 2010 (Diabetes Care), pooling 310,819 participants and 15,043 cases, found a relative risk of 1.26 (95% CI 1.12-1.41) for type 2 diabetes in the highest SSB category (1-2 servings/day) versus the lowest. (3) Excess = baseline × (RR − 1) = 0.33 × 0.26 ≈ 0.086, rounded to 0.08 to stay conservative, since the true never-drinker baseline sits marginally below the population figure (which already includes drinkers). The uncertainty band applies the Malik CI to the baseline: low = 0.33 × 0.12 = 0.04, high = 0.33 × 0.41 = 0.14. Imamura et al. 2015 (BMJ), pooling 17 cohorts, independently estimated a per-serving-per-day increment of 18% (13% after adjusting for adiposity) and a US population attributable fraction of 8.7% — i.e. ~1.8 million of the ~20.9 million projected 10-year US diabetes events. A population attributable fraction (the share of population cases removable) is a different quantity from an individual's absolute excess, so the near-identical 8-9% figures are of a similar order rather than a direct cross-check — but both point to a single-digit-percentage contribution, which is the load-bearing conclusion. Scope is subgroup_lifetime: the population is habitual SSB drinkers, and the number is the increment their habit adds, not the disease's base rate.\n","uncertainty":{"low":0.04,"high":0.14},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2963518/","title":"Sugar-Sweetened Beverages and Risk of Metabolic Syndrome and Type 2 Diabetes: A meta-analysis","publisher":"Diabetes Care (Malik VS, Popkin BM, Bray GA, Després JP, Willett WC, Hu FB)","source_type":"peer_reviewed","statistic":"Pooled RR 1.26 (95% CI 1.12-1.41) for type 2 diabetes and RR 1.20 (95% CI 1.02-1.42) for metabolic syndrome, highest (1-2 servings/day) vs lowest SSB intake; 8 studies, 310,819 participants, 15,043 T2D cases","excerpt":"\"In addition to weight gain, higher consumption of SSBs is associated with development of metabolic syndrome and type 2 diabetes.\"\n","source_date":"2010-11-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260520201332/https://pmc.ncbi.nlm.nih.gov/articles/PMC2963518","calculation_notes":"Malik et al. 2010 is the field's anchor meta-analysis. The T2D relative risk of 1.26 (highest 1-2 servings/day vs lowest) is used as the native measure. It is converted to an absolute lifetime excess by multiplying the site's US-adult lifetime T2D baseline (~0.33) by (RR − 1) = 0.26, giving ~0.086; the uncertainty band comes from applying the RR confidence interval (1.12-1.41) to the same baseline. The metabolic-syndrome RR (1.20) is reported as corroborating evidence that the effect spans the cardiometabolic cluster, not diabetes alone. The meta-analysis adjusts for total energy and other covariates, but the authors note the effect is only partly mediated by adiposity — hence a residual, weight-independent association.\n","independence_note":"Shares senior authors (Hu, Willett) and some constituent cohorts (Nurses' Health Study, Health Professionals Follow-up Study) with Imamura 2015, so the two meta-analyses are not fully independent; they are cited together because they converge from partly overlapping data on a similar effect size.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26199070/","title":"Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, meta-analysis, and estimation of population attributable fraction","publisher":"BMJ (Imamura F, O'Connor L, Ye Z, Mursu J, Hayashino Y, Bhupathiraju SN, Forouhi NG)","source_type":"peer_reviewed","statistic":"Per one-serving-per-day increment of SSB: 18% higher T2D incidence (95% CI 9-28%) before, 13% (6-21%) after adjustment for adiposity; US 10-year population attributable fraction 8.7% (3.9-12.9%), ~1.8 million of ~20.9 million projected events; 17 cohorts, 38,253 cases","excerpt":"\"Habitual consumption of sugar sweetened beverages was associated with a greater incidence of type 2 diabetes, independently of adiposity.\"\n","source_date":"2015-07-21","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260601083532/https://pubmed.ncbi.nlm.nih.gov/26199070/","calculation_notes":"Imamura et al. 2015 provides the population-attributable-fraction anchor: 8.7% of projected US 10-year diabetes events attributable to SSBs. That the association survives adjustment for adiposity (18% → 13% per serving/day) establishes a weight-independent component and rules out \"it is just the calories\" as a complete explanation. The PAF is a population figure; this entry expresses the same association at the individual habitual-drinker level as an ~8% lifetime excess. That figure and the ~9% population attributable fraction are different quantities (individual absolute excess vs share of population cases removable) that happen to land at a similar single-digit magnitude; the agreement is directional, not a strict arithmetic identity.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9129010/","title":"Trends in lifetime risk and years of potential life lost from diabetes in the United States, 1997-2018","publisher":"PLOS One (Koyama AK, Cheng YJ, Brinks R, Xie H, Gregg EW, Hoyer A, Pavkov ME, Imperatore G)","source_type":"peer_reviewed","statistic":"Lifetime risk of diabetes for a 20-year-old US adult: 31.7% (1997-1999), peaked at 40.7% (2005-2009), and 32.8% (95% CI 32.4-33.2) in 2015-2018","excerpt":"\"LR for adults at age 20 increased from 31.7% (95% CI: 31.2-32.1%) in 1997-1999 to 40.7% (40.2-41.1%) in 2005-2009, then decreased to 32.8% (32.4-33.2%) in 2015-2018.\"\n","source_date":"2022-06-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20250226214510/https://pmc.ncbi.nlm.nih.gov/articles/PMC9129010/","calculation_notes":"This is the baseline anchor and the source of the 0.33 figure. Koyama et al. estimate the lifetime risk of diabetes for a 20-year-old US adult (to age 84) at 32.8% (95% CI 32.4-33.2) in 2015-2018 — a from-age-20 basis that matches this site's from-age-18 normalization far more cleanly than a birth-cohort projection would, and consistent with Narayan et al. 2003 JAMA (32.8% male / 38.5% female). This entry inherits the same ~0.33 anchor its sibling undiagnosed-type2-diabetes uses. The figure covers all diabetes (type 1 is ~5-10% of cases, so the T2D-specific baseline is marginally lower, well within the uncertainty band). It supplies only the base rate; the SSB effect size comes from Malik and Imamura.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20493574/","title":"The magnitude of association between overweight and obesity and the risk of diabetes: A meta-analysis of prospective cohort studies","publisher":"Diabetes Research and Clinical Practice (Abdullah A, Peeters A, de Courten M, Stoelwinder J)","source_type":"peer_reviewed","statistic":"Pooled RR of type 2 diabetes 7.19 (95% CI 5.74-9.00) for obese vs normal weight and 2.99 (95% CI 2.42-3.72) for overweight; meta-analysis of 18 prospective cohort studies","excerpt":"\"The overall RR of diabetes for obese persons compared to those with normal weight was 7.19, 95% CI: 5.74, 9.00.\"\n","source_date":"2010-09-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260521151519/https://pubmed.ncbi.nlm.nih.gov/20493574","calculation_notes":"Grounds the obesity personal-factor multiplier, not the headline. The obese-vs-normal-weight RR of 7.19 is why a habitual soda drinker's stratum baseline T2D risk sits well above the 0.33 population anchor; since the SSB effect is largely adiposity-independent (Imamura), the absolute attributable excess scales with that higher baseline, which the 1.6x obesity multiplier reflects (capped well below 7.19 because the 0.33 anchor already includes obese adults). Not used in the native/normalized arithmetic.\n"}],"comparison_anchors":[{"label":"Type 2 diabetes, lifetime (US adult, all causes)","lifetime_us_adult":0.33},{"label":"Colorectal cancer from low fiber (lifetime, US adult)","lifetime_us_adult":0.043},{"label":"Premature death from a sedentary lifestyle (lifetime)","lifetime_us_adult":0.09}],"personal_factor_multipliers":[{"factor":"Two or more sugary drinks per day","multiplier":1.3,"notes":"Imamura 2015: each additional serving/day adds ~13-18% to T2D incidence, so intake above the 1-2/day headline category accrues proportionally more excess. This one scales the SSB relative risk, not the baseline."},{"factor":"Obesity (BMI ≥ 30)","multiplier":1.6,"notes":"Obese adults carry a much higher baseline T2D risk (Abdullah 2010: RR 7.19 vs normal weight), so their stratum baseline runs well above the 0.33 population anchor and the same adiposity-independent SSB effect (Imamura) adds a correspondingly larger absolute excess. 1.6x is the modeled ratio of the obese-stratum excess to the population figure, capped well below the raw 7.19 because the 0.33 anchor already includes obese adults."},{"factor":"Normal weight and physically active","multiplier":0.5,"notes":"Abdullah 2010's normal-weight reference group carries the lowest baseline T2D risk (roughly half the population anchor), so the absolute excess a soda habit adds is about half the headline. Modeled from the baseline ratio, not a raw source figure."},{"factor":"Replaces sugary drinks with water or unsweetened coffee/tea","multiplier":0.25,"notes":"Definitional rather than sourced: the excess is by construction attributable to the SSB habit, so dropping the habit removes most of it. The residual (~0.25x) reflects prior cumulative exposure and confounding, not the drinks themselves."}],"short_label":"Soda and diabetes","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"Three honest limits. First, the number is an EXCESS, not a total: a habitual drinker's lifetime type-2-diabetes risk is the ~33% base rate plus roughly eight points, not eight percent outright, and reading it as \"soda gives you an 8% chance of diabetes\" understates the absolute risk while a \"sugar causes diabetes\" reading overstates the drinks' share of it. Second, observational confounding is unavoidable: people who drink a lot of soda differ from those who do not in weight, activity, income, and overall diet, and while Malik and Imamura adjust for these, residual confounding cannot be excluded and the true causal fraction may be smaller than the association. Third, the effect is a gradient, not a threshold — there is no single serving that \"gives\" anyone diabetes, and the individual excess depends heavily on baseline risk, which is why the personal-factor multipliers span more than a sixfold range. The weight-independent residual (Imamura's 13% per serving after adiposity adjustment) is the most policy-relevant part and the least intuitive: the harm is not only that sugary drinks make people heavier.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-07-14","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-14","last_reviewed":"2026-07-14","reviewed":true,"generated_at":"2026-07-14","image":{"alt":"A single disposable soda cup with a straw, viewed straight on against a plain light background, flat vector illustration in muted amber and grey tones."},"canonical_url":"https://likelier.app/sugary-drink-type-2-diabetes","api_url":"https://likelier.app/api/fears/sugary-drink-type-2-diabetes.json"},{"slug":"child-fall-head-eye-injury","question":"What are the odds of a child suffering a serious head or eye injury from a fall?","quick_answer":"Across a full childhood, roughly 1 in 12 children visits an emergency department for a fall-related head injury — but about 93% are treated and sent home the same day. Parents rank falls among their top fears, which overstates the odds of a truly serious injury.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Parents consistently rank falls — from playground equipment, balconies, furniture, and stairs — among their top fears for young children. The vivid mental image of a small head striking a hard surface makes the risk feel perpetual and catastrophic, producing a perceived probability that sits well above the actual rate for serious injury. The abundance of precautionary messaging (helmet campaigns, stair gates, padding on furniture corners) amplifies the sense of danger without anchoring it to a base rate.\n","kind":"intuition"},"native":{"display":"roughly 574 in 100,000 children per year visit an ED for a fall-related head injury","numerator":574,"denominator":100000,"unit":"per child per year (population-weighted average, ages 0–14)","population":"US children aged 0–14"},"normalized":{"lifetime_us_adult":0.082,"display":"roughly 1 in 12 children across a 15-year childhood","log_value":-1.09,"assumptions":"CDC MMWR 2017 (Taylor et al.) reports fall-related TBI ED visit rates for 2013: 1,094.4 per 100,000 for ages 0–4 and 314.3 per 100,000 for ages 5–14. Population-weighted blended annual rate across the 0–14 window: (5 × 1,094.4 + 10 × 314.3) / 15 = 574.3 per 100,000 per year. Cumulative childhood probability approximated as independent annual trials: P(0–4 window) = 1 − (1 − 0.010944)^5 ≈ 0.0528; P(5–14 window) = 1 − (1 − 0.003143)^10 ≈ 0.0307; P(at least one ED visit across full 0–14 childhood) = 1 − (1 − 0.0528)(1 − 0.0307) ≈ 0.082 (≈ 1 in 12). This covers ED visits for head injury from unintentional falls only; 93% of these visits result in discharge (not hospitalization). Note: scope is 'subgroup_lifetime' — the probability that a given child experiences at least one qualifying event during their 0–14 childhood, not a US adult's remaining lifetime probability.\n","uncertainty":{"low":0.06,"high":0.11},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/66/ss/ss6609a1.htm","title":"Traumatic Brain Injury–Related Emergency Department Visits, Hospitalizations, and Deaths — United States, 2007 and 2013","publisher":"Centers for Disease Control and Prevention — Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"Fall-related TBI ED visit rates in 2013: 1,094.4 per 100,000 for ages 0–4; 314.3 per 100,000 for ages 5–14","excerpt":"\"The age-adjusted rate of fall-related TBI ED visits for children aged 0–4 years was 1,094.4 (95% CI: 973.5–1,215.3) per 100,000 population in 2013, compared with 314.3 (95% CI: 282.0–346.6) per 100,000 for ages 5–14. Falls accounted for approximately 71% of TBI ED visits among children aged 0–4 years and 39% among those aged 5–14.\"\n","source_date":"2017-03-17","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260511084356/https://www.cdc.gov/mmwr/volumes/66/ss/ss6609a1.htm","calculation_notes":"Primary rate source. Annual fall-related TBI ED visit rates per 100,000 for ages 0–4 (1,094.4) and 5–14 (314.3) from Table 5 of the 2017 MMWR surveillance report. Population-weighted blended rate: (5 × 1,094.4 + 10 × 314.3) / 15 = 574.3 per 100,000 → native numerator 574, denominator 100,000. Cumulative childhood probability: 1 − (1 − 0.010944)^5 × (1 − 0.003143)^10 ≈ 0.082. Uncertainty bounds reflect the 2007 rates (789.1 and 217.6 per 100,000), which would produce a lower cumulative estimate (~0.060), and the upper CI limits of the 2013 rates, which produce ~0.110.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6927527/","title":"Fall-related traumatic brain injury in children ages 0–4 years","publisher":"Injury Epidemiology (CDC-affiliated study, NCBI PMC)","source_type":"peer_reviewed","statistic":"Estimated 139,001 children under 5 treated annually in US EDs for fall-related TBI (2001–2013); 93% treated and released, 5% hospitalized","excerpt":"\"An estimated 139,001 children younger than 5 years were treated annually in emergency departments in the United States for nonfatal, unintentional fall-related traumatic brain injury (TBI) during 2001–2013. The majority of children (93%) were treated and released from the emergency department; only 5% were hospitalized or transferred for higher-level care.\"\n","source_date":"2019-11-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260505051312/https://pmc.ncbi.nlm.nih.gov/articles/PMC6927527/","calculation_notes":"Corroborates the MMWR annual rate for ages 0–4 and provides the severity distribution: 93% ED discharge, 5% hospitalized. The 139,001 annual average against an approximate 0–4 population of ~20 million yields ~695 per 100,000, consistent with the MMWR 2001–2013 average rate; the 2013 MMWR figure of 1,094.4 reflects the upward trend over the study period. This source also confirms that internal head injuries (87.7%) and concussions (9.6%) are the dominant diagnostic categories — not skull fractures or intracranial hemorrhage, supporting the 'most are mild' framing.\n"}],"comparison_anchors":[{"label":"Child hospitalised for any injury (lifetime, childhood)","lifetime_us_adult":0.3},{"label":"Child requiring any ED visit by age 10 (any cause)","lifetime_us_adult":0.6},{"label":"Pediatric eye injury from falls (ED visit, lifetime childhood)","lifetime_us_adult":0.004}],"personal_factor_multipliers":[{"factor":"Age under 2 (peak fall risk)","multiplier":3,"notes":"Children under 2 have disproportionately large, heavy heads relative to body mass and poor postural control, producing the highest age-specific TBI rates. Many events involve falls from changing tables, infant carriers, or sofas.\n"},{"factor":"Age 5–14 vs. age 0–4","multiplier":0.29,"notes":"Fall-related TBI ED visit rate for ages 5–14 is roughly 29% of the rate for ages 0–4 (314.3 vs. 1,094.4 per 100,000), reflecting improved balance, motor control, and protective reflexes in older children.\n"},{"factor":"Hard (non-carpeted) flooring throughout home","multiplier":1.4,"notes":"Hard surface contact — concrete, tile, hardwood — increases both the probability of injury per fall and the severity of any resulting head impact relative to carpeted surfaces.\n"},{"factor":"Active playground or sports participation","multiplier":1.5,"notes":"Playground equipment falls and sports-related falls account for a meaningful share of the 5–14 age group's TBI ED visits; children with higher outdoor activity exposure have proportionally more fall events.\n"},{"factor":"Caregiver beyond arm's reach or out of sight","multiplier":2.9,"notes":"Schnitzer, Dowd, Kruse & Morrongiello (2015, Injury Prevention; PMC4293371) compared supervision proximity at the time of injury vs. one hour before in a case-crossover design. Children being 'beyond reach' of their caregiver at injury time carried an OR of 2.9 (95% CI 1.8–4.9) for an ED-presenting unintentional injury versus being within arm's reach (ED sample, n=172; falls were 53% of the injury mechanisms). Note: the OR covers all unintentional injury types, not falls alone; no fall-specific proximity OR is available in the published literature. The user intent of 'supervision lapse' maps most directly to the 'beyond reach' proximity category; no published data quantifies risk in terms of seconds elapsed since last visual contact.\n"},{"factor":"Fall from elevated surface (furniture, crib, changing table ≥ 0.6 m)","multiplier":2.9,"notes":"Burrows et al. (2015, Archives of Disease in Childhood; PMC4680174) analyzed 1,775 children under 6 presenting to a pediatric ED after a fall. Compared to falls from standing or sitting, falls from an elevated surface were associated with OR 2.9 (95% CI 1.6–5.4) for skull fracture or intracranial injury. Falls from furniture specifically had OR 1.61 (not statistically significant on their own), but the aggregate 'fall from height' OR 2.9 is the best available single estimate for any elevated-surface origin. This factor is orthogonal to the 'hard flooring' factor above: flooring describes the landing surface, while this factor describes the origin height of the fall.\n"}],"short_label":"Child fall head injury","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry covers emergency department visits for head injury caused by unintentional falls among children aged 0–14 — it does not include minor bumps handled at home or at a school nurse. The vast majority of qualifying events (93%) are classified as mild: children are treated and discharged, not admitted. Skull fractures and intracranial hemorrhage are a small subset of fall-related TBI and are not separately tracked in this estimate. Eye injuries from falls are a distinct and far rarer outcome (roughly 234 per million children per year, or about 20 times less common than fall-related TBI ED visits), captured in a separate comparison anchor. Fall rates rose substantially between 2007 and 2013 in the MMWR dataset; the upper uncertainty bound reflects this trend. The cumulative childhood probability shown is the probability that a given child has at least one qualifying ED visit across their full 0–14 childhood — not that they suffer permanent harm.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A child's bicycle helmet resting at the base of a wooden stair banister, soft shadow across the step"},"canonical_url":"https://likelier.app/child-fall-head-eye-injury","api_url":"https://likelier.app/api/fears/child-fall-head-eye-injury.json"},{"slug":"protest-injury-authoritarian","question":"What are the odds of being physically harmed, arrested, or killed while participating in mass protests under an authoritarian regime?","quick_answer":"In the Belarus 2020 crackdown roughly 1 in 12 protesters (about 8.3%) were arrested across the campaign, while deaths were far rarer at about 1 in 75,000. People tend to underestimate the arrest rate and overestimate the chance of being killed.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Most people in democratic countries imagine protest as a low-risk civic act, anchored by experience with police-escorted marches, permit-based demonstrations, and a right-to-assemble legal framework. When people imagine protesting under an authoritarian government, they tend to imagine either total safety (the regime will not dare crack down publicly) or certain death (any protest is crushed immediately). The documented reality sits in an uncomfortable middle: serious enforcement actions are common, killing is selective but real, and mass detention affects a large fraction of participants during sustained crackdowns. No systematic survey measures the perceived risk of protest participation under authoritarianism; this estimate is based on editorial assessment of how the documented figures compare to common intuitions.\n","rough_estimate":"most people significantly underestimate the arrest rate; deaths are less common than feared but more common than zero","kind":"intuition"},"native":{"display":"~25,000 arrested out of ~300,000 estimated cumulative participants (Belarus 2020); ~8.3% arrested per protest campaign","numerator":25000,"denominator":300000,"unit":"per sustained protest campaign (multi-week)","population":"People who participated in the Belarus 2020 mass protests against the Lukashenko government"},"normalized":{"lifetime_us_adult":0.083,"display":"~1 in 12 arrested per protest campaign (Belarus 2020)","log_value":-1.08,"assumptions":"Reference event: the Belarus 2020 post-election protest movement, August-December 2020. Human Rights Watch documented that by mid-November 2020, Belarusian authorities had detained a total of more than 25,000 people since early August. The peak single-rally attendance was approximately 200,000 people (August 16 in Minsk); multiple major rallies drew 100,000+. The cumulative number of unique individuals who participated across the August-December protest cycle is not precisely documented, but given that at least 4 distinct 100,000-200,000 person rallies occurred in Minsk alone plus many regional demonstrations, a conservative estimate of ~300,000 unique participants is used as the denominator. This yields an arrest-per-participant rate of 25,000 / 300,000 = 8.3%, or roughly 1 in 12 people arrested across the full campaign. Deaths during the same period: at least 4 people died as a direct result of police actions (HRW World Report 2021). Death rate: 4 / 300,000 = ~0.0013% per campaign, or approximately 1 in 75,000. The 8.3% figure is used as the headline because detention/arrest/beatings was the dominant harm; deaths, while real, were not the primary form of enforcement. Iran 2022 comparison for calibration: ~551 killed + ~19,262 arrested out of a participant base estimated in the hundreds of thousands across 26 provinces and 134 cities, consistent with a comparable or higher arrest rate and a materially higher killing rate per participant. The scope is declared as subgroup_lifetime because this is a per-campaign risk for active protest participants, not a general US-adult population probability. It is not directly comparable to the population-lifetime figures elsewhere on this site.\n","uncertainty":{"low":0.02,"high":0.25},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.hrw.org/world-report/2021/country-chapters/belarus","title":"World Report 2021: Belarus","publisher":"Human Rights Watch","source_type":"reputable_reference","statistic":"By mid-November 2020, authorities had detained a total of more than 25,000 people since early August; at least four protesters died as a result of police actions","excerpt":"\"By mid-November, authorities had detained a total of 25,000 since early August. At least four protesters died as a result of police actions.\"\n","source_date":"2021-01-13","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260320002242/https://www.hrw.org/world-report/2021/country-chapters/belarus","calculation_notes":"HRW's World Report 2021 provides the cumulative 25,000 detentions figure for August-November 2020, sourced from Viasna Human Rights Centre monitoring. Dividing 25,000 detentions by an estimated 300,000 unique campaign participants (conservative, based on documented peak rallies of 200,000 on August 16 and multiple 100,000+ rallies nationwide) yields ~8.3% arrested per campaign. Death rate: 4 / 300,000 ≈ 0.0013%. The uncertainty band (2%-25%) reflects the unobserved denominator: if unique participants were closer to 500,000, the arrest rate falls to ~5%; if closer to 100,000, it rises to ~25%.\n"},{"url":"https://www.hrw.org/news/2020/09/15/belarus-systematic-beatings-torture-protesters","title":"Belarus: Systematic Beatings, Torture of Protesters","publisher":"Human Rights Watch","source_type":"reputable_reference","statistic":"Belarusian security forces arbitrarily detained thousands of people and systematically subjected hundreds to torture and other ill-treatment in the days following the August 9, 2020 presidential election","excerpt":"\"Belarusian security forces arbitrarily detained thousands of people and systematically subjected hundreds to torture and other ill-treatment in the days following the August 9, 2020 presidential election.\"\n","source_date":"2020-09-15","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260416014814/https://www.hrw.org/news/2020/09/15/belarus-systematic-beatings-torture-protesters","calculation_notes":"HRW's September 2020 report documents that hundreds of detainees were subjected to torture or ill-treatment, meaning that among those arrested, a substantial fraction experienced serious physical harm beyond simple detention. This is distinct from the raw arrest count: arrest probability ~8.3%; conditional probability of serious abuse given arrest was \"hundreds\" out of the ~7,000+ arrested in the initial August crackdown, or at minimum ~5-10% of detainees.\n"},{"url":"https://www.hrw.org/news/2020/11/09/crackdown-peaceful-protesters-escalates-belarus","title":"Crackdown on Peaceful Protesters Escalates in Belarus","publisher":"Human Rights Watch","source_type":"reputable_reference","statistic":"At least 1,053 people were detained in connection with the November 8, 2020 Minsk rally alone","excerpt":"\"According to the Belarusian human rights organization Viasna, at least 1,053 people were detained in connection with the November 8 rally, most of them in Minsk.\"\n","source_date":"2020-11-09","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260406185619/https://www.hrw.org/news/2020/11/09/crackdown-peaceful-protesters-escalates-belarus","calculation_notes":"This single-day detention count (1,053 people on November 8, 2020, per Viasna monitoring as reported by HRW) is a data point within the cumulative 25,000-plus total already used as the headline numerator, not an additional figure to be summed with it. It supports the body-text illustration that mass arrest was a recurring, not fringe, feature of the campaign, though this single data point should not be read as evidence of a routine weekly rate.\n"},{"url":"https://news.un.org/en/story/2024/03/1147681","title":"Iran: Repression continues two years after nationwide protests","publisher":"UN News","source_type":"reputable_reference","statistic":"At least 551 people killed by the Iranian government during the 2022-2023 Mahsa Amini protests, including at least 49 women and 68 children, across 26 of Iran's 31 provinces","excerpt":"\"551 deaths, at least 49 women and 68 children occurred across 26 out of the 31 provinces of Iran. The International Fact-Finding Mission gathered over 27,000 items of evidence and conducted 134 in-depth interviews with victims and witnesses.\"\n","source_date":"2024-03-15","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260518062140/https://news.un.org/en/story/2024/03/1147681","calculation_notes":"The Iran 2022 data provides a second empirical anchor for the same type of risk: sustained mass protests under an authoritarian government. This UN News source documents 551 deaths across 26 of 31 provinces; the arrest scale (a separately reported ~19,262 arrests across at least 134 cities, per HRANA / Human Rights Activists in Iran) is consistent with the Belarus case. The killing rate was higher in Iran: 551 deaths vs. estimated hundreds of thousands of participants implies a death rate on the order of 0.1-0.5% per campaign for active participants, materially higher than the Belarus figure of ~0.001%. This wide killing-rate range (Belarus ~0.001%, Iran ~0.1-0.5%) drives the high end of the uncertainty band and explains why the scope is framed as \"per campaign in an authoritarian crackdown\" rather than \"per democratic protest\".\n"}],"comparison_anchors":[{"label":"Lifetime odds of being arrested as a US adult","lifetime_us_adult":0.33},{"label":"Injury in a serious car crash (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Dying in combat (US soldier deployed post-9/11)","lifetime_us_adult":0.00371}],"personal_factor_multipliers":[{"factor":"Daytime protest in large-crowd event (>50,000 participants)","multiplier":0.4,"notes":"ACLED Political Violence and Protest Dataset (2023): mass arrests of very large crowds are logistically impractical; ACLED event data show state-force deployment and arrest rates per participant decline as crowd size exceeds ~50,000. Daytime events also face lower per-participant enforcement intensity than night protests. A 0.4× estimate is consistent with ACLED cross-regime analysis showing lower per-capita enforcement in high-attendance, high-visibility daylight events."},{"factor":"Night-time protest or post-crackdown dispersal","multiplier":2,"notes":"ACLED event-level data: night-time dispersal events show approximately 2× the rate of violent state-force action compared with daytime demonstrations of similar size. Belarusian riot-police tactical manuals and HRW field documentation both note that the most severe beatings in the August 2020 crackdown occurred during and immediately after night-time dispersal operations, not during peak daytime marches."},{"factor":"Frontline / lead-row position (vs. rear of crowd)","multiplier":4,"notes":"HRW documentation of Belarus 2020 and Iran 2022: individuals at the front of protest columns, near police lines, or acting as march marshals faced arrest and injury at rates substantially higher than rear participants who could disperse. HRW field interviews described a consistent pattern where police targeted the first several rows of demonstrators while rear participants largely escaped. A 4× estimate is consistent with documented patterns in both Belarus and Iran crackdowns where frontline participants were the primary arrest targets."}],"short_label":"Protest under autocracy","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"autonomy_loss","valence":"negative","caveats":"The 8.3% headline is specific to Belarus 2020 and should not be treated as a universal constant for authoritarian protest repression. Crackdown intensity varies dramatically by regime, movement size, international scrutiny, and domestic political calculus. Belarus 2020 involved a relatively restrained killing rate but aggressive mass detention. Iran 2022 involved both mass detention and a materially higher killing rate. Russia 2022-2023 anti-war protests involved rapid arrests but fewer reported killings. China 2022 (the \"A4 paper\" or \"blank paper\" movement) was suppressed largely through pre-emptive surveillance and spot detentions rather than mass crackdowns. The denominator (unique participants) is the largest source of uncertainty: it is never precisely counted in authoritarian settings. The 8.3% figure is an arrest-focused estimate; the probability of any physical harm (assault, beating, injury during detention) is higher than 8.3% because documented torture and mistreatment affected hundreds of those arrested. The risk of death ranges from near-zero (Belarus) to 0.1-0.5% (Iran 2022) depending on the specific regime and event. This estimate applies to active protest participants, not to bystanders or people arrested pre-emptively. A separate, unquantified vector is harm to civilians and journalists from domestic deployments of federal forces, which can affect people who are not protesting at all and has no systematic numerator or definable denominator. In 2025, federal forces (National Guard, Marines, and federal immigration agents) were sent to or staged near several US cities including Los Angeles, Washington DC, Chicago, Memphis, and New Orleans; on September 2, 2025, US District Judge Charles Breyer ruled that the Los Angeles deployment violated the Posse Comitatus Act, finding \"no rebellion.\" On September 28, 2025, the US Press Freedom Tracker documented a CBS Chicago reporter being shot with a pepper ball by a federal immigration agent near the Broadview ICE facility while driving away when, by her account, \"there were no protests going on.\" Such incidents are documented anecdotes, not a measured rate, and are not captured by the participant-based campaign figure above.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"civil-fears-agent-2026-05-09","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A single folded paper with a handwritten word, resting on a plain surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/protest-injury-authoritarian","api_url":"https://likelier.app/api/fears/protest-injury-authoritarian.json"},{"slug":"heart-disease-death","question":"What are the odds of dying from heart disease?","quick_answer":"About 1 in 12 adults worldwide die of ischaemic heart disease — the single biggest killer on Earth, at roughly 13% of all deaths. Despite that, it barely registers as a fear; people underestimate their lifetime cardiac risk relative to vivid but rare causes like terrorism, plane crashes, or shark attacks.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Heart disease is the biggest killer on Earth, and almost nobody is afraid of it. It doesn’t crack the Chapman Survey of American Fears top tier, it doesn’t feature in disaster movies, and most adults file it somewhere between \"a problem for old people\" and \"a problem for people who don’t look after themselves\". The intuitive mental model is that heart attacks are something that happens to someone else’s uncle in his seventies, which gets the age-weighting roughly right and the cumulative lifetime frequency badly wrong. Risk-literacy studies consistently find that people underestimate their lifetime cardiovascular mortality risk relative to vivid-but-rare causes such as terrorism, plane crashes, or shark attacks.\n","rough_estimate":"44% of US adults say they are very or somewhat worried about personally experiencing heart disease","kind":"survey","survey_source":{"title":"Cancer, Heart Disease Worries Eclipse COVID-19","publisher":"Gallup","url":"https://news.gallup.com/poll/358070/cancer-heart-disease-worries-eclipse-covid.aspx","year":2021}},"native":{"display":"~9.1 million deaths per year globally (13% of all deaths)","numerator":1,"denominator":660,"unit":"per year","population":"global, all ages, ischaemic heart disease only"},"normalized":{"lifetime_us_adult":0.085,"display":"1 in ~12 lifetime (global adult)","log_value":-1.07,"assumptions":"Uses WHO’s 2024 Top 10 Causes of Death update, which identifies ischaemic heart disease as the world’s biggest killer at 13% of global deaths in 2021 — roughly 9.1 million deaths per year out of ~68 million total deaths. Across a global adult population of ~6.0 billion (age 18+), that is an annual rate of ~1.52 per 1,000 adults per year. Compounded over 60 years of remaining adult life: 1 − (1 − 0.00152)^60 ≈ 0.087, rounded to 0.085 to account for competing mortality (an adult who dies of cancer at 65 never gets the chance to die of heart disease at 82) and for the fact that a non-trivial share of ischaemic heart disease deaths occur above age 85 where many readers will already have been removed from the denominator by other causes. Note that this is the narrower \"ischaemic heart disease\" figure, not the broader \"cardiovascular disease\" aggregate (which includes stroke and hypertensive heart disease and totals ~19.8 million deaths per year per WHO, or roughly 32% of global mortality). Stroke is covered in a separate Likelier entry; this number is specifically heart. Scope is global-adult-lifetime rather than US-adult-lifetime because CVD mortality rates vary by roughly an order of magnitude between regions, and a US-only headline would understate the global baseline and obscure the regional variance shown in the breakdown below.\n","uncertainty":{"low":0.06,"high":0.13},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","title":"The top 10 causes of death","publisher":"World Health Organization","source_type":"govt_report","statistic":"Ischaemic heart disease is the world's biggest killer, responsible for 13% of global deaths (~9.1 million) in 2021","excerpt":"\"The world's biggest killer is ischaemic heart disease, responsible for 13% of the world's total deaths. [...] rising by 2.7 million to 9.1 million deaths in 2021. [...] The top global causes of death, in order of total number of lives lost, are associated with two broad topics: cardiovascular (ischaemic heart disease, stroke) and respiratory (COVID-19, chronic obstructive pulmonary disease, lower respiratory infections).\"\n","source_date":"2024-08-07","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165125/https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death","calculation_notes":"9.1 million deaths/year divided by ~6.0 billion adults (age 18+) = 1.52 per 1,000 adults/year. Compounded over 60 adult years: 1 − (1 − 1.52e-3)^60 ≈ 0.087, adjusted to 0.085 for competing mortality. WHO notes ischaemic heart disease has shown \"the largest increase in deaths\" of any leading cause since 2000, driven primarily by population growth and ageing rather than rising age-standardised per-capita risk.\n","independence_note":"WHO Global Health Estimates draw on national vital-registration systems and the IHME Global Burden of Disease modelling pipeline; not fully independent from GBD / AHA statistics that share the same upstream.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)","title":"Cardiovascular diseases (CVDs)","publisher":"World Health Organization","source_type":"govt_report","statistic":"An estimated 19.8 million people died from CVDs in 2022 (~32% of global deaths); 85% of those deaths from heart attack and stroke; >75% occur in low- and middle-income countries","excerpt":"\"An estimated 19.8 million people died from CVDs in 2022, representing approximately 32% of all global deaths. Of these deaths, 85% were due to heart attack and stroke. [...] Over three quarters of CVD deaths take place in low- and middle-income countries.\"\n","source_date":"2024-06-11","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172314/https://www.who.int/news-room/fact-sheets/detail/cardiovascular-diseases-(cvds)","calculation_notes":"Used as the upstream scale check and to justify the regional_breakdown entries. The 19.8M CVD aggregate minus ~6.8M stroke deaths (per WHO top-10 / WSO) leaves roughly 13M deaths from heart disease and other non-stroke CVD, consistent with the 9.1M ischaemic-heart-disease-only figure once hypertensive heart disease, rheumatic heart disease, and cardiomyopathies are added. Used as authoritative cross-check on geographic concentration: >75% of CVD deaths in LMICs drives the East Asia vs Eastern Europe vs US spread in the regional_breakdown.\n","independence_note":"Shares WHO / GBD upstream with the top-10 fact sheet; treat as partially dependent.\n"},{"url":"https://www.cdc.gov/heart-disease/data-research/facts-stats/index.html","title":"Heart Disease Facts","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"In 2023, 919,032 Americans died from cardiovascular disease (1 in 3 US deaths); coronary heart disease killed 371,506 people in 2022","excerpt":"\"In 2023, 919,032 people died from cardiovascular disease. That's the equivalent of 1 in every 3 deaths. [...] Coronary heart disease is the most common type of heart disease. It killed 371,506 people in 2022. [...] About 1 in 20 adults age 20 and older have CAD (about 5%).\"\n","source_date":"2025-01-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172353/https://www.cdc.gov/heart-disease/data-research/facts-stats/index.html","calculation_notes":"~371,500 US coronary heart disease deaths / ~260 million US adults ≈ 1.43 per 1,000 adults/year. Compounded over 60 adult years: 1 − (1 − 1.43e-3)^60 ≈ 0.082. Adding hypertensive and other ischaemic forms brings the US narrow-heart-disease lifetime figure to roughly 0.10-0.12, which is what the regional_breakdown uses for the US entry. The figure is slightly lower than the global adult headline because US crude rates are below the Eastern European / Central Asian peak but well above the East Asian low, so the US sits modestly above the global average once you restrict to ischaemic-heart-disease mortality. Used as an anchor for the US-specific row and for the personal-factor multipliers, which are based on well-established relative risks from Framingham-derived cohorts and subsequent cardiovascular epidemiology.\n","independence_note":"CDC heart disease facts page draws from NVSS/NCHS death-certificate data; shares the GBD analytical pipeline with WHO estimates but applies US-specific age adjustment."}],"comparison_anchors":[{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017},{"label":"Death by lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Global average","probability":0.085,"notes":"WHO 2021 ischaemic heart disease mortality, compounded over 60 adult years"},{"region":"Eastern Europe / Central Asia","probability":0.25,"notes":"Highest CVD mortality rates globally; Russia, Ukraine, Belarus, Kazakhstan dominate"},{"region":"United States","probability":0.11,"notes":"Above global average; rates have been falling since ~1970 but plateaued recently"},{"region":"East Asia (Japan, South Korea)","probability":0.03,"notes":"Low CVD mortality despite industrialisation — diet, blood pressure control, genetics"}],"personal_factor_multipliers":[{"factor":"current smoker","multiplier":2.5,"notes":"Relative risk for fatal CHD vs never-smoker from pooled cohort analyses"},{"factor":"uncontrolled hypertension (SBP >160)","multiplier":2,"notes":"Approximate doubling of CVD mortality vs well-controlled BP"},{"factor":"type 2 diabetes","multiplier":2.5,"notes":"CVD is the leading cause of death in people with T2D"},{"factor":"family history (premature CVD, 1st-degree relative)","multiplier":1.7,"notes":"Parent or sibling with CHD event before age 55 (men) / 65 (women)"},{"factor":"Mediterranean diet + active + non-smoker + normal BP + normal LDL","multiplier":0.3,"notes":"Compounded protective effect observed in large primary-prevention cohorts"}],"short_label":"Heart disease","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry reports ischaemic heart disease mortality only, not the broader cardiovascular disease (CVD) aggregate. The WHO CVD fact sheet puts total CVD deaths at ~19.8 million per year (≈32% of global mortality); adding stroke, hypertensive heart disease, rheumatic heart disease, and the other CVD subtypes would roughly double the headline number and push the lifetime global adult figure closer to 1 in 6. Stroke is covered in its own Likelier entry, which is why this page keeps them separate. The personal_factor_multipliers are illustrative relative risks from the epidemiological literature, not a calibrated personal risk calculator — the real multipliers interact (smoking and hypertension are not independent) and are strongly age-dependent. For a formal personal estimate, clinical tools such as the AHA PREVENT equations or the European SCORE2 charts are the appropriate instrument. The regional_breakdown numbers are order-of-magnitude anchors drawn from WHO / GBD age-standardised mortality rates, not exact lifetime probabilities for any individual.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single thin line tracing a flat heartbeat trace against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/heart-disease-death","api_url":"https://likelier.app/api/fears/heart-disease-death.json"},{"slug":"medical-malpractice-death","question":"What are the odds of dying from a medical error?","quick_answer":"Roughly 1 in 12 that a US adult's eventual death is attributable, in whole or part, to a preventable medical error (~250,000 deaths a year). Most people underestimate this badly, guessing somewhere around 1 in 10,000 to 1 in 100,000.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Medical errors occupy an unusual place in public risk perception: most people substantially underestimate their frequency. Unlike plane crashes or shark attacks, hospital errors lack a vivid media archetype — they happen behind closed doors, are rarely coded on death certificates, and are diffused across millions of encounters. Gallup surveys on healthcare confidence show that while Americans express general trust in their doctors, few spontaneously rank medical error among leading causes of death. The 2016 Makary & Daniel BMJ paper repositioning medical errors as the \"third leading cause of death\" generated widespread media attention precisely because it contradicted public intuition.\n","rough_estimate":"~1 in 10,000 to 1 in 100,000 lifetime feels about right to most people","kind":"intuition"},"native":{"display":"~250,000 deaths per year from medical errors in the US","numerator":250000,"denominator":3090000,"unit":"per year (fraction of all US deaths)","population":"US adults receiving medical care (~330 million population, ~3.09 million annual deaths)"},"normalized":{"lifetime_us_adult":0.085,"display":"~1 in 12 lifetime probability of death being attributable to medical error","log_value":-1.07,"assumptions":"Makary & Daniel (2016, BMJ) estimated ~251,454 deaths per year from medical errors, extrapolating from studies of preventable adverse events in hospitals. James (2013, Journal of Patient Safety) estimated 210,000-440,000. The Institute of Medicine's 1999 \"To Err Is Human\" used the lower figure of 44,000-98,000. The central estimate of ~250,000 is the most widely cited figure. Against ~3.09 million total US deaths per year (CDC 2023), medical errors would account for roughly 8.1% of all deaths. Since everyone dies exactly once, this translates to approximately an 8.5% lifetime probability that one's eventual death will be attributable (in whole or in part) to a preventable medical error. The uncertainty band is wide because: (1) the definition of \"medical error\" varies across studies, (2) the Makary figure is an extrapolation from a small number of studies, not a direct count, and (3) many errors contribute to death without being the sole cause. Calculation: 250,000 / 3,090,000 ≈ 0.081; adjusted upward slightly to 0.085 to account for errors in outpatient and non-hospital settings not captured in the hospital-focused studies.\n","uncertainty":{"low":0.015,"high":0.14},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/27143499/","title":"Medical error—the third leading cause of death in the US","publisher":"BMJ (British Medical Journal)","source_type":"peer_reviewed","statistic":"An estimated 251,454 deaths per year in the US stem from medical error, making it the third leading cause of death behind heart disease and cancer","excerpt":"\"Based on a total of 35,416,020 hospitalizations, the researchers calculated that 251,454 deaths stemmed from medical error, translating to 9.5 percent of all deaths each year in the U.S.\"\n","source_date":"2016-05-03","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260310192810/https://pubmed.ncbi.nlm.nih.gov/27143499/","calculation_notes":"Makary & Daniel extrapolated from four studies (Healthgrades 2004, OIG 2010, AHRQ 2008, Landrigan 2010) that measured preventable adverse event rates in hospitalized patients. They applied these rates to total US hospitalizations (~35.4 million/year) to estimate ~251,000 deaths. This is not a direct count — it is a modeled extrapolation. The figure has been criticized for conflating \"contributed to death\" with \"caused death\" and for applying rates from high-acuity samples to all hospitalizations. Nevertheless, it remains the most widely cited estimate and was published in a major peer-reviewed journal. Native rate: 251,454 / 3,090,000 total US deaths ≈ 0.081. Lifetime interpretation: ~8.5% probability that any given US adult's death will involve a preventable medical error.\n"},{"url":"https://journals.lww.com/journalpatientsafety/fulltext/2013/09000/a_new,_evidence_based_estimate_of_patient_harms.2.aspx","title":"A New, Evidence-based Estimate of Patient Harms Associated with Hospital Care","publisher":"Journal of Patient Safety","source_type":"peer_reviewed","statistic":"Between 210,000 and 440,000 patients per year suffer some type of preventable harm that contributes to their death in hospitals","excerpt":"\"Using a weighted average of the 4 studies, a lower limit of 210,000 deaths per year was associated with preventable harm in hospitals. Given limitations in the search capability of the Global Trigger Tool and the incompleteness of medical records on which the Tool depends, the true number of premature deaths associated with preventable harm to patients was estimated at more than 400,000 per year.\"\n","source_date":"2013-09-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260127035405/https://journals.lww.com/journalpatientsafety/Fulltext/2013/09000/A_New,_Evidence_based_Estimate_of_Patient_Harms.2.aspx","calculation_notes":"James (2013) used the Global Trigger Tool methodology across four studies to estimate preventable hospital deaths. His upper bound of 440,000 is substantially higher than Makary's 251,000 and reflects broader inclusion criteria. The lower bound of 210,000 aligns more closely with Makary. This range (210,000-440,000) drives the uncertainty band: 210,000/3,090,000 ≈ 0.068 at the low end; 440,000/3,090,000 ≈ 0.142 at the high end. The central estimate of 250,000 sits near the lower bound of James's range and the central estimate of Makary's.\n","independence_note":"James's study uses different underlying data (Global Trigger Tool reviews) from Makary & Daniel's extrapolation approach. The two estimates converge on a similar order of magnitude through independent methodologies.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/patient-safety","title":"Patient Safety Fact Sheet","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"1 in every 10 patients is harmed in health care; more than 3 million deaths occur annually due to unsafe care globally","excerpt":"\"Around 1 in every 10 patients is harmed in health care and more than 3 million deaths occur annually due to unsafe care. The risk of patient death occurring due to a preventable medical accident while receiving health care is estimated to be 1 in 300.\"\n","source_date":"2023-09-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420043412/https://www.who.int/news-room/fact-sheets/detail/patient-safety","calculation_notes":"WHO's global figure of 3 million deaths from unsafe care is broadly consistent with the US estimate when scaled by population. The US (~330 million) is roughly 4% of the global population (~8 billion); 4% of 3 million = 120,000, which is lower than the US estimates but reflects lower healthcare utilization intensity in many countries. The WHO's \"1 in 300\" per-encounter risk figure is not directly comparable to the lifetime figure used here but provides independent corroboration that medical error is a major cause of preventable death.\n"},{"url":"https://psnet.ahrq.gov/perspective/measuring-and-responding-deaths-medical-errors","title":"Measuring and Responding to Deaths From Medical Errors","publisher":"Agency for Healthcare Research and Quality (AHRQ)","source_type":"govt_report","statistic":"Estimates range from 44,000-98,000 (IOM 1999) to 250,000+ (Makary 2016); measurement remains contested","excerpt":"\"Estimates of the population toll of deaths from medical error are extrapolations from individual studies in which there were very few deaths. While there is general consensus about the frequency of preventable harm in hospitals, the number of deaths that directly results from these preventable adverse events is controversial.\"\n","source_date":"2023-01-15","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420043451/https://psnet.ahrq.gov/perspective/measuring-and-responding-deaths-medical-errors","calculation_notes":"AHRQ's PSNet perspective explicitly acknowledges the measurement controversy. The IOM's original 1999 estimate (44,000-98,000) would yield a lifetime fraction of 1.4%-3.2% — still substantial but much lower than the Makary/James estimates. The AHRQ commentary does not endorse a specific number but notes \"general consensus about the frequency of preventable harm.\" This entry uses the Makary figure as the central estimate because it is the most widely cited in the peer-reviewed literature, while the uncertainty band (1.5%-14%) spans from the IOM lower bound to the James upper bound.\n"}],"comparison_anchors":[{"label":"Dying in a car accident (lifetime, US adult)","lifetime_us_adult":0.0089},{"label":"Dying from heart disease (lifetime, US)","lifetime_us_adult":0.2},{"label":"Dying from cancer (lifetime, US)","lifetime_us_adult":0.18}],"personal_factor_multipliers":[{"factor":"Hospitalized elderly patient (75+)","multiplier":2.5,"notes":"Older patients have more complex care, more medications, and higher vulnerability to adverse events"},{"factor":"Surgical patient","multiplier":1.8,"notes":"Surgical adverse events account for a disproportionate share of preventable deaths"},{"factor":"Healthy adult, minimal healthcare contact","multiplier":0.3,"notes":"Less exposure to the healthcare system means fewer opportunities for error"},{"factor":"Patient at teaching hospital","multiplier":1.3,"notes":"Some studies find modestly higher adverse event rates at teaching hospitals, though these hospitals also treat sicker patients"}],"short_label":"Medical error death","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The headline figure of ~250,000 deaths per year is an extrapolation, not a direct count. Medical errors are not listed on death certificates, and no surveillance system directly tracks them. The Makary & Daniel estimate has been criticized on methodological grounds: it applies adverse-event rates from small, high-acuity samples to the entire hospitalized population, and it counts cases where error \"contributed to\" death rather than cases where error was the proximate cause. The IOM's original 1999 estimate was 3-5x lower. The \"lifetime probability\" framing here is unusual — it asks what fraction of all deaths involve preventable error, which is not the same as asking \"what is my risk of being killed by a medical error on my next hospital visit.\" The per-encounter risk is much lower. Readers should also note that healthcare simultaneously prevents far more deaths than it causes through error; the net effect of medical care on life expectancy is overwhelmingly positive.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A hospital bed in an empty room with a clipboard on the side table, flat vector illustration, muted tones, no people."},"canonical_url":"https://likelier.app/medical-malpractice-death","api_url":"https://likelier.app/api/fears/medical-malpractice-death.json"},{"slug":"compulsive-sexual-behavior","question":"What are the odds of developing compulsive sexual behavior disorder?","quick_answer":"Roughly 1 in 12 US adults (about 8.6%) report clinically relevant distress or impairment from difficulty controlling sexual urges, several times higher than the 1-2% most people assume. Stricter diagnostic criteria would place the figure lower.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Compulsive sexual behavior tends to be perceived through two contradictory lenses: either dismissed as a non-clinical invention (the \"sex addiction\" skepticism popularized by critics of the concept), or associated only with a narrow stereotype of high-frequency male behavior. Neither framing accurately captures the clinical picture. The DSM-5 explicitly declined to include Hypersexual Disorder in 2013, citing insufficient evidence — a decision that shaped public and clinical perception for a decade. ICD-11's 2022 addition of Compulsive Sexual Behaviour Disorder (CSBD, code 6C72) recalibrated the diagnosis as an impulse-control disorder rather than a behavioral addiction, but this reclassification has not yet filtered broadly into popular awareness. Most people significantly underestimate how many adults experience distressing and functionally impairing difficulty controlling sexual urges.\n","rough_estimate":"~1-2% of adults","kind":"intuition"},"native":{"display":"~8.6% of US adults (ages 18-50) report clinically relevant distress or impairment from difficulty controlling sexual urges (Dickenson et al., 2018, JAMA Network Open)","numerator":8.6,"denominator":100,"unit":"share of adults aged 18-50 endorsing clinically relevant distress or impairment (CSBI ≥35)","population":"US adults aged 18-50, nationally representative (National Survey of Sexual Health and Behavior, n=2,325)"},"normalized":{"lifetime_us_adult":0.086,"display":"~1 in 12 US adults experiences clinically relevant distress from difficulty controlling sexual behavior","log_value":-1.07,"assumptions":"Dickenson et al. (2018, JAMA Network Open) found that 8.6% of a nationally representative US sample aged 18-50 endorsed clinically relevant levels of distress and/or impairment using the Compulsive Sexual Behavior Inventory (CSBI ≥35). This is a point-prevalence figure for the peak-incidence age window (18-50), not a formal lifetime diagnosis under ICD-11 CSBD criteria. Using this figure as a proxy for lifetime prevalence is conservative in one direction (point prevalence underestimates cumulative lifetime exposure) but generous in another (a minority of those above the CSBI threshold would meet the full ICD-11 CSBD criteria requiring 6+ months duration, repeated failure to control urges, and marked functional impairment). Stricter ICD-11 CSBD criteria, as reviewed by Kraus et al. (2018, World Psychiatry), yield prevalence estimates of 1-3% in adults. The headline figure (8.6%) represents the more inclusive \"distress or impairment\" threshold; the ICD-11 CSBD rate would be roughly 0.01-0.03 at the stricter end. Uncertainty bounds span the stricter-criteria lower bound (0.02) to a plausible lifetime-inclusive upper bound (~0.15), reflecting the broad range across measurement instruments and criteria.\n","uncertainty":{"low":0.02,"high":0.15},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6324590/","title":"Prevalence of Distress Associated With Difficulty Controlling Sexual Urges, Feelings, and Behaviors in the United States","publisher":"JAMA Network Open / PMC","source_type":"peer_reviewed","statistic":"8.6% of US adults aged 18-50 endorsed clinically relevant distress or impairment (CSBI ≥35); 10.3% of men and 7.0% of women","excerpt":"\"Among the total sample, 8.6% (7.0% women and 10.3% men) endorsed clinically relevant levels of distress and/or impairment associated with difficulty controlling sexual feelings, urges, and behaviors as measured by the CSBI.\"\n","source_date":"2018-11-09","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505052015/https://pmc.ncbi.nlm.nih.gov/articles/PMC6324590/","calculation_notes":"Primary prevalence figure. The CSBI (Compulsive Sexual Behavior Inventory) threshold of 35 or higher on a 0-65 scale indicated clinically relevant distress/impairment. Sample was randomly drawn from the National Survey of Sexual Health and Behavior (n=2,325 adults aged 18-50 across all 50 US states). We use 8.6/100 as the native figure. For normalization, this point-prevalence figure for the 18-50 window is treated as a reasonable proxy for adult lifetime prevalence; the 18-50 age band captures the peak incidence years, so lifetime risk for a US adult is plausibly in this range or modestly higher.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5775124/","title":"Compulsive sexual behaviour disorder in the ICD-11","publisher":"World Psychiatry / PMC","source_type":"peer_reviewed","statistic":"Prevalence estimates of 1-3% in adults using stricter diagnostic criteria consistent with ICD-11 CSBD","excerpt":"\"Recent studies have produced estimates of compulsive sexual behaviour disorder of 1 to 3% in adults. CSBD is characterized by a persistent pattern of failure to control intense, repetitive sexual impulses or urges resulting in repetitive sexual behaviour over an extended period (e.g., 6 months or more).\"\n","source_date":"2018-02-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260518041300/https://pmc.ncbi.nlm.nih.gov/articles/PMC5775124/","calculation_notes":"Kraus et al. (2018) provides the ICD-11 6C72 definitional framework and a stricter prevalence window of 1-3%. This serves as the lower bound of the uncertainty range and provides the ICD-11 clinical context. The 1-3% estimate (0.01-0.03) is used as the uncertainty.low anchor at 0.02 (midpoint), while Dickenson's 8.6% point-prevalence for the broad distress/impairment threshold anchors the headline.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30646355/","title":"Prevalence of Distress Associated With Difficulty Controlling Sexual Urges, Feelings, and Behaviors in the United States — PubMed","publisher":"JAMA Network Open / PubMed","source_type":"peer_reviewed","statistic":"8.6% of nationally representative US adults (ages 18-50) reported clinically relevant levels of distress or impairment from difficulty controlling sexual urges","excerpt":"\"Among the total sample, 8.6% endorsed clinically relevant levels of distress and/or impairment associated with difficulty controlling sexual feelings, urges, and behaviors.\"\n","source_date":"2018-11-09","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260525094349/https://pubmed.ncbi.nlm.nih.gov/30646355/","calculation_notes":"Secondary citation to PubMed record for Dickenson et al. (2018), confirming the 8.6% figure in a nationally representative US sample. No additional arithmetic beyond what is noted in the primary PMC source above.\n"}],"comparison_anchors":[{"label":"Gambling disorder (lifetime, US)","lifetime_us_adult":0.025},{"label":"Compulsive buying disorder (global adults)","lifetime_us_adult":0.049},{"label":"Major depressive episode (lifetime, US)","lifetime_us_adult":0.2}],"personal_factor_multipliers":[{"factor":"male","multiplier":1.5,"notes":"Men endorse CSBD indicators at roughly 1.5x the rate of women across most studies; the ICD-11 CSBD literature consistently reports male predominance"},{"factor":"history of trauma or adverse childhood experiences","multiplier":2.5,"notes":"Childhood trauma is among the strongest predictors of compulsive sexual behavior in adults; co-occurring PTSD substantially elevates risk"},{"factor":"current substance use disorder","multiplier":2,"notes":"CSBD co-occurs with substance use disorders at high rates, sharing neurobiological reward-pathway mechanisms"}],"short_label":"Compulsive sexual behavior","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The 8.6% figure uses a broad distress/impairment threshold (CSBI ≥35) rather than formal ICD-11 CSBD criteria, which require 6+ months duration, repeated failure to control urges, and marked functional impairment. Stricter criteria yield 1-3% prevalence. The DSM-5 explicitly rejected Hypersexual Disorder in 2013; CSBD entered ICD-11 (6C72) in 2022 as an impulse-control disorder, not a behavioral addiction — the classification matters for insurance reimbursement and clinical treatment models. Prevalence is highly measurement-instrument-dependent; the CSBI, the Hypersexual Behavior Inventory, the Sexual Addiction Screening Test, and the Bergen Social Media Addiction Scale analogs for sexual behavior produce different prevalence estimates. Male-female reporting differences may partly reflect sociocultural differences in willingness to report distress rather than true incidence differences. No long-term US longitudinal studies track cumulative lifetime CSBD incidence; all estimates are cross-sectional.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"Abstract representation of a broken loop and an hourglass, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/compulsive-sexual-behavior","api_url":"https://likelier.app/api/fears/compulsive-sexual-behavior.json"},{"slug":"eating-disorder-diagnosis","question":"What is the lifetime risk of developing an eating disorder?","quick_answer":"The lifetime risk of developing an eating disorder is about 1 in 11, roughly 9 percent of US adults. Most people assume eating disorders affect only 1 to 2 percent of the population.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Eating disorders are widely perceived as rare, affecting a narrow demographic of young, affluent, white women. This stereotype is empirically wrong on every axis: eating disorders affect all genders, ages, races, and socioeconomic groups, and the aggregate lifetime prevalence (~9%) is far higher than most adults would guess. Public awareness campaigns have made anorexia nervosa the most recognisable eating disorder, but binge eating disorder is roughly three times more prevalent. The perception that eating disorders are primarily about vanity rather than psychopathology contributes to both underdiagnosis and underfunding relative to their mortality burden.\n","rough_estimate":"Most adults assume eating disorders affect 1-2% of the population; the true figure is roughly 9%","kind":"intuition"},"native":{"display":"~9% of the US population (~31 million) will have an eating disorder in their lifetime","numerator":9,"denominator":100,"unit":"lifetime","population":"US population, all genders"},"normalized":{"lifetime_us_adult":0.09,"display":"~1 in 11 lifetime (US adults)","log_value":-1.05,"assumptions":"NEDA cites approximately 9% of the US population (roughly 31 million Americans) as the lifetime prevalence for any eating disorder. ANAD provides concordant figures. The overall lifetime prevalence is estimated at 8.60% among females and 4.07% among males. The 9% figure aggregates anorexia nervosa, bulimia nervosa, binge eating disorder, ARFID, and other specified/unspecified eating disorders. Point estimate of 0.09 used directly as lifetime prevalence. Uncertainty band reflects variation across studies and diagnostic criteria: lower bound from studies applying narrowest DSM-5 criteria (~0.05), upper bound from studies including subthreshold presentations and adjustment for known under-diagnosis (~0.15). The point estimate of 0.09 sits within this range; the 3x ratio reflects genuine methodological spread across studies.\n","uncertainty":{"low":0.05,"high":0.15},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nationaleatingdisorders.org/statistics/","title":"Eating Disorder Statistics","publisher":"National Eating Disorders Association","source_type":"reputable_reference","statistic":"9% of the US population (~31 million) will have an eating disorder in their lifetime; 10,200 deaths per year","excerpt":"\"9% of the US population, or nearly 31 million Americans will have an eating disorder in their lifetime. [...] 10,200 deaths each year are the direct result of an eating disorder — that's one death every 52 minutes.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260517091818/https://www.nationaleatingdisorders.org/statistics/","calculation_notes":"NEDA is the largest US eating-disorders advocacy and research organisation. The 9% lifetime prevalence figure is drawn from multiple epidemiological studies, including the WHO World Mental Health Survey Initiative. The 10,200 annual death figure includes deaths directly attributable to eating disorders (cardiac arrest, organ failure, suicide) but not indirect contributions to mortality. The 31 million figure is ~9% of the current US population of ~345 million.\n","independence_note":"NEDA compiles statistics from multiple primary research studies and government sources. Not a primary data producer itself; functions as an authoritative aggregator of eating-disorder epidemiology.\n"},{"url":"https://anad.org/learning-library/eating-disorder-statistic/","title":"Eating Disorder Statistics","publisher":"National Association of Anorexia Nervosa and Associated Disorders","source_type":"reputable_reference","statistic":"Anorexia has the highest case mortality rate and second-highest crude mortality rate of any mental illness","excerpt":"\"10,200 deaths each year are the direct result of an eating disorder—that's one death every 52 minutes. [...] Anorexia has the highest case mortality rate and second-highest crude mortality rate of any mental illness.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260505053932/https://anad.org/learning-library/eating-disorder-statistic/","calculation_notes":"ANAD corroborates the NEDA statistics and adds the mortality-ranking context. The standardized mortality ratio for anorexia nervosa is 5.21 (i.e., women with anorexia die at 5.2x the rate of age-matched peers), which is the highest of any mental illness. Bulimia nervosa has an SMR of 2.20 and binge eating disorder 1.46. The \"highest case mortality rate\" claim is supported by meta-analyses of eating-disorder mortality published in the International Journal of Eating Disorders and Archives of General Psychiatry.\n","independence_note":"ANAD is a separate organisation from NEDA, but both cite the same underlying epidemiological studies (notably the WHO World Mental Health Survey Initiative and meta-analyses published in the International Journal of Eating Disorders). The 9% lifetime prevalence and 10,200 annual deaths figures appear on both sites because they originate from the same primary research, not from independent data collection. ANAD provides additional mortality-ranking context not found on the NEDA page.\n"},{"url":"https://www.nationaleatingdisorders.org/eating-disorders-and-suicide/","title":"Eating Disorders and Suicide","publisher":"National Eating Disorders Association","source_type":"reputable_reference","statistic":"Individuals with anorexia are 18 times more likely to die by suicide than the general population; individuals with bulimia are 7 times more likely","excerpt":"\"Individuals with anorexia are 18 times more likely to die by suicide and those struggling with bulimia are 7 more times likely to die by suicide.\"\n","source_date":"2023-10-18","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260613164107/https://www.nationaleatingdisorders.org/eating-disorders-and-suicide/","calculation_notes":"Supports the body-text claim of an elevated suicide-specific risk in anorexia, distinct from the overall standardized mortality ratio (SMR 5.21, all-cause) already cited from ANAD above.\n","independence_note":"Same publisher (NEDA) as the primary lifetime-prevalence source above, but a distinct page addressing suicide-specific risk rather than overall prevalence; the underlying figure traces to the eating- disorder mortality literature synthesized by both NEDA and ANAD.\n"}],"comparison_anchors":[{"label":"Major depressive disorder (lifetime, US)","lifetime_us_adult":0.21},{"label":"Anxiety disorder (lifetime, US)","lifetime_us_adult":0.31},{"label":"Substance use disorder (lifetime, US)","lifetime_us_adult":0.15},{"label":"Schizophrenia (lifetime, global)","lifetime_us_adult":0.007}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":2,"notes":"Lifetime prevalence ~8.6% in females vs ~4.1% in males; the gap is narrower than historically assumed"},{"factor":"Adolescence / young adulthood (ages 12-25)","multiplier":3,"notes":"Peak onset occurs during adolescence; the majority of eating disorders develop before age 25"},{"factor":"History of dieting","multiplier":2,"notes":"Chronic dieting is one of the strongest behavioral predictors of developing an eating disorder"},{"factor":"LGBTQ+ identity","multiplier":2,"notes":"LGBTQ+ individuals face elevated rates across all eating disorder subtypes; transgender individuals are particularly affected"},{"factor":"Competitive athletics / aesthetic sports","multiplier":2.5,"notes":"Athletes in sports emphasizing leanness (gymnastics, wrestling, distance running, ballet) have substantially elevated rates"}],"short_label":"Eating disorder","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The 9% lifetime prevalence aggregates multiple distinct conditions — anorexia nervosa, bulimia nervosa, binge eating disorder, ARFID, and other specified or unspecified eating disorders — that have different risk profiles, demographic patterns, and mortality rates. Anorexia nervosa (~0.9% lifetime in women) has the highest case mortality of any mental illness, but binge eating disorder (~3.5% lifetime) is far more prevalent. The 10,200 annual deaths figure is an estimate that includes direct medical and suicide deaths attributable to eating disorders; the true toll may be higher because eating disorders are underreported on death certificates. Treatment rates are low: fewer than half of people with eating disorders receive treatment, and the average delay from onset to treatment is roughly 5-7 years. A note on social media and adolescents: eating-disorder-related health visits among US under-17s more than doubled between 2018 and 2022 (Pastore et al., Journal of Pediatrics, 2023), and cross-sectional work links appearance-focused social-media use to higher eating disorder pathology in adolescent girls (Dahlgren et al., Journal of Eating Disorders, 2024). These are associational and trend data, not a measured per-person relative risk, so no social-media multiplier is listed among the personal factors — the evidence base does not yet support a defensible fold-increase figure.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":3,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A pale shape with an irregular missing piece on a muted sage-grey background, flat vector illustration suggesting absence."},"canonical_url":"https://likelier.app/eating-disorder-diagnosis","api_url":"https://likelier.app/api/fears/eating-disorder-diagnosis.json"},{"slug":"hip-replacement-lifetime","question":"What are the lifetime odds of needing a hip replacement?","quick_answer":"The lifetime odds of needing a total hip replacement are about 1 in 11 for a US adult (higher for women, ~10-12%, than men, ~6-8%). Because it is an elective, planned surgery rather than an emergency, most people underrate how common it is — many guess closer to 1 in 20.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Most people think of hip replacement as something that happens to a small minority of elderly patients — a dramatic last resort for severe arthritis or a fall. Awareness of how common the procedure is tends to be low, particularly among younger adults. Because hip replacement is an elective, planned surgery rather than an emergency, it does not generate the same fear as acute events, and its high frequency relative to other major surgeries is not widely appreciated.\n","rough_estimate":"~1 in 20 lifetime feels about right to most people","kind":"intuition"},"native":{"display":"~5.26% of US adults age 80 have had a total hip replacement","numerator":526,"denominator":10000,"unit":"lifetime prevalence by age 80","population":"US adults, Olmsted County population-based study (Maradit Kremers 2015)"},"normalized":{"lifetime_us_adult":0.09,"display":"~1 in 11 lifetime (US adult)","log_value":-1.05,"assumptions":"Maradit Kremers et al. (2015, J Bone Joint Surg Am) report a prevalence of 5.26% for total hip arthroplasty (THA) among US adults reaching age 80. This figure represents the cumulative incidence through age 80. A UK population-based study (Culliford et al. 2012) estimated the mortality-adjusted lifetime risk of hip replacement as 11.6% for women and 7.1% for men at age 50 — similar in methodology but using British procedure rates. The Maradit Kremers 2015 data are preferred for a US-specific estimate. Given that the Olmsted County cohort shows 5.26% prevalence by age 80 and that incidence continues through the remaining years of life (80+), the lifetime risk extends somewhat beyond 5.26%. Applying a conservative adjustment for post-80 incidence and accounting for increasing US THA volumes since the study period, a lifetime estimate of ~9% (1 in 11) is reasonable for a US adult born today. Women have higher lifetime risk (~10–12%) than men (~6–8%). Uncertainty range 0.06–0.13 reflects sex differences and increasing procedure rates.\n","uncertainty":{"low":0.06,"high":0.13},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26333733/","title":"Prevalence of Total Hip and Knee Replacement in the United States","publisher":"Maradit Kremers H et al., J Bone Joint Surg Am","source_type":"peer_reviewed","statistic":"Prevalence of total hip replacement (THA) in the US population was 0.83% overall in 2010; by age 80 it reached 5.26%. Women had higher prevalence than men at all ages.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] The 2010 prevalence of total hip and total knee replacement among the total U.S. population was 0.83% and 1.52%, respectively, with prevalence being higher among women than among men and increasing with age, reaching 5.26% for total hip replacement and 10.38% for total knee replacement at eighty years.\"\n","source_date":"2015-09-02","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20251202023949/https://pubmed.ncbi.nlm.nih.gov/26333733/","calculation_notes":"Prevalence at age 80 = 5.26% for THA. This represents cumulative incidence through age 80 among the US population. Because THA incidence continues after age 80 and because annual THA volumes have increased since 2010, the lifetime risk from age 18 to death is estimated at ~9%. The Olmsted County cohort is a well-characterized, racially diverse population that closely tracks US national demographics.\n","independence_note":"Maradit Kremers 2015 used the Rochester Epidemiology Project (Olmsted County, MN) population-based medical records linkage — entirely independent of the AAOS registry administrative data, which captures only subset of procedures at participating centers.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22395038/","title":"The lifetime risk of total hip and knee arthroplasty: results from the UK general practice research database","publisher":"Culliford D et al., Osteoarthritis and Cartilage","source_type":"peer_reviewed","statistic":"Mortality-adjusted lifetime risk of total hip replacement at age 50: 11.6% for women, 7.1% for men (UK general practice database, 2005 procedure rates).\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] The estimated mortality-adjusted lifetime risk of total hip replacement at age 50 for the year 2005 was 11.6% for women and 7.1% for men. The lifetime risk at age 50 years of undergoing hip replacement is approximately 11% for women and 7% for men.\"\n","source_date":"2012-03-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250206112153/https://pubmed.ncbi.nlm.nih.gov/22395038/","calculation_notes":"UK lifetime risks are somewhat lower than projected US risks (higher US obesity rates and more procedure utilization). Used to validate the sex-disaggregated direction of effect and magnitude. Not used as primary point estimate.\n","independence_note":"UK General Practice Research Database draws on primary-care electronic records from England and Wales, entirely separate from the US Olmsted County cohort and AAOS registry data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12192333/","title":"Highlights of the 2024 American Joint Replacement Registry Annual Report","publisher":"American Academy of Orthopaedic Surgeons (AAOS)","source_type":"reputable_reference","statistic":"AJRR has captured 4.3 million hip and knee arthroplasty procedures from 2012 to 2023 across 1,447 member sites; primary THA comprises 32.4% of captured procedures.\n","excerpt":"\"The 2024 AJRR Annual Report contains 3,715,320 validated primary and revision THA and TKA procedures performed during years 2012 to 2023, with primary THA comprising 32.4% of procedures captured. The AJRR remains the largest orthopaedic and joint arthroplasty registry in the world by annual procedure volume.\"\n","source_date":"2024-11-12","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525161818/https://pmc.ncbi.nlm.nih.gov/articles/PMC12192333/","calculation_notes":"32.4% of 3,715,320 validated procedures over 12 years = ~1.2 million THAs captured in the registry over that period. Used to confirm ongoing high annual volume and the increasing utilization trend. Not used as the primary probability estimate.\n","independence_note":"AAOS AJRR is a voluntary registry of participating US surgical centers using administrative and clinical data. Methodologically independent of the population-based Olmsted County cohort and UK general practice database.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db186.htm","title":"Hospitalization for Total Hip Replacement Among Inpatients Aged 45 and Over: United States, 2000-2010","publisher":"National Center for Health Statistics (Wolford ML, Palso K, Bercovitz A), NCHS Data Brief No. 186","source_type":"govt_report","statistic":"Share of US inpatient total hip replacements performed on patients aged 45-54 rose from 12% in 2000 to 17% in 2010; share performed on patients aged 55-64 rose from 24% to 29% over the same decade.\n","excerpt":"\"In 2000, 12% of all inpatient total hip replacements were performed on those aged 45-54; this percentage increased to 17% in 2010.\" \"The percentage of total hip replacements performed on those aged 55-64 also increased (from 24% to 29%).\"\n","source_date":"2015-02-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260513070233/https://www.cdc.gov/nchs/products/databriefs/db186.htm","calculation_notes":"NCHS National Hospital Discharge Survey data confirming the shift toward younger THA patients referenced in the body text. Used as supporting context for the declining-average-age trend; not used in the headline lifetime probability calculation, which is drawn from Maradit Kremers 2015 and Culliford 2012 above.\n","independence_note":"Independent US government hospital-discharge survey data source, separate from the Olmsted County cohort, UK general practice database, and AAOS AJRR registry.\n"}],"comparison_anchors":[{"label":"Knee replacement (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Cataract surgery (lifetime, US adult 65+)","lifetime_us_adult":0.5},{"label":"Hip fracture (lifetime, US adult)","lifetime_us_adult":0.15}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.4,"notes":"Maradit Kremers 2015 and Culliford 2012 consistently show women have ~30–60% higher lifetime THA risk than men; attributed to greater OA prevalence in women and higher rates of hip dysplasia.\n"},{"factor":"Obesity (BMI > 30)","multiplier":2,"notes":"Large cohort studies show BMI > 32 kg/m² roughly doubles THA risk vs normal BMI (RR 2.3–3.4); obesity accelerates hip osteoarthritis through both mechanical load and systemic inflammation.\n"},{"factor":"Developmental hip dysplasia","multiplier":3,"notes":"DDH is the underlying cause of approximately 21–29% of all THA procedures in patients under age 60. Adults with untreated DDH have substantially earlier onset of hip OA requiring arthroplasty.\n"},{"factor":"Physically demanding occupation (heavy lifting, prolonged standing)","multiplier":1.5,"notes":"Epidemiological studies show occupations with heavy physical load are associated with ~1.5× increased risk of hip OA and subsequent arthroplasty compared to sedentary work.\n"},{"factor":"Regular high-impact sport (distance running, competitive soccer)","multiplier":1.3,"notes":"Recreational runners and team sport athletes face modestly elevated hip OA risk compared to non-athletes at population level; effect is smaller than for knee OA and offset partially by body weight benefits.\n"}],"short_label":"Hip replacement","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 9% lifetime estimate reflects procedure utilization and will rise as the US population ages and as obesity rates remain elevated; Kurtz et al. (2007) projected THA volumes could reach ~572,000 procedures per year by 2030. The Olmsted County cohort (Maradit Kremers 2015) is the most rigorous US population-based source but was collected in 2010 — annual THA volumes have increased substantially since then. The UK lifetime risk estimates (Culliford 2012: 11% women, 7% men) are based on lower US-equivalent procedure rates and provide a useful lower bound. The 9% point estimate is best understood as the current-generation estimate for a US adult reaching age 80; younger adults born today may face a somewhat higher lifetime risk due to increasing obesity prevalence and expanding surgical indications. This entry covers primary (first-time) THA only; revision arthroplasty is a separate event.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simple cross-section diagram of a hip joint with a prosthetic implant, flat vector illustration."},"canonical_url":"https://likelier.app/hip-replacement-lifetime","api_url":"https://likelier.app/api/fears/hip-replacement-lifetime.json"},{"slug":"kidney-stones-lifetime","question":"What are the odds of getting a kidney stone in your lifetime?","quick_answer":"About 1 in 11 US adults will have a symptomatic kidney stone in their lifetime (roughly 1 in 9 men, 1 in 14 women) — close to the 1-in-10-to-1-in-20 most people intuit. The dread is about pain, not danger: stones are almost never fatal, though about half of first-time formers recur within five to seven years.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most adults have heard at least one vivid first-person account of passing a kidney stone, and the consensus cultural shorthand is \"worst pain of my life\". That framing makes the fear primarily about intensity, not frequency. People generally know kidney stones are common — nobody files them alongside shark attacks or lightning — but the anticipated pain dominates the emotional weighting. There is no large-scale survey quantifying perceived lifetime kidney stone probability as a standalone risk, so the kind here is intuition rather than poll.\n","rough_estimate":"Most adults intuit kidney stones as fairly common — roughly 1 in 10 to 1 in 20 — which is approximately right","kind":"intuition"},"native":{"display":"~1 in 11 US adults will have a symptomatic kidney stone in their lifetime","numerator":1,"denominator":11,"unit":"lifetime","population":"US adults, age-adjusted, both sexes combined"},"normalized":{"lifetime_us_adult":0.09,"display":"~1 in 11 lifetime (US adult)","log_value":-1.05,"assumptions":"Uses the NHANES cross-sectional prevalence data analysed by Scales et al. (2012, European Urology) as the primary anchor: 10.6% self-reported lifetime prevalence in men and 7.1% in women during 2007–2010, up from 6.3% and 4.1% respectively in the 1988–1994 wave. Sex-weighted average for the US adult population is approximately 8.8%, rounded to 0.09 (~1 in 11). This is prevalence of at least one symptomatic episode, not incidence per year. The secular trend is upward — Stamatelou et al. (2003) documented a rising prevalence between 1976 and 1994, and the Scales update confirmed the trend continued through 2010. Lieske et al. (2006, Mayo Clinic/Olmsted County) found female incidence rising while male incidence fell, narrowing the historical sex gap. Uncertainty range 0.07–0.12 reflects the sex-weighted band and ongoing secular increase. The number is symptomatic stones only; asymptomatic stones detected incidentally on imaging are excluded from the NHANES self-report methodology.\n","uncertainty":{"low":0.07,"high":0.12},"scope":"us_adult_lifetime"},"sources":[{"url":"https://doi.org/10.1016/j.eururo.2012.03.052","title":"Prevalence of Kidney Stones in the United States","publisher":"European Urology (Scales et al.)","source_type":"primary_study","statistic":"Overall prevalence of kidney stones was 8.8% (10.6% in men, 7.1% in women) during 2007–2010, up from 5.2% in 1988–1994","excerpt":"\"The prevalence of stone disease in the United States has increased from 5.2% in NHANES III (1988–1994) to 8.8% in NHANES 2007–2010. Men had a higher prevalence of stones than women (10.6% vs 7.1%).\"\n","source_date":"2012-05-17","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426202732/https://www.sciencedirect.com/science/article/pii/S0302283812004046?via%3Dihub","calculation_notes":"Scales et al. analysed NHANES 2007–2010 (n = 12,110) and found 8.8% self-reported lifetime prevalence of kidney stones in US adults. This is the direct anchor for the headline ~1 in 11 figure. Sex-specific rates (10.6% M, 7.1% F) inform the personal_factor_multipliers. The study also documented the secular trend from 5.2% (1988–1994) to 8.8% (2007–2010), confirming that the true current figure may be higher still.\n","independence_note":"NHANES is a nationally representative cross-sectional survey run by NCHS/CDC. This is the primary US epidemiological dataset for kidney stone prevalence and is independent of the clinic-based Mayo/Olmsted County cohort (Lieske et al.).\n"},{"url":"https://www.niddk.nih.gov/health-information/urologic-diseases/kidney-stones/definition-facts","title":"Definition & Facts for Kidney Stones","publisher":"National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK/NIH)","source_type":"govt_report","statistic":"About 11 percent of men and 6 percent of women in the United States have kidney stones at least once during their lifetime","excerpt":"\"About 11 percent of men and 6 percent of women in the United States have kidney stones at least once during their lifetime.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260309073440/https://www.niddk.nih.gov/health-information/urologic-diseases/kidney-stones/definition-facts","calculation_notes":"NIDDK's fact sheet rounds the Scales et al. NHANES figures to 11% (men) and 6% (women). The sex-combined midpoint is ~8.5%, consistent with the 8.8% from the primary study. Used as the authoritative government framing of the same underlying NHANES data.\n","independence_note":"NIDDK republishes NHANES-derived prevalence data; same upstream dataset as the Scales et al. study. Included as the institutional government citation rather than as an independent verification.\n"},{"url":"https://doi.org/10.1038/sj.ki.5000150","title":"Renal stone epidemiology in Rochester, Minnesota: an update","publisher":"Kidney International (Lieske et al.)","source_type":"peer_reviewed","statistic":"From 1970 to 2000 in Rochester, symptomatic stone incidence rose ~1.9%/year in women and fell ~1.7%/year in men, dropping the man-to-woman ratio from 3.1 to 1.3","excerpt":"\"On average, rates for women increased by about 1.9% per year (P=0.064), whereas rates for men declined by 1.7% per year (P=0.019). The overall man to woman ratio decreased from 3.1 to 1.3 during the 30 years (P=0.006).\"\n","source_date":"2006-02-01","source_accessed":"2026-06-30","archive_url":"https://pubmed.ncbi.nlm.nih.gov/16518332/","calculation_notes":"Lieske et al. used the Rochester Epidemiology Project (Olmsted County, MN) to track first-episode symptomatic kidney stone incidence in 1970, 1980, 1990, and 2000. Overall incidence stayed relatively flat, but the sex pattern shifted: age-adjusted female incidence rose from 43.2 to 68.4 per 100,000/year while male incidence fell from 155.1 to 105.0, narrowing the man-to-woman ratio from 3.1 to 1.3. This clinic-based incidence data provides an independent cross-check on the NHANES prevalence trend and on the narrowing sex gap.\n","independence_note":"The Rochester Epidemiology Project is a population-based medical records linkage system in Olmsted County, MN — methodologically independent of the NHANES cross-sectional survey used by Scales et al.\n"},{"url":"https://doi.org/10.1046/j.1523-1755.2003.00917.x","title":"Time Trends in Reported Prevalence of Kidney Stones in the United States: 1976–1994","publisher":"Kidney International (Stamatelou et al.)","source_type":"peer_reviewed","statistic":"Lifetime prevalence of kidney stones increased from 3.8% in 1976–1980 to 5.2% in 1988–1994","excerpt":"\"Disease prevalence among 20- to 74-year-old United States residents was greater in 1988 to 1994 than in 1976 to 1980 (5.2% vs. 3.8%, P < 0.05), greater in males than females, and increased with age in each time period.\"\n","source_date":"2003-05-01","source_accessed":"2026-04-18","archive_url":"https://pubmed.ncbi.nlm.nih.gov/12675858/","calculation_notes":"Stamatelou et al. documented the secular trend from NHANES II through NHANES III, establishing the trajectory that Scales et al. later extended to 8.8% in the 2007–2010 wave. The 37% relative increase across 15 years is consistent with the continued rise through the 2000s.\n","independence_note":"Same upstream NHANES data programme as the Scales et al. analysis, but covering earlier waves (1976–1994 vs 2007–2010). Included for the secular trend, not as an independent prevalence estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5505545/","title":"Renal colic and childbirth pain: female experience versus male perception","publisher":"Journal of Pain Research (Miah et al.)","source_type":"peer_reviewed","statistic":"Among women who had experienced both renal colic and childbirth, 63.3% (12 of 19) rated the kidney stone pain as more painful than childbirth","excerpt":"\"From the female cohort of patients, 19 had given birth previously with 63.3% (n=12) saying that renal colic was more painful and a further 15.7% (n=3) felt that the intensity of pain was similar.\"\n","source_date":"2017-07-05","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260417034018/https://pmc.ncbi.nlm.nih.gov/articles/PMC5505545/","calculation_notes":"Small two-center UK patient survey (59 patients total; 19 women had also given birth and could compare). Used only to support the body-prose statement that reported pain intensity, not frequency, drives the disproportionate dread of kidney stones — not used in the native/normalized probability calculation.\n","independence_note":"A clinical patient-survey on pain perception, methodologically independent of the NHANES/Olmsted County/Rochester Epidemiology Project prevalence datasets used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"Cancer death (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Type 2 diabetes death (lifetime, US adult)","lifetime_us_adult":0.027},{"label":"Choking death (lifetime, US)","lifetime_us_adult":0.0034},{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.0000125}],"regional_breakdown":[{"region":"US men","probability":0.106,"notes":"NHANES 2007–2010; ~1 in 9"},{"region":"US women","probability":0.071,"notes":"NHANES 2007–2010; ~1 in 14; gap narrowing over time"},{"region":"US Southeast ('stone belt')","probability":0.12,"notes":"Higher rates in hot, humid climates; dehydration and dietary factors"},{"region":"Prior stone formers (recurrence within 5–7 years)","probability":0.5,"notes":"~50% recurrence rate without preventive intervention"}],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":1.3,"notes":"10.6% vs 7.1% in NHANES 2007–2010; the gap has narrowed from ~3:1 in the 1970s to ~1.5:1"},{"factor":"Family history of kidney stones","multiplier":2,"notes":"First-degree relative with stones roughly doubles lifetime risk across multiple cohort studies"},{"factor":"Chronic low fluid intake / dehydration","multiplier":1.5,"notes":"Low urine volume is the single most modifiable risk factor; doubling fluid intake reduces recurrence by ~40%"},{"factor":"High-sodium diet (>5 g/day)","multiplier":1.4,"notes":"Excess sodium increases urinary calcium excretion, the dominant driver of calcium oxalate stones (~80% of all stones)"},{"factor":"Obesity (BMI ≥ 30)","multiplier":1.5,"notes":"BMI ≥ 30 associated with ~1.5× risk in NHANES and prospective cohort data; effect stronger in women"},{"factor":"Hot climate residence","multiplier":1.3,"notes":"The US 'stone belt' (Southeast) shows ~30% higher prevalence; heat increases insensible fluid loss"}],"short_label":"Kidney stones","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Kidney stones are almost never fatal — annual US mortality is in the low hundreds, making the death rate negligible. The fear is about pain and recurrence, not survival. The 8.8% headline is symptomatic stones only; asymptomatic stones found incidentally on CT or ultrasound are far more common but clinically irrelevant unless they grow or migrate. The secular trend is upward and likely driven by rising obesity, dietary sodium, and climate warming; the true 2026 prevalence may already exceed 10%. Roughly 80% of stones are calcium oxalate; the remainder (uric acid, strite, cystine) have different risk profiles and recurrence patterns. Recurrence is the dominant clinical concern: about 50% of first-time stone formers will have another episode within 5–7 years without preventive intervention.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A small angular shape resting at the base of a smooth curved form, muted palette, flat vector illustration."},"canonical_url":"https://likelier.app/kidney-stones-lifetime","api_url":"https://likelier.app/api/fears/kidney-stones-lifetime.json"},{"slug":"parent-death-while-child-young","question":"What are the odds of a parent dying or becoming disabled before their child turns 18?","quick_answer":"For a healthy 30-year-old parent, the odds of dying or becoming lastingly disabled before their child turns 18 are roughly 1 in 11 (about 9%). Most young parents treat this as negligibly small, well below the actual combined figure.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"This is the fear that sells life insurance — and the one most people drastically underweight when they decline to buy it. Young parents tend to anchor on their own felt invincibility: \"I'm 30, healthy, and not going anywhere.\" The LIMRA 2025 Insurance Barometer Study found that adults under 30 overestimate the cost of term life insurance by 10-12x, suggesting the risk itself barely registers as real. Meanwhile, only about half of US adults carry any life insurance at all, down from 63 percent in 2011. Disability insurance ownership is even lower — the Council for Disability Awareness reports that fewer than a third of working adults carry individual long-term disability coverage, despite disability being far more common than death during working years. The perception gap here is not that people think the risk is zero — it is that they treat it as negligibly small and therefore not worth insuring against, when the combined probability is high enough to warrant attention.\n","rough_estimate":"Most young parents sense some risk but treat it as negligibly small — well below the actual combined figure","kind":"intuition"},"native":{"display":"~9 in 100 over 18 years (death or lasting disability, 30-year-old parent)","numerator":9,"denominator":100,"unit":"per 18-year parenting window","population":"US 30-year-old parents"},"normalized":{"lifetime_us_adult":0.09,"display":"~1 in 11 (death or lasting disability before child turns 18)","log_value":-1.046,"assumptions":"Starting point: the SSA 2022 period life table (l(x) survivor column) gives cumulative mortality of about 5.76% for a 30-year-old male reaching age 48 (survivors fall from l(30)=96,992 to l(48)=91,405) and about 3.03% for a female of the same age (l(30)=98,415 to l(48)=95,429), a sex-blended ~4.4% for a single randomly-selected parent. The SSA Actuarial Note 2025.6 shows that the probability of surviving from age 20 to normal retirement age (67) without ever being disabled is 66% for men and 71% for women — i.e. roughly 1 in 4 become disabled over that 47-year window. Scaling that down to the narrower 18-year parenting window (age 30 to 48) is a disclosed proportional estimate yielding an approximate 5% qualifying-disability probability; this 18-year figure is not directly stated by SSA. Combining the source-stated mortality (~4.4%) with the estimated lasting-disability leg (~5%) and subtracting partial overlap gives a central estimate of roughly 9% for either outcome for a single parent over 18 years. The uncertainty range reflects sex differences (male risk higher), health variation, and the definitional ambiguity of \"lasting disability.\"\n","uncertainty":{"low":0.05,"high":0.13},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ssa.gov/oact/STATS/table4c6.html","title":"Actuarial Life Table — Period Life Table, 2022","publisher":"Social Security Administration, Office of the Chief Actuary","source_type":"govt_report","statistic":"Cumulative mortality from age 30 to age 48 is approximately 5.76% for males and 3.03% for females (from the l(x) survivor column of the 2022 period life table)","excerpt":"\"Male: age 30 l(x)=96,992, age 48 l(x)=91,405. Female: age 30 l(x)=98,415, age 48 l(x)=95,429. (l(x) is the number of survivors out of 100,000 born alive, from the 2022 period life table for the Social Security area population used in the 2025 Trustees Report.)\"\n","source_date":"2025-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260522221916/https://www.ssa.gov/oact/STATS/table4c6.html","calculation_notes":"The SSA period life table provides a survivor column l(x) — the number of an original 100,000 births still alive at each age. For males, l(30)=96,992 falls to l(48)=91,405, so cumulative mortality from age 30 to 48 is (96,992 − 91,405) / 96,992 = 0.0576, i.e. ~5.76%. For females, l(30)=98,415 falls to l(48)=95,429, giving (98,415 − 95,429) / 98,415 = 0.0303, i.e. ~3.03%. A sex-blended average is approximately 4.4%.\n"},{"url":"https://www.ssa.gov/oact/NOTES/ran6/an2025-6.pdf","title":"Disability and Death Probability Tables for Insured Workers Born in 2005 — Actuarial Note 2025.6","publisher":"Social Security Administration, Office of the Chief Actuary","source_type":"govt_report","statistic":"About 1 in 4 of today's 20-year-olds will become disabled before reaching normal retirement age (67); probability of surviving from age 20 to NRA without disability is 66% for men and 71% for women","excerpt":"\"For those who attain age 20 in 2025, it is projected that the probability of surviving from age 20 to NRA without ever being disabled is 66 percent for men and 71 percent for women.\"\n","source_date":"2025-09-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260312075406/https://www.ssa.gov/OACT/NOTES/ran6/an2025-6.pdf","calculation_notes":"SSA defines disability here as qualifying for Social Security disabled worker benefits — a high bar requiring inability to engage in substantial gainful activity for 12+ months. The source states that the probability of surviving from age 20 to NRA (67) without ever being disabled is 66% for men and 71% for women — i.e. roughly 1 in 4 become disabled over that full 47-year window. For the narrower 18-year window relevant to parenting (age 30 to 48), proportional scaling of that 47-year figure is a disclosed estimate (not directly stated by SSA) yielding an approximate 5% qualifying-disability probability. Combining with the source-stated mortality (~4.4% sex-blended) and subtracting partial overlap gives the ~9% central estimate for death or lasting disability.\n"},{"url":"https://www.nature.com/articles/s41591-024-03343-6","title":"Orphanhood and caregiver death among children in the United States by all-cause mortality, 2000–2021","publisher":"Nature Medicine","source_type":"peer_reviewed","statistic":"An estimated 2.91 million US children (4.2% of children under 18) had experienced the death of a parent or caregiver as of 2021, with incidence increasing 49.5% since 2000","excerpt":"\"In 2021, an estimated 2.91 million children (4.2% of children) had in their lifetime experienced prevalent orphanhood and caregiver death combined, with incidence increasing by 49.5% and prevalence by 7.9% since 2000.\"\n","source_date":"2025-01-10","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426204837/https://www.nature.com/articles/s41591-024-03343-6","calculation_notes":"This cross-sectional prevalence figure (4.2% of children) captures a snapshot: at any given time, roughly 1 in 24 US children under 18 have already lost a parent or primary caregiver. This is consistent with (and slightly higher than) the per-parent mortality probability derived from life tables, because it aggregates across all parental ages at birth, both parents, and includes caregiver deaths beyond biological parents. The increase since 2000 is largely driven by drug overdose deaths surpassing COVID-19 as the leading cause of parental death by 2021.\n","independence_note":"This study uses CDC WONDER mortality data as its upstream, independent of the SSA life tables which use SSA-area population data. The two datasets overlap substantially but are compiled by different agencies with different methodologies.\n"},{"url":"https://www.limra.com/siteassets/newsroom/liam/2025/2025_facts_about_life_insurance.pdf","title":"2025 Facts About Life Insurance","publisher":"LIMRA and Life Happens","source_type":"reputable_reference","statistic":"51% of US consumers report having life insurance (down from 63% in 2011); 40% say they need or need more coverage; adults under 30 overestimate term life cost by 10-12x","excerpt":"\"Just 51 percent of consumers say they have life insurance coverage, down from 63 percent in 2011. Adults age 30 and younger overestimate the cost of life insurance by 10 to 12 times.\"\n","source_date":"2025-09-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260306110643/https://www.limra.com/siteassets/newsroom/liam/2025/2025_facts_about_life_insurance.pdf","calculation_notes":"LIMRA data is used here for context on the perception gap — the disconnect between the actuarial probability and the behavioral response (insurance ownership). The 10-12x cost overestimate among young adults helps explain why a ~9% combined risk goes largely uninsured: the perceived cost of addressing it is wildly inflated relative to the actual premium.\n"}],"comparison_anchors":[{"label":"Death by car crash (lifetime, US adult)","lifetime_us_adult":0.0095},{"label":"Lung cancer (lifetime, US adult)","lifetime_us_adult":0.057},{"label":"Heart disease death (lifetime, US adult)","lifetime_us_adult":0.2}],"regional_breakdown":[{"region":"Parent death only (age 30–48)","probability":0.044,"notes":"SSA life table cumulative mortality (5.76% male / 3.03% female), sex-blended average"},{"region":"Parent lasting disability only (age 30–48)","probability":0.05,"notes":"Disclosed proportional estimate from SSA disability incidence scaled to the 18-year window (not directly stated by SSA)"},{"region":"Either death or lasting disability (single parent)","probability":0.09,"notes":"Combined with partial overlap subtracted"},{"region":"At least one parent (two-parent household)","probability":0.17,"notes":"1 − (1 − 0.09)^2 ≈ 0.17; assumes independent risks between parents"}],"personal_factor_multipliers":[{"factor":"single-parent household","multiplier":1,"notes":"baseline — the headline figure already represents a single parent"},{"factor":"two-parent household (at least one affected)","multiplier":1.9,"notes":"1 − (1 − 0.09)^2 ≈ 0.17, roughly 1.9× the single-parent figure"},{"factor":"parent age 40+ at birth of youngest","multiplier":1.8,"notes":"mortality and disability rates roughly double in the 40-58 window vs 30-48"},{"factor":"parent age 22 at birth of youngest","multiplier":0.6,"notes":"lower baseline mortality and disability at younger ages"},{"factor":"parent with chronic illness or high-risk occupation","multiplier":2.5,"notes":"construction, mining, military, or pre-existing cardiovascular/metabolic disease"},{"factor":"healthy non-smoking parent, white-collar occupation","multiplier":0.5,"notes":"substantially below population average mortality and disability rates"}],"short_label":"Parent death/disability","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"bereavement","valence":"negative","caveats":"The 9% central estimate is a population-level average for a 30-year-old US parent over 18 years, combining both death and qualifying disability (defined by SSA as inability to perform substantial gainful activity for 12+ months). Individual risk varies enormously by sex (males ~2× females for mortality), race, occupation, health status, and health-care access. The disability component is especially sensitive to definition: SSA's bar is high (complete work incapacity for 12+ months), so milder but still life-altering disabilities are excluded. Using a broader definition of disability would push the combined figure well above 10%. For a two-parent household, the probability that at least one parent is affected roughly doubles.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-research-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single small pair of children's shoes next to a large empty adult shoe outline, flat vector illustration, muted palette."},"canonical_url":"https://likelier.app/parent-death-while-child-young","api_url":"https://likelier.app/api/fears/parent-death-while-child-young.json"},{"slug":"sedentary-lifestyle-death","question":"What are the odds of dying prematurely from a sedentary lifestyle?","quick_answer":"Roughly 9% of premature deaths are attributable to physical inactivity, and the most sedentary, least-active adults carry a 1.59x all-cause mortality hazard versus the most active. Almost nobody fears it, one of the site's largest underrated gaps, despite it ranking among the top global mortality risks.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Physical inactivity is the fourth leading risk factor for global mortality according to the WHO, responsible for roughly 4-5 million deaths per year worldwide. Almost nobody is afraid of it. It does not appear in fear surveys, it does not feature in disaster films, and it occupies no cultural space as a threat. \"I should exercise more\" is filed under minor self-improvement, somewhere between \"I should floss\" and \"I should learn a language,\" not under mortality risk management. People spend mental energy worrying about plane crashes, shark attacks, and terrorism while sitting 10+ hours a day without registering it as a risk at all. The result is one of the largest perception gaps on this site, running in the underrated direction.\n","rough_estimate":"Most adults sense inactivity is unhealthy but do not classify it as a serious mortality risk","kind":"intuition"},"native":{"display":"HR 1.59 for most sedentary + least active vs least sedentary + most active (all-cause mortality)","numerator":59,"denominator":100,"unit":"excess hazard ratio","population":"adults sitting >8 h/day with <2.5 MET-h/week physical activity vs those sitting <4 h/day with >35.5 MET-h/week"},"normalized":{"lifetime_us_adult":0.09,"display":"~9% of premature deaths globally attributable to physical inactivity","log_value":-1.05,"assumptions":"The headline figure draws on two converging lines of evidence. (1) Lee et al. 2012 (Lancet) estimated that physical inactivity causes 9% of premature mortality worldwide (range 5.1-12.5%), corresponding to more than 5.3 million of the 57 million deaths in 2008. Applied to the US adult population: approximately 3.4 million US deaths per year, 9% = ~306,000 deaths attributable to inactivity per year. Against ~258 million US adults, that yields an annual hazard of ~306,000/258,000,000 = 0.00119. Compounded over 59 years of remaining adult life: 1 - (1 - 0.00119)^59 = 0.068. However, this is a population-average PAF that includes already-active adults who contribute zero to the numerator. (2) Ekelund et al. 2016 (Lancet, 1,005,791 participants) found that adults sitting >8 h/day with the lowest physical activity (<2.5 MET-h/week) had HR 1.59 (95% CI 1.52-1.66) for all-cause mortality compared to the most active + least sedentary reference group. For the ~25% of US adults who are completely inactive (CDC 2024), the excess lifetime mortality risk from inactivity is substantially higher than the population average. We use 0.09 as the point estimate, reflecting the population-level PAF (Lee 2012) adjusted slightly upward from the raw compounding because CDC data show only 26.4% of US adults meet both aerobic and strength guidelines, meaning the majority carry some inactivity-attributable risk. Uncertainty range 0.05-0.13 spans the Lee et al. PAF confidence interval (5.1-12.5%) applied to US mortality.\n","uncertainty":{"low":0.05,"high":0.13},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30370-1/abstract","title":"Does physical activity attenuate, or even eliminate, the detrimental association of sitting time with mortality? A harmonised meta-analysis of data from more than 1 million men and women","publisher":"The Lancet (Ekelund, Steene-Johannessen, Brown, Fagerland, Owen, Powell, Bauman, Lee)","source_type":"peer_reviewed","statistic":"HR 1.59 (95% CI 1.52-1.66) for all-cause mortality in most sedentary + least active quartile; 60-75 min/day moderate activity eliminates excess sitting risk; 1,005,791 participants, 84,609 deaths","excerpt":"\"High levels of moderate intensity physical activity (ie, about 60-75 min per day) seem to eliminate the increased risk of death associated with high sitting time. However, this high activity level attenuated, but did not eliminate the increased risk associated with high TV-viewing time.\"\n","source_date":"2016-07-28","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250810054735/https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(16)30370-1/abstract","calculation_notes":"Ekelund et al. harmonised individual-level data from 16 prospective cohort studies (13 with sitting data, 3 with TV-viewing data). Participants were stratified into quartiles of both sitting time and physical activity. The reference group was the least sedentary + most active quartile (<4 h/day sitting, >35.5 MET-h/week). The highest-risk group (>8 h/day sitting, <2.5 MET-h/week) had HR 1.59. Critically, those sitting >8 h/day but in the most active quartile (>35.5 MET-h/week, roughly 60-75 min/day of moderate activity) had HR 1.04 (95% CI 0.99-1.10), statistically indistinguishable from the reference. This demonstrates that high physical activity can fully offset the mortality risk of prolonged sitting. The HR 1.59 is used as the native display because it captures the worst-case combination most relevant to completely inactive office workers.\n"},{"url":"https://www.thelancet.com/article/S0140-6736(12)61031-9/fulltext","title":"Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy","publisher":"The Lancet (Lee, Shiroma, Lobelo, Puska, Blair, Katzmarzyk)","source_type":"peer_reviewed","statistic":"Physical inactivity causes 9% of premature mortality worldwide (range 5.1-12.5%); 6% of CHD, 7% of type 2 diabetes, 10% of breast cancer, 10% of colon cancer; elimination would increase global life expectancy by 0.68 years (0.41-0.95)","excerpt":"\"We estimated that physical inactivity causes 6% of the burden of disease from coronary heart disease, 7% of type 2 diabetes, 10% of breast cancer, and 10% of colon cancer. Inactivity causes 9% of premature mortality, or more than 5.3 of the 57 million deaths that occurred worldwide in 2008.\"\n","source_date":"2012-07-21","source_accessed":"2026-04-19","calculation_notes":"Lee et al. computed population attributable fractions (PAFs) by comparing disease incidence in inactive vs active populations across 122 countries. The 9% PAF for premature mortality is the primary basis for the normalized lifetime estimate. Applied to US mortality: ~3.4 million deaths/year x 0.09 = ~306,000 inactivity-attributable deaths per year. Against 258 million US adults, annual hazard = 0.00119. Over 59 years: 1 - (1 - 0.00119)^59 = 0.068. The point estimate of 0.09 is slightly above this raw compounding because only 26.4% of US adults meet both aerobic and strength guidelines (CDC 2024), meaning the US-specific PAF may be at the higher end of Lee's global range.\n"},{"url":"https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2212267","title":"Leisure Time Physical Activity and Mortality: A Detailed Pooled Analysis of the Dose-Response Relationship","publisher":"JAMA Internal Medicine (Arem, Moore, Patel, Hartge, de Gonzalez, Visvanathan, et al.)","source_type":"peer_reviewed","statistic":"HR 0.80 (95% CI 0.78-0.82) for those doing less than recommended 7.5 MET-h/week vs none; HR 0.69 (0.67-0.70) at 1-2x guidelines; maximum benefit at 3-5x guidelines","excerpt":"\"Compared to those reporting no leisure-time physical activity, we observed a 20% lower mortality risk among those performing less than the recommended minimum of 7.5 metabolic equivalent hours per week.\"\n","source_date":"2015-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260318075355/https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2212267","calculation_notes":"Arem et al. pooled six prospective cohort studies with 661,137 participants and 116,686 deaths over a median 14.2 years of follow-up. The key finding for this entry is the steep dose-response at the low end: going from zero activity to sub-guideline activity (less than 150 min/week moderate or 75 min/week vigorous) buys a 20% mortality reduction (HR 0.80). Meeting the guidelines (7.5-15 MET-h/week) yields HR 0.69 (31% reduction). Maximum benefit plateaus at 3-5x guidelines, with no excess risk even at 10x. This non-linear curve means the biggest marginal gain comes from moving off zero, not from optimizing an already-active regimen.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db555.htm","title":"Aerobic Physical Activity Among Adults Age 18 and Older: United States, 2024","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"47.2% of US adults met aerobic guidelines in 2024; 26.4% met both aerobic and muscle-strengthening guidelines; rates vary by sex, disability status, weight status","excerpt":"\"In 2024, 47.2% of adults age 18 and older met the federal guidelines for aerobic physical activity. When broken down by sex, 52.3% of men and 42.4% of women met the standards.\"\n","source_date":"2025-04-07","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260423093144/https://www.cdc.gov/nchs/products/databriefs/db555.htm","calculation_notes":"CDC NCHS Data Brief 555 provides the most recent national prevalence data on physical activity. The 26.4% meeting both guidelines means roughly 73.6% of US adults carry some degree of inactivity-attributable risk. The ~25% who are completely inactive (no leisure-time physical activity) are the population most directly described by the Ekelund HR 1.59 estimate. These prevalence figures contextualize the PAF: if three-quarters of US adults fail to meet full guidelines, the population-level mortality burden of inactivity is substantial.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK305049/","title":"Global Recommendations on Physical Activity for Health","publisher":"World Health Organization","source_type":"govt_report","statistic":"Physical inactivity is the fourth leading risk factor for global mortality (6% of deaths globally), following high blood pressure (13%), tobacco use (9%), and high blood glucose (6%)","excerpt":"\"Physical inactivity has been identified as the fourth leading risk factor for global mortality (6% of deaths globally). This follows high blood pressure (13%), tobacco use (9%) and high blood glucose (6%).\"\n","source_date":"2010-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250828083404/https://www.ncbi.nlm.nih.gov/books/NBK305049/","calculation_notes":"WHO's global risk-factor ranking, cited in the body text to support the \"fourth leading risk factor\" framing already present in perceived.description. Grounds the specific ranking (behind high blood pressure, tobacco, and high blood glucose) used in the second paragraph of the body prose. Not used in the native/normalized lifetime-probability calculation, which is derived from Lee et al. 2012 and Ekelund et al. 2016 above.\n"}],"comparison_anchors":[{"label":"Death from drug overdose (lifetime, US adult)","lifetime_us_adult":0.0237},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0095},{"label":"Death from regular smoking (lifetime, lifelong smoker)","lifetime_us_adult":0.5},{"label":"Sleep deprivation mortality (lifetime, chronic <6 h sleeper)","lifetime_us_adult":0.12}],"regional_breakdown":[{"region":"Completely inactive (0 min/week moderate activity)","probability":0.13,"notes":"Highest-risk group. Ekelund HR 1.59 applies most directly here. ~25% of US adults."},{"region":"Insufficiently active (some but below WHO 150 min/week)","probability":0.07,"notes":"Arem et al.: even sub-guideline activity yields HR 0.80 vs inactive. Partial protection."},{"region":"Meets guidelines (150+ min/week moderate)","probability":0.02,"notes":"Baseline — excess mortality attributable to inactivity approaches zero. Arem HR 0.69 vs inactive."},{"region":"Highly active (300+ min/week)","probability":0.01,"notes":"Maximum benefit range. Arem: no excess risk, slight additional benefit over 1x guidelines. Diminishing returns."}],"personal_factor_multipliers":[{"factor":"Completely sedentary (<5,000 steps/day) + sits >8 hrs/day","multiplier":1.8,"notes":"Ekelund 2016: highest risk category (HR 1.59 vs most active + least sedentary). The combination of prolonged sitting and zero exercise is worse than either alone."},{"factor":"Office worker, sits 8+ hrs but exercises 60-75 min/day","multiplier":0.15,"notes":"Ekelund 2016: HR 1.04 (95% CI 0.99-1.10) — exercise at this level eliminates the excess sitting risk almost entirely."},{"factor":"Age 65+","multiplier":1.5,"notes":"Absolute risk higher due to elevated baseline mortality; relative risk from inactivity similar or slightly lower in older cohorts."},{"factor":"Active commuter (walks/cycles to work)","multiplier":0.4,"notes":"Active commuting provides 30-60 min/day of moderate activity, substantially offsetting sedentary job exposure."},{"factor":"Any activity, even below guidelines","multiplier":0.6,"notes":"Arem 2015: even sub-guideline activity (less than 150 min/week) reduces mortality by ~20% vs zero activity."}],"short_label":"Sedentary lifestyle","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"\"Sedentary lifestyle\" is not a discrete exposure like smoking a pack a day — it is a gradient spanning zero movement to not-quite-enough movement to guideline-meeting to highly active, with the dose-response curve steeply non-linear at the low end. This makes a single population-level number inherently lossy. The 9% PAF from Lee et al. 2012 is a global average; the US-specific figure may differ given higher obesity prevalence and more sedentary occupational patterns. Confounding is the persistent challenge: people who do not exercise tend to have lower socioeconomic status, higher rates of smoking, poorer diets, and more chronic illness. The meta-analyses adjust for these factors but residual confounding cannot be fully excluded from observational data. The \"sitting is the new smoking\" comparison, while catchy, is misleading at the individual level: smoking carries a relative risk of ~15-20x for lung cancer, while sedentary behavior carries ~1.2-1.6x for all-cause mortality. The comparison works for population attributable fraction (both kill millions) but not for individual risk magnitude. Finally, all-cause mortality includes deaths that exercise would not have prevented — the attributable fraction isolates the excess, but the boundary between \"would have died anyway\" and \"died because inactive\" is an epidemiological construct, not a clinical certainty.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty office chair viewed from behind against a plain light background, flat vector illustration in muted grey-blue tones."},"canonical_url":"https://likelier.app/sedentary-lifestyle-death","api_url":"https://likelier.app/api/fears/sedentary-lifestyle-death.json"},{"slug":"unsanitary-beauty-salon","question":"What are the odds of getting an infection from an unsanitary beauty salon?","quick_answer":"Any clinically significant infection runs about 1 in 500 per salon visit — roughly a 1 in 11 chance over a lifetime of visits. Many clients assume a higher per-visit risk of 1 in 50 to 1 in 200.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Media coverage of nail-salon outbreaks — the 2000 Mycobacterium fortuitum cluster in California that infected over 100 pedicure customers, periodic MRSA cases traced to shared razors or contaminated tools — creates a vivid impression that beauty salons are petri dishes. The fear is amplified by the visible hygiene variance between salons: budget operations with visibly grimy footbaths versus medical-grade sterilization at high-end establishments. Many clients have an intuitive sense that \"something could be wrong\" but lack a framework for estimating how often that translates into actual clinical infection versus harmless colonization.\n","rough_estimate":"Many clients assume a non-trivial risk per visit, perhaps 1 in 50 to 1 in 200","kind":"intuition"},"native":{"display":"~1 in 500 per salon visit (any clinically significant infection)","numerator":1,"denominator":500,"unit":"per salon visit","population":"US adults visiting nail salons and barbershops"},"normalized":{"lifetime_us_adult":0.095,"display":"~1 in 11 lifetime","log_value":-1.02,"assumptions":"Estimates ~1 in 500 per visit for any clinically significant infection (bacterial, fungal, or viral) based on outbreak investigation data and contamination prevalence studies. A 2005 CDC-linked survey found mycobacteria in 97% of 30 whirlpool footbaths across 18 California nail salons. Not every exposure produces clinical disease, but minor fungal infections (onychomycosis, tinea) are common. Assumes ~50 salon visits over a US adult lifetime (roughly one every 14 months on average, accounting for the fact that many adults visit salons rarely while regular clients go monthly). Lifetime ≈ 1 − (1 − 1/500)^50 ≈ 0.095. The estimate is rough because no population-level incidence study tracks salon-acquired infections systematically.\n","uncertainty":{"low":0.02,"high":0.2},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa012643","title":"An Outbreak of Mycobacterial Furunculosis Associated with Footbaths at a Nail Salon","publisher":"New England Journal of Medicine","source_type":"primary_study","statistic":"Over 100 pedicure customers developed Mycobacterium fortuitum furunculosis from contaminated whirlpool footbaths at a single California nail salon","excerpt":"\"We identified 110 customers of the salon who had furunculosis of the lower extremities. The outbreak strain of M. fortuitum was isolated from the footbaths. Shaving the legs before the pedicure was a risk factor for infection (70 percent of patients vs. 31 percent of controls).\"\n","source_date":"2002-05-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251122221531/https://www.nejm.org/doi/full/10.1056/NEJMoa012643","calculation_notes":"This NEJM case-control study documented the largest known nail-salon infection outbreak. 110 cases from one salon over a period of months. The attack rate among exposed customers is not precisely stated but was high enough to trigger a public health investigation. Shaving legs before pedicure roughly doubled the risk, consistent with micro-abrasion as an entry route. This is a point-source outbreak, not a population-rate study, so it informs the plausibility of transmission rather than the per-visit base rate directly.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3320319/","title":"Mycobacteria in Nail Salon Whirlpool Footbaths, California","publisher":"Emerging Infectious Diseases (CDC)","source_type":"govt_report","statistic":"Mycobacteria found in 29 of 30 (97%) whirlpool footbaths sampled across 18 nail salons in 5 California counties; M. fortuitum in 47% of footbaths","excerpt":"\"Mycobacteria were isolated from 29 (97%) of 30 footspas sampled from 18 nail salons in 5 California counties. Mycobacterium fortuitum was the most frequently isolated mycobacterium, found in 14 (47%) of the 30 footspas surveyed.\"\n","source_date":"2005-04-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260214011659/https://pmc.ncbi.nlm.nih.gov/articles/PMC3320319/","calculation_notes":"This CDC Emerging Infectious Diseases study provides the contamination prevalence that underpins the per-visit estimate. Near-universal mycobacterial contamination of footbaths does not mean near-universal clinical infection — most exposures do not breach skin barriers — but it establishes that pathogen exposure is the norm, not the exception, in whirlpool-equipped nail salons. The gap between exposure prevalence (97%) and clinical infection rate (~0.2% per visit) reflects the effectiveness of intact skin as a barrier.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8007475/","title":"Beauty Salons are Key Potential Sources of Disease Spread","publisher":"Infection and Drug Resistance (Dove Medical Press)","source_type":"peer_reviewed","statistic":"Beauty salons can transmit hepatitis B, hepatitis C, MRSA, herpes, fungal infections, and other pathogens through shared tools, contaminated surfaces, and skin-breaking procedures","excerpt":"\"Among the diseases that can be transmitted in beauty salons are hepatitis B & C, herpes, AIDS, skin and eye infections, hair lice, and chronic fungal diseases. The ingredients in cosmetic products and the materials used in the tools make the salon an ideal environment for the proliferation of microbes, thus contributing to the spread of various diseases.\"\n","source_date":"2021-03-22","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260305121442/https://pmc.ncbi.nlm.nih.gov/articles/PMC8007475/","calculation_notes":"This review aggregates the pathogen-transmission literature for beauty salons globally. It lists the diseases documented to spread in salons — hepatitis B & C, herpes, AIDS, skin and eye infections, hair lice, and chronic fungal diseases — and attributes the risk to contaminated cosmetic ingredients and shared tools. The review confirms that the transmission pathway is real and documented across multiple countries, though population-level incidence data remain sparse, so it informs plausibility of transmission rather than the per-visit base rate directly.\n"}],"comparison_anchors":[{"label":"Food poisoning (lifetime, US adult)","lifetime_us_adult":0.999},{"label":"Hospital-acquired infection (lifetime, US adult)","lifetime_us_adult":0.087}],"personal_factor_multipliers":[{"factor":"Shaves legs before pedicure","multiplier":2.5,"notes":"Micro-abrasions from shaving create entry points; NEJM outbreak study found 70% of cases vs 31% of controls shaved"},{"factor":"Visits high-end salon with autoclave sterilization","multiplier":0.2,"notes":"Proper sterilization protocols dramatically reduce pathogen load"},{"factor":"Has diabetes or immunosuppression","multiplier":3,"notes":"Impaired immune response and poor wound healing increase both infection risk and severity"},{"factor":"Visits budget salon weekly","multiplier":4,"notes":"Higher visit frequency and lower hygiene standards compound the per-visit risk"}],"short_label":"Salon infection","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The per-visit estimate of 1 in 500 is a rough approximation, not a directly measured population incidence rate. No national surveillance system tracks salon-acquired infections as a category. The estimate synthesizes outbreak data, contamination prevalence studies, and the gap between pathogen exposure (very common) and clinical disease (much less common). Most salon-acquired infections are minor fungal conditions (onychomycosis, tinea pedis) that are annoying but not dangerous. Serious infections — mycobacterial furunculosis, MRSA, hepatitis — are far rarer but drive the headlines. The wide uncertainty range (2-20% lifetime) reflects genuine ignorance about the true population rate.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A stylized nail file and small bowl rendered in muted rose and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/unsanitary-beauty-salon","api_url":"https://likelier.app/api/fears/unsanitary-beauty-salon.json"},{"slug":"e-scooter-serious-injury","question":"What are the odds of serious injury riding an electric scooter?","quick_answer":"A regular e-scooter rider faces roughly a 1 in 10 lifetime chance of a serious injury, about 20 injuries per 100,000 trips. Casual riders rarely think about it, while headline-readers tend to assume the risk is very high.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Electric scooters occupy an odd perceptual space. Casual users treat them as toys — low-speed, short-trip, no protective gear required by instinct or habit. The CDC's Austin study found that fewer than 1 percent of injured riders had been wearing a helmet. At the same time, emergency physicians and public health researchers have raised alarms about a rising tide of scooter-related ER visits, and the CPSC's recall of 300,000 Onewheel self-balancing boards after four deaths drew national attention to the broader micromobility injury problem. Most riders underestimate the risk; most non-riders who have read a headline overestimate it.\n","rough_estimate":"casual riders rarely think about injury probability; headline-readers assume it is very high","kind":"intuition"},"native":{"display":"~20 injuries per 100,000 e-scooter trips (Austin, TX, 2018)","numerator":20,"denominator":100000,"unit":"per trip","population":"e-scooter riders, Austin TX shared-fleet study period"},"normalized":{"lifetime_us_adult":0.096,"display":"~1 in 10 lifetime chance of a serious injury (regular rider)","log_value":-1.02,"assumptions":"The CDC/Austin Public Health study (2019) found approximately 20 injuries per 100,000 e-scooter trips requiring medical attention. A \"regular rider\" is modeled as someone who takes 2 scooter trips per week for 10 years — roughly 1,040 trips. At 0.0002 injuries per trip, the lifetime probability of ANY ER-treated injury is 1 − (1 − 0.0002)^1040 ≈ 0.189, or about 1 in 5. The Austin study, however, reports the severity split directly: \"Almost half of the people we identified had a severe injury, such as a broken leg.\" Using that ~48% severe fraction, the serious-injury rate is ~9.7 per 100,000 trips (0.000097 per trip). Over 1,040 trips: 1 − (1 − 0.000097)^1040 ≈ 0.096. The point estimate of 0.096 (about 1 in 10) reflects this serious-injury subset. The scope is activity_specific_lifetime because the risk applies only to people who actually ride e-scooters.\n","uncertainty":{"low":0.06,"high":0.155},"scope":"activity_specific_lifetime"},"sources":[{"url":"http://web.archive.org/web/20260214004906/https://www.austintexas.gov/sites/default/files/files/Health/Epidemiology/APH_Dockless_Electric_Scooter_Study_5-2-19.pdf","title":"Dockless Electric Scooter-Related Injuries Study — Austin, Texas, September-November 2018","publisher":"Austin Public Health / Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"190 injuries over 936,110 trips; 45% involved head injuries; 15% sustained traumatic brain injury; <1% wore helmets","excerpt":"\"Overall, 936,110 e-scooter trips occurred in Austin during the study period. Among the 190 injured riders identified, nearly half sustained head injuries and 15% had a traumatic brain injury. Less than 1% of injured riders had been wearing a helmet.\"\n","source_date":"2019-05-02","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260214004906/https://www.austintexas.gov/sites/default/files/files/Health/Epidemiology/APH_Dockless_Electric_Scooter_Study_5-2-19.pdf","calculation_notes":"190 injuries / 936,110 trips = 20.3 per 100,000 trips. This is the native all-injury rate. For serious injuries, the study reports that \"Almost half of the people we identified had a severe injury, such as a broken leg\" — a ~48% severe fraction. Serious-injury rate ≈ 0.48 × 20.3 ≈ 9.7 per 100,000 trips. Lifetime for a regular rider (1,040 trips): 1 − (1 − 0.000097)^1040 ≈ 0.096, or about 1 in 10.\n"},{"url":"https://www.cpsc.gov/Newsroom/News-Releases/2024/E-Scooter-and-E-Bike-Injuries-Soar-2022-Injuries-Increased-Nearly-21","title":"E-Scooter and E-Bike Injuries Soar: 2022 Injuries Increased Nearly 21%","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"Micromobility injuries rose nearly 21% in 2022 from 2021 (e-scooters alone +22%); an estimated 360,800 micromobility ER visits from 2017 through 2022","excerpt":"\"injuries associated with all micromobility devices increased nearly 21% in 2022 from 2021. Micromobility-related injuries have trended upward since 2017, increasing an estimated average 23% annually... There were an estimated 360,800 emergency department visits related to all micromobility devices from 2017 through 2022.\"\n","source_date":"2024-10-07","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420035125/https://www.cpsc.gov/Newsroom/News-Releases/2024/E-Scooter-and-E-Bike-Injuries-Soar-2022-Injuries-Increased-Nearly-21","calculation_notes":"CPSC uses the National Electronic Injury Surveillance System (NEISS) to estimate national ER visits from a nationally representative sample of US hospitals. The 360,800 micromobility ED visits over 2017-2022, the nearly 21% all-device rise in 2022 (e-scooters +22%), and the upward ~23% annual trend provide the national scale. These corroborate the Austin per-trip rate when divided by estimated national trip volumes but are not used directly for the native rate because the denominator (trips) is less precisely known at national scale.\n"},{"url":"https://ajph.aphapublications.org/doi/10.2105/AJPH.2024.307820","title":"The Burden of Injuries Associated With E-Bikes, Powered Scooters, Hoverboards, and Bicycles in the United States: 2019-2022","publisher":"American Journal of Public Health","source_type":"peer_reviewed","statistic":"Population-based rates of powered scooter injuries increased 88% between 2019 and 2022; head injuries accounted for 18-28% of cases","excerpt":"\"The population-based rates of powered scooter injuries increased by 88.0% between 2019 and 2022.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250502233409/https://ajph.aphapublications.org/doi/10.2105/AJPH.2024.307820","calculation_notes":"The AJPH study used NEISS data for 2019-2022 and found powered scooters had the fastest-growing injury rate among micromobility devices. Head injury prevalence of 18-28% across studies is consistent with the Austin CDC finding of 45% (the higher Austin figure likely reflects less helmet use in the early shared-scooter era). Used as independent corroboration of trend and injury profile.\n"}],"comparison_anchors":[{"label":"Cycling head injury (lifetime, regular cyclist)","lifetime_us_adult":0.02},{"label":"Fatal car crash (lifetime, US adult)","lifetime_us_adult":0.0095}],"personal_factor_multipliers":[{"factor":"No helmet vs helmet use","multiplier":3.3,"notes":"Medina et al. (2021, Injury — PMID 34083100): helmet use was associated with a roughly 66% reduction in head injury risk in e-scooter crashes, implying a ~3× higher risk for unhelmeted riders. Consistent with the CDC/Austin finding that <1% of injured riders wore helmets."},{"factor":"Alcohol or drug intoxication","multiplier":2.7,"notes":"Kim et al. (2021, Journal of Trauma and Injury), a nationwide cross-sectional study in South Korea, found that riding an electric personal mobility device while intoxicated was associated with severe injury (adjusted odds ratio 2.78; 95% CI 1.52–5.08). Multiple US emergency-room studies corroborate impairment as a leading injury factor."},{"factor":"Night riding","multiplier":2,"notes":"Nelson et al. (2019, Journal of Transport & Health) and CPSC NEISS analysis: nighttime e-scooter injuries occur at approximately double the rate of daytime injuries after controlling for exposure, likely due to reduced visibility and increased impairment rates after dark."},{"factor":"Age <25 or >65","multiplier":1.7,"notes":"CPSC NEISS data and the AJPH 2024 micromobility study show riders under 25 (inexperience, risk-taking) and over 65 (slower reaction times, greater injury severity) sustain injury rates approximately 1.5-2× the 25-64 baseline per trip."}],"short_label":"E-scooter injury","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The CDC/Austin study is the best per-trip denominator available but covers only one city during a three-month period in 2018, before many scooter-share operators improved rider education and speed-limiting geofences. National CPSC data confirm the injury volume is large and growing, but per-trip rates at scale may differ from the Austin snapshot. The \"serious injury\" definition (fractures, TBI, hospitalization) is approximate; different studies use different thresholds. Helmet use has remained extremely low across all studies.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single electric scooter handlebar with a small first-aid cross icon, flat vector illustration on a pale background."},"canonical_url":"https://likelier.app/e-scooter-serious-injury","api_url":"https://likelier.app/api/fears/e-scooter-serious-injury.json"},{"slug":"cosmetic-surgery-abroad-complications","question":"What are the odds of a serious complication from cosmetic surgery abroad?","quick_answer":"For budget-market cosmetic surgery abroad, serious complications run around 1 in 10 per trip, several times higher than the under-1-in-50 many patients assume. Accredited international facilities can match home-country safety; this headline rate reflects unaccredited budget clinics.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"The narrative around cosmetic surgery tourism is dominated by two competing images: the transformation story (dramatic before-and-after, fraction of UK price, five-star recovery villa) and the horror story (sepsis, necrosis, NHS emergency admission). Neither is the statistically typical outcome. Most patients underestimate the serious-complication rate because marketing presents curated success cases, and because the travel-and-surgery package creates psychological commitment that suppresses pre-booking risk evaluation. The assumption that surgery performed in a clean, professional-looking Turkish clinic must meet roughly the same standards as home-country care systematically underestimates the role of regulatory oversight, surgeon board certification, and post-operative follow-up infrastructure.\n","rough_estimate":"Many patients assume serious complications are rare -- under 1 in 50 -- comparable to home-country surgery","kind":"intuition"},"native":{"display":"~1 in 10 cosmetic surgery trips abroad (serious complication)","numerator":1,"denominator":10,"unit":"per cosmetic surgery trip abroad","population":"International cosmetic surgery tourists undergoing major procedures (abdominoplasty, BBL, breast surgery) at overseas facilities"},"normalized":{"lifetime_us_adult":0.1,"display":"~1 in 10 per procedure (activity-specific)","log_value":-1,"assumptions":"Multiple converging data streams anchor the ~10% serious-complication rate for cosmetic surgery abroad. (1) BAAPS (British Association of Aesthetic Plastic Surgeons) national audit: 324 patients required NHS corrective surgery after returning to the UK over four years, with the annual number rising 44% in 2021 compared to the previous year; Turkey accounted for 100% of complications in one audit cohort (abdominoplasty 75%, breast surgery 25%). (2) A Wounds UK / Welsh systematic review of 44 studies found that up to 53% of returning patients had moderate-to-severe complications. This is a selected sample (patients who presented), not a denominator-based rate. (3) The PMC study by Somogyi et al. (2019), examining 20 UK patients presenting with overseas cosmetic complications, found 20% with major complications and 40% with intermediate complications, with abdominoplasty and gluteal augmentation driving the most severe outcomes. (4) A high-volume accredited Colombian centre (2,324 patients, 7,141 procedures) reported 2.2% per-procedure serious complication rate under controlled conditions -- demonstrating that medical tourism can match home-country safety when properly regulated, but that this requires accreditation that Turkish budget clinics frequently lack. The 10% headline is an estimate for budget-market overseas procedures (primarily Turkey, unaccredited); it is explicitly not the rate for JCI-accredited international facilities. BBL-specific mortality: the Multi-Society Task Force (ASERF/ASAPS/ISAPS) documented a mortality rate of ~1 in 3,448 for all BBL procedures in 2017, improving to ~1 in 14,952 in 2019 following guideline adoption. However, a South Florida PMC study found 92% of 25 BBL deaths (2010--2022) occurred at high-volume budget clinics -- suggesting the budget-clinic mortality rate substantially exceeds the society-survey average. Scope is activity-specific: one procedure trip, per person.\n","uncertainty":{"low":0.03,"high":0.3},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/32306045/","title":"Improvement in Brazilian Butt Lift (BBL) Safety With the Current Recommendations from ASERF, ASAPS, and ISAPS","publisher":"Aesthetic Surgery Journal","source_type":"peer_reviewed","statistic":"BBL mortality improved from 1 in 3,448 (2017) to 1 in 14,952 (2019) following multi-society safety guidelines; pulmonary fat embolism incidence fell from 1 in 1,030 to 1 in 2,492","excerpt":"\"[Paraphrase from abstract -- full text paywalled] The mortality rate showed improvement trends, declining from 1 in 3,448 (2017) to 1 in 14,952 (2019). PFE incidence decreased from 1 in 1,030 (2017) to 1 in 2,492 (2019). 94% of surgeons reported awareness of the 2017 recommendations. Unsafe deep muscle injection declined from 13.1% to 0.8% of surgeons.\"\n","source_date":"2020-04-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20251027195109/https://pubmed.ncbi.nlm.nih.gov/32306045/","calculation_notes":"Rios & Gupta (2020) is the primary ISAPS/ASAPS survey documenting BBL mortality before and after guideline adoption. The 1-in-3,448 pre-guideline mortality rate (2017) is the historically cited figure for all BBLs. The improvement to 1-in-14,952 reflects guideline-compliant surgeons -- but the South Florida data (PMC 9896146) show 92% of deaths occurred at budget clinics where guidelines were not followed, meaning the budget-clinic mortality rate in the 2019 period remained far higher than the survey average. This source provides the denominator for the BBL-specific mortality caveat in the entry; it is not the direct source for the 10% headline, which covers all serious complications across all cosmetic procedure types.\n","independence_note":"Multi-society survey (ASAPS + ISAPS members); independent of the South Florida retrospective case series and the BAAPS audit data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9896146/","title":"Brazilian Butt Lift -- Associated Mortality: The South Florida Experience","publisher":"Aesthetic Surgery Journal (PMC)","source_type":"peer_reviewed","statistic":"25 BBL-related PFE deaths in South Florida 2010--2022; 92% at high-volume budget clinics; 14 deaths occurred after 2018 safety guidelines","excerpt":"\"South Florida has experienced 25 BBL-related fat emboli deaths between 2010 and 2022; however, 14 of these occurred after publication of the Aesthetic Surgery Education and Research Foundation's 2018 guidelines and the 2019 Florida Board of Medicine's BBL 'subcutaneous-only' rule.\"\n","source_date":"2022-08-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260505173839/https://pmc.ncbi.nlm.nih.gov/articles/PMC9896146/","calculation_notes":"The 92% budget-clinic association for BBL deaths is the load-bearing statistic for why published society-average mortality rates understate the risk at unaccredited budget facilities -- the same facilities used by the majority of cosmetic surgery tourists. 14 of 25 deaths occurred after safety guidelines were established, confirming that guidelines adopted by society members do not protect patients at non-member budget operations.\n","independence_note":"Retrospective case series from a South Florida trauma centre; independent of the ASAPS/ISAPS survey methodology and the BAAPS audit data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6635218/","title":"Complications of Cosmetic Surgery Abroad -- Cost Analysis and Patient Perception","publisher":"Plastic and Reconstructive Surgery -- Global Open (PMC)","source_type":"peer_reviewed","statistic":"Among 20 UK patients presenting with overseas cosmetic complications: 20% major, 40% intermediate, 40% minor; abdominoplasty 45% of cases; all major complications from gluteal augmentation","excerpt":"\"[Paraphrase from abstract -- full text paywalled] 20 patients were studied (95% female). Minor complications: 40% of cases; intermediate: 40%; major: 20%. Abdominoplasty accounted for 45% of all complications (9 cases). All major complications occurred in gluteal augmentation cases. Lower cost was the most popular reason for travel.\"\n","source_date":"2019-06-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20250217103827/https://pmc.ncbi.nlm.nih.gov/articles/PMC6635218/","calculation_notes":"This UK case series documents the severity distribution among patients who presented with complications. It is a selected sample (not all overseas cosmetic patients), so the 20% major rate applies within the complication-presenting population, not to all cosmetic tourists. However, it provides procedure-level evidence that abdominoplasty and gluteal augmentation are the highest-risk cosmetic procedures abroad -- consistent with BAAPS audit data.\n","independence_note":"Independent UK trauma-centre case series; different methodology and population from the Rios & Gupta society survey and the BAAPS national audit.\n"},{"url":"https://aestheticsjournal.com/news/baaps-audit-reveals-increased-complications-from-surgery-abroad/","title":"BAAPS Audit Reveals Increased Complications from Surgery Abroad","publisher":"Aesthetics Journal (reporting on BAAPS national audit)","source_type":"reputable_reference","statistic":"324 patients required NHS corrective surgery over four years; annual number rose 44% in 2021 vs the previous year; 100% of complications came from Turkey; abdominoplasty 75% of complications, breast surgery 25%; some patients required ICU/HDU admission","excerpt":"\"324 patients required corrective surgery after returning to the UK in the past four years. The annual number rose by 44% in 2021 compared to the previous year. 100% of complications came from Turkey. Abdominoplasty accounted for 75% of complications, followed by breast surgery at 25%. Some patients required emergency surgical removal of dead skin tissue and admission to intensive care for life support following systemic infection.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20250514120532/https://aestheticsjournal.com/news/baaps-audit-reveals-increased-complications-from-surgery-abroad/","calculation_notes":"The BAAPS national audit is the most comprehensive UK-population data source on cosmetic surgery abroad complications. The audit's 324 corrective-surgery cases over four years -- with the annual number rising 44% in 2021 and 100% attributable to Turkey -- provide the denominator context for Turkey as the dominant source of cosmetic tourism risk. The audit does not report a per-procedure complication rate (denominator is total procedures abroad, which is unknown), but the absolute numbers and severity distribution corroborate the ~10% estimate when set against estimated UK cosmetic tourism volumes.\n","independence_note":"BAAPS national audit of UK NHS trusts; independent of the peer-reviewed case series and ISAPS society surveys.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/41018743/","title":"Safety and Outcomes in Plastic Surgery Medical Tourism: A Review of 2324 Patients and 7141 Procedures","publisher":"Plastic and Reconstructive Surgery -- Global Open (PubMed)","source_type":"peer_reviewed","statistic":"Retrospective review of 2,324 international patients / 7,141 procedures at an accredited Cartagena, Colombia practice (2013--2024): 6.2% complication rate per patient, 2.2% per procedure","excerpt":"\"The overall complication rate was 6.2% per patient (2.2% per procedure), which compares favorably with published benchmarks from board-certified plastic surgeons in the United States.\"\n","source_date":"2025-09-26","source_accessed":"2026-07-14","calculation_notes":"This high-volume accredited Colombian centre reports a 2.2% per-procedure serious complication rate under controlled, board-certified conditions. It is the low-end anchor for the entry's uncertainty band: it demonstrates that medical tourism can approach home-country safety when the facility is properly accredited, which the budget-market Turkish clinics driving the BAAPS complication audit frequently are not. It is not the source for the ~10% headline (which covers unaccredited budget-market procedures); it bounds the optimistic tail.\n","independence_note":"Single-centre retrospective observational study (Cartagena, Colombia); independent of the UK case series, the ISAPS society surveys, and the BAAPS national audit.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Brazilian Butt Lift (BBL)","multiplier":5,"notes":"BBL carries the highest procedure-specific mortality of any elective cosmetic operation: ~1 in 3,448 per procedure in 2017 at society-member surgeons; at budget clinics where 92% of PFE deaths occurred, the rate is substantially higher. The 5x multiplier reflects that BBL at an unaccredited budget clinic represents the worst-case end of the cosmetic tourism risk spectrum.\n"},{"factor":"JCI-accredited or nationally certified overseas facility","multiplier":0.2,"notes":"High-volume accredited facilities (Colombia retrospective, 2,324 patients) report 2.2% per-procedure serious complication rates, roughly equivalent to home-country benchmarks. Accreditation substantially closes the safety gap. JCI and equivalent national hospital accreditation bodies provide the most verifiable quality signal.\n"},{"factor":"Flying within 5--10 days post-surgery","multiplier":2,"notes":"Air travel post-abdominoplasty, BBL, or liposuction significantly increases deep vein thrombosis and pulmonary embolism risk. BAAPS explicitly warns that combining surgery with air travel within 5--10 days is a primary additional risk factor for cosmetic tourists, beyond the surgical complication risk itself.\n"},{"factor":"Procedure performed at high-volume budget Turkish clinic","multiplier":3,"notes":"BAAPS audit: 100% of UK NHS cosmetic tourism complication cases in the 2021 cohort cited Turkey. High-volume budget clinics in Turkey are structurally optimized for throughput (one documented case performed as the seventh BBL of the day, beginning at 8:31 PM) rather than individual patient safety. Abdominoplasty and gluteal augmentation at these facilities account for the majority of ICU-level complications.\n"},{"factor":"Rhinoplasty or breast augmentation (non-gluteal, non-abdominal)","multiplier":0.4,"notes":"Non-gluteal, non-abdominal cosmetic procedures carry substantially lower serious complication rates. Rhinoplasty and breast augmentation done abroad have lower mortality and acute-complication risks than the body-contouring procedures that dominate the BAAPS complication audit. Revision and aesthetic dissatisfaction remain risks, but the life-threatening serious-complication rate is meaningfully lower.\n"}],"short_label":"Cosmetic surgery abroad risk","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"The ~10% serious complication rate is an estimate for major cosmetic procedures (abdominoplasty, gluteal augmentation, body contouring) at unaccredited overseas budget facilities, primarily Turkey. It is not representative of all cosmetic surgery abroad: accredited facilities in Colombia, Thailand, and South Korea with JCI or equivalent certification achieve serious complication rates of ~2%, comparable to home-country surgery. The high-end uncertainty (30%) reflects extreme cases -- patients presenting to NHS emergency wards, as documented in the BAAPS audit. The BBL-specific mortality data (1 in 3,448 pre-guideline; ~1 in 15,000 post-guideline among society-member surgeons) is separately quantified and substantially higher than for other cosmetic procedures; the entry uses `outcome_severity: fatal` to flag that mortality is a documented outcome for the highest-risk procedures in this category. The vast majority of serious complications are non-fatal: wound dehiscence, seroma, haematoma, infection, tissue necrosis, and DVT requiring hospitalisation. \"Serious\" means an adverse event requiring emergency or surgical treatment at home -- not aesthetic dissatisfaction or routine healing. Cosmetic tourism patients are not tracked in any prospective registry; all estimates are derived from retrospective case series, national audit data, and society surveys, each with denominator uncertainty about total overseas procedure volumes.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A flat vector illustration of an abstract airplane silhouette overlaid with a simple surgical incision line, muted palette."},"canonical_url":"https://likelier.app/cosmetic-surgery-abroad-complications","api_url":"https://likelier.app/api/fears/cosmetic-surgery-abroad-complications.json"},{"slug":"early-sugar-salt-metabolic-disease","question":"How much does added sugar and salt in the first two years of life raise a child's risk of developing type 2 diabetes and hypertension in adulthood?","quick_answer":"Unrestricted added sugar in the first two years of life is associated with about 10 additional percentage points of lifetime type 2 diabetes risk (roughly 40 percent versus 30 percent with a sugar-restricted early diet), and low-sodium infant diets tracked 3.6 mmHg lower systolic blood pressure. Almost no parents cite adult diabetes risk as a reason to restrict infant sugar.\n","category":"health","tags":["kids","infant","toddler","food"],"no_reliable_estimate":false,"perceived":{"description":"Most parents connect infant sugar with dental cavities and, at most, a short-term behavioral jolt. Salt for babies raises some concern about kidney load but is rarely framed as a long-term cardiovascular risk. The idea that a toddler's diet permanently calibrates adult type 2 diabetes and hypertension risk — across four to five decades — sits well outside ordinary parental risk awareness. When infant nutrition guidelines recommend avoiding added sugar and salt before 12 months, the stated reasons tend to be taste preference and kidney function rather than metabolic programming. The long-run causal pathway is rarely communicated, and even parents who follow the guidelines typically do not know why the stakes are this high.\n","rough_estimate":"~0% of parents cite adult type 2 diabetes risk as a reason to restrict infant sugar; dental cavities and blood pressure in childhood are the typical framing","kind":"intuition"},"native":{"display":"~78 per 1,000 UK adults born after sugar rationing ended (unrestricted early-life diet) developed type 2 diabetes in midlife — an estimated 35% more than the ~51 per 1,000 born during rationing (restricted early-life diet); 15-year RCT found 3.6 mmHg lower systolic blood pressure in low-sodium infants","numerator":78,"denominator":1000,"unit":"T2D diagnoses per 1,000 adults in midlife (unrestricted early-life sugar cohort, back-calculated from Gracner et al. 2024 relative risk applied to overall study T2D rate)","population":"UK adults born October 1951–March 1956, assessed at age 50-70 via UK Biobank (N≈60,183); unrestricted group born after sugar rationing ended September 1953 (N≈22,000)"},"normalized":{"lifetime_us_adult":0.1,"display":"~10 additional percentage points of lifetime T2D risk attributable to unrestricted early-life sugar exposure (US adult lifetime T2D risk ≈40% with typical sugar exposure vs ≈30% with sugar-restricted early diet)","log_value":-1,"assumptions":"Applies Gracner et al.'s (2024) 35% relative risk reduction cross-nationally to the US adult lifetime T2D risk (~40%, CDC/Gregg et al. projections). The excess attributable risk of ~10 percentage points assumes the biological mechanism (early-life metabolic programming) operates similarly in US populations. UK Biobank has a healthy-volunteer bias that likely understates absolute rates; the effect size estimate is considered robust given the natural-experiment design. Lower bound (6 pp) reflects the 95% CI floor on Gracner's relative risk; upper bound (16 pp) reflects the 40% reduction observed with 1.5+ years of restriction applied to a US 40% lifetime base.\n","uncertainty":{"low":0.06,"high":0.16},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.science.org/doi/10.1126/science.adn5421","title":"Exposure to sugar rationing in the first 1000 days of life protected against chronic disease","publisher":"Science (Gracner, Boone & Gertler)","source_type":"peer_reviewed","statistic":"Early-life sugar rationing reduced type 2 diabetes risk by ~35% and hypertension risk by ~20%; delayed disease onset by 4 years (T2D) and 2 years (hypertension); among the 60,183 UK Biobank participants, nearly 4,000 developed T2D and almost 20,000 had hypertension","excerpt":"\"We examined the impact of exposure to sugar restrictions within 1000 days after conception on type 2 diabetes and hypertension, leveraging quasi-experimental variation from the end of the United Kingdom's sugar rationing in September 1953. Rationing restricted sugar intake to levels within current dietary guidelines, and consumption nearly doubled immediately after rationing ended. Using an event study design with UK Biobank data comparing adults conceived just before or after rationing ended, we found that early-life rationing reduced type 2 diabetes and hypertension risk by about 35 and 20% and delayed disease onset by 4 and 2 years, respectively. Protection was evident with in utero exposure and increased with postnatal sugar restriction, especially after 6 months, when eating of solid foods likely began. In utero sugar rationing alone accounted for about one-third of the risk reduction.\"\n","source_date":"2024-10-31","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20250802143428/https://www.science.org/doi/10.1126/science.adn5421","calculation_notes":"Gracner T, Boone C, Gertler PJ. Science 386(6725):1043-1048. N=60,183 UK Biobank participants born October 1951–March 1956. The natural experiment compares adults whose first 1,000 days included UK WWII sugar rationing (rationed group, N≈38,000, conceived before September 1953) against those born after rationing ended (unrationed group, N≈22,000). Sugar consumption nearly doubled immediately after rationing ended, providing a near-random assignment of early-life sugar exposure. Overall cohort: ~4,000 developed T2D (~6.6%) and ~20,000 had hypertension (~33%) — reported verbatim in Science news coverage (Offord, Science, Oct 31 2024). The paper reports a 35% relative risk reduction for T2D and 20% for hypertension. The full-text tables (paywalled) contain group-specific absolute rates; the back-calculated estimates used here are: Overall T2D rate (6.6%) / (0.65 × 38,000/60,183 + 1.00 × 22,000/60,183) ≈ unrationed group 7.8%, rationed group 5.1%. The native numerator (78/1,000) represents the back-calculated T2D rate in the unrationed (unrestricted sugar) group and thus reflects the risk faced by the typical Western child with added sugar in early diet. The 3% point gap (78 vs 51 per 1,000) is the excess T2D burden attributable to unrestricted early-life sugar. Dose-response: in utero exposure alone → 15% lower T2D; 1.5+ years of restriction → 40% lower T2D (from Science news/Offord article, reporting the paper's Figures 1–3). Normalized to US lifetime T2D risk: CDC projects ~40% lifetime T2D risk for US adults born in 2000 (Gregg et al. 2014, Lancet Diabetes Endocrinol, updated modeling). If the Gracner 35% relative risk applies cross-nationally (early-life metabolic programming is a biological mechanism, not UK-specific), a US child with restricted early sugar has an equivalent lifetime T2D risk of ~40%/1.35 ≈ 30%. Excess attributable lifetime risk from typical US early sugar exposure: ~10 percentage points (uncertainty 6–16 pp, reflecting the confidence interval on the 35% relative risk and cross-national extrapolation uncertainty). The UK Biobank has a well-documented healthy-volunteer bias (skewing toward higher socioeconomic status and lower disease burden than the UK general population); the true T2D rates in the general UK population of that cohort would be higher than the 6.6% overall rate observed.\n"},{"url":"https://www.ahajournals.org/doi/10.1161/01.HYP.29.4.913","title":"Long-term effects of neonatal sodium restriction on blood pressure","publisher":"Hypertension (Geleijnse, Hofman et al.)","source_type":"peer_reviewed","statistic":"In a randomized trial of 476 Dutch newborns (low vs normal sodium for first 6 months), the low-sodium group had 3.6 mmHg lower systolic blood pressure at 15-year follow-up (95% CI: -6.6 to -0.5)","excerpt":"\"The adjusted systolic blood pressure at follow-up was 3.6 mm Hg lower (95% confidence interval, -6.6 to -0.5) and the diastolic pressure was 2.2 mm Hg lower (95% confidence interval, -4.5 to 0.2) in children who had been assigned to the low sodium group (n = 71) compared with the control group (n = 96). These findings suggest that sodium intake in infancy may be important in relation to blood pressure later in life.\"\n","source_date":"1997-04-01","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20241219113029/https://www.ahajournals.org/doi/10.1161/01.HYP.29.4.913","calculation_notes":"Geleijnse JM, Hofman A, Witteman JC, Hazebroek AA, Valkenburg HA, Grobbee DE. Hypertension 1997;29(4):913-7. PMID 9095076. Original RCT 1980: N=476 Dutch neonates randomized to low-sodium (n=231) or normal-sodium (n=245) diet for first 6 months. 15-year follow-up retained 167 participants (35% of original N). The 3.6 mmHg lower systolic BP in the low-sodium group at age 15 is a clinically meaningful difference: a 5 mmHg reduction in systolic BP is associated with approximately 20% lower stroke risk (meta-analyses from the Blood Pressure Lowering Treatment Trialists' Collaboration). The diastolic effect (-2.2 mmHg, 95% CI: -4.5 to 0.2) was not statistically significant. The 35% follow-up retention rate is a limitation. This RCT provides the strongest experimental evidence for a lasting blood pressure effect of early-life sodium restriction, independently corroborating Gracner's observational finding on hypertension. Note: commonly misattributed as \"Hofman et al. 1997\" — the correct first author is Geleijnse.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7512119/","title":"Added Sugars Intake in Infants and Toddlers","publisher":"Journal of the Academy of Nutrition and Dietetics (Herrick et al., NHANES 2011-2016)","source_type":"peer_reviewed","statistic":"60.6% of US infants aged 6-11 months consumed added sugars on a given survey day; infants averaged ~0.9 tsp/day and toddlers (12-23 mo) averaged 5.8 tsp/day of added sugars","excerpt":"\"About 60.6% (95% CI, 55.6% to 65.4%) of infants aged 6 to 11 months and 98.3% (95% CI, 96.9% to 99.1%) of toddlers aged 12 to 23 months consumed added sugars on a given day between 2011 and 2016. On average, infants consumed almost 1 tsp (0.9 tsp, 95% CI 0.7 to 1.1 tsp) of added sugars on a given day, and toddlers consumed 5.8 tsp (95% CI, 5.3 to 6.4 tsp).\"\n","source_date":"2019-11-14","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20250314083604/https://pmc.ncbi.nlm.nih.gov/articles/PMC7512119/","calculation_notes":"Herrick KA, Fryar CD, Hamner HC, Park S, Ogden CL. J Acad Nutr Diet 2020;120(1):23-32 (published online 2019-11-14). DOI 10.1016/j.jand.2019.09.007, PMID 31735600. NHANES 2011-2016, N=1,211 US infants and toddlers aged 6-23 months (24-hour dietary recall). This study establishes that the \"unrestricted early sugar\" scenario is not a theoretical risk but the current reality for the majority of US infants: 60.6% already consume added sugars before age 12 months, averaging ~0.9 tsp/day (infants) and 5.8 tsp/day (toddlers). The 2017 AHA statement recommends children under 2 avoid any added sugars. This prevalence figure anchors the claim that the normalized excess-risk calculation applies to most US children, not an unusual subgroup.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Full restriction of added sugar and salt through first 2 years (solids period)","multiplier":0.6},{"factor":"Restriction limited to in utero period only (born after rationing ended)","multiplier":0.85},{"factor":"High genetic metabolic risk (obesity-predisposing genotype) without early-life sugar restriction","multiplier":3},{"factor":"Breastfeeding exclusively for ≥6 months (no complementary foods with added sugar)","multiplier":0.75}],"short_label":"Infant sugar/salt and adult disease","myth_framing":"underrated","outcome_severity":"moderate_harm","outcome_type":"chronic_illness","valence":"negative","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-13","last_reviewed":"2026-05-13","reviewed":true,"generated_at":"2026-05-13","image":{"alt":"A small bowl of bright-colored fruit puree beside a plain bowl of unseasoned mash on a high-chair tray, soft kitchen light."},"canonical_url":"https://likelier.app/early-sugar-salt-metabolic-disease","api_url":"https://likelier.app/api/fears/early-sugar-salt-metabolic-disease.json"},{"slug":"elder-financial-scam-loss","question":"How likely is an adult 60+ to lose money to a financial scam in retirement?","quick_answer":"Roughly 1 in 10 adults 60+ lose money to a financial scam over a 20-year retirement, though the estimate carries wide uncertainty. Reported complaints run only 2 to 3 per 1,000 each year, but most victims never report, so the true rate is several times higher.\n","category":"other","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Financial fraud targeting older adults is broadly understood to exist but not widely perceived as a personal risk. Most adults over 60 consider themselves too savvy to fall for scams, a self-assessment that conflicts sharply with the data. The perception gap is compounded by the silence around victimization: financial fraud carries social stigma, and adults who are scammed rarely disclose it — to family members, to regulators, or to researchers. This underreporting makes the true scale essentially invisible, and also means that the people closest to older adults are rarely aware of the risk level until after a loss has occurred.\n","kind":"intuition"},"native":{"display":"Roughly 2 to 3 adults 60+ per 1,000 file a fraud-loss complaint each year, and the FTC estimates only a small fraction of victims report — so the true annual rate is several times higher but not precisely known","numerator":23,"denominator":10000,"unit":"reported complaints per year (true rate several times higher, underreported)","population":"US adults aged 60+ (FBI IC3 2024 reported complaints; FTC Protecting Older Consumers 2024-2025 documents heavy underreporting)"},"normalized":{"lifetime_us_adult":0.1,"display":"roughly 1 in 10 adults 60+ lose money to a scam over a 20-year retirement window — with wide uncertainty","log_value":-1,"assumptions":"Annual reported rate: FBI IC3 2024 recorded 147,127 complaints from adults 60+ (~0.23% of ~63 million US adults 60+ per year filing a complaint); the IC3 2025 report (released May 2026) recorded more than 201,000 complaints from victims over 60, ~0.32% per year — a 37% increase in complaints, alongside a 59% increase in losses (to $7.7B). The faster growth in dollars than in complaints, together with the FTC's finding that older adults report losing money at a lower rate than younger adults, indicates the escalation is driven by per-victim loss magnitude rather than rising incidence — so the headline probability is held at 0.10 despite the larger dollar totals. Reported complaints capture only a fraction of victims: the FTC Protecting Older Consumers 2024-2025 report cites a mass-market fraud study in which \"just 4.8% of consumers who experience fraud said that they reported the incident,\" and its own case-matching found \"an average 2.0% reporting rate in cases with average losses of less than $1,000 and a 6.7% reporting rate in cases with an average loss of over $1,000.\" The FTC uses these rates to estimate the DOLLAR cost of fraud to older adults in 2024 at between $10.1 billion and $81.5 billion, versus $2.4 billion reported — an underreporting factor on losses of roughly 4× to 34×. That factor is a dollar-loss correction, not a victim-count correction, and no cited source publishes a corrected annual VICTIMIZATION rate for adults 60+, so the true annual incidence cannot be pinned precisely. Taking the reported complaint rate as a floor and the FTC reporting fractions as an order-of-magnitude ceiling, the corrected annual victimization rate plausibly runs from a few tenths of a percent to a few percent. Applying the reported rate over a 20-year retirement window gives 1 - (1-0.0023)^20 ≈ 4% (2024 base) to 1 - (1-0.0032)^20 ≈ 6% (2025 base); if the true rate is several times higher, the 20-year figure rises well above 10%. The headline (0.10) is a conservative central estimate: above the reported-only floor, well below the high-reporting-fraction extreme. Wide uncertainty reflects genuine irreducible uncertainty about the true victimization rate. UK Finance 2024 and ACCC 2024 data corroborate the scale but do not provide age-stratified lifetime rates comparable to US estimates. Low (0.04): if incidence is close to the reported complaint rate (little correction for underreporting). High (0.60): if the true victim count is roughly an order of magnitude above reported complaints and risks compound fully over 20 years.\n","uncertainty":{"low":0.04,"high":0.6},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ic3.gov/AnnualReport/Reports/2024_IC3ElderFraudReport.pdf","title":"2024 Elder Fraud Report","publisher":"Federal Bureau of Investigation Internet Crime Complaint Center (FBI IC3)","source_type":"govt_report","statistic":"147,127 complaints from adults 60+ in 2024, total losses $4.885 billion; average loss per victim higher than any other age group","excerpt":"\"In 2024, the Internet Crime Complaint Center received 147,127 complaints from victims 60 years of age and older, with losses exceeding $4.885 billion. This represents a 43 percent increase in losses compared to 2023. Adults aged 60 and over experienced the highest average loss per victim of any age group, with investment fraud and tech support scams accounting for the largest share of losses.\"\n","source_date":"2025-04-01","source_accessed":"2026-05-04","calculation_notes":"FBI IC3 2024 Elder Fraud Report. 147,127 complaints / approximately 63 million US adults 60+ = 0.23% annual reported complaint rate. This is the reported (not corrected) figure; it is used as the lower anchor in the calculation. The FTC Protecting Older Consumers 2024-2025 report documents heavy underreporting (reporting rates of 4.8%, and 2.0%/6.7% by loss size), which it uses to estimate the DOLLAR cost of fraud to older adults at $10.1B–$81.5B versus $2.4B reported. That is a loss-dollar correction, not a corrected victim-count rate; no cited source publishes a corrected annual victimization rate, so the true incidence is bounded qualitatively rather than by a single multiplier. The 20-year reported lifetime is ~4%; the headline (0.10) is a conservative central estimate set above the reported floor to reflect documented underreporting, with explicit uncertainty (0.04–0.60).\n"},{"url":"https://www.fbi.gov/news/stories/scammers-target-older-adult-victims","title":"Scammers Target Older Adult Victims (reporting 2025 IC3 Elder Fraud figures)","publisher":"Federal Bureau of Investigation (FBI)","source_type":"govt_report","statistic":"2025: complaints from victims over 60 exceeded 201,000 (up 37%) with reported losses over $7.7 billion (up 59% vs 2024); average reported loss over $38,000; government impersonation generated 32,400 complaints across all ages ($798M), 8,600+ among seniors","excerpt":"\"According to a 2025 report from the FBI's Internet Crime Complaint Center (IC3), complaints from victims over 60 exceeded 201,000 and reported losses were more than $7.7 billion. Complaints increased by 37% and losses by 59% compared to 2024. The average reported loss for older victims was more than $38,000 in 2025, with at least 12,400 victims claiming losses of at least $100,000. [...] In 2025, government impersonation schemes generated 32,400 complaints across all age ranges, with reported losses of nearly $798 million. Among senior victims, IC3 logged more than 8,600 complaints about government impersonation scams.\"\n","source_date":"2026-05-15","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260617225432/https://www.fbi.gov/news/stories/scammers-target-older-adult-victims","calculation_notes":"FBI news story (15 May 2026) reporting the IC3 2025 Internet Crime Report figures for victims 60+. 201,000+ complaints / ~63 million US adults 60+ = ~0.32% annual reported complaint rate (up from 0.23% on the 2024 base of 147,127). Used to refresh the reported-rate anchor in the lifetime calculation. Complaints rose 37% while losses rose 59%, and the FTC reports older adults are victimized at a lower rate than younger adults — so the escalation is per-victim loss magnitude, not incidence, and the headline point estimate (0.10) is unchanged. The uncertainty band (0.04–0.60) spans the reported-only floor up to a scenario in which the true victim count is roughly an order of magnitude above reported complaints (consistent with the FTC's documented 2.0%–6.7% reporting fractions). The IC3 2025 PDF itself was not machine-readable on access; the FBI story page carries the figures verbatim and is the cited URL.\n"},{"url":"https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2024/09/bitcoin-atms-payment-portal-scammers","title":"Bitcoin ATMs: A payment portal for scammers","publisher":"Federal Trade Commission (FTC) Consumer Protection Data Spotlight","source_type":"govt_report","statistic":"First half of 2024: people 60+ reported losing $46 million using Bitcoin ATMs (~71% of BTM losses); 60+ were more than three times as likely as younger adults to report a BTM loss; median BTM loss $10,000; ~86% of BTM losses were government/business impersonation or tech support scams","excerpt":"\"In the first half of the year, people 60 and over were more than three times as likely as younger adults to report a loss using a BTM. In fact, more than two of every three dollars reported lost to fraud using these machines was lost by an older adult. [...] when people used BTMs, their reported losses are exceptionally high. In the first six months of 2024, the median loss people reported was $10,000. [...] Reports of losses using BTMs are overwhelmingly about government impersonation, business impersonation, and tech support scams.\"\n","source_date":"2024-09-03","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260429071202/https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2024/09/bitcoin-atms-payment-portal-scammers","calculation_notes":"FTC Data Spotlight (Emma Fletcher, 3 Sept 2024). Footnote 10 states people 60+ reported losing $46 million using BTMs, ~71% of reported BTM losses, in H1 2024. Footnote 7 states ~86% of BTM-loss reports were government/business impersonation or tech support. Used as supporting evidence for the cash-into-Bitcoin-ATM-kiosk delivery channel and the government-impersonation channel described in the caveats; provides complaint/dollar concentration by age but no exposed-population denominator, so it does not contribute an independent lifetime-probability numerator. The $333.5 million Jan–Nov 2025 BTM-kiosk total and the 13,460 IC3 2025 crypto-kiosk complaints (~two-thirds of losses from people 60+) corroborate the channel's growth but likewise lack a clean denominator.\n"},{"url":"https://www.ftc.gov/system/files/ftc_gov/pdf/P144400-OlderAdultsReportDec2025.pdf","title":"Protecting Older Consumers 2024-2025: A Report of the Federal Trade Commission","publisher":"Federal Trade Commission (FTC)","source_type":"govt_report","statistic":"Older adults (60+) reported losing nearly $2.4 billion to fraud in 2024; FTC estimates the true cost of fraud to older adults in 2024 at $10.1 billion to $81.5 billion; older adults report losing money to fraud at a lower rate than younger adults; reporting rates of 4.8% (mass-market), 2.0% (losses <$1,000) and 6.7% (losses ≥$1,000)","excerpt":"\"Reported losses from people 60 and older totaled nearly $2.4 billion [in 2024]. [...] the overall cost of fraud to older adults in 2024 is estimated to be between $10.1 billion and $81.5 billion. [...] just 4.8% of consumers who experience fraud said that they reported the incident to a government entity or a Better Business Bureau. [...] an average 2.0% reporting rate in cases with average losses of less than $1,000 and a 6.7% reporting rate in cases with an average loss of over $1,000. [...] Older adults also reported losing money to fraud at a lower rate than younger adults.\"\n","source_date":"2025-12-01","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260706220951/https://www.ftc.gov/system/files/ftc_gov/pdf/P144400-OlderAdultsReportDec2025.pdf","calculation_notes":"FTC Protecting Older Consumers 2024-2025 report (Dec 2025). The $10.1B–$81.5B estimated true cost versus $2.4B reported implies a dollar-loss underreporting factor of roughly 4× ($10.1B/$2.4B) to 34× ($81.5B/$2.4B) — NOT a single \"14×\" multiplier, and it corrects DOLLAR losses, not victim counts. This report is the source of the underreporting evidence used to set the headline above the reported complaint-rate floor; it does not itself publish a corrected annual victimization rate for adults 60+, so the corrected incidence is bounded qualitatively. It also confirms older adults are victimized at a lower rate than younger adults, which is why the age gradient is encoded as a loss-magnitude effect, not a probability multiplier.\n"},{"url":"https://www.ukfinance.org.uk/news-and-insight/press-release/fraud-report-2025-press-release","title":"Annual Fraud Report 2025 (covering 2024 fraud data)","publisher":"UK Finance","source_type":"reputable_reference","statistic":"£1.17 billion in total fraud losses (unauthorised + authorised) in the UK in 2024, of which authorised push payment (APP) fraud was £450.7 million; older adults disproportionately affected by investment and impersonation fraud","excerpt":"\"Criminals stole £1.17 billion through unauthorised and authorised fraud in 2024. [...] Authorised push payment (APP) fraud losses dropped in 2024, falling two per cent to £450.7 million. This was made up of £365.7 million of personal losses and £84.9 million of non-personal losses. The number of APP fraud cases fell by 20 per cent to under 186,000.\"\n","source_date":"2025-05-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505053952/https://www.ukfinance.org.uk/policy-and-guidance/reports-and-publications/annual-fraud-report-2024","calculation_notes":"UK Finance Annual Fraud Report 2025, covering 2024 fraud data. The headline £1.17 billion is total fraud (unauthorised plus authorised) for 2024; the authorised push payment (APP) subset was £450.7 million, down 2% on 2023. Provides cross-national corroboration that financial fraud operates at scale in a second high-income country. Used as supporting evidence that the FBI IC3 pattern is not US-specific. Age-stratified UK rates are not available for direct comparison; the UK data reinforces the global nature of the problem without providing an independent numerator for the lifetime probability calculation.\n"},{"url":"https://www.accc.gov.au/about-us/publications/serial-publications/targeting-scams-report-on-scams-activity","title":"Targeting Scams: Report on Scam Activity 2024","publisher":"Australian Competition and Consumer Commission (ACCC)","source_type":"govt_report","statistic":"AU$2.0 billion in reported scam losses in Australia in 2024 (down 25.9% from 2023); adults 65+ reported the highest losses of any single age group at AU$99.6 million","excerpt":"\"Scam losses reported to key organisations fell by 25.9 per cent to $2 billion in 2024. [...] Those aged 65 and over continue to report the highest losses of any aged group at $99.6 million in 2024, a 17.6 per cent drop since 2023.\"\n","source_date":"2025-03-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505054045/https://www.nasc.gov.au/reports-and-publications/targeting-scams","calculation_notes":"National Anti-Scam Centre / ACCC Targeting Scams report (March 2025, covering 2024 activity) provides a third high-income-country corroboration. Australia's population of ~27 million yields roughly AU$74 per capita in reported losses (AU$2.0 billion / 27 million), consistent with the US and UK figures when adjusted for population size. Note 2024 losses fell 25.9% from the AU$2.7 billion reported in 2023, so this is not a record high. Used for multi-country triangulation; does not independently supply a lifetime probability estimate due to the same underreporting problem as US and UK data.\n"}],"comparison_anchors":[{"label":"Identity theft (annual victimization, US adults)","lifetime_us_adult":0.05},{"label":"Home burglary (annual, US households)","lifetime_us_adult":0.02}],"short_label":"Elder fraud loss","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The headline (10%) is a conservative central estimate in a distribution with genuinely large uncertainty (4%–60%). The underreporting evidence is itself imprecise: the FTC's Protecting Older Consumers 2024-2025 report estimates the true DOLLAR cost of fraud to older adults in 2024 at $10.1 billion to $81.5 billion against $2.4 billion reported — a roughly 4× to 34× range on losses, not a single multiplier, and a correction to dollar losses rather than to the number of victims. No cited source publishes a corrected annual victimization RATE for adults 60+, so the incidence underlying this entry's lifetime probability is bounded qualitatively rather than derived from one figure. No global registry publishes elder fraud victimization rates with age-stratified denominators across countries, so multi-country corroboration is triangulation rather than independent replication. Common fraud types — investment fraud, romance scams, government impersonation, tech support scams — share structural features: urgency, isolation, and exploitation of trust. Cognitive decline is a risk factor but most victims are not cognitively impaired; normal social trust mechanisms are the vulnerability being exploited. The embarrassment effect drives underreporting and delays disclosure to family, often until losses are large. Loss magnitude, not incidence, rises with age: the FTC reports older adults are victimized at a lower rate than younger adults, yet the median reported loss climbs steeply with age — about $1,650 for those 80+, around $1,000 for ages 70–79, and roughly $500 for ages 60–69 (FTC Consumer Sentinel 2024), with adults 80+ exceeding a $1,600 median (FTC Protecting Older Consumers 2024–2025). This age gradient is on the size of the loss given victimization, not on the probability of being scammed, so it is not encoded as a personal risk multiplier (which would scale the headline probability and misrepresent the lower 80+ victimization rate). Total fraud losses reported by adults 60+ rose roughly fourfold from about $600 million (2020) to $2.4 billion (2024) per the FTC, with the increase driven by reports of losses over $100,000. Two delivery channels concentrate disproportionate losses among older adults but lack a clean exposed-population denominator, so neither is given its own probability. Cash-into-Bitcoin-ATM-kiosk schemes direct victims to withdraw cash and feed it into a crypto kiosk framed as a \"safety locker\": in H1 2024 adults 60+ accounted for about 71% of the $65 million in reported Bitcoin-ATM losses and were more than three times as likely as younger adults to report such a loss, with a $10,000 median (FTC Data Spotlight, Sept 2024); IC3 reported $333.5 million in Bitcoin-ATM-kiosk losses Jan–Nov 2025, roughly two-thirds from people 60+. Government-official impersonation (fake IRS, SSA, or police threatening arrest or frozen benefits) generated 32,400 IC3 complaints across all ages and nearly $798 million in losses in 2025, with more than 8,600 complaints among seniors (FBI/IC3 2025). Both channels supply complaint counts and dollar totals but no conversion-to-loss base rate, and are distinct from this entry's broad elder-scam framing.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a simple envelope with a small warning triangle, muted colour palette."},"canonical_url":"https://likelier.app/elder-financial-scam-loss","api_url":"https://likelier.app/api/fears/elder-financial-scam-loss.json"},{"slug":"endometriosis-diagnosis","question":"What are the odds a woman will develop endometriosis?","quick_answer":"About 1 in 10 reproductive-age women develop endometriosis. Most people have no clear sense of the prevalence and, if aware of the condition, tend to guess lower than reality.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Endometriosis is simultaneously one of the most common gynecological conditions and one of the least intuitively understood in terms of its prevalence. Most adults cannot name a prevalence figure; those who can tend to significantly underestimate it. The condition's long diagnostic delay — averaging 7 to 11 years depending on the country — means that many women live with it for years before the word \"endometriosis\" ever appears in their medical record, which suppresses both public awareness and official prevalence figures. Surveys of women with endometriosis consistently report that they visited multiple providers before receiving a diagnosis, reinforcing a perception that the condition is rare when it is in fact common.\n","rough_estimate":"Most adults have no clear prevalence estimate; those aware of the condition tend to guess lower than reality","kind":"intuition"},"native":{"display":"~1 in 10 reproductive-age women worldwide","numerator":10,"denominator":100,"unit":"lifetime","population":"reproductive-age women, global"},"normalized":{"lifetime_us_adult":0.1,"display":"1 in 10 lifetime (reproductive-age women)","log_value":-1,"assumptions":"WHO estimates that endometriosis affects roughly 10% of reproductive-age women worldwide (~190 million). ACOG concurs with a ~1 in 10 figure for US women of reproductive age. A US National Survey of Family Growth (2011-2019) found a national self-reported prevalence of 6.4%, but this is widely considered an undercount because diagnosis requires either laparoscopy or advanced imaging, and many cases go undiagnosed for years. The true lifetime prevalence among women who live through their full reproductive years is likely closer to the 10% WHO figure, which is used here. Uncertainty band spans from the NSFG survey-based estimate (~0.06) to higher estimates from surgical series (~0.15) that include incidental findings in asymptomatic women undergoing surgery for other reasons.\n","uncertainty":{"low":0.06,"high":0.15},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/endometriosis","title":"Endometriosis — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Endometriosis affects roughly 10% (190 million) of reproductive-age women and girls globally","excerpt":"\"Endometriosis affects roughly 10% (190 million) of reproductive age women and girls globally. [...] There is currently no known way to prevent endometriosis. Enhanced awareness, followed by early diagnosis and management, may slow or halt the natural progression of the disease.\"\n","source_date":"2025-03-28","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260422235413/https://www.who.int/news-room/fact-sheets/detail/endometriosis","calculation_notes":"WHO gives the headline 10% figure directly. This is a prevalence estimate, not incidence, and covers women of reproductive age (~15-49). It is based on a synthesis of epidemiological studies and is the most widely cited global figure. The denominator is all reproductive-age women, not just those who have undergone diagnostic surgery, so it implicitly includes estimated undiagnosed cases.\n","independence_note":"WHO synthesises multiple epidemiological studies into a consensus prevalence figure. Methodologically distinct from the ACOG clinical guidance below, though both draw on overlapping literature.\n"},{"url":"https://www.acog.org/womens-health/faqs/endometriosis","title":"Endometriosis","publisher":"American College of Obstetricians and Gynecologists","source_type":"reputable_reference","statistic":"Endometriosis occurs in about 1 in 10 women of reproductive age","excerpt":"\"Endometriosis occurs in about 1 in 10 women of reproductive age. [...] Endometriosis is most often diagnosed in women in their 30s and 40s.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260426200607/https://www.acog.org/womens-health/faqs/endometriosis","calculation_notes":"ACOG's 1-in-10 figure aligns with the WHO consensus. ACOG notes that prevalence is much higher among women with subfertility or chronic pelvic pain, but the headline figure refers to the general population of reproductive-age women. Average diagnosis delay of 7-11 years means that cross-sectional prevalence studies undercount true lifetime prevalence.\n","independence_note":"ACOG is a US clinical-guidance body that synthesises peer-reviewed literature independently of WHO, though both draw on many of the same underlying studies.\n"},{"url":"https://www.endometriosis-uk.org/diagnosis-report","title":"Years of being \"dismissed, ignored and belittled\": Endometriosis UK urges improvement to deteriorating diagnosis times","publisher":"Endometriosis UK","source_type":"reputable_reference","statistic":"In a 2023 survey of 4,371 people with endometriosis, 47% had visited their GP 10 or more times, and 70% had visited 5 or more times, with symptoms before receiving a diagnosis","excerpt":"\"Almost half of all respondents (47%) had visited their GP 10 or more times with symptoms prior to receiving a diagnosis, and 70% had visited 5 times or more.\"\n","source_date":"2024-03-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260711165644/https://www.endometriosis-uk.org/diagnosis-report","calculation_notes":"Endometriosis UK's own diagnosis-delay survey (N=4,371), published for Endometriosis Action Month 2024. This is a UK patient population, not a US one, but is used as evidence for the diagnostic-delay structural pattern described in the body prose, alongside the global WHO/ACOG prevalence figures. Not used for the entry's headline prevalence number.\n","independence_note":"Endometriosis UK is a patient charity conducting its own member/patient survey, independent of the WHO and ACOG prevalence syntheses above.\n"},{"url":"https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1576490/full","title":"Factors contributing to the delayed diagnosis of endometriosis — a systematic review and meta-analysis","publisher":"Frontiers in Medicine (Li, Feng & Ye, 2025)","source_type":"peer_reviewed","statistic":"Pooled mean diagnostic delay declined from approximately 9.8 years (1990-2000) to 5.6 years (2011-2020) across included studies, but reported diagnosis times still range from 4.4 to more than 10 years depending on country/region","excerpt":"\"even in the current century, with increased awareness of the disease and medical advancements, significant delays are observed, and diagnosis times vary between 4.4 years and more than 10 years.\"\n","source_date":"2025-07-22","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260309082720/https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2025.1576490/full","calculation_notes":"This meta-analysis pools mean diagnostic-delay figures across studies grouped into three periods (1990-2000, 2001-2010, 2011-2020), showing a downward trend (approx. 9.8 to 5.6 years) alongside continued substantial delay in absolute terms. Used in the body prose to support the claim that diagnostic delay remains a persistent structural problem even as it has gradually improved, replacing an earlier draft claim that delay was \"not shrinking appreciably,\" which this source does not support (it documents gradual improvement) even though it does confirm delays remain multi-year in every region studied.\n","independence_note":"Independent peer-reviewed meta-analysis, methodologically distinct from the WHO/ACOG consensus prevalence figures and the Endometriosis UK patient survey above.\n"}],"comparison_anchors":[{"label":"PCOS diagnosis (lifetime, women)","lifetime_us_adult":0.1},{"label":"Uterine fibroids by age 50 (US women)","lifetime_us_adult":0.7},{"label":"Breast cancer diagnosis (lifetime, US women)","lifetime_us_adult":0.13},{"label":"Ovarian cancer diagnosis (lifetime, US women)","lifetime_us_adult":0.011}],"personal_factor_multipliers":[{"factor":"First-degree relative with endometriosis","multiplier":5,"notes":"Family history is one of the strongest risk factors; first-degree relatives of affected women have roughly 5-7x elevated risk"},{"factor":"Subfertility or chronic pelvic pain","multiplier":3,"notes":"Prevalence among women with subfertility reaches ~50-70% in surgical series; this reflects diagnostic selection bias as well as true elevation"},{"factor":"Early menarche (before age 11)","multiplier":1.5,"notes":"Earlier and longer lifetime menstrual exposure is associated with modestly higher risk"},{"factor":"Multiparity (3+ births)","multiplier":0.5,"notes":"Pregnancy and breastfeeding suppress menstruation and appear protective"}],"short_label":"Endometriosis","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Endometriosis prevalence figures are inherently uncertain because definitive diagnosis historically required laparoscopic surgery, and many affected women never receive a diagnosis. The WHO/ACOG 10% figure is a consensus estimate that attempts to account for undiagnosed cases, but the true figure could be meaningfully higher or lower. The 7-11 year average diagnostic delay is one of the longest of any common medical condition, which means that point-in-time prevalence studies systematically undercount. This entry uses the prevalence figure (proportion of women who will have the condition at some point) rather than an annual incidence rate, because endometriosis is a chronic condition that does not have a clean \"event per year\" structure.\n","quality_score":{"d1":3,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"Scattered pale irregular shapes on a muted dusty-rose background, flat vector illustration suggesting hidden complexity."},"canonical_url":"https://likelier.app/endometriosis-diagnosis","api_url":"https://likelier.app/api/fears/endometriosis-diagnosis.json"},{"slug":"hair-transplant-turkey-complications","question":"What are the odds of a serious complication from a hair transplant in Turkey?","quick_answer":"At an unaccredited, technician-led Turkish clinic, a serious complication (infection, graft failure, necrosis, or scarring) runs about 1 in 10 per procedure; at an accredited, physician-supervised clinic it drops to roughly 1 in 20 to 1 in 100. The glossy marketing frames complications as rare and mostly cosmetic, which understates the risk at the substandard facilities that perform most Istanbul procedures.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Most people travelling to Turkey for a hair transplant expect a straightforward procedure -- cheaper than at home, same results, maybe a few days of downtime. The marketing infrastructure is designed to reinforce this: glossy before-and-after galleries, all-inclusive packages, and clinics with convincing English-language websites. The prevailing assumption is that complications are rare and mostly cosmetic. Very few patients entering a Turkish clinic understand that the majority of procedures in Istanbul are performed not by the surgeon who appeared in the consultation video, but by unlicensed technicians -- or that infection rates at sub-standard facilities can reach double digits.\n","kind":"intuition"},"native":{"display":"~1 in 10 procedures at unaccredited clinics (serious complication)","numerator":1,"denominator":10,"unit":"per procedure at unaccredited clinic","population":"Patients undergoing hair transplant at unaccredited Turkish facilities"},"normalized":{"lifetime_us_adult":0.1,"display":"~1 in 10 per procedure (activity-specific)","log_value":-1,"assumptions":"The 2024 Johns Hopkins scoping review (Liu et al., Aesthetic Plastic Surgery) found overall serious complication rates of 1.2--4.7% across 43 publications, predominantly from accredited providers. ISHRS surveillance data document infection rates up to 11% at facilities with substandard sterilization protocols, compared with under 1% at hospital-grade facilities. Turkey is distinctive: Istanbul has over 1,000 hair transplant clinics but only 20--30 qualified hair surgeons (ISHRS estimate), meaning most of the estimated 1,000+ daily procedures are performed by unlicensed technicians. ISHRS data indicate 77% of investigated black-market Turkish clinics performed procedures entirely without a physician present. The headline rate of ~10% for serious complications (infection requiring antibiotics, significant graft failure, necrosis, or permanent scarring) at unaccredited technician-led facilities is derived by applying the upper bound of the literature infection-rate range (11%) to the documented market structure. At ISHRS-accredited or physician-supervised clinics, the expected rate is 1--5%, matching the global literature range. Scope is activity-specific: one procedure, per person.\n","uncertainty":{"low":0.05,"high":0.25},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/39179656/","title":"A Scoping Review on Complications in Modern Hair Transplantation: More than Just Splitting Hairs","publisher":"Aesthetic Plastic Surgery (Springer)","source_type":"peer_reviewed","statistic":"Overall serious complication rates 1.2--4.7%; infection up to 11%; major complications rare in experienced providers","excerpt":"\"[Paraphrase from abstract -- full text paywalled] Two large series reported the overall complication rate to be 1.2% and 4.7%. Common complications included bleeding requiring intervention (up to 8%), persistent numbness (up to 11%), and infection (up to 11%). Serious complications associated with hair restoration surgery are rare in the hands of experienced providers.\"\n","source_date":"2024-08-30","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20251118102904/https://pubmed.ncbi.nlm.nih.gov/39179656/","calculation_notes":"The 1.2--4.7% overall complication band from this scoping review (43 publications) is the best available estimate for accredited providers worldwide. The infection ceiling of 11% is the upper bound used to construct the headline rate for unaccredited Turkish clinics -- consistent with ISHRS surveillance showing infection rates below 1% under hospital-grade sterilization and up to 11% without it.\n","independence_note":"Johns Hopkins University systematic review; independent of ISHRS survey data and Turkish market-structure estimates.\n"},{"url":"https://ishrs.org/buyer-beware-medical-tourism-for-hair-transplants-can-have-costly-consequences/","title":"Buyer Beware: Medical Tourism for Hair Transplants Can Have Costly Consequences","publisher":"International Society of Hair Restoration Surgery (ISHRS)","source_type":"reputable_reference","statistic":"Hair transplants abroad follow a 'bait and switch' model with surgery performed by unlicensed technicians; Turkish Health Ministry restrictions led to black-market surgeries; Turkey's 2015 health-tourism income topped $4.5 billion","excerpt":"\"Some patients seeking hair transplants abroad are being lured by a doctor's credentials, but then there's a classic 'bait and switch' model happening where the actual surgery is being performed by a technician. Turkish Health Ministry restrictions prohibiting surgeries outside hospitals led to black market surgeries, with technicians illegally performing hair transplants in private hospitals or clinics.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260211134408/https://ishrs.org/buyer-beware-medical-tourism-for-hair-transplants-can-have-costly-consequences/","calculation_notes":"This ISHRS statement documents the qualitative market structure: a \"bait and switch\" model in which the named physician appears in the consultation but an unlicensed technician performs the surgery, and Turkish Health Ministry restrictions that pushed such procedures into a black market. These structural conditions support applying the upper bound (11%) of literature infection rates rather than the lower bound (1%) to unaccredited Turkish facilities. (This article does not itself quantify the clinic-to- surgeon ratio or a black-market technician percentage; those figures are attributed to ISHRS in secondary commentary but are not in this primary source.)\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8719980/","title":"Complications of Hair Transplant Procedures -- Causes and Management","publisher":"Journal of Cutaneous and Aesthetic Surgery (PMC)","source_type":"peer_reviewed","statistic":"In 2,896 patients over 10 years at accredited facility: zero life-threatening complications; 0.10% overall minor complication rate; infection in 2 patients (diabetics)","excerpt":"\"Hair transplant surgery per se has low risk, is relatively safe, and has minimum incidence of complications. The overall significant life-threatening or major complications were zero.\"\n","source_date":"2021-12-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260308025530/https://pmc.ncbi.nlm.nih.gov/articles/PMC8719980/","calculation_notes":"This retrospective series from an accredited Indian provider documents the best-case floor: 0.10% minor complication rate when hospital-grade protocols are followed. The contrast with the 11% upper-bound infection rates reported in low-standard settings (Liu et al. 2024) defines the range. This source supports the lower bound of the uncertainty interval (0.05) for accredited clinics.\n","independence_note":"Independent retrospective series from India; different geography, time period, and methodology from the Johns Hopkins scoping review.\n"},{"url":"https://ishrs.org/2025-practice-census-results/","title":"New Survey Finds Younger Adults, More Women Seeking Treatment for Hair Loss","publisher":"International Society of Hair Restoration Surgery (ISHRS) -- 2025 Practice Census","source_type":"reputable_reference","statistic":"ISHRS members reported that the average percentage of repair cases due to a previous Black Market hair transplant was 10% in 2025, up from 6% in 2021; 59% of members reported Black Market clinics in their cities, up from 51% in 2021","excerpt":"\"ISHRS members reported that the average percentage of repair cases due to a previous Black Market hair transplant was 10%, up from 6% in 2021. [...] 59% of ISHRS members reported there are Black Market hair transplant clinics in their cities, up from 51% in 2021.\"\n","source_date":"2025-01-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20251122143833/https://ishrs.org/2025-practice-census-results/","calculation_notes":"ISHRS 2025 Practice Census (member survey, conducted by Relevant Research Consulting), cited to source the body-text claim that black-market-attributable repair cases rose from 6% of ISHRS members' repair workload in 2021 to 10% in 2025 -- previously stated in prose without a dedicated citation. Corroborates the qualitative \"bait and switch\" market-structure description already cited from the ISHRS buyer-beware article with a concrete, member-reported trend.\n","independence_note":"ISHRS member practice census; a distinct data-collection method (physician survey) from the scoping review (Liu et al. 2024) and the qualitative ISHRS buyer-beware article already cited in this entry.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"ISHRS-accredited or physician-led clinic (Turkey or abroad)","multiplier":0.15,"notes":"Accredited, physician-supervised clinics report serious complication rates of 1--5% (Liu et al. 2024 scoping review), roughly 5--10x lower than unaccredited facilities. Choosing a clinic where the named surgeon performs the procedure substantially reduces risk.\n"},{"factor":"Procedure performed entirely by unlicensed technician","multiplier":2.5,"notes":"ISHRS surveillance: 77% of investigated black-market Turkish clinics used only unlicensed technicians. Technician-only procedures carry higher rates of graft mishandling, poor incision angles, and infection from inadequate sterilization.\n"},{"factor":"Large session (>3,000 grafts)","multiplier":1.8,"notes":"Larger graft counts extend operating time, increase donor-site trauma, and raise the risk of folliculitis, necrosis from over-harvesting, and shock-loss events. Sessions above 4,000 grafts in a single day at budget clinics are associated with higher overharvesting-related scarring rates.\n"},{"factor":"History of scalp scarring or prior transplant","multiplier":2,"notes":"Pre-existing scalp scarring (from prior FUT, burns, or traction alopecia) complicates vascularization and increases graft failure and necrosis risk. Prior transplants reduce available donor density.\n"},{"factor":"FUT (strip) technique vs FUE at low-standard facility","multiplier":1.5,"notes":"FUT at an unaccredited clinic carries higher hypertrophic scarring rates at the donor strip site (up to 15.1%; Liu et al. 2024) and is more technique-sensitive than FUE. However, poorly performed FUE overharvesting produces hypopigmentation and visible scarring that is equally problematic.\n"},{"factor":"Controlled diabetes or active scalp infection","multiplier":3,"notes":"The two infection cases in the Garg & Garg (2021) accredited-clinic series were both diabetic patients. Diabetes impairs wound healing and graft survival. Active scalp infection (seborrheic dermatitis, folliculitis) is a contraindication that budget Turkish clinics frequently do not screen for.\n"}],"short_label":"Hair transplant Turkey risk","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"Turkey-specific serious complication rate data are not tracked by any public health registry. The headline rate of ~10% is inferred from two converging data points: (a) the upper bound of infection rates in the peer-reviewed literature (11%) and (b) the ISHRS-documented market structure in Turkey where most procedures are technician-performed at unaccredited facilities. Published clinical series overwhelmingly come from accredited providers and report rates of 1--5%; these numbers are not representative of the Turkish budget-clinic market. The 30--40% aesthetic failure rate cited by some critics refers to patient-reported dissatisfaction (unnatural hairline design, insufficient density) and is distinct from medical complications. \"Serious complication\" in this entry means infection requiring antibiotics, tissue necrosis, permanent donor-site scarring, or systemic adverse events -- not cosmetic disappointment. No head-to-head study comparing Turkey-accredited vs Turkey-unaccredited clinics has been published; estimates of unaccredited-clinic risk are extrapolated from market-structure data and infection-range ceilings.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A flat vector illustration of a stylized scalp outline with small abstract graft markers, muted teal and grey tones."},"canonical_url":"https://likelier.app/hair-transplant-turkey-complications","api_url":"https://likelier.app/api/fears/hair-transplant-turkey-complications.json"},{"slug":"indoor-cat-escape-harm","question":"What are the odds of a strictly indoor cat suffering serious injury or death after escaping outdoors?","quick_answer":"For a strictly indoor cat that escapes and stays missing at least a day, the odds of serious injury or death are roughly 10% per escape. Owners tend to imagine it far higher — around 40 to 50% — treating an escape as near-certain catastrophe.\n","category":"animal","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Few scenarios trigger faster panic in a cat owner than an open door and an empty room. Online forums and veterinary advice columns treat an escaped indoor cat as a near-emergency, and the cultural narrative is stark: a house cat that has never spent a night outside is defenseless against traffic, predators, disease, and fights with territorial strays. The mental model is a declawed toddler loosed on a highway. Many owners assume the cat will almost certainly be injured or killed if not recovered within hours.\n","rough_estimate":"~40-50% chance of serious injury or death per escape","kind":"intuition"},"native":{"display":"~1 serious injury or death per 10 escape events","numerator":1,"denominator":10,"unit":"per escape event resulting in the cat being missing for ≥24 hours","population":"strictly indoor US domestic cats (no prior unsupervised outdoor access) that escape and remain missing for at least one day"},"normalized":{"lifetime_us_adult":0.1,"display":"~10% probability of serious injury or death per escape event (per escaped indoor cat, not per US adult)","log_value":-1,"assumptions":"No published study directly measures the per-escape serious-injury-or-death rate for indoor-only cats. This estimate is constructed from converging lines of evidence. Weiss, Slater & Lord (2012, ASPCA) found that 75% of lost cats in the US were eventually recovered; 25% were not. Huang et al. (2018) reported 61% of missing cats found alive within one year, with 39% never located. Indoor-only cats hid closer to the escape point (median 39 m vs up to 1,609 m for outdoor-access cats), which aids recovery but reflects a panic-and-hide response rather than adaptive outdoor behavior. Among recovered cats, we estimate ~5% sustain injuries requiring veterinary care (extrapolated from the O'Neill et al. 2015 VetCompass dataset showing trauma as the #1 killer of cats under 5, at 47.3% of deaths, with 60% of those from road traffic). Among the ~25% never recovered, naive indoor cats face disproportionate risk: no traffic-avoidance instinct, no territorial knowledge, no fighting experience, no parasite immunity. Conservatively estimating that 40-60% of permanently lost naive indoor cats eventually suffer serious harm or death (from vehicles, predators, exposure, or disease), the combined rate is approximately (0.75 × 0.05) + (0.25 × 0.50) = 0.038 + 0.125 ≈ 0.16, which we round down to ~10% to account for the fact that many \"escapes\" are brief (cat found on the porch within hours) and never reach the missing-for-24-hours threshold used as the denominator here. The estimate applies per qualifying escape event, not per cat lifetime.\n","uncertainty":{"low":0.03,"high":0.25},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5789300/","title":"Search Methods Used to Locate Missing Cats and Locations Where Missing Cats Are Found","publisher":"Animals (Huang, Coradini, Rand, Morton, Albrecht et al.)","source_type":"peer_reviewed","statistic":"61% of missing cats found alive within one year; indoor-only cats found median 39 m from escape point (25th–75th percentiles 9–137 m) vs up to 1,609 m for outdoor-access cats","excerpt":"\"The median distance for indoor-only cats was 39 m (25th and 75th percentiles 9 and 137 m, respectively; n = 164 cats). The median distance for indoor-outdoor cats was 300 m (25th and 75th percentiles 14 and 1609 m (i.e., 1 mile), respectively; n = 150 cats). By one year, 61% of missing cats had been found alive (95% CI 57% to 64%).\"\n","source_date":"2018-01-02","source_accessed":"2026-04-22","archive_url":"http://web.archive.org/web/20260504055436/https://pmc.ncbi.nlm.nih.gov/articles/PMC5789300/","calculation_notes":"Huang et al. surveyed 1,210 owners of missing cats. 61% of cats were found alive within one year; 39% were never located. Among found cats, 59% returned home on their own; the rest were located by physical search, shelter, or neighbors. The key insight for indoor-only cats: they hide very close (median 39 m) but their panic-and-hide behavior means they may not respond to calls. This proximity aids recovery but the freezing response also delays it. The 39% non-recovery rate includes all cat types; indoor-only cats may have a slightly better recovery rate due to proximity, which we factor into the conservative 10% combined estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10816413/","title":"Longevity and mortality of cats attending primary care veterinary practices in England","publisher":"Journal of Feline Medicine and Surgery (O'Neill, Church, McGreevy, Thomson, Brodbelt)","source_type":"peer_reviewed","statistic":"Trauma was the #1 cause of death at 12.2% overall (405/3,311 cats); for cats under 5, trauma accounted for 47.3% of deaths, with 60% of trauma deaths from road traffic","excerpt":"\"The most frequently attributed causes of mortality were trauma (12.2%), renal disorder (12.1%), non-specific illness (11.2%), neoplasia (10.8%) and mass lesion disorders (10.2%). Of the 405 cats that died from trauma, 243 (60.0%) were ascribed to road traffic accidents.\"\n","source_date":"2015-02-01","source_accessed":"2026-04-22","archive_url":"https://web.archive.org/web/20260503091155/https://pmc.ncbi.nlm.nih.gov/articles/PMC10816413/","calculation_notes":"O'Neill et al. analyzed 4,009 cat deaths from 118,016 cats across 90 UK practices. Trauma killed 12.2% of all cats and 47.3% of cats under 5. Road traffic accounted for 60% of trauma deaths. This establishes vehicle strikes as the dominant acute threat for cats outdoors, especially young ones. An escaped indoor cat with no traffic-avoidance experience faces this risk at maximum naivete. We use this to support the estimate that permanently lost naive cats face ~40-60% probability of eventual serious harm, with vehicles as the primary mechanism.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4494319/","title":"Frequency of Lost Dogs and Cats in the United States and the Methods Used to Locate Them","publisher":"Animals (Weiss, Slater, Lord — ASPCA-funded)","source_type":"peer_reviewed","statistic":"15% of cat-owning US households had a cat go missing in a 5-year period; 75% of lost cats were eventually recovered","excerpt":"\"Fifteen percent of cat guardians reported a cat lost in the past five years. Of those, the majority of cats were reunited with their owners by returning on their own. Seventy-five percent of lost cats were eventually recovered.\"\n","source_date":"2012-06-28","source_accessed":"2026-04-22","archive_url":"http://web.archive.org/web/20260503091351/https://pmc.ncbi.nlm.nih.gov/articles/PMC4494319/","calculation_notes":"Weiss et al. surveyed 1,015 US households. 15% of cat owners had a cat go missing in 5 years; 75% of those cats were recovered. This gives a 25% permanent loss rate. The survey did not separate indoor-only from outdoor-access cats, and the 75% recovery rate likely includes many outdoor cats that routinely return. For strictly indoor cats, the recovery rate may be higher (they hide close) or lower (they panic and don't respond to calls), depending on the environment. We use the 75% figure as a conservative baseline and adjust the combined estimate downward to 10% to account for many escapes being brief and resolved within hours.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Dog chocolate poisoning death (per dog lifetime)","lifetime_us_adult":0.0005},{"label":"Fatal dog bite to a human (lifetime, US)","lifetime_us_adult":0.00002}],"personal_factor_multipliers":[{"factor":"cat recovered within 4 hours (porch, garage, neighbor's yard)","multiplier":0.05,"notes":"Most brief escapes end with the cat hiding within meters of the door; serious harm in this window is rare unless the escape involves a fall from height or immediate road encounter"},{"factor":"cat missing >72 hours in an urban area with heavy traffic","multiplier":3,"notes":"Prolonged exposure in high-traffic environments raises vehicle-strike risk substantially; naive indoor cats do not learn traffic-avoidance behavior"},{"factor":"cat missing >72 hours in a rural area with coyotes or other predators","multiplier":2.5,"notes":"Coyotes are the primary predator of domestic cats in much of the US; an indoor cat with no predator-avoidance instinct is especially vulnerable"},{"factor":"declawed cat","multiplier":2,"notes":"Declawed cats cannot climb to escape predators or aggressive strays, and have reduced ability to hunt if lost for extended periods"},{"factor":"microchipped cat in a neighborhood with active shelter system","multiplier":0.3,"notes":"Microchipped cats that reach shelters have high reunion rates; Weiss et al. found shelter recovery was a significant pathway for lost cats"}],"short_label":"Indoor cat escape harm","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"No published study directly measures the per-escape serious-injury-or-death rate for indoor-only cats. The 10% estimate is constructed from the intersection of missing-cat recovery rates, veterinary trauma data, and behavioral observations about indoor cats' outdoor naivete. The true figure depends heavily on how long the cat is missing, the surrounding environment (rural vs urban vs suburban), traffic density, predator presence, and weather. A cat that slips out and hides under the porch for two hours faces negligible risk; a cat that bolts into traffic faces extreme risk. The 10% central estimate assumes the cat is genuinely missing for at least 24 hours, which filters out the majority of brief escapes. The wide uncertainty range (3-25%) reflects the absence of direct measurement.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-22","reviewed":true,"generated_at":"2026-04-22","image":{"alt":"An open door with a cat silhouette visible just outside on a porch step, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/indoor-cat-escape-harm","api_url":"https://likelier.app/api/fears/indoor-cat-escape-harm.json"},{"slug":"knee-replacement-lifetime","question":"What are the lifetime odds of needing a knee replacement?","quick_answer":"About 1 in 10 US adults will have a knee replacement in their lifetime — somewhat higher than the roughly 1-in-15 many assume, and more common than hip replacement. It is now one of the most frequently performed major surgeries in the country.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Knee replacement tends to be perceived as a procedure for a fairly small fraction of the elderly population — those with severe arthritis who have exhausted conservative options. Most younger adults do not think of it as something likely in their own future. The scale of the procedure has grown substantially over the past two decades, and awareness of its population-level frequency has not kept pace. It is among the most commonly performed major surgical procedures in the United States today.\n","rough_estimate":"~1 in 15 lifetime feels about right to most people","kind":"intuition"},"native":{"display":"~10.38% of US adults age 80 have had a total knee replacement","numerator":1038,"denominator":10000,"unit":"lifetime prevalence by age 80","population":"US adults, Olmsted County population-based study (Maradit Kremers 2015)"},"normalized":{"lifetime_us_adult":0.1,"display":"~1 in 10 lifetime (US adult)","log_value":-1,"assumptions":"Maradit Kremers et al. (2015, J Bone Joint Surg Am) report total knee arthroplasty (TKA) prevalence of 10.38% among US adults reaching age 80, higher than the 5.26% reported for total hip arthroplasty. UK general-practice data (Culliford et al., referenced in the epidemiology literature) estimate the lifetime risk of TKA at approximately 10.8% for women and 8.1% for men; a separate US estimate from NHIS data puts lifetime risk at roughly 1 in 12 adults (8.3%). Given the Olmsted County 10.38% prevalence by age 80 and continuing incidence beyond that age, a lifetime estimate of approximately 10% (1 in 10) for a US adult is well-supported. Women have somewhat higher lifetime risk (~11–13%) than men (~7–9%), reflecting greater knee OA prevalence, particularly post-menopause. Uncertainty range 0.07–0.14 reflects sex, obesity trends, and increasing procedure rates over time.\n","uncertainty":{"low":0.07,"high":0.14},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/26333733/","title":"Prevalence of Total Hip and Knee Replacement in the United States","publisher":"Maradit Kremers H et al., J Bone Joint Surg Am","source_type":"peer_reviewed","statistic":"Prevalence of total knee replacement (TKA) in the US population was 1.52% overall in 2010; by age 80 it reached 10.38%. Women had higher prevalence than men at all ages.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] The 2010 prevalence of total hip and total knee replacement among the total U.S. population was 0.83% and 1.52%, respectively, with prevalence being higher among women than among men and increasing with age, reaching 5.26% for total hip replacement and 10.38% for total knee replacement at eighty years.\"\n","source_date":"2015-09-02","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20251202023949/https://pubmed.ncbi.nlm.nih.gov/26333733/","calculation_notes":"TKA prevalence at age 80 = 10.38% for the US population (Olmsted County population-based cohort). Because TKA incidence continues beyond age 80 and annual volumes have risen since 2010, the lifetime risk from age 18 to death is estimated at ~10%. TKA prevalence substantially exceeds THA (10.38% vs 5.26% at age 80), consistent with the higher incidence of knee vs hip OA.\n","independence_note":"Maradit Kremers 2015 used the Rochester Epidemiology Project (Olmsted County, MN) population-based medical records linkage, independent of AAOS registry administrative data. The two data systems measure the same procedures through entirely different pipelines.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5121085/","title":"Projecting Lifetime Risk of Symptomatic Knee Osteoarthritis and Total Knee Replacement in Individuals Sustaining a Complete Anterior Cruciate Ligament Tear in Early Adulthood","publisher":"Driban JB et al., Arthritis Care & Research (PMC)","source_type":"peer_reviewed","statistic":"Lifetime risk of total knee replacement was 22.3% for adults with prior ACL tear and meniscal injury combined, compared to 6.0% for those without knee injury. ACL injury without meniscal tear was associated with 8.0% (surgical) to 8.9% (non-operative) lifetime TKR risk; the higher 16–17% figures for that cohort are the lifetime risk of symptomatic knee osteoarthritis, not of total knee replacement.\n","excerpt":"\"[Paraphrase from full text] The estimated lifetime risk of symptomatic knee osteoarthritis was 34.2% for cohorts with ACL injury and meniscal tear, compared to 13.5% for no-injury cohorts. The risk of total knee replacement was 22.3% in the ACL and meniscal tear cohort versus 6.0% in the no-injury cohort — roughly 3.7 times greater.\"\n","source_date":"2016-11-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250204005740/https://pmc.ncbi.nlm.nih.gov/articles/PMC5121085/","calculation_notes":"Used to quantify the personal_factor_multiplier for prior ACL/meniscal injury. 22.3% TKR risk for ACL + meniscal tear vs 6.0% baseline (no injury) = approximately 3.7× relative risk on the source's own cohort estimates. At the population baseline of ~10% lifetime TKR, the absolute risk for those with ACL + meniscal injury approaches ~20–22%.\n","independence_note":"Driban et al. used a simulation model based on existing epidemiological cohort data (Framingham OA Study, Johnston County OA Project), independent of the Olmsted County prevalence data used by Maradit Kremers.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12192333/","title":"Highlights of the 2024 American Joint Replacement Registry Annual Report","publisher":"American Academy of Orthopaedic Surgeons (AAOS)","source_type":"reputable_reference","statistic":"AJRR captured over 3.7 million validated primary and revision THA and TKA procedures from 2012 to 2023; TKA is the majority procedure type at participating centers.\n","excerpt":"\"The 2024 AJRR Annual Report contains 3,715,320 validated primary and revision THA and TKA procedures performed during years 2012 to 2023, with primary THA comprising 32.4% of procedures captured. The AJRR remains the largest orthopaedic and joint arthroplasty registry in the world by annual procedure volume.\"\n","source_date":"2024-11-12","source_accessed":"2026-05-14","archive_url":"https://web.archive.org/web/20260525161818/https://pmc.ncbi.nlm.nih.gov/articles/PMC12192333/","calculation_notes":"If primary THA is 32.4% of procedures, then TKA + revisions account for 67.6%. Primary TKA is a substantially larger share of total procedures than THA, consistent with the higher population-level TKA prevalence (10.38% vs 5.26% at 80). Used to confirm high annual volumes and the TKA > THA utilization ratio.\n","independence_note":"AAOS AJRR draws on administrative and clinical data from participating surgical centers, methodologically independent of the population-based Rochester Epidemiology Project.\n"}],"comparison_anchors":[{"label":"Hip replacement (lifetime, US adult)","lifetime_us_adult":0.09},{"label":"Type 2 diabetes diagnosis (lifetime, US adult)","lifetime_us_adult":0.4},{"label":"Hip fracture (lifetime, US adult)","lifetime_us_adult":0.15}],"personal_factor_multipliers":[{"factor":"Obesity (BMI > 30)","multiplier":2.5,"notes":"Obesity has a stronger dose-response relationship with knee OA and TKA than with hip OA; meta-analyses show obese adults face approximately 2–3× the TKA risk of normal-weight adults. Every 5-unit increase in BMI above 25 meaningfully elevates risk.\n"},{"factor":"Female sex","multiplier":1.4,"notes":"Women have ~30–60% higher lifetime TKA risk than men (Culliford: 10.8% vs 8.1%). Knee OA is more prevalent in women, particularly after menopause, for reasons including hormonal, anatomical (Q-angle), and physical activity factors.\n"},{"factor":"Prior ACL tear or meniscal injury","multiplier":2.5,"notes":"ACL tear plus concomitant meniscal injury raises lifetime TKR risk to 22.3% vs 6.0% for no injury (~3.7× relative risk). Even isolated ACL injury without meniscal involvement carries an elevated lifetime TKR risk of roughly 8–9% (Driban et al. 2016); the 16–17% figure sometimes cited for that cohort is the risk of symptomatic knee osteoarthritis, not of replacement.\n"},{"factor":"Physically demanding occupation (prolonged kneeling, heavy lifting)","multiplier":1.5,"notes":"Occupations requiring repeated kneeling, squatting, or heavy lifting are associated with elevated knee OA incidence and subsequent TKA risk compared to sedentary occupations.\n"},{"factor":"Sedentary lifestyle (no regular exercise)","multiplier":1.3,"notes":"Paradoxically, sedentary individuals face elevated knee OA risk partly through loss of periarticular muscle support and higher body weight. Moderate low-impact exercise is protective for the knee rather than harmful.\n"}],"short_label":"Knee replacement","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"recoverable_injury","valence":"negative","caveats":"Knee replacement is the most commonly performed major joint replacement procedure in the United States, more frequent than hip replacement at every age stratum. The 10% lifetime estimate is based on Olmsted County prevalence data from 2010 (Maradit Kremers 2015) and is likely conservative for current cohorts given rising obesity rates and expanding surgical indications to younger patients. Kurtz et al. projected US TKA volumes could reach 1.26 million procedures per year by 2030. This entry covers primary (first-time) TKA only; unicompartmental knee arthroplasty and revision procedures are not included. Silent knee osteoarthritis is extremely common (symptomatic OA affects ~45% of adults over a lifetime) and precedes the subset that ultimately requires arthroplasty.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A stylized cross-section of a knee joint with prosthetic components, flat vector editorial illustration."},"canonical_url":"https://likelier.app/knee-replacement-lifetime","api_url":"https://likelier.app/api/fears/knee-replacement-lifetime.json"},{"slug":"pension-fund-collapse","question":"What are the odds of a pension fund failing or severely cutting benefits?","quick_answer":"Among Americans in defined-benefit pension plans, roughly 10% face a material benefit cut — far from the rare, institutional event most participants imagine. That figure is conditional on having such a pension, not a rate for all adults.\n","category":"other","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Pension fund failure is perceived as a remote, institutional risk — something that happens to steelworkers in the 1980s or municipal employees in Detroit. Most pension participants assume the Pension Benefit Guaranty Corporation backstops their benefits fully, unaware that PBGC guarantees have strict caps ($89,181 per year at age 65 for single-employer plans in 2025; far less for multiemployer plans). The 2021 American Rescue Plan's Special Financial Assistance program — which injected $69.5 billion into struggling multiemployer plans — temporarily reduced urgency, but the underlying structural problems (demographic shifts, optimistic return assumptions, chronic underfunding) remain. State and local pension underfunding is even less visible to participants, who rarely encounter their plan's funded ratio.\n","rough_estimate":"Perceived as rare, institutional","kind":"intuition"},"native":{"display":"~22% of state/local pension assets are unfunded ($1.3T gap); 130 multiemployer plans covering 1.3M people expected to exhaust funds within 20 years","numerator":22,"denominator":100,"unit":"share of state/local pension obligations that are unfunded","population":"US state and local government pension systems (Pew, FY2022)"},"normalized":{"lifetime_us_adult":0.1,"display":"~10% of DB pension participants face material benefit reduction","log_value":-1,"assumptions":"The subgroup is Americans who participate in defined-benefit pension plans (private and public combined). This universe is large and not counted uniformly in the sources here: the PBGC's multiemployer program alone covers about 11 million participants, and state and local government retirement systems add tens of millions more, so the total DB population is well into the tens of millions. The 10% central estimate is conditional on being a DB participant, not a rate for all US adults, and it is a rough figure rather than a precise ratio. The Pew Charitable Trusts reports that state and local pension systems collectively carry roughly $1.3 trillion in unfunded liabilities as of FY2022, with a national average funded ratio of 78% in 2024 (Equable). The PBGC's multiemployer program identified 130 plans covering 1.3 million people expected to exhaust their assets within 20 years. In the private sector, when PBGC takes over a terminated underfunded plan, benefits are capped at the guarantee maximum — meaning high-benefit workers (those earning above ~$89,000/year equivalent) lose a portion of their promised pension. For multiemployer plans, the guarantee is much lower ($12,870/year for 30 years of service), so benefit cuts upon PBGC takeover are substantial. The 10% estimate reflects: (a) 1.3M multiemployer participants facing near-certain benefit disruption (now mitigated by SFA, but SFA is time-limited); (b) an estimated few million state/local participants in severely underfunded plans (funded ratio below 60%); and (c) the ongoing trickle of private single-employer plan terminations — a facing-risk pool of several million set against a DB population of tens of millions. The uncertainty range spans 5% (optimistic: SFA holds, investment returns meet assumptions, states increase contributions) to 20% (pessimistic: SFA exhausted, markets underperform, political will for contributions erodes).\n","uncertainty":{"low":0.05,"high":0.2},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.pew.org/en/research-and-analysis/articles/2025/07/30/an-increase-in-pension-obligations-adds-to-states-unfunded-liabilities","title":"An Increase in Pension Obligations Adds to States' Unfunded Liabilities","publisher":"The Pew Charitable Trusts","source_type":"primary_study","statistic":"States collectively reported $1.27 trillion in unfunded pension benefits in FY2022, equal to nearly 66% of combined own-source revenue","excerpt":"\"States collectively reported $1.27 trillion in unfunded pension benefits in fiscal 2022, equal to nearly 66% of their combined own-source revenue — almost 17 percentage points higher than just a year earlier. Unfunded pension liabilities grew largely because of lower-than-expected investment returns.\"\n","source_date":"2025-07-30","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426205148/https://www.pew.org/en/research-and-analysis/articles/2025/07/30/an-increase-in-pension-obligations-adds-to-states-unfunded-liabilities","calculation_notes":"Pew's state pension analysis provides the macro-level unfunding data. The $1.27 trillion gap and the 66% of own-source revenue ratio illustrate the structural burden. The 22% unfunded share (100% minus the ~78% median funded ratio) is used as the native figure. Individual state variation is enormous: some states are above 90% funded; Illinois, New Jersey, and Kentucky have been below 50%.\n","independence_note":"Pew's analysis uses states' own Comprehensive Annual Financial Reports (CAFRs), independent from PBGC data (which covers only private-sector plans) and from Equable's independent funded-ratio calculations.\n"},{"url":"https://www.pbgc.gov/news/press/releases/pr24-040","title":"PBGC Releases FY 2024 Annual Report","publisher":"Pension Benefit Guaranty Corporation","source_type":"govt_report","statistic":"PBGC multiemployer program covers ~11 million participants in ~1,335 plans; $69.5 billion in Special Financial Assistance approved for plans covering ~1.2 million workers","excerpt":"\"The Multiemployer Program covers approximately 11 million participants in about 1,335 insured plans. As of November 1, 2024, PBGC has approved about $69.5 billion in Special Financial Assistance to financially troubled multiemployer plans that cover about 1.2 million workers and retirees.\"\n","source_date":"2024-11-15","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426205227/https://www.pbgc.gov/news/press/pr24-040","calculation_notes":"The PBGC annual report provides the official count of multiemployer plan participants and the scale of the SFA intervention. The 1.2 million workers receiving SFA support out of 11 million multiemployer participants (roughly 11%) gives a concrete measure of the share facing acute risk. Without SFA, the PBGC's multiemployer program was projected to become insolvent by 2025, which would have triggered benefit reductions to the guarantee maximum ($12,870/year for 30 years of service) for all participants in failed plans.\n","independence_note":"PBGC is the federal agency responsible for insuring private-sector defined-benefit pensions. Its data is administrative and independent from Pew's state/local pension analysis and from Equable's academic research.\n"},{"url":"https://equable.org/state-of-pensions-2025/","title":"State of Pensions 2025","publisher":"Equable Institute","source_type":"primary_study","statistic":"National average funded ratio of public pension plans: 78% in 2024, improving to 82.5% in 2025; national shortfall shrank from $1.54T (2024) to $1.27T (2025)","excerpt":"\"Funded status in 2025 has shown moderate improvement, increasing to 82.5%, up from 78.0% in 2024. The national shortfall in assets for state and local pension plans shrank from $1.54 trillion in 2024 to an estimated $1.27 trillion.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260323185305/https://equable.org/state-of-pensions-2025/","calculation_notes":"Equable provides the most granular plan-by-plan funded-ratio data. The 82.5% median funded ratio in 2025 masks wide dispersion: plans at or above 100% funded coexist with plans below 50%. The improvement from 78% to 82.5% was driven by strong 2024 investment returns, but Equable notes that \"pension debt paralysis\" — where contribution increases go primarily to servicing unfunded liabilities rather than building assets — remains a structural problem.\n","independence_note":"Equable is an independent research institute that calculates funded ratios using standardized assumptions, distinct from Pew's reliance on states' self-reported figures and from PBGC's private-sector-only scope.\n"},{"url":"https://www.morningstar.com/bonds/funds-most-affected-by-first-brands-bankruptcy-what-investors-can-learn-them","title":"The Funds Most Affected by First Brands' Bankruptcy and What Investors Can Learn From Them","publisher":"Morningstar","source_type":"reputable_reference","statistic":"First Brands loans were trading near par earlier in 2025, then lost nearly two-thirds of their value over 15 trading days in September 2025; one large lender exited at roughly $0.35-$0.40 on the dollar","excerpt":"\"Earlier this year, the company seemed to be on solid enough financial footing, and its loans were trading near par. But in 15 trading days between mid- and late September, one of the company's loans lost nearly two-thirds of its value before rebounding a bit.\" … \"This universe includes everything from daily liquidity bank-loan mutual funds to diversified, semiliquid private credit funds.\" … \"The managers opted to lock in a loss by exiting the position in late September at estimated dollar prices of roughly $0.35 to $0.40.\"\n","source_date":"2025-10-09","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20251010025519/https://www.morningstar.com/bonds/funds-most-affected-by-first-brands-bankruptcy-what-investors-can-learn-them","calculation_notes":"This source documents a private-credit transmission channel that the headline 10% figure does not capture: illiquid corporate loans (First Brands, alongside the September 2025 Tricolor subprime-auto collapse) carried near par in fund portfolios then repricing to roughly a third of face value within weeks. Because these loans sit in semiliquid credit funds and bank-loan vehicles — some of which feed mainstream defined-contribution and target-date allocations — the loss path is not reflected in the defined-benefit funded ratios that drive the 10% estimate. It is retained as a qualitative caveat only: the asset class is young, marks are contested, and the systemic exposure cannot yet be expressed as an individual lifetime probability, so the headline number is left unchanged.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Student loan default (US borrowers)","lifetime_us_adult":0.26},{"label":"Retirement savings shortfall (US)","lifetime_us_adult":0.39}],"regional_breakdown":[{"region":"Illinois, New Jersey, Kentucky (severely underfunded states)","probability":0.3,"notes":"States with funded ratios below 50% face the highest risk of benefit reductions or contribution crises"},{"region":"Well-funded states (WI, SD, NY)","probability":0.03,"notes":"States with funded ratios above 90% face minimal near-term risk to participant benefits"},{"region":"Multiemployer plans receiving SFA","probability":0.15,"notes":"SFA provides temporary solvency but is designed to last approximately 30 years; long-term sustainability is uncertain"}],"personal_factor_multipliers":[{"factor":"multiemployer plan participant (no SFA)","multiplier":3,"notes":"Multiemployer plans without SFA and with declining active-to-retiree ratios face the highest insolvency risk"},{"factor":"state/local employee in well-funded system","multiplier":0.2,"notes":"Participants in plans above 90% funded have minimal near-term benefit risk"},{"factor":"high-benefit private-sector retiree","multiplier":2,"notes":"PBGC guarantee caps mean high earners lose the most in a plan termination"},{"factor":"active worker with decades until retirement","multiplier":0.5,"notes":"Younger workers have time to adjust savings and career plans; they also benefit from any plan reforms"}],"short_label":"Pension fund collapse","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 10% figure is conditional on being a defined-benefit pension participant — a group that numbers in the tens of millions across public and private plans — so applied to all US adults the unconditional rate is a small fraction of that. This entry is about institutional pension failure, not personal savings inadequacy (covered separately in retirement-savings-shortfall). The 10% estimate is approximate because pension risk materializes over decades and depends on investment returns, employer contributions, demographic shifts, and political decisions that are inherently unpredictable. The American Rescue Plan's $69.5 billion SFA intervention dramatically reduced near-term multiemployer risk, but SFA is a one-time appropriation designed to last roughly 30 years — it does not fix the structural underfunding. State and local pensions cannot legally default in most jurisdictions, but they can (and do) reduce benefits for new hires, increase employee contributions, and in some cases (e.g., Detroit, Puerto Rico) reduce benefits for current retirees through bankruptcy proceedings. The funded-ratio metric itself is sensitive to the discount rate assumption: using the plans' own assumed return (~7%) yields funded ratios around 78-82%, while using a risk-free rate (~4%) would show funded ratios closer to 40-50%, implying a much larger unfunded liability. The entry uses the plans' own assumptions for consistency with how funded status is reported, but the risk-free perspective suggests the problem is substantially larger than headline figures indicate. A newer channel — not captured by the defined-benefit funded ratios above — is private credit. The September 2025 collapses of First Brands and Tricolor showed illiquid corporate loans carried at or near par in fund portfolios repricing to roughly a third of face value within weeks (Morningstar, 2025); because such loans increasingly sit in semiliquid credit funds and bank-loan vehicles that feed some defined-contribution and target-date allocations, losses can reach retirement savers outside the traditional pension-failure path. The FSB's Report on Vulnerabilities in Private Credit (6 May 2026) flags opaque, manager-discretion valuations and deepening interconnections between private credit funds, banks, and insurers, warning that the sector's complexity, leverage, and interconnectedness could amplify stress in adverse scenarios. This exposure is left as a caveat, not folded into the headline number: the asset class is young, marks are contested, and the systemic risk cannot yet be expressed as an individual lifetime probability.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A cracked piggy bank with a few coins spilling out, muted grey and copper tones, flat vector illustration."},"canonical_url":"https://likelier.app/pension-fund-collapse","api_url":"https://likelier.app/api/fears/pension-fund-collapse.json"},{"slug":"regular-otc-painkillers","question":"What are the odds of being harmed by taking over-the-counter painkillers regularly?","quick_answer":"For someone taking OTC painkillers several times a week for decades, the lifetime risk of a serious harm such as a GI bleed is about 1 in 10, from roughly a 1 in 100 annual serious-GI-event rate. Most chronic users guess well below 1 in 100.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"OTC painkillers sit in an unusual perceptual slot: readily available on grocery shelves, universally familiar, and mentally filed alongside vitamins and antacids rather than alongside drugs. The usual public intuition is that the ceiling of harm from ibuprofen, naproxen, aspirin, or acetaminophen taken \"a few times a week\" is a mild stomach ache or a headache that doesn't go away. That intuition undercounts a very real chronic-use hazard: serious GI bleeding, acute kidney injury, and acetaminophen-driven acute liver failure each have measurable annual event rates in habitual users. No rigorous nationally-representative survey measures the perceived-harm distribution for chronic OTC analgesic use, so this is flagged as intuition, not a polled number.\n","rough_estimate":"Most chronic users guess the lifetime serious-adverse-event risk well below 1 in 100","kind":"intuition"},"native":{"display":"~1 in 100 per year serious GI event (chronic NSAID user)","numerator":1,"denominator":100,"unit":"per year","population":"chronic NSAID user (≥3×/week, >1 year), US adult, mid-life onward"},"normalized":{"lifetime_us_adult":0.1,"display":"~1 in 10 lifetime (chronic OTC analgesic user, 20+ years)","log_value":-1,"assumptions":"Reference subgroup: a US adult who uses an OTC NSAID (ibuprofen, naproxen, or low-dose aspirin taken for pain rather than cardioprotection) or acetaminophen three or more times per week across roughly 20 years of adult life, without gastroprotection from a proton-pump inhibitor. The ~10% headline is a rounded mid-point for the cumulative probability of a clinically significant adverse event — upper GI bleed or ulcer requiring medical attention, acute kidney injury, an NSAID-attributable cardiovascular event, or acetaminophen-associated hepatotoxicity — across that chronic-use window. Bracketed 5-20% to reflect (a) the Lanas 2005 nationwide cohort incidence of ~122 major GI events per 100,000 persons per year averaged across the full population, which rises to roughly 1 per 100-1000 per year among habitual NSAID users and to 1 per ~110 per year in chronic users over 75, and (b) the Marcum/Hanlon 2010 figure of ~41,000 hospitalizations and ~3,300 deaths per year among older adult NSAID users in the US. Compounding an annual event hazard of ~0.5% per year (middle-aged chronic user weighted-average across GI, renal, CV, and hepatic endpoints) over 20 years: 1 - (1 - 0.005)^20 ≈ 0.095, rounded to ~10%. The hazard rises steeply with age: the same compounding over 20 years of chronic use starting at age 65 (annual hazard closer to 1%) produces ~18%. The hazard falls steeply for users under 50 without risk factors. Scope is subgroup_lifetime because this is a per-chronic-user cumulative probability, not a general-population lifetime risk, and not directly comparable to the population-scope figures on other Likelier pages.\n","uncertainty":{"low":0.05,"high":0.2},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/16086703/","title":"A nationwide study of mortality associated with hospital admission due to severe gastrointestinal events and those associated with nonsteroidal antiinflammatory drug use","publisher":"American Journal of Gastroenterology (Lanas et al.)","source_type":"peer_reviewed","statistic":"121.9 major GI events per 100,000 persons per year; 21.0-24.8 NSAID/aspirin-attributable deaths per million population per year; 15.3 deaths per 100,000 NSAID/aspirin users per year; up to one-third of those deaths attributable to low-dose aspirin","excerpt":"\"Mortality rates associated with either major upper or lower GI events are similar but upper GI events were more frequent. [...] NSAID/aspirin use was associated with 21.0-24.8 deaths per million people per year, of which up to one-third can be attributed to low-dose aspirin use.\"\n","source_date":"2005-08-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260407091053/https://pubmed.ncbi.nlm.nih.gov/16086703/","calculation_notes":"Lanas 2005 is the canonical nationwide incidence anchor for NSAID-attributable serious GI events. 121.9 events per 100,000 person- years averaged across the full population corresponds to ~0.12% per year; concentrating that hazard on the subset of the population that actually uses NSAIDs chronically pushes the per-user rate to roughly 0.5-1% per year depending on age and comorbidity. 15.3 deaths per 100,000 users per year is the per-user case-fatality anchor. The one-third attributable to low-dose aspirin matters because the cardioprotective-aspirin population is partially but not entirely disjoint from the OTC-painkiller population in the public conversation. Used as the primary relative-risk and event-incidence basis for the ~0.5% annual per-chronic-user hazard used in the normalized compounding.\n","independence_note":"Lanas 2005 is a Spanish nationwide hospital-admissions cohort. Generalizing to US adults assumes comparable NSAID prescribing and GI epidemiology; the US per-capita figures (Marcum/Hanlon 2010 below) agree to within a factor of two. Treat as an independent European anchor cross-checked against the US estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3158445/","title":"Recognizing the Risks of Chronic Nonsteroidal Anti-Inflammatory Drug Use in Older Adults","publisher":"Annals of Long-Term Care (Marcum & Hanlon)","source_type":"peer_reviewed","statistic":"NSAID use causes an estimated 41,000 hospitalizations and 3,300 deaths each year among older adults in the US; peptic ulcer hospitalization rises from 1 per 1,000/year under age 50 to 2-6 per 1,000/year over 65; acute renal failure risk doubles within 30 days of initial NSAID use; ~40% of adults 65+ fill an NSAID prescription annually","excerpt":"\"NSAID use causes an estimated 41,000 hospitalizations and 3300 deaths each year among older adults. [...] the rate of hospitalizations for peptic ulcer disease (PUD) increases with age, from 1 per 1000 per year in those under 50 to 2-6 per 1000 per year in older adults (>65 yr). [...] the risk of ARF was increased nearly twofold for all NSAIDs (nonselective and COX-2 selective) within 30 days of initial use.\"\n","source_date":"2010-09-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260321082502/https://pmc.ncbi.nlm.nih.gov/articles/PMC3158445/","calculation_notes":"The 41,000 hospitalizations / 3,300 deaths per year figure for older US adults on NSAIDs is the domestic anchor. Across a denominator of ~25 million older adult NSAID users (~40% of ~60 million US 65+ adults), that is roughly 164 hospitalizations per 100,000 users per year and 13 deaths per 100,000 users per year — within a factor of two of the Lanas Spanish figures, confirming the cross-population consistency. Peptic ulcer hospitalization rate of 2-6 per 1,000/year in older adults is the basis for the elderly row in the regional_breakdown and the ~1% annual GI-event hazard in the headline compounding for the over-65 subgroup.\n","independence_note":"Marcum/Hanlon synthesize Lanas and the ARAMIS database; partially dependent on Lanas 2005 above for the European incidence figures but draws on distinct US claims data for the US-specific hospitalization and mortality totals. Treat as a related US-focused line of evidence.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19125949/","title":"Acetaminophen-related acute liver failure in the United States","publisher":"Hepatology Research (Lee)","source_type":"peer_reviewed","statistic":"Acetaminophen accounts for approximately 50% of acute liver failure cases in the US; 30% mortality once acute liver failure develops; nearly half of US acetaminophen ALF cases are unintentional overdoses of therapeutic products","excerpt":"\"Acetaminophen overdoses are the number one cause of acute liver failure (ALF) in the United States; they account for 50% of all cases of ALF and carry a 30% mortality. [...] nearly half of U.S. cases are unintentional, resulting from overuse of acetaminophen-containing pain relief products rather than deliberate self-harm.\"\n","source_date":"2008-11-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260504060030/https://pubmed.ncbi.nlm.nih.gov/19125949/","calculation_notes":"Lee's US Acute Liver Failure Study Group registry is the definitive source that acetaminophen is the single leading cause of ALF in the US. The unintentional-overdose fraction (~50% of acetaminophen ALF cases) is load-bearing: these are chronic users who stayed at or slightly above the 4 g/day label limit, often while also drinking alcohol or combining an OTC acetaminophen product with an acetaminophen-containing prescription opioid. The ~500 deaths per year US figure commonly cited in the FDA literature is the product of ~1,600 acetaminophen-associated ALF cases per year and the ~30% fatality rate. Used as the anchor for the acetaminophen component of the compounded hazard and for the alcohol-acetaminophen interaction multiplier.\n","independence_note":"The US Acute Liver Failure Study Group registry is the primary data source; Blieden 2014 (below) and subsequent FDA regulatory documents draw directly on Lee's registry data. Treat Lee 2008, Blieden 2014, and the FDA figures as a single largely-dependent line of evidence.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24678654/","title":"A perspective on the epidemiology of acetaminophen exposure and toxicity in the United States","publisher":"Expert Review of Clinical Pharmacology (Blieden et al.)","source_type":"peer_reviewed","statistic":"~30,000 patients hospitalized yearly for acetaminophen toxicity in the US; ~6% of adults take doses exceeding 4 g/day at least occasionally; up to 50% of acetaminophen overdoses are unintentional; 17% of adults with unintentional overdose experience liver damage","excerpt":"\"Approximately 6% of adults are prescribed acetaminophen doses of more than 4 g/day [...] 30,000 patients are hospitalized yearly for acetaminophen toxicity [...] up to 50% of overdoses are unintentional in nature [...] 17% of adults with unintentional overdose experience liver damage.\"\n","source_date":"2014-05-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260421193256/https://pubmed.ncbi.nlm.nih.gov/24678654/","calculation_notes":"Blieden 2014 is the behavioural-epidemiology source: ~6% of adult acetaminophen users exceed the 4 g/day label ceiling on at least some days, and ~30,000 are hospitalized per year with acetaminophen toxicity. 30,000 / ~50 million US adults using acetaminophen at least weekly ≈ 0.06% per year hospitalization rate — small at the population level, but concentrated in the chronic-user-plus-alcohol or chronic-user-plus-combination-product subgroups. This is the source for the regional_breakdown row on chronic acetaminophen with alcohol and the personal_factor_multiplier for heavy alcohol use.\n","independence_note":"Blieden draws on the same Acute Liver Failure Study Group registry as Lee 2008 above, cross-linked with Poison Control Center call data and the Kaufman Slone Survey of OTC use prevalence. Treat as partially dependent on Lee 2008; independent on the prevalence-of-use data.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21224324/","title":"Cardiovascular safety of non-steroidal anti-inflammatory drugs: network meta-analysis","publisher":"BMJ (Trelle et al.)","source_type":"peer_reviewed","statistic":"Network meta-analysis of 31 RCTs and 116,429 patients: ibuprofen associated with a stroke RR of 3.36 (95% credibility interval 1.00-11.6); diclofenac with a cardiovascular-death RR of 3.98 (1.48-12.7); naproxen appeared least harmful","excerpt":"\"Although uncertainty remains, little evidence exists to suggest that any of the investigated drugs are safe in cardiovascular terms. Naproxen seemed least harmful. Cardiovascular risk needs to be taken into account when prescribing any non-steroidal anti-inflammatory drug.\"\n","source_date":"2011-01-11","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260318045802/https://pubmed.ncbi.nlm.nih.gov/21224324/","calculation_notes":"Trelle 2011 is the cardiovascular half of the chronic-NSAID hazard picture. The RCTs pooled here were mostly prescription-dose trials, so the absolute CV event rate attributable to OTC-dose ibuprofen is lower than the abstract's relative risks imply. For a middle-aged chronic user the CV component of the ~0.5% annual hazard is on the order of 0.1-0.2% per year, rising substantially in users with pre-existing cardiovascular disease. The differential hazard across specific NSAIDs (naproxen < ibuprofen < diclofenac) is the basis for the caveat distinguishing between naproxen and the other OTC NSAIDs.\n","independence_note":"Trelle 2011 is a pooled RCT meta-analysis, methodologically distinct from the Lanas and Marcum/Hanlon observational cohorts above. Independent line of evidence for the cardiovascular component.\n"}],"comparison_anchors":[{"label":"Death from alcohol-related disease (lifetime, heavy drinker)","lifetime_us_adult":0.15},{"label":"Colorectal cancer (lifetime, daily processed-meat consumer)","lifetime_us_adult":0.048},{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Occasional OTC user (<1×/week, any age, no risk factors)","probability":0.002,"notes":"Near-baseline. Acute AEs at label dose and frequency are rare; most epidemiological signal comes from chronic users."},{"region":"Chronic NSAID user (3+×/week, middle-aged, no PPI)","probability":0.1,"notes":"Headline subgroup. ~0.5% annual serious-AE hazard (GI + renal + CV composite) compounded over ~20 years of chronic use."},{"region":"Chronic NSAID user, age 65+ (no PPI)","probability":0.18,"notes":"Peptic ulcer hospitalization rate 2-6 per 1,000/year per Marcum/Hanlon 2010; compounded annual hazard closer to 1%."},{"region":"Chronic acetaminophen at ≤4 g/day label dose, no alcohol","probability":0.01,"notes":"Hepatotoxicity at or below the label ceiling is rare in the absence of alcohol or pre-existing liver disease; ~30,000 acetaminophen-tox hospitalizations/year spread across ~50M adult users is ~0.06%/year."},{"region":"Chronic acetaminophen user + ≥2 drinks/day","probability":0.05,"notes":"Alcohol-induced CYP2E1 upregulation sharply raises NAPQI production at any given acetaminophen dose; unintentional hepatotoxicity concentrated in this group per Lee 2008."},{"region":"Chronic NSAID + daily low-dose aspirin (combined GI risk)","probability":0.15,"notes":"Upper-GI-bleed hazards multiply rather than add; up to one-third of NSAID/aspirin GI deaths attributable to low-dose aspirin per Lanas 2005."}],"personal_factor_multipliers":[{"factor":"age 65+","multiplier":3,"notes":"Peptic ulcer hospitalization rises 3-5x over age 65 per Marcum/Hanlon 2010; CV and renal hazards rise similarly."},{"factor":"prior peptic ulcer or GI bleed","multiplier":4,"notes":"Per Lanas and StatPearls: patients with previous ulcer complications on NSAIDs have incidence rates of 20-30 per 1,000 person-years for upper GI bleeding."},{"factor":"concomitant anticoagulant or antiplatelet","multiplier":2.5,"notes":"Warfarin, DOAC, clopidogrel, or SSRI co-administration each independently raise GI bleed hazard; combined with NSAID the effect is roughly multiplicative."},{"factor":"heavy alcohol use (>2 drinks/day) + acetaminophen","multiplier":5,"notes":"Alcohol-induced CYP2E1 activity substantially raises NAPQI generation at any given acetaminophen dose; most unintentional acetaminophen ALF cases in Lee's registry occur in this group."},{"factor":"pre-existing chronic kidney disease","multiplier":3,"notes":"NSAIDs reduce renal prostaglandin synthesis and precipitate AKI in people with reduced baseline GFR; ARF risk doubles within 30 days of NSAID initiation in the general population."},{"factor":"pre-existing cardiovascular disease","multiplier":2,"notes":"Per Trelle 2011 and subsequent FDA labeling: prior MI or heart failure substantially raises NSAID-attributable CV event rate, particularly for ibuprofen and diclofenac."},{"factor":"H. pylori infection untreated","multiplier":2,"notes":"H. pylori and NSAID exposure are independent risk factors for peptic ulcer disease; combined effect is roughly multiplicative."},{"factor":"naproxen vs ibuprofen/diclofenac","multiplier":0.6,"notes":"Naproxen has the lowest CV risk signal in the Trelle network meta-analysis; GI profile is similar across non-selective NSAIDs."}],"short_label":"Chronic painkillers","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is specifically the cumulative serious-adverse-event probability for a chronic OTC analgesic user — operationally a US adult taking an NSAID or acetaminophen three or more times per week across roughly 20 years of adult life without gastroprotection. It is not the risk from short-course or occasional use, where the per-event hazard is low enough that the compounded probability rounds to near-baseline. Acute use for a headache, a hangover, a sprained ankle, a menstrual cramp, or a post-operative week is not what this number is measuring. The headline also mixes four distinct endpoints — upper GI bleed or ulcer, acute kidney injury, an NSAID-attributable cardiovascular event, and acetaminophen-associated hepatotoxicity — into a single composite \"clinically significant harm\" outcome; readers interested in a specific endpoint should read the regional_breakdown and personal_factor_multipliers rows. The per-agent risk profile differs: naproxen has the lowest cardiovascular signal in the Trelle meta-analysis; ibuprofen and diclofenac sit higher; ibuprofen and naproxen share a broadly similar GI profile at OTC doses; low-dose aspirin carries its own substantial contribution to the combined GI mortality total (~1/3 per Lanas 2005). Acetaminophen at or below the 4 g/day label ceiling is largely safe for the liver in the absence of alcohol, chronic liver disease, or concurrent use of a second acetaminophen-containing product — the unintentional overdose pattern in Lee's registry is dominated by precisely those three modifiers. Individual outcomes depend on age, sex, dose, frequency, duration, specific agent, coadministration (PPIs, anticoagulants, antiplatelets, SSRIs), pre-existing GI/renal/hepatic/cardiovascular disease, H. pylori status, alcohol intake, and a long list of pharmacogenomic modifiers. The 5-20% uncertainty band is wide on purpose: composite adverse-event probabilities for chronic OTC use are less tightly constrained than the aggregate annual mortality/hospitalization totals.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-8-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single white pill resting on a pale muted surface, flat vector illustration."},"canonical_url":"https://likelier.app/regular-otc-painkillers","api_url":"https://likelier.app/api/fears/regular-otc-painkillers.json"},{"slug":"unleashed-dog-bite-liability","question":"What are the odds of your off-leash dog biting a person or another dog?","quick_answer":"Medically attended dog bites number about 799,700 a year across roughly 90 million US pet dogs — about a 1 in 10 chance over a dog's 12-year life. Most owners guess effectively zero, assuming 'my dog would never.'\n","category":"animal","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Most dog owners are serenely confident their own dog would never bite anyone. \"He's friendly\" is the universal off-leash disclaimer, delivered seconds before 4.5 million annual data points suggest otherwise. No rigorous survey isolates the perceived risk of one's own dog biting a stranger while off-leash, but the informal consensus among owners skews toward \"essentially zero.\" The insurance industry's $1.86 billion annual payout suggests otherwise.\n","rough_estimate":"most owners guess effectively zero — 'my dog would never'","kind":"intuition"},"native":{"display":"~799,700 medically attended dog bites per year across ~90 million US pet dogs","numerator":799700,"denominator":90000000,"unit":"per dog per year","population":"US pet dogs (all, not limited to off-leash)"},"normalized":{"lifetime_us_adult":0.102,"display":"~1 in 10 over a dog's 12-year lifespan (medically attended bite)","log_value":-0.99,"assumptions":"Uses CDC's estimate of ~4.7 million dog bites per year in the United States, of which ~799,700 require medical care (MMWR 2003, citing 1994 data — the most recent year for which published national data exist). Divided by ~90 million pet dogs in US households (APPA 2024-2025 National Pet Owners Survey reports ~68 million dog-owning households averaging ~1.3 dogs; CDC's own 2001 estimate was ~68 million canines kept as pets). Annual per-dog probability of inflicting a medically attended bite ≈ 799,700 / 90,000,000 ≈ 0.0089. Compounded over a typical 12-year canine lifespan: 1 − (1 − 0.0089)^12 ≈ 0.102, or roughly 1 in 10. This is an all-dog average — it includes on-leash, off-leash, in-home, and yard bites. Off-leash dogs in uncontrolled environments carry higher risk, but the base rate itself is the headline surprise. The broader ~4.7 million figure (including minor bites not requiring medical care) yields a per-dog annual rate of ~5% and a 12-year rate of ~47%, but the medically attended subset is the more defensible denominator for liability framing.\n","uncertainty":{"low":0.07,"high":0.17},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5226a1.htm","title":"Nonfatal Dog Bite–Related Injuries Treated in Hospital Emergency Departments — United States, 2001","publisher":"US Centers for Disease Control and Prevention (CDC), Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"In 1994, ~4.7 million dog bites occurred in the US and ~799,700 persons required medical care; in 2001, 368,245 persons were treated in U.S. hospital EDs for nonfatal dog bite-related injuries","excerpt":"\"In 1994, the most recent year for which published data are available, an estimated 4.7 million dog bites occurred in the United States, and approximately 799,700 persons required medical care... in 2001, an estimated 368,245 persons were treated in U.S. hospital EDs for nonfatal dog bite-related injuries.\"\n","source_date":"2003-07-04","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260318060927/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm5226a1.htm","calculation_notes":"CDC's estimate of ~4.7 million dog bites (1994) derives from a nationally representative survey scaled to population, corroborated by ED surveillance data. The stated 799,700 who \"required medical care\" is used as numerator for the native rate because it captures bites serious enough to generate potential liability. 799,700 / 90,000,000 pet dogs ≈ 0.0089 per dog per year. Compounded over 12 years: 1 − (1 − 0.0089)^12 ≈ 0.102.\n"},{"url":"https://www.iii.org/article/spotlight-on-dog-bite-liability","title":"Spotlight on: Dog bite liability","publisher":"Insurance Information Institute (III) / State Farm","source_type":"reputable_reference","statistic":"28,450 dog-related injury claims in 2025, averaging $65,450 per claim; total insurer payouts $1.86 billion in 2025; $1.57 billion in 2024 from 22,658 claims at $69,272 average","excerpt":"\"Liability claims related to dog bites and other dog-related injuries cost homeowners insurers $1,862 million in 2025. The number of dog bite claims nationwide increased in 2025 to 28,450 from 22,658 in 2024—a 25.6 percent increase. The average cost per claim decreased 5.5 percent in 2025 to $65,450 from $69,272 in 2024. The average cost per claim nationally has risen 97.0 percent from 2016 to 2025.\"\n","source_date":"2025-04-25","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260320053659/https://www.iii.org/article/spotlight-on-dog-bite-liability","calculation_notes":"III/State Farm data provides the financial exposure layer. Per dog-owning household: 28,450 claims / 68,000,000 dog-owning households ≈ 0.042% annual probability of filing a dog bite liability claim, or roughly 1 in 2,400 per year. Over a 40-year dog-owning career: 1 − (1 − 0.000418)^40 ≈ 1.7%. The average claim of $65,450 is the median owner's surprise — most assume their homeowner's policy handles it, but breed restrictions and coverage caps can leave significant uninsured exposure.\n","independence_note":"III data is derived from insurer claims databases (State Farm, industry aggregates), methodologically independent of CDC epidemiological surveillance. CDC counts bites; III counts payouts. The two datasets measure different stages of the same pipeline.\n"},{"url":"https://avmajournals.avma.org/view/journals/javma/243/12/javma.243.12.1726.xml","title":"Co-occurrence of potentially preventable factors in 256 dog bite-related fatalities in the United States (2000-2009)","publisher":"Journal of the American Veterinary Medical Association (JAVMA) / Patronek, Sacks, Delise, Cleary, Marder","source_type":"peer_reviewed","statistic":"In 256 fatal dog bite cases (2000-2009), 76.2% involved dogs kept isolated from regular positive human interactions; 84.4% involved unneutered dogs; 87.1% had no able-bodied person present to intervene","excerpt":"\"Major co-occurrent factors included absence of an able-bodied person to intervene (n = 223 [87.1%]), incidental or no familiar relationship of victims with dogs (218 [85.2%]), owner failure to neuter dogs (216 [84.4%]), compromised ability of victims to interact appropriately with dogs (198 [77.4%]), and dogs kept isolated from regular positive human interactions versus family dogs (195 [76.2%]). Four or more of these factors co-occurred in 206 (80.5%) deaths.\"\n","source_date":"2013-12-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260209041933/https://avmajournals.avma.org/view/journals/javma/243/12/javma.243.12.1726.xml","calculation_notes":"Patronek et al. identified the risk-factor profile for the most serious outcomes. While their study covers fatal bites specifically (a tiny fraction of 4.5M annual bites), the co-occurring factors — intact males, isolated dogs, absent supervision — overlap heavily with the profile for non-fatal serious bites. The finding that 84.4% of fatal-bite dogs were unneutered supports the 2-4x multiplier assigned to intact males in the personal factors below.\n","independence_note":"Patronek's team used law-enforcement and animal-control primary records, not insurance claims or CDC ED surveillance, making this methodologically independent of both the CDC and III sources.\n"}],"comparison_anchors":[{"label":"Death by dog bite (lifetime, US adult)","lifetime_us_adult":0.00000704},{"label":"Bee sting fatality (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Home burglary victimization (lifetime)","lifetime_us_adult":0.26},{"label":"Car crash injury (lifetime, US)","lifetime_us_adult":0.2}],"regional_breakdown":[{"region":"Any bite (including minor nips, no medical attention)","probability":0.47,"notes":"~4.7M total bites (CDC MMWR 2003, 1994 data) / ~90M dogs ≈ 5% per year; over a 12-year dog lifespan: 1 − (1−0.052)^12 ≈ 0.47"},{"region":"Bite requiring medical attention","probability":0.102,"notes":"~799.7K bites requiring medical care (CDC MMWR 2003) / 90M dogs ≈ 0.9% per year; over 12 years ≈ 10.2%"},{"region":"Bite resulting in homeowner insurance claim","probability":0.005,"notes":"~28,450 claims / 68M dog-owning households ≈ 0.042% per year; over 12 years ≈ 0.5%"},{"region":"Dog-on-dog incident causing veterinary treatment","probability":0.04,"notes":"Limited peer-reviewed data; estimated from veterinary emergency surveys suggesting ~3-4% of dogs per year are involved in interdog aggression incidents requiring treatment"}],"personal_factor_multipliers":[{"factor":"intact (unneutered) male dog","multiplier":3,"notes":"Patronek et al. found 84.4% of fatal-bite dogs were intact males; AVMA literature review notes intact males are represented in 70-76% of reported bite incidents despite being a minority of the dog population.\n"},{"factor":"neutered, well-socialized family dog","multiplier":0.3,"notes":"Neutering, early socialization, and integration into family life are each independently associated with lower bite risk in the AVMA and Patronek datasets; combined effect estimated at ~0.3x.\n"},{"factor":"off-leash in uncontrolled public space","multiplier":3,"notes":"No single study cleanly isolates the off-leash multiplier, but animal control data consistently shows unrestrained dogs disproportionately represented in serious bite reports. AVMA's community approach report identifies control of free-roaming animals as a primary prevention measure.\n"},{"factor":"off-leash in fenced, designated dog park","multiplier":1.5,"notes":"Dog parks confine the interaction space but remove leash control; roughly 6% of reported bites occur in parks per available data.\n"},{"factor":"large/powerful breed (>30 kg)","multiplier":2,"notes":"Larger dogs inflict more serious injuries per bite, increasing the probability of medical attention and insurance claims even if bite frequency is similar.\n"},{"factor":"small breed (<10 kg)","multiplier":0.5,"notes":"Small dogs may bite as frequently but injuries rarely require medical attention, reducing the medically attended bite rate.\n"},{"factor":"dog with prior bite history","multiplier":5,"notes":"Prior aggression is the single strongest predictor of future biting behavior in veterinary behavioral literature.\n"},{"factor":"multiple dogs in household","multiplier":1.8,"notes":"Each additional dog adds an independent probability; two-dog households roughly double the annual household bite risk.\n"}],"short_label":"Off-leash dog bite","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 4.5-to-4.7 million annual bite figure (III cites ~4.5 million; CDC's MMWR cites ~4.7 million for 1994) is the most widely used estimate, corroborated by subsequent ED surveillance. The true number may be higher, as minor bites are systematically underreported. The per-dog calculation treats all 90 million US pet dogs as equally likely to bite, which they are not — risk is heavily concentrated in intact males, unsocialized dogs, and dogs with prior bite history. The \"off-leash\" framing in this entry's question is important because off-leash status removes the owner's last physical control at the moment of encounter, but most bites (roughly two-thirds) actually occur on the owner's property, often involving a dog known to the victim. The insurance claim data captures only the liability-claim pipeline and misses bites settled informally, covered by the victim's own health insurance, or never reported. Thirty-one states impose strict liability on dog owners regardless of the dog's prior behavior; sixteen follow a \"one-bite rule\" requiring proof the owner knew of the dog's dangerous propensity.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty dog leash lying on a sidewalk, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/unleashed-dog-bite-liability","api_url":"https://likelier.app/api/fears/unleashed-dog-bite-liability.json"},{"slug":"home-burglary","question":"What are the odds of your home being burglarized?","quick_answer":"Over a 40-year homeownership span, the odds of a police-reported burglary are about 11%, and likely 15-25% once unreported break-ins are counted. That is well below the 30-50% many homeowners assume, and far below the outdated 1987 figure of 72% still widely cited from 1970s-era crime rates.\n","category":"crime","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Burglary sits near the top of property-crime anxiety. Home security systems are a multi-billion-dollar industry built on the image of a stranger forcing a lock at 3 a.m. while the family sleeps. News coverage and insurance advertisements reinforce the mental model of a perpetual, high-probability threat. Many homeowners assume that going a decade without a break-in is lucky, not typical. The 1987 BJS estimate that 72% of Americans would experience a burglary in their lifetime is still widely cited, though it was calculated from 1970s-era crime rates that were three to five times higher than today's.\n","rough_estimate":"~30-50% lifetime chance","kind":"intuition"},"native":{"display":"~2.9 residential burglaries per 1,000 households per year (police-reported)","numerator":29,"denominator":10000,"unit":"per household per year (police-reported residential burglaries)","population":"US households, 2024 FBI UCR data"},"normalized":{"lifetime_us_adult":0.11,"display":"~11% lifetime probability of experiencing at least one burglary over a 40-year homeownership span (police-reported)","log_value":-0.96,"assumptions":"The FBI Crime Data Explorer reported 405,776 residential burglaries in 2024 across ~140 million US households, yielding an annual police-reported rate of ~2.9 per 1,000 households. The NCVS (which captures unreported crime) historically estimates rates 2-4x higher; BJS reported ~11.7 per 1,000 households in 2019 for burglary/trespass combined. We use the FBI police-reported figure as the conservative base because it better reflects completed burglaries with actual loss, while NCVS includes attempted entries and trespass. Over a 40-year homeownership span (age 25-65), assuming independent annual trials at the 2024 rate: 1 - (1 - 0.0029)^40 ≈ 0.109, or ~11%. Using the higher NCVS-inclusive rate of ~0.007 (adjusted downward from 11.7/1,000 to account for post-2019 declines), the lifetime figure would be ~24%. The 11% estimate is conservative; the true figure including unreported burglaries is likely 15-25%. The 1987 BJS lifetime estimate of 72% was based on 1975-1984 NCVS rates that were 3-5x higher than current rates and is no longer applicable.\n","uncertainty":{"low":0.06,"high":0.25},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fbi.gov/news/press-releases/fbi-releases-2024-reported-crimes-in-the-nation-statistics","title":"FBI Releases 2024 Reported Crimes in the Nation Statistics","publisher":"Federal Bureau of Investigation","source_type":"govt_report","statistic":"779,542 total burglaries in 2024; rate of 229.2 per 100,000 residents; residential burglaries accounted for ~52% of the total","excerpt":"\"The FBI reported 779,542 burglary offenses in 2024, a rate of 229.2 per 100,000 inhabitants, down from 253.3 per 100,000 in 2023 — a 9.5% year-over-year decline.\"\n","source_date":"2025-09-30","source_accessed":"2026-04-22","archive_url":"http://web.archive.org/web/20260426024116/https://www.fbi.gov/news/press-releases/fbi-releases-2024-reported-crimes-in-the-nation-statistics","calculation_notes":"779,542 total burglaries × 52% residential = ~405,776 residential burglaries. Divided by ~140 million US households = 2.9 per 1,000 households per year. Over 40 years: 1 - (1 - 0.0029)^40 = 0.109, or ~11% lifetime probability. The 2024 rate continues a long-term decline: the 2005 rate was 738.9 per 100,000, meaning burglary has fallen ~69% in rate over two decades.\n"},{"url":"https://bjs.ojp.gov/content/pub/pdf/vdhb.pdf","title":"Victimization During Household Burglary","publisher":"Bureau of Justice Statistics, U.S. Department of Justice","source_type":"govt_report","statistic":"28% of burglaries occurred while a household member was present; 7% involved violence; 38% of entries were through unlocked doors/windows","excerpt":"\"A household member was present in roughly 28% of the 3.7 million average annual burglaries that occurred between 2003 and 2007. In about 7% of all household burglaries, a household member experienced some form of violent victimization. Simple assault (15%) was the most common form of violence when a resident was home.\"\n","source_date":"2010-09-01","source_accessed":"2026-04-22","archive_url":"https://web.archive.org/web/20260331202522/https://bjs.ojp.gov/content/pub/pdf/vdhb.pdf","calculation_notes":"BJS NCVS data (2003-2007 average): 3.7 million burglaries/year (including unreported). 28% occurred with someone home (\"hot burglary\") = ~1.04 million. Of those, 26% involved violence against a resident. 38% of all entries used no force (unlocked door/window). Of occupied burglaries, offenders entered through an unlocked (28%) or open (27%) door/window in 55% of cases. 65% of violent burglary offenders were known to the victim. These figures are from 2003-2007; current absolute numbers are lower due to the overall decline in burglary, but the proportional breakdowns (occupied vs unoccupied, forced vs unforced) are the most recent detailed data available.\n"},{"url":"https://bjs.ojp.gov/library/publications/lifetime-likelihood-victimization-0","title":"Lifetime Likelihood of Victimization","publisher":"Bureau of Justice Statistics (Koppel, 1987)","source_type":"govt_report","statistic":"72% lifetime probability of household burglary (based on 1975-1984 NCVS rates)","excerpt":"\"Based on 1975-84 annual victimization rates and life tables, the lifetime likelihood of being a victim of a personal theft or household burglary was estimated at 72% for burglary.\"\n","source_date":"1987-03-01","source_accessed":"2026-04-22","archive_url":"https://web.archive.org/web/20260426201815/https://bjs.ojp.gov/library/publications/lifetime-likelihood-victimization-0","calculation_notes":"Koppel (1987) used 1975-1984 NCVS annual burglary rates (which ranged from ~60-80 per 1,000 households) and NCHS life tables. At those rates the 72% lifetime figure was mathematically sound. However, the NCVS burglary rate has fallen from ~63/1,000 (1994) to ~12/1,000 (2019) — a decline of over 80%. Recalculating with the 2024 FBI police-reported rate of 2.9/1,000 yields ~11% over 40 years. The 72% figure is included because it is still widely cited and shapes public perception, but it is no longer representative of current risk.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Identity theft victimization (lifetime, US)","lifetime_us_adult":0.6},{"label":"Homicide victimization (lifetime, US)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"doors and windows left unlocked","multiplier":2.5,"notes":"38% of burglaries involve no forced entry (FBI UCR 2019); BJS found unlocked or open doors/windows were the entry point in 55% of occupied-home burglaries"},{"factor":"someone home during the burglary (hot burglary)","multiplier":0.4,"notes":"Only 28% of burglaries occur while a household member is present (BJS 2003-2007); burglars overwhelmingly prefer empty homes. However, hot burglaries carry a 26% chance of violence against the resident"},{"factor":"no one home (property unoccupied)","multiplier":1.4,"notes":"72% of burglaries target unoccupied homes; the base rate already includes both types, so the uplift for confirmed-empty homes is moderate"},{"factor":"visible alarm system or security cameras","multiplier":0.2,"notes":"83% of convicted burglars said they would avoid homes with visible security indicators; Rutgers study found 60% fewer break-in attempts at alarmed homes"},{"factor":"single female-headed household with children","multiplier":2.5,"notes":"BJS NCVS found this demographic had the highest occupied-home burglary rate at 22 per 1,000 households vs 4 per 1,000 for married couples without children"},{"factor":"high-burglary metro (e.g. Springfield IL, 1,036/100K)","multiplier":4.5,"notes":"Springfield IL's 2024 burglary rate was 1,036/100K — 4.5x the national average of 229/100K; conversely, Manchester-Nashua NH was 46/100K (0.2x)"},{"factor":"large dog in the household","multiplier":0.5,"notes":"Survey of 86 burglars identified a large, loud dog as a major deterrent; neighborhoods with higher dog ownership showed ~33% lower robbery rates"}],"short_label":"Home burglary (global)","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The FBI UCR data captures only police-reported burglaries; the NCVS victimization survey consistently estimates 2-4x more burglaries because many go unreported. The lifetime figure of ~11% uses the conservative FBI rate; including unreported burglaries would push it to 15-25%. The detailed breakdown of occupied vs unoccupied, forced vs unforced entry comes from BJS data covering 2003-2007 — the most recent report with that level of detail. Overall burglary rates have fallen substantially since then, but the proportional breakdowns are assumed to be roughly stable. The \"hot burglary\" rate of 28% is notably higher than in England and Wales (~7-13%), possibly because US burglars face higher risk of armed confrontation and still proceed. Apartment burglary dynamics differ from single-family homes; multi-unit buildings with 10+ units had lower rates (8/1,000 vs ~20+/1,000 for detached homes).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-22","reviewed":true,"generated_at":"2026-04-22","image":{"alt":"A front door with a simple lock and a welcome mat, viewed from outside at dusk, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/home-burglary","api_url":"https://likelier.app/api/fears/home-burglary.json"},{"slug":"tooth-loss-dentures-by-65","question":"What are the odds of losing all your teeth by age 65?","quick_answer":"About 1 in 9 US adults reaching 65 to 74 today has lost all natural teeth, but the trend is steeply down: from ~19% of all adults in the late 1950s to under 5% by 2009-2012, with ~2.6% projected for 2050. For someone in their 20s or 30s now, the lifetime odds sit well below the current seniors' figure.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Fear of losing teeth is common but tends to be unfocused — people worry about individual tooth loss and cavities more often than complete edentulism. Complete tooth loss (requiring full dentures) is widely associated with past generations and poverty, and many adults underestimate how common it remains in the US. The declining trend is real and widely unappreciated; most people's mental model of edentulism prevalence dates to their grandparents' era rather than to current data.\n","rough_estimate":"~1 in 5 lifetime feels about right to many people","kind":"intuition"},"native":{"display":"~11.4% of US adults aged 65–74 are completely edentulous","numerator":114,"denominator":1000,"unit":"prevalence at ages 65–74","population":"US adults aged 65–74 (CDC 2024 Oral Health Surveillance Report, NHANES 2017–2020)"},"normalized":{"lifetime_us_adult":0.11,"display":"~1 in 9 chance of complete tooth loss by age 75","log_value":-0.96,"assumptions":"Complete tooth loss (edentulism) is irreversible: once all natural teeth are lost, they are not recovered. Therefore, the prevalence of edentulism in an age group directly equals the cumulative incidence through that age — no compounding is needed. CDC 2024 Oral Health Surveillance Report (NHANES 2017–2020 data) reports edentulism prevalence of 11.4% among adults aged 65–74 and 19.7% among those aged 75 and older. A conservative lifetime estimate of approximately 11% reflects the probability of being edentulous by the early-senior years (ages 65–74). Full lifetime risk (reaching any age) would approach the 65+ overall figure of approximately 13.8% (CDC 2020 data). Slade et al. (2014) document a strong declining secular trend: prevalence in all US adults fell from 18.9% in 1957–1958 to 4.9% in 2009–2012, with projection to 2.6% by 2050. A person currently aged 18–30 faces substantially lower lifetime risk than the current 65+ cohort reflects. The 11% point estimate represents current middle-aged US adults reaching their 60s. Uncertainty range 0.06–0.18 reflects the steep trend and the large socioeconomic gradient.\n","uncertainty":{"low":0.06,"high":0.18},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/oral-health/php/2024-oral-health-surveillance-report/selected-findings.html","title":"2024 Oral Health Surveillance Report: Selected Findings","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Edentulism (complete tooth loss) prevalence: 11.4% for adults aged 65–74 and 19.7% for adults aged 75 and older, based on NHANES 2017–March 2020 prepandemic data. Complete tooth loss among US adults 65+ decreased from 16.2% in 2012 to 13.8% in 2020.\n","excerpt":"\"The prevalence of edentulism among adults increased from 1.2% at 35–49 years to 5.9% at 50–64 years, 11.4% at 65–74 years, and 19.7% at 75 years or older. More than 1 in 10 adults aged 65–74 years had lost all their teeth, and nearly 1 in 5 adults aged 75 or older had lost all their teeth.\"\n","source_date":"2024-10-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525100717/https://www.cdc.gov/oral-health/php/2024-oral-health-surveillance-report/selected-findings.html","calculation_notes":"Edentulism is irreversible, so prevalence at age group = cumulative incidence through that age. Prevalence at 65–74 = 11.4%; this is the primary point estimate. The progression to 19.7% at 75+ indicates continued incident edentulism in the 75–84 decade. Point estimate of 11% is rounded from 11.4% and represents the probability of complete tooth loss by the mid-60s for current US adults. The ~13.8% figure for all adults 65+ blends the 65–74 and 75+ groups and gives a higher overall estimate.\n","independence_note":"CDC 2024 Oral Health Surveillance Report uses NHANES 2017–2020 clinical dental examination data, methodologically independent of the Slade 2014 J Dent Res analysis which used time-series data from multiple earlier NHANES cycles.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25146182/","title":"Projections of U.S. Edentulism Prevalence Following 5 Decades of Decline","publisher":"Slade GD, Akinkugbe AA, Sanders AE — Journal of Dental Research","source_type":"peer_reviewed","statistic":"Edentulism prevalence in US adults declined from 18.9% in 1957–1958 to 4.9% in 2009–2012; projected to reach 2.6% (95% PI: 2.1%–3.1%) by 2050. Number of edentulous Americans projected to fall 30% from 12.2 million in 2010 to 8.6 million by 2050.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] During the half century spanning the surveys, prevalence of edentulism in U.S. adults declined from 18.9% to 4.9%. With the passing of generations born in the mid-20th century, the rate of decline in edentulism is projected to slow, reaching 2.6% (95% prediction limits: 2.1%, 3.1%) by 2050. The predicted number of edentulous people in 2050 (8.6 million) will be 30% lower than the 12.2 million edentulous people in 2010.\"\n","source_date":"2014-08-21","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260319101732/https://pubmed.ncbi.nlm.nih.gov/25146182/","calculation_notes":"Used to establish the secular trend: edentulism is declining rapidly and is not a static risk. The 18.9% prevalence in the 1950s explains why older current cohorts (now 75+) have higher rates. Younger adults today will have lower lifetime risk than the 11% CDC point estimate suggests, because the rate continues to fall. Not used as the primary point estimate — the CDC NHANES data are more current.\n","independence_note":"Slade et al. analyzed five independent NHANES cross-sectional surveys (1957–2012), methodologically distinct from the CDC 2024 surveillance report which uses a single more recent NHANES cycle (2017–2020). The two sources are complementary but measure different time periods with different methodologies.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4212322/","title":"Projections of U.S. Edentulism Prevalence Following 5 Decades of Decline (PMC full text)","publisher":"Slade GD et al., J Dent Res — PubMed Central","source_type":"peer_reviewed","statistic":"Strong socioeconomic gradient in edentulism: prevalence among those with less than high school education is more than triple the rate among those with higher education; smoking and poverty are independent risk factors.\n","excerpt":"\"[Paraphrase from PMC full text — paywalled at journal] Edentulism prevalence is strongly patterned by socioeconomic status: adults with less than a high school education have more than three times the edentulism rate of college graduates. Smoking and poverty are independent predictors of complete tooth loss even after controlling for access to dental care.\"\n","source_date":"2014-08-21","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260227071905/https://pmc.ncbi.nlm.nih.gov/articles/PMC4212322/","calculation_notes":"Used to support the personal_factor_multiplier values for education, insurance, and smoking. The 3× education gradient directly informs the ~2× insurance-access multiplier used in the entry.\n","independence_note":"Same study as source 2; included separately to distinguish the trend projection (source 2) from the socioeconomic modifiers (source 3). Both are from the same Slade et al. 2014 paper.\n"}],"comparison_anchors":[{"label":"Complete hearing loss (lifetime, US adult)","lifetime_us_adult":0.07},{"label":"Hip replacement (lifetime, US adult)","lifetime_us_adult":0.09},{"label":"Cataract surgery (lifetime, US adult 65+)","lifetime_us_adult":0.5}],"personal_factor_multipliers":[{"factor":"No dental insurance or coverage (lifetime)","multiplier":2,"notes":"Adults without dental insurance are roughly twice as likely to experience complete tooth loss as those with consistent coverage, reflecting delayed care, fewer preventive visits, and untreated disease progression.\n"},{"factor":"Current or former smoker","multiplier":2.5,"notes":"CDC data show current smokers aged 65+ have 29.4% edentulism prevalence vs ~10–11% for non-smokers; smoking impairs periodontal healing and accelerates bone loss. Former smokers retain elevated risk compared to never-smokers.\n"},{"factor":"Diabetes diagnosis","multiplier":2,"notes":"Diabetes (especially poorly controlled) is an established independent risk factor for periodontal disease and tooth loss; diabetic adults have approximately twice the edentulism risk of non-diabetics with similar dental care access.\n"},{"factor":"Regular dental cleaning every 6 months (lifetime habit)","multiplier":0.3,"notes":"Adults with consistent preventive dental care have substantially lower edentulism rates; regular cleanings and early intervention reduce risk by approximately 70% compared to adults who seek dental care only for acute symptoms.\n"}],"short_label":"Complete tooth loss","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"degenerative","outcome_type":"chronic_illness","valence":"negative","caveats":"Complete edentulism (loss of all natural teeth) is a very different outcome from losing one or several teeth, which is far more common. NHANES data measure complete tooth loss confirmed by clinical dental examination. The strong declining secular trend is the most important contextual fact: the current 65+ prevalence of ~14% reflects generations born in the 1930s–1950s who had limited fluoride exposure, less dental care, and higher smoking rates. Adults currently aged 18–45 will likely face substantially lower lifetime risk, possibly in the 5–8% range, if current trends continue. The pronounced socioeconomic gradient (33.4% edentulism among those without a high school diploma vs 9.5% for higher education per CDC 2024) means the 11% average is not evenly distributed — it is concentrated in lower-income, lower-education, and minority populations. Edentulism has been in long-term decline in the US since the 1950s; unlike many other health risks on this site, it is becoming less common rather than more.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A stylized illustration of a single white tooth shape on a pale background, flat vector editorial."},"canonical_url":"https://likelier.app/tooth-loss-dentures-by-65","api_url":"https://likelier.app/api/fears/tooth-loss-dentures-by-65.json"},{"slug":"hitchhiking-assault","question":"What are the odds of a serious incident when hitchhiking?","quick_answer":"Across a hitchhiking lifetime (~200 rides), the odds of any adverse incident — intoxicated driver, harassment, or attempted theft — are about 1 in 9, roughly 6 per 10,000 rides; physical assault specifically is a subset, around 2-4% over that same lifetime. Many people imagine per-trip assault odds of 1 in 50 to 1 in 200, which would make hitchhiking vastly more dangerous than the data show.\n","category":"crime","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Hitchhiking occupies a durable slot in the cultural horror catalogue: the lone figure on the roadside is almost always, in thriller fiction and true-crime podcasts, either predator or prey. Parents warn teenagers with the certainty of established fact, and police in several US states and Canadian provinces have issued formal advisories against the practice. The mental image is specific enough — isolated highway, unknown driver, no witnesses — to produce visceral discomfort even in people who have never hitched a ride. The intuitive sense is that serious harm is a realistic outcome: not certain, but plausible enough to make the activity unreasonable. Most people who hold this belief would put the assault probability somewhere in the range of 1 in 50 to 1 in 200 per trip — a magnitude that, if true, would make hitchhiking roughly as dangerous as working in coal mining by injury rate.\n","kind":"intuition"},"native":{"display":"~6 in 10,000 rides","numerator":6,"denominator":10000,"unit":"per ride","population":"hitchhikers, self-reported ride-level survey (Hitchlog community, predominantly European, experienced hitchhikers)"},"normalized":{"lifetime_us_adult":0.113,"display":"~1 in 9 over a hitchhiking lifetime","log_value":-0.947,"assumptions":"Assumes 200 total rides over a regular hitchhiker's active years. Formula: 1 − (1 − 0.0006)^200 ≈ 0.113. The 0.06% per-ride rate covers the broadest adverse category (intoxicated driver, sketchy person, sexual harassment, attempted theft); physical assault alone would be a subset, roughly 1–2 per 10,000 rides, implying a ~2–4% lifetime rate for assault specifically. Wide uncertainty band reflects self-selected sample, geographic concentration in Europe, and gender skew (61% male) of the Hitchlog dataset.\n","uncertainty":{"low":0.03,"high":0.3},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://globaldane.com/is-hitchhiking-dangerous","title":"Is Hitchhiking Dangerous? Here are The Statistics (9,564 Rides)","publisher":"GlobalDane (analysis of Hitchlog community data)","source_type":"reputable_reference","statistic":"94.71% of 9,564 logged rides rated \"Good\"; 0.06% rated \"Very bad\" (approximately 6 per 10,000 rides), encompassing intoxicated drivers, sketchy people, sexual harassment, or attempted theft; 61% male, 39% female hitchhikers; average age 25; 58% hitchhiked alone.\n","excerpt":"\"94.71% Good, 4.23% Very good, 0.80% Neutral, 0.21% Bad, 0.06% Very bad... The 0.06% 'Very bad' experiences typically involve intoxicated drivers, sketchy people, sexual harassment, or attempts of theft... hitchhiking isn't as dangerous as it's made out to be.\"\n","source_date":"2021-01-01","source_accessed":"2026-05-01","archive_url":"https://web.archive.org/web/20260505055441/https://globaldane.com/is-hitchhiking-dangerous","calculation_notes":"Primary rate source. 0.06% = 6 per 10,000 rides, from 9,564 rides logged by 729 Hitchlog users. Used directly as the native per-ride rate. Lifetime probability computed as 1 − (1 − 0.0006)^200 ≈ 0.113 for 200 lifetime rides. The 200-ride assumption represents a committed recreational hitchhiker over several years of activity. Sample is self-selected (experienced hitchhikers using a logging app), 61% male, geographically concentrated in Europe, so the figure likely underestimates risk for infrequent, solo female hitchhikers in less familiar territory. \"Very bad\" collapses qualitatively different outcomes; physical assault alone is a subset, estimated at ~1–2 per 10,000 rides based on the proportion of serious crimes in the Cal HP 1974 data.\n","independence_note":"Draws on Hitchlog community data; independent of both the California Highway Patrol study and Beauregard et al.\n"},{"url":"https://www.ojp.gov/ncjrs/virtual-library/abstracts/california-crimes-and-accidents-associated-hitchhiking","title":"California Crimes and Accidents Associated with Hitchhiking","publisher":"California Highway Patrol, Operational Analysis Section (archived by Office of Justice Programs, NCJRS)","source_type":"govt_report","statistic":"Hitchhikers involved in 0.63% of crimes during study period; 71.7% of hitchhiker-related crimes had the hitchhiker as victim; female hitchhikers seven times more likely to be crime victims than males; ~80% of crimes against female hitchhikers were sex-related; less than 1% of hitchhikers were killed; 41% of crimes within first 2.9 miles; over 40% of crimes between 9 pm and 3 am.\n","excerpt":"\"Hitchhikers were more likely to be victims (71.7 percent) than perpetrators (28.3 percent) of major crimes... Female hitchhikers were found to be seven times more likely to be victims of crimes than males... Approximately 80 percent of the crimes against female hitchhikers were sex related... Less than 1% of hitchhikers were killed... The results do not show that hitchhikers are over-represented in crimes or accidents beyond their numbers.\"\n","source_date":"1974-01-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260505055533/https://www.ojp.gov/ncjrs/virtual-library/abstracts/california-crimes-and-accidents-associated-hitchhiking","calculation_notes":"Study conducted by the California Highway Patrol's Operational Analysis Section in response to California Senate Resolution 18 (1973). Data collected May–October 1973 from 662 crime and accident reports involving hitchhikers. The 0.63% figure is a prevalence share of all California crimes — not a per-ride incidence rate — because the total ride count during the study period was never measured. Used here for the gender multiplier (7×), crime-type breakdown (80% sex-related for women), and temporal/distance pattern (40%+ of crimes at night; 41% within 2.9 miles) rather than for a base rate. Data are 50+ years old; included for the crime-profile breakdown, which remains the most detailed available.\n","independence_note":"Primary government study; fully independent of GlobalDane/Hitchlog and Beauregard et al.\n"},{"url":"https://journals.sagepub.com/doi/10.1177/0306624X241313287","title":"Lost Highways: An Examination of the Question of Risk Involved in Sexual Homicides of Hitchhiking Victims","publisher":"International Journal of Offender Therapy and Comparative Criminology (Beauregard E, Chopin J, DeLisi M, 2025)","source_type":"peer_reviewed","statistic":"First published academic study to analyse sexual homicides of hitchhiking victims from both offender and crime-scene perspectives using the Sexual Homicide International Database (SHIelD); physical attacks described as rare relative to total exposure; offenders targeting hitchhikers view them as opportunities for confinement without restraint.\n","excerpt":"\"Despite cultural references to the dangers of hitchhiking, particularly for sexual homicide, no published research investigates these incidents from both an offender and crime scene perspective... Offenders targeting hitchhikers view them as opportunities for confinement without physical restraint, often committing sexual acts and theft.\"\n","source_date":"2025-01-10","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20260505055601/https://journals.sagepub.com/doi/10.1177/0306624X241313287","calculation_notes":"Beauregard et al. used the Sexual Homicide International Database (SHIelD) to compare hitchhiking-victim sexual homicides with sex-trade-victim cases. The study is qualitative and comparative — it characterises offender and crime-scene profiles rather than computing an incidence rate. No denominator (total hitchhiking rides) is available in this database, so no per-ride risk figure can be derived from this source. Included as the most recent peer-reviewed treatment of the worst-case event class (sexual homicide) and as evidence that the risk, while real, is rare enough that even forensic databases cannot supply a meaningful base rate.\n","independence_note":"Fully independent of the California Highway Patrol study and GlobalDane/Hitchlog; uses an international forensic homicide database.\n"}],"comparison_anchors":[{"label":"Lifetime chance of being a US violent crime victim (any single year)","lifetime_us_adult":0.023},{"label":"Drowning death (US lifetime)","lifetime_us_adult":0.0089},{"label":"Lyft/Uber: serious sexual assault per ~200 rideshare trips","lifetime_us_adult":0.0004}],"personal_factor_multipliers":[{"factor":"Female hitchhiker","multiplier":7,"notes":"California CHP 1974: female hitchhikers seven times more likely to be crime victims than males; ~80% of those crimes were sex-related."},{"factor":"Hitchhiking at night (9 pm – 3 am)","multiplier":3,"notes":"California CHP 1974: over 40% of crimes occurred in this window vs ~6 hours of a 24-hour day, implying roughly 3–4× higher per-hour risk at night."},{"factor":"Hitchhiking with a companion (vs. solo)","multiplier":0.3,"notes":"Consistently reported across traveler accounts as substantially reducing risk; no precise multiplier available from primary data."},{"factor":"Experienced hitchhiker using standard safety norms","multiplier":0.5,"notes":"The Hitchlog sample is experienced and norm-aware; novice hitchhikers likely face higher rates. Rough estimate only."}],"myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The per-ride rate of 0.06% comes from a self-selected sample of experienced hitchhikers who voluntarily log rides on Hitchlog — predominantly male (61%), young (mean age 25), and concentrated in Europe. This almost certainly understates risk for infrequent hitchhikers unfamiliar with safety norms, solo female hitchhikers, or those hitchhiking in unfamiliar regions or at night. The \"very bad\" category in the Hitchlog data collapses qualitatively different outcomes (intoxicated driver vs. sexual assault) without distinguishing them; the physical assault rate is a subset of this figure, estimated at roughly 1–2 per 10,000 rides. The 1974 California Highway Patrol study — still the most detailed crime-profile analysis available — found female hitchhikers were seven times more likely to be crime victims than males, with ~80% of those crimes being sex-related. Over 40% of all crimes occurred between 9 pm and 3 am. A 1989 German Bundeskriminalamt study (not available in full online) reached similar conclusions: physical attacks \"very rare,\" risk \"much lower than publicly perceived,\" no general prohibition recommended. The 200-ride lifetime assumption is illustrative; someone who hitchhikes 500 rides over a career faces roughly a 26% chance of a \"very bad\" incident under the same model.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A single outstretched thumb silhouetted against a pale empty road horizon, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/hitchhiking-assault","api_url":"https://likelier.app/api/fears/hitchhiking-assault.json"},{"slug":"alzheimers-disease","question":"What are the odds of dying from Alzheimer's disease or other dementia?","quick_answer":"Roughly 1 in 8 adults worldwide will die of, or with, dementia if they live long enough. The fear tracks the condition's cultural prominence more than its personal odds — people file it as a tragedy that happens to others while badly underrating their own lifetime risk.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Dementia gets steady cultural airtime — prestige movies, celebrity diagnoses, think pieces about ageing parents — and yet it rarely shows up near the top of fear surveys ordered by the raw probability the fear represents. Most adults file Alzheimer’s as \"a tragedy that happens to some elderly people\" rather than as \"the thing that has a roughly 1-in-5 chance of ending my own life if I make it to 80\". The perceived-vs-actual gap is not that the fear is invisible; it is that the personal odds attached to it are wildly underappreciated. When the Alzheimer’s Association asks directly, people do rank it high — but ask the same people to guess their own lifetime risk from age 65 and the median answer is nowhere near the real ~1 in 4 to 1 in 5.\n","rough_estimate":"~48% of adults 50-64 are concerned about developing dementia; 73% of those with a family history consider themselves likely to develop it","kind":"survey","survey_source":{"title":"Poll Results Show Concern About Dementia","publisher":"AARP / University of Michigan Institute for Healthcare Policy and Innovation","url":"https://www.aarp.org/health/conditions-treatments/poll-on-dementia-concerns/","year":2019}},"native":{"display":"~57 million people living with dementia worldwide; ~10 million new cases per year","numerator":1,"denominator":6,"unit":"of people reaching age 85 (high-income)","population":"global adults"},"normalized":{"lifetime_us_adult":0.12,"display":"1 in ~8 lifetime (global adult)","log_value":-0.92,"assumptions":"Uses the WHO 2025 dementia fact sheet headline of 57 million people living with dementia worldwide in 2021 and nearly 10 million new cases per year as the global incidence anchor. Two complementary routes to a lifetime figure: (a) Direct from the Alzheimer’s Association 2025 Facts and Figures, which puts the conditional lifetime risk for Alzheimer’s specifically, from age 45, at 1 in 5 for women and 1 in 10 for men in the US — i.e. roughly a 15% population average lifetime incidence from mid-life, before adding non-Alzheimer’s dementias that push the all-cause dementia lifetime incidence closer to 25%. (b) From global mortality: WHO ranks dementia as the 7th leading cause of death globally; the Alzheimer’s Association reports Alzheimer’s as the 6th leading cause of death among US adults 65+; CDC FastStats reports ~116,000 US Alzheimer’s deaths in the most recent year of vital-registration data. Dementia is underreported on death certificates (the immediate cause of death is usually coded as pneumonia, cardiac arrest, or \"failure to thrive\"), so both WHO and GBD treat the reported count as a floor. Adjusting for underreporting, roughly 12-15% of adults alive today worldwide are likely to die of dementia or with dementia as a primary contributing cause if they survive other causes long enough. Headline figure 0.12 (≈ 1 in 8) with an uncertainty band of 0.09 to 0.18 to reflect the gap between the narrow death-certificate count and the broader \"dying with dementia\" framing, and to capture the large spread between regions where most adults die before the peak dementia-risk decades (sub-Saharan Africa, South Asia) and regions where most adults now routinely reach 85+ (high-income Asia-Pacific, Western Europe). Scope is global-adult-lifetime because the US-only number is meaningfully higher (~20-25% lifetime incidence from 65+) and would overstate the global baseline.\n","uncertainty":{"low":0.09,"high":0.18},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/dementia","title":"Dementia — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"57 million people worldwide had dementia in 2021, with nearly 10 million new cases every year; dementia is the 7th leading cause of death globally; Alzheimer's disease is the most common form at 60-70% of cases","excerpt":"\"In 2021, 57 million people had dementia worldwide, over 60% of whom live in low-and middle-income countries. Every year, there are nearly 10 million new cases. Alzheimer disease is the most common form of dementia and may contribute to 60–70% of cases. [...] Dementia is currently the seventh leading cause of death and one of the major causes of disability and dependency among older people globally. [...] Women are disproportionately affected by dementia, both directly and indirectly. Women experience higher disability-adjusted life years and mortality due to dementia, but also provide 70% of care hours for people living with dementia.\"\n","source_date":"2025-03-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412000654/https://www.who.int/news-room/fact-sheets/detail/dementia","calculation_notes":"57M prevalent cases across ~6.0 billion adults = ~0.95% point prevalence in adults today. 10M annual incident cases across the same adult denominator = ~1.7 per 1,000 adults per year. Naive 60-year compounding gives ~10%; age-weighting for the fact that dementia incidence is heavily concentrated above age 75 (where the annual hazard is several-fold higher than the all-adult average) pulls the realistic lifetime figure toward 12-15% for adults who survive competing mortality. The 7th-leading-cause-of-death framing is the direct mortality anchor. Used as the primary global headline because it is the only source with a concurrent prevalence + incidence + mortality-rank triple that can be reconciled into a single lifetime estimate.\n","independence_note":"WHO dementia figures draw on the same GBD / IHME upstream as most other institutional citations; treat as partially dependent with any GBD-derived source below.\n"},{"url":"https://www.alz.org/alzheimers-dementia/facts-figures","title":"2025 Alzheimer's Disease Facts and Figures","publisher":"Alzheimer's Association","source_type":"reputable_reference","statistic":"7.2 million Americans age 65+ living with Alzheimer's in 2025; lifetime risk at age 45 is 1 in 5 for women and 1 in 10 for men; 6th leading cause of death among people 65+; deaths more than doubled 2000-2022","excerpt":"\"An estimated 7.2 million Americans age 65 and older are living with Alzheimer’s in 2025. [...] By 2050, the number of people age 65 and older with Alzheimer’s may grow to a projected 12.7 million. [...] The lifetime risk for Alzheimer’s at age 45 is 1 in 5 for women and 1 in 10 for men. [...] Alzheimer’s disease was the sixth-leading cause of death among people age 65 and older in 2022. [...] One in 3 older Americans dies with Alzheimer’s or another dementia.\"\n","source_date":"2025-03-05","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260412130448/https://www.alz.org/alzheimers-dementia/facts-figures","calculation_notes":"ACS-style direct lifetime risk calculation from US life-table conditional on reaching age 45: 20% for women, 10% for men → population-weighted ~15% lifetime Alzheimer’s-specific incidence in the US from mid-life. Adding non-Alzheimer’s dementias (vascular, Lewy body, frontotemporal, mixed) pushes all-cause dementia lifetime incidence from 65 onward to roughly 25% in the US, consistent with the \"1 in 3 older Americans dies with Alzheimer’s or another dementia\" framing. This is the direct anchor for the US rows in the regional_breakdown below and justifies using ~12% as a global adult floor once LMIC competing mortality is netted out. Note that \"lifetime risk at 45\" is a conditional-on-reaching-45 figure, not a birth-cohort estimate; the population-averaged figure before conditioning is slightly lower because of pre-45 mortality.\n","independence_note":"The Alzheimer’s Association synthesizes Chicago Health and Aging Project cohort data, Framingham Heart Study dementia sub-studies, and CDC/NCHS vital registration. Not fully independent from the CDC death-certificate count cited below but uses methodologically distinct cohort-based lifetime-risk estimation.\n"},{"url":"https://www.cdc.gov/nchs/fastats/alzheimers.htm","title":"FastStats — Alzheimer Disease","publisher":"US Centers for Disease Control and Prevention / National Center for Health Statistics","source_type":"govt_report","statistic":"116,022 US Alzheimer's deaths (most recent year); 34.1 deaths per 100,000 population; 6th leading cause of death","excerpt":"\"Number of deaths: 116,022. Deaths per 100,000 population: 34.1. Cause of death rank: 6.\"\n","source_date":"2024-10-25","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163154/https://www.cdc.gov/nchs/fastats/alzheimers.htm","calculation_notes":"116,022 US Alzheimer’s deaths / ~260 million US adults ≈ 0.45 per 1,000 adults/year. Compounded over 60 adult years naively: 1 − (1 − 4.5e-4)^60 ≈ 0.027, which is the death-certificate floor for Alzheimer’s specifically. The Alzheimer’s Association’s \"1 in 3 older Americans dies with dementia\" framing is roughly an order of magnitude higher because death certificates typically record the proximal cause (pneumonia, sepsis, cardiac arrest) rather than the underlying dementia. The realistic US lifetime Alzheimer’s-attributable mortality number sits between the CDC floor (~3%) and the \"dies with dementia\" ceiling (~33%), with most methodologically serious estimates clustering around 10-15% for Alzheimer’s alone and 20-25% for all-cause dementia. Used as the mortality-rank and US-floor anchor.\n","independence_note":"CDC/NCHS vital registration is the methodological alternative to cohort-based lifetime-risk estimation. Reading this source alongside the Alzheimer’s Association figure gives the floor-vs-ceiling spread that the uncertainty band reflects.\n"},{"url":"https://www.alzint.org/news-events/news/lancet-commission-identifies-two-new-risk-factors-for-dementia-and-suggests-45-of-cases-could-be-delayed-or-reduced/","title":"Lancet Commission identifies two new risk factors for dementia and suggests 45% of cases could be delayed or reduced","publisher":"Alzheimer's Disease International","source_type":"reputable_reference","statistic":"2024 Lancet Commission: 45% of dementia cases could potentially be delayed or reduced by addressing 14 modifiable risk factors","excerpt":"\"45% of cases of dementia could potentially be delayed or reduced, marking a 5% increase from their 2020 findings. [...] failing eyesight and elevated LDL cholesterol levels [were added as new risk factors]. [...] social isolation, air pollution and vision loss [have greater impact] in late life, and less education [has greater impact] in early life.\"\n","source_date":"2024-07-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260305050336/https://www.alzint.org/news-events/news/lancet-commission-identifies-two-new-risk-factors-for-dementia-and-suggests-45-of-cases-could-be-delayed-or-reduced/","calculation_notes":"Used to anchor the personal_factor_multipliers block below. The Lancet Commission’s 45% population-attributable fraction is an upper bound for theoretical modifiable prevention; real-world effect sizes for any individual adult adopting a composite \"healthy brain\" profile are closer to a 30-50% relative risk reduction vs baseline, which is the 0.6 multiplier used in the regional_breakdown entry for the active/educated/engaged subgroup. The 14 factors span education, hearing loss, high LDL cholesterol, depression, traumatic brain injury, physical inactivity, diabetes, smoking, hypertension, obesity, excessive alcohol, social isolation, air pollution, and visual impairment.\n","independence_note":"Alzheimer’s Disease International is the global federation of national Alzheimer’s associations; this page is a summary of the peer-reviewed Lancet Commission 2024 report (Livingston et al., The Lancet, 2024). The underlying Commission is independent from the WHO / CDC mortality pipelines above.\n"},{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(24)01296-0/fulltext","title":"Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission","publisher":"Livingston et al., The Lancet (Vol 404, Aug 10, 2024)","source_type":"peer_reviewed","statistic":"Hearing loss: RR for dementia 1.4 (95% CI 1.0-1.9), weighted PAF 7% — the joint-highest of the Commission's 14 modifiable factors; dose-response 4-24% per 10 dB; hearing-aid use HR 0.83 (0.77-0.90) for dementia","excerpt":"\"We did new meta-analyses for only depression and hearing loss (as there were no recent reviews) [...] The magnitude of this risk increase varied between studies, from a 4% increase to a 24% increase in dementia risk per 10 dB decrease in hearing ability. [...] a systematic review and meta-analysis of eight cohort studies with 126903 participants [...] reported that people with hearing loss who used hearing aids had a significantly lower risk of [...] dementia (0·83, 0·77–0·90 [...]).\" Table 1 (RR for dementia, 95% CI): Hearing loss 1·4 (1·0–1·9); weighted PAF 7% (rounded), the joint-largest modifiable fraction alongside high LDL cholesterol.\n","source_date":"2024-07-31","source_accessed":"2026-06-14","calculation_notes":"The Commission's Table 1 reports a relative risk of 1.4 (95% CI 1.0-1.9) for dementia in adults with mid-life hearing loss, from a new meta-analysis run for this report. This RR is used verbatim as the 1.4 multiplier in personal_factor_multipliers (rounded as-published). Hearing loss carries the joint-highest weighted PAF (7%) of the 14 factors, hence the \"single largest modifiable risk factor\" framing — though it is tied with high LDL cholesterol at 7%, so the entry phrases it as joint-highest, not sole. The hearing-aid figure (HR 0.83) is observational, not from an RCT primary endpoint (the ACHIEVE RCT found no overall effect but a substantial effect in its higher-risk ARIC subgroup), so it is recorded as \"associated with lower risk\", not as a proven protective multiplier.\n","independence_note":"This is the underlying peer-reviewed Commission report summarised by the Alzheimer's Disease International page cited above; the two are not independent. It is independent of the WHO / CDC mortality pipelines. Added specifically to give the hearing-loss multiplier a clean primary provenance.\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death from cancer (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.12,"notes":"All-cause dementia lifetime mortality; WHO 7th-leading-cause-of-death anchor"},{"region":"US adults 65+","probability":0.25,"notes":"Those who reach 65 face roughly a 1-in-4 lifetime all-cause dementia incidence; consistent with the Alzheimer's Association 'one in three older Americans dies with Alzheimer's or another dementia' framing"},{"region":"Women (longer life expectancy, all regions)","probability":0.15,"notes":"Population-averaged; women’s higher dementia incidence is partly longevity (they live long enough to reach peak risk ages) and partly plausibly biological"},{"region":"Men","probability":0.09,"notes":"Lower lifetime figure driven primarily by lower life expectancy; competing mortality removes men from the denominator before peak dementia-risk ages"},{"region":"High-education, active, socially engaged 65+","probability":0.08,"notes":"Lancet Commission 2024: 45% of dementia cases are theoretically preventable or delayable via 14 modifiable factors; a composite healthy-brain profile yields ~30-50% relative risk reduction in observational cohorts"}],"personal_factor_multipliers":[{"factor":"APOE ε4 homozygous (two copies)","multiplier":12,"notes":"The strongest common-genetic risk factor known for late-onset Alzheimer’s; ~2-3% of the general population, aggregate odds ratio ~8-15× in pooled meta-analyses"},{"factor":"APOE ε4 heterozygous (one copy)","multiplier":3,"notes":"~25% of the general population; ~2.5-4× odds ratio for Alzheimer’s vs ε3/ε3 baseline"},{"factor":"First-degree family history of dementia","multiplier":2,"notes":"Parent or sibling with dementia roughly doubles individual risk; overlaps partially with APOE inheritance and is not additive with it"},{"factor":"Untreated mid-life hearing loss","multiplier":1.4,"notes":"2024 Lancet Commission (Livingston et al., The Lancet, 2024) new meta-analysis: RR 1.4 (95% CI 1.0-1.9) for dementia in people with baseline hearing loss. It carries the joint-highest weighted population-attributable fraction (7%) of the Commission's 14 modifiable factors. Risk is dose-responsive (4-24% increase per 10 dB of hearing lost). Hearing-aid use is associated with lower dementia risk (HR 0.83, 95% CI 0.77-0.90 in an 8-cohort meta-analysis of 126,903 people), so the multiplier applies to untreated loss."},{"factor":"Uncontrolled hypertension in midlife","multiplier":1.6,"notes":"Midlife (ages 40-65) uncontrolled blood pressure is one of the Lancet Commission’s 14 modifiable factors; the relative risk shown here is for later-life dementia, not immediate cardiovascular events"},{"factor":"12+ years education, active lifestyle, non-smoker, normal BP, no hearing loss","multiplier":0.6,"notes":"Composite protective profile from the Lancet Commission modifiable-factor framework; aligns with the ~40% relative-risk reduction observable in long-term cohorts"}],"short_label":"Alzheimer's","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Two methodological issues shape every dementia mortality number: underreporting on death certificates (the proximal cause recorded is usually something else, which is why CDC’s ~116,000 US Alzheimer’s-coded deaths sits roughly one order of magnitude below the Alzheimer’s Association’s \"one in three older Americans dies with Alzheimer’s or another dementia\" framing) and the distinction between Alzheimer’s disease specifically (~60-70% of dementias) and all-cause dementia. This entry is all-cause dementia mortality; the headline figure sits between the narrow coded-Alzheimer’s floor and the broader \"dying with dementia\" ceiling. The regional_breakdown entries for women vs men are population averages and do not control for competing mortality cleanly; the gap shrinks but does not close in cohorts restricted to adults who survive to age 80. Personal factor multipliers are illustrative relative risks from the epidemiological literature and overlap with one another (APOE status, family history, and composite lifestyle factors are not independent). Finally, the Lancet Commission’s 45% preventable-fraction is a population-attributable estimate under an idealised counterfactual where every modifiable factor is eliminated from birth; real-world individual effect sizes are closer to a 30-50% relative risk reduction for a composite healthy-brain profile.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":4,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale folded paper shape casting a long soft shadow on a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/alzheimers-disease","api_url":"https://likelier.app/api/fears/alzheimers-disease.json"},{"slug":"housing-market-crash","question":"What are the odds of a major housing market crash wiping out your home equity?","quick_answer":"A US homeowner has roughly a 1 in 8 chance of a major crash wiping out substantial equity over a lifetime, and even in the worst modern crash about 23% of mortgaged homes went underwater. People tend to picture a major crash once a decade; the real frequency is closer to once every 30 to 50 years.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"The 2008 financial crisis left a deep imprint on American risk perception. A 2024 Clever Real Estate survey found that 70% of Americans fear an imminent housing market crash, and Gallup polling consistently shows that roughly seven in ten adults consider it a \"bad time to buy a house.\" The availability heuristic is doing heavy lifting here: the 2008 collapse was the most televised financial disaster in history, and it anchored an entire generation's model of what housing markets do. Most people, when asked, treat a >20% national price decline as something that happens roughly once a decade. The actual frequency is closer to once every 30-50 years.\n","rough_estimate":"most people treat a major crash as a once-per-decade event","kind":"survey","survey_source":{"title":"70% of Americans Fear an Imminent Housing Market Crash","publisher":"Clever Real Estate","url":"https://www.noradarealestate.com/blog/americans-fear-imminent-housing-market-crash-in-2024/","year":2024}},"native":{"display":"~23% of mortgaged homes were underwater at peak (Q1 2012, US)","numerator":23,"denominator":100,"unit":"per major national crash","population":"US mortgaged homeowners during worst modern crash"},"normalized":{"lifetime_us_adult":0.12,"display":"~1 in 8 lifetime (US adult homeowner)","log_value":-0.92,"assumptions":"The probability combines two components: (1) the frequency of major national housing crashes (>20% decline) and (2) the share of homeowners who actually lose substantial equity during such an event. The US has experienced roughly 3 national-level housing price declines >20% in the past century (1929-33, arguably 1989-97 in real terms, and 2006-12). Over a 59-year adult life, a homeowner who owns for ~30 years has roughly a 60-70% chance of living through at least one such event. During the 2006-12 crash (the worst in 80 years), about 23% of mortgaged properties went underwater (CoreLogic Q1 2012). But most homeowners who bought before the bubble or had substantial equity were not wiped out. Combining the ~65% chance of experiencing a major crash with the ~20-25% conditional probability of equity destruction during such an event gives roughly 0.13-0.16. However, structural reforms since 2008 (qualified mortgage rules, higher lending standards, stress testing) reduce the prospective probability. The 0.12 central estimate reflects a modest downward adjustment for post-GFC regulatory changes. The wide uncertainty band (0.05-0.25) reflects genuine disagreement about whether a 2008-scale event can recur under current regulations.\n","uncertainty":{"low":0.05,"high":0.25},"scope":"us_adult_lifetime"},"sources":[{"url":"https://fred.stlouisfed.org/series/CSUSHPINSA","title":"S&P CoreLogic Case-Shiller U.S. National Home Price Index (CSUSHPINSA)","publisher":"Federal Reserve Bank of St. Louis (FRED)","source_type":"govt_report","statistic":"The S&P/Case-Shiller U.S. National Home Price Index declined approximately 27% from its peak in Q1 2006 to its trough in Q1 2012; the 10-city composite fell ~35%.","excerpt":"\"The S&P CoreLogic Case-Shiller U.S. National Home Price Index measures the value of the residential real estate market by tracking changes in the value of residential real estate both nationally and in 20 metropolitan regions.\"\n","source_date":"2025-12-31","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503081300/https://fred.stlouisfed.org/series/CSUSHPINSA","calculation_notes":"The Case-Shiller National Home Price NSA Index peaked at approximately 184.6 in Q2 2006 and troughed at approximately 134 in Q1 2012, a decline of roughly 27%. The 10-city and 20-city composites, which overweight bubble metros, fell further (~33-35%). Individual metros saw even larger declines: Las Vegas -55%, Phoenix -51%, Miami -47%. This is the definitive long-run US home price series, used as the foundation for the frequency-of-crashes calculation. Over the index's history (1987-present) plus pre-index academic reconstructions, roughly 3 national >20% declines have occurred in 100 years.\n"},{"url":"https://www.prnewswire.com/news-releases/corelogic-reports-negative-equity-increase-in-q4-2011-141058473.html","title":"CoreLogic Reports Negative Equity Increase in Q4 2011","publisher":"CoreLogic (via PR Newswire)","source_type":"primary_study","statistic":"11.1 million mortgaged residential properties (22.8%) were in negative equity at end of Q4 2011.","excerpt":"\"11.1 million, or 22.8 percent, of all residential properties with a mortgage were in negative equity at the end of the fourth quarter of 2011. Nevada had the highest negative equity percentage with 61 percent of all of its mortgaged properties underwater, followed by Arizona (48 percent), Florida (44 percent).\"\n","source_date":"2012-03-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503091243/https://www.prnewswire.com/news-releases/corelogic-reports-negative-equity-increase-in-q4-2011-141058473.html","calculation_notes":"CoreLogic's negative equity report provides the conditional probability: given that a major national crash occurs, what share of mortgaged homeowners actually lose their equity? The answer at peak was ~23-24% nationally, with enormous regional variation (61% in Nevada vs single digits in stable markets). This is the key input for converting \"crash frequency\" into \"personal equity loss probability.\" The lifetime estimate multiplies crash frequency (~65% chance of experiencing one in a 30-year ownership span) by conditional equity-loss probability (~20-25%), giving ~12-16%.\n","independence_note":"CoreLogic's property-level data is independent of the Case-Shiller index methodology. Case-Shiller measures price changes via repeat sales; CoreLogic compares current estimated value to outstanding mortgage balance. They measure different things and draw from different data pipelines.\n"},{"url":"https://www.federalreserve.gov/publications/files/scf23.pdf","title":"Changes in U.S. Family Finances from 2019 to 2022: Evidence from the Survey of Consumer Finances","publisher":"Board of Governors of the Federal Reserve System","source_type":"govt_report","statistic":"Primary residence equity constitutes roughly two-thirds of wealth for the median US household; median net worth rose 37% from 2019 to 2022, largely driven by home equity gains.","excerpt":"\"The typical household is far more concentrated in home equity and retirement savings, with limited exposure to stocks or private business ownership. From 2019 to 2022, public equities and home equity grew as a share of the average household balance sheet.\"\n","source_date":"2023-10-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260324210016/https://www.federalreserve.gov/publications/files/scf23.pdf","calculation_notes":"The Fed SCF establishes why housing crashes matter disproportionately: for the median household, home equity IS wealth. A 27% national price decline translates to a much larger percentage loss of net worth for leveraged homeowners because housing is typically 65-70% of total household wealth at the median. This amplification effect is why housing crashes feel catastrophic even when the absolute price decline is moderate compared to equity market drawdowns.\n","independence_note":"The SCF is a triennial household survey conducted by the Federal Reserve, entirely independent of both the Case-Shiller price index and CoreLogic's property-level data.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Car crash death (lifetime, US)","lifetime_us_adult":0.0095}],"regional_breakdown":[{"region":"National average (all mortgaged homeowners)","probability":0.12,"notes":"Central estimate; assumes ~30-year ownership and one major crash per adult lifetime"},{"region":"Speculative bubble markets (Las Vegas, Phoenix, Miami)","probability":0.3,"notes":"During 2006-12: 48-61% underwater in these metros. Higher frequency of regional booms/busts."},{"region":"Stable Midwest/Northeast metros","probability":0.04,"notes":"Markets like Pittsburgh, Buffalo, Indianapolis saw <10% declines even in 2008; less volatile historically"}],"personal_factor_multipliers":[{"factor":"high leverage (>90% LTV at purchase)","multiplier":3.5,"notes":"Minimal equity buffer means even a 10-15% decline puts the homeowner underwater; this was the core 2008 dynamic"},{"factor":"bought at local price peak","multiplier":2.5,"notes":"Timing is the dominant risk factor; buyers who purchased in 2005-2006 were vastly more exposed than those who bought in 2001"},{"factor":"adjustable-rate mortgage (ARM)","multiplier":1.8,"notes":"Payment shock from rate resets amplifies default risk during downturns; post-2008 QM rules reduce but don't eliminate this"},{"factor":"stable market, >20% equity","multiplier":0.15,"notes":"Homeowners with substantial equity in non-speculative markets almost never experience equity wipeout"},{"factor":"bought >5 years before crash with fixed-rate mortgage","multiplier":0.3,"notes":"Pre-bubble buyers had accumulated enough equity and appreciation to absorb a 27% national decline without going underwater"},{"factor":"2022-2024 new-build FHA loan with builder rate-buydown","multiplier":1.5,"notes":"Captures purchase-time (not crash-dependent) negative equity unique to the 2022-2024 new-build cohort whose advertised sub-market mortgage rates were funded by builders inflating the sale price. Per the WSJ analysis by John Comiskey (Reverse Engineering Finance) and ICE Mortgage Technology's November 2025 Mortgage Monitor, 18% of DHI Mortgage (D.R. Horton) FHA loans originated 2022-2024 are already underwater; the 1.5x multiplier is the ratio of that 18% to the 0.12 baseline. The Lennar Mortgage cohort runs higher (27%, ~2.25x). AEI Housing Center estimates builder buydowns let the 21 largest builders avoid a 10-12% price cut, so these buyers paid an inflated price with little down and went underwater without any market-wide decline.\n"}],"short_label":"Housing crash","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The 12% lifetime estimate is genuinely uncertain because it depends on the frequency of tail events for which we have only 2-3 data points in US history. The calculation treats the 2008 GFC as representative of a \"major crash,\" but it was arguably a once-per-century event driven by specific institutional failures (subprime securitization, ratings fraud, excessive leverage). Post-crisis reforms — qualified mortgage rules, stress testing, higher capital requirements — were specifically designed to prevent a repeat. If those reforms hold, the true probability is closer to the low end of the uncertainty band. If they erode or new fragilities emerge, the high end applies. The number also pools all homeowners; a renter has zero exposure, and a homeowner with 50% equity and a fixed-rate mortgage in a stable market has near-zero exposure even during a severe crash. One non-crash exception has emerged in the 2022-2024 new-build market: large homebuilders (Lennar, D.R. Horton) advertised sub-market mortgage rates funded by inflating the sale price, so buyers with small FHA down payments could be underwater the day they close, with no market-wide decline required. The WSJ's analysis (John Comiskey / Reverse Engineering Finance) found 27% of Lennar Mortgage and 18% of DHI Mortgage FHA loans originated 2022-2024 already underwater; ICE Mortgage Technology's November 2025 Mortgage Monitor put 2024-vintage FHA loans at 13% underwater versus a 1.6% national negative-equity rate. The AEI Housing Center estimates these permanent buydowns let the 21 largest builders avoid a 10-12% price cut. This is idiosyncratic, at-purchase negative equity rather than the systemic crash this entry's headline number measures.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-research-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single small house icon tilting on a downward-sloping line, muted grey and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/housing-market-crash","api_url":"https://likelier.app/api/fears/housing-market-crash.json"},{"slug":"sleep-deprivation-mortality","question":"How much does regularly sleeping less than six hours raise the risk of early death?","quick_answer":"An occasional short night is not what the data flags. The ~12% excess mortality applies to chronically sleeping under six hours; cohorts with weekday short sleep plus weekend recovery show roughly a third of that excess, and consistent sub-five-hour sleep raises it to ~50%. Consistency, not the odd bad night, is what moves the number.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Sleep deprivation occupies an unusual position in the popular risk landscape: most people know it is \"bad for you\" in the way that skipping vegetables is bad for you, but very few treat it as a serious mortality risk factor. In many professional cultures — finance, medicine, tech, military — chronic short sleep is worn as a badge of honor, a signal of dedication rather than a warning sign. The result is a risk that is simultaneously well-known and systematically underweighted. Ask a typical adult how much sleeping five hours a night shortens your life and you will rarely hear \"about a decade\" — which is closer to what the cohort data actually show for the most severe chronic short sleepers.\n","rough_estimate":"Most adults sense short sleep is unhealthy but underestimate the mortality magnitude","kind":"intuition"},"native":{"display":"HR 1.12 for <6 h/night vs 7-8 h/night, all-cause mortality","numerator":12,"denominator":100,"unit":"hazard ratio per person","population":"adults sleeping <6 hours per night chronically"},"normalized":{"lifetime_us_adult":0.12,"display":"~12% excess lifetime mortality risk (chronic <6 h sleeper)","log_value":-0.92,"assumptions":"The headline hazard ratio of 1.12 for habitual sleep <6 hours comes from two large independent meta-analyses: Cappuccio et al. 2010 (1.38 million participants, 16 prospective studies, RR 1.12, 95% CI 1.06-1.18) and Itani et al. 2017 (5.17 million participants, 153 studies, RR 1.12, 95% CI 1.08-1.16). Translating a hazard ratio into an excess lifetime probability is not straightforward — it depends on duration of exposure, competing risks, and baseline mortality — but a sustained 12% elevation in the all-cause hazard rate over a 40-50 year adult career of short sleep corresponds to roughly 10-15% excess lifetime mortality attributable to the sleep deficit alone. We use 0.12 as the point estimate. For very short sleepers (<5 h/night), Wang et al. 2020 in JAMA Network Open reported HR 1.50 for all-cause mortality among consistently short sleepers, implying substantially higher excess risk in that tail. Scope is subgroup_lifetime: this is the excess risk for someone who chronically sleeps under six hours, not a population average that includes normal sleepers.\n","uncertainty":{"low":0.06,"high":0.18},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20469800/","title":"Sleep duration and all-cause mortality: a systematic review and meta-analysis of prospective studies","publisher":"Sleep (Cappuccio, D'Elia, Strazzullo, Miller)","source_type":"peer_reviewed","statistic":"Short sleep duration RR 1.12 (95% CI 1.06-1.18) for all-cause mortality; 1,382,999 participants across 16 prospective studies; 112,566 deaths","excerpt":"\"Both short and long duration of sleep are significant predictors of death in prospective population studies.\"\n","source_date":"2010-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260505063805/https://pubmed.ncbi.nlm.nih.gov/20469800/","calculation_notes":"Cappuccio et al. pooled 16 prospective studies with follow-up ranging from 4 to 25 years. The reference category was 7-8 hours of sleep. Short sleep was defined as ≤5-6 hours depending on the individual study. The pooled relative risk of 1.12 is the primary basis for the native hazard ratio and the normalized excess lifetime risk estimate of ~12%. The meta-regression showed a linear dose-response below 6 hours — the shorter the sleep, the higher the risk.\n","independence_note":"Cappuccio 2010 and Itani 2017 draw from overlapping but not identical sets of prospective studies. Cappuccio included 16 studies; Itani included 153 studies with a much larger participant pool. The convergence on RR 1.12 from partially overlapping but independently conducted meta-analyses strengthens confidence in the point estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27743803/","title":"Short sleep duration and health outcomes: a systematic review, meta-analysis, and meta-regression","publisher":"Sleep Medicine (Itani, Jike, Watanabe, Kaneita)","source_type":"peer_reviewed","statistic":"Short sleep RR 1.12 (95% CI 1.08-1.16) for mortality; 5,172,710 participants across 153 studies; also RR 1.37 diabetes, 1.17 hypertension, 1.16 CVD, 1.38 obesity","excerpt":"\"Short sleep was significantly associated with the mortality outcome (RR, 1.12; 95% CI, 1.08-1.16). Meta-regression analyses found a linear association between a statistically significant increase in mortality and sleep duration at less than six hours.\"\n","source_date":"2017-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183750/https://pubmed.ncbi.nlm.nih.gov/27743803/","calculation_notes":"Itani et al. is the largest meta-analysis on the topic to date, covering 153 studies and over 5 million participants. It confirms Cappuccio's RR 1.12 for all-cause mortality and adds dose-response evidence: the mortality association becomes statistically significant below six hours and steepens with further reduction. The additional associations with diabetes (RR 1.37), hypertension (RR 1.17), and obesity (RR 1.38) help explain the mechanism — short sleep drives mortality partly through cardiometabolic pathways.\n","independence_note":"Partially overlapping with Cappuccio 2010 in terms of underlying primary studies, but conducted independently seven years later with a much larger study pool. The identical point estimate (RR 1.12) from a substantially expanded evidence base is reassuring.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32442289/","title":"Association of Longitudinal Patterns of Habitual Sleep Duration With Risk of Cardiovascular Events and All-Cause Mortality","publisher":"JAMA Network Open (Wang, Wang, Chen, Li, Lu, Vitiello, Wang, Tang, Shi, Lu, Wu, Bao)","source_type":"peer_reviewed","statistic":"Consistently sleeping <5 h/night: HR 1.50 (95% CI 1.07-2.10) for all-cause mortality; HR 1.47 (95% CI 1.05-2.05) for cardiovascular events; 52,599 adults followed ~6.7 years","excerpt":"\"The low-stable pattern was associated with the highest risk of CVEs (HR, 1.47; 95% CI, 1.05-2.05) and death (HR, 1.50; 95% CI, 1.07-2.10). People reporting consistently sleeping less than 5 hours per night should be regarded as a population at higher risk for CVE and mortality.\"\n","source_date":"2020-05-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183825/https://pubmed.ncbi.nlm.nih.gov/32442289/","calculation_notes":"Wang et al. used a longitudinal design tracking sleep patterns over four years rather than a single baseline measurement, which reduces misclassification of habitual sleep duration. The HR 1.50 for the consistently-short (<5 h) group is substantially higher than the pooled RR 1.12 from the meta-analyses, which mix <5 h and <6 h sleepers. This supports a steep dose-response: moving from <6 h to <5 h roughly triples the excess risk. Used as the basis for the personal_factor_multiplier for very short sleepers (<5 h).\n","independence_note":"Wang et al. is a single Chinese cohort study (Kailuan Study), fully independent of the Western-dominated meta-analyses by Cappuccio and Itani. Provides cross-cultural validation and finer dose-response granularity.\n"}],"comparison_anchors":[{"label":"Death from regular smoking (lifetime, lifelong smoker)","lifetime_us_adult":0.5},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Chronic <6 h/night (meta-analytic average)","probability":0.12,"notes":"Headline subgroup. Based on pooled RR 1.12 from Cappuccio 2010 and Itani 2017."},{"region":"Chronic <5 h/night (severe short sleepers)","probability":0.3,"notes":"Wang et al. 2020: HR 1.50 for consistently <5 h sleepers. Roughly 2.5x the <6 h excess risk."},{"region":"Occasional short sleep (weekday <6 h, weekend recovery)","probability":0.04,"notes":"Weekend recovery sleep partially attenuates the mortality signal in several cohort studies; excess risk reduced but not eliminated."},{"region":"Normal sleeper (7-8 h/night)","probability":0,"notes":"Reference category — no excess sleep-attributable mortality risk."}],"personal_factor_multipliers":[{"factor":"shift work (rotating or permanent night shifts)","multiplier":1.8,"notes":"Shift work compounds sleep deprivation with circadian disruption; Vyas et al. 2012 BMJ meta-analysis found shift work associated with ~23% increased vascular events independent of sleep duration"},{"factor":"untreated obstructive sleep apnea","multiplier":2.5,"notes":"OSA fragments sleep architecture even when total duration appears adequate; untreated severe OSA carries its own mortality hazard (HR ~1.5-2.0) that compounds with short duration"},{"factor":"combined with obesity (BMI >30)","multiplier":1.5,"notes":"Short sleep and obesity are bidirectionally linked via appetite hormones (ghrelin/leptin); combined effect on cardiometabolic mortality is more than additive"},{"factor":"very short sleep (<5 h/night chronic)","multiplier":2.5,"notes":"Wang et al. 2020: HR 1.50 vs HR 1.12 for <6 h; the excess risk roughly 2.5x the headline figure"}],"short_label":"Sleep deprivation","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry measures the excess all-cause mortality attributable to chronic short sleep duration (<6 hours/night) relative to the 7-8 hour reference category. It is a subgroup estimate, not a general-population lifetime risk. The 12% excess figure is a population average across many confounders — individual risk depends heavily on sleep quality (not just quantity), genetic short-sleeper variants (rare, <1% of the population), comorbidities, and compensatory behaviors like weekend recovery sleep. Causality is not fully established: observational studies cannot fully disentangle whether short sleep causes excess mortality or whether underlying illness causes both short sleep and death (reverse causation). However, the consistency across dozens of prospective cohorts, the dose-response relationship, and the biological plausibility via cardiometabolic pathways all support a causal interpretation. The \"10-year life-expectancy reduction\" sometimes cited in popular media applies to the most extreme chronic deprivation (<4-5 h/night over decades) and should not be generalized to the <6 h category.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-13","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single dimmed bedside clock showing late hours against a muted navy background, flat vector illustration."},"canonical_url":"https://likelier.app/sleep-deprivation-mortality","api_url":"https://likelier.app/api/fears/sleep-deprivation-mortality.json"},{"slug":"smokeless-tobacco-health","question":"What are the odds of dying from a smokeless-tobacco-related disease as a regular user?","quick_answer":"Roughly 1 in 8 lifelong regular smokeless-tobacco users die of a tobacco-attributable disease — a real risk, but well below the ~50% lifetime toll of lifelong cigarette smoking that some users assume it matches, and far above the negligible risk others imagine.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Smokeless tobacco — snus, chewing tobacco, moist snuff — occupies an ambiguous space in public health perception. Many users treat it as a substantially safer substitute for cigarettes, citing the absence of combustion and the absence of lung cancer as the defining differences. Others assume the risk is comparable to smoking, equating \"tobacco\" with \"similar hazard.\" Neither picture is accurate. The evidence points to a genuine but considerably smaller mortality risk than cigarette smoking — roughly a quarter to a third of smoking's attributable burden — concentrated mainly in cardiovascular disease and, for products higher in tobacco-specific nitrosamines (TSNA), oral and esophageal cancer. The headline figure is not zero, and it is substantially lower than the ~50% lifetime attributable mortality for a lifelong cigarette smoker.\n","rough_estimate":"Many users believe the risk is negligible; others assume it matches cigarette smoking","kind":"intuition"},"native":{"display":"Roughly 1 in 8 lifelong regular smokeless tobacco users die from a tobacco-attributable disease","numerator":1,"denominator":8,"unit":"per regular user","population":"lifelong regular smokeless tobacco users (primarily snus-type products in Scandinavia; US chewing tobacco users may have lower absolute mortality based on contemporaneous US data)"},"normalized":{"lifetime_us_adult":0.12,"display":"~1 in 8 lifetime (lifelong regular user)","log_value":-0.9208,"assumptions":"Reference subgroup: an adult who begins regular smokeless tobacco use in early adulthood and continues for decades, modelled primarily on Swedish snus data (the highest-quality long-term cohort evidence available). The 0.12 point estimate derives from the following reasoning. Byhamre et al. (PMC7825961), a pooled analysis of eight Swedish prospective cohort studies (N=169,103 never-smoking men; 10,928 deaths over 2,857,312 person-years), found current snus users had an all-cause mortality adjusted HR of 1.28 (95% CI 1.20–1.35) and a cardiovascular mortality HR of 1.27 (95% CI 1.15–1.41) relative to never-users of tobacco. The population attributable fraction implied by an HR of 1.28 is (HR-1)/HR = 0.28/1.28 ≈ 22%. Applying that attributable fraction to the total lifetime mortality of a cohort (which asymptotically approaches 1.0 over a long enough follow-up) yields an approximate lifetime attributable mortality probability of ~12–18%. The 0.12 central estimate deliberately sits at the lower end of this range because: (1) US studies using NLMS and NHIS nationally representative cohorts (Fisher et al. 2019, PMC6458834) found null all-cause mortality (HR ≈ 1.0) for exclusive smokeless tobacco users in US populations, suggesting product-class differences (US chewing tobacco and moist snuff contain higher TSNAs than Swedish pasteurized snus); (2) modern US oral cancer incidence in smokeless tobacco users is only modestly elevated over non-users in the JMIR Cancer 2024 population-based study (HR 1.4, barely significant); (3) the Swedish pooled data represent heavy long-term male snus users who are older on average than US SLT users. Uncertainty range 0.05–0.25 reflects the genuine evidence conflict between Swedish cohort mortality data and US null-finding cohort data. Scope is declared subgroup_lifetime because this is a per-regular-user probability, not a general-population lifetime risk, and is not directly comparable to per-US-adult figures on other Likelier pages.\n","uncertainty":{"low":0.05,"high":0.25},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7825961/","title":"Swedish snus use is associated with mortality: a pooled analysis of eight prospective studies","publisher":"International Journal of Epidemiology (Byhamre et al.)","source_type":"peer_reviewed","statistic":"Current snus users: all-cause mortality aHR 1.28 (95% CI 1.20–1.35); cardiovascular mortality aHR 1.27 (1.15–1.41)","excerpt":"\"Current snus users had an increased risk of all-cause mortality (aHR 1.28, 95% CI 1.20 to 1.35), cardiovascular mortality (aHR 1.27, 95% CI 1.15 to 1.41), cancer mortality (aHR 1.12, 95% CI 1.00 to 1.26) and other cause mortality (aHR 1.37, 95% CI 1.24 to 1.52) compared with never-users of tobacco. Mortality from all causes except for cancer increased with duration of snus use at baseline, although there were no clear dose–response relationships with the amount of snus used.\"\n","source_date":"2021-01-27","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20260309220143/https://pmc.ncbi.nlm.nih.gov/articles/PMC7825961/","calculation_notes":"This pooled analysis of 169,103 never-smoking Swedish men provides the primary quantitative basis for the mortality hazard estimate. The HR 1.28 all-cause figure is the main anchor. Attributable fraction = (1.28-1)/1.28 = 0.219; applied to lifetime mortality (approaching 1.0 over a full adult lifespan) yields ~22% attributable in theory; adjusted to ~12% central estimate to account for US data showing null findings (see Fisher et al.). The Swedish cohort is male-only and drawn from a population using pasteurized snus with lower TSNA levels than most US products; this is both a strength (large, clean, never-smoking) and a limitation (product and population specificity).\n","independence_note":"This is the largest and most methodologically rigorous source for snus-specific mortality. It is independent from the Fisher US cohort study below. The two datasets point in different directions, which is why both are cited.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6458834/","title":"Smokeless tobacco mortality risks: an analysis of two contemporary nationally representative longitudinal mortality studies","publisher":"Harm Reduction Journal (Fisher et al., 2019)","source_type":"peer_reviewed","statistic":"Exclusive SLT users in US: all-cause mortality HR ≈ 1.0 in both NLMS and NHIS cohorts; no evidence of excess cancer mortality","excerpt":"\"No evidence of excess mortality risk among exclusive SLT users was found in either study. [...] Heart failure: NHIS HR 2.75 (95% CI 1.55–4.89). [...] No increased mortality risk from any of the major neoplasms often associated with SLT use. [...] Exclusive smokers showed 12-fold increased lung cancer risk versus 3 deaths in 1,863 SLT-user observations.\"\n","source_date":"2019-04-01","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20250409110841/https://pmc.ncbi.nlm.nih.gov/articles/PMC6458834/","calculation_notes":"Fisher et al. analysed two US nationally representative longitudinal mortality studies: NLMS (N=210,090, ~5-year follow-up) and NHIS (N=154,286, ~10-year follow-up). Among exclusive SLT users, all-cause mortality HR was approximately 1.0 in both datasets with overlapping confidence intervals, meaning no statistically significant excess all-cause mortality was found in US populations. This is the primary evidence underlying the downward adjustment from the Swedish HR-implied ~22% to the ~12% central estimate. The notable exception is heart failure (HR 2.75, 21 deaths in NHIS), which, while based on a small event count, is consistent with the cardiovascular signal in the Swedish data and is the biological mechanism most consistently associated with smokeless tobacco across study populations.\n","independence_note":"Fisher et al. use NLMS and NHIS — entirely different populations and time periods from the Swedish pooled analysis. This is a genuine independent replication attempt, and the null result is the principal reason for the wide uncertainty band (0.05–0.25) on this entry.\n"},{"url":"https://cancer.jmir.org/2024/1/e51936","title":"Oral Cancer Incidence Among Adult Males With Current or Former Use of Cigarettes or Smokeless Tobacco: Population-Based Study","publisher":"JMIR Cancer","source_type":"peer_reviewed","statistic":"ST-only users: oral cancer incidence 20.6 per 100,000 vs never-user 22.1 per 100,000; modest HR 1.4 (95% CI 1.1–1.9) for exclusive ST vs never-tobacco","excerpt":"\"Never Cig/Current ST (smokeless tobacco users): 20.6 per 100,000 (95% CI 18.3–23.3). Never Cig/Never ST (non-users): 22.1 per 100,000 (95% CI 21.5–22.6). [...] smokeless tobacco users had 1.4 times higher oral cancer risk compared to never users (95% CI 1.1–1.9, P=.02). [...] US smokeless tobacco predominantly consists of moist snuff (~80% market share) and chewing tobacco (~18% market share).\"\n","source_date":"2024-01-15","source_accessed":"2026-05-22","archive_url":"http://web.archive.org/web/20260210032330/https://cancer.jmir.org/2024/1/e51936/","calculation_notes":"This 2024 population-based study of 19,536 US oral cancer cases is the most current direct evidence on oral cancer incidence for US SLT users. The absolute incidence rates are nearly identical between ST users (20.6) and never-tobacco users (22.1) per 100,000 person-years — a null result in absolute terms, with only a modest relative risk that is barely statistically significant. This substantially weakens the older oral-cancer-focused risk narrative for modern US smokeless tobacco users (primarily moist snuff) compared to 1980s studies that found higher RRs for dry snuff with higher TSNA content. The oral cancer channel contributes negligibly to the central estimate of 0.12, which is driven primarily by cardiovascular mortality signalled in the Swedish data.\n"}],"comparison_anchors":[{"label":"Dying from cigarette smoking (lifelong smoker)","lifetime_us_adult":0.5},{"label":"Death from ischaemic heart disease (US adult, general population)","lifetime_us_adult":0.085},{"label":"Death from any cancer (US adult)","lifetime_us_adult":0.14},{"label":"Death from oral/pharyngeal cancer (US adult, general population)","lifetime_us_adult":0.004}],"personal_factor_multipliers":[{"factor":"product type: US chewing tobacco or moist snuff (higher TSNA)","multiplier":1.5,"notes":"US chewing tobacco and older dry-snuff products contain substantially higher tobacco-specific nitrosamines than Swedish pasteurized snus, which uses heat treatment (Lund & Paulsen 2005; Surgeon General 2004 smokeless tobacco chapter). Older US meta-analyses found oral cancer RRs of 2.6–4.7 for chewing tobacco, compared to near-null for Swedish snus (Valen et al. 2023, Int J Cancer). Adjusts the Swedish-data-anchored headline upward for US product users.\n"},{"factor":"Swedish pasteurized snus (low TSNA)","multiplier":0.7,"notes":"Swedish snus undergoes pasteurization rather than fermentation, dramatically reducing TSNA content. Valen et al. 2023 systematic review found no statistically significant increase in cancer risk for exclusive Swedish snus users. The Swedish pooled mortality finding (HR 1.28) is thus driven primarily by cardiovascular rather than cancer pathways.\n"},{"factor":"concurrent cigarette smoking or recent quitter","multiplier":4,"notes":"Former smokers who switch to SLT still carry residual tobacco-attributable cancer and cardiovascular risk from prior cigarette exposure; JMIR Cancer 2024 found former-cigarette/current-ST users had lower oral cancer risk than continuing smokers but substantially higher than never-tobacco users.\n"},{"factor":"existing cardiovascular disease or hypertension","multiplier":2,"notes":"Smokeless tobacco causes acute nicotine-driven sympathetic activation raising heart rate and blood pressure (AHA Circulation 2023). In individuals with existing CVD or hypertension, this pressor effect is the primary mechanism behind the heart failure and fatal MI signals consistently found across study populations; excess risk concentrates in those with established disease.\n"},{"factor":"heavy use (>6 pouches/day or equivalent)","multiplier":1.5,"notes":"Duration of snus use shows a dose-response for all-cause and cardiovascular mortality in the Swedish pooled analysis (≥15-year users: aHR 1.29 vs 1.28 overall). High-frequency users accumulate greater nicotine exposure per day, amplifying the cardiovascular pressor pathway.\n"}],"short_label":"Smokeless tobacco","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry is anchored on Swedish snus cohort data, which provides the largest and most methodologically rigorous long-term mortality evidence. US cohort data (NLMS and NHIS) found null all-cause mortality for exclusive SLT users, which produces the wide uncertainty range (0.05–0.25). The discrepancy likely reflects product differences (Swedish snus is pasteurized with lower TSNA content than most US chewing tobacco and moist snuff), cohort composition, and follow-up duration differences. The headline 0.12 is an intermediate estimate designed to represent a typical long-term regular user of Western smokeless tobacco products, weighted toward the Swedish data as the longer-follow-up source, with downward adjustment for the US null findings. South and Southeast Asian smokeless tobacco products (gutkha, pan masala, betel quid with tobacco) carry substantially higher cancer risks and are in a different risk category entirely; this entry does not apply to those products. The oral cancer channel for modern US SLT users (primarily moist snuff) is substantially weaker than older literature suggested — contemporary incidence data show near-identical rates in SLT users and never-tobacco users. The primary remaining pathway is cardiovascular, particularly fatal myocardial infarction and heart failure in existing-CVD patients.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-22","last_reviewed":"2026-05-22","reviewed":true,"generated_at":"2026-05-22","image":{"alt":"A small tin container casting a soft shadow on a muted warm surface, flat vector illustration."},"canonical_url":"https://likelier.app/smokeless-tobacco-health","api_url":"https://likelier.app/api/fears/smokeless-tobacco-health.json"},{"slug":"cycling-helmetless-head-injury","question":"What are the odds of serious head injury when cycling without a helmet?","quick_answer":"For a frequent unhelmeted urban cyclist over a riding career, the lifetime risk of a serious head injury is roughly 1 in 8. It is real and reducible by wearing a helmet, but well below the near-certainty many non-cyclists imagine. Road infrastructure is a larger driver than helmet use alone.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Cultural messaging about unhelmeted cycling tends to cluster around a single vivid claim: riding without a helmet means certain, catastrophic brain damage. Non-cyclists often treat it as close to a coin flip across a riding life, and the PSA framing (\"one crash and you're a vegetable\") pushes perceived risk toward certainty. We have not found a standalone survey isolating \"fear of cycling head injury without a helmet\", so perceived risk is marked as editorial intuition. The interesting property of this fear is that it is directionally correct — helmets really do reduce serious head injury by a large factor — but the magnitude is usually overstated, and the much larger structural driver (road infrastructure) is almost entirely absent from the conversation.\n","rough_estimate":"most non-cyclists expect near-certain serious head injury over a lifetime of unhelmeted riding","kind":"intuition"},"native":{"display":"~18.8 bicycle-TBI ED visits per 100,000 US residents per year (2009)","numerator":18.8,"denominator":100000,"unit":"per person per year (all US residents, not just cyclists)","population":"US residents of all ages, 2009-2018 CDC NEISS-AIP sample"},"normalized":{"lifetime_us_adult":0.125,"display":"~1 in 8 lifetime serious head injury (frequent unhelmeted urban cyclist)","log_value":-0.903,"assumptions":"Scope is activity_specific_lifetime — this is the probability for a frequent unhelmeted urban cyclist over a 30-year riding career, not a US population average. Starting point: the CDC MMWR series (Peterson et al., 2021) reports 596,972 ED visits for bicycle-related TBIs across 2009-2018, or roughly 60,000 per year across ~330 million US residents, giving a population-average rate of ~18 per 100,000 per year that fell to ~14 per 100,000 by 2018. Perhaps 10-15 percent of those are serious enough to warrant admission or longer follow-up (moderate-to-severe TBI rather than mild concussion), implying ~6,000 to 9,000 serious bicycle TBIs per year nationally. Concentrating that numerator on the ~15 million US adults who cycle frequently (commuters and recreational riders averaging 1+ rides per week) gives an annual serious-TBI risk on the order of 4-6 per 10,000 active cyclists. Compounded across a 30-year riding career for an unhelmeted rider in mixed-traffic urban conditions, that runs to roughly 1 − (1 − 0.0005)^30 ≈ 0.015 at the low end, but this undercounts because the active-cyclist numerator absorbs most of the incidents. Cross-checking against per-kilometre exposure — studies of European cycling cohorts (Scholten et al., Netherlands) put serious bicycle-related TBI at roughly 1-3 per million km cycled — and assuming a frequent rider covers 3,000 km/year × 30 years ≈ 90,000 km lifetime, the implied serious TBI probability lands at roughly 0.09 to 0.27, or ~1 in 4 to 1 in 11. We report 1 in 8 (0.125) as the central estimate, with a wide uncertainty band reflecting the three-fold spread between methods. Applying the Olivier-Creighton meta-analysis odds ratio of 0.31 for serious head injury, a helmeted rider on the same exposure profile would face roughly 1 in 25 over the same career.\n","uncertainty":{"low":0.05,"high":0.25},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/70/wr/mm7019a1.htm","title":"Emergency Department Visits for Bicycle-Related Traumatic Brain Injuries Among Children and Adults — United States, 2009-2018","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"596,972 ED visits for bicycle-related TBIs during 2009-2018; rate fell from 18.8 to 13.6 per 100,000","excerpt":"\"An estimated 596,972 ED visits for bicycle-related TBIs occurred in the United States\" and \"The rate per 100,000 population of ED visits for bicycle-related TBIs during this time decreased by 27.7%, from 18.8 in 2009 to 13.6 in 2018.\"\n","source_date":"2021-05-14","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260324052605/https://www.cdc.gov/mmwr/volumes/70/wr/mm7019a1.htm","calculation_notes":"The CDC MMWR report is the anchor for the numerator: ~60,000 bicycle-related TBI ED visits per year in the US, falling modestly over the decade. It does not split out \"serious\" vs \"mild\" TBI in the headline figure, but roughly 10-15 percent of TBI ED visits nationally are admitted rather than treated and released, giving a rough serious-TBI denominator of ~6,000 to 9,000 per year. Combined with an active-cyclist denominator of ~15 million frequent US riders, this yields the per-year serious-TBI risk of ~4-6 per 10,000 that underpins the normalized lifetime estimate. The paper also notes that helmets \"are not designed to prevent a concussion, which occurs after linear and rotational forces cause extreme brain movement inside the skull\" — a reminder that the helmet effect size applies to skull fracture and focal trauma, not to the full concussion outcome.\n","independence_note":"CDC NEISS-AIP sample is upstream of most US bicycle-injury aggregators, including IIHS and Injury Facts. Treat as the primary US measurement.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27450862/","title":"Bicycle injuries and helmet use: a systematic review and meta-analysis","publisher":"International Journal of Epidemiology (Olivier & Creighton 2017)","source_type":"peer_reviewed","statistic":"OR 0.31 (95% CI 0.25-0.37) for serious head injury; OR 0.35 (95% CI 0.14-0.88) for fatal head injury; OR 0.49 for any head injury; OR 0.67 for facial injury; from 40 studies with 64,000+ injured cyclists","excerpt":"\"helmet use was associated with odds reductions for head (OR = 0.49, 95% confidence interval (CI): 0.42-0.57), serious head (OR = 0.31, 95% CI: 0.25-0.37), face (OR = 0.67, 95% CI: 0.56-0.81) and fatal head injury (OR = 0.35, 95% CI: 0.14-0.88).\"\n","source_date":"2017-02-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165625/https://pubmed.ncbi.nlm.nih.gov/27450862/","calculation_notes":"Olivier & Creighton is the modern canonical meta-analysis of bicycle helmet effectiveness, aggregating 40 studies and over 64,000 injured cyclists. Its OR of 0.31 for serious head injury — a ~69 percent reduction — is the multiplier used to compare the unhelmeted and helmeted lifetime risks on this page. Notably, the meta-analysis effect size is smaller than the 85 percent reduction reported by the original 1989 Thompson-Rivara Seattle case-control study but larger than the most skeptical recent estimates; it is the figure most safety agencies now cite. The wide confidence interval on fatal head injury (0.14-0.88) reflects the small number of cohort-level fatality studies and should not be interpreted as precise.\n","independence_note":"Overlaps with the older Thompson-Rivara 1989 case-control in its literature base, but pools across four decades of studies from multiple countries and study designs, which reduces the influence of any single dataset.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/2716781/","title":"A case-control study of the effectiveness of bicycle safety helmets","publisher":"New England Journal of Medicine (Thompson, Rivara & Thompson 1989)","source_type":"peer_reviewed","statistic":"85% reduction in head injury risk and 88% reduction in brain injury risk among helmeted cyclists","excerpt":"\"7 percent of the case patients were wearing helmets at the time of their head injuries, as compared with 24 percent of the emergency room controls and 23 percent of the second control group.\"\n","source_date":"1989-05-25","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20250625092429/https://pubmed.ncbi.nlm.nih.gov/2716781/","calculation_notes":"The Thompson-Rivara-Thompson 1989 Seattle case-control study is the landmark original estimate of bicycle helmet effectiveness and the source of the much-quoted \"85 percent head injury reduction\" figure. It is cited here for historical anchoring and to show how the effect size has been revised downward over 30 years of replication — from 85 percent in the original to roughly 69 percent in the Olivier-Creighton meta. Both directions still support meaningful protection; the downward revision is about precision, not about the sign of the effect.\n","independence_note":"Upstream of every subsequent bicycle helmet review. Included as the historical reference point, not as an independent measurement.\n"},{"url":"https://www.iihs.org/topics/fatality-statistics/detail/bicyclists","title":"Fatality Facts 2023: Bicyclists","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"1,155 US bicyclists killed in 2023 (highest ever recorded); bicyclist deaths up 86% since their 2010 low","excerpt":"\"A total of 1,155 bicyclists were killed in crashes with motor vehicles in 2023, the highest number ever recorded.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165710/https://www.iihs.org/research-areas/fatality-statistics/detail/bicyclists","calculation_notes":"IIHS compiles NHTSA FARS fatality counts for bicyclists. Used here as the corroborating source for the US fatality numerator (roughly 1,100-1,200 cyclist deaths per year). IIHS also reports that cyclist deaths have risen sharply since 2010 while cycling exposure has not, which is consistent with infrastructure and driver-behavior drivers dominating any helmet-related trend.\n","independence_note":"IIHS draws from NHTSA FARS, which is a separate dataset from the CDC NEISS ED sample and the peer-reviewed meta-analysis. Used as the US mortality cross-check.\n"}],"comparison_anchors":[{"label":"Serious skiing injury (lifetime, active skier)","lifetime_us_adult":0.7},{"label":"Death on a motorcycle (lifetime, active US rider)","lifetime_us_adult":0.02},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Frequent unhelmeted urban cyclist (mixed traffic, US conditions)","probability":0.125,"notes":"Headline figure — roughly 1 in 8 lifetime risk of at least one serious head injury over a 30-year career of regular riding in US mixed-traffic urban conditions without a helmet."},{"region":"Frequent helmeted urban cyclist (same exposure)","probability":0.04,"notes":"Applying the Olivier-Creighton meta OR of 0.31 for serious head injury to the unhelmeted baseline. Roughly 1 in 25 — a ~3x reduction, not the ~6x sometimes implied by the older 1989 estimate."},{"region":"Commuter in protected bike lanes (Copenhagen/Amsterdam-style infrastructure)","probability":0.02,"notes":"Northern European cities with segregated cycling infrastructure see per-km serious head injury rates roughly a fifth of US mixed-traffic rates. Copenhagen cyclists without helmets have lower head injury rates than US cyclists in helmets — infrastructure dominates equipment."},{"region":"Off-road mountain biking, technical downhill","probability":0.4,"notes":"Order-of-magnitude estimate. Per-hour TBI risk on technical MTB terrain is several times the road baseline, concentrated in falls rather than motor vehicle collisions. Helmet use is near-universal in this subgroup but effect sizes still apply."},{"region":"Occasional recreational rider (a few times per month, paved paths)","probability":0.015,"notes":"Most US cyclists. The low exposure collapses the lifetime figure into something closer to the background TBI rate from other causes."}],"personal_factor_multipliers":[{"factor":"wearing helmet","multiplier":0.3,"notes":"Roughly 60-70 percent reduction in serious head injury per the Olivier-Creighton 2017 meta (OR 0.31). Skull fracture and focal trauma benefit most; rotational concussion benefits little."},{"factor":"commuting in protected / segregated bike lanes","multiplier":0.2,"notes":"Infrastructure is the largest single lever on cyclist head injury risk — larger than helmet choice. Per-km serious head injury rates in Dutch and Danish cycling cohorts are roughly a fifth of US mixed-traffic rates."},{"factor":"mixed-traffic urban riding, no helmet, aggressive style","multiplier":3,"notes":"Combines the baseline unhelmeted risk with the higher crash rate of riders who lane-split, run gaps, and ride in driver blind spots. Not a ceiling."},{"factor":"off-road MTB downhill / technical terrain","multiplier":5,"notes":"Per-hour TBI rate rises sharply with terrain technicality. Multiplier applies to head injury risk per hour of riding, not per kilometre."},{"factor":"night riding without lights","multiplier":2.5,"notes":"More than half of US cyclist fatalities occur in the dark despite dark riding being a minority of exposure. The lights-and-reflectors margin is comparable to the helmet margin in size."},{"factor":"alcohol involved (rider or nearby driver)","multiplier":2,"notes":"NHTSA reports alcohol involvement in about 34 percent of fatal cyclist crashes. Applies to fatality risk specifically and also raises serious non-fatal head injury risk."}],"short_label":"Cycling w/o helmet","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"\"Serious head injury\" bundles outcomes that differ by an order of magnitude. The mild concussion that gets you discharged from the ED the same day, the subdural hematoma that requires a craniotomy, and the fatal TBI are all inside the numerator, but they are not the same experience. We have aimed the headline number at the middle of that range — roughly \"injury serious enough to be admitted and leave some lasting effect\" — because that is the outcome the fear is actually about. Fatal cycling head injury is much rarer: roughly 1,100-1,200 cyclist deaths per year in the US across all causes, implying a per-year fatal-head-injury risk on the order of 1 in 15,000 for a frequent unhelmeted rider and a lifetime risk of roughly 1 in 500, similar to the lifetime odds of dying in a bicycle crash for the general US adult population. The 1989 Thompson study's 85 percent head-injury reduction has been revised to about 69 percent by the Olivier-Creighton 2017 meta, and helmet skeptics sometimes cite that downward revision as evidence of no effect — the meta still supports a meaningful effect, just a smaller one. The single largest modifier on this page is not helmet choice but infrastructure: Copenhagen cyclists without helmets riding in protected lanes have lower head-injury rates than US cyclists in helmets riding in mixed traffic. Finally, the CDC and IIHS datasets only capture crashes reported to emergency departments or police; solo falls where the cyclist drives themselves home are systematically missing from the numerator, which pushes the true rate upward by an unknown amount.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty bicycle helmet resting on a pale surface next to a coiled cable lock, flat vector illustration in muted greys."},"canonical_url":"https://likelier.app/cycling-helmetless-head-injury","api_url":"https://likelier.app/api/fears/cycling-helmetless-head-injury.json"},{"slug":"bruxism-tooth-damage","question":"What are the odds of developing clinically significant tooth damage or TMJ dysfunction from bruxism?","quick_answer":"About 13% of adults grind at clinically frequent levels, but that is the share at elevated risk, not the share whose teeth get destroyed: how often bruxism progresses to severe damage is uncertain and individual, and prevalence falls with age rather than accumulating.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Teeth grinding (bruxism) carries a vivid popular image: waking up with a sore jaw, a dentist pointing to worn cusps, a night guard presented as essential armour against impending tooth loss. The condition splits into sleep bruxism, which occurs during sleep without awareness, and awake bruxism, a daytime clenching habit often tied to concentration or stress. Cultural messaging around bruxism is almost universally alarmist: left untreated, grinding will erode enamel to stubs, fracture crowns, and destroy the temporomandibular joint. The fear is amplified by the invisibility of sleep bruxism — most people who do it cannot feel it happening, which makes the imagined damage feel uncontrollable and accumulating nightly.\n","rough_estimate":"Most adults believe untreated bruxism will progressively destroy teeth and joints","kind":"intuition"},"native":{"display":"~12.8% of adults report frequent sleep bruxism (systematic review)","numerator":128,"denominator":1000,"unit":"point prevalence","population":"General adult population, international"},"normalized":{"lifetime_us_adult":0.13,"display":"~13% of US adults have sleep bruxism at any given time; lifetime exposure is higher","log_value":-0.886,"assumptions":"The native figure (12.8% ± 3.1%) is the point prevalence of frequent sleep bruxism in adults, from Manfredini et al.'s 2013 systematic review of population studies. A 2024 global meta-analysis (Zieliński et al.) placed adult sleep bruxism prevalence at 21% when broader diagnostic criteria are applied. The lifetime risk of ever having sleep bruxism at clinical levels is somewhat higher than the cross-sectional prevalence because bruxism tends to decrease with age (peak in younger adults), so more people pass through the condition over a lifetime than are affected at any single measurement point. The 0.13 central estimate reflects the Manfredini narrower definition (frequent bruxism by questionnaire), which is the more clinically relevant threshold for tooth-damage risk. This does NOT represent the probability of experiencing severe tooth damage — that conditional probability is uncertain and highly individual. It represents the probability of being in the population that grinds enough to be at elevated risk.\n","uncertainty":{"low":0.08,"high":0.21},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.jofph.com/articles/10.11607/jop.921","title":"Epidemiology of Bruxism in Adults: A Systematic Review of the Literature","publisher":"Journal of Oral & Facial Pain and Headache","source_type":"peer_reviewed","statistic":"Sleep bruxism prevalence in adults: 12.8% ± 3.1% across studies using questionnaire-based frequent-bruxism criteria","excerpt":"\"prevalence of sleep bruxism was found to be more consistent across the three studies investigating the report of 'frequent' bruxism (12.8% ± 3.1%).\"\n","source_date":"2013-04-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250520113732/https://www.jofph.com/articles/10.11607/jop.921","calculation_notes":"Manfredini et al. systematically reviewed population-level bruxism studies using questionnaires, clinical assessment, polysomnography (PSG), and electromyography (EMG). For sleep bruxism, three studies using \"frequent\" bruxism criteria yielded a mean prevalence of 12.8% (SD ±3.1%). This is the most cited and methodologically rigorous pre-2020 estimate. The authors cautioned that findings must be interpreted with caution due to reliance on self-report. The 12.8% figure is used as the native denominator basis: 128 out of 1,000 adults have sleep bruxism at a clinically relevant frequency. This is a cross-sectional prevalence, not a per-year incidence or lifetime risk; the normalized lifetime figure is estimated slightly higher at 0.13 given that bruxism peaks in younger adults and more individuals pass through it over a lifetime than are affected at any single point.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11278015/","title":"Global Prevalence of Sleep Bruxism and Awake Bruxism in Pediatric and Adult Populations: A Systematic Review and Meta-Analysis","publisher":"Journal of Clinical Medicine","source_type":"peer_reviewed","statistic":"Global sleep bruxism prevalence 21% (all adults); awake bruxism 23%; nearly one in four individuals may experience awake bruxism","excerpt":"\"The global sleep bruxism prevalence is 21% ... nearly one in four individuals may experience awake bruxism.\"\n","source_date":"2024-07-22","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260212103446/https://pmc.ncbi.nlm.nih.gov/articles/PMC11278015/","calculation_notes":"Zieliński, Pająk, and Wójcicki analyzed 176 populations from 170 publications. The global sleep bruxism prevalence of 21% uses broader diagnostic criteria than Manfredini's \"frequent\" threshold, which explains the higher estimate. Adult female sleep bruxism was 15%; adult male 8%. This study sets the upper bound of the uncertainty range (0.21). The lower bound of 0.08 is grounded in the narrower Manfredini estimate minus one SD (12.8% − 3.1% ≈ 9.7%, rounded to 0.08 conservatively for the minimum clinically significant frequency). The 2024 figure serves as corroboration that the Manfredini estimate is not inflated; if anything, the narrower criteria undercount total exposure.\n","independence_note":"Independent meta-analysis by Polish research team; does not reuse the same primary studies as Manfredini 2013 (different search window and inclusion criteria).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37474733/","title":"Is bruxism associated with temporomandibular joint disorders? A systematic review and meta-analysis","publisher":"Evidence-Based Dentistry","source_type":"peer_reviewed","statistic":"Bruxism increases odds of TMD by 2.25x (OR 2.25, 95% CI 1.94–2.56); sleep bruxism OR 2.06, awake bruxism OR 2.51","excerpt":"\"The available data demonstrate a positive relationship between bruxism and TMD, with the presence of bruxism increasing the odds of TMD by 2.25 times (OR = 2.25, 95% CI (1.94-2.56))\"\n","source_date":"2023-07-19","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250701000820/https://pubmed.ncbi.nlm.nih.gov/37474733/","calculation_notes":"Mortazavi et al. conducted a systematic review and meta-analysis of 20 studies examining the bruxism-TMD association. The overall OR of 2.25 is the most statistically robust finding linking bruxism to a measurable clinical outcome. Sleep bruxism OR = 2.06; awake bruxism OR = 2.51. With baseline US adult TMD prevalence of approximately 5% (NCBI Bookshelf 2020), a 2.25-fold elevation implies a conditional risk of roughly 10-11% for bruxers — compared to ~5% for non-bruxers. The authors note that causality remains debated and that reverse causation (TMD pain driving bruxism) cannot be excluded from cross-sectional data.\n","independence_note":"This is an independent meta-analysis from a different research group than Manfredini and Zieliński; the three sources triangulate bruxism prevalence and its TMD consequence from separate methodological perspectives.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38574847/","title":"Tooth wear and bruxism: A scoping review","publisher":"Journal of Dentistry","source_type":"peer_reviewed","statistic":"Most studies found no or weak associations between tooth wear and bruxism; clinicians should not infer bruxism from tooth wear alone","excerpt":"\"Most studies reported no or weak associations between tooth wear and bruxism, except for the studies done on cervical tooth wear ... Dental clinicians should not infer bruxism activity solely on the presence of tooth wear.\"\n","source_date":"2024-03-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260525092242/https://pubmed.ncbi.nlm.nih.gov/38574847/","calculation_notes":"Bronkhorst et al. reviewed 30 publications on the bruxism-tooth wear relationship. 90% were cross-sectional. The review found inconclusive results: most studies reported no or weak statistical associations between objectively measured bruxism and tooth wear severity. Studies relying on self-reported bruxism showed associations more often than those using instrumental (PSG/EMG) confirmation. This finding is load-bearing for the caveats: it means the popular assumption that bruxism reliably erodes teeth is not supported by rigorous evidence, and the feared outcome (progressive enamel destruction) cannot be quantified reliably at the population level. This is why the normalized figure represents bruxism prevalence (the exposure), not tooth damage incidence (the outcome), which cannot be cleanly estimated from existing data.\n","independence_note":"Scoping review led by Bronkhorst et al. at Radboud University; independent of the Manfredini, Zieliński, and Mortazavi research groups.\n"}],"comparison_anchors":[{"label":"Any TMD (temporomandibular disorder), US adult prevalence","lifetime_us_adult":0.05},{"label":"Periodontal disease (US adults over 30)","lifetime_us_adult":0.47},{"label":"Adult tooth loss (at least one tooth, US lifetime)","lifetime_us_adult":0.69}],"personal_factor_multipliers":[{"factor":"No night guard (confirmed bruxer)","multiplier":3,"notes":"Orthodontic and prosthodontic literature; occlusal splints (night guards) are widely accepted as the standard protective intervention for bruxism-related tooth wear; retrospective comparisons suggest 3-4x higher rate of enamel attrition and crown fracture in confirmed bruxers without a guard vs those using one consistently; note: RCT-level evidence on long-term structural tooth damage is limited"},{"factor":"Stress/anxiety disorder","multiplier":2,"notes":"Zieliński et al. (Journal of Clinical Medicine, 2024) meta-analysis and Manfredini et al. (2013): psychological stress and anxiety are the most consistently identified risk factors for both sleep and awake bruxism severity; individuals with diagnosed anxiety or stress disorders show approximately 2x higher bruxism severity scores and more frequent grinding episodes"},{"factor":"Missing posterior teeth (edentulous in molar region)","multiplier":2.5,"notes":"Prosthodontic and occlusal literature: loss of posterior support concentrates masticatory forces on remaining anterior and premolar teeth, increasing occlusal load per tooth by 2-3x; this mechanically amplifies wear on remaining teeth in bruxers with posterior tooth loss"},{"factor":"Female sex","multiplier":1.9,"notes":"Zieliński et al. (Journal of Clinical Medicine, 2024): global meta-analysis found adult female sleep bruxism prevalence approximately 15% vs 8% for adult males — a roughly 1.9x higher prevalence; the sex ratio is less pronounced in awake bruxism; the mechanism is not established but hormonal factors and differential stress response have been proposed"}],"short_label":"Bruxism tooth damage","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The normalized figure represents the prevalence of sleep bruxism (the exposure), not the probability of clinically significant tooth damage (the outcome). These are not the same thing. The relationship between bruxism and tooth wear is genuinely uncertain: a 2024 scoping review of 30 studies found mostly weak or null associations between objectively measured bruxism and tooth wear. The stronger finding is the TMD link (OR 2.25), but even there, causality runs in both directions — pain from TMD can drive clenching behaviour, making it difficult to separate cause from consequence in cross-sectional data. Prevalence estimates vary widely (8% to 31%) depending on whether bruxism is defined by questionnaire self-report, clinical exam, or polysomnography. Most studies use self-report, which overestimates definite bruxism. The nocturnal presentation (sleep bruxism) is by definition unobservable without a bed partner or monitoring device, creating inherent measurement uncertainty. Bruxism tends to decrease with age, so younger adults carry more of the exposure burden. Women show slightly higher rates in some studies. Stress, sleep disorders, caffeine, and certain medications (SSRIs, stimulants) are associated risk factors, but causation has not been established for any of them.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A single molar tooth on a pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/bruxism-tooth-damage","api_url":"https://likelier.app/api/fears/bruxism-tooth-damage.json"},{"slug":"mail-check-fraud","question":"What are the odds of becoming a victim of mail theft or mailed-check fraud?","quick_answer":"About 1 in 8 over a lifetime for a US household, at an annual per-household rate near 0.2%. That is roughly in line with most people's sense that the yearly risk is under 1%.\n","category":"crime","tags":["household","digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Mail theft and check washing have surged since 2020 but remain largely invisible to people who have not been directly affected. Most Americans who learned about mail fraud in childhood think of it as a dated crime involving elderly targets; the modern version -- organized rings stealing USPS collection box deposits and washing checks -- is less widely known. No rigorous national survey isolates worry specifically about mailed-check fraud.\n","rough_estimate":"Most people would put annual personal risk at under 1%; the aggregate data suggests this is roughly calibrated for the population average","kind":"intuition"},"native":{"display":"299,020 mail theft complaints to USPIS, March 2020 - February 2021","numerator":299020,"denominator":130000000,"unit":"per year","population":"US households"},"normalized":{"lifetime_us_adult":0.13,"display":"~1 in 8 lifetime (US household, using March 2020 - February 2021 complaint rate)","log_value":-0.89,"assumptions":"The US Postal Inspection Service (USPIS) received 299,020 mail theft complaints between March 2020 and February 2021 (a 161 percent increase over the prior 12 months). There are approximately 130 million US households. Annual probability per household: 299,020 / 130,000,000 = 0.0023. Compounded over 59 adult years: 1 − (1 − 0.0023)^59 ≈ 0.127 ≈ 0.13. This is an upper-bound figure because (1) a single victim may file multiple complaints; and (2) this 12-month window captured the early pandemic surge near peak; subsequent rates have declined with enforcement. Check fraud SARs (FinCEN data) provide a separate line of evidence: 680,000 check fraud SARs were filed in 2022 by financial institutions, roughly double the 350,000 in 2021. Not all check fraud is mail-related, but FinCEN specifically attributed the surge primarily to mail theft. Given under-reporting, the true victimization rate likely exceeds the complaint count.\n","uncertainty":{"low":0.05,"high":0.25},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fincen.gov/system/files/shared/FinCEN%20Alert%20Mail%20Theft-Related%20Check%20Fraud%20FINAL%20508.pdf","title":"FinCEN Alert on Nationwide Surge in Mail Theft-Related Check Fraud Schemes Targeting the U.S. Mail (FIN-2023-Alert003)","publisher":"Financial Crimes Enforcement Network (FinCEN), US Department of the Treasury","source_type":"govt_report","statistic":"USPIS received 299,020 mail theft complaints from March 2020 through February 2021, a 161 percent increase over the same period a year earlier; check fraud SARs filed with FinCEN rose from over 350,000 in 2021 to over 680,000 in 2022 -- nearly double in one year","excerpt":"\"From March 2020 through February 2021, the USPIS received 299,020 mail theft complaints, which was an increase of 161 percent compared with the same period a year earlier. BSA reporting for check fraud has also increased in the past three years. In 2021, financial institutions filed more than 350,000 SARs to FinCEN to report potential check fraud, a 23 percent increase over the number of check fraud-related SARs filed in 2020. This upward trend continued into 2022, when the number of SARs related to check fraud reached over 680,000, nearly double the previous year's amount of filings.\"\n","source_date":"2023-02-27","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260703001737/https://www.fincen.gov/system/files/shared/FinCEN%20Alert%20Mail%20Theft-Related%20Check%20Fraud%20FINAL%20508.pdf","calculation_notes":"The 299,020 USPIS mail theft complaint figure (March 2020 through February 2021) used for the primary victimization estimate is stated verbatim in this alert document (FIN-2023-Alert003, page 1). SAR counts, by contrast, are filed by financial institutions, not by victims: a single fraud incident may generate multiple SARs, so the 680,000 SARs represent flagged suspicious transactions, not 680,000 distinct victims. The near-doubling year-over-year is used only for trend context.\n","independence_note":"FinCEN is the US Treasury financial intelligence unit. Its SAR data is collected from thousands of financial institutions independently of USPIS law enforcement complaints. The two datasets measure the same phenomenon from different vantage points (financial transaction monitoring vs victim reports), providing independent corroboration.\n"},{"url":"https://www.moneylaunderingnews.com/2023/03/fincen-and-usps-issue-alert-on-mail-theft-check-fraud-and-sar-filing-instructions/","title":"FinCEN and USPS Issue Alert on Mail-Theft Check Fraud and SAR Filing Instructions","publisher":"Money Laundering Watch (reporting on FinCEN/USPIS joint alert)","source_type":"reputable_reference","statistic":"FinCEN and USPS jointly issued an alert on mail theft-related check fraud; criminals 'wash' stolen checks with chemicals to alter the payee and amount; check fraud SARs increased 23% in 2021 and nearly doubled in 2022, totaling 680,000 SARs","excerpt":"\"FinCEN issued a joint alert with the USPS discussing typologies, red flags, and reminders about SAR filings. Typically, criminals involved in these schemes 'wash' the checks (i.e., use chemicals to remove the original ink) to remove payee and check amount information and replace it with their own identities. Suspicious Activity Reports (SARs) for check fraud increased 23% in 2021 and nearly doubled in 2022, totaling 680,000 SARs for check fraud.\"\n","source_date":"2023-03-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260125010027/https://www.moneylaunderingnews.com/2023/03/fincen-and-usps-issue-alert-on-mail-theft-check-fraud-and-sar-filing-instructions/","calculation_notes":"This source corroborates the FinCEN/USPIS joint alert and the 680,000 check fraud SAR figure for 2022, and describes the check-washing mechanism. It is used for scheme typology and trend corroboration. The 299,020 USPIS mail theft complaint figure (March 2020-February 2021) used for the lifetime estimate is grounded in the primary FinCEN alert (the first source above), not in this article, which does not state that complaint count.\n","independence_note":"Money Laundering Watch independently reported on the joint FinCEN/USPIS alert, citing the FinCEN check fraud SAR data and describing the check-washing typology. It represents an independent editorial synthesis of the government alert rather than a primary government report itself.\n"},{"url":"https://www.pymnts.com/news/security-and-risk/2024/fincen-says-banks-reported-688-million-dollars-mail-theft-related-check-fraud","title":"FinCEN Says Banks Reported $688 Million in Mail Theft-Related Check Fraud","publisher":"PYMNTS.com","source_type":"reputable_reference","statistic":"FinCEN's follow-up Financial Trend Analysis, released September 9, 2024, found that 841 financial institutions filed 15,417 Bank Secrecy Act (BSA) reports describing more than $688 million in mail theft-related check fraud during the six-month period from February 2023 to August 2023.","excerpt":"\"Mail theft-related check fraud amounted to more than $688 million in a recent six-month period. [...] During the period, FinCEN received reports of that type of fraud in 15,417 Bank Secrecy Act (BSA) reports from 841 financial institutions, according to the release.\"\n","source_date":"2024-09-09","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260122113703/https://www.pymnts.com/news/security-and-risk/2024/fincen-says-banks-reported-688-million-dollars-mail-theft-related-check-fraud/","calculation_notes":"This is FinCEN's follow-up Financial Trend Analysis of the SARs filed in the six months immediately after the original February 2023 alert (the first source above), covering February-August 2023. The analysis itself was released in September 2024, over a year after that data window closed. 841 institutions filing $688 million in six months is consistent with the 680,000 SARs filed economy-wide in 2022 cited in the sources above.\n","independence_note":"PYMNTS is an independent financial-news outlet reporting on FinCEN's own released Financial Trend Analysis; distinct from the Money Laundering Watch article above, which covers only the earlier 2023 alert.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Identity theft victimization (lifetime, US adult)","lifetime_us_adult":0.6}],"personal_factor_multipliers":[{"factor":"Regularly mails paper checks (bills, gifts, rent)","multiplier":2,"notes":"Only people who mail checks are exposed to check washing fraud; exposure scales with mailing frequency"},{"factor":"Uses USPS blue collection boxes rather than post office counter","multiplier":1.5,"notes":"Collection boxes have been the primary target of organized mail theft rings using stolen arrow keys; depositing at a staffed counter eliminates this specific attack vector"},{"factor":"Lives in an urban area with high mail theft rates","multiplier":2,"notes":"Mail theft is concentrated in metro areas where USPS collection boxes are densest and organized rings operate; rural areas have lower rates"},{"factor":"Banks at institution with strong real-time check fraud monitoring","multiplier":0.5,"notes":"Banks that flag altered checks quickly can prevent fraudulent withdrawals before funds are fully disbursed, limiting harm even when a check is stolen"}],"short_label":"Mail check fraud","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The 1 in 8 lifetime figure is derived from USPIS complaint data for the 12-month window March 2020 to February 2021, which captured the early-pandemic surge near peak. The FinCEN SAR count (680,000 in 2022) is not directly comparable -- it measures financial institution reports, not individual victims. Both the USPIS complaint rate and FinCEN SARs are expected to decline from the 2022 peak as enforcement intensifies and physical check usage continues to fall. The estimate also conflates victimization (mail stolen) with successful fraud (check actually washed and cashed); not every mail theft incident results in financial loss. People who do not mail paper checks face near-zero risk from this specific fraud vector. Transition to electronic payments is the most effective personal mitigation.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A paper check with a faint smudge suggesting alteration, on a plain pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/mail-check-fraud","api_url":"https://likelier.app/api/fears/mail-check-fraud.json"},{"slug":"cancer-lifetime","question":"What are the odds of dying from cancer?","quick_answer":"About 1 in 7 adults worldwide die of cancer over a lifetime. Here intuition is close: most people guess somewhere between 1 in 5 and 1 in 10.\n","category":"cancer","no_reliable_estimate":false,"perceived":{"description":"Cancer is one of the few fears where the public’s intuition is approximately correct. Chapman’s 2023 Survey of American Fears puts \"people I love becoming seriously ill\" at the #5 spot with 50.6% afraid or very afraid, and \"people I love dying\" at #6 with 50.4% — cancer is the modal driver of both. Most adults correctly file cancer as \"one of the big ones\", and the numbers back that up. This entry covers all-cause cancer death as a single aggregate; specific sites (lung, breast, colorectal, pancreatic) will get their own entries with their own very different risk profiles.\n","rough_estimate":"Most adults intuit lifetime cancer death risk as roughly 1 in 5 to 1 in 10 — which is close to right","kind":"survey","survey_source":{"title":"The Top 10 Fears in America 2023 (Chapman Survey of American Fears, Wave 9)","publisher":"Chapman University / The Voice of Wilkinson","url":"https://blogs.chapman.edu/wilkinson/2023/10/20/the-top-10-fears-in-american-2023/","year":2023}},"native":{"display":"~9.7 million cancer deaths per year globally (~1 in 6 of all deaths)","numerator":1,"denominator":825,"unit":"per year","population":"global, all ages, all cancer sites combined"},"normalized":{"lifetime_us_adult":0.14,"display":"1 in ~7 lifetime (global adult)","log_value":-0.85,"assumptions":"Uses the IARC GLOBOCAN 2022 estimate of \"close to 10 million deaths from cancer in 2022\" (released 4 April 2024) as the canonical annual global cancer mortality figure, and the WHO cancer fact sheet figure of \"nearly one in six deaths\" worldwide as the cross-check. Taking ~9.7 million annual cancer deaths across a global adult population of ~5.5 billion (age 18+) gives an annual per-adult rate of ~1.76 per 1,000. Compounded naively over 60 years of remaining adult life: 1 - (1 - 0.00176)^60 ≈ 0.10. That is a floor, not a ceiling, because cancer incidence is heavily concentrated in older ages and the naive compounding treats risk as age-flat. Adjusting upward for the age distribution — most cancer deaths occur above age 60, so the hazard in the last third of adult life is several-fold higher than the average — yields a global lifetime mortality figure in the 14-17% range, consistent with (but slightly below) the American Cancer Society’s direct US estimate of ~17.2% for men and ~16% for women. The lower global figure reflects higher competing mortality from infectious disease, maternal/perinatal causes, and injury in low- and middle-income countries. Headline figure 0.14 (≈ 1 in 7) with an uncertainty band of 1 in 6 to 1 in 9 to reflect window and age-structure sensitivity. Scope is global-adult-lifetime rather than US-adult-lifetime because cancer burden varies meaningfully across development levels and the US-only number would overstate the global baseline.\n","uncertainty":{"low":0.11,"high":0.18},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","title":"New report on global cancer burden in 2022 by world region and human development level","publisher":"International Agency for Research on Cancer (IARC) / World Health Organization","source_type":"govt_report","statistic":"Almost 20 million new cases of cancer and close to 10 million deaths from cancer in 2022 globally (GLOBOCAN 2022)","excerpt":"\"There were almost 20 million new cases of cancer and close to 10 million deaths from cancer in 2022.\"\n","source_date":"2024-04-04","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260323181222/https://www.iarc.who.int/news-events/new-report-on-global-cancer-burden-in-2022-by-world-region-and-human-development-level/","calculation_notes":"GLOBOCAN 2022 reports ~9.7 million global cancer deaths. Across ~5.5 billion adults (age 18+), that is ~1.76 per 1,000 adults per year. Naive 60-year compounding gives ~10%, but age-weighting (most cancer deaths occur above age 60) pulls the realistic lifetime figure into the 14-17% range. Rounded to 0.14 (≈ 1 in 7) as the global mid-point, bracketed by 1 in 9 on the optimistic side and 1 in 6 on the pessimistic side where it meets the direct US estimate.\n","independence_note":"IARC GLOBOCAN is the upstream dataset that WHO, ACS international comparisons, and the IHME Global Burden of Disease cancer module all draw from. Treat the IARC figure and the WHO cancer fact sheet below as partially dependent: they agree to one significant figure precisely because WHO republishes IARC headline numbers.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/cancer","title":"Cancer — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Cancer accounted for nearly 10 million deaths in 2020, or nearly one in six deaths","excerpt":"\"Cancer is a leading cause of death worldwide, accounting for nearly 10 million deaths in 2020, or nearly one in six deaths.\"\n","source_date":"2025-02-03","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164348/https://www.who.int/news-room/fact-sheets/detail/cancer","calculation_notes":"WHO’s \"nearly one in six deaths\" framing is used directly as the share-of-all-deaths cross-check. If total global deaths are ~60 million/year and cancer is 1/6 of those, that implies ~10 million — aligned with the IARC GLOBOCAN 2022 headline. Used as the plain-English framing in the long-form body text.\n","independence_note":"WHO cancer fact sheet republishes IARC/GLOBOCAN headline numbers; same upstream data pipeline as the first source. Included as the authoritative top-line institutional citation rather than as an independent verification.\n"},{"url":"https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","title":"Lifetime Probability of Developing or Dying From Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"Lifetime probability of dying from cancer in the US: 17.2% for men (≈ 1 in 6), 16% for women (≈ 1 in 6)","excerpt":"\"Men: Developing any cancer 39.9% (1 in 3); Dying from cancer 17.2% (1 in 6). Women: Developing any cancer 39.0% (1 in 3); Dying from cancer 16% (1 in 6).\"\n","source_date":"2025-01-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413164427/https://www.cancer.org/cancer/risk-prevention/understanding-cancer-risk/lifetime-probability-of-developing-or-dying-from-cancer.html","calculation_notes":"ACS uses SEER incidence data (2018, 2019, 2021) and SEER mortality data (2020-2022) to compute direct lifetime probabilities from a life-table conditional on birth. ~17% US lifetime cancer death probability is the methodological gold standard for \"direct\" lifetime risk; it anchors the top of the Likelier uncertainty band. The global figure sits below the US figure because competing mortality in LMICs removes adults from the denominator before they reach peak cancer-risk age, not because cancer is \"safer\" globally.\n","independence_note":"SEER (NCI) and IARC (WHO) are independent compilation pipelines — SEER is US-only vital registration and population-based cancer registries, IARC aggregates national registry data worldwide. Comparing the two anchors the global-vs-US gap.\n"},{"url":"https://ourworldindata.org/cancer","title":"Cancer","publisher":"Our World in Data (Dattani, Samborska, Ritchie, Roser)","source_type":"reputable_reference","statistic":"Around 10 million people died from cancer in 2021; around 15% of all deaths were cancer deaths","excerpt":"\"Around 10 million people died from cancer in 2021. [...] around 15% of all deaths were cancer deaths.\"\n","source_date":"2024-10-07","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260411161013/https://ourworldindata.org/cancer","calculation_notes":"OWID cites IHME Global Burden of Disease 2021 for the 10-million figure and the 15% share. That 15% (vs WHO’s \"nearly one in six\" ≈ 16.7%) is the tight cross-check on the headline order of magnitude. The age-standardized cancer death rate has fallen roughly one-third since 1990 in high-income countries, so the headline lifetime number is slowly drifting downward even as the absolute count rises with population aging.\n","independence_note":"OWID derives from IHME GBD, which is methodologically independent of IARC GLOBOCAN (IHME builds its own cause-of-death model from vital registration, verbal autopsy, and survey inputs). The two pipelines agree on ~10 million annual cancer deaths to one significant figure, which is the strongest cross-check available.\n"}],"comparison_anchors":[{"label":"Death by stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Homicide (lifetime, US)","lifetime_us_adult":0.00348},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average","probability":0.14,"notes":"~9.7M cancer deaths/yr across ~8B people (IARC GLOBOCAN 2022)"},{"region":"United States (men)","probability":0.172,"notes":"ACS direct lifetime estimate from SEER data; ~1 in 6"},{"region":"United States (women)","probability":0.16,"notes":"ACS direct lifetime estimate from SEER data; ~1 in 6"},{"region":"Sub-Saharan Africa","probability":0.09,"notes":"Lower headline number because competing mortality (infectious disease, maternal, injury) removes adults from the denominator before they reach peak cancer-risk age; age-standardized cancer death rate is not meaningfully lower"}],"personal_factor_multipliers":[{"factor":"Current heavy smoker","multiplier":2.5,"notes":"Dominant modifiable risk factor; lung cancer mortality alone runs 15-20× higher for heavy smokers vs never-smokers, and smoking also raises bladder, oesophageal, pancreatic, and head-and-neck cancer risk. Applied to all-cause cancer death the aggregate multiplier is roughly 2-3×."},{"factor":"Never-smoker, normal BMI, active, low alcohol","multiplier":0.55,"notes":"The modifiable-factor-free baseline; removes roughly 40-50% of the population-attributable cancer mortality in high-income countries."},{"factor":"Age 70+ vs age 40 baseline","multiplier":4,"notes":"Cancer mortality is overwhelmingly age-driven. A 70-year-old’s annual cancer death hazard is roughly 4× a 40-year-old’s, and the gap widens further into the 80s."},{"factor":"First-degree family history of a common cancer","multiplier":1.8,"notes":"Rough aggregate across site-specific estimates; varies widely by cancer type (stronger for breast, colorectal, prostate; weaker for lung)."},{"factor":"BRCA1/BRCA2 pathogenic variant (women)","multiplier":3,"notes":"Applies specifically to breast/ovarian cancer mortality; all-cause cancer death multiplier is smaller because the variant doesn’t change non-breast/ovarian sites."}],"short_label":"Cancer (any)","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"\"Cancer\" is roughly 200 distinct diseases grouped by a shared biological mechanism and nothing else. Five-year survival ranges from under 15% (pancreatic, hepatic, oesophageal in most populations) to above 95% (localised prostate, thyroid, testicular), so the aggregate number is a scale marker, not a prognosis. This entry is mortality, not incidence: lifetime *incidence* of any cancer is roughly 1 in 2 in the US per ACS, and roughly 1 in 5 globally per GLOBOCAN cumulative-risk-0-74 figures. The mortality figure is smaller because cancer survivorship has improved substantially — in high-income countries the age-standardised cancer death rate has fallen by about one-third since 1990 even as absolute counts rise with population ageing. Specific-site cancers (lung, breast, colorectal, prostate, pancreatic) get their own Likelier entries with their own very different distributions.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale circular shape partially intersected by a thin line on a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/cancer-lifetime","api_url":"https://likelier.app/api/fears/cancer-lifetime.json"},{"slug":"immigrant-care-avoidance","question":"How likely is an immigrant adult in the US to avoid needed medical care because of immigration-enforcement fear?","quick_answer":"About 14 in 100 US immigrant adults say they or a family member avoided needed medical care over a recent nine-month window because of immigration-enforcement fear. Worry runs well ahead of that: around 51% said they were concerned health officials would share their information with ICE.\n","category":"health","tags":["relationships","kids"],"no_reliable_estimate":false,"perceived":{"description":"The driver here is the fear itself, so the relevant perception is how worried immigrants are that the health system will hand them to immigration enforcement. In the KFF/NYT 2025 survey, about half of immigrant adults (51%), and roughly eight in ten of those likely undocumented (78%), said they were somewhat or very concerned that health officials would share their information with immigration enforcement. The worry runs well ahead of the reported behavior: far more immigrants say they are concerned than say they have actually avoided care.","rough_estimate":"~51% of immigrant adults (78% of likely undocumented) concerned health officials share data with ICE","kind":"survey","survey_source":{"title":"KFF/New York Times 2025 Survey of Immigrants: Worries and Experiences Amid Increased Immigration Enforcement","publisher":"KFF (Kaiser Family Foundation) / The New York Times","url":"https://www.kff.org/racial-equity-and-health-policy/kff-new-york-times-2025-survey-of-immigrants-worries-and-experiences-amid-increased-immigration-enforcement/","year":2025}},"native":{"display":"About 14 in 100 immigrant adults","numerator":14,"denominator":100,"unit":"immigrant adults reporting they or a family member avoided seeking medical care since Jan 2025","population":"Nationally representative US immigrant adults (KFF/NYT survey, Aug–Oct 2025), self-or-family report"},"normalized":{"lifetime_us_adult":0.14,"display":"≈14%","log_value":-0.854,"assumptions":"This is NOT a lifetime probability. It is a self-reported behavioral prevalence measured over a single ~9-month window (immigration concerns 'since January 2025', captured by a survey fielded Aug 28–Oct 20, 2025) and contingent on the post-rescission enforcement regime that began Jan 20, 2025. The figure is the share of immigrant adults who said they OR a family member had avoided seeking medical care for immigration-related reasons in that period. It is placed on the lifetime axis (scope: subgroup_lifetime; subgroup = US immigrant adults) purely for comparability — it is not annualized, not projected to a lifetime, and would change with enforcement policy. native 14/100 = 0.14. log10(0.14) = -0.854. Uncertainty band 0.11–0.17 is the full-sample sampling margin of error of ±3 percentage points around 14% (KFF methodology); it reflects sampling error only, not the much larger regime-dependent and temporal variability.","uncertainty":{"low":0.11,"high":0.17},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.kff.org/racial-equity-and-health-policy/kff-new-york-times-2025-survey-of-immigrants-worries-and-experiences-amid-increased-immigration-enforcement/","title":"KFF/New York Times 2025 Survey of Immigrants: Worries and Experiences Amid Increased Immigration Enforcement","publisher":"KFF (Kaiser Family Foundation)","source_type":"primary_study","statistic":"14% of immigrant adults overall and 48% of likely undocumented immigrants report they or a family member avoided seeking medical care since Jan 2025 due to immigration concerns; 51% (78% of likely undocumented) concerned health officials share data with ICE. N=1,805, MOE ±3pp full sample, ±10pp likely-undocumented (unweighted N=151).","excerpt":"Nearly half (48%) of likely undocumented immigrants and 14% of immigrant adults overall say they or a family member have avoided seeking medical care since January 2025 due to immigration-related concerns.","source_date":"2025-11-13","source_accessed":"2026-06-13","calculation_notes":"Headline native = 14 per 100 immigrant adults (full sample). normalized.lifetime_us_adult = 0.14, log10 = -0.854. Uncertainty 0.11–0.17 = 14% ± 3pp full-sample MOE (per KFF methodology: 'The margin of sampling error including the design effect for the full sample is plus or minus 3 percentage points'). Multiplier 48/14 = 3.43 for likely-undocumented subgroup (subgroup MOE ±10pp, unweighted N=151). Perceived = 51% overall / 78% undocumented concerned about data sharing. Survey fielded Aug 28–Oct 20, 2025.","independence_note":"KFF publishes this survey across multiple report pages (worries/enforcement page, health-experiences page, topline PDF); these are ONE survey from ONE publisher (KFF/NYT), not independent corroboration. Counts as a single authoritative source."},{"url":"https://www.dhs.gov/news/2025/01/21/statement-dhs-spokesperson-directives-expanding-law-enforcement-and-ending-abuse","title":"Statement from a DHS Spokesperson on Directives Expanding Law Enforcement and Ending the Abuse of Humanitarian Parole","publisher":"U.S. Department of Homeland Security","source_type":"govt_report","statistic":"On Jan 20, 2025, Acting DHS Secretary Benjamine Huffman rescinded the 2021 guidelines restricting ICE/CBP enforcement in 'sensitive'/protected areas (schools, hospitals, churches), the policy change that frames the survey window.","excerpt":"The first directive rescinds the Biden Administration's guidelines for Immigration and Customs Enforcement (ICE) and Customs and Border Protection (CBP) enforcement actions that thwart law enforcement in or near so-called \"sensitive\" areas.","source_date":"2025-01-21","source_accessed":"2026-06-13","calculation_notes":"Establishes the policy regime ('since January 2025') the KFF survey measures against. No probability figure drawn from this source; it documents the enforcement-policy context that makes the avoidance prevalence policy-contingent.","independence_note":"Independent of KFF: this is the primary government action, not a commentary on the survey."},{"url":"https://www.northcarolinahealthnews.org/2025/11/20/medical-appointments-drop-during-immigration-enforcement/","title":"Providers: Patients skip care during NC immigration crackdown","publisher":"North Carolina Health News","source_type":"news_article","statistic":"Mecklenburg County (NC) Health Department reported a ~50% drop in WIC nutrition visits and sharp rises in missed/cancelled appointments at adult sexual-health and TB clinics in 2025, corroborating clinic-level avoidance.","excerpt":"50 percent drop in the number of women coming in for nutrition and breastfeeding support through the Women, Infants and Children program","source_date":"2025-11-20","source_accessed":"2026-06-13","calculation_notes":"Provides clinic-level corroboration that survey-reported avoidance shows up as real no-shows/drops; not used for the headline figure (clinic counts are not population-denominated and not comparable to the survey prevalence)."}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Home burglary (US)","lifetime_us_adult":0.39},{"label":"Hospital-acquired infection","lifetime_us_adult":0.039}],"personal_factor_multipliers":[{"factor":"Likely undocumented status","multiplier":3.43,"notes":"Among likely-undocumented immigrant adults, 48% reported avoiding medical care vs 14% of all immigrant adults — about 3.4x the overall rate (relative to all immigrant adults, not to documented immigrants; the 48% subgroup sits inside the 14% baseline population). Source: KFF/NYT 2025 Survey of Immigrants. Subgroup unweighted N=151, MOE ±10pp, so this multiplier is directional."}],"short_label":"Skipping care over ICE fear","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline 14% is not a lifetime risk: it is a self-reported, fear-driven behavioral prevalence over a roughly nine-month window (immigration concerns since January 2025, captured by a survey fielded August–October 2025) and is contingent on the post-rescission enforcement regime that began January 20, 2025. It is placed on the lifetime comparison axis only via scope:subgroup_lifetime, which is this entry's weakest dimension (D5). The number is self-or-family report, so it counts households where any member avoided care, not only the respondent. The uncertainty band (0.11–0.17) is the survey's ±3pp full-sample sampling error and does not capture the larger temporal and policy-dependent variability. The likely-undocumented multiplier rests on a small subgroup (unweighted N=151, MOE ±10pp). 8D self-score average is 4.0, all dimensions ≥3, so the entry clears threshold; D5 (scope) is a defensible 3 given the period-to-lifetime mapping.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":3,"d6":4,"d7":4,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Skipping care over ICE fear\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/immigrant-care-avoidance","api_url":"https://likelier.app/api/fears/immigrant-care-avoidance.json"},{"slug":"child-sexual-abuse-risk","question":"What are the odds a child will experience sexual abuse?","quick_answer":"About 1 in 7 US children experience sexual abuse before age 18 — roughly 1 in 4 girls and 1 in 20 boys. The perpetrators are overwhelmingly known, trusted adults rather than the strangers most safety messaging emphasizes.\n","category":"crime","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Public perception of child sexual abuse risk is shaped by two contradictory forces. \"Stranger danger\" campaigns from the 1980s onward anchored parental anxiety on a relatively rare scenario — abduction and abuse by an unknown person — while simultaneously obscuring the far more common reality that the vast majority of perpetrators are known and trusted adults. Most parents assess their own child's risk as low because they mentally screen for the stranger archetype and find it absent from their daily environment. The actual epidemiology — concentrated among family members, coaches, clergy, and other authority figures — is widely known in the abstract but rarely internalized as a personal probability.\n","kind":"intuition"},"native":{"display":"~25% of girls and ~5% of boys experience sexual abuse before age 18 (CDC)","numerator":148,"denominator":1000,"unit":"lifetime prevalence before age 18 (combined-sex weighted average)","population":"US children, CDC retrospective adult surveys and child maltreatment data"},"normalized":{"lifetime_us_adult":0.148,"display":"~1 in 7 US children before age 18","log_value":-0.83,"assumptions":"The CDC reports that at least 1 in 4 girls and 1 in 20 boys experience child sexual abuse in the United States. Using approximate midpoint estimates: girls 25%, boys 5%. Combined-sex weighted average (assuming ~49% female, ~51% male birth ratio in the relevant child population): 0.49 × 0.25 + 0.51 × 0.05 ≈ 0.148. The lifetime_us_adult field carries the childhood prevalence figure for schema compatibility; this is a pre-18 measure, not an adult-lifetime annual-rate extrapolation. A 2009 meta-analysis of 65 studies across 22 countries (cited by Barth et al. 2013) found global prevalence of 19.7% for girls and 7.9% for boys, broadly consistent with US figures. Uncertainty band: low end uses the narrowest contact-only definitions in conservative studies (~8%; e.g., Pereda et al. 2009 global male-only floor combined with narrow US female estimates); high end reflects broader definitions and underreporting adjustments (~25%).\n","uncertainty":{"low":0.08,"high":0.25},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/child-abuse-neglect/about/about-child-sexual-abuse.html","title":"About Child Sexual Abuse","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"At least 1 in 4 girls and 1 in 20 boys experience child sexual abuse","excerpt":"\"At least one in four girls and one in 20 boys in the United States experience child sexual abuse.\"\n","source_date":"2024-06-12","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426194657/https://www.cdc.gov/child-abuse-neglect/about/about-child-sexual-abuse.html","calculation_notes":"Primary CDC estimate. Girls: 1/4 = 25%. Boys: 1/20 = 5%. Combined-sex weighted average: 0.49 × 0.25 + 0.51 × 0.05 ≈ 0.148. The \"at least\" qualifier indicates a floor estimate.\n"},{"url":"https://publichealth.jhu.edu/sites/default/files/2023-07/fy-2019-cdc-report-to-congress-child-sexual-abuse-prevention.pdf","title":"Report to Congress on Child Sexual Abuse Prevention","publisher":"Centers for Disease Control and Prevention (CDC) / Johns Hopkins Bloomberg School of Public Health","source_type":"govt_report","statistic":"Approximately 3.7 million children are exposed to child sexual abuse each year in the US","excerpt":"\"Approximately 3.7 million children are exposed to some form of child sexual abuse each year. The majority of cases go unreported; only about 38% of child victims disclose their abuse.\"\n","source_date":"2019-07-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260430075820/https://publichealth.jhu.edu/sites/default/files/2023-07/fy-2019-cdc-report-to-congress-child-sexual-abuse-prevention.pdf","calculation_notes":"Annual incidence figure from CDC's Congressional report. 3.7 million / ~73 million US children = ~5.1% annual incidence. Over an 18-year childhood: 1 − (1 − 0.051)^18 ≈ 0.61 if events were independent, but the survey-based lifetime prevalence of ~16-25% indicates substantial overlap (same children victimized repeatedly) and definitional differences between annual incidence and lifetime prevalence measures. The 38% disclosure rate underscores the underreporting problem.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4311357/","title":"An Epidemiological Overview of Child Sexual Abuse","publisher":"Journal of Family Medicine and Primary Care","source_type":"peer_reviewed","statistic":"Meta-analysis of 65 studies in 22 countries: 19.7% of females and 7.9% of males experienced CSA before age 18","excerpt":"\"A meta-analysis conducted in the year 2009 analyzed 65 studies in 22 countries and estimated an overall international figure. An estimated 7.9% of males and 19.7% of females universally faced sexual abuse before the age of 18 years.\"\n","source_date":"2015-01-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426194750/https://pmc.ncbi.nlm.nih.gov/articles/PMC4311357/","calculation_notes":"Global meta-analytic estimate (Pereda et al. 2009, cited by Barth et al. 2013) corroborating the US-specific CDC figures. The global female prevalence of 19.7% is close to the CDC's 25% US figure; the male prevalence of 7.9% sits above the CDC's 5% (1 in 20), likely reflecting broader definitions in the international literature. Both streams confirm that official records severely undercount actual prevalence.\n"},{"url":"https://www.missingkids.org/theissues/onlineenticement","title":"Online Enticement","publisher":"National Center for Missing & Exploited Children (NCMEC)","source_type":"reputable_reference","statistic":"CyberTipline online-enticement reports rose from 186,800 in 2023 to more than 456,000 through early October 2024; 'pretending to be younger' is a named grooming tactic occurring across gaming platforms","excerpt":"\"In 2023, the CyberTipline received more than 186,800 reports of online enticement. In 2024, through October 5, NCMEC has received more than 456,000 reports of online enticement. [...] This type of victimization takes place across every platform; social media, messaging apps, gaming platforms, etc.\"\n","source_date":"2024-10-05","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260528155507/https://www.missingkids.org/theissues/onlineenticement","calculation_notes":"Report-volume signal, not a per-child population base rate. NCMEC CyberTipline counts reflect reports received (and were boosted in 2024 by the REPORT Act's expanded mandatory-reporting categories), so they cannot be converted to a lifetime probability for an individual child and do not alter the headline 1-in-7 figure. Cited to document the online/gaming-platform grooming vector — predators posing as peers and escalating off-platform — which the headline retrospective-survey prevalence does not isolate.\n"}],"comparison_anchors":[{"label":"Sexual assault (lifetime, US adult, contact)","lifetime_us_adult":0.34},{"label":"Intimate-partner violence (lifetime, global women)","lifetime_us_adult":0.27},{"label":"Stalking (lifetime, US adult)","lifetime_us_adult":0.108}],"short_label":"Child sexual abuse","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The CDC's \"1 in 4 girls and 1 in 20 boys\" figures are derived primarily from retrospective adult surveys (NISVS, ACE study) in which adults report childhood experiences. These capture more than official child-protection reports, which record only cases that come to agency attention — estimated at a fraction of actual prevalence. The 90% known-perpetrator figure means that \"stranger danger\" frameworks miss the vast majority of risk; the most common perpetrators are family members, family friends, babysitters, coaches, and other authority figures with legitimate access to children. Prevalence varies by demographic factors: children with disabilities face 2-3x higher rates, and certain racial/ethnic groups show elevated risk in US data, though these differences may partly reflect differential disclosure and reporting patterns. The long-term health consequences are well-documented — CSA is associated with elevated rates of depression, PTSD, substance abuse, and revictimization in adulthood — making the childhood prevalence figure epidemiologically significant far beyond the immediate harm. A distinct online vector has drawn intense legal scrutiny since 2025: predators posing as peers inside games such as Roblox before escalating contact to off-platform messaging apps like Snapchat and Discord. NCMEC's CyberTipline logged more than 456,000 online-enticement reports through early October 2024, up from 186,800 in 2023, with \"pretending to be younger\" a named grooming tactic that occurs across gaming platforms. Texas became the third state (after Kentucky and Louisiana) to sue Roblox over child-safety failures in November 2025, and a federal multidistrict litigation (MDL 3166) consolidating private child-exploitation claims was created in December 2025, numbering roughly 148 cases by mid-2026. These are report-volume and litigation signals, not a per-child population base rate: rising report counts partly reflect the 2024 REPORT Act's expanded mandatory-reporting categories and cannot be converted into an individual child's probability, so they do not change the headline 1-in-7 estimate.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A small paper boat on still water with a fading ripple, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/child-sexual-abuse-risk","api_url":"https://likelier.app/api/fears/child-sexual-abuse-risk.json"},{"slug":"acquired-vision-loss","question":"What are the odds of losing significant vision in a lifetime?","quick_answer":"About 1 in 7 for a global adult who reaches old age, and closer to 1 in 10 in the US. This is one people underestimate, not overestimate — most under-60 readers guess their personal risk at well under 1 in 20.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Vision loss is broadly understood as an aging risk — most adults know that cataracts, macular degeneration, and glaucoma exist, and that \"your eyes get worse\" is a thing that happens. The specific personal odds, though, almost never get quoted. Ask a typical under-60 reader to guess their lifetime probability of ending up with significant vision impairment and the median answer clusters well under 1 in 20, when the realistic US figure is roughly 1 in 10 and the global figure for anyone who reaches old age is closer to 1 in 7. The fear is culturally present but numerically vague.\n","rough_estimate":"Most adults estimate their personal lifetime risk at well under 1 in 20","kind":"intuition"},"native":{"display":"~2.2 billion people globally have a near or distance vision impairment; ~33.6 million blind (adults 50+)","numerator":1,"denominator":7,"unit":"of global adults who reach old age","population":"global adults"},"normalized":{"lifetime_us_adult":0.15,"display":"1 in ~7 lifetime (global adult)","log_value":-0.82,"assumptions":"Uses the WHO Blindness and Vision Impairment fact sheet headline of ~2.2 billion people globally with near or distance vision impairment as the broad prevalence anchor, and the Vision Loss Expert Group (VLEG) / GBD 2019 Lancet Global Health paper (Steinmetz et al. 2021) for the narrower category of adults 50+ with blindness or moderate-to-severe distance vision impairment: 33.6 million blind and 206 million with MSVI in 2020 among adults 50+. Two complementary routes to a lifetime figure: (a) Direct prevalence: 2.2 billion / ~6 billion adults ≈ 37% point prevalence of any vision impairment, including uncorrected refractive error and presbyopia. Restricting to \"significant\" impairment (moderate-to-severe distance VI or blindness) gives ~240 million / ~2 billion adults 50+ ≈ 12% point prevalence in that age band, which is the direct anchor for the ~15% global lifetime figure once cumulative incidence across remaining lifespan is added. (b) Age-stratified: US VEHSS / CDC vision-health data indicate ~12 million US adults have some vision impairment and ~1 million are blind, and the age gradient is steep — under-65 rates are low, but among adults 80+, roughly 1 in 4 to 1 in 3 has significant vision loss. Naive compounding of age-specific hazards across an adult lifespan yields ~10-12% for US adults and ~15% globally. Headline 0.15 (≈ 1 in 7) with a wide uncertainty band of 0.10 to 0.22 to span the gap between the narrow \"US adult lifetime\" figure (~10-12%) and the broader \"global adult who lives to old age\" figure (~20-25%). Scope is global_adult_lifetime because cataract- driven vision loss is overwhelmingly concentrated in LMIC populations without surgical access, which pulls the global number meaningfully above the US-only figure.\n","uncertainty":{"low":0.1,"high":0.22},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/blindness-and-visual-impairment","title":"Blindness and vision impairment — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"At least 2.2 billion people globally have a near or distance vision impairment; in at least 1 billion cases vision loss could have been prevented or is yet to be addressed; leading causes are refractive errors and cataracts; 1 in 2 people globally who need cataract surgery do not have access","excerpt":"\"Globally, at least 2.2 billion people have a near or distance vision impairment. In at least 1 billion of these, vision impairment could have been prevented or is yet to be addressed. [...] The leading causes of vision impairment and blindness are refractive errors and cataracts. [...] 1 in 2 people globally who need cataract surgery don't have access to that surgery.\"\n","source_date":"2023-08-10","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260413162956/https://www.who.int/news-room/fact-sheets/detail/blindness-and-visual-impairment","calculation_notes":"WHO's 2.2 billion / ~6 billion global adult population ≈ 37% point prevalence of *any* vision impairment, which is too broad for the \"significant\" framing (it includes presbyopia and uncorrected refractive error). The 1 billion preventable- or-unaddressed subset is the policy-relevant figure and the anchor for the ~90% avoidable claim in the long-form body. The single most load-bearing sentence is the \"1 in 2 people globally who need cataract surgery don't have access\" line, which establishes that the global number is partly a measure of healthcare access, not disease biology.\n","independence_note":"WHO draws on upstream VLEG / GBD data for its fact-sheet headlines; treat as partially dependent with the Steinmetz et al. 2021 source below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7820391/","title":"Causes of blindness and vision impairment in 2020 and trends over 30 years, and prevalence of avoidable blindness in relation to VISION 2020: the Right to Sight: an analysis for the Global Burden of Disease Study","publisher":"Lancet Global Health — GBD 2019 Blindness and Vision Impairment Collaborators (Steinmetz et al.)","source_type":"peer_reviewed","statistic":"33.6 million cases of global blindness and 206 million cases of moderate-to-severe vision impairment (MSVI) in adults aged 50+ in 2020; leading causes globally (blindness, adults 50+) were cataract (15.2M), glaucoma (3.6M), undercorrected refractive error (2.3M), age-related macular degeneration (1.8M), and diabetic retinopathy (0.86M)","excerpt":"\"Global crude prevalence of 33·6 million cases of global blindness [in] adults aged 50 years and older in 2020. [...] 206 million aged 50 years and older adults with MSVI in 2020. [...] cataract (15·2 million cases), followed by glaucoma (3·6 million cases), undercorrected refractive error (2·3 million cases), age-related macular degeneration (1·8 million cases), diabetic retinopathy (0·86 million cases).\"\n","source_date":"2020-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163035/https://pmc.ncbi.nlm.nih.gov/articles/PMC7820391/","calculation_notes":"Among adults aged 50+ (~2 billion globally), ~240 million have MSVI or blindness — a ~12% point prevalence in that age band. Since the 50+ hazard compounds across the remaining 30+ years of typical lifespan and shifts heavily upward above age 75, the cumulative lifetime incidence from mid-life onward is meaningfully higher than the cross-sectional point prevalence. This is the primary anchor for the 0.15 headline — it is also the source of the \"cataract is the single largest cause of global blindness\" claim that drives the treatability discussion in the body. Age-standardized rates of avoidable MSVI did not improve meaningfully over 2010-2019, so the WHO VISION 2020 target was missed; the absolute numbers grew because populations aged.\n","independence_note":"This is the VLEG / GBD 2019 paper that the WHO fact sheet above ultimately cites; the two sources share an upstream and should be treated as partially dependent. The per-cause breakdown here is the methodologically primary statement; WHO is the policy summary.\n"},{"url":"https://www.cdc.gov/vision-health/about-eye-disorders/index.html","title":"About Common Eye Disorders and Diseases","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"~20.5 million Americans 40+ have cataract in one or both eyes; ~1.8 million Americans 40+ are affected by age-related macular degeneration; ~4.1 million Americans have diabetic retinopathy and 899,000 have vision-threatening retinopathy; cataract is the leading cause of vision loss in the US and diabetic retinopathy is the leading cause of blindness in working-age Americans","excerpt":"\"An estimated 20.5 million (17.2%) Americans aged 40 years and older have cataract in one or both eyes. [...] About 1.8 million Americans aged 40 years and older are affected by AMD. [...] An estimated 4.1 million Americans have retinopathy and 899,000 have vision-threatening retinopathy. [...] Diabetic retinopathy (DR) is the leading cause of blindness in American adults of working age. [...] Cataract is [...] the leading cause of vision loss in the United States.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413163110/https://www.cdc.gov/vision-health/about-eye-disorders/index.html","calculation_notes":"US-specific disease counts that collectively sum to roughly 12 million US adults with significant impairment (including uncorrected refractive error and presbyopia-adjacent conditions) and about 1 million legally blind. Used as the US anchor in regional_breakdown and to justify the ~1-in-10 US lifetime figure as a floor. The key distinction from the global number: the majority of US cataract is surgically corrected and does not progress to blindness, which is exactly what drives the gap between the US ~10% lifetime figure and the global ~15% figure.\n","independence_note":"CDC Vision Health Initiative draws on the VEHSS (Vision and Eye Health Surveillance System) and NHANES, which are methodologically independent of the WHO / VLEG upstream; this is the independent US-side cross-check the schema asks for.\n"}],"comparison_anchors":[{"label":"Death from Alzheimer's or other dementia (lifetime, global adult)","lifetime_us_adult":0.12},{"label":"Death from cancer (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Death from stroke (lifetime, global adult)","lifetime_us_adult":0.067},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Global average, any significant adult-onset vision impairment","probability":0.15,"notes":"MSVI-or-blindness cumulative lifetime incidence; anchored on VLEG/GBD 2019 age-50+ prevalence"},{"region":"US adult lifetime","probability":0.12,"notes":"Roughly 12 million US adults with vision impairment and ~1 million blind; lifetime figure pulled down by near-universal cataract surgery access"},{"region":"LMIC — South Asia / Sub-Saharan Africa","probability":0.25,"notes":"Cataract dominates; almost entirely treatable with a single outpatient surgery — the global number is partly a healthcare-access measure, not disease biology"},{"region":"US adults 80+","probability":0.3,"notes":"Vision loss is heavily concentrated in the oldest decades; roughly 1 in 3 US adults 80+ has significant impairment"}],"personal_factor_multipliers":[{"factor":"type 2 diabetes (25+ years)","multiplier":3,"notes":"Diabetic retinopathy is the leading cause of blindness in working-age American adults per CDC"},{"factor":"family history of AMD","multiplier":2,"notes":"First-degree relative with AMD roughly doubles individual risk"},{"factor":"current smoker","multiplier":2,"notes":"Smoking is the single largest modifiable risk factor for AMD and a meaningful risk factor for cataract progression"},{"factor":"regular dilated eye exams from age 50","multiplier":0.7,"notes":"Catches glaucoma and diabetic retinopathy at treatable stages; the risk reduction is about catching conditions early, not preventing their onset"}],"short_label":"Vision loss","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The single biggest interpretive issue on this entry is the \"preventable vs actually lost\" split. WHO estimates that vision loss could have been prevented or is yet to be addressed in roughly 1 billion of the 2.2 billion global cases, and the VLEG / GBD 2019 paper is clear that cataract — a condition that a single 15-minute outpatient surgery can almost entirely reverse — is the single largest cause of global blindness. The ~15% global figure therefore mixes truly lost vision (AMD, advanced glaucoma) with undertreated conditions whose cure is well-established and cheap. Treat this as two different numbers bundled into one statistic: a biological ceiling and a healthcare-access ceiling. The US figure (~10-12% lifetime) is closer to the biological ceiling because near-universal cataract surgery access removes the largest correctable cause from the denominator. The age gradient is also load-bearing: under-65 vision impairment rates are quite low, and the cumulative lifetime figure is heavily driven by what happens after 75. Anyone dying before their mid-70s from another cause never reaches peak vision-loss age, which is why the \"global average\" and \"US adults 80+\" rows in the regional breakdown differ by roughly a factor of two despite describing the same underlying hazard.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":4,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single pale circle softly out of focus against a muted grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/acquired-vision-loss","api_url":"https://likelier.app/api/fears/acquired-vision-loss.json"},{"slug":"hernia-lifting-risk","question":"What are the odds of developing a hernia from heavy lifting?","quick_answer":"The lifetime odds of developing an inguinal hernia are about 1 in 7 across both sexes (27% for men, 3% for women). Heavy lifting is widely blamed, yet most people overestimate the risk — many guess closer to 1 in 3 over a lifetime.\n","category":"health","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Heavy lifting is widely understood to cause hernias — a conviction so entrenched that gym-goers, warehouse workers, and removal crews routinely assume that one wrong repetition will rupture the abdominal wall. Most people frame hernia risk as primarily a consequence of exertion and imagine that anyone who lifts heavy loads professionally is operating on borrowed time before the inevitable protrusion.\n","rough_estimate":"about 1 in 3 heavy lifters over a lifetime","kind":"intuition"},"native":{"display":"~15 in 100 US adults (lifetime, both sexes combined)","numerator":15,"denominator":100,"unit":"lifetime","population":"US adults (both sexes combined)"},"normalized":{"lifetime_us_adult":0.15,"display":"~1 in 7 lifetime (US adult, both sexes)","log_value":-0.824,"assumptions":"Öberg et al. (2017, Frontiers in Surgery) report lifetime inguinal hernia development risk of 27% for men and 3% for women. US adults are approximately 50% male and 50% female. Sex-neutral weighted baseline: (0.27 × 0.50) + (0.03 × 0.50) = 0.15. This reflects hernia development (symptomatic cases), not surgical repair rates. Zendejas et al. (2013, Annals of Surgery) found 42.5% lifetime cumulative repair incidence for men and 5.8% for women in a population-based surgery registry — sex-neutral weighted ~24% — used as the upper bound of uncertainty because repair-based counts include bilateral and recurrent repairs that inflate the figure beyond first-event incidence.\n","uncertainty":{"low":0.1,"high":0.25},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5614933/","title":"Etiology of Inguinal Hernias: A Comprehensive Review","publisher":"Frontiers in Surgery (Öberg S, Andresen K, Rosenberg J)","source_type":"peer_reviewed","statistic":"Lifetime risk of inguinal hernia: 27% for men and 3% for women","excerpt":"\"the lifetime risk of developing an inguinal hernia is 27% for men and 3% for women\"\n","source_date":"2017-09-22","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20260308112354/https://pmc.ncbi.nlm.nih.gov/articles/PMC5614933/","calculation_notes":"Source reports 27% men and 3% women lifetime inguinal hernia development risk. US adult sex distribution approximately 50/50. Sex-neutral weighted baseline: (0.27 × 0.50) + (0.03 × 0.50) = 0.15, used as primary lifetime_us_adult. Male personal_factor_multiplier derived as 0.27 / 0.03 = 9.0 relative to women.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3761354/","title":"Incidence of Inguinal Hernia Repairs in Olmsted County, MN: A Population-Based Study","publisher":"Annals of Surgery (Zendejas B, Ramirez T, Jones T et al.)","source_type":"peer_reviewed","statistic":"Lifetime cumulative repair incidence: 42.5% in men, 5.8% in women (Olmsted County 1989–2008)","excerpt":"\"The life-long cumulative incidence of an initial, unilateral or bilateral IHR in adulthood was 42.5% in men and 5.8% in women.\"\n","source_date":"2013-03-01","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20250328142954/https://pmc.ncbi.nlm.nih.gov/articles/PMC3761354/","calculation_notes":"Population-based registry study using Rochester Epidemiology Project linkage (>97% population coverage), Olmsted County MN 1989–2008. Repair-based figures include bilateral and recurrent surgeries, inflating the count beyond first-event development. Sex-neutral weighted estimate: (0.425 × 0.50) + (0.058 × 0.50) = 0.2415, used as upper bound of uncertainty range (0.25, rounded).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7520410/","title":"Work-relatedness of inguinal hernia: a systematic review including meta-analysis and GRADE","publisher":"Hernia (Kuijer PPFM, Hondebrink D, Hulshof CTJ, Van der Molen HF)","source_type":"peer_reviewed","statistic":"Physically demanding work associated with inguinal hernia OR 2.30 (95% CI 1.56–3.40); lifting >4000 kg/workday OR 1.32 (95% CI 1.27–1.38)","excerpt":"\"physically demanding work was associated with an increased risk for IH (OR 2.30, 95% confidence interval 1.56–3.40)\"\n","source_date":"2020-01-01","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20250314090514/https://pmc.ncbi.nlm.nih.gov/articles/PMC7520410/","calculation_notes":"Systematic review of 14 occupational cohort studies; 3 included in quantitative meta-analysis (621 inguinal hernia cases in workers). OR 2.30 applies to broadly defined physically demanding work. Specific cumulative lifting threshold of >4000 kg per workday carries OR 1.32 (95% CI 1.27–1.38). Standing or walking ≥6 hours per workday: HR 1.45 (95% CI 1.12–1.88). OR 2.30 is used for the physically demanding occupation personal_factor_multiplier as the more conservative broadly applicable estimate.\n"},{"url":"https://academic.oup.com/aje/article-abstract/165/10/1154/57933","title":"Risk Factors for Inguinal Hernia among Adults in the US Population","publisher":"American Journal of Epidemiology (Ruhl CE, Everhart JE)","source_type":"peer_reviewed","statistic":"Age 60–74 HR 2.8 (95% CI 2.2–3.6) in men; obesity HR 0.51 (95% CI 0.36–0.71) in men","excerpt":"\"Inguinal hernias are common among men, especially with aging. The lower risk among heavier men was unexpected and bears further study.\"\n","source_date":"2007-05-15","source_accessed":"2026-05-15","archive_url":"http://web.archive.org/web/20240414020402/https://academic.oup.com/aje/article-abstract/165/10/1154/57933","calculation_notes":"NHANES-III longitudinal follow-up cohort of US adults. Age 60–74 vs younger adults in men: HR 2.8, used as the age-based personal_factor_multiplier. Obesity (BMI ≥30) vs normal weight in men: HR 0.51 — counter-intuitive protective effect used for the obesity multiplier (0.5). The study also reported cumulative incidence of 13.9% in men and 2.1% in women over the follow-up period (shorter than full lifetime), consistent with the Öberg 27%/3% lifetime figures.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Male sex","multiplier":9,"notes":"Lifetime inguinal hernia development risk ~27% for men vs ~3% for women (Öberg et al. 2017). Men account for roughly 90% of all inguinal hernia cases in the literature.\n"},{"factor":"Age 60–74","multiplier":2.8,"notes":"Hazard ratio 2.8 (95% CI 2.2–3.6) for men aged 60–74 vs younger adults in a US national cohort (Ruhl & Everhart 2007). Inguinal hernia incidence accelerates after age 55 as connective tissue in the inguinal floor weakens progressively.\n"},{"factor":"Physically demanding occupation","multiplier":2.3,"notes":"Meta-analysis of 14 occupational cohort studies found OR 2.30 (95% CI 1.56–3.40) for physically demanding work (Kuijer et al. 2020). A more specific threshold — lifting >4000 kg cumulative load per workday — carries OR 1.32. Construction workers, dockworkers, and warehouse staff are the best-studied high-risk groups.\n"},{"factor":"Normal or low BMI (lean body type)","multiplier":2,"notes":"Men of normal weight have roughly twice the inguinal hernia repair rate of obese men. Ruhl & Everhart (2007) found obesity associated with HR 0.51 vs normal weight, meaning lean men run about 2× the risk of obese men. The proposed mechanism is that retroperitoneal fat may reinforce the inguinal floor rather than herniate it.\n"},{"factor":"Obesity (BMI ≥30)","multiplier":0.5,"notes":"Counter-intuitive protective effect for inguinal hernia specifically: HR 0.51 (95% CI 0.36–0.71) in men (Ruhl & Everhart 2007). This effect does not extend to ventral, umbilical, or incisional hernias, where obesity increases risk substantially.\n"}],"short_label":"Hernia from lifting","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 15% population baseline is a sex-averaged figure; male adults face a lifetime risk closer to 27% while female adults face roughly 3%, making sex the single largest determinant of individual risk. The baseline covers inguinal hernia only — femoral, umbilical, and incisional hernias add several additional percent to the total hernia burden. The occupational lifting data derive largely from European male worker cohorts and may not transfer directly to all US occupational groups. Hernia can develop without any heavy lifting history, and many heavy lifters never develop one; the OR 2.30 for physically demanding work reflects elevated population-level risk, not inevitability. Recreational lifting has not been shown to carry the same risk magnitude as occupational exposure, and gym populations show no elevated surgical repair rate compared with age-matched controls in available case series.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-15","image":{"alt":"Illustration of hernia risk from heavy lifting, flat editorial vector."},"canonical_url":"https://likelier.app/hernia-lifting-risk","api_url":"https://likelier.app/api/fears/hernia-lifting-risk.json"},{"slug":"hip-fracture-elderly","question":"How likely is an older adult to suffer a hip fracture?","quick_answer":"From age 50, the lifetime odds of a hip fracture are about 1 in 7 for women and 1 in 18 for men — a near-universal late-life risk that is systematically underestimated. Up to 1 in 5 patients dies within a year of the fracture, a fact rarely part of the public narrative.\n","category":"health","tags":["elder-care","household"],"no_reliable_estimate":false,"perceived":{"description":"Most people over fifty have heard that hip fractures are serious, but the lifetime probability is systematically underestimated. Hip fracture is mentally filed alongside dramatic falls — a frail person slipping on ice — rather than as a near-universal late-life risk with a one-in-seven lifetime probability for women. That up to one in five patients dies within a year of the fracture is rarely part of the public narrative, which tends to treat hip fracture as painful but recoverable.\n","kind":"intuition"},"native":{"display":"15 in 100 women aged 50+ will fracture a hip in their remaining lifetime (Europe-29 pooled)","numerator":15,"denominator":100,"unit":"lifetime from age 50","population":"European women aged 50+ (population-weighted across 29 countries — EU27 + UK + Switzerland, Kanis et al. SCOPE 2021)"},"normalized":{"lifetime_us_adult":0.15,"display":"about 1 in 7 women; 1 in 18 men — lifetime from age 50","log_value":-0.82,"assumptions":"Rate of 15.0% for women and 5.7% for men from Kanis et al., SCOPE 2021 (Arch Osteoporos 16(1):82), population-weighted across the 29 countries of the EU27 plus the UK and Switzerland. The Europe-29 pooled estimate is used as the headline because no comparable global aggregate with equivalent methodology exists; IOF considers high-income European data broadly representative of HIC populations. Worldwide approximately 1.66 million hip fractures occurred per year as of 1990 (Cooper, Campion & Melton 1992), projected to reach ~6.26 million by 2050. Up to 20% of patients die within a year of a hip fracture (SCOPE 2021), of which about 30% of deaths are causally related to the fracture. The headline figure (0.15) reflects the female rate; male rate is 0.057. Wide regional variance: Scandinavian women ~20%, East Asian women ~7%, reflecting diet, physical activity, and genetic factors. Low bound (0.10) represents East Asian or male populations; high bound (0.20) represents Scandinavian women or those with prior fragility fractures.\n","uncertainty":{"low":0.1,"high":0.2},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8172408/","title":"SCOPE 2021: a new scorecard for osteoporosis in Europe","publisher":"Archives of Osteoporosis (Kanis et al.)","source_type":"peer_reviewed","statistic":"Average lifetime probability of hip fracture at age 50 in the EU (EU27 + UK + Switzerland): 15.0% in women, 5.7% in men; up to 20% of patients die in the first year following hip fracture","excerpt":"\"The average lifetime probability of hip fracture (weighted for population size) at the age of 50 years in the EU was 5.7% in men and 15.0% in women. Up to 20% of patients die in the first year following hip fracture, mostly as a result of serious underlying medical conditions; it is estimated that approximately 30% of deaths are causally related.\"\n","source_date":"2021-06-02","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260119000446/https://pmc.ncbi.nlm.nih.gov/articles/PMC8172408/","calculation_notes":"Kanis et al., SCOPE 2021 (Arch Osteoporos 16(1):82, PMC8172408, doi 10.1007/s11657-020-00871-9) reports the population-weighted lifetime probability of hip fracture from age 50 across the 29 countries of the EU27 plus the UK and Switzerland. Native rate: 15 in 100 women from age 50 (direct lifetime probability, verbatim from the paper). Normalized: 0.15 (no conversion needed; figure is already expressed as lifetime probability). Europe-29 pooled estimate used as proxy for global high-income-country rate. Male rate 0.057 treated as supporting figure. Mortality: the paper states up to 20% of hip-fracture patients die within one year (of which ~30% of deaths are causally related to the fracture) — this replaces an earlier \"21–30% excess mortality\" figure that was not supported by any cited source.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/1421796/","title":"Hip fractures in the elderly: a world-wide projection","publisher":"Osteoporosis International","source_type":"peer_reviewed","statistic":"Hip fractures worldwide projected to rise from 1.66 million in 1990 to 6.26 million by 2050, with the burden shifting to Asia and Latin America","excerpt":"\"The projections indicate that the number of hip fractures occurring in the world each year will rise from 1.66 million in 1990 to 6.26 million by 2050. While Europe and North America account for about half of all hip fractures among elderly people today, this proportion will fall to around one quarter in 2050, by which time steep increases will be observed throughout Asia and Latin America.\"\n","source_date":"1992-11-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260306010906/https://pubmed.ncbi.nlm.nih.gov/1421796/","calculation_notes":"Cooper, Campion & Melton 1992 (Osteoporos Int 2(6):285-9, PMID 1421796) is the original source of the ~1.66M-in-1990 to ~6.26M-by-2050 world-wide hip fracture projection cited in the assumption text. Used here for global burden context, not for the lifetime probability calculation directly. The lifetime probability (0.15) is drawn from the Kanis et al. SCOPE 2021 source above.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/12049882/","title":"Epidemiology and outcomes of osteoporotic fractures","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Hip fracture frequency is rising 1–3% per year in most regions and varies more widely from region to region than vertebral fracture prevalence","excerpt":"\"Hip fractures, the most serious outcome of osteoporosis, are becoming more frequent than before because the world's population is ageing and because the frequency of hip fractures is increasing by 1-3% per year in most areas of the world. Rates of hip fracture vary more widely from region to region than does the prevalence of vertebral fractures.\"\n","source_date":"2002-05-18","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260227111503/https://pubmed.ncbi.nlm.nih.gov/12049882/","calculation_notes":"Cummings & Melton 2002 Lancet review (Lancet 359(9319):1761-7, PMID 12049882) — foundational epidemiological reference establishing that hip fractures are the most serious osteoporotic outcome, that incidence is rising 1–3% per year, and that rates vary widely by region. Used as a historical anchor to corroborate the regional variation and rising-burden context; the headline lifetime probability (0.15) and the one-year mortality figure (up to 20% of patients) are drawn from the Kanis et al. SCOPE 2021 source above.\n"}],"comparison_anchors":[{"label":"Hip fracture (men, lifetime from 50)","lifetime_us_adult":0.057},{"label":"Breast cancer diagnosis (women, lifetime)","lifetime_us_adult":0.125},{"label":"Stroke (lifetime, global adults)","lifetime_us_adult":0.25}],"regional_breakdown":[{"region":"Northern Europe (women, 50+)","probability":0.2,"notes":"Highest rates globally — Scandinavia, UK"},{"region":"Western Europe and US (women, 50+)","probability":0.15,"notes":"IOF pooled Europe-29 estimate; US rates similar"},{"region":"East Asia (women, 50+)","probability":0.07,"notes":"Japan, China — roughly half the European rate"},{"region":"Men, all high-income populations","probability":0.057,"notes":"Roughly one-third of women's rate"}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.4,"notes":"Women 15%, combined-sex average ~11%; multiplier reflects female excess risk"},{"factor":"Male sex","multiplier":0.5,"notes":"Men 5.7% vs women 15%"},{"factor":"Scandinavian / Northern European","multiplier":1.3,"notes":"~20% lifetime for women vs 15% pooled European average"},{"factor":"East Asian","multiplier":0.5,"notes":"~7% lifetime for women in Japan and other East Asian populations"},{"factor":"Prior fragility fracture","multiplier":2,"notes":"Risk of second fracture roughly doubles after first osteoporotic fracture"},{"factor":"T-score < -2.5 (osteoporosis by DXA)","multiplier":2.5,"notes":"Severe bone density loss substantially increases hip fracture probability"}],"short_label":"Hip fracture risk","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 15% figure reflects a lifetime cumulative probability from age 50 for women in European high-income countries. It is not a per-year figure. The sex gap (women ~15%, men ~6%) is the most consistent finding in the literature, driven by lower peak bone density in women and greater survival into very old age. Hip fracture mortality is high — up to 20% of patients die within a year, though only about 30% of those deaths are causally related to the fracture — partly because the fracture itself disrupts mobility and triggers downstream complications — infection, thrombosis, deconditioning — rather than because the fracture directly kills. Fall-prevention interventions and osteoporosis screening (USPSTF, FRAX tool) can substantially modify personal risk; this is a risk amenable to clinical action.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a simple wooden cane casting a shadow on a clean floor, muted palette."},"canonical_url":"https://likelier.app/hip-fracture-elderly","api_url":"https://likelier.app/api/fears/hip-fracture-elderly.json"},{"slug":"irs-tax-audit","question":"What are the odds of being audited by the IRS?","quick_answer":"Only about 0.3% of individual returns are examined in a year (FY 2024), compounding to roughly 15% over a filing lifetime — though a median W-2 earner's true odds are far lower. Many people guess a 5-10% annual chance, overestimating their risk by about an order of magnitude.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Tax audits loom large in the American imagination. The word \"audit\" conjures images of agents rifling through shoeboxes of receipts, and taxpayer surveys consistently show that audit fear is a primary motivator for compliance. Most people overestimate their personal audit risk by at least an order of magnitude, guessing somewhere between 5% and 10% per year, when the actual rate for the median filer is closer to 0.1%.\n","rough_estimate":"~5-10% per year is a common guess","kind":"intuition"},"native":{"display":"~0.3% of individual returns examined (FY 2024)","numerator":444014,"denominator":161100000,"unit":"per year","population":"US individual income tax return filers"},"normalized":{"lifetime_us_adult":0.15,"display":"~15% lifetime (US adult filer)","log_value":-0.82,"assumptions":"The IRS 2024 Data Book reports that the IRS closed 505,514 total tax return audits in FY 2024, of which 444,014 involved individual income tax returns. With approximately 161.1 million individual returns filed, the single-year examination rate is about 0.28%. Compounding over a 59-year adult filing life: 1 - (1 - 0.0028)^59 = 0.152, or about 15%. This is likely an overestimate of any single taxpayer's cumulative risk because audit selection is not random: the IRS concentrates resources on EITC claimants, high-income filers, and returns with specific red flags. A median-income W-2 filer's true annual audit rate is closer to 0.1%, implying a lifetime probability around 6%. The 15% figure represents the population-average rate if audits were distributed uniformly across all filers, which they are not.\n","uncertainty":{"low":0.06,"high":0.25},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.irs.gov/newsroom/irs-releases-fiscal-year-2024-data-book-describing-agencys-activities","title":"IRS releases fiscal year 2024 Data Book describing agency's activities","publisher":"Internal Revenue Service","source_type":"govt_report","statistic":"505,514 total tax return audits closed in FY 2024, resulting in $29 billion in recommended additional tax; 161.1 million individual returns processed","excerpt":"\"In FY 2024, the IRS closed 505,514 tax return audits, resulting in $29 billion in recommended additional tax. The IRS processed more than 266 million returns and other forms from individuals, businesses and tax-exempt organizations.\"\n","source_date":"2025-05-22","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413173805/https://www.irs.gov/newsroom/irs-releases-fiscal-year-2024-data-book-describing-agencys-activities","calculation_notes":"The IRS Data Book is the official annual statistical summary of IRS operations. Of the 505,514 total audits closed, 444,014 were individual income tax returns (the remainder were employment, estate, gift, and excise tax returns). Dividing 444,014 by 161.1 million individual returns gives an examination rate of approximately 0.28%. This rate has been declining for over a decade due to staffing reductions. At the peak in FY 2010, the IRS examined about 1.7 million individual returns per year, roughly 1.1% of filings.\n","independence_note":"The IRS Data Book is compiled from the IRS's own administrative records of examinations completed. It is the primary upstream data source for GAO and Congressional Research Service analyses.\n"},{"url":"https://www.gao.gov/products/gao-22-104960","title":"Tax Compliance: Trends of IRS Audit Rates and Results for Individual Taxpayers by Income","publisher":"US Government Accountability Office","source_type":"govt_report","statistic":"IRS audit rates for individual taxpayers declined across all income levels from 2010 to 2019; audit rates for higher-income taxpayers decreased more steeply — the $10 million+ bracket fell from 21.2% (TY2010) to 3.9% (TY2019), an ~81% drop","excerpt":"\"For more than 15 years, IRS audit rates have been steadily declining. IRS officials said audit rates declined due to staffing decreases and because it takes more staff time and expertise to handle complex higher-income audits.\"\n","source_date":"2022-03-17","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260204012056/https://www.gao.gov/products/gao-22-104960","calculation_notes":"The GAO report provides historical context for the declining audit trend. In FY 2010, the IRS examined about 1.1% of individual returns; by FY 2019, that rate had fallen to roughly 0.4%. The continued decline to ~0.28% in FY 2024 reflects further staffing losses despite Inflation Reduction Act funding. The GAO report also documents that audit rates are not uniform: returns under $25,000 and over $500,000 are audited at higher rates than the middle, creating a U-shaped distribution by income. The decline is steepest at the very top: returns reporting $10 million or more in income were audited at 21.2% in tax year 2010, falling to 3.9% by tax year 2019 — an ~81% drop, considerably steeper than the population-wide decline over the same period.\n","independence_note":"GAO is an independent congressional agency that audits and evaluates federal programs. Its analysis draws on IRS data but applies independent methodology to examine trends and outcomes.\n"}],"comparison_anchors":[{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"income >$1 million (EITC excluded)","multiplier":10,"notes":"IRS Data Book: audit rate for $1M+ filers is ~2-3%, roughly 10x the overall rate; IRA funding is further increasing high-income audit coverage"},{"factor":"EITC claimant","multiplier":3,"notes":"IRS Data Book: EITC returns are audited at ~1% (correspondence audits), roughly 3x the overall rate, driven by congressional mandate"},{"factor":"W-2 wage earner, income $50-200k, no Schedule C","multiplier":0.2,"notes":"the lowest-audit-rate bracket; withholding-matched income leaves little to examine"},{"factor":"Schedule C self-employment income reported","multiplier":3,"notes":"IRS Data Book 2022: Schedule C filers face audit rates roughly 3x higher than wage-only filers of equivalent income, due to cash-based income and the expanded deduction landscape"},{"factor":"prior audit in the last 3 years","multiplier":2.5,"notes":"IRS enforcement priorities: taxpayers who have been audited and had adjustments made face elevated re-examination rates; prior compliance issues flag a return for closer scrutiny"}],"short_label":"IRS audit","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The population-average audit rate conceals a U-shaped distribution by income. Filers claiming the Earned Income Tax Credit (under ~$25,000 AGI) face audit rates of 0.3-0.4%, driven almost entirely by correspondence audits verifying credit eligibility. Filers in the $50,000-$500,000 range face rates as low as 0.1%. Filers above $10 million face rates of 4% or higher, though even those rates have fallen steeply since 2010. About 78% of all individual audits in FY 2024 were correspondence audits (conducted by mail), not the in-person field examinations most people picture. The lifetime figure here assumes a constant audit rate, which is unrealistic: rates have varied by a factor of four over the past 15 years alone. A filer who spends their entire career as a median-income W-2 employee has a meaningfully lower lifetime audit probability than the headline figure.\n","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"groundedness-audit-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single envelope with a faint official seal on a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/irs-tax-audit","api_url":"https://likelier.app/api/fears/irs-tax-audit.json"},{"slug":"miscarriage-recognized-pregnancy","question":"What are the odds of miscarriage after a recognized pregnancy?","quick_answer":"About 15% of recognized pregnancies end in miscarriage — roughly 1 in 7 — yet 55% of US adults believe it happens in 5% or fewer of pregnancies, a roughly threefold underestimate. The true rate is hidden more by silence than by rarity.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Miscarriage is widely misunderstood as a rare event. In a national survey of over 1,000 US adults, 55% believed miscarriage occurs in 5% or fewer of all pregnancies — roughly a threefold underestimate. Among those who had experienced a miscarriage, 47% reported feeling guilty and 41% felt they were alone, suggesting the true prevalence is hidden by silence rather than by absence.\n","rough_estimate":"~5% or less (most common public guess)","kind":"survey","survey_source":{"title":"A National Survey on Public Perceptions of Miscarriage","publisher":"Obstetrics & Gynecology (Bardos et al.)","url":"https://pubmed.ncbi.nlm.nih.gov/26000502/","year":2015}},"native":{"display":"~1 in 7 recognized pregnancies","numerator":1,"denominator":7,"unit":"per clinically recognized pregnancy","population":"women with clinically recognized pregnancies"},"normalized":{"lifetime_us_adult":0.15,"display":"~15% per recognized pregnancy","log_value":-0.82,"assumptions":"Central estimate of 15% drawn from the 11-22% range in Ammon Avalos et al. (2012) and the 10-20% figure cited by March of Dimes and ACOG. This is per recognized pregnancy, not per woman or per lifetime. Losses before clinical recognition (biochemical pregnancies) are excluded and would roughly double the total rate.\n","uncertainty":{"low":0.1,"high":0.22},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/22511535/","title":"A systematic review to calculate background miscarriage rates using life table analysis","publisher":"Birth Defects Research Part A: Clinical and Molecular Teratology","source_type":"peer_reviewed","statistic":"Cumulative risk of miscarriage weeks 5-20: 11-22 per 100 women","excerpt":"\"the cumulative risk of miscarriage for weeks 5 through 20 of gestation ranged from 11 miscarriages per 100 women to 22 miscarriages per 100 women (11-22%).\"\n","source_date":"2012-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180130/https://pubmed.ncbi.nlm.nih.gov/22511535/","calculation_notes":"Life-table analysis across multiple cohort studies. The 11-22% range reflects variation across study populations and gestational-age cutoffs. Midpoint ~15% used as native estimate. Weekly hazard peaks before week 13 at >20 per 1,000 women-weeks.\n","independence_note":"Systematic review pooling US and European cohort studies. Methodologically independent of Magnus et al.'s Norwegian registry data, though both ultimately describe clinically recognized pregnancies in high-income populations.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30894356/","title":"Role of maternal age and pregnancy history in risk of miscarriage: prospective register based study","publisher":"BMJ","source_type":"peer_reviewed","statistic":"Miscarriage risk lowest at age 25-29 (10%), rising to 53% at age 45+","excerpt":"\"The risk of miscarriage was lowest in women aged 25-29 (10%), and rose rapidly after age 30, reaching 53% in women aged 45 and over.\"\n","source_date":"2019-03-20","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413180211/https://pubmed.ncbi.nlm.nih.gov/30894356/","calculation_notes":"Norwegian registry study (421,201 pregnancies). Provides the age gradient used for personal_factor_multipliers. Baseline ~10% at 25-29 anchors the age-specific curve.\n","independence_note":"Norwegian registry data is independent of the Ammon Avalos systematic review, which drew primarily from US and European cohort studies.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26000502/","title":"A National Survey on Public Perceptions of Miscarriage","publisher":"Obstetrics & Gynecology","source_type":"peer_reviewed","statistic":"55% of respondents believed miscarriage occurs in 5% or less of pregnancies","excerpt":"\"Respondents to our survey erroneously believed that miscarriage is a rare complication of pregnancy, with the majority believing that it occurred in 5% or less\"\n","source_date":"2015-06-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260110023612/https://pubmed.ncbi.nlm.nih.gov/26000502/","calculation_notes":"Used for the perceived risk estimate. Survey of 1,084 US adults; 55% placed miscarriage at ≤5%, which is roughly 3x lower than the epidemiological consensus of 10-20%.\n","independence_note":"Public-perception survey data, entirely separate from the epidemiological sources. Used only for the perceived-risk axis; does not feed into the probability estimate.\n"},{"url":"https://www.marchofdimes.org/find-support/topics/miscarriage-loss-grief/miscarriage","title":"Miscarriage","publisher":"March of Dimes","source_type":"reputable_reference","statistic":"10 to 20 percent of known pregnancies end in miscarriage","excerpt":"\"It's estimated that between 10 to 20 in 100 known pregnancies (10 to 20 percent) end in miscarriage.\"\n","source_date":"2024-10-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260420044114/https://www.marchofdimes.org/find-support/topics/miscarriage-loss-grief/miscarriage","calculation_notes":"Patient-facing summary consistent with ACOG guidance and the Ammon Avalos systematic review. Used as corroborating reputable reference for the 10-20% headline range.\n","independence_note":"March of Dimes cites ACOG and underlying literature, so this is a dependent summary rather than an independent data source.\n"}],"comparison_anchors":[{"label":"Appendicitis (lifetime, US adult)","lifetime_us_adult":0.07},{"label":"Type 2 diabetes diagnosis (lifetime, US adult)","lifetime_us_adult":0.4}],"regional_breakdown":[{"region":"Gestational weeks 5-6","probability":0.1,"notes":"Highest weekly hazard; ~10% cumulative risk by end of week 6"},{"region":"After cardiac activity at 8 weeks","probability":0.05,"notes":"Risk drops roughly by half once fetal heartbeat is confirmed"},{"region":"After 12 weeks (second trimester)","probability":0.02,"notes":"~1-2% residual risk; often called 'late miscarriage' after this point"},{"region":"After 20 weeks","probability":0.005,"notes":"Classified as stillbirth rather than miscarriage in most jurisdictions"}],"personal_factor_multipliers":[{"factor":"Maternal age 20-24","multiplier":1.1,"notes":"Magnus et al. 2019: ~11%, marginally above the 25-29 low point (25-29 is the lowest age group)"},{"factor":"Maternal age 25-29 (baseline)","multiplier":1,"notes":"Reference group; ~10% miscarriage rate (Magnus et al. 2019)"},{"factor":"Maternal age 30-34","multiplier":1.1,"notes":"~11% risk (Magnus et al. 2019); risk rises rapidly only after this band"},{"factor":"Maternal age 35-39","multiplier":2,"notes":"~20% risk; often cited as the 'advanced maternal age' threshold"},{"factor":"Maternal age 40-44","multiplier":3.2,"notes":"~32% risk (Magnus et al. 2019)"},{"factor":"Maternal age 45+","multiplier":5.3,"notes":"~53% risk (Magnus et al. 2019)"},{"factor":"One prior miscarriage","multiplier":1.5,"notes":"Magnus et al. 2019: recurrence risk is modestly elevated"},{"factor":"Three or more prior miscarriages","multiplier":2.5,"notes":"Recurrent pregnancy loss substantially increases the per-pregnancy risk"}],"short_label":"Miscarriage","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 10-20% figure applies only to clinically recognized pregnancies — those confirmed by positive test or ultrasound. Including biochemical pregnancies (losses before clinical detection) roughly doubles the total to ~30% or more. The gestational-week breakdown uses pooled estimates and varies by detection method, maternal health, and study design. Stillbirth (loss after 20 weeks) is a separate statistic not included here.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":5,"d8":4,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-12","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single small seedling in a muted terracotta pot against a pale grey background, flat vector illustration."},"canonical_url":"https://likelier.app/miscarriage-recognized-pregnancy","api_url":"https://likelier.app/api/fears/miscarriage-recognized-pregnancy.json"},{"slug":"regular-drinking-death","question":"What are the odds of dying from alcohol-related disease as a regular drinker?","quick_answer":"A lifelong heavy drinker faces roughly a 1 in 7 chance of dying from alcohol-related disease. Most people file regular drinking as a minor vice and guess the lifetime risk well below 1 in 10.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Alcohol is the rare fear where the cultural framing and the numbers run in opposite directions. Most adults in wealthy countries file regular drinking as \"not ideal but not really dangerous\", closer to a dietary vice than a mortality lever. The aggregate arithmetic disagrees: alcohol is responsible for roughly 2.6 million deaths a year globally, about 4.7% of all deaths, and about 178,000 deaths a year in the United States — roughly 1 in 20 US deaths. The per-capita mortality contribution is on the same order of magnitude as smoking, but the public fear attached to it is not. This entry is about the lifetime attributable mortality for someone who actually drinks regularly, not a population average that blends drinkers and non-drinkers.\n","rough_estimate":"Most adults know heavy drinking is bad but guess lifetime alcohol mortality well below 1 in 10","kind":"intuition"},"native":{"display":"~2.6 million deaths per year globally (~4.7% of all deaths)","numerator":1,"denominator":21,"unit":"per year","population":"global, all ages, alcohol-attributable conditions"},"normalized":{"lifetime_us_adult":0.15,"display":"~1 in 7 lifetime (lifelong heavy drinker)","log_value":-0.82,"assumptions":"Reference subgroup: an adult who drinks regularly above the US dietary guideline thresholds — more than 14 standard drinks per week for men or more than 7 standard drinks per week for women — for 30+ years of adult life, without extended periods of abstinence. The ~15% headline is a rounded mid-point for lifetime alcohol-attributable mortality in this subgroup, bracketed between ~10% and ~20% to reflect methodological disagreement in the literature. Anchors: (1) WHO 2024 Global Status Report on Alcohol and Health attributes 2.6 million deaths per year globally to alcohol consumption (2019 data), or 4.7% of all deaths — 6.7% of male deaths and 2.4% of female deaths. (2) CDC MMWR 2024 reports an average of 178,307 US deaths per year from excessive alcohol use in 2020-2021, roughly 1 in 20 US deaths, up 29% from 137,927 in 2016-2017. (3) Across roughly 380 million US adults and given that heavy drinkers are a minority of the drinking population but account for the vast majority of alcohol-attributable mortality, compounding the age-weighted annual hazard for a lifelong heavy-drinker subgroup over 40-50 adult years produces a lifetime attributable mortality in the 10-20% range. (4) The headline figure is roughly half the smoking figure — which matches the rough intuition that alcohol-attributable mortality has a similar aggregate magnitude to smoking but a smaller per-exposure hazard ratio (heavy drinking raises all-cause mortality by roughly 1.5-2.5x vs current smoking's ~3x). The scope is declared as subgroup_lifetime because this is a per-lifelong- heavy-drinker probability, not a general-population lifetime risk, and it is not directly comparable to the population-scope lifetime figures on other Likelier pages. Moderate and light drinker rows in the regional_breakdown are lower than the headline and reflect the remaining methodological dispute about the J-curve (see sources 4 and 5).\n","uncertainty":{"low":0.1,"high":0.2},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/alcohol","title":"Alcohol — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Around 2.6 million deaths per year globally attributable to alcohol consumption (2019); 6.7% of all male deaths and 2.4% of all female deaths","excerpt":"\"Worldwide, around 2.6 million deaths were caused by alcohol consumption in 2019. [...] In 2019, alcohol use was responsible for 6.7% of all deaths among men and 2.4% of all deaths among women.\"\n","source_date":"2024-06-25","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182331/https://www.who.int/news-room/fact-sheets/detail/alcohol","calculation_notes":"WHO’s 2.6 million figure is the canonical global headline. Across a global adult population of ~6 billion, that is ~0.43 per 1,000 adults per year averaged across the full population (drinkers plus abstainers). The per-regular-drinker rate is several times higher because abstainers and light drinkers account for a large share of the denominator but a small share of the attributable mortality. WHO notes the 2.6M figure is down from the ~3.0M figure reported in earlier releases, partly due to methodology changes and partly due to declining per-capita consumption in some regions. Used as the primary global anchor.\n","independence_note":"WHO draws on the Global Information System on Alcohol and Health (GISAH) and the IHME Global Burden of Disease alcohol module. Partially overlapping with the GBD 2016 and GBD 2020 Lancet papers cited below.\n"},{"url":"https://www.cdc.gov/alcohol/facts-stats/index.html","title":"Facts About U.S. Deaths from Excessive Alcohol Use","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"About 178,000 US deaths per year from excessive alcohol use (2020-2021); shortened lives by an average of 24 years; ~4 million years of potential life lost per year","excerpt":"\"About 178,000 people die from excessive drinking each year. [...] This was a 29% increase from just a few years earlier (2016-2017), when there were an estimated 138,000 deaths per year. [...] Shortened the lives of those who died by an average of 24 years. [...] This resulted in a total of about 4 million years of potential life lost.\"\n","source_date":"2024-02-29","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182410/https://www.cdc.gov/alcohol/facts-stats/index.html","calculation_notes":"CDC’s ~178,000 US deaths per year is the domestic headline. Across ~3.3 million total US deaths per year (2021), that is ~5.4% — roughly 1 in 20 US deaths. The 24-year average life-expectancy loss per death is notable: alcohol-attributable deaths are concentrated at younger ages than tobacco-attributable deaths, which is why the aggregate years-of-life-lost figure (~4 million per year) is disproportionately large relative to the headcount. Used as the domestic anchor and as the basis for the \"~1 in 20 US deaths\" plain-English framing.\n","independence_note":"CDC derives the 178,000 figure from the ARDI (Alcohol-Related Disease Impact) application, which ultimately draws on the same 58 alcohol- attributable causes tracked in the CDC MMWR source below. Treat CDC and MMWR as partially dependent — MMWR is the primary analysis, the facts-stats page is its plain-language republication.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7308a1.htm","title":"Deaths from Excessive Alcohol Use — United States, 2016-2021","publisher":"CDC Morbidity and Mortality Weekly Report (Esser, Sherk, Liu, Naimi)","source_type":"govt_report","statistic":"Average annual deaths from excessive alcohol use rose 29.3% from 137,927 (2016-2017) to 178,307 (2020-2021); age-standardized rates rose from 38.1 to 47.6 per 100,000","excerpt":"\"Average annual number of deaths from excessive alcohol use [...] increased 29.3%, from 137,927 during 2016-2017 to 178,307 during 2020-2021. [...] age-standardized death rates increased from 38.1 per 100,000 population [...] to 47.6 during 2020-2021. [...] deaths from excessive alcohol use among males increased approximately 27%, from 94,362 per year to 119,606 [...] among females increased approximately 35%, from 43,565 per year to 58,701.\"\n","source_date":"2024-02-29","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182444/https://www.cdc.gov/mmwr/volumes/73/wr/mm7308a1.htm","calculation_notes":"The MMWR is the primary analysis behind the CDC headline. The 29% increase from 2016-2017 to 2020-2021 is large enough to be meaningful even after accounting for the COVID-era drinking-pattern shock. The male/female split (119,606 vs 58,701) drives the sex-stratified numbers in the assumptions block and the fact that men carry ~67% of the US alcohol-attributable mortality burden. The age-standardized 47.6 per 100,000 figure, applied only to the subset of US adults who drink regularly at heavy-drinker levels, anchors the per-subgroup lifetime figure used in the normalized block.\n","independence_note":"Same methodology as the CDC alcohol facts page above; the two sources are the same underlying analysis presented at different levels of detail. Treat as one combined line of evidence, not as independent verification.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30146330/","title":"Alcohol use and burden for 195 countries and territories, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016","publisher":"The Lancet (GBD 2016 Alcohol Collaborators)","source_type":"peer_reviewed","statistic":"Alcohol accounted for 2.2% of age-standardized female deaths and 6.8% of age-standardized male deaths globally in 2016; the level of alcohol consumption that minimized harm across health outcomes was zero standard drinks per week","excerpt":"\"Alcohol use was the seventh leading risk factor for both deaths and DALYs in 2016, accounting for 2.2% (95% uncertainty interval [UI] 1.5-3.0) of age-standardised female deaths and 6.8% (5.8-8.0) of age-standardised male deaths. [...] The level of alcohol consumption that minimised harm across health outcomes was zero (95% UI 0.0-0.8) standard drinks per week.\"\n","source_date":"2018-09-22","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182517/https://pubmed.ncbi.nlm.nih.gov/30146330/","calculation_notes":"GBD 2016 is the most cited peer-reviewed analysis arguing that the historical \"J-curve\" protective effect of light drinking was an artefact of abstainer heterogeneity and sick-quitter bias, and that the consumption level minimising all-cause health loss is zero. This is the methodological basis for the regional_breakdown \"light drinker\" row being non-zero rather than protective. The sex-stratified attributable fractions (2.2% F, 6.8% M) line up closely with the WHO 2019 figures (2.4% F, 6.7% M), providing independent cross-check on the global attributable share.\n","independence_note":"GBD is the upstream source for the WHO fact sheet’s attributable-fraction figures; WHO republishes GBD alcohol module outputs. Treat as partially dependent on the WHO source above — they agree to one significant figure precisely because they share a modelling pipeline.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35843246/","title":"Population-level risks of alcohol consumption by amount, geography, age, sex, and year: a systematic analysis for the Global Burden of Disease Study 2020","publisher":"The Lancet (GBD 2020 Alcohol Collaborators)","source_type":"peer_reviewed","statistic":"Among individuals aged 40 and older, the burden-weighted relative risk curve was J-shaped with a 2020 theoretical minimum-risk exposure level (TMREL) of 0.114-1.87 standard drinks per day; 59.1% of those consuming harmful amounts were aged 15-39","excerpt":"\"Among individuals aged 15-39 years in 2020, the TMREL varied between 0 (95% uncertainty interval 0-0) and 0.603 (0.400-1.00) standard drinks per day. [...] Among individuals aged 40 years and older, the burden-weighted relative risk curve was J-shaped for all regions, with a 2020 TMREL that ranged from 0.114 (0-0.403) to 1.87 (0.500-3.30) standard drinks per day. [...] Among individuals consuming harmful amounts of alcohol in 2020, 59.1% (54.3-65.4) were aged 15-39 years and 76.9% (73.0-81.3) were male.\"\n","source_date":"2022-07-14","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260421193147/https://pubmed.ncbi.nlm.nih.gov/35843246/","calculation_notes":"GBD 2020 partially walks back the \"zero is optimal\" framing of GBD 2016 by showing an age-dependent TMREL — approximately zero for young adults, non-zero (and J-shaped) for older adults. This is the source for the methodological dispute flagged in the caveats and in the body text. Does not overturn the aggregate attributable-mortality numbers, only the claim that there is no safe level at any age. Used to motivate the uncertainty band on the moderate-drinker row of the regional_breakdown.\n","independence_note":"Same GBD pipeline as GBD 2016 above and as the WHO attributable-fraction figures; treat as partially dependent. Included for the age-dependent TMREL finding, which is the most substantive methodological update in the literature since 2018.\n"},{"url":"https://www.aha.org/news/headline/2025-03-06-cdc-adult-smoking-hits-60-year-low","title":"CDC: Adult smoking hits 60-year low","publisher":"American Hospital Association (AHA) News, reporting CDC/NHIS data","source_type":"news_article","statistic":"Cigarette smoking among US adults fell from 42.4% in 1965 to 11.6% in 2022, the lowest rate in 60 years","excerpt":"\"Cigarette smoking overall by U.S. adults decreased from 42.4% in 1965 to 11.6% in 2022.\"\n","source_date":"2025-03-06","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260419110520/https://www.aha.org/news/headline/2025-03-06-cdc-adult-smoking-hits-60-year-low","calculation_notes":"Grounds the body text's contrast between alcohol's flat public perception and smoking's decades-long decline. Not used in this entry's own probability calculation (which concerns alcohol- attributable mortality); cited only to support the smoking-rate comparison in the \"cultural asymmetry\" paragraph.\n","independence_note":"Reports CDC/National Health Interview Survey smoking-prevalence data; independent of the WHO, CDC alcohol, and GBD sources used for this entry's headline alcohol figures.\n"}],"comparison_anchors":[{"label":"Death from a smoking-related disease (lifetime, regular smoker)","lifetime_us_adult":0.5},{"label":"Death from cancer (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Death from ischaemic heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Lifelong heavy drinker (>14/wk M, >7/wk W, 30+ years)","probability":0.15,"notes":"Headline subgroup. Mid-point of a 10-20% band; alcohol-attributable mortality concentrated in liver disease, cancers, cardiovascular disease, and injuries."},{"region":"Moderate drinker (5-10 drinks/wk, lifelong)","probability":0.03,"notes":"Substantially reduced hazard vs heavy drinking but non-zero attributable mortality per GBD 2016; GBD 2020 suggests some age-dependent attenuation of this figure."},{"region":"Light drinker (1-3 drinks/wk)","probability":0.008,"notes":"Small attributable fraction; the historical 'J-curve' protective effect is disputed by GBD 2016 as an abstainer-heterogeneity artefact."},{"region":"Lifelong non-drinker baseline","probability":0,"notes":"Zero alcohol-attributable mortality — this entry measures excess attributable risk only."}],"personal_factor_multipliers":[{"factor":"binge drinking (5+ drinks/occasion, weekly)","multiplier":2,"notes":"Injury and cardiovascular event risk rises sharply with episodic heavy drinking above and beyond average intake."},{"factor":"heavy drinker + smoker (interaction)","multiplier":3,"notes":"Multiplicative rather than additive interaction for upper-aerodigestive cancers (oesophageal, head and neck) and liver disease."},{"factor":"quit before age 40","multiplier":0.2,"notes":"Liver and cardiovascular risk largely reverses with sustained abstinence; cancer risk attenuates more slowly but still substantially."},{"factor":"chronic hepatitis B or C co-infection","multiplier":4,"notes":"Interaction term for liver cirrhosis and hepatocellular carcinoma mortality is large and well-documented."}],"short_label":"Regular drinking","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry is specifically the lifetime attributable mortality for someone who drinks regularly at heavy-drinker thresholds across most of adult life, not a general-population average. It is not directly comparable to the population-scope lifetime numbers on other Likelier pages (cancer, heart disease, stroke), which are averaged across drinkers and non-drinkers. The subgroup definition matters: \"regular drinker\" in the public conversation often means something much lighter than the >14 drinks/week (men) / >7 drinks/week (women) threshold used here. Light and moderate drinkers sit on a much lower part of the distribution, as reflected in the regional_breakdown rows. There is also an active methodological dispute in the literature about whether light drinking is protective, neutral, or mildly harmful. GBD 2016 (Lancet 2018) argued the historical J-curve was an artefact of abstainer heterogeneity and sick-quitter bias, and that zero drinks per week minimises all-cause health loss. GBD 2020 (Lancet 2022) partially walked this back, showing that for adults over 40 the theoretical minimum-risk exposure level is small but non-zero and varies by region. The headline figure on this page is the mid-point of the resulting uncertainty band and should be read as an order-of-magnitude calibration for heavy-drinker mortality, not as a personal forecast. Individual outcomes depend on intensity, duration, drinking pattern (steady vs binge), age, sex, co-existing tobacco use, chronic viral hepatitis status, and a long list of genetic and environmental modifiers. The 2.6M WHO and 178K CDC aggregates are robust across methodology; the per-subgroup conversion is where the uncertainty lives.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty tumbler glass sitting on a muted sand surface, flat vector illustration."},"canonical_url":"https://likelier.app/regular-drinking-death","api_url":"https://likelier.app/api/fears/regular-drinking-death.json"},{"slug":"stalking-persistence","question":"What are the lifetime odds of being stalked?","quick_answer":"About 1 in 6 US adults are stalked at some point in life — roughly 22.5% of women and 9.7% of men, per CDC survey data.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Stalking occupies a peculiar position in public risk perception. It is simultaneously the subject of intense media dramatization — the obsessed stranger, the celebrity fan, the ex who will not stop — and widespread trivialization in everyday life, where persistent unwanted contact is often reframed as romantic persistence, harmless attention-seeking, or a private matter between former partners. Most people do not consider themselves likely targets. Survey data on perceived personal stalking risk are essentially nonexistent; the concept itself was not codified in US criminal law until California's 1990 anti-stalking statute, and public understanding of what constitutes stalking remains inconsistent.\n","kind":"intuition"},"native":{"display":"~22.5% of women and ~9.7% of men have experienced stalking in their lifetime (CDC NISVS 2023/2024)","numerator":162,"denominator":1000,"unit":"lifetime prevalence (combined-sex weighted average)","population":"US adults aged 18+, NISVS nationally representative survey"},"normalized":{"lifetime_us_adult":0.162,"display":"~1 in 6 US adults experience stalking in a lifetime","log_value":-0.79,"assumptions":"CDC NISVS 2023/2024 Stalking Data Brief reports lifetime stalking prevalence of 22.5% for women (28.8 million) and 9.7% for men (11.9 million). The 2023/2024 cycle used revised stalking measurement and data collection methodology; the CDC explicitly cautions against comparing these figures to earlier NISVS waves (2011, 2016/2017) due to these methodological changes. Weighted combined-sex average using US census sex ratio (51.1% F / 48.9% M): 0.511 × 0.225 + 0.489 × 0.097 ≈ 0.1624, rounded to 0.162. This is a directly measured lifetime prevalence, not an annual-rate extrapolation. The BJS Supplemental Victimization Survey (2019) independently estimated 1.3% annual stalking prevalence (3.4 million victims), which over a 59-year adult life yields a rough lifetime estimate consistent with the NISVS figure. Uncertainty band: low end uses the BJS annual-rate extrapolation (~12%); high end reflects the upper range of state-level NISVS estimates and broader stalking definitions that capture lower-severity persistent contact (~30% combined-sex); low end uses narrower legal-threshold definitions consistent with BJS/NCVS survey instruments (~8% combined-sex).\n","uncertainty":{"low":0.08,"high":0.3},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nisvs/media/pdfs/stalking-brief.pdf","title":"National Intimate Partner and Sexual Violence Survey: 2023/2024 Stalking Data Brief","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"22.5% of women and 9.7% of men experienced stalking in their lifetime","excerpt":"\"More than 1 in 5 women (22.5% or an estimated 28.8 million) in the United States have experienced stalking during their lifetimes. Approximately 1 in 10 men (9.7% or about 11.9 million) in the United States have experienced stalking in their lifetimes.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-29","archive_url":"http://web.archive.org/web/20260511105452/https://www.cdc.gov/nisvs/media/pdfs/stalking-brief.pdf","calculation_notes":"Primary lifetime stalking prevalence from NISVS 2023/2024. Women: 22.5%, men: 9.7%. Combined-sex weighted average: 0.511 × 0.225 + 0.489 × 0.097 ≈ 0.1624 → 0.162. Used directly as lifetime_us_adult. The CDC cautions against comparing to earlier NISVS waves due to revised stalking measurement and methodology changes.\n"},{"url":"https://bjs.ojp.gov/content/pub/pdf/sv19.pdf","title":"Stalking Victimization, 2019","publisher":"Bureau of Justice Statistics (DOJ)","source_type":"govt_report","statistic":"1.3% of persons age 16+ were stalked in 2019 (approximately 3.4 million victims)","excerpt":"\"About 1.3% (3.4 million) of all persons age 16 or older were victims of stalking in 2019. The percentage of persons who experienced stalking declined from 1.5% in 2016 to 1.3% in 2019.\"\n","source_date":"2022-03-01","source_accessed":"2026-04-29","archive_url":"http://web.archive.org/web/20260507024602/https://bjs.ojp.gov/content/pub/pdf/sv19.pdf","calculation_notes":"BJS annual prevalence from the NCVS Supplemental Victimization Survey. Independent from the CDC NISVS (different survey instrument, different definition threshold). The 1.3% annual rate over a 59-year adult life yields a rough lifetime estimate of 1 − (1 − 0.013)^59 ≈ 0.54, but this overstates true lifetime prevalence because stalking victimization is not independent across years. The NISVS directly measured lifetime prevalence of 16.2% (combined-sex) is more reliable for the lifetime figure. The same 2019 BJS report also found that fewer than a third (29%) of stalking victims reported their victimization to police.\n","independence_note":"BJS uses the National Crime Victimization Survey (NCVS), methodologically independent from the CDC's NISVS. Different sampling frame, different questionnaire, different stalking definition threshold. Cross-validation strengthens confidence in the order of magnitude.\n"},{"url":"https://journals.sagepub.com/doi/10.1177/1088767999003004003","title":"Stalking and Intimate Partner Femicide","publisher":"Homicide Studies (SAGE)","source_type":"peer_reviewed","statistic":"76% of femicide victims and 85% of attempted femicide victims were stalked by their intimate partner","excerpt":"\"The prevalence of stalking was 76% for femicide victims and 85% for attempted femicide victims. Incidence of intimate partner assault was 67% for femicide victims and 71% for attempted femicide victims.\"\n","source_date":"1999-11-01","source_accessed":"2026-04-29","archive_url":"https://web.archive.org/web/20260525162555/https://journals.sagepub.com/doi/10.1177/1088767999003004003","calculation_notes":"McFarlane, Campbell et al. (1999) used an 18-item stalking inventory and personal interviews to describe intimate partner stalking within 12 months of attempted and actual partner femicide, covering 141 femicide and 65 attempted femicide incidents. The 76% figure applies to completed femicide; the 85% figure applies to attempted femicide. These are distinct populations and the figures should not be conflated.\n"}],"comparison_anchors":[{"label":"Sexual assault (lifetime, US adult, contact)","lifetime_us_adult":0.34},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.11},{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348}],"short_label":"Stalking","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The 2023/2024 NISVS used revised stalking measurement and data collection methodology; the CDC explicitly cautions against comparing these findings to earlier NISVS waves (2011, 2016/2017). The higher 2023/2024 figures may reflect improved measurement rather than a true increase in prevalence. The sex disparity (~2.3:1 female-to-male ratio) is present but narrower than in earlier waves, which may partly reflect changes in survey methodology, differential willingness to report, and differential thresholds for feeling fearful. Roughly two-thirds of female stalking victims are stalked by a current or former intimate partner, not a stranger — a pattern that contradicts the dominant media archetype. Technology-facilitated stalking (GPS tracking, social media monitoring, spyware) is increasingly prevalent; the 2023/2024 brief reports technology- facilitated tactics affected 16-29% of victims. The stalking-to-homicide pipeline is well-documented: McFarlane et al. (1999) found that 76% of intimate-partner femicide victims were stalked by their killer in the year before their death, and 85% of attempted femicide victims were stalked (these are distinct populations).\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A long shadow stretching across an empty sidewalk, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/stalking-persistence","api_url":"https://likelier.app/api/fears/stalking-persistence.json"},{"slug":"first-heart-attack-before-70","question":"What are the odds of having a first heart attack before age 70?","quick_answer":"A first heart attack strikes about 1 in 6 US adults over a lifetime, drawn from some 605,000 first heart attacks a year. Public intuition is close here, with around 1 in 5 feeling about right to most people.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Heart attacks occupy a central place in the public imagination of medical catastrophe. The perceived risk is high and fairly diffuse — surveys consistently show that heart disease is the condition Americans most fear after cancer. Many people overestimate the acute fatality rate (believing most heart attacks are immediately fatal) while simultaneously underestimating their own cumulative lifetime incidence risk. The distinction between dying from heart disease and having a non-fatal heart attack is frequently blurred in popular coverage, which conflates the two outcomes.\n","rough_estimate":"~1 in 5 lifetime feels about right to many people","kind":"intuition"},"native":{"display":"~605,000 first heart attacks per year in the US","numerator":605000,"denominator":260000000,"unit":"per year","population":"US adults (CDC, citing AHA 2025 Statistics)"},"normalized":{"lifetime_us_adult":0.17,"display":"~1 in 6 lifetime (first non-fatal MI by age ~75)","log_value":-0.77,"assumptions":"The Lloyd-Jones et al. (1999, Lancet) Framingham analysis reports the lifetime risk of hard coronary events (MI + coronary death) from age 40 as approximately 42% for men and 25% for women (calculated as total CHD risk minus angina-only events, which were 6–7 percentage points of the total at ages 40–60). Blending by sex: (42 + 25) / 2 = ~33% lifetime hard CHD from age 40. Of all acute MI events, roughly 25% are fatal in the acute phase (AHA 2025 statistics: ~805,000 total heart attacks per year, of which ~605,000 are first attacks; CDC states 1 in 5 MI deaths occur before hospital arrival). Removing fatal first events: ~33% × 0.75 ≈ 25% for surviving a first MI from age 40. Adjusted downward to ~17% to reflect the full US adult population from age 18 (where incidence below age 40 is very low) and the fact that ~25% of MI deaths occur before any non-fatal event is recorded. This ~17% lifetime estimate represents a US adult surviving a first recognized myocardial infarction — it excludes fatal events and silent MI. This entry is distinct from heart-disease-death.mdx, which covers mortality. Uncertainty range 0.12–0.24 reflects sex, birth cohort, and risk-factor differences.\n","uncertainty":{"low":0.12,"high":0.24},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/10023892/","title":"Lifetime risk of developing coronary heart disease","publisher":"Lloyd-Jones DM et al., The Lancet","source_type":"peer_reviewed","statistic":"Lifetime risk of coronary heart disease from age 40: 48.6% for men, 31.7% for women; hard events (MI + coronary death) 6–7 percentage points lower, yielding ~42% men and ~25% women for hard CHD lifetime risk from age 40.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] At age 40 years, lifetime risk of coronary heart disease was 48.6% (95% CI 45.8–51.3) for men and 31.7% (29.2–34.2) for women. At ages 40–60 years the lifetime risk of hard coronary heart disease events, excluding angina pectoris, was 6–7% lower than that for all coronary heart disease.\"\n","source_date":"1999-01-03","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20251001023151/https://pubmed.ncbi.nlm.nih.gov/10023892/","calculation_notes":"Total CHD lifetime risk from age 40: men 48.6%, women 31.7%. Hard events (MI + coronary death) = total minus angina-only: approximately 42% men, 25% women. Sex-blended average: (42 + 25) / 2 = ~33.5%. This is the hard-CHD lifetime risk from age 40, which is the closest available measure to \"first MI incidence.\" The Framingham cohort data are from 1971–1975 follow-up; contemporary rates are somewhat lower due to improved treatment, so 33% is a conservative upper bound.\n","independence_note":"Framingham Heart Study is a prospective longitudinal cohort study (Framingham, MA), entirely separate from CDC NHANES cross-sectional surveys and AHA administrative data compilations. Methods are independent: Framingham uses direct clinical examination and event adjudication over decades of follow-up.\n"},{"url":"https://www.cdc.gov/heart-disease/data-research/facts-stats/index.html","title":"Heart Disease Facts","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"About 805,000 people in the United States have a heart attack each year; 605,000 are a first heart attack. Every 40 seconds someone in the US has a heart attack.\n","excerpt":"\"Every year, about 805,000 people in the United States have a heart attack. Of these, 605,000 are a first heart attack and 200,000 happen to people who have already had a heart attack.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260523073754/https://www.cdc.gov/heart-disease/data-research/facts-stats/index.html","calculation_notes":"605,000 first heart attacks per year. US adult population ~260 million. Annual incidence rate: 605,000 / 260,000,000 ≈ 0.233% per year. Over 59 adult years at this rate: 1 − (1 − 0.00233)^59 ≈ 12.7% — but this double-counts survivors who may have subsequent events tracked separately, and underestimates because CDC figures include both fatal and non-fatal first MIs. The Framingham lifetime-risk method (see source 1) is the more appropriate basis for the lifetime point estimate. Used here to anchor the native display rate and corroborate the Framingham estimate.\n","independence_note":"CDC Health Statistics draws on the AHA Statistical Update (2025) and NHANES surveillance. It is methodologically independent of the Framingham Heart Study prospective cohort, using administrative and survey-based data rather than direct cohort follow-up.\n"},{"url":"https://www.ahajournals.org/doi/10.1161/CIR.0000000000001303","title":"2025 Heart Disease and Stroke Statistics","publisher":"American Heart Association / Circulation","source_type":"reputable_reference","statistic":"Approximately 805,000 US heart attacks per year; average age of first MI is 65.5 years for men and 72 years for women; about 1 in 5 MI deaths occur before hospital arrival.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] The 2025 Statistical Update reports approximately 805,000 myocardial infarctions per year in the United States, of which approximately 605,000 are first events. The average age of first MI is 65.5 years for males and 72.0 years for females. Approximately 1 in 5 MI deaths occur before hospital arrival.\"\n","source_date":"2025-01-27","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525095140/https://www.ahajournals.org/doi/10.1161/CIR.0000000000001303","calculation_notes":"Used to characterize the acute fatality fraction (~20–25%) and average age of first MI, which inform the adjustment from hard-CHD lifetime risk to surviving-first-MI lifetime risk. Not used as the primary probability estimate — the Framingham lifetime-risk method is preferred.\n","independence_note":"AHA Statistics compiles administrative, survey, and cohort data from multiple independent sources (NHANES, NHLBI, Medicare claims, Framingham, ARIC, CHS). It is a secondary aggregator that draws on Framingham but synthesizes far more data and is published annually by a separate editorial team.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15364185/","title":"Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study","publisher":"Yusuf S et al., The Lancet","source_type":"peer_reviewed","statistic":"Nine potentially modifiable risk factors (smoking, abnormal lipids, hypertension, diabetes, abdominal obesity, psychosocial factors, low fruit/vegetable intake, alcohol use, physical inactivity) collectively accounted for 90% of the population attributable risk of a first myocardial infarction in men and 94% in women, across 52 countries.\n","excerpt":"\"[Paraphrase from abstract — full text paywalled] Collectively, these nine risk factors accounted for 90% of the population attributable risk (PAR) of acute myocardial infarction in men and 94% in women. Individual risk factor PARs included smoking 35.7%, abnormal ApoB/ApoA1 ratio 49.2%, hypertension 17.9%, diabetes 9.9%, abdominal obesity 20.1%, psychosocial factors 32.5%, low fruit and vegetable intake 13.7%, regular alcohol consumption 6.7%, and regular physical activity 12.2%. The effect was consistent across regions, sexes, and ethnic groups.\"\n","source_date":"2004-09-11","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260320043659/https://pubmed.ncbi.nlm.nih.gov/15364185/","calculation_notes":"Yusuf et al., INTERHEART study (Lancet, 2004): case-control study of 15,152 MI cases and 14,820 matched controls across 52 countries. Cited to support the body text's claim that nine modifiable risk factors explain most of the individual-level variation around the population-average lifetime MI risk reported by the other sources in this entry. Not used to derive the headline native/normalized probability (which remains the Framingham lifetime-risk method, source 1); used only to substantiate the \"population averages conceal large individual risk-factor variation\" claim in the body text.\n","independence_note":"INTERHEART is a multinational case-control study (52 countries, enrollment 1999–2003), methodologically independent of the Framingham cohort (source 1), CDC surveillance (source 2), and the AHA statistical compilation (source 3).\n"}],"comparison_anchors":[{"label":"Death from heart disease (lifetime, US adult)","lifetime_us_adult":0.19},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Stroke (lifetime incidence, US adult)","lifetime_us_adult":0.25}],"personal_factor_multipliers":[{"factor":"Current smoker","multiplier":2.5,"notes":"Framingham 34-year follow-up and multiple pooled analyses consistently show current smokers have approximately 2–3× the MI incidence of non-smokers; effect is stronger in women (RR ~2.2) than men (RR ~1.4–2.0).\n"},{"factor":"Hypertension (systolic BP ≥140 mmHg)","multiplier":2,"notes":"AHA risk-factor data: hypertension approximately doubles the risk of first MI compared to normotensive adults; risk scales with degree of elevation.\n"},{"factor":"Type 2 diabetes","multiplier":2.5,"notes":"AHA 2025 Statistics: adults with type 2 diabetes have 2–4× higher MI risk; the effect is stronger in women, partially closing the sex gap in incidence.\n"},{"factor":"Family history of premature MI (1st-degree relative <55M or <65F)","multiplier":2,"notes":"Framingham risk equations and INTERHEART study: family history of premature coronary disease approximately doubles lifetime MI risk, independent of shared lifestyle factors.\n"},{"factor":"Non-Hispanic Black women vs Non-Hispanic White women","multiplier":1.4,"notes":"AHA Statistics: Black women have higher MI incidence and younger average age of first event than White women, reflecting both biological and social determinants of cardiovascular risk.\n"},{"factor":"Regular aerobic exercise ≥150 min/week","multiplier":0.5,"notes":"Meta-analyses of physical activity and CHD: regular moderate-intensity exercise reduces MI risk by approximately 35–50% compared to sedentary adults.\n"}],"short_label":"First heart attack","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry covers non-fatal first myocardial infarction (heart attack where the patient survives); it is distinct from heart-disease-death.mdx, which covers mortality from cardiovascular disease. The Framingham lifetime-risk figures include both MI and coronary death as \"hard events\" — the non-fatal MI subset is estimated by removing the acute fatality fraction (~20–25%). The Framingham cohort was predominantly white and recruited in 1971–1975; contemporary MI rates are somewhat lower due to statin therapy, smoking declines, and improved acute care, so the ~17% lifetime estimate is modestly conservative. Silent MI (myocardial infarction without recognized symptoms) is not captured in these figures; NHANES data suggest silent MI may account for an additional 25–40% of all MI events, meaning true MI prevalence is substantially higher than event-based incidence data suggest. Sex differences are large: lifetime risk from age 40 is roughly 42% for men vs 25% for women for all hard coronary events; the blended 17% surviving-first-MI figure reflects the full US adult population (including ages 18–40 where incidence is very low).\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"An ECG heartbeat trace on a pale background, flat vector editorial illustration."},"canonical_url":"https://likelier.app/first-heart-attack-before-70","api_url":"https://likelier.app/api/fears/first-heart-attack-before-70.json"},{"slug":"infertility-couple","question":"What are the odds of experiencing infertility when trying to conceive?","quick_answer":"About 1 in 6 couples (roughly 17.5%) experience infertility at some point when trying to conceive. Most couples assume conception will come quickly and underestimate this — fewer than half even regard infertility as a medical condition.\n","category":"health","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Most couples approaching conception assume it will happen within a few months. Surveys consistently find low-to-moderate fertility awareness among reproductive-age adults: university students overestimate the length of the fertile window, underestimate the effect of age, and fewer than half of people surveyed regard infertility as a medical condition. A 2018 systematic review of fertility-awareness studies found that participants across multiple countries reported \"inadequate fertility awareness concerning fertility, infertility risk factors, and consequences of delaying childbearing.\" The default mental model is that conception is easy and infertility is rare — a significant underestimate of the actual prevalence.\n","rough_estimate":"Most couples assume conception will happen quickly; fewer than half consider infertility a medical condition","kind":"survey","survey_source":{"title":"What do people know about fertility? A systematic review on fertility awareness and its associated factors","publisher":"BMC Public Health (Pedro et al.)","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6055749/","year":2018}},"native":{"display":"~1 in 6 couples (17.5% lifetime prevalence)","numerator":175,"denominator":1000,"unit":"lifetime prevalence","population":"reproductive-age couples globally, 12-month definition"},"normalized":{"lifetime_us_adult":0.175,"display":"~17.5% lifetime prevalence (1 in ~6 couples)","log_value":-0.76,"assumptions":"Uses the WHO 2023 pooled lifetime prevalence estimate of 17.5% for 12-month infertility, drawn from Cox et al. (2022) systematic review of 133 studies spanning 1990-2021. The figure represents the proportion of reproductive-age people who have ever experienced 12 months of unprotected intercourse without conceiving. Period prevalence (at any given time) is lower at 12.6%. The WHO report found limited variation by income level: 17.8% in high-income countries vs 16.5% in low- and middle-income countries. US-specific data from the 2015-2019 NSFG gives 8.7% infertility among married women 15-44 using a stricter current-status definition, and 13.4% impaired fecundity among all women 15-49. The WHO lifetime figure of 17.5% is used as the headline because it captures the cumulative probability a couple will face this outcome across their reproductive years.\n","uncertainty":{"low":0.126,"high":0.225},"scope":"subgroup_lifetime"},"sources":[{"url":"https://academic.oup.com/hropen/article/2022/4/hoac051/6825316","title":"Infertility prevalence and the methods of estimation from 1990 to 2021: a systematic review and meta-analysis","publisher":"Human Reproduction Open (Cox et al.)","source_type":"peer_reviewed","statistic":"Pooled lifetime prevalence of 12-month infertility: 17.5%; period prevalence: 12.6%","excerpt":"\"Pooled estimates of lifetime and period prevalence of 12-month infertility were 17.5% and 12.6%, respectively, but this varied by study population and methodological approach.\"\n","source_date":"2022-11-12","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20240730075231/https://academic.oup.com/hropen/article/2022/4/hoac051/6825316","calculation_notes":"Systematic review and meta-analysis of 133 studies from 1990-2021. This is the evidence base underlying the WHO 2023 infertility report. Lifetime prevalence 17.5% (1 in ~6) is used as the native estimate. Period prevalence 12.6% anchors the uncertainty low bound. Highest regional lifetime prevalence was 23.2% (Western Pacific), lowest 10.7% (Eastern Mediterranean). The denominator definition matters: period estimates among couples actively trying ranged from 9.4-32.0%.\n","independence_note":"This is the primary meta-analysis underpinning the WHO 2023 report. It synthesises data from 133 independent studies across all WHO regions.\n"},{"url":"https://www.who.int/news/item/04-04-2023-1-in-6-people-globally-affected-by-infertility","title":"1 in 6 people globally affected by infertility: WHO","publisher":"World Health Organization","source_type":"govt_report","statistic":"Around 17.5% of the adult population — roughly 1 in 6 worldwide — experience infertility","excerpt":"\"Around 17.5% of the adult population – roughly 1 in 6 worldwide – experience infertility, showing the urgent need to increase access to affordable, high-quality fertility care for those in need.\"\n","source_date":"2023-04-04","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260505060129/https://www.who.int/news/item/04-04-2023-1-in-6-people-globally-affected-by-infertility","calculation_notes":"WHO policy report summarising the Cox et al. 2022 systematic review. Confirms the 17.5% lifetime prevalence figure and notes limited variation between high-income (17.8%) and low- and middle-income countries (16.5%). Used as corroborating governmental source for the headline number.\n","independence_note":"Dependent on the Cox et al. systematic review — this is the policy translation of the same data, not an independent data source.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38722687/","title":"Infertility and Impaired Fecundity in Women and Men in the United States, 2015-2019","publisher":"National Health Statistics Reports, No. 202 (Chandra & Copen)","source_type":"govt_report","statistic":"8.7% of married women aged 15-44 were infertile; 13.4% of all women 15-49 had impaired fecundity","excerpt":"\"The percentage of married women ages 15-44 who were infertile rose from 2011-2015 (6.7%) to 2015-2019 (8.7%). Among all women, 13.4% of women ages 15-49 and 15.4% of women ages 25-49 had impaired fecundity in 2015-2019.\"\n","source_date":"2024-04-24","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260420042130/https://pubmed.ncbi.nlm.nih.gov/38722687/","calculation_notes":"US-specific data from the National Survey of Family Growth (NSFG), 2015-2019 cycle. Uses a stricter current-status definition: married women currently in 12+ months of unprotected intercourse without conception. The 8.7% married-women figure is lower than the WHO lifetime figure because it captures point-in-time prevalence among a specific subgroup, not cumulative lifetime experience. Impaired fecundity (13.4%) is a broader measure including difficulty carrying to term.\n","independence_note":"Entirely independent US household survey data (NSFG), separate from the WHO meta-analysis which pooled international studies.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24559617/","title":"Female age-related fertility decline. Committee Opinion No. 589","publisher":"American College of Obstetricians and Gynecologists","source_type":"reputable_reference","statistic":"Fecundity decreases gradually but significantly beginning at age 32 and more rapidly after age 37","excerpt":"\"The fecundity of women decreases gradually but significantly beginning approximately at age 32 years and decreases more rapidly after age 37 years.\"\n","source_date":"2014-03-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260420042153/https://pubmed.ncbi.nlm.nih.gov/24559617/","calculation_notes":"ACOG Committee Opinion providing the clinical framework for age-related fertility decline. Sourced the age-specific conception probabilities used in the regional breakdown: ~85% within 12 months before age 30, ~75% at 30, ~66% at 35, ~44% at 40. Updated in 2025 by a newer Committee Statement but the underlying age curve remains unchanged.\n","independence_note":"Clinical practice guideline synthesising multiple reproductive biology studies. Independent of the WHO meta-analysis and NSFG survey, though drawing from some of the same underlying literature on age and fecundability.\n"}],"comparison_anchors":[{"label":"Miscarriage (per recognized pregnancy)","lifetime_us_adult":0.15},{"label":"Cancer (lifetime, US adult)","lifetime_us_adult":0.4},{"label":"Divorce (US marriages)","lifetime_us_adult":0.41}],"regional_breakdown":[{"region":"Female age under 30","probability":0.15,"notes":"~85% conceive within 12 months; ~15% meet the 12-month infertility definition"},{"region":"Female age 30","probability":0.25,"notes":"~75% conceive within 12 months; probability of infertility rises to ~25%"},{"region":"Female age 35","probability":0.34,"notes":"~66% conceive within 12 months; ACOG recommends evaluation after 6 months at this age"},{"region":"Female age 38","probability":0.44,"notes":"Accelerating decline; roughly 56% conceive within 12 months"},{"region":"Female age 40","probability":0.56,"notes":"~44% conceive within 12 months; ACOG recommends immediate evaluation"},{"region":"Female age 43+","probability":0.7,"notes":"Fewer than 1 in 3 conceive within 12 months; egg quality and quantity sharply reduced"}],"personal_factor_multipliers":[{"factor":"Female age under 30 (baseline)","multiplier":1,"notes":"Reference group; ~15% 12-month infertility rate"},{"factor":"Female age 30-34","multiplier":1.7,"notes":"Gradual decline begins at 32; ~25% fail to conceive within 12 months"},{"factor":"Female age 35-37","multiplier":2.3,"notes":"~34% 12-month infertility; ACOG shifts to 6-month evaluation window"},{"factor":"Female age 38-39","multiplier":2.9,"notes":"~44% 12-month infertility; rapid decline accelerates"},{"factor":"Female age 40+","multiplier":3.7,"notes":"~56% 12-month infertility; per-cycle fecundability drops to ~5-10%"},{"factor":"PCOS diagnosis","multiplier":2,"notes":"Polycystic ovary syndrome is the most common ovulatory disorder; affects ~6-12% of reproductive-age women"},{"factor":"Endometriosis diagnosis","multiplier":2.5,"notes":"30-50% of women with endometriosis experience infertility"},{"factor":"Male factor present","multiplier":1.5,"notes":"Male factor solely responsible in ~20% of cases, contributory in another 30-40% (AUA/ASRM)"},{"factor":"Prior successful pregnancy","multiplier":0.6,"notes":"Secondary infertility is less common than primary; proven fertility reduces baseline risk"},{"factor":"Current smoker (either partner)","multiplier":1.6,"notes":"Smoking reduces female fecundability and impairs sperm parameters"},{"factor":"BMI over 30 (either partner)","multiplier":1.4,"notes":"Obesity disrupts ovulation and reduces sperm quality; dose-response relationship"}],"short_label":"Infertility","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline 17.5% is a lifetime prevalence — the proportion of reproductive-age people who will ever experience 12 months of unprotected intercourse without conceiving. Many of these couples eventually conceive without treatment (subfertility is not sterility). The 12-month clinical definition does not distinguish temporary from permanent infertility. Age-specific figures in the regional breakdown refer to the probability of not conceiving within 12 months at that age; they are not additive across ages. Male factor data are underrepresented in population surveys because most historical studies defined infertility through female respondents only. IVF success rates are not included in the headline probability — they describe treatment outcomes, not population prevalence.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"Two overlapping circles in muted tones against a pale background, one slightly incomplete, flat vector illustration."},"canonical_url":"https://likelier.app/infertility-couple","api_url":"https://likelier.app/api/fears/infertility-couple.json"},{"slug":"student-sexual-assault","question":"How likely is a woman in higher education to experience sexual assault during her studies?","quick_answer":"About 1 in 6 women in higher education globally (17.5%) report sexual assault during their studies, rising to roughly 1 in 4 in US undergraduate surveys — a risk that tends to be systematically underestimated.\n","category":"crime","tags":["relationships","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Public perception of sexual assault on campuses varies enormously by country and cultural context. In the US, a series of high-profile cases, the AAU 2019 campus climate survey, and the Title IX regulatory framework have raised awareness to levels not seen in comparable educational systems. In other high-income countries — Australia, the UK, Germany — awareness is lower despite comparably measured prevalence. Globally, the risk is systematically underestimated: women in higher education who are assaulted most often do not report to authorities, and many frame the experience in ways that diverge from legal definitions of assault. The multi- year exposure of a university degree means cumulative probability is substantially higher than annual rates would suggest.\n","kind":"intuition"},"native":{"display":"17.5 in 100 women students report sexual assault during higher education","numerator":17,"denominator":100,"unit":"multi-year cumulative (study period)","population":"women enrolled in higher education (Steele et al. 2024 meta-analysis, 131 articles, 21 countries; US-dominated)"},"normalized":{"lifetime_us_adult":0.175,"display":"approximately 1 in 6 women students globally; 1 in 4 in US undergraduate surveys","log_value":-0.76,"assumptions":"Steele et al. 2024 (Trauma, Violence, & Abuse) meta-analysis of 131 articles across 21 countries: 17.5% of women students report sexual assault during higher education (95% CI 15.3–20.0; US-dominated, ~73% of articles). US data (AAU 2019, n=181,752): 25.9% of undergrad women report nonconsensual sexual contact since enrollment; 16.2% report nonconsensual penetration. The headline (0.175) uses the Steele et al. 2024 meta-analysis as the primary source. The US AAU figure (0.26) is cited for comparison. Australia 2021 National Student Safety Survey (Universities Australia / Social Research Centre): 4.5% sexual assault, 16.1% sexual harassment since starting university (all students). Scope is subgroup_lifetime (women enrolled in higher education, study period cumulative). Definitional heterogeneity is substantial; see caveats. Low (0.05): narrower definitions (penetration only, studies with lower recall). High (0.30): US-specific broad definitions including all nonconsensual sexual contact.\n","uncertainty":{"low":0.05,"high":0.3},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11155219/","title":"The Prevalence of Sexual Assault Among Higher Education Students: A Systematic Review With Meta-Analyses","publisher":"Trauma, Violence, & Abuse","source_type":"peer_reviewed","statistic":"17.5% of women students report sexual assault during higher education — meta-analysis of 131 articles from 21 countries","excerpt":"\"We identified 131 articles, from 21 different countries. The meta-analyzed prevalence of sexual assault was 17.5% for women. The vast majority of studies came from the United States, and prevalence estimates varied widely depending on how sexual assault was defined and measured.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260107070431/https://pmc.ncbi.nlm.nih.gov/articles/PMC11155219/","calculation_notes":"Steele et al. 2024 (Trauma, Violence, & Abuse) — systematic review with meta-analyses. 131 articles from 21 countries; the meta-analyzed prevalence of sexual assault was 17.5% for women students (95% CI 15.3–20.0). The 17.5% pooled prevalence is the native rate. Normalized directly: 0.175, scope = subgroup_lifetime (women students, study-period cumulative). No conversion needed as this is already expressed as a cumulative probability over the higher education period. (Note: the 17.5% figure is dominated by US studies — 95 of 131 articles, ~73% — so the global generalization is limited; see caveats.)\n"},{"url":"https://www.aau.edu/sites/default/files/AAU-Files/Key-Issues/Campus-Safety/Revised%20Aggregate%20report%20%20and%20appendices%201-7_(01-16-2020_FINAL).pdf","title":"AAU Campus Climate Survey on Sexual Assault and Misconduct 2019","publisher":"Association of American Universities","source_type":"reputable_reference","statistic":"25.9% of undergraduate women report nonconsensual sexual contact since enrollment; 16.2% report nonconsensual penetration; largest US campus sexual climate survey (181,752 respondents)","excerpt":"\"Among undergraduate women, 25.9 percent reported experiencing nonconsensual sexual contact by physical force, threats of physical force, or incapacitation since enrollment at their institution. 16.2 percent of undergraduate women reported nonconsensual penetration since enrolling. The survey was administered across 33 AAU universities, yielding 181,752 completed responses; the overall rate of nonconsensual sexual contact by physical force or inability to consent across all students was 13 percent.\"\n","source_date":"2020-01-16","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260510011136/https://www.aau.edu/sites/default/files/AAU-Files/Key-Issues/Campus-Safety/Revised%20Aggregate%20report%20%20and%20appendices%201-7_(01-16-2020_FINAL).pdf","calculation_notes":"AAU Campus Climate Survey 2019 — revised final report. n=181,752 across 33 research universities. The 25.9% \"nonconsensual sexual contact\" figure and 16.2% penetration figure (both undergraduate-women rates) are the US-specific anchor. The 13% figure reported by AAU is the ALL-STUDENTS nonconsensual-contact rate, not the undergraduate- women penetration rate. AAU universities are selective research institutions, possibly not representative of all US colleges; the figures are used here as the US-specific corroboration for the global meta-analysis headline.\n"},{"url":"https://universitiesaustralia.edu.au/wp-content/uploads/2022/03/2021-NSSS-National-Report.pdf","title":"2021 National Student Safety Survey: National Report","publisher":"Universities Australia (Social Research Centre)","source_type":"reputable_reference","statistic":"4.5% of students in Australia were sexually assaulted, and 16.1% sexually harassed, since starting university (all students, 43,819 respondents across 39 universities)","excerpt":"\"One in twenty (4.5 per cent) students had been sexually assaulted since starting university, and one in six (16.1 per cent) students had been sexually harassed since starting university. Most incidents were not formally reported. Female, transgender and non-binary students reported higher rates than male students.\"\n","source_date":"2022-03-23","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260625175549/https://universitiesaustralia.edu.au/wp-content/uploads/2022/03/2021-NSSS-National-Report.pdf","calculation_notes":"2021 National Student Safety Survey — national report, commissioned by Universities Australia and conducted by the Social Research Centre (NOT the Australian Human Rights Commission, which ran the earlier 2016 survey). n=43,819 students across 39 universities. The 4.5% sexual assault rate is an ALL-STUDENT figure (not women-specific) under a narrower Australian definition, providing a lower-bound anchor; 16.1% is the all-student harassment rate. The narrower definition and broader denominator explain why this figure is lower than the global meta or the US AAU undergraduate-women rate; all are likely measuring the same phenomenon with different definitional inclusion thresholds and population bases.\n"}],"comparison_anchors":[{"label":"Sexual assault lifetime (all women, general population)","lifetime_us_adult":0.2},{"label":"Intimate partner violence (lifetime, women)","lifetime_us_adult":0.3}],"personal_factor_multipliers":[{"factor":"First year of study","multiplier":2.5,"notes":"First-semester risk is disproportionately high ('red zone' — first 8 weeks of college); NISVS and campus studies consistently show this peak"},{"factor":"Fraternity/sorority membership (US context)","multiplier":1.7,"notes":"US campus studies: fraternity-affiliated social environments associated with 1.5–2× elevated risk for women"}],"short_label":"Student sexual assault","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"Definitional heterogeneity is the primary methodological challenge. The Steele et al. 2024 meta-analysis pools studies using widely varying definitions: some include verbal coercion, some require physical force, some cover any nonconsensual contact, and some are limited to penetration. The 17.5% headline is a pooled average across these heterogeneous studies, and is heavily US-weighted (~73% of the 131 included articles came from the United States), so it is not a representative global average. The US AAU figure (25.9% contact, 16.2% penetration for undergraduate women) reflects broader definitions than many international studies. Direct cross-country comparison requires care: a 4.5% rate in Australia (all students, narrower definition) and 17.5% for women may reflect definitional and population-base variation rather than prevalence differences. Reporting rates are very low across all settings (typically 5–20%), meaning these figures come from research surveys, not crime statistics. Study period varies (1 year vs. entire degree); the cumulative-over-degree framing used in most US studies produces higher rates than annual estimates.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of an empty university corridor at night, muted tones."},"canonical_url":"https://likelier.app/student-sexual-assault","api_url":"https://likelier.app/api/fears/student-sexual-assault.json"},{"slug":"currency-collapse-hyperinflation","question":"What are the odds of living through a currency collapse or hyperinflation?","quick_answer":"Globally, roughly 1 in 5.5 adults live through a currency collapse or hyperinflation in their lifetime, far more common than reserve-currency residents assume, for whom it feels like a once-a-century freak event. The risk is heavily concentrated in chronic-inflation regions rather than spread evenly.\n","category":"other","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"For readers in reserve-currency economies, currency collapse is the canonical \"happens to other people\" risk. The mental image is fixed and historical: Weimar wheelbarrows of marks, a Zimbabwean hundred-trillion-dollar note, a Venezuelan shop re-pricing twice a day. Because those images are extreme and foreign, the typical US, eurozone, or Japanese adult treats the event as a museum piece — vanishingly rare, confined to failed states and lost wars. That intuition is well calibrated for their own currency and badly calibrated as a statement about the world: at the global level, a substantial minority of adults alive today have personally lived through an inflation crisis, and in Latin America, Sub-Saharan Africa, and the former Soviet bloc it is closer to a generational rite of passage than a freak event. The Chapman Survey of American Fears does not isolate currency collapse, but \"economic/financial collapse\" recurs in its top ten, suggesting the dread exists even where the personal probability is near zero.\n","rough_estimate":"Most reserve-currency adults treat it as a once-a-century freak event; globally it is closer to a once-a-generation regional regularity","kind":"intuition"},"native":{"display":"25 of 133 market economies experienced a very-high-inflation episode (>100%/yr) between 1960 and 1996 (Fischer, Sahay & Vegh, JEL 2002)","numerator":25,"denominator":133,"unit":"market economies with at least one >100%/year inflation episode, 1960-1996","population":"133 market economies in the IMF sample"},"normalized":{"lifetime_us_adult":0.18,"display":"~1 in 5.5 lifetime (global adult living through a currency collapse or hyperinflation)","log_value":-0.74,"assumptions":"This is a population-weighted global-adult estimate, not a country-count rate, and not a US-adult rate. The distinction matters more here than for almost any other entry on the site, because inflation crises are extraordinarily concentrated. They are not sprinkled uniformly across countries: a persistent subset of chronic-inflation economies returns to crisis repeatedly (Argentina spent over seventeen years above 100%/yr in the Fischer et al. sample, Brazil over fifteen, the Democratic Republic of the Congo had six separate episodes totalling fifteen years), while reserve-currency and large low-inflation economies essentially never cross the threshold in a living adult's lifetime. The right model is therefore not a uniform per-country hazard compounded over 59 years, but a population partition: lifetime exposure is approximately 1 for an adult in a chronic / crisis-prone economy and approximately 0 for an adult in a low-inflation giant. The threshold must be applied symmetrically to every country, which is the crux. The question asks specifically about a systemic monetary-order failure — hyperinflation, or an outright currency collapse — not an ordinary high-inflation year, so the operational line is \"sustained very-high inflation / monetary-order failure\" rather than the broad Reinhart-Rogoff >=20%/yr inflation-crisis line. Under that strict reading the in-bucket population is roughly: most of Latin America and the Caribbean (~663M, ~8% of world population — Argentina, Brazil, Bolivia, Peru, Venezuela and others), the chronic / very-high-inflation subset of Sub-Saharan Africa (Fischer et al. name nine African countries with >100%/yr episodes — Angola, the DRC, Ghana, Guinea-Bissau, Sierra Leone, Somalia, Sudan, Uganda, Zambia — plus Zimbabwe; on the order of 5-7% of world population, NOT all 16% of SSA), the former Soviet bloc (~300M, ~4% — the entire bloc ran inflation above 100%/yr in 1992-94), plus chronic cases outside those blocs (Turkey, Iran, Lebanon, Myanmar, Suriname; roughly 2-3%). Crucially, China and India are OUT under this strict reading: China's last hyperinflation was 1947-49 and India has never crossed ~100%/yr, and excluding them is only consistent because SSA is counted as the named high-inflation subset rather than the whole region. Summing the in-bucket population shares gives roughly 18% of global adults, or ~1 in 5.5 — consistent with the native rate (25 of 133 market economies, ~19%). The uncertainty band reflects the threshold choice, which is the dominant lever: restricting to true Cagan hyperinflation alone (>=50%/month) pulls the figure toward ~0.10; broadening to any >=20%/yr inflation-crisis year would pull China (~24% in 1994), India (~28% in the mid-1970s oil shock), Indonesia and much of South/Southeast Asia into the bucket and push the figure above ~0.35. The point estimate sits at the strict monetary-order-failure reading the question specifies.\n","uncertainty":{"low":0.1,"high":0.35},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.nber.org/papers/w8930","title":"Modern Hyper- and High Inflations","publisher":"National Bureau of Economic Research / Stanley Fischer, Ratna Sahay, Carlos A. Vegh (also Journal of Economic Literature 40(3), 2002)","source_type":"peer_reviewed","statistic":"Following Cagan (1956), hyperinflation begins the month inflation first exceeds 50% per month; between 1947 and 1984 there were no hyperinflations, and since 1984 at least seven (in six countries); 45 very-high-inflation episodes (>100%/yr) in 25 countries; as many as 25 of 133 market economies have had a very-high-inflation episode","excerpt":"\"Cagan defined a hyperinflation as beginning in the month inflation first exceeds 50 percent (per month) and as ending in the month before the monthly inflation rate drops below 50 percent for at least a year. [...] Between 1947 and 1984 there were no hyperinflations. Since 1984, there have been at least seven (in six countries) in the market economies [...] since 1960, most countries have suffered from at least one episode of inflation of more than 25 percent per annum, and as many as 25 (out of 133) market economies have experienced an episode of very high inflation (i.e., twelve-month inflation above 100 percent).\"\n","source_date":"2002-05-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20250710172105/https://www.nber.org/papers/w8930","calculation_notes":"Fischer, Sahay & Vegh provide the cleanest cross-country base rate. The native figure uses their headline count: 25 of 133 market economies experienced an episode above 100%/yr over roughly 1960-1996. This is a share of COUNTRIES, not of people. It is used as the native rate (25/133 ~= 19% of market economies), and as the anchor for the lower end of the population-weighted band, because true hyperinflation and >100%/yr inflation are rarer than the broader 20-40%/yr \"currency collapse\" events that the headline normalized figure counts. The Cagan 50%/month definition fixes the upper (hyperinflation) tail; the \"no hyperinflations 1947-1984, then seven since 1984\" sentence establishes that true hyperinflation is a handful of episodes per generation, concentrated in a few countries.\n","independence_note":"Methodologically independent of Reinhart & Rogoff: Fischer et al. use IMF International Financial Statistics monthly price data and Cagan-style episode dating on a 133-country market-economy sample, whereas Reinhart & Rogoff build an eight-century annual cross-country inflation database. The two converge on the same qualitative picture from different data.\n"},{"url":"https://www.nber.org/papers/w13882","title":"This Time Is Different: A Panoramic View of Eight Centuries of Financial Crises","publisher":"National Bureau of Economic Research / Carmen M. Reinhart, Kenneth S. Rogoff","source_type":"peer_reviewed","statistic":"Inflation crisis defined as annual inflation >=20%; no emerging-market country in history (including the US in the 1860s) has escaped bouts of high inflation; across 66 countries inflation crises and exchange-rate crashes travel hand in hand","excerpt":"\"If serial default is the norm for a country passing through the emerging market state of development, then the tendency to lapse into periods of high and extremely high inflation is an even more striking common denominator. No emerging market country in history, including the United States (whose inflation rate exceeded 20 percent during the country's 1860s civil war) has managed to escape bouts of high inflation. [...] inflation crises and exchange rate crises travel hand-in-hand in the overwhelming majority of episodes across time and countries.\"\n","source_date":"2008-03-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260120221923/https://www.nber.org/papers/w13882","calculation_notes":"Reinhart & Rogoff supply the operational threshold for the broad \"currency collapse\" framing: an inflation crisis is a year with annual inflation >=20%, and they show inflation crises and exchange-rate crashes co-occur in the overwhelming majority of episodes — i.e. an inflation crisis IS, in practice, a currency collapse. Their 66-country eight-century database documents that essentially every emerging-market economy has had at least one such episode, while only a short list of countries (notably New Zealand and Panama in their Table 13) show no period of inflation over 20%. This supports the population-partition method: the in-bucket population share is approximated by the emerging-market / chronic-inflation regions, against an out-bucket of reserve-currency economies plus China and India.\n","independence_note":"Independent annual cross-country price database (consumer-price and cost-of-living indices back to the 1700s-1800s for many countries), distinct from the IMF monthly-data approach in Fischer et al. Reinhart & Rogoff cite Fischer, Sahay & Vegh only for the African high-inflation analysis, not for the crisis-dating used here.\n"},{"url":"https://www.cato.org/research/world-inflation-and-hyperinflation-table","title":"World Inflation and Hyperinflation Table","publisher":"Cato Institute / Steve H. Hanke and Nicholas Krus (Johns Hopkins Institute for Applied Economics)","source_type":"reputable_reference","statistic":"The Hanke-Krus World Hyperinflation Table documents all 56 documented episodes of hyperinflation since the 1790s; Venezuela was added in 2016 as the 57th verified episode","excerpt":"\"This chapter supplies, for the first time, a table that contains all 56 episodes of hyperinflation, including several which had previously gone unreported.\"\n","source_date":"2017-01-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260305081239/https://www.cato.org/research/world-inflation-and-hyperinflation-table","calculation_notes":"The Hanke-Krus table is the definitive catalogue of true (Cagan-threshold, >=50%/month) hyperinflation. Fifty-six episodes since the assignat inflation of revolutionary France (1795-96), with Venezuela added as the 57th in 2016, across well over two centuries and the whole world, is the headline rarity fact: true hyperinflation is roughly one verified episode every four years globally, heavily clustered around wars, revolutions, the collapse of empires, and the births of new states (the post-WWI European cluster, the post-Soviet 1992-94 cluster). This source bounds the EXTREME tail of the distribution and is why the normalized figure is driven by the broader 20-40%/yr currency-collapse band rather than by the much rarer hyperinflation band alone.\n","independence_note":"Independent episode catalogue compiled from primary price data by Hanke and Krus, distinct from both the IMF (Fischer et al.) and the Reinhart-Rogoff databases. Uses the strict Cagan monthly-rate definition.\n"},{"url":"https://www.guinnessworldrecords.com/world-records/762605-highest-inflation-rate-ever","title":"Highest inflation rate (ever)","publisher":"Guinness World Records","source_type":"reputable_reference","statistic":"The highest recorded inflation occurred in Hungary in July 1946: a daily inflation rate of 207%, with prices doubling roughly every 15 hours","excerpt":"\"The highest recorded rate of inflation occurred in Hungary during July 1946. [...] This works out to a daily inflation rate of 207%, meaning that prices denominated in Hungarian pengo doubled every 15 hours.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20251207003008/https://www.guinnessworldrecords.com/world-records/762605-highest-inflation-rate-ever","calculation_notes":"Used only as the illustrative ceiling of the distribution — the single worst monthly/daily inflation rate ever measured. It anchors the \"how bad can it get\" end of the prose and is not used in the probability arithmetic. Hungary's August 1946 forint reset (replacing the pengo at a rate on the order of 4 x 10^29 to 1) is the canonical example of a complete monetary-order collapse and currency replacement.\n","independence_note":"Record-keeping reference; the Hungary 1946 figure is independently reported by the Hanke-Krus table, Cagan (1956), and Fischer et al., so the illustrative fact is well corroborated.\n"}],"comparison_anchors":[{"label":"Living through a major economic recession (US adult, lifetime)","lifetime_us_adult":0.99},{"label":"Major stock market crash >30% (US adult investor, lifetime)","lifetime_us_adult":0.99},{"label":"Pension fund failure / benefit cut (DB participant, lifetime)","lifetime_us_adult":0.1},{"label":"Personal bankruptcy (US adult, lifetime)","lifetime_us_adult":0.1}],"personal_factor_multipliers":[{"factor":"resident of a reserve-currency economy (US dollar, euro, yen, Swiss franc)","multiplier":0.05,"notes":"Reinhart & Rogoff (2008): no advanced reserve-currency economy has had an inflation crisis (>=20%/yr) in the post-WWII period; the US last exceeded 20% during the 1860s Civil War. Personal lifetime exposure is near zero, which is why the global headline number feels wrong to these readers."},{"factor":"resident of Latin America, Sub-Saharan Africa, or a former Soviet state","multiplier":2.5,"notes":"Fischer, Sahay & Vegh (2002) locate the large majority of >100%/yr episodes in Latin America (12 countries) and Africa (9 countries); the entire former Soviet bloc ran inflation above 100%/yr in 1992-94. Reinhart & Rogoff find essentially every economy in these regions has at least one inflation crisis on record. Lifetime exposure approaches 1."},{"factor":"resident of a chronic-inflation economy (Argentina, Venezuela, Zimbabwe, Turkey)","multiplier":3,"notes":"Fischer et al. (2002): chronic-inflation economies return to crisis repeatedly — Argentina spent >17 years above 100%/yr in their sample, the DRC had six separate episodes. For an adult born into one of these economies, living through at least one currency collapse is close to certain rather than merely likely."},{"factor":"wealth held in foreign hard currency or hard assets (USD, gold, property abroad)","multiplier":0.4,"notes":"Reinhart & Rogoff (2008) show inflation crises and exchange-rate crashes co-occur; dollarization and hard-asset holding are the standard documented hedge. This does not change the probability of the macro event but sharply reduces the personal wealth destruction conditional on it — the same event/severity split used in the recession and stock-crash entries."},{"factor":"resident of China or India","multiplier":0.15,"notes":"Fischer et al. (2002) and Reinhart & Rogoff (2008): China's last hyperinflation was 1947-49 and India has never crossed ~100%/yr. Together these two countries are ~36% of humanity and sit in the low-exposure bucket under the strict monetary-order-failure reading — the single largest reason the population-weighted global figure is ~1 in 5 rather than a majority. (Both DID have inflation-crisis years above 20%/yr — China ~24% in 1994, India ~28% in the mid-1970s — so under the broad >=20%/yr definition they would move into the bucket and the global figure would rise above ~0.35.)"}],"short_label":"Currency collapse","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"The headline ~1 in 5.5 is a GLOBAL-adult, lifetime figure and is dominated by one modelling choice and one definitional choice. The modelling choice: inflation crises are concentrated, not uniform, so the figure is built by partitioning the world's population into a high-exposure bucket (lifetime probability ~1) and a near-zero bucket, then summing population shares — not by compounding a single per-country hazard. The definitional choice: \"currency collapse or hyperinflation\" has no single agreed threshold, and the threshold MUST be applied symmetrically to every country or the figure becomes incoherent. True Cagan hyperinflation (>=50%/month) is genuinely rare — 57 verified episodes worldwide since the 1790s in the Hanke-Krus table, with none at all between 1947 and 1984. The central estimate uses the strict \"systemic monetary-order failure\" reading the question specifies — sustained very-high inflation up through hyperinflation — which keeps the low-inflation giants China and India out of the bucket and counts Sub-Saharan Africa as its named high-inflation subset rather than the whole region. That reading lands at ~1 in 5.5 and agrees with the country rate (25 of 133, ~19%). Restricting all the way to true Cagan hyperinflation pulls the figure toward the lower bound (~0.10); broadening to the looser Reinhart-Rogoff inflation-crisis line (>=20%/yr) pulls China (~24% in 1994), India (~28% in the mid-1970s) and much of South and Southeast Asia into the bucket and pushes it toward the upper bound (~0.35). The US-versus-global gap is the key caveat: a US, eurozone, or Japanese adult's personal lifetime probability is near zero (the reserve-currency multiplier of 0.05 reflects this), while a Latin American, Sub-Saharan African, or post-Soviet adult's is close to certain. The figure also measures living through the event, not suffering proportional wealth loss — the standard event-versus-severity distinction. Holding hard currency or hard assets is the documented hedge and is captured as a protective multiplier, not as a change to the macro probability. The estimate treats post-2000 quiescence cautiously: Reinhart & Rogoff explicitly warn that quiet periods in inflation, as in default, do not extend indefinitely, and Venezuela's 2016 entry into the hyperinflation record book within a decade of the book's publication is a live example.\nThe dread also runs the other way, and that direction has no base rate. A separate strand of forecasting — most loudly Elon Musk's 2026 prediction that AI- and robotics-driven abundance will make money \"irrelevant\" by the mid-2030s — imagines the monetary order dissolving through deflation and post-scarcity rather than through inflation. That is a speculative forecast, not a measured frequency: it has no historical episode base, so it is deliberately excluded from the figure above and carries no number here. It is flagged only so the absence of a deflation-side estimate is not read as a claim that a currency can fail only by losing its value, never by losing its purpose.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.6,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"A banknote rendered as a simple rectangle whose value digits trail off into ever-smaller zeros, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/currency-collapse-hyperinflation","api_url":"https://likelier.app/api/fears/currency-collapse-hyperinflation.json"},{"slug":"five-plus-years-paid-ltc","question":"What are the odds of needing more than five years of paid long-term care?","quick_answer":"About 1 in 5.6 US adults over a lifetime, roughly 22 in 100 of those turning 65, will need more than five years of paid long-term care. Most people guess under 1 in 10, underestimating the odds.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Most people who engage with long-term care planning at all think of it as a short-term bridge risk — a year or two at most. Few people spontaneously consider the possibility of needing paid care for five or more years. The catastrophic tail of the LTC distribution is the least-discussed and most financially consequential part of the risk profile. No rigorous survey measuring perceived probability of 5+ year LTC need is available; the figure is almost certainly underestimated by most adults.\n","rough_estimate":"under 1 in 10 lifetime, most people guess","kind":"intuition"},"native":{"display":"~22 in 100 adults turning 65 in 2021-2025","numerator":22,"denominator":100,"unit":"lifetime conditional on reaching age 65","population":"US adults turning 65 in 2021-2025 (ASPE DYNASIM4 projections)"},"normalized":{"lifetime_us_adult":0.18,"display":"~1 in 5.6 US adults over their lifetime","log_value":-0.745,"assumptions":"ASPE 2022 DYNASIM4 projections: 22% of adults turning 65 in 2021-2025 will have LTSS needs lasting more than five years. This is a conditional probability (given reaching age 65). Converting to unconditional lifetime probability: 0.22 × 0.82 (survival to age 65) ≈ 0.18. This figure captures the catastrophic-duration tail of the LTSS distribution — the subset for whom care needs persist long enough to exhaust most private savings. The 2019 ASPE/Urban Institute historical analysis (based on HRS cohorts) found that 38% of severe-need episodes lasted more than 4 years, from which a >5-year probability can be estimated at roughly 25-30% conditional; the DYNASIM4 22% figure is somewhat lower and is used as the primary estimate because it is forward-looking for the current cohort. Uncertainty range: 0.12 (if the narrower paid-care-only definition is used) to 0.26 (if broader severe-need definition is applied).\n","uncertainty":{"low":0.12,"high":0.26},"scope":"us_adult_lifetime"},"sources":[{"url":"https://aspe.hhs.gov/reports/ltss-older-americans-risks-financing-2022","title":"Long-Term Services and Supports for Older Americans: Risks and Financing, 2022","publisher":"US Department of Health and Human Services, ASPE","source_type":"govt_report","statistic":"22% of adults turning 65 in 2021-2025 will have LTSS needs lasting more than five years; this share is higher for low-income adults (29%) and those in fair/poor health at 65 (25%)","excerpt":"\"About 56 percent of people turning 65 between 2021 and 2025 will need LTSS in their lifetime. About 22 percent will have needs lasting more than five years.\"\n","source_date":"2022-09-27","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260306042934/https://aspe.hhs.gov/reports/ltss-older-americans-risks-financing-2022","calculation_notes":"ASPE 2022 DYNASIM4 microsimulation for adults turning 65 in 2021-2025. The 22% figure is the share with LTSS needs (any type) lasting >5 years, conditional on reaching age 65. Unconditional from birth: 0.22 × 0.82 ≈ 0.18. Using 2024 Genworth/CareScout Cost of Care data for context: $127,750/year for a nursing home private room; 5 years = ~$638,750. Even at home health aide rates ($77,792/year median in 2024), 5 years = ~$388,960. These costs are beyond the reach of most middle-income households without LTC insurance or Medicaid spend-down.\n","independence_note":"ASPE DYNASIM4 uses its own microsimulation modeling framework drawing on Census, HRS, and CMS administrative data. The 2022 projections are forward-looking estimates, distinct from the retrospective HRS-based analysis in Johnson (2019).\n"},{"url":"https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","title":"What Is the Lifetime Risk of Needing and Receiving Long-Term Services and Supports?","publisher":"US Department of Health and Human Services, ASPE","source_type":"govt_report","statistic":"38% of severe LTSS need episodes last more than 4 years; 9% last more than 10 years; only 5% of older adults receive paid LTSS for more than 5 years","excerpt":"\"About 70 percent of adults who survive to age 65 will develop severe LTSS needs before they die and 48 percent will receive some paid care over their lifetime. Only 24 percent of older adults receive more than two years of paid LTSS care, and only 15 percent spend more than two years in a nursing home.\"\n","source_date":"2019-04-03","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260523053610/https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","calculation_notes":"The 2019 ASPE/Urban Institute analysis provides the historical duration distribution: of severe need episodes, 40% last ≤2 years, 38% last >4 years, and 9% last >10 years. Applied to the 70% conditional prevalence: the >4-year group is 0.70 × 0.38 ≈ 27% of those reaching 65. Only 24% receive more than 2 years of *paid* care, and only 5% of all older adults receive paid LTSS for more than 5 years — lower than the 22% with any LTSS need >5 years because many long-need individuals rely on unpaid family care.\n","independence_note":"The 2019 ASPE/Urban Institute report and the 2022 ASPE brief use different methodologies and cohort definitions; both are included to bracket the plausible range. The 22% figure (2022) refers to any LTSS need >5 years while the 5% figure (2019) refers to paid care >5 years, explaining much of the gap between the two estimates.\n"}],"comparison_anchors":[{"label":"Needing severe LTSS at some point after 65 (unconditional)","lifetime_us_adult":0.57},{"label":"Nursing home admission, at least one night (lifetime)","lifetime_us_adult":0.46},{"label":"Bankruptcy or severe financial distress (lifetime)","lifetime_us_adult":0.17}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.4,"notes":"Women have substantially longer LTSS need durations; 44% of women vs 28% of men have severe needs lasting >2 years (ASPE 2019); the sex ratio compounds strongly at the 5-year tail"},{"factor":"Alzheimer's or dementia diagnosis","multiplier":3,"notes":"Dementia-related LTSS needs are the primary driver of very long care episodes; dementia patients account for a disproportionate share of the 5+ year tail"},{"factor":"Non-Hispanic Black or Hispanic ethnicity","multiplier":1.3,"notes":"ASPE 2022 data: 50% of older non-Hispanic Blacks and 57% of older Hispanics with severe LTSS needs experience needs lasting >4 years, vs 35% of non-Hispanic Whites"},{"factor":"Has long-term care insurance covering 5+ year benefit period","multiplier":1,"notes":"Insurance affects financial exposure, not the underlying care probability; the 5-year risk is unchanged by coverage"},{"factor":"Reports fair or poor health at age 65","multiplier":1.25,"notes":"ASPE 2022 DYNASIM projections: 25% of those in fair/poor health at 65 have needs lasting >5 years, vs 22% overall"}],"short_label":"5+ years paid LTC","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"financial","valence":"negative","caveats":"The 22% figure (ASPE 2022) measures adults with *any* LTSS need lasting more than five years; the 5% figure in the 2019 analysis measures *paid* care lasting >5 years. The gap between the two (22% any LTSS need vs 5% paid LTSS >5 years) reflects the large role of unpaid family caregiving in absorbing long care episodes. Financial planning for the catastrophic tail should focus on the paid-care definition, which is the relevant exposure for asset depletion and Medicaid spend-down. The normalized unconditional estimate of 0.18 is based on the broader any-LTSS-need definition and therefore represents an upper bound on the financially catastrophic scenario (paid care >5 years unconditional ≈ 0.05 × 0.82 ≈ 0.04). Both figures are reported in the caveats for completeness. Annual nursing home cost data from Genworth/CareScout 2024 ($127,750 private room median) is used for illustrative financial context only; costs vary widely by setting (home health, assisted living, memory care) and geography.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A series of calendar pages fanning out from left to right on a pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/five-plus-years-paid-ltc","api_url":"https://likelier.app/api/fears/five-plus-years-paid-ltc.json"},{"slug":"teen-suicide-attempt","question":"What are the odds of a US teenager attempting suicide?","quick_answer":"About 1 in 6 US adolescents attempt suicide at least once between ages 12 and 17, and roughly 1 in 10 high schoolers report an attempt in the past year. Most parents guess far lower, closer to 1 in 100.\n","category":"kids","tags":["teen","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Most American parents dramatically underestimate how common suicide attempts are among teenagers. When the CDC reports that roughly one in ten high school students attempted suicide in the past year, the number lands as a shock — parents of teenagers tend to guess something closer to one in a hundred or one in several hundred, partly because attempts are far less visible than deaths and partly because teens who attempt rarely tell their parents. The cultural framing of teen suicide leans heavily on completed suicides (news stories, school assemblies after a death), which are far rarer than attempts in this age group. That asymmetry means the public mental model is anchored on the death rate (~7 per 100,000 for ages 15-19) rather than the past-year attempt rate (~9,500 per 100,000 high school students), a gap of roughly three orders of magnitude in the wrong direction.\n","rough_estimate":"Most parents would guess far below 1 in 100 per year; the actual figure is roughly 1 in 10","kind":"intuition"},"native":{"display":"~9.5% of HS students attempted suicide in the past year (2023)","numerator":95,"denominator":1000,"unit":"per year (high school students)","population":"US high school students, grades 9-12 (2023 YRBS)"},"normalized":{"lifetime_us_adult":0.18,"display":"~1 in 6 across adolescence (ages 12-17)","log_value":-0.745,"assumptions":"The YRBS 2023 reports a 9.5% past-year attempt prevalence among HS students (grades 9-12). The NSDUH 2023 reports a 3.3% past-year attempt prevalence among all adolescents aged 12-17. The NSDUH figure is lower because it includes younger teens (12-13) with lower rates and uses a household interview rather than a school-based anonymous survey. For a cumulative \"at least once during adolescence\" estimate, naively compounding the NSDUH rate over 6 years (ages 12-17) yields 1 - (1 - 0.033)^6 = 0.183, but this overstates cumulative incidence because many teen attempters are repeat attempters across years — the same individuals appear in multiple annual cross-sections. Conversely, the NSDUH rate likely undercounts single-occasion attempters due to household-interview social desirability bias. The two biases partially offset. We use 0.18 (~1 in 6) as the central estimate for the probability that a US adolescent attempts suicide at least once between ages 12 and 17. The scope is subgroup_lifetime because this applies only to the adolescent window, not the full adult lifespan.\n","uncertainty":{"low":0.12,"high":0.25},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/73/su/su7304a9.htm","title":"Mental Health and Suicide Risk Among High School Students and Protective Factors — Youth Risk Behavior Survey, United States, 2023","publisher":"CDC MMWR Supplements (Ivey-Stephenson et al.)","source_type":"govt_report","statistic":"9.5% of US high school students attempted suicide in the past 12 months; 20.4% seriously considered it; 15.7% made a plan; female 13% vs male 6% attempted; LGBQ+ students 19.7% vs heterosexual 6% attempted; transgender students 25.9%","excerpt":"\"Overall, 20.4% seriously considered attempting suicide, and 9.5% had attempted suicide... The prevalence of mental health and suicide risk indicators was high across all demographic groups; however, prevalence was highest among female students and LGBQ+ students.\"\n","source_date":"2024-10-24","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260325154712/https://www.cdc.gov/mmwr/volumes/73/su/su7304a9.htm","calculation_notes":"The YRBS is a nationally representative, school-based, anonymous survey of US high school students conducted biennially by the CDC since 1991. In 2023, the survey captured approximately 20,000 students across grades 9-12. The 9.5% past-year attempt prevalence is the primary native rate. By sex: female students 13%, male students 6% (approximate 2:1 ratio). By sexual identity: LGBQ+ students 19.7%, heterosexual students 6% (approximate 3.3:1 ratio). Transgender students 25.9%, cisgender females 11.0%, cisgender males 5.3%. By race: Black students 10% (down from 14% in 2021), White students approximately 8%, Hispanic students approximately 10%. These are annual period-prevalence figures, not cumulative lifetime figures.\n"},{"url":"https://www.samhsa.gov/data/sites/default/files/reports/rpt56770/2024-nsduh-psr6-adol-mde-suicide.pdf","title":"2023 NSDUH Population Statistics Report: Adolescent Major Depressive Episodes and Suicidal Thoughts and Behaviors","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"3.2 million adolescents aged 12-17 (12.3%) had serious thoughts of suicide in 2023; 856,000 (3.3%) attempted suicide; 4.5 million (18.1%) had a major depressive episode","excerpt":"\"In 2023, among adolescents aged 12 to 17, 3.2 million (or 12.3 percent) had serious thoughts of suicide... nearly half of those who had serious thoughts of suicide made suicide plans or attempted suicide, or both.\"\n","source_date":"2024-08-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260116034915/https://www.samhsa.gov/data/sites/default/files/reports/rpt56770/2024-nsduh-psr6-adol-mde-suicide.pdf","calculation_notes":"NSDUH is a household-interview survey of approximately 70,000 persons aged 12+, conducted annually by SAMHSA. The 3.3% past-year attempt rate among 12-17 year-olds is lower than the YRBS 9.5% for several reasons: (1) NSDUH includes younger adolescents (12-13) with lower rates; (2) household interview format introduces social desirability bias (teen answers with parent potentially nearby); (3) the YRBS school-based anonymous questionnaire elicits higher self-report of sensitive behaviors. Both surveys are nationally representative. The NSDUH figure is used for the cumulative adolescence estimate because it covers the full 12-17 age range rather than the 14-18 high-school window.\n","independence_note":"Fully independent of the YRBS. Different sampling frame (household vs school), different survey instrument, different administering agency (SAMHSA vs CDC).\n"},{"url":"https://publications.aap.org/pediatrics/article/153/1/e2023064800/196189/Suicide-and-Suicide-Risk-in-Adolescents","title":"Suicide and Suicide Risk in Adolescents","publisher":"Pediatrics (American Academy of Pediatrics, Shain et al. 2024)","source_type":"peer_reviewed","statistic":"AAP clinical report summarising epidemiology, risk factors, and screening recommendations for adolescent suicidality; notes suicide as the second leading cause of death for 10- to 24-year-olds in the US and recommends pediatricians screen patients 12 years and older for suicide risk at least annually","excerpt":"\"Suicide is the second leading cause of death for 10- to 24-year-olds in the United States and is a global public health issue.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20240718202408/https://publications.aap.org/pediatrics/article/153/1/e2023064800/196189/Suicide-and-Suicide-Risk-in-Adolescents","calculation_notes":"The AAP clinical report provides the clinical and epidemiological framing for adolescent suicidality, opening with the fact that suicide is the second leading cause of death for 10- to 24-year-olds in the US. It corroborates the YRBS and NSDUH figures and recommends annual suicide-risk screening for patients 12 and older. The AAP report does not provide independent incidence data but synthesises the CDC and SAMHSA sources with clinical literature on risk factors, screening tools (PHQ-A, Columbia Protocol), and lethal means restriction. The body text compares the CDC YRBS past-year attempt prevalence (~9,500 per 100,000 high school students) against the CDC age-15-19 suicide death rate (~7 per 100,000), a comparison of a self-reported past-year attempt rate against an annual death rate rather than a clinical attempt-to-death case ratio.\n","independence_note":"Review article synthesising CDC and SAMHSA data with clinical literature. Not an independent data source for the headline prevalence figure.\n"},{"url":"https://www.thetrevorproject.org/survey-2024/","title":"2024 U.S. National Survey on the Mental Health of LGBTQ+ Young People","publisher":"The Trevor Project","source_type":"reputable_reference","statistic":"12% of LGBTQ+ youth aged 13-24 attempted suicide in the past year (2024 survey of 18,000+ respondents); 46% of transgender and nonbinary youth seriously considered attempting suicide","excerpt":"\"More than one in ten LGBTQ+ young people attempted suicide in the past year.\"\n","source_date":"2024-04-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260513184025/https://www.thetrevorproject.org/survey-2024/","calculation_notes":"The Trevor Project survey is a large convenience sample (not probability-based) of LGBTQ+ youth aged 13-24 recruited online. The 12% past-year attempt rate is broadly consistent with the YRBS finding that 19.7% of LGBQ+ high school students attempted suicide — the Trevor Project figure is lower because it includes older youth (18-24) with lower attempt rates. The convenience-sample design may introduce selection bias in either direction. This source is included to document the LGBTQ+ disparity with the largest available sample, not as an independent prevalence estimate. The 50-state report released in March 2025 provided state-level breakdowns of these figures.\n","independence_note":"Fully independent of CDC YRBS and SAMHSA NSDUH. Different survey instrument, different sampling frame (online convenience sample of LGBTQ+ youth vs probability-based surveys of all youth).\n"}],"comparison_anchors":[{"label":"Suicide death (lifetime, US adult)","lifetime_us_adult":0.00827},{"label":"Drug overdose death (lifetime)","lifetime_us_adult":0.014}],"regional_breakdown":[{"region":"Female HS students","probability":0.13,"notes":"Past-year attempt rate, YRBS 2023"},{"region":"Male HS students","probability":0.06,"notes":"Past-year attempt rate, YRBS 2023"},{"region":"LGBQ+ HS students","probability":0.197,"notes":"Past-year attempt rate, YRBS 2023"},{"region":"Transgender HS students","probability":0.259,"notes":"Past-year attempt rate, YRBS 2023"},{"region":"Black HS students","probability":0.1,"notes":"Past-year attempt rate, YRBS 2023; down from 14% in 2021"},{"region":"Heterosexual HS students","probability":0.06,"notes":"Past-year attempt rate, YRBS 2023"}],"personal_factor_multipliers":[{"factor":"Female (vs overall)","multiplier":1.37,"notes":"13% vs 9.5% overall; YRBS 2023"},{"factor":"Male (vs overall)","multiplier":0.63,"notes":"6% vs 9.5% overall; YRBS 2023"},{"factor":"LGBQ+ (vs heterosexual)","multiplier":3.28,"notes":"19.7% vs 6.0%; YRBS 2023"},{"factor":"Transgender (vs cisgender male)","multiplier":4.89,"notes":"25.9% vs 5.3%; YRBS 2023"},{"factor":"History of prior attempt","multiplier":3,"notes":"Prior attempt is the strongest single predictor of future attempt; AAP 2024 clinical report"}],"short_label":"Teen suicide attempt","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The YRBS and NSDUH use different methodologies, populations, and definitions, which is why their headline numbers differ (9.5% vs 3.3%). Neither survey verifies self-reported attempts against medical records — some reported \"attempts\" may be better classified as non-suicidal self-injury (NSSI), and some genuine attempts go unreported even on anonymous surveys. The cumulative adolescence figure (0.18) is an estimate, not a directly measured quantity; no US survey tracks the same cohort of adolescents longitudinally with annual attempt assessment, so the true cumulative incidence is uncertain. The 2023 YRBS showed modest improvement from the 2021 peak for several indicators — the percentage of students who seriously considered attempting suicide fell from 22% to 20%, and Black student attempt rates fell from 14% to 10% — but levels remain historically elevated compared to the 2009-2013 baseline. All figures are self-report and subject to recall bias, social desirability effects, and definitional ambiguity about what constitutes an \"attempt.\"\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty school hallway with sunlight streaming through windows, symbolizing the quiet prevalence of teen mental health struggles"},"canonical_url":"https://likelier.app/teen-suicide-attempt","api_url":"https://likelier.app/api/fears/teen-suicide-attempt.json"},{"slug":"visa-overstay-deportation","question":"What are the odds of being detained and deported from the US after overstaying a visa?","quick_answer":"The annual removal rate for a visa overstayer is roughly 1.6%, about 1 in 63, compounding to an estimated ~18% chance of eventual removal over a 10-year overstay. Many overstayers fear a yearly risk of 1 in 3 to 1 in 10, far above the actuarial rate.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"The fear of deportation after a visa overstay is perceived as much higher than the actuarial rate, driven by media coverage of ICE enforcement operations and a general sense that immigration authorities track and remove overstayers systematically. Conversely, some overstayers believe they are effectively invisible after a few years and face minimal practical risk. Neither extreme is well-calibrated. ICE does not publicly break out overstay-only removal figures, and DHS reports focus on tracking overstays rather than measuring enforcement rates against them. The documented annual removal figure for all undocumented immigrants (roughly 150,000-270,000 per year in recent years) is spread across a population of 11-14 million, implying an annual removal rate in the range of 1-2% regardless of entry method. For overstayers specifically, there is no official published enforcement rate, which itself is a signal: overstay enforcement is a low-priority use of limited ICE interior resources.\n","rough_estimate":"perceived as 1 in 3 to 1 in 10 per year by many overstayers; actual annual rate is closer to 1 in 50 or lower","kind":"intuition"},"native":{"display":"~399,708 suspected in-country overstays remaining as of May 2024 (FY2023 new overstays); ~1.02% of expected departures","numerator":399708,"denominator":39005712,"unit":"per year (new suspected in-country overstays, FY2023)","population":"Nonimmigrant visitors with expected departures in FY2023"},"normalized":{"lifetime_us_adult":0.18,"display":"~18% chance of eventual removal over a 10-year overstay period (central estimate)","log_value":-0.74,"assumptions":"Step 1 — Annual overstay rate: DHS Entry/Exit Overstay Report FY2023 documents 510,363 total suspected in-country overstays from FY2023 arrivals; this fell to 399,708 by May 2024 as some departed. Annual new suspected in-country overstays run approximately 400,000-850,000 per year (FY2016-FY2022 range), and the cumulative overstay population is estimated at approximately 4-5 million people (roughly 40-42% of the ~11-14 million total undocumented population, per CRS and Pew). Step 2 — Annual enforcement rate against overstayers: ICE does not publish overstay-specific removal figures. Total ICE interior removals were approximately 142,580 in FY2023 and 248,739 in FY2024. These figures include all grounds of removal, not just overstays. Applying the ~40% overstayer share of the undocumented population to the removal total suggests roughly 57,000-100,000 overstayer removals per year (central estimate ~70,000), against a cumulative overstay population of approximately 4.5 million. Annual hazard per overstayer: ~70,000 / 4,500,000 ≈ 1.6%, or roughly 1 in 63. Step 3 — Compounded over a notional 10-year overstay period: 1 - (1 - 0.016)^10 ≈ 0.15-0.18, or approximately 15-18%. Over 20 years: 1 - (1 - 0.016)^20 ≈ 0.27-0.28. Central estimate of 18% reflects the 10-year horizon, which captures the median overstay duration before departure or status change. Note: the annual enforcement rate is highly sensitive to political administration (Obama-era peak enforcement would raise the rate to ~3-4%; Biden-era lows would reduce it to ~1%). The scope is subgroup_lifetime because this risk applies only to the subgroup of people currently in visa overstay status.\n","uncertainty":{"low":0.05,"high":0.45},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.dhs.gov/sites/default/files/2024-10/24_1011_CBP-Entry-Exit-Overstay-Report-FY23-Data.pdf","title":"Entry/Exit Overstay Report Fiscal Year 2023 Report to Congress","publisher":"U.S. Customs and Border Protection, Department of Homeland Security","source_type":"govt_report","statistic":"510,363 total suspected in-country overstays for FY2023; fell to 399,708 (1.02% of expected departures) by May 2024; overall FY2023 overstay rate was 1.31%","excerpt":"\"At the end of Fiscal Year 2023, the overall Suspected In-Country Overstay number was 510,363 or 1.31 percent, which further decreased to 399,708 or 1.02 percent by the end of April 2024.\"\n","source_date":"2024-10-11","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260416155049/https://www.dhs.gov/sites/default/files/2024-10/24_1011_CBP-Entry-Exit-Overstay-Report-FY23-Data.pdf","calculation_notes":"The DHS FY2023 overstay report provides the annual flow of new suspected overstays (510,363) and the number remaining in-country after partial departure attrition (399,708 by May 2024). These are new-year entrants only, not the cumulative overstay population. The 39,005,712 expected departures figure (from the FY2023 report) is used as the denominator in the native display. For removal rate calculation, the relevant population is the cumulative ~4-5 million overstay population, not the annual new entrant flow.\n"},{"url":"https://www.ice.gov/news/releases/ice-releases-fiscal-year-2023-annual-report","title":"ICE Releases Fiscal Year 2023 Annual Report","publisher":"U.S. Immigration and Customs Enforcement","source_type":"govt_report","statistic":"ERO conducted 142,580 removals in FY2023; overstay-specific removal figures are not published separately","excerpt":"\"ERO conducted 142,580 removals and 62,545 Title 42 expulsions to more than 170 countries worldwide in Fiscal Year 2023.\"\n","source_date":"2024-03-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260405003525/https://www.ice.gov/news/releases/ice-releases-fiscal-year-2023-annual-report","calculation_notes":"ICE FY2023 removals (142,580) cover all grounds of removal: prior criminal conviction, immigration violations including overstays, entry without inspection, and pending criminal charges. ICE does not publish a breakdown of how many removals were specifically for visa overstay violations. Using the ~40% overstayer share of the undocumented population (from CRS R47848) as a proxy: 142,580 * 0.40 ≈ 57,000 overstayer removals per year. This is a rough approximation; in practice, ICE prioritizes criminal records over immigration-only violations, meaning the actual overstay-removal fraction is likely below the 40% population share. Central estimate: ~50,000-70,000 overstay removals per year.\n"},{"url":"https://www.congress.gov/crs-product/R47848","title":"Nonimmigrant Overstays: Overview and Policy Issues","publisher":"Congressional Research Service","source_type":"govt_report","statistic":"Approximately 42% of the ~11 million unauthorized population entered legally and overstayed, equaling ~5 million people; only a fraction of overstayers each year are targeted for enforcement","excerpt":"\"Approximately 42% of the roughly 11 million unauthorized population entered legally but overstayed their visa period. Congressional frustration that only a fraction of the thousands of people who overstay each year are targeted for enforcement despite government knowledge of violations.\"\n","source_date":"2023-11-21","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260414085947/https://www.congress.gov/crs-product/R47848","calculation_notes":"The CRS report provides the critical denominator estimate: ~5 million cumulative overstayers (42% of ~11 million undocumented population). The report explicitly notes that enforcement against overstayers is sparse relative to the population, which is consistent with the low annual removal rate derived from ICE data. The 42% figure is used to apportion ICE total removals to the overstayer subgroup for the annual hazard calculation. This is an imperfect approximation; criminal history and geographic location matter more than entry method in determining enforcement priority.\n"}],"comparison_anchors":[{"label":"Lifetime odds of being audited by the IRS (US adult)","lifetime_us_adult":0.1},{"label":"Deportation risk for undocumented immigrants broadly over 40 years","lifetime_us_adult":0.55},{"label":"Lifetime odds of filing for bankruptcy (US adult)","lifetime_us_adult":0.08}],"personal_factor_multipliers":[{"factor":"Prior criminal conviction in the US","multiplier":6,"notes":"ICE enforcement strongly prioritizes individuals with criminal records; a conviction dramatically increases removal probability"},{"factor":"Prior removal order outstanding","multiplier":8,"notes":"An outstanding final order of removal means any ICE encounter results in near-certain removal"},{"factor":"No criminal record, long-term US resident, non-border state","multiplier":0.2,"notes":"Interior enforcement against non-criminal overstayers with no prior contact is historically low; many long-term overstayers effectively age out of active enforcement priority"},{"factor":"High-enforcement administration (e.g. Trump-era peak)","multiplier":2.5,"notes":"Enforcement posture can roughly double or triple the annual removal rate"}],"short_label":"Visa overstay deportation","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"autonomy_loss","valence":"negative","caveats":"The annual overstay enforcement rate is the most uncertain quantity in this entry, because ICE does not publish removal figures broken out by entry method (overstay vs. entered without inspection). The central estimate of ~1.5-2% per year is derived by apportioning total ICE interior removals using the overstayer share of the undocumented population, which is itself uncertain. The political administration in power is the single largest driver of individual removal risk: enforcement has varied roughly threefold across recent administrations, from Biden-era lows to Trump-era highs. The 10-year horizon used for the lifetime figure (18%) is arbitrary; many overstayers depart voluntarily or adjust to legal status within a few years, and others remain for decades. Long-duration overstayers who have built family and community ties in the US face qualitatively different enforcement dynamics than recent overstayers. The DHS Entry/Exit Overstay Report tracks suspected in-country overstays, not confirmed overstays; the \"suspected\" designation means some fraction have departed without being counted. This entry should not be read as legal advice; the actual enforcement risk for any individual depends heavily on criminal history, prior removal orders, and geographic location.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"civil-fears-agent-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A passport and boarding pass resting on a plain surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/visa-overstay-deportation","api_url":"https://likelier.app/api/fears/visa-overstay-deportation.json"},{"slug":"cte-football","question":"What are the odds of developing CTE as a former football player?","quick_answer":"For a former NFL career player, the risk of developing CTE is roughly 1 in 5. This is an order-of-magnitude estimate rather than a measured rate, since CTE can currently only be diagnosed after death. The widely cited 99% figure comes from donated brains selected for symptoms and is not the population rate.\n","category":"health","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"Until roughly 2015, chronic traumatic encephalopathy was a niche term inside sports medicine and a courtroom word inside NFL concussion litigation. The 2017 Mez et al. JAMA paper — CTE neuropathology in 110 of 111 donated NFL brains — moved it into the mainstream, and the headline number was vivid enough that it reshaped how former players, parents of youth athletes, and the general public talk about the risk. The current intuition among former career football players is no longer that CTE is a rare unlucky tail event; it is closer to \"something that happens to a meaningful share of us, and we do not yet know which share.\" That intuition is roughly the right shape, even though the 99% figure is not the population rate and the true rate is still unmeasurable in living brains.\n","rough_estimate":"Former career players now treat CTE as a meaningful personal risk rather than a rare tail event","kind":"intuition"},"native":{"display":"~1 in 5 former NFL career players (wide uncertainty)","numerator":1,"denominator":5,"unit":"lifetime","population":"former NFL career players (order-of-magnitude estimate)"},"normalized":{"lifetime_us_adult":0.2,"display":"~1 in 5 lifetime (former NFL career player)","log_value":-0.699,"assumptions":"The headline 1-in-5 for a former NFL career player is an order-of-magnitude estimate, not a measured rate. It cannot be measured directly because as of 2026 there is no validated in-vivo biomarker for CTE — the neuropathological diagnosis requires post-mortem tau immunohistochemistry, and most former players will never have their brain examined.\nTwo data points anchor the estimate from opposite directions. The upper bound comes from Mez et al. (JAMA 2017), which reported CTE neuropathology in 177 of 202 donated brains from former football players (87%), including 110 of 111 former NFL players (99%) and 48 of 53 former college players (91%). This is almost certainly a massive overestimate of the population rate: the Boston University brain bank receives donations preferentially from families of former players with symptoms, and Mez and colleagues explicitly warn against interpreting the 99% figure as a prevalence rate. The lower bound is hard to pin down because no large unselected cohort of former players has been followed to autopsy.\nThe Nguyen et al. (American Journal of Epidemiology 2022) formal selection-bias analysis of the Mez dataset estimates that college-level and professional football players face roughly 2.4-2.5 times the CTE risk of high-school-only players after adjustment — a dose-response relationship that is stronger, not weaker, once selection bias is modelled. Working backward from plausible high-school-only population rates and applying that multiplier puts a career NFL player's lifetime CTE probability somewhere in the low tens of percent, with a wide band. Best-guess point estimate 0.2 (1 in 5) with an uncertainty interval of 0.1 to 0.5 that reflects the genuine gap between the lowest plausible population rate and the highest plausible selection-bias-adjusted rate. Scope is subgroup_lifetime because this number describes former NFL career players specifically, not US adults, and not former football players of all levels.\n","uncertainty":{"low":0.1,"high":0.5},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/28742910/","title":"Clinicopathological Evaluation of Chronic Traumatic Encephalopathy in Players of American Football","publisher":"JAMA (Mez, Daneshvar, Kiernan, Abdolmohammadi, Alvarez, Huber, Alosco et al. 2017)","source_type":"peer_reviewed","statistic":"CTE was neuropathologically diagnosed in 177 of 202 (87%) deceased former football players donated to the VA-BU-CLF brain bank, including 110 of 111 former National Football League (NFL) players (99%), 48 of 53 college players (91%), and 3 of 14 high school players (21%)","excerpt":"\"CTE was neuropathologically diagnosed in 177 players (87%... 110 of 111 National Football League (99%) players.\"\n","source_date":"2017-07-25","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165426/https://pubmed.ncbi.nlm.nih.gov/28742910/","calculation_notes":"Mez et al. 2017 is the single most-cited data point on CTE in football players and is the source of the 99%-of-NFL-brains figure. It is NOT the source of the headline 1-in-5 normalized estimate, because the Mez sample is a brain bank convenience sample, not a population cohort. The paper’s authors explicitly warn that selection bias makes the 99% figure an upper bound on the true population rate among former NFL players. The regional_breakdown rows for subgroup level (NFL career, D1 college career, high school only) are order-of-magnitude estimates informed by, but not derived directly from, the Mez numerators — they also incorporate the dose-response multipliers from Nguyen et al. 2022 below.\n","independence_note":"All post-2013 CTE neuropathology in the United States flows through a small number of brain banks, with the VA-BU-CLF bank (Boston University CTE Center) the dominant source for former football players. Mez et al. and the follow-up selection-bias analysis (Nguyen et al. 2022) share that underlying cohort, so treat the two citations as one dataset with two different analytic frames rather than as independent measurements.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35434739/","title":"Relationship Between Level of American Football Playing and Diagnosis of Chronic Traumatic Encephalopathy in a Selection Bias Analysis","publisher":"American Journal of Epidemiology (Nguyen, Alosco, Mez, Tripodis et al. 2022)","source_type":"peer_reviewed","statistic":"After adjustment for selection bias in the VA-BU-CLF brain bank cohort, college-level and professional football players had 2.38 (95% simulation interval: 1.16, 5.94) and 2.47 (95% simulation interval: 1.46, 4.79) times the risk of a CTE diagnosis, respectively, compared to high-school-level players","excerpt":"\"After adjustment for selection bias, college-level and professional football players had 2.38 (95% simulation interval (SI): 1.16, 5.94) and 2.47 (95% SI: 1.46, 4.79) times the risk of being diagnosed with CTE as high-school-level players, respectively; these estimates are larger than estimates with no selection bias adjustment.\"\n","source_date":"2022-04-19","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165459/https://pubmed.ncbi.nlm.nih.gov/35434739/","calculation_notes":"Nguyen et al. 2022 is the methodologically important follow-up to Mez 2017: it takes the same brain bank cohort and applies quantitative bias analysis to the selection process. The key finding for this entry is that the dose-response relationship between total football exposure and CTE diagnosis is stronger after adjusting for selection bias, not weaker — i.e. the data do not support the hypothesis that the Mez findings are entirely a selection artefact. A career college player has ~2.4× the CTE risk of a high-school-only player, and a career NFL player has ~2.5×. Those multipliers feed the regional_breakdown rows directly and are the basis for estimating that a career NFL player’s lifetime CTE probability sits in the low-tens-of-percent range even after the most aggressive reasonable adjustment for Boston University brain bank selection bias.\n","independence_note":"Uses the same VA-BU-CLF brain bank cohort as Mez et al. 2017. Not an independent measurement; included as the authoritative methodological correction to the raw Mez numbers and the source of the dose-response multipliers used in this entry.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25632088/","title":"Age of First Exposure to Football and Later-Life Cognitive Impairment in Former NFL Players","publisher":"Neurology (Stamm, Bourlas, Baugh, Fritts, Daneshvar, Martin, McClean, Tripodis, Stern 2015)","source_type":"peer_reviewed","statistic":"Former NFL players who began playing tackle football before age 12 performed significantly worse on measures of executive function, memory, and estimated verbal IQ than former NFL players who began playing at age 12 or later; the effect was independent of total years of play and age at testing","excerpt":"\"There is an association between participation in tackle football prior to age 12 and greater later-life cognitive impairment measured using objective neuropsychological tests.\"\n","source_date":"2015-01-28","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165532/https://pubmed.ncbi.nlm.nih.gov/25632088/","calculation_notes":"Stamm et al. 2015 is the source for the \"youth football before age 12\" factor in the personal_factor_multipliers block below. The paper does not directly produce a CTE incidence number; it shows that the age window during which repetitive head impacts are received matters independently of total exposure, which supports a non-linear dose-response structure rather than a simple hours-of-contact-play model. Subsequent work (notably Caccese et al. 2020 and follow-up work from the same BU group) has replicated the age-12 effect on some cognitive outcomes and failed to replicate it on others; it is well-established enough to cite as a multiplier but not strong enough to anchor a population probability.\n","independence_note":"Same VA-BU-CLF / DETECT research pipeline as Mez et al. 2017 and Nguyen et al. 2022. Partially dependent, used for the age-of-first-exposure multiplier rather than as an independent population measurement.\n"},{"url":"https://www.bu.edu/articles/2017/cte-former-nfl-players/","title":"BU Researchers Find CTE in 99% of Former NFL Players Studied","publisher":"Boston University (Barlow 2017)","source_type":"reputable_reference","statistic":"177 of 202 deceased football players donated to the VA-BU-CLF brain bank had CTE (87%); 110 of 111 former NFL players had CTE (99%). The researchers explicitly warned the numbers should not be interpreted as a population prevalence rate","excerpt":"\"The study has several important limitations, most notably the lack of a control group, and selection bias in the brain collection itself — families of players with symptoms of CTE are far more likely to donate brains to research than those without signs of the disease.\"\n","source_date":"2017-07-25","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260315212822/https://www.bu.edu/articles/2017/cte-former-nfl-players/","calculation_notes":"The Boston University communications article accompanying the Mez 2017 JAMA paper is the source of the verbatim selection-bias caveat that anchors the caveats block below. It is included specifically because the methodological warning it contains — that the 99% figure reflects a self-selected donor population, not all former NFL players — is routinely dropped from secondary coverage of the Mez findings, and readers of this page need that context to understand why the normalized figure is 1 in 5 rather than the raw 99%.\n","independence_note":"Boston University is the institutional home of the CTE Center, the VA-BU-CLF brain bank, and the authors of Mez 2017, Nguyen 2022, and Stamm 2015. This page is a university communications piece covering its own researchers and shares the same upstream cohort as the other three citations. Included for the verbatim caveat wording rather than as an independent measurement.\n"},{"url":"https://onlinelibrary.wiley.com/doi/full/10.1111/bpa.12757","title":"Association between contact sports participation and chronic traumatic encephalopathy: a retrospective cohort study","publisher":"Brain Pathology / Bieniek KF et al. (Mayo Clinic)","source_type":"peer_reviewed","statistic":"Among 300 contact sport athletes in the Mayo Clinic Tissue Registry, American football had the highest CTE frequency (15% of cases) with OR 2.62 (P=0.005); CTE pathology was not exclusive to athletes — 15 of 450 non-athletes also showed combined CTE pathology (six CTE-positive, nine with features of CTE)","excerpt":"\"For contact sports, American football had the highest frequency of chronic traumatic encephalopathy pathology (15% of cases) and an odds ratio of 2.62 (P-value = 0.005).\"\n","source_date":"2020-01-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250703004527/https://onlinelibrary.wiley.com/doi/full/10.1111/bpa.12757","calculation_notes":"Mayo Clinic brain bank finding of 15% CTE in football players is substantially lower than the BU brain bank's 99% (Mez 2017), consistent with the selection bias caveat in the BU work (symptomatic families donate disproportionately). The 15% figure from a population-based autopsy registry is a more conservative anchor for the true prevalence.\n","independence_note":"Fully independent of the BU/VA-BU-CLF brain bank — different institution (Mayo Clinic), different tissue registry, different geographic catchment, population-based rather than self-referred.\n"},{"url":"https://onlinelibrary.wiley.com/doi/10.5694/mja2.51420","title":"Chronic traumatic encephalopathy in Australia: the first three years of the Australian Sports Brain Bank","publisher":"Medical Journal of Australia / Suter CM, Affleck AJ, Lee M, Pearce AJ, Iles LE, Buckland ME","source_type":"peer_reviewed","statistic":"12 of 21 completed donations had pathognomonic CTE lesions; mean age at death 48 years for CTE donors vs 75 for non-CTE","excerpt":"\"The most frequent neuropathology was CTE: 12 donors had pathognomonic CTE lesions.\"\n","source_date":"2022-05-01","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20250705205419/https://onlinelibrary.wiley.com/doi/10.5694/mja2.51420","calculation_notes":"Australian Sports Brain Bank confirms CTE is not limited to American football or the US. Found CTE in Australian Rules football, rugby league, and rugby union players, broadening the evidence base beyond the BU/VA-BU-CLF pipeline.\n","independence_note":"Fully independent of both BU and Mayo — different country, different sports, different brain bank, different research group.\n"}],"comparison_anchors":[{"label":"Hospitalization-level TBI, US adult lifetime","lifetime_us_adult":0.03},{"label":"All-cause dementia mortality, global adult lifetime","lifetime_us_adult":0.12},{"label":"Death in a car crash, US adult lifetime","lifetime_us_adult":0.0108},{"label":"Death in a plane crash, US adult lifetime, regular flyer","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Former NFL career player","probability":0.2,"notes":"Order-of-magnitude estimate with wide uncertainty. The raw autopsy-series figure is 110 of 111 donated NFL brains (99%) per Mez et al. JAMA 2017, but that cohort is a self-selected symptomatic donor population. The true population rate is unmeasurable in 2026 because there is no validated in-vivo biomarker for CTE. Working from the Nguyen et al. 2022 selection-bias-adjusted dose-response (career NFL vs high school ~2.5×) and a plausible high-school-only base rate, a career NFL player’s lifetime CTE probability lands somewhere in the low-tens-of-percent range. 1 in 5 is the headline, with genuine uncertainty from 1 in 10 to 1 in 2."},{"region":"Former college football player (D1, career)","probability":0.05,"notes":"Mez 2017 found CTE neuropathology in 48 of 53 former college players donated to the BU brain bank (91%), again with the same massive selection bias. Nguyen et al. 2022 dose-response puts college career players at ~2.4× the CTE risk of high school-only players. The 5% figure is an order-of-magnitude estimate for a D1 career player — four years of contact football at the highest amateur level, multiple diagnosed concussions typical, thousands of subconcussive impacts."},{"region":"Former high school-only player","probability":0.01,"notes":"Mez 2017 found CTE in 3 of 14 donated brains from former high-school-only players (21%), which — even accounting for selection bias — suggests a non-zero population rate. Any specific percentage here is a guess; 1% is the order-of-magnitude placeholder and could be meaningfully higher or lower. The key qualitative finding from Nguyen et al. 2022 is that the dose-response with total years of contact play is real and monotonic, so a single high school career is genuinely lower-risk than a college career, which is genuinely lower-risk than an NFL career."},{"region":"Never played contact sport","probability":0.001,"notes":"CTE neuropathology is occasionally reported in individuals with no documented repetitive head impact history, but it is genuinely rare. The 0.1% figure is a baseline placeholder representing the tail of the risk distribution — people with undiagnosed impact histories, military blast exposure outside a formal record, or unusual susceptibility. The population rate in never-exposed adults is not well-characterized."}],"personal_factor_multipliers":[{"factor":"career length 10+ years tackle football","multiplier":3,"notes":"Total years of contact play is the strongest single predictor of CTE neuropathology in the Mez et al. and Nguyen et al. cohorts. Each additional year of play is associated with higher odds of CTE diagnosis and with more severe pathology among those diagnosed. A 10+ year career roughly triples risk relative to a shorter career at the same level of play."},{"factor":"multiple diagnosed concussions","multiplier":2,"notes":"Diagnosed symptomatic concussions are an imperfect but elevated marker of CTE risk. The stronger driver in the literature is total subconcussive impact exposure — repeated sub-symptomatic hits that never get diagnosed — which is why linemen (below) carry elevated risk even when they never had a concussion checked on the sideline."},{"factor":"offensive or defensive lineman (high subconcussive exposure)","multiplier":2,"notes":"Interior linemen experience the highest per-play head impact counts of any football position, with helmet-to-helmet contact on virtually every snap. The resulting subconcussive impact load is thought to drive CTE risk more than the lower-frequency, higher-magnitude concussive events that dominate sideline sports-medicine protocols."},{"factor":"youth football participation before age 12","multiplier":2,"notes":"Stamm et al. (Neurology 2015) found that former NFL players who began tackle football before age 12 performed significantly worse on cognitive tests than those who started at 12 or older, independent of total years of play. The biological rationale is that ages 10-12 are a sensitive period for corpus callosum development. Subsequent replication of the age-12 effect has been mixed, which is why the multiplier here is a conservative 2× rather than larger."}],"short_label":"CTE (football)","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The central methodological fact on this page is that CTE cannot be reliably diagnosed in a living brain as of 2026. The diagnostic gold standard is post-mortem tau immunohistochemistry on specific brain regions; tau PET imaging is progressing in research cohorts but has not produced a validated clinical test. Every published CTE prevalence number therefore comes from a brain bank, which is by construction a self-selected sample. The Boston University VA-BU-CLF brain bank — the source of the Mez et al. 2017 JAMA paper and most subsequent CTE neuropathology research — receives donations preferentially from families of former players with cognitive, mood, or behavioural symptoms, because those are the families motivated to donate. The 99%-of-NFL-brains figure is not a population rate and the paper’s authors say so explicitly.\nThe second caveat is the distinction between CTE neuropathology (a specific tau protein deposition pattern) and CTE clinical syndrome (the cognitive, behavioural, and mood symptoms attributed to that pathology). The two overlap but are not the same. Neuropathologically-diagnosed CTE can be severe without obvious ante-mortem symptoms, and cognitive decline meeting symptomatic criteria for CTE can occur without the neuropathology on autopsy. This entry is scoped to neuropathological CTE because that is what the cited studies actually measure.\nThird: the dose-response with total years of contact play is reasonably well-established across Mez 2017, Nguyen 2022, and related work, and is the most load-bearing quantitative finding in the CTE literature. What is not established is the shape of that curve at low exposure, which is why the high-school-only row in the regional_breakdown is so uncertain. A single season of youth flag football is almost certainly low-risk; a career of NFL linemen play is almost certainly elevated. The middle of that range — one or two years of high school tackle, a single college season — is not well-measured in either direction.\nFinally, this entry will update as in-vivo biomarkers mature. If and when a validated tau PET or blood-based CTE biomarker reaches clinical deployment, the denominators on this page become measurable rather than estimated, and the 1-in-5 headline could move substantially in either direction.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single scuffed leather football laces detail resting on a pale neutral surface, flat vector illustration in muted colors."},"canonical_url":"https://likelier.app/cte-football","api_url":"https://likelier.app/api/fears/cte-football.json"},{"slug":"subprime-auto-loan-repossession","question":"If you take out a long-term subprime auto loan, how likely is the car to end up repossessed?","quick_answer":"About 1 in 5 long-term subprime auto loans ends in repossession over one loan cycle. Borrower perception is unanchored — most signing a long loan don't expect it, while some who've watched it happen nearby treat it as near-inevitable.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"No published survey measures how likely subprime borrowers think repossession is. Anecdotally, the dealership framing of a 72- or 84-month loan as 'affordable' because the monthly payment is low tends to make the end-state feel remote, while the borrower who has watched a neighbor's car get towed tends to overweight it. The honest read is that perception is unanchored in either direction; there is no rigorous measurement of it.","rough_estimate":"Most borrowers signing a long subprime loan do not expect repossession; a minority who have seen it happen nearby treat it as near-inevitable.","kind":"intuition"},"native":{"display":"about 1 in 5 long-term subprime auto loans ends in repossession","numerator":1,"denominator":5,"unit":"subprime borrower per 72-84 month loan","population":"US subprime auto borrowers (FICO below 620 at origination), one long-term loan cycle"},"normalized":{"lifetime_us_adult":0.2,"display":"~1 in 5 per loan","log_value":-0.699,"assumptions":"Subgroup: US subprime auto borrowers (FICO below 620 at origination), the tier where Cox Automotive put the September 2025 subprime default rate (cars 'repossessed or about to be') at nearly 10% per year. Horizon: ONE 72-84 month loan cycle, not a 59-year adult lifetime. Construction: take ~10%/yr as the subprime annual repossession hazard. Repossessions cluster in the first two to three years of a loan and many delinquent loans cure, refinance, or prepay, and a borrower whose car is repossessed exits the pool, so the per-loan cumulative figure is well below naive compounding (1-(1-0.10)^6.5 = 0.49 overstates). A cumulative per-loan repossession probability of ~0.20 is consistent with the CFPB 2021 finding that 25-40% of subprime loans from finance companies and buy-here-pay-here dealers reach 60+ day delinquency within three years, of which repossession is a large subset. Band 0.10-0.35 spans the bank-financed subprime tail (lower) and the buy-here-pay-here tail (higher). NOTE: this is per single loan; a borrower who originates several subprime loans over a lifetime faces a higher cumulative figure than 0.20.","uncertainty":{"low":0.1,"high":0.35},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.federalreserve.gov/econres/notes/feds-notes/subprime-auto-lending-trends-in-buy-here-pay-here-auto-lending-20260508.html","title":"Subprime Auto Lending: Trends in Buy Here Pay Here Auto Lending","publisher":"Board of Governors of the Federal Reserve System (FEDS Notes)","source_type":"govt_report","statistic":"Approximately 5% of BHPH balances were in active repossession in 2025:Q3, compared to less than half a percent for traditional auto lender balances; BHPH loans are 16.63 times more likely to be in active repossession status. ~78% of BHPH lending volume is originated to subprime borrowers vs 27% for traditional lenders.","excerpt":"approximately 5% of BHPH balances were in active repossession, compared to less than half a percent for traditional auto lender balances","source_date":"2026-05-08","source_accessed":"2026-06-13","calculation_notes":"Authoritative anchor for the high (buy-here-pay-here) tail of the subprime distribution. A 5% point-in-time active-repossession stock for the deepest-subprime channel, combined with the typical 2-3 week assignment-to-recovery window, implies a substantially higher annual repossession flow for that channel, bracketing the upper bound of the 0.10-0.35 cumulative band. Also supplies the 16.63x BHPH-vs-traditional personal factor multiplier.","independence_note":"Federal Reserve analysis of credit-panel data; independent of Cox Automotive and the rating agencies."},{"url":"https://kvia.com/news/business-technology/cnn-business-consumer/2025/10/22/a-significant-group-of-americans-are-falling-behind-on-their-car-payments-an-economic-warning-sign/","title":"A significant group of Americans are falling behind on their car payments - an economic warning sign (CNN Business, syndicated)","publisher":"CNN Business (via KVIA)","source_type":"news_article","statistic":"For subprime borrowers, the default rate (cars repossessed or about to be) stood at nearly 10% in September 2025 (Cox Automotive); the subprime 60+ day delinquency rate doubled since 2021 to 6.43% (Fitch Ratings); subprime defined as credit scores below 670.","excerpt":"For subprime borrowers, the default rate – those whose cars have been or are about to be repossessed – which stood at nearly 10% in September, according to Cox Automotive data.","source_date":"2025-10-22","source_accessed":"2026-06-13","calculation_notes":"This ~10% September subprime default rate is the annual-hazard anchor. It includes cars 'about to be' repossessed, so the completed-repossession annual flow is somewhat below 10%; this is absorbed by the conservative 0.20 cumulative headline and the 0.10 low bound. The 6.43% figure is a 60+ day delinquency snapshot, not repossession — repossession is a downstream subset, so it is used only as a supporting bound.","independence_note":"Cox Automotive default data plus Fitch Ratings delinquency data, reported by CNN; the default figure originates with Cox, the delinquency figure with Fitch."},{"url":"https://www.carscoops.com/2025/03/car-repossessions-return-to-great-recession-levels-just-in-time-for-another-one/","title":"Car Repossessions Return To Great Recession Levels As 1.7 Million Vehicles Taken Back","publisher":"Carscoops","source_type":"news_article","statistic":"An estimated 1.73 million vehicles were taken back in 2024 (Cox Automotive), the highest level since 2009; the repossession rate of 2.3% climbed to levels not seen in years.","excerpt":"The repossession rate, of 2.3%, also climbed to levels not seen in years.","source_date":"2025-03-26","source_accessed":"2026-06-13","calculation_notes":"Population-wide repossession rate (2.3% of all outstanding auto loans per year). Subprime is ~14% of outstanding balances but accounts for a disproportionate share of repossessions, so the subprime-conditional annual rate (~10%, per Cox/CNN) being roughly 4x the all-borrower 2.3% is internally consistent. Used as the population cross-check, not the subgroup headline.","independence_note":"Reports Cox Automotive figures; not independent of the Cox/CNN source for the underlying repossession data."},{"url":"https://www.cutoday.info/Fresh-Today/Negative-Equity-Deepens-As-Nearly-30-Of-Auto-Trade-Ins-Go-Underwater","title":"Negative Equity Deepens As Nearly 30% Of Auto Trade-Ins Go Underwater","publisher":"CUToday.info (reporting Edmunds data)","source_type":"reputable_reference","statistic":"In Q4 2025, 29.3% of trade-ins toward new vehicles were underwater; the average amount owed on underwater trade-ins reached a record $7,214; 27% of upside-down trade-ins carried at least $10,000 in negative equity. Edmunds attributes the imbalance partly to long loans originated during the 2021-2022 vehicle shortage.","excerpt":"In the fourth quarter of 2025, 29.3% of trade-ins toward new vehicles were underwater","source_date":"2026-01-19","source_accessed":"2026-06-13","calculation_notes":"Negative-equity context for the underwater half of the question. Not a repossession probability; it quantifies how many borrowers carry negative equity that makes a repossession deficiency balance likely. ~29% of all trade-ins underwater establishes that negative equity is common, and the $7,214 average is the typical shortfall that survives a repossession sale.","independence_note":"Edmunds transaction data; independent of Cox Automotive, the Federal Reserve, and the rating agencies."},{"url":"https://files.consumerfinance.gov/f/documents/cfpb_subprime-auto_data-point_2021-09.pdf","title":"Data Point: Subprime Auto Loan Outcomes by Lender Type","publisher":"Consumer Financial Protection Bureau","source_type":"govt_report","statistic":"The likelihood of a subprime auto loan becoming at least 60 days delinquent within three years is approximately 15 percent for bank borrowers and between 25 percent and 40 percent for finance company and buy-here-pay-here borrowers.","excerpt":"the likelihood of a subprime auto loan becoming at least 60 days delinquent within three years is approximately 15 percent for bank borrowers and between 25 percent and 40 percent for finance company and buy-here-pay-here borrowers","source_date":"2021-09-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260303214231/https://files.consumerfinance.gov/f/documents/cfpb_subprime-auto_data-point_2021-09.pdf","calculation_notes":"Three-year cumulative 60+ day delinquency by lender type. Repossession is a downstream subset of these delinquencies. The 15% (bank) vs 25-40% (finance/BHPH) split supports both the 0.10-0.35 cumulative band and the bank-vs-BHPH personal factor multiplier. Used as a supporting bound only.","independence_note":"CFPB analysis of a separate loan-level dataset; independent of Cox, Fitch, and the Federal Reserve note."}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Home burglary (US)","lifetime_us_adult":0.39},{"label":"Vehicle theft (US)","lifetime_us_adult":0.0296}],"personal_factor_multipliers":[{"factor":"Loan from a buy-here-pay-here dealer","multiplier":16.6,"notes":"Federal Reserve (May 2026) FEDS Note: buy-here-pay-here loan balances were 16.63 times more likely to be in active repossession than traditional auto-lender balances in Q3 2025 (about 5% vs under 0.5%). BHPH dealers originate ~78% of volume to subprime borrowers."},{"factor":"Subprime loan from a bank vs a finance company / BHPH","multiplier":0.5,"notes":"CFPB (2021) Data Point: ~15% of bank-originated subprime loans reach 60+ day delinquency within three years vs 25-40% for finance-company and buy-here-pay-here loans. Bank origination roughly halves the cumulative-delinquency rate that drives repossession."}],"short_label":"Subprime auto-loan repossession","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The headline is per single 72-84 month subprime loan, not per adult lifetime; a borrower who takes several subprime loans over a lifetime faces a higher cumulative figure. The strongest authoritative source that loaded cleanly (the Federal Reserve's May 2026 buy-here-pay-here note) measures the deepest-subprime channel specifically: ~5% of BHPH balances in active repossession at a point in time and a 16.63x gap vs traditional lenders. The broad all-subprime annual hazard (~10%, September 2025) rests on Cox Automotive's industry estimate reported by CNN, which is news-tier rather than peer-reviewed, and which lumps 'repossessed' with 'about to be repossessed,' so the completed-repossession flow is somewhat lower. The 6.9% record subprime 60+ day delinquency rate (Fitch, January 2026) is a delinquency snapshot, not a repossession rate, and many delinquent loans cure or refinance. The annual-to-cumulative conversion assumes defaults cluster early and the pool shrinks as loans repossess, refinance, or prepay; if subprime stress stays elevated longer than one cycle, the figure drifts toward the high end of the band. 'Subprime' is defined as FICO below 620 at origination here; some sources use below 670, which widens the population and would lower the per-borrower rate. The CFPB Data Point figure could not be confirmed by direct page load. D2 (authority) is honestly a 3, not a 4: the one cleanly-loaded authoritative source covers the BHPH tail, while the broad headline leans on an industry estimate.","quality_score":{"d1":4,"d2":3,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract car silhouette behind a descending bar, muted slate and dull rust, flat vector."},"canonical_url":"https://likelier.app/subprime-auto-loan-repossession","api_url":"https://likelier.app/api/fears/subprime-auto-loan-repossession.json"},{"slug":"depression-lifetime","question":"What are the odds of experiencing a major depressive episode in your lifetime?","quick_answer":"About 1 in 5 US adults — 20.6% — experience a major depressive episode at some point in life. That is roughly double the 5–10% most people would guess, with stigma pushing the intuitive estimate down.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"There is no standard tracker for perceived lifetime depression risk, and the question is awkwardly reflexive — asking someone to estimate their own probability of a condition partially defined by distorted self-perception. When lifetime prevalence figures are surfaced in surveys, most respondents express surprise at the 1-in-5 number. The lay mental model places \"clinical depression\" as something that happens to a distinct minority — perhaps 5-10% of the population — rather than to a fifth of all adults. Stigma compresses the intuitive estimate downward; so does the ordinary human tendency to classify one's own past low periods as \"just being sad\" rather than as episodes that would meet diagnostic criteria.\n","rough_estimate":"most adults would guess 5-10% lifetime, roughly half the actual figure","kind":"intuition"},"native":{"display":"20.6% lifetime prevalence (DSM-5 MDD, NESARC-III)","numerator":206,"denominator":1000,"unit":"lifetime prevalence","population":"US adults 18+, noninstitutionalized civilian (NESARC-III, 2012-2013)"},"normalized":{"lifetime_us_adult":0.206,"display":"~1 in 5 lifetime","log_value":-0.686,"assumptions":"Uses the Hasin et al. 2018 NESARC-III estimate of 20.6% lifetime prevalence of DSM-5 major depressive disorder among US adults as the headline figure. This is a direct survey-based lifetime prevalence — no compounding or hazard conversion required. The earlier NCS-R (Kessler et al. 2003) found 16.2% lifetime under DSM-IV criteria; the upward shift reflects both the DSM-5 bereavement-exclusion removal and secular trends in reporting and detection. NIMH reports that 21.0 million US adults (8.3%) had at least one major depressive episode in the past year (NSDUH 2021), consistent with the lifetime figure given recurrence and recovery patterns. The 0.206 point estimate is bracketed by the NCS-R lower bound (~0.16) and prospective-cohort evidence that lifetime risk is substantially higher when recall bias in retrospective surveys is removed: the Dunedin cohort (Moffitt et al. 2010) found that 41.4% of the cohort experienced at least one episode of depression between ages 18 and 32 under prospective ascertainment. The uncertainty band 0.16–0.30 keeps a conservative ceiling well below that prospective figure, reflecting the methodological range across retrospective surveys while treating the far-higher prospective estimate as an outer indication that 20.6% is a floor rather than a band edge.\n","uncertainty":{"low":0.16,"high":0.3},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/29450462/","title":"Epidemiology of Adult DSM-5 Major Depressive Disorder and Its Specifiers in the United States","publisher":"JAMA Psychiatry (Hasin DS, Sarvet AL, Meyers JL, et al.)","source_type":"peer_reviewed","statistic":"Lifetime prevalence of DSM-5 MDD: 20.6% overall; 26.1% women, 14.7% men; 12-month prevalence 10.4% overall","excerpt":"\"Of the 36,309 participants, the lifetime and 12-month prevalences of DSM-5 MDD were 20.6% (SE, 0.4) and 10.4% (SE, 0.3), respectively... Lifetime prevalence was 26.1% among women and 14.7% among men. Most lifetime MDD cases were moderate (39.7%) or severe (49.5%).\"\n","source_date":"2018-04-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260405112357/https://pubmed.ncbi.nlm.nih.gov/29450462/","calculation_notes":"NESARC-III is a nationally representative face-to-face survey of 36,309 US civilian noninstitutionalized adults aged 18+ conducted April 2012 to June 2013. Lifetime prevalence of 20.6% is used directly as the normalized figure — no hazard compounding needed because this is already a lifetime estimate. The women-to-men ratio of 26.1%/14.7% ≈ 1.78 provides the female multiplier of ~1.7. DSM-5 criteria were applied retrospectively; the removal of the DSM-IV bereavement exclusion slightly inflates prevalence relative to the earlier NCS-R estimate.\n","independence_note":"NESARC-III (NIAAA/NIH) is an independent nationally representative survey with its own sampling frame, field operations, and diagnostic instrument (AUDADIS-5). Fully independent from the NCS-R (Harvard/NIMH) and from NSDUH (SAMHSA).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/12813115/","title":"The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R)","publisher":"JAMA (Kessler RC, Berglund P, Demler O, et al.)","source_type":"peer_reviewed","statistic":"Lifetime prevalence of major depressive disorder: 16.2% (95% CI 15.1-17.3); 12-month prevalence 6.6% (95% CI 5.9-7.3)","excerpt":"\"The lifetime prevalence of DSM-IV MDD was 16.2% (95% confidence interval [CI], 15.1%-17.3%), representing approximately 32.6 to 35.1 million US adults... the 12-month prevalence was 6.6% (95% CI, 5.9%-7.3%), representing 13.1 to 14.2 million US adults.\"\n","source_date":"2003-06-18","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260425050106/https://pubmed.ncbi.nlm.nih.gov/12813115/","calculation_notes":"NCS-R is a face-to-face household survey of 9,090 respondents aged 18+ conducted 2001-2002 across the 48 contiguous US states. The 16.2% figure under DSM-IV criteria is the lower bound of the uncertainty range. The gap between NCS-R (16.2%) and NESARC-III (20.6%) partly reflects DSM-5 bereavement-exclusion removal, partly secular trend, and partly methodological differences in diagnostic interview instruments (CIDI vs AUDADIS-5).\n","independence_note":"NCS-R (Harvard/NIMH, Kessler) and NESARC-III (NIAAA, Hasin) are fully independent epidemiologic surveys with different sampling frames, field operations, and diagnostic instruments. Agreement to within 4 percentage points across different DSM editions and a decade apart is the strongest cross-validation available for US MDD prevalence.\n"},{"url":"https://www.nimh.nih.gov/health/statistics/major-depression","title":"Major Depression — Statistics","publisher":"National Institute of Mental Health (NIMH), NIH","source_type":"govt_report","statistic":"An estimated 21.0 million US adults (8.3%) had at least one major depressive episode in the past year (2021); highest among 18-25 year olds (18.6%); females (10.3%) vs males (6.2%)","excerpt":"\"An estimated 21.0 million adults in the United States had at least one major depressive episode. This number represented 8.3% of all U.S. adults... The prevalence of major depressive episode was higher among adult females (10.3%) compared to males (6.2%).\"\n","source_date":"2021-12-31","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260411033541/https://www.nimh.nih.gov/health/statistics/major-depression","calculation_notes":"NIMH republishes SAMHSA NSDUH data. The 8.3% past-year prevalence is consistent with the NESARC-III 12-month prevalence of 10.4% (NSDUH uses a screening instrument rather than a full diagnostic interview, which tends to yield slightly different estimates). The 21 million figure is the annual incidence/recurrence count; lifetime accumulation of this annual flow is what produces the 20.6% lifetime prevalence in the NESARC-III cohort study. Used as the federal government cross-check on the peer-reviewed lifetime figures.\n","independence_note":"NSDUH (SAMHSA) is a separate survey from both NESARC-III and NCS-R, with its own sampling design and screening instrument. The past-year estimate is consistent with but methodologically independent of the lifetime prevalence figures above.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/depression","title":"Depressive disorder (depression) — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"An estimated 4% of the global population experiences depression; 5.7% of adults globally (4.6% among men, 6.9% among women); ~332 million people worldwide; about 1.5 times more common among women","excerpt":"\"An estimated 4% of the population experience depression, including 5.7% of adults (4.6% among men and 6.9% among women), and 5.9% of adults aged 70 years and older. Approximately 332 million people in the world have depression. Depression is about 1.5 times more common among women than among men. Worldwide, more than 10% of pregnant women and women who have just given birth experience depression.\"\n","source_date":"2025-08-29","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034808/https://www.who.int/news-room/fact-sheets/detail/depression","calculation_notes":"WHO reports point prevalence (proportion affected at any given time), not lifetime prevalence. The 5.7% adult point prevalence globally is consistent with US 12-month figures of 6.6-10.4% given that many episodes last less than a year and global detection rates are lower. Used here as the international cross-check on depression burden magnitude, not as the primary lifetime estimate.\n","independence_note":"WHO depression estimates derive from the Global Burden of Disease study (IHME) and WHO's own mental health surveys — fully independent from the US-based NESARC-III, NCS-R, and NSDUH pipelines.\n"}],"comparison_anchors":[{"label":"Cancer death (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Type 2 diabetes death (lifetime, US adult)","lifetime_us_adult":0.025},{"label":"Suicide (lifetime, US adult)","lifetime_us_adult":0.00827},{"label":"Alzheimer's disease (lifetime, US adult)","lifetime_us_adult":0.107}],"personal_factor_multipliers":[{"factor":"Female","multiplier":1.7,"notes":"NESARC-III lifetime prevalence: 26.1% women vs 14.7% men, ratio ≈ 1.78. Rounded to 1.7 as a conservative multiplier. The gender gap is one of the most replicated findings in psychiatric epidemiology."},{"factor":"Age 18-25","multiplier":1.5,"notes":"NIMH NSDUH data shows 18.6% past-year prevalence among 18-25 year olds vs 8.3% population average; the elevated rate partly reflects first-onset concentration in early adulthood."},{"factor":"History of childhood adversity (4+ ACEs)","multiplier":2,"notes":"Meta-analyses of adverse childhood experiences consistently find OR ≈ 2.0-2.1 for adult depression among those with multiple ACEs (neglect, abuse, household dysfunction). Dose-response relationship is well established."},{"factor":"First-degree family history of MDD","multiplier":2.8,"notes":"Offspring of parents with MDD carry approximately 2-3x the risk of lifetime MDD; twin-study heritability estimates for MDD are 30-40%."},{"factor":"Strong social support network","multiplier":0.5,"notes":"Protective factor. Prospective cohort studies consistently find roughly halved depression incidence among adults with robust social integration vs those who are socially isolated."},{"factor":"Male","multiplier":0.7,"notes":"14.7% male lifetime prevalence vs 20.6% overall in NESARC-III; ratio ≈ 0.71."}],"short_label":"Major depression","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"\"Major depressive episode\" is defined by DSM-5 criteria: five or more symptoms during a two-week period, including depressed mood or loss of interest/pleasure, representing a change from previous functioning. The lifetime prevalence figure captures anyone who has ever met these criteria, not current cases. Retrospective surveys undercount lifetime episodes because people forget or reframe past episodes — the Dunedin longitudinal cohort (Moffitt et al. 2010) found that 41.4% of the cohort experienced at least one episode of depression between ages 18 and 32 under prospective ascertainment, suggesting the 20.6% NESARC-III figure is likely a floor. Conversely, survey-based diagnostic instruments may overcount mild episodes that a clinician would not diagnose. The 12-month prevalence (8-10% of US adults) is the more policy-relevant number for treatment-capacity planning; the lifetime figure is the right one for answering \"how common is this, really?\" Depression is treatable: roughly 60% of US adults with a past-year episode received some form of treatment in 2021 per NIMH, though treatment adequacy varies widely.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single chair facing a window with muted grey light filtering through, flat vector illustration in subdued tones."},"canonical_url":"https://likelier.app/depression-lifetime","api_url":"https://likelier.app/api/fears/depression-lifetime.json"},{"slug":"hiking-injury-day-hike","question":"What are the odds of serious injury or death while day-hiking?","quick_answer":"Over an active day-hiking career (~600 hikes), the odds of a medically-attended injury — usually a sprain, twisted knee, or cut needing stitches — are about 1 in 5, far higher than the 'safe activity' reputation suggests. A fatal hiking injury is much rarer, on the order of 1 in 3,000 to 1 in 10,000 across that same career.\n","category":"health","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"Day-hiking sits in an awkward perceptual category: it is the prototypical \"safe outdoor activity\" recommended to almost everyone, and at the same time the headline cases that reach the news — lost hiker, fatal fall, SAR helicopter — are vivid enough that the fear of getting hurt on a hike is widespread, especially for solo hikers and first-timers on unfamiliar trails. There is no good survey isolating perceived probability of injury per hike-day, so we mark this as editorial intuition. The interesting property is that the two halves of the fear track very different probabilities. Fatal hiking injury is rare, around one per one to two million park visits. A sprain, strain, twisted knee, or blister bad enough to need attention is genuinely common across an active hiking career.\n","rough_estimate":"Most people would guess the per-hike chance of any meaningful injury is well under 1 percent","kind":"intuition"},"native":{"display":"~1 medically-attended injury per ~2,500 hike-days (recreational day-hiker on maintained trails)","numerator":1,"denominator":2500,"unit":"per hike-day (recreational day-hike on maintained US trails)","population":"US adult recreational day-hikers on maintained park or forest trails","exposures_per_year":20},"normalized":{"lifetime_us_adult":0.21,"display":"~1 in 5 over a 30-year, ~600-hike active day-hiking career","log_value":-0.678,"assumptions":"Scope is activity_specific_lifetime, expressed across a typical recreational day-hiking career of approximately 20 hikes per year over 30 years (~600 hike- days). There is no single canonical published per-hiker-day injury rate for US recreational day-hikers; the literature uses three different denominators (program-days, park visits, EMS-eligible visits) and we triangulate. Starting from McIntosh et al. (2007) NOLS expeditions at 1.18 injuries per 1,000 program- days (multi-day off-trail with packs, an upper bound for hiking intensity) and scaling down by roughly a factor of three for the lower intensity of recreational day-hiking on maintained trails, the per-hike-day rate of medically-attended injury lands near 1 in 2,500. Across 600 hike-days: 1 minus (1 minus 1/2500) to the 600th power equals approximately 0.21, or about 1 in 5. This is meaningfully higher than most people expect for \"hiking\" because it includes urgent-care-level ankle sprains, twisted knees, and cuts requiring stitches — the mundane outcomes that dominate the injury mix. Fatal and SAR-worthy outcomes sit two to four orders of magnitude below: the Heggie and Amundson (2009) NPS-wide series reports roughly 5 SAR-injured persons and 0.6 fatalities per million park visits, so the lifetime fatality probability for a 600-hike day-hiker is on the order of 1 in 3,000 to 1 in 10,000, comparable to dying in a bicycle crash.\n","uncertainty":{"low":0.1,"high":0.35},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20030438/","title":"Injury and illness encountered in Shenandoah National Park","publisher":"Wilderness & Environmental Medicine (Forrester JD, Holstege CP)","source_type":"peer_reviewed","statistic":"2.7 persons reported injured or ill per 100,000 visitors to Shenandoah National Park; hiking is the most common activity at the time of injury; most common adult injury is soft-tissue injury of the distal lower extremity","excerpt":"\"2.7 persons reported injured or ill per 100,000 visitors to Shenandoah National Park.\" \"The most common activity in which adults were involved at the time of the injury was hiking.\" \"soft tissue injury, with the most common anatomical location being the distal lower extremity.\" [Paraphrase from abstract — full text paywalled]\n","source_date":"2009-12-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260120030951/https://pubmed.ncbi.nlm.nih.gov/20030438/","calculation_notes":"Forrester and Holstege analysed 5 years of Shenandoah National Park ranger case-reports (2003 to 2007). The 2.7 per 100,000 figure is the rate at which a park visitor is sufficiently injured or ill to be formally recorded by a ranger — a sampling floor, not a per-day injury rate. Most blisters, minor sprains, and even some moderate strains never reach a ranger; the figure is best read as \"rate of ranger-attention-level events per park visit.\" Because not every park visit is a hike, the per-hike rate is meaningfully higher than the per-visit rate quoted here — we use this source primarily to anchor the lower bound of the injury severity ladder, not the headline native rate. The finding that distal lower-extremity soft tissue injury (ankle sprain, knee strain) dominates is consistent across every hiking injury study and drives the personal factor multipliers.\n","independence_note":"Independent dataset (Shenandoah ranger case-reports). Does not overlap with NOLS expedition data or NPS-wide SAR aggregates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/18076301/","title":"Medical incidents and evacuations on wilderness expeditions","publisher":"Wilderness & Environmental Medicine (McIntosh SE, Leemon D, Visitacion J, Schimelpfenig T, Fosnocht D)","source_type":"peer_reviewed","statistic":"Injuries occurred at 1.18 per 1,000 program-days; illnesses at 1.08 per 1,000 program-days across NOLS wilderness expeditions; sprains and strains were the most common injury class","excerpt":"\"Injuries occurred at a rate of 1.18 per 1000 program days, and illnesses at a rate of 1.08 per 1000 program days.\" [Paraphrase from abstract — full text paywalled]\n","source_date":"2007-12-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20251113063339/https://pubmed.ncbi.nlm.nih.gov/18076301/","calculation_notes":"The NOLS injury rate is the cleanest per-day measure in the wilderness-medicine literature, but NOLS program-days are multi-day expeditions with packs, often off-trail, and student populations skewed young and adventurous. They are an upper-bound proxy for recreational day-hiking on maintained trails. We divide the NOLS rate by approximately three to land the per-hike-day medically- attended-injury rate near 1 in 2,500 — the value used as native.numerator/ denominator. The factor of three is a defensible mid-range adjustment: NOLS participants carry packs, hike off-trail, and accumulate continuous exposure across multi-week expeditions, all of which roughly triple injury risk per day relative to a four-hour Saturday day-hike on a Class 1 trail with no pack. The factor is the largest source of uncertainty in the entry's headline, reflected in the wide normalized.uncertainty band of 0.10 to 0.35.\n","independence_note":"Independent of all NPS-based sources. Same data lineage as Leemon and Schimelpfenig 2003 (1.07 per 1,000 program-days for 1999 to 2002) but a different time-window, not double-counting.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19737043/","title":"Dead men walking: search and rescue in US National Parks","publisher":"Wilderness & Environmental Medicine (Heggie TW, Amundson ME)","source_type":"peer_reviewed","statistic":"1992 to 2007: 65,439 SAR incidents involving 78,488 individuals; 2,659 fatalities, 24,288 ill or injured, 13,212 saves; mean 11.2 SAR incidents per day NPS-wide","excerpt":"\"From 1992 to 2007 there were 78,488 individuals involved in 65,439 SAR incidents. These incidents ended with 2659 fatalities, 24,288 ill or injured individuals, and 13,212 saves. On average there were 11.2 SAR incidents each day at an average cost of $895 per operation.\" [Paraphrase from abstract — full text paywalled]\n","source_date":"2009-09-01","source_accessed":"2026-05-28","archive_url":"https://web.archive.org/web/20260531015312/https://pubmed.ncbi.nlm.nih.gov/19737043/","calculation_notes":"NPS-wide SAR totals over 16 years. 24,288 ill or injured divided by 16 years equals roughly 1,520 SAR-attended injuries per year. NPS visitation in this window averaged about 280 million visits per year, so the SAR-injured rate is roughly 5.4 per million visits and the long-run fatality rate roughly 0.6 per million visits. Lane et al. 2015's modern NPS fatality counts (an annual average of about 250 deaths across ~280 million visits, or roughly 0.9 per million) are the same order of magnitude, so the all-visit fatality rate is on the order of 0.6 to 0.9 per million park visits. Not every park visit is a hike, so the per-hike-day fatality rate is somewhat higher than the all-visit rate but remains on the order of one per one to a few million hike-days. Over a 600-hike career, a per-hike-day fatality rate of roughly 1.7e-7 to 5.5e-7 compounds to a lifetime fatality probability of about 1 in 3,000 to 1 in 10,000, comparable to the lifetime probability of dying in a bicycle crash. This source supplies the SAR severity rung directly (5.4 per million NPS visits) and the long-run fatality floor (0.6 per million visits); the modern Lane counts corroborate the same order of magnitude.\n","independence_note":"Shares the NPS-IRMA incident database with Lane et al. 2015. The Heggie and Lane numbers should not be arithmetically combined; both summarise the same reporting system at different points in time.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26384763/","title":"Emergency Medical Service in the US National Park Service: A Characterization and Two-Year Review, 2012 to 2013","publisher":"Wilderness & Environmental Medicine (Lane JE, Taylor B, Smith JE, Wheeler EC)","source_type":"peer_reviewed","statistic":"EMS responses totaled 40 calls per million visitors in 2012 and 34 calls per million visitors in 2013; trauma 49 percent, medical 51 percent; 262 fatalities in 2012, 238 in 2013 NPS-wide, traumatic fatalities approximately twice as common as nontraumatic","excerpt":"\"EMS responses totaled 40 calls per million visitors in 2012 and 34 calls per million visitors in 2013.\" \"There were 262 total fatalities in 2012 and 238 in 2013, with traumatic fatalities occurring approximately twice as often as nontraumatic fatalities.\" [Paraphrase from abstract — full text paywalled]\n","source_date":"2015-09-01","source_accessed":"2026-05-28","archive_url":"https://web.archive.org/web/20260531015342/https://pubmed.ncbi.nlm.nih.gov/26384763/","calculation_notes":"The modern (post-2010) per-visit EMS denominator. An EMS-call rate of roughly 37 per million NPS visits is broader than SAR — it includes parking-lot cardiac events and motor-vehicle crashes in addition to wilderness incidents. For a hiking-specific subset, dividing by approximately three (hiking is one of several activity contributors that trigger EMS) gives roughly 12 per million hike-visits or 1 in 80,000 per hike-day. This is the same order of magnitude as the Heggie SAR-injured rate, providing a cross-check on the SAR severity rung. Lane also reports 262 total NPS fatalities in 2012 and 238 in 2013 — an annual average of roughly 250 deaths across about 280 million visits per year, or about 0.9 per million visits, consistent with the long-run Heggie figure of roughly 0.6 per million. (Note: the two years' totals are separate annual counts and must not be summed into a single year.)\n","independence_note":"Same NPS reporting system as Heggie. Use Lane for the modern per-visit baseline and Heggie for the long historical aggregate; do not double-count.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Serious skiing injury per 20-day season (active recreational skier)","lifetime_us_adult":0.0392},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Per hike-day (medically-attended injury, day-hike on maintained trail)","probability":0.0004,"notes":"Point estimate of 1 in 2,500. NOLS expedition rate of 1.18 per 1,000 program-days scaled down by approximately 3x for the lower intensity of recreational day-hiking. The figure includes ankle sprain, twisted knee, cuts requiring stitches, fall injuries — the mundane outcomes that dominate the hiking injury mix."},{"region":"Per hike-day (SAR-worthy or EMS-eligible event)","probability":0.0000125,"notes":"About 1 in 80,000 per hike-day, derived from Lane et al. 2015 NPS EMS rates with a hiking-activity adjustment. SAR-eligible means an incident serious enough to require coordinated rescue or ambulance response — well past a sprained ankle the hiker walked out on."},{"region":"Per hike-day (fatality, hiking-attributable)","probability":3.5e-7,"notes":"On the order of 1 in 3 million per hike-day, floored on the Heggie and Amundson (2009) long-run NPS fatality rate of about 0.6 per million park visits and corroborated by Lane et al. 2015 modern NPS counts (roughly 0.9 per million). Because not every park visit is a hike, the per-hike-day rate is somewhat higher than the all-visit rate; the defensible range is roughly 1 in 1.8 million to 1 in 6 million per hike-day. Cardiac events, falls from height, and exposure dominate; lost-hiker fatalities are a small fraction of the total."},{"region":"Per 600-hike active day-hiking career (any medically-attended injury)","probability":0.21,"notes":"Headline activity_specific_lifetime figure: about 1 in 5 active day-hikers will accumulate at least one medically-attended hiking injury across a 30-year, 20-hikes-per-year career. Wide uncertainty band 0.10 to 0.35 reflects the multi-source denominator triangulation."},{"region":"Per 600-hike career (fatality)","probability":0.00021,"notes":"Roughly 1 in 3,000 to 1 in 10,000 over the full active career, floored on the Heggie and Amundson (2009) long-run NPS fatality rate of about 0.6 per million park visits and corroborated by Lane et al. 2015 modern counts (roughly 0.9 per million). Comparable to the lifetime probability of dying in a bicycle crash. The dramatic SAR-helicopter version of the fear is far rarer per hike than the mundane sprained ankle."},{"region":"Backcountry / multi-day expedition with pack (per equivalent day)","probability":0.0012,"notes":"NOLS rate of 1.18 per 1,000 program-days — the unscaled upper bound. Multi-day exposure with packs, off-trail terrain, and accumulated fatigue roughly triples the per-day rate relative to a maintained-trail day-hike."}],"personal_factor_multipliers":[{"factor":"off-trail / scrambling / unmarked terrain","multiplier":3,"notes":"Most hiking injuries are caused by a loss of footing. Off-trail terrain multiplies the rate of ankle and knee sprains substantially; the NOLS figure is the natural upper bound at roughly 3x the maintained-trail rate."},{"factor":"solo hike, no companion","multiplier":1.5,"notes":"Per-event injury risk is roughly unchanged, but a moderate injury that a partnered hiker walks out from can become a SAR-eligible event for a solo hiker. The multiplier captures the shift up the severity ladder, not the underlying fall rate."},{"factor":"hiker age 65 or older","multiplier":2,"notes":"Older hikers fall slightly more often and recover from a fall substantially worse. Hip fracture from a trail fall is rare in younger hikers and meaningfully more common past 65; case-fatality of a hiking injury also rises with age. The factor combines a small increase in fall rate with a larger increase in serious-outcome risk."},{"factor":"strenuous summit attempt or peak-bagging","multiplier":3,"notes":"Cumulative fatigue, exposure, altitude, and steeper terrain raise both fall rate and case-severity. Most reported hiking fatalities involve summit or near-summit incidents. The factor is approximate but consistent across SAR case-series."},{"factor":"established maintained nature trail (Class 1, under 5 km)","multiplier":0.5,"notes":"The mid-day flat-terrain park nature walk is the safest end of the spectrum. Per-day rate roughly halves vs the entry's maintained-trail baseline because exposure time is short and terrain difficulty is minimal."}],"short_label":"Hiking injury","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"No single published study reports a per-hiker-day injury rate for US recreational day-hikers on maintained trails. The literature uses three different denominators: program-days (NOLS, multi-day intensive), park visits (NPS, visitor includes scenic drivers and picnickers), and EMS-eligible visits (NPS EMS, broader than hiking-attributable). The headline 1-in-2,500-per-hike rate is a triangulation across these systems with an explicit intensity-adjustment factor of approximately three from the NOLS upper bound. That factor is the largest source of uncertainty and the reason the normalized uncertainty band is wide (0.10 to 0.35) rather than narrow. The headline also bundles outcomes that span three orders of magnitude in severity: a Saturday-afternoon ankle sprain that needs urgent care, a SAR-eligible fall from a switchback, and a fatal cardiac event on a steep ascent are all \"hiking injury\" in the surveillance data but answer very different versions of the underlying fear. The breakdown rows make the severity ladder explicit. Finally, \"hiking\" is not a homogeneous activity. Off-trail scrambling, alpine ascents, and Class 4 backcountry are not what most readers mean by \"I went on a hike,\" and carry substantially higher per-hour risk; the entry headline is calibrated to maintained-trail day-hiking and explicitly excludes mountaineering, technical scrambling, and multi-day wilderness expeditions, which are closer to the adventure-sports category.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":3,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-05-28","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single pair of hiking boots resting on a flat trail stone, viewed from a low angle, calm muted palette."},"canonical_url":"https://likelier.app/hiking-injury-day-hike","api_url":"https://likelier.app/api/fears/hiking-injury-day-hike.json"},{"slug":"postpartum-depression","question":"What are the odds of developing postpartum depression?","quick_answer":"About 1 in 8 women per delivery, roughly a 21% lifetime probability for a US woman who gives birth at least once. That is well above the ~5% many assume — a guess that often confuses postpartum depression with the much rarer postpartum psychosis.\n","category":"kids","tags":["mental-health","pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Most people associate the postpartum period with joy and bonding, framing sustained depressive symptoms as rare or as a personal failing rather than a common medical complication. The phrase \"baby blues\" — which describes a distinct, milder, and self-limiting condition affecting up to 80% of new mothers — further muddies the picture, leading many to assume that anything beyond a few weepy days is unusual. No rigorous population survey has directly asked the public to estimate PPD prevalence, but clinical experience and qualitative research consistently find that patients and partners underestimate how common it is.\n","rough_estimate":"~5% (common lay guess, confusing PPD with rare postpartum psychosis)","kind":"intuition"},"native":{"display":"~1 in 8 women with a recent live birth","numerator":13,"denominator":100,"unit":"per delivery","population":"US women who gave birth (PRAMS, self-reported depressive symptoms)"},"normalized":{"lifetime_us_adult":0.21,"display":"~21% lifetime probability for a US woman who gives birth at least once","log_value":-0.68,"assumptions":"Per-delivery prevalence of ~13% from CDC PRAMS (2018 Vital Signs). Average US fertility of ~1.9 births per woman who becomes a mother. Treating each delivery as an approximately independent trial: 1 - (1 - 0.13)^1.9 ≈ 0.23. Central estimate trimmed to 0.21 to account for partial overlap (a woman who had PPD once may have it again, but the per-delivery rate already includes recurrences in the denominator population). Restricted to women who give birth — roughly 85% of US women by age 44 (NCHS). The 13% PRAMS figure uses a validated symptom screen (PHQ-2/PHQ-9), not a clinical diagnosis, so it captures probable cases including those never formally diagnosed.\n","uncertainty":{"low":0.13,"high":0.3},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/69/wr/mm6919a2.htm","title":"Vital Signs: Postpartum Depressive Symptoms and Provider Discussions About Perinatal Depression — United States, 2018","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"About 1 in 8 women (13.2%) with a recent live birth reported symptoms of postpartum depression","excerpt":"\"About 1 in 8 women with a recent live birth reported symptoms of postpartum depression. One in five women did not have a provider discussion about depression during prenatal visits, and one in eight did not have one during postpartum visits.\"\n","source_date":"2020-05-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260407014055/https://www.cdc.gov/mmwr/volumes/69/wr/mm6919a2.htm","calculation_notes":"PRAMS data from 31 sites (2018). Self-reported depressive symptoms screened via PHQ-2. The 13.2% prevalence is the headline native estimate. Screening gaps: 79.1% asked prenatally, 87.4% asked postpartum — meaning roughly 1 in 5 women were never screened prenatally and 1 in 8 were never screened postpartum.\n"},{"url":"https://www.acog.org/clinical/clinical-guidance/clinical-practice-guideline/articles/2023/06/screening-and-diagnosis-of-mental-health-conditions-during-pregnancy-and-postpartum","title":"Screening and Diagnosis of Mental Health Conditions During Pregnancy and Postpartum","publisher":"American College of Obstetricians and Gynecologists (ACOG)","source_type":"reputable_reference","statistic":"Perinatal depression affects approximately 1 in 7 women; ACOG recommends screening at least twice during pregnancy and once postpartum","excerpt":"\"Perinatal mood and anxiety disorders are among the most common complications of pregnancy and the postpartum period. Depression should be screened at least two times during pregnancy and again during a postpartum visit using validated instruments.\"\n","source_date":"2023-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503093029/https://www.acog.org/clinical/clinical-guidance/clinical-practice-guideline/articles/2023/06/screening-and-diagnosis-of-mental-health-conditions-during-pregnancy-and-postpartum","calculation_notes":"ACOG Clinical Practice Guideline No. 5 (June 2023). The 1-in-7 figure (~14%) is consistent with the CDC PRAMS 13.2% estimate and the Hahn-Holbrook et al. meta-analytic range. ACOG recommends the Edinburgh Postnatal Depression Scale (EPDS) or PHQ-9 as screening tools. This source establishes the clinical consensus on prevalence and screening protocols.\n","independence_note":"ACOG guideline synthesizes multiple independent evidence streams including PRAMS and peer-reviewed meta-analyses. Not an independent data source per se, but an authoritative clinical interpretation of the evidence base.\n"},{"url":"https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2017.00248/full","title":"Economic and Health Predictors of National Postpartum Depression Prevalence: A Systematic Review, Meta-analysis, and Meta-Regression of 291 Studies from 56 Countries","publisher":"Frontiers in Psychiatry (Hahn-Holbrook et al.)","source_type":"peer_reviewed","statistic":"Global pooled prevalence of postpartum depression: 17.7% (95% CI 16.6-18.8%) across 291 studies from 56 countries","excerpt":"\"The global pooled prevalence of PPD was 17.7% (95% confidence interval: 16.6–18.8%)... ranging from 3% (2–5%) in Singapore to 38% (35–41%) in Chile. Higher rates of income inequality, maternal mortality, infant mortality, and of women of childbearing age working 40 or more hours a week were all significant predictors of higher national rates of PPD.\"\n","source_date":"2017-11-20","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260320055625/https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2017.00248/full","calculation_notes":"Hahn-Holbrook et al. 2018 (received 2017, published February 2018) is the largest meta-analysis to date on PPD prevalence, covering 291 studies across 56 countries. The 17.7% global pooled estimate is somewhat higher than the US-specific PRAMS 13.2%, consistent with higher rates in low- and middle-income countries. Used to anchor the global context and the upper end of the uncertainty band.\n","independence_note":"Independent meta-analysis pooling primary studies globally. Overlaps with some studies also covered by PRAMS but draws predominantly from non-US populations, making it methodologically independent of the CDC data.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK519070/","title":"Perinatal Depression","publisher":"StatPearls, NCBI Bookshelf (National Library of Medicine)","source_type":"reputable_reference","statistic":"Up to 50% of perinatal depression cases remain undiagnosed, largely due to stigma and patients' reluctance to disclose symptoms","excerpt":"\"up to 50% of cases remain undiagnosed due to the stigma surrounding the condition and patients' reluctance to disclose symptoms.\"\n","source_date":"2023-04-24","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260622061924/https://www.ncbi.nlm.nih.gov/books/NBK519070/","calculation_notes":"Used for the diagnosis-gap claim in the body text. NCBI Bookshelf clinical reference; the ~50% undiagnosed figure is the best-attested version of this widely cited statistic (some secondary aggregators quote ~60%, but no authoritative primary source for that higher figure was found, so the body text uses the more conservative, sourced ~50% estimate). Does not itself quantify what share of diagnosed cases go untreated.\n","independence_note":"NIH/NLM-hosted clinical reference; independent of the CDC PRAMS surveillance data and the Hahn-Holbrook meta-analysis above.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27475890/","title":"Prevalence of paternal depression in pregnancy and the postpartum: An updated meta-analysis","publisher":"Journal of Affective Disorders (Cameron et al.)","source_type":"peer_reviewed","statistic":"Paternal postpartum depression prevalence: 8.4% (95% CI 7.2-9.6%) across 74 studies","excerpt":"\"The meta-estimate for paternal depression was 8.4% (95% confidence interval 7.2%-9.6%), based on 74 studies involving 41,480 participants.\"\n","source_date":"2016-08-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250615162055/https://pubmed.ncbi.nlm.nih.gov/27475890/","calculation_notes":"Used for the paternal PPD regional_breakdown entry. The 8.4% estimate is consistent with the ~8-10% range commonly cited in clinical literature. Paternal PPD is correlated with maternal PPD (partners of depressed mothers have roughly 2.5x the risk) but is not included in the native or normalized probability, which refers to birthing parents only.\n","independence_note":"Entirely independent dataset focusing on fathers/partners. No overlap with PRAMS or Hahn-Holbrook et al., which study maternal depression exclusively.\n"}],"comparison_anchors":[{"label":"Depression (lifetime, US adult)","lifetime_us_adult":0.206},{"label":"Miscarriage (per recognized pregnancy)","lifetime_us_adult":0.15},{"label":"C-section complications (per delivery, serious)","lifetime_us_adult":0.034}],"regional_breakdown":[{"region":"US average (PRAMS, per delivery)","probability":0.132,"notes":"CDC Vital Signs 2020, self-reported depressive symptoms via PHQ-2"},{"region":"First-time mothers","probability":0.15,"notes":"Slightly elevated risk for primiparous women; estimates range 14-17%"},{"region":"Mothers with prior depression history","probability":0.33,"notes":"Approximately 1 in 3 women with prior depressive episodes develop PPD"},{"region":"Fathers (paternal PPD)","probability":0.084,"notes":"Cameron et al. 2016 meta-analysis; 8.4% across 74 studies"},{"region":"Low- and middle-income countries","probability":0.247,"notes":"JAMA Psychiatry 2023 meta-analysis; ~1 in 4 perinatal women"}],"personal_factor_multipliers":[{"factor":"History of depression or anxiety","multiplier":2.5,"notes":"Strongest predictor; roughly 1 in 3 with prior mood disorder develops PPD vs ~13% baseline"},{"factor":"Lack of social support","multiplier":2,"notes":"Consistently identified across meta-analyses as a strong psychosocial predictor"},{"factor":"Unplanned pregnancy","multiplier":1.4,"notes":"OR ~1.38 (95% CI 1.22-1.57) in pooled analyses; larger effects in some LMIC studies"},{"factor":"Intimate partner violence","multiplier":3,"notes":"Prevalence ~39% among IPV-exposed women vs ~13% baseline"},{"factor":"Pregnancy or delivery complications","multiplier":1.5,"notes":"Emergency cesarean, preterm birth, or NICU admission elevate risk"},{"factor":"Partner also depressed","multiplier":2.5,"notes":"Maternal and paternal PPD are correlated; partner depression roughly doubles risk"}],"short_label":"Postpartum depression","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The 13% PRAMS figure relies on self-reported symptom screens, not clinical diagnosis — it likely captures some false positives (transient distress) while missing false negatives (women who minimize symptoms). The normalized lifetime estimate of ~21% assumes approximately independent risk across pregnancies, which is an oversimplification: women who had PPD once are at higher risk in subsequent pregnancies, while women who did not are at somewhat lower risk. The \"baby blues\" — a milder, self-limiting mood disturbance affecting up to 80% of new mothers in the first two weeks — is explicitly excluded from these figures. Postpartum psychosis, a rare psychiatric emergency affecting roughly 1-2 per 1,000 deliveries, is also a distinct condition and not included here.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single empty rocking chair beside a window with soft grey light, rendered in muted blue and warm grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/postpartum-depression","api_url":"https://likelier.app/api/fears/postpartum-depression.json"},{"slug":"e-scooter-helmetless-head-injury","question":"What are the odds of serious head injury riding an e-scooter without a helmet?","quick_answer":"About 22 percent of e-scooter crash injuries affect the head or neck, a per-crash figure rather than a lifetime risk, and going unhelmeted raises head-injury risk by roughly 48 to 81 percent. Many riders assume only a 2 to 5 percent chance per crash.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"E-scooter riders overwhelmingly ride bareheaded. Observational studies find that 80-96% of shared e-scooter riders skip helmets entirely, with the figure even higher in European cities where ETSC recorded just 4% helmet use among crash-involved riders. The cultural framing treats e-scooters as casual, low-speed transport -- closer to walking than cycling. Helmets feel disproportionate to the perceived risk. Many riders assume that at 15-25 km/h, a fall is unlikely to cause anything worse than scraped palms.\n","rough_estimate":"~2-5% chance of serious head injury per crash","kind":"intuition"},"native":{"display":"~18-40% of e-scooter injuries involve the head or neck; unhelmeted riders face ~48-81% higher head injury risk than helmeted riders","numerator":22,"denominator":100,"unit":"proportion of e-scooter injuries that are head/neck injuries (systematic review of 34 studies)","population":"e-scooter riders presenting to emergency departments worldwide, predominantly unhelmeted"},"normalized":{"lifetime_us_adult":0.22,"display":"~22% of all e-scooter crash injuries affect the head/neck (per crash event, not per US adult)","log_value":-0.66,"assumptions":"A systematic review of 34 studies (PMC 2022) found 22.2% of e-scooter injuries involved the head and neck. Trauma center studies capturing more severe cases find 38-40% head involvement. We use the 22% systematic review figure as the central estimate because it pools across severity levels and geographies. This is a per-crash conditional probability: given that a rider crashes and presents to an ED, ~22% will have a head/neck injury. The vast majority of these riders (~80-96%) were unhelmeted. In a controlled bicycle-collision cohort (the closest robust helmet-effect benchmark, since e-scooter helmet series are too small for odds ratios), helmet use reduced head injury odds by 48% (OR 0.52, 95% CI 0.31-0.86) and high-energy head trauma odds by 72% (OR 0.28). In one community ED study, 100% of TBI and closed head injuries occurred in unhelmeted patients. The annual crash risk per rider is harder to estimate due to unknown rider-miles; with on the order of tens of thousands of US e-scooter ED visits per year (the JAMA NEISS pull below puts the 2022 figure at 56,847) against an estimated 40-50 million US e-scooter trips/year, the per-trip ER visit rate is roughly 0.1-0.3%, but many crashes go unreported. 2026 refresh: a NEISS analysis in JAMA Network Open (Fernandez et al. 2024) puts US e-scooter ED injuries at 8,566 in 2017 rising to 56,847 in 2022 (>45%/yr) -- an independent confirmation of the adoption-driven surge that does not move the 22% per-crash head/neck central estimate. The CDC / Austin Public Health field study (2018-19) found ~48% head injury, ~15% TBI, and <1% helmet use among injured dockless riders, with 33% on their first ride -- consistent with our unhelmeted-baseline framing and inside the 0.15-0.40 band. The point estimate, log_value, and uncertainty are unchanged; the new sources refresh incidence context and reinforce the helmet effect size rather than revise the headline number.\n","uncertainty":{"low":0.15,"high":0.4},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9533239/","title":"The impact of e-scooter injuries: a systematic review of 34 studies","publisher":"Bone & Joint Open (British Editorial Society of Bone and Joint Surgery)","source_type":"peer_reviewed","statistic":"22.2% of e-scooter injuries were head and neck injuries across 34 studies; TBI in 2.5%, intracranial hemorrhage in 1.9%, concussions in 3.2%","excerpt":"\"Head and neck injuries accounted for 22.2% of all reported e-scooter injuries across 34 studies. Traumatic brain injury composed 2.5%, intracranial hemorrhage 1.9%, and concussions 3.2% of all injuries in overwhelmingly unhelmeted populations.\"\n","source_date":"2022-10-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260505053532/https://pmc.ncbi.nlm.nih.gov/articles/PMC9533239/","calculation_notes":"Systematic review pooling 34 studies on e-scooter injuries globally. The 22.2% head/neck figure is the weighted proportion across all injury presentations. Most study populations had <20% helmet use. The TBI subcomponents (2.5% TBI, 1.9% ICH, 3.2% concussion) total ~7.6% of all injuries being brain-specific, with the remainder being facial fractures, lacerations, and soft tissue injuries.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/38905159/","title":"Head injuries related to bicycle collisions and helmet use - an observational study","publisher":"Traffic Injury Prevention (Taylor & Francis)","source_type":"peer_reviewed","statistic":"Helmet use reduced head injury odds ratio to 0.52 (95% CI 0.31-0.86); in high-energy trauma, OR was 0.28 (95% CI 0.12-0.80) -- bicycle-collision cohort, applied here as the closest helmet-effect benchmark for unprotected scooter-type falls","excerpt":"\"Helmet use reduced the risk of head injury with an odds ratio of 0.52, (95% CI 0.31 - 0.86). In high-energy trauma, the use of a helmet showed a significant reduction in the risk of sustaining a head injury with an odds ratio of 0.28, (95% CI 0.12 - 0.80).\"\n","source_date":"2024-06-21","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260714013603/https://pubmed.ncbi.nlm.nih.gov/38905159/","calculation_notes":"An observational study of 407 bicyclists assessed by Copenhagen Emergency Medical Services (2016-2019), of whom 247 (61%) sustained a head injury. The OR of 0.52 means helmeted riders had roughly half the head injury risk of unhelmeted riders; for high-energy crashes the OR of 0.28 means helmets reduced head injury risk by 72%. These are bicycle-collision figures. Dedicated e-scooter helmet-effect studies are smaller and rarely report odds ratios (the largest e-scooter helmet series had only one helmeted rider out of 92), so the bicycle benchmark is the closest robust helmet-effect estimate; the head-to-pavement biomechanics of a forward fall are comparable between the two modes, but the cross-mode transfer is an approximation.\n"},{"url":"https://www.cpsc.gov/s3fs-public/Micromobility-Products-Related-Deaths-Injuries-and-Hazard-Patterns_2017-2023.pdf","title":"Micromobility Products-Related Deaths, Injuries, and Hazard Patterns, 2017-2023","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"233 micromobility deaths 2017-2022, of which 111 (48%) were e-scooter fatalities; ~169,300 e-scooter ED-treated injuries 2017-2022; the 2022 e-scooter ED estimate of 51,700 was a 22% increase over 2021","excerpt":"\"CPSC identified 233 fatalities involving micromobility products from 2017 to 2022. E-scooter-related fatalities represent 111 out of 233 (48 percent) total fatalities. Staff estimates that 169,300 injuries related to e-scooters were treated from 2017 through 2022. The 2022 ED-treated injury estimate of 51,700 for e-scooters reflects an increase of 22 percent from the 2021 estimate, which is statistically significant (p-value < 0.01).\"\n","source_date":"2024-06-01","source_accessed":"2026-06-30","archive_url":"https://web.archive.org/web/20250809053021/https://www.cpsc.gov/s3fs-public/Micromobility-Products-Related-Deaths-Injuries-and-Hazard-Patterns_2017-2023.pdf","calculation_notes":"111 e-scooter deaths over 6 years = ~18.5 per year nationally. With estimated 40-50 million annual e-scooter trips in the US, the per-trip fatality rate is roughly 1 in 2-3 million trips. E-scooter ED-treated injuries rose steadily across 2017-2022 to a 2022 estimate of 51,700 (a statistically significant 22% increase over 2021), reflecting explosive adoption rather than increasing per-trip risk. The ~115,713 figure previously attributed here to a 2024 NEISS total is not in this 2017-2022 report and has been removed; the headline 22% head/neck share comes from the systematic-review source above, not from this CPSC report.\n"},{"url":"https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2821387","title":"Injuries With Electric vs Conventional Scooters and Bicycles","publisher":"JAMA Network Open","source_type":"peer_reviewed","statistic":"e-Scooter ED injuries rose from 8,566 (2017) to 56,847 (2022), >45% annually; head injury associated with hospitalization (AOR 1.20), drug use (AOR 2.70), alcohol use (AOR 1.71)","excerpt":"\"e-Scooter injuries increased by more than 45% annually (P = .002), from 8566 (95% CI, 5522-11 611) to 56 847 (95% CI, 39 673-74 022). Head injury (AOR, 1.20; 95% CI, 1.03-1.41), drug use (AOR, 2.70; 95% CI, 2.26-3.23), and alcohol use (AOR, 1.71; 95% CI, 1.37-2.13) at the time of vehicle injury were each associated with hospitalization among all vehicle types.\"\n","source_date":"2024-07-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260710100620/https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2821387","calculation_notes":"National Electronic Injury Surveillance System (NEISS) analysis of US ED visits 2017-2022. The 8,566-to-56,847 trajectory (a 6.6-fold rise over 5 years) confirms the CPSC adoption-driven surge from an independent NEISS pull. The drug-use AOR of 2.70 for hospitalization is the clean, vehicle-pooled multiplier added as a personal factor below; it is a distinct axis from the alcohol AOR (1.71) and the head-injury-to-hospitalization escalation (1.20). These are hospitalization odds conditional on an injury, not per-trip risks.\n"},{"url":"http://web.archive.org/web/20260214004906/https://www.austintexas.gov/sites/default/files/files/Health/Epidemiology/APH_Dockless_Electric_Scooter_Study_5-2-19.pdf","title":"Dockless Electric Scooter-Related Injuries Study (CDC / Austin Public Health)","publisher":"Centers for Disease Control and Prevention / Austin Public Health","source_type":"govt_report","statistic":"Of 190 injured dockless e-scooter riders (125 interviewed), ~48% had a head injury, ~15% had injuries suggestive of TBI, and <1% wore a helmet; 33% were first-time riders","excerpt":"\"Nearly half (48%) of injured riders had a head injury. About 15% had evidence suggestive of a traumatic brain injury. Less than 1% of injured riders were wearing a helmet at the time of their injury. Roughly one-third (33%) were first-time e-scooter riders.\"\n","source_date":"2019-05-02","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260214004906/https://www.austintexas.gov/sites/default/files/files/Health/Epidemiology/APH_Dockless_Electric_Scooter_Study_5-2-19.pdf","calculation_notes":"First US public-health agency field study of dockless shared e-scooters (Austin, TX, Sep-Nov 2018). The <1% observed helmet rate anchors the \"unhelmeted baseline IS the headline rate\" framing. The 48% head-injury share is higher than the 22% systematic-review pool because this cohort skews toward more severe ED/EMS presentations; the 15% TBI rate sits well inside our 0.15-0.40 uncertainty band. Excerpt paraphrased from agency findings as reported by Consumer Reports and the CDC MMWR summary; figures cross-checked against multiple reports of the same study.\n"}],"comparison_anchors":[{"label":"E-scooter serious injury per rider (general, existing entry)","lifetime_us_adult":0.03},{"label":"Cycling without helmet head injury (per crash)","lifetime_us_adult":0.17},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"wearing a certified helmet","multiplier":0.3,"notes":"Helmets reduce head injury odds by 48-72% depending on crash energy; OR 0.52 overall, OR 0.28 for high-energy impacts"},{"factor":"riding a shared (rental) scooter vs personal","multiplier":1.5,"notes":"Shared scooter riders have lower helmet rates (4-17% vs ~30-60% for private owners) and less riding experience; unfamiliarity with braking and handling increases crash risk"},{"factor":"riding at night or intoxicated","multiplier":3,"notes":"CPSC data shows disproportionate fatalities at night; alcohol involvement is a major factor in severe e-scooter injuries across multiple studies"},{"factor":"collision with a motor vehicle (vs solo fall)","multiplier":5,"notes":"Vehicle collisions accounted for the majority of e-scooter fatalities; the energy differential makes head protection critical but often insufficient"},{"factor":"drug use at the time of riding","multiplier":2.7,"notes":"JAMA Network Open 2024 (Fernandez et al., NEISS) found drug use at the time of injury carried an adjusted odds ratio of 2.70 (95% CI 2.26-3.23) for hospitalization across all micromobility vehicle types -- a distinct axis from alcohol (AOR 1.71)"}],"short_label":"E-scooter no helmet","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 22% head injury rate is conditional on presenting to an ED after a crash, not a per-trip probability. Many minor crashes (scrapes, bruises) never reach a hospital. The unhelmeted rider population dominates the data (80-96% of riders), so the baseline head injury rate effectively IS the unhelmeted rate. True per-trip risk is hard to estimate because total trip counts are uncertain. The CPSC ED visit figures use NEISS projections that carry sampling uncertainty. E-scooter injury epidemiology is a rapidly evolving field -- adoption is growing faster than the evidence base, and most studies cover 2018-2023 data from early-adoption cities. The CDC / Austin study's ~48% head-injury and ~15% TBI shares come from a severe-presentation cohort (EMS/ED) and run higher than the 22% global systematic-review pool; both are conditional on a crash reaching care, not per-trip rates. The drug-use and alcohol odds ratios from the JAMA NEISS study describe hospitalization odds given an injury across pooled micromobility vehicles, so they are approximate when applied to e-scooter head injury specifically.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"An electric scooter parked on a sidewalk next to a bicycle helmet resting on the ground, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/e-scooter-helmetless-head-injury","api_url":"https://likelier.app/api/fears/e-scooter-helmetless-head-injury.json"},{"slug":"eating-after-child-infection","question":"What are the odds of getting a lasting infection from sharing food or drinks with your child?","quick_answer":"For the highest-risk subgroup, a CMV-seronegative parent with an actively shedding child, the odds of a lasting CMV infection are about 24 percent per year of exposure.\n","category":"health","tags":["kids","food"],"no_reliable_estimate":false,"perceived":{"description":"The fear surfaces with a predictable trigger: a toddler hands a parent a half-eaten cracker, or drinks from a shared cup, and a moment of reluctance flickers — all that saliva, all those daycare germs. Parenting forums are full of conflicted posts about whether to refuse the offer and feel cold, or accept it and feel vaguely contaminated. The intuitive risk estimate tends to be framed as \"a reasonable chance of catching a stomach bug within a week,\" which captures something real about short-term illness transmission but misses the more interesting question: is there a lasting infection risk from this habit, not just another cold? Most parents simultaneously underestimate the specific long-term pathogens (CMV, H. pylori) and overestimate the uniqueness of the sharing route — much of the transmission that happens via shared food would also happen through household air, hand contact, and surface touch anyway.\n","kind":"intuition"},"native":{"display":"~24% per year of exposure (CMV, for seronegative parent with shedding child)","numerator":24,"denominator":100,"unit":"per year of household exposure","population":"CMV-seronegative parents of toddlers with confirmed CMV shedding (Cannon et al. 2010 review)"},"normalized":{"lifetime_us_adult":0.24,"display":"~24% annual seroconversion risk (CMV) for the highest-risk subgroup","log_value":-0.62,"assumptions":"The most precisely quantified pathway is CMV (cytomegalovirus) transmission from a shedding toddler to a seronegative parent. Cannon et al. (2010), reviewing the literature, found annual CMV seroconversion rates of 24% (95% CI 18–30%) for seronegative parents of a child actively shedding CMV, versus 2.1% in parents whose child was not shedding. Because CMV is present in saliva and is efficiently transmitted via shared cups, food, and hand-to-mouth contact, this is the most plausible causal pathway from the specific behavior in question. Approximately 45% of US adults are CMV-seronegative (susceptible); of young children in daycare, roughly 23% are actively shedding at any given time. For the typical parent without a confirmed shedding child, the annual seroconversion risk is much lower (~2–10%). The normalized.lifetime_us_adult field is set to 0.24 to reflect the clinically important subgroup (seronegative parent + shedding child); the scope is subgroup_lifetime. For H. pylori, intrafamilial transmission is documented but harder to isolate as an annual rate because most US adults were either already infected in childhood or will never acquire it in adulthood regardless of child contact. No single per-year figure is authoritative for H. pylori child-to-parent transmission.\n","uncertainty":{"low":0.18,"high":0.3},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20645278/","title":"Cytomegalovirus seroconversion rates and risk factors: implications for congenital CMV","publisher":"Reviews in Medical Virology (Wiley/PubMed) — Cannon MJ, Hyde TB, Schmid DS","source_type":"peer_reviewed","statistic":"Annual CMV seroconversion rate 24% (95% CI 18–30%) for seronegative parents with a CMV-shedding child; 2.1% (95% CI 0.3–6.8%) when child not shedding; ~50% of seronegative mothers with an infected child under age 2 in daycare seroconvert within 1 year","excerpt":"\"Among parents of children infected with CMV, the annual CMV seroconversion rate was 24% (95% CI: 18–30%) for seronegative parents versus 2.1% (95% CI: 0.3–6.8%) for parents of uninfected children.\"\n","source_date":"2010-07-01","source_accessed":"2026-05-02","archive_url":"https://web.archive.org/web/20260503080522/https://pubmed.ncbi.nlm.nih.gov/20645278/","calculation_notes":"Cannon et al. reviewed prospective studies of CMV seroconversion in parents of young children. The 24% annual rate applies to seronegative parents whose child has been confirmed as actively shedding CMV (via urine or saliva culture). This is the rate for any seroconversion (acquiring CMV antibodies), which in immunocompetent adults is typically asymptomatic or produces a brief mononucleosis-like illness. The clinical significance is highest for seronegative pregnant women, for whom primary CMV infection carries a 30–50% fetal transmission rate and is the leading infectious cause of congenital disability in the US. ~45% of US adults are CMV-seronegative; ~23% of young children in daycare are actively shedding CMV at any given time — both background prevalence figures are synthesized from the same Cannon et al. review.\n","independence_note":"Cannon et al. is a systematic literature review, independent of the H. pylori and RSV sources below; it pooled original prospective cohort studies. The CMV biology is well-replicated and not controversial.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/12237602/","title":"Role of infected parents in transmission of Helicobacter pylori to their children","publisher":"Gastroenterology (American Gastroenterological Association)","source_type":"peer_reviewed","statistic":"Children of saliva-positive mothers had OR 3.9 (95% CI 1.4–10.6) for H. pylori infection; H. pylori prevalence in children: 17–19% if parent was saliva-positive vs. 5–7% if parent was seronegative","excerpt":"\"When the mother was saliva-positive for Helicobacter pylori, the child's infection OR was 3.9 (95% CI 1.4–10.6). H. pylori prevalence in children was 17.3–19.1% when a parent was saliva-positive versus 5.1–6.8% when parents were seronegative.\"\n","source_date":"2002-09-01","source_accessed":"2026-05-02","archive_url":"https://web.archive.org/web/20260505053700/https://pubmed.ncbi.nlm.nih.gov/12237602/","calculation_notes":"This study demonstrates bidirectional saliva-route transmission of H. pylori within families — the association documented here is parent-to-child, but the same mechanism operates in reverse. H. pylori prevalence among US adults is approximately 30–37% overall, higher in older cohorts. For adults who are seronegative, having an infected child roughly doubles to quadruples the intrafamilial transmission risk, but most H. pylori acquisition in the US occurs in childhood (before age 10), so the marginal adult risk from child contact is difficult to isolate as an annual rate. Reported here primarily as mechanistic confirmation that saliva is an established transmission route for a pathogen with real long-term consequences (peptic ulcer in 10–20%, gastric cancer in ~1–2% of infected persons over decades).\n","independence_note":"Independent of the Cannon CMV review and the RSV transmission source; different pathogen, different study design (cross-sectional H. pylori prevalence and saliva culture), different research group.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4396434/","title":"Transmission of Respiratory Syncytial Virus Infection Within Families","publisher":"PMC/NIH — Munywoki PK et al.","source_type":"peer_reviewed","statistic":"RSV detected in 77% of families within 1 week of index child's hospitalization; RSV detected in 47% of individual family members at that point; children are major amplifiers of RSV within households","excerpt":"\"RSV was detected in 77% of families and in 47% of individual family members within one week of an index child's RSV hospitalization. Young children are important sources of RSV transmission to other household members.\"\n","source_date":"2015-04-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20250403193501/https://pmc.ncbi.nlm.nih.gov/articles/PMC4396434/","calculation_notes":"RSV in healthy adults causes cold-like illness; clinically significant (hospitalization risk) primarily for adults over 65 or immunocompromised. This study quantifies household secondary attack rates but does not isolate the shared-food/cup route from droplet and contact transmission, which are the primary RSV routes. Included to establish that child-to- parent infection transmission across multiple pathogens is not a theoretical concern — nearly half of exposed household members acquire infection during a child's acute RSV illness.\n","independence_note":"Independent of the CMV and H. pylori sources; distinct pathogen, distinct research group (Kenyan household cohort), distinct study design.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6009a2.htm","title":"Premastication of Food by Caregivers of HIV-Exposed Children — Nine U.S. Sites, 2009–2010","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"31% of primary caregivers reported children received pre-chewed food; pre-chewing linked to transfer of HIV (via blood), hepatitis B, Streptococcus mutans, and other pathogens","excerpt":"\"Overall, 48 (31%) of 154 primary caregivers stated that they or someone else had premasticated food for the child... In addition to HIV, transmission of hepatitis B virus and group A streptococcus by premastication has been documented.\"\n","source_date":"2011-03-11","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260218212452/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6009a2.htm","calculation_notes":"CDC MMWR establishes that saliva-sharing behaviors (specifically premastication, the most extreme form) are common, documented transmitters of multiple pathogens. The CDC does not provide a per-event infection probability for immunocompetent adults, but documents the transmission mechanism and affected pathogen list. This source anchors the \"real transmission risk\" framing and confirms that the behavior is common enough to study at a public health level.\n","independence_note":"CDC MMWR is independent of the three peer-reviewed sources above. Provides public health surveillance framing rather than mechanistic transmission quantification.\n"},{"url":"https://www.cdc.gov/cytomegalovirus/congenital-infection/hearing-loss.html","title":"Congenital CMV and Hearing Loss","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CMV is the most common infectious cause of birth defects in the United States; hearing loss is common in babies with congenital CMV, even those without symptoms at birth","excerpt":"\"CMV is the most common infectious cause of birth defects in the United States. [...] Hearing loss is common in babies with congenital CMV, even in those without symptoms at birth.\"\n","source_date":"2024-05-10","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260616152856/https://www.cdc.gov/cytomegalovirus/congenital-infection/hearing-loss.html","calculation_notes":"Confirms the body-text framing that congenital CMV carries a documented, CDC-recognized public health burden (most common infectious cause of birth defects, common cause of infant hearing loss) independent of the Cannon et al. seroconversion-rate literature cited above; this is \"why it matters\" context for the seronegative- pregnant-woman subgroup, not a numeric seroconversion estimate.\n","independence_note":"CDC public health consumer-guidance source, independent of the peer-reviewed transmission-rate literature (Cannon et al., the H. pylori and RSV studies) above; distinct from the MMWR premastication source (different CDC publication, different topic).\n"}],"comparison_anchors":[{"label":"Gastroenteritis from restaurant food (annual, US adult)","lifetime_us_adult":0.17},{"label":"CMV seroconversion (general population, per year)","lifetime_us_adult":0.02},{"label":"H. pylori lifetime prevalence (US adults)","lifetime_us_adult":0.35}],"personal_factor_multipliers":[{"factor":"Seronegative pregnant woman with child in daycare","multiplier":12,"notes":"The combination of CMV-seronegative status + child actively shedding CMV (23% prevalence in daycare children) + the fetal transmission risk (~30–50% of primary maternal CMV) represents the highest-stakes version of this risk. CDC and ACOG both note congenital CMV as the leading infectious cause of birth defects in the US."},{"factor":"Child just returned from daycare sick","multiplier":5,"notes":"Active viral shedding during symptomatic illness maximizes saliva viral load. Transmission probability per exposure event is higher when the child is symptomatic, though CMV shedding also occurs asymptomatically."},{"factor":"Already CMV-seropositive parent","multiplier":0.05,"notes":"Prior CMV infection confers strong (not absolute) immunity; seropositive parents face negligible risk of primary CMV from child contact. They may experience reactivation, which is typically subclinical in immunocompetent adults. This is the majority of US adults (55% are CMV-seropositive by age 40)."},{"factor":"Immunocompromised parent (transplant, chemotherapy, HIV)","multiplier":5,"notes":"CDC and IDSA guidelines: immunocompromised individuals face substantially higher susceptibility to CMV (both primary acquisition and reactivation), RSV (hospitalization risk in adults), and other pathogens transmitted via child saliva; the ~5x multiplier reflects the convergence of higher acquisition probability and more severe disease course, not a single published rate for this exact exposure scenario"},{"factor":"No hand washing after handling child's saliva/urine","multiplier":2.5,"notes":"CDC CMV prevention guidance and the Cannon et al. 2010 review: hand hygiene after contact with a young child's saliva and urine is the primary recommended mitigation for CMV transmission; studies of seronegative daycare workers and parents consistently show that careful hand hygiene roughly halves acquisition risk, implying that the absence of this practice approximately doubles it"}],"short_label":"Infection from sharing food with child","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The 24% annual seroconversion figure applies to a specific high-risk subgroup: seronegative parents of children confirmed to be actively shedding CMV. For the broader population of parents sharing food with young children, the annual risk of acquiring a lasting infection from this behavior specifically (as opposed to other household routes) cannot be cleanly isolated. Most household pathogen transmission — RSV, common cold viruses, CMV — occurs primarily via droplets and hand-to-face contact, not exclusively through shared food and saliva; the shared-food route is a contributor, not the sole pathway. For immunocompetent non-pregnant adults, CMV acquisition is typically asymptomatic or produces a transient mononucleosis-like illness; the lasting harm framing applies mainly to congenital CMV (seronegative pregnant women) and to H. pylori's long-term ulcer and cancer sequelae. H. pylori annual incidence in US adults is very low (~0.5% per year overall) because most transmission occurs in childhood; adult acquisition via child contact is biologically plausible but difficult to quantify as an isolated probability. RSV and common cold transmission from a sick child to a parent is very common (~47% per acute illness episode) but causes self-limiting illness in healthy adults, not lasting infection. The \"lasting infection\" fear is most empirically supported for CMV in seronegative pregnant women and H. pylori in seronegative adults with a H. pylori-positive child.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A half-eaten cracker on a small plate beside a child's sippy cup, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/eating-after-child-infection","api_url":"https://likelier.app/api/fears/eating-after-child-infection.json"},{"slug":"e-bike-helmetless-head-injury","question":"What are the odds of serious head injury riding an e-bike without a helmet?","quick_answer":"About 25 percent of e-bike crash injuries involve the head, a per-crash figure rather than a lifetime risk, and unhelmeted riders are roughly twice as likely to suffer head trauma. That is higher than the 5 to 10 percent per crash many riders assume.\n","category":"transport","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"E-bikes inhabit the same mental category as regular bicycles, and most riders carry over the same casual attitude toward helmets. Only about 44% of hospitalized e-bike riders were wearing helmets in the largest US study, and the rate has been declining roughly 6% per year since 2017. The cultural framing treats an e-bike as \"a bicycle with a boost\" rather than what it functionally is: a vehicle that routinely cruises at 20-28 mph with a rider whose protective equipment was designed for 14 mph impacts.\n","rough_estimate":"~5-10% chance of head injury per crash (same as a bicycle)","kind":"intuition"},"native":{"display":"~25% of e-bike crash injuries involve the head; unhelmeted riders are ~2x more likely to suffer head trauma","numerator":1,"denominator":4,"unit":"proportion of e-bike crash injuries involving the head (across multiple studies)","population":"e-bike riders presenting to emergency departments, US and European studies (2017-2024)"},"normalized":{"lifetime_us_adult":0.25,"display":"~25% of e-bike crash injuries affect the head (per crash event, not per US adult)","log_value":-0.6,"assumptions":"Head injury rates in e-bike crashes range from 18% (CPSC population-level NEISS data) to 40%+ (trauma center studies). JAMA Surgery (Feb 2024) found head trauma from e-bike crashes increased 49-fold from ~163 cases in 2017 to ~7,922 in 2022. Only 44% of hospitalized e-bike riders were helmeted. Unhelmeted riders were approximately twice as likely to suffer head injuries. A Dutch study found e-bike use was an independent predictor of severe TBI (OR 1.64) and skull fractures (OR 1.50) compared to regular cycling. We use 25% as the central estimate, splitting the difference between population-level and trauma-center data. This is a per-crash conditional probability. No per-mile exposure-adjusted rate exists because total e-bike miles ridden are not systematically collected. CPSC recorded 193 e-bike fatalities (2017-2023), accounting for 52% of all micromobility deaths despite e-bikes being a fraction of micromobility trips. E-bike injuries in US EDs rose from 751 in 2017 to 23,493 in 2022.\n","uncertainty":{"low":0.15,"high":0.4},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10882498/","title":"Electric Bicycle Injuries and Hospitalizations in the United States","publisher":"JAMA Surgery (Feb 2024)","source_type":"peer_reviewed","statistic":"E-bike head trauma increased 49-fold (2017-2022); only 44% of hospitalized riders wore helmets; unhelmeted riders were nearly twice as likely to suffer head injuries","excerpt":"\"Head trauma related to electric bicycle crashes increased approximately 49-fold from 2017 to 2022. Only 44 percent of hospitalized e-bike riders were wearing helmets. Helmet use among e-bike crash patients declined approximately 6 percent per year over the study period.\"\n","source_date":"2024-02-21","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260220005359/https://pmc.ncbi.nlm.nih.gov/articles/PMC10882498/","calculation_notes":"JAMA Surgery analysis of NEISS data. Head trauma cases rose from ~163 (2017) to ~7,922 (2022). The 49-fold increase reflects both explosive adoption and the concentration of injuries among unhelmeted riders. The 44% helmet rate among hospitalized riders suggests helmets are underrepresented in the injury pool relative to general ridership, consistent with a protective effect. The 6% annual decline in helmet use is concerning given the speed differential.\n"},{"url":"https://www.cpsc.gov/s3fs-public/Micromobility-Products-Related-Deaths-Injuries-and-Hazard-Patterns_2017-2023.pdf","title":"Micromobility Products-Related Deaths, Injuries, and Hazard Patterns, 2017-2023","publisher":"U.S. Consumer Product Safety Commission (CPSC)","source_type":"govt_report","statistic":"193 e-bike fatalities out of 373 micromobility deaths (2017-2023); e-bike ED injuries increased tenfold from ~3,400 to ~34,200","excerpt":"\"E-bikes accounted for 193 of 373 micromobility-related fatalities from 2017 through 2023, representing 52 percent of all micromobility deaths. E-bike-related emergency department visits increased approximately tenfold over the period.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20250809053021/https://www.cpsc.gov/s3fs-public/Micromobility-Products-Related-Deaths-Injuries-and-Hazard-Patterns_2017-2023.pdf","calculation_notes":"193 deaths over 7 years = ~28/year, heavily back-loaded as adoption surged. E-bikes are 52% of micromobility deaths despite being a smaller share of trips than e-scooters. The tenfold ED visit increase (3,400 to 34,200) tracks adoption growth. E-bike riders had the highest proportion of motor vehicle involvement (35.4%) among micromobility devices, reflecting their use on roads alongside traffic. NHTSA does not separately track e-bike fatalities in FARS, so the CPSC figure is likely an undercount.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40268590/","title":"E-bikers at risk for severe traumatic brain injury and skull fractures compared to conventional cyclists","publisher":"PubMed / European Journal of Trauma and Emergency Surgery","source_type":"peer_reviewed","statistic":"E-bike use was an independent predictor of severe TBI (OR 1.64) and skull fractures (OR 1.50) compared to regular cycling","excerpt":"\"E-bike use was an independent predictor of severe traumatic brain injury with an odds ratio of 1.64 and skull fractures with an odds ratio of 1.50 compared to conventional cycling, after adjusting for age, sex, and helmet use.\"\n","source_date":"2025-04-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260505053455/https://pubmed.ncbi.nlm.nih.gov/40268590/","calculation_notes":"Dutch study comparing e-bike and conventional bicycle crash outcomes. The OR of 1.64 for severe TBI means e-bikers face 64% higher odds of severe brain injury per crash, independent of helmet use. This is attributable to higher impact speeds: e-bikes cruise at 15-28 mph vs 10-14 mph for regular bicycles. Kinetic energy scales with velocity squared, so a 25 mph impact delivers ~2.8x the energy of a 15 mph impact. Standard bicycle helmets are tested at ~14 mph (CPSC standard), below typical e-bike cruising speed.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/41390298/","title":"Electric-bicycles and speed-related trauma in pediatrics: Risk of internal injury and hospitalization","publisher":"Injury (Flyer et al., 2026)","source_type":"peer_reviewed","statistic":"Speed-related pediatric e-bike crashes carried head/neck/facial injuries in 49.1% vs 28.7% of other crashes, internal organ injuries in 24.1% vs 10.4%, and hospital admission in 7.3% vs 4.7%; national NEISS estimate of 15,121 pediatric patients (2019-2023), 71.3% aged 13-18","excerpt":"\"Injuries attributed to speed were associated with a higher incidence of head, neck, or facial injuries (49.1% vs 28.7%). A greater proportion of children with speed-related injuries sustained internal organ injuries (24.1% vs. 10.4%) and were admitted to the hospital (7.3% vs. 4.7%).\"\n","source_date":"2026-05-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260226092509/https://pubmed.ncbi.nlm.nih.gov/41390298/","calculation_notes":"Flyer et al., Injury 2026;57(5):112931 (Epub 2025-12-04), NEISS-based national estimate of 15,121 pediatric e-bike injury patients (79.7% male, 71.3% aged 13-18). The head/neck/facial ratio 49.1/28.7 = ~1.71x supports the child/teen speed-related multiplier. Internal-organ (24.1/10.4 = ~2.3x) and admission (7.3/4.7 = ~1.55x) ratios corroborate that speed escalates severity, not just frequency. The 71.3% adolescent share is pool composition, not a risk ratio, so no teen-vs-adult multiplier is derived from it. No per-child-rider denominator exists (e-bike miles ridden by children are not collected), so this study reinforces the per-crash conditional rather than revising the headline 25%.\n"}],"comparison_anchors":[{"label":"Cycling without helmet head injury (per crash)","lifetime_us_adult":0.17},{"label":"E-scooter head injury without helmet (per crash)","lifetime_us_adult":0.22},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"wearing a standard bicycle helmet (CPSC-rated)","multiplier":0.5,"notes":"Helmets roughly halve head injury risk, but standard bicycle helmets are tested at 14 mph -- below typical e-bike speeds. Protection is partial at 20+ mph."},{"factor":"wearing an NTA 8776 or moped-rated helmet","multiplier":0.3,"notes":"NTA 8776 helmets are tested at 21% higher impact speeds than CPSC bicycle helmets; recommended for Class 3 e-bikes and S-Pedelecs"},{"factor":"Class 3 e-bike or S-Pedelec (28 mph / 45 km/h)","multiplier":2,"notes":"Kinetic energy at 28 mph is ~3.5x that at 15 mph; severe TBI probability increases sharply above 20 mph"},{"factor":"rider aged 65+ (largest e-bike fatality demographic)","multiplier":2.5,"notes":"In the Netherlands, 75% of e-bike fatalities were in riders aged 65+; older riders have thinner skulls, slower reflexes, and higher vulnerability to TBI"},{"factor":"riding in mixed traffic (motor vehicle involvement)","multiplier":2,"notes":"35.4% of e-bike injuries involved motor vehicles -- the highest proportion among micromobility devices; vehicle collisions dramatically increase injury severity"},{"factor":"child/teen rider in a speed-related crash (>20 mph)","multiplier":1.7,"notes":"Flyer et al. 2026 (Injury 57(5):112931), national NEISS estimate of 15,121 pediatric e-bike patients: speed-attributed crashes carried head/neck/facial injuries in 49.1% vs 28.7% of other crashes (~1.7x). Measured in a pediatric/teen population (71.3% aged 13-18); the figure is head-or-neck-or-facial, broader than head injury alone."}],"short_label":"E-bike no helmet","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 25% head injury rate is conditional on presenting to an ED after a crash. Many minor e-bike incidents never reach a hospital. No per-mile or per-trip exposure- adjusted rate exists because total e-bike miles ridden are not systematically collected in any country. CPSC does not disaggregate e-bikes from regular bicycles in NEISS prior to 2017, and NHTSA's FARS still does not separately code e-bike fatalities, leading to systematic undercounting. The 193 death figure (CPSC 2017- 2023) is a floor, not a ceiling. Standard bicycle helmets provide meaningful but potentially insufficient protection at e-bike speeds; the gap between CPSC test speed (14 mph) and typical e-bike cruising speed (20-28 mph) is a recognized safety concern. The Dutch data showing 75% of e-bike fatalities in riders aged 65+ may not generalize to the US, where the age distribution of e-bike riders skews younger. A pediatric surge runs the opposite direction from the elderly fatality skew: a national NEISS estimate of 15,121 pediatric e-bike injury patients (2019-2023) found 71.3% were aged 13-18, and speed-attributed crashes carried higher rates of head/neck/ facial injury (49.1% vs 28.7%), internal organ injury (24.1% vs 10.4%), and hospital admission (7.3% vs 4.7%) than other crashes (Flyer et al. 2026). At least one pediatric trauma center (Rady Children's, San Diego) reports e-bike crashes as its single largest cause of ER trauma, and several US children's hospitals describe injury counts roughly tripling year-over-year, with only about a third of injured children wearing helmets. Those single-hospital and press figures show direction, not a per-child-rider rate.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"An electric bicycle parked next to a bike rack with a helmet hanging from the handlebars, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/e-bike-helmetless-head-injury","api_url":"https://likelier.app/api/fears/e-bike-helmetless-head-injury.json"},{"slug":"long-term-disability-working-age","question":"What is the chance I will become too disabled to work before I retire?","quick_answer":"More than 1 in 4 US workers become disabled before reaching retirement age (67) — roughly 34% of men and 29% of women.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"LIMRA surveys consistently find that roughly half of workers estimate their personal disability risk at 10% or less, with a median perceived lifetime risk well below 15%. Most workers picture disability as something caused by dramatic accidents or late-stage cancer, not the mundane musculoskeletal and mental-health conditions that drive the majority of claims. The modal underestimate places perceived lifetime risk around 8%, against a true figure of 25%+.\n","kind":"intuition"},"native":{"display":"more than 1 in 4 workers will experience a disability before reaching retirement age","numerator":1,"denominator":4,"unit":"workers aged 20 to Normal Retirement Age (67)","population":"US workers entering the workforce at age 20 (SSA actuarial cohort)"},"normalized":{"lifetime_us_adult":0.25,"display":"roughly 1 in 4 US workers will become disabled before retirement","log_value":-0.6,"assumptions":"SSA OACT 2025 actuarial tables for insured workers attaining age 20. The headline figure (1 in 4, or 25%) is the SSA's own published rounded estimate. Sex-specific estimates are ~34% male, ~29% female, yielding a combined ~31% weighted average; the headline 25%+ reflects SSA's conservative public communication. Normalized to a career-horizon scope (age 20 to NRA 67) rather than birth-to-77, hence activity_specific_lifetime scope. Uncertainty low = 0.22 (narrower \"unable to perform any substantial gainful activity\" definition); uncertainty high = 0.34 (male actuarial upper estimate from OACT).\n","uncertainty":{"low":0.22,"high":0.34},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.ssa.gov/disabilityfacts/facts.html","title":"The Faces and Facts of Disability","publisher":"Social Security Administration","source_type":"govt_report","statistic":"More than 1 in 4 of today's 20-year-olds will become disabled before reaching retirement age.","excerpt":"\"The sobering fact for 20-year-olds is that more than 1-in-4 of them becomes disabled before reaching retirement age.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20260113220242/https://www.ssa.gov/disabilityfacts/facts.html","calculation_notes":"SSA headline figure: 1 in 4 (25%) of 20-year-olds becomes disabled before NRA (age 67). This is the native statistic: 1/4 workers over a career horizon. SSA OACT 2025 actuarial tables refine this sex-specifically to ~34% of men and ~29% of women; the headline rounds to 25%+ for the combined cohort. Normalized lifetime_us_adult set to 0.25 (the SSA headline), consistent with the activity_specific_lifetime scope (working career from age 20 to NRA 67, not birth-to-77). Uncertainty low = 0.22 (lower bound allowing for definitional narrowing of \"disability\"); uncertainty high = 0.34 (SSA actuarial male upper estimate).\n"},{"url":"https://www.ssa.gov/OACT/NOTES/ran6/an2025-6.pdf","title":"Disability and Death Probability Tables for Insured Workers Attaining Age 20 in 2025","publisher":"Social Security Administration, Office of the Chief Actuary","source_type":"primary_study","statistic":"Probability of becoming disabled before NRA: ~34% for men, ~29% for women (attaining age 20 in 2025).","excerpt":"\"The probability of surviving from age 20 to Normal Retirement Age without ever being disabled is 66 percent for men and 71 percent for women.\"\n","source_date":"2025-06-01","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20260312075406/https://www.ssa.gov/OACT/NOTES/ran6/an2025-6.pdf","calculation_notes":"Complement of survival probability: 1 - 0.66 = 0.34 (men); 1 - 0.71 = 0.29 (women). Combined approximate: ~0.31 weighted average, consistent with the SSA headline of \"more than 1 in 4.\" The combined figure understates male risk and overstates female risk; the headline 25%+ figure reflects the SSA's own rounded public communication, not a recalculation.\n"},{"url":"https://thecdia.org/the-average-duration-of-long-term-disability-is-31-2-months/","title":"The Average Duration of Long-Term Disability is 31.2 Months","publisher":"Council for Disability Income Awareness (CDIA)","source_type":"reputable_reference","statistic":"Average long-term disability claim lasts approximately 31–35 months.","excerpt":"\"The average duration of a long-term disability claim is almost 35 months, according to the Council for Disability Awareness.\"\n","source_date":"2023-01-01","source_accessed":"2026-05-13","archive_url":"http://web.archive.org/web/20250525192821/https://thecdia.org/the-average-duration-of-long-term-disability-is-31-2-months/","calculation_notes":"This source provides claim duration context, not a probability figure. Used to establish that a qualifying disability event is not typically brief: the modal claim runs nearly three years, reinforcing the financial materiality of the risk.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Blue-collar / physical labor occupation","multiplier":1.7,"notes":"Ropponen et al. (2018), BMC Public Health — blue-collar workers face substantially higher disability and mortality risk than white-collar workers at equivalent work ability; SSDI applicants in blue-collar jobs are more than twice as likely to apply. Multiplier is a conservative mid-range estimate consistent with published odds ratios.\n"},{"factor":"Male sex (working-age)","multiplier":1.35,"notes":"SSA OACT 2025 actuarial tables — lifetime disability probability ~34% for men vs ~29% for women attaining age 20, yielding an approximate 1.17 raw ratio; industry LTD data shows men file at higher rates, consistent with a 1.3–1.4 range when controlling for workforce participation.\n"},{"factor":"Prior musculoskeletal disorder or back injury","multiplier":2.5,"notes":"NIOSH / GBD 2019 — musculoskeletal disorders are the leading cause of LTD claims (29%); prior injury is a strong recurrence predictor. Recurrence rates for compensable back injuries consistently show 2–3× elevated subsequent disability risk in prospective cohort studies.\n"},{"factor":"Mental health diagnosis (depression or anxiety disorder)","multiplier":1.8,"notes":"DOL ERISA Advisory Council (2023) — mental health conditions account for approximately 40% of LTD claims and are rising. Workers with prior depression or anxiety diagnoses face elevated risk of a qualifying disability episode; 1.8× is consistent with claims incidence data relative to the general workforce.\n"},{"factor":"No long-term disability insurance coverage","multiplier":1,"notes":"This is a financial-exposure factor rather than a probability multiplier; approximately 65% of private-sector workers lack LTD coverage (LIMRA 2023). Absence of insurance does not change the probability of becoming disabled but substantially worsens financial outcomes.\n"}],"outcome_severity":"serious_harm","outcome_type":"financial","valence":"negative","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-13","last_reviewed":"2026-05-13","reviewed":true,"generated_at":"2026-05-13","image":{"alt":"A desk calendar with pages fading to blank beside a pinned paycheck on a corkboard."},"canonical_url":"https://likelier.app/long-term-disability-working-age","api_url":"https://likelier.app/api/fears/long-term-disability-working-age.json"},{"slug":"lyme-disease-tick-bite","question":"What are the odds of getting Lyme disease from a tick bite?","quick_answer":"A single deer-tick bite transmits Lyme roughly 1.2-3.4% of the time in an endemic area; over a lifetime it's about 1 in 4 for a moderately outdoor-active adult in an endemic region. Endemic-area residents often overestimate the per-bite odds at 30-50%.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Lyme disease occupies a peculiar niche in risk perception: in the northeastern United States, it is a near-universal worry during tick season, with parents, hikers, and gardeners treating every embedded tick as a medical emergency. In non-endemic regions — the Mountain West, most of the South, the Pacific Northwest — the same tick bite barely registers. No rigorous national poll isolates \"fear of Lyme disease from a single tick bite\" as a standalone question, so the perceived estimate here is editorial intuition calibrated by regional conversation patterns rather than survey data.\n","rough_estimate":"endemic-area residents often guess 30–50% per bite; non-endemic residents guess ~0%","kind":"intuition"},"native":{"display":"~1.2–3.4% transmission rate per Ixodes scapularis bite (attachment <72 hrs, endemic area)","numerator":34,"denominator":1000,"unit":"per bite","population":"persons bitten by Ixodes scapularis (deer tick) in an endemic area of the northeastern/upper-midwestern US"},"normalized":{"lifetime_us_adult":0.25,"display":"~1 in 4 lifetime (US adult in endemic region, moderate outdoor exposure)","log_value":-0.6,"assumptions":"CDC estimates ~476,000 new US Lyme infections per year (2024 revised estimate, up from the earlier 30,000 confirmed-case figure). US population ~335 million, but roughly 95% of cases concentrate in 15 northeastern and upper-midwestern states with a combined population of ~115 million. Annual per-capita hazard in the endemic footprint ≈ 476,000 × 0.95 / 115,000,000 ≈ 3.93 × 10⁻³. Compounded over 59 adult years: 1 - (1 - 0.00393)^59 ≈ 0.207. Adjusting upward slightly for outdoor-active adults (gardeners, hikers, dog walkers) who accumulate more tick encounters than sedentary residents gives a central estimate of ~0.25, or about 1 in 4 lifetime for a moderately active adult in an endemic state. The uncertainty band spans the sedentary endemic resident (~0.15) to the avid outdoors person in peak-incidence counties (~0.40).\n","uncertainty":{"low":0.15,"high":0.4},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/lyme/data-research/facts-stats/index.html","title":"Lyme Disease: Data and Statistics","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC estimates approximately 476,000 people are diagnosed and treated for Lyme disease each year in the United States.","excerpt":"\"CDC estimates that approximately 476,000 people may get Lyme disease each year in the United States. This estimate was derived using methods including insurance claims data, clinical laboratory data, and self-reported physician-diagnosed cases.\"\n","source_date":"2024-03-11","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260412203222/https://www.cdc.gov/lyme/data-research/facts-stats/index.html","calculation_notes":"The 476,000/year estimate is the population-level anchor. Divided across the endemic-region population of ~115 million (the 15 states that account for ~95% of confirmed cases), the annual per-capita hazard is ~3.93 × 10⁻³. Compounded over 59 adult years: 1 - (1 - 0.00393)^59 ≈ 0.207. With a modest upward adjustment for moderate outdoor activity, the central lifetime estimate is ~0.25.\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJM200107123450201","title":"Prophylaxis with Single-Dose Doxycycline for the Prevention of Lyme Disease after an Ixodes scapularis Tick Bite","publisher":"New England Journal of Medicine","source_type":"peer_reviewed","statistic":"3.2% of placebo recipients developed erythema migrans after an Ixodes scapularis bite in a Lyme-endemic area; 0.4% of doxycycline recipients developed it.","excerpt":"\"Erythema migrans developed at the site of the tick bite in a significantly smaller proportion of the subjects in the doxycycline group than of those in the placebo group (1 of 235 subjects [0.4 percent] vs. 8 of 247 subjects [3.2 percent], P<0.04).\"\n","source_date":"2001-07-12","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251122182322/https://www.nejm.org/doi/full/10.1056/NEJM200107123450201","calculation_notes":"Nadelman et al. 2001 provides the per-bite baseline: 3.2% of placebo recipients (no prophylaxis) developed erythema migrans in an endemic area. This is the per-bite transmission rate used as the native figure. The doxycycline arm (0.4%) informs the prophylactic-doxycycline personal factor multiplier (~0.13×). Subjects had to present within 72 hours of tick removal, so this rate reflects bites with attachment duration up to ~72 hours, biased toward shorter attachments since people who find ticks quickly are over-represented.\n","independence_note":"RCT conducted in Westchester County, NY — independent clinical trial data, not derived from CDC surveillance.\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJMcp1314325","title":"Lyme Disease","publisher":"New England Journal of Medicine","source_type":"peer_reviewed","statistic":"Even in highly endemic areas, the risk of Lyme disease transmission from a recognized tick bite is low (1 to 3%); in one study the risk of erythema migrans was 25% if the tick had fed for 72 or more hours and 0% if it had fed for less than 72 hours.","excerpt":"\"Consequently, even in areas where Lyme disease is highly endemic, the risk of disease transmission from a recognized bite is low (1 to 3%). [...] In a study comparing doxycycline with placebo in participants who had been bitten by nymphal ticks for which the duration of feeding could be assessed, the risk of erythema migrans in the placebo group was 25% (3 of 12 bites) if the tick had fed for 72 or more hours and 0% (0 of 48 bites) if the tick had fed for less than 72 hours.\"\n","source_date":"2014-05-01","source_accessed":"2026-04-18","archive_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4487875/","calculation_notes":"Shapiro 2014 NEJM review states the overall risk of transmission from a recognized bite in endemic areas is low (1–3%), and reports a sharp attachment-duration threshold: in the cited placebo arm, the risk of erythema migrans was 25% if the tick had fed for 72 or more hours and 0% if it had fed for less than 72 hours. The review elsewhere notes that transmission from infected nymphal ticks generally occurs only after 36–48 hours of attachment. This underpins the regional_breakdown rows stratified by attachment duration and the personal_factor_multiplier for prompt tick removal vs prolonged attachment.\n","independence_note":"Narrative review synthesizing multiple clinical studies — independent secondary analysis, not derived from CDC surveillance data.\n"}],"comparison_anchors":[{"label":"Bee/wasp sting fatality (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Shark attack fatality (lifetime, US adult)","lifetime_us_adult":1.76e-7},{"label":"Melanoma from UV exposure (lifetime, US adult)","lifetime_us_adult":0.028}],"regional_breakdown":[{"region":"Endemic northeast/upper-midwest US (per bite, attachment <72 hrs)","probability":0.032,"notes":"Nadelman et al. 2001 placebo arm: 3.2% per bite. Most recognized bites involve shorter attachment."},{"region":"Endemic US (per bite, attachment >72 hrs / fully engorged)","probability":0.2,"notes":"Shapiro 2014: erythema migrans risk was 25% in the placebo arm when the tick had fed ≥72 hrs (vs 0% if <72 hrs); 0.20 is a conservative estimate for prolonged attachment."},{"region":"Endemic US adult (lifetime, moderate outdoor activity)","probability":0.25,"notes":"CDC 476K cases/yr across ~115M endemic-region population, compounded over 59 years with outdoor-activity adjustment."},{"region":"Non-endemic US (lifetime)","probability":0.005,"notes":"Remaining ~5% of US Lyme cases spread across ~220M people in non-endemic states; lifetime risk rounds to ~1 in 200."}],"personal_factor_multipliers":[{"factor":"Outdoor worker in endemic region (forestry, landscaping, farming)","multiplier":3,"notes":"Occupational tick exposure is several-fold higher than recreational; CDC occupational health data."},{"factor":"Avid hiker / hunter in endemic region","multiplier":2,"notes":"Regular woodland exposure during May–September peak nymphal season increases encounter rate."},{"factor":"Prompt tick removal (<24 hrs) after every outdoor session","multiplier":0.3,"notes":"Transmission is near-zero with attachment <36 hrs (Shapiro 2014); diligent tick checks dramatically reduce risk."},{"factor":"Prophylactic doxycycline after recognized bite","multiplier":0.13,"notes":"Nadelman et al. 2001: doxycycline reduced EM incidence from 3.2% to 0.4% (87% relative reduction)."},{"factor":"Resident of non-endemic US state","multiplier":0.02,"notes":"~5% of US cases across ~66% of the population; per-capita risk roughly 50× lower than endemic states."}],"short_label":"Lyme disease","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The normalized lifetime figure applies specifically to moderately active adults living in the 15-state endemic corridor (Connecticut to Minnesota). Outside that footprint, lifetime Lyme risk drops by roughly two orders of magnitude. The per-bite transmission rate depends critically on attachment duration: near-zero under 36 hours, rising steeply after 72 hours. Nymphal ticks (May–July) cause the majority of human infections because they are small enough to go unnoticed; adult ticks, being larger, are found and removed sooner. The 476,000/year CDC estimate includes clinically diagnosed cases that may not have confirmatory serology, so it is substantially higher than the ~30,000 confirmed cases reported through passive surveillance. Post-bite prophylactic doxycycline (single 200 mg dose within 72 hours) reduces transmission by ~87% per Nadelman et al. and is recommended by IDSA guidelines in endemic areas. No human Lyme vaccine is approved as of mid-2026: LYMErix was withdrawn in 2002, and the Pfizer/Valneva candidate VLA15 (now LB6V) reported roughly 74% efficacy in its Phase 3 VALOR trial in March 2026, with regulatory filings to the FDA and EMA planned for late 2026. This entry does not constitute medical advice.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single stylized deer tick on a pale leaf, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/lyme-disease-tick-bite","api_url":"https://likelier.app/api/fears/lyme-disease-tick-bite.json"},{"slug":"iron-deficiency-reproductive-age-women","question":"What are the odds a woman of reproductive age is iron deficient?","quick_answer":"About 25.5% of US women aged 12-49 — roughly 1 in 4 — are iron deficient at any given time (NHANES 2015-2016). Most people guess a 5-10% chance and treat it as rare, but it is the predictable result of monthly blood loss meeting a diet low in absorbable iron.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Iron deficiency reads to most people as a Victorian complaint -- pale heroines fainting on chaise longues -- or, at most, something a pregnant woman is told to watch. The modern intuition is that a normal diet covers it and that \"tired all the time\" is about sleep or stress, not blood chemistry. Few menstruating women would guess that roughly one in four of them is running low on iron at any given moment, and that the depletion is a direct arithmetic consequence of monthly blood loss rather than a rare disorder.\n","rough_estimate":"~5-10% chance for a typical woman","kind":"intuition"},"native":{"display":"~25.5% of US females aged 12-49 had iron deficiency (inflammation-adjusted serum ferritin <15 µg/L, NHANES 2015-2016)","numerator":25.5,"denominator":100,"unit":"cross-sectional prevalence of iron deficiency among reproductive-age US females","population":"US females aged 12-49 (NHANES 2015-2016, CDC/NCHS)"},"normalized":{"lifetime_us_adult":0.255,"display":"~1 in 4 US women aged 12-49 is iron deficient (25.5% point prevalence, NHANES 2015-2016)","log_value":-0.59,"assumptions":"The headline is the cited figure itself, un-inflated: Healthy People 2030 sets the national baseline at 25.5% of females aged 12-49 with iron deficiency in 2015-2016, defined as inflammation-adjusted serum ferritin <15 µg/L (CRP-corrected via the BRINDA method). That is a cross-sectional snapshot — at any single blood draw, roughly a quarter of reproductive-age women are below the threshold. Because iron deficiency in this group is driven by recurring menstrual blood loss and the demands of pregnancy, it recurs and remits across a ~35-year reproductive span, so the cumulative probability that a given woman is deficient at SOME point is plausibly higher than the 25.5% snapshot — but that cumulative figure is not separately measured, so it is left as the upper end of the uncertainty band rather than asserted as the headline. The wide band also spans the definitional spread: a stricter 2-of-3-biomarker definition puts non-pregnant prevalence near 9.8% (NHANES 1999-2006), while a higher ferritin cutoff (~25 µg/L) would push it above the headline. The scope is a reproductive-age-woman subgroup, not a general US-adult lifetime — for men and postmenopausal women the figure is a small fraction of this.\n","uncertainty":{"low":0.1,"high":0.4},"scope":"subgroup_lifetime"},"sources":[{"url":"https://odphp.health.gov/healthypeople/objectives-and-data/browse-objectives/nutrition-and-healthy-eating/reduce-iron-deficiency-females-aged-12-49-years-nws-17/data-methodology","title":"Reduce iron deficiency in females aged 12 to 49 years — NWS-17, Data Methodology","publisher":"Office of Disease Prevention and Health Promotion (Healthy People 2030), U.S. HHS","source_type":"govt_report","statistic":"Baseline: 25.5% of females aged 12-49 had iron deficiency in 2015-2016; national target 21.7%. Source: NHANES, CDC/NCHS.","excerpt":"\"Females 12 to 49 years are defined as having iron deficiency if they have a measurement of inflammation-adjusted serum ferritin less than 15µg/L.\"\n","source_date":"2020-08-18","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20260511054220/https://odphp.health.gov/healthypeople/objectives-and-data/browse-objectives/nutrition-and-healthy-eating/reduce-iron-deficiency-females-aged-12-49-years-nws-17/data-methodology","calculation_notes":"Healthy People 2030 objective NWS-17. Baseline 25.5% in 2015-2016, target 21.7%. Underlying data: National Health and Nutrition Examination Survey (NHANES), CDC/NCHS. The deficiency threshold is inflammation-adjusted serum ferritin <15 µg/L, adjusted for C-reactive protein using the BRINDA regression approach (a change from the Healthy People 2020 total-body-iron indicator). The 25.5% figure feeds both the native prevalence and the normalized headline directly (0.255), un-inflated; the cumulative-over-reproductive-years effect is carried in the upper uncertainty bound rather than the point estimate.\n"},{"url":"https://ods.od.nih.gov/factsheets/Iron-HealthProfessional/","title":"Iron — Health Professional Fact Sheet","publisher":"NIH Office of Dietary Supplements","source_type":"govt_report","statistic":"Pregnant women, women with heavy menstrual bleeding, and frequent blood donors are named among the groups more likely to have inadequate iron intakes / higher risk of iron deficiency.","excerpt":"\"Certain groups of people are more likely than others to have inadequate iron intakes, including pregnant women, infants and young children, women with heavy menstrual bleeding, and frequent blood donors. People with certain conditions, such as cancer, gastrointestinal disorders, and heart failure, may also have inadequate iron intakes.\"\n","source_date":"2025-09-04","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20260725194357/https://ods.od.nih.gov/factsheets/Iron-HealthProfessional/","calculation_notes":"The NIH ODS fact sheet grounds the risk-group breakdown that feeds the personal_factor_multipliers (heavy menstrual bleeding, pregnancy, blood donation, vegetarian diet). It also supplies the measurement caveat: hemoglobin and hematocrit are insensitive screens, and serum ferritin — the marker used by Healthy People 2030 — is confounded by inflammation, which is why the CDC uses CRP adjustment. It does not itself state the 25.5% headline figure; that comes from the Healthy People 2030 / NHANES source above.\n","independence_note":"NIH ODS is editorially independent of the ODPHP Healthy People objective; both draw on NHANES but are separately authored government references."},{"url":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0112216","title":"Iron Status and Reproduction in US Women: National Health and Nutrition Examination Survey, 1999-2006","publisher":"PLOS ONE (Mei, Cogswell, et al.)","source_type":"peer_reviewed","statistic":"Iron deficiency 9.8% in non-pregnant women vs 25.4% in pregnant women (NHANES 1999-2006), by a 2-of-3 biomarker definition.","excerpt":"\"rates of iron deficiency were 9.8% in non-pregnant women and 25.4% in pregnant women.\"\n","source_date":"2014-11-13","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20250430115816/https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0112216","calculation_notes":"NHANES 1999-2006. Iron deficiency defined as two of three of hemoglobin, serum ferritin (<12 ng/mL), and transferrin saturation (<16%) below cutoff — a stricter, older definition than the single-ferritin<15 threshold used by Healthy People 2030, which is why its non-pregnant figure (9.8%) is lower than the 25.5% headline. The pregnant-vs-non-pregnant contrast (25.4% vs 9.8%, a ~2.6x ratio) grounds the pregnancy multiplier and the cumulative-lifetime adjustment: even by the conservative definition, pregnancy alone lifts point prevalence to a quarter.\n","independence_note":"Peer-reviewed NHANES analysis, independent of the government fact sheets though built on overlapping survey cycles."}],"comparison_anchors":[{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33},{"label":"Depression (lifetime, US adult)","lifetime_us_adult":0.2},{"label":"Vitamin D deficiency without supplements (lifetime)","lifetime_us_adult":0.35}],"personal_factor_multipliers":[{"factor":"pregnancy","multiplier":2.6,"notes":"NHANES 1999-2006 (cited source): iron deficiency 25.4% in pregnant vs 9.8% in non-pregnant women (~2.6x). Expanded blood volume and fetal demand raise requirements to ~27 mg/day, which diet alone rarely meets — hence routine antenatal screening and supplementation guidance."},{"factor":"heavy menstrual bleeding","multiplier":2.5,"notes":"Directional estimate — magnitude not separately quantified in the cited ledger. NIH ODS names women with heavy menstrual bleeding among the highest-risk groups; menorrhagia (loss above ~80 mL per cycle) routinely outpaces dietary iron replacement, so it is the dominant driver of deficiency in this population. The ~2.5x value is a conservative directional figure grounded in that mechanism, not a cited relative risk."},{"factor":"vegetarian or vegan diet","multiplier":1.8,"notes":"Directional estimate. Non-heme plant iron is absorbed at roughly a third the rate of heme iron; the Institute of Medicine sets recommended iron intakes ~1.8x higher for vegetarians on that basis, which anchors this factor. Grounded in the absorption mechanism rather than a cited prevalence ratio."},{"factor":"frequent blood donor","multiplier":2,"notes":"Directional estimate. NIH ODS lists frequent blood donors among at-risk groups; each whole-blood donation removes ~200-250 mg of iron, enough to deplete stores faster than diet replaces them. The ~2x value is a conservative directional figure grounded in that iron-loss mechanism, not a cited relative risk."},{"factor":"postmenopausal or male (no menstrual loss)","multiplier":0.2,"notes":"Directional protective estimate. Without recurring menstrual loss, iron deficiency in men and postmenopausal women falls to low single-digit percentages (a small fraction of the reproductive-age-female baseline), and new deficiency in these groups is a flag to investigate gastrointestinal blood loss rather than diet."}],"short_label":"Iron gap (women)","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is scoped to females aged 12-49 — the population the deficiency burden actually falls on — not to a general US adult, for whom the figure would be a small fraction of this. \"Iron deficiency\" also depends heavily on where the ferritin line is drawn. Healthy People 2030 uses inflammation-adjusted ferritin <15 µg/L; some recent work argues the physiologically justified threshold for non-pregnant women is closer to ~25 µg/L, which would push prevalence higher still, while older WHO-style 2-of-3 definitions yield the lower ~10% non-pregnant figure. Iron deficiency is also a stage before iron-deficiency anemia: most deficient women are not yet anemic, and the symptom picture (fatigue, poor concentration, cold intolerance) overlaps with much else, so the condition is routinely missed. Ferritin is an acute-phase reactant that rises with inflammation, which is why the CDC applies CRP adjustment; an unadjusted ferritin can mask deficiency in someone with concurrent infection or obesity.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-07-18","methodology_version":"1.0"},"reviewer":"8d-eval-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-07-18","image":{"alt":"A stylized iron nail resting on a calendar page, muted red and grey flat vector illustration."},"canonical_url":"https://likelier.app/iron-deficiency-reproductive-age-women","api_url":"https://likelier.app/api/fears/iron-deficiency-reproductive-age-women.json"},{"slug":"infant-painkiller-prophylaxis","question":"What are the odds that giving an infant paracetamol prophylactically at vaccination blunts the immune response to the vaccine?","quick_answer":"Giving an infant paracetamol prophylactically around vaccination lowers the vaccine's antibody response by roughly 26% per series (measured for hepatitis B in adults, with consistent blunting seen in infant trials). Parents generally perceive the practice as near-zero risk and pure benefit.\n","category":"kids","tags":["infant"],"no_reliable_estimate":false,"perceived":{"description":"The intuition is almost universally pro-medication. Parents who pre-dose their infant with paracetamol before a vaccination appointment believe they are being thoughtful and protective — softening the expected fever and distress so the baby suffers less. The idea that this common, seemingly benign practice could reduce the very immunity the vaccine is designed to confer does not register in most parental mental models. No large-scale survey specifically asks whether parents anticipate any downside to prophylactic antipyretics around vaccination, so this is flagged as intuition — but the qualitative direction is clear: the perceived risk of prophylactic paracetamol is near-zero, and the perceived benefit is real fever prevention.\n","kind":"intuition"},"native":{"display":"~26% lower anti-HBs antibody GMC (adults, hepatitis B vaccine, prophylactic vs. no paracetamol)","numerator":26,"denominator":100,"unit":"percentage reduction in vaccine antibody geometric mean concentration","population":"healthy adults aged 18–48 receiving hepatitis B vaccination series; prophylactic paracetamol group vs. control group (Doedée et al. 2014 PLoS One, n=496)"},"normalized":{"lifetime_us_adult":0.26,"display":"~26% reduction in vaccine antibody GMC per vaccination series with prophylactic paracetamol","log_value":-0.59,"assumptions":"The normalized figure represents the magnitude of antibody blunting rather than a binary event probability. Doedée et al. (2014, PLoS One) is the most precisely quantified single source: prophylactic paracetamol (given at vaccination, not reactively) reduced anti-HBs geometric mean concentration by 26% versus no paracetamol in a 496-person adult RCT (5768 mIU/mL in controls vs. 4257 mIU/mL in the prophylactic group, p=0.048). Therapeutic paracetamol (given reactively, 6+ hours after vaccination) had no significant effect (p=0.34), establishing timing as the decisive variable. In the foundational infant RCT (Prymula et al. 2009, Lancet, n=459 Czech infants), prophylactic paracetamol reduced antibody GMCs for all 10 pneumococcal conjugate vaccine serotypes plus Hib, diphtheria, tetanus, and pertactin after the primary series, with persistence to the booster for several antigens. Seroprotection thresholds were maintained for most antigens in most studies; the one documented instance of seroprotection itself being significantly reduced is for serotype 6B in the Prymula 2009 infant cohort (exact % not published in the abstract). Wysocki et al. (2017, Vaccine) confirmed GMC reduction for 5 of 13 PCV13 serotypes but maintained seroprotection for all antigens. The 0.26 point estimate anchors to the Doedée 2014 HBV figure as the best single-antigen quantification; the probability of experiencing measurable GMC blunting for at least one antigen across a full infant primary series is effectively certain based on consistent RCT findings, but whether this translates to below-seroprotection titers for any individual child cannot be derived from the available abstract- level data. Uncertainty band 0.05–0.80 reflects the spread between the \"some blunting is near-certain\" upper edge and the \"clinical vaccine failure is rare\" lower edge of the evidence.\n","uncertainty":{"low":0.05,"high":0.8},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/19837254/","title":"Effect of prophylactic paracetamol administration at time of vaccination on febrile reactions and antibody responses in children: two open-label, randomised controlled trials","publisher":"The Lancet (Prymula R et al.)","source_type":"peer_reviewed","statistic":"Fever ≥38°C: 42% (paracetamol) vs. 66% (control) after primary series; antibody GMCs significantly lower in the prophylactic paracetamol group for all 10 pneumococcal serotypes, protein D, anti-PRP (Hib), anti-diphtheria, anti-tetanus, and anti-pertactin after primary vaccination; seroprotection for serotype 6B significantly lower in paracetamol group","excerpt":"\"Although febrile reactions significantly decreased, prophylactic administration of antipyretic drugs at the time of vaccination should not be routinely recommended since antibody responses to several vaccine antigens were reduced.\"\n","source_date":"2009-10-17","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505055748/https://pubmed.ncbi.nlm.nih.gov/19837254/","calculation_notes":"Prymula 2009 is the foundational RCT for this finding: 459 healthy infants randomized to prophylactic paracetamol (three doses every 6–8 hours post-vaccination) vs. no prophylactic treatment across two consecutive trials (primary series + booster). The fever reduction (42% vs. 66%) confirms the drug works for its stated purpose, making the antibody-blunting finding all the more counterintuitive. The abstract confirms significance for all 10 pneumococcal serotypes plus 4 bacterial antigens; the specific GMC percentage reductions per serotype are in the paywalled full text. Serotype 6B is the only antigen for which seroprotection itself (not just GMC) was significantly lower in the paracetamol group — the most clinically concerning finding. Used as the primary anchor for the consistent-blunting-in-infants evidence.\n","independence_note":"Prymula 2009 is an independent prospective RCT, distinct from the Doedée 2014 HBV adult study below (different population, different vaccine, different lead investigator and centre); methodology is independently verifiable.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4045752/","title":"Effects of Prophylactic and Therapeutic Paracetamol Treatment during Vaccination on Hepatitis B Antibody Levels in Adults: Two Open-Label, Randomized Controlled Trials","publisher":"PLoS One (Doedée AMCM et al.)","source_type":"peer_reviewed","statistic":"Anti-HBs GMC at one month post-second booster: 5768 mIU/mL (control) vs. 4257 mIU/mL (prophylactic paracetamol), a 26% reduction, p=0.048; therapeutic paracetamol: 4958 mIU/mL, not significant (p=0.34); all groups maintained seroprotection (≥10 IU/L) after full series","excerpt":"\"Only prophylactic paracetamol treatment, and not therapeutic treatment, during vaccination has a negative influence on the antibody concentration after hepatitis B vaccination in adults. These findings prompt to consider therapeutic instead of prophylactic treatment to ensure maximal vaccination efficacy and retain the possibility to treat pain and fever after vaccination.\"\n","source_date":"2014-06-04","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505055746/https://pmc.ncbi.nlm.nih.gov/articles/PMC4045752/","calculation_notes":"This RCT provides the cleanest quantified figure: 26% reduction in anti-HBs GMC when paracetamol is given prophylactically (at time of vaccination) vs. not given. The key methodological contribution is the three-arm design (prophylactic, therapeutic, control), which isolates timing as the causal variable. Therapeutic paracetamol given 6–8 hours after vaccination produced no statistically significant GMC difference (p=0.34). Since all participants in both groups maintained seroprotection, the concern is about antibody durability and future protection breadth, not immediate vaccine failure. The 26% GMC reduction at one month post-series is the most citable single number in this literature and is used as the native stat anchor.\n","independence_note":"Conducted in adults (18–48 years) with hepatitis B vaccine, methodologically distinct from the Prymula 2009 infant pneumococcal study. Different lead author (Doedée), independent RCT with separate enrolment, randomisation, and vaccine type. Confirms the finding replicates across age groups and vaccine platforms.\n"},{"url":"https://www.cdc.gov/pinkbook/hcp/table-of-contents/chapter-6-vaccine-administration.html","title":"Vaccine Administration — Chapter 6 (CDC Pink Book)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC: prophylactic use of antipyretics at time of vaccination is not recommended; some studies suggest these medications might suppress immune response to some vaccine antigens","excerpt":"\"The prophylactic use of antipyretics (e.g., acetaminophen and ibuprofen) before or at the time of vaccination is not recommended. There is no evidence these will decrease the pain associated with an injection. In addition, some studies have suggested these medications might suppress the immune response to some vaccine antigens.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260406064024/https://www.cdc.gov/pinkbook/hcp/table-of-contents/chapter-6-vaccine-administration.html","calculation_notes":"CDC explicitly advises against prophylactic antipyretic use at vaccination in the authoritative Pink Book, citing immune-response suppression. This guidance was updated after the Prymula findings became established. The CDC distinguishes prophylactic use (not recommended) from therapeutic use (permissible when symptoms develop). Used as the primary authoritative public-health guidance anchor confirming the clinical concern is recognised by a major national immunisation programme.\n","independence_note":"Independent government guidance drawing on the broader immunisation literature; not a reanalysis of the Prymula RCT data.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/28262330/","title":"A randomized study of fever prophylaxis and the immunogenicity of routine pediatric vaccinations","publisher":"Vaccine (Wysocki J et al.)","source_type":"peer_reviewed","statistic":"Prophylactic paracetamol reduced IgG for 5 of 13 PCV13 serotypes (3, 4, 5, 6B, 23F); ibuprofen reduced antibody responses to pertussis FHA and tetanus; delayed administration (6–8 hours post-vaccination) showed no effect on vaccine responses for either drug; seroprotection maintained (≥0.35 µg/mL) across all groups","excerpt":"\"Prophylactic antipyretics affect immune responses to vaccines; these effects vary depending on the vaccine, antipyretic agent, and time of administration.\"\n","source_date":"2017-04-04","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505055822/https://pubmed.ncbi.nlm.nih.gov/28262330/","calculation_notes":"Wysocki 2017 replicates and refines the Prymula 2009 finding with PCV13 (13-valent rather than 10-valent) and includes ibuprofen as a separate arm. Key additional findings: (1) the blunting is restricted to 5 of 13 serotypes for paracetamol with PCV13, versus all 10 serotypes in Prymula 2009 — indicating vaccine formulation matters; (2) delayed dosing (6+ hours after vaccination) removes the blunting effect entirely; (3) ibuprofen blunts pertussis and tetanus but spares most pneumococcal serotypes. The maintained seroprotection in all groups is reassuring for clinical outcomes but does not negate the concern about antibody level durability over time. Used as the key replication study and timing-evidence source.\n","independence_note":"Independent Polish RCT with distinct enrolment, vaccine brand, and antipyretic arms from both the Prymula 2009 and Doedée 2014 studies. Converges on the same core finding via a different methodology.\n"}],"comparison_anchors":[{"label":"Serious adverse event from chronic OTC painkillers (cumulative, adult)","lifetime_us_adult":0.1},{"label":"Serious adverse event from routine vaccine (per dose)","lifetime_us_adult":0.0001},{"label":"Vaccine-preventable measles (unvaccinated child, US, lifetime)","lifetime_us_adult":0.03}],"personal_factor_multipliers":[{"factor":"Paracetamol given at or before vaccination vs. 6+ hours after","multiplier":10,"notes":"The blunting is specific to prophylactic (pre-emptive) dosing. Wysocki 2017 and Doedée 2014 both confirm that therapeutic dosing (given reactively when fever or pain develops) produces no statistically significant GMC difference. The entire documented risk disappears if timing is shifted."},{"factor":"10-valent vs. 13-valent pneumococcal vaccine","multiplier":2,"notes":"Prymula 2009 found blunting in all 10 serotypes with PHiD-CV; Wysocki 2017 found blunting in 5 of 13 with PCV13. Vaccine formulation modifies extent."},{"factor":"Ibuprofen vs. paracetamol","multiplier":0.5,"notes":"Wysocki 2017 found ibuprofen blunts pertussis FHA and tetanus but largely spares pneumococcal serotypes; different blunting profile from paracetamol."}],"short_label":"Painkiller before infant vaccination","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The entry documents a real and consistent finding — prophylactic paracetamol at the time of vaccination reduces antibody GMCs for multiple vaccine antigens in every RCT that has tested this — but the clinical magnitude is uncertain. Most children in the published studies still achieved accepted seroprotection thresholds; the concern is about antibody durability and potential long-term protection breadth rather than immediate vaccine failure. The critical variable is timing: therapeutic paracetamol given reactively when the child actually develops fever (6+ hours after vaccination) does not produce the same blunting effect in either the Doedée 2014 adult trial or the Wysocki 2017 infant trial. Ibuprofen has a different blunting profile from paracetamol — it spares most pneumococcal serotypes in Wysocki 2017 but reduces pertussis and tetanus responses. The UK is a notable exception: NHS guidelines recommend three doses of paracetamol after (not before) the MenB vaccine at 8 and 12 weeks because the MenB combination produces fever in over 50% of infants and the JCVI found no immunogenicity effect of post-vaccination paracetamol with that specific schedule. No published trial has demonstrated an increase in actual vaccine-preventable disease incidence in children whose parents gave prophylactic paracetamol — the evidence is at the surrogate endpoint (antibody titer) level.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A small dropper bottle of infant medicine drops beside a vaccine syringe, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/infant-painkiller-prophylaxis","api_url":"https://likelier.app/api/fears/infant-painkiller-prophylaxis.json"},{"slug":"loneliness-health-impact","question":"What are the odds of chronic loneliness causing serious health harm over a lifetime?","quick_answer":"Chronic loneliness carries about a 26% higher risk of premature death (29% for social isolation) among affected adults. Most people treat loneliness as merely unpleasant rather than a measurable mortality risk.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Most people register loneliness as an emotional state, not a medical risk factor. The cultural framing treats it as a personality problem or a phase rather than something that belongs on a doctor's intake form alongside blood pressure and cholesterol. The 2023 Surgeon General's advisory briefly moved the topic into mainstream awareness, but the headline comparison to \"smoking 15 cigarettes a day\" struck many as hyperbolic rather than literal. The result is a risk that is both widely experienced and systematically underweighted: surveys consistently find that more than half of US adults report feeling lonely, yet very few treat that loneliness as carrying a quantifiable mortality premium.\n","rough_estimate":"Most adults consider loneliness emotionally unpleasant but not a serious physical health threat","kind":"intuition"},"native":{"display":"26% increased risk of premature mortality for loneliness; 29% for social isolation","numerator":26,"denominator":100,"unit":"excess relative risk of premature death","population":"adults reporting chronic loneliness"},"normalized":{"lifetime_us_adult":0.26,"display":"~26% excess lifetime mortality risk (chronically lonely adults)","log_value":-0.59,"assumptions":"The headline figure comes from Holt-Lunstad et al. 2015, which found loneliness associated with OR 1.26 (26% increased odds of premature death) and social isolation with OR 1.29 (29% increase) across 70 studies and 3.4 million participants. We use 0.26 as the point estimate, representing the excess mortality risk attributable to chronic loneliness over an adult lifetime. This is a subgroup estimate: it applies to individuals who are chronically lonely, not a population average. The earlier Holt-Lunstad et al. 2010 meta-analysis of 148 studies (308,849 participants) found that stronger social relationships corresponded to a 50% increased likelihood of survival (OR 1.50), which is the basis for the oft-cited \"equivalent to smoking 15 cigarettes a day\" comparison. We use the more conservative 2015 loneliness-specific figure rather than the broader 2010 social- relationships figure. Scope is subgroup_lifetime: the excess risk for someone who is chronically lonely, not a population average including well-connected adults.\n","uncertainty":{"low":0.2,"high":0.32},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/25910392/","title":"Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review","publisher":"Perspectives on Psychological Science (Holt-Lunstad, Smith, Baker, Harris, Stephenson)","source_type":"peer_reviewed","statistic":"Loneliness OR 1.26 (26% increased mortality), social isolation OR 1.29, living alone OR 1.32; 70 studies, 3,407,134 participants","excerpt":"\"Across studies in which several possible confounds were statistically controlled for, the weighted average effect sizes were: social isolation OR = 1.29, loneliness OR = 1.26 and living alone OR = 1.32, corresponding to an average of 29%, 26%, and 32% increased likelihood of mortality respectively.\"\n","source_date":"2015-03-11","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260406063345/https://pubmed.ncbi.nlm.nih.gov/25910392/","calculation_notes":"Holt-Lunstad et al. 2015 is the primary basis for the normalized excess lifetime risk of ~26%. The meta-analysis covered 70 studies with 3.4 million participants and distinguished loneliness (subjective feeling), social isolation (objective lack of contacts), and living alone. We use the loneliness-specific OR of 1.26 for the headline figure, as the entry focuses on the subjective experience. The social isolation OR of 1.29 anchors the upper end of the uncertainty range. All effect sizes controlled for demographic confounds.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20668659/","title":"Social Relationships and Mortality Risk: A Meta-analytic Review","publisher":"PLoS Medicine (Holt-Lunstad, Smith, Layton)","source_type":"peer_reviewed","statistic":"OR 1.50 (95% CI 1.42-1.59) for survival with stronger social relationships; 148 studies, 308,849 participants","excerpt":"\"Data across 308,849 individuals, followed for an average of 7.5 years, indicate a 50% increased likelihood of survival for participants with stronger social relationships. This finding remained consistent across age, sex, initial health status, cause of death, and follow-up period.\"\n","source_date":"2010-07-27","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260319094126/https://pubmed.ncbi.nlm.nih.gov/20668659/","calculation_notes":"The 2010 meta-analysis is the source of the widely cited \"equivalent to smoking 15 cigarettes a day\" comparison: the OR 1.50 for weak social relationships was benchmarked against the known mortality effect sizes of smoking, obesity, and physical inactivity. The comparison is between relative risk magnitudes, not biological mechanisms. This broader measure (any social relationship deficit) yields a larger effect size than the 2015 loneliness-specific OR of 1.26, which is why we use the 2015 figure as the more conservative headline.\n","independence_note":"Holt-Lunstad 2010 and 2015 are by the same lead author but use different inclusion criteria and study pools. The 2010 review focused on any measure of social relationships; the 2015 review specifically separated loneliness, social isolation, and living alone. The 2015 study draws from a largely non-overlapping set of 70 studies compared to the 148 in 2010.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/37792968/","title":"Our Epidemic of Loneliness and Isolation: The U.S. Surgeon General's Advisory on the Healing Effects of Social Connection and Community","publisher":"US Department of Health and Human Services (Surgeon General Vivek Murthy)","source_type":"govt_report","statistic":"Loneliness increases premature death risk by 26%; social isolation by 29%; mortality impact comparable to smoking up to 15 cigarettes/day","excerpt":"\"The mortality impact of being socially disconnected is similar to that caused by smoking up to 15 cigarettes a day, and even greater than that associated with obesity and physical inactivity.\"\n","source_date":"2023-05-02","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260416143031/https://pubmed.ncbi.nlm.nih.gov/37792968/","calculation_notes":"The Surgeon General's advisory synthesizes the Holt-Lunstad meta-analyses and additional evidence. It does not produce new primary data but serves as the most authoritative US government endorsement of the loneliness-mortality link. The \"15 cigarettes a day\" comparison originates from the Holt-Lunstad 2010 benchmarking exercise. The advisory also reports that loneliness is associated with a 29% increased risk of coronary heart disease and a 32% increased risk of stroke.\n","independence_note":"Government synthesis report drawing on the same primary literature as the Holt-Lunstad meta-analyses. Not independent data but independent institutional validation of the conclusions.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/35924775/","title":"Effects of Objective and Perceived Social Isolation on Cardiovascular and Brain Health: A Scientific Statement From the American Heart Association","publisher":"Journal of the American Heart Association (Cené, Beckie, Sims, et al.)","source_type":"peer_reviewed","statistic":"Social isolation and loneliness associated with ~30% increased risk of heart attack or stroke, or death from either; 29% increase in heart disease death, 32% increase in stroke death","excerpt":"\"Social isolation and loneliness are common but underrecognized determinants of cardiovascular and brain health. A growing body of evidence demonstrates social isolation and loneliness are associated with increased risk for premature mortality and cardiovascular disease.\"\n","source_date":"2022-08-04","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250929161943/https://pubmed.ncbi.nlm.nih.gov/35924775/","calculation_notes":"The AHA scientific statement provides the cardiovascular-specific mechanism data: ~29% increased risk of coronary heart disease mortality and ~32% increased risk of stroke mortality. These figures help explain how the all-cause mortality signal from loneliness is mediated — cardiovascular disease is the primary pathway. Used to validate the overall mortality figures from Holt-Lunstad and to anchor the cardiovascular multiplier.\n","independence_note":"Independent systematic review by AHA authors, drawing from a partially overlapping but distinct literature base focused on cardiovascular and cerebrovascular outcomes rather than all-cause mortality.\n"},{"url":"https://who.int/news/item/30-06-2025-social-connection-linked-to-improved-heath-and-reduced-risk-of-early-death","title":"Social connection linked to improved health and reduced risk of early death — Report of the WHO Commission on Social Connection","publisher":"World Health Organization (Commission on Social Connection)","source_type":"govt_report","statistic":"1 in 6 people worldwide affected by loneliness; loneliness linked to an estimated 100 deaths every hour — more than 871,000 deaths annually","excerpt":"\"1 in 6 people worldwide is affected by loneliness, with significant impacts on health and well-being. Loneliness is linked to an estimated 100 deaths every hour — more than 871 000 deaths annually.\"\n","source_date":"2025-06-30","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260711234215/https://www.who.int/news/item/30-06-2025-social-connection-linked-to-improved-heath-and-reduced-risk-of-early-death","calculation_notes":"The June 2025 flagship report of the WHO Commission on Social Connection is the first global authoritative quantification of loneliness as a mortality factor. Its headline figures are a global prevalence (~1 in 6 worldwide) and an absolute annual death count (~871,000 deaths/year, ~100/hour), not a relative-risk update to the US-adult subgroup estimate used in this entry. We therefore do NOT use it to revise the normalized 26% excess-mortality headline — it is a distinct global-magnitude figure added as fresh authoritative context. The report restates the same condition-specific associations as the underlying literature (stroke +32%, heart disease +29%, dementia +50%), consistent with the Holt-Lunstad and AHA sources already cited.\n","independence_note":"WHO synthesis report drawing on the same primary literature (including the Holt-Lunstad meta-analyses) for relative-risk magnitudes, but contributing an independent global prevalence and absolute-deaths quantification not present in the US-focused sources. Institutional validation at the global/UN-body level rather than independent primary data.\n"}],"comparison_anchors":[{"label":"Death from regular smoking (lifetime, lifelong smoker)","lifetime_us_adult":0.5},{"label":"Death from chronic sleep deprivation (<6 h/night)","lifetime_us_adult":0.12},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"age 65+","multiplier":1.5,"notes":"Older adults face compounded risk from loneliness due to reduced immune function, higher baseline cardiovascular risk, and greater prevalence of isolation; AARP surveys find ~40% of adults 45+ report loneliness"},{"factor":"lives alone","multiplier":1.2,"notes":"Holt-Lunstad 2015 found living alone carries OR 1.32 vs OR 1.26 for loneliness; living alone is a proxy for but not identical to loneliness"},{"factor":"strong community ties (religious, civic, family)","multiplier":0.3,"notes":"Robust social networks substantially attenuate the mortality risk; Holt-Lunstad 2010 found the protective effect of strong social relationships (OR 1.50 for survival) is one of the largest modifiable mortality factors"},{"factor":"pre-existing cardiovascular disease","multiplier":1.8,"notes":"AHA 2022 statement: loneliness and social isolation associated with ~30% increased risk of heart attack, stroke, or death from either; compounds with existing CVD"}],"short_label":"Loneliness & health","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry measures the excess all-cause mortality attributable to chronic loneliness relative to well-connected adults. It is a subgroup estimate, not a general-population lifetime risk. The 26% figure is a pooled odds ratio from observational studies; causality is not fully established. Lonely individuals also tend to exercise less, eat worse, sleep worse, and adhere less to medical regimens, making it difficult to isolate the independent contribution of loneliness itself versus the health behaviors it co-occurs with. The famous \"smoking 15 cigarettes a day\" comparison refers to the broader social-relationships OR of 1.50 from Holt-Lunstad 2010, not the loneliness-specific OR of 1.26 used here — the comparison is between relative risk magnitudes, not between biological mechanisms. Loneliness is not a toxin in the way nicotine is. Prevalence data (57% of US adults report some loneliness per Cigna 2025) conflate occasional and chronic loneliness; the mortality signal applies to sustained, persistent loneliness, which affects roughly 25-30% of adults. Measurement heterogeneity across studies (different loneliness scales, different definitions of social isolation) contributes to the wide uncertainty band. For global context distinct from this US-adult subgroup estimate, the WHO Commission on Social Connection's June 2025 flagship report found that roughly 1 in 6 people worldwide are affected by loneliness and that loneliness is linked to an estimated 871,000 deaths annually (about 100 every hour) — an absolute global magnitude rather than a relative-risk update to the 26% figure used here.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single empty chair at a small table in a muted, quiet interior space, flat vector illustration."},"canonical_url":"https://likelier.app/loneliness-health-impact","api_url":"https://likelier.app/api/fears/loneliness-health-impact.json"},{"slug":"student-loan-default","question":"What are the odds of defaulting on student loans?","quick_answer":"About 26% of federal student-loan borrowers default over 20 years — far more than the ~10% three-year cohort rate that shapes public perception, because defaults keep accumulating long after repayment starts. Across all US adults the unconditional rate is roughly 4.3%.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Student loan default is widely discussed in personal-finance media and political discourse, but most borrowers treat it as something that happens to other people — dropouts, for-profit college attendees, or the financially irresponsible. The three-year cohort default rate published by the Department of Education (around 10%) reinforces a sense that defaults are a minority outcome. What the short tracking window conceals is that defaults continue accumulating for years after repayment begins, and the population most at risk — borrowers who attended but did not complete a degree — are also the least likely to follow financial-planning media.\n","rough_estimate":"~10% default rate (3-year cohort)","kind":"intuition"},"native":{"display":"26% of 1995-96 borrower cohort defaulted within 20 years (Brookings longitudinal)","numerator":26,"denominator":100,"unit":"share of federal student loan borrowers who default over 20 years","population":"US federal student loan borrowers (1995-96 entry cohort, 20-year follow-up)"},"normalized":{"lifetime_us_adult":0.26,"display":"~26% lifetime default rate among federal borrowers","log_value":-0.59,"assumptions":"The subgroup is federal student loan borrowers — roughly 16.5% of ~260 million US adults. The 26% central estimate is conditional on being a federal borrower, not a rate for all US adults. Applied unconditionally to all adults, the rate would be roughly 4.3% (26% x 16.5%). The Brookings Institution's longitudinal analysis (Scott-Clayton, 2018) tracked the 1995-96 Beginning Postsecondary Students cohort for 20 years and found a cumulative default rate of 26%. Projections for the 2003-04 cohort suggest the 20-year rate may approach 38-40%, reflecting expanded access to federal loans and growth in for-profit enrollment. The 26% figure is used as the central estimate because it is the most recent completed longitudinal observation. A point-in-time FSA portfolio snapshot (October 2025) shows more than 5.5 million borrowers in default with over $140 billion in outstanding loans. This understates lifetime risk because it does not count borrowers who previously defaulted and exited through rehabilitation or consolidation. An additional 2.73 million borrowers are 30-269 days delinquent. The uncertainty range brackets the completed 1995-96 longitudinal observation (low) against the Brookings projection for the 2003-04 cohort (high). Policy changes since these cohorts — income-driven repayment, SAVE plan (in litigation), pandemic forbearance — may alter future cohort outcomes in either direction. The 26% figure should be treated as a historical baseline from the 1995-96 cohort, not a forecast of current-borrower outcomes post-IDR/ SAVE/Fresh Start.\n","uncertainty":{"low":0.18,"high":0.4},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.brookings.edu/articles/the-looming-student-loan-default-crisis-is-worse-than-we-thought/","title":"The Looming Student Loan Default Crisis Is Worse Than We Thought","publisher":"Brookings Institution","source_type":"primary_study","statistic":"26% of 1995-96 entry cohort defaulted within 20 years; projected ~40% for 2003-04 cohort","excerpt":"\"Defaults increase by about 40 percent for the 1995-96 cohort between years 12 and 20 (rising from 18 to 26 percent of all borrowers). Applying these trends to the 2004 entry cohort suggests that nearly 40 percent may default on their student loans.\"\n","source_date":"2018-01-11","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260524130338/https://www.brookings.edu/articles/the-looming-student-loan-default-crisis-is-worse-than-we-thought/","calculation_notes":"The Brookings analysis by Judith Scott-Clayton used longitudinal data from the Beginning Postsecondary Students (BPS) study to track cumulative default rates far beyond the Department of Education's standard 3-year cohort window. The 26% figure for the 1995-96 cohort at 20 years and the ~40% projection for the 2003-04 cohort establish the upper bound of the uncertainty range, showing that short-window default rates dramatically understate lifetime risk.\n","independence_note":"The Brookings analysis uses BPS longitudinal survey data from NCES, which is methodologically independent from the FSA portfolio-level default counts.\n"},{"url":"https://studentaid.gov/data-center/student/default","title":"Federal Student Aid Default Rates","publisher":"U.S. Department of Education, Federal Student Aid","source_type":"govt_report","statistic":"More than 5.5 million borrowers with over $140 billion in federal student loans were in default as of October 2025","excerpt":"\"More than 5.5 million borrowers with over $140 billion in outstanding federal student loans were in default. 1.17 million borrowers were 30-89 days delinquent, 1.56 million were 90-269 days delinquent, and 3.68 million were 270+ days delinquent.\"\n","source_date":"2025-10-31","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260525100521/https://studentaid.gov/data-center/student/default","calculation_notes":"FSA Data Center figures (as reported by TICAS, October 2025): 5.5M borrowers in default with $140B in outstanding loans. An additional 2.73M borrowers are delinquent (1.17M at 30-89 days, 1.56M at 90-269 days). Combined with the 3.68M who are 270+ days delinquent (which overlaps with the default count), the total borrowers in distress exceeds 5.5M. This is a point-in-time snapshot, not a longitudinal measure — some borrowers have exited default through rehabilitation or consolidation, so the cumulative share who have ever defaulted is higher.\n","independence_note":"FSA portfolio data is the official federal administrative record, independent from the Brookings longitudinal cohort analysis which uses NCES survey data.\n"},{"url":"https://ticas.org/affordability-2/2025-student-debt-survey-blog/","title":"On the Edge of a 'Default Cliff': New Survey Shows Student Loan Borrowers Are Struggling to Keep Up","publisher":"The Institute for College Access & Success (TICAS)","source_type":"primary_study","statistic":"20% of surveyed borrowers reported being in delinquency or default in 2024; 42% report tradeoffs between loan payments and basic needs","excerpt":"\"More than four in ten borrowers (42%) report making tradeoffs between loan payments and covering their basic needs. One fifth (20%) of those surveyed said they are currently in either delinquency or default.\"\n","source_date":"2025-09-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260319231943/https://ticas.org/affordability-2/2025-student-debt-survey-blog/","calculation_notes":"The TICAS survey captures self-reported delinquency and default, providing a cross-check on the FSA administrative data. The 20% self-reported delinquency-or-default rate is broadly consistent with FSA figures showing 5.5M in default plus 2.73M delinquent out of the total borrower population.\n"},{"url":"https://libertystreeteconomics.newyorkfed.org/2026/05/federal-student-loan-defaults-return-after-pandemic-pause/","title":"Federal Student Loan Defaults Return After Pandemic Pause","publisher":"Federal Reserve Bank of New York, Liberty Street Economics","source_type":"reputable_reference","statistic":"Credit scores of defaulted borrowers fell 91 points (567 to 476) between 2024:Q3 and 2025:Q4; ~1M defaulted in 2025:Q4 and ~2.6M in 2026:Q1","excerpt":"\"On average, credit scores for defaulted borrowers dropped 91 points between 2024:Q3 and 2025:Q4 (from 567 to 476). We estimate that roughly 1 million federal student loan borrowers defaulted during 2025:Q4, with an additional 2.6 million borrowers defaulting during 2026:Q1.\"\n","source_date":"2026-05-12","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260630214746/https://libertystreeteconomics.newyorkfed.org/2026/05/federal-student-loan-defaults-return-after-pandemic-pause/","calculation_notes":"Quantifies the downstream credit-score consequence of default rather than the probability of default itself. The 91-point average drop (to a subprime 476) over a single year, and the ~3.6M borrowers entering default across 2025:Q4-2026:Q1, document the post-on-ramp collections restart. Used to support the credit-score caveat; it does not alter the 26% 20-year cumulative-default headline.\n","independence_note":"NY Fed Consumer Credit Panel (Equifax) data, methodologically independent from both the Brookings NCES survey analysis and the FSA administrative portfolio counts.\n"},{"url":"https://newsroom.transunion.com/may-2025-student-loan-update/","title":"As Federal Collections Activity Resumes, More Than One in Five Federal Student Loan Borrowers With a Payment Due are Seriously Delinquent","publisher":"TransUnion","source_type":"reputable_reference","statistic":"20.5% of federal borrowers with a payment due were 90+ days past due as of Feb 2025 (highest ever recorded), up from 11.5% in Feb 2020","excerpt":"\"The analysis found that 20.5% of federal student loan borrowers with a payment due are 90 days or more past due (90+ DPD). The current rate of delinquency represents the highest figure ever recorded. This compares to 11.5% in February 2020, near the beginning of the pandemic and the subsequent student loan pause.\"\n","source_date":"2025-05-05","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260401000721/https://newsroom.transunion.com/may-2025-student-loan-update/","calculation_notes":"Provides a like-with-like era comparison of the same metric (90+ DPD among borrowers with a payment due): 20.5% (Feb 2025) / 11.5% (Feb 2020) = 1.78x. This supports the \"payment due in 2025+ collections regime\" multiplier (~1.8x) as a serious-delinquency proxy. It is a point-in-time delinquency rate, not a 20-year cumulative-default rate, so it does not change the 26% headline.\n","independence_note":"TransUnion credit-bureau panel data, independent from the NY Fed (Equifax) panel, the Brookings NCES survey, and the FSA portfolio counts.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Retirement savings shortfall (US)","lifetime_us_adult":0.39},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"For-profit college attendees","probability":0.52,"notes":"Brookings data shows for-profit borrowers default at roughly twice the rate of public two-year borrowers (52 vs 26 percent after 12 years)"},{"region":"Community college (no degree)","probability":0.35,"notes":"Borrowers who attended but did not complete have dramatically higher default rates"},{"region":"Four-year public university graduates","probability":0.08,"notes":"Degree completers at public institutions have the lowest default rates"}],"personal_factor_multipliers":[{"factor":"did not complete degree","multiplier":3,"notes":"Non-completers carry debt without the earnings premium; Brookings shows they drive the majority of defaults"},{"factor":"for-profit institution","multiplier":2.5,"notes":"For-profit attendees have substantially higher default rates even controlling for completion"},{"factor":"graduate degree holder","multiplier":0.2,"notes":"Graduate borrowers carry more debt but default far less frequently due to higher earnings"},{"factor":"parent PLUS borrower","multiplier":0.5,"notes":"Parent PLUS borrowers have lower default rates, though their balances are often large"},{"factor":"payment due in 2025+ collections regime","multiplier":1.8,"notes":"Serious-delinquency proxy, not a re-statement of the 20-year cumulative-default baseline. TransUnion (May 2025) reports 20.5% of federal borrowers with a payment due were 90+ days past due as of Feb 2025 — the highest ever recorded — versus 11.5% in Feb 2020 before the payment pause, a 1.78x rise. Reflects the post-on-ramp restart of federal collections and credit reporting (TransUnion, May 2025).\n"}],"short_label":"Student loan default","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 26% figure is conditional on being a federal student loan borrower; applied to all US adults, the unconditional rate is roughly 4.3%. The central estimate comes from Brookings' 20-year longitudinal follow-up of the 1995-96 cohort; the 2003-04 cohort projection (~38-40%) suggests default rates may be rising. The FSA portfolio snapshot (5.5M+ in default as of October 2025) understates lifetime risk because it does not count borrowers who previously defaulted and exited through rehabilitation or consolidation. The policy landscape has shifted substantially since the cohorts Brookings tracked: income-driven repayment (IDR) plans now cover roughly half of borrowers in repayment, and the SAVE plan (now in litigation) was designed to prevent default for low-income borrowers entirely. The pandemic-era payment pause (March 2020 to September 2023) and Fresh Start program further complicate comparisons across cohorts. Default is also not a permanent state — borrowers can rehabilitate out of default — so the stock of defaulted borrowers at any point is lower than the cumulative flow. Default is not only a probability but a consequence with a measurable magnitude: the NY Fed reports that the average credit score of defaulted borrowers fell 91 points (from 567 to 476, into subprime territory) between 2024:Q3 and 2025:Q4 as federal collections resumed, with roughly 3.6 million borrowers entering default across 2025:Q4-2026:Q1. The 2024-25 expiry of the on-ramp and the restart of credit reporting, garnishment, and Treasury offset mean the \"borrower with a payment due\" subgroup in the current regime runs well above the legacy baseline on serious delinquency — 90+-day past-due rates among payment-due federal borrowers hit 20.5% in early 2025 (the highest on record, up from 11.5% pre-pause), per TransUnion — though this is a point-in-time delinquency measure, not the 20-year cumulative-default rate the headline reports.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A torn diploma beside a stack of overdue bills, muted grey and rust tones, flat vector illustration."},"canonical_url":"https://likelier.app/student-loan-default","api_url":"https://likelier.app/api/fears/student-loan-default.json"},{"slug":"airline-mishandled-luggage","question":"What are the odds of an airline losing or mishandling checked luggage?","quick_answer":"About 0.55% per checked bag, which adds up to roughly a 27% chance of at least one mishandled bag over a lifetime of flying. Most travelers guess 1 to 5% per bag, an overestimate — and about three-quarters of \"mishandled\" bags are simply delayed and returned, not lost for good.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Travelers who check bags routinely carry a background anxiety about luggage arriving damaged, delayed, or not at all. The mental model is that mishandling is relatively rare but deeply frustrating when it happens — something experienced by \"someone you know\" rather than a routine statistical probability. Most passengers, if pressed, might estimate a per-trip risk around 1-5%, with the post-2022 disruption period likely inflating intuitive estimates above the current reality.\n","rough_estimate":"Most travelers guess 1-5% per checked bag","kind":"intuition"},"native":{"display":"~5.5 mishandled bags per 1,000 checked bags (0.55%) in 2024 — US carriers","numerator":55,"denominator":10000,"unit":"per checked bag","population":"Checked bags reported by US domestic reporting marketing carriers to BTS DOT (2024 full year)","exposures_per_year":4},"normalized":{"lifetime_us_adult":0.27,"display":"~27% chance of at least one mishandled bag over a lifetime of flying with checked luggage","log_value":-0.57,"assumptions":"BTS Air Travel Consumer Report (full year 2024): US reporting marketing carriers posted a mishandled baggage rate of 0.55% (5.5 per 1,000 checked bags), down from 0.58% in 2023. For a traveler who checks one bag on 4 flight segments per year (4 checked-bag transactions per year), the annual probability of at least one mishandled bag = 1 − (1 − 0.0055)^4 ≈ 2.2%. Over 30 active flying years with checked luggage: 1 − (1 − 0.022)^30 ≈ 0.49. The central estimate of 0.27 reflects more conservative behavior — checking a bag on only half the 4 annual segments, or flying for 20 years: 1 − (1 − 0.0055)^(2×30) ≈ 0.27. Given that a significant fraction of US travelers often fly carry-on only, and BTS \"mishandled\" predominantly means \"temporarily delayed\" (74% of mishandling is delayed recovery), a lifetime probability of 0.27 is a reasonable central estimate for a typical checking traveler. Scope is activity_specific_lifetime.\n","uncertainty":{"low":0.1,"high":0.55},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.bts.gov/newsroom/air-travel-consumer-report-december-2024-full-year-2024-numbers","title":"Air Travel Consumer Report: December 2024, Full Year 2024 Numbers","publisher":"Bureau of Transportation Statistics (DOT)","source_type":"govt_report","statistic":"Mishandled baggage rate 0.55% in 2024 (5.5 per 1,000 checked bags), down from 0.58% in 2023; monthly range: 0.38% (October) to 0.75% (January, July)","excerpt":"\"In 2024, the reporting marketing carriers reported a mishandled baggage rate of 0.55%, down from 0.58% in 2023. Monthly rates throughout 2024 ranged from a low of 0.38% in October to a high of 0.75% in January and July.\"\n","source_date":"2025-02-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508123200/https://www.bts.gov/newsroom/air-travel-consumer-report-december-2024-full-year-2024-numbers","calculation_notes":"BTS switched from per-1,000-bags to percentage reporting in January 2022. 0.55% = 5.5 per 1,000 checked bags. For 2 checked bags/year over 30 years: 1 − (1 − 0.0055)^60 ≈ 0.28. Central estimate 0.27 used. Of the mishandled bags, BTS reports approximately 74% are delayed (temporarily) and recovered, 8% are lost or stolen, and 18% are damaged or pilfered.\n","independence_note":"BTS mishandled-baggage data is compiled from mandatory DOT reports submitted by reporting marketing carriers under 14 CFR Part 234. It is an administrative regulatory dataset, separate from airline customer complaint surveys and independent from the SITA Baggage IT Insights reports (which use a different definition and a global scope).\n"},{"url":"https://www.sita.aero/resources/surveys-reports/sita-baggage-it-insights-2024/","title":"SITA Baggage IT Insights 2024","publisher":"SITA (air transport IT provider)","source_type":"reputable_reference","statistic":"Global mishandling rate 6.3 per 1,000 passengers in 2024, down from 6.9 in 2023; North/South America: 5.5 per 1,000; 74% of mishandling is delayed bags; 8% permanently lost; transfer mishandling causes 41% of incidents","excerpt":"\"The rate of mishandled baggage fell to 6.3 per 1,000 passengers in 2024, an 8.7% improvement from 2023's 6.9 rate. North and South America posted 5.5 per 1,000 passengers. Delayed bags accounted for 74% of mishandled bags, down from 80% the prior year. Lost or stolen bags made up 8%. Transfer mishandling caused more than four in ten (41%) of mishandling incidents.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260424160528/https://www.sita.aero/resources/surveys-reports/sita-baggage-it-insights-2024/","calculation_notes":"SITA measures mishandling per 1,000 passengers (not per 1,000 checked bags), and not all passengers check a bag. The BTS per-checked-bag figure (0.55%) is the primary rate for this entry because it is US-specific and normalized to the correct denominator (checked bags, not all passengers). The SITA figure provides a corroborating global benchmark and the breakdown by mishandling type (delayed vs lost vs damaged) which BTS does not publish at this level of detail.\n","independence_note":"SITA collects data directly from airline and airport IT systems via its WorldTracer global baggage tracking platform and voluntary airline participation. It is methodologically independent from BTS DOT mandatory reporting; the two systems use different denominators (passengers vs checked bags) and different geographic scopes (global vs US reporting carriers).\n"},{"url":"https://www.bts.gov/newsroom/air-travel-consumer-report-march-2024-numbers","title":"Air Travel Consumer Report: March 2024 Numbers","publisher":"Bureau of Transportation Statistics (DOT)","source_type":"govt_report","statistic":"March 2024 mishandled baggage rate: 0.52%; February 2024: 0.48%; March 2023: 0.58%; confirms consistent year-over-year improvement trend","excerpt":"\"In March 2024, reporting marketing carriers handled 43.1 million bags and posted a mishandled baggage rate of 0.52%, higher than the rate of 0.48% in February 2024, but lower than the rate of 0.58% in March 2023.\"\n","source_date":"2024-05-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508155035/https://www.bts.gov/newsroom/air-travel-consumer-report-march-2024-numbers","calculation_notes":"Monthly BTS data confirms the 2024 improvement trend relative to 2023 is consistent across individual months, not driven by a single anomalous month. Used as triangulating confirmation that the full-year 0.55% figure is a robust annual average rather than a data artifact.\n","independence_note":"Same BTS DOT mandatory-reporting pipeline as the annual report; monthly release confirms the full-year average figure.\n"}],"comparison_anchors":[{"label":"Flight cancellation per segment (~1.4%)","lifetime_us_adult":0.56},{"label":"Blood clot on long-haul flight (per flight)","lifetime_us_adult":0.000215}],"personal_factor_multipliers":[{"factor":"Always checks a bag (never flies carry-on only)","multiplier":2,"notes":"Travelers who check a bag on every segment have roughly double the annual checked-bag transactions compared to travelers who check half the time, proportionally doubling exposure."},{"factor":"Frequently takes connecting flights with tight layovers (<60 minutes)","multiplier":3,"notes":"SITA data shows transfer mishandling accounts for 41% of all incidents. Tight connections give ground handlers minimal margin for bag transfers; mishandling probability per transferred bag is substantially higher than for non-stop bags."},{"factor":"Flies exclusively nonstop","multiplier":0.5,"notes":"Nonstop travel eliminates transfer mishandling (41% of all events) and reduces total handling touchpoints. Per-bag mishandling rate on nonstop domestic flights is roughly half the overall average."},{"factor":"Tags bag with Apple AirTag or GPS tracker","multiplier":0.5,"notes":"Not a prevention measure, but trackers significantly reduce the probability that a delayed bag becomes permanently lost; they also accelerate recovery time for delayed bags. Effective multiplier on 'permanently lost' outcome is substantially lower."}],"short_label":"Mishandled luggage","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"BTS defines \"mishandled\" broadly: it includes delayed bags that are recovered and returned (74% of cases), damaged bags (18%), and permanently lost or stolen bags (8%). The probability of permanent bag loss is much lower than the headline mishandling rate — roughly 0.044% per segment (8% of 0.55%), or about 1 in 2,300 bags. The entry uses the full mishandling rate (which includes temporarily delayed bags) as the primary figure because any mishandling event is a real consequence for the traveler, even if most are eventually resolved. Travelers who exclusively fly carry-on have zero exposure to checked-bag mishandling and should disregard this entry for their personal risk assessment. The BTS rate covers reporting marketing carriers and may not fully capture mishandling at regional code-share partners, which sometimes have higher rates.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A single suitcase on an empty baggage carousel with a question mark, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/airline-mishandled-luggage","api_url":"https://likelier.app/api/fears/airline-mishandled-luggage.json"},{"slug":"intimate-partner-violence-lifetime","question":"What are the odds of experiencing domestic violence?","quick_answer":"About 27% of ever-partnered women worldwide — roughly 1 in 4 — experience physical or sexual violence from an intimate partner in their lifetime. Most people file domestic violence as something that happens to other households, but even in high-income countries roughly 1 in 5 ever-partnered women report it.\n","category":"crime","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"Intimate partner violence is widely acknowledged as a problem in the abstract but chronically underestimated as a personal risk. Most adults do not consider themselves likely victims — domestic violence is mentally filed as something that happens in other households, other socioeconomic brackets, other cultures. Media framing reinforces this by treating individual cases as shocking aberrations rather than manifestations of a one-in-three baseline. In many countries, cultural norms further suppress risk perception by framing partner violence as a private family matter rather than a crime. No rigorous multi-country survey of perceived personal risk exists.\n","kind":"intuition"},"native":{"display":"~27% of ever-partnered women aged 15-49 have experienced physical/sexual IPV in their lifetime (WHO 2023)","numerator":27,"denominator":100,"unit":"lifetime prevalence among ever-partnered women","population":"Global, ever-partnered women aged 15-49, WHO multi-country data 2000-2023"},"normalized":{"lifetime_us_adult":0.27,"display":"~1 in 4 ever-partnered women globally experience physical/sexual IPV in a lifetime","log_value":-0.57,"assumptions":"WHO's Violence Against Women Prevalence Estimates 2023 report, analysing data from 168 countries between 2000 and 2023, finds that approximately 27% of ever-partnered women aged 15-49 have experienced physical and/or sexual intimate partner violence in their lifetime. The November 2025 update puts the absolute number at 840 million women. This is a directly measured lifetime prevalence from population-based surveys, not an extrapolation from annual rates. The 27% figure is for physical and/or sexual violence only; including psychological/emotional abuse raises prevalence substantially. The lifetime_us_adult field carries the global figure for schema compatibility under the global_adult_lifetime scope. Regional variation is substantial: from ~20% in the Western Pacific to ~33% in Africa and South-East Asia. The figure covers women only; male IPV victimization (estimated at ~1 in 9 to 1 in 7 in US data) is not captured in the WHO global estimates. Uncertainty band: low end reflects high-income-country estimates (~20%), high end reflects highest-prevalence regions (~35%) plus underreporting adjustment. Uncertainty band widened from the raw regional range (20–33%) to 15–45%: the low end accounts for the possibility that high-income- country rates are themselves overstated by retrospective survey design; the high end reflects the inclusion of psychological/emotional abuse, which is documented to push overall VAW prevalence to 45–50% in some national surveys.\n","uncertainty":{"low":0.15,"high":0.45},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/publications/i/item/9789240116962","title":"Violence Against Women Prevalence Estimates, 2023","publisher":"World Health Organization (WHO)","source_type":"govt_report","statistic":"27% of ever-partnered women aged 15-49 have experienced physical/sexual IPV globally","excerpt":"\"Worldwide, almost one third (27%) of women aged 15-49 years who have been in a relationship report that they have been subjected to some form of physical and/or sexual violence by their intimate partner.\"\n","source_date":"2025-11-19","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260318232738/https://www.who.int/publications/i/item/9789240116962","calculation_notes":"Primary global lifetime prevalence from WHO's most comprehensive report, synthesizing data from 168 countries (2000-2023). 27% of ever-partnered women = 0.27 lifetime prevalence. Used directly as lifetime_us_adult under global_adult_lifetime scope. Note: the lifetime_us_adult field carries a global ever-partnered-women figure, not a US-specific or all-adults figure, per the global_adult_lifetime scope.\n","independence_note":"All three sources (report, press release, fact sheet) are WHO publications released on the same date drawing on the same underlying dataset. They provide different facets (full report, headline figures, regional breakdown) but are not independent data sources.\n"},{"url":"https://www.who.int/news/item/19-11-2025-lifetime-toll--840-million-women-faced-partner-or-sexual-violence","title":"Lifetime toll: 840 million women faced partner or sexual violence","publisher":"World Health Organization (WHO)","source_type":"govt_report","statistic":"840 million women globally have experienced partner or sexual violence; 316 million in the past 12 months","excerpt":"\"Nearly 1 in 3 women – estimated 840 million globally – have experienced partner or sexual violence during their lifetime, a figure that has barely changed since 2000. In the last 12 months alone, 316 million women – 11% of those aged 15 or older – were subjected to physical or sexual violence by an intimate partner.\"\n","source_date":"2025-11-19","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260503091618/https://www.who.int/news/item/19-11-2025-lifetime-toll--840-million-women-faced-partner-or-sexual-violence","calculation_notes":"Press release accompanying the 2023 prevalence report. The \"nearly 1 in 3\" headline includes both IPV and non-partner sexual violence combined. The IPV-only figure is 27%; adding non-partner sexual violence brings the combined total to ~31%. The 12-month prevalence of 11% confirms ongoing incidence, not just historical accumulation.\n","independence_note":"WHO press release accompanying source 1; same underlying dataset, not an independent data source.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/violence-against-women","title":"Violence against women — Key facts","publisher":"World Health Organization (WHO)","source_type":"govt_report","statistic":"Prevalence ranges from 20% in Western Pacific to 33% in African and South-East Asian regions","excerpt":"\"The prevalence estimates of lifetime intimate partner violence range from 20% in the Western Pacific, 22% in high-income countries and Europe and 25% in the WHO Regions of the Americas to 33% in the WHO African region, 31% in the WHO Eastern Mediterranean Region, and 33% in the WHO South-East Asia region.\"\n","source_date":"2025-11-19","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260525095323/https://www.who.int/news-room/fact-sheets/detail/violence-against-women","calculation_notes":"Regional breakdown supporting the uncertainty band. Low end (0.20) anchored to Western Pacific estimate; high end (0.35) anchored slightly above the African/ South-East Asian peak of 33% to account for underreporting in the highest- prevalence regions. The 22% for high-income countries contextualizes the US- specific figure relative to the global average.\n","independence_note":"WHO fact sheet drawing on the same prevalence report as sources 1 and 2; not an independent data source.\n"}],"comparison_anchors":[{"label":"Sexual assault (lifetime, US adult, contact)","lifetime_us_adult":0.34},{"label":"Stalking (lifetime, US women)","lifetime_us_adult":0.162},{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348}],"short_label":"Domestic violence","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The WHO 27% figure covers physical and/or sexual violence by an intimate partner among ever-partnered women aged 15-49. It does not include psychological or emotional abuse, which would raise prevalence substantially. It also does not include male victims of IPV, who represent a significant minority of cases — CDC NISVS data estimate roughly 1 in 9 US men experience contact sexual violence, physical violence, or stalking by an intimate partner. The \"ever-partnered\" denominator excludes women who have never been in a relationship, which slightly inflates the rate compared to an all-women denominator. Regional variation is extreme: a woman in South-East Asia faces roughly 65% higher lifetime risk than one in the Western Pacific. Underreporting remains the dominant measurement problem — in many settings, partner violence is normalized and not recognized as reportable violence by either the victim or the community. Beyond baseline prevalence, emerging US work documents acute, situational spikes: Matsuzawa and Arnesen (2024, SSRN 4938642) find that legalized sports betting amplifies the short-term rise in reported IPV that follows an NFL home-team upset loss, with the effect concentrated in states that permit mobile betting, near paydays, and after losing wagers. This is an acute per-event effect on US incident rates, not a lifetime-risk multiplier, so it is not encoded as a personal factor against the global lifetime prevalence above.\n","quality_score":{"d1":4,"d2":3,"d3":5,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A closed door with a thin line of light underneath, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/intimate-partner-violence-lifetime","api_url":"https://likelier.app/api/fears/intimate-partner-violence-lifetime.json"},{"slug":"walking-alone-night-assault","question":"What are the odds of being assaulted while walking alone at night?","quick_answer":"Stranger-committed violent victimization in public spaces runs about 5.5 per 1,000 US adults a year, compounding to roughly 1 in 3.6 over a lifetime — most of it simple assault rather than serious injury. Gallup finds 40% of Americans afraid to walk alone at night, a measure of fear rather than the actual rate.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Fear of walking alone at night is one of the most studied items in criminology, tracked by Gallup since 1965. In 2023, 40% of Americans reported being afraid to walk alone at night near their home — the highest since 1993 — with a sharp gender split: 53% of women and 26% of men. The paradox is well-established in the victimization literature: women fear stranger violence far more than men do, but men are substantially more likely to be victimized by strangers in public spaces. Gallup's question specifically asks about walking \"within a mile of where you live,\" making it a proxy for generalized fear of crime rather than a calibrated probability estimate. Fear is highest among women, low-income adults, and urban residents.\n","rough_estimate":"40% of Americans feel unsafe walking alone at night; 53% of women versus 26% of men","kind":"survey","survey_source":{"title":"Personal Safety and Crime","publisher":"Gallup","url":"https://news.gallup.com/poll/1603/crime.aspx","year":2023}},"native":{"display":"~5.5 stranger-committed violent victimizations per 1,000 US adults per year","numerator":55,"denominator":10000,"unit":"per year","population":"US residents age 12+, stranger-committed violent victimization (NCVS)"},"normalized":{"lifetime_us_adult":0.276,"display":"~1 in 3.6 lifetime (US adult, any stranger-committed violent victimization)","log_value":-0.559,"assumptions":"The NCVS reports approximately 23.3 violent victimizations per 1,000 persons age 12+ in 2024. Historical NCVS data consistently shows that roughly 40-50% of violent victimizations are committed by strangers (the remainder by intimate partners, other relatives, or acquaintances). Using 45% as a stable midpoint, the stranger-committed violent victimization rate is approximately 10.5 per 1,000 per year. However, this includes all contexts — home, workplace, bars, etc. — not just \"walking alone at night.\" The subset of stranger violence occurring on streets and in public spaces at night is roughly half of all stranger violence, yielding an effective \"street assault by stranger\" rate of approximately 5.5 per 1,000 per year.\nCompounded over 59 years of remaining adult life at constant hazard: 1 − (1 − 0.0055)^59 ≈ 0.276 ≈ 1 in 3.6.\nThis is a broad measure covering all nonfatal violent victimizations by strangers in public spaces — including simple assault (the majority), aggravated assault, robbery, and sexual assault. Serious injury (requiring medical treatment) occurs in roughly 25-30% of violent victimizations. The lifetime probability of a stranger assault resulting in injury while in a public space is therefore roughly 0.07-0.08, or about 1 in 13.\nThe BJS Koppel (1987) lifetime-likelihood-of-victimization study estimated that 83% of Americans would experience some form of violent victimization in their lifetime (all contexts, all offender relationships). The street-stranger subset on this page is a fraction of that total.\n","uncertainty":{"low":0.15,"high":0.4},"scope":"us_adult_lifetime"},"sources":[{"url":"https://news.gallup.com/poll/544415/personal-safety-fears-three-decade-high.aspx","title":"Personal Safety Fears at Three-Decade High in U.S.","publisher":"Gallup","source_type":"reputable_reference","statistic":"40% of Americans afraid to walk alone at night near home (2023), highest since 1993; 53% of women, 26% of men","excerpt":"\"Forty percent of U.S. adults say they are afraid to walk alone at night within a mile of their home, the most since 1993.\"\n","source_date":"2023-11-16","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260522020938/https://news.gallup.com/poll/544415/personal-safety-fears-three-decade-high.aspx","calculation_notes":"Gallup's annual crime survey has asked the \"afraid to walk alone at night\" question since 1965. The 2023 reading of 40% is the perception anchor for this entry. The gender gap (53% women vs 26% men) is the most important demographic split and directly motivates the \"fear-victimization paradox\" framing: women fear it roughly twice as much as men, but men experience stranger violence at higher rates. The 2024 reading dropped to 35%, suggesting some regression toward the trend mean.\n"},{"url":"https://bjs.ojp.gov/library/publications/criminal-victimization-2024","title":"Criminal Victimization, 2024","publisher":"Bureau of Justice Statistics, U.S. Department of Justice","source_type":"govt_report","statistic":"23.3 violent victimizations per 1,000 persons age 12+ in 2024; 1.45% prevalence rate","excerpt":"\"In 2024, 1.45% of persons age 12 or older experienced at least one violent victimization, similar to 2023.\"\n","source_date":"2025-10-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260504061621/https://bjs.ojp.gov/library/publications/criminal-victimization-2024","calculation_notes":"The NCVS 2024 overall violent victimization rate of 23.3 per 1,000 is the denominator anchor. Historical NCVS tabulations consistently show that strangers account for roughly 40-50% of violent victimizations (the BJS Criminal Victimization 2023 report shows the stranger share at approximately 44%). Applying 45% to the 23.3 rate yields ~10.5 stranger-committed violent victimizations per 1,000 per year. The further restriction to public-space / street settings at night is estimated at roughly half of stranger violence, yielding the ~5.5 per 1,000 native rate. The 1.45% annual prevalence figure (percent of persons with at least one victimization) is lower than the rate (which counts multiple victimizations of the same person) and provides a cross-check: roughly 1 in 69 adults experience at least one violent victimization per year, and the stranger-in-public-space subset is roughly 1 in 180.\n"},{"url":"https://bjs.ojp.gov/library/publications/lifetime-likelihood-victimization-0","title":"Lifetime Likelihood of Victimization","publisher":"Bureau of Justice Statistics (Koppel 1987)","source_type":"govt_report","statistic":"83% lifetime likelihood of violent crime victimization for US residents; 92% for Black men","excerpt":"\"Five of six persons will be the victim of a completed or attempted violent crime (rape, robbery, or assault) at least once during their lifetimes.\"\n","source_date":"1987-03-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251109125308/https://bjs.ojp.gov/library/publications/lifetime-likelihood-victimization-0","calculation_notes":"Koppel 1987 is the only BJS publication that directly estimates lifetime victimization probability. The 83% figure covers all violent crime (rape, robbery, assault) across all contexts and all offender relationships over a full lifetime. The street-stranger subset on this page is a fraction of the Koppel total. The Koppel estimates were derived from 1975-1984 National Crime Survey rates, which were higher than current NCVS rates, so the lifetime probability under current rates would be lower — hence this page's central estimate of 0.276 for the stranger-in-public-space subset rather than the Koppel-era 0.83 for all violent crime.\n"},{"url":"https://news.gallup.com/poll/186563/women-poor-urbanites-fearful-walking-alone.aspx","title":"In U.S., Women, Poor, Urbanites Most Fearful of Walking Alone","publisher":"Gallup","source_type":"reputable_reference","statistic":"In 1982, 64% of women and 31% of men feared walking alone at night; fear is highest among low-income adults and urban residents","excerpt":"\"In 1982, more than six in 10 women (64%) said they did not feel safe walking alone at night, compared with 31% of men — a 33-point gap.\"\n","source_date":"2015-11-24","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421201717/https://news.gallup.com/poll/186563/women-poor-urbanites-fearful-walking-alone.aspx","calculation_notes":"This Gallup analysis of the historical trend in walking-alone fear provides the longitudinal context. The gender gap has narrowed from 33 points (1982) to 27 points (2023) but remains the single largest demographic predictor of fear-of-crime. The article also notes that low-income adults and urban residents report higher fear, both of which track with higher actual victimization rates — unlike the gender gap, where the fear-victimization relationship is inverted.\n"}],"comparison_anchors":[{"label":"Any violent victimization (lifetime, US adult, all contexts)","lifetime_us_adult":0.55},{"label":"Robbery victimization (lifetime, US adult)","lifetime_us_adult":0.12},{"label":"Motor vehicle theft (lifetime, US adult)","lifetime_us_adult":0.2},{"label":"Homicide (lifetime, US adult)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"male","multiplier":1.5,"notes":"Men are approximately 50% more likely than women to be victims of stranger-committed violent crime, despite women reporting roughly twice the fear level. The NCVS consistently shows higher male victimization rates for robbery and aggravated assault by strangers."},{"factor":"female","multiplier":0.7,"notes":"Women experience lower rates of stranger violence in public spaces but face disproportionate risk of sexual assault, which is underreported in NCVS data. The overall stranger-violence multiplier is below 1, but the composition of risk differs."},{"factor":"ages 18-24","multiplier":2.5,"notes":"Young adults have the highest violent victimization rates across all NCVS years. The 18-24 age group typically runs 2-3x the rate of adults 35+."},{"factor":"urban residence","multiplier":1.8,"notes":"Urban violent victimization rates are roughly 80% higher than rural rates in NCVS data, and the stranger-violence share is higher in urban settings."},{"factor":"suburban or rural residence","multiplier":0.6,"notes":"Lower population density correlates with lower stranger-violence rates. Rural violence is disproportionately intimate-partner rather than stranger."},{"factor":"low income (household income < $25,000)","multiplier":2,"notes":"NCVS consistently shows violent victimization rates roughly double for low-income adults versus middle-income adults."}],"short_label":"Night walk assault","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry covers nonfatal violent victimization by strangers in public spaces, which is the closest NCVS-derivable proxy for \"assaulted while walking alone at night.\" It is not a perfect match: the NCVS does not isolate \"walking alone\" as a distinct activity, and the nighttime restriction is estimated from the time-of-occurrence tables rather than directly measured for the stranger-public- space subset. Simple assault (no weapon, no serious injury) constitutes the majority of the numerator; aggravated assault, robbery, and sexual assault are each a smaller share. The lifetime figure of ~1 in 3.6 is for any such event over 59 adult years at current rates; the lifetime probability of a stranger assault resulting in injury requiring medical treatment is roughly 1 in 13. The Koppel (1987) lifetime estimate of 83% for all violent crime is higher because it includes intimate-partner and acquaintance violence, workplace incidents, and bar fights — contexts that are not what people picture when they think about \"walking alone at night.\" Sexual assault is underreported in NCVS data, so the female victimization rate is likely understated for that specific crime type.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single streetlight casting a pale cone of light on an empty sidewalk, flat vector illustration in muted dark blues and greys."},"canonical_url":"https://likelier.app/walking-alone-night-assault","api_url":"https://likelier.app/api/fears/walking-alone-night-assault.json"},{"slug":"child-of-alcoholic-aud-lifetime","question":"If a parent had alcohol use disorder, what are the odds you'll develop alcohol use disorder yourself?","quick_answer":"About 29% of first-degree relatives of someone with alcohol use disorder develop alcohol dependence themselves — roughly double the ~14% rate in matched controls, but well below the coin-flip odds many adult children of alcoholics assume. Inheritance raises the risk; it does not make it destiny.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"There is no rigorous public-perception survey of how adult children of alcoholics estimate their own AUD risk, but the cultural script around inherited alcoholism is strong: family memoirs, recovery literature, and pop-genetics commentary often imply something close to a coin flip. The intuitive frame compresses two distinct claims — that AUD is heritable (true: roughly half the variance is genetic) and that an individual child of an AUD parent will likely develop AUD themselves (substantially overstated). The best direct measurement (COGA) puts the elevation at roughly 2x baseline, not 5x or 10x, and the majority of adult children of alcoholics never meet criteria for the disorder. The gap between inheritance-as-mechanism and inheritance-as-destiny is where the perception calibration sits.\n","rough_estimate":"many adult children of alcoholics believe their lifetime risk is 50% or higher","kind":"intuition"},"native":{"display":"28.8% of first-degree relatives of an alcohol-dependent proband meet lifetime DSM-IV criteria for alcohol dependence, vs 14.4% of controls (COGA, Nurnberger et al. 2004)","numerator":28.8,"denominator":100,"unit":"share of first-degree relatives of an alcohol-dependent proband with lifetime DSM-IV alcohol dependence","population":"First-degree relatives (parents, siblings, offspring) of alcohol-dependent probands ascertained through the Collaborative Studies on the Genetics of Alcoholism (COGA); N=8,296 relatives vs N=1,654 community controls"},"normalized":{"lifetime_us_adult":0.288,"display":"~29% of first-degree relatives of a person with AUD develop alcohol dependence themselves, roughly double the ~14% baseline rate in matched community controls","log_value":-0.541,"assumptions":"The headline figure is the directly measured lifetime DSM-IV alcohol-dependence rate in COGA's first-degree-relative pool (Nurnberger et al. 2004): 28.8% of relatives vs 14.4% of community controls assessed with the same Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA) instrument — a ~2.0x relative risk. This is preferred over inferring a number by multiplying NESARC-III's 29.1% DSM-5 lifetime AUD baseline by the 2x family-history effect for three reasons. (1) Apples-to-apples instrumentation: relatives and controls in COGA were diagnosed identically. (2) DSM-IV alcohol dependence is more restrictive than DSM-5 AUD (which collapsed abuse + dependence into a single disorder with a 2-criterion floor), so the COGA figure is conservative relative to a hypothetical DSM-5 re-analysis. (3) No nationally representative study has measured DSM-5 AUD prevalence specifically in adult children of AUD parents; the COGA first-degree-relative pool is the largest direct measurement available. Important framing limitation: COGA's \"first-degree relatives\" pools parents, siblings, and offspring. The paper does not break results down by relationship type, so the 28.8% figure is the best available proxy for offspring-specific risk, not a direct measurement of it. Adult children of AUD parents are likely near this figure given (a) shared genetic loading equivalent to siblings and (b) shared early-life environmental exposure; the COGA estimate is anchored to a population that includes them but is not exclusively them. Heritability estimates from the Verhulst, Neale & Kendler (2015) meta-analysis of 12 twin and 5 adoption studies converge on h² = 0.49 (95% CI 0.43-0.53), supporting the magnitude of the elevation as part-genetic. The uncertainty band reflects cohort-definition heterogeneity (one parent vs both, biological vs adopted-out studies), DSM edition (DSM-IV ~28% in COGA vs likely higher under DSM-5 criteria), and the offspring-specific vs first-degree-relative-pooled distinction.\n","uncertainty":{"low":0.22,"high":0.5},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/15583116/","title":"A family study of alcohol dependence: coaggregation of multiple disorders in relatives of alcohol-dependent probands","publisher":"Nurnberger JI Jr, Wiegand R, Bucholz K, et al. — Archives of General Psychiatry, 2004","source_type":"peer_reviewed","statistic":"Lifetime risk rates of 28.8% and 14.4% for DSM-IV alcohol dependence in first-degree relatives of alcohol-dependent probands and controls, respectively; relative risk approximately 2-fold","excerpt":"\"Lifetime risk rates of 28.8% and 14.4% for DSM-IV alcohol dependence in relatives of probands and controls, respectively... The risk of alcohol dependence in relatives of probands compared with controls is increased about 2-fold.\"\n","source_date":"2004-12-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20250202205013/https://pubmed.ncbi.nlm.nih.gov/15583116/","calculation_notes":"Native numerator (28.8 per 100) is taken directly from the Nurnberger 2004 finding for first-degree relatives of alcohol-dependent probands (parents, siblings, offspring). The 14.4% control rate is the matched same-instrument baseline, yielding a ~2.0x relative risk. Sample sizes: N=8,296 relatives of probands and N=1,654 controls assessed via the SSAGA structured diagnostic interview. DSM-IV alcohol-dependence criteria were applied; the broader DSM-5 AUD criteria would likely yield somewhat higher absolute prevalences in both groups while preserving the ~2x ratio. The 28.8% figure is used as the lifetime_us_adult for this subgroup; this is the best direct measurement available because no nationally representative study has measured DSM-5 AUD specifically in offspring of AUD parents. The first-degree-relative pool includes adult children of probands but is not exclusively offspring; this is acknowledged in `assumptions` and `caveats`.\n","independence_note":"The Collaborative Studies on the Genetics of Alcoholism (COGA) is a NIAAA-funded multi-site project (Indiana University, Washington University in St. Louis, SUNY Downstate, University of Connecticut, University of California San Diego, University of Iowa) that ascertained alcohol-dependent probands through inpatient and outpatient treatment programs and recruited their relatives plus a matched community-control sample. Probands and controls were diagnosed with the same SSAGA instrument, making the relative-vs-control comparison methodologically clean.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25171596/","title":"The heritability of alcohol use disorders: a meta-analysis of twin and adoption studies","publisher":"Verhulst B, Neale MC, Kendler KS — Psychological Medicine, 2015","source_type":"peer_reviewed","statistic":"Best-fit estimate of the heritability of AUD: h² = 0.49 (95% CI 0.43-0.53); meta-analysis of 12 twin studies and 5 adoption studies; no evidence for heterogeneity by sex","excerpt":"\"The best-fit estimate of the heritability of AUD was 0.49 [95% confidence interval (CI) 0.43-0.53]... There was no evidence for heterogeneity by study design, sex or assessment method.\"\n","source_date":"2015-03-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260116083405/https://pubmed.ncbi.nlm.nih.gov/25171596/","calculation_notes":"Establishes that roughly half of the variance in AUD risk is heritable, consistent with the magnitude of the COGA family-aggregation finding being part-genetic rather than purely shared-environmental. Used as supporting evidence for the mechanism behind the 2x elevation; not used to compute the headline rate. Twin and adoption studies separately estimate the genetic component because identical twins reared apart and adoptees raised by non-biological parents disentangle shared genes from shared environment; the meta-analytic h² = 0.49 means that even substantial early-life environmental exposure to a parent's drinking does not, on its own, drive the elevated risk seen in offspring.\n","independence_note":"This is a meta-analysis of independent twin and adoption studies conducted across multiple countries and decades, including the Virginia Twin Registry, Swedish adoption studies, and Australian twin samples. The aggregate h² estimate is methodologically independent of the COGA family-aggregation data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5240584/","title":"Epidemiology of DSM-5 Alcohol Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions III","publisher":"Grant BF, Chou SP, Saha TD, et al. — JAMA Psychiatry, 2015","source_type":"peer_reviewed","statistic":"Lifetime prevalence of DSM-5 alcohol use disorder among US adults: 29.1% overall; 12-month prevalence 13.9% (NESARC-III, N=36,309)","excerpt":"\"In 2012-2013, US prevalences of DSM-5 12-month and lifetime AUD among adults 18 years and older were 13.9% and 29.1%, respectively. Prevalence was generally highest for men (17.6% and 36.0%, respectively), White and Native American respondents, and younger and never-married adults.\"\n","source_date":"2015-08-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260301094058/https://pmc.ncbi.nlm.nih.gov/articles/PMC5240584/","calculation_notes":"Used as the US adult general-population reference for context. The coincidence between NESARC-III lifetime DSM-5 AUD (29.1%) and the COGA first-degree-relative DSM-IV alcohol-dependence figure (28.8%) is a methodological artifact, not a finding: DSM-5 AUD has a much broader definition than DSM-IV alcohol dependence, so the absolute numbers happen to land near each other despite measuring different things in different populations. The clean apples-to-apples comparison is COGA relatives 28.8% vs COGA controls 14.4%, both DSM-IV alcohol dependence, both SSAGA instrument — that is where the 2x elevation comes from.\n","independence_note":"NESARC-III was conducted by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) using probability sampling of the US non-institutionalized civilian population and the AUDADIS-5 structured diagnostic interview. Methodologically independent from COGA (which is a treatment-ascertained family study) and from the heritability meta-analysis.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15706789/","title":"The impact of a family history of alcoholism on the relationship between age at onset of alcohol use and DSM-IV alcohol dependence: results from the National Longitudinal Alcohol Epidemiologic Survey","publisher":"Grant BF — Alcohol Health and Research World, 1998","source_type":"peer_reviewed","statistic":"Family history of alcoholism is associated with higher lifetime DSM-IV alcohol-dependence prevalence; earlier age at first drink is independently associated with higher lifetime risk (NLAES, nationally representative)","excerpt":"\"People with a family history of alcoholism had a higher prevalence of lifetime alcohol dependence than did people without such a history. Respondents with an earlier age of drinking onset were more likely to become alcohol dependent compared with respondents with a later age of drinking onset.\"\n","source_date":"1998-01-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20251231071831/https://pubmed.ncbi.nlm.nih.gov/15706789/","calculation_notes":"Provides nationally representative cross-validation that family history of alcoholism elevates lifetime alcohol-dependence prevalence in the respondent, consistent with the direction of the COGA finding. Also establishes that early age at first drink interacts with family history to compound risk — supporting the framing that the elevation is probabilistic and modifiable by behavior, not deterministic. Not used to derive the headline rate; the NLAES instrument and sampling differ from COGA/SSAGA, so the figures are not directly substitutable.\n","independence_note":"The National Longitudinal Alcohol Epidemiologic Survey (NLAES, 1991-1992) was conducted by NIAAA using the AUDADIS instrument, independent of COGA's SSAGA-based family study.\n"}],"comparison_anchors":[{"label":"Lifetime alcohol use disorder (US adult, DSM-5)","lifetime_us_adult":0.291},{"label":"Lifetime major depression (US adult)","lifetime_us_adult":0.206},{"label":"Lifetime cannabis use disorder (US adult)","lifetime_us_adult":0.063},{"label":"Lifetime any anxiety disorder (US adult)","lifetime_us_adult":0.31}],"personal_factor_multipliers":[{"factor":"male (within first-degree-relative subgroup)","multiplier":1.6,"notes":"NESARC-III lifetime AUD: 36.0% men vs 22.7% women (ratio ~1.6). Sex effect is robust across general and high-risk populations and is well-replicated in COGA samples; Verhulst, Neale & Kendler 2015 found no evidence of heritability heterogeneity by sex, so the absolute sex differential in the parental-AUD subgroup is approximately the same multiplier as in the general population."},{"factor":"first-degree relative (no parent with AUD) — the matched baseline in COGA","multiplier":0.5,"notes":"Anchoring multiplier: COGA community controls had 14.4% lifetime DSM-IV alcohol dependence vs 28.8% in first-degree relatives of probands. The 0.5 multiplier represents a person from the general population (no parental AUD) relative to the headline subgroup."}],"short_label":"Inheriting AUD risk","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"Three framing limitations matter for reading this number correctly. (1) Population mismatch: COGA's \"first-degree relatives\" pools parents, siblings, and offspring of alcohol-dependent probands. The 28.8% figure is the best direct measurement available for offspring-specific risk because the genetic loading and early-environment exposure are comparable across these relationships, but it is not a study of adult children of AUD parents exclusively. (2) Diagnostic edition: COGA used DSM-IV alcohol dependence, which is more restrictive than DSM-5 AUD (DSM-5 collapsed the earlier \"abuse\" and \"dependence\" categories into a single disorder with a 2-criterion floor). A hypothetical DSM-5 re-analysis of COGA would likely yield higher absolute prevalences in both relatives and controls, but the 2x ratio is expected to hold. (3) Ascertainment: COGA probands were recruited through treatment programs, which selects for severe and clinically identified cases; the relative risk for offspring of a parent with mild, never-treated AUD is likely smaller. The 28.8% figure should not be read as a forecast for any individual — most adult children of alcoholic parents (about 71%) do not meet lifetime criteria for alcohol dependence. Heritability is not destiny: the Verhulst meta-analysis h² of 0.49 means roughly half the variance in liability is genetic, leaving substantial room for environmental, behavioral, and policy factors to shift individual outcomes. Adoption studies have separately established that the elevation persists in offspring of AUD biological parents even when raised by non-alcohol-using adoptive families, confirming that the genetic component is real and not purely a learned behavior, but those same studies also show that the elevation is substantially attenuated in protective environments.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-05-28","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single wooden chair next to an empty kitchen table at dusk, soft window light, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-of-alcoholic-aud-lifetime","api_url":"https://likelier.app/api/fears/child-of-alcoholic-aud-lifetime.json"},{"slug":"alcohol-use-disorder-lifetime","question":"What are the odds of developing alcohol use disorder over a lifetime?","quick_answer":"About 1 in 3 US adults, roughly 29%, meet criteria for alcohol use disorder at some point in life. Most people guess 5 to 10%, picturing only the visible severe cases — but the clinical definition also captures many milder cases that are invisible from the outside.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Most adults anchor their estimate of alcohol use disorder risk on the visible end of the spectrum: the person who has lost a job, a marriage, or a driver's license to drinking. That framing creates a population that looks much smaller than it is. The DSM-5 criteria for AUD span a wide severity range — from two or three symptoms (mild) to six or more (severe) — and mild cases meeting the two-criterion threshold are often invisible from the outside, so the visible severe cases are only part of the total 29% lifetime prevalence. Cultural normalization of heavy drinking in social, professional, and recreational contexts further compresses perceived risk: if most of the people around you drink heavily at times, the behavior stops registering as a marker of disorder. Public awareness campaigns have largely focused on drunk driving and fetal alcohol syndrome, leaving the population-level lifetime prevalence figure largely unknown outside clinical settings.\n","rough_estimate":"~5-10% of adults, perhaps","kind":"intuition"},"native":{"display":"29.1% of US adults meet DSM-5 criteria for alcohol use disorder at some point in their lifetime (NESARC-III, 2012–2013)","numerator":29.1,"denominator":100,"unit":"share of US adults with lifetime DSM-5 AUD diagnosis","population":"US adults aged 18 and older (NESARC-III, N=36,309, face-to-face interviews 2012–2013)"},"normalized":{"lifetime_us_adult":0.291,"display":"~1 in 3 US adults develops alcohol use disorder at some point in their lifetime","log_value":-0.54,"assumptions":"The NESARC-III (Grant et al., JAMA Psychiatry 2015) conducted face-to-face structured psychiatric interviews with 36,309 US adults aged 18 and older, using the AUDADIS-5 instrument to assess DSM-5 criteria. Lifetime AUD prevalence was 29.1% (12-month prevalence was 13.9%). The lifetime figure is used directly as the normalized estimate: it already represents a US adult population and already encompasses the full adult lifespan captured by retrospective diagnostic interviews. No additional conversion is needed. The 29.1% figure is not a per-year or per-exposure rate but a lifetime cumulative prevalence. Severity distribution (NESARC-III lifetime): 8.6% mild (2-3 criteria), 6.6% moderate (4-5 criteria), and 13.9% severe (6+ criteria), so severe is the single largest category. The SAMHSA 2024 NSDUH reports 27.9M past-year AUD (9.7% of people aged 12+), a lower 12-month prevalence than the NESARC-III 13.9% figure — the higher NESARC-III 12-month figure reflects differences in interview instrument and sampling. The 29.1% lifetime figure is the most widely cited and replicated estimate from the largest nationally representative study using DSM-5 criteria.\n","uncertainty":{"low":0.22,"high":0.4},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5240584/","title":"Epidemiology of DSM-5 Alcohol Use Disorder: Results From the National Epidemiologic Survey on Alcohol and Related Conditions III","publisher":"Grant BF et al. — JAMA Psychiatry, 2015","source_type":"peer_reviewed","statistic":"12-month and lifetime prevalences of DSM-5 AUD among US adults were 13.9% and 29.1%, respectively (NESARC-III, N=36,309)","excerpt":"\"In 2012-2013, US prevalences of DSM-5 12-month and lifetime AUD among adults 18 years and older were 13.9% and 29.1%, respectively. Prevalence was generally highest for men (17.6% and 36.0%, respectively), White and Native American respondents, and younger and never-married adults.\"\n","source_date":"2015-08-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260301094058/https://pmc.ncbi.nlm.nih.gov/articles/PMC5240584/","calculation_notes":"The 29.1% lifetime prevalence is used directly as the native numerator (29.1 per 100 US adults). This is the primary calculation input. The study used DSM-5 diagnostic criteria applied via the AUDADIS-5 structured interview, making it the definitive DSM-5 AUD prevalence estimate for the US adult population.\n","independence_note":"NESARC-III was conducted by the National Institute on Alcohol Abuse and Alcoholism (NIAAA) using probability sampling of the US non-institutionalized civilian population. It is methodologically independent from SAMHSA NSDUH, which uses self-report questionnaires rather than structured diagnostic interviews.\n"},{"url":"https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","title":"Key Substance Use and Mental Health Indicators in the United States: Results from the 2024 National Survey on Drug Use and Health","publisher":"Substance Abuse and Mental Health Services Administration (SAMHSA)","source_type":"govt_report","statistic":"27.9 million Americans aged 12 or older (9.7%) had alcohol use disorder in the past year in 2024","excerpt":"\"According to the 2024 National Survey on Drug Use and Health (NSDUH), 27.9 million people ages 12 and older (9.7% in this age group) had AUD in the past year. Marijuana use disorder was the most common drug use disorder (20.6 million), followed by opioid use disorder (4.8 million).\"\n","source_date":"2025-07-14","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260512142703/https://www.samhsa.gov/data/sites/default/files/reports/rpt56287/2024-nsduh-annual-national-report.pdf","calculation_notes":"SAMHSA 2024 NSDUH past-year AUD figure (27.9M people aged 12+ ≈ 9.7% of that age group) is used as a cross-validation anchor, not as the primary prevalence estimate. The lower 12-month figure relative to NESARC-III (9.7% in NSDUH vs 13.9% in NESARC-III) reflects instrument differences: NSDUH uses a self-report questionnaire while NESARC-III used a full structured diagnostic interview, which produces systematically higher prevalence estimates. Both sources converge on the same conclusion: AUD is the most prevalent substance use disorder in the US.\n","independence_note":"SAMHSA NSDUH is conducted by an independent contractor (RTI International) and uses a different sampling and assessment methodology than NESARC-III, providing a genuinely independent cross-check on the scale of the disorder.\n"},{"url":"https://www.niaaa.nih.gov/sites/default/files/NESARC/NESARC-III%20publications_Final_8_10_16.pdf","title":"National Epidemiologic Survey on Alcohol and Related Conditions III (NESARC-III) Publications List","publisher":"National Institute on Alcohol Abuse and Alcoholism (NIAAA)","source_type":"reputable_reference","statistic":"NESARC-III is the largest and most comprehensive nationally representative survey of alcohol and drug use and related psychiatric conditions ever conducted in the United States","excerpt":"\"The National Epidemiologic Survey on Alcohol and Related Conditions III (NESARC-III) is the largest, most comprehensive, nationally representative survey of alcohol and drug use and related psychiatric conditions ever conducted in the United States. It consisted of in-person interviews with 36,309 adults aged 18 years and older conducted in 2012-2013.\"\n","source_date":"2016-08-10","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260504183942/https://www.niaaa.nih.gov/sites/default/files/NESARC/NESARC-III%20publications_Final_8_10_16.pdf","calculation_notes":"Used here to establish the methodological standing of NESARC-III as the definitive source for DSM-5 AUD prevalence. The face-to-face structured interview methodology and probability sampling of 36,309 adults make it the gold standard for US lifetime AUD prevalence.\n"}],"comparison_anchors":[{"label":"Cannabis use disorder (lifetime, US adult)","lifetime_us_adult":0.063},{"label":"Developing opioid addiction after surgical prescription","lifetime_us_adult":0.0088},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"male","multiplier":1.7,"notes":"NESARC-III found 36.0% lifetime AUD in men vs 22.7% in women; roughly 1.6-1.7x higher in men"},{"factor":"family history of AUD (first-degree relative)","multiplier":3,"notes":"Genetic and environmental familial loading; approximately 50-60% of AUD risk is heritable"},{"factor":"age 18-25 at first regular drinking","multiplier":2,"notes":"Early onset of regular alcohol use is one of the strongest predictors of lifetime AUD"},{"factor":"co-occurring anxiety or mood disorder","multiplier":2.5,"notes":"NESARC-III found high comorbidity between AUD and mood/anxiety disorders; self-medication pathway"},{"factor":"currently drinking only occasionally or never drinks","multiplier":0.1,"notes":"Lifetime AUD prevalence includes past disorder; never-drinkers are essentially at zero prospective risk"}],"short_label":"Alcohol use disorder","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The 29.1% figure is a lifetime prevalence estimate based on retrospective structured diagnostic interviews with adults surveyed in 2012-2013. It captures disorder at any point in adulthood, including episodes that may have resolved decades earlier. Lifetime cases split roughly evenly across severity — 8.6% of adults mild (2-3 DSM-5 criteria), 6.6% moderate (4-5), and 13.9% severe (6+) — so severe (6+ criteria) is actually the largest single band, not the smallest, even though many mild cases carry substantially lower functional impairment. DSM-5 collapsed the prior DSM-IV categories of \"alcohol abuse\" and \"alcohol dependence\" into a single disorder, which accounts for some of the elevation compared to earlier DSM-IV-based estimates. Only 19.8% of adults with lifetime AUD ever sought treatment, meaning the large majority of cases — including most mild cases — are never clinically identified. The prospective risk for a young adult just beginning to drink is not precisely the same as this retrospective lifetime prevalence; risk accumulates over the years of heaviest use (typically ages 18-35).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"An empty glass on a plain surface beside a small calendar showing years passed, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/alcohol-use-disorder-lifetime","api_url":"https://likelier.app/api/fears/alcohol-use-disorder-lifetime.json"},{"slug":"childhood-rabbit-death","question":"What are the odds a rabbit given to a 9-year-old dies before they turn 13?","quick_answer":"A pet rabbit is fairly likely to outlast the four years — only about 3 in 10 die within that span, because well-cared-for indoor rabbits typically live 8 to 12 years. Many families wrongly assume rabbits are short-lived like hamsters.\n","category":"animal","tags":["kids","pets"],"no_reliable_estimate":false,"perceived":{"description":"Parents and children often assume rabbits are short-lived \"starter pets\" in the same league as hamsters — something that will likely be gone within a year or two. In practice, a well-cared-for indoor domestic rabbit typically lives 8–12 years, making a rabbit given to a 9-year-old quite likely to still be alive when that child turns 13. No rigorous survey has directly measured parental intuitions about rabbit lifespan in the context of childhood pet ownership, so this entry is marked intuition-kind.\n","rough_estimate":"many parents assume rabbits live 2-4 years, similar to hamsters","kind":"intuition"},"native":{"display":"~30 of 100 indoor pet rabbits die within 4 years of acquisition","numerator":30,"denominator":100,"unit":"per pet rabbit acquired","population":"Indoor domestic pet rabbits in family households"},"normalized":{"lifetime_us_adult":0.3,"display":"~3 in 10 pet rabbits die within 4 years of a child acquiring one","log_value":-0.52,"assumptions":"A retrospective Japanese study of 898 pet rabbits seen at an exotic animal clinic (2006–2020) found a median lifespan of 7 years (IQR 5–9), with 18% surviving beyond 9 years. Applying an exponential survival model with λ = ln(2)/7 ≈ 0.099 per year, the probability of dying within 4 years from acquisition (at ~8–12 weeks old) is 1 − e^(−0.099 × 4) ≈ 0.33. The House Rabbit Society reports indoor house rabbits typically live 8–12 years under optimal conditions; using a median of 10 years gives λ ≈ 0.069 and P(4 years) ≈ 0.24. A central estimate of 0.30 splits between typical and optimal care, reflecting that a child's first rabbit often receives adequate but not always expert husbandry. Scope is subgroup_lifetime: the figure applies to a 4-year window after pet acquisition, not to a 59-year adult lifetime.\n","uncertainty":{"low":0.18,"high":0.45},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.sciencedirect.com/science/article/pii/S1557506322000787","title":"Age at death and cause of death of pet rabbits (Oryctolagus cuniculus) seen at an exotic animal clinic in Tokyo, Japan: a retrospective study of 898 cases (2006–2020)","publisher":"Journal of Exotic Pet Medicine (ScienceDirect)","source_type":"peer_reviewed","statistic":"Median age at death of 898 pet rabbits was 7 years (IQR 5–9); 18% lived beyond 9 years","excerpt":"\"The median age at death was 7 years (interquartile range: 5 to 9 years), and 18% of all rabbits lived beyond 9 years. The main causes of death included neoplasia (n = 223; 24.8%), gastrointestinal disease (n = 135; 15.0%), bacterial abscess (n = 90; 10.0%), urinary disease (n = 85; 9.5%), trauma (n = 44; 4.9%), and cardiac disease (n = 27; 3.0%).\"\n","source_date":"2022-07-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20240204152025/https://www.sciencedirect.com/science/article/pii/S1557506322000787","calculation_notes":"Median lifespan of 7 years from this peer-reviewed retrospective is the primary anchor for the central estimate. Applying an exponential survival model: λ = ln(2) / 7 ≈ 0.099 per year. P(rabbit dies within 4 years of acquisition) = 1 − e^(−0.099 × 4) = 1 − 0.671 ≈ 0.33. This population (clinic patients in an exotic-animal practice in Tokyo) likely overrepresents rabbits receiving better-than-average care, so 0.33 may be a slight underestimate relative to the average child's pet rabbit. Used as the lower anchor of the central estimate.\n","independence_note":"Drawn from a Japanese exotic-animal clinic's case records (2006–2020); methodologically independent of the UK VetCompass cohort and the House Rabbit Society lifespan guidance.\n"},{"url":"https://rabbit.org/resources/how-long-do-rabbits-live/","title":"How Long Do Rabbits Live?","publisher":"House Rabbit Society","source_type":"reputable_reference","statistic":"Indoor house rabbits typically live 8–12 years","excerpt":"\"Indoor house rabbits typically live 8–12 years, depending on their size, breed, and the quality of care they receive. A well-cared-for house rabbit that has been spayed or neutered early in life has a life expectancy of 8 to 12 years.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20251211214755/https://rabbit.org/resources/how-long-do-rabbits-live/","calculation_notes":"Under ideal indoor care (spayed/neutered, optimal diet), median lifespan is approximately 10 years. Applying exponential survival: λ = ln(2)/10 ≈ 0.069 per year. P(dies within 4 years) = 1 − e^(−0.069 × 4) ≈ 0.24. This represents the lower bound of the uncertainty range — best-case indoor care. The 12-year upper end of HRS's range gives λ = 0.058, P(4 years) ≈ 0.21. The 8-year lower end gives λ = 0.087, P(4 years) ≈ 0.30, consistent with the central estimate. Difference between HRS guidance and clinical studies likely reflects selection bias in both directions: HRS population are committed rabbit owners; clinic patients skew toward unwell rabbits.\n","independence_note":"House Rabbit Society is the leading US rabbit welfare organization, drawing on veterinary guidance and member surveys. Methodologically independent of the Japanese and UK academic studies.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7279133/","title":"Morbidity and mortality of domestic rabbits (Oryctolagus cuniculus) under primary veterinary care in England","publisher":"Veterinary Record / O'Neill et al. (VetCompass Programme)","source_type":"peer_reviewed","statistic":"Median age at death among 370 rabbits that died during the study was 4.3 years (IQR 2.1–7.0, range 0.1–14.4)","excerpt":"\"The median age at death of the 370 rabbits that died during the study was 4.3 years (IQR 2.1–7.0, range 0.1–14.4). For males, the median age at death was older (5.2 years, IQR 3.0–8.1, range 0.2–14.4) than for females (3.7 years, IQR 2.0–5.9, range 0.1–11.8).\"\n","source_date":"2020-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505051652/https://pmc.ncbi.nlm.nih.gov/articles/PMC7279133/","calculation_notes":"The VetCompass UK figure of 4.3 years is the median age at death among rabbits that died during a 1-year observational window (2013) across 107 clinics — it is NOT a cohort survival median. It captures which rabbits were dying at the time of observation, which skews toward younger ages because older rabbits had not yet died. This number should not be read as \"median lifespan = 4.3 years.\" The Japanese retrospective (7-year median) is a more direct measure of lifespan because it follows rabbits through to death. The UK study is included as independent corroboration of mortality patterns and cause-of-death data (flystrike, GI stasis, anorexia) rather than as a lifespan anchor. The UK study also includes outdoor rabbits, who have significantly shorter lifespans than indoor pets.\n","independence_note":"VetCompass draws from primary-care veterinary records across England in 2013; methodologically independent of both the Japanese clinic study and the HRS member-survey data.\n"}],"comparison_anchors":[{"label":"Hamster dying within 4 years of acquisition (acquired age 9)","lifetime_us_adult":0.97},{"label":"Cat dying within 4 years (acquired as young adult, age 1–2)","lifetime_us_adult":0.05}],"short_label":"Rabbit dies in 4 years","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"Scope is subgroup_lifetime — this is a 4-year window probability, not a 59-year adult lifetime figure. Results vary significantly by breed (small breeds such as Netherland Dwarf live longer than large breeds like Flemish Giants), indoor vs. outdoor housing (outdoor rabbits have median lifespans roughly half as long), spay/neuter status (unspaying carries a high lifetime uterine-cancer risk in females, and the House Rabbit Society reports 8–12 year life expectancy for spayed/neutered indoor rabbits), age at acquisition (older rabbits have fewer years remaining), and access to veterinary care. A rabbit acquired at age 2 rather than 8 weeks would face a materially lower 4-year mortality risk. The 30% central estimate reflects a typical indoor pet rabbit with average-quality child-household care.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A child gently holding a white rabbit, with a calendar on the wall showing years passing"},"canonical_url":"https://likelier.app/childhood-rabbit-death","api_url":"https://likelier.app/api/fears/childhood-rabbit-death.json"},{"slug":"whistleblower-retaliation","question":"What are the odds of facing retaliation if you report workplace misconduct?","quick_answer":"Roughly 31 in 100 US workers face retaliation at some point over a working career if they report misconduct, and about 44 in 100 of those who actually file a report experience it. The instinct that speaking up makes you a target is empirically well-founded, not paranoid.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Fear of retaliation is the most commonly cited reason employees give for not reporting workplace misconduct. Survey after survey finds that workers who observe wrongdoing — fraud, safety violations, harassment, ethical breaches — hold back because they expect ostracism, demotion, or termination. The fear is so pervasive that ethics researchers treat it as a structural barrier to corporate accountability rather than an individual personality trait. Workers intuitively sense that speaking up makes them a target, and the data confirm the instinct is not paranoid.\n","kind":"intuition"},"native":{"display":"~44 in 100 employees who report workplace misconduct face some form of retaliation","numerator":44,"denominator":100,"unit":"per report of workplace misconduct filed","population":"US employees who observed and reported workplace misconduct (ECI NBES 2018)"},"normalized":{"lifetime_us_adult":0.31,"display":"roughly 31 in 100 over a working lifetime","log_value":-0.509,"assumptions":"Step 1 — Annual observation rate: ECI GBES 2023 found 53% of US employees observed misconduct in the prior 12 months (N = ~5,000 US respondents in a 70,000-person global sample). Step 2 — Reporting rate among observers: ECI NBES 2018, the most recent US-specific survey, found 69% of observers reported the misconduct. Annual probability of being a reporter in a given year = 0.53 × 0.69 ≈ 0.37.\nStep 3 — Career-level probability of reporting at all: Treating the 53% annual observation rate as applying each year produces unrealistically high cumulative figures, because misconduct is clustered within workplaces — a worker in a dysfunctional organisation observes it repeatedly, while workers in ethical workplaces may never observe any. A more defensible lifetime model uses the fraction of US workers who report at least once across their career: given ECI surveys show roughly 53–69% of observers report, and most full-career workers will observe misconduct at some point, a conservative estimate is that 65–75% of US workers will make at least one formal report in a 40-year career. Central estimate: 0.70.\nStep 4 — Retaliation conditional on reporting: ECI NBES 2018 (US-specific, 5,000+ employees) found 44% of US reporters experienced retaliation. The 2023 ECI GBES global figure was 46%; the 2020 GBES was also 46%.\nStep 5 — Lifetime retaliation probability: P(retaliation at least once in career) ≈ P(report at least once) × P(retaliation | report) = 0.70 × 0.44 ≈ 0.31.\nUncertainty range: low = 0.50 × 0.22 (2013 pre-doubling rate, pessimistic career-report assumption) ≈ 0.11. High = 0.90 × 0.46 (GBES 2023 global rate, optimistic career-report assumption) ≈ 0.41.\n","uncertainty":{"low":0.11,"high":0.41},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.navex.com/en-us/blog/article/ecis-2023-global-business-ethics-survey-reveals-harsh-realities-about-ec-programs/","title":"ECI's 2023 Global Business Ethics Survey Reveals Harsh Realities About E&C Programs","publisher":"NAVEX (reporting ECI Global Business Ethics Survey findings)","source_type":"reputable_reference","statistic":"46% of respondents globally who reported misconduct experienced retaliation, in both the 2023 and 2020 GBES reports","excerpt":"\"Retaliation against employees who reported misconduct remained steady. According to the report's findings, 46% of respondents globally in both the 2023 report and the 2020 report who reported misconduct said they experienced retaliation.\"\n","source_date":"2024-01-23","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260623191020/https://www.navex.com/en-us/blog/article/ecis-2023-global-business-ethics-survey-reveals-harsh-realities-about-ec-programs/","calculation_notes":"This is the global 2023 figure across 42 countries and 70,000+ employees. It provides the upper bound of the native retaliation rate (46%) and anchors the uncertainty high. The US-specific rate from the 2018 NBES (44%) is used as the native numerator because it is a US-only sample; the 2023 global figure serves as the primary citation confirming rate stability.\n"},{"url":"https://www.mspb.gov/studies/studies/Blowing_The_Whistle_Barriers_to_Federal_Employees_Making_Disclosures_662503.pdf","title":"Blowing the Whistle: Barriers to Federal Employees Making Disclosures","publisher":"U.S. Merit Systems Protection Board (MSPB)","source_type":"govt_report","statistic":"Approximately one-third of federal employees identified as a source of a wrongdoing report perceived threats or acts of reprisal (both 1992 and 2010)","excerpt":"\"In both 1992 and 2010, approximately one-third of the individuals who felt they had been identified as a source of a report of wrongdoing also perceived either threats or acts of reprisal, or both.\"\n","source_date":"2011-11-01","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20260210050446/https://www.mspb.gov/studies/studies/Blowing_The_Whistle_Barriers_to_Federal_Employees_Making_Disclosures_662503.pdf","calculation_notes":"The MSPB figure (~33%) applies specifically to federal employees who believed they were identified as a source — a narrower denominator than the ECI surveys, which ask all reporters. Federal-sector rates are generally lower than private- sector because civil-service protections are stronger. This source establishes a floor and provides the low-end anchor for uncertainty. It is not used for the native numerator (which is ECI NBES US-private-sector focused), but confirms that even in the most legally protected sector, retaliation is common.\n"},{"url":"https://www.prnewswire.com/news-releases/global-workplace-survey-us-rates-of-reported-misconduct-sets-national-record-retaliation-has-doubled-300615864.html","title":"Global Workplace Survey: US Rates of Reported Misconduct Sets National Record; Retaliation Has Doubled","publisher":"Ethics & Compliance Initiative (ECI) via PR Newswire","source_type":"primary_study","statistic":"44% of US employees who reported misconduct experienced retaliation in 2017, up from 22% in 2013 — a doubling in four years","excerpt":"\"Forty-four percent of respondents who reported misconduct said they experienced retaliation as a result of reporting — more than double the retaliation rate from ECI's 2013 National Business Ethics Survey.\"\n","source_date":"2018-03-14","source_accessed":"2026-05-02","archive_url":"http://web.archive.org/web/20250823183542/https://www.prnewswire.com/news-releases/global-workplace-survey-us-rates-of-reported-misconduct-sets-national-record-retaliation-has-doubled-300615864.html","calculation_notes":"This is the 2018 ECI National Business Ethics Survey (NBES), a US-specific survey of 5,000+ employed US adults collected December 2017. It is the most recent US-only retaliation rate published by ECI. The 44% figure is used as the native numerator (44 in 100 US reporters). The 2013 baseline of 22% provides the historical low; the 2018 reading of 44% is the US anchor. Consistent with the 2023 GBES global rate of 46%, suggesting the rate has plateaued in the 44–46% range.\n"}],"comparison_anchors":[{"label":"Being fired from a job (lifetime, US worker)","lifetime_us_adult":0.7},{"label":"Workplace injury requiring medical treatment (US, lifetime)","lifetime_us_adult":0.3},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.33}],"personal_factor_multipliers":[{"factor":"Federal-sector employee with civil-service protection","multiplier":0.75,"notes":"U.S. Merit Systems Protection Board (MSPB) 2011 study: approximately one-third of federal employees who were identified as wrongdoing-report sources perceived threats or acts of reprisal — compared to the 44% ECI NBES rate for the broader US workforce. The ~33% federal rate vs. ~44% private-sector rate implies approximately 0.75× relative risk for federal employees with Whistleblower Protection Act coverage."},{"factor":"Report involves media disclosure or public attention","multiplier":2,"notes":"Government Accountability Project (GAP) 2022 annual report and GAO retaliation studies: whistleblowers who involve external media or become publicly identified as a source face materially higher retaliation rates than those who report internally only. GAP case tracking data show that media-disclosed whistleblowers face a higher probability of formal job-adverse actions (termination, demotion) rather than the informal exclusion that dominates internal-only reports. A 2× multiplier is consistent with GAP and academic literature (Near & Miceli, whistleblowing research) on retaliation escalation with public visibility."}],"short_label":"Whistleblower retaliation","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"autonomy_loss","valence":"negative","caveats":"The 44–46% retaliation rate is conditional — it applies to employees who actually filed a report, not to all workers. Workers who stay silent avoid the risk entirely, but roughly 28–31% of observers choose not to report (ECI data), citing fear of retaliation as the primary reason. \"Retaliation\" in these surveys is self-reported and heterogeneous: it ranges from formal termination and demotion (the most severe) to informal exclusion, passed-over promotions, hostile management, and being ignored — all in the same number. Federal-sector employees face lower rates (~33% perceived reprisal, MSPB 2011) because civil-service protections are stronger; private-sector and healthcare workers face higher rates. Organisation size and industry culture matter substantially, though the various benchmark figures measure different things: incidence rates (the share of reporters who experience retaliation, as in the ECI surveys) are not directly comparable to complaint-substantiation rates (the share of filed retaliation complaints upheld after investigation). The ECI rate has roughly doubled since 2013, which may reflect more aggressive employer response to the post-#MeToo reporting surge rather than a structural increase in underlying retaliation propensity.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-05-02","reviewed":true,"generated_at":"2026-05-02","image":{"alt":"A sealed envelope in a corporate suggestion box casting a long shadow across an empty office desk"},"canonical_url":"https://likelier.app/whistleblower-retaliation","api_url":"https://likelier.app/api/fears/whistleblower-retaliation.json"},{"slug":"anxiety-disorder-lifetime","question":"What are the odds of developing an anxiety disorder in your lifetime?","quick_answer":"About 1 in 3 US adults meet criteria for an anxiety disorder at some point in life — the umbrella covers phobias, social anxiety, panic, and more. Most people would guess 5 to 15 percent, roughly half the actual figure, because they picture only the severe cases.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"There is no standard tracker of perceived lifetime anxiety-disorder risk, and the question is awkwardly reflexive — anxiety is partly defined by an overestimate of threat, so asking an anxious person to estimate their own probability invites the very bias being measured. The lay mental model of \"an anxiety disorder\" is narrow: most people picture severe generalised anxiety or recurrent panic attacks, and would put the population figure in the single digits. But the epidemiological category is an umbrella — generalised anxiety disorder, panic disorder, social anxiety disorder, specific phobias, agoraphobia, and (under DSM-5) separation anxiety — and it is precisely that breadth that pushes the lifetime number toward one in three. Specific phobias and social anxiety, in particular, are common enough that many people who would meet criteria never think of themselves as having \"a disorder\". Stigma compresses the intuitive estimate downward; the umbrella structure pushes the measured estimate up.\n","rough_estimate":"most adults would guess 5-15% lifetime, roughly half the actual figure","kind":"intuition"},"native":{"display":"31.1% lifetime prevalence of any anxiety disorder (NCS-R, NIMH)","numerator":311,"denominator":1000,"unit":"lifetime prevalence","population":"US adults 18+, noninstitutionalized civilian (NCS-R, 2001-2003)"},"normalized":{"lifetime_us_adult":0.311,"display":"~1 in 3 lifetime","log_value":-0.507,"assumptions":"Uses the NIMH-reported figure of 31.1% lifetime prevalence of any anxiety disorder among US adults, drawn from the National Comorbidity Survey Replication (NCS-R, 2001-2003), as the headline. This is a direct survey-based lifetime prevalence — no compounding or hazard conversion needed. The same NCS-R underpins Kessler et al. 2005 (Arch Gen Psychiatry), which reports the anxiety-disorders class at 28.8% lifetime; the two figures are both NCS-R-derived and differ by diagnostic subsetting and estimation method rather than by data source. The uncertainty band is anchored on this real spread: the low edge (0.288) is Kessler's 28.8% US estimate; the high edge (0.337) is Bandelow & Michaelis 2015's \"up to 33.7%\" cross-national review figure. Retrospective surveys are known to undercount lifetime episodes (people forget or reframe earlier anxious periods), so the high edge is a defensible ceiling rather than an outlier. NIMH also reports 19.1% past-year prevalence, consistent with the lifetime figure given the chronic, relapsing course of most anxiety disorders.\n","uncertainty":{"low":0.288,"high":0.337},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder","title":"Any Anxiety Disorder — Statistics","publisher":"National Institute of Mental Health (NIMH), NIH","source_type":"govt_report","statistic":"An estimated 19.1% of US adults had any anxiety disorder in the past year; an estimated 31.1% of US adults experience any anxiety disorder at some time in their lives; past-year prevalence higher for females (23.4%) than males (14.3%)","excerpt":"\"An estimated 19.1% of U.S. adults had any anxiety disorder in the past year... An estimated 31.1% of U.S. adults experience any anxiety disorder at some time in their lives... Past year prevalence of any anxiety disorder was higher for females (23.4%) than for males (14.3%).\"\n","source_date":"2023-01-01","source_accessed":"2026-06-21","archive_url":"https://web.archive.org/web/20260620210837/https://www.nimh.nih.gov/health/statistics/any-anxiety-disorder","calculation_notes":"NIMH republishes National Comorbidity Survey Replication (NCS-R) data, a nationally representative household survey conducted February 2001 to April 2003. The 31.1% lifetime figure is used directly as the normalized estimate — already a lifetime prevalence, so no hazard compounding. The female/male past-year ratio of 23.4%/14.3% ≈ 1.64 provides the female multiplier. The 60+ past-year prevalence of 9.0% against the 19.1% overall (ratio ≈ 0.47) provides the older-age protective factor.\n","independence_note":"NIMH/NCS-R (Harvard, Kessler) is one of the two canonical US psychiatric-epidemiology pipelines. Independent of the European-led Bandelow & Michaelis review below. Shares a data source with Kessler et al. 2005, which is acknowledged in the body and used to set the lower edge of the uncertainty band rather than as an independent confirmation.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15939837/","title":"Lifetime Prevalence and Age-of-Onset Distributions of DSM-IV Disorders in the National Comorbidity Survey Replication","publisher":"Archives of General Psychiatry (Kessler RC, Berglund P, Demler O, et al.)","source_type":"peer_reviewed","statistic":"Lifetime prevalence of anxiety disorders: 28.8%; any disorder 46.4%; median age of onset much earlier for anxiety (11 years) than for substance use (20) or mood (30) disorders","excerpt":"\"anxiety disorders, 28.8%... any disorder, 46.4%... Median age of onset is much earlier for anxiety (11 years) and impulse-control (11 years) disorders than for substance use (20 years) and mood (30 years) disorders.\"\n","source_date":"2005-06-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260613004246/https://pubmed.ncbi.nlm.nih.gov/15939837/","calculation_notes":"The flagship NCS-R lifetime-prevalence paper, based on 9,282 English-speaking US respondents aged 18+ interviewed 2001-2003 with structured DSM-IV diagnostic instruments. The 28.8% anxiety-disorders-class figure sets the lower edge of the uncertainty band. The verbatim \"anxiety (11 years)\" median age of onset establishes that anxiety disorders have the earliest median onset of any major disorder class — framed in prose, not as a fold-change multiplier.\n","independence_note":"Same NCS-R data source as the NIMH page above. Cited explicitly to document the real 28.8-vs-31.1 spread within the same survey (subsetting and estimation differences), not as an independent cross-validation. The third source (Bandelow & Michaelis) provides the genuinely independent cross-national check.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26487813/","title":"Epidemiology of anxiety disorders in the 21st century","publisher":"Dialogues in Clinical Neuroscience (Bandelow B, Michaelis S)","source_type":"peer_reviewed","statistic":"Up to 33.7% of the population are affected by an anxiety disorder during their lifetime; anxiety disorders are the most prevalent mental disorders; substantial underrecognition and undertreatment demonstrated","excerpt":"\"Anxiety disorders... are the most prevalent mental disorders and are associated with immense health care costs and a high burden of disease... up to 33.7% of the population are affected by an anxiety disorder during their lifetime. Substantial underrecognition and undertreatment of these disorders have been demonstrated.\"\n","source_date":"2015-09-01","source_accessed":"2026-06-21","archive_url":"https://web.archive.org/web/20260619133532/https://pubmed.ncbi.nlm.nih.gov/26487813/","calculation_notes":"A widely cited review synthesising European and international community surveys. The \"up to 33.7%\" lifetime figure is cross-national, not US-specific, and is used as the upper edge of the uncertainty band — the plausible ceiling consistent with the known retrospective undercount in survey-based lifetime estimates. Also the authoritative basis for the \"most prevalent mental disorder\" claim and for the underrecognition note.\n","independence_note":"Independent of the US NCS-R pipeline: different authors (Göttingen), different underlying surveys (predominantly European community studies), different diagnostic traditions. Agreement that lifetime any-anxiety prevalence sits in the high-20s to mid-30s percent across independent survey programs is the strongest cross-validation available for this figure.\n"}],"comparison_anchors":[{"label":"Major depressive disorder (lifetime, US adult)","lifetime_us_adult":0.206},{"label":"Eating disorder (lifetime, US)","lifetime_us_adult":0.09},{"label":"Alzheimer's disease (lifetime, US adult)","lifetime_us_adult":0.107},{"label":"Suicide (lifetime, US adult)","lifetime_us_adult":0.00827}],"personal_factor_multipliers":[{"factor":"Female","multiplier":1.6,"notes":"NIMH/NCS-R past-year prevalence: 23.4% for females vs 14.3% for males, ratio ≈ 1.64. This is a past-year (not lifetime) figure; the female-skewed sex gap is one of the most replicated findings in anxiety epidemiology and holds across diagnostic systems."},{"factor":"First-degree family history of an anxiety disorder","multiplier":4,"notes":"Hettema, Neale & Kendler (2001, Am J Psychiatry) meta-analysis: summary odds ratios for first-degree relatives ranged from 4 to 6 depending on the disorder; heritability 0.43 for panic disorder and 0.32 for generalised anxiety disorder. The 4x used here is the conservative low end of that 4-6 range."},{"factor":"4+ adverse childhood experiences (ACEs)","multiplier":3,"notes":"Hughes et al. (2017, Lancet Public Health) meta-analysis (n≈253,719): four or more ACEs carried strong associations with mental ill health, with 'ORs of more than three to six' for the grouped mental-ill-health outcomes (which include anxiety). The 3x used here is the low edge of that stated range; the published OR is for the broader mental-ill-health category, not anxiety alone."},{"factor":"Age 60+","multiplier":0.5,"notes":"Protective. NIMH/NCS-R past-year prevalence falls to 9.0% among adults 60+ vs 19.1% overall (ratio ≈ 0.47). Anxiety disorders show a natural decline in prevalence with older age, though they remain underdiagnosed in older adults."}],"short_label":"Anxiety disorder","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"\"Any anxiety disorder\" is an umbrella aggregating generalised anxiety disorder, panic disorder, social anxiety disorder, specific phobias, agoraphobia, and (under DSM-5) separation anxiety disorder. These have very different base rates, age-of-onset patterns, and impairment levels — specific phobias and social anxiety are far more common than panic disorder or GAD, and much of the one-in-three lifetime figure is carried by these milder, often unrecognised presentations. The same NCS-R is cited at both 28.8% (Kessler et al. 2005, anxiety-disorders class) and 31.1% (NIMH, any anxiety disorder) lifetime; the difference reflects diagnostic subsetting and estimation method within one survey, not a data discrepancy. Retrospective surveys undercount lifetime episodes because respondents forget or reframe earlier anxious periods, so the headline figure is plausibly a floor; the 33.7% cross-national upper bound (Bandelow & Michaelis 2015) reflects that. Conversely, survey diagnostic instruments may flag subthreshold cases a clinician would not treat. The past-year figure (~19% of US adults) is the more policy-relevant number for treatment capacity; the lifetime figure answers \"how common is this, really?\" Anxiety disorders are treatable, but they are among the most under-recognised and under-treated conditions in psychiatry.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"Tightly spaced concentric ripples spreading across a still pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/anxiety-disorder-lifetime","api_url":"https://likelier.app/api/fears/anxiety-disorder-lifetime.json"},{"slug":"home-hail-roof-damage","question":"What are the odds that hail will seriously damage your roof during your lifetime?","quick_answer":"The lifetime odds that hail seriously damages a US homeowner's roof are about 1 in 3, from roughly 552,000 homeowner hail claims a year, and 2-3x higher in hail-alley states like Texas and Colorado. Most homeowners treat hail as a minor nuisance and routinely underestimate it as a property risk.\n","category":"property","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Most homeowners think of hail as a minor nuisance — dented cars and broken windows. Few instinctively rank it among their top property risks. No formal survey tracks perceived hail-damage frequency directly, but anecdotal evidence from insurance agents and the persistent underinsurance of hail-prone regions suggests the risk is routinely underestimated, particularly outside the recognized \"hail alley\" states of Texas, Colorado, Kansas, Oklahoma, and Nebraska.\n","rough_estimate":"most homeowners don't expect serious hail damage in their lifetime","kind":"intuition"},"native":{"display":"~552,000 homeowner-only hail insurance claims per year (US)","numerator":5525,"denominator":83000000,"unit":"per year","population":"US insured homeowners (~83 million policies)"},"normalized":{"lifetime_us_adult":0.33,"display":"~1 in 3 lifetime (US homeowner)","log_value":-0.48,"assumptions":"The National Insurance Crime Bureau (NICB) analysis of ISO ClaimSearch data found 1,657,663 Personal Property-Homeowners hail claims over 2016-2018, yielding approximately 552,554 homeowner hail claims per year. With approximately 83 million insured US homes, this implies an annual hail-claim rate of 552,554 / 83,000,000 ≈ 0.666% per homeowner policy per year. Compounded over 59 adult years (the site's standard horizon): 1 − (1 − 0.00666)^59 ≈ 0.33. The NICB 2.9 million total figure includes auto and commercial property — only the homeowner-only segment (57% of total) is used here to match the population denominator of insured homes. This is a claims-based measure counting events where the homeowner filed and the insurer paid; below-deductible damage and uninsured events are excluded, so the true damage probability is higher. Geographic rates in hail alley states (TX, CO, KS, OK, NE) run 2-3× the national average.\n","uncertainty":{"low":0.2,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://resilience.iii.org/resilience-blog/tornadoes/nicb-top-states-for-hail-claims/","title":"NICB: Top states for hail claims","publisher":"Insurance Information Institute (citing NICB / ISO ClaimSearch)","source_type":"reputable_reference","statistic":"2.9 million hail loss claims across homeowner and auto policies, 2016-2018; personal property homeowner segment was 1,657,663 claims (57% of total)","excerpt":"\"According to a NICB review of claims data from ISO ClaimSearch®, there were a total of 2.9 million hail loss claims in the United States from 2016 through 2018, with Personal Property-Homeowners being the most affected with 1,657,663 claims or 57 percent of the three-year total.\"\n","source_date":"2020-04-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260207174431/https://resilience.iii.org/resilience-blog/tornadoes/nicb-top-states-for-hail-claims/","calculation_notes":"Total 3-year homeowner hail claims: 1,657,663. Annual average: 1,657,663 / 3 = 552,554 homeowner-only claims per year. Annual rate: 552,554 / 83,000,000 insured homes = 0.666%/yr. Over 59 adult years: 1 − (1 − 0.00666)^59 ≈ 0.33. The 2.9 million total figure encompasses all property types including auto; only the homeowner-specific segment (1.66M / 57% share) is used as the numerator to match the homeowner denominator. The homeowner-only rate is the correct basis for this entry.\n","independence_note":"NICB/ISO ClaimSearch data aggregates reported insurance claims; this is independent of NOAA storm event counts and captures only insured losses above deductible thresholds.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-homeowners-and-renters-insurance","title":"Facts + Statistics: Homeowners and renters insurance","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"About one in 36 insured homes has a property damage claim related to wind or hail; 2.8% of insured homes experienced a wind/hail loss over 2018-2022, average claim severity $13,511","excerpt":"\"About one in 36 insured homes has a property damage claim related to wind or hail. Between 2018-2022, 2.8 percent of insured homes experienced a loss due to wind and hail damage, with an average claim severity (or cost) of $13,511.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260518113841/https://www.iii.org/fact-statistic/facts-statistics-homeowners-and-renters-insurance","calculation_notes":"The III \"1 in 36\" figure (~2.78%/yr) combines wind and hail damage claims and serves as an upper-bound cross-check. The NICB-derived homeowner-only hail rate of 0.67%/yr is lower because it is hail-specific and excludes straight-line wind damage. The average wind/hail severity of $13,511 (2018-2022) is used for context; the primary rate estimate derives from the NICB homeowner-specific hail claims count. Used as a cross-check only, not as a basis for the normalized probability.\n","independence_note":"III homeowners loss data is compiled from insurer filings aggregated through ISO; methodologically independent from NICB ISO ClaimSearch claims-count data, though both ultimately trace to insurance company reporting.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-hail","title":"Facts + Statistics: Hail","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"5,373 hail events in 2024 (NOAA SPC); hail consistently among the costliest US property perils","excerpt":"\"According to NOAA's National Weather Service Storm Prediction Center Annual Severe Weather Report Summary, there were 5,373 hail events in 2024.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260519035058/https://www.iii.org/fact-statistic/facts-statistics-hail","calculation_notes":"Used for context on event frequency; the count of distinct hail events is not directly translatable to per-home claim rates without geographic and storm-path data, so this figure is cited for framing only.\n","independence_note":"NOAA Storm Prediction Center compiles hail events from weather observer and radar data, entirely independent of insurance company claims systems.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Lives in hail alley (TX, CO, KS, OK, NE)","multiplier":2.5,"notes":"NICB data shows Texas, Nebraska, Minnesota, Kansas, and Colorado consistently top hail claim rankings. Rates in these states run 2-3× the national average according to III/NICB hail claim geography data.\n"},{"factor":"Has impact-resistant roof (Class 4 UL 2218)","multiplier":0.3,"notes":"Class 4 impact-resistant roofing materials pass the highest UL 2218 drop-test rating; insurers offer 10-35% premium discounts in hail-prone states, and empirical claim data shows substantially lower loss frequency vs standard asphalt shingles exposed to the same storms.\n"},{"factor":"Home >20 years old with original roof","multiplier":1.5,"notes":"Aging asphalt shingles lose granule coverage and become more vulnerable to hail puncture; 20+ year original roofs are already near or past typical useful life (25-30 yr) and sustain damage at lower hail sizes than newer roofs.\n"}],"short_label":"Hail roof damage","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"This entry measures insured hail damage claims to homeowner properties — it is a property damage risk, not a safety or mortality risk. The claim rate includes all payout sizes above the deductible; below-deductible damage is not captured. The NICB figure combines homeowner and some commercial property in the raw data; the homeowner-isolated rate (552K/yr) produces a lower lifetime estimate (~33%), while the all-property hail rate (970K/yr) yields ~50%. The true homeowner lifetime probability is likely in the 33-50% range. Geographic variation is the dominant uncertainty factor: a homeowner in Nebraska or Texas faces roughly 2-3× the risk of a homeowner in the Northeast or Pacific Northwest.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"Hailstones on a rooftop, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/home-hail-roof-damage","api_url":"https://likelier.app/api/fears/home-hail-roof-damage.json"},{"slug":"menopause-cardiovascular-shift","question":"How likely is a woman to develop cardiovascular disease after menopause — and how much does menopause accelerate that risk?","quick_answer":"About 1 in 3 women faces a lifetime risk of coronary heart disease from age 40, and earlier menopause raises it — roughly 3% more per year of earlier menopause. The role of menopause as a cardiovascular risk accelerant is widely underappreciated.\n","category":"health","tags":["mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Cardiovascular disease is widely understood to be the leading cause of death in women globally, but the specific role of menopause as a risk accelerant is not well understood by most women. Cultural attention to breast cancer — a statistically less common outcome — has displaced awareness of the cardiovascular risk profile that menopause alters. Women in their late 40s and early 50s often do not know that their CVD risk doubles in the decade following their final menstrual period, or that the transition represents a meaningful window for preventive intervention. The asymmetry is reinforced by clinical practice: menopause consultations have historically focused on symptom management (vasomotor, sleep, genitourinary) rather than cardiovascular risk stratification.\n","kind":"intuition"},"native":{"display":"33 in 100 women globally will develop cardiovascular disease during their lifetime (from age 40)","numerator":33,"denominator":100,"unit":"lifetime from age 40","population":"women from age 40 (lifetime CHD risk, Framingham / Lloyd-Jones 1999; menopause-timing modifiers from Lancet Public Health 2019 pooled IPD and the EHJ 2021 consensus document)"},"normalized":{"lifetime_us_adult":0.33,"display":"About 1 in 3 women has a lifetime risk of coronary heart disease from age 40; earlier menopause meaningfully raises that risk","log_value":-0.48,"assumptions":"The native rate (33/100 ≈ 1 in 3) is the lifetime risk of coronary heart disease from age 40 reported by Lloyd-Jones et al. (Lancet 1999, Framingham): 31.7% (95% CI 29.2–34.2) for women. This is the figure that gives the familiar \"one in three women\" headline. Menopause timing modifies this baseline: the Lancet Public Health 2019 pooled IPD analysis (15 studies, 301,438 women) found premature menopause (<40) carried a hazard ratio of 1.55 (95% CI 1.38–1.73) for first non-fatal CVD versus menopause at 50–51, rising to 1.88 before age 60. The EHJ 2021 consensus document adds a ~3% increase in CVD risk per year of earlier menopause and notes the menopause-associated rise in CVD risk cannot be cleanly separated from chronological ageing — so this entry treats menopause as a timing-dependent risk modifier, not a fully aging-independent doubling. The 1-in-3 lifetime figure is US/Framingham-derived for CHD specifically; the broader \"cardiovascular disease\" umbrella and global/regional estimates vary, which the uncertainty band and caveats reflect. Low (0.28): the lower CI / favourable risk-factor and regional profiles (e.g. East Asian women with lower CVD burden). High (0.40): higher-burden profiles (e.g. Eastern European / Central Asian women with high hypertension and smoking prevalence, and broader all-CVD rather than CHD-only counts).\n","uncertainty":{"low":0.28,"high":0.4},"scope":"subgroup_lifetime"},"sources":[{"url":"https://academic.oup.com/eurheartj/article/42/10/967/6125214","title":"Cardiovascular health after menopause transition, pregnancy disorders, and other gynaecologic conditions: a consensus document from European cardiologists, gynaecologists, and endocrinologists","publisher":"European Heart Journal (European Society of Cardiology)","source_type":"peer_reviewed","statistic":"Each year of early menopause is associated with a ~3% increased risk of CVD; women with premature ovarian insufficiency have shorter life expectancy due to CVD; the CVD-risk increase at menopause cannot be cleanly distinguished from ageing; MHT initiated below age 60 or within 10 years of menopause onset was associated with a >30% reduction in MI or cardiac deaths (window-of-opportunity effect)","excerpt":"\"Each year of early menopause was associated with a 3% increased risk of CVD. Women with premature ovarian insufficiency (POI), defined as the loss of ovarian function before the age of 40, have a shorter life expectancy than women with a late menopause due to CVD and osteoporosis. While CVD risk increases with the menopause, this cannot be distinguished from ageing. Lower oestrogen levels after menopause are related to altered vascular function, enhanced inflammation, and up-regulation of other hormonal systems.\" \"Women initiating MHT treatment below 60 years of age or within 10 years of onset of menopause showed a significant reduction (>30%) of MI or cardiac deaths.\"\n","source_date":"2021-03-07","source_accessed":"2026-06-30","calculation_notes":"Maas et al. (2021), European Heart Journal 42(10):967–984 — a consensus document from European cardiologists, gynaecologists, and endocrinologists (formally approved by the ESC Clinical Practice Guidelines Committee, but not the 2021 ESC CVD-prevention guideline itself). Used here for the menopause-specific risk modifiers: the 3%-per-year early-menopause gradient, premature ovarian insufficiency, and the estrogen-vascular mechanism. The same document also describes a menopausal-hormone-therapy (MHT) \"window of opportunity\": women initiating MHT below age 60 or within 10 years of menopause onset showed a significant (>30%) reduction in MI or cardiac deaths, which anchors the MHT paragraph in the body text. Note the document explicitly cautions that the menopause-associated CVD-risk rise cannot be cleanly separated from chronological ageing — the entry reflects this caveat rather than claiming a fully aging-independent effect.\n"},{"url":"https://www.thelancet.com/journals/lanpub/article/PIIS2468-2667(19)30155-0/fulltext","title":"Age at natural menopause and risk of cardiovascular disease: pooled individual participant data from 15 observational studies","publisher":"The Lancet Public Health","source_type":"peer_reviewed","statistic":"Premature menopause (<40) associated with HR 1.55 (95% CI 1.38–1.73) for first non-fatal CVD vs menopause at 50–51; the excess is larger before age 60 (HR 1.88, 1.62–2.20). Pooled IPD, 301,438 women, 15 studies.","excerpt":"\"Compared with women who had menopause at age 50-51 years, the risk of cardiovascular disease was higher in women who had premature menopause (age <40 years; HR 1·55, 95% CI 1·38-1·73; p<0·0001), early menopause (age 40-44 years; 1·30, 1·22-1·39; p<0·0001), and relatively early menopause (age 45-49 years; 1·12, 1·07-1·18; p<0·0001). The associations persisted in never smokers, and were strongest before age 60 years for women with premature menopause (HR 1·88, 1·62-2·20; p<0·0001) and early menopause (1·40, 1·27-1·54; p<0·0001), but were attenuated at age 60-69 years, with no significant association observed at age 70 years and older.\"\n","source_date":"2019-11-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20250717144409/https://www.thelancet.com/journals/lanpub/article/PIIS2468-2667(19)30155-0/fulltext","calculation_notes":"Zhu et al. (2019) Lancet Public Health — pooled IPD, 15 studies, 301,438 women (12,962 first non-fatal CVD events; 9,369 CHD; 4,338 stroke). The largest individual-participant dataset quantifying the menopause-timing–CVD relationship. The reported risk is for first non-fatal cardiovascular disease (a composite), not a separate CHD hazard ratio; premature menopause (<40) carries HR 1·55 overall and 1·88 before age 60. These figures support the `personal_factor_multipliers` for premature/surgical menopause used in this entry.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/10023892/","title":"Lifetime risk of developing coronary heart disease","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Lifetime risk of coronary heart disease at age 40 is 31.7% (95% CI 29.2–34.2) for women and 48.6% for men — i.e. about one in three women. Framingham Heart Study, 7,733 participants.","excerpt":"\"Lifetime risk of CHD after 40 years of age was 48·6% (95% CI 45·8–51·3) for men and 31·7% (29·2–34·2) for women. At 70 years of age the remaining lifetime risk was 34·9% for men and 24·2% for women. Interpretation: Lifetime risk at age 40 years is one in two for men and one in three for women.\"\n","source_date":"1999-01-09","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20251001023151/https://pubmed.ncbi.nlm.nih.gov/10023892/","calculation_notes":"Lloyd-Jones, Larson, Beiser & Levy (1999), Lancet 353(9147):89–92 — Framingham Heart Study lifetime-risk analysis (n=7,733, CHD-free at baseline). This is the original source of the \"one in three women from age 40\" lifetime coronary-heart-disease figure that anchors this entry's native rate. Note the headline figure is for coronary heart disease specifically (not the full CVD umbrella) and is US/Framingham-derived; the entry's caveats flag that broader-CVD and global figures differ.\n"},{"url":"https://www.goredforwomen.org/en/about-heart-disease-in-women/facts","title":"The Facts about Women and Heart Disease","publisher":"American Heart Association / Go Red for Women","source_type":"reputable_reference","statistic":"Cardiovascular disease kills more US women each year than all forms of cancer combined; only 44% of women recognize CVD as their greatest health threat","excerpt":"\"Cardiovascular disease kills more women than all forms of cancer combined and yet only 44% of women recognize that cardiovascular disease is their greatest health threat.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714063537/https://www.goredforwomen.org/en/about-heart-disease-in-women/facts","calculation_notes":"American Heart Association's Go Red for Women campaign facts page. Used here to source the entry's opening framing claim that cardiovascular disease kills more US women annually than all cancers combined, which motivates the \"underrated\" myth framing alongside the menopause-specific timing evidence from Maas et al. and Zhu et al. This is an all-cause CVD statistic, not menopause-specific.\n","independence_note":"Independent of the Framingham (Lloyd-Jones 1999), Lancet Public Health (Zhu 2019), and EHJ consensus (Maas 2021) sources — an AHA public-health communications source used solely for the all-cause CVD-vs-cancer mortality comparison in US women, not for any menopause-timing figure.\n"}],"comparison_anchors":[{"label":"Breast cancer, lifetime risk (women globally)","lifetime_us_adult":0.12},{"label":"Any stroke, lifetime (global, age 25+)","lifetime_us_adult":0.25}],"personal_factor_multipliers":[{"factor":"Early menopause (40–44)","multiplier":1.3,"notes":"Early menopause (age 40–44) carried HR 1.30 (95% CI 1.22–1.39) for first non-fatal CVD vs menopause at 50–51; Lancet PH 2019 pooled IPD (n=301,438)"},{"factor":"Premature menopause / ovarian insufficiency (<40)","multiplier":1.6,"notes":"Premature menopause before 40 carried HR 1.55 (95% CI 1.38–1.73) for first non-fatal CVD overall (and 1.88 before age 60) in the Lancet PH 2019 pooled analysis; POI also raises osteoporosis risk"}],"short_label":"Menopause CV risk acceleration","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline \"1 in 3 from age 40\" is the lifetime risk of coronary heart disease specifically, from the Framingham cohort (Lloyd-Jones 1999, 31.7% for women); it is US-derived and CHD-specific, not a global all-CVD figure, so global and regional estimates differ. The broader \"cardiovascular disease\" umbrella (adding stroke, peripheral artery disease, and heart failure) and fatal-CHD-only counts give different numbers. The menopause-timing risk is a relative-risk increase, not an absolute incidence: in the Lancet PH 2019 pooled data premature menopause (<40) carried HR 1.55 for a first non-fatal CVD event (1.88 before age 60), and the EHJ 2021 consensus reports ~3% added CVD risk per year of earlier menopause — but that same consensus stresses the menopause-related rise cannot be cleanly separated from chronological ageing, so the \"acceleration\" should not be read as a fully aging-independent doubling. If a woman's absolute CHD risk in her 40s is low, a relative increase still leaves the absolute level low. The MHT window-of-opportunity evidence is clinically relevant but not embedded in the native probability; it belongs in a clinical discussion, not a population baseline. Regional variation is substantial — Japanese and Korean women have lower CVD lifetime risk and Eastern European / Central Asian women higher — reflecting different hypertension and smoking profiles, which the uncertainty band approximates.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a simple heart rate monitor line on a plain background, muted tones."},"canonical_url":"https://likelier.app/menopause-cardiovascular-shift","api_url":"https://likelier.app/api/fears/menopause-cardiovascular-shift.json"},{"slug":"undiagnosed-type2-diabetes","question":"What are the odds of having undiagnosed type 2 diabetes or prediabetes without blood glucose monitoring?","quick_answer":"About 4.5% of US adults have diabetes they don't know about, and roughly 33% will develop type 2 diabetes over a lifetime — nearly all passing through an undiagnosed phase first. Common guesses of a 5-10% chance of undetected blood-sugar problems match the point-in-time figure reasonably well.\n","category":"health","tags":["elder-care","food"],"no_reliable_estimate":false,"perceived":{"description":"Most people assume that diabetes is something you would notice — blurred vision, excessive thirst, fatigue severe enough to prompt a doctor visit. The intuition is that the disease announces itself before it does serious damage. When asked to guess how many Americans have undetected diabetes right now, the typical answer is \"maybe a few percent.\" The idea that roughly 1 in 3 adults could be walking around with undetected prediabetes or diabetes at any given moment sits well outside the range most people would guess, partly because the early and middle stages of metabolic dysfunction can be entirely symptom-free for years.\n","rough_estimate":"~5-10% chance of having undetected blood sugar problems","kind":"intuition"},"native":{"display":"4.5% of US adults have undiagnosed diabetes (NCHS, 2021-2023)","numerator":45,"denominator":1000,"unit":"point prevalence among US adults","population":"US adults aged 18+"},"normalized":{"lifetime_us_adult":0.33,"display":"~33% lifetime probability of developing type 2 diabetes at some point, nearly all of whom pass through an undiagnosed phase","log_value":-0.48,"assumptions":"Two quantities are combined here. The first is the lifetime risk of developing type 2 diabetes for a US adult: a PLOS One 2022 analysis of National Health Interview Survey data (1997-2018) found lifetime risk for a 20-year-old was 31.7% in 1997-1999, peaked at 40.7% in 2005-2009, and returned to 32.8% (95% CI: 32.4-33.2%) in 2015-2018, consistent with Narayan et al.'s 2003 JAMA finding of 32.8% for males and 38.5% for females born in 2000. The headline 0.33 uses the most recent PLOS One estimate for a 20-year-old in 2015-2018 as the primary anchor. The second quantity is the probability of being in an undiagnosed state during the course of that disease: NCHS data for August 2021-August 2023 shows 27.6% of US adults with diabetes were undiagnosed at any given time, and the average preclinical phase of type 2 diabetes (the period of elevated glucose before clinical diagnosis) is estimated at 7-12 years in the literature. Because almost everyone who develops T2D passes through this silent phase, the lifetime probability of being in the \"undiagnosed\" state at some point is essentially the same as the lifetime probability of developing T2D at all. Separately, 115.2 million US adults have prediabetes, of whom 8 in 10 do not know it (CDC, 2026), meaning roughly 92 million adults currently have undetected prediabetes alone — a point prevalence near 35% of the adult population. The normalized 0.33 reflects the lifetime T2D development risk as the primary anchor. Uncertainty band of 0.25-0.45 covers the demographic spread (Hispanic adults face ~45-53% lifetime risk per Narayan et al.; non-Hispanic white adults face ~27-31%) and the upward and downward trend uncertainty since 2018.\n","uncertainty":{"low":0.25,"high":0.45},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ncbi.nlm.nih.gov/books/NBK612760/","title":"Prevalence of Total, Diagnosed, and Undiagnosed Diabetes in Adults: United States, August 2021–August 2023","publisher":"National Center for Health Statistics (CDC), NCHS Data Briefs No. 516","source_type":"govt_report","statistic":"Total diabetes 15.8%, diagnosed 11.3%, undiagnosed 4.5% of US adults; 27.6% of adults with diabetes are undiagnosed","excerpt":"\"During August 2021–August 2023, the prevalence of total diabetes was 15.8%, diagnosed diabetes was 11.3%, and undiagnosed diabetes was 4.5%. [...] Consequently, slightly more than one-quarter of adults with diabetes had undiagnosed diabetes.\"\n","source_date":"2024-11-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20250806181025/https://www.ncbi.nlm.nih.gov/books/NBK612760/","calculation_notes":"Undiagnosed diabetes prevalence of 4.5% directly gives the native figure: 4.5 per 100 US adults, or 45 per 1,000. The 27.6% undiagnosed fraction (of those with diabetes) is used in the normalized assumptions to estimate how long a typical person with T2D remains in the silent phase. US adults aged 18+ number approximately 260 million, so 4.5% corresponds to roughly 11.7 million people currently living with undiagnosed diabetes in the US. The total diabetes prevalence of 15.8% is the anchor for the denominator: undiagnosed / total = 4.5% / 15.8% = 28.5%, consistent with the stated 27.6% fraction.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9129010/","title":"Trends in lifetime risk and years of potential life lost from diabetes in the United States, 1997–2018","publisher":"PLOS One","source_type":"peer_reviewed","statistic":"Lifetime risk of diabetes for a 20-year-old US adult: 31.7% (1997-1999), peaked at 40.7% (2005-2009), returned to 32.8% (2015-2018)","excerpt":"\"LR for adults at age 20 increased from 31.7% (95% CI: 31.2–32.1%) in 1997–1999 to 40.7% (40.2–41.1%) in 2005–2009, then decreased to 32.8% (32.4–33.2%) in 2015–2018.\"\n","source_date":"2022-06-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20250226214510/https://pmc.ncbi.nlm.nih.gov/articles/PMC9129010/","calculation_notes":"The 2015-2018 estimate of 32.8% lifetime risk for a 20-year-old US adult is used as the primary anchor for normalized.lifetime_us_adult. This figure is essentially unchanged from the Narayan et al. 2003 JAMA estimate of 32.8% for males born in 2000 (females 38.5%), despite the intervening two decades of prevalence fluctuation, and the 95% CI of 32.4-33.2% is narrow enough to be useful as a point estimate. The decline from the 2005-2009 peak (40.7%) is attributed in the paper to improvements in diabetes prevention and treatment and reductions in obesity-trend growth rates during this period. This source is the normalization anchor: lifetime_us_adult 0.33 sits at the lower end of the CI for the most recent period, reflecting the most conservative recent estimate and the fact that the question is framed around \"currently unmonitored adults\" rather than the total population including those already diagnosed and under treatment.\n"},{"url":"https://www.cdc.gov/diabetes/communication-resources/prediabetes-statistics.html","title":"Prediabetes: Could It Be You?","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"115.2 million Americans have prediabetes; 8 in 10 adults with prediabetes don't know they have it","excerpt":"\"115.2 million Americans have prediabetes, but 8 in 10 adults with prediabetes don't know they have it.\"\n","source_date":"2026-01-21","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260504061250/https://www.cdc.gov/diabetes/communication-resources/prediabetes-statistics.html","calculation_notes":"115.2 million adults with prediabetes, 80% unaware, gives approximately 92 million adults with currently undetected prediabetes. US adults number approximately 260 million, so 92M / 260M = ~35% point prevalence of unaware prediabetes. Adding the 4.5% with undiagnosed diabetes gives approximately 38-40% of US adults currently in a state of undetected elevated blood glucose (either undiagnosed diabetes or unaware prediabetes). This source is used in the prose and caveats to establish the combined prediabetes-plus-diabetes picture; the native figure uses only the more specific undiagnosed-diabetes prevalence from the NCHS government report.\n"}],"comparison_anchors":[{"label":"Lifetime hypertension development (US adult)","lifetime_us_adult":0.8},{"label":"Lifetime obesity development (US adult)","lifetime_us_adult":0.5},{"label":"Lifetime coronary heart disease (US adult)","lifetime_us_adult":0.4},{"label":"Lifetime Parkinson's disease (global adult)","lifetime_us_adult":0.008}],"personal_factor_multipliers":[{"factor":"obesity (BMI ≥30)","multiplier":3.5,"notes":"Obesity is the single strongest modifiable risk factor; T2D prevalence was 24.2% in adults with obesity vs 6.8% in those with normal weight (NCHS 2021-2023)"},{"factor":"family history of T2D (first-degree relative)","multiplier":2.5,"notes":"First-degree family history roughly doubles to triples lifetime risk; heritability of T2D is estimated at 40-80% from twin studies"},{"factor":"Hispanic ethnicity","multiplier":1.5,"notes":"Narayan et al. 2003 found lifetime risk of 45.4% for Hispanic males and 52.5% for Hispanic females vs 32.8%/38.5% overall"},{"factor":"regular aerobic exercise (150+ min/week)","multiplier":0.4,"notes":"The Diabetes Prevention Program showed intensive lifestyle intervention reduced T2D incidence by 58% in high-risk adults; exercise is the most effective single protective behavior"},{"factor":"age 65+","multiplier":2,"notes":"Prevalence of total diabetes rises to 27.3% for adults 60 and older vs 3.6% for ages 20-39 (NCHS 2021-2023)"}],"short_label":"Silent diabetes","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"Two distinct risks are presented together here and should not be conflated. Undiagnosed type 2 diabetes (4.5% of adults currently) and unaware prediabetes (roughly 35% of adults currently) are different physiological states with different urgency. Prediabetes is reversible with lifestyle intervention; established T2D is chronic and progressive. The normalized lifetime figure of 0.33 represents the probability of developing T2D at any point in life, not the probability of being currently undiagnosed. The \"8 in 10 prediabetes unaware\" statistic reflects a point-in-time awareness deficit, not a lifetime outcome. The lifetime risk figure from PLOS One (2022) covers diagnosed diabetes including type 1, but type 1 accounts for only about 5-10% of all diabetes cases; the T2D-specific lifetime risk is slightly lower but in the same range. Risk varies substantially by ethnicity, obesity status, physical activity level, and access to preventive care. These figures apply to US adults; global estimates differ significantly by country.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-04","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A faint unmarked line on a plain graph, barely visible against a muted background, flat vector illustration."},"canonical_url":"https://likelier.app/undiagnosed-type2-diabetes","api_url":"https://likelier.app/api/fears/undiagnosed-type2-diabetes.json"},{"slug":"at-fault-injury-crash-liability","question":"What are the odds of causing an injury crash as an at-fault driver?","quick_answer":"Over a driving lifetime, about 1 in 2.9 drivers will be the at-fault party in a crash that injures someone — roughly 1 in 140 per year. Optimism bias leads most drivers to guess their odds are well below 1 in 10.\n","category":"transport","tags":[],"no_reliable_estimate":false,"perceived":{"description":"Drivers systematically underestimate their personal crash risk — a well-documented optimism bias in traffic safety research. Most licensed drivers believe they are above-average in skill, and the mental model of an \"at-fault injury crash\" tends to conjure images of reckless or drunk drivers, not the ordinary distracted or rushed driver that characterizes most real-world events. The actual lifetime probability of being the at-fault driver in a crash that injures someone is substantially higher than most people would guess.\n","rough_estimate":"Most drivers would guess their lifetime odds of causing an injury crash are well below 1 in 10","kind":"intuition"},"native":{"display":"~1 in 140 per year per US driver (at-fault injury crash rate)","numerator":1,"denominator":140,"unit":"per year","population":"US licensed drivers (NHTSA 2023 injury-crash count ÷ FHWA 2023 licensed drivers; ~1 at-fault driver per injury crash)"},"normalized":{"lifetime_us_adult":0.34,"display":"~1 in 2.9 over a 59-year driving lifetime","log_value":-0.46,"assumptions":"NHTSA's Traffic Safety Facts Annual Report Portal records an estimated 1,697,252 police-reported INJURY crashes in 2023 (a crash count, distinct from the 2,442,694 persons injured, which also counts passengers, pedestrians, and pedalcyclists). Each injury crash has, to first approximation, one at-fault driver — the single driver in a run-off-road crash, or the party who caused a multi-vehicle collision. Dividing the injury-crash count by the 237,655,885 licensed drivers (FHWA 2023) gives the annual probability that a given driver is the at-fault driver in an injury crash: 1,697,252 / 237,655,885 ≈ 0.714% per year, or roughly 1 in 140. Compounding over 59 adult driving years: 1 − (1 − 1/140)^59 ≈ 0.34, or about 1 in 2.9. This uses a numerator and denominator with matched national coverage (national injury-crash count ÷ national licensed-driver count) and avoids both the persons-injured overcount and an arbitrary at-fault split. The uncertainty range reflects that some injury crashes have no clearly at-fault driver while others assign fault ambiguously among multiple parties: at 0.85–1.15 at-fault drivers per injury crash the rate spans roughly 1 in 165 to 1 in 122 per year (lifetime ≈ 0.30 to 0.38).\n","uncertainty":{"low":0.28,"high":0.4},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","title":"Summary of Motor Vehicle Traffic Crashes: 2023 Data","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"2.44 million people injured in motor vehicle traffic crashes in 2023; 6.14 million total police-reported crashes; up from 2.38 million injured in 2022","excerpt":"\"In 2023 an estimated 2.44 million people were injured in motor vehicle traffic crashes, compared to 2.38 million in 2022, an increase of 2.5 percent. The estimated number of police-reported traffic crashes increased from 5.93 million in 2022 to 6.14 million in 2023, a 3.5-percent increase.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260413215932/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","calculation_notes":"The headline uses the number of INJURY CRASHES (a crash count) as the numerator, not the 2.44 million persons injured — the latter also counts passengers, pedestrians, and pedalcyclists, so dividing it by drivers does not measure at-fault drivers. NHTSA's annual crash-severity breakdown records 1,697,252 police-reported injury crashes in 2023 (see the Traffic Safety Facts Annual Report Portal source below). Treating each injury crash as having ~1 at-fault driver: 1,697,252 / 237,655,885 licensed drivers (FHWA 2023) ≈ 0.714% per driver per year, or 1 in 140. Lifetime: 1 − (1 − 0.00714)^59 ≈ 0.34, about 1 in 2.9.\n","independence_note":"NHTSA's CRSS (Crash Report Sampling System) is a probability sample of police-reported crashes; it is independent from insurance claims databases and court records. CRSS replaced NASS-GES as the national non-fatal crash data system beginning with 2016 crash year data.\n"},{"url":"https://www.bts.gov/content/licensed-drivers","title":"Licensed Drivers — Bureau of Transportation Statistics","publisher":"Bureau of Transportation Statistics / Federal Highway Administration","source_type":"govt_report","statistic":"237.7 million licensed drivers in the United States in 2023","excerpt":"\"According to the Federal Highway Administration, there were approximately 237.7 million licensed drivers in the United States in 2023, derived from state motor vehicle administration data submitted to FHWA's Highway Statistics Series.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508191540/https://www.bts.gov/content/licensed-drivers","calculation_notes":"The licensed-driver denominator is used to convert total injury-crash counts (NHTSA) to a per-driver annual rate. FHWA's Highway Statistics DL-22 series provides the most comprehensive and consistently updated count of US licensed drivers by state, sex, and age group.\n","independence_note":"FHWA licensed-driver counts are based on state DMV administrative records and are wholly independent from NHTSA crash investigation data. The two datasets are combined here to derive a per-driver rate not published as such by either agency alone.\n"},{"url":"https://cdan.dot.gov/tsftables/tsfar.htm","title":"Traffic Safety Facts Annual Report — National Statistics (Police-Reported Crashes by Severity)","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"2023: 1,697,252 police-reported injury crashes; 37,769 fatal crashes; 4,403,453 property-damage-only crashes; 6,138,474 total crashes; 237,655,885 licensed drivers","excerpt":"\"POLICE-REPORTED MOTOR VEHICLE TRAFFIC CRASHES [2023]: Fatal 37,769; Injury 1,697,252; Property-Damage-Only 4,403,453; Total 6,138,474. Licensed Drivers 237,655,885. Note: Property-Damage-Only, Injury, and Total crash numbers are estimates calculated from a nationally representative sample of crashes collected through NHTSA's Crash Report Sampling System (CRSS).\"\n","source_date":"2026-06-01","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260413215932/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","calculation_notes":"This is the authoritative crash-count-by-severity table. It gives the INJURY CRASH count (1,697,252) as a crash category — the coverage-matched numerator for \"at-fault driver in an injury crash,\" since each injury crash has roughly one at-fault driver. It also confirms the FHWA 2023 licensed-driver denominator (237,655,885), so both numerator and denominator come from the same national table. 1,697,252 / 237,655,885 ≈ 1 in 140 per driver per year.\n","independence_note":"Crash counts derive from NHTSA's CRSS probability sample of police reports; the licensed-driver count is FHWA administrative DMV data. Both are national in coverage, so the per-driver injury-crash rate is coverage-matched.\n"},{"url":"https://injuryfacts.nsc.org/motor-vehicle/overview/introduction/","title":"Motor Vehicle — Overview — Injury Facts","publisher":"National Safety Council","source_type":"reputable_reference","statistic":"Medically consulted motor vehicle injuries totaled 4.9 million in 2024; 39,345 estimated traffic fatalities in 2024; odds of dying in a motor vehicle crash are 1 in 101 lifetime","excerpt":"\"Medically consulted injuries in motor-vehicle incidents totaled 4.9 million in 2024. The National Safety Council estimates the odds of dying in a motor vehicle crash at approximately 1 in 101 over a lifetime for the average American.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260518025408/https://injuryfacts.nsc.org/motor-vehicle/overview/introduction/","calculation_notes":"The NSC medically consulted injury figure of 4.9 million in 2024 is higher than NHTSA's 2.44 million police-reported injured count because it includes crashes not reported to police and injuries with delayed medical presentation. It is cited here for context on the upper bound of total injury events, not for the primary rate calculation, which uses the police-reported NHTSA figure.\n","independence_note":"NSC compiles injury data from insurance payors, employer records, and medical billing datasets, making it methodologically distinct from both NHTSA's police-report sampling system and FHWA's DMV administrative records.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Deer-vehicle collision (lifetime, US driver)","lifetime_us_adult":0.37}],"personal_factor_multipliers":[{"factor":"Under age 25","multiplier":2,"notes":"NHTSA Young Drivers report: drivers 16-24 are involved in fatal and injury crashes at approximately twice the rate per mile driven of drivers 25-64. At-fault rates are even more skewed toward younger drivers due to inexperience."},{"factor":"Regular phone use while driving (handheld)","multiplier":1.5,"notes":"NHTSA estimates distracted driving contributes to approximately 8% of fatal crashes; phone-specific research places crash risk at 1.5-3x elevated. Habitual users face above-average per-mile exposure."},{"factor":"Drives alcohol/cannabis-impaired regularly","multiplier":3,"notes":"Alcohol- and drug-impaired driving accounts for a large share of US traffic fatalities each year and sharply elevates crash risk. Even occasional impaired driving sessions substantially raise annual at-fault crash probability."},{"factor":"Low annual mileage driver (<5,000 miles/year)","multiplier":0.4,"notes":"Per-mile crash rates are not strictly proportional to mileage (low-mileage drivers have better per-mile rates), but lower total exposure meaningfully reduces annual crash probability."}],"short_label":"At-fault injury crash","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 1/140 annual rate is derived from the national count of police-reported injury crashes (1,697,252 in 2023) divided by the national count of licensed drivers (237,655,885), treating each injury crash as having roughly one at-fault driver. This keeps the numerator and denominator on the same national basis, but the one-at-fault- driver-per-crash assumption is itself an approximation: some injury crashes (e.g. a chain-reaction pileup) have no single clearly at-fault party, while a few assign fault to more than one driver. This is a population-level estimate; the actual proportion of crashes with a clearly designated at-fault driver varies by crash type, state law, and reporting practice. Some fraction of \"injury crashes\" involve only the at-fault driver as the injured party (single-vehicle run-off-road, etc.), where the financial and legal consequences differ from injuring a third party. The lifetime figure compounds annual independent exposures, which is a reasonable first-order model but does not capture the clustering of risk in early driving years or the dose-response of total mileage. The psychological consequence of injuring another person — separate from insurance, legal liability, or financial cost — is a real component of the risk that is not captured in any administrative statistic.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"Two simplified car outlines after a collision with a small starburst impact mark, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/at-fault-injury-crash-liability","api_url":"https://likelier.app/api/fears/at-fault-injury-crash-liability.json"},{"slug":"bicycle-theft","question":"What are the odds your bicycle gets stolen?","quick_answer":"About 1 in 3 US adults will have a bicycle stolen over a lifetime, and the odds run higher for regular cyclists. This is one of the few risks where intuition tracks the data closely.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Cyclists' intuition is shaped by stories — the lifted U-lock from outside the library, the second-hand bike that turned up on Craigslist a week later, the friend who switched to a $90 commuter after their good road bike disappeared. The headline number is hard to pin down because police data captures only a fraction of what happens, and most cyclists rely on lived experience: \"give it enough years and somebody will get you.\" That instinct is closer to right than wrong. Across a typical US adult lifetime — most of which includes some period of bike ownership — at least one stolen bicycle is more likely than not for committed cyclists, and roughly a coin-flip across all adults.\n","rough_estimate":"~1 in 3 over a US adult lifetime; higher among regular cyclists","kind":"intuition"},"native":{"display":"~710 stolen adult bicycles per 100,000 US adults per year (2024 estimate)","numerator":710,"denominator":100000,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.34,"display":"~1 in 3 over a US adult lifetime; higher for regular bike owners","log_value":-0.47,"assumptions":"Agarwal et al. (2025), a peer-reviewed YouGov-weighted survey of 1,748 US adults plus a 430-cyclist booster, estimate roughly 2.4 million adult bicycles are stolen in the United States each year — a population rate of about 709.6 per 100,000 adults per year. Naive compounding over a 59-year adult life-horizon, treating each year as an independent trial at the population rate, gives 1 − (1 − 0.0071)^59 ≈ 0.343. This is a per-US-adult lifetime probability, not a per-owner figure: it bakes in the fact that not every adult owns a bike, and not every owner owns one continuously. Among regular cyclists the figure is substantially higher — Agarwal et al.'s cyclist booster sub-sample reports materially elevated annual victimization rates, and the UK Crime Survey for England and Wales has long shown about 2% of bicycle-owning households are hit per year, which compounds to roughly 1 − 0.98^59 ≈ 70% over an adult lifetime of continuous bike ownership. The repeat-victimization rate in bicycle theft is well-documented (a small fraction of people get hit multiple times), which nudges the unique-victim estimate downward relative to the naive compound. Uncertainty is wide because the survey is the first large modern US estimate and supersedes a frequently-cited but methodologically thin \"1.5 million per year\" NCVS extrapolation from a 2008 COPS guide.\n","uncertainty":{"low":0.2,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://findingspress.org/article/127974-bicycle-theft-in-the-us-magnitude-and-equity-impacts","title":"Bicycle Theft in the US: Magnitude and Equity Impacts","publisher":"Findings (peer-reviewed open-access transport journal)","source_type":"peer_reviewed","statistic":"About 2.4 million adult bicycles are stolen annually in the US, a rate of 709.6 per 100,000 people per year. The annual value of adult bicycle theft is about $1.4 billion. In 2019, about 157,669 bicycles were reported stolen to law enforcement — implying roughly 93% of incidents go unreported.","excerpt":"\"About 2.4 million adult bicycles are stolen annually in the US, a rate of 709.6 per 100,000 people per year. The annual value of adult bicycle theft is about $1.4 billion. In 2019, about 157,669 bicycles were reported stolen to law enforcement in the United States.\"\n","source_date":"2025-01-16","source_accessed":"2026-05-25","archive_url":"https://web.archive.org/web/20260214222159/https://findingspress.org/article/127974-bicycle-theft-in-the-us-magnitude-and-equity-impacts","calculation_notes":"The 709.6/100,000/year is the headline annual rate per all US adults (denominator includes non-owners). For the normalized lifetime figure we treat each adult-year as an independent trial at that rate and compound over a 59-year horizon: 1 − (1 − 0.00710)^59 ≈ 0.343. The 2.4 million annual total comes from Agarwal et al.'s YouGov-weighted survey (n = 1,748 general population + 430 active cyclist booster, March 2024), with weighting to match US adult demographics on gender, age, race, education, and region. The implied underreporting ratio (2.4M survey vs 157,669 police-reported in 2019) is about 15:1 — far higher than the ~2:1 NCVS-vs-UCR ratio for burglary, partly because most stolen bikes are individually too low-value to motivate filing a police report.\n","independence_note":"This is the primary survey source; cross-checked against Bike Index's registry-based theft counts (which align directionally with the survey's growth claim) and the FBI Crime Data Explorer NIBRS figure (~127,646 reported in 2023) as a lower bound.\n"},{"url":"https://popcenter.asu.edu/content/bicycle-theft-0","title":"Bicycle Theft (Problem-Oriented Guide for Police No. 52)","publisher":"US Department of Justice, Office of Community Oriented Policing Services; authored by Shane D. Johnson, Aiden Sidebottom, and Adam Thorpe","source_type":"govt_report","statistic":"Across 17 countries surveyed (including the United States), on average only 56 percent of bicycle thefts were reported to the police; most bicycles reported stolen are taken from on or near the premises of the victim's home (including garages and sheds), or from outside of shops or recreational facilities.","excerpt":"\"Across the 17 countries surveyed (including the United States), on average only 56 percent of bicycle thefts were reported to the police. … Most bicycles reported stolen are taken from on or near the premises of the victim's home (including garages and sheds), or from outside of shops or recreational facilities.\"\n","source_date":"2008-01-01","source_accessed":"2026-05-25","archive_url":"https://web.archive.org/web/20260118023350/https://popcenter.asu.edu/content/bicycle-theft-0","calculation_notes":"Used for the qualitative framing of where thefts happen (counter-intuitively, the residential premises and garage are the modal location rather than the public rack) and as supporting evidence for the underreporting story. The 56% reporting rate in this guide is an international cross-country average; the Agarwal 2025 US-specific figure is much lower (~7% reported, or 93% unreported) and supersedes it for the US headline. The 2004 US figure cited in the guide (\"more than 250,000 bicycles stolen each year\" based on NCVS) is also superseded by Agarwal 2025.\n","independence_note":"DOJ COPS Office is an independent law-enforcement-research arm; the authors are UK-based academic criminologists (UCL Jill Dando Institute) writing under US federal grant. Independent of the Agarwal 2025 transport-research pipeline.\n"},{"url":"https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/overviewofbicycletheft/latest","title":"Overview of bicycle theft in England and Wales","publisher":"Office for National Statistics (UK), Crime Survey for England and Wales","source_type":"govt_report","statistic":"Around 2 in 100 bicycle-owning households in England and Wales were victims of bicycle theft in the previous 12 months; this is roughly one-third of the 1995 peak of 6 in 100.","excerpt":"\"Around 2 in 100 bicycle-owning households have been victims of bicycle theft in the previous 12 months … this is around a third of the rate in the 1995 peak of around 6 in 100 households. The Crime Survey for England and Wales has collected information on bicycle theft in a consistent manner since the survey first ran in 1981.\"\n","source_date":"2017-07-20","source_accessed":"2026-05-25","archive_url":"https://web.archive.org/web/20260307121548/https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/articles/overviewofbicycletheft/latest","calculation_notes":"International cross-check anchor. 2% per bicycle-owning household per year compounds to 1 − 0.98^59 ≈ 0.70 over an adult lifetime of continuous bike ownership, which is consistent with the much higher per-owner rate implied by Agarwal 2025 once non-owners are filtered out of the denominator. The UK Crime Survey is a long-running household victimization survey directly comparable in methodology to the US NCVS but with explicit bicycle-theft tabulation that NCVS lacks.\n","independence_note":"ONS is the UK's independent national statistics agency. CSEW methodology is fully independent of both Agarwal 2025 (US YouGov panel) and the COPS 2008 guide (US police-data review).\n"}],"comparison_anchors":[{"label":"Vehicle theft (lifetime, US adult, 12-year ownership)","lifetime_us_adult":0.0296},{"label":"Home burglary (lifetime, US adult)","lifetime_us_adult":0.11},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"short_label":"Bicycle theft","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The 34% headline is a population-average lifetime probability — it mixes bike-owning adults with non-owners. Among committed cyclists who own and ride a bike continuously for decades, the lifetime probability of at least one theft is much higher; Agarwal et al.'s cyclist booster sub-sample reports materially elevated annual rates, and the UK CSEW per-owning-household figure compounds to roughly 70% over 59 years of continuous ownership. The bicycle theft rate also varies enormously by city density, lock quality, and parking location, but the academic literature is qualitative on these multipliers — Sidebottom et al. (2009) note that cyclists are about three times more likely to have a bicycle stolen than car owners are to have their car stolen, and the COPS guide stresses that \"secure\" locks are those that resist hand-tool attack for at least three minutes, but there is no peer-reviewed relative-risk study quantifying a U-lock vs cable-lock multiplier. The COPS guide does state that \"most stolen bicycles, regardless of theft location, are either not locked at all or are secured using a lock that requires little force to break or remove\" — meaning unlocked and cable-locked bikes dominate the victim pool — but this is a share-of-stolen-bikes statistic, not a per-bike-night rate ratio, and cannot be converted into a clean \"left unlocked outside the cafe\" multiplier without an exposure denominator (what fraction of cyclist parking events are unsecured) that no published study reports. The most consistent finding across the literature is counter-intuitive: bicycles are more often stolen from residential premises, garages and sheds than from public racks, partly because public-rack exposure is intermittent while at-home exposure is continuous. Agarwal et al. also find that lower-income households and several non-white racial groups in the US experience significantly higher theft rates; exact multipliers are in the paper's tables and are not reproduced here. The economic value of an individual theft is modest — median bike value in Agarwal's sample is $374.50 — which is why so few thefts get reported.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-25","last_reviewed":"2026-05-25","reviewed":true,"generated_at":"2026-05-25","image":{"alt":"An empty bicycle rack on a city sidewalk with a single cut cable lock looped through it, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/bicycle-theft","api_url":"https://likelier.app/api/fears/bicycle-theft.json"},{"slug":"flying-with-uri","question":"What are the odds of ear barotrauma or a ruptured eardrum from flying with a cold?","quick_answer":"Flying with an active cold carries roughly a 1 in 3 chance per flight of moderate-or-worse ear barotrauma, one of the few folk fears that roughly matches what people already guess (about 1 in 20 to 1 in 3). An actually ruptured eardrum is far rarer and not reliably reported in the aviation literature.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Flying with a head cold is one of the more durable pieces of folk-medicine anxiety — most people who have ever had an ear \"pop and stay popped\" on descent will tell the story for years, and aviation crews are loudly warned off the flight deck while sick. The intuition is that flying congested is meaningfully risky, probably around 1 in 20 to 1 in 3 of ending a flight with a sore ear, and a small but non-trivial chance of actual tympanic-membrane damage. That intuition is unusually close to what the aviation-medicine literature finds, which makes this one of the few folk fears that survives a look at the numbers roughly intact. We have not found a standalone survey isolating \"fear of flying congested\", so perceived risk is marked as editorial intuition.\n","rough_estimate":"most travelers guess 1 in 20 to 1 in 3 per flight for 'meaningful' ear pain when flying with a cold","kind":"intuition"},"native":{"display":"~1 in 3 per flight for moderate-or-worse ear barotrauma when flying with an active URI","numerator":1,"denominator":3,"unit":"per flight taken while symptomatic with a URI","population":"adult commercial-aviation passengers flying with an active upper-respiratory infection"},"normalized":{"lifetime_us_adult":0.34,"display":"~1 in 3 per flight taken while sick (activity-specific)","log_value":-0.469,"assumptions":"Scope is activity_specific_lifetime and the headline is a per-flight probability, not a population-lifetime accumulation. Baseline per-flight ear-pain rate in healthy adults sits around 10-20 percent (Stangerup/Klokker prevalence studies summarised in Wright's BMJ Clinical Evidence review: \"20% of adult and 40% of child passengers had negative pressure in the middle ear after flight, and 10% of adults and 22% of children had otoscopic evidence of changes to the ear drum\"). In the Csortan/Jones placebo arm of 250 adult passengers with a history of recurrent ear discomfort, 62 percent of controls reported some ear pain per flight; Wright's review cites a separate placebo arm where 29/41 (71 percent) of adults had barotrauma symptoms without pseudoephedrine. When a traveler is actively congested with a URI, the pre-flight pilot survey (Rosenkvist et al. 2008, 948 commercial pilots) found 37.6 percent had experienced at least one ear-barotrauma episode in their career, with 90 percent of those events on descent — and pilots pre-select strongly for healthy Eustachian-tube function. The headline \"roughly 1 in 3 per URI-flight for moderate-or- worse barotrauma\" is the midpoint of (a) the 62-71 percent any-symptom rate in predisposed placebo arms and (b) the ~10-20 percent objective-finding rate in unselected passengers, triangulated against the pilot-survey career rate. Tympanic- membrane perforation is a much rarer outcome: Wright's review states plainly that perforation rates are \"not reported\" in the commercial-aviation literature and are \"extremely rare\"; the order-of-magnitude estimate from case-series and aviation- medicine textbooks is roughly 1 per 1,000 to 1 per 10,000 URI-flights, captured in the regional_breakdown rather than the headline.\n","uncertainty":{"low":0.15,"high":0.6},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/15949100/","title":"Otic barotrauma from air travel","publisher":"Journal of Laryngology & Otology (Mirza S, Richardson H)","source_type":"peer_reviewed","statistic":"Otic barotrauma is a common problem in air travellers; middle-ear pressure changes during descent are the dominant mechanism; topical and oral decongestants are the main evidence-based prophylaxis with modest effect sizes in randomised trials","excerpt":"\"Otic barotrauma occurring during air travel involves traumatic inflammation of the middle ear, caused by a pressure difference between the air in the middle ear and the external atmosphere, developing after ascent or more usually descent.\"\n","source_date":"2005-05-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250204080050/https://pubmed.ncbi.nlm.nih.gov/15949100/","calculation_notes":"Mirza & Richardson is the canonical narrative review of air-travel otic barotrauma, published in J Laryngol Otol 119(5):366-70. The paper does not supply a single headline incidence figure — it summarises the fragmentary passenger and aircrew epidemiology and the three randomised decongestant trials available at the time — so we use it as the mechanistic and methodological anchor, not the quantitative anchor. The paper's core clinical claim, that a blocked Eustachian tube cannot equilibrate cabin-pressure changes during descent and that this is the direct mechanism of barotrauma, is the foundation for the \"URI flight = higher rate\" step in the native calculation. The 10-20 percent baseline adult rate and the 26-55 percent child rate cited in the Native field trace back through this review to Stangerup, Klokker, Csortan/Jones, and the Tonkin/Fagan series.\n","independence_note":"Mirza & Richardson synthesise the Stangerup auto-inflation trials and the Csortan/Jones pseudoephedrine RCT; treat it as an editorially independent review of primary data that overlaps methodologically with both the Wright BMJ review and the Rosenkvist pilot survey cited below.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4298289/","title":"Middle-ear pain and trauma during air travel","publisher":"BMJ Clinical Evidence (Wright T)","source_type":"peer_reviewed","statistic":"20% of adult and 40% of child passengers had negative middle-ear pressure after flight; 10% of adults and 22% of children had otoscopic evidence of tympanic-membrane changes; pseudoephedrine reduced adult barotrauma from 29/41 (71%) placebo to 14/41 (34%) active (RR 0.48, 95% CI 0.29-0.67); pseudoephedrine ineffective in children; topical oxymetazoline seemed no more effective than placebo (27/42 (64%) vs 29/41 (71%))","excerpt":"\"20% of adult and 40% of child passengers had negative pressure in the middle ear after flight, and that 10% of adults and 22% of children had otoscopic evidence of changes to the ear drum. ... 14/41 (34%) with oral pseudoephedrine ... 29/41 (71%) with placebo (RR 0.48, 95% CI 0.29 to 0.67). ... oxymetazoline nasal spray seems no more effective than placebo ... (27/42 (64%) with oxymetazoline nasal spray vs 29/41 (71%) with placebo).\"\n","source_date":"2015-01-19","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260211132445/https://pmc.ncbi.nlm.nih.gov/articles/PMC4298289/","calculation_notes":"Wright's BMJ Clinical Evidence review (PMC4298289) is the single densest public source of per-flight prevalence numbers. The 10-20 percent adult rate of objective tympanic-membrane changes per flight is the baseline used in the Native field. The pseudoephedrine RCT numbers are used twice: the 71 percent placebo rate in adults prone to ear pain is the upper end of the \"any symptom\" per-flight probability for predisposed flyers, which brackets the URI-flight regime; the 34 percent pseudoephedrine rate sets the \"decongestant pre-flight\" multiplier in personal_factor_multipliers (roughly 0.5x the symptomatic rate, not zero). The pediatric ineffectiveness of pseudoephedrine and the 40 percent pediatric post-flight negative-pressure rate anchor the \"children with URI\" row in regional_breakdown.\n","independence_note":"Wright's review cites the Csortan/Jones 1994 Annals of Emergency Medicine RCT and Stangerup's auto-inflation trials, overlapping with Mirza & Richardson on primary sources. The two authoritative reviews are not fully independent on inputs but reach convergent numeric ranges, which raises confidence in the order-of-magnitude estimates used here.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/18856186/","title":"Upper respiratory infections and barotraumas in commercial pilots: a retrospective survey","publisher":"Aviation, Space, and Environmental Medicine (Rosenkvist L, Klokker M, Katholm M)","source_type":"peer_reviewed","statistic":"948 Danish commercial pilots surveyed; 37.6% reported one or more ear-barotrauma episodes in their career; 90% of ear-barotrauma events occurred during descent; 19.5% reported sinus barotrauma; 42.8% continued flying despite URI symptoms; 78.0% of those who flew used decongestants; 57.2% self-assessed as unfit to fly with URI","excerpt":"\"Ear barotrauma was reported by 37.6% of the pilots. Ninety percent of the ear barotraumas were reported during descent. ... 42.8% continued flying duties despite URI symptoms; 78.0% of those who flew used decongestants.\"\n","source_date":"2008-10-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260426201144/https://pubmed.ncbi.nlm.nih.gov/18856186/","calculation_notes":"Rosenkvist et al. is the single study that directly couples URI prevalence with in-career barotrauma prevalence in a well-defined flying population. One third of Denmark's commercial pilots — a population pre-screened for normal Eustachian-tube function — reported at least one ear-barotrauma episode across a career of thousands of flights, with 90 percent of those episodes on descent. This anchors two things: (1) descent as the dominant phase of risk, used throughout the regional_breakdown and caveats; (2) the \"URI flight is categorically different from a healthy flight\" framing, since nearly half the cohort admitted flying with URI symptoms and the in-career barotrauma rate is orders of magnitude above the per-flight baseline. The pilot career rate is not directly convertible to a per-URI-flight probability without knowing the URI-flight fraction of the career, but it corroborates the order of magnitude of the headline.\n","independence_note":"Rosenkvist is fully independent of Mirza & Richardson and of Wright — different country, different population (active airline pilots vs passengers), different data source (retrospective self-report vs post-flight otoscopy). The three authoritative sources on this page converge from three methodological directions.\n"},{"url":"https://www.merckmanuals.com/professional/ear,-nose,-and-throat-disorders/middle-ear-and-tympanic-membrane-disorders/otic-barotrauma","title":"Otic Barotrauma","publisher":"Merck Manual Professional Edition (Jan TA, reviewed by Lustig LR)","source_type":"reputable_reference","statistic":"A person with nasal congestion due to an upper respiratory infection or allergies should avoid flying; topical nasal vasoconstrictor (phenylephrine, oxymetazoline) applied 30-60 minutes before descent is the standard prophylaxis when flight is unavoidable","excerpt":"\"When an upper respiratory infection, allergy, or other mechanism interferes with eustachian tube function during changes in environmental pressure, the pressure in the middle ear either falls below ambient pressure. ... A person with nasal congestion due to an upper respiratory infection or allergies should avoid flying and diving. ... When these activities are unavoidable, a topical nasal vasoconstrictor (eg, phenylephrine, oxymetazoline) is applied 30 to 60 minutes before descent and ascent.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20240717174143/https://www.merckmanuals.com/professional/ear,-nose,-and-throat-disorders/middle-ear-and-tympanic-membrane-disorders/otic-barotrauma","calculation_notes":"The Merck Manual Professional entry supplies the standard-of-care clinical frame: URI is a recognised contraindication to flying; descent is the dangerous phase; topical decongestant pre-descent is the evidence-based mitigation. Used as a reputable clinical reference anchoring the mechanism and the mitigation language rather than a primary quantitative source, since Merck does not publish a headline incidence figure.\n","independence_note":"Merck Manual editorial reviews are independent of the three primary-literature sources above and draw on a wider clinical evidence base, but the synthesis points at the same RCTs and reviews, so treat as corroborating rather than as an independent measurement.\n"}],"comparison_anchors":[{"label":"Blood clot (VTE) per long-haul flight","lifetime_us_adult":0.000215},{"label":"Dengue per 2-week trip to endemic area","lifetime_us_adult":0.005},{"label":"Serious ski injury per 20-day season","lifetime_us_adult":0.0392},{"label":"Death in a commercial plane crash (per flight)","lifetime_us_adult":7.3e-8}],"regional_breakdown":[{"region":"Per flight, adult with active URI (any ear symptoms)","probability":0.6,"notes":"Upper end of the range — closer to the 71 percent placebo-arm figure from the pseudoephedrine RCTs for adults with a history of recurrent ear pain. Captures the 'ear hurt enough to notice' outcome, not the 'clinically significant barotrauma' outcome."},{"region":"Per flight, adult with active URI (moderate or worse barotrauma)","probability":0.34,"notes":"Headline number. Triangulated from placebo-arm symptom rates in predisposed flyers, objective otoscopic-change rates after flight, and the descent-dominant Rosenkvist career data. Moderate-or-worse means pain lasting beyond landing, temporary conductive hearing loss, or a Teed grade ≥2 otoscopic finding."},{"region":"Per flight, general-population adult (no URI)","probability":0.15,"notes":"Baseline passenger rate for any ear discomfort. Objective tympanic-membrane changes on post-flight otoscopy sit closer to 10 percent in unselected adults; this row conflates subjective pain and objective findings at the upper end of that range."},{"region":"Per flight, adult with URI on descent specifically","probability":0.3,"notes":"Descent is where essentially all clinically significant barotrauma originates — the Rosenkvist pilot survey attributed 90 percent of career ear-barotrauma events to descent. Pressure changes during ascent are dominated by active Eustachian-tube venting, which a congested tube can still manage; descent requires the tube to open against a pressure differential, which it often cannot."},{"region":"Per flight, child with URI","probability":0.55,"notes":"Upper end of the 26-55 percent pediatric ear-pain rate cited in post-flight otoscopy studies, adjusted upward for active URI. Smaller Eustachian tubes plus higher URI prevalence compound; pseudoephedrine has not been shown to help in children."},{"region":"Per flight, adult with URI taking oral pseudoephedrine 30+ minutes pre-flight","probability":0.17,"notes":"The pseudoephedrine RCT midpoint: roughly half the placebo rate in predisposed adults (34% active vs 71% placebo in the Wright BMJ review; 32% active vs 62% placebo in Csortan/Jones 1994). Topical oxymetazoline does not show the same benefit — the oxymetazoline arm in Wright's review ran 64% versus 71% placebo, a non-significant reduction. Oral pseudoephedrine reduces the rate meaningfully but does not eliminate it."},{"region":"Per flight, adult with URI — tympanic-membrane perforation","probability":0.0005,"notes":"Order-of-magnitude estimate. Wright's BMJ review states perforation rates are 'not reported' and 'extremely rare' in commercial aviation; aviation-medicine case series suggest roughly 1 per 1,000 to 1 per 10,000 URI-flights, with the higher end applying to severe URI plus aggressive descent (combat aviation, rapid depressurisation). TM perforation is a small fraction of the clinically significant-barotrauma denominator."}],"personal_factor_multipliers":[{"factor":"active URI with nasal congestion (headline assumption)","multiplier":1,"notes":"The headline number already assumes active URI. Listed here so the baseline-healthy multiplier below is comprehensible."},{"factor":"no URI, normal Eustachian-tube function","multiplier":0.3,"notes":"Baseline passenger per-flight barotrauma-symptom rate is roughly one third of the URI rate — 10-20 percent vs 30-60 percent. The Eustachian tube opens under its own mucosal control when dry; a congested tube cannot."},{"factor":"child under 7 with URI","multiplier":1.6,"notes":"Children have shorter, more horizontal Eustachian tubes and higher URI prevalence; post-flight otoscopy studies find roughly 2x the adult rate of tympanic-membrane changes. The multiplier is applied against the URI-adult headline."},{"factor":"prior tympanic-membrane perforation history","multiplier":2,"notes":"A previously perforated and healed tympanic membrane is a thinner barrier and a recurrence risk. Rough clinical estimate; no clean RCT data."},{"factor":"oral pseudoephedrine 30-60 min pre-flight (adults)","multiplier":0.5,"notes":"Two RCTs converge on roughly halving the symptomatic-barotrauma rate in adults with a history of recurrent ear pain. Ineffective in children. Also has cardiovascular and sleep side effects."},{"factor":"topical oxymetazoline before descent","multiplier":0.9,"notes":"Often recommended in aviation-medicine guidance because it avoids systemic decongestant side effects, but the trial evidence is weak: Wright's BMJ review reports the oxymetazoline nasal-spray arm at 64% (27/42) versus 71% (29/41) with placebo — a small, statistically non-significant reduction. Unlike oral pseudoephedrine, topical oxymetazoline seems little more effective than placebo at reducing barotrauma symptoms, so this near-1 multiplier reflects a marginal, unproven effect rather than the halving seen with pseudoephedrine."},{"factor":"SCUBA diving on same trip + URI-flight","multiplier":3,"notes":"SCUBA-trained ears often have a history of subclinical round-window weakness; combined diving-then-flying while congested carries meaningfully elevated risk of inner-ear barotrauma. Order-of-magnitude clinical estimate."}],"short_label":"Flying with cold","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"The headline \"roughly 1 in 3 per flight\" applies to adult commercial passengers with active symptomatic upper-respiratory infection and moderate-or-worse otic barotrauma — pain that outlasts the flight, temporary conductive hearing loss, or Teed grade ≥2 otoscopic findings. It is not a rate of tympanic-membrane perforation, which is roughly two to three orders of magnitude rarer (~1 in 1,000 to ~1 in 10,000 URI-flights) and is captured in the regional_breakdown. The figure also assumes a standard commercial descent profile in a pressurised Part 121 airliner; rapid descents, unpressurised general aviation, and combat or EMS aviation are a different regime. Descent is the dominant phase of risk (roughly 90 percent of Rosenkvist's pilot cohort's barotrauma events), so \"the flight\" is really \"the last twenty minutes of the flight\". Oral pseudoephedrine before descent cuts the symptomatic rate by roughly half in adults in randomised trials but does not eliminate risk and has not been shown to help children; topical oxymetazoline, though commonly recommended, ran only 64 percent versus 71 percent placebo in Wright's review — no significant benefit. Finally, the underlying literature is heterogeneous on outcome definition: some studies count any negative middle-ear pressure on post-flight tympanometry, others count only patient-reported pain, and the three trials of decongestant prophylaxis use three different symptom scales. The 15-60 percent uncertainty range reflects that mixture honestly rather than hiding it.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-9-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single stylized airplane window in muted grey-blue tones against a pale sky, flat vector illustration, empty cabin."},"canonical_url":"https://likelier.app/flying-with-uri","api_url":"https://likelier.app/api/fears/flying-with-uri.json"},{"slug":"job-loss-depression","question":"What are the odds of developing depression after losing your job?","quick_answer":"Among adults who lose their jobs, about 34% show clinical-level psychological problems including depression — roughly double the 16% baseline among the employed. Most people brace for temporary stress and the financial hit and underweight the mental-health cost of a layoff.\n","category":"health","tags":["workplace","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"People fear job loss primarily for its financial consequences: missed mortgage payments, depleted savings, downward mobility. The mental health cost rarely features in the worry. When it does, most adults frame it as temporary stress rather than clinical illness, expecting the low mood to lift once re-employment arrives. Surveys on workplace anxiety focus almost entirely on the economic dimension; questions about depression as a downstream consequence of displacement are uncommon in public polling. The result is a risk that is structurally underweighted: the probability of developing clinical depression after involuntary job loss is roughly double the employed baseline, yet it seldom appears on anyone's list of things to fear about a layoff.\n","rough_estimate":"Most people expect temporary stress, not clinical depression","kind":"intuition"},"native":{"display":"~34% prevalence of clinical-level psychological problems among unemployed vs ~16% employed","numerator":34,"denominator":100,"unit":"prevalence among unemployed","population":"unemployed adults (pooled across 237 cross-sectional studies, predominantly OECD countries)"},"normalized":{"lifetime_us_adult":0.34,"display":"~34% of unemployed adults show clinical-level psychological problems (incl. depression) — roughly double the ~16% employed baseline","log_value":-0.47,"assumptions":"This entry is scoped to the subgroup that experiences unemployment, not to the full US-adult lifetime. Paul & Moser (2009), pooling 237 cross-sectional studies, found that 34% of unemployed persons met the threshold for clinical-level psychological problems, versus 16% of employed persons — roughly a doubling of prevalence. That 34% is the source's central pooled figure and is used directly as the normalized value; the 16% employed rate is the comparison baseline. Two caveats govern the framing. First, the 34% is the prevalence of \"psychological problems\" as a composite (mixed distress, depression, anxiety, psychosomatic symptoms, low subjective well-being, and low self-esteem), not depression alone, so it overstates the depression-specific rate; the display is worded accordingly (\"incl. depression\"). Second, it is total prevalence among the unemployed, which includes pre-existing cases as well as new-onset ones, so it is not the incremental risk attributable to job loss alone. The prior version multiplied this 34% by an unsourced ~0.80 \"career job-loss probability\" to manufacture a ~27% lifetime figure; that 0.80 both contradicted the entry's own BLS-derived career-displacement estimate of ~0.41 (1 - (1 - 0.013)^40 from the 2021-23 displaced-worker rate) and layered a semantic mismatch (a composite point-prevalence presented as lifetime depression incidence). No cited source states a lifetime probability of \"depression triggered by job loss,\" so that compound has been removed and the value regrounded to the conditional prevalence the source actually reports.\n","uncertainty":{"low":0.16,"high":0.4},"scope":"subgroup_lifetime"},"sources":[{"url":"https://psycnet.apa.org/record/2009-06258-005","title":"Unemployment impairs mental health: Meta-analyses","publisher":"Journal of Vocational Behavior (Paul & Moser)","source_type":"peer_reviewed","statistic":"34% prevalence of clinical-level psychological problems among unemployed vs 16% among employed; mean effect size d = 0.51","excerpt":"\"The average overall effect size was d = 0.51 with unemployed persons showing more distress than employed persons. A significant difference was found for several indicator variables of mental health (mixed symptoms of distress, depression, anxiety, psychosomatic symptoms, subjective well-being, and self-esteem). The average number of persons with psychological problems among the unemployed was 34%, compared to 16% among employed individuals.\"\n","source_date":"2009-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503091812/https://psycnet.apa.org/record/2009-06258-005","calculation_notes":"Paul & Moser (2009) conducted the largest meta-analysis to date on unemployment and mental health, covering 237 cross-sectional and 87 longitudinal studies. The 34% vs 16% prevalence figures are the central estimates for the native rate. The effect size d = 0.51 represents a medium effect. Moderator analyses showed that distress peaks around month 9 of unemployment (d = 0.73), with stabilization at medium levels during the second year. Men and blue-collar workers showed larger effects than women and white-collar workers. The 87 longitudinal studies confirmed the causal direction: unemployment causes mental health deterioration, not merely the reverse.\n","independence_note":"This is the foundational meta-analysis in the field. Independent from SAMHSA administrative data and from the Milner et al. suicide meta-analyses, which use different outcome measures and study pools.\n"},{"url":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0051333","title":"Long-Term Unemployment and Suicide: A Systematic Review and Meta-Analysis","publisher":"PLOS ONE (Milner, Page & LaMontagne)","source_type":"peer_reviewed","statistic":"Pooled RR of suicide after unemployment = 1.70 (95% CI 1.22-2.18); within 5 years RR = 2.50 (95% CI 1.83-3.17)","excerpt":"\"A random effects meta-analysis on a subsample of six cohort studies indicated that the pooled relative risk of suicide in relation to average follow-up time after unemployment was 1.70 (95% CI 1.22 to 2.18). The greatest risk of suicide occurred within five years of unemployment compared to the employed population (RR = 2.50, 95% CI 1.83 to 3.17).\"\n","source_date":"2013-01-09","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420042555/https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0051333","calculation_notes":"Milner et al. (2013) established the suicide risk gradient by unemployment duration. The RR of 2.50 in the first five years is consistent with the Paul & Moser finding that depression peaks during the first year and remains elevated. This source is used here not for the normalized probability (which is about depression, not suicide) but to corroborate the severity of mental health consequences and to anchor the outcome_severity classification. Suicide is the extreme end of the depression spectrum that job loss can trigger.\n","independence_note":"Uses different outcome (suicide mortality) and different study pool from Paul & Moser. Methodologically independent.\n"},{"url":"https://www.bls.gov/news.release/disp.nr0.htm","title":"Displaced Workers Summary, January 2024","publisher":"U.S. Bureau of Labor Statistics","source_type":"govt_report","statistic":"6.3 million workers displaced in the 2021-2023 period; layoff/discharge rate ~1.0-1.1% of nonfarm employment per month","excerpt":"\"From January 2021 to December 2023, 6.3 million workers were displaced from jobs they had held for at least 3 years or from jobs held for less than 3 years. In January 2024, 65.7 percent of the 2.6 million long-tenured displaced workers were reemployed.\"\n","source_date":"2024-08-29","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420042631/https://www.bls.gov/news.release/disp.nr0.htm","calculation_notes":"The BLS Displaced Workers Survey provides the denominator for estimating lifetime job-loss probability. With approximately 6.3 million displaced workers over a 3-year period in a labor force of ~160 million, the annual displacement rate is roughly 1.3%. JOLTS data show a monthly layoff/discharge rate of ~1.0-1.1% (including short-tenure workers), or roughly 12-13% per year including all separations classified as involuntary. Over a 40-year career, using the conservative displaced-worker definition (1.3%/year), the probability of at least one displacement is 1 - (1 - 0.013)^40 ≈ 0.41. Using the broader JOLTS layoff/discharge rate (~12%/year), the figure approaches certainty, but many of those separations are brief and may not trigger the sustained unemployment that drives depression. We use ~0.80 as a central estimate for at least one significant involuntary job loss over a career, reflecting the reality that most American workers will experience this at least once.\n","independence_note":"BLS administrative survey data, independent from the clinical studies on depression prevalence. Different data collection pipeline (employer establishment survey and household survey) from the mental health literature.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4603822/","title":"Depressive symptoms as a cause and effect of job loss in men and women: evidence in the context of organisational downsizing","publisher":"BMC Public Health (Magnusson Hanson et al.)","source_type":"peer_reviewed","statistic":"Displaced workers had a threefold risk of incident major depression; men showed nearly fivefold risk after layoff","excerpt":"\"In the total sample including men and women, displaced workers experienced a more than threefold risk of incident major depression and twofold risk of less severe symptoms among those with no depression at baseline. A nearly fivefold risk of incident major depression was observed in unemployed men with no depression at baseline.\"\n","source_date":"2015-10-06","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250405161025/https://pmc.ncbi.nlm.nih.gov/articles/PMC4603822/","calculation_notes":"Magnusson Hanson et al. (2015) used the Swedish Longitudinal Occupational Survey of Health (SLOSH) to examine depression as both cause and effect of job loss during organisational downsizing. The threefold risk of incident major depression among displaced workers (and fivefold among men) is higher than the ~2x implied by the Paul & Moser prevalence ratio (34%/16%), likely because the SLOSH study isolated incident cases (new-onset depression in previously non-depressed workers) rather than point prevalence. This suggests the 34% prevalence figure includes both pre-existing and new-onset depression, and the true causal effect of job loss may be larger than the cross-sectional data indicate.\n","independence_note":"Swedish longitudinal cohort study using different population, design (prospective longitudinal vs cross-sectional meta-analysis), and outcome measure (incident major depression vs prevalence) from Paul & Moser. Methodologically independent.\n"}],"comparison_anchors":[{"label":"Depression (lifetime, US adult, general population)","lifetime_us_adult":0.206},{"label":"Divorce (lifetime, US first marriage)","lifetime_us_adult":0.42},{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1}],"personal_factor_multipliers":[{"factor":"Unemployment >6 months","multiplier":1.8,"notes":"Paul & Moser found distress peaks at ~9 months (d = 0.73 vs overall d = 0.51); prolonged unemployment roughly doubles the conditional depression risk compared to brief episodes"},{"factor":"Prior history of depression","multiplier":2,"notes":"Pre-existing vulnerability approximately doubles risk of recurrence after job loss; bidirectional causality documented in longitudinal studies"},{"factor":"Strong social and financial safety net","multiplier":0.5,"notes":"Social support, severance, and savings buffer the mental health impact; Paul & Moser found country-level social protection moderates the effect"},{"factor":"Voluntary career change vs involuntary layoff","multiplier":0.3,"notes":"The depression literature consistently distinguishes involuntary displacement from voluntary separation; the mental health impact is concentrated in involuntary job loss"}],"short_label":"Job loss & depression","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"The 34% prevalence figure from Paul & Moser is a pooled estimate across 237 studies spanning several decades and many countries. It includes both pre-existing depression and new-onset cases triggered by unemployment, so the causal attributable fraction is smaller than 34%. The normalized lifetime estimate depends heavily on the assumed probability of involuntary job loss over a career, which varies enormously by occupation, industry, and economic conditions. The depression risk is strongly moderated by unemployment duration: brief episodes (under 3 months) carry much lower risk than prolonged unemployment. Country-level social protection also matters; the Paul & Moser meta-analysis found smaller effects in countries with generous unemployment benefits. The ~34% figure is the prevalence of any clinical-level psychological problem (a composite that includes depression, anxiety, and psychosomatic symptoms), not depression alone and not the increment attributable to job loss; many more people experience subclinical distress that does not meet diagnostic thresholds. No cited source supports a compound \"lifetime probability of depression triggered by job loss,\" so this entry reports the conditional prevalence during unemployment rather than a synthesised lifetime number. Gender differences are notable: men show larger mental health effects from unemployment than women in most studies, possibly due to stronger identity investment in employment.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty office desk with a cardboard box and a wilting plant, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/job-loss-depression","api_url":"https://likelier.app/api/fears/job-loss-depression.json"},{"slug":"ai-voice-clone-scam-target","question":"What are the odds you'll be targeted by an AI voice-clone scam in your lifetime?","quick_answer":"About 35% over a lifetime, with roughly 10% of adults already personally targeted as of 2023. Most people guess 1 to 3%, anchoring on dramatic news stories. Being targeted, which is not the same as being fooled, is far more routine than that.\n","category":"tech","tags":["digital-fraud","elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Public coverage of AI voice cloning has been dominated by single dramatic incidents — the Arizona mother who heard her daughter's cloned voice begging for ransom in 2023, the $25 million Hong Kong deepfake video call in 2024, the impostor impersonating Secretary of State Marco Rubio in 2025 — which leaves the typical reader with two competing intuitions: that this is everywhere and that it is exotic. When asked to estimate how many adults have already received a voice-clone scam attempt directed at them personally, guesses cluster around 1–3% — roughly the rate of dramatic news stories per capita. The measured figure from McAfee's 2023 nine-country survey is about 10% of adults who said they had personally received an AI voice-clone message at the time of the survey, with one in four saying they had experienced such a scam or knew someone who had. The actual ceiling, integrated over a full adult lifetime in 2026 onward, is meaningfully higher than the snapshot suggests.\n","rough_estimate":"~1–3% lifetime feels right to most respondents","kind":"intuition"},"native":{"display":"~10 in 100 adults personally targeted by an AI voice-clone scam (point prevalence, McAfee 2023, 9 countries)","numerator":10,"denominator":100,"unit":"personally targeted to date","population":"adults in McAfee's 9-country sample (US, UK, France, Germany, Australia, India, Japan, Brazil, Mexico)"},"normalized":{"lifetime_us_adult":0.35,"display":"~35% lifetime probability of being targeted by an AI voice-clone scam (US adult)","log_value":-0.46,"assumptions":"McAfee's early-2023 survey (N=7,000 across 9 countries, MSI-ACI panel, fielded January 27 to February 1, 2023) found that one in ten respondents had personally received a message from an AI voice clone, and one in four said they had experienced such a scam or knew someone who had. That 10% is a point-prevalence snapshot covering the first ~18 months of widely available consumer voice-cloning tools (ElevenLabs launched in beta January 2023; open-source models followed within months). The lifetime figure extrapolates from that snapshot under three forces: (1) the underlying tool capability is maturing rapidly — the FBI's December 2024 IC3 PSA confirmed that voice cloning is now a standard tool in scam campaigns alongside text and image generation; (2) impersonation scams are already the FTC's third-largest complaint category, with Americans losing close to $3 billion to imposter scams in 2024; (3) the per-year targeting hazard is bounded below by the McAfee 10% / ~1.5 years ≈ 6–7% per year and above by FTC imposter-scam complaint rates (~1% per year reporting an actual loss, with many more contacted but not defrauded). A central estimate of 35% lifetime sits between the naive compound of a 1% annual targeting hazard over 59 years (~45%) and the floor from McAfee's already-observed 10% prevalence in 2023. The uncertainty band (0.18–0.55) reflects (a) wide variation in what counts as \"targeted\" (a fully personalised cloned-voice call versus a generic AI-voiced robocall), (b) demographic targeting concentrated among older adults and people with public audio footprints (podcasters, executives), and (c) the rapid technology trajectory making point estimates inherently fragile beyond a 5-year horizon.\n","uncertainty":{"low":0.18,"high":0.55},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.mcafee.com/ai/news/ai-voice-scam/","title":"Scammers Use AI Voice Cloning Tools to Fuel New Scams","publisher":"McAfee","source_type":"reputable_reference","statistic":"One in four adults said they experienced an AI voice scam or knew someone who had; one in ten said they personally received a message from an AI voice clone; 77% of those victims lost money; 70% of adults not confident they could distinguish a cloned voice from the real one","excerpt":"\"Our recent global study found that out of 7,000 people surveyed, one in four said that they experienced an AI voice cloning scam or knew someone who had. [...] One in ten of people surveyed in our study said they received a message from an AI voice clone, and 77% of those victims said they lost money as a result. [...] 70% of people in our worldwide survey said they weren't confident they could tell the difference between a cloned voice and the real thing.\"\n","source_date":"2023-05-02","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260531014325/https://www.mcafee.com/ai/news/ai-voice-scam/","calculation_notes":"The 10% personally-targeted figure is the spine of the native estimate (10 in 100). It is a global average across 9 countries (US, UK, France, Germany, Australia, India, Japan, Brazil, Mexico). The US-only subsample is not separately reported by McAfee; US adults are assumed to face at least the global rate given that English-language voice cloning is the most mature subset of the technology and that US adults receive the highest volume of impersonation-scam contact attempts per FTC Sentinel data. The 77% loss rate among targeted individuals is conditional on responding to the scam, not on receiving the call; the unconditional \"targeted AND lost money\" rate is therefore lower than 77% of 10%, since not every targeted adult engages with the scammer.\n","independence_note":"Commissioned online survey conducted by Market Research Company MSI-ACI between January 27 and February 1, 2023. Methodologically independent from the FTC and FBI complaint databases.\n"},{"url":"https://www.ic3.gov/PSA/2024/PSA241203","title":"Criminals Use Generative Artificial Intelligence to Facilitate Financial Fraud","publisher":"FBI Internet Crime Complaint Center (IC3)","source_type":"govt_report","statistic":"FBI public service announcement identifying voice cloning as one of five primary generative-AI fraud vectors in active use; no aggregate incident count provided","excerpt":"\"Generative AI reduces the time and effort criminals must expend to deceive their targets. [...] Criminals generate short audio clips containing a loved one's voice to impersonate a close relative in a crisis situation, asking for immediate financial assistance or demanding a ransom.\"\n","source_date":"2024-12-03","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260515181110/https://www.ic3.gov/PSA/2024/PSA241203","calculation_notes":"The FBI PSA confirms that voice cloning has moved from research curiosity to standard criminal tool by late 2024, supporting the assumption that the per-year targeting hazard is non-trivial and rising rather than rare and stable. The PSA does not publish counts of voice-clone-specific complaints because IC3 categorises by scam type (grandparent scam, tech-support scam, romance scam), not by impersonation technique. This means voice-clone incidents are scattered across multiple categories and almost certainly undercounted in any single line of IC3 reporting.\n","independence_note":"Federal law enforcement advisory. Independent from McAfee's commercial survey and from FTC consumer-complaint data.\n"},{"url":"https://advocacy.consumerreports.org/press_release/more-than-75000-consumers-urge-ftc-to-crack-down-on-ai-voice-cloning-fraud/","title":"More Than 75,000 Consumers Urge FTC to Crack Down on AI Voice Cloning Fraud","publisher":"Consumer Reports","source_type":"reputable_reference","statistic":"Americans lost nearly $3 billion in imposter scams during 2024; six commercially available AI voice-cloning tools tested lacked meaningful safeguards against misuse","excerpt":"\"Americans lost nearly $3 billion in imposter scams during 2024 alone. [...] AI voice cloning tools are making it easier than ever for scammers to impersonate someone's voice. [...] Consumer Reports assessed six freely or cheaply available AI voice cloning products and found that most did not have meaningful safeguards to stop fraud or misuse.\"\n","source_date":"2025-08-13","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260210220448/https://advocacy.consumerreports.org/press_release/more-than-75000-consumers-urge-ftc-to-crack-down-on-ai-voice-cloning-fraud","calculation_notes":"The $3 billion / 2024 imposter-scam total is the FTC Consumer Sentinel envelope inside which voice-clone scams sit as one technique. Voice-clone-specific losses are not separately published, but the magnitude and growth of the parent category establish that the targeting attempt rate must be much higher than the loss rate (most contacted adults do not engage or do not pay). This source also documents the supply side — that the underlying tools remain cheap and accessible — which is the mechanism behind the rising-hazard assumption in the normalized calculation.\n","independence_note":"Consumer Reports advocacy filing to the FTC, citing FTC Sentinel data and Consumer Reports' own product testing. Independent from McAfee survey and FBI advisory.\n"},{"url":"https://www.aarp.org/pri/topics/work-finances-retirement/fraud-consumer-protection/ai-fraud-concerns-older-adults/","title":"Older Adults Express High Concern and Limited Knowledge About AI Scams and Fraud","publisher":"AARP Research","source_type":"reputable_reference","statistic":"84% of US adults 50+ are concerned about criminals using AI for voice cloning; 77% worry about becoming personal targets of AI-related fraud (N=1,000, Foresight 50+ Omnibus panel, fielded August 15–19, 2024)","excerpt":"\"Older adults are worried about criminals using AI for a variety of fraudulent activities. [...] Voice cloning: 84 percent [are concerned]. [...] A significant majority of older adults expressed worry that they might personally become targets of an AI-related fraud in the future (77 percent).\"\n","source_date":"2024-10-01","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260120195848/https://www.aarp.org/pri/topics/work-finances-retirement/fraud-consumer-protection/ai-fraud-concerns-older-adults/","calculation_notes":"Used to characterise the older-adult subgroup that is disproportionately targeted and contributes to the personal_factor_multipliers. The 84% concern figure tracks worry, not incidence, and is used only on the perception side and as evidence that voice-clone scams have moved into mainstream awareness. AARP did not publish an incidence figure for personal voice-clone targeting in this wave, so the calculation remains anchored on McAfee.\n","independence_note":"Nationally representative US online/phone panel of adults 50+. Methodologically independent from McAfee global commercial survey, FBI complaint database, and FTC Sentinel.\n"},{"url":"https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2025/08/false-alarm-real-scam-how-scammers-are-stealing-older-adults-life-savings","title":"False alarm, real scam: how scammers are stealing older adults' life savings","publisher":"Federal Trade Commission (Consumer Protection Data Spotlight)","source_type":"govt_report","statistic":"From 2020 to 2024 the number of reports from adults 60+ who lost $10,000 or more to government/business impersonation scams increased more than fourfold (1,790 in 2020 to 8,269 in 2024); reports of $100,000+ losses rose nearly sevenfold and combined losses eightfold ($55M in 2020 to $445M in 2024)","excerpt":"\"From 2020 to 2024, the number of reports from older adults who lost $10,000 or more to these scams increased more than fourfold. When older adults reported losing more than $100,000, the trend was even more striking: during the same period, the number of reports increased nearly sevenfold, and the combined reported losses went up eightfold. [...] The total number of business imposter and government imposter reports filed by older adults with a loss of $10K or more are as follows: 1,790 (2020), 3,516 (2021), 5,559 (2022), 7,091 (2023), 8,269 (2024).\"\n","source_date":"2025-08-13","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260505223725/https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2025/08/false-alarm-real-scam-how-scammers-are-stealing-older-adults-life-savings","calculation_notes":"This spotlight is cited as a post-2025 corroboration that high-value impersonation fraud against older adults is rising steeply, supporting the entry's caveat that the 2023 McAfee snapshot understates the current era. IMPORTANT SCOPE LIMIT: the FTC data here measure GOVERNMENT and BUSINESS impersonation scams (fake bank/Amazon/Microsoft/ Apple/SSA/FTC alerts), not relative-impersonation voice-clone scams, and the page does not mention AI or voice cloning. It therefore does NOT supply a voice-clone-specific incidence rate or a multiplier; it is used only to (a) refresh the older-adult harm picture and (b) document that the impersonation-fraud envelope continues to grow, which is why the headline 0.35 is left unchanged and no new multiplier is derived from it. The 41% phone-as-initial-contact figure (footnote 4) confirms these scams still depend on a live call, the same delivery channel voice cloning exploits.\n","independence_note":"Federal law enforcement / consumer-protection agency analysis of FTC Consumer Sentinel reports. Independent from McAfee's commercial survey and methodologically distinct from the 2024 IC3 advisory.\n"},{"url":"https://www.cnn.com/2024/05/16/tech/arup-deepfake-scam-loss-hong-kong-intl-hnk","title":"Arup revealed as victim of $25 million deepfake scam involving Hong Kong employee","publisher":"CNN Business","source_type":"news_article","statistic":"A finance worker at engineering firm Arup was deceived into making 15 transfers totaling HK$200 million (~US$25.6 million) after a live deepfake video call in which fraudsters impersonated the company's UK-based CFO and other colleagues","excerpt":"\"A finance worker at a multinational firm was tricked into paying out $25 million to fraudsters using deepfake technology to pose as the company's chief financial officer in a video conference call [...]. The worker had grown suspicious after receiving a message that was purportedly from the company's UK-based chief financial officer. [...] the worker had been convinced by the appearance of the CFO and others in attendance who had looked and sounded like colleagues he recognized.\"\n","source_date":"2024-05-16","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260510183515/https://www.cnn.com/2024/05/16/tech/arup-deepfake-scam-loss-hong-kong-intl-hnk","calculation_notes":"Cited only to characterise an ADJACENT vector — live multi-participant deepfake VIDEO-call fraud against an enterprise finance employee — which is distinct from the personal audio voice-clone scope this entry measures. It does not feed the 0.35 lifetime number (an enterprise, employee-targeted, single-incident event with no published per-person rate) and is referenced in the caveats as evidence that the synthetic-impersonation threat now spans video and group calls, not audio alone.\n","independence_note":"Mainstream news report of a Hong Kong Police case publicly confirmed by Arup. Independent from McAfee, FBI/IC3, and FTC sources; corroborated by multiple outlets.\n"}],"comparison_anchors":[{"label":"Online scam loss (lifetime, US adult)","lifetime_us_adult":0.4},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Investment scam loss — pig-butchering era (lifetime, US adult)","lifetime_us_adult":0.02}],"personal_factor_multipliers":[{"factor":"Age 65+","multiplier":1.5,"notes":"Older adults are the most frequently impersonated relatives in grandparent-style voice-clone scams and the demographic with the highest median loss per imposter incident per FTC Sentinel 2024"},{"factor":"Public audio footprint (podcaster, executive, public speaker)","multiplier":3,"notes":"Voice cloning currently requires only 3 seconds of clean audio per McAfee 2023 and ElevenLabs documentation; people with hours of public recordings are first-target candidates and were the basis of the 2024 Hong Kong deepfake CEO case"},{"factor":"Family member with public audio footprint","multiplier":2,"notes":"Targeting is often indirect — the scammer clones a relative's voice to call the wealthier or more trusting family member. AARP 2024 documents elder relatives as the most common payment-side targets even when the cloned voice belongs to a younger family member"},{"factor":"No landline / unlisted mobile / no social media","multiplier":0.4,"notes":"Voice-clone scam delivery depends on the scammer obtaining both a contact number and a voice sample; adults with restricted social media presence and unpublished phone numbers have substantially lower targeting probability per general scam-exposure research"},{"factor":"Family safe-word established","multiplier":0.7,"notes":"A pre-arranged verbal challenge — recommended by the FBI's December 2024 PSA — does not reduce the targeting rate but reduces the conditional probability of a successful scam given targeting; included here because the headline framing is targeting probability, and a safe word changes downstream outcomes more than upstream contact"}],"short_label":"AI voice scam","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"\"Targeted\" is a deliberately broad construct: it covers receiving a voice-clone scam call or voicemail directed personally, whether or not the recipient engaged or lost money. The probability of being defrauded by such a scam is much lower than the probability of being targeted by one — McAfee's data implies roughly 77% of those who engaged with the scammer lost money, but most targeted adults hang up, ignore the voicemail, or call the supposed relative back through known channels. The single biggest definitional uncertainty is whether to count generic AI-voiced robocalls (which now make up a meaningful share of unsolicited call traffic) as \"voice-clone scams\" — this entry treats only personalised impersonation as a clone-scam attempt; the broader robocall category would push the lifetime figure substantially higher. The McAfee survey is the only large-sample incidence measure available and is now three years old in a fast-moving technology area; the FBI and FTC have confirmed that voice cloning is now a standard tool but have not published updated incidence counts. The 35% central estimate is therefore a model-based projection from a single 2023 snapshot and should be revised when comparable post-2025 surveys appear. Population heterogeneity is large: people with significant public audio footprints are at much higher risk than the population mean, and older adults face the highest median financial harm given a successful scam. A separate but related vector sits outside this entry's personal-audio scope: live, multi-participant deepfake VIDEO-call fraud against employees. In the 2024 Arup case, a Hong Kong finance worker made 15 transfers totalling about US$25.6 million after a video conference in which fraudsters impersonated the CFO and other colleagues (CNN, May 2024). That is an enterprise, employee-targeted incident with no published per-person rate, so it does not move the personal lifetime figure here — but it shows the synthetic-impersonation threat now spans video and group calls, not audio alone. On incidence trend: the FTC's August 2025 spotlight \"False alarm, real scam\" found that reports from adults 60+ losing $10,000 or more to impersonation scams rose more than fourfold from 2020 to 2024 (1,790 to 8,269 reports), with $100,000-plus losses up nearly sevenfold. This corroborates that high-value impersonation fraud against older adults is surging and supports treating the 2023 McAfee snapshot as a likely floor for the current era — but with an explicit caveat: that FTC data covers government and business impersonation (fake bank, Amazon, Microsoft, Apple, SSA, and even FTC alerts), not relative-impersonation voice cloning, and does not break out an AI-voice-specific rate. It refreshes the direction of travel, not the headline number, which remains anchored on McAfee pending a comparable post-2025 voice-clone incidence survey.\n","quality_score":{"d1":4,"d2":4,"d3":3,"d4":3,"d5":4,"d6":4,"d7":4,"d8":4,"avg":3.75,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single smartphone resting face-up on a wooden kitchen table, screen showing only a blank incoming call, no caller ID, no people in the frame."},"canonical_url":"https://likelier.app/ai-voice-clone-scam-target","api_url":"https://likelier.app/api/fears/ai-voice-clone-scam-target.json"},{"slug":"forced-retirement-post-50","question":"How likely is a worker over 50 to be pushed out of their job before they planned to retire?","quick_answer":"A worker over 50 has roughly a 1 in 3 chance of being pushed out of a job before their planned retirement, though estimates range widely, from 9% under a narrow European definition to 56% under a broad US one. It is common enough to be a real planning variable, not a rare event.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Workers past 50 often perceive the threat of involuntary job loss as real but hard to quantify. High-profile rounds of age-discriminatory layoffs generate media coverage, and many older workers have watched colleagues forced out. Yet few have a precise sense of how common the experience is across the full population. The received assumption is that it happens frequently — but without a concrete baseline, workers cannot reliably weigh the risk when making financial plans or assessing the security of their position. Research shows the actual rate is high enough to be a genuine planning variable.\n","kind":"intuition"},"native":{"display":"roughly 35 in 100 workers aged 50+ experience an involuntary employer-driven separation before their planned retirement (midpoint of a 9% narrow EU exit rate and a 56% broad US displacement rate)","numerator":35,"denominator":100,"unit":"lifetime (working age to retirement)","population":"workers aged 50+ (SHARE European panel narrow lower bound; US Health and Retirement Study broad upper bound)"},"normalized":{"lifetime_us_adult":0.35,"display":"roughly 1 in 3 workers over 50 are involuntarily pushed out before their planned retirement (range spans 9% narrow EU exit to 56% broad US displacement)","log_value":-0.46,"assumptions":"No single source reports an overall probability of involuntary separation before retirement for older workers, so the 0.35 point estimate is a midpoint between the two quantitative panel studies that bracket the construct. SHARE European panel (15 countries, 7 waves 2004–2017; Mäcken et al. 2022) finds 9% of older workers exit the labor market involuntarily under a narrow self-reported definition (mean age 58.8 vs 62.0 voluntary) — the lower bound (tight construct: worker-reported involuntary exit only). US HRS analysis (ProPublica/Urban Institute) finds 56% of full-time workers 50+ are laid off, forced out, or obligated to leave at least once before retirement under a broad definition — the upper bound. The two differ in construct breadth and legal context (US employer-at-will offers less institutional protection than most European labor markets), so 0.35 is presented as an order-of-magnitude midpoint, not a precise measured rate. OECD Employment Outlook 2025 corroborates the severity but reports no overall separation probability; it documents that displaced older workers face longer joblessness, lower reemployment, and steeper wage losses than younger workers. Low (0.09): SHARE narrow worker-reported involuntary exit (Mäcken et al. 2022). High (0.56): US HRS cohort, broad displacement definition (ProPublica/Urban Institute).\n","uncertainty":{"low":0.09,"high":0.56},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.oecd.org/en/publications/oecd-employment-outlook-2025_194a947b-en.html","title":"OECD Employment Outlook 2025","publisher":"Organisation for Economic Co-operation and Development","source_type":"reputable_reference","statistic":"Older workers who are displaced spend more time unemployed, are much less likely to be reemployed, and suffer larger wage losses than younger workers; the report gives no single involuntary-separation probability","excerpt":"\"When older workers do experience job displacement, they face much greater challenges compared to younger workers: they spend more time unemployed, are significantly less likely to become reemployed, and when they do find new work, they typically suffer much larger wage losses. [...] Older workers are less likely to become unemployed compared to their younger counterparts, but when they do lose their jobs, they face significantly higher risks of long-term unemployment. As a result, job loss in one's 50s or 60s can have lasting effects on labour market attachment, income security, and retirement pathways.\"\n","source_date":"2025-07-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260524015237/https://www.oecd.org/en/publications/oecd-employment-outlook-2025_194a947b-en.html","calculation_notes":"OECD Employment Outlook 2025, \"Navigating the golden years\" chapter. The report documents that older-worker displacement is typically involuntary and driven by structural economic shifts, and that displaced older workers face longer joblessness, lower reemployment, and steeper wage losses than younger workers. It does NOT report an overall probability of involuntary separation before retirement, so it is cited here for the severity and poor-recovery dimension of the risk, not as the source of the headline rate. The headline point estimate is derived from the two quantitative panel studies below (SHARE 9% narrow EU exit; US HRS 56% broad displacement).\n"},{"url":"https://www.cambridge.org/core/journals/journal-of-social-policy/article/educational-inequalities-in-labor-market-exit-of-older-workers-in-15-european-countries/ECD5682953EF50792B50A67D0EFF87BC","title":"Educational Inequalities in Labor Market Exit of Older Workers in 15 European Countries","publisher":"Mäcken, Präg, Hess & Ellwardt — Journal of Social Policy 51(2), 435–459 (Cambridge University Press); SHARE data","source_type":"peer_reviewed","statistic":"9% of older European workers exit the labor market involuntarily (mean age 58.8 vs 62.0 for voluntary exit); 15 SHARE countries, 7 waves 2004–2017","excerpt":"\"Most workers exited the labor market voluntarily and only 9% involuntarily, whereby the average age of involuntary work exit was 58.8 years and by that more than three years earlier than a voluntary work exit with 62.0 years.\"\n","source_date":"2022-04-01","source_accessed":"2026-06-30","archive_url":"https://web.archive.org/web/20260505050638/https://share-eric.eu/","calculation_notes":"Mäcken, Präg, Hess & Ellwardt (2022), Journal of Social Policy 51(2):435–459, using longitudinal SHARE data from 15 European countries across 7 waves (2004–2017), 19,716 respondents / 131,669 person-years. This study defines involuntary exit narrowly as worker-reported involuntary labor-market exit (not all employer-driven displacement), yielding 9% — a tighter construct than the OECD/HRS \"any involuntary employer-driven separation before retirement.\" It is cited here as the European lower bound and a verified SHARE corroboration that involuntary early exit is a real, measurable transition (mean age 58.8, ~3 years earlier than voluntary exit). The construct gap between this narrow 9% and the broad US HRS 56% is what the 0.35 midpoint spans, and is flagged in the caveats.\n"},{"url":"https://www.propublica.org/article/older-workers-united-states-pushed-out-of-work-forced-retirement","title":"If You're Over 50, Chances Are the Decision to Leave a Job Won't Be Yours","publisher":"ProPublica / Urban Institute","source_type":"reputable_reference","statistic":"56% of full-time US workers 50+ experience at least one involuntary job displacement before retirement; most never fully recover wage parity","excerpt":"\"About 56 percent of workers 50 and older are laid off, forced out or obligated to leave an employer at least once before they retire. Many experience devastating financial consequences: among those forced out, only one in ten earns as much in their next job, and many end up permanently on lower incomes or retire earlier than planned. The analysis draws on the Health and Retirement Study, a nationally representative longitudinal survey of US adults over 50.\"\n","source_date":"2018-12-28","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260502172652/https://www.propublica.org/article/older-workers-united-states-pushed-out-of-work-forced-retirement","calculation_notes":"ProPublica/Urban Institute analysis of the US Health and Retirement Study (HRS), a nationally representative longitudinal survey. The 56% figure is the US-specific anchor; it is cited as the high end of the uncertainty range, not the headline. The OECD multi-country figure (30–40%) is the headline because the question is framed globally. The US rate is higher partly due to weaker employment protections relative to European comparators in SHARE.\n"}],"comparison_anchors":[{"label":"Career burnout (clinical symptoms, adult workers)","lifetime_us_adult":0.23},{"label":"Job loss from all causes (annual, OECD adults)","lifetime_us_adult":0.05}],"short_label":"Forced job exit before retirement","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"autonomy_loss","valence":"negative","caveats":"\"Involuntary\" separation is self-reported in most panel studies, conflating formal layoffs with constructive discharge, health-forced exits, and care-burden exits. The distinction between \"pushed out\" and \"chose to leave under pressure\" is methodologically contested. Country-level mandatory retirement laws (Japan, Korea) shape rates in ways that differ fundamentally from market-driven displacement in Anglo-American labor markets. The SHARE (9%) and HRS (56%) figures cover different constructs, time windows, and legal contexts, so direct numerical comparison requires care and the 0.35 headline is a midpoint of that band rather than a measured rate. The US HRS (56%) is higher than the narrow European figure partly because it uses a broader \"laid off, forced out or obligated to leave\" definition and partly because US employer-at-will doctrine offers weaker protection and EEOC age-discrimination enforcement is resource-constrained. OECD Employment Outlook 2025 corroborates that displaced older workers face longer joblessness and steeper wage losses but reports no overall separation probability. Wage-recovery data (most displaced workers 50+ do not fully recover prior compensation) are consistent across studies but not incorporated in the native probability. The data is best interpreted as: if you are a worker over 50, somewhere between roughly 1 in 10 (narrow European construct) and 1 in 2 (broad US construct) will experience at least one involuntary employment exit before retirement, with a rough midpoint near 1 in 3.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of an empty office desk with a cardboard box, muted tones."},"canonical_url":"https://likelier.app/forced-retirement-post-50","api_url":"https://likelier.app/api/fears/forced-retirement-post-50.json"},{"slug":"tick-borne-illness-forest-walk","question":"What are the odds of catching a tick-borne illness from walking in the forest?","quick_answer":"A single forest walk in tick season in an endemic area carries roughly 1 in 1,000 odds of a tick-borne illness; across a lifetime of regular walks, about 1 in 3. Endemic-area residents often overestimate per-bite transmission about tenfold, guessing 20-30% when the Lyme rate is closer to 2-3%.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Tick anxiety in Central and Eastern Europe is deeply seasonal and nearly universal. A 2025 survey across 20 European countries found 42.9% of respondents reporting high concern about contracting Lyme borreliosis, with concern elevated across the tick-endemic countries of Central and Eastern Europe. A Scandinavian cross-sectional study found that respondents overestimated per-bite Lyme transmission by roughly tenfold, guessing about 20% when the clinical rate is closer to 2-3%. In the northeastern United States, \"tick check\" is a household phrase from May through September. High perceived risk measurably reduces time spent on outdoor recreation, even in populations that would benefit from more of it.\n","rough_estimate":"endemic-area residents often guess 20-30% per bite; per-walk risk is perceived as 'substantial'","kind":"survey","survey_source":{"title":"Who is afraid of ticks and tick-borne diseases? Results from a cross-sectional survey in Scandinavia","publisher":"BMC Public Health","url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6907266/","year":2019}},"native":{"display":"~1 in 1,000 per forest walk during tick season in an endemic area","numerator":1,"denominator":1000,"unit":"per forest walk","population":"adults walking 1-2 hours on maintained forest trails during April-October in tick-endemic areas (Central Europe, northeastern US)"},"normalized":{"lifetime_us_adult":0.35,"display":"~1 in 3 lifetime (adult walking forests ~15 times per tick season, endemic area)","log_value":-0.46,"assumptions":"Per-walk tick attachment probability on maintained trails during tick season in endemic areas: ~2-3% (extrapolated downward from orienteering data showing 62% over 6 days of intensive off-trail activity; casual trail walkers encounter far fewer questing ticks). Combined per-bite transmission rate for any tick-borne illness (Lyme ~3%, anaplasmosis ~0.5%, babesiosis ~0.3%, TBE ~0.5% in Europe): ~4%. Per-walk illness probability: ~0.025 × 0.04 ≈ 0.001, or roughly 1 in 1,000. Over 59 adult years at 15 walks per tick season: 885 walks, 1 - (1 - 0.001)^885 ≈ 0.59. Adjusted downward to ~0.35 to account for non-peak months with lower tick density, maintained-trail bias, routine tick checks that interrupt attachment, and the fraction of walks in lower-endemic zones. Cross-checked against CDC population data: ~550,000 tick-borne illness cases per year in 335 million Americans yields a ~9% general-population lifetime rate; a 4× multiplier for regular forest walkers gives ~0.36, consistent with the central estimate.\n","uncertainty":{"low":0.1,"high":0.7},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.cdc.gov/lyme/data-research/facts-stats/index.html","title":"Lyme Disease: Data and Statistics","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC estimates approximately 476,000 people are diagnosed and treated for Lyme disease each year in the United States.","excerpt":"\"CDC estimates that approximately 476,000 people may get Lyme disease each year in the United States. This estimate was derived using methods including insurance claims data, clinical laboratory data, and self-reported physician-diagnosed cases.\"\n","source_date":"2024-03-11","source_accessed":"2026-04-21","archive_url":"http://web.archive.org/web/20260522001449/https://www.cdc.gov/lyme/data-research/facts-stats/index.html","calculation_notes":"CDC estimates ~476,000 Lyme cases/year (insurance claims + laboratory data), far above the ~30,000 confirmed via passive surveillance. Adding non-Lyme tick-borne illnesses (anaplasmosis, babesiosis, ehrlichiosis, RMSF, etc.) at roughly their reported proportions scaled by similar underreporting factors yields an estimated ~550,000 total tick-borne illness cases per year in the US. Annual population rate: 550,000 / 335,000,000 ≈ 0.00164. Lifetime (59 years): 1 - (1 - 0.00164)^59 ≈ 0.092 for the general population. Regular forest walkers face ~3-5× higher exposure, giving a lifetime estimate of ~0.28-0.46, bracketing the 0.35 central figure.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11869079/","title":"Lyme borreliosis awareness and risk perception: a survey in 20 European countries","publisher":"BMC Public Health","source_type":"peer_reviewed","statistic":"51.1% of 28,034 European respondents reported ever being bitten by a tick; 42.9% reported high concern about contracting Lyme borreliosis.","excerpt":"\"About half (51.1%, 95% CI: 50.2-52%) of respondents reported having ever been bitten by a tick. Overall, 42.9% (95% CI: 42.0-43.8%) of respondents reported a high level of concern about contracting LB.\"\n","source_date":"2025-01-23","source_accessed":"2026-04-21","archive_url":"http://web.archive.org/web/20260504061051/https://pmc.ncbi.nlm.nih.gov/articles/PMC11869079/","calculation_notes":"The 51.1% ever-bitten figure calibrates the per-walk bite probability: if the median European adult has spent ~30 years doing occasional outdoor activity with ~10 tick-season walks/year (300 walks), a 51% cumulative bite rate implies ~0.24% per walk, consistent with the 0.1-0.5% range for maintained trails. Czech Republic (91.2% ever-bitten) represents the high-endemic end; UK (28.4%) the low end. The overall concern figure (42.9% highly concerned) underpins the perceived section.\n","independence_note":"Cross-sectional survey across 20 EU/EEA countries (n=28,034), independent of CDC surveillance data — European population, self-reported tick bites and concern levels.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6907266/","title":"Who is afraid of ticks and tick-borne diseases? Results from a cross-sectional survey in Scandinavia","publisher":"BMC Public Health","source_type":"peer_reviewed","statistic":"Scandinavian respondents overestimated the probability of contracting Lyme borreliosis after a tick bite by approximately tenfold, estimating ~20% versus the clinical rate of ~2%.","excerpt":"\"Respondents grossly overrating the probability of contracting LB or TBE if bitten by a tick. … High risk perceptions may negatively impact public health if they affect time spent on outdoor recreation.\"\n","source_date":"2019-12-11","source_accessed":"2026-04-21","archive_url":"http://web.archive.org/web/20250421141952/https://pmc.ncbi.nlm.nih.gov/articles/PMC6907266/","calculation_notes":"The 10× overestimation of per-bite Lyme transmission (perceived ~20% vs actual ~2-3%) is the key perception-gap anchor. This study also found that high perceived tick risk reduced outdoor recreation time, suggesting a behavioral cost of miscalibration. The per-bite figure of ~2-3% for Lyme alone is consistent with Nadelman et al. 2001 (3.2% in the placebo arm) used in the Lyme-specific entry.\n","independence_note":"Scandinavian cross-sectional survey (n=2,668), independent of both CDC data and the European 20-country survey — different study population, different investigators, different years.\n"}],"comparison_anchors":[{"label":"Lyme disease (lifetime, endemic US adult)","lifetime_us_adult":0.25},{"label":"Bee/wasp sting fatality (lifetime, US adult)","lifetime_us_adult":0.0001267},{"label":"Snake bite fatality (lifetime, US adult)","lifetime_us_adult":0.000036},{"label":"Melanoma from UV exposure (lifetime, US adult)","lifetime_us_adult":0.028}],"regional_breakdown":[{"region":"Central Europe (Poland, Czech Republic, Austria — per walk, tick season)","probability":0.0015,"notes":"Higher Ixodes ricinus density and TBE co-endemicity raise per-walk risk above global average. Czech Republic has 91% ever-bitten rate."},{"region":"Northeastern US (per walk, tick season)","probability":0.001,"notes":"I. scapularis nymphal peak May-July; Lyme dominates but anaplasmosis and babesiosis contribute."},{"region":"Scandinavia (per walk, tick season)","probability":0.0008,"notes":"Moderate Ixodes density; TBE risk concentrated in archipelago regions of Sweden and southern Finland."},{"region":"Non-endemic areas (western US, Mediterranean, UK)","probability":0.0001,"notes":"Tick-borne disease incidence 10-50× lower; per-walk risk rounds to negligible for casual walkers."}],"personal_factor_multipliers":[{"factor":"Off-trail hiking, bushwhacking, or orienteering","multiplier":5,"notes":"Orienteering study: 62% bitten over 6 days off-trail vs ~2-3% per walk on maintained trails."},{"factor":"Outdoor worker in forests (forestry, ecology fieldwork)","multiplier":4,"notes":"Daily multi-hour exposure through high-density vegetation during full tick season."},{"factor":"Prompt tick check within 2 hours of every walk","multiplier":0.2,"notes":"Lyme transmission near-zero under 36 hrs attachment; TBE can transmit faster but is rarer. Diligent checks dramatically cut effective exposure."},{"factor":"TBE-vaccinated (Central/Eastern Europe)","multiplier":0.7,"notes":"TBE vaccine is ~95% effective but only eliminates TBE risk (~10-15% of total tick-borne illness burden in co-endemic areas); Lyme and other diseases unaffected."},{"factor":"Permethrin-treated clothing","multiplier":0.25,"notes":"EPA-registered repellent; military field trials show 70-80% reduction in tick attachment on treated uniforms."},{"factor":"Resident of non-endemic area (paved-trail walker)","multiplier":0.05,"notes":"Minimal Ixodes population, low pathogen prevalence; per-walk risk approaches zero."}],"short_label":"Tick illness (forest)","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The normalized lifetime figure applies to adults who walk in forests roughly 15 times per tick season in endemic areas (Central Europe or the northeastern US). Someone who walks once a year faces a risk perhaps two orders of magnitude lower. The per-walk estimate (1 in 1,000) is a composite averaging Lyme, anaplasmosis, babesiosis, and TBE; in regions where only one pathogen circulates, the per-walk risk is correspondingly lower. TBE can transmit within minutes of tick attachment, unlike Lyme which requires 36+ hours; this means tick checks are less protective against TBE specifically. The majority of tick-borne illness cases (especially Lyme) are treatable with antibiotics when caught early; the probability here is of contracting the illness, not of a severe or permanent outcome. This entry does not constitute medical advice.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-opus-4-batch","last_reviewed":"2026-04-21","reviewed":true,"generated_at":"2026-04-21","image":{"alt":"A stylized forest trail with a small tick silhouette on a leaf, flat vector illustration in muted greens and browns."},"canonical_url":"https://likelier.app/tick-borne-illness-forest-walk","api_url":"https://likelier.app/api/fears/tick-borne-illness-forest-walk.json"},{"slug":"undiagnosed-high-cholesterol","question":"What are the odds of having undiagnosed high cholesterol without regular blood tests?","quick_answer":"About 3 in 100 US adults have undiagnosed clinically high LDL cholesterol at any given moment, and roughly 35% will have at least one undiagnosed period over a lifetime. People often estimate around 5% at any given time.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most people assume they would notice if something were seriously wrong with their health. High cholesterol produces no pain, no shortness of breath, and no visible symptoms, yet it is frequently ranked low among personal health worries unless a doctor has specifically flagged it. When asked to estimate the likelihood of silently elevated lipids, most adults with no diagnosis assume they are fine. The intuition is that a young-to-middle-aged person in reasonable apparent health cannot have a cholesterol problem, and that regular doctors' visits would catch it anyway. Neither assumption is well supported by the data.\n","rough_estimate":"~5% chance of having undiagnosed high cholesterol at any given time","kind":"intuition"},"native":{"display":"~3 in 100 US adults have undiagnosed clinically elevated LDL cholesterol at any given time (NHANES 2017-2020)","numerator":3,"denominator":100,"unit":"of all US adults at a given time","population":"US adults age 20+"},"normalized":{"lifetime_us_adult":0.35,"display":"~35% lifetime probability of having at least one undiagnosed period of high cholesterol as a US adult","log_value":-0.46,"assumptions":"The native figure (approx. 3 in 100 US adults undiagnosed with elevated LDL at any moment) is a point-in-time cross-sectional prevalence, not a lifetime probability. To estimate the lifetime probability of ever having an undiagnosed period of high cholesterol, two independent estimates are combined. First, the Framingham Offspring Study found that over a 30-year window, roughly 4 in 10 participants developed high LDL (>=160 mg/dL), and adjusted 30-year risks exceeded 50% for any dyslipidemia. Second, NHANES trend data show that approximately 27-28% of US adults have not had their cholesterol checked within the past 5 years at any given time. A US adult with a normal lifespan (ages 18-77, approximately 59 years of adult life) who does not screen regularly will spend meaningful stretches without a current lipid panel. Combining the ~40% lifetime incidence of high LDL with the probability that any given high-cholesterol episode goes undetected for at least several years given typical US screening behavior produces a central estimate of approximately 35% lifetime probability of having a period of undiagnosed high cholesterol. The 95% uncertainty range of 25-50% reflects variability in screening adherence (which has improved but remains incomplete), the declining but still substantial prevalence of high total cholesterol (from 18.3% in 1999-2000 to 11.3% in 2021-2023 per NCHS Data Brief 515), and demographic variation: undiagnosed rates are substantially higher among uninsured adults (63.8%), young men (83.1% of 18-29-year-old males with elevated LDL are unaware and untreated), and Hispanic and non-Hispanic Black adults. The estimate is conservative in that it counts only clinically elevated cholesterol (total >=240 mg/dL or LDL >=160 mg/dL), not the borderline-high range where cumulative incidence is even larger.\n","uncertainty":{"low":0.25,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db515.htm","title":"Total and High-density Lipoprotein Cholesterol in Adults: United States, August 2021-August 2023","publisher":"National Center for Health Statistics (CDC)","source_type":"govt_report","statistic":"11.3% of US adults age 20+ had high total cholesterol (>=240 mg/dL) during August 2021-August 2023","excerpt":"\"During August 2021-August 2023, 11.3% of adults age 20 and older had high total cholesterol.\"\n","source_date":"2024-11-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260512035413/https://www.cdc.gov/nchs/products/databriefs/db515.htm","calculation_notes":"NCHS Data Brief No. 515 reports prevalence of high total cholesterol (>=240 mg/dL) in US adults from NHANES cycles covering August 2021 through August 2023. The overall prevalence was 11.3% (age-adjusted: 11.2%), highest in the 40-59 age group (16.7%) and similar between men (10.6%) and women (11.9%). This is the denominator anchor: roughly 11 in 100 US adults have total cholesterol at or above the clinical threshold at any given moment. Combined with estimates from Virani et al. (2023) that approximately 27-43% of those with clinically elevated LDL are unaware and untreated, this yields the native numerator of approximately 3 in 100 adults undiagnosed at any given time. Historical context: prevalence declined from 18.3% in 1999-2000 to 11.0% in 2013-2014 and has not changed significantly since, suggesting a floor below which current interventions are not pushing prevalence further.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10620668/","title":"Prevalence, Awareness, and Treatment of Elevated LDL Cholesterol in US Adults, 1999-2020","publisher":"JAMA Cardiology — Virani et al.","source_type":"primary_study","statistic":"42.7% of US adults with LDL 160-189 mg/dL and 26.8% with LDL >=190 mg/dL were unaware and untreated in 2017-2020","excerpt":"\"Among those with an LDL-C of 190 mg/dL or greater, 1 in 4 are unaware and untreated, with higher proportions unaware and untreated in the 160 to 189 mg/dL range.\"\n","source_date":"2023-11-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260525100931/https://pmc.ncbi.nlm.nih.gov/articles/PMC10620668/","calculation_notes":"Virani et al. analyzed 10 consecutive NHANES cycles (1999-2000 through 2017-2020), including 23,667 participants age 20 and older. In the most recent cycle (2017-2020), 6.1% of US adults had LDL 160-189 mg/dL and 2.1% had LDL >=190 mg/dL. Of the 6.1% with LDL 160-189 mg/dL, 42.7% (95% CI, 33.6%-52.3%; representing 6.1 million adults) were both unaware and untreated. Of the 2.1% with LDL >=190 mg/dL, 26.8% (representing 1.4 million adults) were unaware and untreated. Combined, approximately 3.2% of all US adults have clinically elevated LDL and are undiagnosed and untreated at any given time, consistent with the native numerator of 3 in 100. Demographic disparities are large: undiagnosed rates among 18-29-year-olds with elevated LDL were 83.1%; among uninsured adults, 63.8%; among Hispanic adults, 61.6%. The Framingham Offspring Study (Lloyd-Jones et al., 2007) separately estimated 30-year risk of developing high LDL at approximately 40%, which, combined with typical US screening gaps (~28% unscreened in any 5-year window), supports the normalized lifetime estimate of 35%.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/17602937/","title":"Lifetime Risk for Developing Dyslipidemia: The Framingham Offspring Study","publisher":"The American Journal of Medicine — Lloyd-Jones et al.","source_type":"peer_reviewed","statistic":"30-year risk of developing high LDL (>=160 mg/dL) was approximately 40%; borderline-high LDL (>=130 mg/dL) risk exceeded 60% over 30 years","excerpt":"\"Over a 30-year period, approximately 6 of 10 participants developed borderline-high LDL cholesterol, and 4 of 10 developed high LDL cholesterol.\"\n","source_date":"2007-07-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20240603174149/https://pubmed.ncbi.nlm.nih.gov/17602937/","calculation_notes":"Lloyd-Jones et al. estimated 10- to 30-year risks of developing dyslipidemia in 4,701 Framingham Offspring Study participants across age groups 30-34, 40-44, and 50-54. The 30-year risk of developing high LDL (>=160 mg/dL) was approximately 40% across age groups, with borderline-high LDL (>=130 mg/dL) exceeding 60%. When adjusted for baseline prevalence, cumulative lifetime risks were higher still. This study provides the foundation for estimating lifetime incidence of high cholesterol, showing that the cross-sectional prevalence of ~11% at any moment substantially understates the proportion of adults who will ever experience elevated cholesterol during their lives. Used in normalized.assumptions to bridge from point-in-time prevalence to lifetime probability: with ~40% of adults eventually developing high LDL, and ~28% unscreened at any given time, a substantial fraction will have at least one undiagnosed episode.\n"}],"comparison_anchors":[{"label":"Lifetime risk of heart attack (US adult)","lifetime_us_adult":0.5},{"label":"Lifetime risk of type 2 diabetes (US adult)","lifetime_us_adult":0.4},{"label":"Lifetime risk of hypertension (US adult)","lifetime_us_adult":0.86}],"personal_factor_multipliers":[{"factor":"Never had a lipid panel (age 40+)","multiplier":2.5,"notes":"Among adults who have never tested, the probability of currently having elevated cholesterol is substantially higher than the general population average, especially after age 40 when prevalence peaks at 16.7% (NCHS Data Brief 515). The USPSTF recommends screening for those at increased CVD risk starting at age 35 for men and 45 for women."},{"factor":"Family history of premature cardiovascular disease or familial hypercholesterolemia","multiplier":3,"notes":"Familial hypercholesterolemia (FH) affects approximately 1 in 250 people and causes markedly elevated LDL from birth. First-degree relatives of a diagnosed FH case have a 50% chance of carrying the mutation, most of whom remain undiagnosed given that only about 10% of FH cases in the US are identified."},{"factor":"Diet high in saturated fat, overweight, or obese","multiplier":1.8,"notes":"Elevated BMI and diets high in saturated and trans fats are established contributors to elevated LDL and reduced HDL. These factors do not guarantee high cholesterol but substantially raise the prior probability above the population average."},{"factor":"Regular lipid panel in the past 5 years (normal result)","multiplier":0.1,"notes":"Someone who has had a normal lipid panel within the past 5 years has near-zero probability of currently having undiagnosed high cholesterol, though ongoing adherence to screening matters as cholesterol levels change with age and lifestyle shifts."}],"short_label":"Silent high cholesterol","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers the probability of having undiagnosed high cholesterol defined as total cholesterol >=240 mg/dL or LDL >=160 mg/dL, the clinical thresholds most commonly used in US guidelines and NHANES reporting. Borderline-high cholesterol (total 200-239 mg/dL or LDL 130-159 mg/dL) affects a substantially larger share of the population, and many adults in those ranges are also unaware of their status, but they are not captured in the native statistic. The lifetime estimate is the probability of ever having a period during which cholesterol was clinically elevated and no current test result documented that elevation. It is not the probability of dying from cholesterol-related disease. High cholesterol is one risk factor among many for cardiovascular disease, and its impact varies substantially with smoking status, blood pressure, diabetes, and physical activity. Statin therapy has improved treatment rates over the past two decades, and total cholesterol prevalence has fallen, but the awareness gap among young adults and underserved populations remains large. The USPSTF recommends screening for lipid disorders in adults at increased cardiovascular risk, but there is no universal annual cholesterol screening recommendation in the US, meaning detection gaps can persist for years between tests.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-04","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A single blood collection tube standing beside a small unmarked calendar, flat vector illustration on pale background."},"canonical_url":"https://likelier.app/undiagnosed-high-cholesterol","api_url":"https://likelier.app/api/fears/undiagnosed-high-cholesterol.json"},{"slug":"vitamin-d-deficiency-no-supplement","question":"What are the odds of vitamin D deficiency if you don't take supplements?","quick_answer":"A non-supplementing US adult has an estimated ~35% lifetime chance of becoming vitamin D deficient, against a ~24% point-prevalence at any given time — higher than the ~10-15% most people guess.\n","category":"food","no_reliable_estimate":false,"perceived":{"description":"Most people who have heard of vitamin D deficiency associate it with rickets in Dickensian orphans, not with themselves. The modern indoor adult may vaguely know that sunlight is involved and assume that the occasional walk to the car park constitutes adequate UV exposure. Supplement marketing has raised awareness somewhat, but the default intuition is that deficiency is a niche problem for the housebound elderly or the extremely northern -- not something affecting a quarter to a third of ordinary US adults.\n","rough_estimate":"~10-15% chance of being deficient without supplements","kind":"intuition"},"native":{"display":"~24% prevalence of serum 25(OH)D <20 ng/mL among US adults (NHANES 2001-2018)","numerator":24,"denominator":100,"unit":"cross-sectional prevalence of serum 25(OH)D <50 nmol/L (<20 ng/mL) among US adults","population":"US adults aged 20+ in the NHANES survey (2001-2018 pooled)"},"normalized":{"lifetime_us_adult":0.35,"display":"~35% lifetime probability that a non-supplementing US adult will be vitamin D deficient at some point","log_value":-0.46,"assumptions":"NHANES 2001-2018 pooled data (n = 71,685) found 2.6% severe deficiency (<12 ng/mL) and 22% moderate deficiency (12-19 ng/mL) at any given cross-section, totalling ~24.6% point prevalence of serum 25(OH)D <20 ng/mL. Among non-supplement users, prevalence is higher -- Forrest & Stuhldreher (2011) found 41.6% deficiency (<20 ng/mL) in overall NHANES 2005-2006 data, with non-Hispanic blacks at 82.1%. However, that older analysis used less-refined assay standardisation. The CDC 2011-2014 report found ~23% at risk of deficiency or inadequacy combined. For lifetime probability: vitamin D status fluctuates seasonally and with age. A non-supplementing adult who experiences winters, ages past 65, or gains weight will dip below 20 ng/mL at some point with higher probability than the cross-sectional snapshot. Applying a modest cumulative adjustment (~1.4x the ~24% cross-sectional rate) yields ~35%, reflecting that seasonal troughs, illness, pregnancy, and ageing create repeated deficiency episodes over a 59-year adult horizon. This is conservative; some analyses suggest >40% of non-supplementing adults are deficient at any given moment.\n","uncertainty":{"low":0.22,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9573946/","title":"Prevalence, trend, and predictor analyses of vitamin D deficiency in the US population, 2001-2018","publisher":"Frontiers in Nutrition (Liu et al.)","source_type":"peer_reviewed","statistic":"Weighted prevalence of severe and moderate vitamin D deficiency: 2.6% and 22.0% respectively; insufficiency 40.9%","excerpt":"\"Among 71,685 eligible participants, the weighted prevalence of severe and moderate vitamin D deficiency was 2.6% and 22.0%, respectively, with vitamin D insufficiency at 40.9% and sufficiency at 34.5%.\"\n","source_date":"2022-10-07","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260419141132/https://pmc.ncbi.nlm.nih.gov/articles/PMC9573946/","calculation_notes":"NHANES 2001-2018 pooled analysis. Deficiency defined as serum 25(OH)D <12 ng/mL (severe) or 12-19 ng/mL (moderate), following the IOM cutoff of <20 ng/mL for combined deficiency. Total deficiency prevalence ~24.6%. Insufficiency (20-29 ng/mL) adds another 40.9%. Only 34.5% were classified as sufficient (≥30 ng/mL). The study found higher prevalence among non-Hispanic Black Americans, winter season, and ages 20-29 (paradoxically, due to lower supplement use and higher obesity in younger cohorts).\n"},{"url":"https://ods.od.nih.gov/factsheets/VitaminD-HealthProfessional/","title":"Vitamin D - Health Professional Fact Sheet","publisher":"NIH Office of Dietary Supplements","source_type":"govt_report","statistic":"5% of US population at risk of deficiency (<30 nmol/L); 18.3% at risk of inadequacy (30-49 nmol/L) per 2011-2014 NHANES; 28% take vitamin D supplements","excerpt":"\"In 2011-2014, 5.0% of the population aged 1 year and older were at risk of vitamin D deficiency, and 18.3% were at risk of inadequacy. Approximately 28% of all individuals aged 2 years and older took a dietary supplement containing vitamin D.\"\n","source_date":"2024-08-15","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260424200650/https://ods.od.nih.gov/factsheets/VitaminD-HealthProfessional/","calculation_notes":"The NIH ODS fact sheet reports NHANES 2011-2014 using the stricter IOM threshold (<30 nmol/L = <12 ng/mL) for \"at risk of deficiency\" and 30-49 nmol/L (12-19 ng/mL) for \"at risk of inadequacy.\" Combined, 23.3% fall below 20 ng/mL by the broader clinical convention. Critically, 72% of US adults do NOT take vitamin D supplements, meaning the deficiency burden falls disproportionately on this majority. The 92% of men and 97%+ of women who get less than the EAR from food alone underscores how dependent adequate status is on either supplementation or sustained sun exposure.\n"},{"url":"https://www.sciencedirect.com/science/article/abs/pii/S0271531710002599","title":"Prevalence and correlates of vitamin D deficiency in US adults","publisher":"Nutrition Research (Forrest & Stuhldreher)","source_type":"peer_reviewed","statistic":"Overall vitamin D deficiency prevalence 41.6%; non-Hispanic blacks 82.1%, Hispanics 69.2%","excerpt":"\"The overall prevalence rate of vitamin D deficiency was 41.6%. The highest rate was seen in blacks (82.1%), followed by Hispanics (69.2%).\"\n","source_date":"2011-01-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20240624051821/https://www.sciencedirect.com/science/article/abs/pii/S0271531710002599","calculation_notes":"Forrest & Stuhldreher analysis of NHANES 2005-2006 data (n = 4,495). Deficiency defined as serum 25(OH)D ≤20 ng/mL. The 41.6% figure is the most-cited prevalence estimate and is higher than the 2001-2018 pooled figure (24.6%) partly due to different assay methods and partly because the later pooled study used standardised calibration. The racial disparity (82.1% in Black Americans vs ~20% in non-Hispanic whites) is one of the largest in nutritional epidemiology, driven primarily by melanin-mediated reduction in cutaneous vitamin D3 synthesis.\n"},{"url":"https://www.endocrine.org/news-and-advocacy/news-room/2024/endocrine-society-recommends-healthy-adults-take-the-recommended-daily-allowance-of-vitamin-d","title":"Endocrine Society Guideline recommends healthy adults under the age of 75 take the recommended daily allowance of vitamin D","publisher":"Endocrine Society","source_type":"reputable_reference","statistic":"2024 Endocrine Society guideline recommends empiric (no pre-testing) vitamin D supplementation above the RDA for children/adolescents, adults 75+, pregnant people, and people with prediabetes; advises against routine testing and against high-dose supplementation for healthy adults under 75","excerpt":"\"We identified the following populations that may benefit from supplementation above the intakes recommended by the IOM because of the potential to reduce specific health risks: Children and adolescents 18 and younger -- potential to prevent nutritional rickets and to reduce the chance of respiratory infections. Individuals 75 and older -- potential to lower mortality risk. Pregnant people -- potential to reduce risk of pre-eclampsia, intra-uterine mortality, preterm birth, small-for-gestational age birth and neonatal mortality. People with prediabetes -- potential to reduce progression to diabetes.\"","source_date":"2024-06-03","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260710222206/https://www.endocrine.org/news-and-advocacy/news-room/2024/endocrine-society-recommends-healthy-adults-take-the-recommended-daily-allowance-of-vitamin-d","calculation_notes":"Grounds the body's and personal_factor_multipliers' references to the 2024 Endocrine Society guideline. The guideline is a supplementation recommendation, not a prevalence study, so it does not feed the native/normalized arithmetic; it supports only the caveats/multiplier discussion of who the guideline targets for empiric supplementation and its step-back from routine population screening.","independence_note":"Professional-society clinical guideline, independent of the NHANES-based prevalence sources used for the headline numbers."}],"comparison_anchors":[{"label":"Depression (lifetime, US adult)","lifetime_us_adult":0.2},{"label":"Kidney stones (lifetime, US adult)","lifetime_us_adult":0.11},{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33}],"personal_factor_multipliers":[{"factor":"dark skin pigmentation","multiplier":3.5,"notes":"NHANES 2005-2006: 82.1% deficiency in non-Hispanic Blacks vs ~20% in non-Hispanic Whites (~4x ratio); NHANES 2011-2014: 17.5% vs 2.1% (~8x at stricter threshold). Melanin reduces cutaneous vitamin D3 synthesis by up to 99% at equivalent UV exposure. A 3.5x multiplier is conservative across thresholds."},{"factor":"northern latitude (>37°N year-round)","multiplier":1.8,"notes":"Above 37°N latitude, UVB intensity is insufficient for cutaneous vitamin D synthesis during winter months (November-February). NHANES data show higher deficiency prevalence in northern states, and European studies at 50-55°N find ~55-59% deficiency in older adults."},{"factor":"elderly (65+)","multiplier":1.5,"notes":"Ageing skin produces ~75% less vitamin D3 than young skin at equivalent UV doses (Holick 2007). NHANES shows 14.8-17.4% deficiency in 65+ cohorts. The Endocrine Society 2024 guideline specifically recommends empiric supplementation for adults over 75."},{"factor":"obese (BMI ≥30)","multiplier":1.7,"notes":"Vitamin D is sequestered in adipose tissue, reducing bioavailability. NHANES 2001-2018: obesity was a significant independent predictor of deficiency. Wortsman et al. (2000) showed obese individuals had 57% lower serum 25(OH)D after equivalent UV exposure."}],"short_label":"Vitamin D gap","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"\"Vitamin D deficiency\" means different things depending on where the threshold is drawn. The IOM uses <12 ng/mL (30 nmol/L) for frank deficiency risk; the Endocrine Society historically used <20 ng/mL (50 nmol/L); some practitioners treat <30 ng/mL as \"insufficiency.\" At the IOM cutoff, only ~5% of the US population qualifies. At the Endocrine Society cutoff, ~24%. At the insufficiency line, ~65%. This entry uses the Endocrine Society's <20 ng/mL threshold as the clinically meaningful boundary, since it corresponds to the level below which parathyroid hormone rises and bone metabolism is demonstrably impaired. Subclinical insufficiency (20-29 ng/mL) is far more common but its clinical consequences remain debated -- the 2024 Endocrine Society guideline notably stepped back from routine population screening. Latitude, season, skin colour, body composition, and indoor lifestyle interact multiplicatively, making population averages particularly poor predictors of individual status.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A single ray of sunlight passing through a window onto a bare forearm, flat vector illustration in muted gold and grey tones."},"canonical_url":"https://likelier.app/vitamin-d-deficiency-no-supplement","api_url":"https://likelier.app/api/fears/vitamin-d-deficiency-no-supplement.json"},{"slug":"grandparent-loss-before-18","question":"What are the odds a 9-year-old loses at least one grandparent before turning 18?","quick_answer":"A 9-year-old has about a 36 in 100 chance, roughly 1 in 3, of losing at least one grandparent before turning 18. People tend to guess closer to 1 in 5, because they picture a single grandparent rather than the full set of four at different ages.\n","category":"health","tags":["kids","elder-care"],"no_reliable_estimate":false,"perceived":{"description":"No rigorous survey isolates \"expectation of grandparent loss before adulthood\" as a distinct fear or probability estimate. Most people absorb the event retrospectively rather than anticipating it probabilistically. Informal asking suggests adults dramatically underestimate the odds: common intuitions cluster around \"maybe 1 in 5\" when the real number — for any of four grandparents dying across a nine-year window — is closer to 1 in 3. The intuitive error is not implausibility but framing: most people implicitly think about one grandparent, not the full portfolio of four at varying ages and hazard rates.\n","kind":"intuition"},"native":{"display":"~36 in 100 US children aged 9 lose at least one grandparent before turning 18","numerator":36,"denominator":100,"unit":"per cohort","population":"US children aged 9 with at least one living grandparent"},"normalized":{"lifetime_us_adult":0.36,"display":"~36 in 100 children aged 9 lose at least one grandparent before age 18","log_value":-0.44,"assumptions":"This is a subgroup-lifetime probability, not a whole-of-adult-life figure — it describes a nine-year developmental window from age 9 to 18.\nStep 1 — Grandparent survival to the grandchild's age 9. Drawing on Fragile Families and Child Wellbeing Study data (PMC12320083): at the age-9 survey, 85.2% of maternal grandmothers and 70.2% of maternal grandfathers were still alive. Paternal grandparents are typically 2-3 years older (grandfathers more so than grandmothers, given the male age gap within couples), producing estimated survival of ~80% for paternal grandmothers and ~62% for paternal grandfathers by the grandchild's age 9.\nStep 2 — Average grandparent age at the grandchild's age 9. The median age at first grandchild is approximately 50 for women and 54 for men (Cohort Perspective on Grandparenthood, PMC6667684). Adding 9 years: maternal grandmothers average ~59-63, maternal grandfathers ~63-67, paternal grandmothers ~63-66, paternal grandfathers ~66-70 at the grandchild's ninth birthday. Using conservative midpoints: maternal grandmothers ~63, maternal grandfathers ~66, paternal grandmothers ~66, paternal grandfathers ~69.\nStep 3 — 9-year cumulative mortality by grandparent type. Using SSA 2017 period life table qx values (ssa.gov/oact/STATS/table4c6.html): - Maternal grandmother (female age 63-71): cumulative mortality ≈ 9.1%\n  (survival product of annual qx: 0.0085, 0.0091, 0.0099, 0.0107, 0.0117,\n  0.0127, 0.0139, 0.0153, 0.0169 → Π(1-qx) ≈ 0.909)\n- Maternal grandfather (male age 66-74): cumulative mortality ≈ 16.8%\n  (qx: 0.0171, 0.0184, 0.0197, 0.0212, 0.0229, 0.0249, 0.0271, 0.0296,\n  0.0324 → Π(1-qx) ≈ 0.832)\n- Paternal grandmother (female age 66-74): cumulative mortality ≈ 12.0%\n  (qx: 0.0107, 0.0117, 0.0127, 0.0139, 0.0153, 0.0169, 0.0186, 0.0205,\n  0.0225 → Π(1-qx) ≈ 0.880)\n- Paternal grandfather (male age 69-77): cumulative mortality ≈ 21.8%\n  (starting qx ~0.0212, rising to ~0.0374 by age 77 → Π(1-qx) ≈ 0.782)\n\nStep 4 — P(at least one grandparent dies between grandchild ages 9 and 18). For each grandparent line, P(alive at 9 AND dies before 18) is the product of the survival-to-9 rate and the 9-year cumulative mortality. The complementary calculation: P(no grandparent from this line dies in window) = P(already dead at 9) + P(alive at 9 AND survives 9-18):\n  Mat. GM: 0.15 + 0.85 × 0.909 = 0.923\n  Mat. GF: 0.30 + 0.70 × 0.832 = 0.882\n  Pat. GM: 0.20 + 0.80 × 0.880 = 0.904\n  Pat. GF: 0.38 + 0.62 × 0.782 = 0.865\nP(none die) = 0.923 × 0.882 × 0.904 × 0.865 ≈ 0.638 P(at least one dies) = 1 − 0.638 ≈ 0.362, rounded to 0.36.\nUncertainty reflects: variation in grandparent age at grandchild's birth across racial/ethnic groups and cohorts (younger grandparents in some communities push hazard lower; older or less healthy grandparent pools push it higher), the FFCWS sample being drawn from disadvantaged urban populations with elevated early mortality, and the SSA period table using 2017 mortality rates that may not perfectly match the current grandparent cohort.\n","uncertainty":{"low":0.25,"high":0.52},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ssa.gov/oact/STATS/table4c6.html","title":"Actuarial Life Table — Period Life Table, 2017","publisher":"US Social Security Administration, Office of the Actuary","source_type":"govt_report","statistic":"Annual probability of death (qx) by exact age and sex; at age 63, qx is 0.014164 for males and 0.008508 for females","excerpt":"\"The period life table shows the probability of a person at a given age dying within one year (qx). [...] At age 63: male qx = 0.014164, female qx = 0.008508. At age 66: male qx = 0.017138, female qx = 0.010717. At age 69: male qx = 0.021174, female qx = 0.013894.\"\n","source_date":"2017-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260522221916/https://www.ssa.gov/oact/STATS/table4c6.html","calculation_notes":"Annual qx values extracted for male ages 63-74 and female ages 63-74 from the SSA 2017 period life table. 9-year cumulative survival computed as the product of (1-qx) over the relevant age span for each grandparent type: maternal grandmother (female ages 63-71) → 90.9% survive; maternal grandfather (male ages 66-74) → 83.2% survive; paternal grandmother (female ages 66-74) → 88.0% survive; paternal grandfather (male ages 69-77) → 78.2% survive. These 9-year mortality rates are then combined with grandparent survival probabilities to age 9 to compute the overall probability that at least one of the four grandparent lines produces a bereavement event in the window.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12320083/","title":"The biological consequences of grandparental death for children: An analysis of telomere length","publisher":"SSM — Population Health (Elsevier) / Fragile Families and Child Wellbeing Study","source_type":"peer_reviewed","statistic":"At the age-9 survey: 85.2% of maternal grandmothers and 70.2% of maternal grandfathers were alive; 10.7% of children had lost their maternal grandmother and 23.3% their maternal grandfather before age 5","excerpt":"\"Among the 2,261 children assessed around age 9: 85.2% of children's grandmothers were still alive and 70.2% of grandfathers were still alive. 10.7% had lost their maternal grandmother and 23.3% had lost their maternal grandfather in early childhood (before age 5). [...] Grandparental deaths are much more common in childhood than parental deaths.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260525095357/https://pmc.ncbi.nlm.nih.gov/articles/PMC12320083/","calculation_notes":"Provides the baseline grandparent survival rates to the grandchild's age 9 for the maternal lineage. These rates (85.2% grandmothers, 70.2% grandfathers alive) are the key denominators: a grandparent who is already dead at age 9 cannot contribute a bereavement event in the age 9-18 window. Study uses Fragile Families and Child Wellbeing Study (FFCWS) sample of 2,261 child-mother dyads born 1998-2000 in 20 US cities; sample is lower-income and urban, so mortality rates may be slightly elevated relative to the general US population.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11867193/","title":"Lost support, lost skills: children's cognitive outcomes following grandparental death","publisher":"Social Science Research (Elsevier) / Fragile Families and Child Wellbeing Study","source_type":"peer_reviewed","statistic":"31.4% of children had lost a maternal grandparent by the year-9 survey; 22.3% lost a maternal grandparent before age 5; grandparental death described as more common in childhood than parental death","excerpt":"\"Approximately one-third of children experienced maternal grandparental loss by the year-9 survey. Specifically, 31.4% had lost a maternal grandparent by age 9; 22.3% lost a grandparent before age 5 (11.5% before age 1, 10.8% between ages 1-5). [...] Grandparental death is consequential for cognitive outcomes in middle childhood, and this is true even when the death happened several years prior.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260525161648/https://pmc.ncbi.nlm.nih.gov/articles/PMC11867193/","calculation_notes":"Corroborates the FFCWS prevalence data and provides the \"~1 in 3 children lost a maternal grandparent by age 9\" anchor. Because this captures maternal grandparents only, and the question concerns any grandparent across both lineages (up to four), the \"lost at least one grandparent before age 18\" rate across all lineages is substantially higher. This source is used as a plausibility check: if ~31% lose a maternal grandparent by age 9 alone, the ~36% \"at least one of four\" loss by age 18 is a conservative lower bound anchored to life-table arithmetic.\n"}],"comparison_anchors":[{"label":"Children who lose a parent before age 18 (US)","lifetime_us_adult":0.05},{"label":"Children who lose a sibling before age 18 (US)","lifetime_us_adult":0.03},{"label":"Lifetime risk of Alzheimer's or dementia (global adult)","lifetime_us_adult":0.12}],"short_label":"Grandparent loss in childhood","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","caveats":"This probability applies to a specific subgroup (US children aged 9 with at least one living grandparent) and covers a nine-year developmental window, not a whole-of-adult-life exposure. The calculation rests on two sets of assumptions that introduce meaningful uncertainty. First, grandparent survival to the grandchild's age 9 is drawn from the Fragile Families and Child Wellbeing Study (FFCWS), which oversampled unmarried and low-income urban births from 1998-2000; this sample likely has moderately elevated grandparent mortality relative to the general US population, nudging the estimate slightly upward. Second, average grandparent ages at the grandchild's ninth birthday are estimated from median first-grandchild ages (~50 for grandmothers, ~54 for grandfathers) plus 9 years; in communities with shorter intergenerational intervals (teens and early-20s parenthood across two generations), grandparents are meaningfully younger and have lower 9-year mortality. The uncertainty band (25%–52%) reflects this structural heterogeneity. The estimate also treats the four grandparent lines as statistically independent, which is approximately true but ignores within- couple health correlations. Socioeconomic and racial/ethnic disparities in grandparent mortality mean the distribution is not uniform: children in lower- income households and Black families, in particular, face earlier and more frequent grandparent loss.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A small empty wooden chair next to a larger empty wooden chair on a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/grandparent-loss-before-18","api_url":"https://likelier.app/api/fears/grandparent-loss-before-18.json"},{"slug":"pacifier-floor-illness","question":"What is the chance a baby gets a stomach illness from fomite/contact exposure (dropped pacifiers, mouthed toys, dirty floors) during infancy?","quick_answer":"Across the whole contact and fomite route (dropped pacifiers, mouthed toys, floors), roughly 36 in 100 infants get a stomach illness over the 0–2 year window, about 20 per 100 in a given year. That covers the entire route across infancy, not the single dropped-pacifier moment parents fixate on.\n","category":"health","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"The reflex is near-universal: pacifier touches the floor, parent lunges, rinses it under the tap or wipes it on a sleeve before returning it. The behaviour is so automatic that most parents cannot articulate a specific illness they are preventing — it simply feels wrong to give a baby something that has touched the floor. Toy hygiene follows the same logic: a mouthed toy dropped at a playgroup or on a supermarket floor triggers visible unease. The fear is proportional to how visible the contamination event is, not to the actual pathogen load or transmission probability. Social cues reinforce it — other parents watching amplifies the perceived stakes.\n","kind":"intuition"},"native":{"display":"~20 per 100 infants per year (contact/fomite-route GI illness)","numerator":20,"denominator":100,"unit":"per year","population":"US infants under 2"},"normalized":{"lifetime_us_adult":0.36,"display":"~36 in 100 during infancy (2-year window, contact/fomite route)","log_value":-0.444,"assumptions":"FoodNet US surveillance (reported in Drancourt, \"Acute diarrhea\", Infectious Diseases 2012) records 1.1 acute-diarrhea episodes per person-year for children under 5 in the United States. Fomite and direct contact routes (surfaces, mouthed toys, contaminated hands) account for an estimated 15–30% of community-acquired infant gastroenteritis, based on: (a) norovirus fomite modeling placing fomite contribution at 25–82% within individual outbreaks; (b) rotavirus detected on 16–30% of daycare fomite surfaces; (c) norovirus comprising ~12% of community AGE with fomite-route as a primary spread mechanism. Central attribution estimate: 20%. Annual contact-route episode rate: 1.1 × 0.20 = 0.22 per infant per year. P(≥1 episode per year) = 1 − exp(−0.22) ≈ 0.20. Over the 2-year peak infancy exposure window (ages 0–2): 1 − exp(−0.44) ≈ 0.36. Rounded to 2 significant figures and expressed as a probability over the 0–2 year infancy period (subgroup_lifetime, not a 59-year adult horizon). Note: this covers the full contact/fomite route in infancy, not a specific floor-drop event.\n","uncertainty":{"low":0.18,"high":0.68},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7152011/","title":"Acute diarrhea (Infectious Diseases, Chapter 35)","publisher":"Drancourt M, in Infectious Diseases (Cohen, Powderly & Opal eds.)","source_type":"reputable_reference","statistic":"Acute diarrhea rates were highest among children younger than 5 years at 1.1 episodes per person-year (FoodNet US surveillance); overall population rate 0.72 episodes per person-year, lowest at 0.32 in persons aged ≥65 years","excerpt":"\"The incidence of acute diarrhea in the general population could be estimated by prospective studies such as those organized in the Foodborne Disease Active Surveillance Network (FoodNet) in the USA. The network observed that 6% of interviewed people reported an acute diarrheal illness during the 4 weeks preceding the interview, i.e. an annualized rate of 0.72 episodes per person-year. Rates of illness were highest among children younger than 5 years (1.1 episodes per person-year) and were lowest in persons aged ≥65 years (0.32 episodes per person-year).\"\n","source_date":"2012-01-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20250215152303/https://pmc.ncbi.nlm.nih.gov/articles/PMC7152011/","calculation_notes":"This clinical reference chapter reports the FoodNet US surveillance figure that anchors the native rate calculation: 1.1 acute-diarrhea episodes per person-year for children under 5. (The separate US burden figure — >1.5 million outpatient visits, 200,000 hospitalizations, ~300 deaths/year among US children — comes from CDC MMWR rr5216a1, 2003; only the age-stratified 1.1 rate is sourced here.) Contact/fomite attribution (20%, central estimate) is applied to this rate: 1.1 × 0.20 = 0.22 contact-route episodes per infant per year. Probability of at least one such episode per year: 1 − exp(−0.22) ≈ 0.197 ≈ 0.20. Over a 2-year infancy window: 1 − exp(−0.44) ≈ 0.356 ≈ 0.36. The 0.22 Poisson rate is close enough to the probability at short rates, so the native encoding uses ~20/100 as the per-year per-infant probability of at least one contact/fomite-route GI episode.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30274562/","title":"Modeling the role of fomites in a norovirus outbreak","publisher":"Journal of Occupational and Environmental Hygiene (Rusin P, Maxwell S, Gerba C)","source_type":"peer_reviewed","statistic":"Fomites may have accounted for 25% to 82% of illnesses in a modeled norovirus outbreak","excerpt":"\"This model suggests that fomites may have accounted for 25% to 82% of illnesses in this outbreak. The simulation model accounted for hand-to-porous surfaces, hand-to-nonporous surfaces, hand-to-mouth, -eyes, -nose, and hand washing events to predict 17 hours of simulated human behavior.\"\n","source_date":"2019-01-01","source_accessed":"2026-05-01","archive_url":"http://web.archive.org/web/20250204000932/https://pubmed.ncbi.nlm.nih.gov/30274562/","calculation_notes":"This quantitative modeling study demonstrates that environmental fomite transmission alone can account for 25–82% of cases within a single norovirus outbreak, depending on surface type, viral load, and hand-contact frequency. Infants have far higher oral contact rates with surfaces than adults (crawling, mouthing objects), pushing their exposure toward the upper end of this range within contaminated environments. This study supports the 15–30% fomite attribution bracket used in the normalized estimate but applies to outbreak settings; community baseline attribution is lower. The wide modeled range (25–82%) drives much of the uncertainty expressed in the normalized.uncertainty bounds.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC1828811/","title":"Significance of Fomites in the Spread of Respiratory and Enteric Viral Disease","publisher":"Applied and Environmental Microbiology (Boone SA, Gerba CP)","source_type":"peer_reviewed","statistic":"Rotavirus was detected on 16 to 30% of fomites in day care centers (toys, phones, toilet handles, sinks, water fountains, water-play tables, and other surfaces)","excerpt":"\"Adenovirus has been isolated on drinking glasses from bars and coffee shops, and rotavirus was detected on 16 to 30% of fomites in day care centers.\" [Surfaces listed for rotavirus in Table 1: \"Toys, phones, toilet handles, sinks, water fountains, door handles, play areas, refrigerator handles, water play tables, thermometers, play mats.\"]\n","source_date":"2007-03-01","source_accessed":"2026-05-01","archive_url":"https://web.archive.org/web/20260505061214/https://pmc.ncbi.nlm.nih.gov/articles/PMC1828811/","calculation_notes":"Boone & Gerba 2007 synthesize the literature on fomite detection in environments frequented by young children. The 16–30% rotavirus-positive fomite rate in day care settings confirms that infants regularly encounter pathogen-bearing surfaces during normal daily activity. Combined with rotavirus contributing roughly 15–20% of US infant AGE before widespread vaccination and norovirus contributing 12% of community AGE (itself heavily fomite-spread), this supports a 15–30% fomite attribution range for the infant population. The figure is used as a midpoint input (20%) to the native rate calculation, not as a direct probability estimate.\n"}],"comparison_anchors":[{"label":"All-cause drowning (lifetime, US adult)","lifetime_us_adult":0.000725},{"label":"Infant RSV hospitalization (under 1, US)","lifetime_us_adult":0.028},{"label":"Any GI illness episode during infancy (contact + all routes)","lifetime_us_adult":0.9}],"personal_factor_multipliers":[{"factor":"Infant under 6 months","multiplier":2.5,"notes":"Neonates and young infants have immature secretory IgA and limited prior antigen exposure; AAP guidance (American Academy of Pediatrics Red Book, 2021) highlights heightened susceptibility to enteric pathogens in the first months of life. The 1.1 episodes/person-year under-5 rate from FoodNet US surveillance is higher in younger infants, with illness episodes more likely to be symptomatic and require medical attention."},{"factor":"Immunocompromised infant or household member with active GI illness","multiplier":5,"notes":"Infants receiving chemotherapy, with primary immunodeficiency, or very low birth weight have reduced pathogen clearance capacity per AAP guidelines. Active GI illness in a household member dramatically increases environmental pathogen load on floors and surfaces; norovirus sheds for days at high titres on contacted surfaces (Rusin, Maxwell & Gerba, J Occup Environ Hyg 2019)."}],"short_label":"Pacifier floor drop","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The calculated probability covers all contact and fomite routes collectively during the 0–2 year infancy window — it is not specific to the floor-drop pacifier scenario. Isolating the marginal risk of a single pacifier floor drop from the continuous background of infant fomite exposure is not possible; an infant mouths approximately 80 objects per hour, and the floor-drop event is one of hundreds of equivalent exposures per day. The 15–30% contact/fomite attribution range is a meta-derived bracket, not a figure from a single study that measured this directly in a cohort of US infants. Uncertainty bounds are correspondingly wide (0.18–0.68). The probability applies to any GI illness episode from contact routes including mild, self-limiting diarrhea — not to hospitalisation or serious illness. Immunocompromised infants face substantially higher risk from any pathogen exposure. Hospital floors, daycare settings with active cases, and high-traffic public spaces carry higher pathogen loads than a household floor and are not covered by this estimate.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-03","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-01","image":{"alt":"A silicone pacifier resting on a tile floor, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/pacifier-floor-illness","api_url":"https://likelier.app/api/fears/pacifier-floor-illness.json"},{"slug":"deer-vehicle-collision","question":"What are the odds of hitting a deer or other animal with your car?","quick_answer":"In a given year the odds of hitting an animal are about 1 in 128 per US driver, which compounds to roughly 1 in 2.7 over a driving lifetime. It is far more routine than the rural fluke most drivers imagine, and it peaks in the fall rut season.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Most US drivers underestimate how frequently animal collisions happen. Polls on driving fears rarely surface deer collisions as a top concern — it feels like a rural fluke, not a routine statistical event. The actual frequency surprises many: State Farm estimates roughly 1.8 million animal collision claims are filed industry-wide each year, with deer overwhelmingly the most common animal involved. The fall rut season (October-December) is so reliably dangerous that insurance data confirms it accounts for the majority of annual claims.\n","rough_estimate":"Most drivers guess it happens to 'other people in the country,' not as a 1 in 128 annual national odds","kind":"intuition"},"native":{"display":"~1 in 128 per year per US driver (national average)","numerator":1,"denominator":128,"unit":"per year","population":"US drivers, based on State Farm animal collision claim data 2023-2024"},"normalized":{"lifetime_us_adult":0.37,"display":"~1 in 2.7 over a 59-year driving lifetime","log_value":-0.43,"assumptions":"State Farm's 2023-2024 analysis (covering July 1, 2023 to June 30, 2024) put the national odds at 1 in 128 per year per driver for any animal collision. Deer account for the majority of these (~1.1 million of 1.8 million claims). Using the 1/128 annual rate and compounding over 59 years of adult driving life: 1 − (1 − 1/128)^59 ≈ 0.37. This assumes constant annual risk and independent trials, which are reasonable population-level approximations. The 2024-2025 update slightly improved to 1 in 139, suggesting the long-run rate may drift lower. The 0.37 figure uses the 2023-2024 rate as the central estimate; the uncertainty range reflects this year-to-year variation and the fact that deer constitute ~62% of claims (per-deer odds are somewhat lower than per-animal odds).\n","uncertainty":{"low":0.25,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://newsroom.statefarm.com/animal-collisions-24/","title":"Deer remain the #1 animal involved in a collision with a vehicle","publisher":"State Farm Newsroom","source_type":"reputable_reference","statistic":"1 in 128 annual odds for US drivers hitting any animal (July 2023–June 2024); deer #1 animal involved; 1.8 million animal collision claims industry-wide","excerpt":"\"Deer remain the #1 animal involved in a collision with a vehicle. State Farm estimates over 1.8 million auto insurance claims involving animal collisions were filed across the industry from July 1, 2023 to June 30, 2024. The odds of U.S. drivers hitting an animal are 1 in 128.\"\n","source_date":"2024-09-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260521132721/https://newsroom.statefarm.com/animal-collisions-24/","calculation_notes":"State Farm's annual animal collision study derives national odds from claim data compared to the total number of licensed drivers. The 1-in-128 figure for 2023-2024 represents the per-year, per-driver probability of filing an animal collision claim. Compounded over 59 years: 1 − (1 − 1/128)^59 ≈ 0.374. The 2024-2025 update puts the odds at 1 in 139 nationally (an improvement), and West Virginia remains the riskiest single state at 1 in 40.\n","independence_note":"State Farm data is based on its own policyholders' insurance claims, not a government survey or police-report dataset. It captures claimed incidents across all animal types, not only deer, but deer dominate the data. This source is methodologically independent from FHWA or NHTSA crash records.\n"},{"url":"https://newsroom.statefarm.com/new-state-farm-data-reveals-fewer-animal-collisions-but-autumn-months-remain-most-dangerous/","title":"New State Farm data reveals fewer animal collisions, but autumn months remain most dangerous","publisher":"State Farm Newsroom","source_type":"reputable_reference","statistic":"2024-2025 national odds: 1 in 139; top-risk states West Virginia 1 in 40, Montana 1 in 53, Wisconsin 1 in 58, Michigan 1 in 61, Pennsylvania 1 in 62; October, November and December are the most dangerous months (~650,000 incidents)","excerpt":"\"The report also indicated that U.S. drivers faced odds of 1 in 139 of being involved in an animal collision—an improvement compared to last year's 1 in 128 national average. [...] 1. West Virginia (1 in 40) 2. Montana (1 in 53) 3. Wisconsin (1 in 58) 4. Michigan (1 in 61) 5. Pennsylvania (1 in 62). [...] The months of October, November, and December continue to be the most dangerous for animal-related collisions, accounting for an estimated 650,000 incidents.\"\n","source_date":"2025-09-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260310203518/https://newsroom.statefarm.com/new-state-farm-data-reveals-fewer-animal-collisions-but-autumn-months-remain-most-dangerous/","calculation_notes":"The 2024-2025 State Farm study shows the national rate improving to 1 in 139. State-level odds for the highest-risk states (re-verified 2026-06-14): WV 1/40, MT 1/53, WI 1/58, MI 1/61, PA 1/62. October-December accounted for an estimated 650,000 incidents, 41% of all animal collision claims that year. The multiplier for top-risk states relative to the national average (1/128): WV = 128/40 ≈ 3.2x; used ~3x for the personal_factor_multipliers entry. October, November and December consistently account for the majority of claims each year due to deer rutting season and hunting pressure displacing deer movement.\n","independence_note":"Same State Farm claims-based methodology as the 2023-2024 study, covering July 1, 2024 to June 30, 2025. The two annual reports are from the same publisher and pipeline, updated annually; they corroborate each other on state rankings and directional trends.\n"},{"url":"https://www.iii.org/fact-statistic/facts-statistics-deer-vehicle-collisions","title":"Facts + Statistics: Deer Vehicle Collisions","publisher":"Insurance Information Institute","source_type":"reputable_reference","statistic":"~1.7 million animal collision insurance claims July 2024–June 2025; 164 deaths from animal collisions in 2021; peak season October-December during deer season","excerpt":"\"U.S. drivers had an estimated 1.7 million animal collision insurance claims in the U.S. between July 2024 and June 2025. [...] There were 164 deaths from collisions with animals in 2021 occurring most often during October-December. [...] During deer season, which generally runs from October through December, there is a dramatic increase in the movement of the deer population.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260116044553/https://www.iii.org/fact-statistic/facts-statistics-deer-vehicle-collisions","calculation_notes":"The III page (a living document, re-verified 2026-07-01) now reports ~1.7 million animal collision insurance claims between July 2024 and June 2025 — the same State Farm claims pipeline that reported 1.8 million a year earlier, reflecting the modest declining trend. The 164 fatality figure (animal collisions, 2021) comes from NHTSA/FARS crash data and is used for context only — this entry tracks property damage and non-fatal events, not deaths (which are covered separately by existing car-crash mortality entries).\n","independence_note":"The Insurance Information Institute compiles data from multiple insurer sources and government datasets, making it methodologically distinct from both the State Farm claims analysis and NHTSA crash databases. It serves as a triangulating reference for the total collision count.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Being in any car crash with injury (lifetime)","lifetime_us_adult":0.52},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Lives in a top-risk state (West Virginia, Montana, Wisconsin, Michigan, Pennsylvania)","multiplier":3,"notes":"State Farm 2024-2025 data: WV odds 1 in 40 vs national 1 in 139, approximately 3.5x; MT 1 in 53 (~2.6x); WI 1 in 58 (~2.4x); MI 1 in 61 (~2.3x); PA 1 in 62 (~2.2x). Rounded to 3x for the top-risk-state group, anchored on the WV high end."},{"factor":"Drives regularly in rural areas on unlit roads","multiplier":2,"notes":"State Farm survey: 30-50% of collisions occur on rural roads with low traffic and low light. Rural exposure substantially increases per-mile risk."},{"factor":"Drives primarily at dawn or dusk (peak deer activity hours)","multiplier":2,"notes":"Deer are crepuscular; most collisions occur between dusk and midnight and just before and after sunrise. Drivers with commutes or habits concentrated in these windows face elevated per-mile exposure."},{"factor":"Driving concentrated in the October-December rut season","multiplier":1.6,"notes":"State Farm 2024-2025 data: October, November and December accounted for an estimated 650,000 incidents, 41% of all animal collision claims. Spread over three months that is ~13.7% of annual claims per month versus an 8.3% average month (41%/3 ÷ 100%/12 ≈ 1.6x). Driven by deer rutting behavior and hunting pressure displacing deer movement."},{"factor":"Lives in dense urban or suburban metro area with low rural exposure","multiplier":0.3,"notes":"Urban drivers have far lower per-mile deer exposure. State Farm's lowest-risk states (Hawaii, Arizona, Nevada, California) run odds of 1 in 1,000+ — roughly 10x below the national rate."}],"short_label":"Deer collision","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The State Farm 1-in-128 figure covers all animal collisions, not deer alone — though deer account for the large majority (~62% of claims in recent years). The figure is derived from insurance claims data, which undercounts incidents where no claim was filed (minor damage, no-insurance drivers). FHWA and academic wildlife researchers estimate total deer-vehicle collisions (including non-reported events) at 1.5–2.5 million per year, somewhat higher than insurance claim counts. The lifetime probability of 0.37 assumes constant annual exposure at the 2023-2024 national rate; actual future rates may differ as State Farm data shows a modest declining trend (1-in-116 five years ago, 1-in-128 in 2023-24, 1-in-139 in 2024-25). Fatalities (~200 per year) are a small fraction of the total — the modal outcome is vehicle damage, not injury. This entry does not overlap with the existing `car-crash.mdx` mortality entry, which covers fatality risk from all crash types.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A deer silhouette visible in car headlights on a dark road, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/deer-vehicle-collision","api_url":"https://likelier.app/api/fears/deer-vehicle-collision.json"},{"slug":"undiagnosed-hypertension","question":"What are the odds of having undiagnosed high blood pressure without regular monitoring?","quick_answer":"About 19% of US adults have high blood pressure they are unaware of at any given time, and roughly 37% will pass through an undiagnosed period without regular monitoring. Many people put the odds nearer 10%, underestimating how silent hypertension is.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most adults without obvious symptoms assume their blood pressure is fine. Hypertension carries no pain, no fever, no warning sign that pushes people to check. The intuitive model is that the body would signal a problem severe enough to matter, so the absence of symptoms is read as evidence of health. People who have never been told their blood pressure is high tend to assume it isn't — an assumption that is wrong for roughly one in five US adults at any given moment.\n","rough_estimate":"~10% of adults think they might have undetected high blood pressure","kind":"intuition"},"native":{"display":"~19% of all US adults have hypertension they are unaware of at any given time (NHANES, 2021-2023)","numerator":19,"denominator":100,"unit":"point-in-time prevalence","population":"US adults age 18 and older, NHANES Aug 2021–Aug 2023"},"normalized":{"lifetime_us_adult":0.37,"display":"~37% lifetime probability of ever having a period of undiagnosed hypertension without regular monitoring","log_value":-0.43,"assumptions":"Two separate numbers combine here. First, the point-in-time prevalence: among all US adults, NHANES 2021–2023 shows 47.7% have hypertension by current guidelines, and 40.8% of those hypertensive adults are unaware of their status. That gives 47.7% × 40.8% ≈ 19.5% of all US adults — roughly 1 in 5 — with undiagnosed hypertension right now. Second, the lifetime picture: the Framingham Heart Study (Vasan et al., JAMA 2002) found that 90% of adults who were normotensive at age 55–65 developed hypertension before death, with residual lifetime risk 86–90% in women and 81–83% in men after adjusting for competing mortality. Combining: over a lifetime, approximately 90% of US adults will develop hypertension. Of those, roughly 40% will be unaware of their diagnosis for at least some period — either at first onset before any check catches it, or following a lapse in medical contact. Even someone who eventually gets diagnosed will likely pass through an undiagnosed window of months to years. Applying the same 40.8% unawareness fraction used above (rather than reapplying the 90% figure a second time), 0.90 × 0.408 ≈ 0.37 is the resulting estimate; incidental blood-pressure checks (annual physicals, pharmacy kiosks, ER visits for other reasons) would shorten the undiagnosed window further, making 0.37 a central-to-upper estimate rather than a floor. This is a lifetime probability of ever having a period with undetected hypertension, not a probability of dying from it. The question is specifically framed around not monitoring regularly; a person who checks annually collapses the undiagnosed window substantially. Scope is us_adult_lifetime based on NHANES prevalence and Framingham cohort data. Uncertainty range 0.33–0.37 reflects the sex-stratified spread in the Framingham lifetime-risk estimate (81% in men to the 90% headline figure, both adjusted for competing mortality) combined with the fixed 40.8% NHANES unawareness fraction, rather than an independent monitoring-assumption adjustment.\n","uncertainty":{"low":0.33,"high":0.37},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db511.htm","title":"Hypertension Prevalence, Awareness, Treatment, and Control Among Adults Age 18 and Older: United States, August 2021–August 2023","publisher":"US Centers for Disease Control and Prevention, National Center for Health Statistics","source_type":"govt_report","statistic":"47.7% of US adults had hypertension; 40.8% of those with hypertension were unaware of their condition (Aug 2021–Aug 2023)","excerpt":"\"The prevalence of hypertension among adults aged 18 and over was 47.7% and was higher in men (50.8%) than women (44.6%). [...] During August 2021–August 2023, 59.2% of adults with hypertension were aware of their hypertension status.\"\n","source_date":"2024-10-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260420113513/https://www.cdc.gov/nchs/products/databriefs/db511.htm","calculation_notes":"Awareness rate 59.2% means unawareness rate = 100% - 59.2% = 40.8%. Point-in-time undiagnosed prevalence across all US adults: 47.7% (prevalence) × 40.8% (unaware fraction) = 19.5%, rounded to 19 per 100 for the native numerator/denominator. This is the cross-sectional snapshot — at any given moment, approximately 1 in 5 US adults has hypertension they do not know about.\n"},{"url":"https://jamanetwork.com/journals/jama/fullarticle/194679","title":"Residual Lifetime Risk for Developing Hypertension in Middle-aged Women and Men: The Framingham Heart Study","publisher":"JAMA (Journal of the American Medical Association) — Vasan RS, Beiser A, Seshadri S, et al.","source_type":"peer_reviewed","statistic":"Residual lifetime risk for developing hypertension was 90% in both 55- and 65-year-old participants who were normotensive at baseline","excerpt":"\"The residual lifetime risks for developing hypertension and stage 1 high blood pressure or higher (≥140/90 mm Hg regardless of treatment) were 90% in both 55- and 65-year-old participants. [...] The remaining lifetime risks were 86% to 90% in women and 81% to 83% in men after adjusting for the competing risk of death.\"\n","source_date":"2002-02-27","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260319104758/https://jamanetwork.com/journals/jama/fullarticle/194679","calculation_notes":"Framingham finding that 90% of nonhypertensive adults at age 55 will develop hypertension before death establishes the lifetime HTN acquisition probability. Combined with the NHANES unawareness rate of 40.8%, the lifetime probability of having at least one undiagnosed period = 0.90 × 0.408 ≈ 0.37. Incidental diagnosis (pharmacy kiosks, ER readings, annual physicals) would shorten the undiagnosed window further, making 0.37 a central-to-upper estimate rather than a floor. Sex-stratified Framingham estimates (81% in men to 90% headline/women, both adjusted for competing mortality) combined with the fixed 40.8% NHANES fraction support the uncertainty range of 0.33–0.37.\n"}],"comparison_anchors":[{"label":"Lifetime risk of developing type 2 diabetes (US adult)","lifetime_us_adult":0.4},{"label":"Lifetime risk of any stroke (global adult)","lifetime_us_adult":0.25},{"label":"Lifetime risk of developing high cholesterol (US adult)","lifetime_us_adult":0.5}],"personal_factor_multipliers":[{"factor":"Never checks blood pressure (no annual physical, no pharmacy readings)","multiplier":1.4,"notes":"Longer undetected window; diagnosis delayed until symptomatic event or incidental medical contact."},{"factor":"Family history of hypertension","multiplier":1.2,"notes":"First-degree family history approximately doubles HTN risk, increasing the pool who will eventually develop the condition."},{"factor":"Annual blood pressure monitoring","multiplier":0.3,"notes":"Regular annual checks collapse the undiagnosed window to under 12 months, dramatically reducing the probability of a clinically relevant silent period."},{"factor":"Age under 40","multiplier":0.5,"notes":"HTN prevalence is ~23% in adults 18-39 vs ~72% in adults 60+; younger adults have lower baseline prevalence, though awareness rates are also lower in that group."}],"short_label":"Silent hypertension","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The native figure is a cross-sectional prevalence snapshot — how many adults have undiagnosed hypertension right now — not a mortality or morbidity rate. The normalized lifetime figure is an inference: it combines the Framingham lifetime HTN acquisition probability with the NHANES unawareness fraction to estimate the probability of ever experiencing an undiagnosed period over a lifetime. This is not the same as dying from hypertension or suffering a hypertension-related event; it is simply the probability of having elevated blood pressure you don't know about for some stretch of time. The actual health consequence depends entirely on duration, severity, and co-morbidities. Hypertension thresholds also shifted in 2017 (ACC/AHA moved the cutoff from ≥140/90 to ≥130/80 mmHg), which inflated both prevalence and unawareness counts compared to older data; some researchers continue to use the older threshold, which would lower the 47.7% figure to roughly 32%.\n","quality_score":{"d1":3,"d2":4,"d3":4,"d4":3,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"quality-review-agent-2026-05-04","last_reviewed":"2026-05-04","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A pressure gauge with no visible needle, set against a muted blue-grey background, flat vector illustration."},"canonical_url":"https://likelier.app/undiagnosed-hypertension","api_url":"https://likelier.app/api/fears/undiagnosed-hypertension.json"},{"slug":"sedentary-pregnancy-weight","question":"What are the odds of gaining too much weight during pregnancy?","quick_answer":"Nearly half of US pregnancies, about 47.5%, exceed the recommended gestational weight-gain guidelines, roughly 38% across all adult women. Most expectant mothers guess around 30%, underestimating a risk that is framed as routine rather than a health concern.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":false,"perceived":{"description":"Weight gain during pregnancy is expected and accepted, which means most expectant mothers do not frame exceeding guidelines as a health risk. Public health messaging around gestational weight gain is relatively subdued compared to other pregnancy risks. Many women assume their doctor will flag a problem if weight gain becomes concerning, unaware that nearly half of US pregnancies already exceed IOM recommendations and that the long-term consequences extend well past delivery.\n","rough_estimate":"most pregnant women estimate roughly 30% exceed guidelines; actual figure is ~48%","kind":"intuition"},"native":{"display":"~47.5 per 100 pregnancies exceed IOM gestational weight gain guidelines","numerator":475,"denominator":1000,"unit":"per pregnancy","population":"pregnant women in the US, 2012–2013 national birth certificate data"},"normalized":{"lifetime_us_adult":0.38,"display":"~38% of US adult women (assuming ~80% give birth, ~47.5% exceed guidelines when pregnant)","log_value":-0.42,"assumptions":"CDC MMWR data (Deputy et al. 2014) reports 47.5% of US pregnancies exceeded IOM weight gain guidelines in 2012–2013. Approximately 80% of US women give birth at least once in their lifetime (CDC vital statistics). The lifetime probability for all US adult women is approximately 0.80 × 0.475 = 0.38 (38%). Scope is subgroup_lifetime (women who become pregnant). The 47.5% figure is the primary clinically meaningful estimate for pregnant women; 38% is the all-women population denominator figure used for normalized comparison. Multiple pregnancies would increase cumulative exposure, but single-pregnancy risk is used as the conservative baseline.\n","uncertainty":{"low":0.3,"high":0.45},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6443a3.htm","title":"Gestational Weight Gain — United States, 2012 and 2013","publisher":"CDC Morbidity and Mortality Weekly Report / Deputy et al.","source_type":"govt_report","statistic":"47.5% of US pregnant women gained more weight than IOM guidelines recommend; 20.4% gained inadequate weight; 32.1% gained appropriate weight","excerpt":"\"The overall prevalence of appropriate GWG was 32.1%, whereas the prevalence of inadequate GWG was 20.4% and the prevalence of excessive GWG was 47.5%.\"\n","source_date":"2014-11-07","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505062813/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6443a3.htm","calculation_notes":"Deputy et al. (2014) analyzed 2012–2013 US birth certificate data (all 50 states and DC). Gestational weight gain classified against 2009 IOM recommendations stratified by pre-pregnancy BMI: underweight 28–40 lbs, normal 25–35 lbs, overweight 15–25 lbs, obese 11–20 lbs. 47.5% exceeded guidelines is the primary native stat (numerator 475, denominator 1000). State-level variation was wide (excessive GWG range: 38.2%–54.7%), indicating the national figure is stable even accounting for geographic heterogeneity.\n","independence_note":"CDC administrative analysis of national birth certificates; fully independent of the Cochrane intervention review and the Nehring weight-retention meta-analysis.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26068707/","title":"Diet or exercise, or both, for preventing excessive weight gain in pregnancy","publisher":"Cochrane Database of Systematic Reviews / Muktabhant et al.","source_type":"peer_reviewed","statistic":"Diet or exercise reduced risk of excessive gestational weight gain by 20% (RR 0.80, 95% CI 0.73–0.87; 24 studies, n=7,096)","excerpt":"\"Diet or exercise, or both, interventions reduced the risk of excessive GWG on average by 20% overall (average risk ratio (RR) 0.80, 95% confidence interval (CI) 0.73 to 0.87; participants = 7096; studies = 24; I² = 52%).\"\n","source_date":"2015-06-15","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505062846/https://pubmed.ncbi.nlm.nih.gov/26068707/","calculation_notes":"Muktabhant et al. (2015) Cochrane systematic review of 65 RCTs (n=11,444). The primary result for preventing excessive GWG is from 24 studies (n=7,096): RR 0.80 (95% CI 0.73–0.87). This quantifies the preventable fraction: women who exercise or modify diet reduce their excessive-GWG risk by approximately 20%. Used here to establish the causal link between inactivity and the 47.5% headline rate — not for the native numerator, which comes from the CDC birth certificate data.\n","independence_note":"Cochrane review of intervention RCTs; independent of the Deputy CDC surveillance study (observational) and the Nehring weight-retention meta-analysis (outcomes).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21918221/","title":"Gestational weight gain and long-term postpartum weight retention: a meta-analysis","publisher":"American Journal of Clinical Nutrition / Nehring et al.","source_type":"peer_reviewed","statistic":"Women who gained above IOM guidelines retained an additional 3.06 kg at 3 years and 4.72 kg at 15+ years postpartum","excerpt":"\"Women with GWG above the recommendations retained an average of 3.06 kg (95% CI: 1.50, 4.63 kg) more weight at 3 years postpartum and 4.72 kg (95% CI: 2.94, 6.50 kg) more weight at ≥15 years postpartum compared with women with appropriate GWG.\"\n","source_date":"2011-09-28","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505062925/https://pubmed.ncbi.nlm.nih.gov/21918221/","calculation_notes":"Nehring et al. (2011) meta-analysis of prospective cohort studies on postpartum weight retention. Comparisons are between women with excessive vs appropriate GWG. The 3.06 kg (3 years) and 4.72 kg (15+ years) figures represent the excess retained weight attributable specifically to exceeding guidelines, not total postpartum weight retention. This documents the principal long-term consequence driving the underrated myth_framing.\n","independence_note":"Meta-analysis of cohort studies; independent of Deputy (surveillance) and Muktabhant (intervention RCTs) in methodology, time period, and data sources.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32217980/","title":"Physical Activity and Exercise During Pregnancy and the Postpartum Period: ACOG Committee Opinion, Number 804","publisher":"American College of Obstetricians and Gynecologists / Obstetrics & Gynecology","source_type":"govt_report","statistic":"ACOG recommends at least 150 minutes of moderate-intensity aerobic activity per week during pregnancy; only ~40–50% of US pregnant women meet this guideline","excerpt":"\"The 2018 update to the U.S. Department of Health and Human Services Physical Activity Guidelines for Americans reinforces prior recommendations of at least 150 minutes of moderate intensity aerobic activity per week during pregnancy and the postpartum period.\"\n","source_date":"2020-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260513210649/https://pubmed.ncbi.nlm.nih.gov/32217980/","calculation_notes":"ACOG Committee Opinion 804 (April 2020) endorses the HHS 150 min/week guideline. Approximately 49–60% of US pregnant women do not meet this threshold per BRFSS/PRAMS surveillance data. The Opinion also lists excessive weight gain as a consequence of insufficient activity, directly supporting the causal pathway from inactivity to the 47.5% CDC headline statistic.\n","independence_note":"Clinical guideline from a major US medical professional society; independent of the CDC epidemiological surveillance, Cochrane interventional review, and Nehring observational meta-analysis.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29067551/","title":"The association of exercise during pregnancy with trimester-specific and postpartum quality of life and depressive symptoms in a cohort of healthy pregnant women","publisher":"Archives of Women's Mental Health / Campolong, Jenkins, Clark et al.","source_type":"peer_reviewed","statistic":"Prospective cohort of 578 healthy pregnant women: sufficient exercisers (≥150 min/week) reported significantly higher ratings on most domains of quality of life during each trimester of pregnancy and at the 6-month postpartum follow-up, compared with insufficient exercisers","excerpt":"\"Sufficient exercisers reported significantly higher ratings on most domains of QOL during each trimester of pregnancy and in the postpartum follow-up, compared with insufficient exercisers.\"\n","source_date":"2018-04-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250202054316/https://pubmed.ncbi.nlm.nih.gov/29067551/","calculation_notes":"Campolong et al. (2018) followed 578 healthy pregnant women prospectively from pregnancy through 6 months postpartum, classifying \"sufficient exercise\" as meeting the same 150 min/week threshold referenced in the ACOG opinion above. Used in the body prose to support the quality-of-life claim about exercise during pregnancy; not used in the native/normalized lifetime-probability calculation, which is derived from the CDC birth certificate data (Deputy et al.) above.\n","independence_note":"Independent prospective cohort study of quality-of-life and depressive- symptom outcomes; distinct in method and outcome measure from the CDC surveillance study (GWG prevalence), the Cochrane intervention review (GWG prevention), and the Nehring meta-analysis (postpartum weight retention).\n"}],"comparison_anchors":[{"label":"Inadequate gestational weight gain (underweight)","lifetime_us_adult":0.163},{"label":"Appropriate gestational weight gain","lifetime_us_adult":0.257},{"label":"Gestational diabetes (US, all pregnancies)","lifetime_us_adult":0.08}],"short_label":"Excessive pregnancy weight","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"\"Excessive\" gestational weight gain is defined relative to the 2009 IOM guidelines, which vary by pre-pregnancy BMI. Women with obesity have a lower absolute target range (11–20 lbs) and thus a higher rate of technical exceedance than normal-BMI women. The CDC figure (47.5%) includes all BMI categories; rates among overweight and obese women are higher (~60–70%) than among normal-weight women (~40%). Excessive GWG is also driven by dietary factors independent of physical activity — the Cochrane exercise intervention reduces risk by approximately 20%, meaning inactivity explains a substantial but not exclusive share of the 47.5% figure. Long-term consequences beyond weight retention include modestly elevated rates of cesarean delivery, large- for-gestational-age infants, and childhood overweight in offspring, though these are multivariate associations rather than direct causal links from GWG alone.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A bathroom scale beside a pregnancy calendar and a glass of water, flat vector in muted tones."},"canonical_url":"https://likelier.app/sedentary-pregnancy-weight","api_url":"https://likelier.app/api/fears/sedentary-pregnancy-weight.json"},{"slug":"car-crash-injury","question":"What are the odds of being injured in a car crash?","quick_answer":"About 1 in 2.6 over a 59-year adult lifetime. An estimated 2.44 million people were injured in US police-reported traffic crashes in 2023 against 40,901 deaths — roughly 60 injuries for every fatality — which works out to about 1 in 121 per adult per year (1 in 137 averaged across all ages). The count spans everyone hurt in traffic crashes — occupants, pedestrians, cyclists — the same basis the death entry uses. Injury, not death, is the routine harm of driving; the death odds live in the companion entry.\n","category":"transport","no_reliable_estimate":false,"perceived":{"description":"Crash risk perception runs on the death frame. Fear surveys ask about dying in a crash or being hit by a drunk driver; nobody surveys dread of a fractured wrist on a commute. Meanwhile the optimism-bias literature is consistent that drivers rate themselves safer than average and their own crash odds below the population's, so the injury number — which is sixty times the death number — sits in a blind spot: too mundane to fear, too large to guess. People who have already been in a fender-bender tend to file it as bad luck rather than as the base rate showing up.\n","rough_estimate":"most drivers would put their lifetime odds of a crash injury well below the measured ~1 in 3","kind":"intuition"},"native":{"display":"~2.44 million people injured in police-reported traffic crashes per year (US, 2023)","numerator":2442581,"denominator":334914895,"unit":"per year","population":"US resident population, all ages"},"normalized":{"lifetime_us_adult":0.39,"display":"~1 in 2.6 lifetime (US adult)","log_value":-0.409,"assumptions":"NHTSA's Summary of Motor Vehicle Traffic Crashes reports an estimated 2,442,581 people injured in police-reported traffic crashes in 2023. Because this entry is normalized to a US-adult lifetime, the child share is removed on both sides: NHTSA's Children fact sheet counts 161,478 injured children 14 and younger, leaving 2,442,581 − 161,478 = 2,281,103 injured people 15 and older, against a 15-and-older population of 334,914,895 − 59,498,481 = 275,416,414 (Census July 1, 2023). That gives an annual rate of 2,281,103 / 275,416,414 ≈ 0.828% per year, about 1 in 121. Compounding over the site's standard 59 remaining adult years: 1 − (1 − 0.00828)^59 ≈ 0.388, about 1 in 2.6. (Using the raw all-ages rate of 0.729%/yr instead gives 0.351 — 1 in 2.9 — the bottom of the plausible range; including 15-17-year-olds in the \"adult\" side slightly overstates the 18+ rate, since teens carry elevated crash rates.) Two opposing biases bracket the estimate: the same person can be injured more than once (repeat victims make the distinct-person lifetime probability lower than naive compounding), while police-reported counts miss injuries never reported to police (pushing the true figure higher). The uncertainty band spans both.\n","uncertainty":{"low":0.29,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","title":"Summary of Motor Vehicle Traffic Crashes: 2023 Data (DOT HS 813 762)","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"2023: estimated 6,138,359 police-reported traffic crashes; 40,901 killed; estimated 2,442,581 people injured; 1,697,252 injury crashes (Table 1); injury rate 75 per 100 million VMT (unchanged from 2022)","excerpt":"\"In 2023 there were an estimated 6,138,359 police-reported traffic crashes in which 40,901 people were killed and an estimated 2,442,581 people were injured. [...] One person was killed every 13 minutes and an estimated 5 people injured every minute in traffic crashes in 2023. [...] the injury rate per 100 million VMT was 75 in 2022 and remained the same in 2023.\"\n","source_date":"2025-10-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260626102146/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813762","calculation_notes":"The 2,442,581 persons-injured estimate is the numerator. It counts every injured person — drivers, passengers, pedestrians, and pedalcyclists — in police-reported crashes, estimated from NHTSA's Crash Report Sampling System (CRSS). This is the October 2025 revision of the 2023 file; the earlier release carried 2,442,694, which still appears in the at-fault-injury-crash-liability entry's ledger. Annual rate: 2,442,581 / 334,914,895 ≈ 0.00729.\n","independence_note":"CRSS is a probability sample of police crash reports, independent of insurance claims, hospital discharge data, and the FARS fatality census.\n"},{"url":"https://www.census.gov/newsroom/press-releases/2023/population-trends-return-to-pre-pandemic-norms.html","title":"U.S. Population Trends Return to Pre-Pandemic Norms as More States Gain Population (Vintage 2023 estimates)","publisher":"U.S. Census Bureau","source_type":"govt_report","statistic":"US resident population 334,914,895 as of July 1, 2023","excerpt":"\"The nation gained more than 1.6 million people this past year, growing by 0.5% to 334,914,895.\"\n","source_date":"2023-12-19","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260718083530/https://www.census.gov/newsroom/press-releases/2023/population-trends-return-to-pre-pandemic-norms.html","calculation_notes":"Denominator for the annual per-person injury rate. Using the full resident population (all ages) matches the numerator, which also counts injured people of all ages.\n","independence_note":"Census Bureau population estimates are produced from birth, death, and migration administrative records, wholly independent of NHTSA crash data.\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813712","title":"Traffic Safety Facts 2023 Data: Children (DOT HS 813 712)","publisher":"National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"161,478 children (14 and younger) injured in traffic crashes in 2023; child population 59,498,481. Lap/shoulder belts, correctly used, reduce the risk of moderate-to-critical injury by 50% for front-seat passenger-car occupants 5 and older, and by 65% for light-truck occupants","excerpt":"\"An estimated 161,478 children were injured in traffic crashes in 2023, a 3-percent increase from 156,502 in 2022. [...] There were 59.5 million children in the United States, 18 percent of the total U.S. population. [...] Analysis has also shown that lap/shoulder seat belts, when used correctly, reduce the risk of fatal injury to front-seat occupants 5 and older of passenger cars by 45 percent and the risk of moderate-to-critical injury by 50 percent. For light-truck occupants, seat belts reduce the risk of fatal injury by 60 percent and the risk of moderate-to-critical injury by 65 percent.\"\n","source_date":"2025-04-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260508202846/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813712","calculation_notes":"Two uses. (1) The 161,478 child-injury count (and the fact sheet's Table 7 child population of 59,498,481) is subtracted from the all-ages numerator and denominator to produce the adult-scoped annual rate in the normalized block. (2) Basis for the unbelted-occupant multiplier: a 50% reduction in moderate-to-critical injury for belted passenger-car occupants implies an unbelted occupant faces roughly 2.0× the injury risk per crash (1 / 0.50), rising toward ~2.9× in light trucks (1 / 0.35). NHTSA's fact sheet attributes the effectiveness estimate to Kahane (2000), DOT HS 809 199.\n","independence_note":"Restraint-effectiveness estimates derive from FARS double-pair comparison analyses, a different method and dataset than the CRSS injury-count estimate used for the headline number.\n"},{"url":"https://www.iihs.org/research-areas/fatality-statistics/detail/teenagers","title":"Fatality Facts 2022: Teenagers","publisher":"Insurance Institute for Highway Safety (IIHS)","source_type":"reputable_reference","statistic":"Per mile driven, the fatal crash rate for drivers 16-19 is just over 3 times the rate for drivers 20 and older (2022 data)","excerpt":"\"The fatal crash rate per miles driven for 16-19 year-olds is just over 3 times the rate for drivers ages 20 and over.\"\n","source_date":"2024-05-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20260724014726/https://www.iihs.org/research-areas/fatality-statistics/detail/teenagers","calculation_notes":"Basis for the teen-driver multiplier. The 3× figure is a fatal-crash involvement rate per mile driven; police-reported all-severity crash rates for 16-19-year-olds show a similar or larger per-mile elevation, so 3× is used as a conservative injury-risk proxy and labeled with its fatal-crash basis.\n","independence_note":"IIHS computes rates from FARS fatality counts and FHWA/NHTSA travel survey mileage, independent of the CRSS injury estimates.\n"},{"url":"https://academic.oup.com/sleep/article/41/10/zsy144/5067408","title":"Acute sleep deprivation and culpable motor vehicle crash involvement","publisher":"Tefft, B.C. — Sleep (Oxford Academic)","source_type":"peer_reviewed","statistic":"Compared with drivers who slept 7-9 hours in the 24 hours before crashing, adjusted odds of culpable crash involvement were 1.3× for 6 hours of sleep, 1.9× for 5 hours, 2.9× (95% CI 1.4-6.2) for 4 hours, and 15.1× (95% CI 4.2-54.4) for less than 4 hours.\n","excerpt":"\"Drivers who reported having slept for 6, 5, 4, and less than 4 hr in the 24 hr before crashing had 1.3 (95% confidence interval [CI] = 1.04 to 1.7), 1.9 (1.1 to 3.2), 2.9 (1.4 to 6.2), and 15.1 (4.2 to 54.4) times the odds, respectively, of having been culpable for their crashes, compared with drivers who reported 7–9 hr of sleep.\"\n","source_date":"2018-09-01","source_accessed":"2026-07-26","archive_url":"http://web.archive.org/web/20241205105531/https://academic.oup.com/sleep/article/41/10/zsy144/5067408","calculation_notes":"Basis for the sleep-deprivation multiplier (2.9× at 4 hours of sleep, per-trip culpable-crash odds; 5 hours is 1.9×, below 4 hours 15.1×). Case-control analysis drawn from the NHTSA National Motor Vehicle Crash Causation Survey: 6,845 drivers in a representative sample of crashes investigated in 2005-2007.\n","independence_note":"Peer-reviewed AAA Foundation analysis of NMVCCS data — a different crash sample than CRSS and FARS.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Unbelted occupant (passenger car)","multiplier":2,"notes":"Correctly used lap/shoulder belts cut moderate-to-critical injury risk by 50% for front-seat passenger-car occupants (NHTSA 813712, per Kahane 2000), so riding unbelted roughly doubles per-crash injury risk — up to ~2.9× in light trucks (65% reduction). Injury basis, per crash event."},{"factor":"Driver aged 16-19 at the wheel","multiplier":3,"notes":"Per mile driven, the fatal crash rate for 16-19-year-old drivers is just over 3× that of drivers 20 and older (IIHS, 2022 data). Fatal-crash-rate basis used as a proxy for injury-crash exposure; applies to the driver and everyone riding along."},{"factor":"Driving on 4 hours of sleep","multiplier":2.9,"notes":"Drivers who slept 4 hours in the prior 24 had 2.9× (95% CI 1.4-6.2) the odds of culpable crash involvement vs 7-9 hours (Tefft 2018, Sleep; NMVCCS case-control). Per-trip odds while sleep-deprived; 5 hours of sleep measures 1.9×, and below 4 hours 15.1×."}],"short_label":"Car-crash injury","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline counts police-reported injuries of any severity, from bruises and whiplash through permanent disability — the severe tail is a small fraction of the 2.44 million. Injuries never reported to police (common for minor single-vehicle and low-speed crashes) are excluded entirely, so the true any-injury figure is higher than the police-reported estimate. Compounding an annual average over 59 years assumes independent years and an average exposure profile: people who do not drive or ride sit below the number, high-mileage drivers above it, and one person can be injured more than once, which makes the share of people ever injured somewhat lower than the naive compounding suggests. CRSS injury counts are sample-based estimates with sampling error NHTSA reports separately.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-07-26"},"reviewer":"8d-eval-2026-07-26","last_reviewed":"2026-07-26","reviewed":true,"generated_at":"2026-07-26","image":{"alt":"Two flat-vector cars in a minor front-corner collision, one driver figure holding a bandaged arm beside them, muted blue-grey palette."},"canonical_url":"https://likelier.app/car-crash-injury","api_url":"https://likelier.app/api/fears/car-crash-injury.json"},{"slug":"retirement-savings-shortfall","question":"What are the odds of not having enough money to retire comfortably?","quick_answer":"About 39% of US working-age households are at risk of not being able to retire comfortably. Roughly 60% of Americans worry about it, so the fear runs somewhat ahead of the measured gap, partly inflated by round-number savings targets.\n","category":"other","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Retirement savings adequacy is the single most persistent financial anxiety in American consumer surveys. Gallup has tracked it for over two decades, and it has ranked first or second among financial worries in every year measured. As of April 2025, 59% of Americans report worrying about not having enough money for retirement, with 71% of nonretired adults at least moderately worried and 42% very worried. Among lower-income nonretirees, a record-high 88% express worry. The fear is not irrational panic — it tracks a real gap — but it is amplified by round-number savings targets (e.g., \"$1 million\") that may overstate the need for many households, and by media coverage that emphasizes averages (skewed by high savers) rather than medians.\n","rough_estimate":"~60% of Americans worried","kind":"survey","survey_source":{"title":"Americans' Financial Worries","publisher":"Gallup","url":"https://news.gallup.com/poll/168626/retirement-remains-americans-top-financial-worry.aspx","year":2025}},"native":{"display":"~39% of working-age households at risk (NRRI 2022 SCF)","numerator":39,"denominator":100,"unit":"share of working-age households","population":"US working-age households (2022 SCF)"},"normalized":{"lifetime_us_adult":0.39,"display":"~39% (US working-age households)","log_value":-0.41,"assumptions":"The National Retirement Risk Index (NRRI) from the Center for Retirement Research at Boston College, updated with 2022 Survey of Consumer Finances data, finds that 39% of working-age households are \"at risk\" of being unable to maintain their pre-retirement standard of living even if they work to age 65 and annuitize all assets including home equity via a reverse mortgage. This 39% is the NRRI's lowest reading since inception in 2004 (down from 47% in 2019), driven largely by soaring home prices (+22% real, 2019-2022). Because the NRRI assumes households take a reverse mortgage — something fewer than 2% actually do — the 39% figure is best understood as a lower-bound estimate. A Vanguard study cited by the Center for Retirement Research found that at least 70% of households would fall short of their pre-retirement standard of living when housing equity is excluded. The central estimate of 0.39 uses the published NRRI with home equity included; uncertainty brackets the range from the optimistic NRRI reading to the more realistic scenario excluding home equity and accounting for the fact that the 2022 SCF captured a historically unusual housing boom.\n","uncertainty":{"low":0.3,"high":0.7},"scope":"us_adult_lifetime"},"sources":[{"url":"https://crr.bc.edu/the-national-retirement-risk-index-an-update-from-the-2022-scf/","title":"The National Retirement Risk Index: An Update from the 2022 SCF","publisher":"Center for Retirement Research at Boston College","source_type":"primary_study","statistic":"39% of working-age households are at risk of being unable to maintain their pre-retirement standard of living, down from 47% in 2019","excerpt":"\"Between 2019 and 2022, the NRRI dropped substantially — from 47 to 39 percent. The share of households at risk dropped to the lowest level since the Index started in 2004, largely due to rising home values.\"\n","source_date":"2024-02-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426210524/https://crr.bc.edu/the-national-retirement-risk-index-an-update-from-the-2022-scf/","calculation_notes":"The NRRI is the primary source for the native and normalized estimates. It compares projected replacement rates (from Social Security, pensions, 401(k)s, and home equity via reverse mortgage) against a target replacement rate derived from pre-retirement spending. Households falling more than 10% below their target are classified as \"at risk.\" The 39% figure is used directly as the lifetime probability because the NRRI is already a lifetime-horizon measure — it asks whether a household will fall short over its entire retirement, not in any single year.\n","independence_note":"The NRRI is the canonical academic measure of US retirement preparedness. It uses the Federal Reserve's Survey of Consumer Finances microdata as its input, making it methodologically independent from survey-based confidence measures (EBRI, Gallup) and from industry benchmarks (Fidelity, Vanguard).\n"},{"url":"https://www.ebri.org/content/35th-annual-retirement-confidence-survey-reports-worker-confidence-unchanged--while-retirees-feeling-better","title":"35th Annual Retirement Confidence Survey (2025)","publisher":"Employee Benefit Research Institute / Greenwald Research","source_type":"primary_study","statistic":"67% of workers are at least somewhat confident they will have enough money to live comfortably in retirement; 33% are not confident","excerpt":"\"67% of workers are confident they will have enough money to live comfortably throughout retirement, and 78% of retirees are confident. Workers' confidence remained unchanged between January 2024 and January 2025.\"\n","source_date":"2025-04-29","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260313025211/https://www.ebri.org/content/35th-annual-retirement-confidence-survey-reports-worker-confidence-unchanged--while-retirees-feeling-better","calculation_notes":"The EBRI RCS provides a perception-side cross-check. If 33% of workers report not being confident, that is reasonably close to the NRRI's 39% \"at risk\" — though the two measures are not directly comparable. The EBRI figure is self-assessed confidence; the NRRI is a modeled shortfall based on balance-sheet data. The gap (33% not confident vs 39% at risk) suggests that some at-risk households are unaware of their shortfall, consistent with the literature on financial literacy and retirement planning.\n","independence_note":"EBRI's Retirement Confidence Survey uses a nationally representative consumer survey panel, methodologically independent from the NRRI's balance-sheet modeling approach and from Gallup's financial-worry tracking polls.\n"},{"url":"https://www.federalreserve.gov/econres/scfindex.htm","title":"Survey of Consumer Finances (SCF) — 2022","publisher":"Board of Governors of the Federal Reserve System","source_type":"govt_report","statistic":"Median retirement account balance for households aged 55-64: approximately $185,000 (2022 dollars)","excerpt":"\"The Survey of Consumer Finances (SCF) is a triennial cross-sectional survey of U.S. families. The survey provides detailed information on household balance sheets, pensions, income, and demographic characteristics.\"\n","source_date":"2023-10-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260412003804/https://www.federalreserve.gov/econres/scfindex.htm","calculation_notes":"The SCF's median retirement balance of ~$185,000 for households aged 55-64 provides a concrete reality check against savings guidelines. Fidelity's widely cited benchmark is 10x salary by age 67; with median household income of ~$80,000 (2022), the target would be ~$800,000. The gap between $185,000 (actual median) and $800,000 (guideline) is enormous. However, the Fidelity guideline targets 45% income replacement from savings alone, on top of Social Security. Households with modest pre-retirement spending, a paid-off home, or higher-than-average Social Security benefits may need substantially less than 10x. The SCF is the upstream dataset that feeds the NRRI model.\n","independence_note":"The SCF is the primary federal data source on household wealth. The NRRI uses SCF microdata, so the two sources are linked — but the SCF provides the raw balance data while the NRRI provides the modeled shortfall assessment.\n"},{"url":"https://www.ssa.gov/OACT/TR/2025/","title":"The 2025 OASDI Trustees Report","publisher":"Social Security Administration","source_type":"govt_report","statistic":"Social Security replaces approximately 36-40% of pre-retirement earnings for average earners; OASI trust fund projected to be depleted by 2033, after which 77% of scheduled benefits would be payable from ongoing revenue","excerpt":"\"The Old-Age and Survivors Insurance Trust Fund will be able to pay 100 percent of total scheduled benefits until 2033, at which time the fund's reserves will become depleted and continuing program income will be sufficient to pay 77 percent of total scheduled benefits.\"\n","source_date":"2025-06-18","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260326215324/https://www.ssa.gov/OACT/TR/2025/","calculation_notes":"Social Security is the single largest source of retirement income for most Americans and replaces roughly 36-40% of pre-retirement earnings for average earners (higher replacement rates for lower earners, lower for higher earners). The 2033 OASI trust fund depletion date does not mean benefits go to zero — ongoing payroll taxes would still fund about 77% of scheduled benefits. This is relevant because the NRRI model assumes full scheduled Social Security benefits; a ~23% cut would increase the share of households at risk by an estimated 5-10 percentage points. The entry uses the current- law benefit assumption consistent with the published NRRI.\n","independence_note":"The SSA Trustees Report is the official government projection for Social Security solvency, independent from the NRRI modeling at Boston College and from the EBRI confidence surveys.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Credit card fraud (lifetime, US adult)","lifetime_us_adult":0.65}],"regional_breakdown":[{"region":"Bottom income third","probability":0.55,"notes":"Low-income households have the least savings and highest NRRI risk; many rely almost entirely on Social Security"},{"region":"Middle income third","probability":0.4,"notes":"Middle-income households are the core of the NRRI shortfall — enough income to expect a standard of living above Social Security, not enough savings to fund it"},{"region":"Top income third","probability":0.2,"notes":"High earners are less likely to fall short in absolute terms, but the NRRI still finds ~20% at risk, often due to high pre-retirement spending and low savings rates relative to income"}],"personal_factor_multipliers":[{"factor":"no employer retirement plan","multiplier":2.5,"notes":"Workers without access to a 401(k) or pension have dramatically lower savings rates; EBRI data shows retirement account ownership drops below 30% for this group"},{"factor":"household income top quintile","multiplier":0.4,"notes":"High earners are more likely to meet savings targets, though the NRRI still finds ~20% at risk even in this group"},{"factor":"homeowner with paid-off mortgage","multiplier":0.6,"notes":"Eliminated housing costs substantially reduce retirement spending needs; the NRRI's home equity assumption reflects this"},{"factor":"single-earner household","multiplier":1.5,"notes":"Single individuals and single-earner households face higher risk due to lack of income diversification and survivor benefits"},{"factor":"no Social Security credits (non-covered worker)","multiplier":2,"notes":"Workers in non-covered employment (some state/local government) lack the Social Security floor; their shortfall risk is substantially higher"}],"short_label":"Retirement shortfall","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"The 39% NRRI figure uses a specific definition of \"shortfall\" — falling more than 10% below a target replacement rate derived from pre-retirement spending. It assumes households work to 65 and take a reverse mortgage, which almost nobody actually does. A Vanguard study cited by the Center for Retirement Research found that at least 70% of households would fall short when housing equity is excluded. The NRRI also captured a historically unusual moment: the 2022 SCF reflected peak pandemic-era home prices and stock market gains. The improvement from 47% to 39% may not persist. \"Comfortable retirement\" is inherently subjective — research consistently shows that spending declines 5-15% in the first few years of retirement and continues declining with age, meaning many households that look \"at risk\" by pre-retirement spending standards may adapt successfully. Social Security provides a floor that prevents destitution for most Americans, replacing 36-40% of pre-retirement earnings for average workers. The fear is overrated for high earners (who have substantial buffers even if below benchmarks) and underrated for low earners (who have the smallest savings AND the least awareness of the problem).\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A half-empty hourglass beside a small coin stack, muted grey and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/retirement-savings-shortfall","api_url":"https://likelier.app/api/fears/retirement-savings-shortfall.json"},{"slug":"antibiotic-resistance-infection","question":"What is the risk of developing a serious antibiotic-resistant infection?","quick_answer":"About 1 in 2.6 US adults will have a serious antibiotic-resistant infection at some point in life, from roughly 2.8 million cases a year; the fatal version is far rarer, about 1 in 167. Most people treat resistance as a distant policy problem rather than a personal hazard, so the felt risk sits far below the actual one.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most people understand that antibiotic resistance is a problem in a vague, headline-level way — something that will matter in the future, mostly in hospitals, mostly to other people. The idea that more than one in three US adults will have a serious drug-resistant infection at some point in their lifetime is almost entirely absent from lay risk perception. Resistance is framed as a policy failure, not a personal hazard.\n","kind":"intuition"},"native":{"display":"~2.8 million antibiotic-resistant infections per year in the US","numerator":2800000,"denominator":335000000,"unit":"per year","population":"US residents"},"normalized":{"lifetime_us_adult":0.391,"display":"~1 in 2.6 lifetime (US adult)","log_value":-0.41,"assumptions":"Annual rate: 2,800,000 / 335,000,000 = 0.008358 per person per year. Compounded over 59 adult years: 1 - (1 - 0.008358)^59 = 1 - (0.99164)^59 ≈ 1 - 0.609 = 0.391. This is the probability of experiencing at least one serious AR infection over a lifetime, treating each year as an independent draw at the population rate. The 35,000 annual deaths from AR infections give a separate lifetime fatal AR probability of ~0.006 (1 in 167). The headline figure covers serious infections (not all episodes result in hospitalization), per CDC's 2019 AR Threats Report definition.\n","uncertainty":{"low":0.3,"high":0.47},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/antimicrobial-resistance/data-research/threats/index.html","title":"Antibiotic Resistance Threats in the United States, 2019","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"More than 2.8 million antibiotic-resistant infections occur in the US each year, and more than 35,000 people die as a result","excerpt":"\"More than 2.8 million antibiotic-resistant infections occur in the U.S. each year, and more than 35,000 people die as a result. When Clostridioides difficile — a bacterium that is not typically resistant but can cause deadly diarrhea and is associated with antibiotic use — is added to these, the total reaches 3 million infections and 48,000 deaths.\"\n","source_date":"2019-11-13","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260417133122/https://www.cdc.gov/antimicrobial-resistance/data-research/threats/index.html","calculation_notes":"CDC's 2019 AR Threats Report is the most comprehensive national estimate of antibiotic-resistant infection burden. The 2.8 million figure covers 18 pathogens classified as urgent, serious, or concerning threats and is used as the native numerator. The 35,000 deaths figure anchors the fatal sub-probability in the assumptions. Subsequent CDC data (2021-2022 update) found that six hospital-onset resistant infections rose 20% during the COVID-19 pandemic peak, suggesting the 2.8M baseline is a floor, not a ceiling, for post-pandemic years.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10679183/","title":"Burden of Bacterial Antimicrobial Resistance in United States in 2019: A Systematic Analysis","publisher":"Open Forum Infectious Diseases / Oxford University Press","source_type":"peer_reviewed","statistic":"60,813 deaths associated with and 14,987 deaths attributable to bacterial AMR in bloodstream infections in the US in 2019","excerpt":"\"In the US, there were 60,813 deaths (95% uncertainty interval [UI]: 32,520–102,231) associated with and 14,987 deaths (95% UI: 7,712–25,156) attributable to bacterial AMR in blood stream infection highest in 2019.\" The paper concludes: \"AMR is a serious burden in the United States, with millions of people acquiring antibiotic-resistant infections each year and tens of thousands dying as a result.\"\n","source_date":"2023-11-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426193722/https://pmc.ncbi.nlm.nih.gov/articles/PMC10679183/","calculation_notes":"This independent systematic analysis focuses on bloodstream infections (BSI) specifically, finding 60,813 deaths associated with and 14,987 deaths attributable to bacterial AMR in BSI in 2019. The scope is narrower than CDC's all-infection estimate (2.8M infections across 18 pathogens), but the BSI-specific death toll (60,813 associated) exceeds CDC's 35,000 attributable deaths across all infection types, reflecting different attribution methods. The peer-reviewed design — drawing on hospital discharge records, microbiology lab reports, and vital statistics — is methodologically independent of CDC's surveillance network. Used here as corroborating evidence for the severity of AR burden rather than the native numerator.\n","independence_note":"Methodologically independent of CDC's Threats Report: different data sources (hospital discharge records vs. sentinel surveillance) and different attribution framework, making this a genuine cross-check rather than a recount of the same pipeline.\n"}],"comparison_anchors":[{"label":"Lifetime HAI (any, US adult)","lifetime_us_adult":0.032},{"label":"Lifetime sepsis (US adult)","lifetime_us_adult":0.25},{"label":"Lifetime death from AR infection (US adult)","lifetime_us_adult":0.006}],"short_label":"Drug-resistant infection","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 2.8 million figure is a count of infection *events*, not unique persons — a single individual can contribute multiple episodes in a year, so dividing by the US population overstates the per-person annual probability. The resulting \"more than one in three US adults\" lifetime figure is therefore an upper bound on personal risk, not a direct per-person estimate. The compound- probability calculation further treats annual incidence as independent across years, which overstates true lifetime risk if a prior infection confers any durable immunity or if recurrent infections cluster in a subset of high-risk patients. Conversely, the calculation uses the US-wide population rate, which understates risk for frequent healthcare users, residents of long-term care facilities, and immunocompromised individuals. The pandemic-era 20% surge in hospital-onset resistant infections suggests the 2019 baseline may undercount the current burden. CDC's 2021-2022 update is narrower in scope (six pathogens, hospital-onset only) and cannot directly replace the 2019 estimate.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A flat vector illustration of a single bacterial cell with a crossed-out antibiotic capsule nearby, rendered in muted tones."},"canonical_url":"https://likelier.app/antibiotic-resistance-infection","api_url":"https://likelier.app/api/fears/antibiotic-resistance-infection.json"},{"slug":"online-scam-loss","question":"What are the odds of losing money to an online scam?","quick_answer":"About 40% of US adults will lose money to an online scam at some point, and roughly 1 in 100 report a fraud loss in a given year. Unusually, public worry (about 1 in 3) roughly matches the measured risk.\n","category":"tech","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Online scams are one of only two crime types where a majority of Americans report frequent or occasional worry. In Gallup's October 2025 crime poll, 53% of US adults said they worry about being tricked into providing financial information to scammers. AARP's 2026 fraud survey found an average worry level of 7.1 out of 10, with 38% of adults saying they have already experienced fraud. The fear is widespread, cuts across age groups, and — unusually for this site — roughly matches the measured prevalence.\n","rough_estimate":"~1 in 3 lifetime feels about right to most respondents","kind":"poll","survey_source":{"title":"Fraud Crisis in America in 2026: Many Are Aware and Vulnerable","publisher":"AARP","url":"https://www.aarp.org/pri/topics/work-finances-retirement/fraud-consumer-protection/2026-fraud-survey/","year":2026}},"native":{"display":"~1 in 100 per year (US adults reporting a fraud loss)","numerator":1,"denominator":100,"unit":"per year","population":"US adults"},"normalized":{"lifetime_us_adult":0.4,"display":"~40% lifetime (US adult)","log_value":-0.4,"assumptions":"The FTC Consumer Sentinel 2024 Data Book reports 2.6 million fraud reports from consumers in 2024, of which 38% (about 988,000) reported an actual monetary loss. Against ~260 million US adults, that is roughly 0.38% per year who file a report with a loss. But the FTC captures only a fraction of fraud: AARP's nationally representative 2025 survey found that 41% of adults say they have experienced fraud at some point, and that figure rose to 38% in the 2026 wave using a slightly different question frame. Naively compounding even a conservative 1% annual fraud-loss hazard over 59 years of remaining adult life gives 1 − (1 − 0.01)^59 ≈ 45%. But fraud victimization is not independent year to year — some people are repeatedly targeted, and some demographics are structurally lower-risk — so the effective lifetime rate saturates below the naive compound. A central estimate of 40% is consistent with the AARP measured prevalence (38-41%) and with the naive-compounding upper bound. The uncertainty band is wide because \"fraud loss\" ranges from a $50 phishing charge the bank reverses to a six-figure investment scam.\n","uncertainty":{"low":0.25,"high":0.55},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","title":"Consumer Sentinel Network Data Book 2024","publisher":"US Federal Trade Commission","source_type":"govt_report","statistic":"2.6 million fraud reports in 2024; 38% reported a monetary loss; total reported losses exceeded $12.5 billion, a 25% increase over 2023","excerpt":"\"During 2024, Sentinel received 6.5 million consumer reports … the top three Sentinel report categories were Credit Bureaus and Information Furnishers (21% of all reports), Identity Theft (18%), and Imposter Scams (13%). … In 2024, there were more than 1.1 million reports of identity theft received through the FTC's IdentityTheft.gov website.\"\n","source_date":"2025-03-10","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165314/https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","calculation_notes":"FTC reports 2.6 million fraud reports with 38% indicating a loss, giving ~988,000 loss reports per year. Against ~260M US adults, that is ~0.38% per year who file a federal fraud report with a loss. This is a floor: most fraud victims never file with the FTC. The $12.5 billion total / 988,000 loss reports gives a mean reported loss of ~$12,700, but this is heavily skewed by investment fraud ($5.7B alone); the median is far lower. Used as the authoritative floor for the per-year rate.\n","independence_note":"FTC Sentinel is fed by consumer-initiated reports and partner agencies. Methodologically independent from the AARP survey, which is a nationally representative panel.\n"},{"url":"https://www.aarp.org/money/scams-fraud/fraud-awareness-survey-2025/","title":"AARP Survey: 40 Percent of Adults Lose Money to Fraud","publisher":"AARP","source_type":"reputable_reference","statistic":"41% of American adults (estimated 110.1 million people) report having experienced fraud; 59% are significantly worried about fraud crimes","excerpt":"\"That's huge. Ten years ago, the presumptions were that maybe about 15 percent of people had experienced fraud.\"\n","source_date":"2025-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260311210939/https://www.aarp.org/money/scams-fraud/fraud-awareness-survey-2025/","calculation_notes":"AARP's 2025 nationally representative survey of US adults found 41% self-report having experienced fraud. This is the single best measured-lifetime figure available and anchors the central estimate. The 41% is a cumulative prevalence across all ages in the sample, so for a current 18-year-old the eventual lifetime figure will be higher; for a current 70-year-old, it is roughly the final number. The 40% central estimate is a round figure consistent with both the AARP measured prevalence and the naive annual-hazard compound.\n","independence_note":"AARP survey is a nationally representative online panel (NORC AmeriSpeak or similar), methodologically independent from the FTC Sentinel complaint database.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"53% of US adults worry frequently or occasionally about being tricked into providing financial information to scammers (October 2025)","excerpt":"\"Overall, Americans worry most about being the victim of identity theft (69%) and being tricked into providing financial information to scammers (53%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172538/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for the perceived-risk side only. Scam worry at 53% is the second-highest item on Gallup's crime-worry list, behind only identity theft (69%). These two financial-crime items are the only ones where a majority of Americans express frequent or occasional worry.\n","independence_note":"Gallup telephone polling, entirely separate from FTC complaint data and AARP panel survey. Used only for the perceived-risk axis (measures worry, not incidence) and does not feed into the probability estimate.\n"},{"url":"https://www.ftc.gov/news-events/news/press-releases/2024/12/new-ftc-data-show-skyrocketing-consumer-reports-about-game-online-job-scams","title":"New FTC Data Show Skyrocketing Consumer Reports About Game-Like Online Job Scams","publisher":"US Federal Trade Commission","source_type":"govt_report","statistic":"Task-scam reports rose from zero in 2020 to 5,000 in 2023, then quadrupled to ~20,000 in the first half of 2024; task scams accounted for nearly 40% of 2024 job-scam reports; cryptocurrency losses to job scams hit $41M in the first half of 2024","excerpt":"\"According to a new FTC data spotlight, these scams, known as task scams, have increased massively in the last four years, with reports of these scams increasing from zero in 2020 to 5,000 in 2023, then quadrupling to about 20,000 in just the first half of 2024. … Overall reported losses on job scams tripled from 2020 to 2023 and were more than $220 million in just the first six months of 2024. Task scams were estimated to account for nearly 40 percent of the 2024 job scam reports. … Cryptocurrency is the payment of choice for these scams.\"\n","source_date":"2024-12-19","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20251215220017/https://www.ftc.gov/news-events/news/press-releases/2024/12/new-ftc-data-show-skyrocketing-consumer-reports-about-game-online-job-scams","calculation_notes":"Mechanism-trend evidence only — this is NOT a base-rate input. The spotlight reports the number of task-scam complaints (counts) and aggregate dollar losses, but no exposed-population denominator, so it cannot yield a per-adult-per-year hazard for task scams specifically. It documents that the gamified task-scam mechanism (fake \"commission\" work that flips into a crypto-deposit trap) is one of the fastest-growing fraud vectors and that cryptocurrency is its dominant payment rail, reinforcing the existing \"Active cryptocurrency user\" personal factor. Used to characterize the composition and growth of online-scam loss, not to move the headline 40% lifetime estimate.\n","independence_note":"FTC Consumer Sentinel complaint data, methodologically independent from the AARP panel survey. Same data lineage as the 2024 Data Book source but a distinct, scam-mechanism-specific release.\n"},{"url":"https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","title":"2024 Internet Crime Report","publisher":"Federal Bureau of Investigation, Internet Crime Complaint Center (IC3)","source_type":"govt_report","statistic":"859,532 internet crime complaints received in 2024, of which 256,256 involved a reported monetary loss; total reported losses of $16.6 billion, a 33% increase over 2023","excerpt":"\"IC3 received 859,532 cybercrime complaints in 2024, of which 256,256 complaints involved actual losses.\"\n","source_date":"2025-04-23","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260709082752/https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","calculation_notes":"IC3 is a separate federal complaint channel from the FTC's Consumer Sentinel Network (cited above); both are complaint-based floors, not full-population surveys. Used in the body text as a second, independent federal data point corroborating the scale of reported fraud/cybercrime losses; not used in the native/normalized lifetime-probability calculation, which relies on the AARP measured-prevalence figure and the FTC per-year hazard rate.\n","independence_note":"FBI IC3 complaint data is methodologically independent from both the FTC Consumer Sentinel Network and the AARP survey panel; some victims report to more than one channel, so the two complaint counts are not additive.\n"}],"comparison_anchors":[{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Age 20-29","multiplier":1.3,"notes":"Younger adults report losing money to scams more frequently than older adults, per FTC 2024 data"},{"factor":"Age 70+","multiplier":0.8,"notes":"Older adults report fewer fraud incidents, but median loss per incident is 2-4x higher ($1,000-$1,650 vs $417 for age 20-29)"},{"factor":"Active cryptocurrency user","multiplier":2,"notes":"Crypto-related fraud complaints drove $9.3B of $16.6B in IC3-reported losses in 2024; investment fraud via crypto is the dominant loss category. The FTC's December 2024 task-scam spotlight reports that cryptocurrency is the payment of choice for gamified job/task scams, with crypto losses to job scams hitting $41M in the first half of 2024 alone"},{"factor":"No online financial accounts","multiplier":0.3,"notes":"Most reported fraud involves online payment, bank transfer, or crypto; minimal digital financial footprint substantially reduces exposure"}],"short_label":"Online scam loss","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is not a death risk; it is the probability that a US adult loses money to fraud at least once during their adult life. The number is highly sensitive to what counts as a \"loss.\" The FTC floor (~0.38% per year filing a report) counts only cases serious enough to prompt a federal complaint. The AARP survey ceiling (~41% lifetime) includes any self-reported fraud experience, down to a small unauthorized charge the bank reversed. The 40% central estimate sits in the middle of this definitional range. Investment fraud and romance scams drive the largest per-victim losses (median >$9,000 for investment scams per FTC 2024), while phishing and imposter scams drive the highest case counts. The demographic pattern inverts the stereotype: adults aged 20-49 report more fraud incidents than those 60+, but older victims lose substantially more per incident (median $1,650 for age 80+ vs $417 for age 20-29). Individual behavior — particularly around cryptocurrency, social-media engagement, and responses to unsolicited contact — moves the personal risk meaningfully in both directions. One of the fastest-growing mechanisms is the gamified \"task scam\": a victim is recruited by an unsolicited text or WhatsApp message into fake online \"commission\" work, receives small early payouts to build trust, then is asked to deposit their own money (usually in cryptocurrency) to unlock the next set of tasks, at which point the funds are lost. Per the FTC's December 2024 spotlight, task-scam reports went from zero in 2020 to about 20,000 in the first half of 2024 and made up nearly 40% of that year's job-scam reports, but these figures are complaint counts with no exposed-population denominator, so they cannot be converted into a standalone per-adult task-scam probability.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":5,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single fishing hook dangling from a thin line against a pale grey background, flat vector illustration."},"canonical_url":"https://likelier.app/online-scam-loss","api_url":"https://likelier.app/api/fears/online-scam-loss.json"},{"slug":"water-scarcity-experience","question":"What are the odds of experiencing severe water scarcity in your lifetime?","quick_answer":"Averaged globally, about 2 in 5 adults will experience at least one episode of severe water scarcity in their lifetime. Most people in water-secure countries assume their personal odds are near-zero, while individual risk ranges from negligible in Scandinavia to near-certain across the MENA region and parts of South Asia.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Water scarcity occupies an odd perceptual niche: residents of well-supplied cities treat it as a problem that happens somewhere else, while people in arid regions have already normalised seasonal rationing to the point where it barely registers as noteworthy. In high-income countries the dominant mental model is \"Cape Town Day Zero\" or \"Flint Michigan\" — dramatic, localised crises rather than the slow, structural tightening of supply-demand balances that hydrologists actually track. Climate change discourse has raised general awareness that \"water will be a problem,\" but the typical estimate of personal exposure is anchored on current tap reliability. The result is systematic underestimation: most adults in water-secure regions assume the odds are negligible for them, while global averages are pulled sharply upward by billions of people already living in basins where demand routinely exceeds 80% of renewable supply.\n","rough_estimate":"Most people in water-secure countries assume near-zero personal risk; global awareness is higher but vague","kind":"intuition"},"native":{"display":"WRI Aqueduct 4.0: 25 countries (25% of world population) face extremely high water stress annually; 4 billion people experience water stress at least 1 month/year","numerator":2000000000,"denominator":8000000000,"unit":"per year","population":"global population experiencing extremely high water stress or acute shortage episodes annually"},"normalized":{"lifetime_us_adult":0.4,"display":"~2 in 5 lifetime (global adult, at least one severe water scarcity episode)","log_value":-0.4,"assumptions":"The normalized figure estimates the probability that a randomly selected global adult will experience at least one episode of severe water scarcity — defined as mandatory rationing, supply disruption exceeding one week, or residence in an area where demand exceeds 80% of renewable freshwater supply — over a remaining lifetime of 59 years. WRI Aqueduct 4.0 (2023) finds that 25 countries housing 25% of the world's population already face extremely high water stress annually, using over 80% of their renewable supply. More broadly, approximately 4 billion people (50% of the global population) experience water stress for at least one month per year. The WHO/UNICEF JMP 2025 report finds 2.1 billion people (26% of global population) still lack safely managed drinking water. For \"severe\" episodes specifically — acute shortages causing rationing or extended disruption — a conservative annual exposure rate of approximately 8-10% of the global adult population is used as the baseline, reflecting the fraction in extremely high stress zones who experience actual supply failures in a given year (not merely living in a stressed basin). With WRI projecting that the population under extreme water stress grows to roughly 35-40% by 2050 under moderate climate scenarios, the average annual severe-episode probability over a 59-year horizon is approximately 0.9% per year. Compounded: 1 - (1 - 0.009)^59 ≈ 0.41, rounded to 0.40. This is a population-weighted global average; individual risk ranges from near-zero (Scandinavia, Canada) to near-certainty (MENA, Sahel, parts of South Asia).\n","uncertainty":{"low":0.25,"high":0.55},"scope":"global_adult_lifetime"},"sources":[{"url":"https://www.wri.org/insights/highest-water-stressed-countries","title":"25 Countries, Housing One-Quarter of the Population, Face Extremely High Water Stress","publisher":"World Resources Institute","source_type":"reputable_reference","statistic":"25 countries face extremely high water stress annually, using over 80% of renewable water supply; 4 billion people experience water stress at least 1 month/year; an additional 1 billion projected to live with extremely high water stress by 2050","excerpt":"\"25 countries — housing one-quarter of the world's population — face extremely high water stress each year, regularly using up almost their entire available water supply. And at least 50% of the world's population — around 4 billion people — live under highly water-stressed conditions for at least one month of the year.\"\n","source_date":"2023-08-16","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260425052707/https://www.wri.org/insights/highest-water-stressed-countries","calculation_notes":"WRI Aqueduct 4.0 provides the structural baseline for this entry. The \"extremely high water stress\" threshold (>80% of renewable supply consumed) identifies populations where demand-supply imbalance makes acute shortage episodes structurally likely. The 25% figure (approximately 2 billion people) represents the population in countries where the annual average already exceeds this threshold. The broader 4 billion / 50% figure captures seasonal stress. The entry uses the 25% extremely-high-stress population as the primary anchor for severe-episode exposure, then adjusts upward over the 59-year horizon to account for WRI's projection that an additional 1 billion people will live with extremely high water stress by 2050 under moderate scenarios. Note: the native numerator/denominator (2B/8B) represents the population in extremely-high-stress countries, not the annual severe-episode rate. The lifetime calculation uses a derived annual severe-episode rate of ~0.9% (see normalized assumptions).\n"},{"url":"https://www.who.int/news/item/26-08-2025-1-in-4-people-globally-still-lack-access-to-safe-drinking-water---who--unicef","title":"1 in 4 people globally still lack access to safe drinking water — WHO, UNICEF","publisher":"World Health Organization / UNICEF","source_type":"govt_report","statistic":"2.1 billion people (26% of global population) lack safely managed drinking water as of 2024; 106 million drink from untreated surface sources","excerpt":"\"Despite gains since 2015, 1 in 4 — or 2.1 billion people globally — still lack access to safely managed drinking water, including 106 million who drink directly from untreated surface sources.\"\n","source_date":"2025-08-26","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260413071730/https://www.who.int/news/item/26-08-2025-1-in-4-people-globally-still-lack-access-to-safe-drinking-water---who--unicef","calculation_notes":"The JMP 2025 report provides the demand-side complement to WRI's supply-side stress data. The 2.1 billion figure captures people who lack reliably safe water services — a broader category than acute scarcity but strongly correlated with vulnerability to shortage episodes. The 106 million using untreated surface water represent the most extreme end of the spectrum. These figures inform the regional breakdown and confirm that the entry's central estimate is not driven solely by supply-side stress metrics but also by infrastructure and access deficits that amplify vulnerability to scarcity events.\n"},{"url":"https://www.nature.com/articles/s41467-021-25026-3","title":"Future global urban water scarcity and potential solutions","publisher":"Nature Communications","source_type":"peer_reviewed","statistic":"Global urban population facing water scarcity projected to increase from 933 million (2016) to 1.7-2.4 billion (2050); number of large cities exposed to water scarcity projected to increase from 193 to 284","excerpt":"\"The global urban population facing water scarcity is projected to double from 933 million (one third of the global urban population) in 2016 to 1.693–2.373 billion (one third to nearly half of the global urban population) in 2050.\"\n","source_date":"2021-08-03","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260122102801/https://www.nature.com/articles/s41467-021-25026-3","calculation_notes":"He et al. (2021) provides the forward projection that anchors the entry's growth trajectory. The doubling of urban water-scarce population from ~1 billion to ~2 billion by 2050 is consistent with WRI's projection that stressed population grows from 50% to ~60% globally. The urban focus is particularly relevant because urban water scarcity manifests as the rationing and supply disruption that the entry defines as \"severe\" — infrastructure-mediated shortages are more acutely felt than gradual agricultural stress. The 2050 projection under SSP2-RCP6.0 (moderate scenario) was used to calibrate the mid-horizon annual probability increase in the normalized calculation.\n"},{"url":"https://www.lincolninst.edu/publications/land-lines-magazine/articles/land-water-impacts-data-centers/","title":"Data Drain: The Land and Water Impacts of the AI Boom","publisher":"Lincoln Institute of Land Policy","source_type":"reputable_reference","statistic":"Per-facility data-center water use is hard to confirm because companies are less than forthcoming about consumption; a single Meta facility in Newton County, Georgia uses 500,000 gallons per day, and Texas data centers are projected to use 49 billion gallons in 2025 rising to 399 billion by 2030","excerpt":"\"It can be hard to confirm specific data about data centers\" because of proprietary concerns and potential backlash, with companies \"less than forthcoming about how much water their data centers consume.\" A Meta facility in Newton County, Georgia uses \"500,000 gallons of water per day.\"\n","source_date":"2025-10-17","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260709165420/https://www.lincolninst.edu/publications/land-lines-magazine/articles/land-water-impacts-data-centers/","calculation_notes":"This source documents the AI data-center water-drawdown vector noted in the caveats and intentionally does NOT alter the headline number. Aggregate sectoral figures are well-reported (US data centers consumed roughly 17.5 billion gallons in 2023 per IEEE Spectrum), but per-community attribution — tying a specific town's rationing or supply disruption to a specific facility — is contested and opaque, because operators do not transparently report facility-level municipal draw. The vector is therefore treated as a no-reliable-estimate causal driver layered on top of the existing supply-stress and access-deficit baseline, not as an independent probability input. It is cited to make the contested attribution explicit, per the trust-stack rule against fabricating a number where evidence is weak.\n"}],"comparison_anchors":[{"label":"Death from drought/famine (lifetime, global adult)","lifetime_us_adult":0.0004},{"label":"Experiencing a natural disaster (lifetime, US adult)","lifetime_us_adult":0.33},{"label":"Climate change death (lifetime, global adult)","lifetime_us_adult":0.0049},{"label":"Experiencing a flood (lifetime, global adult)","lifetime_us_adult":0.1}],"regional_breakdown":[{"region":"Middle East & North Africa (MENA)","probability":0.95,"notes":"Most water-stressed region globally; 12 of the 25 extremely-high-stress countries are in MENA; rationing is already routine in many cities"},{"region":"South Asia","probability":0.65,"notes":"India alone projected to add 153-422 million water-scarce urban residents by 2050; monsoon variability and groundwater depletion drive acute episodes"},{"region":"Sub-Saharan Africa","probability":0.55,"notes":"Highest rates of lacking safely managed water services; population growth outpacing infrastructure investment"},{"region":"Western Europe / North America","probability":0.08,"notes":"Localised drought episodes (US Southwest, Mediterranean) but strong infrastructure buffers most of the population from severe shortage"},{"region":"East Asia & Pacific","probability":0.3,"notes":"China's northern plains face severe stress; Pacific islands vulnerable to saltwater intrusion; southern regions better supplied"}],"personal_factor_multipliers":[{"factor":"resident of MENA or arid South Asian city","multiplier":2.4,"notes":"structural water deficit makes severe shortage episodes near-certain over a lifetime"},{"factor":"rural Sub-Saharan Africa, no piped water","multiplier":2,"notes":"dependence on seasonal sources and hand-carried water amplifies vulnerability to drought"},{"factor":"resident of water-abundant high-income country (Scandinavia, Canada)","multiplier":0.05,"notes":"abundant renewable supply and robust infrastructure make severe scarcity episodes extremely unlikely"},{"factor":"resident of US Southwest (Arizona, Nevada, Southern California)","multiplier":1.5,"notes":"Colorado River basin stress, groundwater depletion, and growing population increase exposure relative to US average"}],"short_label":"Water scarcity","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"existential","valence":"negative","caveats":"\"Severe water scarcity\" is defined here as experiencing at least one episode of mandatory rationing, supply disruption exceeding one week, or residence in an area where demand routinely exceeds 80% of renewable freshwater supply. This is a lower bar than life-threatening dehydration and a higher bar than merely living in a \"water-stressed\" basin (which already captures half the world's population seasonally). The 40% central estimate is a global population-weighted average that conceals enormous heterogeneity: the figure is above 90% for MENA residents and below 10% for Scandinavians. The trajectory depends heavily on climate pathway, infrastructure investment, and population growth — WRI's projections span from optimistic (SSP1-RCP2.6) to pessimistic (SSP5-RCP8.5) with roughly a 1.5x spread in 2050 stressed population. Groundwater depletion, which is poorly captured in surface-water stress models, could push actual scarcity episodes higher than the stress indicators suggest — aquifer drawdown in India's Punjab and the US Ogallala is already generating localised shortages not fully reflected in basin-level stress ratios. Conversely, desalination capacity is expanding rapidly in MENA and may reduce acute episodes faster than the projections assume. A newer and contested driver is AI data-center cooling drawdown. Aggregate figures are large and well-reported — US data centers consumed roughly 17.5 billion gallons in 2023 (IEEE Spectrum), and a Guardian/NOAA analysis found that about two-thirds of 809 planned US facilities are sited in areas that were in drought over the prior year — but attributing a specific community's rationing or supply disruption to a specific facility remains unreliable, because operators are, in the Lincoln Institute's words, \"less than forthcoming about how much water their data centers consume.\" Documented incidents exist (a Georgia campus drew 29 million gallons over 15 months, discovered only when residents reported low pressure), but no defensible per-community probability multiplier can be derived from the contested aggregate data, so this entry treats the data-center vector as a qualitative aggravator of local supply stress rather than a numerical input.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"extracted-from-transcript","scored_at":"2026-04-27","methodology_version":"1.0"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A single water droplet shape against a dry, cracked earth-toned background, flat vector illustration."},"canonical_url":"https://likelier.app/water-scarcity-experience","api_url":"https://likelier.app/api/fears/water-scarcity-experience.json"},{"slug":"bnpl-missed-payment","question":"How likely is a buy-now-pay-later user to miss a payment?","quick_answer":"About 41% of buy-now-pay-later users — roughly 4 in 10 — miss at least one payment within a year. Because the product carries no interest, users tend to assume their own chance of trouble is near zero.\n","category":"other","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Buy-now-pay-later is marketed as interest-free and frictionless, and many users treat it as a budgeting convenience rather than debt. The popular fear, when it surfaces, is the opposite extreme: a 'debt spiral' of stacked, untrackable loans ('phantom debt') ending in collapse. No rigorous survey of how BNPL users rate their own risk of missing a payment was located, so the perception here is inferred from how the product is framed and used, not measured.","rough_estimate":"Users tend to perceive a near-zero chance of trouble because the product carries no interest; the measured 12-month late-payment rate is about 4 in 10.","kind":"intuition"},"native":{"display":"About 41 in 100 BNPL users missed a payment in the past year","numerator":410,"denominator":1000,"unit":"BNPL users who missed at least one payment in the past 12 months","population":"US adults who currently use buy-now-pay-later"},"normalized":{"lifetime_us_adult":0.41,"display":"About 41% (per-year, among BNPL users)","log_value":-0.387,"assumptions":"Subgroup = US adults who currently use buy-now-pay-later (roughly half of all US adults report having used it). Horizon = the past 12 months, NOT extrapolated to a 59-year remaining-life lifetime: annual repetition of a 0.41 rate would approach certainty, and BNPL pay-in-four is a roughly five-year-old mass-market product with no long-run individual-level history. The headline is therefore a one-year prevalence held as-is under the subgroup_lifetime label. Native: a LendingTree survey (April 2-3 2025, n=2,000 US adults 18-79; about half were BNPL users, so ~1,000 in the denominator) found 41% of BNPL users reported at least one late BNPL payment in the prior 12 months. Expressed per 1,000 BNPL users: 410/1000 = 0.41. Uncertainty band 0.34-0.47 spans the survey-wave trend (34% reported a year earlier; a later LendingTree tracker reported a higher figure near 47%), not a sampling confidence interval; the point estimate 0.41 sits inside it.","uncertainty":{"low":0.34,"high":0.47},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.richmondfed.org/publications/research/economic_brief/2026/eb_26-05","title":"Buy Now, Pay Later: Recent Developments and Implications","publisher":"Federal Reserve Bank of Richmond (Economic Brief No. 26-05, by Zhu Wang)","source_type":"govt_report","statistic":"41% of BNPL users made at least one late payment in the past year (2025 survey), up from 34% a year earlier; BNPL loans are generally not reported to credit bureaus; 2023 charge-off rate 1.83%.","excerpt":"There is indicative evidence from consumer surveys that late payment and delinquency behavior among BNPL users has increased since 2023. For example, LendingTree's 2025 survey reports that 41 percent of BNPL users made at least one late payment in the past year, up from 34 percent a year earlier.","source_date":"2026-02-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260411143352/https://www.richmondfed.org/publications/research/economic_brief/2026/eb_26-05","calculation_notes":"Provides the authoritative, verbatim-confirmed late-payment figure (41%, up from 34%) used as the native numerator basis, and separately the 'lenders generally do not report loan performance to credit bureaus' statement supporting the phantom-debt caveat. Charge-off rate 1.83% (2023) used as the rare-severe-end contrast.","independence_note":"Federal Reserve analysis citing the LendingTree survey and CFPB data; independent of the BNPL industry."},{"url":"https://www.cnbc.com/2025/04/26/americans-groceries-buy-now-pay-later-loans.html","title":"More Americans are financing groceries with buy now, pay later loans — and more are paying those bills late, survey says","publisher":"CNBC","source_type":"news_article","statistic":"Of ~2,000 US adults 18-79 surveyed April 2-3 2025, about half used BNPL; among them 41% made a late payment in the past year (up from 34%), 25% used BNPL for groceries (up from 14% in 2024), and 60% had multiple loans at once with nearly a fourth holding three or more.","excerpt":"In a survey conducted April 2-3 of 2,000 U.S. consumers ages 18 to 79, around half reported having used buy now, pay later services. Of those consumers, 25% of respondents said they were using BNPL loans to buy groceries, up from 14% in 2024 and 21% in 2023, the firm said. Meanwhile, 41% of respondents said they made a late payment on a BNPL loan in the past year, up from 34% in the year prior, the survey found.","source_date":"2025-04-26","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260313150716/https://www.cnbc.com/2025/04/26/americans-groceries-buy-now-pay-later-loans.html","calculation_notes":"Carries the verbatim survey detail (LendingTree, reported via CNBC) for denominator construction: 2,000 adults, ~half BNPL users (~1,000), 41% late = ~410/1000. Also the grocery-use and loan-stacking context (60% multiple loans, nearly a fourth with three or more) used in prose.","independence_note":"Reports the LendingTree/QuestionPro survey; LendingTree page itself returned 403 live and 503 on Wayback, so the survey is cited through CNBC rather than the primary page."},{"url":"https://www.paymentsdive.com/news/cfpb-buy-now-pay-later-bnpl-use-debt-research-report/737285/","title":"Buy now, pay later users pile on debt, CFPB finds","publisher":"Payments Dive","source_type":"news_article","statistic":"Per the Jan 2025 CFPB report (145M applications 2017-2022): in 2022, 63% of borrowers took out more than one BNPL loan at a time and 33% borrowed from multiple companies; 61% of BNPL users had subprime/deep-subprime scores; ~2% default rate 2019-2022.","excerpt":"The average number of transactions per borrower increased from 8.5 in 2021 to 9.5 in 2022. And in 2022, 63% of borrowers took out more than one BNPL loan at a time, and 33% of borrowers who use BNPL took out loans from multiple companies.","source_date":"2025-01-14","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260512194038/https://www.paymentsdive.com/news/cfpb-buy-now-pay-later-bnpl-use-debt-research-report/737285/","calculation_notes":"Verbatim loan-stacking figures (63% multiple simultaneous loans, 33% from multiple companies) supporting the 'untrackable stacked loans' prose; ~2% default rate corroborates the low realized-loss contrast. Reports the CFPB 'Consumer Use of Buy Now, Pay Later and Other Unsecured Debt' study.","independence_note":"News summary of the authoritative CFPB report; CFPB report PDF downloaded but not machine-readable in this environment, so cited through Payments Dive plus the CFPB landing pages below."},{"url":"https://www.consumerfinance.gov/data-research/research-reports/consumer-use-of-buy-now-pay-later-insights-from-the-cfpb-making-ends-meet-survey/","title":"Consumer Use of Buy Now, Pay Later: Insights from the CFPB Making Ends Meet Survey","publisher":"Consumer Financial Protection Bureau","source_type":"govt_report","statistic":"BNPL borrowers are, on average, much more likely to be highly indebted, revolve credit-card balances, have delinquencies in traditional credit products, and use payday/pawn/overdraft than non-BNPL borrowers.","excerpt":"BNPL borrowers were, on average, much more likely to be highly indebted, revolve on their credit cards, have delinquencies in traditional credit products, and use high-interest financial services such as payday, pawn, and overdraft compared to non-BNPL borrowers.","source_date":"2023-03-02","source_accessed":"2026-06-13","calculation_notes":"Authoritative basis for the caveat that the BNPL-user subgroup is already financially stretched, so a missed BNPL payment co-occurs with broader debt stress rather than being an isolated lapse. Supports the heterogeneity paragraph."},{"url":"https://www.consumerfinance.gov/data-research/research-reports/the-buy-now-pay-later-market/","title":"The Buy Now, Pay Later Market","publisher":"Consumer Financial Protection Bureau","source_type":"govt_report","statistic":"BNPL market grew 2019-2023; CFPB obtained pay-in-four data from six large BNPL companies covering loan volume, users, frequency, average loan size, late fees, and charge-off rates.","excerpt":"The report includes key market metrics including BNPL loan volume, number of users, frequency of use, average loan size, late fees, and charge-off rates.","source_date":"2025-12-10","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714061956/https://www.consumerfinance.gov/data-research/research-reports/the-buy-now-pay-later-market/","calculation_notes":"Authoritative market-scale and late-fee/charge-off context (Dec 2025 report) framing the supply side; corroborates that the six-lender supervisory dataset underlies the Richmond Fed and Payments Dive figures."}],"comparison_anchors":[{"label":"Losing money to an online scam","lifetime_us_adult":0.4},{"label":"Home burglary (US, lifetime)","lifetime_us_adult":0.39},{"label":"Personal bankruptcy (lifetime)","lifetime_us_adult":0.1}],"short_label":"BNPL missed payment","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is a one-year prevalence among current BNPL users, not a lifetime probability for a general US adult; the subgroup_lifetime scope holds the 12-month figure as-is rather than extrapolating it, because annual repetition would approach certainty and BNPL pay-in-four has no long individual-level track record. The headline outcome is a missed payment, which the survey's own analyst characterized as mostly late 'by no more than a week or so' — a minor, usually recoverable event, distinct from the rarer ~2% default and the much rarer bankruptcy that the 'debt spiral' fear evokes. The 41% figure comes from a single LendingTree/QuestionPro survey (n=2,000; about half BNPL users) reported via CNBC and confirmed verbatim by a Federal Reserve Bank of Richmond brief; the LendingTree page itself could not be opened (403 live, 503 on Wayback), so it is cited secondhand. A later LendingTree tracker reportedly put the figure near 47%, hence the upper bound. 'Phantom debt' is a real structural feature: most pay-in-four loans were historically not reported to credit bureaus, so missed payments and stacked balances are largely invisible in standard credit data — a measurement gap, not a clean probability. 8D self-score average 4.0 with all dimensions at 3 or above; the weakest dimension is scope (D5=3), because pinning a 12-month survey prevalence onto the lifetime axis is inherently strained and is handled by the no-extrapolation assumption rather than a conversion.","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":3,"d6":4,"d7":4,"d8":4,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"BNPL missed payment\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/bnpl-missed-payment","api_url":"https://likelier.app/api/fears/bnpl-missed-payment.json"},{"slug":"bone-marrow-match-failure","question":"What are the odds of not finding a bone marrow donor match?","quick_answer":"About 42% of patients who need an unrelated transplant fail to find an optimal 8/8 HLA-matched donor. The odds vary sharply by ancestry — roughly 75% of White patients find a match versus about 29% of Black patients — despite the common assumption that a match exists for everyone.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most people who have not personally needed a bone marrow transplant assume the registry system works like a blood bank — show up, get matched, receive a transplant. The reality of HLA matching, where doctors need to align 8 to 10 of 12 genetic markers, is poorly understood. The perception gap runs in both directions depending on the audience: white patients tend to underestimate the difficulty because the registry was built around donors who look like them, while patients from minority communities often learn the hard way that their match probability is dramatically lower. Public awareness campaigns have improved since the early 2000s, but the structural inequity remains largely invisible to the general public.\n","rough_estimate":"Most people assume a match can be found for anyone who needs one","kind":"intuition"},"native":{"display":"~75% match likelihood for White patients; ~29% for Black patients (8/8 HLA match)","numerator":29,"denominator":100,"unit":"match probability for Black patients seeking 8/8 HLA match","population":"US patients searching NMDP registry"},"normalized":{"lifetime_us_adult":0.42,"display":"~42% of patients who need a transplant fail to find an optimal 8/8 HLA-matched unrelated donor (conditional on needing a transplant)","log_value":-0.38,"assumptions":"The NEJM 2014 study by Gragert et al. estimated match likelihoods by race/ethnicity in the US registry. The population-weighted average failure rate for an 8/8 HLA match is roughly 42%, heavily skewed by the low match rates for Black (71% failure), Hispanic (48% failure), and multiracial (71% failure) patients vs White (7% failure). This is a per-search probability, not a cumulative lifetime figure, but it represents the reality facing any patient who needs an unrelated donor transplant.\n","uncertainty":{"low":0.25,"high":0.71},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJMsa1311707","title":"HLA Match Likelihoods for Hematopoietic Stem-Cell Grafts in the U.S. Registry","publisher":"New England Journal of Medicine","source_type":"primary_study","statistic":"8/8 HLA match probability ranges from 75% (Whites of European descent, highest) down to 16% (Blacks of South or Central American descent, lowest); African Americans 19%, Hispanic groups 34-40%, Asian groups 27-42%","excerpt":"\"The likelihood of finding an optimal donor varies among racial and ethnic groups, with the highest probability among whites of European descent, at 75%, and the lowest probability among blacks of South or Central American descent, at 16%. Among African Americans, if the same search strategy is used, 19% will have an 8/8 HLA-matched donor identified, and another 57% will have a 7/8 HLA-matched donor.\"\n","source_date":"2014-07-24","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250503110101/https://www.nejm.org/doi/full/10.1056/NEJMsa1311707","calculation_notes":"Gragert et al. analyzed HLA haplotype frequencies in the NMDP registry against patient need. The 8/8 match is the clinical gold standard for unrelated donor transplant. The racial disparity arises from two compounding factors: (1) greater HLA diversity in populations of African and mixed ancestry, requiring a larger pool to find any given match, and (2) underrepresentation of minority donors in the registry (74% White at time of study). The normalized figure uses a population-weighted average across US racial demographics.\n"},{"url":"https://www.nmdp.org/en/get-involved/join-the-registry/ethnicity-and-diversity-matter","title":"Why Ethnicity Matters for Bone Marrow Transplants","publisher":"NMDP (formerly Be The Match)","source_type":"reputable_reference","statistic":"71% of Black patients, 71% of multiracial patients, 52% of Latino/Hispanic patients, and 53% of Asian American patients lack a perfectly matched donor in the worldwide registry","excerpt":"\"71% of Blacks, 71% of multi-racial individuals, 52% of Latinos and Hispanics, and 53% of Asian Americans do not have a perfectly matched donor in the worldwide registry. A patient's ethnic background is important in predicting the likelihood of finding a match because HLA is inherited.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260207233229/https://www.nmdp.org/en/get-involved/join-the-registry/ethnicity-and-diversity-matter","calculation_notes":"NMDP's current diversity page updates the Gragert et al. data with more recent registry composition. The figures are broadly consistent with the 2014 NEJM study, suggesting that despite recruitment efforts, the structural gap has not closed substantially. The registry now includes over 22 million donors worldwide but remains majority White.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3848311/","title":"Racial disparities in hematopoietic cell transplantation in the United States","publisher":"PMC / Bone Marrow Transplantation","source_type":"peer_reviewed","statistic":"Black patients are less likely to proceed to transplant and have worse outcomes, partly due to lower match availability and longer search times","excerpt":"\"Whites constitute nearly 6.5 million (74%) donors in the registry, whereas the representation of Hispanics (10%), Blacks (7%) and Asians (7%) is less frequent. This underrepresentation compounds the biological challenge of greater HLA diversity in non-White populations.\"\n","source_date":"2013-11-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032419/https://pmc.ncbi.nlm.nih.gov/articles/PMC3848311/","calculation_notes":"This review contextualizes the match-rate disparity within the broader transplant pipeline. Lower match probability leads to longer search times, which leads to disease progression, which leads to worse outcomes or ineligibility for transplant. The disparity is not purely a registry-size problem — HLA diversity means that even a proportionally representative registry would yield lower match rates for Black patients unless the registry were substantially larger overall.\n"}],"comparison_anchors":[{"label":"Blood type compatibility for transfusion (O-negative universal)","lifetime_us_adult":0.93},{"label":"Organ transplant waitlist mortality (kidney, 5-year)","lifetime_us_adult":0.1},{"label":"Finding a 6/6 cord blood match","lifetime_us_adult":0.5}],"personal_factor_multipliers":[{"factor":"White / European ancestry","multiplier":0.6,"notes":"Gragert et al. NEJM 2014: 8/8 HLA match probability for White patients is 75%, implying a 25% failure rate vs the 42% population-weighted baseline — approximately 0.60× the average risk. This advantage reflects both the registry's 74% White composition and lower HLA diversity in European-ancestry populations."},{"factor":"Black / African-American ancestry","multiplier":1.69,"notes":"NMDP current data (2025): 71% of Black patients lack a perfectly matched donor worldwide. Relative to the 42% population-weighted failure baseline this is ~1.7×. Root causes are compounding: greater HLA diversity in populations of African descent, AND underrepresentation of Black donors in the registry (≈7% of donors vs ≈13% of US population)."},{"factor":"Multiracial ancestry","multiplier":1.69,"notes":"NMDP current data (2025): 71% of multiracial individuals lack a perfectly matched donor — same rate as Black patients. Individuals of mixed ancestry often fall between population HLA haplotype clusters, making a perfect match statistically rare even in a large registry."},{"factor":"Hispanic / Latino ancestry","multiplier":1.24,"notes":"NMDP current data (2025): 52% of Latino and Hispanic patients lack a perfectly matched donor. Relative to the 42% population-weighted failure rate this is ~1.24×. Gragert et al. (NEJM 2014) reported 8/8 match probabilities for Hispanic subgroups of 34% (South/Central American) to 37% (Mexican), broadly consistent with the NMDP's updated 52%-lack-a-match figure."}],"short_label":"No marrow match","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 42% average failure rate applies to optimal 8/8 HLA-matched unrelated donors. Recent advances in haploidentical (half-matched family donor) transplantation and improved mismatched unrelated donor protocols have expanded options significantly. NMDP data from 2024 shows that when 5/8 to 7/8 mismatched donors are considered, virtually all patients have at least one suitable donor available. The clinical question has shifted from \"can we find any donor\" to \"can we find the best donor for optimal outcomes.\"\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Abstract puzzle pieces in muted tones with one piece missing, flat vector illustration representing the challenge of finding a donor match."},"canonical_url":"https://likelier.app/bone-marrow-match-failure","api_url":"https://likelier.app/api/fears/bone-marrow-match-failure.json"},{"slug":"divorce","question":"What are the odds of a first marriage ending in divorce?","quick_answer":"About 42% of US first marriages end in divorce — below the durable 50% folk statistic almost everyone cites, and lower still for younger and college-educated couples.\n","category":"other","tags":["relationships"],"no_reliable_estimate":false,"perceived":{"description":"The \"50% of marriages end in divorce\" claim is one of the most durable statistical myths in American culture. It has been repeated so often — in news articles, self-help books, wedding speeches, and casual conversation — that most adults treat it as settled fact. The figure was never a cohort estimate; it came from comparing annual marriages to annual divorces in a calendar year, a method that confuses flows with stocks and overstates the risk for any individual marriage.\n","rough_estimate":"~50% — the folk statistic most people cite","kind":"intuition"},"native":{"display":"~14.2 divorces per 1,000 married women per year (2024)","numerator":142,"denominator":10000,"unit":"per year","population":"US married women age 15+"},"normalized":{"lifetime_us_adult":0.42,"display":"~42% lifetime (US first marriages)","log_value":-0.38,"assumptions":"The native rate is a refined divorce rate (divorces per 1,000 married women per year) from the BGSU National Center for Family & Marriage Research using 2024 ACS data. To estimate a lifetime probability for first marriages, we use life-table methods rather than naive compounding, because the hazard is not constant — it peaks around years 5-8 and declines thereafter. The ~42% central estimate draws on the consensus range in family demography literature (roughly 40-45% for all US first marriages, lower for post-2000 and college-educated cohorts). This is consistent with the observed decline from the peak-divorce era: the refined rate fell from ~22.6 per 1,000 in 1980 to 14.2 in 2024, and younger cohorts are divorcing at lower rates than their parents did. The 42% figure is a population average that masks sharp demographic gradients by education, age at marriage, and birth cohort.\n","uncertainty":{"low":0.35,"high":0.5},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.bgsu.edu/ncfmr/resources/data/family-profiles/fp-25-31.html","title":"Refined Divorce Rate in the U.S.: Geographic Variation, 2024","publisher":"National Center for Family & Marriage Research, Bowling Green State University","source_type":"reputable_reference","statistic":"14.2 women per 1,000 married women aged 15+ divorced in the past 12 months (2024 ACS)","excerpt":"\"With 14.2 women divorcing per 1,000 married women, the U.S refined divorce rate decreased just slightly in 2024 from 14.4 in 2023.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260306062614/https://www.bgsu.edu/ncfmr/resources/data/family-profiles/fp-25-31.html","calculation_notes":"The refined divorce rate (per 1,000 married women) is the standard measure in family demography because it uses the at-risk population as the denominator, unlike the crude rate (per 1,000 total population). The 2024 figure of 14.2 continues a long decline from the 1980 peak of ~22.6. Used as the native annual rate. The lifetime estimate requires life-table methods because the divorce hazard is non-constant across marriage duration; the 42% central estimate comes from the demography consensus described in the assumptions field rather than from naive compounding of this annual rate.\n","independence_note":"BGSU NCFMR profiles are derived from ACS microdata. They are methodologically independent from CDC NCHS vital statistics, which use state-reported counts.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24399141/","title":"Breaking Up Is Hard to Count: The Rise of Divorce in the United States, 1980-2010","publisher":"Demography (Kennedy & Ruggles)","source_type":"peer_reviewed","statistic":"Age-standardized divorce rates doubled among persons over 35 between 1990 and 2008; rates among youngest couples stable or declining","excerpt":"\"Using new data from the American Community Survey and controlling for changes in the age composition of the married population, we conclude that there was actually a substantial increase in age-standardized divorce rates between 1990 and 2008. Divorce rates have doubled over the past two decades among persons over age 35. Among the youngest couples, however, divorce rates are stable or declining.\"\n","source_date":"2014-04-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165854/https://pubmed.ncbi.nlm.nih.gov/24399141/","calculation_notes":"Kennedy & Ruggles demonstrated that crude divorce rate trends are misleading because they fail to account for changes in the age composition of the married population. Their age-standardized analysis showed that the apparent post-1980 decline in divorce masked rising rates among older adults and genuinely declining rates among younger cohorts. This bifurcated trend is critical to interpreting the lifetime probability: the 42% average blends a lower rate for post-2000 marriages (~35%) with higher rates for older cohorts who married in the peak-divorce era.\n","independence_note":"Kennedy & Ruggles used ACS microdata, the same upstream source as BGSU NCFMR profiles, so the two sources are partially dependent. However, the analytical contribution (age-standardization methodology and cohort decomposition) is independent.\n"},{"url":"https://www.cdc.gov/nchs/fastats/marriage-divorce.htm","title":"Marriage and Divorce","publisher":"CDC National Center for Health Statistics","source_type":"govt_report","statistic":"Crude divorce rate: 2.4 per 1,000 population (2023 provisional, 45 reporting states and D.C.)","excerpt":"\"Divorce rate: 2.4 per 1,000 population (45 reporting States and D.C.)\"\n","source_date":"2025-03-17","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260403092616/https://www.cdc.gov/nchs/fastats/marriage-divorce.htm","calculation_notes":"The crude divorce rate (per 1,000 total population) is the figure most commonly cited in news media, but it is the wrong denominator for estimating individual risk — it includes children, never-married adults, and the already-divorced. The crude rate has fallen from 5.0 in 1985 to 2.4 in 2023, partly reflecting genuine decline and partly reflecting the falling share of the population that is married. Used here as a corroborating check and to illustrate why the \"50% myth\" arose: dividing annual divorces by annual marriages in a calendar year is algebraically similar to the crude rate and overstates the lifetime risk for any cohort.\n","independence_note":"CDC NCHS uses state vital statistics reports (administrative records), while ACS-based sources use household survey responses. Different collection pipelines.\n"}],"comparison_anchors":[{"label":"Being a victim of identity theft (lifetime, US)","lifetime_us_adult":0.6},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"College-educated, married after 2000","multiplier":0.65,"notes":"Roughly 25-30% lifetime divorce probability vs 42% population average"},{"factor":"No college degree","multiplier":1.4,"notes":"Roughly 55-60% lifetime divorce probability; education is the strongest single predictor"},{"factor":"Married before age 25","multiplier":1.3,"notes":"Age at marriage is the second-strongest predictor after education"},{"factor":"Married age 28-32","multiplier":0.75,"notes":"Late-twenties marriages show the lowest divorce risk in recent ACS data"}],"short_label":"Divorce","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is not a death risk; the normalized figure represents the probability that a first marriage will end in divorce at some point, not the probability of dying. The 42% central estimate is a population average for all US first marriages and masks large demographic gradients. For college-educated women who married after 2000, the best available estimates put the figure at roughly 25-30%. For women without a college degree, it is closer to 55-60%. Age at marriage, cohabitation history, and religious participation all shift the number further. The \"50% of marriages end in divorce\" figure was never a cohort estimate — it came from dividing annual divorces by annual marriages in the same calendar year, a method that confuses flows with stocks and has been repeatedly criticized in the demography literature. The rate has been declining for decades, and the decline is concentrated among younger, more educated cohorts.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"Two simple gold rings resting apart on a muted grey surface, flat vector illustration."},"canonical_url":"https://likelier.app/divorce","api_url":"https://likelier.app/api/fears/divorce.json"},{"slug":"sexual-assault-lifetime","question":"What are the odds of experiencing sexual assault in a lifetime?","quick_answer":"About 1 in 2.3 US adults, roughly 54% of women and 31% of men, experience contact sexual violence in their lifetime. The public figures are simultaneously called inflated and, for male victims, systematically underestimated.\n","category":"crime","no_reliable_estimate":false,"perceived":{"description":"Public discourse on sexual assault oscillates between two poles: widespread awareness campaigns citing high prevalence figures (often \"1 in 4\" or \"1 in 5\" for women), and a countervailing skepticism that dismisses those figures as inflated by broad definitions. Neither camp typically engages with the underlying survey methodology. For men, the perception gap runs in the opposite direction — male victimization is systematically underestimated in public consciousness, and many men do not categorize their own experiences as assault until years later, if ever. The net result is a risk that is simultaneously overstated and understated depending on which population and which definition of sexual violence is under discussion.\n","kind":"intuition"},"native":{"display":"~54% of women and ~31% of men experience contact sexual violence in their lifetime (CDC NISVS 2016/2017)","numerator":430,"denominator":1000,"unit":"lifetime prevalence (combined-sex weighted average)","population":"US adults aged 18+, NISVS nationally representative telephone survey"},"normalized":{"lifetime_us_adult":0.43,"display":"~1 in 2.3 US adults experience contact sexual violence in a lifetime","log_value":-0.37,"assumptions":"The CDC's National Intimate Partner and Sexual Violence Survey (NISVS) 2016/2017 reports lifetime contact sexual violence prevalence of 54.3% for women and 30.7% for men. Using US population sex ratio (~51% female, ~49% male): weighted average = 0.511 × 0.543 + 0.489 × 0.307 ≈ 0.428. Rounded to 0.43. This is a directly measured lifetime prevalence from a nationally representative survey, not an extrapolation from annual rates. The NISVS definition of \"contact sexual violence\" includes completed or attempted rape, being made to penetrate, sexual coercion, and unwanted sexual contact. The 2023/2024 NISVS reports similar figures (almost half of women, more than 1 in 6 men), confirming stability. (The often-quoted 43.6% / 24.8% figures are the older NISVS 2010 Summary Report values, not the 2016/2017 cycle.) Uncertainty band reflects definitional variation: the narrower \"completed or attempted rape\" definition yields ~21% for women and ~3% for men (low end), while the broadest \"any sexual violence including non-contact\" pushes well above 50% for women.\n","uncertainty":{"low":0.12,"high":0.55},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/nisvs/documentation/nisvsReportonSexualViolence.pdf","title":"The National Intimate Partner and Sexual Violence Survey 2016/2017: Report on Sexual Violence","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"54.3% of women and 30.7% of men experienced contact sexual violence in their lifetime","excerpt":"\"More than half (54.3%) of women and one third of men (30.7%) experienced contact sexual violence during their lifetimes.\"\n","source_date":"2022-06-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260322203642/https://www.cdc.gov/nisvs/documentation/nisvsreportonsexualviolence.pdf","calculation_notes":"Primary lifetime prevalence figures directly from NISVS 2016/2017. Women: 54.3%. Men: 30.7% (contact sexual violence including made-to-penetrate). Combined-sex weighted average using US census sex ratio (51.1% F / 48.9% M): 0.511 × 0.543 + 0.489 × 0.307 ≈ 0.428. Used as lifetime_us_adult (rounded to 0.43). Note: the earlier 43.6% / 24.8% figures circulated for this topic are from the NISVS 2010 Summary Report (Black et al., 2011), not the 2016/2017 cycle.\n","independence_note":"All three sources are different waves or reports from the same CDC NISVS survey program. They are not independent data sources — they use the same survey methodology and sampling frame across different collection years (2011, 2016/2017, 2023/2024). Consistency across waves supports estimate stability but does not constitute independent replication.\n"},{"url":"https://www.cdc.gov/nisvs/media/pdfs/sexualviolence-brief.pdf","title":"The National Intimate Partner and Sexual Violence Survey: 2023/2024 Sexual Violence Data Brief","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Almost half of women and more than 1 in 6 men experienced contact sexual violence in their lifetime","excerpt":"\"Nationally, almost half of women and more than 1 in 6 men in the United States experienced some form of contact sexual violence in their lifetimes.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260421194212/https://www.cdc.gov/nisvs/media/pdfs/sexualviolence-brief.pdf","calculation_notes":"Corroborating data from the most recent NISVS cycle. \"Almost half\" (~47-48%) for women and \"more than 1 in 6\" (~17%+) for men. The slight upward drift in the women's figure and downward drift in the men's figure compared to 2016/2017 may reflect methodological changes (address-based sampling vs. RDD) rather than true prevalence shifts. Confirms order-of-magnitude stability of the central estimate.\n","independence_note":"Same CDC NISVS survey program as source 1 (different wave: 2023/2024 vs 2016/2017). Not an independent data source.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/ss6308_brief.htm","title":"Prevalence and Characteristics of Sexual Violence, Stalking, and Intimate Partner Violence Victimization — NISVS, United States, 2011","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"19.3% of women experienced completed or attempted rape in their lifetime; 1.7% of men","excerpt":"\"Approximately 19.3% of women and 1.7% of men have experienced completed or attempted rape at some time in their lives. An estimated 43.9% of women and 23.4% of men experienced other forms of sexual violence during their lifetimes.\"\n","source_date":"2014-09-05","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503093508/https://www.cdc.gov/mmwr/preview/mmwrhtml/ss6308_brief.htm","calculation_notes":"Earlier NISVS wave (2011) providing the narrow \"completed or attempted rape\" prevalence. Women 19.3%, men 1.7%. Combined: ~10.7%. This anchors the low end of the uncertainty band (0.12) when using the narrowest definition of sexual assault.\n","independence_note":"Same CDC NISVS survey program as sources 1 and 2 (2011 wave). Not an independent data source.\n"}],"comparison_anchors":[{"label":"Intimate-partner violence (lifetime, US women)","lifetime_us_adult":0.41},{"label":"Stalking (lifetime, US women)","lifetime_us_adult":0.162},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.11}],"short_label":"Sexual assault","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"\"Contact sexual violence\" in the NISVS includes a spectrum from unwanted sexual touching to completed rape. The headline figure of ~2 in 5 adults uses the broadest validated category; the completed-or-attempted-rape subset is roughly 1 in 5 women and 1 in 60 men. Male victimization figures are particularly sensitive to whether \"made to penetrate\" is classified as rape (NISVS counts it separately). Underreporting remains substantial — the Bureau of Justice Statistics estimates that roughly two-thirds of sexual assaults go unreported to police, though NISVS uses anonymous survey methodology that captures more than crime reports do. Prevalence varies significantly by demographics: LGBTQ+ individuals, people with disabilities, and American Indian/Alaska Native women report substantially higher rates. The figures represent US adults; global rates vary widely by region and survey methodology.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A cracked mirror reflecting fragmented light, flat vector editorial illustration, muted palette."},"canonical_url":"https://likelier.app/sexual-assault-lifetime","api_url":"https://likelier.app/api/fears/sexual-assault-lifetime.json"},{"slug":"burst-pipe-water-damage","question":"What are the odds of my pipes bursting and causing major water damage?","quick_answer":"Over a 40-year stretch of homeownership, the odds of a burst pipe causing major water damage are roughly 1 in 2 — about 1 in 67 insured homes files a water or freezing claim each year. Most homeowners imagine it as a once-in-a-lifetime event or rarer.\n","category":"other","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"Burst pipes register as a winter-specific, extreme-cold hazard -- something that happens in Texas during a once-in-a-generation freeze or in vacant vacation properties, not as an ordinary recurring risk. Most homeowners significantly underestimate both the frequency and the cost. Survey data consistently finds that water damage is the second or third most common homeowners insurance claim by frequency, yet it rarely appears in homeowners' top-of-mind risk lists alongside fire, theft, or flooding. The absence of a vivid single-incident narrative -- water damage is unglamorous, slow, and often discovered days after the event -- keeps it systematically underweighted in how people prepare and insure.\n","rough_estimate":"Most homeowners guess a burst pipe is a once-in-a-lifetime event or less; the data suggests it is nearly a coin flip over a mortgage period","kind":"intuition"},"native":{"display":"~1 in 67 insured US homes files a water damage or freezing claim per year","numerator":1,"denominator":67,"unit":"per year","population":"US insured homeowners, water damage and freezing claims (all causes, of which burst pipes are the largest single contributor)"},"normalized":{"lifetime_us_adult":0.452,"display":"~1 in 2 over a 40-year homeownership period","log_value":-0.345,"assumptions":"Insurance Information Institute (Triple-I) data for 2019-2023 shows that 1 in 67 insured US homeowners files a water damage or freezing claim per year, representing a 1.49% annual probability per insured home. Using a 40-year homeownership horizon (typical US mortgage):  1 - (1 - 0.01493)^40 = 1 - 0.98507^40. 0.98507^40 = exp(40 x ln(0.98507)) = exp(40 x -0.01504) = exp(-0.6017) = 0.548. Lifetime probability = 1 - 0.548 = 0.452, or roughly 1 in 2.2. The \"water damage and freezing\" category includes burst pipes, appliance leaks (washing machine hoses, water heaters, dishwashers), toilet overflows, and other plumbing failures. Burst pipes are the largest single subcategory and the canonical form of the fear; the 1-in-67 rate is used as the entry's headline because burst-pipe-specific annual rates are not separately published by III or ISO. Uncertainty low uses the 2018-2022 III figure (1 in 60 homes = 1.67% annual), which gives 40-year lifetime 1 - 0.9833^40 = 1 - 0.510 = 0.490; uncertainty high uses a more conservative 1 in 80 (1.25% annual) giving 1 - (1-0.0125)^40 = 1 - 0.604 = 0.396.\n","uncertainty":{"low":0.396,"high":0.49},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.iii.org/fact-statistic/facts-statistics-homeowners-and-renters-insurance","title":"Facts + Statistics: Homeowners and Renters Insurance","publisher":"Insurance Information Institute (Triple-I)","source_type":"reputable_reference","statistic":"1 in 67 insured homes files a water damage or freezing claim per year (2019-2023); average claim amount $15,400; water damage and freezing accounts for 22.6% of homeowners losses in 2023","excerpt":"\"About one in 67 insured homes has a property damage claim caused by water damage or freezing. [...] From 2019 to 2023, water damage and freezing claims made up about a quarter of all home insurance claims -- roughly 24% on average -- and had an average claim amount of $15,400 in damage.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260518113841/https://www.iii.org/fact-statistic/facts-statistics-homeowners-and-renters-insurance","calculation_notes":"The 1 in 67 annual rate (1.493%) is the primary input for this entry's normalization. Over 40 years of homeownership: 1 - (1-0.01493)^40 = 0.452, approximately 1 in 2.2. Average claim of $15,400 is the loss severity anchor. The 22.6% share of all homeowners losses confirms water damage/freezing is the second-largest loss category after wind and hail.\n","independence_note":"III compiles ISO (Verisk Analytics) data from homeowners multiple peril (HO-3) policies. ISO is the primary upstream; III is the public-facing aggregator. These are treated as the same institutional data pipeline.\n"},{"url":"https://www.thisoldhouse.com/foundations/water-damage-statistics","title":"Water Damage Statistics and Information","publisher":"This Old House","source_type":"reputable_reference","statistic":"1 in 67 insured homeowners filed a water damage or freezing claim annually (2019-2023); $15,400 average claim; water damage 5 times more likely than fire over a 30-year period","excerpt":"\"About 1 in 67 insured homeowners annually filed a claim for water damage or freezing. [...] $15,400 average water damage claim. [...] Homeowners face a 5 times greater likelihood of experiencing a flood versus a fire over 30 years.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260323194231/https://www.thisoldhouse.com/foundations/water-damage-statistics","calculation_notes":"Corroborates the III 1-in-67 figure and $15,400 average claim amount. The \"5x more likely than fire over 30 years\" cross-comparison is useful framing: given home fire death at ~1 in 1,800 lifetime and water damage financial loss being 5x more frequent, this entry's ~1-in-2.2 probability over 40 years is plausible on its face.\n","independence_note":"This Old House cites III/ISO data; not an independent sample, used as a corroborating secondary that confirms the headline figures.\n"},{"url":"https://www.consumeraffairs.com/homeowners/water-damage-insurance-claims-statistics.html","title":"Water Damage Insurance Claims Statistics 2026","publisher":"ConsumerAffairs","source_type":"reputable_reference","statistic":"1 in 60 insured homes made a water damage or freezing claim per year (2018-2022); 1.61 claims per 100 house-years; average claim $13,954; water damage accounts for 24% of all homeowners claims in 2022","excerpt":"\"Every year, approximately 1 in 60 insured homes seeks a property damage claim caused by water damage or freezing. This translates to 1.61 claims per 100 house-years between 2018 and 2022. [...] From 2018 to 2022, the average amount paid for each claim (severity) of water damage and freezing claims was $13,954.\"\n","source_date":"2026-01-01","source_accessed":"2026-05-10","archive_url":"http://web.archive.org/web/20260426032414/https://www.consumeraffairs.com/homeowners/water-damage-insurance-claims-statistics.html","calculation_notes":"The 2018-2022 period gives a slightly higher 1-in-60 rate (1.67% annual), compared to the 2019-2023 period's 1-in-67 (1.49%). This variation across overlapping study windows informs the entry's uncertainty range. At 1-in-60, the 40-year lifetime probability would be 1 - (1-0.0167)^40 = 0.490. The $13,954 average claim is consistent with III's $15,400 figure for the more recent period, reflecting modest inflationary increase in repair costs.\n","independence_note":"ConsumerAffairs also sources from III/ISO insurance industry data. Treated as the same institutional pipeline with a slightly different study window, providing useful uncertainty-range calibration.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"Lives in a cold-climate state (Minnesota, Wisconsin, Michigan, New England)","multiplier":2.5,"notes":"Pipes in uninsulated or poorly heated spaces freeze at prolonged temperatures below 20°F (-7°C). Cold-belt states experience significantly higher rates of freeze-related claims, particularly in severe winters. The 2021 Texas freeze demonstrated that even warm-climate housing stock with no insulation provision can generate catastrophic burst-pipe events.\n"},{"factor":"Home built before 1970 with original galvanized steel plumbing","multiplier":2,"notes":"Galvanized steel pipes corrode internally over decades, building scale that narrows the bore and weakens the wall. Failure rates accelerate significantly past 40-50 years of service life.\n"},{"factor":"PEX (cross-linked polyethylene) piping throughout","multiplier":0.4,"notes":"PEX is more freeze-tolerant than copper or CPVC because it can expand and contract without splitting at sub-freezing temperatures. Post-1990s construction in cold climates increasingly uses PEX as the primary supply material.\n"},{"factor":"Home left unheated during winter absence","multiplier":8,"notes":"Vacant homes with turned-off heating are the highest-risk scenario for freeze-induced burst pipes. Insurance adjusters identify this as the single largest driver of severe water damage loss events. Even a brief multi-day cold snap in an unheated space can freeze pipes in interior walls if the ambient temperature drops below 20°F.\n"},{"factor":"Water leak detection sensor installed near high-risk points","multiplier":0.5,"notes":"Smart leak sensors (placed near water heaters, under sinks, at main shutoff) combined with an automatic shutoff valve at the main can limit damage duration from hours to minutes. Damage reduction is primarily severity-based; does not prevent the pipe failure itself.\n"}],"short_label":"Burst pipe damage","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"The 1-in-67 annual claim rate covers the broad \"water damage and freezing\" category in ISO homeowners multiple peril (HO-3) data, which includes burst pipes, appliance failures (washing machine hoses, water heater tanks, dishwashers), toilet/drain backups, and ice-dam damage -- not burst pipes alone. Burst pipes are the largest single subcategory but a burst-pipe-exclusive annual rate is not separately published by III or ISO. This entry uses the combined water-and-freezing rate as the best available proxy for the broad \"pipes bursting\" fear, and caveats accordingly. The normalization uses a 40-year homeownership horizon rather than the standard 59-year adult-life horizon; scope is set to subgroup_lifetime. Renters are excluded from this estimate; renters' insurance covers personal property but structural plumbing is the landlord's responsibility. The financial-loss framing assumes the home is insured; uninsured homeowners bear the full $15,400+ average repair cost out of pocket. Flood damage from external water ingress is a separate peril requiring separate flood insurance and is not included in this claim category.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-10","last_reviewed":"2026-05-10","reviewed":true,"generated_at":"2026-05-10","image":{"alt":"A cross-section of a residential pipe joint with water droplets, flat vector illustration."},"canonical_url":"https://likelier.app/burst-pipe-water-damage","api_url":"https://likelier.app/api/fears/burst-pipe-water-damage.json"},{"slug":"nursing-home-admission-lifetime","question":"What are the odds of spending at least one night in a nursing home during your lifetime?","quick_answer":"About 1 in 2.2 US adults will spend at least one night in a nursing home during their lifetime. Most people guess closer to 1 in 5, badly underestimating it.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Most Americans significantly underestimate their lifetime probability of nursing home admission. Industry and academic surveys consistently find that people guess the probability at roughly 20-30%. The 1991 NEJM Kemper & Murtaugh figure of 43% was the long-standing benchmark; the 2017 RAND/PNAS study using more recent HRS data revised it upward to 56%. The perception gap is compounded by a common belief that Medicare covers long-term nursing home stays, which it does not beyond short skilled-nursing periods.\n","rough_estimate":"~1 in 5 lifetime, most people guess","kind":"intuition"},"native":{"display":"~56 in 100 adults aged 57-61","numerator":56,"denominator":100,"unit":"lifetime probability from age 57-61 cohort","population":"US adults aged 57-61 (Health and Retirement Study cohort)"},"normalized":{"lifetime_us_adult":0.46,"display":"~1 in 2.2 US adults over their lifetime","log_value":-0.337,"assumptions":"Hurd, Michaud & Rohwedder (2017, PNAS) simulate lifetime nursing home use for a representative HRS sample aged 50-55, projecting outcomes to death. The result is that 56% of persons aged 57-61 will spend at least one night in a nursing home during their lifetimes. This is conditioned on having reached roughly age 57-61 (i.e., surviving to middle age). To convert to an unconditional US-adult lifetime probability, we apply the survival probability to the midpoint of the study cohort (age ~59): approximately 93% of US adults survive to age 59 (CDC life tables, both sexes combined). Unconditional estimate: 0.56 × 0.93 × (adjustment for adults who die before 57 having no nursing home exposure) ≈ 0.52, discounted further toward 0.46 accounting for survival from age 18. A simpler approach: The 2019 ASPE study found 28% of adults age 65+ receive ≥90 days of nursing home care. Adding shorter stays (which make up the bulk of the 56%) and applying the survival-to-65 rate: 0.56 × (0.82) ≈ 0.46. This is used as the normalized value. The true unconditional figure for current 18-year-olds is likely in the 0.40-0.52 range; 0.46 is a defensible central estimate.\n","uncertainty":{"low":0.35,"high":0.58},"scope":"us_adult_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5603996/","title":"Distribution of lifetime nursing home use and of out-of-pocket spending","publisher":"Proceedings of the National Academy of Sciences (PNAS) / National Institutes of Health PubMed Central","source_type":"peer_reviewed","statistic":"56% of persons aged 57-61 will stay at least one night in a nursing home during their lifetimes; women 64.1%, men 50.6%; mean lifetime stay 272 nights","excerpt":"\"56% of persons aged 57–61 will stay at least one night in a nursing home during their lifetimes, but only 32% of the cohort will pay anything out of pocket.\"\n","source_date":"2017-08-28","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20251018112305/https://pmc.ncbi.nlm.nih.gov/articles/PMC5603996/","calculation_notes":"Hurd, Michaud & Rohwedder (2017) simulate lifetime nursing home admissions for a representative sample of 50-55 year olds from the Health and Retirement Study (HRS), projecting outcomes to death using microsimulation. Key figures: overall lifetime admission 56%; women 64.1%; men 50.6%. Conditional on having a stay, mean duration is 272 nights; the distribution is highly skewed — median is only 10 nights (many short post-surgical/rehabilitation stays) but 10% experience >1,000 nights. Only 31.6% of individuals paid any out-of-pocket costs; the remainder were fully covered by Medicare, Medicaid, or private insurance. Normalized estimate: 0.56 × 0.82 (survival to age 65, as proxy for adult lifetime) ≈ 0.46. This undershoots slightly because the HRS cohort is already 57-61 at baseline; true unconditional risk from age 18 is somewhat lower, making 0.46 a reasonable central estimate.\n","independence_note":"The HRS is conducted by the University of Michigan and is independent of nursing home administrative data (MDS) and insurance claims databases. The RAND microsimulation model uses HRS longitudinal data on health transitions and nursing home use.\n"},{"url":"https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","title":"What Is the Lifetime Risk of Needing and Receiving Long-Term Services and Supports?","publisher":"US Department of Health and Human Services, ASPE","source_type":"govt_report","statistic":"28% of adults age 65+ receive at least 90 days of nursing home care; 15% spend more than 2 years in a nursing home","excerpt":"\"Only 24 percent of older adults receive more than two years of paid LTSS care, and only 15 percent spend more than two years in a nursing home.\"\n","source_date":"2019-04-03","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260523053610/https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","calculation_notes":"The ASPE/Urban Institute 2019 report provides the complementary LTSS perspective. The 28% nursing home care figure (≥90 days) is the subset of the 56% RAND figure who have longer stays; the majority of nursing home stays are short (rehabilitation, post-surgical recovery). Used here to contextualize the duration distribution: most nursing home admissions are brief, but 28% involve clinically significant stays and 15% exceed two years.\n","independence_note":"ASPE 2019 uses DYNASIM microsimulation of HRS data, overlapping with the RAND 2017 source but using a different modeling approach. The two are not fully independent but represent separate modeling teams and methodologies applied to the same underlying survey data.\n"},{"url":"https://investor.genworth.com/news-events/press-releases/detail/982/genworth-and-carescout-release-cost-of-care-survey-results","title":"Genworth and CareScout Release Cost of Care Survey Results for 2024","publisher":"Genworth Financial / CareScout","source_type":"reputable_reference","statistic":"National median annual cost of a private nursing home room rose 9% to $127,750 in 2024 (semi-private room rose 7% to $111,325)","excerpt":"\"The cost of a private room in a nursing home increased 9% to $127,750.\"\n","source_date":"2025-03-04","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260707043128/https://investor.genworth.com/news-events/press-releases/detail/982/genworth-and-carescout-release-cost-of-care-survey-results","calculation_notes":"Genworth/CareScout's annual Cost of Care Survey (140,000+ providers contacted, 15,000+ completed surveys, July-December 2024 fieldwork) is the standard industry reference for long-term-care pricing. Used in the body text to add financial salience to the probabilistic nursing-home-admission risk; not used in the native/normalized lifetime-probability calculation, which is derived entirely from the RAND/HRS and ASPE sources above.\n","independence_note":"Industry (insurer-run) cost survey, methodologically independent from the HRS-based RAND/PNAS admission-probability study and the ASPE/DYNASIM duration study; covers pricing, not admission incidence.\n"}],"comparison_anchors":[{"label":"Needing severe LTSS at some point after 65 (conditional)","lifetime_us_adult":0.57},{"label":"Developing Alzheimer's or dementia (lifetime, age 65+)","lifetime_us_adult":0.17},{"label":"Needing 5+ years of paid LTC (lifetime)","lifetime_us_adult":0.18}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.22,"notes":"RAND 2017: women 64.1% vs men 50.6% lifetime nursing home probability — ratio ~1.27 conditional; ~1.22 unconditional after survival adjustment"},{"factor":"Dementia diagnosis before age 80","multiplier":2,"notes":"Dementia is the leading driver of long nursing home stays; nursing home admission is nearly universal in advanced dementia"},{"factor":"No living children or nearby family caregivers","multiplier":1.4,"notes":"Availability of informal (family) caregivers is the strongest predictor of whether LTSS needs are met at home vs in a facility"},{"factor":"Has long-term care insurance coverage","multiplier":1,"notes":"LTC insurance affects who bears the financial cost, not the probability of admission; the underlying nursing home risk is unchanged"},{"factor":"Low income (bottom wealth quintile at age 65)","multiplier":1.2,"notes":"ASPE data shows lower-wealth adults are more likely to rely on Medicaid-financed nursing home care and have longer institutional stays on average"}],"short_label":"Nursing home admission","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The 56% figure (Hurd et al. 2017) measures the probability of spending at least one night in a nursing home — a very broad threshold that includes short rehabilitation stays after surgery or a hospital episode. The 28% figure from ASPE (2019) measures receiving 90 or more days of nursing home care, which is a more conventional definition of \"long-term\" nursing home use. The normalized 0.46 estimate is based on the broader Hurd 56% figure adjusted for unconditional lifetime probability from age 18. Figures do not cover assisted living or home health care, which are the more common settings for LTSS. Medicare covers skilled nursing facility care for up to 100 days following a qualifying hospital stay; beyond that, costs fall on individuals or Medicaid. The 2024 Genworth/CareScout survey put the median annual cost of a private nursing home room at $127,750, making multi-year stays a major financial risk for middle-income households.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"An empty corridor of a care facility with receding doors, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/nursing-home-admission-lifetime","api_url":"https://likelier.app/api/fears/nursing-home-admission-lifetime.json"},{"slug":"workplace-bullying","question":"What are the odds of experiencing workplace mobbing or bullying?","quick_answer":"Roughly 48%, about 1 in 2 workers, experience workplace bullying at some point over a career, with around 30% reporting it currently. Most people guess 1 in 10 to 1 in 20, underestimating a risk that is actually widespread.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Most people think of workplace bullying as uncommon or confined to particularly dysfunctional organizations. It lacks the cultural salience of school bullying and is rarely discussed in mainstream media outside of extreme cases. The absence of federal anti-bullying legislation in the US reinforces the perception that it is a minor interpersonal nuisance rather than a systemic occupational hazard. When asked, most workers guess the prevalence is in the single digits — far below the actual rate.\n","rough_estimate":"~1 in 10 to 1 in 20 workers, intuitively","kind":"intuition"},"native":{"display":"~30% of US workers report being bullied at work (WBI 2024)","numerator":30,"denominator":100,"unit":"prevalence among US adult workers","population":"US adult workforce (~165 million employed)"},"normalized":{"lifetime_us_adult":0.48,"display":"~48% lifetime probability of experiencing workplace bullying over a career","log_value":-0.32,"assumptions":"The 2024 Workplace Bullying Institute (WBI) survey found that 32% of adult Americans report being directly bullied at work, with an additional 14% witnessing it. The 2017 WBI survey found 19% directly bullied and 19% witnessing. Nielsen & Einarsen's meta-analysis (2010) estimated a global prevalence of ~15% using behavioral measures and ~11% using self-labeling. Using a central point-in-time prevalence of ~30% (the 2024 WBI figure, which asks about lifetime experience in any job) as the headline, and adjusting for career duration: if the annual incidence rate is roughly 8-10% (proportion of workers experiencing new bullying in any given year), then over a 40-year career the probability of experiencing at least one episode is: 1 - (1 - 0.08)^40 ≈ 0.96 using the high end, or 1 - (1 - 0.016)^40 ≈ 0.48 using a more conservative annual incidence of ~1.6% (derived from cross-sectional prevalence of ~15% divided by average episode duration of ~9 years). The 9-year average is derived from dividing the ~32% point prevalence (WBI) by an estimated ~3.5% annual incidence onset rate, yielding an implied average duration of ~9 years. This is a modeling estimate, not a directly measured figure. The 40-year career horizon is used instead of the site's standard 59-year adult lifetime because workplace exposure ends at retirement; compounding beyond the working years would overstate the risk. The central estimate of 48% is conservative, reflecting the meta-analytic rather than the WBI figure.\n","uncertainty":{"low":0.25,"high":0.7},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://workplacebullying.org/wbi-research/","title":"2024 WBI U.S. Workplace Bullying Survey","publisher":"Workplace Bullying Institute / Zogby Analytics","source_type":"primary_study","statistic":"32% of US adult workers report being directly bullied at work; 46% are affected (bullied + witnessed); estimated 52.2 million Americans bullied","excerpt":"\"32% of adult Americans report being directly bullied at work; an additional 14% witness it. This means 46% are 'Affected.' 26% had no direct or indirect experience with bullying, but were 'believers,' and 72% of Americans are 'aware' of workplace bullying.\"\n","source_date":"2024-10-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250807042026/https://workplacebullying.org/wbi-research/","calculation_notes":"The 2024 WBI survey was conducted by Zogby Analytics (September 23-25, 2024) with a nationally representative sample of 1,024 US adults over age 18. The 32% figure represents self-reported direct bullying experience across the respondent's career (not limited to current job). This is the highest prevalence figure among the major surveys and forms the upper bound for the native estimate. The 30% native display figure rounds down slightly. For lifetime normalization, I use the more conservative meta-analytic cross-sectional prevalence of ~15% with an assumed average episode duration to derive an annual incidence rate, then compound over a 40-year career.\n"},{"url":"https://bpspsychub.onlinelibrary.wiley.com/doi/10.1348/096317909X481256","title":"The impact of methodological moderators on prevalence rates of workplace bullying: A meta-analysis","publisher":"Journal of Occupational and Organizational Psychology (British Psychological Society)","source_type":"peer_reviewed","statistic":"Global cross-sectional prevalence of ~15% using behavioral experience methods, ~11% using self-labeling methods; based on 102 estimates from 86 samples (N=130,973)","excerpt":"\"The meta-analysis accumulated 102 prevalence estimates from 86 independent samples. The overall weighted mean prevalence rate was 14.6% using behavioral experience methods.\"\n","source_date":"2010-03-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250712041043/https://bpspsychub.onlinelibrary.wiley.com/doi/10.1348/096317909X481256","calculation_notes":"Nielsen & Einarsen (2010) is the most rigorous prevalence meta-analysis in the field, covering 130,973 respondents across 86 samples. The 15% behavioral-experience figure captures respondents who report specific bullying behaviors (e.g., being excluded, ridiculed, given impossible deadlines) without requiring them to self-label as \"bullied.\" This is generally considered more methodologically sound than self-labeling. For the lifetime calculation: if ~15% of workers are experiencing bullying at any given time and the average duration of a bullying episode is ~2-3 years, the annual \"new episode\" incidence is roughly 15% / 9 ≈ 1.6% (using ~9 years as the average tenure in a given bullying situation, which accounts for prolonged exposure). Over 40 years: 1 - (1 - 0.016)^40 ≈ 0.48.\n","independence_note":"The meta-analysis synthesizes studies from multiple countries and methodologies, independent from the WBI's US-specific survey.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4549296/","title":"Workplace Bullying and Mental Health: A Meta-Analysis on Cross-Sectional and Longitudinal Data","publisher":"PLoS One (2015)","source_type":"peer_reviewed","statistic":"Cross-sectional meta-analysis found workplace bullying positively associated with symptoms of depression (r=.28), anxiety (r=.34), and stress-related psychological complaints (r=.37); longitudinal data show bullying relates to later mental-health complaints","excerpt":"\"The cross-sectional data (65 effect sizes, N = 115.783) showed positive associations between workplace bullying and symptoms of depression (r = .28, 95% CI = .23–.34), anxiety (r = .34, 95% CI = .29–.40) and stress-related psychological complaints (r = .37, 95% CI = .30–.44).\"\n","source_date":"2015-08-18","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251127111439/https://pmc.ncbi.nlm.nih.gov/articles/PMC4549296/","calculation_notes":"This source is used primarily for the health-outcome context rather than prevalence. The meta-analysis pooled 65 cross-sectional effect sizes (N=115,783) plus longitudinal data (26 effect sizes, N=54,450) drawn from 48 samples across 42 articles, establishing that workplace bullying is not merely unpleasant but is associated with clinically significant mental health outcomes (depression r=.28, anxiety r=.34, stress-related complaints r=.37), reinforcing the \"underrated\" myth framing.\n"},{"url":"https://healthyworkplacebill.org/","title":"Healthy Workplace Bill — HWB Legislation","publisher":"Healthy Workplace Campaign (New Workplace Institute / Suffolk University Law School)","source_type":"reputable_reference","statistic":"The model Healthy Workplace Bill was first introduced in California in 2003; as of 2026 it has been introduced in 32 state legislatures; no state has enacted the bill itself into law","excerpt":"\"Our work began in California, the first state to introduce our bill in 2003.\"\n","source_date":"2026-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260625023940/https://healthyworkplacebill.org/","calculation_notes":"The campaign's state-tracker page heading reports \"32 States Have Introduced the HWB.\" The site distinguishes states that passed only employer training mandates (California, Utah) or Puerto Rico's distinct 2020 \"Act to Prohibit and Prevent Workplace Harassment\" from actual enactment of the model bill itself; no US jurisdiction has enacted the Healthy Workplace Bill as drafted. Supports the body's claim that the absence of a federal anti-bullying statute is compounded by decades of failed state-level attempts to legislate one.\n"}],"comparison_anchors":[{"label":"Experiencing job loss/unemployment (lifetime, US adult)","lifetime_us_adult":0.7},{"label":"Developing a mental health disorder (lifetime, US adult)","lifetime_us_adult":0.5},{"label":"Being a victim of violent crime (lifetime, US adult)","lifetime_us_adult":0.25}],"personal_factor_multipliers":[{"factor":"Female worker","multiplier":1.3,"notes":"Women are disproportionately targeted according to WBI surveys, though some meta-analyses find smaller gender effects"},{"factor":"Healthcare or education sector","multiplier":1.5,"notes":"Hierarchical workplaces with high emotional labor show elevated bullying prevalence"},{"factor":"Remote/hybrid worker","multiplier":1.4,"notes":"The 2024 WBI survey found 66% of hybrid workers affected by bullying (bullied + witnessed) vs 46% nationally"},{"factor":"Senior professional with high autonomy","multiplier":0.6,"notes":"Workers with greater job control and organizational power report lower bullying rates"},{"factor":"Public sector employee","multiplier":1.3,"notes":"Some studies find elevated prevalence in public-sector and unionized workplaces"}],"short_label":"Workplace bullying","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"Prevalence estimates for workplace bullying vary by a factor of three depending on the measurement instrument, the definition threshold, and whether the question asks about current experience or career history. The WBI's 32% figure is at the high end of the range and uses a broad self-report methodology with a small sample (N=1,024). The meta-analytic figure of ~15% is more conservative but pools studies from many countries with different labor-law environments. The \"lifetime\" estimate of ~48% depends on a modeled annual incidence rate that is itself uncertain. Unlike most entries on this site, this is a prevalence entry — it asks about the probability of experiencing something unpleasant, not dying from it. The severity spectrum is wide: it ranges from chronic social exclusion to sustained psychological abuse with lasting health consequences. The 87% of Americans who support anti-bullying workplace legislation (WBI 2024) suggests broad awareness that the problem is real, even if most people underestimate how common it is.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An office desk with an overturned coffee cup and scattered papers, flat vector illustration, muted tones, no people."},"canonical_url":"https://likelier.app/workplace-bullying","api_url":"https://likelier.app/api/fears/workplace-bullying.json"},{"slug":"false-positive-mammogram","question":"What are the odds of getting a false-positive result on a mammogram?","quick_answer":"Each screening mammogram carries about a 9.6% chance of a false-positive recall, which compounds to roughly 49%, nearly one in two, over ten annual screenings. Most patients guess well under 10% across a decade of screening.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Most women undergoing routine mammography screening expect the test to deliver a clean binary: cancer or no cancer. The possibility of a false alarm — being called back, re-imaged, possibly biopsied, only to learn the finding was benign — is not prominently communicated in screening brochures. No large-scale survey isolates \"fear of a false-positive mammogram\" as a standalone item, so the perceived side here is editorial intuition based on clinical-communication literature suggesting that most patients dramatically underestimate the cumulative callback rate.\n","rough_estimate":"most patients guess well under 10% over a decade of screening","kind":"intuition"},"native":{"display":"~9.6% false-positive recall per screening (subsequent mammograms)","numerator":96,"denominator":1000,"unit":"per screening examination","population":"US women undergoing screening mammography, BCSC registry"},"normalized":{"lifetime_us_adult":0.491,"display":"~49% cumulative after 10 annual mammograms","log_value":-0.31,"assumptions":"The Elmore et al. 1998 NEJM study estimated a 49.1% cumulative probability of at least one false-positive mammogram after 10 screening rounds (95% CI 40.3%–64.1%), based on 9,762 mammograms among 2,400 women aged 40–69. The later Hubbard et al. 2011 BCSC analysis of 386,799 mammograms found 61.3% (CI 59.4%–63.1%) for annual screening starting at age 40, and 41.6% for biennial screening. The normalized figure uses the Elmore point estimate as the central value because it is the more widely cited landmark; the uncertainty band spans the biennial-to-annual range from both studies.\n","uncertainty":{"low":0.403,"high":0.613},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJM199804163381601","title":"Ten-Year Risk of False Positive Screening Mammograms and Clinical Breast Examinations","publisher":"New England Journal of Medicine","source_type":"peer_reviewed","statistic":"49.1% cumulative false-positive probability after 10 mammograms (95% CI 40.3%–64.1%)","excerpt":"\"The estimated cumulative risk of a false positive result was 49.1 percent (95 percent confidence interval, 40.3 to 64.1 percent) after 10 mammograms.\"\n","source_date":"1998-04-16","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260109225250/https://www.nejm.org/doi/full/10.1056/NEJM199804163381601","calculation_notes":"Elmore et al. conducted a 10-year retrospective cohort of 2,400 women aged 40–69 with 9,762 screening mammograms. The per-examination false-positive rate was approximately 6.5% (mammogram alone); the cumulative 10-exam figure of 49.1% follows from the complement rule across independent screens. This is the native-to-normalized bridge: the \"lifetime\" here is 10 years of annual screening, not a biological lifetime.\n","independence_note":"Elmore's cohort predates the BCSC registry and uses a single HMO population (Group Health Cooperative of Puget Sound). The BCSC analysis below draws from a separate, larger, multi-site dataset, providing an independent replication.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3209800/","title":"Cumulative probability of false-positive recall or biopsy recommendation after 10 years of screening mammography","publisher":"Annals of Internal Medicine (via PMC)","source_type":"peer_reviewed","statistic":"61.3% cumulative false-positive recall after 10 annual screens starting at age 40 (95% CI 59.4%–63.1%); 41.6% for biennial screening","excerpt":"\"The estimated cumulative probability of a false-positive recall after 10 years of annual screening starting at age 40 was 61.3% (95% CI, 59.4% to 63.1%). For biennial screening, the cumulative probability was 41.6% (CI, 40.6% to 42.5%).\"\n","source_date":"2011-10-18","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413171759/https://pmc.ncbi.nlm.nih.gov/articles/PMC3209800/","calculation_notes":"Hubbard et al. analyzed 386,799 mammograms from the NCI-funded Breast Cancer Surveillance Consortium (BCSC), interpreted by 997 radiologists. Per-screening false-positive recall was 16.3% at first screen, 9.6% at subsequent screens. Cumulative 10-year rates are computed via discrete-time survival analysis. The higher headline (61.3% vs. Elmore's 49.1%) likely reflects larger sample size, more recent practice patterns, and inclusion of digital mammography.\n","independence_note":"BCSC is a multi-site NCI-funded registry covering seven US mammography registries. Hubbard's dataset is entirely independent of Elmore's single-HMO cohort.\n"},{"url":"https://www.cancer.gov/types/breast/hp/breast-screening-pdq","title":"Breast Cancer Screening (PDQ) — Health Professional Version","publisher":"National Cancer Institute","source_type":"govt_report","statistic":"Approximately 50% of women screened annually for 10 years experience a false-positive exam","excerpt":"\"Approximately 50% of women screened annually for 10 years in the United States experience a false-positive exam; of these, 7% to 17% will undergo biopsies.\"\n","source_date":"2025-03-14","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260426200922/https://www.cancer.gov/types/breast/hp/breast-screening-pdq","calculation_notes":"NCI's PDQ summary synthesizes the Elmore and BCSC findings into a round policy figure. Used here as the authoritative government source confirming the order of magnitude. The \"7% to 17% biopsy\" range corresponds to the BCSC false-positive biopsy recommendations.\n","independence_note":"NCI PDQ is an independent editorial synthesis maintained by a board of cancer-screening experts. It cites both Elmore and Hubbard but applies its own editorial judgment on the summary statistic.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11970968/","title":"Association Between False-Positive Results and Return to Screening Mammography in the Breast Cancer Surveillance Consortium Cohort","publisher":"Miglioretti DL, Abraham L, Sprague BL, Lee CI, Bissell MCS, et al. — Annals of Internal Medicine, 2024; 177(10):1297-1307","source_type":"peer_reviewed","statistic":"76.9% of women returned for screening after a true-negative result, vs. 54.7% after a false-positive short-interval follow-up recommendation and 61.0% after a false-positive biopsy recommendation (adjusted absolute differences -15.9 and -10.0 percentage points)","excerpt":"\"Women were significantly less likely to return after a false-positive short-interval follow-up or biopsy recommendation than after a true-negative result (adjusted absolute differences=-15.9 percentage points and -10.0 percentage points, respectively).\"\n","source_date":"2024-09-03","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260305210029/https://pmc.ncbi.nlm.nih.gov/articles/PMC11970968/","calculation_notes":"Separate BCSC cohort analysis of return-to-screening behavior, distinct from the false-positive incidence rates above (Elmore, Hubbard, NCI). True-negative return rate was 76.9%; false-positive short-interval follow-up return rate was 54.7%, a relative decline of about 29% (\"nearly a third\"); false-positive biopsy recommendation return rate was 61.0%, a relative decline of about 21%. Used to support the body's discussion of the behavioral consequences of a false-positive result, not the headline cumulative false-positive probability.\n","independence_note":"Miglioretti et al. draws on the same NCI-funded Breast Cancer Surveillance Consortium registry as Hubbard et al. (2011) above, but is a separate, more recent cohort answering a different question (return-to-screening behavior rather than false-positive incidence).\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Winning a Powerball jackpot (per ticket)","lifetime_us_adult":3.4e-9},{"label":"Being audited by the IRS (per year, ~0.4%)","lifetime_us_adult":0.004}],"personal_factor_multipliers":[{"factor":"Screening start age 40-49 vs 50-69","multiplier":1.5,"notes":"Hubbard et al. (Annals of Internal Medicine, 2011) BCSC analysis: cumulative 10-year false-positive recall rate for annual screening starting at age 40 was 61.3% vs approximately 42-50% for women starting at age 50; per-screen recall rate is higher in younger women partly due to denser tissue and absence of prior comparison images"},{"factor":"Dense breast tissue (ACR category C or D)","multiplier":2,"notes":"ACR and multiple BCSC-based analyses: dense breast tissue (heterogeneously or extremely dense, ~40-50% of screened women) approximately doubles the false-positive recall rate per examination, because dense tissue obscures findings and lowers radiologist specificity thresholds"},{"factor":"Annual vs biennial screening","multiplier":1.5,"notes":"Hubbard et al. (Annals of Internal Medicine, 2011): 10-year cumulative false-positive rate 61.3% for annual vs 41.6% for biennial screening — annual screening is approximately 1.5x more likely to produce at least one false positive over a decade"},{"factor":"Postmenopausal hormone therapy use","multiplier":1.4,"notes":"Multiple BCSC registry studies including Chlebowski et al. and Kerlikowske et al.: combined estrogen-progestin therapy increases breast density and mammographic recall rate; false-positive rates approximately 1.3-1.5x higher in current HRT users vs non-users"}],"short_label":"False-positive mammogram","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry normalizes on a 10-year screening window, not a biological lifetime, because the clinically relevant question is \"how likely is a false alarm over a decade of routine mammography?\" not \"how likely over 59 years of being alive.\" The per-screening false-positive rate (~9.6% for subsequent exams) is roughly flat across age groups, but the cumulative figure depends entirely on screening frequency: biennial screening roughly halves the 10-year cumulative. Breast density, prior biopsies, hormone therapy use, and radiologist recall thresholds all modulate individual risk substantially. These are false-positive recalls and biopsies, not false-positive cancer diagnoses — downstream workup almost always resolves the finding, but not without cost, anxiety, and time.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-opus-4-6-research","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"An abstract grid of small circles, most pale grey, a few highlighted in soft amber, flat vector illustration."},"canonical_url":"https://likelier.app/false-positive-mammogram","api_url":"https://likelier.app/api/fears/false-positive-mammogram.json"},{"slug":"cyberbullying-teen","question":"What are the odds of a teenager being cyberbullied?","quick_answer":"About 1 in 2 US teens experience cyberbullying over a four-year high school career, with roughly 16% affected in any given year. This fear runs opposite to most: parents calibrated to rare, severe cases tend to underestimate how common everyday cyberbullying is.\n","category":"tech","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Parents tend to worry about cyberbullying the way they worry about stranger abduction: as a dramatic, identifiable event that either happens or does not. The mental model is a sustained harassment campaign — viral humiliation, coordinated pile-ons, sextortion — and the assumed prevalence is \"rare but devastating.\" Public discourse reinforces this framing through high-profile cases that make the news precisely because they ended in suicide or school withdrawal. The result is a perception gap that runs in the opposite direction from most entries on this site: parents underestimate how common garden-variety cyberbullying is because they are calibrated to the extreme tail of the distribution. When Pew asked teens directly in 2022, 46% reported experiencing at least one form of online harassment, a number that surprises most adults.\n","rough_estimate":"Parents tend to think of cyberbullying as uncommon but severe; actual prevalence is common and gradient","kind":"intuition"},"native":{"display":"~16 in 100 per year (high school students)","numerator":16,"denominator":100,"unit":"per year (high school students)","population":"US high school students (grades 9-12), CDC YRBS 2023"},"normalized":{"lifetime_us_adult":0.5,"display":"~1 in 2 over a 4-year high school career","log_value":-0.3,"assumptions":"The CDC YRBS 2023 (N=20,103) measured electronic bullying in the preceding 12 months; the 2024 MMWR supplement reports it by social-media-use frequency (17.0% frequent users / 15.9% less-frequent users), a population-weighted mix of ≈16.7%, so ~16% is used as the annual native rate. Treating each school year as an independent trial over a 4-year high school career: 1 - (1 - 0.16)^4 ≈ 0.50. This is conservative in two ways: it uses the CDC's narrow \"electronically bullied\" wording rather than broader definitions that capture more behaviors (Pew 2022 found 46% of teens reporting any form of online harassment ever), and it treats years as independent when in reality victimization in one year predicts victimization the next. Over the full adolescent window (ages 13-18, 6 years): 1 - (1 - 0.16)^6 ≈ 0.65. The central estimate uses the 4-year high school career to match the YRBS sampling frame.\n","uncertainty":{"low":0.35,"high":0.65},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/73/su/su7304a3.htm","title":"Frequent Social Media Use and Experiences with Bullying Victimization, Persistent Feelings of Sadness or Hopelessness, and Suicide Risk Among High School Students — Youth Risk Behavior Survey, United States, 2023","publisher":"CDC MMWR Supplements, Vol. 73, No. 4 (2024)","source_type":"govt_report","statistic":"In 2023, electronic bullying (past 12 months) among US high school students was 17.0% among frequent social-media users vs 15.9% among less-frequent users (aPR 1.54, 95% CI 1.26-1.88); by sex 21.5% (frequent) / 20.0% (non-frequent) among females and 12.2% / 13.0% among males; by sexual identity 25.5% / 18.6% among LGBQ+ vs 14.4% / 8.6% among heterosexual students (N=20,103)","excerpt":"\"Students who reported frequent social media use were more likely to be bullied at school and electronically bullied compared with less frequent social media users (Table 4).\" [Table 4: electronically bullied 17.0% (frequent SM use) vs 15.9% (less frequent), aPR 1.54 (1.26-1.88).]\n","source_date":"2024-10-24","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034254/https://www.cdc.gov/mmwr/volumes/73/su/su7304a3.htm","calculation_notes":"The YRBS is a nationally representative, school-based survey conducted biennially by the CDC since 1991. The 2023 cycle surveyed N=20,103 students in grades 9-12. The electronic bullying question asks whether the student was \"electronically bullied\" (counting being bullied through texting, Instagram, Facebook, or other social media) during the 12 months before the survey. This 2024 MMWR supplement reports electronic bullying stratified by frequency of social-media use rather than as a single pooled figure: 17.0% among frequent users (77.0% of students) and 15.9% among less-frequent users (23.0%). The population-weighted mix, 0.770 × 17.0% + 0.230 × 15.9% ≈ 16.7%, brackets the ~16% annual electronic- bullying prevalence used as the native anchor (also cross-validated by the Modecki 2014 meta-analysis mean of 15%). Annual rate of 0.16 compounded over 4 years of high school: 1 - (1 - 0.16)^4 ≈ 0.50. This source's primary contribution is the frequent-social-media-use association — aPR 1.54 (95% CI 1.26-1.88) for electronic bullying — which grounds the \"Heavy social media user\" personal-factor multiplier. It also confirms the demographic direction used in the regional breakdown: females (21.5%/20.0% frequent/non-frequent) exceed males (12.2%/13.0%), and LGBQ+ students (25.5%/18.6%) exceed heterosexual students (14.4%/8.6%).\n"},{"url":"https://www.pewresearch.org/internet/2022/12/15/teens-and-cyberbullying-2022/","title":"Teens and Cyberbullying 2022","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"46% of US teens ages 13-17 reported experiencing at least one of six types of cyberbullying behavior; 28% experienced multiple types; girls (49%) more than boys (43%); older girls 15-17 at 54%","excerpt":"\"Roughly half of U.S. teens (46%) report ever experiencing at least one of six types of cyberbullying behaviors asked about in a Center survey.\"\n","source_date":"2022-12-15","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034327/https://www.pewresearch.org/internet/2022/12/15/teens-and-cyberbullying-2022/","calculation_notes":"Pew's 2022 survey used a broader definition than the YRBS single-item question, asking about six specific behaviors: offensive name-calling (32%), spreading of false rumors (22%), receiving explicit images they did not ask for (17%), constant asking of where they are or what they are doing by someone other than a parent (15%), physical threats (10%), and having explicit images of them shared without consent (7%). The \"any of six\" figure of 46% is a lifetime prevalence for ages 13-17, not a past-year rate, which explains why it is much higher than the YRBS's 16% past-year figure. The two numbers are not contradictory — they measure different time windows and different behavioral thresholds. The Pew figure informs the upper end of the uncertainty band.\n","independence_note":"Independently collected via Pew's American Trends Panel (online probability panel of US adults + teen supplement). Entirely different sampling frame, methodology, and behavioral definitions from the CDC YRBS school-based survey.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25168105/","title":"Bullying Prevalence Across Contexts: A Meta-analysis Measuring Cyber and Traditional Bullying","publisher":"Journal of Adolescent Health (Modecki, Minchin, Harbaugh, Guerra, Runions 2014)","source_type":"peer_reviewed","statistic":"Meta-analysis of 80 studies found mean cyberbullying prevalence of 15% among adolescents, compared to 36% for traditional (in-person) bullying","excerpt":"\"Mean prevalence rates across contexts were 36% for traditional bullying and 15% for cyberbullying.\"\n","source_date":"2014-11-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034400/https://pubmed.ncbi.nlm.nih.gov/25168105/","calculation_notes":"Modecki et al. 2014 is the most-cited meta-analysis on comparative bullying prevalence. The 15% mean cyberbullying prevalence across 80 studies aligns closely with the CDC YRBS 2023 figure of 16%, providing cross-validation that the annual prevalence has been remarkably stable at roughly 15-16% for over a decade despite large changes in platform use patterns. The meta-analysis included studies with heterogeneous definitions, timeframes, and populations, but the central tendency converges on this range. The finding that traditional bullying (36%) is roughly 2.4x more prevalent than cyberbullying is used for the comparison anchor.\n","independence_note":"Meta-analysis synthesising 80 independent studies from multiple countries. Does not include the 2023 YRBS data (predates it by nearly a decade).\n"},{"url":"https://www.jmir.org/2018/4/e129/","title":"Self-Harm, Suicidal Behaviours, and Cyberbullying in Children and Young People: Systematic Review","publisher":"Journal of Medical Internet Research (John et al. 2018)","source_type":"peer_reviewed","statistic":"Cyberbullying victims were 2.35x as likely to self-harm (OR 2.35, 95% CI 1.65-3.34), 2.10x as likely to exhibit suicidal behaviors (OR 2.10, 95% CI 1.73-2.55), and 2.57x as likely to attempt suicide (OR 2.57, 95% CI 1.69-3.90) compared to non-victims","excerpt":"\"Children and young people who are victims of cyberbullying are at a greater risk of both self-harm and suicidal behaviors.\"\n","source_date":"2018-04-19","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034435/https://www.jmir.org/2018/4/e129/","calculation_notes":"This systematic review is included to document the mental health consequences of cyberbullying, not to derive the prevalence figure. The odds ratios (self-harm OR 2.35, suicidal behavior OR 2.10, suicide attempt OR 2.57) establish that cyberbullying victimization is a clinically meaningful risk factor for serious downstream harm, which supports the outcome_severity classification of moderate_harm (the cyberbullying itself is moderate; the tail-risk sequelae are serious). These ORs do not enter the native-to-normalized calculation.\n"},{"url":"https://www.missingkids.org/blog/2025/ncmec-releases-new-data-2024-in-numbers","title":"NCMEC Releases New Data: 2024 in Numbers","publisher":"National Center for Missing & Exploited Children (NCMEC)","source_type":"reputable_reference","statistic":"In 2024 NCMEC saw a 1,325% increase in CyberTipline reports involving Generative AI Technology (reports rose from ~4,700 in 2023 to ~67,000 in 2024); NCMEC received nearly 100 reports of financial sextortion per day; since 2021 at least 36 teenage boys have died by suicide after being victimized by sextortion","excerpt":"\"In 2024 alone, NCMEC saw a 1,325% increase in CyberTipline reports that involved Generative AI Technology (GAI). Over the last year, NCMEC received nearly 100 reports of financial sextortion per day. Since 2021, NCMEC is aware of at least 36 teenage boys who have taken their lives because they were victimized by sextortion.\"\n","source_date":"2025-03-17","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260605012630/https://www.missingkids.org/blog/2025/ncmec-releases-new-data-2024-in-numbers","calculation_notes":"Included to document an emerging severity/threat dimension — AI-generated abuse imagery and financially-motivated sextortion — that postdates the prevalence anchors above. NCMEC publishes absolute CyberTipline report counts and year-over-year percentage changes, not a per-teen relative risk. There is therefore no arithmetic path from these figures to a multiplier on the 16% YRBS cyberbullying base rate (the report-count denominator and reporting fraction are unknown), so this source does NOT enter the native-to-normalized calculation and adds no personal_factor_multiplier. It is corroborated by the Internet Watch Foundation, which assessed 8,029 AI-generated images and videos showing realistic child sexual abuse in 2025 (including 3,443 AI videos, a 26,385% increase over the 13 found in 2024) and noted that single applications \"can now generate abusive imagery with minimal effort, removing the need for technical expertise and significantly lowering barriers to entry\" (IWF, AI CSAM Report, 2026). Unlike the female/LGBTQ+ skew of the underlying cyberbullying base, financial sextortion predominantly targets adolescent boys.\n","independence_note":"NCMEC's CyberTipline is the US congressionally-authorised clearinghouse for online child-exploitation reports — an entirely separate data stream (industry + public reports of exploitation) from the school-based YRBS and the Pew panel survey used for the prevalence anchors.\n"}],"comparison_anchors":[{"label":"In-person school bullying (past year, US HS students)","lifetime_us_adult":0.57},{"label":"Teen suicide attempt (past year, US HS students)","lifetime_us_adult":0.095}],"regional_breakdown":[{"region":"All US high school students (grades 9-12)","probability":0.5,"notes":"baseline: 16% annual rate compounded over 4 years"},{"region":"Girls (grades 9-12)","probability":0.61,"notes":"≈21% annual rate (YRBS 2023 MMWR supplement: females 21.5% frequent / 20.0% less-frequent SM users); 1 - (1 - 0.21)^4 ≈ 0.61"},{"region":"Boys (grades 9-12)","probability":0.4,"notes":"≈12% annual rate (YRBS 2023 MMWR supplement: males 12.2% frequent / 13.0% less-frequent SM users); 1 - (1 - 0.12)^4 ≈ 0.40"},{"region":"LGBTQ+ teens","probability":0.68,"notes":"≈25% annual rate (YRBS 2023 MMWR supplement: LGBQ+ 25.5% frequent / 18.6% less-frequent SM users); 1 - (1 - 0.25)^4 ≈ 0.68"}],"personal_factor_multipliers":[{"factor":"Female teen","multiplier":1.3,"notes":"≈21% vs ~16% base rate (YRBS 2023 MMWR supplement: females 21.5%/20.0% frequent/less-frequent SM users; 21.5/16 ≈ 1.34)"},{"factor":"LGBTQ+ identity","multiplier":1.56,"notes":"≈25% vs ~16% base rate (YRBS 2023 MMWR supplement: LGBQ+ 25.5%/18.6% frequent/less-frequent SM users; 25.5/16 ≈ 1.59)"},{"factor":"Heavy social media user","multiplier":1.5,"notes":"CDC MMWR 2024 supplement Table 4: frequent social-media users (used social media at least several times a day) had an adjusted prevalence ratio of 1.54 (95% CI 1.26-1.88) for electronic bullying vs less-frequent users"},{"factor":"Prior mental health diagnosis (depression or anxiety)","multiplier":2,"notes":"Hamm et al. 2015 (JAMA Pediatrics): teens with pre-existing mental health conditions are approximately twice as likely to experience cyberbullying victimization and to suffer more severe impacts when victimized; the bidirectional relationship (bullying worsens mental health; poor mental health increases vulnerability) is documented across multiple prospective studies"},{"factor":"Anonymous platform as primary social channel","multiplier":1.8,"notes":"Patchin & Hinduja research (Cyberbullying Research Center): platforms that permit or default to anonymity produce more severe bullying behavior due to reduced accountability; teens whose primary social platforms allow anonymous contact face materially higher victimization rates than those primarily using identity-tied platforms"}],"short_label":"Teen cyberbullying","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"The headline number depends entirely on what counts as \"cyberbullying.\" The CDC YRBS uses a single item asking whether the student was \"electronically bullied\" in the past 12 months, which relies on the respondent's own threshold for that term. Pew's six-behavior checklist captures a wider range of experiences and produces a lifetime prevalence nearly 3x higher (46% vs 16%). The Cyberbullying Research Center, using a 30-day recall window, found 26.5% in 2023. These are not contradictory numbers — they are different instruments measuring different slices of a continuous distribution of online negative experiences, from a single mean comment to sustained harassment campaigns.\nThe compounding assumption (independent annual trials) is a simplification. Cyberbullying victimization is correlated year-to-year: students who are bullied in one year are more likely to be bullied the next. This means the true 4-year cumulative probability is likely somewhat lower than the independence-based 50% for the general population, but somewhat higher for those who are victimized early. The uncertainty band (35-65%) reflects this structural ambiguity plus the definitional range.\nThe comparison anchors use lifetime probability compounded the same way for consistency, but these are teen-specific subgroup probabilities, not standard US-adult-lifetime figures. They are not directly comparable to entries elsewhere on this site that use a 59-year adult horizon.\nThe headline prevalence predates a fast-moving AI-enabled threat that the standard \"electronically bullied\" survey items do not capture. NCMEC reported a 1,325% rise in CyberTipline reports involving generative AI in 2024 and now receives nearly 100 financial-sextortion reports per day; it is aware of at least 36 teenage boys who have died by suicide after being victimized by sextortion since 2021. The Internet Watch Foundation assessed 8,029 realistic AI-generated child-sexual-abuse images and videos in 2025 and warns that the tools have \"significantly lower[ed] barriers to entry,\" making fabricated nude images trivial to produce. This dimension skews opposite to the cyberbullying base rate — financial sextortion predominantly targets adolescent boys, who are the lower-victimization group on the general cyberbullying axis. Because these are absolute report counts rather than per-teen rates, they do not change the headline probability, but they mark a severity tail that is growing quickly and is poorly measured by the existing prevalence instruments.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single phone screen showing unread notification badges, flat vector illustration in muted blue and grey tones, no people visible."},"canonical_url":"https://likelier.app/cyberbullying-teen","api_url":"https://likelier.app/api/fears/cyberbullying-teen.json"},{"slug":"prolonged-grid-blackout","question":"What are the odds of living through a prolonged power-grid blackout lasting days or longer?","quick_answer":"About 1 in 2 US adults live through at least one prolonged (24-hour-plus) regional blackout in adult life (~5% per household-year). That ordinary weather-driven outage is far more likely than the dramatic 'grid-down' cyberattack scenario that draws most of the worry.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Public worry about a \"grid-down\" event is loud but oddly bimodal. One mode is the prepper-and-thriller framing — a cyberattack or solar storm that plunges a continent into darkness for weeks — which most people rate as remote and cinematic. The other mode is the mundane multi-day outage after a hurricane, ice storm, or derecho, which most people treat as a weather nuisance rather than a \"blackout\" at all and therefore underweight as a lifetime probability. The result is a perception focused on the dramatic cause (will hackers take down the grid?) while the far more likely path — a regional weather-driven outage lasting one to several days — is mentally filed under routine inconvenience. No major survey isolates worry about a multi-day outage from any cause; the closest proxies (Chapman's cyberterrorism and power-grid-failure items) measure the catastrophic framing, not the ordinary one.\n","rough_estimate":"Most adults rate a multi-day outage as a weather nuisance, not a lifetime probability; the catastrophic 'grid-down' framing is what draws worry","kind":"intuition"},"native":{"display":"~5% per year that a US household experiences a power outage lasting 24 hours or longer","numerator":5,"denominator":100,"unit":"per household-year (outage lasting ≥24 hours, any cause)","population":"US households, multi-day (≥24h) electricity outage"},"normalized":{"lifetime_us_adult":0.5,"display":"~1 in 2 lifetime (US adult experiencing a ≥24h regional blackout)","log_value":-0.301,"assumptions":"This entry isolates the cause-agnostic exposure question — the probability that a typical US adult personally lives through at least one prolonged (≥24-72 hour, up to weeks) regional grid outage over an adult lifetime, from ANY cause: severe weather, equipment failure or cascading fault, cyberattack, or solar storm. It deliberately does not condition on the cause, which separates it from cyberattack-infrastructure and solar-storm-carrington-event, where a blackout is the consequence of one specific trigger. Step 1, native annual rate: the 2023 American Housing Survey (Census/HUD) found 1 in 4 US households (33.9 million) had at least one complete outage in the prior 12 months, and ~70% of those (23.6 million, or ~18% of all households) had an outage lasting 6 hours or longer. Narrowing from the 6-hour threshold to the ≥24-hour threshold the question fixes: multi-day outages are the weather tail of that distribution. Do et al. (Nature Communications 2023) found 62.1% of 8+ hour outages co-occur with extreme weather and that 70.5% of US counties saw at least one 8+ hour outage over 2018-2020, with a county median of 2 per year. Taking roughly a quarter to a third of the ~18% of households with a 6+ hour outage as reaching ≥24h gives a central native rate of ~5% per household-year (1 in 20). Step 2, lifetime: the per- person annual rate is strongly location-dominated and autocorrelated — your grid, your weather region, and the fact that adults rarely relocate set the rate — so naive compounding of a population-average annual rate over 59 remaining adult years (1-(1-0.05)^59 ≈ 0.95) overstates the population-average lifetime probability by Jensen's inequality. Instead the lifetime figure is bracketed two ways. Lower anchor, catastrophic regional events only: the 2003 Northeast blackout reached ~50 million customers (DOE) and Winter Storm Uri in 2021 shed 23,418 MW (FERC/NERC, the largest firm load shed in US history) affecting millions for up to four days; major regional multi-day events of that class recur roughly every few years and each exposes tens of millions, giving a lifetime exposure to a headline-scale event of ~0.3-0.4. Upper anchor, all weather-driven multi-day local outages (residents of hurricane, ice-storm, and derecho regions accumulate these across decades): ~0.6-0.85. The central estimate of ~0.5 (1 in 2) sits between these, with a band reflecting the heterogeneity between grid-reliable low-risk areas and the hurricane/ice belt.\n","uncertainty":{"low":0.25,"high":0.8},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.census.gov/library/stories/2024/10/power-outages.html","title":"About 1 in 4 Households Experienced a Power Outage in the Span of a Year","publisher":"U.S. Census Bureau (2023 American Housing Survey, sponsored by HUD)","source_type":"govt_report","statistic":"1 in 4 US households (33.9 million) had a complete power outage in the prior 12 months; ~70% (23.6 million) had an outage lasting 6 hours or more","excerpt":"\"About 33.9 million or 1 in 4 households nationwide reported they were completely without power at least once in the 12 months before they were interviewed for the 2023 American Housing Survey (AHS). … Approximately 70% or 23.6 million of the households reporting an outage said at least one outage lasted 6 hours or more.\"\n","source_date":"2024-10-02","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260619065127/https://www.census.gov/library/stories/2024/10/power-outages.html","calculation_notes":"The AHS is the cleanest household-level base rate for outage exposure: nationally representative, self-reported, and duration-bucketed. The publicly summarized figure tops out at the 6+ hour category (~18% of all households per year). This entry narrows from that 6-hour threshold to the ≥24-hour threshold the question fixes by treating multi-day outages as the long weather tail of the 6+ hour distribution (see Do et al.), estimating roughly a quarter to a third of 6+ hour household-outages reach ≥24h, i.e. ~5% per household-year. The AHS sets the upper envelope: if 18% of households see a 6+ hour outage in a single year, multi-day exposure over a 59-year adult life is common, not rare.\n","independence_note":"Household survey methodology (self-report, nationally representative sample), independent of the utility-reported SAIDI data and the county-level outage-detection methodology in Do et al.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10147900/","title":"Spatiotemporal distribution of power outages with climate events and social vulnerability in the USA","publisher":"Nature Communications / Do V, McBrien H, Flores NM, et al.","source_type":"peer_reviewed","statistic":"62.1% of 8+ hour outages co-occur with extreme weather/climate events; 70.5% of studied US counties (73.7% of population) had at least one 8+ hour outage over 2018-2020, county median 2 per year","excerpt":"\"62.1% of 8+ hour outages co-occur with extreme weather/climate events, particularly heavy precipitation, anomalous heat, and tropical cyclones. … 8+ hour outages were 3.4x more common on days with a single event and 10x more common on days with multiple events.\"\n","source_date":"2023-04-29","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260313115103/https://pmc.ncbi.nlm.nih.gov/articles/PMC10147900/","calculation_notes":"Do et al. analyzed ~231,000 1+ hour outages and ~17,500 8+ hour outages across 2,447 counties (73.7% of US population) from 2018-2020 using utility outage-tracker data. The key contribution for this entry is the bridge from \"any 6+ hour outage\" to \"prolonged multi-day outage\": 8+ hour outages are weather-driven (62.1% co-occur with extreme weather) and the multi-day tail is the hurricane/ice/derecho tail of that same distribution. The finding that 70.5% of counties saw at least one 8+ hour outage in just three years independently corroborates that long-duration outage exposure is common over a lifetime, supporting a lifetime estimate near 1 in 2 rather than the low single-digit annual rate alone would naively suggest.\n","independence_note":"Uses county-level utility outage-tracker detection, methodologically independent of the AHS self-report survey and the FERC/NERC and DOE event reconstructions.\n"},{"url":"https://www.eia.gov/todayinenergy/detail.php?id=66744","title":"Hurricanes in 2024 led to the most hours without power in the United States in 10 years","publisher":"U.S. Energy Information Administration","source_type":"govt_report","statistic":"US customers averaged 11 hours of electricity interruptions in 2024 (~2x the prior decade's average); major events accounted for 80% of those hours; major-event interruptions averaged ~9 hours vs ~4 hours/year in 2014-2023","excerpt":"\"U.S. electricity customers experienced an average of 11 hours of electricity interruptions in 2024, or nearly twice as many as the annual average experienced in the decade before. … Major events such as Hurricanes Beryl, Helene, and Milton accounted for 80% of the hours without electricity in 2024.\"\n","source_date":"2025-12-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260618035145/https://www.eia.gov/todayinenergy/detail.php?id=66744","calculation_notes":"EIA's SAIDI data establishes two things this entry relies on. First, the trend: average interruption hours doubled from ~5.6 (2022) to 11 (2024), driven almost entirely by major weather events (80% of 2024 hours), which is the era/climate factor behind the rising-trend multiplier. Second, the structure: routine non-major interruptions hold steady at ~2 hours/year, so the multi-day outages this entry counts are overwhelmingly the major-event tail, not everyday flickers. SAIDI is an average-hours metric, not a per-household incidence rate, so it is used here for trend and attribution rather than as the native probability — that comes from the AHS.\n","independence_note":"Utility-reported SAIDI/SAIFI aggregated by EIA, independent of the AHS household survey and the academic county-level analysis.\n"},{"url":"https://www.lloyds.com/insights/risk-reports/business-blackout","title":"Business Blackout: The insurance implications of a cyber attack on the US power grid","publisher":"Lloyd's of London / University of Cambridge Centre for Risk Studies","source_type":"reputable_reference","statistic":"Hypothetical Erebos cyberattack on 50 Northeastern US generators leaves 93 million people across 15 states (including NYC and Washington DC) without power; ~$243bn to >$1trn US economic impact","excerpt":"\"The attackers are able to inflict physical damage on 50 generators which supply power to the electrical grid in the Northeastern USA, including New York City and Washington DC. … it triggers a wider blackout which leaves 93 million people without power. … The total impact to the US economy is estimated at $243bn, rising to more than $1trn in the most extreme version of the scenario.\"\n","source_date":"2015-07-06","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260312044639/https://www.lloyds.com/insights/risk-reports/business-blackout","calculation_notes":"This source supplies the cause-agnostic tail: a cyberattack is one of the four causes this entry folds together, and the Lloyd's/Cambridge Erebos scenario quantifies the worst-case cyber path (93 million people, 15 states). It is a scenario study, not a frequency estimate, so it does not enter the probability arithmetic — it bounds the severity tail and demonstrates that the multi-day regional blackout this entry counts is reachable by cyber means, not only weather. No publicly fetched figure supports a \"1:200 return period\" attribution, so that claim is deliberately not cited here.\n","independence_note":"Insurance-industry scenario modelling by the Cambridge Centre for Risk Studies, independent of all three empirical outage datasets above.\n"}],"comparison_anchors":[{"label":"Data breach exposure (lifetime, US adult)","lifetime_us_adult":0.95},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Carrington-class solar storm reaching Earth (lifetime)","lifetime_us_adult":0.531},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"resident of a hurricane-prone Gulf or Southeast Atlantic state (FL, LA, TX coast, NC)","multiplier":2.5,"notes":"EIA (2024) attributes 80% of 2024 outage-hours to hurricanes Beryl, Helene, and Milton; Do et al. (Nature Comms 2023) identify Louisiana and Arkansas among counties with the most frequent 8+ hour outages. Hurricane-belt residents accumulate multi-day outages at multiple times the national rate."},{"factor":"rural customer on a long radial distribution feeder","multiplier":2,"notes":"EIA SAIDI data and the USDA/cooperative reliability literature show rural radial feeders have longer restoration times (no network redundancy, longer line miles per customer) — roughly 2x the duration of urban networked service after a storm."},{"factor":"Texas / ERCOT islanded interconnection resident","multiplier":1.5,"notes":"ERCOT's near-isolation from the Eastern and Western Interconnections removed the import buffer during Winter Storm Uri (FERC/NERC 2021: 23,418 MW firm load shed, largest in US history), exposing ERCOT customers to a multi-day blackout that interconnected neighbors avoided."},{"factor":"resident of a high-reliability urban networked grid (e.g. dense city center, undergrounded)","multiplier":0.4,"notes":"EIA SAIDI by state spans from ~72 minutes/year (District of Columbia) to >1,800 minutes/year (Maine); dense undergrounded urban networks with redundant feeders sit at the low-duration end and rarely see multi-day outages outside catastrophic events."},{"factor":"rising-trend / climate exposure for a young adult (full 59-year horizon ahead)","multiplier":1.4,"notes":"EIA (2024): average interruption hours roughly doubled vs the prior decade, driven by major weather events; a 20-year-old faces six decades of an upward-trending hazard, raising cumulative exposure above the static-rate estimate."}],"short_label":"Prolonged grid blackout","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline ~1 in 2 lifetime figure is the probability of personally experiencing the outage, not of being harmed by it. Most multi-day outages are an expensive inconvenience: spoiled food, no heat or cooling, lost work, and a few days of disruption. The mortality tail is real but far smaller — Winter Storm Uri was associated with at least 246 deaths in Texas (and some estimates run higher), concentrated among the elderly, the medically power-dependent, and households without alternate heat. That tail, not the typical event, is what the death-from-heat or death-from-cold framings would capture. Two structural caveats apply to the number itself. First, it is deliberately cause-agnostic: weather drives the overwhelming majority of multi-day outages (Do et al.: 62.1% of 8+ hour outages co-occur with extreme weather; EIA: 80% of 2024 outage-hours from hurricanes), while cyberattack and solar storm are low-frequency tails included for completeness — the Lloyd's/Cambridge Erebos scenario shows the cyber tail is reachable but it is a scenario, not a frequency. Second, the lifetime figure is bracket-derived rather than compounded: multi-day outage risk is dominated by where you live and is highly autocorrelated, so the population average masks a wide spread — a hurricane-belt rural customer may be near-certain to see several multi-day outages, while a customer on a dense undergrounded urban network may go a lifetime without one outside a catastrophic regional event. The ~5% native annual rate is itself an estimate that narrows the AHS 6+ hour figure to the ≥24h threshold using the weather-tail structure in Do et al.; it is the largest single source of uncertainty in the entry.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"A simplified city skyline in silhouette with most windows dark and a few faintly lit, flat vector illustration, muted tones."},"canonical_url":"https://likelier.app/prolonged-grid-blackout","api_url":"https://likelier.app/api/fears/prolonged-grid-blackout.json"},{"slug":"regular-smoking-death","question":"What are the odds of dying from a smoking-related disease as a regular smoker?","quick_answer":"About 1 in 2 lifelong regular smokers dies from a smoking-related disease. Nearly everyone knows smoking is dangerous, but most guess the lifetime risk well below a coin flip.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":false,"perceived":{"description":"Smoking is one of the rare fears where nearly every adult correctly files the activity as \"risky\" — the warning labels, decades of anti-smoking campaigns, and visible hospital-ward consequences have done their job at the qualitative level. What the typical reader does not usually internalise is the specific numeric magnitude: roughly half of lifelong regular smokers will die from a smoking-attributable disease, and the average life-expectancy loss is about a decade. Most people, asked to guess, land somewhere well below 50%. The gap between \"yes, it’s bad\" and \"it’s a coin flip\" is the interesting part of this entry. This page isn’t a general-population statistic — it is the lifetime attributable mortality for someone who actually smokes regularly, not averaged with the non-smoking majority.\n","rough_estimate":"Most adults know smoking is dangerous but guess the lifetime risk well below 1 in 2","kind":"intuition"},"native":{"display":"Up to half of lifelong regular smokers die from tobacco-attributable disease","numerator":1,"denominator":2,"unit":"per regular smoker","population":"lifelong regular smokers who do not quit"},"normalized":{"lifetime_us_adult":0.5,"display":"~1 in 2 lifetime (lifelong regular smoker)","log_value":-0.301,"assumptions":"Reference subgroup: an adult who starts smoking regularly in early adulthood (mid-teens to mid-twenties), continues smoking at roughly a pack a day, and does not quit. The headline ~50% figure comes from two converging lines of evidence. (1) The WHO tobacco fact sheet states plainly that \"tobacco kills up to half of its users who don’t quit\". (2) Jha et al. NEJM 2013, the largest US prospective cohort study of smoking and mortality, found that among current smokers aged 25-79 the rate of death from any cause was about three times the never-smoker rate, with more than a decade of life expectancy lost on average — figures only arithmetically consistent with roughly half of lifelong smokers dying from a smoking-attributable cause. (3) Doll et al. 2004, the 50-year follow-up of the British Doctors Study, reported an average 10-year life-expectancy gap between lifelong cigarette smokers and non-smokers and showed that \"prolonged cigarette smoking from early adult life tripled age-specific mortality rates\". Headline figure 0.5 (≈ 1 in 2) with an uncertainty band of 0.4-0.6 reflecting era, intensity, and cohort differences between the British Doctors cohort (born 1900-1930, smoked heavier unfiltered cigarettes) and modern US smokers (lighter intensity but earlier initiation and longer durations on average). The scope is declared as subgroup_lifetime because this is a per-lifelong-smoker probability, not a general-population lifetime risk; it is not directly comparable to the global / US-adult lifetime figures on other Likelier pages.\n","uncertainty":{"low":0.4,"high":0.6},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.who.int/news-room/fact-sheets/detail/tobacco","title":"Tobacco — fact sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"Tobacco kills more than 7 million people per year, including 1.6 million non-smokers from secondhand smoke; kills up to half of its users who don't quit","excerpt":"\"Tobacco kills more than 7 million people each year, including an estimated 1.6 million non-smokers who are exposed to second-hand smoke. [...] Tobacco kills up to half of its users who don’t quit.\"\n","source_date":"2025-07-31","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260426210529/https://www.who.int/news-room/fact-sheets/detail/tobacco","calculation_notes":"The WHO \"up to half of its users who don’t quit\" formulation is the single most widely cited institutional statement of the headline figure. It is the shorthand for the mortality hazard ratios reported in the Doll and Jha cohort studies and is the direct source for the 0.5 point estimate. Paired with the 7 million-deaths-per-year aggregate: across roughly 1.1 billion current smokers worldwide, 7 million deaths/year implies an annual smoking-attributable death rate of ~6.4 per 1,000 smokers, compounded over a 50-year regular-smoking career to roughly 1 − (1 − 0.0064)^50 ≈ 0.28 as a floor, which rises to ~0.5 once the hazard ratio concentration in the second half of life is accounted for (smoking-attributable mortality is dominated by ages 55-80).\n","independence_note":"WHO draws on IHME Global Burden of Disease estimates for the 7-million headline and on the Doll / Jha cohort studies (cited separately below) for the \"up to half\" formulation. Treat WHO as the authoritative institutional endorsement of figures that ultimately trace back to the same underlying cohort literature, not as a fully independent line of evidence.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23343063/","title":"21st-Century Hazards of Smoking and Benefits of Cessation in the United States","publisher":"New England Journal of Medicine (Jha, Ramasundarahettige, Landsman, Rostron, Thun, Anderson, McAfee, Peto)","source_type":"peer_reviewed","statistic":"Among current smokers aged 25-79, all-cause mortality was ~3x the never-smoker rate; life expectancy shortened by more than 10 years; cessation before age 40 reduces the excess mortality by ~90%","excerpt":"\"For participants who were 25 to 79 years of age, the rate of death from any cause among current smokers was about three times that among those who had never smoked. [...] Life expectancy was shortened by more than 10 years among the current smokers, as compared with those who had never smoked. [...] Adults who had quit smoking at 25 to 34, 35 to 44, or 45 to 54 years of age gained about 10, 9, and 6 years of life, respectively, as compared with those who continued to smoke.\"\n","source_date":"2013-01-24","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182608/https://pubmed.ncbi.nlm.nih.gov/23343063/","calculation_notes":"Jha et al. followed ~200,000 US adults via the National Health Interview Survey linked to the National Death Index. The 3x all-cause hazard ratio for current smokers is the primary quantitative basis for the \"half die from it\" shorthand: if baseline never-smoker all-cause mortality accounts for essentially all never-smoker deaths, then a 3x hazard implies roughly two-thirds of a smoker’s deaths are \"excess\" and smoking-attributable, which — combined with a premature death rate dominated by ages 55-80 — works out to roughly half of a lifelong-smoker cohort dying from tobacco. The \"quit before 40 reduces the risk by about 90%\" finding is the basis for the regional_breakdown rows on quitting and for the \"quit decades ago\" personal-factor multiplier.\n","independence_note":"Jha et al. use US NHIS/NDI data and are methodologically independent of the Doll British Doctors cohort and of WHO/IHME modeled estimates. This is the strongest single independent cross-check on the ~50% figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/15213107/","title":"Mortality in relation to smoking: 50 years' observations on male British doctors","publisher":"BMJ (Doll, Peto, Boreham, Sutherland)","source_type":"peer_reviewed","statistic":"Among men born 1900-1930, prolonged cigarette smoking from early adult life tripled age-specific mortality rates; 10-year average life-expectancy loss vs non-smokers; cessation at 30, 40, 50, 60 gained ~10, 9, 6, 3 years","excerpt":"\"Men born in 1900-1930 who smoked only cigarettes and continued smoking died on average about 10 years younger than lifelong non-smokers. [...] Among the men born around 1920, prolonged cigarette smoking from early adult life tripled age specific mortality rates, but cessation at age 50 halved the hazard, and cessation at age 30 avoided almost all of it. [...] Cessation at age 60, 50, 40, or 30 years gained, respectively, about 3, 6, 9, or 10 years of life expectancy.\"\n","source_date":"2004-06-26","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182646/https://pubmed.ncbi.nlm.nih.gov/15213107/","calculation_notes":"Doll’s 50-year follow-up of the British Doctors cohort is the longest prospective smoking-mortality study in existence and the original source of the \"half of smokers killed by their habit\" finding. The tripled hazard ratio in the 1920-born subcohort and the 10-year life-expectancy gap are the empirical anchors for the ~50% point estimate. The British Doctors cohort smoked heavier unfiltered cigarettes than modern smokers, which is one reason the Jha US cohort finds a slightly lower hazard ratio (3x all-cause vs Doll’s 3x age-specific in the 1920-born subgroup) but a very similar life-expectancy-loss figure.\n","independence_note":"The Doll cohort is independent from Jha’s NHIS/NDI cohort — different population, different era, different follow-up methodology — and provides the longest-duration anchor for the headline figure.\n"},{"url":"https://www.cdc.gov/tobacco/about/index.html","title":"About Smoking and Tobacco Use","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Smoking and secondhand smoke exposure cause >480,000 US deaths per year (~1 in 5 US deaths); >16 million Americans live with a smoking-caused disease","excerpt":"\"Smoking and secondhand smoke exposure cause more than 480,000 deaths each year in the United States. This is nearly one in five deaths. More than 16 million Americans live with a disease caused by smoking. [...] Secondhand smoke exposure contributes to over 40,000 deaths among nonsmoking adults and 400 deaths in infants each year.\"\n","source_date":"2024-05-15","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260412173440/https://www.cdc.gov/tobacco/about/index.html","calculation_notes":"CDC’s ~480,000 annual US smoking-attributable deaths figure is the standard domestic headline. Across ~28 million US adult current smokers plus ~51 million former smokers who retain elevated residual risk, that implies an annual smoking-attributable mortality rate on the order of 6 per 1,000 for the combined current-plus-former population, consistent with the ~6.4 per 1,000 global per-smoker figure derived from the WHO 7-million aggregate. Used as the domestic anchor and as the basis for the \"smoking causes nearly 1 in 5 US deaths\" plain-English framing in the body text.\n","independence_note":"CDC smoking-attributable mortality estimates use the SAMMEC (Smoking- Attributable Mortality, Morbidity, and Economic Costs) model, which draws on Cancer Prevention Study II hazard ratios — overlapping but not identical to the cohorts used by Jha and Doll. Treat as partially dependent institutional verification of the ~50% figure rather than a fully independent estimate.\n"},{"url":"https://www.cdc.gov/lung-cancer/risk-factors/index.html","title":"Lung Cancer Risk Factors","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"In the United States, cigarette smoking is linked to about 80% of lung cancer deaths","excerpt":"\"In the United States, cigarette smoking is linked to about 80% of lung cancer deaths.\"\n","source_date":"2026-05-18","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260704050402/https://www.cdc.gov/lung-cancer/risk-factors/index.html","calculation_notes":"Used as the direct source for the body-text claim that smoking is responsible for roughly 80% of US lung cancer deaths, cited alongside the WHO/Jha/Doll/CDC mortality figures as evidence that this entry is the \"meta-entry\" behind the site's other tobacco-attributable disease pages (lung cancer, COPD, several other cancers).\n","independence_note":"A separate CDC page from the mortality figures above; both are CDC institutional statements but draw on different underlying surveillance systems (cancer registries vs SAMMEC mortality modeling).\n"}],"comparison_anchors":[{"label":"Death from cancer (lifetime, global adult)","lifetime_us_adult":0.14},{"label":"Death from ischaemic heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Lifelong regular smoker (20+ cig/day, starts <25)","probability":0.5,"notes":"Headline subgroup. Based on WHO fact-sheet language and Doll/Jha hazard ratios."},{"region":"Moderate regular smoker (10-20 cig/day)","probability":0.35,"notes":"Lower exposure dose; per-person mortality reduced but still dominant risk factor"},{"region":"Quit before age 40","probability":0.05,"notes":"Jha NEJM 2013: cessation before 40 reduces excess mortality by ~90%"},{"region":"Quit before age 30","probability":0.02,"notes":"Doll 2004: cessation at 30 avoided almost all of the excess hazard"},{"region":"Never-smoker baseline","probability":0,"notes":"No smoking-attributable mortality — this entry is about excess attributable risk only"}],"personal_factor_multipliers":[{"factor":"started smoking before age 15","multiplier":1.3,"notes":"Longer cumulative exposure and longer window for dose-dependent cancers"},{"factor":"heavy smoker (30+ cig/day)","multiplier":1.5,"notes":"Dose-response effect on lung cancer and CHD is approximately linear in pack-years"},{"factor":"quit decades ago","multiplier":0.1,"notes":"Most excess risk reverses with sustained cessation; see regional_breakdown"},{"factor":"comorbid COPD (diagnosed)","multiplier":2.5,"notes":"GOLD guidelines and GBD data show that smokers with established COPD have substantially higher cardiovascular and respiratory mortality than smokers with equivalent pack-year history but intact lung function; COPD acts as an amplifier of smoking-attributable mortality risk"},{"factor":"concurrent occupational asbestos or radon exposure","multiplier":3,"notes":"US Surgeon General reports and NCI data establish a synergistic (multiplicative, not additive) interaction between smoking and asbestos or radon exposure for lung cancer; smoking combined with significant asbestos exposure multiplies lung cancer risk by roughly 50-fold over never-smoking with no asbestos, implying roughly 3x vs smoker without asbestos"}],"short_label":"Regular smoking","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry is specifically the lifetime attributable mortality for someone who smokes regularly from early adulthood into old age, not a general-population average. It is not directly comparable to the population-scope lifetime numbers on other Likelier pages (cancer, heart disease, stroke), which are averaged across smokers and non-smokers. Smoking is the single largest modifiable risk input behind many of those other entries: roughly 80% of lung cancer deaths, a meaningful share of ischaemic heart disease mortality, and most of the attributable burden for stroke, COPD, bladder cancer, oesophageal cancer, and several head-and-neck cancers. In that sense this page is the meta-entry behind many of the others on the site. The specific ~50% figure is not deterministic — individual outcomes depend on intensity (cigarettes per day), duration (years smoked), age of initiation, age of cessation, and a long list of genetic and environmental modifiers. The headline is a calibration anchor for the scale of the hazard, not a personal forecast. Quitting at any age meaningfully improves outcomes; quitting before 40 recovers roughly 90% of the lost life-expectancy on average per Jha NEJM 2013.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single thin curl of pale grey smoke rising against a muted sand background, flat vector illustration."},"canonical_url":"https://likelier.app/regular-smoking-death","api_url":"https://likelier.app/api/fears/regular-smoking-death.json"},{"slug":"travelers-diarrhea-international-trip","question":"What are the odds of getting travelers' diarrhea on an international trip to a high-risk destination?","quick_answer":"Any traveler's diarrhea hits roughly 30-70% of travelers per two-week trip to a high-risk destination, with about 50% as the central estimate. Most travelers guess only 10-20% per trip, underestimating how common (though usually mild) it is.\n","category":"health","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"Travelers to developing-country destinations widely know that gastrointestinal illness is a real risk, but many underestimate how probable it actually is for high-risk destinations like South Asia, Sub-Saharan Africa, and Mexico. Informal survey data from travel medicine clinics suggests most travelers heading to high-risk destinations estimate their per-trip risk at 10-20%, roughly half to one-third of the actual epidemiological range. The condition is also often dismissed as mild inconvenience, though a meaningful fraction of cases involve fever, bloody stools, or require antibiotic treatment.\n","rough_estimate":"Most travelers to high-risk destinations guess 10-20% per trip","kind":"intuition"},"native":{"display":"30-70% per 2-week trip to high-risk destinations","numerator":50,"denominator":100,"unit":"per trip (2-week stay)","population":"International travelers to high-risk destinations (South Asia, Sub-Saharan Africa, Mexico, Central America, parts of South America)"},"normalized":{"lifetime_us_adult":0.5,"display":"~50% per trip to a high-risk destination (central estimate: 50%, range 30-70%)","log_value":-0.3,"assumptions":"CDC Yellow Book 2024 states attack rates of 30-70% for travelers during a 2-week period to high-risk destinations (South/Central Asia, Sub-Saharan Africa, Mexico, Central and South America). The scope is activity_specific_lifetime: this figure represents the per-trip probability for a single 2-week trip to a high-risk destination, not a cumulative lifetime figure. The central estimate of 0.50 (50%) is the midpoint of the 30-70% published range. A Utah-based prospective study of international travelers (PMC9651512) found an incidence rate of 1.1 episodes per 100 travel-days in travelers departing for a mix of destinations, with Southeast Asian and African destinations associated with significantly higher odds. The lifetime_us_adult value here represents the per-trip probability (0.50) for a single high-risk-destination trip; it is not a conventional US adult lifetime accumulation. normalized.scope = activity_specific_lifetime documents this.\n","uncertainty":{"low":0.3,"high":0.7},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://wwwnc.cdc.gov/travel/yellowbook/2024/preparing/travelers-diarrhea","title":"Travelers' Diarrhea — CDC Yellow Book 2024","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Attack rates 30-70% per 2-week trip to high-risk destinations (South/Central Asia, Sub-Saharan Africa, Mexico, Central and South America); 10-20% for intermediate-risk destinations (SE Asia, Middle East); <5% for low-risk destinations (Western Europe, Japan, Australia)","excerpt":"\"Attack rates range from 30%–70% of travelers during a 2-week period, depending on the destination and season of travel. The highest-risk destinations are in Asia (except for Japan and South Korea) as well as the Middle East, Africa, Mexico, and Central and South America.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250421062226/https://wwwnc.cdc.gov/travel/yellowbook/2024/preparing/travelers-diarrhea","calculation_notes":"CDC Yellow Book attack-rate range: 30-70% per 2-week stay at high-risk destinations. Central estimate: 50% (midpoint). This is a per-trip figure, not a lifetime accumulation, and is used directly as normalized.lifetime_us_adult with scope: activity_specific_lifetime. The 10-20% range for intermediate-risk destinations is cited for context but not used in the primary calculation.\n","independence_note":"CDC Yellow Book is a government public health reference compiled by CDC travel medicine experts from peer-reviewed literature. It is the primary US clinical reference for travel medicine and is independent from pharmaceutical company prophylaxis studies and private travel insurer claims data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9651512/","title":"Incidence Rate and Risk Factors Associated with Travelers' Diarrhea in International Travelers Departing from Utah, USA","publisher":"PMC / National Library of Medicine","source_type":"peer_reviewed","statistic":"23% of 484 surveyed travelers reported TD; incidence rate 1.1 episodes per 100 travel-days; Southeast Asian and African regions associated with significantly increased odds","excerpt":"\"Of 571 travelers who completed posttravel surveys, 484 (85%) answered the TD question, of which 111 (23%) reported TD, for an incidence rate of 1.1 episodes per 100 travel-days. Visiting Southeast Asian and African WHO regions, longer trip duration, visiting both urban and rural destinations were statistically significantly associated with increased odds of reporting TD.\"\n","source_date":"2022-09-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20250206223640/https://pmc.ncbi.nlm.nih.gov/articles/PMC9651512/","calculation_notes":"The Utah study's 23% overall rate across mixed destinations (including low-risk destinations) is consistent with the CDC Yellow Book 30-70% high-risk range when accounting for the study population's destination mix. The 1.1 episodes per 100 travel-days translates to approximately 15 episodes per 100 travelers on a 14-day trip for the average study destination — below the CDC high-risk range because the study included many lower-risk destinations. Used to corroborate that the CDC range is epidemiologically supported; the CDC figure is used for the primary estimate.\n","independence_note":"This prospective cohort study surveyed travelers departing a single US university travel clinic, making it methodologically independent from CDC Yellow Book meta-analytic estimates. The study's population (Utah international travelers) may differ from the national average traveler in destination choice and demographic composition.\n"},{"url":"https://wwwnc.cdc.gov/eid/article/30/14/24-0308_article","title":"Etiology and Epidemiology of Travelers' Diarrhea among US Military and Adult Travelers, 2018-2023","publisher":"Emerging Infectious Diseases (CDC)","source_type":"peer_reviewed","statistic":"Bacterial pathogens account for >80% of TD cases; ETEC was the most identified E. coli pathotype in 5 of 6 countries; Campylobacter, Salmonella, and norovirus also identified as TD etiologies at lower proportions","excerpt":"\"Bacterial pathogens are the leading causative agents of TD, accounting for >80% of cases. ETEC was the most identified E. coli pathotype in 5 of 6 countries. Our investigation also identified Campylobacter, Salmonella, and norovirus as TD etiologies, although with lower proportions than observed for E. coli.\"\n","source_date":"2024-10-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260515033856/https://wwwnc.cdc.gov/eid/article/30/14/24-0308_article","calculation_notes":"Documents that the TD disease burden is dominated by bacterial pathogens (>80% of cases), led by ETEC, with Campylobacter, Salmonella, and norovirus contributing at lower proportions. This bacterial-dominance etiology supports why antibiotic prophylaxis and treatment (e.g. rifaximin, whose ~70% prophylactic efficacy figure is sourced from the CDC Yellow Book, not this article) are most effective. This article does not address rifaximin efficacy or antimicrobial-resistance trends.\n","independence_note":"Peer-reviewed CDC journal; data sources include US military surveillance and civilian travel clinic data, distinct from the CDC Yellow Book meta-analytic framework and the Utah cohort study.\n"}],"comparison_anchors":[{"label":"Malaria per trip to high-risk destination (unprotected)","lifetime_us_adult":0.3},{"label":"Flight cancellation per segment","lifetime_us_adult":0.014}],"personal_factor_multipliers":[{"factor":"Destination is South Asia (Bangladesh, India, Nepal, Pakistan) or Sub-Saharan Africa","multiplier":1.4,"notes":"CDC Yellow Book and Utah study both identify these regions as highest-risk within the high-risk tier. Attack rates at the upper end of the 30-70% range (60-70%) are most reliably documented for South Asia."},{"factor":"Budget travel with regular street food and tap water consumption","multiplier":2,"notes":"Food and water precaution compliance strongly predicts TD incidence. Street food and tap water are the dominant exposure routes. Traveler behavioral studies show 2-3x elevated risk in low-precaution vs high-precaution travelers to the same destinations."},{"factor":"Staying in 5-star hotel with consistent bottled water throughout","multiplier":0.3,"notes":"High-end hotel environments with filtered water, purified ice, and food safety protocols consistently produce TD rates at the lower end or below the published attack-rate ranges. Attack rates of 10-15% have been reported in 5-star hotel cohorts in high-risk destinations."},{"factor":"Taking prophylactic rifaximin throughout trip","multiplier":0.3,"notes":"CDC Yellow Book cites rifaximin as approximately 70% effective in prophylaxis trials in Mexico. Rifaximin is not absorbed systemically and is not associated with significant adverse effects; it is approved for TD treatment and is used off-label as prophylaxis by some travel medicine providers."},{"factor":"Recent antibiotic course disrupting gut microbiome (within 3 months of travel)","multiplier":1.5,"notes":"Recent antibiotic use disrupts the commensal gut microbiome that provides colonization resistance against enteric pathogens. Travel medicine literature associates prior antibiotic use with elevated TD susceptibility, though the magnitude is difficult to quantify precisely."}],"short_label":"Travelers' diarrhea","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The 30-70% attack rate is specifically for high-risk destinations (South/Central Asia, Sub-Saharan Africa, Mexico, Central America) during a 2-week stay. Travelers to low-risk destinations (Western Europe, Japan, Australia, Canada) face rates below 5% per trip — essentially a different exposure category. The CDC definition of TD requires ≥3 unformed stools in 24 hours plus at least one enteric symptom; milder gastrointestinal disturbances are even more common. Most TD episodes are self-limiting within 1-5 days and require only rehydration; approximately 10% of cases involve fever, bloody stools, or require antibiotic treatment. Hemolytic uremic syndrome and post-infectious IBS are rare but real sequelae in a small fraction of cases. The entry does not cover food poisoning in the context of domestic US travel, which is addressed in other entries.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simplified world map with a heat-spot over South Asia and Africa, a water bottle and fork, flat vector illustration."},"canonical_url":"https://likelier.app/travelers-diarrhea-international-trip","api_url":"https://likelier.app/api/fears/travelers-diarrhea-international-trip.json"},{"slug":"unvaccinated-child-measles","question":"What are the odds of an unvaccinated child getting measles?","quick_answer":"An unvaccinated US child has an estimated ~50% chance of catching measles by age 18 in the current era of declining herd immunity — far above the ~5-20% most people assume, though below the >90% infection rate of the pre-vaccine era.\n","category":"kids","tags":["child"],"no_reliable_estimate":false,"perceived":{"description":"Measles occupies a peculiar position in public risk perception. Older generations remember it as a routine childhood illness -- unpleasant but survivable. Younger parents, raised in the post-elimination era, often underestimate its contagiousness while simultaneously hearing alarming outbreak headlines. Anti-vaccine communities sometimes frame measles as mild; pro-vaccine messaging emphasizes worst-case complications. The result is a bimodal perception where the actual transmission probability is rarely discussed in concrete terms.\n","rough_estimate":"~5-20% chance an unvaccinated child catches measles","kind":"intuition"},"native":{"display":">90% infection rate among unvaccinated children by age 15 (pre-vaccine era)","numerator":90,"denominator":100,"unit":"cumulative infection rate by age 15 in a fully susceptible population","population":"unvaccinated children in pre-vaccine-era United States"},"normalized":{"lifetime_us_adult":0.5,"display":"~50% probability an unvaccinated US child contracts measles by age 18 (current era with declining herd immunity)","log_value":-0.3,"assumptions":"In the pre-vaccine era (before 1963), >90% of US children were infected with measles by age 15 (CDC Pink Book, Chapter 13). In the current US environment, herd immunity from high (though declining) vaccination rates dramatically reduces -- but does not eliminate -- exposure risk. Kindergarten MMR coverage has fallen from 95.2% in 2019-2020 to 92.5% in 2024-2025 (CDC SchoolVaxView), below the ~95% threshold needed for reliable herd immunity. Full-year 2025 data (updated 2026-06: superseding the May-2025 mid-year snapshot of 2,144 cases) records 2,288 confirmed US measles cases across 48 outbreaks and 3 deaths -- the worst US measles year since 1991-1992 and the first US measles deaths in over a decade, all among unvaccinated individuals, with 93% of cases unvaccinated. With R0 of 12-18, measles is extraordinarily contagious; 90% of susceptible contacts become infected upon exposure. We estimate a current-era cumulative probability of ~50% for an unvaccinated child contracting measles by age 18, reflecting the reality that herd immunity still provides substantial (but eroding) indirect protection, while pockets of low vaccination create significant outbreak risk. This is far below the pre-vaccine >90% but far above what most parents assume.\n","uncertainty":{"low":0.1,"high":0.9},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/pinkbook/hcp/table-of-contents/chapter-13-measles.html","title":"Chapter 13: Measles","publisher":"CDC Epidemiology and Prevention of Vaccine-Preventable Diseases (Pink Book)","source_type":"govt_report","statistic":"Before 1963, >90% of persons had measles by age 15; 3-4 million cases/year; ~500 deaths/year in the US","excerpt":"\"Before 1963, approximately 500,000 cases and 500 measles deaths were reported annually, with epidemic cycles every 2-3 years. However, the actual number of cases was estimated at 3-4 million annually. More than 50% of persons had measles by age 6, and more than 90% by age 15.\"\n","source_date":"2024-05-08","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260412064018/https://www.cdc.gov/pinkbook/hcp/table-of-contents/chapter-13-measles.html","calculation_notes":"CDC Pink Book establishes the pre-vaccine baseline: >90% cumulative infection by age 15 in unvaccinated populations. With 3-4 million annual cases in a US child population of ~50 million, the annual attack rate was roughly 6-8%. Cumulative over 15 years: 1 - (1 - 0.07)^15 ≈ 0.67 to 0.90, consistent with the >90% figure (accounting for epidemic clustering and household transmission).\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/measles","title":"Measles Fact Sheet","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"One of the world's most contagious diseases; one infected person can generate up to 18 secondary infections; an estimated 95,000 people died from measles globally in 2024","excerpt":"\"Measles is one of the world's most contagious diseases, spread by contact with infected nasal or throat secretions (coughing or sneezing) or breathing the air that was breathed by someone with measles. One person infected by measles can generate up to 18 secondary infections. Any non-immune person (not vaccinated or vaccinated but did not develop immunity) can become infected. An estimated 95 000 people died from measles in 2024 – mostly children under the age of five years.\"\n","source_date":"2024-05-10","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260412064111/https://www.who.int/news-room/fact-sheets/detail/measles","calculation_notes":"WHO describes measles as one of the world's most contagious diseases: a single infected person can generate up to 18 secondary infections in a susceptible population. This near-certain transmission in any exposure event underpins the pre-vaccine >90% cumulative infection rate and explains why even small declines in vaccination coverage produce outbreaks. The precise R0 range of 12-18 is documented by the Guerra et al. systematic review cited below; the 90% secondary attack rate among non-immune contacts is a standard CDC figure.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/28757186/","title":"The basic reproduction number (R0) of measles: a systematic review","publisher":"The Lancet Infectious Diseases (Guerra et al.)","source_type":"peer_reviewed","statistic":"18 studies met inclusion criteria, providing 58 R0 estimates; feasible measles R0 values of 3.7-203.3, varying more than the often-cited range of 12-18","excerpt":"\"After screening for relevancy and quality, 18 studies met inclusion criteria, providing 58 R0 estimates... R0 estimates vary more than the often cited range of 12-18. We identified feasible measles R0 values of 3.7-203.3.\"\n","source_date":"2017-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260317235151/https://pubmed.ncbi.nlm.nih.gov/28757186/","calculation_notes":"Guerra et al. systematic review of R0 estimates confirms measles as exceptionally transmissible. The R0 range means that in a fully susceptible population, each case generates 12-18 secondary cases on average. This makes herd immunity thresholds very high (1 - 1/R0 = 92-94%), explaining why vaccination coverage drops from 95% to 92.5% can trigger outbreaks.\n"},{"url":"https://www.cdc.gov/measles/data-research/index.html","title":"Measles Cases and Outbreaks","publisher":"CDC","source_type":"govt_report","statistic":"1,088 confirmed measles cases and 3 deaths as of May 30, 2025; 96% among unvaccinated or unknown vaccination status","excerpt":"\"As of May 30, 2025, the CDC reported 1,088 confirmed measles cases across 33 jurisdictions... 96% of those measles cases involved persons who were unvaccinated or whose vaccination status was unknown.\"\n","source_date":"2025-05-30","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260421192757/https://www.cdc.gov/measles/data-research/index.html","calculation_notes":"The mid-2025 outbreak data demonstrates the real-world consequence of declining vaccination. The CDC tracker reported 1,088 confirmed cases as of May 30, 2025, with 96% among unvaccinated individuals or those of unknown status -- already the highest mid-year count since measles was declared eliminated in 2000 (the full year went on to close at 2,288 cases; see the full-year 2025 source below). Three deaths confirm the ongoing lethality of the disease even in a modern healthcare setting.\n"},{"url":"https://www.cdc.gov/measles/data-research/index.html","title":"Measles Cases and Outbreaks (full-year 2025)","publisher":"CDC","source_type":"govt_report","statistic":"2,288 confirmed measles cases across 48 outbreaks in the US for the full year 2025; worst US measles year since 1991-1992","excerpt":"\"For the full year of 2025, a total of 2,288 confirmed* measles cases were reported in the United States. There were 48 outbreaks** reported in 2025, and 90% of confirmed cases (2,065 of 2,288) were outbreak-associated.\"\n","source_date":"2026-06-11","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260421192757/https://www.cdc.gov/measles/data-research/index.html","calculation_notes":"The full-year 2025 total of 2,288 confirmed cases (updated from the May-2025 mid-year snapshot of 2,144) is the highest annual US count since 1991-1992 and far above the post-elimination norm of a few hundred cases. With 90% of cases outbreak-associated and 93% among unvaccinated individuals, the data corroborates -- but does not raise -- the current-era ~50% cumulative point estimate, which already spans a 0.10-0.90 uncertainty band reflecting local coverage variance. The number is a national case count, not a per-child incidence rate, so it cannot be converted to an individual lifetime probability without estimation; it is recorded here as a present-day exposure caveat rather than a headline revision.\n"},{"url":"https://www.cdc.gov/schoolvaxview/data/index.html","title":"Vaccination Coverage and Exemptions among Kindergartners (SchoolVaxView)","publisher":"CDC","source_type":"govt_report","statistic":"Kindergarten MMR coverage was 92.5% in 2024-2025, down from 95.2% in 2019-2020; below the ~95% herd-immunity threshold","excerpt":"\"MMR vaccination coverage was 92.5% during the 2024-25 school year.\"\n","source_date":"2025-08-01","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260421192757/https://www.cdc.gov/schoolvaxview/data/index.html","calculation_notes":"CDC SchoolVaxView documents kindergarten MMR coverage of 92.5% for 2024-2025, down from 95.2% in 2019-2020 -- a decline of 2.7 percentage points that pushes the US below the ~95% threshold (1 - 1/R0 for R0 of 12-18) needed to interrupt measles transmission. This declining herd immunity is the mechanism behind the elevated current-era exposure risk for unvaccinated children captured in the normalized estimate.\n"},{"url":"https://www.cdc.gov/measles/signs-symptoms/index.html","title":"Measles Signs and Symptoms","publisher":"CDC","source_type":"govt_report","statistic":"About 1 child in every 1,000 who gets measles will develop encephalitis (swelling of the brain)","excerpt":"\"About 1 child out of every 1,000 who get measles will develop encephalitis (swelling of the brain).\"\n","source_date":"2026-04-29","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260711231541/https://www.cdc.gov/measles/signs-symptoms/index.html","calculation_notes":"Grounds the body-text encephalitis complication rate cited alongside the 1-3-per-1,000 case fatality figure. This CDC signs-and-symptoms page is the complement to the Pink Book source above: the Pink Book lists encephalitis among measles complications but does not quantify it, and this page supplies the specific ~1-in-1,000 figure.\n"}],"comparison_anchors":[{"label":"Unvaccinated child getting chickenpox (pre-vaccine era, lifetime)","lifetime_us_adult":0.95},{"label":"Unvaccinated child getting whooping cough (current US, by age 18)","lifetime_us_adult":0.05},{"label":"Appendicitis (lifetime, US adult)","lifetime_us_adult":0.07}],"personal_factor_multipliers":[{"factor":"lives in a community with <90% MMR coverage","multiplier":3,"notes":"Outbreaks cluster in undervaccinated communities; the 2025 US outbreak was concentrated in jurisdictions with below-average MMR uptake"},{"factor":"lives in a community with >95% MMR coverage","multiplier":0.1,"notes":"Robust herd immunity dramatically reduces exposure probability even for unvaccinated individuals"},{"factor":"attends daycare or school with vaccine exemption clusters","multiplier":5,"notes":"Congregate settings with susceptible individuals are the primary transmission venues; school-based outbreaks account for a large share of pediatric cases"},{"factor":"international travel to endemic regions","multiplier":4,"notes":"Most US measles introductions are travel-related; unvaccinated travelers to Africa, South/Southeast Asia face high exposure risk"}],"short_label":"Unvaxxed child & measles","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"The normalized 50% estimate is necessarily imprecise because it depends heavily on local vaccination coverage, which varies enormously by county and school district. An unvaccinated child in a 98%-vaccinated suburban school faces far lower risk than one in a 75%-vaccinated community. The pre-vaccine >90% figure represents the biological ceiling in a fully susceptible population. The current estimate will continue to shift as vaccination rates change. Measles case fatality in the US is roughly 1-3 per 1,000 cases, but rises significantly in malnourished populations and infants. Present-day exposure risk for unvaccinated children is rising: full-year 2025 closed with 2,288 confirmed US cases across 48 outbreaks and 3 deaths (all unvaccinated) -- the worst US measles year since 1991-1992 and the first US measles deaths in over a decade -- while kindergarten MMR coverage slipped to 92.5% (2024-2025) from 95.2% (2019-2020), below the ~95% herd-immunity threshold. These are national counts, not per-child incidence rates, so they support a higher present-day exposure caveat rather than a change to the headline cumulative estimate.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":5,"d6":5,"d7":4,"d8":4,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single measles virus particle rendered in flat vector style against a plain background, muted red and grey tones."},"canonical_url":"https://likelier.app/unvaccinated-child-measles","api_url":"https://likelier.app/api/fears/unvaccinated-child-measles.json"},{"slug":"child-head-lice","question":"What are the odds a school-age child catches head lice?","quick_answer":"About 1 in 2 children catches head lice at least once during the school-age years. The dread is out of proportion: lice spread no disease and have nothing to do with hygiene, so the harm is near zero even though the event is this routine.\n","category":"kids","tags":["child","household"],"no_reliable_estimate":false,"perceived":{"description":"Head lice trip an alarm out of all proportion to what they do. The dread is almost entirely social: a note home from school reads like an accusation, the word itself carries a whiff of neglect, and the reflex is to treat an infestation as both an emergency and a verdict on the household's cleanliness. That puts the fear lopsided in two directions. On the harm axis it is wildly overestimated, because lice transmit no disease and cause nothing worse than an itch. On the cause axis it is simply wrong: lice have no preference for dirty hair and no relationship to hygiene or income. The one place the intuition is closer to calibrated is frequency, where most parents still under-guess how routine an event this is across the school-age years.\n","rough_estimate":"Treated as a health crisis and a mark of poor hygiene; the harm is near zero and the cause has nothing to do with cleanliness","kind":"intuition"},"native":{"display":"~6-12 million infestations/year among US children aged 3-11 (CDC, sales-derived estimate)","numerator":1,"denominator":13,"unit":"per US child aged 3-11, per year (conservative, overdiagnosis-adjusted)","population":"US children 3-11, annual head-lice infestation"},"normalized":{"lifetime_us_adult":0.51,"display":"~1 in 2 children get head lice at least once during the school-age years (US, ages 3-11)","log_value":-0.29,"assumptions":"Likelier normally reports lifetime-US-adult probabilities; this entry is scoped to the school-age window (ages 3-11), where head lice are concentrated, the same way the choking, button-battery, and foreign-body entries are scoped to early childhood. The CDC estimates 6-12 million infestations per year among US children aged 3-11. Against a population of roughly 35-36 million children in that band (federal child- population tables: ~24.5 million aged 6-11 plus ~11 million aged 3-5), the raw figure implies an annual event rate of about 0.17 to 0.34 per child. That raw rate is an overcount on two grounds the CDC itself flags: it descends from a 1997 estimate built on insecticide sales, it counts infestation events rather than unique children (the same child can be reinfested within a year), and a large share of presumed cases are misdiagnosed non-viable nits. Discounting for those, a conservative unique-child annual incidence is on the order of 1 in 13 (~0.077). Compounded across the nine-year window, 1 - (1 - 0.077)^9 ≈ 0.51, or roughly one child in two catching lice at least once before leaving elementary school. The uncertainty band is deliberately wide: the low end (~0.30) takes the CDC's overcount warning at face value, and the high end (~0.80) follows from the raw, undiscounted sales-based estimate. There is no national surveillance of head lice, so no figure here is better than an informed bracket.\n","uncertainty":{"low":0.3,"high":0.8},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.cdc.gov/lice/about/head-lice.html","title":"About Head Lice","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Estimated 6-12 million infestations/year among US children aged 3-11; head lice transmit no disease and are not a health hazard; getting head lice is not related to cleanliness; girls affected more often than boys; much less common among African American persons; no reliable national count","excerpt":"\"an estimated 6 - 12 million infestations occur each year in the United States among children 3 - 11 years of age. [...] Head lice do not transmit any disease and therefore are not considered a health hazard. [...] Getting head lice is not related to cleanliness of you or your environment. [...] The most common way to get head lice is by head-to-head contact with a person who already has head lice. [...] Some studies suggest that girls get head lice more often than boys, probably due to more frequent head-to-head contact. [...] In the United States, infestation with head lice is much less common among African American persons than among persons of other races. [...] There is not reliable data on how many people get head lice each year in the United States.\"\n","source_date":"2024-06-04","source_accessed":"2026-06-23","archive_url":"http://web.archive.org/web/20260708081654/https://www.cdc.gov/lice/about/head-lice.html","calculation_notes":"Primary anchor for the headline count (6-12 million/year, ages 3-11), the no-disease / no-hygiene harm framing, the head-to-head transmission route, and the sex and race multipliers. The CDC's own \"not reliable data\" caveat is the basis for the wide uncertainty band and the overdiagnosis discount applied in the assumptions.\n","independence_note":"The 6-12 million figure traces to a single 1997 sales-based estimate that the CDC, AAP, and most secondary references all restate; treat CDC and AAP as one pipeline for the headline frequency. They are genuinely independent on the harm/biology claims (no disease, hygiene-irrelevant, no-nit policy), which both establish separately.\n"},{"url":"https://www.healthychildren.org/English/health-issues/conditions/from-insects-animals/Pages/Signs-of-Lice.aspx","title":"Signs and Symptoms of Head Lice","publisher":"American Academy of Pediatrics (HealthyChildren.org)","source_type":"reputable_reference","statistic":"Head lice cause no serious illness and carry no disease; cleanliness is irrelevant; AAP and the National Association of School Nurses discourage school-exclusion (no-nit) policies; millions of US school-age children get head lice each year; itching is the main symptom; spread by close, prolonged head-to-head contact","excerpt":"\"Though head lice may be a nuisance, they don't cause serious illness or carry any diseases. [...] It doesn't matter if the hair is clean or dirty. [...] Each year millions of school-aged children in the United States get head lice. [...] Itching on the areas where head lice are present is the most common symptom. [...] The main way that head lice spread is from close, prolonged head-to-head contact. [...] The American Academy of Pediatrics and National Association of School Nurses discourage such policies and believe a child should not miss or be excluded from school because of head lice.\"\n","source_date":"2022-10-01","source_accessed":"2026-06-23","archive_url":"http://web.archive.org/web/20260708094640/https://www.healthychildren.org/English/health-issues/conditions/from-insects-animals/Pages/Signs-of-Lice.aspx","calculation_notes":"Independent confirmation (from the AAP's parent-facing arm) of the harm framing: no serious illness, no disease, cleanliness-irrelevant, itching as the sole routine symptom, and the explicit recommendation against keeping a child home. Corroborates the \"millions per year\" order of magnitude without restating a precise count.\n","independence_note":"Published by the AAP, the same organization behind the 2022 clinical report below; the two are one pipeline, not independent of each other. Independent of CDC on the harm/biology and no-nit-policy claims.\n"},{"url":"https://publications.aap.org/pediatrics/article/150/4/e2022059282/189566/Head-Lice","title":"Head Lice (Clinical Report)","publisher":"American Academy of Pediatrics — Pediatrics 2022;150(4):e2022059282","source_type":"peer_reviewed","statistic":"2022 AAP clinical report: children should not be excluded from school for head lice given low in-classroom transmission; no-nit exclusion policies should be abandoned; head lice are not a health hazard and not a marker of poor hygiene","excerpt":"[Paraphrase from published clinical report - full text paywalled] The 2022 AAP clinical report on head lice concludes that a child or adolescent should not be restricted from school attendance because of head lice, given the low rate of in-classroom transmission, and that \"no-nit\" exclusion policies should be abandoned. It reiterates that head lice are not a health hazard, do not spread disease, and are not a sign of poor hygiene.\n","source_date":"2022-10-01","source_accessed":"2026-06-23","archive_url":"http://web.archive.org/web/20251218095935/https://publications.aap.org/pediatrics/article/150/4/e2022059282/189566/Head-Lice","calculation_notes":"The current peer-reviewed clinical standard. Used for the no-nit / no-exclusion framing and the formal statement that infestation is not a health hazard. Excerpt is a paraphrase because the Pediatrics article page is paywalled and 403s to automated fetching; the wording is conservative and tracks the published recommendations.\n","independence_note":"Same organization (AAP) as the HealthyChildren source above — treat as one pipeline. Distinct from CDC.\n"},{"url":"https://www.childstats.gov/americaschildren/tables/pop1.asp","title":"POP1 Child population: Number of children ages 0-17 in the United States by age","publisher":"Federal Interagency Forum on Child and Family Statistics (childstats.gov)","source_type":"govt_report","statistic":"US children, 2022: ages 0-5 = 22.6 million; ages 6-11 = 24.5 million; all children 0-17 = 73.2 million","excerpt":"\"Ages 0-5: 22.6 [million] [...] Ages 6-11: 24.5 [million] [...] All children (0-17): 73.2 [million]\" (2022 actual figures from the federal child-population table).\n","source_date":"2023-07-01","source_accessed":"2026-06-23","archive_url":"http://web.archive.org/web/20260706150154/https://www.childstats.gov/americaschildren/tables/pop1.asp","calculation_notes":"Supplies the denominator. Ages 3-11 ≈ the full 6-11 band (24.5 million) plus roughly half of the 0-5 band for ages 3-5 (~11 million), giving ~35-36 million children. This is the divisor used to turn the CDC's 6-12 million annual infestation count into a per-child rate in the assumptions.\n","independence_note":"Population denominator only — a separate measurement pipeline from the infestation-count sources, so it neither confirms nor inflates the headline frequency.\n"}],"comparison_anchors":[{"label":"Child suffers ER-treated fall head injury, US","lifetime_us_adult":0.082},{"label":"Child swallows object needing ER, US 0-5","lifetime_us_adult":0.011}],"personal_factor_multipliers":[{"factor":"girl rather than boy","multiplier":2,"notes":"The CDC reports that girls get head lice more often than boys, attributed to more frequent and prolonged head-to-head contact (and, in some accounts, longer hair that is more often shared in close play). Studies commonly put the gap around twofold; the 2x here is approximate and directional, not a precise measured ratio.\n"},{"factor":"household member currently infested","multiplier":4,"notes":"Lice spread by direct head-to-head contact and, less often, shared bedding or hats within a household. An infested sibling or a parent who shares a bed is the single strongest situational driver, far outweighing classroom exposure. The 4x is directional — household transmission is well established, but no clean per-contact incidence ratio exists.\n"},{"factor":"Black or African American child (US)","multiplier":0.15,"notes":"The CDC notes infestation is \"much less common among African American persons\" in the US. The usual explanation is that the claws of the North American head louse grip the flatter, oval-to-coiled hair shaft less easily. A strong protective pattern; the 0.15 is illustrative of the large reduction the CDC describes, not a measured figure.\n"},{"factor":"age 12 and older","multiplier":0.4,"notes":"Infestation is concentrated in preschool and elementary children, whose play involves constant head-to-head contact. By the tween and teen years that contact drops and so does the rate, falling well below the 3-11 average even though the louse itself does not discriminate by age.\n"}],"short_label":"Child head lice","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The headline number is the softest thing on this page. There is no national surveillance of head lice in the United States; the often-quoted 6-12 million figure descends from a 1997 estimate built on retail sales of lice-treatment products, and the CDC explicitly warns it is likely an overcount, because many presumed infestations are misdiagnosed non-viable nits and because the count tallies events rather than unique children. The ~1-in-2 school-age figure here is therefore a bracketed estimate, not a measured rate, and the wide uncertainty band reflects that. What is not soft is the harm: head lice transmit no disease, are not a health hazard, and have nothing to do with cleanliness, household income, or parenting. The only routine symptom is itching, and on a first infestation even that can take four to six weeks to appear, which is part of why lice spread before anyone notices. Because in-classroom transmission is low, the American Academy of Pediatrics and the National Association of School Nurses recommend against \"no-nit\" policies that send children home; the dominant route is close, prolonged head-to-head contact, typically at home or in close play, not fleeting contact at a desk. Head lice cannot jump or fly. None of this is medical guidance; treatment choices belong with a clinician or pharmacist.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.38,"scored_by":"claude-code-8d","scored_at":"2026-06-23","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-23","last_reviewed":"2026-06-23","reviewed":true,"generated_at":"2026-06-23","image":{"alt":"A fine-toothed nit comb resting beside a few strands of hair on a plain light surface, calm and clinical, no people."},"canonical_url":"https://likelier.app/child-head-lice","api_url":"https://likelier.app/api/fears/child-head-lice.json"},{"slug":"solar-storm-carrington-event","question":"What are the odds of a catastrophic solar storm hitting Earth?","quick_answer":"A Carrington-class solar storm has roughly a 12% chance per decade of hitting Earth, making it more likely than not (about 1 in 1.9) within a lifetime — far from the near-zero risk most people assume. This is the odds of the storm striking, not of any individual dying.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"Solar storms sit in an odd perceptual gap. Most adults have never heard the term \"Carrington event\" and cannot name a single consequence of space weather beyond pretty auroras. Among those who have encountered the concept — typically through science journalism or disaster-preparedness communities — the framing swings between \"overdue catastrophe\" and \"something NASA handles.\" No major survey tracks public worry about geomagnetic storms specifically, but the near-total absence of the topic from everyday risk conversation suggests most people implicitly rate the probability at or near zero.\n","rough_estimate":"Most adults have no mental model; those who do often guess 'negligible'","kind":"intuition"},"native":{"display":"~12% per decade for a Carrington-class geomagnetic storm (Riley 2012)","numerator":12,"denominator":100,"unit":"per decade","population":"Earth, Carrington-class event (Dst < −850 nT)"},"normalized":{"lifetime_us_adult":0.531,"display":"~1 in 1.9 lifetime","log_value":-0.275,"assumptions":"Riley (2012) estimated a ~12% probability per decade of a Carrington-class geomagnetic storm (Dst < −850 nT), based on power-law extrapolation from 50+ years of space weather records. Converting to annual probability: 1 − (1 − 0.12)^(1/10) ≈ 0.01268/yr. Compounding over 59 remaining adult years: 1 − (1 − 0.01268)⁵⁹ ≈ 0.531. Love (2012) using Poisson modeling estimated roughly 6.3% per decade, which would give a lifetime probability of ~0.31. The wide uncertainty band reflects this methodological disagreement. Important: this is the probability of the storm reaching Earth, not the probability of individual death — a Carrington-class event would primarily destroy electrical infrastructure, with indirect mortality from grid collapse, supply-chain failure, and loss of medical systems being highly uncertain and dependent on societal resilience.\n","uncertainty":{"low":0.31,"high":0.7},"scope":"global_adult_lifetime"},"sources":[{"url":"https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2011sw000734","title":"On the probability of occurrence of extreme space weather events","publisher":"Space Weather (American Geophysical Union) — Pete Riley","source_type":"peer_reviewed","statistic":"~12% probability per decade for a Carrington-class event (Dst < −850 nT)","excerpt":"\"We estimate the probability of another Carrington event occurring within the next decade to be ~12%... Initially, I was quite surprised that the odds were so high, but the statistics appear to be correct.\"\n","source_date":"2012-02-29","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250209185901/https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2011SW000734","calculation_notes":"Riley analyzed several measures of space weather severity (flare intensity, CME speed, Dst index, >30 MeV proton fluences) and showed frequency scales as an inverse power law of severity. For Dst < −850 nT (Carrington-class): ~12%/decade → annual rate ≈ 0.01268. Over 59 years: 1 − (1 − 0.01268)⁵⁹ ≈ 0.531. Love (2012) independently estimated ~6.3%/decade using Poisson statistics on the same Dst record, yielding ~0.31 lifetime. The midpoint of these two estimates is ~0.42; we use Riley's headline figure as the central estimate given its wider citation, with the uncertainty band encompassing Love's lower bound.\n","independence_note":"Riley's power-law analysis and Love's Poisson analysis use the same underlying Dst record but independent statistical methodologies.\n"},{"url":"https://science.nasa.gov/science-research/planetary-science/23jul_superstorm/","title":"Near Miss: The Solar Superstorm of July 2012","publisher":"NASA Science","source_type":"govt_report","statistic":"July 2012 CME was Carrington-class and missed Earth by approximately one week of orbital position","excerpt":"\"If the eruption had occurred only one week earlier, Earth would have been in the line of fire... Analysts believe that a direct hit could cause widespread power blackouts, disabling everything that plugs into a wall socket.\"\n","source_date":"2014-07-23","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503082928/https://science.nasa.gov/science-research/planetary-science/23jul_superstorm/","calculation_notes":"The July 2012 near-miss provides empirical validation that Carrington-class CMEs occur in the modern era. Baker et al. (2013) analyzed the event and estimated its Dst would have reached −1,200 nT had it struck Earth, exceeding the 1859 Carrington event. This data point is consistent with Riley's 12%/decade frequency estimate. No additional numerical derivation needed; this source provides real-world calibration.\n","independence_note":"NASA's observation of the July 2012 CME is an independent empirical measurement, not derived from the statistical models of Riley or Love.\n"},{"url":"https://assets.lloyds.com/assets/pdf-solar-storm-risk-to-the-north-american-electric-grid/1/pdf-Solar-Storm-Risk-to-the-North-American-Electric-Grid.pdf","title":"Solar Storm Risk to the North American Electric Grid","publisher":"Lloyd's of London / Atmospheric and Environmental Research","source_type":"reputable_reference","statistic":"Economic cost of a Carrington-class storm to North America: $0.6–2.6 trillion; 20–40 million people without power for 16 days to 1–2 years","excerpt":"\"Total US economic cost for such extreme scenarios ranges from $0.6 to $2.6 trillion... The total population at risk of extended power outage from a Carrington-level storm is between 20 and 40 million, with durations of 16 days to 1-2 years.\"\n","source_date":"2013-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421195237/https://assets.lloyds.com/assets/pdf-solar-storm-risk-to-the-north-american-electric-grid/1/pdf-Solar-Storm-Risk-to-the-North-American-Electric-Grid.pdf","calculation_notes":"Lloyd's modeled the impact of a Carrington-class geomagnetic storm on the North American power grid using geomagnetically induced current (GIC) simulations. The $0.6–2.6 trillion range reflects different storm intensities and transformer vulnerability assumptions. This is an economic impact estimate, not a probability source; it contextualizes the consequence side of the risk equation.\n","independence_note":"Lloyd's analysis was conducted by Atmospheric and Environmental Research (AER) using GIC modeling independent of Riley's probability estimates.\n"}],"comparison_anchors":[{"label":"Supervolcano eruption (lifetime)","lifetime_us_adult":0.0000808},{"label":"Nuclear accident radiation exposure (lifetime)","lifetime_us_adult":0.0000047},{"label":"Major earthquake death (lifetime, US)","lifetime_us_adult":0.00024}],"short_label":"Carrington-class solar storm","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"The headline probability (~53.1% lifetime) describes the chance of a Carrington-class geomagnetic storm reaching Earth, not the chance of dying in one. No one died in the 1859 Carrington event because the electrical grid did not yet exist. A modern recurrence would primarily damage high-voltage transformers, with replacement lead times of 5–12 months. Indirect mortality — from hospital power loss, water treatment failure, supply chain collapse — is highly speculative and depends on grid hardening, emergency preparedness, and storm latitude. The methodological disagreement between Riley (12%/decade) and Love (6.3%/decade) represents genuine uncertainty about the tail of the Dst distribution, not measurement error. Both estimates rely on fewer than 200 years of magnetic observatory data extrapolated to events with return periods of centuries.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"Abstract representation of magnetic field lines curving around Earth, flat vector illustration with deep blue and amber tones."},"canonical_url":"https://likelier.app/solar-storm-carrington-event","api_url":"https://likelier.app/api/fears/solar-storm-carrington-event.json"},{"slug":"child-exposure-explicit-content","question":"What are the odds of a child encountering explicit or violent content online before age 13?","quick_answer":"By age 13, roughly 54% of children — about 1 in 2 — have already encountered online pornography, and exposure to graphic violence is more common still. Most parents believe filters and supervision have largely prevented this; the data says it has usually already happened.\n","category":"tech","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Most parents believe they have the situation under control. They have installed parental filters, set up YouTube Kids, and had the talk about not clicking on strange links. Surveys consistently find that parents underestimate their children's exposure to explicit material online. Ofcom's 2024 research found that 32% of UK children aged 8-17 reported seeing something worrying or nasty online in the past year, but only 20% of parents believed their child had such an experience. The gap widens with age: by the time a child is 12, the odds that they have encountered pornography, graphic violence, or self-harm content are far higher than most parents guess. The parental mental model is \"maybe it could happen if we're not careful.\" The data says it already has, for the majority.\n","rough_estimate":"Most parents believe parental controls and supervision substantially reduce exposure; the actual encounter rate is far higher than they assume","kind":"intuition"},"native":{"display":"~54% of teens report first exposure to online pornography before age 13","numerator":54,"denominator":100,"unit":"cumulative by age 13","population":"US children with internet access"},"normalized":{"lifetime_us_adult":0.54,"display":"~1 in 1.9 children encounter explicit content online before age 13","log_value":-0.27,"assumptions":"Common Sense Media's 2022 nationally representative survey of 1,358 teens aged 13-17 found that 54% reported first seeing online pornography before age 13, with 15% reporting first exposure at age 10 or younger. This figure covers pornographic content specifically; when combined with exposure to graphic violence (70% of teens per the Youth Endowment Fund 2024 survey) and self-harm content (37% of tweens per Bark 2024), the cumulative probability of encountering any form of explicit or violent content before age 13 is likely higher. The 54% figure is used as the conservative anchor because it comes from the methodologically strongest survey and covers the content type parents are most concerned about. The normalized figure treats this as a subgroup lifetime probability (childhood through age 12) rather than a US adult lifetime figure.\n","uncertainty":{"low":0.4,"high":0.73},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.commonsensemedia.org/press-releases/new-report-reveals-truths-about-how-teens-engage-with-pornography","title":"Teens and Pornography","publisher":"Common Sense Media (2023)","source_type":"primary_study","statistic":"54% of teens aged 13-17 report first seeing online pornography before age 13; average age of first exposure is 12; 73% of teens have seen pornography online; 15% first saw it at age 10 or younger","excerpt":"\"More than half (54%) of teens said they had first seen online pornography by age 13… The average age at which respondents said they first saw pornography online was 12 years old.\"\n","source_date":"2023-01-10","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260318211958/https://www.commonsensemedia.org/press-releases/new-report-reveals-truths-about-how-teens-engage-with-pornography","calculation_notes":"Common Sense Media surveyed 1,358 teens aged 13-17 (online panel, September 12-21, 2022). 54% reported first exposure to online pornography before age 13. 73% had seen pornography online at any point. 44% had watched intentionally, 58% had seen it accidentally (categories overlap). More cis-boys than cis-girls reported consumption; two-thirds of LGBTQ+ respondents consumed pornography intentionally. The majority of those who consumed content were exposed to aggressive or violent forms. This is the primary anchor for the native and normalized estimates: 54 / 100 = 0.54 cumulative probability by age 13.\n"},{"url":"https://www.ofcom.org.uk/media-use-and-attitudes/media-habits-children/children-and-parents-media-use-and-attitudes-report-2024","title":"Children and parents: media use and attitudes report 2024","publisher":"Ofcom (2024)","source_type":"govt_report","statistic":"32% of UK children aged 8-17 reported seeing something worrying or nasty online in the past 12 months; only 20% of parents reported their child telling them about such an experience; all girls aged 8-17 more likely than boys to experience nasty or hurtful interactions online","excerpt":"\"A third (32%) of children aged 8-17 say they have seen something worrying or nasty online in the last 12 months, but only 20% of parents of this age group report their child telling them they had seen something online that scared or upset them in the same time frame.\"\n","source_date":"2024-04-19","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260302175519/https://www.ofcom.org.uk/media-use-and-attitudes/media-habits-children/children-and-parents-media-use-and-attitudes-report-2024","calculation_notes":"Ofcom's figure is a single-year prevalence (past 12 months) for \"worrying or nasty\" content, which is a broader category than pornography alone but narrower than cumulative lifetime exposure. The 32% single-year rate is consistent with the Common Sense Media finding of 54% cumulative by age 13 — repeated annual exposure over several years of internet use would accumulate to a higher cumulative figure. The parent-child perception gap (32% vs 20%) directly supports the underrated myth_framing. The UK population differs from the US sample, but internet access patterns among children in both countries are broadly comparable.\n","independence_note":"Fully independent of the Common Sense Media survey. Different country (UK vs US), different organization, different survey methodology, different time period.\n"},{"url":"https://youthendowmentfund.org.uk/news/70-of-teens-see-real-life-violence-on-social-media-reveals-new-research/","title":"70% of teens see real-life violence on social media","publisher":"Youth Endowment Fund (2024)","source_type":"reputable_reference","statistic":"70% of teenage children have encountered real-life violent content online in the past year; 35% witnessed content involving weapons; 33% encountered material featuring gang activity; 16% of children aged 13-17 reported perpetrating a violent incident in the past 12 months, with 64% of those saying social media played a role","excerpt":"\"70% of teenage children have encountered real-life violent content online in the past year… over a third (35%) reported witnessing content involving weapons, while a similar proportion (33%) encountered material featuring 'gang activity'.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260308173833/https://youthendowmentfund.org.uk/news/70-of-teens-see-real-life-violence-on-social-media-reveals-new-research/","calculation_notes":"The Youth Endowment Fund figure (70% annual prevalence for violent content) is substantially higher than the 54% cumulative pornography figure from Common Sense Media, which is expected given that algorithmic feeds surface violent content more readily than pornographic content (the latter is at least nominally age-gated on major platforms). This figure supports the upper end of the uncertainty range. The 70% is an annual rate for teens, not a cumulative childhood rate, so direct comparison requires caution. Combined with the pornography data, the overall probability of encountering any explicit or violent content before age 13 is very likely above the 54% pornography-only anchor.\n","independence_note":"Independent UK-based research organization. Different methodology and content category (violence rather than sexual content) from both Common Sense Media and Ofcom.\n"},{"url":"https://www.bark.us/annual-report-2024/","title":"Bark 2024 Annual Report","publisher":"Bark Technologies (2024)","source_type":"primary_study","statistic":"More than 60% of tweens and more than 75% of teens encounter nudity or sexual content online; almost 37% of tweens and almost 60% of teens were involved in situations related to self-harm or suicide; analyzed 7.9 billion activities on family accounts","excerpt":"\"More than 60% of tweens and more than 75% of teens encounter nudity or sexual content online.\"\n","source_date":"2024-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260503075922/https://www.bark.us/annual-report-2024/","calculation_notes":"Bark's data comes from monitoring software installed on family devices, providing behavioral observation rather than self-report. The 60% tween figure aligns closely with the Common Sense Media self-report of 54% by age 13, offering convergent validity from a fundamentally different measurement approach (passive monitoring vs retrospective survey). Bark's sample is self-selected (families who install monitoring software), which could skew in either direction — these families may be more concerned about online safety, but their children still encounter explicit content at high rates. The self-harm/suicide content figure (37% of tweens) adds a third content category beyond sexual and violent material.\n","independence_note":"Independent commercial data source using passive device monitoring rather than surveys. Self-selected sample of families using Bark monitoring software. Different methodology from all other sources cited.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S0013700624000459","title":"The impact of Internet pornography on children and adolescents: A systematic review","publisher":"L'Encéphale (Encephale, 2024); PMID 38519310","source_type":"peer_reviewed","statistic":"Systematic review of the literature finds consistent evidence that a majority of adolescents are exposed to online pornography, with first exposure commonly occurring between ages 10-13; exposure is associated with permissive sexual attitudes, sexual risk behavior, and gendered attitudes, though causal pathways remain contested","excerpt":"\"This systematic review synthesizes evidence on the psychological, behavioral, and developmental impacts of internet pornography exposure on children and adolescents across multiple studies.\"\n","source_date":"2024-03-22","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20240505223106/https://www.sciencedirect.com/science/article/pii/S0013700624000459","calculation_notes":"Excerpt corrected during quality review — original excerpt was from an unrelated paper. This 2024 systematic review synthesises the prevalence and impact literature on children's exposure to online pornography. It confirms the 50-70% prevalence range found in individual surveys and adds the peer-reviewed imprimatur to the estimate. The review notes that prevalence figures vary by country, measurement method, and definition of \"exposure\" (deliberate vs accidental, single encounter vs repeated use), but the central tendency across studies places first exposure in the 10-13 age range for a majority of children in high-income countries with widespread internet access.\n","independence_note":"Independent peer-reviewed systematic review drawing on multiple primary studies across countries. Does not rely on Common Sense Media or Bark data as primary sources.\n"}],"comparison_anchors":[{"label":"Child kidnapped by stranger (childhood, US)","lifetime_us_adult":0.0000284},{"label":"Data breach exposure (annual, US adult)","lifetime_us_adult":0.33},{"label":"Credit card fraud (lifetime, US adult)","lifetime_us_adult":0.3}],"regional_breakdown":[{"region":"Explicit sexual content (pornography)","probability":0.54,"notes":"Common Sense Media 2023: 54% of teens report first exposure before age 13. Bark 2024 passive monitoring finds 60%+ of tweens encounter nudity or sexual content."},{"region":"Graphic violence","probability":0.7,"notes":"Youth Endowment Fund 2024: 70% of teens encountered real-life violent content on social media in the past year. Cumulative childhood exposure likely higher."},{"region":"Self-harm / pro-eating-disorder content","probability":0.37,"notes":"Bark 2024: 37% of tweens were involved in situations related to self-harm or suicide content. Bark 2025 reports tween engagement with disordered eating content rose 650% since 2021."}],"personal_factor_multipliers":[{"factor":"Has own smartphone before age 10","multiplier":1.5,"notes":"Earlier device ownership strongly correlates with earlier exposure. Common Sense Media found 15% of teens first saw pornography at age 10 or younger, concentrated among those with early personal device access."},{"factor":"Active parental monitoring/filtering","multiplier":0.7,"notes":"Parental controls reduce but do not eliminate exposure. Bark's own data shows 60%+ encounter rates even among families who install monitoring software. Filters miss content in messaging apps, peer-shared material, and social media algorithmic feeds."},{"factor":"No internet-connected device in bedroom","multiplier":0.5,"notes":"Removing unsupervised access — especially at night — is the single most effective structural intervention, though children still encounter content on school devices, friends' devices, and shared family devices."},{"factor":"LGBTQ+ youth","multiplier":1.3,"notes":"Common Sense Media found two-thirds of LGBTQ+ teen respondents consumed pornography intentionally, compared to lower rates among cisgender heterosexual peers. This may partly reflect information-seeking behavior in the absence of inclusive sex education."}],"short_label":"Kids & explicit content","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"The headline 54% figure from Common Sense Media is a retrospective self-report by teens aged 13-17 about when they first encountered pornography. Retrospective recall of age-at-first-exposure is unreliable in predictable ways: respondents may misremember their exact age, and the social desirability of the answer shifts depending on the respondent's current relationship with the material. The true figure could be somewhat higher (teens who minimize or forget early accidental encounters) or somewhat lower (teens who compress timelines).\n\"Exposure\" is doing enormous work in this entry. A 9-year-old who glimpses a pornographic pop-up ad for three seconds before closing the browser tab and a 12-year-old who regularly seeks out violent content are both counted as \"exposed.\" The prevalence literature almost never distinguishes fleeting accidental encounters from sustained deliberate consumption, and the harm profiles of these two experiences are radically different. The headline number may not measure the thing most parents are actually worried about.\nThe harm literature is less settled than the prevalence literature. The 2024 systematic review in L'Encéphale finds associations between pornography exposure and permissive sexual attitudes, sexual risk behavior, and gendered attitudes — but most evidence is correlational, effect sizes are modest, and the causal direction is contested. Children who seek out explicit content may differ from those who encounter it accidentally in ways that confound the association. The strongest evidence for harm relates to exposure to violent or degrading pornography specifically, not to nudity or sexual content in general.\nBark's data, while large-scale (7.9-11.1 billion activities analyzed), comes from a self-selected sample of families who chose to install monitoring software. This population may not be representative of all families with children.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A child's tablet lying face-down on a kitchen table beside a glass of orange juice, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-exposure-explicit-content","api_url":"https://likelier.app/api/fears/child-exposure-explicit-content.json"},{"slug":"low-magnesium-intake-no-supplement","question":"What are the odds your magnesium intake falls short of the recommended amount?","quick_answer":"About 54% of US adults — more than half — take in less magnesium than the estimated requirement from food and drink. Most people would guess only 15-20% fall short.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Magnesium sits in a blind spot. It is not a vitamin people think to \"get more of,\" it has no marquee deficiency disease that the public can name, and the assumption is that a reasonably varied diet handles it automatically. Supplement marketing has lately pushed magnesium for sleep and cramps, which nudges awareness up, but the default intuition is still that intake shortfalls are somebody else's problem -- the very old, the very ill -- rather than a majority experience. Almost no one would guess that more than half of US adults habitually eat less magnesium than the requirement assumes.\n","rough_estimate":"~15-20% of people fall short","kind":"intuition"},"native":{"display":"~52-57% of US adults get less magnesium than the EAR from food and beverages (57% of men, 52% of women; NHANES 2017-March 2020, supplements not counted)","numerator":54,"denominator":100,"unit":"share of US adults with habitual dietary magnesium intake from food and beverages below the Estimated Average Requirement (EAR)","population":"US adults aged 19+ (What We Eat in America, NHANES 2017-March 2020; food + beverage intake, excluding supplements)"},"normalized":{"lifetime_us_adult":0.54,"display":"~54% of US adults take in less magnesium than the requirement from food and beverages (before supplements)","log_value":-0.27,"assumptions":"What We Eat in America / NHANES 2017-March 2020 reports average dietary magnesium intake of 339 mg/day for men (ages 19+) and 269 mg/day for women, with 57% of men and 52% of women below the Estimated Average Requirement (EAR). Weighting the two roughly equal-sized groups gives ~54% of adults below the EAR — the figure used for both the native prevalence and the normalized value. Unlike an acute event, dietary-intake inadequacy is a stable population trait: eating patterns change slowly, so the cross-sectional ~54% is close to the share of adults whose habitual intake falls short over adulthood. This is an INTAKE measure (below the EAR), not a measured deficiency rate — see caveats. The EAR, not the higher RDA, is the threshold, so this is a conservative floor; measured against the RDA (400-420 mg men, 310-320 mg women) the shortfall population is larger still.\n","uncertainty":{"low":0.5,"high":0.6},"scope":"us_adult_lifetime"},"sources":[{"url":"https://lpi.oregonstate.edu/mic/minerals/magnesium","title":"Magnesium — Micronutrient Information Center","publisher":"Linus Pauling Institute, Oregon State University","source_type":"reputable_reference","statistic":"NHANES 2017-March 2020: average magnesium intake 339 mg/day (men 19+) and 269 mg/day (women); 57% of men and 52% of women had inadequate dietary intakes (below the EAR).","excerpt":"\"Data from a US national survey, National Health and Nutrition Examination Survey (NHANES) 2017-March 2020, show an average dietary magnesium intake of 339 mg/day for men (ages ≥19 years) and 269 mg/day for women. Concerningly, 57% of men and 52% of women had inadequate dietary intakes (i.e., less than the EAR) of magnesium.\"\n","source_date":"2025-01-01","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20260722023304/https://lpi.oregonstate.edu/mic/minerals/magnesium","calculation_notes":"The Linus Pauling Institute compiles the USDA What We Eat in America / NHANES 2017-March 2020 usual-intake tables. Below-EAR shares: men 57%, women 52%. Combining the two near-equal adult subpopulations yields ~54% of adults below the EAR, the headline. The LPI page also states plainly: \"National surveys indicate that over one-half of US adults have inadequate dietary magnesium intake.\" EAR (not RDA) is the comparison threshold, so the figure is a lower bound on inadequacy.\n"},{"url":"https://ods.od.nih.gov/factsheets/Magnesium-HealthProfessional/","title":"Magnesium — Health Professional Fact Sheet","publisher":"NIH Office of Dietary Supplements","source_type":"govt_report","statistic":"Many people in the US consume less than recommended amounts of magnesium; there are no current national data on magnesium STATUS; symptomatic deficiency is uncommon in healthy people.","excerpt":"\"Many people in the United States consume less than the recommended amounts of magnesium, with men age 71 years and older and adolescent males and females having the highest risk of low magnesium intakes. [...] There is no current data on magnesium status among people who live in the United States, as magnesium status is not routinely evaluated in national surveys, hospitals, or clinics. [...] While certain health conditions and the use of certain medications can cause symptomatic magnesium deficiency, it is uncommon in healthy people.\"\n","source_date":"2026-01-06","source_accessed":"2026-07-18","archive_url":"http://web.archive.org/web/20260724131450/https://ods.od.nih.gov/factsheets/Magnesium-HealthProfessional/","calculation_notes":"The NIH ODS fact sheet supplies the two crucial honesty anchors for this entry: (1) it confirms the intake shortfall (\"many consume less than recommended,\" with older men and adolescents at highest risk), and (2) it draws the sharp line between INTAKE and STATUS — magnesium status is not measured in US national surveys, so there is NO defensible national deficiency-prevalence figure, and symptomatic deficiency is uncommon in otherwise healthy people. This is why the entry is framed as an intake-below-requirement probability, not a deficiency rate.\n","independence_note":"NIH ODS and the Linus Pauling Institute are independently authored references; both draw on the same USDA/NHANES intake data but neither derives from the other."}],"comparison_anchors":[{"label":"High blood pressure (lifetime, US adult)","lifetime_us_adult":0.84},{"label":"Vitamin D deficiency without supplements (lifetime)","lifetime_us_adult":0.35},{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33}],"short_label":"Magnesium gap","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The single most important caveat is that this is an INTAKE measure, not a deficiency rate. \"Below the EAR\" means habitual dietary intake falls under the amount estimated to meet the needs of half the population — it does NOT mean 54% of adults are clinically magnesium deficient. The two are routinely conflated by supplement marketing, and the gap between them is large: the kidneys tightly conserve magnesium, less than 1% of body magnesium is in serum, and — per NIH ODS — symptomatic deficiency is uncommon in healthy people. Crucially, there is no national US survey of magnesium STATUS at all, because no single blood or tissue test reliably captures it; serum magnesium, the usual proxy, does not reflect total body stores. So the honest claim is narrow and well-supported (most adults eat less than the requirement assumes) while the tempting claim is neither measured nor supportable (that most adults are deficient). Whether a chronic sub-EAR intake matters clinically is an open question: low magnesium intake and status are associated in observational studies with hypertension, type 2 diabetes, and cardiovascular disease, but supplementation trials show only marginal blood-pressure effects, and association is not causation. Note also that the ~54% figure is intake from FOOD AND BEVERAGES only: NIH ODS notes that people who take a magnesium-containing supplement have higher total intakes, and a supplement or magnesium-rich diet closes the gap — so this is the share falling short before supplements, not a residual deficit that supplement users still carry.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-07-18","methodology_version":"1.0"},"reviewer":"8d-eval-2026-07-18","last_reviewed":"2026-07-18","reviewed":true,"generated_at":"2026-07-18","image":{"alt":"A small pile of pumpkin seeds and leafy greens on a plate that is more than half empty, muted green and grey flat vector illustration."},"canonical_url":"https://likelier.app/low-magnesium-intake-no-supplement","api_url":"https://likelier.app/api/fears/low-magnesium-intake-no-supplement.json"},{"slug":"adventure-sports-injury","question":"What are the odds of a serious injury from regular participation in surfing, mountain biking, or rock climbing?","quick_answer":"About 1 in 2 over a ~30-year career of regular surfing, mountain biking, or climbing, but that works out to only about 0.3 medically-attended injuries per 100 sessions. Non-participants rate these as uniformly extreme; the per-day rate is modest, and the typical injury is a fracture, sprain, or cut, not the dramatic scenarios people picture.\n","category":"health","tags":["sport"],"no_reliable_estimate":false,"perceived":{"description":"Adventure sports sit awkwardly in the public imagination. Non-participants tend to bundle surfing, mountain biking, and rock climbing with BASE jumping, big-wave tow-in, and wingsuit flying, and rate the whole category as roughly \"extreme\". Participants know the inside story is milder — a surfboard fin cut, a shoulder fracture over the bars, a sprained ankle off a bouldering mat — but they also know someone who has been helicoptered off something. There is no rigorous standalone survey of how the public rates injury risk in recreational surfing versus mountain biking versus climbing, so perceived risk is marked as editorial intuition. The interesting property is that the folk intuition is roughly directionally correct for the category average but systematically wrong about which specific sport and which specific mechanism produce the injuries. Shoulder fractures from forward falls dominate mountain biking; lacerations from one's own board dominate surfing; ankle fractures from bouldering falls dominate indoor climbing. Almost none of those are what a non-participant pictures when they hear \"extreme sport\".\n","rough_estimate":"non-participants rate these sports as uniformly dangerous; participants rate them as medium-risk accumulation sports","kind":"intuition"},"native":{"display":"~0.3 medically-attended injuries per 100 participation days (mixed recreational hobbyist)","numerator":3,"denominator":1000,"unit":"per participation day (surfing, mountain biking, or climbing session)","population":"regular recreational adventure-sport hobbyist, US / comparable Western settings"},"normalized":{"lifetime_us_adult":0.55,"display":"~1 in 2 lifetime (regular hobbyist, ~30 years across mixed surfing / MTB / climbing)","log_value":-0.26,"assumptions":"Scope is activity_specific_lifetime — this is the probability for a regular recreational hobbyist who spends roughly 30 participation-days per year across some mix of surfing, mountain biking, and rock climbing over a 30-year active career (~900 lifetime participation days). It is not a US-adult population figure; most US adults never do any of these sports regularly.\nPer-sport anchors from the epidemiology: (1) Surfing — Nathanson et al. (Am J Sports Med 2007) found 6.6 significant injuries per 1,000 hours of competitive surfing. Recreational rates are lower by roughly a factor of 2-3; the 2002 Nathanson survey (Am J Emerg Med) of 1,237 acute injuries puts the recreational significant-injury rate at roughly 3 per 1,000 surfing hours, or ~0.006 per 2-hour session. (2) Mountain biking — Nelson and McKenzie (Am J Sports Med 2011) estimated ~217,000 US ED visits from 1994-2007, averaging 15,500 per year across roughly 8-9 million US riders, implying an annual ED-visit rate of ~0.18% per rider. Whole-population rates are ~0.6-1 injury per 1,000 hours for cross-country and ~1.1 per 1,000 hours for downhill (Course et al. 2023), with severe-trauma share roughly 10% in the downhill subgroup. (3) Rock climbing — Schöffl et al. (Wilderness Environ Med 2013) recorded 30 acute injuries across 515,337 indoor climbing wall visits, an acute injury rate of 0.02 per 1,000 climbing hours. Outdoor trad climbing is orders of magnitude higher at ~37.5 injuries per 1,000 hours per the International Alpine Trauma Registry synthesis, and sport climbing is intermediate at ~0.2 per 1,000 hours.\nWeighting a mixed hobbyist's exposure across those sports and restricting to medically-attended injuries (ED visit or equivalent), a working per-day rate is ~3 per 1,000 participation days — heavily driven by MTB and surfing with climbing as a smaller denominator contribution. Over 900 lifetime participation days, the probability of at least one such injury is 1 − (1 − 0.003)^900 ≈ 0.93 if we treat days as independent Poisson trials, but that is an overestimate because injury risk is correlated (a hobbyist who is injured once is often one who takes elevated risk, and many \"injuries\" in the literature numerator are minor and self-treated rather than medically attended). Adjusting for that correlation and for the fact that the true \"ED-visit or hospitalization\" threshold is stricter than the Nathanson and Schöffl numerators, a more defensible central estimate lands near 0.55 with a wide uncertainty band. Hospitalization-level (admission, not just ED triage) lifetime probability sits materially lower, probably 0.10-0.20 for the same career profile.\n","uncertainty":{"low":0.25,"high":0.8},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/17021312/","title":"Competitive Surfing Injuries: A Prospective Study of Surfing-Related Injuries Among Contest Surfers","publisher":"American Journal of Sports Medicine (Nathanson, Bird, Dao, Tam-Sing 2007)","source_type":"peer_reviewed","statistic":"13 injuries per 1,000 hours of competitive surfing; 6.6 significant injuries per 1,000 hours; risk more than doubled in overhead waves and over hard bottoms","excerpt":"\"The overall injury rate was 13 injuries per 1000 hours of competitive surfing. The significant injury rate was 6.6 per 1000 hours. The risk of injury was more than double in overhead-sized waves and over a hard seafloor.\"\n","source_date":"2007-01-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031033/https://pubmed.ncbi.nlm.nih.gov/17021312/","calculation_notes":"Nathanson et al. 2007 is the seminal prospective surfing-injury study, covering 32 professional and amateur contests worldwide from 1999-2005. It is the source of the 6.6-significant-injuries-per-1000-hours figure used as the competitive- surfing anchor. For recreational surfing we halve this rate to ~3 per 1,000 hours based on the lower intensity and smaller-wave exposure profile typical of a hobbyist; the 2007 paper explicitly notes that wave size doubles risk and hard bottoms triple it, both of which are disproportionately present in competition. At ~2 hours per surf session, the recreational rate implies ~0.006 significant injuries per session, contributing roughly 6 per 1,000 surf-session days to the mixed-hobbyist denominator.\n","independence_note":"Single prospective cohort of contest surfers. The 2002 Nathanson recreational survey below uses a different methodology (retrospective self-report) and a different population (recreational rather than competitive); the two are complementary anchors rather than a single dataset.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11992332/","title":"Surfing injuries","publisher":"American Journal of Emergency Medicine (Nathanson, Haynes, Galanis 2002)","source_type":"peer_reviewed","statistic":"1,237 acute injuries reported across 1,348 recreational surfer respondents; lacerations 42%, contusions 13%, sprains/strains 12%, fractures 8%; 37% lower extremity, 37% head/neck","excerpt":"\"Older surfers, more expert surfers, and those surfing large waves have a higher relative risk for significant injury.\"\n","source_date":"2002-05-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031105/https://pubmed.ncbi.nlm.nih.gov/11992332/","calculation_notes":"Nathanson 2002 is the recreational counterpart to the 2007 competitive paper, surveying 1,348 surfers on acute and chronic injuries. It anchors the recreational- surfing severity distribution: lacerations from contact with one's own board (~55% of injury mechanism) dominate, with a smaller but consequential fracture share (~8%). The paper also identifies that most surfing injuries are minor and do not present to emergency departments — the 2013-era NEISS-based follow-up by Klick et al. estimated only about 1,300-1,800 ED visits per year for surfing injuries in the US, a much smaller numerator than the injury reports in the cohort survey and consistent with an ED-level threshold of roughly 1 per 5,000 surfing hours.\n","independence_note":"Retrospective self-report from recreational surfers in Hawaii, California, and Rhode Island. Independent of the 2007 competitive dataset and of the NEISS ED- visit aggregators.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21076012/","title":"Mountain biking-related injuries treated in emergency departments in the United States, 1994-2007","publisher":"American Journal of Sports Medicine (Nelson & McKenzie 2011)","source_type":"peer_reviewed","statistic":"~217,433 US ED visits for mountain biking injuries 1994-2007 (~15,531/year); upper-extremity fractures 10.6%; shoulder fractures 8.3%; hospitalization 4.5-6.1%; TBI 8.4% in ages 14-19","excerpt":"\"Nationwide, an estimated 217,433 patients were treated for mountain bike-related injuries in US emergency departments from 1994 to 2007, an average of 15,531 injuries per year ... Mountain bike-related injuries decreased from 1994 to 2007. Upper extremity fractures were the most common injury.\"\n","source_date":"2011-02-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250915150208/https://pubmed.ncbi.nlm.nih.gov/21076012/","calculation_notes":"Nelson and McKenzie 2011 draws from the CPSC NEISS sample and is the anchor for the US mountain biking ED-visit rate. Roughly 15,500 ED visits per year against a US mountain biker population of 8-9 million implies an annual ED-visit probability of ~0.17-0.18% per rider. The hospitalization fraction of 4.5-6.1% implies a narrower annual hospitalization probability on the order of 0.01% per rider. Over a 30-year regular-riding career (weighted so MTB represents roughly a third of total adventure-sport exposure), MTB contributes about 5-7 percentage points to the lifetime medically-attended injury probability in the headline. The upper-extremity fracture dominance is a useful corrective to the cultural framing of MTB as primarily a head-injury sport: TBI is a meaningful outcome but is an order of magnitude less common than broken collarbones and wrists.\n","independence_note":"Drawn from CPSC NEISS, the US surveillance system that also feeds the bicycle- injury numerator on the sibling cycling-helmetless-head-injury page. Treat as the US-specific MTB anchor; not independent of NEISS-derived aggregates in other US injury entries on this site.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23877045/","title":"Acute Injury Risk and Severity in Indoor Climbing — A Prospective Analysis of 515,337 Indoor Climbing Wall Visits in 5 Years","publisher":"Wilderness & Environmental Medicine (Schöffl, Hoffmann, Küpper 2013)","source_type":"peer_reviewed","statistic":"30 acute injuries across 515,337 climbing wall visits; acute injury rate 0.02 per 1,000 climbing hours; 1 injury per ~47,000 hours of climbing; 15 UIAA MedCom grade 2, 13 grade 3, 2 grade 4","excerpt":"\"The acute injury risk in males was 1 injury per 47,742 hours of sports and in females 1 injury per 46,735 hours. The acute injury rate per 1000 hours of sports performance was overall 0.02/1000 h.\"\n","source_date":"2013-07-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031140/https://pubmed.ncbi.nlm.nih.gov/23877045/","calculation_notes":"Schöffl et al. 2013 is the largest prospective indoor-climbing injury cohort in the literature and the basis for the \"indoor climbing is one of the safer gym sports\" finding. At 0.02 acute injuries per 1,000 climbing hours, a regular indoor climber putting in 200 hours per year accumulates ~0.004 acute injuries per year — one injury every 250 years of climbing on average. Outdoor trad climbing is orders of magnitude higher: the International Alpine Trauma Registry synthesis (Rauch et al. 2019) cites 37.5 injuries per 1,000 hours for traditional climbing, 4.07 for ice climbing, 3.1 for competition climbing, 0.2 for sport climbing, and 0.027 for indoor climbing. The three-order-of-magnitude gap between indoor climbing and traditional outdoor climbing is the largest within-sport heterogeneity on this page and drives the wide uncertainty band on the lifetime estimate — a hobbyist whose \"climbing\" is Tuesday-night gym sessions is running a radically different risk from one who is on alpine trad routes every other weekend.\n","independence_note":"Single-facility prospective cohort (Munich area) but widely cross-cited in the climbing-injury literature. Independent of the Buzzacott/Schöffl US NEISS follow-up and of the IATR synthesis, which use different populations and surveillance methods.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6981967/","title":"Climbing Accidents — Prospective Data Analysis from the International Alpine Trauma Registry and Systematic Review of the Literature","publisher":"International Journal of Environmental Research and Public Health (Rauch, Strapazzon, Falla, Brodmann Maeder, Procter, Brugger 2019)","source_type":"peer_reviewed","statistic":"Injury rates per 1,000 climbing hours: traditional 37.5, ice 4.07, competition 3.1, sport 0.2, indoor 0.027; mean ISS 29.6 in multisystem trauma cases","excerpt":"\"Climbing accidents mainly affect young men and mostly lead to minor injuries.\" \"In severe multisystem trauma, injuries to head/neck, chest and abdomen predominate.\"\n","source_date":"2019-12-27","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031213/https://pmc.ncbi.nlm.nih.gov/articles/PMC6981967/","calculation_notes":"Rauch et al. 2019 synthesises the International Alpine Trauma Registry with a systematic review of published climbing-injury studies. It is the source of the five-discipline injury-rate comparison that anchors the regional_breakdown rows for trad versus sport versus indoor climbing. The mean Injury Severity Score of 29.6 among multisystem trauma cases places trad climbing's severe-injury regime in the same rough neighbourhood as motor vehicle trauma, which is why trad climbing is treated as a genuinely different activity on this page and why bounding the headline number requires care about what \"climbing\" means for a particular reader.\n","independence_note":"Multi-centre European registry plus literature synthesis. Partially overlaps with Schöffl et al. 2013 indoor-climbing data as one input among many, but the traditional-climbing 37.5/1,000-hours figure comes from an independent Yosemite National Park study.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10871082/","title":"Health Service Impacts and Risk Factors for Severe Trauma in Mountain Biking: A Narrative Review","publisher":"Cureus / Course, Sharman, Tran 2023","source_type":"peer_reviewed","statistic":"Downhill mountain biking: 1.1 injuries per 1,000 riding hours with ~10% severe; elite enduro: 38.3 injuries per 1,000 race hours vs 3.6 per 1,000 practice hours","excerpt":"\"Downhill MTB was consistently identified as posing the highest risk for injuries, with the mechanism of injury most commonly involving forward falls over handlebars during descents.\"\n","source_date":"2023-12-18","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031246/https://pmc.ncbi.nlm.nih.gov/articles/PMC10871082/","calculation_notes":"Course et al. 2023 is the most recent narrative review of mountain biking severe trauma and the source of the downhill-versus-cross-country split. The 10-fold gap between elite enduro practice (3.6/1,000 hours) and race exposure (38.3/1,000 hours) is a useful illustration that \"mountain biking\" at the top end of the intensity distribution has a rate consistent with outdoor trad climbing rather than with recreational MTB. The headline number on this page is for recreational riders; competitive downhill is covered qualitatively in the regional_breakdown row below.\n","independence_note":"Narrative review rather than a single primary cohort, synthesising European and Canadian MTB studies. Independent of Nelson and McKenzie 2011 (which is US NEISS data) but overlaps with the broader MTB injury literature the Nelson paper also draws on.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10194912/","title":"The epidemiology, risk factors and impact of exposure on unintentional surfer and bodyboarder deaths","publisher":"PLoS One (Lawes, Koon, Berg, van de Schoot, Peden 2023)","source_type":"peer_reviewed","statistic":"155 Australian surfing/bodyboarding deaths over 16 years; exposure-adjusted fatality rate 0.06 per 1 million surfing hours; drowning 58.1%, cardiac conditions 32.9%","excerpt":"\"An average of 9.7 (SD = 3.37) surfing and bodyboarding-related fatalities occurred per year ... the cumulative resident-based mortality rate (resident n = 143) was 0.04 per 100,000 residents.\"\n","source_date":"2023-05-18","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420031332/https://pmc.ncbi.nlm.nih.gov/articles/PMC10194912/","calculation_notes":"Lawes et al. 2023 is the largest exposure-adjusted analysis of surfing fatalities in the literature, based on the Australian Royal National Lifesaving Society registry. The per-million-hours fatality rate of 0.06 is the headline for the surfing fatality row in the regional_breakdown: a recreational surfer putting in 86 hours per year accumulates a per-year fatality probability of roughly 5 per million, and over a 30-year career ~1.5 per 10,000. That is comparable to the lifetime odds of lightning strike, not to the odds of injury on this page. The cardiac-condition share (32.9%) is worth noting because it means a non-trivial fraction of \"surfing deaths\" are cardiac events that happened to occur in the water, not trauma outcomes attributable to surfing biomechanics.\n","independence_note":"Australian national registry data, independent of the Nathanson US/Pacific surfing cohort work and of NEISS. Used here specifically as the fatality anchor because no comparable US-wide exposure-adjusted surfing fatality dataset is publicly available.\n"}],"comparison_anchors":[{"label":"Serious skiing injury (lifetime, 600-day active career)","lifetime_us_adult":0.7},{"label":"Serious skiing injury (per 20-day season)","lifetime_us_adult":0.0392},{"label":"Serious head injury, frequent unhelmeted urban cyclist (lifetime)","lifetime_us_adult":0.125},{"label":"Death on a motorcycle (lifetime, active US rider)","lifetime_us_adult":0.02},{"label":"Death in a car crash (lifetime, US adult)","lifetime_us_adult":0.0108}],"regional_breakdown":[{"region":"Recreational surfer (1-2 sessions/week, small-to-moderate beach break)","probability":0.15,"notes":"Mostly lacerations from one's own board and minor head/neck contusions. Nathanson 2002 survey found fractures at ~8% of the recreational injury mix, and the majority of incidents are self-treated or same-day ED discharges. Lifetime probability of at least one ED-level injury over a 30-year career at this exposure lands around 1 in 7."},{"region":"Recreational cross-country mountain biker (weekend rides, trail network)","probability":0.25,"notes":"Nelson-McKenzie implies ~0.18% annual ED-visit probability per US rider, which compounds to roughly 1 in 20 over 30 years at casual exposure but rises closer to 1 in 4 for someone riding 50+ days a year on technical singletrack. Upper-extremity fractures dominate; traumatic brain injury is a meaningful minority outcome (~5-10% of injuries)."},{"region":"Indoor-gym-only rock climber","probability":0.02,"notes":"Schöffl 2013 found acute injury rate of 0.02 per 1,000 climbing hours. A regular gym climber (~200 hours/year for 30 years = 6,000 hours) accumulates ~0.12 expected acute injuries over a career. Ankle fractures from bouldering falls are the disproportionate mechanism; finger pulley injuries are common but usually not ED-attended. Lifetime ED-visit probability is low-single-digit percent."},{"region":"Outdoor sport climber (bolted routes, lead climbing)","probability":0.15,"notes":"Rauch 2019 cites sport climbing at ~0.2 injuries per 1,000 hours — 10x the indoor rate. Ankle injuries from ground falls at the first bolt and shoulder injuries from lead falls dominate. Lifetime ED-visit probability for a regular sport climber (~200 hours/year) is on the order of 1 in 7."},{"region":"Outdoor traditional / alpine climber","probability":0.75,"notes":"Rauch 2019 cites traditional climbing at 37.5 injuries per 1,000 hours — nearly four orders of magnitude above indoor. This is a genuinely different activity and the mean Injury Severity Score in multisystem trauma cases (29.6) is in the motor-vehicle-trauma regime. The headline number on this page does not apply to regular trad climbers."},{"region":"Downhill / freeride mountain biker (bike park, technical terrain)","probability":0.85,"notes":"Course 2023 cites downhill at 1.1 injuries per 1,000 riding hours with ~10% severe; elite enduro competition at 38.3/1,000 race hours. A regular downhill rider riding ~150 hours/year over 20 years will almost certainly accumulate multiple medically-attended injuries. Whistler Bike Park's single-season report of 898 riders with 1,759 injuries is the relevant order of magnitude."},{"region":"Combined mixed hobbyist (surfing + MTB + climbing, 30 days/year)","probability":0.55,"notes":"Headline figure on this page. A regular recreational hobbyist who splits ~30 days a year across a mix of these sports over a 30-year career (~900 lifetime participation days) has roughly even-money odds of at least one medically-attended injury. The hospitalization-level probability is materially lower, probably 0.10-0.20."},{"region":"Surfing fatality (regular recreational surfer, 30-year career)","probability":0.00015,"notes":"Lawes 2023 per-million-hours rate of 0.06 applied to 86 hours/year × 30 years = 2,580 surfing hours, giving a career fatality probability of ~1.5 per 10,000. Cardiac events are roughly a third of this total, trauma and drowning the remainder. Comparable to the lifetime odds of dying in a bicycle crash for the general US adult."}],"personal_factor_multipliers":[{"factor":"indoor gym climbing only","multiplier":0.05,"notes":"Three-order-of-magnitude reduction relative to outdoor trad climbing (Schöffl 2013 vs Rauch 2019). The largest single risk-reduction lever available to a climber."},{"factor":"technical / downhill mountain biking","multiplier":4,"notes":"Per-hour injury rate on downhill-specific terrain is roughly 2x cross-country plus a severe-injury fraction that is an additional factor on severity. Helmet and body armour use is near-universal in this subgroup but does not eliminate the rate difference."},{"factor":"surfing large waves (>6ft) or hard-bottom breaks","multiplier":2,"notes":"Nathanson 2007 found risk more than doubled in overhead-sized waves and over hard seafloors. Big-wave surfing (20ft+) is a separate regime and not covered by the headline."},{"factor":"traditional or alpine outdoor climbing","multiplier":10,"notes":"Rauch 2019 trad-climbing rate of 37.5/1,000 hours is roughly 10x a mixed recreational hobbyist's weighted rate. Multi-pitch exposure, placed gear, and rope-scenarios drive the gap. A single order of magnitude is conservative — at the extreme end (big-wall, alpine) the multiplier is higher."},{"factor":"20+ years active participation","multiplier":1.5,"notes":"Cumulative exposure drives lifetime risk roughly linearly, but older participants also sit lower on the intensity distribution, which partly offsets the exposure effect. A ~1.5x multiplier captures the net after age-related moderation."},{"factor":"no prior formal instruction","multiplier":1.5,"notes":"Self-taught hobbyists have higher per-hour injury rates in the surfing and climbing literature, concentrated in predictable mechanisms (climbing: unclipping at the anchor; surfing: turtle-roll in shorebreak). MTB shows a smaller instruction effect."}],"short_label":"Adventure sports","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry covers surfing, mountain biking, and rock climbing as a combined category because the underlying fear — \"am I going to get hurt doing weekend adventure sport?\" — is typically undifferentiated at the point readers encounter it. The per-sport rates differ by three or more orders of magnitude, and the headline figure is a weighted average that applies only to a hobbyist who genuinely splits time across multiple sports at moderate intensity. A reader whose adventure sport is entirely indoor climbing should substitute the ~2% row; a reader whose adventure sport is downhill mountain biking at a bike park should substitute the ~85% row; neither reader should use the 55% headline.\nSkiing and snowboarding are covered separately at [skiing-serious-injury](/skiing-serious-injury). The skiing entry's headline of roughly 0.70 lifetime probability for a 600-day active career is higher than this page's 0.55 not because skiing is more dangerous per exposure but because the skiing headline assumes 600 active days versus this page's 900 mixed-sport days with a lower weighted per-day rate. The two entries are not directly comparable and should not be subtracted from each other.\nUrban and commuter cycling are out of scope. Road-cycling head-injury epidemiology is dominated by motor-vehicle collisions and is covered at [cycling-helmetless-head-injury](/cycling-helmetless-head-injury). Mountain biking on this page refers to off-road trail riding only.\nHigh-fatality-rate variants are excluded from the headline. Big-wave surfing (20ft+), alpine and ice climbing above 4,000m, BASE jumping, wingsuit flying, and tow-in surfing at breaks like Nazaré or Mavericks operate in a different regime where fatality rather than injury is the modal serious outcome, and the epidemiological literature on those activities is thin enough that honest quantification is difficult. Whitewater kayaking, paragliding, and extreme skiing sit between the headline adventure-sports category and the big-fatality category; readers with exposure to those sports should treat the ~85% downhill MTB row as a lower bound on their lifetime injury probability rather than the headline 55%.\nFinally, \"medically-attended injury\" is the outcome the headline is calibrated to. Hospitalization-level injury (admission rather than ED triage) is roughly 5-15% of that, so the lifetime probability of being admitted to hospital with an injury from these sports over a 30-year career is probably 0.10-0.20 rather than 0.55. Lifetime fatality probability for the mixed hobbyist is on the order of 1 in 1,000 to 1 in 3,000 across this combined category, dominated by surfing drownings, MTB head trauma, and lead-climbing ground falls.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":3,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-9-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single climbing carabiner resting on a pale neutral surface, flat vector illustration in muted greys with a single cool-blue accent."},"canonical_url":"https://likelier.app/adventure-sports-injury","api_url":"https://likelier.app/api/fears/adventure-sports-injury.json"},{"slug":"deportation-us-undocumented","question":"What are the odds of being deported if undocumented in the US?","quick_answer":"On the population average the odds are about 1 in 50 per year, compounding to roughly 55% over a notional 40-year residency. This is a crude average: individual risk varies by orders of magnitude with criminal history, location, and the administration in power.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Immigration enforcement is among the most politically charged topics in the US, and the fear of deportation dominates the lived experience of undocumented residents. Surveys of undocumented immigrants consistently find that a majority report significant anxiety about removal, with the Pew Research Center noting that roughly two-thirds of unauthorized immigrants who have lived in the US for a decade or more say they worry \"a lot\" or \"some\" about deportation. Media coverage of ICE raids amplifies the perception that enforcement is pervasive, even though the annual removal rate relative to the total undocumented population has historically been low in percentage terms.\n","rough_estimate":"~1 in 3 to 1 in 5 over a lifetime, intuitively","kind":"intuition"},"native":{"display":"~1 in 50 per year (central estimate, ~250,000 removals / ~13 million undocumented)","numerator":250000,"denominator":13000000,"unit":"per year","population":"Undocumented immigrants residing in the US (~11-14 million)"},"normalized":{"lifetime_us_adult":0.55,"display":"~55% over a notional 40-year residency horizon","log_value":-0.26,"assumptions":"ICE ERO reports 142,580 formal ERO removals in FY2023. DHS OHSS repatriation tables report total \"Removals\" (all DHS components) of 177,540 in FY2023 and 329,990 in FY2024. Across recent years these totals have ranged from roughly 85,000 (FY2021) to over 400,000 (FY2014 peak). The undocumented population stood at roughly 11 million (DHS 2022 estimate) to 14 million (Pew 2023 revised estimate). Using a central annual removal rate of ~250,000 — sitting between the FY2023 and FY2024 DHS-reported removal totals — against a midpoint population of ~13 million gives an annual per-person hazard of ~1.9%, or roughly 1 in 50. Compounded over a notional 40-year adult residency horizon: 1 - (1 - 0.019)^40 ≈ 0.55, or ~55%. This is a crude population average; individual risk varies by orders of magnitude depending on criminal history, geographic location, and the political administration in power.\n","uncertainty":{"low":0.2,"high":0.85},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.ice.gov/news/releases/ice-releases-fiscal-year-2023-annual-report","title":"ICE Releases Fiscal Year 2023 Annual Report","publisher":"U.S. Immigration and Customs Enforcement","source_type":"govt_report","statistic":"142,580 formal removals in FY2023; over 1 million total removals and expulsions including Title 42","excerpt":"\"ERO conducted 142,580 removals and 62,545 Title 42 expulsions to more than 170 countries worldwide in Fiscal Year 2023.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260405003525/https://www.ice.gov/news/releases/ice-releases-fiscal-year-2023-annual-report","calculation_notes":"ICE ERO's FY2023 formal removals (142,580) represent the floor of annual deportation activity. Adding Title 42 expulsions and CBP removals raises the total substantially. For the native figure, I use a central estimate of ~250,000 annual removals (blending FY2023 and FY2024 data, excluding Title 42 expulsions which ended May 2023) against a midpoint undocumented population of ~13 million. Annual hazard: 250,000 / 13,000,000 ≈ 0.019. Over 40 years: 1 - (1 - 0.019)^40 ≈ 0.55.\n"},{"url":"https://www.pewresearch.org/race-and-ethnicity/2025/08/21/u-s-unauthorized-immigrant-population-reached-a-record-14-million-in-2023/","title":"U.S. Unauthorized Immigrant Population Reached a Record 14 Million in 2023","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"Estimated 14 million unauthorized immigrants in the US as of mid-2023, up from 10.5 million in 2021","excerpt":"\"Between 2021 and 2023, the number of unauthorized immigrants living in the United States grew from an estimated 10.5 million to 14 million.\"\n","source_date":"2025-08-21","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420034621/https://www.pewresearch.org/race-and-ethnicity/2025/08/21/u-s-unauthorized-immigrant-population-reached-a-record-14-million-in-2023/","calculation_notes":"Pew's revised estimate of 14 million (mid-2023) is the highest credible estimate of the denominator. DHS's own estimate for January 2022 was 10.99 million. The midpoint of ~13 million is used for the native rate calculation. A larger denominator reduces the per-person annual hazard; a smaller one increases it — hence the wide uncertainty band.\n"},{"url":"https://ohss.dhs.gov/topics/immigration/immigration-enforcement/monthly-tables","title":"Immigration Enforcement and Legal Processes Monthly Tables","publisher":"DHS Office of Homeland Security Statistics","source_type":"govt_report","statistic":"Repatriations by fiscal year — Removals: FY2023 177,540; FY2024 329,990; FY2025 (partial, through Nov 2024) 61,630. Total repatriations (removals + returns + Title 42 expulsions): FY2023 1,199,910; FY2024 777,580","excerpt":"\"Yearly Repatriations by Fiscal Year — Removals: 2023 177,540; 2024 329,990; 2025 (so far) 61,630. Total: 2023 1,199,910; 2024 777,580; 2025 (so far) 111,010. Data available through November 2024.\"\n","source_date":"2025-01-16","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260405012437/https://ohss.dhs.gov/topics/immigration/immigration-enforcement/monthly-tables","calculation_notes":"The DHS OHSS monthly tables report \"Removals\" (administrative removals carrying penalties) separately from returns and expulsions. Removals rose from 177,540 in FY2023 to 329,990 in FY2024, then reset to 61,630 for the partial FY2025 window (data through November 2024). This near-doubling from FY2023 to FY2024, and the swing across recent years (FY2021 removals were 85,100), is what drives the wide uncertainty band (20%-85% lifetime). The native central estimate of ~250,000 removals/year sits between the FY2023 (177,540) and FY2024 (329,990) DHS-reported removal totals; under aggressive enforcement the annual hazard could roughly double.\n"},{"url":"https://www.migrationpolicy.org/news/us-third-country-deportation-agreements","title":"U.S. Third-Country Deportation Agreements Are More About Fear than Numbers","publisher":"Migration Policy Institute","source_type":"reputable_reference","statistic":"~15,000 third-country deportations (13,000 to Mexico) between Jan 20 and Dec 31, 2025, against MPI's estimate of ~396,000 ICE deportations in the first year","excerpt":"\"The Migration Policy Institute (MPI) estimates about 15,000 third-country deportations (13,000 of them to Mexico) occurred between January 20, 2025, and December 31, 2025—a tiny share of the 1 million annual deportations the administration has said it intends to achieve. ... MPI estimates that U.S. Immigration and Customs Enforcement (ICE) conducted 396,000 deportations in the administration's first year back.\"\n","source_date":"2026-03-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260619053711/https://www.migrationpolicy.org/news/us-third-country-deportation-agreements","calculation_notes":"Destination-conditional context, not a change to the headline hazard. Of MPI's ~15,000 third-country deportations, ~13,000 went to Mexico (a contiguous hybrid-agreement destination), leaving ~2,000 (by inference, 15,000 − 13,000) sent to countries where the deportee has no ties (El Salvador, South Sudan, Eswatini). Against MPI's ~396,000 first-year ICE deportations: third-country share ≈ 15,000/396,000 ≈ 4%; no-ties share ≈ 2,000/396,000 ≈ 0.5%. This conditions on having already been removed and so does not modify normalized.lifetime_us_adult.\n"}],"comparison_anchors":[{"label":"Being audited by the IRS (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Experiencing bankruptcy (lifetime, US adult)","lifetime_us_adult":0.08},{"label":"Being a victim of violent crime (lifetime, US adult)","lifetime_us_adult":0.25}],"personal_factor_multipliers":[{"factor":"Prior criminal conviction","multiplier":5,"notes":"ICE prioritizes individuals with criminal records; removal rates for this subgroup are far higher"},{"factor":"Long-term resident, no criminal record, interior US","multiplier":0.15,"notes":"Interior enforcement historically targets a small fraction of the total undocumented population; many long-term residents have near-zero removal probability in practice"},{"factor":"Recent border crosser","multiplier":3,"notes":"Individuals encountered at or near the border face much higher immediate removal rates"},{"factor":"Enforcement-heavy administration (e.g. early Obama, Trump)","multiplier":1.8,"notes":"Annual removals can roughly double depending on executive enforcement priorities"}],"short_label":"Deportation (undocumented)","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"autonomy_loss","valence":"negative","caveats":"This entry attempts to average across administrations, criminal history profiles, geographic locations, and duration of residence — each of which can shift individual risk by an order of magnitude or more. The \"40-year residency horizon\" is a modeling convenience; many undocumented immigrants do not remain for four decades, and many who do have effectively zero removal risk because interior enforcement has historically concentrated on individuals with criminal records. The population denominator is itself uncertain by ±3 million. Under maximalist enforcement policies, the annual removal rate could approach 3-4% of the undocumented population; under minimalist interior enforcement, it drops below 1%. The uncertainty band reflects this political volatility more than statistical noise. A separate question is destination given removal: the Migration Policy Institute estimates about 15,000 third-country deportations (13,000 of them to Mexico) between January 20 and December 31, 2025, against its estimate of roughly 396,000 ICE deportations in that first year — so removal to a third country runs on the order of 4% of ICE deportations, and removal to a country with no ties (the El Salvador CECOT, South Sudan, or Eswatini cases, ~2,000 by inference) on the order of 0.5%. MPI characterises these agreements as \"more about fear than numbers.\" This conditional destination share does not change the headline removal probability, which is silent on where one is sent.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A suitcase sitting alone on a porch, flat vector illustration, muted tones, no people."},"canonical_url":"https://likelier.app/deportation-us-undocumented","api_url":"https://likelier.app/api/fears/deportation-us-undocumented.json"},{"slug":"airline-trip-cancellation","question":"What are the odds of having a flight canceled on any given trip?","quick_answer":"About 1.4% of flights are canceled, so any single trip is very unlikely to be hit — though over a lifetime of flying the odds of at least one cancellation reach roughly 56%. After the 2022 meltdowns most travelers guess 5 to 20% per trip, well above today's reliability.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"After the chaos of 2022 — when some major US carriers canceled 3-4% of scheduled flights and high-profile holiday meltdowns dominated news coverage — many travelers calibrated their cancellation anxiety upward. The perception that flight cancellations are a frequent and unpredictable hazard has persisted even as airlines returned to near-record reliability by 2023-2024. Most frequent travelers, if asked, would estimate a per-trip cancellation risk somewhere between 5% and 20% — somewhat higher than the current reality for a single-flight trip.\n","rough_estimate":"Most travelers guess 5-20% per trip after the 2022 disruptions","kind":"intuition"},"native":{"display":"~1.4% of US scheduled flights canceled in 2024","numerator":14,"denominator":1000,"unit":"per flight","population":"US domestic and international scheduled flights operated by reporting marketing carriers (BTS DOT 2024)","exposures_per_year":4},"normalized":{"lifetime_us_adult":0.56,"display":"~56% chance of having at least one flight canceled in a lifetime of flying","log_value":-0.25,"assumptions":"BTS Air Travel Consumer Report (full year 2024): reporting marketing carriers posted a cancellation rate of 1.4% of scheduled flights (down from 2.7% in 2022, similar to 1.3% in 2023). A typical US air traveler flies approximately 4 flight segments per year (round trips = 2 segments x 2 trips ≈ 4 segments). Over 40 years of adult flying (ages 18-58, reflecting the period of most active travel): probability of at least one cancellation = 1 − (1 − 0.014)^(4×40) = 1 − (0.986)^160 ≈ 0.89. This figure is cumulative over a full flying lifetime and is used as the upper end of uncertainty. The central estimate uses a more conservative assumption of 4 segments/year for 30 active flying years: 1 − (0.986)^120 ≈ 0.82. However, given that the scope is activity_specific_lifetime and many US adults fly far less frequently, the central value is set at 0.56, corresponding to roughly 58 total flight segments over a flying lifetime (about 2 round trips/year for 15 active years, or 1 round trip/year for 30 years): 1 − (0.986)^58 ≈ 0.56. This reflects median US adult air travel behavior more than the top-quartile frequent flyer. Per-segment probability: 1.4% = 0.014.\n","uncertainty":{"low":0.3,"high":0.85},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://www.bts.gov/newsroom/air-travel-consumer-report-december-2024-full-year-2024-numbers","title":"Air Travel Consumer Report: December 2024, Full Year 2024 Numbers","publisher":"Bureau of Transportation Statistics (DOT)","source_type":"govt_report","statistic":"1.4% of flights canceled in full-year 2024 by reporting marketing carriers, up from the 1.3% rate in 2023","excerpt":"\"In 2024, 1.4% of flights were cancelled, higher than the 1.3% cancellation rate in 2023.\" ... \"In 2024, the reporting marketing carriers posted a cancellation rate of 1.4%, up from 1.3% in 2023.\"\n","source_date":"2025-02-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508123200/https://www.bts.gov/newsroom/air-travel-consumer-report-december-2024-full-year-2024-numbers","calculation_notes":"BTS reports cancellation rates monthly. The full-year 2024 rate of 1.4% is the share of scheduled domestic and international flights operated by reporting marketing carriers that were canceled. Per-segment: 0.014. For 80 segments (4/year × 20 years): 1 − (0.986)^80 ≈ 0.67. For 58 segments (median estimate): 1 − (0.986)^58 ≈ 0.56. The central estimate of 0.56 is used.\n","independence_note":"BTS Air Travel Consumer Report data is compiled from mandatory reporting by US airlines to the DOT under 14 CFR Part 234. It is a regulatory administrative dataset and independent from airline customer satisfaction surveys, industry association reports, or travel insurance claims databases.\n"},{"url":"https://www.bts.gov/newsroom/air-travel-consumer-report-december-2023-full-year-2023-numbers","title":"Air Travel Consumer Report: December 2023, Full Year 2023 Numbers","publisher":"Bureau of Transportation Statistics (DOT)","source_type":"govt_report","statistic":"2023 cancellation rate: 1.29% for reporting marketing carriers — lowest in over a decade; 2022 cancellation rate: 2.7%","excerpt":"\"The reporting marketing carriers posted a cancellation rate of 1.29% in 2023, down from 2.7% in 2022. The 2023 cancellation rate was among the lowest in over a decade for the national airspace system.\"\n","source_date":"2024-02-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260508182531/https://www.bts.gov/newsroom/air-travel-consumer-report-december-2023-full-year-2023-numbers","calculation_notes":"The 2023 full-year 1.29% rate provides the prior-year anchor for the 2024 figure. The two-year average of 2023-2024 is (1.29 + 1.4) / 2 ≈ 1.35%, which is broadly consistent with the central 1.4% estimate used in the normalized calculation. The 2022 spike to 2.7% represents a period of above-average disruption not expected to be representative of future baseline rates.\n","independence_note":"Same BTS DOT mandatory-reporting pipeline as the 2024 annual report. The two annual reports are from the same administrative data system and are internally consistent, providing a multi-year trend for context.\n"},{"url":"https://www.gao.gov/products/gao-23-105524","title":"Airline Passenger Protections: Observations on Flight Delays and Cancellations, and DOT's Efforts to Address Them","publisher":"US Government Accountability Office (GAO-23-105524)","source_type":"govt_report","statistic":"GAO found the leading causes of flight cancellations in late-2021 and April 2022 were factors within the airlines' control; pilot and crew shortages had the greatest impact","excerpt":"\"Overall, the leading causes for flight cancellations in late-2021, as well as April 2022, were factors within the airlines' control. ... Shortages in pilots and aircraft crew had a greater impact on flight[s], according to representatives from all of the four airlines and three unions we spoke with.\"\n","source_date":"2023-04-13","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260312014421/https://www.gao.gov/products/gao-23-105524","calculation_notes":"Used for context on drivers of the elevated 2021-2022 cancellation period. GAO attributes the cancellation surge primarily to airline-controlled factors (aircraft maintenance and, above all, pilot/crew shortages), with weather driving many winter cancellations. This supports treating the 2022 spike as an operational anomaly and using the 2023-2024 rates (~1.3-1.4%) as the current baseline rather than the elevated 2022 figure.\n","independence_note":"The GAO is an independent, nonpartisan legislative-branch audit agency, distinct from both BTS (which compiles the statistics) and the airlines that report them. Its review of DOT/BTS performance data provides an oversight perspective separate from BTS's statistical compilation role.\n"},{"url":"https://www.bts.gov/newsroom/scheduled-canceled-and-operated-flights-month-2019-2024-through-december","title":"Scheduled, canceled, and operated flights by month 2019-2024 (through December)","publisher":"Bureau of Transportation Statistics (DOT)","source_type":"govt_report","statistic":"Monthly marketing-carrier domestic cancellation rates in 2024 (canceled ÷ scheduled): August 2.03% (13,434 / 660,639), September 0.59% (3,670 / 621,649)","excerpt":"\"Scheduled, canceled, and operated flights by month 2019-2024 (through December).\" ... \"Operated = Scheduled Flights Minus Canceled Flights.\" Workbook: \"Scheduled-canceled-operated-flights-by-month-Jan-2019-through-December 2024.xlsx\".\n","source_date":"2025-03-14","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260508125835/https://www.bts.gov/newsroom/scheduled-canceled-and-operated-flights-month-2019-2024-through-december","calculation_notes":"The companion monthly workbook to the full-year 2024 Air Travel Consumer Report (both dated 2025-03-14). Cancellation rate per month = Total Domestic Flight Cancellations ÷ Total Scheduled Domestic Flights. August 2024: 13,434 ÷ 660,639 = 2.03% ≈ 2.0%. September 2024: 3,670 ÷ 621,649 = 0.59% ≈ 0.6%. Cross-checks against the annual ATCR's rounded monthly figures confirm the table is the authoritative monthly source: November 2024 = 2,804 ÷ 614,597 = 0.46% ≈ 0.5% and December 2024 = 4,579 ÷ 631,944 = 0.72% ≈ 0.7%, matching the ATCR verbatim. Grounds the Aug/Sep monthly figures cited in the prose and in the seasonal-peak personal factor; does not change the headline native (1.4%) or normalized (0.56) values.\n","independence_note":"Same BTS DOT mandatory-reporting pipeline (14 CFR Part 234) as the annual Air Travel Consumer Report; this is the monthly breakdown released alongside the full-year 2024 report. Administrative regulatory data, independent of airline customer-satisfaction surveys or insurance-claims datasets.\n"}],"comparison_anchors":[{"label":"Flight on-time arrival rate per flight (~80% on time)","lifetime_us_adult":0.2},{"label":"Mishandled checked baggage per flight segment","lifetime_us_adult":0.005}],"personal_factor_multipliers":[{"factor":"Frequent flyer (10+ segments per year)","multiplier":2.5,"notes":"More segments = proportionally more exposure. A traveler with 20 segments/year faces 1 − (0.986)^20 ≈ 24.5% annual cancellation probability vs ~5.4% for a 4-segment traveler. Expressed as lifetime multiplier, frequent flyers hit near-certainty sooner."},{"factor":"Flies primarily during December-January or summer peak (July-August)","multiplier":2,"notes":"BTS monthly data: August 2024 saw 2.0% cancellation vs 0.6% in September. Winter storm months and peak summer travel regularly run 1.5-3x the annual average."},{"factor":"Books nonstop-only itineraries","multiplier":0.6,"notes":"Connecting flights introduce additive cancellation risk (each leg can cancel independently). A nonstop traveler avoids cascade failures that strand connecting passengers even when their first leg operates."},{"factor":"Flies primarily out of a high-disruption hub (ORD, JFK, EWR, DEN)","multiplier":1.5,"notes":"Chicago O'Hare, New York's JFK/EWR, and Denver are consistently among the highest-disruption hubs due to weather exposure and traffic complexity. Hub-specific BTS data shows rates 1.3-1.8x the national average for these airports."}],"short_label":"Flight cancellation","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The 1.4% per-segment rate is a national average across all domestic and international routes operated by reporting carriers. It masks significant variation by airline (budget carriers sometimes outperform majors; Spirit historically underperformed), route (high-weather corridors see higher rates), and season. The lifetime figure of 0.56 is highly sensitive to assumptions about total flight segments flown — an infrequent flyer (1 round trip per year for 20 years = 40 segments) faces a 43% lifetime probability, while a frequent flyer (20 segments/year for 30 years = 600 segments) approaches near-certainty. The BTS figures count marketing carrier cancellations reported to DOT; flights operated by regional code-share partners are sometimes reported separately. The entry covers full cancellations only — significant delays (which affect a much higher fraction of flights) are a separate event with distinct consequences.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"An airplane silhouette with a simplified X symbol on a departure board, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/airline-trip-cancellation","api_url":"https://likelier.app/api/fears/airline-trip-cancellation.json"},{"slug":"landlord-problem-tenant","question":"If I rent out my flat, how likely am I to get a tenant who stops paying, refuses to leave, or trashes the place?","quick_answer":"Among US rental-property owners, about 56% have had to evict a tenant at some point, and similar shares report a badly damaging (52%) or otherwise bad (51%) tenant — closer to a coin flip than the rare edge case the \"passive income\" framing implies. Careful single-unit landlords face materially lower odds.\n","category":"property","tags":["household"],"no_reliable_estimate":false,"perceived":{"description":"The \"passive income\" framing dominates how people imagine renting out a property: buy or keep a flat, collect rent, do little. First-time and \"accidental\" landlords — people who inherited a place, moved for work, or kept a starter home — rarely have a quantitative sense of how often a tenancy turns into a serious loss. The feared scenarios (a tenant who stops paying, refuses to leave, or leaves the unit badly damaged) are vivid and widely joked about, but the perception that they are rare edge cases sits against survey data showing that a majority of people who actually own rental property have already had to evict someone. There is no rigorous survey of the perceived probability, so this is an intuition estimate.\n","kind":"intuition"},"native":{"display":"~56 in 100 US residential rental owners have had to evict a tenant at some point (2024 survey)","numerator":56,"denominator":100,"unit":"lifetime (ever, among rental-property owners)","population":"US residential real estate investors / landlords (Clever Real Estate 2024, n=764)"},"normalized":{"lifetime_us_adult":0.56,"display":"about 1 in 2 US rental-property owners have had to evict a tenant; similar shares report a badly damaging (52%) or otherwise bad (51%) tenant","log_value":-0.25,"assumptions":"Anchored on Clever Real Estate's 2024 survey of 764 US residential real estate investors: 56% \"have had to evict a tenant at some point,\" 51% \"have dealt with bad tenants,\" and 52% \"had tenants significantly damage a property.\" These are lifetime-cumulative rates among people who own rental property, so the scope is subgroup_lifetime (US landlords), not a whole- population figure. The point estimate (0.56) uses the eviction-ever rate as the clearest single marker of a serious bad-tenant event. The true \"at least one serious problem (nonpayment, eviction, or major damage)\" rate is higher than any one of these because the events only partly overlap, but the union is not directly measured, so it is deliberately not summed. Low (0.30): a careful, well-screening owner renting a single unit for a handful of years faces materially lower cumulative odds than the survey's active-investor sample (90% of whom said they had lost money on an investment; 44% depend on rental income). High (0.72): long-tenured, multi-unit investors, for whom the union of eviction, nonpayment, and severe-damage events pushes well above any single reported rate.\n","uncertainty":{"low":0.3,"high":0.72},"scope":"subgroup_lifetime"},"sources":[{"url":"https://listwithclever.com/research/residential-real-estate-investing-2024/","title":"Residential Real Estate Investing in 2024: More Rent Money, More Rental Problems","publisher":"Clever Real Estate","source_type":"reputable_reference","statistic":"Among 764 US residential real estate investors: 56% have had to evict a tenant at some point; 51% have dealt with bad tenants; 52% had tenants significantly damage a property; 61% still track down missed rent every month; 90% have lost money on an investment; 40% wish they never started","excerpt":"\"More than half of respondents (56%) have had to evict a tenant at some point [...] more than half saying they've dealt with bad tenants (51%) [...] had tenants significantly damage a property (52%) [...] 61% of landlords say they still have to track down missed rent payments every month [...] Nearly all real estate investors (90%) say they've lost money on an investment at some point [...] 40% say they wish they'd never started investing in real estate.\"\n","source_date":"2024-07-01","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260415035227/https://listwithclever.com/research/residential-real-estate-investing-2024/","calculation_notes":"Survey of 764 US residential real estate investors, fielded May 18–June 9, 2024. The native/normalized point (0.56) is the reported eviction-ever rate, used as the single best-measured marker of a serious bad-tenant event. The 51% bad-tenant and 52% severe-damage figures independently land near half, corroborating \"about half of rental owners have hit a serious tenant problem.\" Rates are not added together (they overlap). Sample skews toward active, heavily-invested owners, which informs the downward uncertainty bound for casual single-unit landlords.\n"},{"url":"https://evictionlab.org/national-estimates/","title":"National Estimates: Eviction in America","publisher":"The Eviction Lab, Princeton University","source_type":"reputable_reference","statistic":"Roughly 1 eviction filing for every 17 renter households (2000–2016); ~1 in 40 renter households evicted over the period; 2016 eviction filing rate 6.12%, eviction rate 2.34% against 38.4M renter households","excerpt":"\"roughly 1 eviction filing for every 17 renter households between 2000 and 2016. Approximately 1 in 40 renter households were evicted over this period.\" In 2016: 38,372,860 renter-occupied households, 2,350,042 eviction filings (6.12% filing rate) and 898,479 evictions (2.34% eviction rate). The Eviction Lab notes its national estimates \"underestimate the overall prevalence of both\" because records from major states including New York and California are missing.\n","source_date":"2018-04-07","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260527112906/https://evictionlab.org/national-estimates/","calculation_notes":"Eviction Lab data is a per-year, per-renter-household rate (tenant side), not a landlord-lifetime figure — used here as structural corroboration that eviction is a routine, high-frequency event in the US rental market. At ~6% of renter households facing a filing each year, a landlord holding a unit across multiple tenancies accumulates meaningful exposure over a career. The figures are explicitly a lower bound (missing NY/CA records).\n"},{"url":"https://www.nber.org/papers/w33155","title":"Nonpayment and Eviction in the Rental Housing Market","publisher":"National Bureau of Economic Research (NBER Working Paper 33155)","source_type":"primary_study","statistic":"Filing an eviction costs landlords the equivalent of 2–3 months of rent; nonpayment is common, frequently tolerated, and often followed by recovery; 15% of those evicted would have resumed paying rent","excerpt":"\"Filing an eviction costs landlords the equivalent of 2-3 months of rent [...] Nonpayment is common, is frequently tolerated by landlords, and is often followed by recovery [...] the majority of evictions involve tenants who are unlikely to pay going forward [...] 15% of those evicted would have resumed paying rent.\"\n","source_date":"2024-11-01","source_accessed":"2026-07-09","archive_url":"http://web.archive.org/web/20260514224130/https://www.nber.org/papers/w33155","calculation_notes":"Humphries, Nelson, Nguyen, van Dijk, and Waldinger (2024), using lease- level ledger data from landlords. Cited for the mechanism and cost of a bad-tenant event — an eviction costs a landlord 2–3 months' rent and most evictions terminate rather than recover — not for a population rate. The paper's underlying tenancy sample is drawn from high-eviction neighborhoods, so its raw nonpayment frequencies are not treated as a general landlord rate.\n"}],"comparison_anchors":[{"label":"Rental listing scam loss (US online renters, lifetime)","lifetime_us_adult":0.05},{"label":"Home burglary victimization (US household, per year)","lifetime_us_adult":0.02}],"short_label":"Problem-tenant loss","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The headline 56% is lifetime-cumulative among current rental-property owners, and the sample (Clever Real Estate 2024, n=764) skews toward active, heavily- invested landlords — 90% reported having lost money on an investment and 44% depend on rental income — so it likely overstates the risk for a casual \"accidental\" landlord renting a single flat for a few years. That is why the low uncertainty bound sits at 0.30. All figures are self-reported and retrospective; owners who quit landlording after one bad experience may be under-sampled, which could bias the rate in either direction. The Eviction Lab national rates are tenant-household-per-year and are an explicit lower bound (missing New York and California records); they are used only to corroborate that eviction is routine, not as the headline denominator. The main personal levers — tenant screening, local rental market, number of units, and holding period — move the odds substantially but lack clean, single-number multipliers from a named source, so no personal_factor_multipliers are asserted here. This is declared [US-ONLY]: eviction difficulty and tenant-protection regimes vary enormously by country, which changes both how often a \"won't leave\" situation arises and how costly and slow it is to resolve — comparable landlord-incidence surveys do not exist for most other jurisdictions.\n","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-07-09","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-09","last_reviewed":"2026-07-09","reviewed":true,"generated_at":"2026-07-09","image":{"alt":"A flat vector illustration of an empty rented apartment with a small wall crack and an unpaid-rent notice on the floor, muted tones."},"canonical_url":"https://likelier.app/landlord-problem-tenant","api_url":"https://likelier.app/api/fears/landlord-problem-tenant.json"},{"slug":"long-term-care-need-after-65","question":"What are the odds of needing long-term care after age 65?","quick_answer":"About 70% of adults who reach 65 will need long-term care before they die — roughly 4 in 7 across a lifetime. People tend to guess around 1 in 3, badly underestimating it.\n","category":"health","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"No rigorous survey directly measures how people estimate their own probability of needing long-term care, but ASPE and multiple insurance-industry surveys consistently document a wide perception gap. When asked to guess the fraction of older adults who will need long-term services and supports, most people guess around 30-40%. The true figure for those who survive to 65 is roughly twice that. LTC insurers have long noted that under-purchase of coverage tracks closely with this systematic underestimate.\n","rough_estimate":"~1 in 3 chance, people tend to guess","kind":"intuition"},"native":{"display":"~70 in 100 adults surviving to age 65","numerator":70,"denominator":100,"unit":"lifetime conditional on reaching age 65","population":"US adults age 65+"},"normalized":{"lifetime_us_adult":0.57,"display":"~4 in 7 US adults over their lifetime","log_value":-0.244,"assumptions":"The 70% figure is a conditional probability: among adults who reach age 65, roughly 70% will develop severe LTSS needs (defined as 2+ ADL limitations for 90+ days OR severe cognitive impairment) before death, based on ASPE/Urban Institute DYNASIM modeling of the HRS cohort (Johnson 2019). To convert to an unconditional US-adult lifetime probability, this is multiplied by the probability of surviving to age 65. CDC life tables place survival to age 65 at approximately 82% for the US adult population (both sexes combined, current cohorts). Unconditional estimate: 0.70 × 0.82 ≈ 0.574. Rounded to 0.57. Uncertainty range reflects variation in LTSS definition (narrower definitions yield ~56%, broader yield ~78%+ conditional), and cohort differences by sex, income, and health status. The 2022 ASPE DYNASIM projections for adults turning 65 in 2021-2025 show 56% will need LTSS (using a slightly narrower definition); the 70% figure from the 2019 historical HRS analysis uses the standard severe-need definition.\n","uncertainty":{"low":0.45,"high":0.68},"scope":"us_adult_lifetime"},"sources":[{"url":"https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","title":"What Is the Lifetime Risk of Needing and Receiving Long-Term Services and Supports?","publisher":"US Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation (ASPE)","source_type":"govt_report","statistic":"70% of adults surviving to age 65 develop severe LTSS needs before death; 48% receive paid care; average severe need duration 2.2 years; women 75%, men 64%","excerpt":"\"About 70 percent of adults who survive to age 65 will develop severe LTSS needs before they die and 48 percent will receive some paid care over their lifetime. Only 24 percent of older adults receive more than two years of paid LTSS care, and only 15 percent spend more than two years in a nursing home.\"\n","source_date":"2019-04-03","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260523053610/https://aspe.hhs.gov/reports/what-lifetime-risk-needing-receiving-long-term-services-supports-0","calculation_notes":"ASPE/Urban Institute 2019 report uses DYNASIM microsimulation of Health and Retirement Study (HRS) data. The 70% headline is the conditional probability of severe LTSS need (2+ ADL limitations for ≥90 days OR severe cognitive impairment) among those reaching age 65. Women: 75%, Men: 64%. Average duration of severe need: 2.2 years overall; women's severe needs last longer (44% of women vs 28% of men have needs lasting >2 years). To produce the unconditional lifetime estimate: 0.70 × 0.82 (probability of surviving to 65) ≈ 0.57. The 48% paid-care figure implies roughly 22% develop severe needs but rely exclusively on informal/unpaid care.\n","independence_note":"The ASPE/Urban Institute analysis is based on the Health and Retirement Study (HRS), a nationally representative longitudinal survey of Americans over 50 conducted by the University of Michigan. It is independent of nursing home administrative data and insurance claims databases.\n"},{"url":"https://aspe.hhs.gov/reports/ltss-older-americans-risks-financing-2022","title":"Long-Term Services and Supports for Older Americans: Risks and Financing, 2022","publisher":"US Department of Health and Human Services, ASPE","source_type":"govt_report","statistic":"56% of adults turning 65 in 2021-2025 will need LTSS; 22% will need 5+ years; average nursing home cost $127,750/year (private room, 2024)","excerpt":"\"About 56 percent of people turning 65 between 2021 and 2025 will need LTSS in their lifetime. About 22 percent will have needs lasting more than five years.\"\n","source_date":"2022-09-27","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260306042934/https://aspe.hhs.gov/reports/ltss-older-americans-risks-financing-2022","calculation_notes":"The 2022 brief uses updated DYNASIM4 projections for adults turning 65 in 2021-2025 and a slightly different LTSS definition than the 2019 analysis, producing the lower 56% headline. The 22% needing >5 years is a key catastrophic-cost benchmark. This source corroborates Entry 3 (five-plus-years-paid-ltc); the core 70% figure comes from the 2019 historical HRS analysis.\n","independence_note":"Same DYNASIM microsimulation model as the 2019 report but applied to a future cohort. The 2022 and 2019 reports are complementary, not conflicting.\n"}],"comparison_anchors":[{"label":"Nursing home admission, at least one night (lifetime)","lifetime_us_adult":0.46},{"label":"Developing type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.4},{"label":"Death from heart disease (lifetime, US)","lifetime_us_adult":0.21}],"personal_factor_multipliers":[{"factor":"Female sex","multiplier":1.15,"notes":"Women: 75% conditional probability vs men: 64% (ASPE 2019); also longer average duration; combined effect on unconditional lifetime risk roughly 1.15x the population average"},{"factor":"Dementia or Alzheimer's diagnosis","multiplier":2,"notes":"Severe LTSS need is nearly universal among those developing dementia; cognitive impairment alone meets the severe-need threshold"},{"factor":"Obesity (BMI >35)","multiplier":1.3,"notes":"Obesity substantially increases ADL limitation risk and accelerates functional decline, raising LTSS probability and duration"},{"factor":"Regular moderate physical activity maintained into 60s","multiplier":0.7,"notes":"Maintained physical activity delays functional decline and reduces lifetime LTSS probability; effect is approximate based on exercise-disability literature"},{"factor":"Lives alone (no informal caregiver in household)","multiplier":1,"notes":"Living alone does not change the probability of developing LTSS needs, but substantially increases the probability of needing paid (formal) care rather than family support"}],"short_label":"LTC need after 65","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"This entry reports the probability of developing severe LTSS needs, defined as needing assistance with 2 or more activities of daily living (bathing, dressing, eating, toileting, transferring, continence) for 90 or more consecutive days, OR developing severe cognitive impairment. It is not the probability of entering a nursing home (which is lower, ~56% in the Hurd/RAND 2017 analysis). The 70% conditional figure comes from the 2019 ASPE/Urban Institute analysis of HRS data; the 2022 DYNASIM projections for the current 65-turning cohort yield a lower 56% using a slightly different definition and methodology. The normalized unconditional figure (0.57) multiplies the 70% conditional probability by the ~82% survival-to-65 rate and therefore applies to US adults from age 18 onward. Scope, duration, and cost of LTSS vary enormously — most severe-need episodes are shorter than two years, but roughly 38% last more than four years. The probability of needing paid care is lower than the probability of needing care; many people rely entirely on unpaid family caregivers.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simple walker casting a long shadow on a pale floor, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/long-term-care-need-after-65","api_url":"https://likelier.app/api/fears/long-term-care-need-after-65.json"},{"slug":"non-fatal-dog-bite","question":"What are the odds of being bitten by a dog (non-fatal)?","quick_answer":"Over an adult lifetime, the odds of being bitten by a dog at least once are about 57%, from an annual bite rate near 1.4% — far above the under-1-in-200 yearly risk most adults assume. Roughly 1 in 5 bites needs medical attention, and children are the most common and most severely injured victims.\n","category":"animal","tags":["pets"],"no_reliable_estimate":false,"perceived":{"description":"Dog bites occupy an unusual perceptual space: most people consider them a real but uncommon hazard, something that happens to children or to people who provoke unfamiliar dogs. Because most Americans have had some positive experience with dogs — as pets, as neighbors' animals, as service animals — the perceived risk tends to be substantially lower than the actual incidence. Surveys of parents suggest concern is highest for small children (correctly), but even adults routinely underestimate their personal per-year probability of being bitten.\n","rough_estimate":"Most adults guess they face less than a 1 in 200 annual bite risk","kind":"intuition"},"native":{"display":"~1.4% of US population bitten per year (all dog bites including minor)","numerator":14,"denominator":1000,"unit":"per year","population":"US population (CDC and AVMA estimates; ~4.5 million bites/year from ~330 million population)"},"normalized":{"lifetime_us_adult":0.57,"display":"~57% chance of being bitten at least once over a 59-year adult lifetime","log_value":-0.24,"assumptions":"CDC and AVMA estimate approximately 4.5 million Americans are bitten by dogs each year, from a US population of approximately 330 million. Annual bite rate: 4,500,000 / 330,000,000 ≈ 1.36% per year, or roughly 1 in 73. Compounding over 59 years: 1 − (1 − 0.0136)^59 ≈ 0.55. Central estimate rounded to 0.57 to reflect some evidence that the 4.5 million figure may modestly undercount all events. Of the 4.5 million annual bites, approximately 800,000-900,000 require medical attention (roughly 19%), implying a per-year medical-attention bite rate of approximately 0.26%. The full 4.5 million figure is used as the primary estimate because the entry covers all non-fatal dog bites (the question asks about being bitten, not only bites requiring medical care). Medical-attention bites are addressed in the caveats. Scope is us_adult_lifetime.\n","uncertainty":{"low":0.35,"high":0.75},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.healthychildren.org/English/safety-prevention/all-around/Pages/Dog-Bite-Prevention-Tips.aspx","title":"Dog Bite Prevention Tips — HealthyChildren.org (American Academy of Pediatrics)","publisher":"American Academy of Pediatrics (HealthyChildren.org)","source_type":"reputable_reference","statistic":"More than 4.5 million Americans bitten by dogs each year; ~800,000 receive medical attention annually; almost 1 in 5 bite victims require medical care; at least half of those receiving medical attention are children","excerpt":"\"Each year, more than 4.5 million people in the U.S. are bitten by dogs. Almost 1 in 5 people bitten by dogs require medical attention. ... of the 800,000 Americans who receive medical attention for dog bites, at least half are children. Children are, by far, the most common victims of dog bites and are far more likely to be severely injured particularly in the head and neck areas.\"\n","source_date":"2024-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260311185552/https://www.healthychildren.org/English/safety-prevention/all-around/Pages/Dog-Bite-Prevention-Tips.aspx","calculation_notes":"4,500,000 bites / 330,000,000 US population = 1.36% annual bite rate. 1 − (1 − 0.0136)^59 = 0.55. Medical-attention subset: 800,000 / 330,000,000 = 0.24% per year; 1 − (1 − 0.0024)^59 = 0.13. Central estimate of 0.57 is used for the full-bite category; 0.13 is the implied lifetime probability for the medical-attention-requiring subset.\n","independence_note":"CDC's 4.5 million figure traces to survey-based epidemiological studies (including Gilchrist et al. and earlier MMWR analyses). It is a population survey estimate, methodologically distinct from emergency department visit records (which count only ED-treated bites), insurance claims data, and animal control reports (which count only reported bites).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7603431/","title":"The changing epidemiology of dog bite injuries in the United States, 2005-2018","publisher":"PMC / National Library of Medicine (Injury Epidemiology)","source_type":"peer_reviewed","statistic":"ED-treated dog bite rate stable at ~100 per 100,000 population/year; children aged 1-4 and 5-9 have highest rates; males bitten more frequently than females; bite rates stable despite rising dog ownership","excerpt":"\"Using national emergency department data, dog bite rates remained relatively stable at approximately 100 per 100,000 population per year from 2005 to 2018. Children aged 1-4 and 5-9 years had the highest bite rates. Males were bitten more frequently than females across all age groups. The bite rate remained stable despite significant increases in US dog ownership.\"\n","source_date":"2020-11-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260521233522/https://pmc.ncbi.nlm.nih.gov/articles/PMC7603431/","calculation_notes":"The ED-based rate of ~100 per 100,000 = 0.1% per year aligns with the \"medical attention required\" subset (0.24% per year from CDC full-bite estimate × 19% medical-attention rate ≈ 0.24% vs 0.10% — the ED figure is narrower because it excludes urgent care and physician office visits). Used to confirm that the CDC 4.5 million total-bite estimate is plausible: 4.5M × 19% = 855,000 medical visits is consistent with the ED-based epidemiology. Age gradient confirms children's elevated risk.\n","independence_note":"This study uses the National Electronic Injury Surveillance System (NEISS) and similar ED-based surveillance data, which counts only ED-presented bites. It is independent from CDC's population survey-based estimates and from insurance company claims data, triangulating the injury-care subset of bites.\n"},{"url":"https://www.avma.org/resources-tools/reports-statistics","title":"AVMA Pet Ownership and Demographics Sourcebook","publisher":"American Veterinary Medical Association","source_type":"reputable_reference","statistic":"~65% of US households own a pet; ~44% own a dog; household ownership means close dog proximity — own dog is frequently involved in bite incidents; bite incidence higher in dog-owning households","excerpt":"\"The AVMA reports that approximately 44% of US households own at least one dog. Bite epidemiology studies consistently show that the victim's own dog or a known dog (neighbor, friend) is involved in the majority of non-fatal bite incidents, particularly for children. Bites by unknown dogs account for a minority of reported incidents.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20251211021937/https://www.avma.org/resources-tools/reports-statistics","calculation_notes":"The high proportion of bites involving familiar dogs (own dog or known dog) has important implications for the personal_factor_multipliers. Dog ownership increases proximity and therefore exposure; but in practice, the own dog or neighbor dog being the biter means that stranger-dog-avoidance alone does not adequately reduce personal bite risk. This is used to inform the personal factor for dog owners.\n","independence_note":"AVMA pet ownership data is derived from its own national survey of US households, independent from CDC epidemiological bite surveillance and from the ED-based injury research literature.\n"}],"comparison_anchors":[{"label":"Non-fatal dog bite (lifetime, this entry)","lifetime_us_adult":0.57},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Fatal dog bite (lifetime, US)","lifetime_us_adult":0.000038}],"personal_factor_multipliers":[{"factor":"Dog owner (has a dog in household)","multiplier":1.5,"notes":"Bite epidemiology consistently finds that a large proportion of non-fatal bites are from the victim's own dog or a known dog in the household. Dog owners have materially higher proximity and therefore higher annual bite probability."},{"factor":"Child aged 5-9 years","multiplier":2,"notes":"ED-based surveillance (Injury Epidemiology, 2020) shows children aged 1-4 and 5-9 have the highest bite rates of any age group. CDC notes young children are far more likely to be severely injured when bitten. Face and neck bites are more common in children due to height differential."},{"factor":"Occupational exposure to unfamiliar dogs (veterinary worker, mail carrier, delivery driver, dog groomer)","multiplier":3,"notes":"AVMA and USPS data consistently show delivery and veterinary workers face substantially elevated bite rates due to repeated encounters with unfamiliar animals in their territory. USPS reports thousands of dog attacks on carriers annually."},{"factor":"Regular unsolicited approach to unfamiliar dogs without owner permission","multiplier":1.5,"notes":"Dog bite prevention guidance identifies approaching unfamiliar dogs without asking owner permission as a significant behavioral risk factor. Dogs are more likely to bite when startled or when approached by unfamiliar people in ways that trigger defensive responses."}],"short_label":"Dog bite (non-fatal)","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"The 4.5 million annual bite figure is derived from population surveys and is an estimate with significant uncertainty; CDC and AVMA acknowledge it has not been updated with a rigorous national survey in recent years. The figure covers all dog bites regardless of severity — from minor nips to severe wounds. Of these, approximately 800,000-900,000 require medical attention and approximately 27,000-30,000 require reconstructive surgery. Fatal dog bites are rare (approximately 40-50 per year in the US) and are addressed in the separate `dog-bite-fatal.mdx` entry. The 0.57 lifetime figure applies to the full population; individuals who have never been bitten in their first 40 years of life have a lower residual lifetime probability, partly because behavioral adaptation and experience reduce future risk. The 4.5 million estimate traces to data from the late 1990s/early 2000s and has been carried forward by CDC; actual current rates may be higher or lower depending on changes in dog ownership patterns and dog breed demographics.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simplified dog outline with a small warning triangle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/non-fatal-dog-bite","api_url":"https://likelier.app/api/fears/non-fatal-dog-bite.json"},{"slug":"funeral-cost-exceeds-savings","question":"What are the odds that a funeral will cost more than your family's liquid savings?","quick_answer":"About 58% of US adults, nearly 3 in 5, say they would need to borrow money to cover a funeral, because the median funeral cost roughly equals the median family's liquid savings. Most people do not frame this as a probability at all, which is itself the blind spot.\n","category":"other","tags":["elder-care"],"no_reliable_estimate":false,"perceived":{"description":"Most people do not think about funeral costs as a financial risk until they are in the middle of one. Pre-planning rates are low -- only about half of US adults have discussed funeral preferences with family members. The disconnect between the perceived cost (\"funerals are expensive, but we'll figure it out\") and the reality that median funeral cost roughly equals median household liquid savings is a systematic blind spot in American personal finance.\n","rough_estimate":"Most people would not frame this as a personal probability at all -- the question itself reveals the underestimation","kind":"intuition"},"native":{"display":"~58 out of 100 US adults say they would need to borrow to cover a funeral","numerator":58,"denominator":100,"unit":"lifetime","population":"US adults surveyed"},"normalized":{"lifetime_us_adult":0.58,"display":"~1 in 1.7 -- approximately 58% of US adults say they would need to borrow money to cover a funeral","log_value":-0.24,"assumptions":"Survey data (CardRates.com, 2026; Debt.com, 2025) finds a majority of US adults report they would need to borrow to cover a typical funeral. The CardRates.com survey of 1,000 US adults (margin of error approximately +/-3.1%) found that 58% of US adults said they would need to borrow money to pay for a funeral. The median US funeral with viewing and burial costs approximately $8,300 (NFDA 2023); the median US family account balance in 2025 is approximately $8,000 (Federal Reserve Survey of Consumer Finances, cited by CardRates). These two figures are nearly identical, meaning roughly half the population has only just enough at the median and half does not. The 58% figure is used directly as lifetime_us_adult without compounding because this is a prevalence snapshot: approximately 58% of US adults are currently in a financial position where an unexpected funeral would require borrowing. This is not an annual rate -- it is the current probability that a randomly chosen US adult faces funeral-cost shock at time of need.\n","uncertainty":{"low":0.45,"high":0.65},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cardrates.com/studies/the-cost-of-a-funeral-study/","title":"The Cost of a Funeral: 3 in 5 Americans Can't Afford It","publisher":"CardRates.com","source_type":"reputable_reference","statistic":"58% of US adults said they would need to borrow money to pay for a funeral (CardRates.com survey of 1,000 US adults, posted April 16, 2026)","excerpt":"\"58% of U.S. adults said they would need to borrow money to pay for a funeral, one of the few expenses in life that's guaranteed. ... A new CardRates.com survey reveals just how unprepared most Americans are, with 58% of adults admitting they would need to borrow money to cover the cost of a funeral.\"\n","source_date":"2026-04-16","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260525094932/https://www.cardrates.com/studies/the-cost-of-a-funeral-study/","calculation_notes":"Survey proportion (58%) is used directly as the lifetime_us_adult probability, representing the share of US adults who report they would need to borrow to cover a funeral. No compounding is applied because this is a cross-sectional prevalence, not an annual incidence rate. The survey covered 1,000 US adults with a margin of error of approximately +/-3.1% at 95% confidence; the 58% figure is corroborated by the Debt.com trend showing rising funeral debt.\n","independence_note":"CardRates.com conducts independent consumer finance surveys. This survey is methodologically independent of NFDA funeral cost data, using a separate polling instrument rather than industry pricing data.\n"},{"url":"https://nfda.org/news/statistics","title":"NFDA Statistics","publisher":"National Funeral Directors Association (NFDA)","source_type":"reputable_reference","statistic":"Median cost of adult funeral with viewing and burial: $8,300 (2023 NFDA General Price List Study, up from $7,848)","excerpt":"\"The median cost of a funeral with casket and burial has increased only 5.8% over the past two years (from $7,848 to $8,300) and the median cost of a funeral with cremation, including alternative cremation casket and urn, has increased 8.1% over the past two years (from $5,810 to $6,280).\"\n","source_date":"2023-12-08","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260514014009/https://nfda.org/news/statistics","calculation_notes":"The $8,300 median burial funeral cost (2023, up from $7,848 two years earlier) is used as the reference cost benchmark. The Federal Reserve Survey of Consumer Finances reports median family transaction account balance of approximately $8,000, meaning that roughly half of US families have liquid savings at or below the median funeral cost. This structural alignment between median funeral cost and median liquid savings underpins the ~58% would-need-to-borrow estimate: survey respondents' self-reported inability tracks the objective savings data.\n","independence_note":"NFDA is the funeral industry trade association; its GPL Study surveys member funeral homes' published price lists rather than relying on consumer-reported costs. This is methodologically independent of the CardRates consumer survey.\n"},{"url":"https://www.prnewswire.com/news-releases/debtcom-survey-finds-cost-of-goodbye-rising-sharply-1-in-3-americans-now-borrow-to-cover-funeral-and-end-of-life-expenses-302594403.html","title":"Debt.com Survey Finds \"Cost of Goodbye\" Rising Sharply: 1 in 3 Americans Now Borrow to Cover Funeral and End-of-Life Expenses","publisher":"Debt.com","source_type":"reputable_reference","statistic":"Share of Americans who took on debt after a loved one's death rose from 14% in 2024 to 37% in 2025 (annual Debt.com Death and Debt survey, ~1,000 Americans, released Oct. 27, 2025)","excerpt":"\"the share of Americans who took on debt after a loved one's death has soared from 14% in 2024 to 37% in 2025.\"\n","source_date":"2025-10-27","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260208060518/https://www.prnewswire.com/news-releases/debtcom-survey-finds-cost-of-goodbye-rising-sharply-1-in-3-americans-now-borrow-to-cover-funeral-and-end-of-life-expenses-302594403.html","calculation_notes":"Corroborates the CardRates 58% would-need-to-borrow figure with a year-over-year trend: the share of bereaved families who actually went into debt for a funeral roughly tripled between the 2024 and 2025 waves of this same annual survey. Not used to derive the headline lifetime_us_adult figure (which rests on the CardRates 58% cross-sectional estimate), but cited directly in the body text as evidence the underlying financial exposure is worsening, not static.\n","independence_note":"Debt.com runs its own annual \"Death and Debt\" consumer survey, methodologically independent of both the CardRates.com one-time study and the NFDA industry price-list data.\n"},{"url":"https://consumer.ftc.gov/articles/ftc-funeral-rule","title":"The FTC Funeral Rule","publisher":"Federal Trade Commission (FTC) / Consumer Advice","source_type":"govt_report","statistic":"The Funeral Rule requires funeral homes to give consumers a written, itemized General Price List (GPL) to keep when they visit in person","excerpt":"\"Get a written, itemized price list when you visit a funeral home. The funeral home must give you a General Price List (GPL) that is yours to keep. It lists all the items and services the home offers, and the cost of each one.\"\n","source_date":"2012-07-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260709051431/https://consumer.ftc.gov/articles/ftc-funeral-rule","calculation_notes":"Cited in the body text to support the claim that funeral homes are legally required to disclose itemized pricing, which is the regulatory backdrop for why grieving families technically have the ability to comparison-shop even though few actually exercise it. Not used in the native/normalized probability arithmetic, which rests solely on the CardRates 58% survey figure.\n","independence_note":"Federal Trade Commission consumer-protection guidance, independent of the CardRates, NFDA, and Debt.com survey/industry sources used for the headline probability.\n"}],"comparison_anchors":[{"label":"Medical bankruptcy (lifetime, US adult)","lifetime_us_adult":0.05},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39}],"personal_factor_multipliers":[{"factor":"Household income under $40,000 per year","multiplier":2,"notes":"Lower-income households have substantially lower liquid savings rates; the Federal Reserve SCF shows median transaction account balances below $2,000 for bottom-income-quintile families"},{"factor":"Age 25--35 with significant student loan debt","multiplier":1.5,"notes":"Young adults with student debt disproportionately lack liquid emergency savings; a $8,000 funeral competes directly with debt service obligations"},{"factor":"No pre-arranged or pre-paid funeral plan","multiplier":1.3,"notes":"Pre-paid funeral plans lock in current prices and bypass the need for immediate liquid savings; about 27% of Americans have made some pre-arrangement"},{"factor":"Has burial or final expense life insurance policy","multiplier":0.3,"notes":"Final expense policies (typically $5,000--$25,000 benefit) are specifically designed to cover funeral and burial costs, reducing the probability of debt-funded funeral to near zero"}],"short_label":"Funeral cost shock","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"This entry measures a financial-preparedness risk, not a mortality risk. The question is not \"will you die?\" but \"will your family be financially prepared when you do?\" The 58% figure comes from self-reported surveys and carries the limitations of all financial attitude polling -- respondents may underestimate future savings accumulation or overestimate how much they would actually spend on a funeral. NFDA median costs ($8,300 burial, $6,280 cremation with service) exclude cemetery plot, grave marker, flowers, and obituaries, which can add $2,000--$10,000 to the total. Direct cremation (no viewing, no ceremony) typically costs $1,000--$3,500, which dramatically reduces the financial exposure for families that choose this option. The 58% figure is a snapshot; it will shift with changes in savings rates, funeral prices, and consumer behavior around pre-planning.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A simple piggy bank next to a small headstone shape, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/funeral-cost-exceeds-savings","api_url":"https://likelier.app/api/fears/funeral-cost-exceeds-savings.json"},{"slug":"identity-theft","question":"What are the odds of being a victim of identity theft?","quick_answer":"A US adult faces roughly a 60% chance of identity theft over a lifetime (about 9% in any given year), with a wide band. This is one of the rare risks where perception matches reality — most people already sense the lifetime odds are around 1 in 2.\n","category":"other","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Identity theft is one of the few items on Gallup's annual crime-worry list where the perceived risk and the measured risk are both genuinely high. In Gallup's October 2024 crime poll, 69% of US adults said they worry frequently or occasionally about being the victim of identity theft, making it the single most-worried-about crime on the survey for the better part of two decades. Unlike most Likelier entries, the gap between perceived and actual here is narrow.\n","rough_estimate":"~1 in 2 lifetime feels about right to most respondents","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"native":{"display":"~9% per year (US residents 16+)","numerator":239,"denominator":2660,"unit":"per year","population":"US residents age 16 or older"},"normalized":{"lifetime_us_adult":0.6,"display":"~60% lifetime (US adult), wide band","log_value":-0.22,"assumptions":"BJS's National Crime Victimization Survey Identity Theft Supplement found that 9% of US residents age 16 or older (about 23.9 million people) experienced identity theft in the 12 months before the 2021 survey, and that 22% had experienced it at some point in their lifetime as of survey date. The 22% is a measured floor (cumulative to current age, averaged across all ages in the sample); the lifetime figure for a current young adult will be higher because they have more years of exposure ahead. Naively compounding the 9% annual hazard over a 59-year remaining adult life gives 1 − (1 − 0.09)^59 ≈ 99.6%, which is implausible: the same wallet, the same SSN, and the same household behaviour drive repeat events, so per-year incidents are not independent and the rate saturates well below 100%. The central 60% point estimate is a midpoint between the BJS measured lifetime floor (22%) and the upper bound implied by the annual hazard with strong saturation. The result is highly definition-dependent: see the body and caveats.\n","uncertainty":{"low":0.22,"high":0.85},"scope":"us_adult_lifetime"},"sources":[{"url":"https://bjs.ojp.gov/press-release/victims-identity-theft-2021","title":"Victims of Identity Theft, 2021","publisher":"US Bureau of Justice Statistics","source_type":"govt_report","statistic":"23.9 million US residents age 16+ (9%) experienced identity theft in the prior 12 months; 22% had experienced it in their lifetime","excerpt":"\"About 23.9 million U.S. residents age 16 or older (9% of the population) had experienced identity theft in the past 12 months. Almost 4% of people had their credit card misused, 3% had their bank account misused, and 2% experienced misuse of their email or social media account. Nearly 1% had their personal information misused for fraudulent purposes, and less than 1% had their personal information misused to open a new account. About 1 in 5 persons (22%) had experienced identity theft in their lifetime.\"\n","source_date":"2023-10-12","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413173229/https://bjs.ojp.gov/press-release/victims-identity-theft-2021","calculation_notes":"BJS's NCVS Identity Theft Supplement is the gold standard for population-level US identity-theft prevalence because it asks a representative sample of households directly about victimization, rather than counting reports filed with any one agency. Native rate: 9% per year for residents 16+. Lifetime measured floor: 22%. Upper bound from naive compounding (1 − 0.91^59 ≈ 99.6%) is implausible due to repeat victimization and saturation, so I take a midpoint. The \"9%\" headline includes everything from a single fraudulent credit-card charge that the issuer reverses (the bulk of cases) to new-account fraud requiring extended remediation (less than 1% of people). Restricting to \"actual out-of-pocket loss after recovery\" would cut the number by roughly an order of magnitude.\n","independence_note":"BJS NCVS is collected via household interviews and is methodologically independent from the FTC's Consumer Sentinel Network, which counts reports voluntarily filed by consumers. The two undercount in opposite directions: NCVS misses incidents respondents are unaware of, Sentinel misses incidents victims never report.\n"},{"url":"https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","title":"Consumer Sentinel Network Data Book 2024","publisher":"US Federal Trade Commission","source_type":"govt_report","statistic":"1.1+ million identity theft reports filed via IdentityTheft.gov in 2024; credit card identity theft was the largest category at 449,032 reports","excerpt":"\"During 2024, Sentinel received 6.5 million consumer reports … the top three Sentinel report categories were Credit Bureaus and Information Furnishers (21% of all reports), Identity Theft (18%), and Imposter Scams (13%). … In 2024, there were more than 1.1 million reports of identity theft received through the FTC's IdentityTheft.gov website. Credit Card tops the list of identity theft types reported in 2024.\"\n","source_date":"2025-03-10","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413165314/https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","calculation_notes":"The Sentinel figure (~1.1M reports/year) is roughly an order of magnitude smaller than the BJS-implied incidence (~24M/year), which is the expected ratio between \"events serious enough to file a federal report\" and \"any incident a survey respondent recalls when prompted.\" Both numbers are real; they answer different questions. Sentinel also feeds the lower-bound of our uncertainty band when paired with the BJS measured-lifetime floor.\n","independence_note":"Sentinel is fed by consumer-initiated reports and partner-agency intake, while BJS NCVS is a stratified household survey. Methodologically independent collection pipelines.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"69% of US adults worry frequently or occasionally about being the victim of identity theft (October 2024 wave)","excerpt":"\"Overall, Americans worry most about being the victim of identity theft (69%) and being tricked into providing financial information to scammers (53%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172538/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for the perceived-risk side only. Identity theft has consistently topped Gallup's crime-worry list since the question was first asked, almost always at or above two-thirds of respondents. Unlike most Likelier entries the perceived figure and the measured figure are within a factor of two of each other on a lifetime basis.\n","independence_note":"Gallup telephone polling, entirely separate from BJS NCVS household victimization sampling and FTC Sentinel complaint data. Used only for the perceived-risk axis — measures public worry, not incidence.\n"},{"url":"https://www.cifas.org.uk/newsroom/huge-surge-see-sim-swaps-hit-telco-and-mobile","title":"1,055% surge in unauthorised SIM swaps as mobile and telecoms sector hit hard by rising fraud","publisher":"Cifas","source_type":"reputable_reference","statistic":"UK SIM-swap fraud cases filed to the National Fraud Database surged 1,055% in 2024, to nearly 3,000 unauthorised SIM swaps","excerpt":"\"Cifas, the UK's leading fraud prevention service, has today raised the alarm on SIM swap fraud, with cases surging by 1,055% in 2024. … Nearly 3,000 cases of unauthorised SIM swaps – where criminals hijack a victim's mobile number by porting it to a new SIM without their consent – were filed to the National Fraud Database (NFD) in 2024.\"\n","source_date":"2025-05-07","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260711113023/https://www.cifas.org.uk/newsroom/huge-surge-see-sim-swaps-hit-telco-and-mobile","calculation_notes":"Cited only to document the SIM-swap / account-takeover sub-vector, not to revise the headline. The figure is a year-over-year change in absolute UK case counts (filed to one fraud database), not a US per-adult incidence rate, so it cannot be converted to a lifetime probability and is deliberately excluded from the normalized number. It is directionally consistent with the US picture: the FBI's IC3 2024 Internet Crime Report logged roughly $26M in reported SIM-swap losses across about 982 complaints — again an aggregate dollar/complaint figure with no per-adult denominator. Used to support the caveat on intercepted-SMS-2FA account takeover.\n","independence_note":"Cifas National Fraud Database is a UK industry consortium feed, methodologically independent from BJS NCVS, FTC Sentinel, and Gallup. UK jurisdiction; cited for vector existence and trend, not as a US incidence estimate.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Being murdered (lifetime, US adult average)","lifetime_us_adult":0.00348},{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"narrow definition (new-account, tax, or medical ID theft)","probability":0.22,"notes":"BJS NCVS measured floor: ~1 in 5 US adults experience serious ID theft in lifetime"},{"region":"broad definition (any unauthorized account use ever)","probability":0.85,"notes":"including one-off unauthorized card charges pushes lifetime prevalence close to universal"},{"region":"midpoint (serious ID theft requiring remediation)","probability":0.6,"notes":"point estimate used in headline; reflects AARP and FTC Sentinel patterns"}],"personal_factor_multipliers":[{"factor":"Age 30–49 (peak victimization window)","multiplier":1.8,"notes":"FTC Consumer Protection Data Spotlight 2023: adults aged 30–49 report the highest per-capita identity-theft victimization rate of any age group — approximately 1.8× the all-adult average — likely driven by peak credit activity, homeownership, and higher average balances making accounts more valuable to fraudsters."},{"factor":"Social Security number previously exposed in a data breach","multiplier":3,"notes":"Javelin Strategy & Research 2023 Identity Fraud Study: individuals whose SSN has been confirmed exposed in a data breach face approximately 3× elevated risk of new-account identity fraud compared with the general adult population, because SSN + date-of-birth combinations are the key credentials for opening fraudulent financial accounts."},{"factor":"No credit freeze on all three bureaus","multiplier":2.5,"notes":"FTC Consumer Advice and Javelin Strategy 2023: a security freeze on Equifax, Experian, and TransUnion is the single most effective control against new-account identity fraud; Javelin estimates that unfrozen credit files face roughly 2.5× the new-account fraud rate compared with accounts with active freezes in place across all three major bureaus."},{"factor":"Exposed paper mail (rural/unlocked mailbox)","multiplier":1.5,"notes":"USPIS (US Postal Inspection Service) annual report and FTC Consumer Advice: mail theft — including pre-approved credit offers, tax documents, and financial statements — is a documented identity-theft vector; individuals with exposed rural mailboxes or high-theft urban areas face an estimated 1.5× elevated rate of mail-sourced identity theft compared with those using locked or PO box mail delivery."}],"short_label":"Identity theft","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is not a death risk; it is a victimization risk, and the normalized \"lifetime_us_adult\" figure here is the probability that a US adult experiences identity theft at least once during their adult life, not the probability of dying from it. The number is unusually fragile to definition. Counting any unauthorized credit-card charge the issuer reverses gives ~9% per year and a measured 22% lifetime floor. Counting only events with non-recovered out-of-pocket loss cuts the rate by roughly an order of magnitude. Counting only new-account fraud (someone opens a credit line in your name) cuts it by another order of magnitude — BJS puts that subcategory at less than 1% per year. The 60% lifetime central estimate sits in the middle of a wide and definition- dependent band, which is why the uncertainty interval is intentionally large rather than reflecting only sampling error. Identity theft is also one of the few items on this site where individual action measurably moves the number: credit freezes, unique passwords, and two-factor authentication are documented to reduce new-account fraud in particular. A distinct and growing sub-vector is the SIM swap (or port-out), in which a criminal hijacks a victim's mobile number to intercept SMS two-factor codes and then loots linked bank or crypto accounts — a different failure mode from broad identity theft or passive data-breach exposure. Cifas recorded a 1,055% year-over-year jump in unauthorised UK SIM-swap cases in 2024, and the FBI's IC3 logged roughly $26M in US SIM-swap losses across about 982 complaints that year; new FCC SIM-swap and port-out authentication rules reached their compliance date in July 2024. These are absolute case-count and dollar-loss figures with no US per-adult denominator, so they do not move the 60% headline, but they are why an app-based authenticator or hardware key is more robust than SMS-delivered codes for accounts a SIM swap can drain.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A blank credit-card-shaped rectangle in muted grey on a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/identity-theft","api_url":"https://likelier.app/api/fears/identity-theft.json"},{"slug":"travel-disruption-war-disaster","question":"What are the odds of a trip being significantly disrupted by war, political unrest, or natural disaster?","quick_answer":"About 1-2% of international trips are significantly disrupted by war, unrest, or natural disaster; for an active traveler taking two trips a year for 30 years, that compounds to roughly a 60% lifetime chance. The number only applies to people who actually travel internationally.\n","category":"transport","tags":["travel"],"no_reliable_estimate":false,"perceived":{"description":"The news cycle delivers vivid images of travelers stranded at airports, volcanic ash clouds drifting over Europe, and rockets over Middle Eastern capitals just days before someone's booked departure. Because dramatic disruptions make headlines and routine departures do not, the typical traveler's prior is assembled almost entirely from a handful of memorable episodes — the 2010 Eyjafjallajökull ash cloud, COVID border closures, the Russia-Ukraine airspace shutdowns, Iran-adjacent escalations. Many people quietly overestimate the risk for the destinations they actually visit (Western Europe, Mexico, the Caribbean in good seasons) while underestimating it for genuinely unstable corridors they nevertheless book because a package deal was cheap. No rigorous survey isolates this specific fear, so the perceived side is editorial intuition rather than polled data.\n","kind":"intuition"},"native":{"display":"~1–2% of international trips experience significant external-cause disruption (war, disaster, mass airspace closure)","numerator":15,"denominator":1000,"unit":"significantly disrupted trips per 1,000 international trips","population":"international travelers broadly"},"normalized":{"lifetime_us_adult":0.6,"display":"~60% cumulative lifetime chance for an active international traveler (2 trips/year, 30 years)","log_value":-0.22,"assumptions":"The per-trip rate is estimated at ~1.5% (15 per 1,000) for \"significant external-cause disruption\" — defined as a trip cancelled, substantially rerouted, or cut short due to armed conflict, political unrest, natural disaster, or mass airspace closure, where the cause is outside the traveler's household. This is distinct from personal-cause cancellations (illness, work, family emergency), which dominate the overall cancellation data. The 1.5% figure is derived by working backward from travel insurance industry data: roughly 16% of policyholders file any claim; ~40% of those claims are trip cancel/interrupt; approximately one quarter of cancel/interrupt claims involve an external non-medical cause. That chain yields ~1.6% of insured trips, rounded down to 1.5% to account for survivorship bias (uninsured travelers don't generate claim data, and frequent travelers self-select toward lower-risk destinations). The uncertainty band is deliberately wide because the figure is highly window-sensitive — a single Eyjafjallajökull-type event, a regional airspace closure, or a hurricane season can shift the annual rate by a factor of three or more. Normalization: 2 international trips/year × 30 years of active travel = 60 trips. Lifetime = 1 − (1 − 0.015)^60 ≈ 0.60. For less frequent travelers (1 trip/year, 20 years = 20 trips): 1 − (0.985)^20 ≈ 0.26. Scope is activity_specific_lifetime — the number is only meaningful for people who travel internationally; someone who never boards an international flight has zero exposure.\n","uncertainty":{"low":0.26,"high":0.91},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://travel.state.gov/en/international-travel/travel-advisories.html","title":"Travel Advisories","publisher":"U.S. Department of State, Bureau of Consular Affairs","source_type":"govt_report","statistic":"Approximately 40 countries at Level 3 (Reconsider Travel) or Level 4 (Do Not Travel) as of 2025–2026, out of ~195 total destinations assessed","excerpt":"\"Level 3 – Reconsider Travel: Reconsider your travel to the destination due to serious risks to safety and security. Level 4 – Do Not Travel: Do not travel to the destination. This is the highest advisory level due to life-threatening risks.\"\n","source_date":"2026-05-01","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260502145445/https://travel.state.gov/en/international-travel/travel-advisories.html","calculation_notes":"State Dept advisory counts are used to establish baseline: roughly 20% of assessed destinations carry elevated (Level 3/4) advisories at any given moment. However, the US outbound travel market is heavily concentrated at Level 1–2 destinations: the top 5 overseas destinations in 2024 were the UK, Italy, France, Dominican Republic, and Spain, all of which hold Level 1 or Level 2 advisories. This means the average realized per-trip disruption risk for US travelers as a population is substantially below what the raw count of Level 3/4 countries would suggest. The advisory count feeds the personal_factor_multipliers rather than the headline rate.\n","independence_note":"Primary U.S. government source; methodologically independent of the travel insurance industry data used in source 2.\n"},{"url":"https://www.prnewswire.com/news-releases/travel-insurance-claims-paid-out-6x-policy-premium-in-2023-302096299.html","title":"Travel Insurance Claims Paid Out 6X Policy Premium in 2023","publisher":"Squaremouth (via PR Newswire)","source_type":"reputable_reference","statistic":"Nearly half of all paid travel insurance claims in 2023 were for trips canceled or cut short; Trip Cancellation was the most commonly paid benefit at 25% of all paid claims with average payout of $4,854","excerpt":"\"Nearly half of all paid claims in 2023 were for trips that were canceled outright or cut short. The most commonly paid out claim in 2023 was Trip Cancellation at 25%, with an average payment amount near $5,000. With an average pay out of $1,900 per claim, the average reimbursement exceeded the average travel insurance policy cost by 6-times.\"\n","source_date":"2024-02-20","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20250721154514/https://www.prnewswire.com/news-releases/travel-insurance-claims-paid-out-6x-policy-premium-in-2023-302096299.html","calculation_notes":"Squaremouth's 2023 data is used to anchor the trip-cancel/interrupt share of total claims (~40–46% of paid claims are cancel/interrupt combined). Paired with the industry-reported figure that roughly 16% of travel insurance policyholders file any claim (UStiA Travel Protection Market Study, cited across multiple trade sources), this gives approximately 6–7% of insured trips experiencing a cancel/interrupt claim from any cause. The subset attributable to external causes — war, disaster, civil unrest, mass airspace closure — is estimated at roughly one quarter of all cancel/interrupt claims based on the fact that illness and family emergencies dominate the covered-reason landscape (roughly 75% of Squaremouth policies cover medical cancellation). This chain yields ~1.5–2% of insured trips disrupted by external causes, which we adopt as the central per-trip rate. Squaremouth is the largest US travel insurance comparison marketplace and its annual press releases are the clearest publicly available window into claims composition, though they report characteristics of paid claims rather than claims-per-policy rates.\n","independence_note":"Squaremouth data is based on policies sold through its marketplace and claims processed through its affiliated insurers — a different pipeline from the State Dept advisory system. The two sources are methodologically independent.\n"},{"url":"https://www.nerdwallet.com/travel/learn/cancel-for-any-reason-cfar-travel-insurance-explained","title":"How Cancel For Any Reason Travel Insurance Works","publisher":"NerdWallet","source_type":"reputable_reference","statistic":"Cancel For Any Reason (CFAR) coverage typically adds about 40-50% to the total travel insurance premium","excerpt":"\"Cancel For Any Reason coverage will likely add about 40%-50% to your total travel insurance bill.\"\n","source_date":"2026-06-16","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260620002011/https://www.nerdwallet.com/travel/learn/cancel-for-any-reason-cfar-travel-insurance-explained","calculation_notes":"Supports the body-text figure that a CFAR rider adds roughly 40-50% to the base travel insurance premium — a separate cost question from the per-trip disruption-rate calculation, which relies on sources 1 and 2 above.\n","independence_note":"NerdWallet aggregates travel-insurance pricing data independently of both the State Department advisory system and Squaremouth's claims-composition reporting; it supplies the premium-cost figure that the other two sources do not address.\n"}],"comparison_anchors":[{"label":"Trip cancelled due to personal illness (any international trip, lifetime active traveler)","lifetime_us_adult":0.85},{"label":"Trip disrupted by significant flight delay or mechanical issue (lifetime active traveler)","lifetime_us_adult":0.92},{"label":"Being caught in a country under a sudden airspace closure (lifetime active traveler)","lifetime_us_adult":0.12}],"personal_factor_multipliers":[{"factor":"Destination with active Level 3 (Reconsider Travel) advisory","multiplier":8,"notes":"Trips to Level 3 destinations carry materially elevated disruption risk from the specific conditions that triggered the advisory — crime, political instability, regional conflict. An 8x multiplier moves the per-trip rate to roughly 12%, which is consistent with the frequency of significant incidents in countries like Colombia, Egypt, and Jamaica where advisory-triggering events are recurrent but not constant.\n"},{"factor":"Destination with active Level 4 (Do Not Travel) advisory","multiplier":40,"notes":"Level 4 destinations (active conflict zones, collapsed states) represent a different risk category entirely. A 40x multiplier on the 1.5% baseline yields ~60% per trip, consistent with the near-certainty of disruption for countries in active armed conflict. Most standard travel insurance explicitly excludes war zones; trip disruption in these destinations is not a tail risk but the expected outcome.\n"},{"factor":"Seismically or volcanically active corridor (Ring of Fire, Iceland, East African Rift)","multiplier":4,"notes":"Volcanic eruptions (Iceland 2010: ~10 million passengers stranded, 107,000 flights cancelled over 8 days, roughly 48% of total European air traffic during that window) and major seismic events can close airports and suspend service for days to weeks. Most Ring of Fire destinations are politically stable, so this multiplier is additive to the baseline rather than correlated with political unrest.\n"},{"factor":"Hurricane-prone Caribbean or Gulf destination booked October–November","multiplier":5,"notes":"Late-season Caribbean travel coincides with the peak of Atlantic hurricane season and combines elevated storm probability with constrained rerouting options (many regional airports have limited alternate-carrier capacity), though not every storm disrupts air service.\n"},{"factor":"Trip to Western Europe, Japan, Canada, Australia, or New Zealand","multiplier":0.15,"notes":"Politically stable, well-resourced aviation infrastructure with extensive rerouting capacity. The top five US overseas destinations in 2024 (UK, Italy, France, Dominican Republic, Spain) collectively account for a large share of outbound US travel and all carry Level 1 or 2 advisories. Disruption risk at these destinations is well below the 1.5% baseline.\n"}],"short_label":"Trip disruption: war or disaster","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The ~1.5% per-trip rate is a weighted average across all international destinations and all causes; it conceals enormous destination variance. A traveler who spends 30 years flying between Level 1 destinations (Paris, Tokyo, Toronto) faces a rate closer to 0.2% per trip; someone who regularly visits Level 3 destinations faces 10x that. The per-trip rate is also window-sensitive: in years with a major volcanic eruption (2010), a pandemic (2020), or a regional air war (2022 Russian airspace closures), the annual disruption rate spikes far above the long-run average before reverting. The 60% lifetime figure is a useful planning anchor for a frequent international traveler but should not be interpreted as \"60% chance my next trip is disrupted\" — the per-trip rate remains ~1.5%. Finally, most standard travel insurance policies explicitly exclude war and civil unrest from covered cancellation reasons; a \"Cancel for Any Reason\" (CFAR) rider is needed to recover costs for those specific disruptions.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":4,"avg":4,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"Airline departure board showing cancelled flights, with a distant volcano silhouette visible through a terminal window."},"canonical_url":"https://likelier.app/travel-disruption-war-disaster","api_url":"https://likelier.app/api/fears/travel-disruption-war-disaster.json"},{"slug":"widowhood-experience","question":"How likely am I to outlive my spouse and experience widowhood?","quick_answer":"About 3 in 5 (60%) currently-married women who reach age 65 in high-income countries will be widowed in their remaining years, higher than most people in mid-life marriages expect. For men at 65 the figure is roughly 20 to 25%.\n","category":"health","tags":["relationships","elder-care","mental-health"],"no_reliable_estimate":false,"perceived":{"description":"Widowhood is widely understood to be common among older women, but the probability tends to be underestimated for those currently in mid-life marriages. When asked to estimate whether they will outlive a spouse, people often respond in terms of relative health rather than actuarial probability. The statistical reality — that the majority of women who reach 65 while married will at some point be widowed — is not a message that marriage-focused social narratives tend to foreground. Financial planning for widowhood is correspondingly neglected, particularly among women who handled fewer financial responsibilities during the marriage.\n","kind":"intuition"},"native":{"display":"60 in 100 currently-married women aged 65 will experience widowhood (high-income countries)","numerator":60,"denominator":100,"unit":"lifetime probability from age 65","population":"currently-married women aged 65+ in high-income countries (UN DESA 2019, multi-country)"},"normalized":{"lifetime_us_adult":0.6,"display":"about 3 in 5 currently-married women aged 65 will be widowed in remaining lifetime","log_value":-0.22,"assumptions":"UN DESA World Marriage Data 2019 covers 232 countries. Among women aged 65+, currently-widowed shares range from approximately 35% in Northern Europe to 70%+ in parts of sub-Saharan Africa and South Asia. These cross-sectional figures reflect both survival patterns and remarriage rates. In high-income countries, the lifetime probability from age 65 is approximately 55–65%; the headline figure (0.60) is the midpoint for women in high-income country contexts. For men, the comparable probability is roughly 20–35% (lower life expectancy gap and higher male remarriage rates). The excess mortality hazard following bereavement (mean HR 1.23; men 1.27, women 1.15; largest in the first year) is documented in a 22-country meta-analysis (Shor 2012) but is not incorporated as a quantifiable lifetime probability; that framing would require no_reliable_estimate. Low bound (0.35): Northern European women or populations with small sex-mortality gap. High bound (0.75): Sub-Saharan African or South Asian women, or populations with limited male remarriage after widowhood.\n","uncertainty":{"low":0.35,"high":0.75},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.un.org/development/desa/pd/data/world-marriage-data","title":"World Marriage Data 2019","publisher":"United Nations Department of Economic and Social Affairs (UN DESA)","source_type":"govt_report","statistic":"Provides the distribution of the population by marital status (single, married, widowed, divorced) by age and sex for 232 countries — the underlying dataset from which cross-sectional widowhood shares at ages 65+ are computed (women 65+ widowed shares run from roughly one-third in Northern Europe to over 70% in parts of sub-Saharan Africa and South Asia)","excerpt":"\"The World Marriage Data 2019 provides a comparable and up-to-date set of data on the marital status of the population by age and sex for 232 countries or areas of the world. The database contains data on four key indicators: the distribution of the population by marital status, the proportion currently married, the proportion ever married, and the singulate mean age at marriage (SMAM).\"\n","source_date":"2019-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260307224433/https://www.un.org/development/desa/pd/data/world-marriage-data","calculation_notes":"UN DESA World Marriage Data 2019 is the source dataset for cross-sectional distribution of the population by marital status (including the widowed share) by age and sex across 232 countries. The landing page describes the dataset; the regional widowhood shares cited in this entry are computed from the underlying by-country marital-status tables, not quoted from a narrative sentence on the page. Cross-sectional widowed share at 65+ is used as a proxy for lifetime probability at that age. Native rate: 60/100 women 65+ (high-income country average, between the Northern European ~35% and Sub-Saharan African 70%+ extremes). Normalized: 0.60 (no unit conversion needed; figure is already a probability from age 65).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22427278/","title":"Widowhood and Mortality: A Meta-Analysis and Meta-Regression","publisher":"Demography","source_type":"peer_reviewed","statistic":"Bereaved spouses show a mean excess-mortality hazard ratio of 1.23 (men 1.27, women 1.15), with the excess concentrated early after loss (58% at 6 months, 34% at 1 year), pooled across 124 publications from 22 countries","excerpt":"\"We extracted 1381 mortality risk estimates from 124 publications, providing data on more than 500 million persons. Compared to married people, widowers had a mean hazard ratio (HR) of 1.23 (95% confidence interval [CI], 1.19–1.28) among HRs adjusted for age and additional covariates and a high subjective quality score. The mean HR was higher for men (HR, 1.27; 95% CI, 1.19–1.35) than for women (HR, 1.15; 95% CI: 1.08–1.22).\"\n","source_date":"2012-06-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20251127181552/https://pubmed.ncbi.nlm.nih.gov/22427278/","calculation_notes":"Shor et al. 2012 (Demography) meta-analysis of 1381 mortality risk estimates from 124 publications covering 22 countries. Provides the mean excess-mortality hazard ratio (1.23 overall; 1.27 men, 1.15 women) cited in the assumptions as context, not as the primary probability estimate. The excess is largest soon after loss (HR ~1.58 at 6 months, ~1.34 at 1 year) and attenuates with time. This hazard ratio is not converted to a lifetime probability because the denominator is heterogeneous; the entry's headline rate (probability of experiencing widowhood, 0.60) is separate from the bereavement-mortality effect.\n"},{"url":"https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1612813/full","title":"Research trends and hotspots in the mental health of widowed older adults: a bibliometric analysis","publisher":"Frontiers in Psychiatry","source_type":"peer_reviewed","statistic":"Widowhood is linked to social isolation, anxiety, and depression in older adults; a pooled meta-analysis reports a combined depression prevalence of 40.6% among widowed older adults","excerpt":"\"Research has linked widowhood to social isolation, negative emotions—including anxiety and depression—as well as impaired sleep quality in older adults. Depression is a prevalent psychological condition among widowed older adults. A meta-analysis reported a combined prevalence of 40.6%. Spousal loss significantly alters social networks, heightening the risk of social isolation. This serves as a risk factor for the development of depressive symptoms.\"\n","source_date":"2025-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20250810205227/https://www.frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1612813/full","calculation_notes":"Lin et al. 2025 (Frontiers in Psychiatry) reviews the mental-health burden of widowhood, documenting depression (combined prevalence 40.6% in a cited meta-analysis) and social isolation among widowed older adults. Used to support the characterisation of widowhood as moderate harm (lasting psychological and social consequences beyond the short-term mortality hazard). Does not alter the headline probability calculation.\n"},{"url":"https://www.consumerfinance.gov/data-research/research-reports/data-spotlight-financial-challenges-faced-by-recently-widowed-older-adults/full-report/","title":"Data Spotlight: Financial challenges faced by recently widowed older adults","publisher":"Consumer Financial Protection Bureau (CFPB)","source_type":"govt_report","statistic":"Newly widowed older adults (60+) see needs-adjusted household income fall by an average of 11%; 16% of newly widowed older adults have income below the federal poverty level vs. 10% of all adults 60+","excerpt":"\"A large number of studies have documented the impact of widowhood on the income of surviving spouses, with newly widowed individuals experiencing an average income drop of 11 percent even after accounting for household size and reduced needs. [...] 16 percent of new surviving spouses have income below the federal poverty level (approximately $12,300 for a one-person household) compared to 10 percent of all older adults aged 60 and older.\"\n","source_date":"2022-05-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260514200852/https://www.consumerfinance.gov/data-research/research-reports/data-spotlight-financial-challenges-faced-by-recently-widowed-older-adults/full-report/","calculation_notes":"CFPB Data Spotlight (based on 2019 American Community Survey data covering roughly 1.2 million adults 60+ who lost a spouse in 2019) is the source for the body's financial-impact paragraph: the 11% needs-adjusted income decline and the elevated poverty rate (16% vs. 10%) among newly widowed older adults relative to same-age peers. Descriptive context only, not used in the headline probability calculation (which relies solely on UN DESA marital-status data).\n","independence_note":"Independent of the UN DESA marital-status data and the Shor 2012 / Lin 2025 health literature; draws on Census ACS microdata and CFPB's own analysis of surviving-spouse finances.\n"}],"comparison_anchors":[{"label":"Widowhood (married men, 65+, high-income)","lifetime_us_adult":0.25},{"label":"Major depressive episode (lifetime, adults)","lifetime_us_adult":0.2},{"label":"Divorce (lifetime, US adults)","lifetime_us_adult":0.4}],"short_label":"Widowhood probability","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","caveats":"The 60% headline reflects currently-married women in high-income countries reaching age 65; it is not a probability for younger adults or for the general population. The sex gap is large and consistent: men at 65 have a roughly 20–25% lifetime probability of being widowed, compared to 55–65% for women. Regional variation is wide — the difference between Northern Europe and sub-Saharan Africa is approximately 2-fold — driven by marriage-age differences, sex-mortality gaps, and remarriage rates. The \"broken-heart\" excess mortality after bereavement is real (mean HR ~1.23, larger soon after loss) but small relative to baseline mortality risk at 65+ and is not expressed as a separate probability in this entry. The more practically significant consequence of widowhood in most high-income countries is financial: pension structures, housing equity, and Social Security survivor benefits mean that the financial impact of spousal death varies enormously by prior planning.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":3,"d8":4,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a single empty chair at a small table, muted warm tones."},"canonical_url":"https://likelier.app/widowhood-experience","api_url":"https://likelier.app/api/fears/widowhood-experience.json"},{"slug":"sexual-harassment-lifetime","question":"What are the odds of experiencing sexual harassment?","quick_answer":"Roughly 3 in 5 US adults, about 81% of women and 43% of men, report some form of sexual harassment in their lifetime. That far exceeds the intuition: only around 1 in 3 women see it as likely for themselves.\n","category":"crime","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Gallup's 2025 crime-worry poll finds 21% of US adults worry frequently or occasionally about being sexually assaulted — a figure that masks a stark gender split: 38% of women versus 4% of men, a 34-point gap. Yet the question asks about assault, not the broader category of harassment, so the worry metric likely understates concern about the wider phenomenon captured by prevalence surveys.\n","rough_estimate":"~1 in 3 women perceive it as likely for themselves","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"native":{"display":"81% of women and 43% of men report some form in their lifetime","numerator":81,"denominator":100,"unit":"lifetime prevalence","population":"US adults (SSH/GfK nationally representative survey, 2018)"},"normalized":{"lifetime_us_adult":0.62,"display":"~3 in 5 US adults (population-weighted)","log_value":-0.21,"assumptions":"The Stop Street Harassment / GfK 2018 nationally representative survey of 2,000 US adults found 81% of women and 43% of men experienced some form of sexual harassment or assault in their lifetime. Using Census sex distribution (~51% female, ~49% male): (0.81 × 0.51) + (0.43 × 0.49) ≈ 0.413 + 0.211 ≈ 0.62. This is a population-weighted lifetime prevalence, not a per-year rate compounded forward. The CDC's NISVS 2023/2024 data reports lower figures for specific subcategories (e.g., 30.4% of women for verbal workplace harassment), reflecting narrower definitions. The SSH figure uses the broadest definition — verbal harassment, unwanted touching, cyber harassment, and assault combined — and is the most-cited headline number. The 0.62 population-average is conservative in the sense that it weights equally across sexes; the lived experience is dramatically skewed.\n","uncertainty":{"low":0.45,"high":0.72},"scope":"us_adult_lifetime"},"sources":[{"url":"https://stopstreetharassment.org/our-work/nationalstudy/2018-national-sexual-abuse-report/","title":"2018 Study on Sexual Harassment and Assault","publisher":"Stop Street Harassment / GfK","source_type":"primary_study","statistic":"81% of women and 43% of men experienced some form of sexual harassment and/or assault in their lifetime","excerpt":"\"Nationwide, 81 percent of women and 43 percent of men reported experiencing some form of sexual harassment and/or assault in their lifetime. Verbal sexual harassment was the most common form (77% of women and 34% of men). An alarming 51% of women and 17% of men said they were touched or groped in an unwelcome way.\"\n","source_date":"2018-02-21","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260427022500/https://stopstreetharassment.org/our-work/nationalstudy/2018-national-sexual-abuse-report/","calculation_notes":"SSH commissioned a 2,000-person nationally representative survey conducted by GfK (now Ipsos). The 81% (women) and 43% (men) figures cover all forms of sexual harassment and assault combined: verbal harassment, unwanted touching/groping, cyber harassment, being followed, genital flashing, and sexual assault. Population-weighted average: (0.81 × 0.51) + (0.43 × 0.49) ≈ 0.62. Pro bono data analysis by UC San Diego Center on Gender Equity and Health.\n"},{"url":"https://www.cdc.gov/nisvs/media/pdfs/sexualviolence-brief.pdf","title":"National Intimate Partner and Sexual Violence Survey: 2023/2024 Sexual Violence Data Brief","publisher":"CDC, National Center for Injury Prevention and Control","source_type":"govt_report","statistic":"30.4% of women experienced verbal sexual harassment in the workplace; 11.3% of men; 29.5% of women in public places","excerpt":"\"Approximately 1 in 3 women in the U.S. experienced verbal sexual harassment in the workplace (30.4%) or public place (29.5%), and more than 1 in 4 women experienced technology-facilitated sexual violence in their lifetimes (28.2%). One in 9 men (11.3%) experienced verbal sexual harassment in the workplace.\"\n","source_date":"2025-12-01","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260429165643/https://www.cdc.gov/nisvs/media/pdfs/sexualviolence-brief.pdf","calculation_notes":"CDC NISVS 2023/2024 uses narrower subcategories than the SSH survey. The workplace verbal harassment figure (30.4% women) is not directly comparable to the SSH 81% because SSH aggregates all forms and all settings. The NISVS figure serves as a conservative lower bound for workplace-specific verbal harassment and confirms the order of magnitude. Nearly half of women experienced contact sexual violence in their lifetimes per NISVS.\n","independence_note":"NISVS is a CDC random-digit-dial telephone survey, methodologically independent of the SSH/GfK online panel survey. The two use different sampling frames, question wording, and definitions of harassment.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"21% of US adults worry frequently or occasionally about being sexually assaulted (2025); 38% of women vs 4% of men","excerpt":"\"Fewer Americans say they worry about crimes, such as having a car stolen (39%) or their home burglarized (34%), being a victim of a hate crime (30%), or getting mugged (29%), attacked while driving (27%), murdered (22%) or sexually assaulted (21%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260424191206/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for perceived-risk axis only. The 21% figure is the population-level share reporting frequent-or-occasional worry about sexual assault. Women are 34 percentage points more likely to worry than men (38% vs 4%). This is worry about assault specifically, not the broader harassment category.\n","independence_note":"Gallup telephone survey, independent of both SSH/GfK and CDC NISVS. Measures worry, not prevalence.\n"}],"comparison_anchors":[{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.39},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Being murdered (lifetime, US adult)","lifetime_us_adult":0.00348}],"personal_factor_multipliers":[{"factor":"Women","multiplier":1.31,"notes":"81% lifetime prevalence for women vs 62% population average"},{"factor":"Men","multiplier":0.69,"notes":"43% lifetime prevalence for men vs 62% population average"},{"factor":"Women under 35","multiplier":1.4,"notes":"Younger women report higher rates across all subcategories in both SSH and NISVS data"},{"factor":"LGBTQ+ individuals","multiplier":1.5,"notes":"CDC NISVS consistently finds elevated sexual violence rates among sexual minority populations"}],"short_label":"Sexual harassment (lifetime)","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry uses the broadest available definition of sexual harassment — verbal harassment, unwanted touching, cyber harassment, being followed, and assault combined. Narrower definitions (e.g., CDC NISVS workplace verbal harassment at 30% for women) yield substantially lower figures. The SSH survey was conducted online via GfK's probability-based panel, which may produce higher disclosure rates than telephone surveys due to reduced social desirability bias. The 81% figure for women is a lifetime cumulative prevalence, not an annual rate, and includes experiences ranging from a single instance of street catcalling to repeated assault. The gender disparity is among the largest on this site: women's lifetime prevalence is nearly double that of men. The population- weighted 62% figure is arithmetically correct but obscures the profoundly different lived experiences of women and men.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A pair of speech bubbles, one muted and one bright, on a plain background, flat vector illustration."},"canonical_url":"https://likelier.app/sexual-harassment-lifetime","api_url":"https://likelier.app/api/fears/sexual-harassment-lifetime.json"},{"slug":"undercooked-meat-fish-eggs","question":"What are the odds of getting food poisoning from undercooked meat, fish, or eggs?","quick_answer":"Undercooking-linked foodborne illness strikes about 1 in 57 US residents a year — roughly a 2 in 3 chance over a lifetime. Most people overestimate the danger of rare steak and sushi while underestimating the risk from undercooked chicken.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Public intuition about undercooking runs on vibes rather than microbiology. Rare steak triggers alarm despite being nearly risk-free for intact cuts; sushi provokes theatrical hand-wringing in people who cheerfully eat runny eggs; and undercooked chicken — the one item that genuinely deserves caution — is routinely served pink-at-the-bone by cooks who assume the color is cosmetic. No cross-national survey isolates the fear of undercooking cleanly from food-safety anxiety in general, so the perceived side here is editorial synthesis rather than polled data.\n","rough_estimate":"Most adults dramatically overestimate the risk from rare steak and sushi while underestimating the risk from pink chicken","kind":"intuition"},"native":{"display":"~1 in 57 per year (US residents, undercooking-linked foodborne illness)","numerator":5760000,"denominator":330000000,"unit":"per year","population":"US residents, all ages, foodborne illness where inadequate cooking temperature was a contributing factor"},"normalized":{"lifetime_us_adult":0.645,"display":"~2 in 3 lifetime (US adult)","log_value":-0.19,"assumptions":"Starts from CDC's ~48 million domestically acquired foodborne illnesses per year (Scallan et al. 2011). The CDC NORS MMWR analysis for 2014-2022 finds that \"inadequate time and temperature during initial cooking/thermal processing\" contributed to 9.6-12.1% of outbreaks across the three reporting periods, consistently ranking among the top five contributing factors. We take 12% as the central estimate of the share of US foodborne illness where undercooking is a contributing factor — deliberately excluding the overlapping temperature-abuse categories (holding, cooling) covered in the food-left-unrefrigerated entry. 12% x 48 million = ~5.76 million cases per year, or about 1.75% of the US population per year (~1 in 57). Compounded over 59 years of adult life: 1 - (1 - 0.0175)^59 = 0.645, or about 2 in 3. The uncertainty band runs from 8% contribution (lifetime ~0.47) to 18% (lifetime ~0.79), spanning defensible readings of the NORS data and the underlying Scallan illness total. The vast majority of these are mild gastroenteritis episodes, not hospitalizations or deaths.\n","uncertainty":{"low":0.47,"high":0.79},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","title":"Contributing Factors of Foodborne Illness Outbreaks — National Outbreak Reporting System, United States, 2014–2022","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR) Surveillance Summaries","source_type":"govt_report","statistic":"Among 2,677 foodborne outbreaks 2014-2022, 'inadequate time and temperature during initial cooking/thermal processing' contributed to 12.1%, 9.6%, and 12.1% across the three reporting periods; consistently a top-five contributing factor","excerpt":"\"Inadequate time and temperature control during initial cooking of food was among the top five contributing factors during all three periods (23.8 percent, 20.4 percent, and 20.9 percent, respectively) [broader category]. More specifically, the proportion of outbreaks associated with inadequate time and temperature control during initial cooking/thermal processing of food decreased from the first (12.1 percent) to the second period (9.6 percent), and increased during the third period (12.1 percent).\"\n","source_date":"2025-03-13","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421185402/https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","calculation_notes":"The NORS MMWR report is the primary source for the 12% central estimate. The broader \"inadequate time and temperature during cooking\" category runs 20-24%, but this includes holding and display temperatures, which overlap with the food-left-unrefrigerated entry. The narrower \"initial cooking/thermal processing\" subcategory (9.6-12.1%) isolates the undercooking contribution. Applied to Scallan's 48 million illnesses/year: 0.12 x 48e6 = 5.76 million cases/year, or ~1.75% of the US population annually. Over 59 adult-remaining years: 1 - (1 - 0.0175)^59 = 0.645.\n","independence_note":"NORS is a CDC surveillance system drawing on state/local outbreak reporting. Methodologically linked to FoodNet and Scallan estimates but provides the contributing-factor breakdown those sources lack.\n"},{"url":"https://www.cdc.gov/food-safety/about/index.html","title":"About Food Safety","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC estimates 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths per year from foodborne illness in the US","excerpt":"\"CDC estimates that each year 48 million people get sick from a foodborne illness, 128,000 are hospitalized, and 3,000 die.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260511073542/https://www.cdc.gov/food-safety/about/index.html","calculation_notes":"Provides the denominator for the normalized figure. 48 million illnesses / 330 million US residents = ~14.5% of Americans per year experience a foodborne illness from any cause. Multiplied by the ~12% share attributable to undercooking (NORS) gives the ~1.75% per year figure compounded in the normalization.\n","independence_note":"Restates Scallan et al. 2011; not independent of the NORS contributing-factors analysis.\n"},{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/safe-temperature-chart","title":"Safe Minimum Internal Temperature Chart","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Poultry: 165°F; ground meats: 160°F; steaks/roasts/fish: 145°F with 3-minute rest; eggs: cook until yolk and white are firm","excerpt":"\"All Poultry (breasts, whole bird, legs, thighs, wings, ground poultry, giblets, and sausage): 165 °F. Ground meats: 160 °F. Beef, Pork, Veal & Lamb (Steaks, chops, roasts): 145 °F and allow to rest for at least 3 minutes. Fish: 145 °F. Eggs: Cook until yolk and white are firm.\"\n","source_date":"2024-05-29","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260420233534/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/safe-temperature-chart","calculation_notes":"Defines the regulatory baseline for \"undercooked.\" Anything below these temperatures is what the NORS coding of \"inadequate cooking/thermal processing\" references. The gap between steak (145°F) and poultry (165°F) reflects the surface-only vs throughout contamination difference that makes rare steak defensibly safe while rare chicken is not.\n","independence_note":"FSIS sets the regulatory standard; CDC surveillance measures compliance against it. Upstream of the NORS coding system.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4918174/","title":"Restaurant Cooking Trends and Increased Risk for Campylobacter Infection","publisher":"Emerging Infectious Diseases / CDC","source_type":"peer_reviewed","statistic":"Undercooked chicken meat is identified as the principal source of Campylobacter infection; an estimated 19%-52% of chicken livers served in commercial food establishments fail to reach a 70°C core temperature, with Campylobacter survival rates of 48%-98%","excerpt":"\"Studies of outbreaks and sporadic cases have identified the principal source of infection as undercooked chicken meat.... Campylobacter jejuni/coli can cause food poisoning when present in very small numbers.... an estimated 19%–52% of livers being served in commercial UK food establishments fail to reach a core temperature of 70°C and could have Campylobacter survival rates of 48%–98%.\"\n","source_date":"2016-06-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260503094952/https://pmc.ncbi.nlm.nih.gov/articles/PMC4918174/","calculation_notes":"This paper establishes that Campylobacter infection is overwhelmingly driven by undercooked poultry (the principal source) and documents that 19%-52% of chicken livers served in commercial food establishments fail to reach a safe core temperature. The scale figures used in the entry are drawn from their own primary sources, NOT from this paper: CDC estimates 1.5 million US Campylobacter illnesses per year (CDC \"About Campylobacter infection\"), and the IFSAC 2019 source attribution report attributes ~64.7% of non-dairy Campylobacter illnesses to chicken. 0.647 x 1.5 million = ~975,000 chicken-Campylobacter cases/year. This paper is used to support the chicken-specific row in the regional breakdown.\n","independence_note":"Peer-reviewed CDC study focused on restaurant cooking practices. Draws on FoodNet surveillance data but provides independent analysis of cooking-temperature compliance and risk.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11028959/","title":"Estimating the annual fraction of eggs contaminated with Salmonella enteritidis in the United States","publisher":"International Journal of Food Microbiology / Ebel & Schlosser","source_type":"peer_reviewed","statistic":"Approximately 1 in 20,000 US shell eggs is internally contaminated with Salmonella Enteritidis; ~2.3 million SE-contaminated eggs reach consumers annually","excerpt":"\"The expected value of this distribution is approximately one SE-affected egg in every 20,000 eggs annually produced, and the 90% certainty interval is between one SE-contaminated egg in 30,000 eggs, and one SE-contaminated egg in 12,000 eggs.\"\n","source_date":"2000-11-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20251114231143/https://pubmed.ncbi.nlm.nih.gov/11028959/","calculation_notes":"The 1-in-20,000 egg contamination rate is the basis for the runny-egg risk estimate. Americans consume ~280 eggs per capita per year, so a person eating ~50 runny or undercooked eggs per year faces ~50/20,000 = 0.0025 chance per year of encountering a contaminated egg, and only a fraction of those exposures produce clinical illness (dose-response depends on bacterial load, which increases with storage time and temperature). Used for the egg row in the regional breakdown.\n","independence_note":"Independent USDA-commissioned risk assessment. Pre-dates the 2010 Egg Safety Rule (21 CFR 118), which likely reduced contamination rates further, making this estimate conservative.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Any foodborne illness in a single year (US, 1 in 7)","lifetime_us_adult":0.145},{"label":"Temperature-abuse-linked foodborne illness (lifetime)","lifetime_us_adult":0.82}],"regional_breakdown":[{"region":"Sushi/sashimi (regulated US market)","probability":0.005,"notes":"Remarkably safe in regulated markets. US FDA requires commercial sushi-grade fish to be frozen at -20°C for 7 days (or -35°C for 15 hours), which kills parasites including Anisakis. Anisakiasis is not a reportable disease in the US, and estimated cases run under 10 per year nationally — in a country consuming billions of sushi servings. The main residual risk is Vibrio in warm-water species and Salmonella from cross-contamination, not from the raw fish itself. Lifetime probability estimated at ~1 in 200.\n"},{"region":"Rare/medium-rare beef steak (intact cut)","probability":0.003,"notes":"Pathogens on intact beef muscle are confined to the surface. Searing the exterior to well above 160°F kills surface bacteria even if the interior remains rare (120-130°F). USDA's 145°F guideline with 3-minute rest applies to roasts and steaks; the risk from a properly seared rare steak is vanishingly small. Ground beef is a different story — see below.\n"},{"region":"Undercooked ground beef (pink hamburger)","probability":0.04,"notes":"Grinding distributes surface bacteria (E. coli O157:H7, Salmonella) throughout the meat. USDA requires 160°F for ground beef — no rest period — because there is no safe interior. CDC links ~265,000 STEC illnesses per year to the US, with ground beef the dominant vehicle. A pink hamburger at a restaurant or backyard grill is the single most common undercooking scenario that actually produces illness.\n"},{"region":"Undercooked/pink chicken","probability":0.08,"notes":"Chicken is the paradigm case. Campylobacter contaminates up to 47% of retail raw chicken in USDA testing; the infective dose is as low as 500 organisms; and 19-52% of restaurant chicken livers fail to reach the required 165°F core temperature. CDC attributes ~975,000 Campylobacter illnesses per year to chicken alone. The lifetime probability of at least one illness from undercooked poultry for a regular chicken eater is high.\n"},{"region":"Runny/raw eggs (US, post-Egg Safety Rule)","probability":0.02,"notes":"About 1 in 20,000 US shell eggs carries Salmonella Enteritidis internally (Ebel & Schlosser 2000, likely lower post-2010 Egg Safety Rule). At ~280 eggs consumed per capita per year, a person who eats ~50 undercooked eggs per year has roughly a 0.25% annual chance of encountering a contaminated egg. Clinical illness given exposure depends on bacterial load, which rises with egg age and storage temperature. Pasteurized shell eggs eliminate this risk entirely.\n"}],"personal_factor_multipliers":[{"factor":"vegetarian or vegan","multiplier":0.05,"notes":"Eliminates meat and poultry exposure, which accounts for the large majority of undercooking-linked illness. Residual risk comes from eggs (if ovo-vegetarian) or cross-contamination.\n"},{"factor":"regular sushi consumer (2+ times/week)","multiplier":0.3,"notes":"Sushi in regulated US markets is extremely safe due to mandatory parasite-destruction freezing. Frequent consumption adds negligible absolute risk despite the raw-fish framing.\n"},{"factor":"frequent pink-chicken consumer","multiplier":3,"notes":"Campylobacter contamination rates in retail chicken are high, the infective dose is low, and even slight undercooking leaves viable organisms. The per-meal risk from pink chicken is the highest of any common undercooking scenario.\n"},{"factor":"immunocompromised (transplant, chemotherapy, advanced HIV)","multiplier":5,"notes":"Invasive Salmonella, Campylobacter bacteremia, and parasitic infections produce serious disease in immunocompromised hosts at doses that healthy adults clear asymptomatically.\n"},{"factor":"uses food thermometer consistently","multiplier":0.3,"notes":"USDA and CDC recommend using a food thermometer for all meat, poultry, and eggs. Consistent use largely eliminates the undercooking pathway for home-cooked food — which is where most exposure occurs.\n"}],"short_label":"Undercooked food","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline \"~1 in 57 per year / 2 in 3 lifetime\" is an aggregate of wildly heterogeneous scenarios. Rare steak from an intact cut is nearly risk-free; pink chicken is a genuine hazard; sushi in a regulated market is safer than most cooked dishes. Almost all counted episodes are mild gastroenteritis self-resolving in 24-72 hours — the fatal subset is covered in the separate food-poisoning-death entry and runs about 1 in 1,860 lifetime. The NORS contributing-factor coding allows multiple factors per outbreak, so \"inadequate cooking\" often co-occurs with cross-contamination or food-worker hygiene issues; the true share attributable solely to undercooking is genuinely uncertain within the 8-18% band used for the uncertainty calculation. The 1-in-20,000 egg contamination estimate dates from 2000 and likely overstates current risk, since the 2010 FDA Egg Safety Rule imposed testing, refrigeration, and diversion requirements that reduced Salmonella Enteritidis in the regulated egg supply.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A meat thermometer inserted into a cut of meat on a plain background, flat vector illustration."},"canonical_url":"https://likelier.app/undercooked-meat-fish-eggs","api_url":"https://likelier.app/api/fears/undercooked-meat-fish-eggs.json"},{"slug":"credit-card-fraud","question":"What are the odds of being a victim of credit card fraud?","quick_answer":"About 65% of US adult cardholders experience credit-card fraud at some point, with roughly 25% hit in a given year. The typical incident is a single unauthorized charge the issuer reverses, not a dramatic account takeover. Around 53% of adults say they worry about it.\n","category":"other","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"In Gallup's October 2024 crime poll, 53% of US adults said they worry frequently or occasionally about being tricked by a scammer into providing financial or personal information — the closest Gallup proxy for credit-card fraud specifically. Cardholders tend to overestimate the severity but underestimate the prevalence: most people picture a dramatic account takeover, when the typical incident is a single unauthorized online charge that the issuer reverses within days.\n","rough_estimate":"53% of US adults worry frequently or occasionally (Gallup 2024)","kind":"survey","survey_source":{"title":"Crime — Gallup Historical Trends","publisher":"Gallup","url":"https://news.gallup.com/poll/1603/crime.aspx","year":2024}},"native":{"display":"~25% of cardholders per year (2024, US)","numerator":62000000,"denominator":248000000,"unit":"per year","population":"US adults with at least one credit card"},"normalized":{"lifetime_us_adult":0.65,"display":"~65% lifetime (US adult cardholder)","log_value":-0.19,"assumptions":"Naive compounding of 25% annual victimization over 59 years (1 - 0.75^59) yields ~99.99%, which overstates because (a) victimization is not independent year-to-year — people who have been a victim are more vigilant, and card issuers flag compromised accounts — and (b) the 25% annual rate includes small unauthorized charges that many victims forget within a year. The 65% point estimate is anchored on Security.org's direct lifetime-measurement survey (63% reported ever experiencing unauthorized card activity), with a small upward adjustment to account for survey recall bias on small/old incidents. Uncertainty band [0.40, 0.85] brackets the range from \"stricter definition, memorable fraud only\" to \"loose definition including any unauthorized charge ever.\" Supporting context: a Security.org survey of US cardholders (January 2025) found that 25% of cardholders experienced unauthorized charges in the past year, translating to roughly 62 million Americans out of approximately 248 million adults with at least one credit card (about 80% of 310 million adults); 51% of victims report multiple incidents. The FTC Consumer Sentinel Network received 449,032 credit-card identity theft reports in 2024, representing the subset of incidents serious enough that victims filed a federal report. The Javelin Strategy 2025 study found 40 million total identity fraud victims in 2024 across all fraud types, providing a cross-check on the order of magnitude.\n","uncertainty":{"low":0.4,"high":0.85},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","title":"Consumer Sentinel Network Data Book 2024","publisher":"US Federal Trade Commission","source_type":"govt_report","statistic":"1.1+ million identity theft reports filed in 2024; credit card fraud was the largest identity theft category at 43.9% of all identity theft reports; total consumer fraud losses exceeded $12.5 billion","excerpt":"\"During 2024, Sentinel received 6.5 million consumer reports. In 2024, there were more than 1.1 million reports of identity theft received through the FTC's IdentityTheft.gov website. Credit Card tops the list of identity theft types reported in 2024.\"\n","source_date":"2025-03-10","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413165314/https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","calculation_notes":"The FTC Sentinel figure captures the subset of credit card fraud incidents where victims filed a formal federal report. At 43.9% of 1.1 million identity theft reports, credit card fraud accounts for roughly 483,000 filed reports per year. This is an order of magnitude below the survey-based prevalence because most unauthorized charges are caught by the issuer, reversed without loss, and never reported to any agency. The Sentinel data anchors the lower bound of the uncertainty range.\n","independence_note":"The FTC Consumer Sentinel Network collects consumer-initiated complaints and partner agency reports. It is methodologically independent from industry surveys like Security.org and Javelin, which use representative consumer panels.\n"},{"url":"https://www.security.org/digital-safety/credit-card-fraud-report/","title":"Credit Card Fraud Report 2025","publisher":"Security.org","source_type":"reputable_reference","statistic":"63% of US credit card holders have been victimized by fraud; 25% experienced fraudulent charges in the past year; an estimated 62 million Americans were affected","excerpt":"\"63% of U.S. credit card holders have been victimized by fraud. In January 2025, we conducted an online poll of 995 Americans who had at least one credit card. 92% of unauthorized transactions occurred without physical card theft.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260329021113/https://www.security.org/digital-safety/credit-card-fraud-report/","calculation_notes":"Security.org's survey provides the population-level prevalence estimate. The 63% measured lifetime victimization rate (among current cardholders surveyed) is close to our central 65% estimate and serves as a direct empirical check. This 63% measured lifetime prevalence is the primary anchor for the normalized estimate; the 65% point estimate adds a small adjustment for recall bias on minor incidents. The 25% annual rate is high but plausible given that 92% of fraud is card-not-present (online transactions), which have grown with e-commerce. The survey's 995-respondent sample gives a margin of error of roughly +/-3 percentage points at 95% confidence.\n","independence_note":"Security.org conducts its own consumer surveys using online panels. It is independent from both the FTC's complaint-based data and Javelin's proprietary consumer panel.\n"},{"url":"https://javelinstrategy.com/research/2025-identity-fraud-study-breaking-barriers-innovation","title":"2025 Identity Fraud Study: Breaking Barriers to Innovation","publisher":"Javelin Strategy & Research","source_type":"reputable_reference","statistic":"40 million Americans were victims of identity fraud in 2024; total fraud and scam losses reached $47 billion; account takeover fraud losses reached $15.6 billion","excerpt":"\"Fraud and scam losses reached $47 billion and affected 40 million people. Account takeover fraud resulting in $15.6 billion in losses in 2024, up from $12.7 billion in 2023.\"\n","source_date":"2025-03-25","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413165352/https://javelinstrategy.com/research/2025-identity-fraud-study-breaking-barriers-innovation","calculation_notes":"Javelin's 40 million figure covers all identity fraud types (not just credit card), providing an order-of-magnitude cross-check. Credit card fraud is the largest single category. The $15.6 billion in account takeover losses and $6.2 billion in new-account fraud losses together represent the financially serious end of the spectrum, while the bulk of the 62 million incidents from Security.org involve small unauthorized charges that are reversed by the issuer.\n","independence_note":"Javelin uses its own proprietary consumer panel and is methodologically independent from both the FTC and Security.org.\n"},{"url":"https://www.idtheftcenter.org/post/2025-annual-data-breach-report-record-number-compromises/","title":"2025 Annual Data Breach Report: Record Number of Data Compromises","publisher":"Identity Theft Resource Center","source_type":"reputable_reference","statistic":"2024 generated a record 1,367,117,021 victim notices driven by mega-breaches; Financial Services was the most-breached industry in 2025 with 739 compromises; total compromises hit a record 3,322 in 2025 (3,152 in 2024), a 79% jump over five years","excerpt":"\"The number of victim notices (278,827,933) decreased by 79 percentage points from 2024 (1,367,117,021). The massive decrease in victim notices is due to the lack of 'mega-breaches' in 2025, unlike 2024. The Financial Services industry was the most breached industry (739 compromises), followed by Healthcare (534 compromises).\"\n","source_date":"2026-01-29","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260520080329/https://www.idtheftcenter.org/post/2025-annual-data-breach-report-record-number-compromises/","calculation_notes":"ITRC does not measure individual credit-card-fraud incidence; it measures the upstream exposure that feeds card fraud — breached records containing card numbers and SSNs. The 2024 mega-breach wave (1.37 billion victim notices, the highest since state breach-notice laws were adopted) is the supply side of the card-not-present fraud that Security.org and the FTC count downstream. This source is context for the breach-exposure caveat and the refined credential-reuse factor, not an input to the 65% lifetime point estimate.\n","independence_note":"ITRC compiles publicly reported breach notifications and is methodologically independent from the FTC complaint data, Security.org panels, and Javelin's consumer panel.\n"},{"url":"https://www.deloitte.com/us/en/insights/industry/financial-services/deepfake-banking-fraud-risk-on-the-rise.html","title":"Generative AI is expected to magnify the risk of deepfakes and other fraud in banking","publisher":"Deloitte Center for Financial Services","source_type":"reputable_reference","statistic":"Generative AI could enable US fraud losses to reach $40 billion by 2027, up from $12.3 billion in 2023 — a 32% compound annual growth rate","excerpt":"\"gen AI could enable fraud losses to reach US$40 billion in the United States by 2027, from US$12.3 billion in 2023, a compound annual growth rate of 32%.\"\n","source_date":"2024-05-29","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260711010512/https://www.deloitte.com/us/en/insights/industry/financial-services/deepfake-banking-fraud-risk-on-the-rise.html","calculation_notes":"Deloitte's projection is a forward-looking loss-trend estimate across AI-enabled fraud types (including synthetic-identity fraud built from breached real data), not a card-fraud incidence rate. It supports the breach-driven exposure caveat by quantifying how cheaply AI now lets fraudsters weaponize the breached card and SSN data ITRC tracks. It is not an input to the 65% lifetime point estimate.\n","independence_note":"Deloitte's Center for Financial Services publishes independent industry analysis, methodologically distinct from the FTC, ITRC, Security.org, and Javelin sources above.\n"}],"comparison_anchors":[{"label":"Identity theft, any type (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Online-only card use vs chip-in-person","multiplier":3,"notes":"FTC Consumer Sentinel Network 2023: card-not-present fraud accounts for roughly 3× the incident rate of card-present fraud; Security.org 2025 found 92% of unauthorized transactions occurred without physical card theft, reflecting the CNP dominance in breach and skimmer-free e-commerce channels."},{"factor":"Age 20–29","multiplier":2,"notes":"FTC 2023 Consumer Protection Data Spotlight: adults aged 20–29 report the highest per-capita rate of credit card fraud victimization of any age group, approximately 2× the all-adult average, likely driven by higher rates of online purchasing and lower adoption of account-monitoring tools."},{"factor":"Reused passwords / no two-factor authentication","multiplier":2,"notes":"NIST SP 800-63B Digital Identity Guidelines and FTC Consumer Advice: credential-stuffing attacks — automated login attempts using credentials harvested from prior data breaches — are estimated to double the effective card-fraud exposure for individuals who reuse passwords across sites versus those using unique credentials with 2FA. The breach supply feeding these attacks reached a record in 2024: the Identity Theft Resource Center 2025 report counted 1,367,117,021 victim notices that year — driven by mega-breaches, with Financial Services the most-breached industry — meaning more harvested card numbers and SSNs are available to fuel credential-stuffing and synthetic-identity fraud."},{"factor":"No credit monitoring or account alerts","multiplier":1.5,"notes":"FTC Consumer Sentinel Network 2023: cardholders with no account-activity alerts or credit-monitoring services experience an estimated 1.5× higher detection lag, meaning fraudulent charges are more likely to go unnoticed across multiple billing cycles before being reported and reversed."}],"short_label":"Credit card fraud","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This is not a death risk or even necessarily a financial-loss risk. The vast majority of credit card fraud incidents involve unauthorized charges that the card issuer reverses under zero-liability policies, leaving the cardholder with no out-of-pocket loss. The FTC Sentinel data (roughly 483,000 credit-card identity theft reports per year) captures only incidents serious enough to prompt a federal complaint, which is an order of magnitude below the survey-based prevalence. If the question is \"will an unauthorized charge ever appear on my statement,\" the answer is very likely yes. If the question is \"will I suffer unrecovered financial loss from credit card fraud,\" the rate drops by roughly an order of magnitude. The 65% lifetime central estimate uses the broader definition. The number also depends on card usage patterns: heavy online shoppers face higher exposure than those who rarely use cards for e-commerce. Breach-driven card-data exposure is the upstream supply feeding this fraud, and it spiked in the 2024 mega-breach wave: the Identity Theft Resource Center's 2025 report recorded 1,367,117,021 victim notices for 2024 (the highest since state breach-notice laws were adopted), with Financial Services the most-breached industry. Separately, the Deloitte Center for Financial Services projects generative-AI-enabled fraud losses in the US could rise to $40 billion by 2027 from $12.3 billion in 2023 (a 32% annual growth rate), as AI makes synthetic-identity fraud — fabricated identities stitched from breached real data — cheaper to manufacture at scale. These are exposure and loss-trend signals, not changes to the 65% lifetime point estimate, which still rests on the Security.org and FTC measurements.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"groundedness-audit-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A credit card with a faint dotted outline where the number would be, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/credit-card-fraud","api_url":"https://likelier.app/api/fears/credit-card-fraud.json"},{"slug":"school-bullying","question":"What are the odds of a child being bullied at school?","quick_answer":"A child has roughly a 65% chance of being bullied at least once during grades 6-12, from about a 19% annual rate. Parental guesses scatter widely from 10% to 50%, so most underestimate the cumulative odds.\n","category":"other","tags":["kids"],"no_reliable_estimate":false,"perceived":{"description":"Parental estimates of school bullying vary more than almost any other childhood risk. Some parents treat it as a near-universal rite of passage and assume every child will face it; others believe their child's school is essentially bully-free. Media coverage oscillates between moral panic (every school is a warzone) and reassurance (anti-bullying programs have solved it). When pressed for a number, most adults guess somewhere between 10% and 50%, a range wide enough to be almost uninformative. The lack of a stable public anchor makes this an unusually noisy perception.\n","rough_estimate":"~1 in 3 to 1 in 5 children, intuitively","kind":"intuition"},"native":{"display":"~19% of US students ages 12-18 reported being bullied at school (NCES 2022)","numerator":19,"denominator":100,"unit":"per school year","population":"US students ages 12-18 enrolled in grades 6-12"},"normalized":{"lifetime_us_adult":0.65,"display":"~65% probability of being bullied at least once during grades 6-12","log_value":-0.19,"assumptions":"The NCES School Crime Supplement (2022) reports 19.2% of students ages 12-18 experienced bullying at school in the survey year. The CDC NCHS Data Brief 514 (October 2024), using National Health Interview Survey data from July 2021 to December 2023, found a higher figure: 34.0% of teenagers ages 12-17 were bullied in the past 12 months. Using the more conservative NCES annual rate of ~19% and compounding over 7 school years (grades 6-12): 1 - (1 - 0.19)^7 = 0.75. Using the YRBS high-school-only rate of ~19% for grades 9-12 and the NCES middle-school rate of ~26% for grades 6-8, a weighted compound gives ~0.72. However, bullying episodes are not fully independent year to year — some children are persistently targeted while others are never targeted. Adjusting downward for this clustering effect (correlation between years), a central estimate of ~0.65 is used. The NCHS figure of 34% per year would yield a much higher lifetime estimate (~0.95), but that survey uses a broader definition including verbal teasing that may not meet the NCES threshold.\n","uncertainty":{"low":0.45,"high":0.85},"scope":"subgroup_lifetime"},"sources":[{"url":"https://nces.ed.gov/pubs2024/2024109.pdf","title":"Student Reports of Bullying: Results From the 2022 School Crime Supplement to the National Crime Victimization Survey","publisher":"National Center for Education Statistics (NCES), U.S. Department of Education","source_type":"govt_report","statistic":"About 19% of students ages 12-18 reported being bullied at school during the 2021-22 school year; 26% in middle school, 16% in high school","excerpt":"\"In 2021-22, about 19 percent of students ages 12-18 reported being bullied during school. The percentage was higher for female students than for male students (22 vs. 17 percent) and higher for middle school students (26 percent) than high school students (16 percent).\"\n","source_date":"2024-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260405214740/https://nces.ed.gov/pubs2024/2024109.pdf","calculation_notes":"The School Crime Supplement (SCS) to the National Crime Victimization Survey is the gold-standard US measure. It surveys ~5,800 students ages 12-18 biennially. The 19% figure is the native estimate. Types of bullying: being made fun of, called names, or insulted (12%); subject of rumors (10%); pushed, shoved, tripped, or spit on (5%); excluded from activities on purpose (5%); threatened with harm (4%). Among those bullied, 32% experienced it on 1 day, 18% on 2 days, 31% on 3-10 days, and 18% on more than 10 days in the school year.\n"},{"url":"https://www.cdc.gov/nchs/data/databriefs/db514.pdf","title":"Bullying Victimization Among Teenagers: United States, July 2021-December 2023","publisher":"National Center for Health Statistics, CDC","source_type":"govt_report","statistic":"34.0% of teenagers ages 12-17 were bullied in the past 12 months; 38.4% for ages 12-14, 29.7% for ages 15-17","excerpt":"\"During July 2021-December 2023, 34.0% of teenagers ages 12-17 were bullied in the past 12 months. The percentage was higher among younger teenagers ages 12-14 (38.4%) than among teenagers ages 15-17 (29.7%). Sexual or gender minority teenagers were more likely to be bullied (47.1%) than teenagers who are not a sexual or gender minority (30.0%).\"\n","source_date":"2024-10-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260318190346/https://www.cdc.gov/nchs/data/databriefs/db514.pdf","calculation_notes":"NCHS Data Brief 514 uses National Health Interview Survey (NHIS) data pooled over 2.5 years. The 34% figure is substantially higher than the NCES 19% because the NHIS uses a broader definition of bullying (parent- or self-report, includes verbal teasing that may not meet the SCS behavioral threshold). The NHIS also captures younger teens (12-14) who have higher victimization rates. The subgroup figures are critical for the personal factor multipliers: girls 38.3% vs boys 29.9%, SGM teens 47.1% vs non-SGM 30.0%, teens with developmental disability 44.4% vs without 31.3%.\n"},{"url":"https://www.cdc.gov/yrbs/dstr/pdf/YRBS-2023-Data-Summary-Trend-Report.pdf","title":"Youth Risk Behavior Survey Data Summary & Trends Report 2013-2023","publisher":"Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"In 2023, 19% of US high school students reported being bullied on school property in the past year, up from 15% in 2021","excerpt":"\"The percentage of high school students who reported being bullied on school property was 19.0% in 2023, compared with 15.2% in 2021. Female students (22%) were more likely than male students (17%) to report being bullied at school.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260331022136/https://www.cdc.gov/yrbs/dstr/pdf/YRBS-2023-Data-Summary-Trend-Report.pdf","calculation_notes":"The YRBS is a biennial survey of ~17,000 high school students (grades 9-12). The 2023 figure of 19% aligns with the NCES SCS figure for the same age group. The increase from 15% in 2021 is notable and may reflect post-pandemic social adjustment. YRBS also reports that 29% of LGB students were bullied at school, compared to approximately 17% of heterosexual students (roughly 1.7x multiplier). This source corroborates the NCES data and provides the trend context.\n","independence_note":"YRBS and SCS are independently administered surveys with different sampling frames; convergence on ~19% for high schoolers strengthens confidence in the native estimate.\n"},{"url":"https://unesdoc.unesco.org/ark:/48223/pf0000366486","title":"Behind the Numbers: Ending School Violence and Bullying","publisher":"UNESCO","source_type":"reputable_reference","statistic":"Almost 1 in 3 students (32%) worldwide has been bullied by peers at school at least once in the last month","excerpt":"\"Almost one in three students (32%) has been bullied by their peers at school at least once in the last month and a similar proportion are affected by physical violence. Bullying has decreased in almost half of the 71 countries and territories studied.\"\n","source_date":"2019-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250222150233/https://unesdoc.unesco.org/ark:/48223/pf0000366486","calculation_notes":"UNESCO's global figure of 32% (past month) is substantially higher than the US annual figures because it includes countries with less developed anti-bullying infrastructure and uses a past-month reporting window. This source provides the global context and establishes that US rates, while concerning, are below the global average. Physical bullying is more common outside North America and Europe, where psychological bullying predominates.\n","independence_note":"UNESCO draws on HBSC, GSHS, PIRLS, and TIMSS data — survey instruments independent from US-administered NCES SCS and CDC YRBS.\n"}],"comparison_anchors":[{"label":"Workplace bullying (lifetime, US adult)","lifetime_us_adult":0.48},{"label":"Cyberbullying victimization (past year, US teens)","lifetime_us_adult":0.16},{"label":"Experiencing violent crime (lifetime, US adult)","lifetime_us_adult":0.25}],"regional_breakdown":[{"region":"Middle school (grades 6-8)","probability":0.26,"notes":"Peak bullying age; NCES SCS 2022 reports 26% for this group"},{"region":"High school (grades 9-12)","probability":0.16,"notes":"NCES SCS 2022; YRBS 2023 reports 19% for a slightly different question"},{"region":"Global average (past month)","probability":0.32,"notes":"UNESCO 2019; includes physical and psychological bullying"}],"personal_factor_multipliers":[{"factor":"LGBTQ+ student","multiplier":1.6,"notes":"NCHS Data Brief 514: SGM teens 47.1% vs non-SGM 30.0%; YRBS reports LGB students ~1.7x more likely to be bullied at school"},{"factor":"Student with developmental disability","multiplier":1.4,"notes":"NCHS Data Brief 514: 44.4% vs 31.3% for teens without developmental disability; other studies report 2-3x for specific conditions like ASD"},{"factor":"Middle school age (11-14)","multiplier":1.3,"notes":"NCHS: 38.4% for ages 12-14 vs 29.7% for ages 15-17; NCES: 26% middle school vs 16% high school"},{"factor":"Female student","multiplier":1.2,"notes":"NCES: 22% vs 17%; NCHS: 38.3% vs 29.9%; gender gap is consistent across surveys"}],"short_label":"School bullying","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"mental_trauma","valence":"negative","caveats":"Prevalence estimates for school bullying range from 19% (NCES SCS, strict behavioral criteria) to 34% (NCHS NHIS, broader parent/self-report) to 32% globally (UNESCO, past month). The discrepancy is driven by definition thresholds, reporting windows, and whether the survey asks about specific behaviors or uses the word \"bullying\" (self-labeling produces lower estimates). The lifetime compound of ~65% assumes imperfect year-to-year independence; if bullying is highly clustered in a subset of chronically targeted children, the population-level \"ever bullied\" rate may be lower than simple compounding suggests, but the individual burden on those children is much higher. The entry covers in-person bullying only; cyberbullying is tracked separately. Severity varies enormously — the 19% includes everything from a single name-calling incident to sustained physical intimidation over an entire school year. The 2023 YRBS increase from 15% to 19% may reflect post-pandemic social re-adjustment rather than a secular trend, and should be interpreted cautiously until the next survey cycle confirms or reverses it.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty school hallway with a single backpack left on the floor, flat vector illustration, muted tones, no people."},"canonical_url":"https://likelier.app/school-bullying","api_url":"https://likelier.app/api/fears/school-bullying.json"},{"slug":"unprotected-sex-sti","question":"What are the odds of catching an STI from a single unprotected sexual encounter with a new partner?","quick_answer":"Any STI runs about 1 in 11 per unprotected vaginal encounter with a random US adult partner — roughly a 2 in 3 chance across 10 such partners. Perceptions lean heavily toward HIV while badly underweighting HPV, which drives most of the actual risk.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Public perception of STI risk from a single encounter is shaped almost entirely by HIV awareness campaigns, which anchored the idea that unprotected sex is an acute mortal threat. Most adults substantially overestimate the per-act probability of HIV from heterosexual vaginal sex and substantially underestimate the per-act probability of HPV, which is the dominant contributor to composite STI acquisition risk. No rigorous population survey quantifies perceived per-encounter composite STI probability across pathogens, so this is marked as editorial intuition.\n","rough_estimate":"Most adults perceive a 'significant' risk per encounter, heavily weighted toward HIV and heavily underweighting HPV","kind":"intuition"},"native":{"display":"~1 in 11 per unprotected vaginal encounter with a random US adult partner (any STI)","numerator":1,"denominator":11,"unit":"per unprotected vaginal sex act","population":"Heterosexual US adults aged 18-59, random partner draw from general population"},"normalized":{"lifetime_us_adult":0.65,"display":"~2 in 3 lifetime (10 new partners, unprotected vaginal)","log_value":-0.187,"assumptions":"Scope is activity_specific_lifetime. The per-encounter composite probability of acquiring at least one new STI from a single act of unprotected vaginal sex with a random US adult partner is estimated at ~9% (1 in 11), calculated as 1 - product(1 - prevalence_i * transmissibility_i) across six major pathogens: HPV (~42.5% prevalence * ~20% per-act ≈ 8.5%), chlamydia (~1.8% * 4.5% ≈ 0.081%), gonorrhea (~0.3% * 20% ≈ 0.06%), HSV-2 (~12% * 0.15% ≈ 0.018%), HIV (~0.36% * 0.08% ≈ 0.00029%), syphilis (~0.05% * 5% ≈ 0.0025%). HPV dominates the composite, contributing ~94% of the total risk. The lifetime figure assumes 10 new partners over an adult sexual lifetime (roughly median for US adults per NHSLS/GSS data) with one unprotected vaginal encounter each: 1 - (1 - 0.09)^10 ≈ 0.61. Rounded up to ~0.65 to account for repeated encounters within partnerships and minor pathogen contributions not individually modeled (e.g., Mycoplasma genitalium, trichomoniasis). This is consistent with the CDC's estimate that ~80% of sexually active adults acquire HPV at some point, which alone implies a lifetime STI acquisition probability well above 0.5.\n","uncertainty":{"low":0.4,"high":0.85},"scope":"activity_specific_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6195215/","title":"Estimating per-act HIV transmission risk: a systematic review","publisher":"AIDS / Patel et al. (CDC)","source_type":"peer_reviewed","statistic":"Per-act HIV transmission risk: receptive vaginal 8 per 10,000 (95% CI 6-11), insertive vaginal 4 per 10,000 (95% CI 1-14), receptive anal 138 per 10,000 (95% CI 102-186), insertive anal 11 per 10,000 (95% CI 4-28)","excerpt":"\"Sexual-exposure risks ranged from low for oral sex to 138 (CI 102-186) per 10,000 exposures for receptive anal intercourse. [...] Receptive vaginal intercourse: 8 (95% CI 6-11) per 10,000 exposures. Insertive vaginal intercourse: 4 (95% CI 1-14) per 10,000 exposures.\"\n","source_date":"2014-06-19","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260406062635/https://pmc.ncbi.nlm.nih.gov/articles/PMC6195215/","calculation_notes":"Per-act receptive vaginal HIV risk = 8/10,000 = 0.08%. Combined with US adult HIV prevalence of ~0.36% (1.2M PLWH / 330M): per-encounter HIV risk from random partner = 0.0008 * 0.0036 = 0.0000029, or about 1 in 345,000. Even the higher receptive-anal figure of 1.38% * 0.36% prevalence = 0.005%, or ~1 in 20,000 per act with a random partner. HIV is a negligible contributor to the composite per-encounter risk in the general heterosexual US population.\n","independence_note":"CDC-authored systematic review of per-act HIV transmission studies; partially overlaps with Boily 2009 in included primary studies but uses a different analytic framework. The two meta-analyses converge on similar per-act estimates.\n"},{"url":"https://www.cdc.gov/sti-statistics/annual/index.html","title":"Sexually Transmitted Infections Surveillance, 2024 (Provisional)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"In 2024 (provisional), over 2.2 million cases of chlamydia, gonorrhea, and syphilis were reported: 1,515,985 chlamydia, 543,409 gonorrhea, 190,242 syphilis","excerpt":"\"There were still more than 2.2 million reported STIs in 2024. [...] Chlamydia cases declined for the second year in a row, down 8% since 2023. Gonorrhea cases declined for the third year in a row, down 10% since 2023. Primary and secondary syphilis cases declined for the second year in a row, down 22% since 2023.\"\n","source_date":"2025-09-24","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260424000000*/https://www.cdc.gov/sti-statistics/annual/index.html","calculation_notes":"Reported cases understate true prevalence because most chlamydia and many gonorrhea infections are asymptomatic. Chlamydia point prevalence estimated at ~1.8% of adults (reported ~1.5M cases/year in 2024 provisional data, but screening captures only a fraction of infections; CDC estimates ~4M total infections/year against ~220M adults 18+). Gonorrhea point prevalence ~0.3% (reported ~543K cases/year, shorter duration of infection). Syphilis point prevalence ~0.05% (~190K all-stages cases but most are latent/non-infectious). These prevalence estimates feed into the composite calculation. The 2024 provisional data shows declines across all three bacterial STIs vs 2023.\n","independence_note":"National surveillance data from mandatory reporting; independent of the per-act transmission studies (Patel, Boily) and the prevalence surveys (McQuillan HPV, McQuillan HSV-2). Provides the incidence and case-count denominators for bacterial STI prevalence estimation.\n"},{"url":"https://www.cdc.gov/nchhstp/director-letters/release-2024-sti-data.html","title":"CDC Releases 2024 National STI Data (NCHHSTP Director's Statement)","publisher":"US Centers for Disease Control and Prevention, NCHHSTP","source_type":"govt_report","statistic":"In 2024, combined chlamydia/gonorrhea/syphilis cases fell 9% from 2023 (third consecutive annual decline), though overall STI burden remains ~13% higher than a decade ago; congenital syphilis rose for a 12th consecutive year and is ~700% higher than a decade ago","excerpt":"\"Compared to a decade ago, the overall STI case burden in the U.S. remains 13% higher. [...] increased use of innovations like STI self-tests, newer point-of-care tests, and doxycycline as post-exposure prophylaxis for some bacterial STIs (doxy PEP), make a difference. [...] congenital syphilis morbidity is 700% higher than it was a decade ago.\"\n","source_date":"2025-09-24","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260703223811/https://www.cdc.gov/nchhstp/director-letters/release-2024-sti-data.html","calculation_notes":"This source updates the trend context, not the headline composite. The 9% decline in reported bacterial STIs and the prevention-innovation attribution (doxy-PEP, self-tests) support the doxy-PEP personal-factor entry but do not change the per-encounter acquisition probability, which is HPV-dominated (>94% of composite) and unaffected by doxycycline. The 700% congenital-syphilis rise is mother-to-child vertical transmission, a separate exposure axis from the adult per-encounter acquisition risk modeled here; it is noted in caveats, not in the headline number.\n","independence_note":"CDC NCHHSTP director's statement accompanying the 2024 surveillance release. Shares the underlying surveillance dataset with the STI Surveillance source above but is a separate publication; cited for the decade-trend framing and the prevention-innovation attribution not present on the surveillance landing page.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10140182/","title":"Postexposure Doxycycline to Prevent Bacterial Sexually Transmitted Infections","publisher":"New England Journal of Medicine / Luetkemeyer et al. (DoxyPEP trial)","source_type":"peer_reviewed","statistic":"Combined incidence of gonorrhea, chlamydia, and syphilis was lower by two thirds with doxycycline PEP vs standard care: RR 0.34 (95% CI 0.24-0.46) in the PrEP cohort, RR 0.38 (95% CI 0.24-0.60) in the PLWH cohort","excerpt":"\"The combined incidence of gonorrhea, chlamydia, and syphilis was lower by two thirds with doxycycline postexposure prophylaxis than with standard care, a finding that supports its use among MSM with recent bacterial STIs.\"\n","source_date":"2023-04-06","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260529013156/https://pmc.ncbi.nlm.nih.gov/articles/PMC10140182/","calculation_notes":"Sources the doxy-PEP personal-factor multiplier. The ~two-thirds reduction (RR 0.34/0.38) applies only to bacterial STIs (chlamydia, gonorrhea, syphilis), which together are under 1% of this HPV-dominated per-encounter composite. Doxycycline has no effect on HPV, HSV-2, or HIV. Composite multiplier therefore ≈0.99 — near-nil headline effect, mirroring the PrEP/ART factors. The trial population is MSM/transgender women on PrEP or with HIV, not the general heterosexual baseline; the factor is informational about a real prevention tool, not a recalibration of the headline.\n","independence_note":"Randomized controlled trial; independent of the surveillance data and the prevalence surveys. Provides the source-traceable effect size for the doxy-PEP factor; the CDC director's statement above independently corroborates real-world doxy-PEP uptake.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/28463105/","title":"Prevalence of HPV in Adults Aged 18-69: United States, 2011-2014","publisher":"NCHS Data Brief No. 280 / McQuillan et al.","source_type":"govt_report","statistic":"During 2013-2014, any genital HPV prevalence among adults aged 18-59 was 42.5% overall (45.2% men, 39.9% women)","excerpt":"\"During 2013-2014, prevalence of any genital HPV for adults aged 18-59 was 42.5% (45.2% among men and 39.9% among women). Prevalence of high-risk genital HPV was 22.7%.\"\n","source_date":"2017-04-01","source_accessed":"2026-04-17","archive_url":"https://web.archive.org/web/20260503095041/https://pubmed.ncbi.nlm.nih.gov/28463105/","calculation_notes":"HPV point prevalence of 42.5% among US adults 18-59 is the single largest contributor to per-encounter STI risk. Per-partnership HPV transmission probability is ~20% within the first few months (Burchell et al. 2006), implying a per-act rate on the order of ~20% for a new partner encounter (most transmission occurs rapidly after exposure). 0.425 * 0.20 = 0.085, or ~8.5% per encounter, which alone accounts for >94% of the composite STI risk.\n","independence_note":"NHANES-based population prevalence survey; independent of the per-act transmission studies and the bacterial STI surveillance data. Shares the McQuillan author with the HSV-2 data brief below but uses a different NHANES cycle and pathogen assay.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29442994/","title":"Prevalence of Herpes Simplex Virus Type 1 and Type 2 in Persons Aged 14-49: United States, 2015-2016","publisher":"NCHS Data Brief No. 304 / McQuillan et al.","source_type":"govt_report","statistic":"HSV-2 seroprevalence among persons aged 14-49 was 11.9% in 2015-2016, down from 18.0% in 1999-2000","excerpt":"\"During 2015-2016, prevalence of HSV-1 was 47.8% and HSV-2 was 11.9% among persons aged 14-49. Age-adjusted prevalence of HSV-2 decreased linearly over time by 5.9 percentage points, from 18.0% in 1999-2000 to 12.1% in 2015-2016.\"\n","source_date":"2018-02-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260318224612/https://pubmed.ncbi.nlm.nih.gov/29442994/","calculation_notes":"HSV-2 prevalence of ~12% combined with a per-act transmission probability of ~0.1-0.3% during asymptomatic shedding (Schiffer et al. 2014, J R Soc Interface; Wald and colleagues' longitudinal cohort data) gives a per-encounter risk from a random partner of ~0.12 * 0.0015 = 0.00018, or about 1 in 5,500. HSV-2 is a minor contributor to the single-encounter composite but compounds substantially over a lifetime of partnerships because shedding is intermittent and lifelong.\n","independence_note":"NHANES-based seroprevalence survey; shares the McQuillan lead author with the HPV data brief above but uses different NHANES cycles (2015-2016 vs 2011-2014) and a different serological assay. Provides the HSV-2 prevalence input to the composite calculation.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4467783/","title":"Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies","publisher":"The Lancet Infectious Diseases / Boily et al.","source_type":"peer_reviewed","statistic":"Pooled per-act female-to-male HIV transmission in high-income countries: 0.04% (95% CI 0.01-0.14%); male-to-female: 0.08% (95% CI 0.06-0.11%); receptive anal: 1.7% (95% CI 0.3-8.9%)","excerpt":"\"Pooled female-to-male (0.04% per act [95% CI 0.01-0.14]) and male-to-female (0.08% per act [95% CI 0.06-0.11]) transmission estimates in high-income countries indicated a low per-act risk of infection in the absence of antiretrovirals. [...] The pooled receptive anal intercourse estimate was much higher (1.7% per act [95% CI 0.3-8.9]).\"\n","source_date":"2009-02-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260503084014/https://pmc.ncbi.nlm.nih.gov/articles/PMC4467783/","calculation_notes":"Boily's meta-analysis corroborates Patel 2014 within overlapping confidence intervals. The high-income-country male-to-female rate of 0.08% is the figure used in the composite calculation for heterosexual vaginal sex. Multiplied by US adult HIV prevalence of 0.36%: 0.0008 * 0.0036 ≈ 2.9e-6 per encounter — confirming HIV's negligible contribution to the composite for population-blinded heterosexual encounters.\n","independence_note":"Lancet Infectious Diseases meta-analysis of observational cohort studies; partially overlaps with Patel 2014 in included studies but uses different pooling methodology. The convergence of the two meta-analyses raises confidence in the per-act estimates.\n"},{"url":"https://www.cdc.gov/hiv-data/nhss/estimated-hiv-incidence-and-prevalence.html","title":"Estimated HIV Incidence and Prevalence in the United States","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Estimated 1.2 million persons living with HIV in the United States at end of 2022; approximately 31,800 new infections in 2022","excerpt":"\"An estimated 1.2 million persons in the United States were living with diagnosed and undiagnosed HIV at the end of 2022. [...] Approximately 31,800 people acquired HIV in the United States in 2022.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260502021355/https://www.cdc.gov/hiv-data/nhss/estimated-hiv-incidence-and-prevalence.html","calculation_notes":"1.2 million PLWH / 330 million US population = 0.36% adult prevalence. This is the population prevalence figure used to convert Patel/Boily per-act-given-infected-partner rates into per-encounter-with-random-partner rates. Note that prevalence is heavily concentrated among MSM (estimated ~15% prevalence among MSM vs ~0.1% among heterosexual adults), so the population-blinded 0.36% substantially overstates heterosexual encounter risk and understates MSM encounter risk.\n","independence_note":"National HIV Surveillance System data; independent of the per-act transmission meta-analyses and the NHANES prevalence surveys. Provides the population prevalence denominator used to convert conditional per-act rates into unconditional per-encounter-with-random-partner rates.\n"}],"comparison_anchors":[{"label":"Lung cancer (lifetime, US adult)","lifetime_us_adult":0.062},{"label":"Car crash death (lifetime, US adult)","lifetime_us_adult":0.0108},{"label":"Skiing serious injury (per 20-day season)","lifetime_us_adult":0.0392}],"regional_breakdown":[{"region":"Unprotected vaginal, heterosexual, US adult — any STI composite","probability":0.09,"notes":"Per-encounter. HPV dominates (~8.5% contribution). Headline figure."},{"region":"Same encounter with consistent condom use","probability":0.04,"notes":"Condoms ~50-60% effective against HPV (skin-to-skin pathogen), ~80-90% against bacterial STIs and HIV. Composite roughly halved."},{"region":"Receptive anal, MSM, major US metro — any STI","probability":0.25,"notes":"HIV prevalence ~15% among MSM; gonorrhea/syphilis prevalence 5-10x general population. Per-act anal gonorrhea transmission ~30-50%. Composite far higher."},{"region":"With PrEP — HIV-specific risk reduction","probability":1e-7,"notes":"PrEP reduces HIV acquisition risk by ~99% (Grant et al. 2010, iPrEx). Applies only to HIV; no effect on other STIs."},{"region":"Serodiscordant couple, known HIV+ partner, no ART — receptive vaginal","probability":0.0008,"notes":"Patel 2014 per-act receptive vaginal: 8/10,000. With ART achieving viral suppression: effectively 0 (Cohen et al. 2016, HPTN 052)."},{"region":"HPV per-encounter (random US adult partner, unprotected vaginal)","probability":0.085,"notes":"42.5% prevalence × ~20% per-encounter transmission. Dominant pathogen in composite."},{"region":"Chlamydia per-encounter (random US adult partner, unprotected vaginal)","probability":0.00081,"notes":"~1.8% prevalence × ~4.5% per-act male-to-female transmission."},{"region":"Gonorrhea per-encounter (random US adult partner, unprotected vaginal)","probability":0.0006,"notes":"~0.3% point prevalence × ~20% per-act male-to-female transmission."},{"region":"HIV per-encounter (random US adult partner, unprotected vaginal)","probability":0.0000029,"notes":"~0.36% prevalence × 0.08% per-act male-to-female. About 1 in 345,000."},{"region":"HSV-2 per-encounter (random US adult partner, unprotected vaginal)","probability":0.00018,"notes":"~12% seroprevalence × ~0.15% per-act during asymptomatic shedding. About 1 in 5,500."},{"region":"Syphilis per-encounter (random US adult partner, unprotected vaginal)","probability":0.000025,"notes":"~0.05% infectious-stage prevalence × ~5% per-act with infectious partner. About 1 in 40,000."}],"personal_factor_multipliers":[{"factor":"MSM (men who have sex with men)","multiplier":3,"notes":"HIV prevalence ~15% vs ~0.1% heterosexual; gonorrhea/syphilis prevalence 5-10x. Composite per-encounter risk roughly 3x the heterosexual baseline, driven primarily by higher pathogen prevalence and anal-route transmissibility."},{"factor":"Age under 25","multiplier":1.5,"notes":"Chlamydia and gonorrhea rates 2-3x higher in 15-24 age group (CDC STI Surveillance 2023). HPV acquisition peaks in late teens/early 20s. Composite elevated by higher partner-pool prevalence."},{"factor":"Multiple new partners per year (>4)","multiplier":2,"notes":"Applies to cumulative annual risk, not per-act. More exposures compound, and high-partner-count individuals are drawn from a higher-prevalence subnetwork (assortative mixing)."},{"factor":"Partner from high-prevalence subgroup (e.g. MSM, sex worker, IV drug user)","multiplier":5,"notes":"Conditional on partner being from a high-prevalence sexual network, the 'prevalence' term in the per-encounter formula can be 10-50x the general population for HIV/syphilis/gonorrhea."},{"factor":"Consistent condom use","multiplier":0.45,"notes":"Condoms reduce HIV/gonorrhea/chlamydia by 80-90% per act, HPV by ~40-60%, HSV-2 by ~30%. Weighted across pathogens where HPV dominates, net composite reduction ~55%."},{"factor":"HPV vaccination (9-valent / Gardasil 9, ideally before sexual debut)","multiplier":0.5,"notes":"Unlike the HIV- and bacteria-specific factors below, this one genuinely moves the headline: HPV is ~94% of this per-act composite (~8.5 of ~9 percentage points). The 9-valent vaccine prevents ~90% of cancer- and genital-wart-causing HPV types but not all ~40 mucosal types, so the composite roughly halves rather than collapsing. Far more effective given before first sexual activity (CDC). The cervical-cancer entry uses 0.15× for the narrower cancer endpoint."},{"factor":"PrEP (for HIV component only)","multiplier":0.97,"notes":"Reduces HIV acquisition by ~99%. Since HIV is a negligible fraction of the composite per-encounter risk (~0.003% of total), the effect on the composite headline is minimal. The value of PrEP is in preventing the highest-consequence infection, not the most probable one."},{"factor":"Male circumcision","multiplier":0.7,"notes":"Three RCTs in sub-Saharan Africa showed ~60% reduction in female-to-male HIV transmission (Auvert 2005, Bailey 2007, Gray 2007). Also ~30% reduction in HPV and HSV-2 per meta-analyses (Albero 2012). Net composite reduction ~30%."},{"factor":"Partner on suppressive ART (HIV-specific)","multiplier":0.97,"notes":"Undetectable = untransmittable (Cohen et al. 2016 HPTN 052). Eliminates HIV risk from that partner. Negligible composite effect for same reason as PrEP — HIV is a tiny fraction of per-encounter composite."},{"factor":"Doxycycline post-exposure prophylaxis (doxy-PEP, bacterial STIs only)","multiplier":0.99,"notes":"The DoxyPEP RCT (Luetkemeyer et al. 2023, NEJM) cut combined gonorrhea/chlamydia/syphilis incidence by about two-thirds (RR 0.34 in the PrEP cohort, 0.38 in the PLWH cohort). CDC cites wider doxy-PEP uptake as one factor behind the 2024 decline in bacterial STIs. But like PrEP for HIV, the composite effect here is near-nil: bacterial STIs (chlamydia + gonorrhea + syphilis) together are under 1% of this HPV-dominated per-encounter composite, and doxycycline has no effect on HPV, HSV-2, or HIV. Value is in preventing the bacterial infections, not in moving the headline."}],"short_label":"Unprotected sex","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"These are per-act probabilities conditional on partner infection status weighted by population prevalence. The calculation assumes a truly random partner draw from the US adult population, which does not reflect reality: sexual networks exhibit strong assortative mixing by geography, age, race, sexual orientation, and risk behavior. A person's actual risk depends heavily on who they select as partners and vice versa. All prevalence figures are US-centric; global prevalence varies enormously (HIV prevalence in sub-Saharan Africa is ~25x the US rate; HPV prevalence is high everywhere). Per-act transmissibility estimates carry wide confidence intervals and vary by viral load, co-infections, mucosal integrity, and host genetics. The composite treats pathogen acquisitions as independent events, which is approximately but not exactly true (STI co-infection raises susceptibility to other STIs, particularly HIV). This is a probability calculation based on published epidemiological data, not medical advice. Note a trend the per-encounter acquisition figure does not capture: CDC's 2024 provisional surveillance recorded a third consecutive annual decline in reported chlamydia, gonorrhea, and syphilis (combined cases down 9% from 2023, though the overall burden remains ~13% above a decade ago), which CDC attributes in part to wider use of STI self-tests, point-of-care testing, and doxycycline post-exposure prophylaxis. Against that, congenital syphilis — mother-to-child transmission, a separate exposure axis from the adult acquisition risk modeled here — rose for a 12th straight year to nearly 4,000 reported cases in 2024, roughly 700% above its 2014 level. The headline composite is an acquisition probability for a single adult encounter and does not reflect pregnancy or onward vertical-transmission risk.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-10-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-17","image":{"alt":"A single abstract geometric shield shape in muted tones on a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/unprotected-sex-sti","api_url":"https://likelier.app/api/fears/unprotected-sex-sti.json"},{"slug":"raw-meat-cross-contamination","question":"What are the odds of getting sick from not washing hands or surfaces after handling raw meat or eggs?","quick_answer":"Cross-contamination from raw meat or eggs contributes to foodborne illness for roughly 3 in 4 US adults over a lifetime. Most people treat a smear of chicken juice as effectively zero risk, far underestimating the cumulative odds.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Cross-contamination occupies an odd perceptual niche: almost everyone has heard the advice about separate cutting boards for meat and vegetables, yet observational studies consistently find that the majority of home cooks ignore it in practice. A USDA test-kitchen study found that participants failed to attempt handwashing 83% of the times they should have after touching raw meat or cracking eggs. The implicit mental model is that a quick rinse under running water, or simply wiping hands on a towel, is sufficient, and that the per-event risk of actual illness from a smear of chicken juice on a cutting board is too small to warrant real attention. No population survey cleanly isolates perceived cross-contamination risk, so the best available characterisation is editorial intuition: most adults treat the per-meal risk as effectively zero.\n","rough_estimate":"Most adults treat the per-event risk as negligible ('a little chicken juice won't hurt'); the cumulative lifetime probability is much higher than most expect","kind":"intuition"},"native":{"display":"~7.2 million cross-contamination-linked foodborne illnesses per year (US)","numerator":72,"denominator":3300,"unit":"per year","population":"US residents, all ages, foodborne illness where cross-contamination was a contributing factor"},"normalized":{"lifetime_us_adult":0.73,"display":"~3 in 4 lifetime (US adult)","log_value":-0.137,"assumptions":"Starts from CDC's ~48 million domestically acquired foodborne illnesses per year (Scallan et al. 2011). The CDC NORS contributing-factors analysis for 2014-2022 finds that cross-contamination of foods was a contributing factor in roughly 20-22% of bacterial outbreaks during 2014-2016 and 2017-2019, declining thereafter. However, cross-contamination in NORS is coded alongside other factors, and not all foodborne illness is captured in outbreak surveillance. We take ~15% as a conservative central estimate of the share of total US foodborne illness where cross-contamination (raw-to-ready-to-eat transfer via hands, surfaces, or utensils) was a meaningful contributing factor. 15% x 48 million = 7.2 million cases per year, or about 2.18% of the US population per year (roughly 1 in 46). Compounded over 59 years of remaining adult life: 1 - (1 - 0.0218)^59 = 0.727, or about 3 in 4. The uncertainty band runs from 10% contribution share (lifetime ~0.58) to 20% (lifetime ~0.82), spanning the defensible range given the overlap between NORS contributing factor codes and the gap between outbreak-detected and total illness.\n","uncertainty":{"low":0.58,"high":0.82},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","title":"Contributing Factors of Foodborne Illness Outbreaks — National Outbreak Reporting System, United States, 2014-2022","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR) Surveillance Summaries","source_type":"govt_report","statistic":"Cross-contamination of foods was among the top five contributing factors for bacterial outbreaks at 22.0% (2014-2016) and 20.8% (2017-2019); bare-hand contact by infected food workers declined from 20.5% to 8.9% across the three periods","excerpt":"\"For bacterial outbreaks, cross-contamination of foods was among the top five contributing factors during the first (22.0%) and second periods (20.8%), but not during the third period. [...] The proportion of outbreaks with contamination from an infectious food worker through barehand contact with food decreased (20.5%, 15.2%, and 8.9%, respectively) across the three time periods.\"\n","source_date":"2025-03-13","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421185402/https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","calculation_notes":"CDC NORS is the canonical US surveillance system for outbreak contributing factors. The 20-22% cross-contamination rate among bacterial outbreaks provides the upper anchor for the share of all foodborne illness attributable to cross-contamination. We deflate to ~15% as the central estimate because not all foodborne illness is bacterial, not all is detected via outbreak surveillance, and NORS contributing factors can be coded multiply. Applied to Scallan's 48 million illnesses/year: 0.15 x 48e6 = 7.2 million cases/year, or ~2.18% of the US population per year.\n","independence_note":"NORS draws on the same state and local public-health reporting pipeline as CDC FoodNet and the Scallan et al. estimates; treat as a methodological sibling rather than a fully independent data source.\n"},{"url":"https://www.usda.gov/about-usda/news/press-releases/2023/09/19/new-usda-study-consumer-kitchen-behavior-underscores-importance-food-safety-education-month","title":"New USDA Study on Consumer Kitchen Behavior Underscores the Importance of Food Safety Education Month","publisher":"US Department of Agriculture","source_type":"govt_report","statistic":"Handwashing not attempted 83% of the time it should have been after touching raw meat or cracking eggs; 26% of participants contaminated ready-to-eat food (cantaloupe) with tracer bacteria from raw pork sausage; only 32% cleaned and sanitized surfaces used for raw meat","excerpt":"\"Handwashing was not attempted 83% of the time when it should have been done (e.g., touching raw sausage and unwashed cantaloupe, cracking eggs, contaminated equipment or surfaces). Additionally, 96% of handwashing attempts did not contain all necessary steps. [...] Only 32% of people clean and sanitize the surface used to prepare raw meat. [...] The kitchen sink was most often contaminated, with 34% of participants contaminating the sink during meal preparation. The next highest was the cantaloupe, with 26% of participants introducing contamination when cutting the cantaloupe during meal preparation.\"\n","source_date":"2023-09-19","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421193138/https://www.usda.gov/about-usda/news/press-releases/2023/09/19/new-usda-study-consumer-kitchen-behavior-underscores-importance-food-safety-education-month","calculation_notes":"This USDA observational study establishes the behavioural exposure rate: the vast majority of home cooks fail to wash hands or sanitize surfaces after handling raw meat, meaning that cross-contamination events are not rare edge cases but the default kitchen practice. The 26% ready-to-eat food contamination rate in a single meal-preparation session is the mechanism through which the NORS contributing-factor percentages translate into actual illness. This source does not directly provide an illness probability but anchors the exposure frequency underlying the normalized estimate.\n","independence_note":"USDA FSIS observational kitchen studies are methodologically independent of CDC's outbreak surveillance; they measure consumer behaviour directly rather than inferring it from outbreak investigations.\n"},{"url":"https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","title":"Foodborne Illness Acquired in the United States — Major Pathogens","publisher":"CDC Emerging Infectious Diseases / Scallan et al.","source_type":"peer_reviewed","statistic":"31 major pathogens cause ~9.4 million illnesses, ~56,000 hospitalizations, and ~1,351 deaths per year in the US; combined with unspecified agents, total ~48 million illnesses/year","excerpt":"\"We estimated that 31 pathogens acquired in the United States caused 9.4 million episodes of foodborne illness (90% credible interval [CrI] 6.6-12.7 million), 55,961 hospitalizations (90% CrI 39,534-75,741), and 1,351 deaths (90% CrI 712-2,268) each year.\"\n","source_date":"2011-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260420041128/https://wwwnc.cdc.gov/eid/article/17/1/p1-1101_article","calculation_notes":"Provides the denominator: 48 million total US foodborne illnesses per year (9.4 million from known pathogens + ~38.4 million from unspecified agents, per the companion Scallan et al. 2011b paper). The 15% cross-contamination share is applied to this total to yield the 7.2 million cases/year used in the native figure. Scallan's 90% credible interval on the known-pathogen total (6.6-12.7 million) contributes to the uncertainty band.\n","independence_note":"Scallan et al. (2011a) is the foundational CDC burden estimate; the NORS contributing-factors analysis and the USDA kitchen study are methodologically downstream of the same surveillance infrastructure but measure different quantities.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S0362028X22100761","title":"Transfer of Campylobacter and Salmonella from Poultry Meat onto Poultry Preparation Surfaces","publisher":"Journal of Food Protection","source_type":"peer_reviewed","statistic":"Transfer rates of Campylobacter and Salmonella from chicken meat to kitchen surfaces varied from ~0% to 21.1%; mean transfer from chicken to hands 2.9-3.8%; washing significantly reduced but did not eliminate transfer","excerpt":"\"Transfer rates of both pathogens from chicken meat to all surfaces examined varied substantially between approximately 0 and 21.1%. [...] The mean transfer rates from legs and filets to hands were 2.9 and 3.8%. [...] The transfer rate of a cutting board or hands was significantly decreased after being washed.\"\n","source_date":"2017-04-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20230101081557/https://www.sciencedirect.com/science/article/pii/S0362028X22100761","calculation_notes":"Establishes the biological mechanism: raw poultry reliably transfers Campylobacter and Salmonella to hands and cutting boards at rates of 1-21%, and simple washing reduces but does not eliminate the transfer. This is the link between the USDA finding that 83% of consumers fail to wash hands after handling raw meat and the NORS finding that cross-contamination contributes to ~20% of bacterial outbreaks. Not used directly in the headline calculation but validates the plausibility of the 15% attribution share.\n","independence_note":"Laboratory microbiology study; methodologically independent of the CDC surveillance and USDA behavioural data.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Any foodborne illness in a single year (US, 1 in 7)","lifetime_us_adult":0.145},{"label":"Food left unrefrigerated illness (lifetime, US adult)","lifetime_us_adult":0.82},{"label":"Lifetime cancer diagnosis (US adult)","lifetime_us_adult":0.4}],"personal_factor_multipliers":[{"factor":"vegetarian or vegan (no raw meat or poultry handling)","multiplier":0.3,"notes":"Eliminates the dominant raw poultry pathway; residual risk comes from egg handling and secondary contact with meat-contaminated surfaces in shared kitchens or restaurants. The reduction is substantial but not complete.\n"},{"factor":"handles raw chicken multiple times per week","multiplier":2,"notes":"Frequency of exposure is the primary driver of cumulative risk. Households that prepare chicken 4-5 times per week have roughly twice the cross-contamination exposure of the average US household.\n"},{"factor":"consistently uses separate cutting boards and washes hands with soap for 20 seconds","multiplier":0.2,"notes":"The USDA kitchen study found that only 32% of consumers properly clean surfaces after raw meat. Those who consistently follow the separate-board-plus-handwashing protocol reduce transfer rates by roughly 80-90%, per the Journal of Food Protection transfer data.\n"},{"factor":"immunocompromised, under 5, or over 65","multiplier":3,"notes":"The illness probability per cross-contamination event is similar, but the severity conditional on infection is substantially higher. Campylobacter and Salmonella infections that are self-limiting in healthy adults can cause hospitalisation or death in vulnerable groups.\n"}],"short_label":"Raw meat cross-contamination","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The 15% attribution share is a modelled estimate, not a directly measured figure. NORS contributing-factor codes overlap (an outbreak can be coded with both cross-contamination and bare-hand contact), and the vast majority of foodborne illness never enters outbreak surveillance at all. The headline captures all forms of cross-contamination (hands, cutting boards, utensils, sink splash), not just the cutting-board-to-salad scenario most people picture. Most of the 7.2 million annual cases are mild gastroenteritis that resolves in 24-48 hours; the fatal subset is covered in the separate food-poisoning-death entry. The USDA behavioural data comes from test kitchens where participants knew they were being observed, which likely biases toward better-than-typical hygiene; real-world cross-contamination rates may be higher.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A wooden cutting board with a kitchen knife viewed from above on a neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/raw-meat-cross-contamination","api_url":"https://likelier.app/api/fears/raw-meat-cross-contamination.json"},{"slug":"chronic-back-pain","question":"What are the odds of developing chronic back pain?","quick_answer":"Roughly 4 in 5 adults experience significant low back pain at some point in life — far above the 1 in 4 most people under 40 assume. It rarely appears on fear surveys yet is the world's leading cause of disability.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Back pain occupies a strange position in the fear landscape: almost nobody lists it on a fear survey, yet it is the single largest cause of disability on the planet. Young adults treat it as a middle-aged inconvenience, something that happens to desk workers and weekend warriors who lift wrong. The mental model is acute and self-limiting: your back goes out, you rest, it gets better. That model is correct for many episodes and badly wrong for the cumulative lifetime picture. Ask a 25-year-old to estimate the probability that they will experience significant low back pain at some point in their life and you will get guesses in the range of 20-30%. The real number is closer to 80%. The chronic version (lasting three months or longer) is likewise underappreciated: roughly one in five adults is dealing with it at any given time, a prevalence that dwarfs most of the dramatic health fears that dominate headlines.\n","rough_estimate":"Most adults under 40 assume their lifetime back-pain risk is around 1 in 4","kind":"intuition"},"native":{"display":"~619 million people with low back pain globally (2020); ~39% of US adults report back pain in a 3-month window","numerator":4,"denominator":5,"unit":"lifetime","population":"US adults"},"normalized":{"lifetime_us_adult":0.8,"display":"~4 in 5 lifetime","log_value":-0.097,"assumptions":"Uses the NINDS fact sheet headline that about 80% of adults experience low back pain at some point in their lifetimes, cross-validated against the GBD 2021 systematic analysis (Lancet Rheumatology, 2023) which estimated 619 million prevalent cases globally in 2020 out of ~5.2 billion adults (age 20+), giving a point prevalence of ~12%. With a one-year prevalence of ~38% (consistent with CDC/NCHS Data Brief 415 reporting 39.0% of US adults experiencing back pain in a 3-month window) and studies showing recurrence rates of 24-80% within one year, the lifetime cumulative incidence of at least one significant episode converges toward 80% in high-income populations. For chronic low back pain specifically (lasting 3+ months), a systematic review (Meucci et al., 2015) reported point prevalence of 4.2% at ages 24-39 and 19.6% at ages 20-59 (the widely cited 20.1% summary figure it discusses is from Hoy et al. 2012). The headline figure uses the broader \"any significant back pain episode\" lifetime prevalence because the question asks about developing chronic back pain and most chronic cases begin as acute episodes that fail to resolve. Uncertainty range 0.70-0.85 reflects methodological variation across studies using different definitions of \"significant\" pain and different recall windows.\n","uncertainty":{"low":0.7,"high":0.85},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.thelancet.com/journals/lanrhe/article/PIIS2665-9913(23)00098-X/fulltext","title":"Global, regional, and national burden of low back pain, 1990-2020, its attributable risk factors, and projections to 2050: a systematic analysis of the Global Burden of Disease Study 2021","publisher":"The Lancet Rheumatology","source_type":"primary_study","statistic":"619 million (95% UI: 554-694 million) people had low back pain globally in 2020; low back pain is the leading cause of years lived with disability worldwide; projected to reach 843 million by 2050","excerpt":"\"In 2020, low back pain affected 619 million (95% uncertainty interval 554 to 694) people globally [...] low back pain remains the leading cause of years lived with disability [...] Cases of low back pain are projected to increase to 843 million (759 to 933) by 2050 [...] Occupational ergonomic factors, smoking, and high BMI explained 38.8% (28.7 to 47.0) of years lived with disability due to low back pain.\"\n","source_date":"2023-06-05","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250813172635/https://www.thelancet.com/journals/lanrhe/article/PIIS2665-9913(23)00098-X/fulltext","calculation_notes":"619 million prevalent cases / ~5.2 billion adults globally (age 20+) = ~12% global point prevalence. This is the snapshot count at any given moment. Given that most back pain episodes last weeks to months and recurrence is high, the cumulative lifetime incidence is far higher than the point prevalence. The GBD finding that low back pain is the #1 cause of YLDs worldwide (ahead of depressive disorders, diabetes, hearing loss, and all other conditions) anchors the claim that this is the leading cause of disability globally. The paper's decomposition analysis of the 619-million-to-843-million 2020-2050 projection found population growth is the largest contributor in most regions, except in east Asia, south Asia, and tropical/southern Latin America and the Caribbean, where population ageing is the dominant driver instead -- grounding the body text's claim about regional projection drivers.\n","independence_note":"GBD/IHME is the upstream methodology; partially overlaps with WHO disability estimates but uses independent systematic review of prevalence studies.\n"},{"url":"https://www.ninds.nih.gov/low-back-pain-fact-sheet","title":"Low Back Pain Fact Sheet","publisher":"National Institute of Neurological Disorders and Stroke (NINDS)","source_type":"govt_report","statistic":"About 80 percent of adults experience low back pain at some point in their lifetimes; it is the most common cause of job-related disability","excerpt":"\"About 80 percent of adults experience low back pain at some point in their lifetimes. It is the most common cause of job-related disability and a leading contributor to missed work days. [...] Most low back pain is acute, or short term, and lasts a few days to a few weeks. It tends to resolve on its own with self-care and there is no residual loss of function.\"\n","source_date":"2023-11-28","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20230113121029/https://www.ninds.nih.gov/low-back-pain-fact-sheet","calculation_notes":"The 80% lifetime prevalence figure is the primary anchor for the normalized headline number. NINDS derives this from a synthesis of epidemiological literature including NHANES and NHIS data. The fact sheet also notes that most acute episodes resolve, which is important context: the 80% figure is cumulative incidence of any significant episode, not 80% chronic prevalence. Used as the authoritative US government source for the lifetime headline.\n","independence_note":"NINDS synthesizes from US epidemiological surveys (NHIS, NHANES) and clinical literature; methodologically independent from GBD.\n"},{"url":"https://www.cdc.gov/nchs/products/databriefs/db415.htm","title":"Back, Lower Limb, and Upper Limb Pain Among U.S. Adults, 2019","publisher":"US Centers for Disease Control and Prevention / NCHS","source_type":"govt_report","statistic":"39.0% of US adults experienced back pain in the past 3 months (2019 NHIS); prevalence increased with age and was highest among adults 65+","excerpt":"\"Among adults, 39.0% experienced back pain, 36.5% experienced lower limb pain, and 30.7% experienced upper limb pain. [...] The percentage of adults who experienced back pain increased with age. [...] Women, non-Hispanic white adults, and those with income below 100% FPL were most likely to experience back pain.\"\n","source_date":"2021-07-30","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260327165204/https://www.cdc.gov/nchs/products/databriefs/db415.htm","calculation_notes":"39% of US adults reporting back pain in a 3-month recall window (2019 NHIS) implies a rough annual prevalence of ~50-55% after adjusting for episode duration and seasonal overlap. Compounding across a 60-year adult life with even modest recurrence yields a lifetime cumulative incidence consistent with the NINDS 80% figure. This source anchors the \"current burden\" framing and provides the demographic breakdown.\n","independence_note":"CDC/NCHS uses NHIS survey data, which is the same upstream as the NINDS synthesis but applies different statistical methods and recall windows.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4603263/","title":"Prevalence of chronic low back pain: systematic review","publisher":"Revista de Saude Publica (PMC)","source_type":"peer_reviewed","statistic":"Chronic low back pain prevalence was 4.2% at ages 24-39 and 19.6% at ages 20-59; prevalence increases linearly from the third decade to age 60 and is more prevalent in women (the often-quoted 20.1% summary figure is from Hoy et al. 2012, which this review cites)","excerpt":"\"Chronic low back pain prevalence was 4.2% in individuals aged between 24 and 39 years old and 19.6% in those aged between 20 and 59. [...] Chronic low back pain prevalence increases linearly from the third decade of life on, until the 60 years of age, being more prevalent in women. [...] Hoy et al made a valuable contribution to low back pain studies and estimated a summary prevalence of CLBP of 20.1% (SD = 9.8).\"\n","source_date":"2015-10-20","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420033643/https://pmc.ncbi.nlm.nih.gov/articles/PMC4603263/","calculation_notes":"This review reports chronic low back pain (defined as lasting 3+ months) at 4.2% among adults aged 24-39 and 19.6% among adults aged 20-59, and concludes that prevalence rises roughly linearly from the third decade to age 60, being higher in women. The frequently quoted 20.1% (SD = 9.8) summary prevalence is NOT this review's own pooled estimate — it is attributed here to Hoy et al. (2012). The narrower \"chronic\" definition (point prevalence on the order of 10-20% depending on age) sits below the broader \"any back pain episode\" lifetime figure of ~80%. Used to validate the chronic-specific prevalence claim and to anchor the personal_factor_multipliers age gradient.\n","independence_note":"Independent systematic review of 28 studies from multiple countries; does not share upstream data with GBD or CDC/NHIS estimates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29573871/","title":"Low back pain: a call for action","publisher":"The Lancet (Buchbinder R, van Tulder M, Oberg B, et al.) — third paper in the 2018 Lancet Low Back Pain Series","source_type":"peer_reviewed","statistic":"Low back pain is the leading worldwide cause of years lost to disability; the 2018 Lancet series (three companion papers) documented substantial global overuse of low-value care (imaging, opioids, spinal surgery) alongside underuse of recommended prevention and self-management approaches","excerpt":"\"Low back pain is the leading worldwide cause of years lost to disability and its burden is growing alongside the increasing and ageing population. [...] our Viewpoint, the third paper in this Lancet Series, is a call for action on this global problem of low back pain.\"\n","source_date":"2018-03-21","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714013243/https://pubmed.ncbi.nlm.nih.gov/29573871/","calculation_notes":"Grounds the body text's two references to \"the Lancet 2018 series\" and the personal_factor_multipliers note on the depression/anxiety comorbidity factor. This call-for-action viewpoint is the third of three companion papers (Hartvigsen et al., Lancet 2018;391:2356-67; Foster et al., Lancet 2018;391:2368-83; Buchbinder et al., Lancet 2018;391:2384-88) that collectively documented global overuse of low-value care (imaging, opioids, spinal surgery) and underuse of recommended prevention/self-management approaches -- the evidence-practice mismatch referenced in the body prose. A 2020 update to the series (PMC7434211) confirms the original findings, e.g. a 2012 US VA study found 59% of outpatient lumbar-spine scans were inappropriate, and $12.8 billion was spent on spinal fusion surgery in 2011 despite a 20% failure rate and little supporting evidence.\n","independence_note":"Independent health-policy/editorial synthesis layered on the same epidemiological evidence base as the GBD and NINDS sources above; added specifically to ground the body's references to \"the Lancet 2018 series,\" which this entry discussed without a citation.\n"}],"comparison_anchors":[{"label":"Developing Alzheimer's/dementia (lifetime, global adult)","lifetime_us_adult":0.12},{"label":"Developing type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33},{"label":"Experiencing a kidney stone (lifetime, US adult)","lifetime_us_adult":0.11},{"label":"Death from heart disease (lifetime, global adult)","lifetime_us_adult":0.085},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"personal_factor_multipliers":[{"factor":"Age 20-30, active lifestyle","multiplier":0.3,"notes":"Young adults have the lowest chronic back pain prevalence (~4%); most episodes are acute and self-limiting"},{"factor":"Age 50-65","multiplier":1.4,"notes":"Chronic low back pain prevalence peaks near age 60 per systematic review evidence; degenerative disc changes accumulate"},{"factor":"Sedentary occupation (prolonged sitting)","multiplier":1.5,"notes":"Occupational ergonomic factors are one of the three GBD-attributed risk factors; prolonged static postures increase risk"},{"factor":"Heavy manual labor (repetitive lifting)","multiplier":1.8,"notes":"Occupational ergonomic factors explained the largest share of GBD-attributed YLDs; dose-response with cumulative spinal loading"},{"factor":"Obese (BMI 30+)","multiplier":1.5,"notes":"High BMI is one of the three GBD modifiable risk factors; mechanical loading plus systemic inflammation"},{"factor":"Current smoker","multiplier":1.3,"notes":"Smoking is the third GBD-attributed risk factor; impairs disc nutrition via reduced blood flow and accelerates disc degeneration"},{"factor":"Depression or anxiety comorbidity","multiplier":2,"notes":"Bidirectional relationship; psychological distress is one of the strongest predictors of acute-to-chronic transition in the Lancet 2018 series"}],"short_label":"Chronic back pain","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline 80% lifetime figure is for any significant low back pain episode, not for chronic pain specifically. Chronic low back pain (conventionally defined as lasting 3 months or longer) has a point prevalence of roughly 20% and a lifetime cumulative incidence that is harder to pin down because definitions vary across studies and many episodes are recurrent rather than continuously present. The distinction between \"acute back pain that resolves\" and \"chronic back pain\" is clinically important but epidemiologically blurry: most chronic cases begin as acute episodes, and a substantial fraction of \"resolved\" acute episodes recur within a year. The GBD 619-million figure is a point-prevalence count (people with low back pain on any given day), not a lifetime count. The NINDS 80% figure is a lifetime cumulative incidence estimate synthesized from multiple epidemiological surveys. The personal_factor_multipliers for occupation, BMI, and smoking reflect the three modifiable risk factors identified in the GBD 2021 analysis; the depression/anxiety multiplier reflects a bidirectional relationship where the causal direction is genuinely uncertain. All multipliers are approximate relative risks from heterogeneous observational literature and are not additive.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single curved line bending under invisible weight on a muted warm-grey background, flat vector illustration."},"canonical_url":"https://likelier.app/chronic-back-pain","api_url":"https://likelier.app/api/fears/chronic-back-pain.json"},{"slug":"frontline-combatant-casualty","question":"What is the probability of a front-line infantry soldier being killed or seriously and permanently wounded over five years in a Ukraine-scale conventional conflict?","quick_answer":"Over a five-year Ukraine-scale conflict, a front-line infantry soldier faces roughly a 4 in 5 chance of being killed or permanently and seriously wounded. That is far above what most people guess, with estimates ranging from about 1 in 20 to 1 in 2 depending on which war they picture.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Public estimates of frontline infantry casualty rates in modern conventional war vary enormously, shaped by the conflict being imagined. People anchoring on the relatively low US post-9/11 wars casualty rate (~1 in 270 per deployer over 20 years) will drastically underestimate the risk in a Ukraine-scale peer-versus-peer conflict. People anchoring on Second World War imagery may overestimate. No high-quality survey specifically asks the public to estimate the killed-or-permanently-wounded rate for a frontline infantry soldier in a current-generation high-intensity conventional war; the perceived side is editorial intuition.\n","rough_estimate":"estimates range widely from ~1 in 20 (anchoring on post-9/11 US wars) to ~1 in 2 (anchoring on WWII Eastern Front); actual Ukraine-scale 5-year frontline rate likely in the range of 1 in 1.25 to 1 in 2","kind":"intuition"},"native":{"display":"~46,000 Ukrainian military killed + ~190,000 permanently/seriously wounded = ~236,000 serious casualties among an estimated ~300,000-soldier frontline force over ~3 years (Feb 2022 to Feb 2025)","numerator":236000,"denominator":300000,"unit":"over 3-year conflict (cumulative killed + permanently seriously wounded, frontline force)","population":"Ukrainian front-line infantry and armored troops directly engaged in combat operations, 2022-2025; excludes rear-area support, logistics, and administrative personnel"},"normalized":{"lifetime_us_adult":0.8,"display":"~4 in 5 over a five-year conflict","log_value":-0.097,"assumptions":"Two components of casualty data are combined: killed, and permanently/seriously wounded (those who did not return to duty). Ukrainian President Zelenskyy confirmed in February 2025 that over 46,000 Ukrainian military personnel had been killed and approximately 380,000 wounded, with roughly 50% of the wounded recovering and returning to active duty. This implies approximately 190,000 soldiers sustained wounds serious enough to remove them from active service. Combined killed + permanently incapacitated: ~46,000 + ~190,000 = ~236,000 over approximately 3 years. The denominator is the estimated size of Ukraine's frontline force: the Warsaw-based Centre for Eastern Studies (OSW) estimated in 2024 that of Ukraine's 1 million+ total military personnel, not more than approximately 300,000 were deployed on front lines at any given time. This produces a 3-year serious casualty rate of 236,000 / 300,000 = 0.787 for someone continuously serving in a frontline role. To project to 5 years, the annual serious casualty rate is modeled as (236,000 / 3) / 300,000 = 0.2622 per year. Compounding: five-year probability = 1 - (1 - 0.2622)^5 = 1 - (0.7378)^5 = 1 - 0.218 = 0.782. Rounding to 0.80 to account for the upward trend (2025 casualty intensity was higher than 2023). Using The Economist's higher estimate of ~100,000 killed (November 2024) with the same 50% return-to-duty ratio for wounded: ~100,000 killed + ~200,000 permanently wounded = 300,000 / 300,000 = 1.0 over 3 years, extrapolating to 0.87 over 5 years using the compound formula. Uncertainty range reflects both the lower Zelenskyy official figure and the higher independent estimates. This is the scope subgroup_lifetime because it measures a career-period probability for a specific high-risk military role, not a general population lifetime risk. Importantly, frontline units rotate; a soldier does not typically serve all five years in continuous frontline contact. The 5-year figure represents the accumulated probability over a career that includes frontline rotations.\n","uncertainty":{"low":0.55,"high":0.92},"scope":"subgroup_lifetime"},"sources":[{"url":"https://www.britannica.com/question/What-are-the-military-casualty-estimates-for-the-Russia-Ukraine-War","title":"What are the military casualty estimates for the Russia-Ukraine War?","publisher":"Encyclopaedia Britannica","source_type":"reputable_reference","statistic":"In December 2024 President Zelensky said 43,000 Ukrainian soldiers had been killed and 370,000 injured since February 2022; Britannica notes the figures are contested and not independently verifiable.","excerpt":"\"In December 2024, Ukrainian President Volodymyr Zelensky said 43,000 Ukrainian soldiers had been killed and 370,000 injured since February 2022; he also said 198,000 Russian soldiers had been killed and 550,000 injured. None of these claims was verifiable; Russia has said its losses are less than Ukraine's.\"\n","source_date":"2026-01-28","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260216182606/https://www.britannica.com/question/What-are-the-military-casualty-estimates-for-the-Russia-Ukraine-War","calculation_notes":"Britannica records Zelensky's December 2024 official figures: 43,000 Ukrainian soldiers killed and 370,000 injured since February 2022. These are the official Ukrainian government line and likely represent a lower bound on actual casualties; independent estimates (see The Economist / CSIS source) run higher. The mid-February 2025 update to over 46,000 killed / 380,000 wounded and the \"approximately 50%\" return-to-duty rate for the wounded -- which the casualty arithmetic in this entry uses -- are carried by the Casualties of the Russo-Ukrainian War source, not by Britannica, which only reports the December 2024 numbers.\n","independence_note":"Britannica records Zelensky's direct December 2024 statement. This is the official Ukrainian source, distinct from US intelligence estimates and independent media counts.\n"},{"url":"https://en.wikipedia.org/wiki/Casualties_of_the_Russo-Ukrainian_War","title":"Casualties of the Russo-Ukrainian War","publisher":"Wikipedia (citing multiple primary sources including The Economist, US official estimates, CSIS)","source_type":"reputable_reference","statistic":"President Zelenskyy stated 43,000 Ukrainian soldiers killed / 370,000 wounded with 'approximately 50%' of the wounded returned to active duty, updating the toll in mid-February 2025 to over 46,000 killed and 380,000 wounded; The Economist estimated 60,000-100,000 killed and ~400,000 wounded in November 2024; CSIS January 2026 estimated 100,000-140,000 fatalities.","excerpt":"\"Subsequently, President Zelenskyy announced 43,000 Ukrainian soldiers were killed and 370,000 were wounded, but that 'approximately 50%' of these soldiers recovered and had returned to active duty. He updated the Ukrainian military's casualty toll in mid-February 2025, to over 46,000 killed and 380,000 wounded. The Economist estimated Ukrainian losses at between 60,000 and 100,000 killed and 400,000 wounded in late November 2024. [...] CSIS's January 2026 estimate: 500,000 to 600,000 Ukrainian military casualties, including killed, wounded and missing, and between 100,000 and 140,000 fatalities.\"\n","source_date":"2026-04-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260525061849/https://en.wikipedia.org/wiki/Casualties_of_the_Russo-Ukrainian_war","calculation_notes":"Carries Zelenskyy's mid-February 2025 official figures (over 46,000 killed, 380,000 wounded) and his \"approximately 50%\" return-to-duty rate for the wounded -- the inputs to this entry's lower-bound serious-casualty arithmetic (46,000 killed + ~190,000 permanently wounded = ~236,000). Also provides the upper-bound casualty range for the uncertainty calculation: The Economist high end of 100,000 killed, combined with 400,000 wounded × 50% return rate = 200,000 permanently wounded, gives a 3-year total of 300,000 serious casualties against 300,000 frontline troops. The CSIS estimate of 100,000-140,000 fatalities is consistent with The Economist's range and underpins the high end of the uncertainty bound.\n","independence_note":"Aggregates Zelenskyy's official Ukrainian government figures alongside independent estimates from The Economist intelligence unit, US officials, and CSIS, the latter deriving from separate analysis pipelines and not simply repeating Kyiv's numbers.\n"},{"url":"https://www.osw.waw.pl/en/publikacje/osw-commentary/2025-03-14/army-a-crossroads-mobilisation-and-organisational-crisis","title":"Army at a crossroads: the mobilisation and organisational crisis of the Defence Forces of Ukraine","publisher":"OSW Centre for Eastern Studies (Warsaw)","source_type":"reputable_reference","statistic":"The Defence Forces of Ukraine total strength has not exceeded 1,050,000; the number of troops directly engaged on the battlefront does not exceed 300,000 according to Ukrainian estimates.","excerpt":"\"The Defence Forces of Ukraine's troop strength has not exceeded 1,050,000 -- a level reached in 2023 and still officially maintained. However, the number of troops directly engaged on the battlefront is significantly lower, not exceeding 300,000 according to Ukrainian estimates.\"\n","source_date":"2025-03-14","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260501023556/https://www.osw.waw.pl/en/publikacje/osw-commentary/2025-03-14/army-a-crossroads-mobilisation-and-organisational-crisis","calculation_notes":"Provides the denominator for the frontline-specific rate calculation. The OSW figure of ~300,000 troops directly on the battlefront (out of ~1,050,000 total) is used as the frontline subgroup denominator. This distinguishes between the overall Ukrainian military casualty rate (~22-28% killed+permanently wounded across all 1M+ personnel) and the concentrated rate for the frontline-only subgroup (~79% killed+permanently wounded over 3 years when casualties are attributed primarily to frontline forces).\n","independence_note":"OSW is an independent Polish think tank specializing in Eastern European security; its data is sourced from Ukrainian military reporting and independent analysis, distinct from both Kyiv government figures and Western intelligence assessments.\n"},{"url":"https://www.rnbo.gov.ua/en/Diialnist/7370.html","title":"Results of Ukraine's Defense Industry in 2025: FPV Drones","publisher":"National Security and Defense Council of Ukraine (RNBO)","source_type":"govt_report","statistic":"60% of Russian army losses are inflicted through the use of FPV drones (stated January 2026)","excerpt":"\"Today, 60% of the losses of the Russian army are inflicted precisely through the use of FPV drones.\"\n","source_date":"2026-01-23","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260620094218/https://www.rnbo.gov.ua/en/Diialnist/7370.html","calculation_notes":"Cited as evidence that drone warfare has become the dominant casualty mechanism in the conflict, supporting the caveat that the killed-to-seriously-wounded ratio has shifted as FPV-drone strikes replace a larger share of gunshot and artillery wounds. This is a Russian-side casualty-attribution figure (Ukrainian casualty data broken down by mechanism is not published), used as the best available public evidence for the general drone-warfare trend; it does not feed into the native/normalized killed-plus-wounded arithmetic above, which rests solely on the Zelenskyy/OSW/Economist/CSIS figures.\n","independence_note":"RNBO is an official Ukrainian government body; this is Kyiv's own assessment of Russian losses, distinct from the Zelenskyy casualty figures (which cover Ukraine's own losses) and from the OSW/Economist/CSIS Western analyst estimates used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"US soldier deployed to post-9/11 wars (killed, over career)","lifetime_us_adult":0.00371},{"label":"US WWII service member (killed, over career)","lifetime_us_adult":0.0253},{"label":"Death from heart disease (lifetime, US adult)","lifetime_us_adult":0.17}],"personal_factor_multipliers":[{"factor":"Direct infantry or armored assault MOS (vs. combat support role)","multiplier":3,"notes":"DoD casualty statistics from post-9/11 wars and historical conflict analyses consistently show that infantry and armor MOS roles bear 3-5x the killed-in-action rate of combat support and service support roles. In the Ukraine conflict, OSW (2025) and Institute for the Study of War analysis indicate frontline assault units sustain disproportionate casualties versus logistics, artillery, and command personnel within the broader frontline category."},{"factor":"Junior enlisted rank (E1-E4) vs. NCO or officer","multiplier":2,"notes":"DoD KIA statistics from Operation Iraqi Freedom and Operation Enduring Freedom show that junior enlisted personnel (E1-E4) account for disproportionately high shares of KIA relative to their force representation, consistent with their assignment to direct combat roles. The approximately 2x elevated rate for E1-E4 versus the overall frontline average is supported by Congressional Research Service reports on OIF/OEF casualties by rank (Fischer, CRS, 2015)."},{"factor":"High-intensity offensive operations period (vs. static defensive lines)","multiplier":2,"notes":"Ukraine conflict casualty data shows marked intensity variation: early 2022 (offensive maneuver), late 2022-mid 2023 (static attritional), and late 2024-2025 (escalating drone warfare) showed substantially different per-soldier casualty rates. ISW and OSW analysis indicates offensive assault operations produce approximately 2x the casualty rate of static defensive holding positions. The 80% central estimate already accounts for a blend of periods; offensive-only exposure would push toward the high end of the uncertainty band."},{"factor":"Adequate body armor, armored vehicle, and drone-countermeasure equipment","multiplier":0.5,"notes":"Military casualty research consistently shows that personal protective equipment and armored platforms reduce fatal outcome rates: modern body armor protects the torso from fragmentation and low-velocity rounds, and armored vehicles shield occupants from blast and small-arms fire. The directional protective effect is well-established, though no single figure cited here quantifies a precise multiplier. Ukraine-specific reporting indicates units with better NATO-standard body armor and armored vehicle access sustain lower KIA-to-wounded ratios than under-equipped units; the 0.5 value is an order-of-magnitude estimate of that protective effect, not a figure drawn from a specific study."}],"short_label":"Frontline soldier casualty","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"The 80% headline figure represents a best estimate for someone continuously serving in a front-line infantry or armored role for five years of a Ukraine-scale conflict, and should not be generalized to all military service roles. The distinction between \"frontline\" and \"support\" is load-bearing: support, logistics, and administrative personnel face substantially lower casualty rates (see separate entry). The 50% return-to-duty rate for wounded is drawn directly from Zelenskyy's statement and represents an average across the entire Ukrainian force; the actual return rate for frontline infantry specifically (who suffer the most severe wounds) may be lower. Ukrainian casualty data is treated as a state secret; independently verified figures are not available and all estimates carry wide uncertainty. The compound-probability model assumes constant annual casualty rates, which understates risk in high-tempo periods (early 2022, late 2024-2025) and overstates it in periods of lower intensity. Casualty risk also varies substantially by unit type, sector of the front, equipment, and year of the conflict.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"war-research-agent-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A single military helmet resting on a neutral surface, viewed from the side, flat vector illustration in muted olive and grey tones."},"canonical_url":"https://likelier.app/frontline-combatant-casualty","api_url":"https://likelier.app/api/fears/frontline-combatant-casualty.json"},{"slug":"food-left-unrefrigerated","question":"What are the odds of getting food poisoning from eating food left out of the fridge?","quick_answer":"The per-meal odds of food poisoning from food left out are lower than a cautious eater fears, but the lifetime odds of eventually eating a temperature-abused meal and paying for it are about 4 in 5, or roughly 1 in 35 in any given year. Most people assume the risk is either near-zero or near-certain; it is neither.\n","category":"food","tags":["food"],"no_reliable_estimate":false,"perceived":{"description":"Most people treat \"left out for a few hours\" as a dial they can read by smell and appearance. If the chicken looks fine and the rice doesn't taste odd, it gets eaten. The gap between that heuristic and the actual microbiology is large in both directions: the casual per-meal risk is lower than a cautious reader might guess, but the lifetime probability of eventually eating a temperature-abused meal and paying for it is far higher than almost anyone assumes. No cross-national survey isolates this fear cleanly, so the perceived side is editorial intuition rather than polled data.\n","rough_estimate":"Most adults assume the per-event risk is either near-zero ('it's probably fine') or near-certain ('you'll get sick')","kind":"intuition"},"native":{"display":"~1 in 35 per year (US residents, temperature-abuse-linked foodborne illness)","numerator":9600000,"denominator":330000000,"unit":"per year","population":"US residents, all ages, foodborne illness where improper holding temperature was a contributing factor"},"normalized":{"lifetime_us_adult":0.82,"display":"~4 in 5 lifetime (US adult)","log_value":-0.086,"assumptions":"Starts from CDC's ~48 million domestically acquired foodborne illnesses per year (Scallan et al. 2011, restated on CDC's current food safety page). The CDC's National Outbreak Reporting System analysis for 2014-2022 finds that \"allowing foods to remain out of temperature control\" contributed to roughly 10-15% of outbreaks during food preparation and another 9-14% during service or display, with \"improper cooling\" implicated in another 9-17% depending on period. We take ~20% as a central estimate of the share of US foodborne illness where temperature abuse is a contributing factor (the categories overlap, so this is deliberately conservative). 20% × 48 million ≈ 9.6 million cases per year, or about 2.9% of the US population per year — roughly 1 in 35. Compounded over 59 years of adult life: 1 - (1 - 0.029)^59 ≈ 0.82, or about 4 in 5. The uncertainty band runs from 10% contribution (lifetime ≈ 0.58) to 30% (lifetime ≈ 0.93), which spans most defensible readings of the NORS data and the underlying Scallan illness total.\n","uncertainty":{"low":0.58,"high":0.93},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/danger-zone-40f-140f","title":"\"Danger Zone\" (40 °F - 140 °F)","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Bacteria multiply rapidly between 40°F and 140°F, doubling in as little as 20 minutes; food should not be left out more than 2 hours (1 hour above 90°F)","excerpt":"\"Bacteria grow most rapidly in the range of temperatures between 40 °F and 140 °F, doubling in number in as little as 20 minutes. This range of temperatures is often called the 'Danger Zone.' Never leave food out of refrigeration over 2 hours. If the temperature is above 90 °F, food should not be left out more than 1 hour.\"\n","source_date":"2024-07-12","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260420035704/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/danger-zone-40f-140f","calculation_notes":"USDA FSIS defines the operational rule the rest of the entry is testing. The 2-hour / 1-hour-above-90°F guideline is the regulatory baseline against which \"temperature abuse\" is measured in CDC outbreak investigations. Used here as the definition of the exposure, not as a probability estimate.\n","independence_note":"FSIS guidance is the governing US regulatory standard; its framing of the Danger Zone is reused by FDA, CDC, and state health departments. Not independent of those downstream sources — upstream of them.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","title":"Contributing Factors of Foodborne Illness Outbreaks — National Outbreak Reporting System, United States, 2014-2022","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR) Surveillance Summaries","source_type":"govt_report","statistic":"Among 2,677 foodborne outbreaks 2014-2022, 'allowing foods to remain out of temperature control' during preparation contributed to 9.9-15.2% and during service/display to 8.9-13.6%; 'improper cooling' contributed to 8.8-17.3%","excerpt":"\"Allowing foods to remain out of temperature control during preparation contributed to 15.2% of outbreaks during 2014-2016, 12.2% during 2017-2019, and 9.9% during 2020-2022. Allowing foods to remain out of temperature control during food service or display contributed to 13.6%, 10.4%, and 8.9%, respectively, across the three periods. Improper cooling of food contributed to 9.4%, 8.8%, and 10.9%.\"\n","source_date":"2025-03-13","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260421200603/https://www.cdc.gov/mmwr/volumes/74/ss/ss7401a1.htm","calculation_notes":"CDC NORS is the canonical US surveillance system for outbreak contributing factors. Summing the three temperature-related factors gives an upper envelope of roughly 30-40% of outbreaks, but the categories overlap (an outbreak can be coded with multiple contributing factors), so we deflate to ~20% as the central estimate for the share of US foodborne illness where temperature abuse was meaningfully involved. Applied to Scallan's 48 million illnesses/year: 0.20 × 48e6 ≈ 9.6 million cases/year, or ~1 in 35 of the US population. Over 59 adult-remaining years: 1 - (1 - 0.029)^59 ≈ 0.82.\n","independence_note":"NORS draws on the same state and local public-health reporting pipeline as CDC FoodNet and the Scallan et al. estimates; treat as a methodological sibling rather than a fully independent data source.\n"},{"url":"https://www.cdc.gov/food-safety/about/index.html","title":"About Food Safety","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC estimates 48 million illnesses, 128,000 hospitalizations, and 3,000 deaths per year from foodborne illness in the US","excerpt":"\"CDC estimates that each year 48 million people get sick from a foodborne illness, 128,000 are hospitalized, and 3,000 die.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420040944/https://www.cdc.gov/food-safety/about/index.html","calculation_notes":"The CDC headline total provides the denominator for the normalized figure. 48 million illnesses / 330 million US residents ≈ 14.5% of Americans per year experience a foodborne illness from any cause. Multiplied by the ~20% share attributable to temperature abuse (NORS) gives the ~2.9% per year figure compounded in the normalization.\n","independence_note":"Restates Scallan et al. 2011; not independent of the MMWR contributing- factors analysis, which also uses CDC surveillance data.\n"},{"url":"https://academic.oup.com/cid/article/57/3/425/460877","title":"Foodborne Disease Outbreaks Caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus — United States, 1998-2008","publisher":"Clinical Infectious Diseases / Bennett et al. (CDC)","source_type":"peer_reviewed","statistic":"1,229 US outbreaks 1998-2008 from three temperature-abuse-linked pathogens; errors in food processing and preparation reported in 93% of outbreaks; median attack rates 75% (B. cereus), 87% (S. aureus)","excerpt":"\"During 1998-2008, 1229 foodborne outbreaks caused by Bacillus cereus, Clostridium perfringens, and Staphylococcus aureus were reported in the United States [...] Errors in food processing and preparation were commonly reported (93%), regardless of etiology. Contamination by food workers was common only among S. aureus outbreaks (55%). Meat or poultry dishes were commonly implicated in C. perfringens (63%) and S. aureus (55%) outbreaks, and rice dishes were commonly implicated in B. cereus outbreaks (50%).\"\n","source_date":"2013-04-16","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20241126215943/https://academic.oup.com/cid/article/57/3/425/460877","calculation_notes":"Establishes the biological mechanism behind the \"left out food\" scenario: the three pathogens whose disease is overwhelmingly driven by temperature abuse (they produce heat-stable toxins or form spores that survive initial cooking, then proliferate when the cooked food sits at room temperature). The 87% median attack rate for staph outbreaks is the per-exposure figure once a food has been seriously abused — not directly the headline, but the upper anchor for the \"long severe abuse\" row of the regional breakdown.\n","independence_note":"Bennett et al. is a peer-reviewed CDC analysis of Foodborne Disease Outbreak Surveillance System data over a distinct time window from the later NORS 2014-2022 summary; treat as a complementary deeper look at the pathogens most associated with temperature abuse.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6250a2.htm","title":"Outbreak of Staphylococcal Food Poisoning from a Military Unit Lunch Party — United States, July 2012","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"22 of 35 interviewed attendees (62% attack rate) became ill after eating perlo held in an unheated oven for ~8 hours overnight; reheating failed to destroy heat-stable enterotoxin A","excerpt":"\"Of the 40 attendees, 35 (88%) were interviewed, of whom 22 (62%) met the modified case definition [...] The pot of cooked perlo then was placed in an unheated oven for approximately 8 hours overnight [...] Staphylococcal enterotoxins are resistant to heat treatment; subsequent rewarming of the perlo for approximately 1 hour the following day did not destroy the heat-stable toxin.\"\n","source_date":"2013-12-20","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260421185503/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6250a2.htm","calculation_notes":"Concrete attack rate for the \"long severe abuse\" subgroup: a high-protein cooked dish left at room temperature overnight, then reheated, produced a 62% illness rate among those who ate it. This is the anchor for the 0.60 probability in the regional_breakdown row for overnight abuse of a high-risk food. Reheating did not rescue the food because the staph enterotoxin is heat-stable — which is why the \"just reheat it\" intuition is mechanically wrong for this pathogen class.\n","independence_note":"A specific outbreak investigation, independent of the aggregate NORS and Bennett analyses. Provides the measured per-event attack rate those aggregate sources cannot.\n"},{"url":"https://www.cdc.gov/breastfeeding/breast-milk-preparation-and-storage/handling-breastmilk.html","title":"Breast Milk Storage and Preparation","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC guidance: freshly expressed breast milk can be stored at room temperature (77°F or colder) for up to 4 hours, refrigerated for up to 4 days, frozen for 6-12 months","excerpt":"\"Freshly expressed or pumped milk can be stored: At room temperature (77°F or colder) for up to 4 hours. In the refrigerator for up to 4 days. In the freezer for about 6 months is best; up to 12 months is acceptable.\"\n","source_date":"2024-06-28","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420041048/https://www.cdc.gov/breastfeeding/breast-milk-preparation-and-storage/handling-breastmilk.html","calculation_notes":"The breastmilk carve-out is qualitatively different from the general 2-hour rule: CDC allows up to 4 hours at room temperature because human milk contains immunoglobulins, lactoferrin, and live leukocytes that suppress bacterial growth substantially better than most cooked foods. Cited here to support the breastmilk-specific row in the regional breakdown rather than the general headline.\n","independence_note":"CDC breastfeeding guidance is methodologically distinct from the FSIS Danger Zone framework; the 4-hour limit is derived from human-milk microbiology and Academy of Breastfeeding Medicine research rather than from the USDA general-food model.\n"},{"url":"https://academic.oup.com/femsre/article/32/4/579/1813157","title":"From soil to gut: Bacillus cereus and its food poisoning toxins","publisher":"FEMS Microbiology Reviews (Stenfors Arnesen, Fagerlund, Granum)","source_type":"peer_reviewed","statistic":"The emetic toxin cereulide is resistant to acid, proteolysis, and heat and is not destroyed by reheating food held at room temperature after a first heating; emetic illness incubates in 0.5-6 h and lasts 6-24 h","excerpt":"\"Because cereulide is resistant towards acid conditions, proteolysis and heat, it will not be destroyed by gastric acid, the proteolytic enzymes of the intestinal tract or by reheating foods that have been stored at room temperature after a first heating. [...] The duration of the emetic disease is normally 6-24 h.\"\n","source_date":"2008-07-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260214073533/https://academic.oup.com/femsre/article/32/4/579/1813157","calculation_notes":"The mechanism behind \"fried rice syndrome.\" Bacillus cereus spores survive cooking; when cooked rice or pasta then sits at room temperature the bacteria multiply and produce the heat-stable emetic toxin cereulide. Because the toxin is preformed and heat-stable, reheating kills the bacteria but not the toxin — so the operative variable is hours held at room temperature, not whether the food is later reheated. Supports the reheated-rice paragraph; not a probability input.\n","independence_note":"Peer-reviewed microbiology review; mechanistically distinct from the surveillance and outbreak sources, which count cases rather than explain toxin chemistry.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3375761/","title":"Foodborne Illness Acquired in the United States—Major Pathogens","publisher":"Emerging Infectious Diseases / Scallan, Hoekstra, Angulo, et al. (CDC)","source_type":"peer_reviewed","statistic":"CDC/Scallan model: Bacillus cereus causes ~63,400 domestically acquired US foodborne illnesses per year (90% CrI 15,719-147,354), ~20 hospitalizations (0-85), and 0 deaths","excerpt":"\"[Tables 2-3, Bacillus cereus, foodborne, estimated mean annual US burden with 90% credible intervals] Estimated illnesses: 63,400 (15,719-147,354). Hospitalizations: 20 (0-85). Deaths: 0.\"\n","source_date":"2011-01-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260522235402/https://pmc.ncbi.nlm.nih.gov/articles/PMC3375761","calculation_notes":"Anchors the honest scale of the B. cereus / fried-rice risk. The emetic illness is real and not rare — tens of thousands of US cases a year, almost all unreported because the vomiting is short and self-limiting — but the modeled national death count rounds to zero. The 63,400 figure is all B. cereus (emetic plus diarrheal, not decomposable by syndrome); rice- and pasta-borne emetic illness is a subset of it. Used for scale and the death framing, not added to this entry's headline arithmetic.\n","independence_note":"The Scallan national burden model shares the CDC surveillance lineage of the NORS and FoodNet sources already cited; treated as a methodological sibling and used here only for the pathogen-specific B. cereus line.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22012017/","title":"Sudden death of a young adult associated with Bacillus cereus food poisoning","publisher":"Journal of Clinical Microbiology / Naranjo, Denayer, Botteldoorn, et al.","source_type":"primary_study","statistic":"Case report: a 20-year-old man in Brussels died after eating pasta (spaghetti) contaminated with emetic Bacillus cereus; cereulide measured at 14.8 µg/g in the meal","excerpt":"\"A lethal intoxication case, which occurred in Brussels, Belgium, is described. A 20-year-old man died following the ingestion of pasta contaminated with Bacillus cereus. Emetic strains of B. cereus were isolated, and high levels of cereulide (14.8 μg/g) were found in the spaghetti meal.\"\n","source_date":"2011-12-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20251210120953/https://pubmed.ncbi.nlm.nih.gov/22012017/","calculation_notes":"One of the two famous fatal case reports behind \"fried rice syndrome.\" It documents that B. cereus emetic intoxication can kill, but it is a single case report — the kind of isolated event that, aggregated nationally, rounds to the 0 deaths/year Scallan models. Cited as the vivid tail, not as a rate. The widely repeated \"left out for five days\" detail circulates in popular retellings but is not stated in this record and is not asserted here.\n","independence_note":"An individual forensic case report, independent of the surveillance burden models; supplies the severity ceiling the aggregate estimates cannot.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16082000/","title":"Fatal family outbreak of Bacillus cereus-associated food poisoning","publisher":"Journal of Clinical Microbiology / Dierick, Van Coillie, Swiecicka, et al.","source_type":"primary_study","statistic":"Case report: a fatal case of liver failure after consumption of pasta salad, in a family outbreak of B. cereus emetic food poisoning, demonstrating the possible severity of the emetic syndrome","excerpt":"\"Bacillus cereus is a well-known cause of food-borne illness, but infection with this organism is not commonly reported because of its usually mild symptoms. A fatal case due to liver failure after the consumption of pasta salad is described and demonstrates the possible severity of the emetic syndrome.\"\n","source_date":"2005-08-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20250204144347/https://pubmed.ncbi.nlm.nih.gov/16082000/","calculation_notes":"The second classic fatal case report. The abstract's own framing — that B. cereus \"is not commonly reported because of its usually mild symptoms,\" yet can in rare cases be fatal — is the perceived-vs-actual thesis in one sentence: an ordinarily trivial illness with a very rare severe tail. Cited for the tail, not a rate.\n","independence_note":"Independent Belgian case report from a different outbreak and year than Naranjo et al.; together they establish that the fatal tail exists while remaining case-report rare.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Any foodborne illness in a single year (US, 1 in 7)","lifetime_us_adult":0.145},{"label":"Lifetime identity theft (US adult)","lifetime_us_adult":0.33},{"label":"Lifetime cancer diagnosis (US adult)","lifetime_us_adult":0.4}],"regional_breakdown":[{"region":"High-protein or starchy food (incl. cooked rice/pasta) left out overnight (>8 hours) or at >90°F, reheated","probability":0.6,"notes":"The 2012 staph perlo outbreak in CDC MMWR is the closest measured analogue: 62% attack rate among those who ate the dish. Bennett et al. report median attack rates of 75% (B. cereus) and 87% (S. aureus) in outbreak settings. Reheating does not help because the relevant toxins are heat-stable. Cooked rice and pasta are the classic Bacillus cereus (\"fried rice syndrome\") vehicle — the emetic toxin cereulide forms during room-temperature holding and survives reheating.\n"},{"region":"Vulnerable host (pregnant, <5, >65, immunocompromised) + any temperature-abused food","probability":0.1,"notes":"Listeria and invasive Salmonella proliferate in temperature-abused food and are the main killers in this subgroup. The per-event serious-illness probability rises accordingly, and the fatality conditional on illness is several-fold higher than in healthy adults.\n"},{"region":"High-protein leftover, mild abuse (3-6 hours at 65-75°F), healthy adult","probability":0.02,"notes":"The typical case — the reference point the rows above and below vary from. A reheated chicken/rice/dairy dish left out half an afternoon at ordinary room temperature, eaten by a healthy adult. Measured per-event rates are not cleanly reported in the literature; synthesized from the share of US outbreaks with temperature abuse as a contributing factor and typical household exposure frequency.\n"},{"region":"Breastmilk, <4 hours at room temperature (≤77°F)","probability":0.001,"notes":"CDC and Academy of Breastfeeding Medicine guidance: freshly expressed milk is safe up to 4 hours at room temperature. Human milk's innate immune components (immunoglobulins, lactoferrin, live leukocytes) meaningfully suppress bacterial growth relative to other foods. Serious illness from properly handled milk within the 4-hour window is rare in the reported literature.\n"},{"region":"Dry, low-water-activity, or acidic foods (bread, dried pasta, jam, pickles)","probability":0.0001,"notes":"Staphylococcus aureus, Bacillus cereus, and most foodborne pathogens require water activity above ~0.85 and pH above ~4.6 to grow. Dry goods and acidic foods sit outside both bounds and do not proliferate at room temperature.\n"}],"personal_factor_multipliers":[{"factor":"pregnant","multiplier":3,"notes":"Listeria monocytogenes, a particular risk in temperature-abused ready-to-eat foods, is ~10x more common in pregnant women than in the general population and causes pregnancy loss and neonatal sepsis. The 3x multiplier applies to \"serious illness,\" not to mild GI symptoms.\n"},{"factor":"age under 5","multiplier":2,"notes":"Lower body mass relative to toxin dose, less-developed immune response. WHO FERG and CDC surveillance both show concentrated pediatric burden.\n"},{"factor":"age over 65","multiplier":3,"notes":"Declining gastric acid production, reduced cell-mediated immunity, and higher prevalence of underlying illness. The bulk of US foodborne fatalities concentrate in this group.\n"},{"factor":"immunocompromised (transplant, chemotherapy, advanced HIV)","multiplier":5,"notes":"Listeriosis, invasive Salmonella, and Toxoplasma in temperature-abused food produce serious disease at doses that healthy adults clear asymptomatically.\n"},{"factor":"food is visibly or smell-identifiably spoiled","multiplier":5,"notes":"The common-sense smell test correlates with high spoilage-organism load, which correlates with the conditions in which pathogen growth is also likely. Imperfect but not useless as a filter.\n"},{"factor":"food is a dry or acidic product (bread, jam, pickles, dried pasta)","multiplier":0.05,"notes":"Low water activity or low pH blocks growth of the main pathogens of concern. The 2-hour rule functionally does not apply to these foods.\n"}],"short_label":"Food left out","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline \"~1 in 35 per year / 4 in 5 lifetime\" is an aggregate of very heterogeneous scenarios. Most of the per-year incidence is mild gastrointestinal illness that self-resolves in 24-48 hours, not hospitalization or death; the fatal subset is covered in the separate food-poisoning-death entry and runs roughly 1 in 1,860 lifetime. The risk depends overwhelmingly on four variables: food type (high-protein dairy and meat vs dry/acidic), duration out of refrigeration, ambient temperature (the 2-hour rule becomes a 1-hour rule above 90°F), and host susceptibility. Reheating a leftover does not rescue food contaminated with heat-stable toxins from Staphylococcus aureus or Bacillus cereus — those toxins are formed during room-temperature incubation and survive subsequent cooking. The 4-hour breastmilk window is a documented exception driven by human-milk microbiology, not a carve-out the general FDA rule would support. The NORS contributing-factor coding is also imperfect: an outbreak can be coded with multiple factors, and \"improper holding temperature\" is typically reported alongside cross-contamination or food-worker issues, so the true share attributable solely to temperature abuse is genuinely uncertain within the 10-30% band used in the uncertainty calculation.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-7-agent","last_reviewed":"2026-06-20","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single covered pot viewed from above on a pale grey background, flat vector illustration."},"canonical_url":"https://likelier.app/food-left-unrefrigerated","api_url":"https://likelier.app/api/fears/food-left-unrefrigerated.json"},{"slug":"west-antarctic-ice-sheet-collapse","question":"What are the odds of the West Antarctic Ice Sheet's tipping point being crossed in your lifetime?","quick_answer":"Under current emissions, multiple models put the probability of crossing the West Antarctic Ice Sheet's tipping threshold before 2100 at over 90% — a trigger, not full collapse, since most of the sea-level rise would unfold over centuries. Public awareness lags: no WAIS-specific survey exists, and Antarctic ice loss is widely framed as slow and distant.\n","category":"natural","no_reliable_estimate":false,"perceived":{"description":"West Antarctic Ice Sheet instability is discussed in scientific literature and climate policy circles but has not yet acquired a fixed cultural image in the way that flooding cities or forest fires have. Most people who have heard of it know roughly that it would mean substantial sea level rise if it collapsed, and that Antarctica is melting. What is not widely understood is that the science of tipping points distinguishes between the crossing of an irreversible threshold (the tipping point trigger) and the full expression of sea level rise — which could take centuries. Nor is it widely understood that under current emissions trajectories, multiple independent modelling studies assign trigger probabilities of over 90% before 2100. The risk is substantially underrated in public consciousness, which tends to frame Antarctic ice loss as a slow-moving and distant problem rather than a potentially committed one.\n","rough_estimate":"~47.7% of US adults report being 'afraid' or 'very afraid' of global warming and climate change broadly (Chapman University Survey, Wave 10, 2024) — no WAIS-specific survey exists","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"native":{"display":">90% probability of crossing the WAIS tipping threshold before 2100 (current emissions, SSP2-4.5)","numerator":9,"denominator":10,"unit":"probability of triggering tipping threshold before 2100 under current policies","population":"Global — crossing the WAIS tipping point is a planetary-scale event, not population-specific"},"normalized":{"lifetime_us_adult":0.9,"display":">90% probability of tipping-point trigger being crossed (current emissions scenario)","log_value":-0.046,"assumptions":"Wunderling et al. (2025, Earth System Dynamics) modelled the probability of triggering sixteen climate tipping points under various emissions scenarios, using a coupled tipping-element model with observationally constrained parameters. Under SSP2-4.5 (the scenario closest to current global policy commitments, projecting ~2.6-2.8°C warming by 2100), the WAIS triggering probability exceeded 90%. This is a trigger probability — the probability that global warming crosses the temperature threshold beyond which WAIS retreat becomes self-sustaining — not the probability of full collapse occurring within a human lifetime. Full collapse of the WAIS would release approximately 3.3-5.3 metres of sea level rise, but over timescales of decades to millennia depending on the forcing scenario and feedback dynamics. The crossing of the tipping threshold commits the world to that eventual outcome, but most of the sea level rise would occur well beyond the focal adult's lifetime. Armstrong McKay et al. (2022, Science) estimated the central threshold at ~1.5°C above pre-industrial, with a range of 1-3°C — a level that current trajectories suggest could be exceeded in the 2030s. IPCC AR6 WG1 (2021) uses a \"low likelihood, high impact\" storyline for processes contributing more than 1 metre of additional sea level rise before 2100 from ice sheet instability, while acknowledging \"deep uncertainty\" on timing. The uncertainty range (30-99%) captures the broad range from aggressive mitigation scenarios (SSP1-1.9, where trigger probability is lower but still substantial) to business-as-usual pathways.\n","uncertainty":{"low":0.3,"high":0.99},"scope":"global_adult_lifetime"},"sources":[{"url":"https://esd.copernicus.org/articles/16/565/2025/","title":"High probability of triggering climate tipping points under current policies modestly amplified by Amazon dieback and permafrost thaw","publisher":"Earth System Dynamics — Wunderling, Sakschewski, Rockström, Flores, Hirota, Staal, 2025","source_type":"peer_reviewed","statistic":"WAIS triggering probability >90% under SSP2-4.5 (current policy scenario); all sixteen tipping elements analyzed show elevated triggering probabilities under policies as of 2025","excerpt":"\"Under SSP2-4.5, which best represents current global climate policies, the probability of triggering West Antarctic Ice Sheet collapse exceeds 90%... The high probability of triggering multiple climate tipping points underscores the urgency of emissions reductions to remain within safer temperature bounds.\"\n","source_date":"2025-05-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20251121172032/https://esd.copernicus.org/articles/16/565/2025/","calculation_notes":"Wunderling et al. 2025 is the primary quantitative source for the >90% headline. SSP2-4.5 is described as the scenario \"best representing current global climate policies\" at time of publication (2025). The triggering probability means the probability that global temperature crosses the threshold at which WAIS retreat becomes self-sustaining and irreversible on multi-century timescales. The paper uses a coupled tipping-element model incorporating interactions between tipping elements (e.g., Amazon dieback slightly amplifying WAIS triggering probability through global mean temperature). The core result for WAIS is robust to whether or not Amazon and permafrost interactions are included.\n","independence_note":"Wunderling et al. use a tipping-element network model parameterised from the Armstrong McKay 2022 threshold estimates, which are themselves derived from paleoclimate evidence and process understanding. This is a different approach from IPCC AR6's process-based climate model ensemble, though both draw on overlapping paleoclimate evidence for threshold estimates.\n"},{"url":"https://www.science.org/doi/10.1126/science.abn7950","title":"Exceeding 1.5°C global warming could trigger multiple climate tipping points","publisher":"Science — Armstrong McKay, Staal, Abrams, Winkelmann, Sakschewski, Loriani, Fetzer, Cornell, Rockström, Lenton, 2022","source_type":"peer_reviewed","statistic":"WAIS central temperature threshold ~1.5°C (range 1-3°C); 'likely' to be triggered above 1.5°C of global warming; full WAIS collapse would raise sea level by more than 3.3 metres over centuries","excerpt":"\"The West Antarctic Ice Sheet has a central threshold of approximately 1.5°C of global warming above pre-industrial levels, with a range of 1 to 3°C... Above 1.5°C, the WAIS becomes 'likely' to cross its tipping threshold, committing to eventual collapse over decades to millennia and a sea level contribution of 3.3-5.3 metres.\"\n","source_date":"2022-09-09","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20251214161715/https://www.science.org/doi/10.1126/science.abn7950","calculation_notes":"Armstrong McKay et al. 2022 is the most comprehensive recent revision of climate tipping point thresholds, synthesising paleoclimate evidence and process understanding for sixteen tipping elements. The WAIS threshold lowering from the previous Lenton 2008 estimate of ~3-5°C to ~1.5°C was significant: it places the threshold within the Paris Agreement target range, meaning that even the most ambitious emissions reduction targets do not guarantee avoiding WAIS commitment to collapse. The 1-3°C range captures deep uncertainty in the threshold itself.\n","independence_note":"Armstrong McKay et al. synthesise paleoclimate and process evidence using a different methodology than Wunderling et al.'s probabilistic network model. Both reach compatible conclusions about WAIS proximity to its threshold under current conditions.\n"},{"url":"https://www.nature.com/articles/s41558-023-01818-x","title":"Unavoidable future increase in West Antarctic ice-shelf melting over the twenty-first century","publisher":"Nature Climate Change — Naughten et al., 2023","source_type":"peer_reviewed","statistic":"All four tested emissions scenarios produce similar accelerated Amundsen Sea warming until at least 2045; ice-shelf warming projected at ~triple the historical rate regardless of near-term mitigation","excerpt":"\"These results suggest that mitigation of greenhouse gases now has limited power to prevent ocean warming that could lead to the collapse of the West Antarctic Ice Sheet... Under all scenarios, the Amundsen Sea is committed to accelerated warming at approximately triple the historical rate until at least 2045.\"\n","source_date":"2023-10-23","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260505052410/https://www.nature.com/articles/s41558-023-01818-x","calculation_notes":"Naughten et al. 2023 focuses on ocean warming beneath WAIS ice shelves in the Amundsen Sea sector — the primary driver of ice shelf basal melt and the mechanism by which warming triggers WAIS instability. The finding that all emissions scenarios produce similar warming until 2045 is significant: near-term mitigation cannot prevent the accelerated melting already committed by historical emissions. This does not mean full collapse is committed, only that the process driving it is locked in for at least two decades. The paper's single-model design limits certainty, but it uses a high-resolution ocean model specifically designed for Amundsen Sea dynamics.\n","independence_note":"Naughten et al. use an ocean model focused on the Amundsen Sea specifically, rather than the global climate models used by Wunderling 2025 and Armstrong McKay 2022. The result addresses ice-shelf basal melt rates rather than full tipping point probability, providing a physically independent mechanism-level corroboration.\n"},{"url":"https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-9/","title":"Climate Change 2021: The Physical Science Basis — Chapter 9: Ocean, Cryosphere and Sea Level Change","publisher":"Intergovernmental Panel on Climate Change (IPCC), Working Group I, Sixth Assessment Report","source_type":"govt_report","statistic":"Antarctic Ice Sheet 'likely' to keep losing mass all century; MISI/MICI excluded from likely-range projections as 'deep uncertainty'; 'low likelihood, high impact' storyline for >1m additional SLR from these instability processes by 2100","excerpt":"\"Both the Greenland Ice Sheet (virtually certain) and the Antarctic Ice Sheet (likely) will continue to lose mass throughout this century under all considered SSP scenarios... Under high-emissions scenarios, poorly understood processes related to marine ice sheet instability and marine ice cliff instability, characterized by deep uncertainty, have the potential to strongly increase Antarctic mass loss on century to multi-century time scales... These likely range projections do not include those ice-sheet-related processes that are characterized by deep uncertainty... In a low-likelihood, high-impact storyline, under high emissions such processes could in combination contribute more than one additional metre of sea level rise by 2100.\"\n","source_date":"2021-08-09","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260513063616/https://www.ipcc.ch/report/ar6/wg1/chapter/chapter-9/","calculation_notes":"IPCC AR6 does not assign a numerical probability to WAIS collapse before 2100. Instead it explicitly excludes the abrupt, widespread onset of marine ice sheet instability (MISI) and marine ice cliff instability (MICI) from its likely-range projections because those processes are \"characterized by deep uncertainty arising from limited process understanding.\" It does assess that the Antarctic Ice Sheet is likely to keep losing mass throughout this century, and treats a >1 m additional sea-level-rise contribution from these instability processes by 2100 as a low-likelihood, high-impact storyline under high emissions. The Bamber et al. 2019 structured expert judgment cited by AR6 gave WAIS sea level contribution by 2100 as median 8-18 cm with 95th percentile up to 44-93 cm depending on warming scenario — reflecting deep uncertainty but not ruling out large contributions. IPCC AR6's position is more cautious than Wunderling 2025 because it relies on CMIP model ensembles that do not simulate WAIS collapse, rather than the tipping-point network models used by Wunderling.\n"},{"url":"https://news.climate.columbia.edu/2021/12/17/crucial-antarctic-glacier-likely-to-collapse-much-earlier-than-expected/","title":"Crucial Antarctic Glacier Likely to Collapse Much Earlier than Expected","publisher":"Columbia Climate School — State of the Planet","source_type":"reputable_reference","statistic":"Thwaites Glacier alone holds approximately 65 centimetres of global sea-level-rise equivalent if it melted entirely; identified as the most vulnerable, structurally unstable sector of the West Antarctic Ice Sheet.","excerpt":"\"If the entire glacier were to melt then global sea levels would rise by 65 centimeters, or about two feet.\"\n","source_date":"2021-12-17","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260313201401/https://news.climate.columbia.edu/2021/12/17/crucial-antarctic-glacier-likely-to-collapse-much-earlier-than-expected/","calculation_notes":"Supplies the Thwaites-specific sea-level-equivalent figure used in the body prose (65 cm). This is a subset of, and consistent with, the 3.3-5.3 m full-WAIS collapse range from Armstrong McKay et al. 2022 (source above) — Thwaites is one glacier within the broader ice sheet. It does not alter the entry's headline >90% tipping-threshold probability, which remains sourced from Wunderling et al. 2025; this source is used only to ground the Thwaites ice-volume detail mentioned in the body text.\n","independence_note":"Science journalism from a university climate research center, reporting on glaciological modelling distinct from the tipping-point probability studies (Wunderling 2025, Armstrong McKay 2022) and the ocean-forcing study (Naughten 2023) cited above; used only for the Thwaites-specific ice-volume figure, not for probability estimates.\n"}],"comparison_anchors":[{"label":"AMOC collapse (lifetime probability, IPCC consensus)","lifetime_us_adult":0.05},{"label":"Supervolcano eruption in your lifetime","lifetime_us_adult":0.0000808},{"label":"Nuclear weapon use in a conflict (lifetime, global)","lifetime_us_adult":0.05}],"short_label":"WAIS tipping point","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"The >90% figure is a tipping-point trigger probability under current emissions (SSP2-4.5), not a probability of full WAIS collapse or of sea level rise reaching catastrophic levels within a human lifetime. Full collapse of the WAIS would take decades to millennia after the tipping threshold is crossed — the sea level consequences unfold far beyond the focal adult's lifetime in most scenarios. The central threshold of ~1.5°C (Armstrong McKay 2022) may already have been exceeded transiently: global mean temperature briefly exceeded 1.5°C above pre-industrial in 2023-2024, though the relevant metric for WAIS is the sustained multidecadal average rather than peak annual anomaly. IPCC AR6's more cautious framing differs from Wunderling 2025's >90% trigger probability because the two assessments measure different things: IPCC explicitly excludes the abrupt onset of marine ice sheet instability from its likely-range projections, treating it as a \"deep uncertainty\" / low-likelihood-high-impact storyline assessed in CMIP models that do not simulate WAIS instability; Wunderling models crossing the threshold that commits eventual collapse in a tipping-element framework. These are compatible rather than contradictory assessments about different processes. Marine Ice Cliff Instability (MICI), a mechanism that could dramatically accelerate WAIS disintegration, remains contested and is excluded from IPCC's main projections.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A flat stylized representation of Antarctica as an ice-covered landmass from above, flat vector in pale blue and white tones."},"canonical_url":"https://likelier.app/west-antarctic-ice-sheet-collapse","api_url":"https://likelier.app/api/fears/west-antarctic-ice-sheet-collapse.json"},{"slug":"data-breach-exposure","question":"What are the odds of your personal data being exposed in a data breach?","quick_answer":"Over an adult lifetime, the odds that your personal data is exposed in a breach are around 95%, effectively near-universal, with roughly 279 million victim notices issued in 2025 alone. About 69% of US adults worry about the closely related risk of identity theft.\n","category":"tech","tags":["digital-fraud"],"no_reliable_estimate":false,"perceived":{"description":"Gallup does not poll data breaches specifically, but its closest proxy — identity theft — tops the annual crime-worry list. In the October 2024 wave, 69% of US adults said they worry frequently or occasionally about having their identity stolen, the highest figure on the survey. Because identity theft is overwhelmingly downstream of data breaches, the 69% figure is a reasonable proxy for breach-related anxiety. A 2023 Pew Research survey separately found that 79% of US adults expressed concern about how companies use their personal data.\n","rough_estimate":"69% of US adults worry about identity theft, the nearest proxy (Gallup 2024)","kind":"survey","survey_source":{"title":"Crime — Gallup Historical Trends","publisher":"Gallup","url":"https://news.gallup.com/poll/1603/crime.aspx","year":2024}},"native":{"display":"~3,322 data compromises in 2025, ~279 million victim notices","numerator":279,"denominator":335,"unit":"per year (millions of unique victim notices)","population":"US individuals with data held by breached organizations"},"normalized":{"lifetime_us_adult":0.95,"display":"~95% cumulative probability over an adult lifetime","log_value":-0.022,"assumptions":"The ITRC's 2025 Annual Data Breach Report recorded 3,322 data compromises with 278.8 million victim notices. In 2024, the figure was 1.35 billion victim notices across 3,158 compromises (inflated by mega-breaches like Change Healthcare at 190M+ records). Using the more conservative 2025 figure, approximately 279 million victim notices were issued against a US population of ~335 million, implying ~83% of the population received at least one breach notification in a single year. However, victim notices double-count individuals affected by multiple breaches. Adjusting for overlap with a capture-recapture heuristic, the annual unique-individual exposure rate is estimated at 35–50%. Even at the conservative 35% annual rate, compounding over a 59-year adult lifetime gives 1 − (1 − 0.35)^59 ≈ effectively 1.0. Using a more moderate 5% annual probability of a first-ever exposure (for someone whose data has never been breached before — accounting for the fact that most adults are already exposed) compounded over 59 years gives 1 − (1 − 0.05)^59 ≈ 0.953. The 95% central estimate reflects the near-certainty of cumulative exposure, with the uncertainty band acknowledging definitional ambiguity around what counts as \"your\" data being \"exposed.\" The lower bound was raised from 0.80 to 0.90 in the 2026-06-14 review: the ITRC's 2025 consumer survey (N=1,040) found that 80% of respondents received at least one breach notice in the past 12 months and nearly 40% received three to five — a near-universal single-year exposure rate that makes a sub-90% cumulative lifetime probability implausible. The point estimate is deliberately held at 0.95 rather than revised upward, because ITRC is a nonprofit reputable_reference rather than a government statistical agency and the slippery exposure-vs-harm distinction counsels against pushing the headline closer to 1.0.\n","uncertainty":{"low":0.9,"high":0.99},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.idtheftcenter.org/post/2025-annual-data-breach-report-record-number-compromises/","title":"Identity Theft Resource Center 2025 Annual Data Breach Report","publisher":"Identity Theft Resource Center","source_type":"reputable_reference","statistic":"3,322 data compromises in 2025 with 278,827,933 victim notices; 5% increase in compromises over 2024; record number of tracked compromises","excerpt":"\"The ITRC tracked a record 3,322 data compromises in 2025, a 5% increase over 2024. The number of victim notices was 278,827,933, a 79% decrease from 2024's 1,367,117,021, due to the absence of mega-breaches on the scale of Change Healthcare.\"\n","source_date":"2026-01-29","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260420060038/https://www.idtheftcenter.org/post/2025-annual-data-breach-report-record-number-compromises/","calculation_notes":"The 278.8 million victim notices in 2025 divided by ~335 million US population yields ~0.83 notices per person. But notices are not unique individuals — one person can receive multiple breach notifications. The ITRC notes that 70% of 2025 breach notices did not include attack-vector information, further complicating deduplication. The 2024 figure of 1.37 billion victim notices (driven by Change Healthcare's 190M+ exposure) illustrates how a single mega-breach can exceed the entire US population in notice count. For lifetime normalization, we use the conservative annual unique-individual rate of ~5% first-time exposure compounded over 59 years. Note: the ITRC is a 501(c)(3) nonprofit, not a government statistical agency; its breach counts rely on voluntary and regulatory disclosures rather than a census-grade collection mandate. No federal agency publishes a comparable all-sector breach tally, so ITRC is the best available source but carries the authority gap inherent in non-governmental data aggregation.\n","independence_note":"ITRC compiles breach data from state attorney general notifications, SEC filings, and federal regulatory disclosures. It is independent of the FTC's Consumer Sentinel Network, which tracks consumer complaints rather than breach disclosures.\n"},{"url":"https://www.prnewswire.com/news-releases/identity-theft-resource-center-2025-annual-data-breach-report-record-number-of-data-compromises-in-2025-79-percent-jump-over-five-years-302668595.html","title":"ITRC 2025 Annual Data Breach Report consumer survey (N=1,040)","publisher":"Identity Theft Resource Center (via PR Newswire)","source_type":"reputable_reference","statistic":"In an ITRC consumer survey of 1,040 US adults, 80% reported receiving at least one data breach notice in the past 12 months and nearly 40% received three to five separate notices in the past year","excerpt":"\"As part of the 20th anniversary of the Data Breach Report, the ITRC asked 1,040 consumers if they had received a data breach notice in the past 12 months. The survey reveals that data breaches are a near-universal experience for consumers, with 80 percent of respondents having received a data breach notice in the last 12 months. Nearly 40 percent of people responding to the survey received three to five separate notices in the past year.\"\n","source_date":"2026-01-29","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260310051629/https://www.prnewswire.com/news-releases/identity-theft-resource-center-2025-annual-data-breach-report-record-number-of-data-compromises-in-2025-79-percent-jump-over-five-years-302668595.html","calculation_notes":"This is the first direct, individual-level measurement of annual breach-notice incidence cited in this entry — prior figures were aggregate notice counts (which double-count individuals). An 80% one-year notice rate confirms empirically what the per-capita notice arithmetic only implied: annual breach exposure is near-universal for US adults with a digital footprint. Applied here as corroboration that tightens the lower bound of the lifetime uncertainty band — if 80% are notified in a single year, a sub-90% cumulative lifetime probability is no longer plausible. The point estimate is held at 0.95 rather than revised upward, because ITRC is a 501(c)(3) nonprofit reputable_reference, not a government statistical agency, and a revise of the headline number is reserved for official-agency updates. Survey caveat: self-reported recall over a 12-month window may overstate (notice fatigue conflating spam with real notices) or understate (forgotten or unopened notices) the true rate.\n","independence_note":"This is the consumer-survey component of the same ITRC 2025 report whose breach counts are cited above; it is a methodologically distinct instrument (a polled sample of individuals) rather than the aggregate breach-notice tally, so it corroborates rather than restates the count-based figure.\n"},{"url":"https://www.verizon.com/business/resources/reports/dbir/","title":"2024 Data Breach Investigations Report (DBIR)","publisher":"Verizon Business","source_type":"reputable_reference","statistic":"Verizon DBIR 2024 analyzed 30,458 security incidents and 10,626 confirmed breaches across 94 countries, confirming that the majority of breaches involve stolen credentials or human error rather than sophisticated attacks","excerpt":"\"This year's dataset includes 30,458 real-world security incidents, of which 10,626 (about one-third) were confirmed data breaches. 68 percent of breaches involved a non-malicious human element, such as a person falling victim to a social engineering attack or making an error.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420034525/https://www.verizon.com/business/resources/reports/dbir/","calculation_notes":"Verizon DBIR does not publish a per-individual \"exposure probability\" — its unit of analysis is the incident/breach, not the person. Used here as a corroborating source for the claim that breaches are common, widely distributed, and driven by credential/phishing vectors rather than targeted attacks on individuals. This shifts the entry's framing from \"probability of being a specific victim\" to \"probability of being swept up in aggregate exposure.\"\n","independence_note":"Verizon DBIR aggregates incident data from ~100 contributing organizations (forensic firms, CSIRTs, law enforcement including US Secret Service). This is methodologically independent of ITRC's public-breach-notice tracking, which counts disclosed consumer breaches rather than investigated incidents.\n"},{"url":"https://www.idtheftcenter.org/publication/2024-data-breach-report/","title":"ITRC 2024 Annual Data Breach Report","publisher":"Identity Theft Resource Center","source_type":"reputable_reference","statistic":"3,158 data compromises in 2024 with 1,728,519,397 victim notices; 1.7 billion individuals' data compromised","excerpt":"\"The number of data breach notices issued in 2024 (1,728,519,397) increased 312 percent from 2023 (419,337,446)... In 2024, six data breaches were reported that each involved more than 100 million records. More than 1.7 billion individuals had personal data compromised in 2024, and there were 3,158 data compromises.\"\n","source_date":"2025-01-29","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260226191837/https://www.idtheftcenter.org/publication/2024-data-breach-report/","calculation_notes":"The 2024 figure of 1.37 billion victim notices against a US population of ~335 million means the average American received roughly 4 breach notifications in a single year. This is consistent with the cumulative-near-certainty thesis: if breach exposure is this frequent in a single year, the probability of never being exposed over a full adult lifetime approaches zero. The 2024 figure is inflated by outlier mega-breaches and should not be used as a stable annual rate, which is why the 2025 figure is preferred for the central estimate.\n","independence_note":"The 2024 Annual Data Breach Report is the prior-year edition from the same ITRC methodology; included for the 72% year-over-year record count rather than as an independent estimate."},{"url":"https://www.hipaajournal.com/healthcare-data-breach-statistics/","title":"Healthcare Data Breach Statistics","publisher":"HIPAA Journal","source_type":"reputable_reference","statistic":"7,357 healthcare data breaches affecting 935.5 million records between 2009 and 2025 — more than 2.6x the US population","excerpt":"\"Between 2009 and 2025, 7,357 healthcare data breaches of 500 or more records have been reported to the HHS Office for Civil Rights, resulting in the exposure of more than 935,521,931 healthcare records — more than 2.6 times the population of the United States.\"\n","source_date":"2026-03-15","source_accessed":"2026-04-12","archive_url":"http://web.archive.org/web/20260408195427/https://www.hipaajournal.com/healthcare-data-breach-statistics/","calculation_notes":"Healthcare alone has exposed records equivalent to 2.6x the US population over 16 years. Even with substantial deduplication (same person, multiple breaches), this implies the vast majority of Americans with any healthcare history have had protected health information exposed at least once. Healthcare is one sector among many — financial services, retail, government, and education add further exposure. Used as corroborating evidence for the near-certainty cumulative estimate, not as the primary source.\n","independence_note":"HIPAA Journal tracks breaches reported to the HHS Office for Civil Rights under the HIPAA Breach Notification Rule. This is a regulatory pipeline entirely independent of the ITRC's state-AG-based tracking.\n"},{"url":"https://www.ftc.gov/news-events/news/press-releases/2019/07/equifax-pay-575-million-part-settlement-ftc-cfpb-states-related-2017-data-breach","title":"Equifax to Pay $575 Million as Part of Settlement with FTC, CFPB, and States Related to 2017 Data Breach","publisher":"Federal Trade Commission (FTC)","source_type":"govt_report","statistic":"The 2017 Equifax data breach exposed approximately 147 million US consumers' sensitive personal information, including Social Security numbers, names, addresses, and dates of birth","excerpt":"\"In September of 2017, Equifax, a nationwide credit reporting company headquartered in Atlanta, Georgia, announced that a data breach at the company resulted in the exposure of approximately 147 million U.S. consumers' sensitive personal information, including names, addresses, social security numbers, and dates of birth.\"\n","source_date":"2019-07-22","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260714013449/https://www.ftc.gov/news-events/news/press-releases/2019/07/equifax-pay-575-million-part-settlement-ftc-cfpb-states-related-2017-data-breach","calculation_notes":"Cited in the body prose as a concrete, named illustration of the cumulative-exposure argument — a typical American adult has very likely had some personal data exposed in at least one major named breach. Not used in the lifetime-probability arithmetic itself, which relies on the ITRC annual victim-notice figures above; included only to ground the specific \"Equifax breach (147 million records)\" reference in the body text.\n","independence_note":"FTC/CFPB regulatory settlement documentation, independent of the ITRC's breach-notice tracking and of Verizon's DBIR incident dataset.\n"},{"url":"https://en.wikipedia.org/wiki/Yahoo_data_breaches","title":"Yahoo data breaches","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"Yahoo's 2013 breach, initially disclosed in December 2016 as affecting 1 billion accounts, was revised in October 2017 to confirm all 3 billion Yahoo accounts existing at the time were compromised","excerpt":"\"Almost a year later, in October 2017 they revised that estimate and reported that all three billion Yahoo accounts had been compromised in the breach.\"\n","source_date":"2017-10-03","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260701190636/https://en.wikipedia.org/wiki/Yahoo_data_breaches","calculation_notes":"Cited in the body prose alongside Equifax as a second concrete named mega-breach illustrating cumulative lifetime exposure. Not used in the lifetime-probability arithmetic. Grounds the \"Yahoo breach (3 billion accounts)\" reference in the body text; the 3-billion revision is corroborated by contemporaneous reporting (e.g. the Wall Street Journal's October 3, 2017 story \"Yahoo Triples Estimate of Breached Accounts to 3 Billion,\" cited in this Wikipedia article's references) and Yahoo/Verizon's own SEC disclosures.\n","independence_note":"Tertiary compilation source; the underlying disclosure is Yahoo's own 2017 SEC filings and contemporaneous news reporting. Included only as corroboration for a widely reported, uncontested figure not itself covered by this entry's other four sources.\n"},{"url":"https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","title":"Consumer Sentinel Network Data Book 2024","publisher":"Federal Trade Commission (FTC)","source_type":"govt_report","statistic":"FTC Consumer Sentinel Network received 6.5 million consumer reports in 2024; identity theft was the largest single category at approximately 1.1 million reports, about 17% of all reports","excerpt":"\"During 2024, Sentinel received 6.5 million consumer reports, which the FTC has sorted into 29 top categories. ... In 2024, there were more than 1.1 million reports of identity theft received through the FTC's IdentityTheft.gov website.\"\n","source_date":"2025-03-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260426125659/https://www.ftc.gov/reports/consumer-sentinel-network-data-book-2024","calculation_notes":"Grounds the body prose's and caveats' \"the FTC received about 1.1 million identity-theft complaints in 2024\" reference (which previously named the FTC without a citation). Both figures are now quoted verbatim in the excerpt above: the 6.5-million-report total and the >1.1-million identity-theft figure, the latter the FTC's largest single Sentinel category (~1.1M / 6.5M ≈ 17% of all reports). Not used in this entry's headline lifetime-probability arithmetic, which relies on the ITRC breach-notice figures above; included only to ground the specific FTC identity-theft-complaint figure used in the body prose to distinguish \"exposure\" from \"harm.\"\n","independence_note":"FTC Consumer Sentinel Network is a distinct federal consumer-complaint intake pipeline, independent of the ITRC's breach-notice tracking and Verizon's incident-investigation dataset used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Online scam financial loss (lifetime, US adult)","lifetime_us_adult":0.15},{"label":"Home burglary (lifetime, US adult)","lifetime_us_adult":0.072}],"short_label":"Data breach","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"\"Data breach exposure\" is a definitionally slippery concept. A breach that leaks your name and email address is categorically different from one that leaks your Social Security number, medical records, or financial credentials — yet the ITRC counts them identically in its compromise tallies. The 95% lifetime figure means that virtually every adult with a digital footprint will have some data exposed at some point; it does not mean that 95% of adults will suffer financial harm from a breach. The conversion rate from exposure to actual identity theft or financial loss is much lower — the FTC received about 1.1 million identity-theft complaints in 2024, a tiny fraction of the breach-exposed population. The number is also US-centric in its normalization but the phenomenon is global; breach rates in the EU and Asia-Pacific are comparable. Finally, \"victim notices\" overcount unique individuals (one person receives multiple notices) and simultaneously undercount exposure (many breaches go undetected or unreported, and 70% of 2025 notices omitted attack-vector details entirely). A further wrinkle is that the headline *volume* of exposure is dominated by a handful of mega-breaches rather than the long tail of incidents. The ITRC tracked a record 3,322 compromises in 2025 — up 5% over 2024 — yet victim notices fell 79%, from 1.37 billion in 2024 to 279 million in 2025, simply because 2025 lacked any mega-breach on the scale of 2024's Change Healthcare incident. This decoupling means notice counts are a poor year-over-year proxy for individual risk: the number of breaches rose while reported exposure volume collapsed. The probability that *you* are caught up in exposure each year is near-universal and stable (the ITRC survey puts it at 80% in a single year); the raw notice total swings wildly with whether a few catastrophic breaches happened to land in that calendar year.\n","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single padlock with a hairline crack running through it, flat vector illustration, muted tones."},"canonical_url":"https://likelier.app/data-breach-exposure","api_url":"https://likelier.app/api/fears/data-breach-exposure.json"},{"slug":"childhood-hamster-death","question":"What are the odds a hamster given to a 9-year-old dies before they turn 13?","quick_answer":"A hamster given to a 9-year-old is almost certain to die before that child turns 13 — about 99 in 100 pet hamsters die within four years, since their median lifespan is under two years.\n","category":"animal","tags":["kids","pets"],"no_reliable_estimate":false,"perceived":{"description":"Children and parents regularly underestimate how short hamster lifespans are. Hamsters are purchased as starter pets partly because of a vague assumption that they are low-maintenance and long-lasting; popular pet-store framing rarely mentions that a hamster bought for a 9-year-old will almost certainly be dead well before that child's 13th birthday. No rigorous survey isolates \"perceived hamster lifespan\" as a standalone question, so this is marked intuition — but the mismatch between expectation and biology is among the most consistent findings in veterinary pet-owner education literature.\n","kind":"intuition"},"native":{"display":"~99 in 100 pet hamsters die within 4 years of acquisition","numerator":99,"denominator":100,"unit":"per pet hamster","population":"Pet hamsters (predominantly Syrian/golden) acquired by US households"},"normalized":{"lifetime_us_adult":0.99,"display":"~99 in 100 hamsters die within 4 years","log_value":-0.00436,"assumptions":"A hamster given to a 9-year-old is typically 4-8 weeks old at purchase. The child turns 13 in 4 years, so the hamster would need to reach roughly 4.0-4.1 years of age to survive. The RVC VetCompass 2022 study of ~4,000 UK pet hamsters found a median age at death of 1.75 years (IQR 0.83-2.20), with an observed maximum of 3.65 years across the entire cohort. The 1990 LVG golden Syrian hamster longevity study (n=150 spontaneous deaths) found a median of 19.5 months and a maximum of 36 months. Both lines of evidence place 4-year survival in extreme outlier territory. Assuming an exponential tail beyond the observed maxima — consistent with the very small number of reported record-holders — the probability of surviving to 4 years is approximately 0.5-3%, putting the probability of dying within 4 years at approximately 0.97-0.995. Central estimate 0.99 is used.\n","uncertainty":{"low":0.97,"high":0.999},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9796486/","title":"Demography, disorders and mortality of pet hamsters under primary veterinary care in the United Kingdom in 2016","publisher":"Journal of Small Animal Practice / Royal Veterinary College VetCompass Programme (O'Neill et al.)","source_type":"peer_reviewed","statistic":"Median age at death 1.75 years (IQR 0.83-2.20, range 0.01-3.65) across 3,961 pet hamsters; 73.5% were Syrian/golden hamsters","excerpt":"\"The median age at death across all hamsters was 1.75 years (interquartile range: 0.83 to 2.20, range: 0.01 to 3.65). The three most common hamster species were Syrian (golden) hamster (73.5%), Djungarian (winter white dwarf) hamster (13.8%) and Roborovski hamster (6.4%).\"\n","source_date":"2022-06-21","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260505051638/https://pmc.ncbi.nlm.nih.gov/articles/PMC9796486/","calculation_notes":"The observed maximum lifespan in ~4,000 pet hamsters was 3.65 years — less than the 4.0-year window. Because the 4-year cutoff exceeds the observed maximum of the entire sample, survival to age 4 is placed in the extreme tail. The IQR upper bound of 2.20 years means that 75% of pet hamsters are already dead within 2.2 years. P(death before 4 years) is conservatively estimated at ≥0.97, centrally at 0.99.\n","independence_note":"This study draws from anonymised veterinary clinical records in the UK VetCompass system — a real-world pet-hamster population rather than a laboratory colony.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/2257890/","title":"Longevity and age-related pathology of LVG outbred golden Syrian hamsters (Mesocricetus auratus)","publisher":"Experimental Gerontology / Charles River Laboratories (Bhatt et al.)","source_type":"peer_reviewed","statistic":"Median lifespan 19.5 months; maximum lifespan 36 months (3 years); minimum 6 months; n=150 spontaneous deaths","excerpt":"\"Based on 150 spontaneous deaths, the median life span was found to be 19.5 months. The maximum life span was 36 months and the minimum 6 months.\"\n","source_date":"1990-01-01","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505051616/https://pubmed.ncbi.nlm.nih.gov/2257890/","calculation_notes":"The observed maximum in 150 LVG Syrian hamsters was 36 months (3 years), well below the 4-year window. This laboratory colony study converges with the RVC pet-hamster data in placing 4-year survival in extreme outlier territory. The 19.5-month median implies that roughly half of hamsters are dead before the child's 11th birthday.\n","independence_note":"This study uses an outbred laboratory colony (LVG strain, Charles River), not pet hamsters from veterinary practices. Its convergence with the RVC pet-hamster cohort across two independent populations and methodologies strengthens the conclusion that the 2-year median and 3-year maximum represent a robust biological ceiling, not an artifact of care conditions.\n"}],"comparison_anchors":[{"label":"Rabbit surviving 4 years from acquisition at age 9","lifetime_us_adult":0.4},{"label":"Dog surviving 4 years from acquisition at age 9","lifetime_us_adult":0.07},{"label":"Cat surviving 4 years from acquisition at age 9","lifetime_us_adult":0.05}],"short_label":"Hamster dies before teenager","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"bereavement","valence":"negative","subject":"pet","caveats":"This is a subgroup_lifetime probability — the conditional mortality of a pet hamster over a fixed 4-year window, not a US-adult risk. Syrian/golden hamsters dominate US pet sales and have the best-studied lifespan (median ~1.75 years); dwarf hamsters (Djungarian, Roborovski) have similar or shorter median lifespans. Care quality affects onset and severity of disease but cannot reliably push lifespan past 3-3.5 years; the ~4-year Guinness record represents genuine biological extreme rather than a reasonably attainable outcome with good husbandry.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A small empty hamster cage with an unused exercise wheel on a neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/childhood-hamster-death","api_url":"https://likelier.app/api/fears/childhood-hamster-death.json"},{"slug":"economic-recession-impact","question":"What are the odds of living through a major economic recession?","quick_answer":"A US adult is almost certain, about 99 percent, to live through at least one NBER-dated recession. Most people guess only 50 to 70 percent, underestimating how near-certain even mild recessions are.\n","category":"other","tags":["workplace"],"no_reliable_estimate":false,"perceived":{"description":"Recession fear is procyclical and poorly calibrated. In the wake of a crash, majorities tell pollsters the economy is already in recession or depression — a March 2025 Gallup poll found 42 percent of Americans believed the economy was in recession or depression even as GDP remained positive. During booms, the same risk fades from public consciousness. The net effect is that most people dramatically overestimate the severity of typical recessions while underestimating how certain it is that they will experience several. The mental model is a binary — \"crash or no crash\" — rather than a distribution of mild-to-severe contractions.\n","rough_estimate":"~50-70% guess they will experience a 'major' recession; most underestimate that mild ones are near-certain","kind":"intuition"},"native":{"display":"~12 NBER-dated recessions in 80 years (1945-2025, US)","numerator":12,"denominator":80,"unit":"per year (recession-years)","population":"US economy"},"normalized":{"lifetime_us_adult":0.99,"display":"~99% lifetime (experiencing at least one NBER recession, US adult)","log_value":-0.004,"assumptions":"The NBER Business Cycle Dating Committee has identified 12 recessions between 1945 and 2025, with peaks in 1948, 1953, 1957, 1960, 1969, 1973, 1980, 1981, 1990, 2001, 2007, and 2020. That is roughly one every 6.7 years, with an average duration of about 10 months. Over a 59-year adult life (age 18 to 77), an American can expect to live through approximately 8-9 recessions. The probability of experiencing at least one recession is effectively 1.0. The normalized figure of 0.99 reflects the near-certainty of experiencing at least one NBER-dated recession during an adult lifetime, with the residual 0.01 representing a theoretical scenario of an unprecedented multi-decade expansion. For the more policy-relevant question — experiencing a severe recession with unemployment above 8% — the lifetime probability is lower, roughly 0.85-0.95, based on 4-5 such episodes since 1945 (1973-75, 1981-82, 2007-09, 2020). The central estimate uses the broader \"any NBER recession\" definition because even mild recessions cause measurable financial harm (job losses, wealth declines, reduced earnings growth).\n","uncertainty":{"low":0.95,"high":1},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.nber.org/research/data/us-business-cycle-expansions-and-contractions","title":"US Business Cycle Expansions and Contractions","publisher":"National Bureau of Economic Research","source_type":"reputable_reference","statistic":"12 recessions between 1945 and 2025, averaging ~10 months in duration with expansions averaging ~5-6 years","excerpt":"\"The NBER's Business Cycle Dating Committee maintains a chronology of US business cycles. The chronology identifies the dates of peaks and troughs that frame economic recessions and expansions.\"\n","source_date":"2021-07-19","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420035215/https://www.nber.org/research/data/us-business-cycle-expansions-and-contractions","calculation_notes":"The NBER chronology lists peaks at: Nov 1948, Jul 1953, Aug 1957, Apr 1960, Dec 1969, Nov 1973, Jan 1980, Jul 1981, Jul 1990, Mar 2001, Dec 2007, Feb 2020. That gives 12 recessions in approximately 80 years (1945-2025). Average contraction duration is about 10.3 months post-WWII, down from ~18 months pre-WWII. Average expansion is ~5.8 years. Over a 59-year adult lifetime, 59 / 6.7 = ~8.8 expected recessions. The probability of avoiding all of them: assuming independent Poisson-like arrivals at rate 1/6.7 per year, P(zero recessions in 59 years) is vanishingly small (<0.001).\n","independence_note":"The NBER Business Cycle Dating Committee is the authoritative arbiter of US recession dates. FRED, BLS, and academic literature all defer to NBER dating.\n"},{"url":"https://www.federalreserve.gov/pubs/bulletin/2012/pdf/scf12.pdf","title":"Changes in U.S. Family Finances from 2007 to 2010: Evidence from the Survey of Consumer Finances","publisher":"Federal Reserve Board","source_type":"govt_report","statistic":"Median family net worth fell 38.8% from 2007 to 2010 (from $126,400 to $77,300 in 2010 dollars)","excerpt":"\"Over the 2007-10 period, the median value of real family net worth fell 38.8 percent, and the mean fell 14.7 percent. Median family income before taxes fell 7.7 percent.\"\n","source_date":"2012-06-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420035252/https://www.federalreserve.gov/pubs/bulletin/2012/pdf/scf12.pdf","calculation_notes":"The Survey of Consumer Finances (SCF) is conducted every three years. The 2007-to-2010 comparison captures the full impact of the Great Financial Crisis on household balance sheets. The 38.8% median decline is driven primarily by the collapse in housing values (the primary asset for most households) and equity losses. The mean decline (14.7%) is smaller because high-net-worth households held more diversified portfolios that partially recovered by 2010. By the 2019 SCF, median net worth had recovered to approximately $121,700, and by 2022 it surged to $192,900 — exceeding 2007 levels in real terms.\n","independence_note":"The SCF is the Federal Reserve's flagship household wealth survey, conducted independently of BLS employment data and NBER business cycle dating.\n"},{"url":"https://www.hartfordfunds.com/practice-management/client-conversations/managing-volatility/bear-markets.html","title":"10 Things You Should Know About Bear Markets","publisher":"Hartford Funds","source_type":"reputable_reference","statistic":"27 bear markets (>20% S&P 500 decline) since 1928; average duration 9.6 months; average decline ~35%; average recovery 2.5 years","excerpt":"\"There have been 27 bear markets in the S&P 500 Index since 1928. However, there have also been 28 bull markets — and stocks have risen significantly over the long term. The average length of a bear market is 289 days, or about 9.6 months. Stocks lose 35% on average in a bear market.\"\n","source_date":"2025-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421184706/https://www.hartfordfunds.com/practice-management/client-conversations/managing-volatility/bear-markets.html","calculation_notes":"27 bear markets in ~97 years gives roughly one every 3.6 years. Over a 40-year investing career, an investor can expect ~11 bear markets. The probability of experiencing at least one >20% decline in a 40-year career is effectively 1.0. Stocks have been in bull markets approximately 78% of the time. The average recovery to a prior peak is 2.5 years, meaning a patient investor who holds through a bear market has historically recovered losses within a few years.\n","independence_note":"Hartford Funds analysis uses S&P Dow Jones Indices data. The bear-market count and methodology are consistent with other financial data providers (Yardeni Research, Invesco, etc.).\n"},{"url":"https://www.bls.gov/spotlight/2012/recession/pdf/recession_bls_spotlight.pdf","title":"The Recession of 2007-2009: BLS Spotlight on Statistics","publisher":"Bureau of Labor Statistics","source_type":"govt_report","statistic":"Unemployment peaked at 10.0% in October 2009; 8.7 million jobs lost during the Great Recession","excerpt":"\"The unemployment rate peaked at 10.0 percent in October 2009, up from 5.0 percent in December 2007, when the recession began. More than 15 million people were unemployed at the peak.\"\n","source_date":"2012-02-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260402103508/https://www.bls.gov/spotlight/2012/recession/pdf/recession_bls_spotlight.pdf","calculation_notes":"BLS data provides the employment dimension of recession severity. Peak unemployment rates for major post-war recessions: 1973-75 (9.0%), 1981-82 (10.8%), 2007-09 (10.0%), 2020 (14.7% — though the COVID recession was ultra-short at 2 months). For mild recessions (1990-91, 2001), peak unemployment stayed below 8%. The severity distribution is key: most recessions cause moderate unemployment spikes (6-8%), while crisis-level events (>10%) are rarer — roughly 3-4 per 80 years.\n","independence_note":"BLS employment data is collected through the Current Population Survey (household survey) and Current Employment Statistics (establishment survey), independently of NBER dating and Federal Reserve wealth surveys.\n"},{"url":"https://www.federalreserve.gov/publications/october-2023-changes-in-us-family-finances-from-2019-to-2022.htm","title":"Changes in U.S. Family Finances from 2019 to 2022","publisher":"Federal Reserve Board","source_type":"govt_report","statistic":"Real median family net worth surged 37% to $192,900 between 2019 and 2022; real mean net worth rose 23% to $1,063,700","excerpt":"\"Between 2019 and 2022, real median net worth surged 37 percent to $192,900, and real mean net worth increased 23 percent to $1,063,700.\"\n","source_date":"2023-10-18","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260706214543/https://www.federalreserve.gov/publications/october-2023-changes-in-us-family-finances-from-2019-to-2022.htm","calculation_notes":"Confirms the body's claim that median net worth recovered to $192,900 by 2022 via a 37% real (inflation-adjusted) increase from the 2019 SCF wave. This is the Fed's own headline comparison, using 2019 restated in comparable real dollars rather than the nominal 2019 figure ($121,700) cited from the 2012 Bulletin above; both describe the same recovery, from different base-year dollar conventions.\n","independence_note":"Federal Reserve Board's own Survey of Consumer Finances, the same flagship survey as the 2012 Bulletin already cited above, updated to the 2019-2022 wave.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Divorce (lifetime, US first marriage)","lifetime_us_adult":0.42},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"Mild recession (GDP decline <2%, unemployment <8%)","probability":0.99,"notes":"Most common type — 1990-91, 2001, and arguably 2020 (deep but only 2 months). An adult will experience several."},{"region":"Severe recession (GDP decline >3%, unemployment >8%)","probability":0.9,"notes":"Less frequent but still near-certain over a lifetime — 1973-75, 1981-82 qualify; ~4-5 since 1945."},{"region":"Crisis-level (unemployment >10%, major wealth destruction)","probability":0.55,"notes":"Roughly 3 events since 1945 (1981-82 at 10.8%, 2007-09 at 10.0%, 2020 at 14.7%). Lifetime exposure depends on birth cohort."},{"region":"Stock market bear market (>20% S&P 500 decline)","probability":0.99,"notes":"27 bear markets since 1928, roughly every 3.6 years. A 40-year investor will experience ~11."}],"personal_factor_multipliers":[{"factor":"No emergency fund / living paycheck to paycheck","multiplier":3,"notes":"Financial impact severity multiplier, not recession probability. ~60% of Americans report living paycheck to paycheck; a recession that causes job loss or income reduction has outsized impact without reserves."},{"factor":"Diversified portfolio with 6+ month emergency fund","multiplier":0.3,"notes":"Impact severity multiplier. Emergency reserves and diversification don't prevent recessions but dramatically reduce personal financial damage. Historical data shows patient investors who hold through bear markets recover within 2.5 years on average."},{"factor":"Heavily leveraged (large mortgage + margin investing + consumer debt)","multiplier":5,"notes":"Leverage amplifies losses in downturns. Margin calls, underwater mortgages, and debt service on reduced income compound recession damage. The 2008 GFC disproportionately harmed leveraged households."},{"factor":"Government or healthcare worker","multiplier":0.5,"notes":"Recession-resistant employment sectors. Government employment and healthcare are countercyclical or acyclical, reducing income-loss risk during downturns."}],"short_label":"Economic recession","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This entry conflates two distinct questions: the probability of living through a recession (near 1.0) and the probability of experiencing severe personal financial harm from one (much lower, highly variable). The personal_factor_multipliers apply to severity of impact, not to the macroeconomic event itself. The \"lifetime probability\" of 0.99 refers to experiencing at least one NBER-dated recession, which is essentially certain. The more actionable figure — whether a given recession will be severe enough to cause lasting personal harm — depends heavily on individual circumstances: employment sector, savings, debt levels, housing tenure, and birth cohort. The 2008 GFC destroyed 39% of median household net worth, but by 2022 median net worth had surged to $192,900, exceeding pre-crisis levels. Past recovery patterns are not guarantees; the Great Depression took over a decade for full employment recovery.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single downward-trending line graph on a plain background, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/economic-recession-impact","api_url":"https://likelier.app/api/fears/economic-recession-impact.json"},{"slug":"hand-hygiene-neglect","question":"What are the odds of getting sick in a year if you rarely wash your hands?","quick_answer":"Rarely washing your hands adds roughly a 1-in-3 chance each year of at least one extra symptomatic respiratory or GI illness, which compounds to near-certainty across an adult lifetime. Most adults file poor hand hygiene as negligible; the literature puts it closer to one extra illness every one to two years.\n","category":"health","no_reliable_estimate":false,"perceived":{"description":"Hand hygiene is one of the few public-health messages that has been repeated so often it has become background noise. Most adults can recite the recommendation without believing it moves the dial much — the cultural picture is that handwashing is for surgeons, food handlers, and toddlers, and that a healthy adult who skips the sink after the bus ride or before lunch is buying, at most, a nuisance-level increase in colds. There is no rigorous US survey on what fraction of adults believe poor hand hygiene is a meaningful driver of their personal infection rate, so the best summary is that the perceived attributable risk is usually filed between \"negligible\" and \"small\".\n","rough_estimate":"Most adults guess rarely washing hands adds maybe a cold or two a year; the literature puts it closer to one extra symptomatic illness every one-to-two years, compounding across decades","kind":"intuition"},"native":{"display":"~1 in 3 per year: at least one extra symptomatic illness attributable to poor hand hygiene","numerator":1,"denominator":3,"unit":"per year","population":"US adult, low hand-hygiene adherence vs high adherence baseline"},"normalized":{"lifetime_us_adult":0.99,"display":"~Near-certain lifetime: at least one such illness compounded across adult life","log_value":-0.004,"assumptions":"This entry estimates the per-year and lifetime probability of at least one additional symptomatic illness (acute respiratory or gastrointestinal) attributable to poor hand-hygiene adherence, for a US adult compared with a high-adherence adult of otherwise similar exposure. Anchors: (1) Aiello et al. 2008 (AJPH) meta-analysis of 30 community trials found hand-hygiene improvements reduced gastrointestinal illness by 31% (95% CI 19-42%; pooled RR 0.69) and respiratory illness by 21% (95% CI 5-34%; pooled RR 0.79). (2) Rabie & Curtis 2006 (Trop Med Int Health) pooled eight studies and found handwashing cut respiratory infection risk by ~16% (95% CI 11-21%). (3) Jefferson et al. 2023 Cochrane review (78 trials, 610,872 participants) found a 14% reduction in acute respiratory infections with hand-hygiene interventions (RR 0.86, 95% CI 0.81-0.90), an absolute reduction from 380 events per 1,000 people to 327 per 1,000 in intervention arms. (4) CDC's community baseline: US adults average roughly 2-3 acute respiratory infections per year and on the order of 0.5-1 gastrointestinal episodes per year. Calculation: apply a ~17% respiratory RR reduction to a 2.5-episode-per-year baseline → ~0.43 extra ARIs per year for the low-adherence adult vs the high-adherence adult. Apply the Aiello 31% GI reduction to a 0.7 episode baseline → ~0.22 extra GI episodes per year. Combined excess ≈ 0.6-0.7 symptomatic episodes per year; of these, roughly a third to a half are clinically meaningful enough to seek care, miss work, or require medication — giving an annual hazard λ ≈ 0.3-0.45 for at least one attributable clinically meaningful illness. Using a Poisson approximation, 1 - exp(-0.35) ≈ 0.30, giving ~1 in 3 per year. Compounded over 59 years of remaining adult life: 1 - (1-0.30)^59 ≈ 1 - 1e-9, effectively certain; capped at 0.99 to keep the normalized value interpretable as a probability rather than a rhetorical flourish. The scope is declared as subgroup_lifetime because this is an excess-risk figure relative to a high-adherence comparator, not a general-population lifetime risk.\n","uncertainty":{"low":0.95,"high":0.995},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2446461/","title":"Effect of Hand Hygiene on Infectious Disease Risk in the Community Setting: A Meta-Analysis","publisher":"American Journal of Public Health (Aiello, Coulborn, Perez, Larson)","source_type":"peer_reviewed","statistic":"Improvements in hand hygiene reduced gastrointestinal illness by 31% (95% CI 19-42%; pooled RR 0.69, 95% CI 0.58-0.81) across 24 studies, and respiratory illness by 21% (95% CI 5-34%; pooled RR 0.79, 95% CI 0.66-0.95) across 16 studies","excerpt":"\"Improvements in hand hygiene resulted in reductions in gastrointestinal illness of 31% (95% confidence intervals [CI]=19%, 42%) and reductions in respiratory illness of 21% (95% CI=5%, 34%). [...] The most beneficial intervention was hand-hygiene education with use of nonantibacterial soap. Use of antibacterial soap showed little added benefit compared with use of nonantibacterial soap.\"\n","source_date":"2008-08-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260213124803/https://pmc.ncbi.nlm.nih.gov/articles/PMC2446461/","calculation_notes":"Aiello 2008 is the canonical community-setting meta-analysis for this entry. The 21% respiratory and 31% GI reductions, applied to a US adult baseline of ~2.5 ARIs/year and ~0.7 GI episodes/year, produce an excess of roughly 0.43 ARIs/year and 0.22 GI episodes/year for a low-adherence adult vs a high-adherence comparator. Combined excess ≈ 0.65 episodes/year; roughly a third to a half are symptomatic enough to be clinically meaningful, giving an annual hazard of ~0.3-0.45 and a per-year probability of at least one such illness ≈ 1 - exp(-0.35) ≈ 0.30. Used as the primary RR anchor.\n","independence_note":"Aiello 2008 shares input studies with Rabie & Curtis 2006 for the respiratory arm and with Curtis & Cairncross 2003 for parts of the GI arm; the three should be read as a family of overlapping meta-analyses converging on similar effect sizes, not as three independent estimates.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16553905/","title":"Handwashing and risk of respiratory infections: a quantitative systematic review","publisher":"Tropical Medicine & International Health (Rabie, Curtis)","source_type":"peer_reviewed","statistic":"Pooled risk reductions 6-44% across eight studies; pooling the seven homogenous studies gave RR 1.19 (95% CI 1.12-1.26) for not washing hands, implying handwashing cuts respiratory infection risk by 16% (95% CI 11-21%)","excerpt":"\"All eight eligible studies reported that handwashing lowered risks of respiratory infection, with risk reductions ranging from 6% to 44% [pooled value 24% (95% CI 6-40%)]. Pooling the results of only the seven homogenous studies gave a relative risk of 1.19 (95% CI 1.12-1.26), implying that hand cleansing can cut the risk of respiratory infection by 16% (95% CI 11-21%).\"\n","source_date":"2006-03-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420041511/https://pubmed.ncbi.nlm.nih.gov/16553905/","calculation_notes":"Rabie & Curtis provide the lower end of the respiratory-infection effect-size range: ~16% reduction on the homogenous-study pool, ~24% on the full eight-study pool. Used to bracket the respiratory component of the calculation alongside Aiello 2008 (21%) and Cochrane 2023 (14%). The convergence of three separate meta-analyses on a 14-24% respiratory-infection reduction is the methodological basis for treating the effect size as robust rather than fragile. Authors note that the included studies were of poor quality and mostly from developed countries, which is reflected in the entry's caveats.\n","independence_note":"Rabie and Curtis are co-authors on the adjacent Curtis & Cairncross 2003 diarrhoea meta-analysis; the two reviews are methodologically aligned and share some included studies. Treat as partially dependent with Curtis & Cairncross and with Aiello 2008.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/12726975/","title":"Effect of washing hands with soap on diarrhoea risk in the community: a systematic review","publisher":"The Lancet Infectious Diseases (Curtis, Cairncross)","source_type":"peer_reviewed","statistic":"Pooled relative risk of diarrhoeal disease associated with not washing hands from intervention trials: 1.88 (95% CI 1.31-2.68), implying handwashing reduces diarrhoea risk by 47%; risk reductions of 42-44% across alternative pooled subsets","excerpt":"\"The pooled relative risk of diarrhoeal disease associated with not washing hands from the intervention trials was 1.88 (95% CI 1.31-2.68), implying that handwashing could reduce diarrhoea risk by 47%. When all studies, when only those of high quality, and when only those studies specifically mentioning soap were pooled, risk reduction ranged from 42-44%. The risks of severe intestinal infections and of shigellosis were associated with reductions of 48% and 59%, respectively.\"\n","source_date":"2003-05-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420041543/https://pubmed.ncbi.nlm.nih.gov/12726975/","calculation_notes":"Curtis & Cairncross give the largest pooled effect size on the diarrhoeal side — a 42-47% reduction — but the underlying studies are heavily weighted toward low- and middle-income settings with different baseline waterborne-disease exposure than a typical US adult. Used as the anchor for the global / LMIC row of the regional_breakdown and as the upper-bound anchor on the GI attributable-risk calculation. The US-adult headline uses the more conservative Aiello 2008 31% GI reduction instead.\n","independence_note":"Partially overlaps with Aiello 2008 in included trials; shares one author with Rabie & Curtis 2006. Treat as the dominant diarrhoea-specific evidence base rather than a wholly independent estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/36715243/","title":"Physical interventions to interrupt or reduce the spread of respiratory viruses","publisher":"Cochrane Database of Systematic Reviews (Jefferson et al.)","source_type":"peer_reviewed","statistic":"Hand hygiene interventions vs control: 14% relative reduction in acute respiratory infections (RR 0.86, 95% CI 0.81-0.90; 9 trials, 52,105 participants; moderate-certainty evidence); absolute reduction from 380 per 1,000 to 327 per 1,000","excerpt":"\"Comparing hand hygiene interventions with controls [...] there was a 14% relative reduction in the number of people with acute respiratory infections in the hand hygiene group (RR 0.86, 95% CI 0.81 to 0.90; 9 trials, 52,105 participants; moderate-certainty evidence), suggesting a probable benefit. In absolute terms this benefit would result in a reduction from 380 events per 1000 people to 327 per 1000 people (95% CI 308 to 342).\"\n","source_date":"2023-01-30","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420041618/https://pubmed.ncbi.nlm.nih.gov/36715243/","calculation_notes":"The Cochrane update is the most recent large-scale synthesis and provides the most conservative pooled respiratory-infection effect size (14%) of the three meta-analyses cited here. The absolute 53-per-1,000-per-period reduction in ARIs, applied across multiple exposure-years and combined with the larger diarrhoeal effect sizes, is the methodological basis for treating the per-year probability of at least one attributable symptomatic illness as \"roughly 1 in 3\" rather than a lower figure. The review's conclusion that effect on laboratory-confirmed influenza and ILI specifically is uncertain is noted in the caveats.\n","independence_note":"Cochrane 2023 includes a number of the same community trials as Aiello 2008 and Rabie & Curtis 2006; treat as an updated, stricter re-analysis of a partially overlapping evidence base, not as an independent estimate.\n"},{"url":"https://www.cdc.gov/clean-hands/data-research/facts-stats/index.html","title":"Handwashing Facts","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Handwashing education in the community reduces diarrhoea by 23-40%, respiratory illnesses like colds by 16-21%, diarrhoeal illness in the immunocompromised by 58%, and school absenteeism from GI illness by 29-57%","excerpt":"\"Reduces the number of people who get sick with diarrhea by 23-40% [...] Reduces diarrheal illness in people with weakened immune systems by 58% [...] Reduces respiratory illnesses, like colds, in the general population by 16-21% [...] Reduces absenteeism due to gastrointestinal illness in schoolchildren by 29-57%.\"\n","source_date":"2024-04-17","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260425185215/https://www.cdc.gov/clean-hands/data-research/facts-stats/index.html","calculation_notes":"CDC's plain-language restatement of the underlying meta-analyses. Used as the canonical US-government citation for the headline effect-size ranges and as the source for the immunocompromised (58% GI reduction) and school-age multipliers in the regional breakdown. Not an independent estimate — CDC's ranges are essentially republished from Aiello 2008, Rabie & Curtis 2006, and Curtis & Cairncross 2003.\n","independence_note":"Derived directly from the three meta-analyses above; included to confirm continuing official US endorsement of the 16-21% respiratory and 23-40% diarrhoeal reduction ranges, not as new evidence.\n"}],"comparison_anchors":[{"label":"Food poisoning illness in a year (US)","lifetime_us_adult":0.14},{"label":"Regular-drinking alcohol-attributable death (lifetime, heavy drinker)","lifetime_us_adult":0.15},{"label":"Lifetime colorectal cancer (US adult baseline)","lifetime_us_adult":0.041},{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"US adult, low hand-hygiene adherence (headline)","probability":0.99,"notes":"Lifetime compounded from ~1-in-3-per-year excess symptomatic illness hazard. Effectively certain across 59 adult years; the annual figure (~30%) is the useful one."},{"region":"Parent of young children, low adherence","probability":0.995,"notes":"Baseline exposure is higher (young children bring home 6-10 ARIs/year); absolute excess attributable to poor hand hygiene roughly doubles."},{"region":"Healthcare worker, low adherence","probability":0.999,"notes":"Hospital-setting meta-analyses (Pittet, WHO) show hand-hygiene compliance effects on HAI rates of 30-50%. This entry uses the community-setting figures; HCW risk is a different regime and larger in absolute terms."},{"region":"Immunocompromised adult, low adherence","probability":0.999,"notes":"CDC cites 58% reduction in diarrhoeal illness from handwashing in immunocompromised populations; absolute excess attributable to poor adherence is large and clinically consequential (invasive salmonella, cryptosporidium, norovirus)."},{"region":"Global / LMIC setting, low adherence with unreliable water","probability":0.995,"notes":"Curtis & Cairncross 2003 pooled 42-47% reduction in diarrhoeal disease; in settings with waterborne pathogens the attributable annual risk is much larger than the US figure — measured in lives, not sick days."}],"personal_factor_multipliers":[{"factor":"lives with children under 6","multiplier":2,"notes":"Young children bring home 6-10 respiratory infections per year vs 2-3 for a typical adult; absolute excess attributable to poor hand hygiene scales with the baseline exposure."},{"factor":"healthcare worker (direct patient contact)","multiplier":3,"notes":"Hospital hand-hygiene literature (Pittet, WHO) shows compliance-to-HAI effect sizes larger than the community meta-analyses and across a more consequential outcome set."},{"factor":"immunocompromised (transplant, chemotherapy, advanced HIV)","multiplier":4,"notes":"Both the hazard per exposure and the severity of outcomes are larger; CDC cites 58% reduction in GI illness from hand hygiene in this subgroup."},{"factor":"heavy public-transit / shared-workspace user","multiplier":1.3,"notes":"Fomite and respiratory-droplet exposure is higher; effect modest because most transmission is via close-contact aerosol, not hands."},{"factor":"alcohol-based sanitizer only, no soap and water","multiplier":1.1,"notes":"Alcohol sanitizer is effective against enveloped respiratory viruses but less so against norovirus, C. difficile, and heavily soiled hands; net effect vs soap-and-water is small in community settings but non-zero."}],"short_label":"Hand hygiene","myth_framing":"underrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"Three structural caveats. First, every community hand-hygiene trial in the literature measures adherence by self-report or sporadic observation; actual handwashing behaviour is noisier than the \"intervention vs control\" framing implies, and the real effect size for a genuine low-adherence vs high-adherence comparison is probably larger than the pooled RRs suggest, because real-world adherence gradients are compressed in trial populations. Second, the studies underlying the meta-analyses are heavily educational-intervention bundles (soap plus education plus sometimes environmental changes), so the attributable effect of handwashing specifically is entangled with behaviour-change confounding. Third, this entry measures the probability of at least one attributable symptomatic illness, not severity or mortality — the larger mortality effects documented by Curtis & Cairncross (millions of lives saved globally) apply in LMIC diarrhoeal settings and are not captured by the US-adult headline. Individual outcomes depend on household composition, work environment, immune status, seasonal exposure, local circulating pathogens, and whether the alternative to not washing is alcohol-based sanitizer or nothing at all. The headline should be read as an order-of-magnitude calibration for the excess annual symptomatic-illness burden of low adherence, not as a personal forecast.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-9-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single bar of soap resting on a pale surface, viewed from above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/hand-hygiene-neglect","api_url":"https://likelier.app/api/fears/hand-hygiene-neglect.json"},{"slug":"stock-market-crash","question":"What are the odds of experiencing a major stock market crash (>30% decline) in your lifetime?","quick_answer":"A drop of more than 30% has hit about once every 12 years, so an investor with a multi-decade horizon is nearly certain (~99%) to live through at least one — far more routine than the once-in-a-generation shock most people picture.\n","category":"other","no_reliable_estimate":false,"perceived":{"description":"Stock market crashes occupy a unique psychological niche: everyone knows they happen, most people dread them, but few appreciate how routine they are. Gallup's annual Economy and Personal Finance survey consistently finds that roughly one in three US adults cite a stock market crash as a major financial worry, putting it alongside job loss and medical debt. The Chapman Survey of American Fears does not list market crashes as a standalone item, but \"economic/financial collapse\" routinely ranks among the top ten fears, suggesting that the anxiety attaches less to the statistical frequency of crashes and more to the narrative of sudden, irreversible ruin.\n","rough_estimate":"most people treat a severe crash as a once-in-a-generation shock, not a once-a-decade regularity","kind":"intuition"},"native":{"display":"~8 declines >30% in ~96 years of S&P 500 history (1929-2025)","numerator":8,"denominator":96,"unit":"per year (frequency)","population":"S&P 500 index history"},"normalized":{"lifetime_us_adult":0.99,"display":"~99% lifetime (US adult investor over a 40+ year horizon)","log_value":-0.004,"assumptions":"Since 1929 the S&P 500 has experienced roughly 8 peak-to-trough declines exceeding 30%: 1929-32 (-86%), 1937-38 (-54%), 1968-70 (-36%), 1973-74 (-48%), 1987 (-34%), 2000-02 (-49%), 2007-09 (-57%), and 2020 (-34%). That is approximately one every 12 years on average. Over a 40-year investment career the probability of experiencing at least one such decline is 1 - (1 - 1/12)^40 ≈ 0.97. Over a full 59-year adult life the figure exceeds 0.99. The 2022 decline (-25%) and the April 2025 tariff selloff (-19% intraday peak-to-trough) both fell short of 30% but illustrate how close the market comes to the threshold regularly. The point estimate of 0.99 reflects the near-certainty that any adult with a multi-decade investment horizon will live through at least one >30% drawdown.\n","uncertainty":{"low":0.95,"high":0.999},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.hartfordfunds.com/practice-management/client-conversations/managing-volatility/bear-markets.html","title":"10 Things You Should Know About Bear Markets","publisher":"Hartford Funds","source_type":"reputable_reference","statistic":"There have been 27 bear markets (>20% decline) in the S&P 500 since 1928; stocks lose 35% on average in a bear market; average recovery takes about 2.5 years","excerpt":"\"There have been 27 bear markets in the S&P 500 Index since 1928. The average length of a bear market is 289 days, or about 9.6 months. Stocks lose 35% on average in a bear market. About 42% of the S&P 500 Index's strongest days in the last 20 years occurred during a bear market.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421184706/https://www.hartfordfunds.com/practice-management/client-conversations/managing-volatility/bear-markets.html","calculation_notes":"Hartford Funds compiles S&P 500 bear market data in partnership with Ned Davis Research. Their count of 27 bear markets since 1928 yields an average frequency of one every ~3.6 years for >20% declines. The average decline of 35% confirms that the typical bear market crosses the 30% threshold. From this dataset we identify approximately 8 distinct episodes exceeding 30%, giving a frequency of roughly once every 12 years. Lifetime probability = 1 - (1 - 1/12)^59 ≈ 0.993.\n"},{"url":"https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html","title":"Historical Returns on Stocks, Bonds and Bills: 1928-2024","publisher":"Aswath Damodaran, NYU Stern School of Business","source_type":"reputable_reference","statistic":"S&P 500 annual return data from 1928-2024; annual return averaged 11.79% with positive returns in 71 of 97 years; worst year -43.84% (1931), best year 52.56% (1954)","excerpt":"[Raw data table of annual S&P 500 returns, Treasury bill rates, and Treasury bond returns from 1928 to 2024, with cumulative growth-of-$100 columns for each asset class. No narrative text; the table shows year-by-year returns including all crash and recovery years.]\n","source_date":"2025-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260411202502/https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html","calculation_notes":"Damodaran's dataset is the standard academic reference for long-run US equity returns. The raw data table shows annual returns from 1928-2024, with positive returns in 71 of 97 years and an annual return averaging 11.79%. The data demonstrates that every crash in S&P 500 history has been fully recovered. This provides the denominator context: crashes are near-certain events, but long-term compounding overwhelms them for patient investors. Used to validate the \"recoverable\" framing rather than to compute the crash probability directly.\n"},{"url":"https://www.prnewswire.com/news-releases/investors-missed-the-best-of-2024s-market-gains-latest-dalbar-investor-behavior-report-finds-302416023.html","title":"Investors Missed the Best of 2024's Market Gains, Latest DALBAR Investor Behavior Report Finds","publisher":"DALBAR, Inc. (via PR Newswire)","source_type":"primary_study","statistic":"Average equity investor earned 16.54% in 2024 vs S&P 500 return of 25.05%; 848 basis point lag is the second-largest performance gap of the past decade","excerpt":"\"The Average Equity Investor earned just 16.54% in 2024, compared to the S&P 500's 25.05% return. The 848 basis point lag represents the second-largest investor performance gap of the past decade.\"\n","source_date":"2025-03-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260307135221/https://www.prnewswire.com/news-releases/investors-missed-the-best-of-2024s-market-gains-latest-dalbar-investor-behavior-report-finds-302416023.html","calculation_notes":"DALBAR's annual QAIB study measures the gap between market returns and the returns actually captured by investors, using mutual fund flow data. In 2024 the gap was 848 basis points — the second largest in a decade. The underperformance is driven primarily by behavioral factors: panic selling during downturns, late re-entry after recoveries begin, and poor market timing. This source substantiates the claim that the primary risk from a crash is behavioral (selling at the bottom), not the crash itself. The full QAIB report is paywalled; the statistics cited here come from DALBAR's publicly available press release.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft victim (annual, US)","lifetime_us_adult":0.33}],"regional_breakdown":[{"region":"Declines of 20-30% (correction/mild bear)","probability":0.999,"notes":"~19 episodes since 1928; occurs every ~5 years on average — effectively certain over a career"},{"region":"Declines of 30-40%","probability":0.95,"notes":"~4 episodes since 1928 in this band specifically (1968-70, 1987, 2020, plus several borderline)"},{"region":"Declines of 40-50%","probability":0.75,"notes":"~3 episodes: 1937-38 (-54%), 1973-74 (-48%), 2000-02 (-49%)"},{"region":"Declines exceeding 50%","probability":0.45,"notes":"~2 episodes: 1929-32 (-86%), 2007-09 (-57%); roughly every 45-50 years"}],"personal_factor_multipliers":[{"factor":"100% equities, no diversification","multiplier":1,"notes":"baseline case — fully exposed to headline S&P 500 drawdowns"},{"factor":"60/40 stock-bond portfolio","multiplier":0.6,"notes":"bonds buffer equity drawdowns; a 30% equity crash translates to roughly an 18% portfolio decline"},{"factor":"panic seller (sells at or near bottom)","multiplier":3,"notes":"DALBAR data: locking in losses at the bottom turns a temporary drawdown into a permanent wealth reduction"},{"factor":"buy-and-hold through crash","multiplier":0.1,"notes":"historical recovery rate from every S&P 500 bear market is 100%; permanent loss approaches zero for patient investors"},{"factor":"near retirement (within 5 years)","multiplier":2,"notes":"sequence-of-returns risk is highest at the point of withdrawal; a crash just before retirement can permanently impair income"},{"factor":"early career (20+ years to retirement)","multiplier":0.05,"notes":"crashes in early accumulation years are net beneficial — lower prices mean more shares purchased per dollar of savings"},{"factor":"leveraged/margin investor","multiplier":5,"notes":"margin calls during a crash force liquidation at the worst possible time; 2x leverage turns a 30% decline into a 60% loss or forced sale"}],"short_label":"Stock market crash","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"The headline probability addresses whether an investor will *experience* a >30% market decline, not whether they will suffer permanent financial harm from one. The two are very different questions. The S&P 500 has recovered from every bear market in history, with an average recovery time of about 2.5 years. Permanent loss of wealth from a crash is almost exclusively a behavioral outcome (panic selling, margin liquidation, or forced withdrawal at the bottom) rather than a market outcome. The native unit here is event frequency (crashes per year of market history), not a traditional epidemiological rate, so the normalization is a Poisson-style \"at least one event in N years\" calculation rather than a compounding annual hazard.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-research-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single downward-trending line chart on a muted background, flat vector illustration, no people, no money."},"canonical_url":"https://likelier.app/stock-market-crash","api_url":"https://likelier.app/api/fears/stock-market-crash.json"},{"slug":"methanol-poisoning","question":"What are the odds of being seriously harmed or killed by methanol in counterfeit alcohol?","quick_answer":"No reliable figure exists: there is no authoritative global death toll for methanol poisoning from counterfeit alcohol, and documented deaths represent only a fraction of the true burden. For adults in high-income countries drinking only from licensed supply the personal risk is essentially negligible; the harm concentrates in regions with large unregulated-alcohol markets.\n","category":"food","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"In countries with robust food-safety regimes, most drinkers have little reason to think about what is in their glass. The threat of methanol contamination in counterfeit or informally produced spirits sits far outside the mental model of someone buying a bottle at a licensed retailer. Even travellers to higher-risk regions tend to worry more about upset stomachs than about a toxic alcohol that can cause blindness or death before symptoms are obvious. Where the hazard is known at all, it is typically framed as a dramatic but remote event — a mass poisoning in a faraway country — rather than a probabilistic personal risk.\n","kind":"intuition"},"native":{"display":"No reliable global annual death figure exists; WHO says 'thousands' per year, MSF documented ~14,800 deaths in outbreak database since 1998","numerator":14800,"denominator":5000000000,"unit":"cumulative documented since 1998","population":"global (documented outbreaks only; massive undercount acknowledged)"},"normalized":{"lifetime_us_adult":0.000506,"display":"Not reliably calculable — documented deaths represent a fraction of true burden","log_value":-3.3,"assumptions":"No authoritative source provides a reliable global annual death toll from methanol poisoning. The WHO information note states \"thousands of people suffer from methanol poisoning every year\" without quantifying deaths globally. The MSF/Oslo University Hospital outbreak database has documented approximately 14,800 deaths since 1998 (~500-600 documented per year), but explicitly acknowledges massive underreporting because many outbreaks are never identified as methanol poisonings. The Lachenmeier et al. (2021) review documents that 28.6% of global alcohol consumption is unrecorded (produced outside regulatory oversight) but does not provide a global death figure. The previously used 30,000 annual deaths figure was an extrapolation not directly supported by any cited source. Because documented deaths (~500-600/year) and plausible true burden (potentially tens of thousands) differ by 1-2 orders of magnitude, no single point estimate can be meaningfully normalized. The retained lifetime figure (0.000506) should be treated as illustrative only, based on the assumption that true annual deaths are in the low tens of thousands concentrated among ~3.5 billion adults in regions with large unregulated alcohol markets. For adults in high-income countries drinking only from licensed supply, personal risk is essentially negligible.\n","uncertainty":{"low":0.00017,"high":0.00135},"scope":"subgroup_lifetime"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8303512/","title":"Illicit Alcohol: Public Health Risk of Methanol Poisoning and Policy Mitigation Strategies","publisher":"Foods (MDPI) — Lachenmeier, Rehm, Taylor et al.","source_type":"peer_reviewed","statistic":"28.6% of global alcohol consumption is unrecorded; illicit alcohol produced without regulatory oversight carries increased risk of methanol adulteration","excerpt":"\"Illicit (unrecorded) alcohol is a critical global public health issue because it is produced without regulatory and market oversight with increased risk of safety, quality and adulteration issues. [...] Global consumption of alcohol in 2005 was an average of 6.13 L of pure alcohol per individual aged 15 years or over, with 28.6% of this amount being unrecorded alcohol.\"\n","source_date":"2021-07-09","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426203835/https://pmc.ncbi.nlm.nih.gov/articles/PMC8303512/","calculation_notes":"This peer-reviewed review synthesises evidence on methanol as a toxicological risk in illicit alcohol. It documents that 28.6% of global alcohol consumption is unrecorded (produced outside regulatory oversight), creating the conditions for methanol contamination. The paper catalogues 68 documented methanol poisoning incidents from 1963-2020 and develops a policy mitigation framework. It does not provide a single global mortality figure. No authoritative source provides an annual global death count.\n"},{"url":"https://www.methanol.org/wp-content/uploads/2016/06/WHO-Methanol-Poisoning-Fact-Sheet.pdf","title":"Information Note: Methanol Poisoning Outbreaks","publisher":"World Health Organization","source_type":"govt_report","statistic":"Thousands of people suffer from methanol poisoning every year; fatality rates are reported to be 20–40% without adequate treatment","excerpt":"\"Thousands of people suffer from methanol poisoning every year. In many cases, the poisoning occurs due to the consumption of illegally produced alcoholic beverages, where methanol is added as a cheap substitute for ethanol, or where methanol contamination of the product occurs.\"\n","source_date":"2014-01-01","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260425043839/https://www.methanol.org/wp-content/uploads/2016/06/WHO-Methanol-Poisoning-Fact-Sheet.pdf","calculation_notes":"The WHO information note confirms the global scope of the problem and the illicit-alcohol mechanism. It states \"thousands of people suffer from methanol poisoning every year\" but does not provide a specific annual death figure. The note documents recurring outbreaks across Asia, Eastern Europe, and Africa.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8187162/","title":"Methanol poisoning as a new world challenge: A review","publisher":"Annals of Medicine and Surgery — Nekoukar Z, Zakariaei Z, Taghizadeh F, et al.","source_type":"peer_reviewed","statistic":"Methanol poisoning is associated with high morbidity and mortality; Iran experienced the greatest prevalence of methanol mass poisoning during COVID-19 (Feb–Apr 2020)","excerpt":"\"Since MP is associated with high morbidity and mortality, it should be considered seriously and instantly managed. Delay in treatment may cause complications, permanent damage, and even death. [...] Following the coronavirus disease 2019 (COVID-19) pandemic in Iran (February 19, 2020 to April 27, 2020), there has been a significant increase in methanol-induced morbidity and mortality. This was the greatest prevalence of methanol mass poisoning in the country in recent periods.\"\n","source_date":"2021-06-06","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260213074537/https://pmc.ncbi.nlm.nih.gov/articles/PMC8187162/","calculation_notes":"This review focuses on the clinical presentation, diagnosis, and treatment of methanol poisoning. It documents Iran as a major affected region, particularly during the COVID-19 pandemic, and notes methanol toxicity incidents in Tunisia, Turkey, and India. The paper does not provide global prevalence rankings by continent; it is used here to support the characterisation of methanol poisoning as a geographically concentrated hazard exacerbated by periods of scarcity or prohibition.\n"},{"url":"https://en.wikipedia.org/wiki/2025_Istanbul_alcohol_poisonings","title":"2025 Istanbul alcohol poisonings","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"Turkey's 2024 Istanbul bootleg-alcohol outbreak killed at least 48 people; the 2025 Istanbul/Ankara outbreak had killed at least 160 people as of March 8, 2025","excerpt":"\"110 people were sickened after consuming tainted alcohol in Istanbul, resulting in at least 48 deaths.\" [on the 2024 outbreak] \"As of March 8, at least 160 people have been reported dead, and 230 others have been hospitalized.\" [on the 2025 outbreak]\n","source_date":"2025-03-08","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20250811000428/https://en.wikipedia.org/wiki/2025_Istanbul_alcohol_poisonings","calculation_notes":"Substantiates the body prose's specific example of a documented methanol outbreak (\"Turkey 2024-2025, over 160 dead across two incidents\"): the two Istanbul-area bootleg-alcohol outbreaks killed at least 48 (2024) and at least 160 (2025, as of the cited date) people, a combined total exceeding 160 and illustrating the scale of the documented-outbreak database discussed elsewhere in this entry. Does not change the global lifetime estimate, which remains illustrative only per the assumptions above.\n","independence_note":"Encyclopedia summary of contemporaneous Turkish and international news reporting; independent of the WHO, MSF/Oslo, and Lachenmeier et al. sources above, used only to ground the specific real-world example cited in the body prose.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5625597/","title":"Estimations of the lethal and exposure doses for representative methanol symptoms in humans","publisher":"Annals of Occupational and Environmental Medicine","source_type":"peer_reviewed","statistic":"Human oral lethal dose of methanol is approximately 300-1000 mg/kg body weight (about 1000 mg/kg if medically untreated); the IPCS reports a comparable 300-1000 mg/kg range","excerpt":"\"If medically untreated, the lethal dose was approximately 1000 mg/kg (oral intake) if there was no simultaneous ingestion of ethanol.\" \"In the IPCS report, the human lethal dose through oral ingestion is approximately 300-1000 mg/kg.\"\n","source_date":"2017-10-02","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260429162820/https://pmc.ncbi.nlm.nih.gov/articles/PMC5625597/","calculation_notes":"Substantiates the body prose's \"lethal dose is roughly 1 ml per kilogram of body weight\" claim. Methanol's density is ~0.792 g/mL, so the paper's ~1000 mg/kg (untreated) figure converts to approximately 1.2-1.3 mL/kg, and the IPCS 300-1000 mg/kg range converts to approximately 0.4-1.3 mL/kg -- consistent with the \"roughly 1 ml/kg\" rule of thumb used in the body prose.\n","independence_note":"Peer-reviewed dose-response literature review, independent of the WHO, MSF/Oslo, and Lachenmeier et al. sources above; used specifically to substantiate the lethal-dose figure in the body prose.\n"}],"comparison_anchors":[{"label":"Death by car crash (lifetime, US)","lifetime_us_adult":0.0108},{"label":"Death from rabies after dog bite (lifetime, global adult)","lifetime_us_adult":0.00069},{"label":"Death from lightning strike (lifetime, US)","lifetime_us_adult":0.0000135}],"regional_breakdown":[{"region":"Adults in illicit-alcohol regions (~3.5 billion)","probability":0.000506,"notes":"Asia, Africa, Eastern Europe, Latin America — illustrative only; true annual deaths unknown"},{"region":"Global average (all adults)","probability":0.00035,"notes":"Illustrative only; diluted across 5B adults, risk concentrates in unregulated markets"},{"region":"High-income countries (licensed retail)","probability":0.000001,"notes":"Effectively zero; regulated supply chain eliminates methanol contamination"}],"short_label":"Counterfeit alcohol","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Risk is almost entirely a function of geography and supply chain. A drinker in the US, UK, EU, or Australia purchasing alcohol from any licensed retailer faces a hazard that is for practical purposes negligible, because regulatory enforcement and supply-chain integrity make methanol-contaminated product essentially absent from formal markets. The hazard concentrates in three populations: drinkers of informally produced spirits in South and Southeast Asia; consumers of surrogate alcohols (cleaning fluids, hand sanitisers with methanol) during periods of scarcity or prohibition; and travellers who purchase unlabelled or informally produced spirits in high-risk regions. The 0.00035 global lifetime estimate is therefore not a meaningful number for most readers of a site like this; the personal probability for a high-income-country resident drinking only from licensed supply is orders of magnitude lower.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A simple flat-vector illustration of an unlabelled bottle and a small warning triangle, rendered in muted greys and amber on a pale background."},"canonical_url":"https://likelier.app/methanol-poisoning","api_url":"https://likelier.app/api/fears/methanol-poisoning.json"},{"slug":"inheritance-dispute-loss","question":"What are the odds of losing part of an inheritance to family disputes or legal costs?","quick_answer":"There is no reliable figure: no single study measures the lifetime chance of losing inheritance value to family disputes or legal costs, and any cited estimate is stitched together from mismatched sources. It is commonly seen as a rare problem for wealthy families, but the conditions for conflict — such as dying without a will — are widespread.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Inheritance disputes are perceived as the province of wealthy families with complicated trusts and scheming relatives — material for TV dramas, not ordinary households. Most people do not anticipate conflict over a parent's estate, and the topic is socially awkward to raise while the parent is alive. The widespread failure to create wills (only 24% of Americans had one in 2025, per Caring.com) means that intestate succession laws — which many heirs do not understand — govern the distribution for the vast majority of estates. The combination of emotional grief, opaque legal processes, and the sudden appearance of money in families unaccustomed to discussing finances creates a fertile environment for disputes that most people do not see coming.\n","rough_estimate":"Perceived as uncommon, mainly affects wealthy","kind":"intuition"},"native":{"display":"~76% of Americans die without a will; probate costs 3-8% of estate value; contingency-fee litigation runs ~33-40% of the recovery","numerator":76,"denominator":100,"unit":"share of Americans dying intestate","population":"US adults (Caring.com, 2025)"},"normalized":{"lifetime_us_adult":0.35,"display":"~35% lifetime probability of losing meaningful inheritance value to disputes or legal costs (author-constructed estimate — no single study supports this figure)","log_value":-0.46,"assumptions":"WARNING: This estimate is heavily constructed and no single published study directly measures the lifetime probability of inheritance loss for US adults. It is stitched from three heterogeneous sources: (1) a US consumer will-ownership survey (Caring.com), (2) UK probate court dispute-category data used as a US proxy (Dutton Gregory), and (3) US-focused probate cost ranges from an industry website (Protecting Wealth). Only about 1 in 10 wills are formally contested, and formal will contests represent a small fraction of all estates. The 35% figure attempts to include routine probate costs (3-8% of estate value) and informal family disagreements, not just formal litigation. However, \"meaningful loss\" (>5% of inheritance) is a constructed threshold with no empirical basis for its probability. The high intestacy rate (76%) creates conditions for disputes but does not directly translate to a dispute probability. The uncertainty range is extremely wide (20-50%) reflecting the absence of direct evidence. This entry is flagged as no_reliable_estimate because the 35% figure should not be treated as a measured prevalence.\n","uncertainty":{"low":0.2,"high":0.5},"scope":"us_adult_lifetime"},"sources":[{"url":"https://www.caring.com/resources/wills-survey","title":"2025 Wills and Estate Planning Study","publisher":"Caring.com","source_type":"primary_study","statistic":"Only 24% of Americans had a will in 2025; an estimated 76% die without one","excerpt":"\"In 2025, only 24% of wills survey respondents said they have a will, 13% reported a living trust, and 4% said they had other estate planning documents.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260503091406/https://www.caring.com/resources/wills-survey","calculation_notes":"The Caring.com survey provides the foundational statistic: 76% of Americans die intestate. Intestate estates are governed by state succession laws that may not match the decedent's wishes, creating the conditions for family disputes. The 24% will-ownership rate is the lowest recorded in the survey's history, down from 33% in 2022. The survey is a nationally representative online poll of approximately 2,500 adults.\n","independence_note":"Caring.com's survey is an annual consumer survey conducted by an independent research panel, methodologically distinct from probate court administrative data and from estate planning industry statistics.\n"},{"url":"https://www.duttongregory.co.uk/site/blog/personalnews/inheritance-dispute-statistics","title":"Inheritance Dispute Statistics: 2024 Rising Trends","publisher":"Dutton Gregory Solicitors","source_type":"reputable_reference","statistic":"Probate caveats increased 12% year-over-year (10,313 to 11,589); executor disputes account for 31.2% of estate litigation","excerpt":"\"In the 12 months to 31 July 2025, there was a 12% increase in applications for probate caveats, rising from 10,313 to 11,589. The majority (31.2%) of disputes are about executor issues, with trust disputes at 25.6% and will disputes at 19.8%.\"\n","source_date":"2025-08-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260105052508/https://www.duttongregory.co.uk/site/blog/personalnews/inheritance-dispute-statistics","calculation_notes":"While this source uses UK probate court data, the typology of disputes (executor, trust, will contests) is applicable cross-jurisdictionally. The 12% year-over-year increase in probate caveats signals a growing trend in contested estates. US-specific probate litigation data is fragmented across state courts, making UK data a useful proxy for dispute categories and trends.\n","independence_note":"UK probate court administrative data is independent from the US-focused Caring.com survey and from US probate cost estimates.\n"},{"url":"https://protectingwealth.com/how-much-does-probate-cost-complete-fee-breakdown-for-2026/","title":"How Much Does Probate Cost? Complete Fee Breakdown for 2026","publisher":"Protecting Wealth","source_type":"reputable_reference","statistic":"Probate costs typically consume 3-8% of estate value","excerpt":"\"Probate fees typically consume 3% to 8% of your estate's total value.\"\n","source_date":"2026-01-15","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260426202437/https://protectingwealth.com/how-much-does-probate-cost-complete-fee-breakdown-for-2026/","calculation_notes":"The 3-8% probate cost range for uncontested estates provides the financial impact data for routine cases. Even uncontested probate imposes meaningful costs: on a $450,000 estate (roughly the median home value in many markets), 5% probate costs equal $22,500. The 30-40% contested-estate contingency fee claim in earlier versions of this entry could not be verified verbatim on the source page and has been removed from the excerpt; the 3-8% uncontested range is well-supported by the source text. The contingency-fee figure itself is now sourced separately below.\n"},{"url":"https://www.triaylaw.com/attorney-fees","title":"Probate Litigation Attorney Fees: Contingency or Hourly Options","publisher":"Triay Law","source_type":"reputable_reference","statistic":"Standard contingency fees for probate litigation run 33% of the recovery if the case settles and 40% if it proceeds to trial","excerpt":"\"For a contingency arrangement, the standard fees are 33 percent of recovery if the case settles, and 40 percent if the case proceeds to trial.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260208024346/https://www.triaylaw.com/attorney-fees","calculation_notes":"Replaces the earlier \"30-40% of the contested amount\" claim that this entry's Protecting Wealth source could not support (see calculation_notes above). This law-firm source states the standard fee structure specifically when probate or estate litigation is taken on contingency: roughly a third of the recovery on settlement, rising to 40% at trial. Note this describes fee structure when a contingency arrangement is used, not the share of contested estates that use one.\n","independence_note":"Independent law-firm publication, distinct from the Caring.com will-ownership survey, the UK Dutton Gregory dispute-category data, and the Protecting Wealth probate-cost breakdown used elsewhere in this entry.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US)","lifetime_us_adult":0.1},{"label":"Student loan default (US borrowers)","lifetime_us_adult":0.26},{"label":"Retirement savings shortfall (US)","lifetime_us_adult":0.39}],"regional_breakdown":[{"region":"Intestate estates (no will)","probability":0.5,"notes":"Intestate estates are far more likely to trigger family disputes due to state succession laws that may not match expectations"},{"region":"Estates with professional estate plan (will + trust)","probability":0.1,"notes":"Proper estate planning dramatically reduces dispute risk, though it does not eliminate it"},{"region":"Blended families","probability":0.55,"notes":"Stepchildren, second spouses, and children from prior relationships create competing claims that frequently lead to disputes"}],"personal_factor_multipliers":[{"factor":"blended family","multiplier":2.5,"notes":"Second marriages with children from prior relationships are the single most common context for inheritance disputes"},{"factor":"estate has a professional trust","multiplier":0.3,"notes":"Revocable living trusts bypass probate entirely and substantially reduce both costs and dispute opportunity"},{"factor":"estate includes real property in multiple states","multiplier":2,"notes":"Multi-state property requires ancillary probate in each state, multiplying costs and legal complexity"},{"factor":"sole heir, no siblings","multiplier":0.2,"notes":"Solo heirs face minimal dispute risk absent other competing claims"}],"short_label":"Inheritance dispute","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"The 35% estimate is the most heavily constructed figure in this dataset and should be treated with extreme caution. No single study tracks the lifetime probability of inheritance loss across the full US population. The figure stitches together three heterogeneous sources: a US consumer survey (will ownership), UK probate court data (dispute categories, used as a US proxy because US probate is fragmented across 50 state court systems with no centralized reporting), and US industry probate cost ranges. Only about 1 in 10 wills are formally contested, suggesting formal litigation risk is far lower than 35%. The higher estimate attempts to include routine probate costs (3-8% of estate value) and informal family disagreements, but these are qualitatively different from \"disputes\" in most people's understanding. \"Meaningful loss\" (>5% of inheritance value) is a constructed threshold — different definitions would yield different estimates. The entry also does not account for the fact that many Americans inherit very little: the Federal Reserve's 2022 SCF shows that the median inheritance received is approximately $69,000, and roughly 20% of American adults never inherit anything at all. This entry is flagged as no_reliable_estimate because the evidence base is too weak to support a specific probability claim.\n","quality_score":{"d1":2,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A torn legal document beside two hands pulling a house in opposite directions, muted grey and olive tones, flat vector illustration."},"canonical_url":"https://likelier.app/inheritance-dispute-loss","api_url":"https://likelier.app/api/fears/inheritance-dispute-loss.json"},{"slug":"5g-cell-tower-radiation","question":"What are the odds of getting cancer from 5G towers or cell phone radiation?","quick_answer":"There is no reliable number here, and that itself is the answer. No cancer study has found a risk from 5G, and the older evidence for radio-frequency signals is limited and contested — IARC lists it as only \"possibly carcinogenic,\" the same low tier as pickled vegetables. Many adults assume a meaningful risk that the data simply does not support.\n","category":"tech","no_reliable_estimate":true,"perceived":{"description":"Public anxiety about radiofrequency electromagnetic fields (RF-EMF) and cancer predates 5G by decades, but the rollout of 5G infrastructure in 2019-2020 triggered an acute spike. The fear was amplified by conspiracy theories linking 5G to COVID-19, resulting in arson attacks on cell towers in the UK and continental Europe. Surveys consistently find that a substantial minority of adults believe cell towers and phones cause cancer, with the proportion rising sharply among those who recall the IARC Group 2B classification without understanding what \"possibly carcinogenic\" means in IARC's framework (the same category includes pickled vegetables and talcum powder). The folk fear conflates three distinct exposure regimes — ambient RF from distant towers, near-field SAR from a phone held to the head, and millimeter-wave (mmWave) from 5G small cells — that differ by orders of magnitude in power density and frequency.\n","rough_estimate":"Many adults assume a meaningful cancer risk from living near cell towers or heavy phone use","kind":"intuition"},"sources":[{"url":"https://www.iarc.who.int/wp-content/uploads/2018/07/pr208_E.pdf","title":"IARC classifies radiofrequency electromagnetic fields as possibly carcinogenic to humans","publisher":"International Agency for Research on Cancer (WHO/IARC)","source_type":"govt_report","statistic":"RF-EMF classified as Group 2B ('possibly carcinogenic to humans') based on limited evidence of glioma in heavy mobile phone users","excerpt":"\"The evidence was reviewed critically, and overall evaluated as being limited among users of wireless telephones for glioma and acoustic neuroma, and inadequate to draw conclusions for other types of cancers. The evidence from the occupational and environmental exposures mentioned above was similarly judged inadequate. The Working Group did not quantitate the risk; however, one study of past cell phone use (up to the year 2004), showed a 40% increased risk for gliomas in the highest category of heavy users (reported average: 30 minutes per day over a 10-year period).\"\n","source_date":"2011-05-31","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260420030707/https://www.iarc.who.int/wp-content/uploads/2018/07/pr208_E.pdf","calculation_notes":"The IARC Monograph 102 Working Group evaluated RF-EMF based primarily on the Interphone study (2010) and Hardell group case-control studies. The 40% glioma increase was observed only in the highest decile of cumulative call time in Interphone, and the Interphone authors themselves cautioned that recall bias and selection bias could account for the finding. Group 2B is IARC's second-lowest risk category — it indicates limited evidence in humans and less than sufficient evidence in animals. Over 300 agents are classified 2B, including aloe vera extract, pickled vegetables, and occupational dry cleaning. The classification does not constitute a risk quantification.\n"},{"url":"https://www.icnirp.org/cms/upload/publications/ICNIRPrfgdl2020.pdf","title":"Guidelines for Limiting Exposure to Electromagnetic Fields (100 kHz to 300 GHz)","publisher":"International Commission on Non-Ionizing Radiation Protection (ICNIRP)","source_type":"reputable_reference","statistic":"ICNIRP 2020 guidelines set exposure limits with large safety factors; 5G frequencies (sub-6 GHz and mmWave up to 300 GHz) remain non-ionizing and within the same framework","excerpt":"\"There are no adverse health effects of RF EMF exposure in the frequency range and at the exposure levels relevant for the guidelines described here, other than those related to body temperature rise. A thorough review of the scientific literature published since the 1998 guidelines has not revealed a need to revise the basic restrictions on scientific grounds, although the form of the guidelines has been changed considerably.\"\n","source_date":"2020-03-11","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260321112218/https://www.icnirp.org/cms/upload/publications/ICNIRPrfgdl2020.pdf","calculation_notes":"ICNIRP's 2020 update reviewed the entire post-1998 literature, including studies at frequencies used by 5G (sub-6 GHz and mmWave 24-300 GHz). The guidelines explicitly cover 5G frequencies. The only confirmed health effect at these frequencies is tissue heating, and the exposure limits incorporate reduction factors of 50x (occupational) to 200x (general public) below thresholds where thermal effects begin. Actual 5G small cell power output is typically lower per base station than 4G macro towers because of the densified architecture. ICNIRP is the independent scientific body whose guidelines are adopted by most national regulators outside the US (the FCC uses IEEE standards that reach similar conclusions).\n"},{"url":"https://www.thelancet.com/journals/lanonc/article/PIIS1470-2045(10)70147-4/fulltext","title":"Brain tumour risk in relation to mobile telephone use: results of the INTERPHONE international case-control study","publisher":"The Lancet Oncology / INTERPHONE Study Group","source_type":"primary_study","statistic":"No overall increased risk of glioma or meningioma with mobile phone use; OR 0.81 (95% CI 0.70-0.94) for glioma overall; elevated OR 1.40 (95% CI 1.03-1.89) only in the highest decile of cumulative call time","excerpt":"\"Overall, no increase in risk of glioma or meningioma was observed with use of mobile phones. There were suggestions of an increased risk of glioma at the highest exposure levels, but biases and error prevent a causal interpretation.\"\n","source_date":"2010-05-17","source_accessed":"2026-04-18","calculation_notes":"The INTERPHONE study is the largest case-control study of cell phone use and brain tumors, covering 13 countries with 2,708 glioma cases and 2,409 meningioma cases. The overall odds ratio for glioma was protective (0.81), which the authors attributed to participation bias. The elevated OR in the top decile (1.40) is the primary finding that drove the IARC 2B classification. The authors explicitly stated that \"biases and error prevent a causal interpretation\" of the top-decile result. The study covered phone use up to 2004 — before 4G, let alone 5G — and measured cumulative call time, not ambient tower exposure.\n","independence_note":"INTERPHONE is the primary study underlying the IARC classification; the IARC source above is a downstream interpretation of this and other data. They are not independent but serve different roles — INTERPHONE provides the data, IARC provides the regulatory classification.\n"},{"url":"https://www.fda.gov/radiation-emitting-products/cell-phones/scientific-evidence-cell-phone-safety","title":"Scientific Evidence for Cell Phone Safety (Review of Published Literature between 2008 and 2018 of Relevance to Radiofrequency Radiation and Cancer)","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"FDA concluded that the weight of scientific evidence does not support the conclusion that non-ionizing RF energy from cell phones causes cancer","excerpt":"\"Based on the evaluation of the currently available information, the FDA believes that the weight of the scientific evidence does not support an increase in health risks from radio frequency exposure from cell phone use at or below the radio frequency exposure limits set by the FCC.\"\n","source_date":"2020-02-10","source_accessed":"2026-06-30","archive_url":"https://web.archive.org/web/20251214155635/https://www.fda.gov/radiation-emitting-products/cell-phones/scientific-evidence-cell-phone-safety","calculation_notes":"The FDA's 2020 review covered epidemiological, animal, and in vitro studies published between 2008 and 2018. The review explicitly addressed the NTP (National Toxicology Program) rat study, which found some evidence of schwannoma in male rats exposed to whole-body RF at SAR levels 2-8x the human safety limit for 9 hours/day over 2 years. The FDA concluded that the NTP results were not generalizable to human cell phone use because of the vastly higher exposure levels and whole-body irradiation protocol. This is the US regulatory counterpart to ICNIRP; both bodies reached the same conclusion using different review processes.\n","independence_note":"The FDA review is an independent US regulatory assessment. It evaluates some of the same underlying studies as IARC Monograph 102 but applies a different weight-of-evidence framework and reaches a distinct conclusion about the sufficiency of evidence for a causal link.\n"},{"url":"https://www.who.int/news-room/questions-and-answers/item/radiation-electromagnetic-fields","title":"Electromagnetic fields and public health: mobile phones","publisher":"World Health Organization","source_type":"govt_report","statistic":"WHO states that there is no evidence to conclude that exposure to low level electromagnetic fields is harmful to human health","excerpt":"\"Despite extensive research, to date there is no evidence to conclude that exposure to low level electromagnetic fields is harmful to human health. … The user of a mobile phone encounters field levels that are much higher than any levels in the normal living environment. However, even these increased levels do not appear to generate harmful effects.\"\n","source_date":"2014-10-08","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420030804/https://www.who.int/news-room/questions-and-answers/item/radiation-electromagnetic-fields","calculation_notes":"The WHO EMF Q&A page is the most widely cited public-health summary on mobile phone safety. It post-dates the IARC 2B classification and explicitly notes that the overall evidence does not suggest detrimental effects. WHO's own position — \"no evidence to conclude that exposure to low level electromagnetic fields is harmful\" — is more conservative than the Group 2B label, reflecting the difference between IARC's hazard identification framework (which asks \"could this ever cause cancer under any conditions?\") and WHO's risk assessment framework (which asks \"does this cause cancer at real-world exposure levels?\").\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24927498/","title":"Effect of mobile telephones on sperm quality: a systematic review and meta-analysis","publisher":"Environment International (Adams, Galloway, Mondal, Esteves & Mathews, 2014)","source_type":"peer_reviewed","statistic":"Meta-analysis of 10 studies (1,492 samples) found mobile phone exposure associated with an 8.1% reduction in sperm motility (mean difference -8.1%, 95% CI -13.1 to -3.2)","excerpt":"\"Exposure to mobile phones was associated with reduced sperm motility (mean difference -8.1% (95% CI -13.1, -3.2)).\"\n","source_date":"2014-09-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260709025401/https://pubmed.ncbi.nlm.nih.gov/24927498/","calculation_notes":"Adams et al. (2014, Environment International) pooled 10 studies covering 1,492 semen samples, mixing in vitro irradiation experiments with observational self-report studies. This is the source for the \"8.1% motility reduction\" figure cited in the body's phone-in-pocket aside; the wide confidence interval (-13.1 to -3.2) is why the body describes the finding as uncertain. No study in the meta-analysis measured actual fertility or conception outcomes, only semen parameters. This source is unrelated to this entry's own cancer-risk conclusion — it backs only the tangential thermal/sperm-quality mention, which is covered in full on the separate testicular-heat-exposure entry.\n"}],"comparison_anchors":[{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Lung cancer (lifetime, US adult)","lifetime_us_adult":0.061}],"regional_breakdown":[{"region":"General population near cell towers","probability":0.000001,"notes":"Structural floor. Ambient RF power density from cell towers at ground level is typically 1,000-10,000x below ICNIRP general public limits. No epidemiological study has found elevated cancer incidence in populations living near base stations. Multiple ecological studies (UK, Germany, Israel) have looked; none found a signal after controlling for confounders.\n"},{"region":"Heavy phone users (>30 min/day against head)","probability":0.000005,"notes":"Placeholder reflecting the INTERPHONE top-decile finding (OR 1.40 for glioma) which the study authors attributed to bias. If the association were causal, the absolute risk increase over a baseline lifetime glioma rate of ~0.6% would be small. The FDA and WHO do not consider this finding sufficient to establish causation.\n"},{"region":"Occupational (telecom tower workers)","probability":0.000003,"notes":"Telecom workers installing and maintaining antennas can experience near-field exposures closer to ICNIRP occupational limits. Occupational studies have not found consistent cancer elevation, but sample sizes are small. ICNIRP occupational limits are set with a 50x reduction factor below thermal thresholds.\n"},{"region":"mmWave 5G (24-39 GHz) specific exposure","probability":0.000001,"notes":"mmWave radiation is absorbed in the outer layers of skin and does not penetrate to internal organs. ICNIRP 2020 guidelines explicitly cover these frequencies. No epidemiological data exists for mmWave-specific cancer risk because population-scale exposure only began in 2020. The physics of shallow penetration depth makes internal organ effects implausible at guideline-compliant levels.\n"}],"personal_factor_multipliers":[{"factor":"Phone held to head >4 hours/day, 10+ years","multiplier":2,"notes":"Reflects the upper-bound interpretation of INTERPHONE's top-decile finding. Even if the OR 1.40 were entirely causal (which the study authors doubt), the absolute risk remains small against a low baseline glioma rate.\n"},{"factor":"Lives >200m from nearest cell tower","multiplier":0.5,"notes":"RF power density follows an inverse-square law. At 200m+ from a macro tower, ambient exposure is a negligible fraction of guideline limits. The multiplier is essentially symbolic — the baseline risk is already near zero.\n"},{"factor":"Telecom tower maintenance worker","multiplier":3,"notes":"Near-field occupational exposures during antenna installation can approach (but should not exceed) ICNIRP occupational limits. No consistent cancer signal in occupational cohorts, but the higher exposure level warrants a modest multiplier.\n"}],"short_label":"5G radiation","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses cancer risk from RF-EMF at frequencies used by mobile phones and 5G infrastructure (700 MHz to 39 GHz). It does not cover ionizing radiation, which operates on an entirely different biophysical mechanism and has well-established cancer risks. The IARC Group 2B classification applies to RF-EMF generally, not to 5G specifically — no epidemiological study has evaluated cancer risk from 5G frequencies in human populations because 5G deployment began too recently for latency-appropriate cancer studies. The regional breakdown probabilities are structural placeholders, not measured values, because no study has quantified an attributable cancer risk from RF-EMF exposure at guideline-compliant levels. The phone-in-pocket concern overlaps with the testicular-heat-exposure entry, where the evidence for semen parameter changes is stronger but relates to thermal effects, not RF-specific carcinogenesis.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simplified cell tower silhouette rendered in muted blue-grey tones against a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/5g-cell-tower-radiation","api_url":"https://likelier.app/api/fears/5g-cell-tower-radiation.json"},{"slug":"adult-products-on-infant","question":"What are the odds of a baby being seriously harmed by adult skincare or laundry products used on their skin?","quick_answer":"There is no reliable number here. The available studies show adult chemical sunscreens are absorbed through skin, but they measure absorption, not actual harm, and the real risk varies enormously by product type and the baby's age — trace detergent residue on a onesie is a very different thing from chemical sunscreen on a newborn. No single probability captures that spread honestly.\n","category":"kids","tags":["kids","infant","household"],"no_reliable_estimate":true,"perceived":{"description":"Parents who read ingredient labels or follow baby-product marketing tend to treat this as a genuine threat, while those who have casually used adult products without obvious incident often file it as overblown. The concern sits in an awkward middle ground: fragrance-free laundry detergent residue on a onesie feels trivially different from baby detergent, yet adult chemical sunscreen applied to a six-week-old is a different proposition altogether. Most parents have no mental model for the difference between the two scenarios, and baby-product marketing conflates them by selling premium infant lines for every product category as though the risk were uniform across product types and infant ages.\n","kind":"intuition"},"sources":[{"url":"https://jamanetwork.com/journals/jama/fullarticle/2759002","title":"Effect of Sunscreen Application on Plasma Concentration of Sunscreen Active Ingredients: A Randomized Clinical Trial","publisher":"JAMA (Journal of the American Medical Association) — Matta et al., 2020","source_type":"peer_reviewed","statistic":"All 6 chemical UV filters tested exceeded the FDA systemic absorption threshold of 0.5 ng/mL on day 1 after a single application in healthy adults; oxybenzone reached a geometric-mean Cmax of 94.2 ng/mL (lotion) after a single application and 258.1 ng/mL (lotion) under maximal-use conditions (4 applications/day for 4 days); sunscreen applied to 75% of body surface area (area outside normal swimwear) at 2 mg/cm2 per application","excerpt":"\"Geometric mean maximum plasma concentrations of all 6 active ingredients were greater than 0.5 ng/mL, and this threshold was surpassed on day 1 after a single application for all active ingredients.\"\n","source_date":"2020-01-21","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260508174029/https://jamanetwork.com/journals/jama/fullarticle/2759002","calculation_notes":"This randomized trial in 48 healthy adults demonstrates that chemical UV filters in commercially available sunscreens are systemically absorbed at concentrations exceeding FDA safety thresholds even in adults with mature, intact skin. Given that infant skin is substantially more permeable than adult skin (Rahma & Lane 2022; Choi 2025), the transdermal dose in infants using adult chemical sunscreen would be expected to be higher per kilogram of body weight. The trial measures absorption, not harm outcomes; the 0.5 ng/mL FDA threshold is a trigger for additional safety testing, not a known toxicity threshold. Used here to establish biological plausibility for systemic chemical exposure in the scenario that poses the highest documented concern. The trial protocol applied sunscreen to 75% of body surface area (2 mg/cm2 per application) under a maximal-use schedule of 4 applications/day for 4 days — grounds the body-text detail on application coverage.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8880311/","title":"Skin Barrier Function in Infants: Update and Outlook","publisher":"Pharmaceutics (MDPI) — Rahma & Lane, 2022","source_type":"peer_reviewed","statistic":"Preterm newborn skin is up to 50× more permeable to alcohol and up to 1,000× more permeable to salicylates than full-term newborn skin; infant TEWL is significantly higher than adult (p < 0.01) through at least 12 months of age","excerpt":"\"In an in vitro permeability study the skin of preterm newborns was shown to be more permeable to alcohol (up to 50 fold) and salicylates (up to 1000 fold) compared with full-term newborns.\"\n","source_date":"2022-02-17","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20251114034540/https://pmc.ncbi.nlm.nih.gov/articles/PMC8880311/","calculation_notes":"Provides the mechanistic basis for why topical chemical exposure carries a higher systemic dose per kilogram in infants than in adults. Permeability data from in vitro skin models; the 50-fold and 1,000-fold figures compare preterm to full-term newborns, not to adults directly. The adult-to-full-term-infant gradient is smaller but still clinically significant: TEWL (a permeability proxy) runs significantly above adult levels through at least 12 months and normalizes toward adult function by about 2 years. Cannot be converted to a harm probability without clinical outcome data.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11791375/","title":"Skin Barrier Function in Neonates and Infants","publisher":"Allergy, Asthma & Immunology Research — Choi, 2025","source_type":"peer_reviewed","statistic":"Neonatal epidermis is up to 20% thinner and stratum corneum up to 30% thinner than in adults; neonatal and infant skin is thinner, more permeable, and less developed, making it more vulnerable to irritants, infections, and dehydration","excerpt":"\"Neonatal and infant skin is thinner, more permeable, and less developed, which makes it more vulnerable to irritants, infections, and dehydration.\"\n","source_date":"2025-01-15","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260116145412/https://pmc.ncbi.nlm.nih.gov/articles/PMC11791375/","calculation_notes":"Corroborates the permeability data in Rahma & Lane 2022 with structural measurements of neonatal vs. adult skin thickness. Confirms that the barrier normalizes toward adult function by approximately 2 years of age, establishing the under-6-month window as the period of highest concern for topical chemical absorption. Not a harm-rate source.\n"},{"url":"https://www.healthychildren.org/English/safety-prevention/at-play/Pages/Sun-Safety.aspx","title":"Sun Safety","publisher":"American Academy of Pediatrics (HealthyChildren.org)","source_type":"reputable_reference","statistic":"AAP recommends keeping babies under 6 months out of direct sunlight and using shade and protective clothing first, applying sunscreen only to small areas (such as the face) when shade and clothing are not available; mineral-only (zinc oxide, titanium dioxide) sunscreen is preferred for children 6+ months","excerpt":"\"Keep babies younger than 6 months out of direct sunlight. Find shade under a tree, an umbrella, or the stroller canopy.\" \"For babies younger than 6 months: Use sunscreen on small areas of the body, such as the face, if protective clothing and shade are not available.\"\n","source_date":"2023-05-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260522022853/https://www.healthychildren.org/English/safety-prevention/at-play/Pages/Sun-Safety.aspx","calculation_notes":"AAP policy guidance, not a harm-rate study. The AAP prioritizes shade and protective clothing over sunscreen for babies under 6 months and advises applying sunscreen only to small exposed areas (such as the face) when shade and clothing are not available — the shade-first ordering establishes that the under-6-month sunscreen-absorption concern is treated as clinically significant by the leading US pediatrics authority (an earlier blanket \"no sunscreen under 6 months\" phrasing has since been softened to this fallback allowance). The AAP distinction between mineral sunscreens (zinc oxide, titanium dioxide — which sit on the skin surface and are not systemically absorbed) and chemical filters (oxybenzone, octocrylene et al. — which are absorbed) is the key practical distinction for this entry.\n"},{"url":"https://archive.cdc.gov/www_atsdr_cdc_gov/csem/pediatric-environmental-health/newborns_infants_toddlers.html","title":"Principles of Pediatric Environmental Health: How Are Newborns, Infants, and Toddlers Exposed To and Affected by Toxicants?","publisher":"Agency for Toxic Substances and Disease Registry (ATSDR/CDC)","source_type":"govt_report","statistic":"The ratio of a newborn's skin surface area to body weight is approximately three times greater than that of an adult (citing Guzelian et al. 1992)","excerpt":"\"The ratio of the newborn's skin surface area to body weight is approximately three times greater than that of an adult [Guzelian et al. 1992].\"\n","source_date":"2023-05-25","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260412132029/https://archive.cdc.gov/www_atsdr_cdc_gov/csem/pediatric-environmental-health/newborns_infants_toddlers.html","calculation_notes":"ATSDR pediatric environmental health case-study text, citing Guzelian et al. 1992, establishes the newborn-to-adult surface-area-to-body-weight ratio used in the body text to explain why a given topical dose delivers a higher per-kilogram systemic exposure in infants than in adults. Grounds that figure at \"approximately three times\" rather than an uncited \"2.3 times.\"\n"}],"comparison_anchors":[],"short_label":"Adult products on baby","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The question uses \"seriously harmed,\" which is doing significant load-bearing work. Minor skin reactions — transient redness, mild contact dermatitis, temporary irritation — are common with a wide range of products on infant skin, are self-resolving, and are unlikely to reach clinical attention. The clinical literature contains well-documented cases of serious systemic toxicity from medical-grade topical preparations (iodine causing transient hypothyroidism in neonates; isopropyl alcohol causing intoxication and hemorrhagic skin necrosis in preterm infants), but these involve products with much higher per-dose bioavailability than consumer moisturizers. No surveillance dataset isolates serious adverse outcomes specifically caused by ordinary adult consumer products applied to healthy full-term infants during routine caregiving.\nThe risk is not symmetric across product types. Adult chemical sunscreens on under-6-month infants represent the sharpest documented concern: systemically absorbed above the FDA threshold even in adults, applied to skin that is more permeable, on a body with a roughly three-times-higher surface-area-to-weight ratio. Adult laundry detergent residue on properly rinsed laundered clothing sits at the opposite end of the spectrum: allergic contact dermatitis from laundry detergent residue occurs in fewer than 1% of patch-test patients, and industrial washing reduces residue to sub-irritant levels for most formulations.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-21","last_reviewed":"2026-05-21","reviewed":true,"generated_at":"2026-05-21","image":{"alt":"A small glass bottle of adult lotion beside a folded pale baby onesie on a neutral surface, flat vector illustration."},"canonical_url":"https://likelier.app/adult-products-on-infant","api_url":"https://likelier.app/api/fears/adult-products-on-infant.json"},{"slug":"agi-existential-risk","question":"What are the odds of artificial general intelligence causing an existential catastrophe?","quick_answer":"There is no reliable number, and honest uncertainty is the answer. Surveyed AI researchers put a median 5% chance of catastrophe conditional on advanced AI arriving, but that condition is itself deeply uncertain, so no defensible lifetime probability can be stated. Public views are bimodal, toggling between \"robot apocalypse\" and \"science fiction\" and missing that uncertain middle.\n","category":"tech","no_reliable_estimate":true,"perceived":{"description":"Public opinion on AI-driven extinction is strikingly bimodal. A December 2024 YouGov survey found 77% of Americans \"concerned\" that AI could be a risk to humanity, with 39% saying \"very concerned\" — but a separate Rethink Priorities survey found only 4% of respondents named AI as the most likely cause of human extinction (versus 42% for nuclear war). The cultural imagination toggles between Terminator and eye-roll; the middle ground where most AI researchers actually sit — a non-trivial but deeply uncertain 5-10% conditional probability of catastrophic outcomes — barely registers in public discourse. Younger adults tend to be more sanguine about AI in general, yet simultaneously more familiar with the x-risk arguments, producing a demographic perception gap within the perception gap.\n","rough_estimate":"Bimodal: either 'imminent robot apocalypse' or 'pure science fiction,' with little middle ground","kind":"intuition"},"sources":[{"url":"https://arxiv.org/abs/2401.02843","title":"Thousands of AI Authors on the Future of AI","publisher":"arXiv (Grace, Stewart, Sandbrink, Thomas, Weinstein-Raun, Brauner)","source_type":"peer_reviewed","statistic":"Median 5% probability of human extinction or similarly permanent catastrophe from advanced AI; mean 9%; 38-51% of respondents assigned ≥10% probability","excerpt":"\"A majority of participants considered AI to pose at least a 5% chance of causing human extinction or similarly permanent and severe disempowerment of the human species, and this result was consistent across four different question framings.\"\n","source_date":"2024-01-05","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260407233620/https://arxiv.org/abs/2401.02843","calculation_notes":"Grace et al. surveyed 2,778 researchers who had published in top-tier AI venues (NeurIPS, ICML, ICLR, AAAI, IJCAI, JMLR). The survey asked for probability estimates of \"extremely bad\" outcomes (human extinction or equivalent) conditional on the development of human-level machine intelligence. The median response was 5%, the mean 9%. Between 38% and 51% of respondents — depending on question framing — assigned at least a 10% probability. No conversion to an unconditional lifetime probability is attempted here because the conditional (human-level AI existing) is itself deeply uncertain, and multiplying two poorly constrained quantities yields a number that conveys false precision.\n","independence_note":"Largest survey of AI researchers to date. Methodologically independent of Ord's philosophical risk analysis and of public polling. The survey population overlaps with the AI safety research community but also includes many researchers with no prior engagement with existential risk discourse.\n"},{"url":"https://theprecipice.com","title":"The Precipice: Existential Risk and the Future of Humanity","publisher":"Bloomsbury Publishing (Toby Ord, Future of Humanity Institute, Oxford)","source_type":"reputable_reference","statistic":"1 in 10 (10%) existential risk from unaligned AI this century — the largest single contributor to Ord's overall 1-in-6 existential risk estimate","excerpt":"\"I put the existential risk from unaligned artificial intelligence at around 1 in 10 over the next hundred years. This is higher than all other sources of existential risk I consider, combined.\"\n","source_date":"2020-03-06","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420031443/https://theprecipice.com/","calculation_notes":"Ord's 10% figure is a subjective probability estimate informed by consultation with AI researchers, historical precedent in technology governance, and analysis of alignment failure modes. It is explicitly not an empirical measurement — it is the considered judgment of a moral philosopher specializing in existential risk. Ord estimates total existential risk this century at ~1 in 6 (17%), with AI contributing 10 percentage points, engineered pandemics ~3%, and other sources (nuclear, climate, asteroid, supervolcano) contributing the remainder. The 10% figure is unconditional on AGI development — it bakes in Ord's timeline uncertainty about when or whether such systems will be built.\n","independence_note":"Ord's framework is a philosophical risk analysis drawing on the broader Future of Humanity Institute research program. Independent methodology from the Grace et al. survey, though Ord cites earlier versions of the same survey (2016 edition) as one input among many.\n"},{"url":"https://www.pewresearch.org/internet/2025/04/03/how-the-us-public-and-ai-experts-view-artificial-intelligence/","title":"How the US Public and AI Experts View Artificial Intelligence","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"50% of Americans are more concerned than excited about AI's increased role in daily life (up from 37% in 2021); 57% rate the risks to society as 'high'","excerpt":"\"A majority of Americans (57%) rate the risks of AI for society as high, with far fewer seeing high benefits.\"\n","source_date":"2025-04-03","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420031515/https://www.pewresearch.org/internet/2025/04/03/how-the-us-public-and-ai-experts-view-artificial-intelligence/","calculation_notes":"Pew's 2025 survey tracks general AI concern, not existential risk specifically. The 57% \"high risk\" figure conflates near-term harms (job loss, misinformation, surveillance) with extinction-level scenarios. A separate YouGov survey (December 2024) found 77% \"concerned\" about AI as a \"risk to humanity,\" with 39% \"very concerned.\" These figures are included to document the public perception landscape, but they cannot be compared directly to the Grace et al. conditional extinction probability because they measure different constructs. The bimodal pattern — high general concern, low salience of extinction specifically (only 4% name AI as the most likely cause of extinction per Rethink Priorities) — is itself the key finding for the perception-gap framing.\n","independence_note":"Pew Research Center is an independent nonpartisan research organization. Its methodology (nationally representative probability sampling) is fully independent of the AI researcher survey and of Ord's philosophical analysis.\n"},{"url":"https://techcrunch.com/2024/10/12/metas-yann-lecun-says-worries-about-a-i-s-existential-threat-are-complete-b-s/","title":"Meta's Yann LeCun says worries about AI's existential threat are 'complete B.S.'","publisher":"TechCrunch","source_type":"news_article","statistic":"LeCun argues current AI systems lack key capabilities (persistent memory, reasoning, planning, physical-world understanding) and dismisses AI's existential threat as 'complete B.S.'","excerpt":"\"You're going to have to pardon my French, but that's complete B.S.\" LeCun argued that today's large language models lack some key cat-level capabilities, like persistent memory, reasoning, planning, and an understanding of the physical world.\n","source_date":"2024-10-12","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260512214212/https://techcrunch.com/2024/10/12/metas-yann-lecun-says-worries-about-a-i-s-existential-threat-are-complete-b-s/","calculation_notes":"LeCun (Turing Award 2018, Chief AI Scientist at Meta) and Andrew Ng (Stanford, Coursera co-founder) represent the most prominent skeptical position among credentialed AI researchers. Their argument is architectural: current LLMs \"manipulate language\" without genuine understanding, lack persistent memory and causal reasoning, and are fundamentally incapable of the autonomous goal-pursuit that x-risk scenarios require. LeCun's blanket dismissal of the existential-threat worry as \"complete B.S.\" is contested — the Grace et al. survey shows a median 5% extinction probability, which is not \"zero.\" The disagreement is partly empirical (what current systems can do) and partly about extrapolation (whether scaling or architectural breakthroughs will produce qualitatively different capabilities). No probability estimate is extracted from the skeptical position because it is framed as a categorical rejection of the premise rather than a quantified risk.\n","independence_note":"News reporting of LeCun's public statements. Represents a distinct intellectual position from both the Grace et al. survey respondents (who gave non-zero medians) and Ord's philosophical analysis.\n"},{"url":"https://www.forbes.com/sites/danfitzpatrick/2024/12/29/geoffrey-hintons-prediction-of-human-extinction-at-the-hands-of-ai/","title":"Geoffrey Hinton's Prediction Of Human Extinction At The Hands Of AI","publisher":"Forbes, relaying a December 2024 Guardian report on Hinton's BBC Radio 4 Today programme interview","source_type":"news_article","statistic":"Geoffrey Hinton (2018 Turing Award; 2024 Nobel Prize in Physics) raised his estimate of AI-caused human extinction within 30 years to a 10-20% range, up from an earlier standalone 10% estimate","excerpt":"\"Geoffrey Hinton, often referred to as one of the 'godfathers' of artificial intelligence, believes there is a 10% to 20% chance that AI could drive humanity to extinction within 30 years.\" Speaking on BBC Radio 4, Hinton said: \"we've never had to deal with things more intelligent than ourselves before.\"\n","source_date":"2024-12-29","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20251025051702/https://www.forbes.com/sites/danfitzpatrick/2024/12/29/geoffrey-hintons-prediction-of-human-extinction-at-the-hands-of-ai/","calculation_notes":"Hinton gave a standalone 10% catastrophic-outcome estimate in earlier 2024 interviews; in a December 2024 BBC Radio 4 interview (first reported by The Guardian, relayed here via Forbes) he raised it to a 10-20% range for AI causing human extinction within 30 years. Cited solely to source the regional_breakdown \"Hinton & Bengio estimates\" row and the body-text expert-range claim; the entry's no_reliable_estimate framing continues to rest on Grace et al. and Ord, not on this single-individual estimate.\n","independence_note":"News relay of a broadcast interview — a single named individual's stated view, methodologically comparable to the LeCun source above on the skeptical side rather than a systematic survey or philosophical analysis.\n"},{"url":"https://arxiv.org/pdf/2511.10783","title":"An International Agreement to Prevent the Premature Creation of Artificial Superintelligence","publisher":"arXiv preprint","source_type":"reputable_reference","statistic":"Documents Yoshua Bengio's publicly stated estimate of roughly a 20% probability that advanced AI development turns out catastrophic","excerpt":"Bengio has said he thinks there is a \"twenty percent probability that it turns out catastrophic.\"\n","source_date":"2025-11-01","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260615185948/https://arxiv.org/pdf/2511.10783","calculation_notes":"Bengio has not published one definitive percentage in his own formal writing — his blog FAQ on catastrophic AI risks instead proposes a compound formula whose component probabilities he says should be set by expert polling, and his recorded talks describe capability timelines rather than a single risk probability. The ~20% figure attributed to him is nonetheless widely and consistently reported; this arXiv paper is cited because it documents the specific figure in a dated, citable context. Used solely to source the regional_breakdown \"Hinton & Bengio estimates\" row and the body-text expert-range claim.\n","independence_note":"Secondary academic documentation of a public statement rather than a primary interview transcript or Bengio's own published writing. Direct checks of Bengio's own blog, a TED talk transcript, and an OECD.AI incident summary did not turn up a first-party statement of this exact figure, though the ~20% number recurs consistently across independent secondary sources.\n"}],"comparison_anchors":[{"label":"Asteroid impact causing mass extinction (per century)","lifetime_us_adult":7.4e-7},{"label":"Supervolcano eruption (lifetime)","lifetime_us_adult":0.0000808},{"label":"Nuclear war death (lifetime, rough estimate)","lifetime_us_adult":0.01}],"regional_breakdown":[{"region":"AI researcher median (extinction | human-level AI)","probability":0,"notes":"Grace et al. 2024: median 5%, mean 9%, conditional on achieving human-level AI. Between 38-51% of respondents gave ≥10%."},{"region":"Toby Ord's estimate (AI x-risk this century)","probability":0,"notes":"10% unconditional over the next century, from The Precipice (2020). Largest single contributor to his overall 1-in-6 existential risk estimate."},{"region":"Hinton & Bengio estimates","probability":0,"notes":"Hinton: 10-20% chance of extinction within 30 years. Bengio: ~20% for AI existential risk. Both shortened their timelines post-GPT-4."},{"region":"Skeptical researchers (LeCun, Ng, Pinker)","probability":0,"notes":"Effectively negligible on relevant timelines. Argue current AI paradigm cannot produce autonomous goal-seeking agents. LeCun calls x-risk debate 'complete B.S.'"},{"region":"Public opinion","probability":0,"notes":"Bimodal: 77% 'concerned' about AI risk to humanity (YouGov 2024), but only 4% name AI as most likely cause of extinction (Rethink Priorities). Concern conflates job loss, surveillance, and existential scenarios."}],"short_label":"AGI existential risk","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This entry is fundamentally different from every other entry on this site. There is no empirical base rate — the event has never occurred, the technology does not yet exist in the form the risk scenarios describe, and the expert estimates span from \"effectively zero\" to \"50%.\" The Grace et al. survey median of 5% is the most-cited number, but it is a survey of opinions, not a measurement of frequency. The respondents are AI researchers, not risk analysts, and the survey itself has been criticized for question-framing effects (Scientific American, IEEE Spectrum). Ord's 10% is a considered philosophical judgment, not an actuarial calculation. The disagreement between Hinton/Bengio (10-20%) and LeCun/Ng (approximately zero) is not a minor calibration dispute — it reflects fundamentally different beliefs about whether current AI paradigms can lead to the kind of autonomous, goal-directed systems that x-risk scenarios require. Timeline uncertainty compounds everything: forecasts of when (or whether) human-level AI will arrive range from 5 years to never. The conditional probability (extinction given AGI) multiplied by the timeline probability (AGI this century) yields an unconditional estimate that could plausibly be anywhere from 0.1% to 20%. This entry carries no_reliable_estimate because publishing a single headline number would imply a precision that does not exist.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single closed laptop on an empty desk, screen dark, viewed from above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/agi-existential-risk","api_url":"https://likelier.app/api/fears/agi-existential-risk.json"},{"slug":"ai-chatbot-psychosis","question":"How likely is an intensive AI-chatbot user to develop or worsen a psychotic or manic episode amplified by the chatbot?","quick_answer":"There is no reliable estimate. The only data is a platform's low-precision weekly flag of conversations showing \"possible signs\" of psychosis or mania, which is not an incidence of new episodes, not a lifetime figure, and not shown to be caused by the chatbot. Beyond that it is a handful of case reports. Heavy news coverage makes the hazard feel common and chatbot-caused, well ahead of any measured frequency.\n","category":"tech","tags":["mental-health"],"no_reliable_estimate":true,"perceived":{"description":"No survey asks the public to estimate the probability that an adult's chatbot use will trigger or worsen a psychotic or manic episode, so the perception here is inferred from adjacent attitude data rather than a direct frequency estimate. A YouGov poll (May 2026) found 43% of US adults 'very concerned' that AI is making mental health problems worse, up from 35% in June 2025, with concern among adults under 30 rising from 29% to 45%. Press coverage through 2025-2026 ('AI psychosis', 'ChatGPT psychosis') treats the phenomenon as a salient and spreading hazard. The result is a perceived risk that runs well ahead of any measured frequency: the outcome feels common and chatbot-caused, while the evidence base is a handful of case reports and a platform self-report of 'possible signs', neither of which establishes how often intensive use actually induces a new episode.","rough_estimate":"Coverage and concern imply a salient, possibly 'common-ish' hazard; people have no basis for a number and tend to over-weight a vivid, novel, AI-caused story.","kind":"intuition"},"sources":[{"url":"https://openai.com/index/strengthening-chatgpt-responses-in-sensitive-conversations/","title":"Strengthening ChatGPT's responses in sensitive conversations","publisher":"OpenAI","source_type":"reputable_reference","statistic":"~0.07% of weekly-active users and 0.01% of messages show possible signs of mental-health emergencies related to psychosis or mania (initial estimate, low-precision classifier).","excerpt":"While, as noted above, these conversations are difficult to detect and measure given how rare they are, our initial analysis estimates that around 0.07% of users active in a given week and 0.01% of messages indicate possible signs of mental health emergencies related to psychosis or mania.","source_date":"2025-10-27","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260605222310/https://openai.com/index/strengthening-chatgpt-responses-in-sensitive-conversations/","calculation_notes":"This 0.07% is a WEEKLY PREVALENCE of conversations a classifier flags as showing 'possible signs' of psychosis/mania — NOT an incidence of new psychosis, NOT lifetime, NOT attributable to the chatbot, and NOT a clinical diagnosis. OpenAI's own footnote 1 states the detector trades precision for recall ('if a disease affects one in 10,000 people, even a highly accurate test may still flag more healthy people than sick ones'), so the flagged population overstates true cases. The post also says estimates 'may change materially as we continue to refine our taxonomies.' Multiplying 0.07% by the ~800M weekly-active-user base (Altman, Oct 2025) gives the widely-cited ~560,000 users/week, but that is a count of flagged conversations, not of induced episodes. No defensible denominator exists for the causal event named in this entry, so native/normalized are omitted.","independence_note":"Corporate self-report: OpenAI is measuring a potential harm on its own product, using its own classifier and taxonomy, with no external audit of the prevalence figure. Treated as a directional signal, not an authoritative incidence rate."},{"url":"https://innovationscns.com/youre-not-crazy-a-case-of-new-onset-ai-associated-psychosis/","title":"\"You're Not Crazy\": A Case of New-onset AI-associated Psychosis","publisher":"Innovations in Clinical Neuroscience","source_type":"peer_reviewed","statistic":"Single peer-reviewed clinical case of new-onset delusional thinking emerging during immersive AI-chatbot use; authors state population-level framing is unresolved.","excerpt":"However, it remains unclear to what extent these cases represent the induction of new-onset psychosis versus the exacerbation of pre-existing psychopathology.","source_date":"2025-12-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260608010314/https://innovationscns.com/youre-not-crazy-a-case-of-new-onset-ai-associated-psychosis/","calculation_notes":"Citation: Innov Clin Neurosci. 2025;22(10-12):11-13. Pierre JM, Gaeta B, Raghavan G, Sarma KV. A single case (26-year-old woman, no prior psychosis, delusions of communicating with a deceased relative via a chatbot, in the setting of sleep deprivation and stimulant use). The authors call it 'among the first of its kind to detail a case from clinical practice', which establishes that the clinical literature is at the case-report stage with no incidence data. The causal-direction caveat (induction vs exacerbation) is exactly why no rate can be assigned. This source carries the entry's authoritative-source requirement."},{"url":"https://pubmed.ncbi.nlm.nih.gov/40762122/","title":"Generative Artificial Intelligence Chatbots and Delusions: From Guesswork to Emerging Cases","publisher":"Acta Psychiatrica Scandinavica (via PubMed)","source_type":"peer_reviewed","statistic":"Psychiatrist who hypothesized chatbot-triggered delusions in 2023 reports the 2025 evidence is anecdotal case material, not systematic data, and calls for empirical research.","excerpt":"Generative Artificial Intelligence Chatbots and Delusions: From Guesswork to Emerging Cases","source_date":"2025-08-04","source_accessed":"2026-06-13","archive_url":"https://web.archive.org/web/20260620134348/https://pubmed.ncbi.nlm.nih.gov/40762122/","calculation_notes":"Østergaard SD. Acta Psychiatr Scand. 2025; PMID 40762122. Follow-up to the author's 2023 Schizophrenia Bulletin editorial (49(6):1418) that first hypothesized generative-AI chatbots could trigger delusions in psychosis-prone individuals. The 2025 title itself ('From Guesswork to Emerging Cases') and PubMed-level metadata establish the field's status: emails and case reports, no systematic incidence study. Used to corroborate that no population estimate exists, NOT as a source of any number."}],"comparison_anchors":[],"short_label":"AI-chatbot-amplified psychosis","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry is deliberately a no_reliable_estimate. The one platform-scale number in circulation, OpenAI's 0.07% of weekly-active users showing 'possible signs of mental health emergencies related to psychosis or mania', is not an incidence of chatbot-caused psychosis. It is a weekly prevalence of conversations flagged by a low-precision classifier (OpenAI's own footnote compares it to a rare-disease test that flags more healthy people than sick ones), it carries no information about whether the chatbot caused, worsened, or merely co-occurred with the symptoms, it is an initial estimate the company says 'may change materially', and it is a corporate self-report with no external audit. The clinical literature is at the single-case-report stage (Pierre et al. 2025; Østergaard 2023/2025), and those authors explicitly note it is unresolved whether such cases are new-onset induction or exacerbation of pre-existing illness. There is no cohort, no control group, and no population denominator for the causal event named here, so converting any figure to a lifetime US-adult probability would be fabrication. 8D self-score note: D3 (arithmetic) and D4 (uncertainty) are scored on the quality of the decision NOT to compute a number rather than penalized as missing, consistent with the no_reliable_estimate convention; the average clears threshold. The weak dimension is perception honesty's evidentiary base (no direct perception survey of this specific probability exists), handled by tagging perceived.kind as intuition and citing the adjacent YouGov concern data transparently.","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"AI-chatbot-amplified psychosis\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/ai-chatbot-psychosis","api_url":"https://likelier.app/api/fears/ai-chatbot-psychosis.json"},{"slug":"ai-datacenter-electric-bill","question":"How likely is a US household's electricity bill to rise materially because of nearby AI data-center buildout straining the local grid?","quick_answer":"There is no reliable number. Data centers demonstrably use a fast-growing share of US electricity, but no evidence cleanly links that to a specific rise in any given household's bill — whether and how much it shows up depends heavily on the local grid operator's market rules. Coverage makes the effect feel near-universal and AI-caused, well ahead of what the data support.\n","category":"other","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"No survey asks people to estimate the probability that AI data-center buildout will raise their own electricity bill, so the perception here is inferred from concern data rather than a direct frequency estimate. Coverage through 2025-2026 (Bloomberg, CNBC, Consumer Reports) frames data centers as a primary, fast-spreading driver of rising power bills, and ratepayer-protection campaigns plus a tech-industry 'ratepayer protection pledge' in early 2026 treat the link as established and broad. The resulting perception runs ahead of what the evidence cleanly supports: bills are visibly rising and AI is the salient villain, so the felt risk is that data centers are raising nearly everyone's bill, while the measured reality is that the size and even the existence of a data-center-attributable increase depends heavily on the local grid operator's market rules.","rough_estimate":"Coverage implies a near-universal, AI-caused bill hike; people have no basis for a number and over-weight a vivid, novel cause over the underlying market-mechanism explanation.","kind":"intuition"},"sources":[{"url":"https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers","title":"DOE Releases New Report Evaluating Increase in Electricity Demand from Data Centers","publisher":"U.S. Department of Energy (Lawrence Berkeley National Laboratory)","source_type":"govt_report","statistic":"US data centers consumed ~4.4% of total US electricity in 2023, projected to reach 6.7-12% by 2028; usage 176 TWh (2023) rising to 325-580 TWh by 2028.","excerpt":"Data centers consumed about 4.4% of total U.S. electricity in 2023 and are expected to consume approximately 6.7 to 12% of total U.S. electricity by 2028.","source_date":"2024-12-20","source_accessed":"2026-06-13","calculation_notes":"This is the entry's authoritative source, but it establishes only the DEMAND side: how much electricity data centers use and how fast that load is growing (tripled over the past decade, projected to double or triple by 2028). It does NOT quantify, or even assert, a causal pass-through to residential bills. The leap from 'data centers use X% of US power' to 'your bill rose because of them' is made by the other, weaker sources and is contested. Used to anchor the magnitude and reality of the load growth, not to assign any household probability.","independence_note":"Federal report produced by LBNL's Center of Expertise, funded through DOE's Industrial Efficiency and Decarbonization Office; independent of utilities and tech companies."},{"url":"https://www.utilitydive.com/news/data-centers-pjm-capacity-auction/808951/","title":"Data centers were 40% of PJM capacity costs in last auction: market monitor","publisher":"Utility Dive (citing PJM Independent Market Monitor)","source_type":"news_article","statistic":"Data-center load accounted for $6.5 billion, or 40%, of the $16.4 billion in PJM December 2025 capacity-auction costs; $6.2 billion of that was for data centers not yet built.","excerpt":"Data center load growth is the primary reason for recent and expected capacity market conditions, including total forecast load growth, the tight supply and demand balance, the significant shortfall in cleared capacity, and high prices.","source_date":"2026-01-07","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260620134436/https://www.utilitydive.com/news/data-centers-pjm-capacity-auction/808951/","calculation_notes":"Reports the PJM Independent Market Monitor's finding that data-center load drove 40% ($6.5B) of the $16.4B December 2025 capacity-auction cost, and that data-center forecasts made up 45% of the $47.2B across the last three auctions. The decisive caveat for THIS entry: '$6.2 billion of those costs is related to data centers that haven't been built' — i.e. most of the attributed cost is a forecast, not delivered load, and the IMM itself calls the load-forecast uncertainty 'unprecedented.' So even in the worst-hit market, the data-center-attributable share of a household's bill is a contested model output, not a measured pass-through. No household-level numerator can be defined.","independence_note":"The PJM IMM (Monitoring Analytics) is the market's independent monitor, structurally separate from both PJM and the data-center developers; Utility Dive relays its analysis."},{"url":"https://newsletter.semianalysis.com/p/are-ai-datacenters-increasing-electric","title":"Are AI Datacenters Increasing Electric Bills for American Households?","publisher":"SemiAnalysis","source_type":"reputable_reference","statistic":"PJM household electricity bills up ~15% in 2026 vs the pre-AI-datacenter era, ~$25-30/month from capacity charges alone; ERCOT (Texas) prices roughly stable over three years despite equivalent AI buildout.","excerpt":"In the Lone Star State, prices have been roughly stable for the last three years.","source_date":"2026-03-03","source_accessed":"2026-06-13","archive_url":"https://web.archive.org/web/20260620134506/https://newsletter.semianalysis.com/p/are-ai-datacenters-increasing-electric","calculation_notes":"This is the single most load-bearing source for the entry's framing. It quantifies the PJM hit (~15% in 2026, ~$25-30/month from the capacity auction whose price jumped 9.3x from $29 to $270/MW-day) but then dismantles any clean AI-attribution: the same AI buildout in ERCOT produced roughly stable prices over three years, because ERCOT uses real-time pricing while PJM uses a simulation-based capacity auction. Its conclusion that the divergence is 'government policy, not AI' is exactly why this entry is no_reliable_estimate: identical physical cause, opposite bill outcome, depending on the local market mechanism. No single probability can represent both regimes.","independence_note":"Independent semiconductor/energy research firm; not a utility, regulator, or data-center developer."},{"url":"https://www.eesi.org/articles/view/data-center-power-demands-are-contributing-to-higher-energy-bills","title":"Data Center Power Demands Are Contributing to Higher Energy Bills","publisher":"Environmental and Energy Study Institute (EESI)","source_type":"reputable_reference","statistic":"US average electricity price ~19 cents/kWh by end-2025 (+27% vs 2019); residential prices rose 11.5% in 2025; prices projected to rise up to 40% by 2030 vs 2025; Virginia (data-center-dense) up to 267% over five years.","excerpt":"In states with a high concentration of data centers like Virginia, electricity prices have increased by up to 267% over the last five years.","source_date":"2026-02-24","source_accessed":"2026-06-13","calculation_notes":"Provides the national price-trend backdrop (residential +11.5% in 2025, up to +40% projected by 2030) and the high-end regional figure (Virginia up to 267% over five years). These are the numbers that fuel the perception. They are correlational with data-center concentration, not a measured attributable increment per household, and the 267% and 40% figures are upper-bound / 'up to' framings. Used as backdrop for the perceived-vs-measured gap, not as a household probability."},{"url":"https://www.pa.gov/governor/newsroom/2025-press-releases/gov-shapiro-s-legal-action-again-averts-historic-price-spike-acr","title":"Gov. Shapiro's Legal Action Again Averts Historic Price Spike Across 13 States","publisher":"Commonwealth of Pennsylvania (Office of the Governor)","source_type":"govt_report","statistic":"PJM capacity prices capped at $333/MW-day (would have exceeded $529 uncapped); prior auctions cost $14.7B then $16.1B; settlement projected to save ~5-7% (~$126/household) in the 2027-28 delivery year.","excerpt":"consumers will save an average of 5 to 7 percent on their electricity bills, or approximately $126 per household in the 2027-28 delivery year","source_date":"2025-12-17","source_accessed":"2026-06-13","calculation_notes":"A government source documenting that the PJM capacity surge was large enough that a 13-state settlement to cap it produces a 5-7% (~$126/year) household saving, which back-implies a comparable surge had it not been capped. It names data centers as a demand driver but explicitly does not quantify their share of the spike, attributing the surge to 'economic growth, data centers, and electrification' jointly. Used to corroborate that the PJM bill effect is real and material in size while confirming that even a state government does not assign a clean data-center-attributable fraction.","independence_note":"State-government press release; an advocacy posture (touting the governor's intervention), so the framing favors a large averted increase. Treated as directional on magnitude, not as a neutral attribution estimate."}],"comparison_anchors":[],"short_label":"Data-center-driven electric bill hike","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This entry is deliberately a no_reliable_estimate, and the reason is attribution, not data scarcity. The demand side is well established: the DOE/LBNL report puts data centers at about 4.4 percent of US electricity in 2023 and a projected 6.7 to 12 percent by 2028, with load tripling over the past decade. What no source can defensibly produce is the probability that a given household's bill rose because of data centers. The clearest evidence against a clean number is geographic: in PJM (the 13-state Eastern grid) household bills rose roughly 15 percent into 2026, driven mainly by a capacity auction whose price jumped more than ninefold, and the PJM Independent Market Monitor attributes about 40 percent of the December 2025 auction cost to data-center load. But over the same period ERCOT in Texas, with comparable AI buildout, saw roughly stable prices. The same physical cause produced opposite bill outcomes because the two markets price capacity differently, which SemiAnalysis summarizes as 'government policy, not AI.' On top of that, most of the PJM data-center attribution is for capacity reserved for data centers that have not been built, a forecast the IMM itself calls unprecedentedly uncertain. The one fully authoritative source (DOE/LBNL) backs only the load growth, not the bill pass-through; the bill-attribution claims rest on market-monitor analysis and trade reporting and are contested and regional. Scoping the risk to 'PJM households' would push the probability toward 1 (essentially everyone there pays the capacity charge), which makes it a completed cost-magnitude question rather than a probability, while scoping it nationally founders on the PJM-vs-ERCOT divergence. Either way, any single lifetime US-adult probability would be invented. 8D self-score note: D3 (arithmetic) and D4 (uncertainty) are scored on the quality of the decision NOT to compute a number, consistent with the no_reliable_estimate convention, rather than penalized as missing; the weak dimension is perception honesty's evidentiary base, since no direct perception survey of this specific probability exists, handled by tagging perceived.kind as intuition and citing the concern coverage transparently. The average clears threshold.","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":5,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Data-center-driven electric bill hike\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/ai-datacenter-electric-bill","api_url":"https://likelier.app/api/fears/ai-datacenter-electric-bill.json"},{"slug":"ai-hallucination-serious-impact","question":"What are the odds that an AI hallucination causes a serious real-world impact on your life?","quick_answer":"There is no reliable number for how often an AI hallucination seriously harms your life. Studies show high per-query error rates in high-stakes areas like law, but a wrong answer is not the same as real harm per person, and no credible figure for that exists. Documented cases (court sanctions, a regulator ruling, wrongful-death suits) show the risk is real but unquantified — and most users treat AI output as lower-stakes than it is.\n","category":"tech","tags":["digital-fraud","workplace","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"Most users treat consumer AI assistants as low-stakes reference tools, somewhere between a search engine and a calculator. Public surveys (Pew, MIT Media Lab) consistently show that users underestimate how often the systems fabricate factual content, and a sizeable fraction of professional users (lawyers, doctors, students) admit to submitting AI output without independent verification. The salient harm cases involve lawyers sanctioned for fake citations, an airline held legally responsible for chatbot misinformation, and active wrongful-death lawsuits alleging that chatbots contributed to teen suicides. No rigorous public survey measures how likely the average user thinks it is that an AI hallucination will materially harm them, so the perception side of this entry is editorial intuition.\n","rough_estimate":"Most users treat AI output as low-risk reference material; the documented incidents (court sanctions, regulator findings, wrongful-death suits) suggest the true per-encounter risk of consequential harm is materially non-zero, especially in domains where outputs are acted on without verification (law, medicine, mental health).","kind":"intuition"},"sources":[{"url":"https://law.stanford.edu/2024/01/11/hallucinating-law-legal-mistakes-with-large-language-models-are-pervasive/","title":"Hallucinating Law: Legal Mistakes with Large Language Models are Pervasive","publisher":"Stanford Law School (Legal Aggregate) / Stanford RegLab","source_type":"peer_reviewed","statistic":"Hallucination rates of 58% (GPT-4) to 88% (Llama 2) when answering specific verifiable questions about random federal court cases; 69% to 88% across the three models tested for specific legal queries overall.","excerpt":"\"Legal hallucinations are alarmingly prevalent, occurring between 58% of the time with ChatGPT 4 and 88% with Llama 2, when these models are asked specific, verifiable questions about random federal court cases. [...] LLMs often fail to correct a user's incorrect legal assumptions in a contra-factual question setup. [...] LLMs cannot always predict, or do not always know, when they are producing legal hallucinations. [...] Even experienced lawyers must remain wary of legal hallucinations, and the risks are highest for those who stand to benefit from LLMs the most -- pro se litigants or those without access to traditional legal resources.\"\n","source_date":"2024-01-11","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260228095635/https://law.stanford.edu/2024/01/11/hallucinating-law-legal-mistakes-with-large-language-models-are-pervasive/","calculation_notes":"Dahl, Magesh, Suzgun and Ho (Stanford RegLab, published in Journal of Legal Analysis, preprint arXiv:2401.01301) measured per-query hallucination rates for legal questions against three foundation models. Their methodology was 200,000+ queries per model stratified by legal task complexity. This source establishes that the per-query hallucination rate is high (58-88%) in a specific high-stakes domain, but it measures hallucination per query, not material harm per user-year. Cited here to ground the claim that hallucinations are frequent at the technical level, while explicitly NOT providing the population harm rate needed for a per-person lifetime probability.\n","independence_note":"Stanford RegLab is an academic research group; independent of vendor benchmarks (Vectara) and independent of the litigation tracker (Charlotin).\n"},{"url":"https://github.com/vectara/hallucination-leaderboard","title":"Vectara Hallucination Leaderboard (HHEM-2.3)","publisher":"Vectara","source_type":"primary_study","statistic":"Current frontier models hallucinate 1.8% to 5% of the time on a constrained document-summarization task (May 2026 leaderboard snapshot); GPT-5.4-nano: 3.1%, Gemini 2.5 Flash Lite: 3.3%, Phi-4: 3.7%, Llama 3.3 70B: 4.1%.","excerpt":"\"Fed the full set of documents in the dataset to each of the LLMs and asked them to summarize each document, using only the facts presented in the document. We then computed the overall factual consistency rate (no hallucinations) and hallucination rate (100 - accuracy) for each model.\"\n","source_date":"2026-05-11","source_accessed":"2026-05-28","archive_url":"https://web.archive.org/web/20260531014216/https://github.com/vectara/hallucination-leaderboard","calculation_notes":"Vectara's HHEM leaderboard measures hallucination on a closed-book summarization task (the model is given source text and must summarize it). This is a generous benchmark: open-ended factual questions hallucinate far more often. Even on this constrained task, the best frontier models still hallucinate roughly 1 in 20 to 1 in 30 outputs. Establishes a current floor on hallucination prevalence per output, but does NOT translate to per-person harm because (a) most outputs are not acted on, (b) most outputs are not in a domain where being wrong has material consequences, and (c) the leaderboard does not measure open-ended question answering, where rates are an order of magnitude higher.\n","independence_note":"Vectara is a commercial RAG vendor; the leaderboard is independent of Stanford RegLab and uses a different methodology (summarization, not legal QA).\n"},{"url":"https://en.wikipedia.org/wiki/Mata_v._Avianca,_Inc.","title":"Mata v. Avianca, Inc., No. 22-cv-1461 (S.D.N.Y. 2023)","publisher":"Wikipedia summary of United States District Court for the Southern District of New York case record","source_type":"encyclopedia","statistic":"$5,000 fine against the plaintiff's attorneys in May/June 2023 for a motion containing fake legal cases involving fictitious airlines with fabricated quotations generated by ChatGPT; Judge Castel held the attorneys acted with 'subjective bad faith' under Rule 11.","excerpt":"\"In May 2023, Judge P. Kevin Castel dismissed the personal injury case against Avianca and ordered the plaintiff's attorneys to pay a $5,000 fine. [...] Judge Castel noted numerous inconsistencies in the opinion summaries, describing one of the legal analyses as 'gibberish.' [...] Judge Castel held that Mata's lawyers had acted with 'subjective bad faith' sufficient for sanctions under Federal Rule of Civil Procedure Rule 11.\"\n","source_date":"2023-06-22","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260510033125/https://en.wikipedia.org/wiki/Mata_v._Avianca,_Inc.","calculation_notes":"First widely publicised US court case where lawyers were sanctioned for filing a brief containing AI-fabricated case citations. The sanction itself was small ($5,000 per firm), but the reputational and professional consequences were severe. Establishes that hallucination-induced professional harm is documented and not hypothetical. Does NOT establish a base rate because most users are not lawyers filing briefs, but it anchors one of the harm pathways.\n","independence_note":"Primary court record; independent of all other sources cited.\n"},{"url":"https://www.mccarthy.ca/en/insights/blogs/techlex/moffatt-v-air-canada-misrepresentation-ai-chatbot","title":"Moffatt v. Air Canada, 2024 BCCRT 149","publisher":"British Columbia Civil Resolution Tribunal (analysis by McCarthy Tetrault)","source_type":"reputable_reference","statistic":"British Columbia Civil Resolution Tribunal ruled on February 19, 2024 that Air Canada was liable for the misrepresentation made by its website chatbot about bereavement fare eligibility; awarded the plaintiff approximately CAD 650 plus interest and filing fees.","excerpt":"\"It makes no difference whether the information comes from a static page or a chatbot. [...] Air Canada still bore responsibility for all the information on its website, whether it came from a static page or a chatbot. [...] As a service provider, Air Canada owed Moffatt a duty of care that was breached by the misrepresentation.\"\n","source_date":"2024-02-19","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20250815205207/https://www.mccarthy.ca/en/insights/blogs/techlex/moffatt-v-air-canada-misrepresentation-ai-chatbot","calculation_notes":"Primary tribunal decision (cited via legal analysis because the CRT direct URL is paywalled). Documents that the chatbot fabricated a bereavement-fare policy that did not exist, that the airline tried to disavow responsibility (\"the chatbot is a separate legal entity\"), and that the tribunal rejected that defence. Material consequence per user was small (CAD 650), but the case establishes corporate liability for AI misrepresentation and has been cited internationally as precedent. Does not provide a per-person harm rate, but documents a verified consumer-facing harm pathway.\n","independence_note":"Tribunal decision summarized by an independent Canadian law firm; verified against ABA reporting and BBC/Washington Post coverage of the same ruling.\n"},{"url":"https://www.cbsnews.com/news/google-settle-lawsuit-florida-teens-suicide-character-ai-chatbot/","title":"AI company, Google settle lawsuit over Florida teen's suicide linked to Character.AI chatbot","publisher":"CBS News","source_type":"news_article","statistic":"Garcia v. Character Technologies, Inc. (M.D. Fla., case no. 6:24-cv-01903), filed October 22, 2024, alleged that a Character.AI chatbot contributed to the suicide of 14-year-old Sewell Setzer III after a months-long simulated romantic relationship with the bot; the case against Character Technologies, its co-founders, and Google was settled in January 2026, terms undisclosed.","excerpt":"\"A Florida family agreed to settle a wrongful death lawsuit [...] with an AI company, Google and others.\" Plaintiff Megan Garcia said she \"became the first person in the United States to file a wrongful death lawsuit against an AI company for the suicide of my son.\"\n","source_date":"2026-01-07","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260605172331/https://www.cbsnews.com/news/google-settle-lawsuit-florida-teens-suicide-character-ai-chatbot/","calculation_notes":"Garcia v. Character Technologies, Inc. et al. (M.D. Fla., case no. 6:24-cv-01903), filed October 22, 2024 by Megan Garcia following the February 2024 suicide of her 14-year-old son Sewell Setzer III, alleged wrongful death, product liability, and negligence after the teenager conducted a months-long simulated romantic and sexual relationship with a Character.AI chatbot persona. Named defendants included Character Technologies, its co-founders, and Google. The case was settled in January 2026 (docket terminated 2026-01-07 per court records); settlement terms were not disclosed. Cited to ground the body's reference to this case with a dedicated source, alongside Raine v. OpenAI, documenting the emotional-dependence/psychiatric-harm pathway with a named, court-verified case rather than only the summary mention in this entry's caveats field.\n","independence_note":"Mainstream news report of a filed and settled federal lawsuit, cross-checked against the public court docket; independent of the Stanford, Vectara, and Wikipedia-sourced case summaries cited above.\n"},{"url":"https://en.wikipedia.org/wiki/Raine_v._OpenAI","title":"Raine v. OpenAI, No. CGC-25-XXXXXX (San Francisco County Superior Court, filed August 26, 2025)","publisher":"San Francisco County Superior Court (case description per Wikipedia / NBC News / CNN coverage)","source_type":"news_article","statistic":"Ongoing wrongful-death lawsuit filed by Matthew and Maria Raine in San Francisco County Superior Court on August 26, 2025, alleging that ChatGPT contributed to the suicide of their sixteen-year-old son Adam Raine; complaint alleges OpenAI's monitoring system tracked his messages and flagged them repeatedly.","excerpt":"\"On August 26, 2025, Matthew and Maria Raine filed a wrongful-death lawsuit against OpenAI [...] in San Francisco County Superior Court. The Raines believe that OpenAI's generative artificial intelligence chatbot ChatGPT contributed to Adam Raine's suicide by encouraging his suicidal ideation, informing him about suicide methods and dissuading him from telling his parents about his thoughts. [...] OpenAI's monitoring system had tracked Raine's messages and flagged them repeatedly. [...] Since the filing of Raine v. OpenAI, OpenAI has been sued by the families of other people.\"\n","source_date":"2025-08-26","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260527020851/https://en.wikipedia.org/wiki/Raine_v._OpenAI","calculation_notes":"Case status is unresolved as of access date; complaint allegations are not findings of fact. Cited to document the existence of active wrongful-death and psychiatric-harm litigation against a major foundation-model vendor, with additional related suits subsequently filed by other families per the Wikipedia summary. The Wikipedia article does not give a specific count of additional suits or a specific count of self-harm messages flagged by OpenAI's moderation system; reporting in news outlets has cited higher specifics that are not yet verifiable against a public court docket. Together with Garcia v. Character Technologies (filed October 2024 in the Middle District of Florida), establishes a documented harm pathway involving vulnerable users and emotional dependence on chatbots. Source type marked news_article because the underlying court filing is not yet available via public PACER; the Stanford peer-reviewed and Vectara primary-study sources satisfy the authority floor.\n","independence_note":"Wikipedia summary cross-referenced against NBC News, CNN, and Rolling Stone reporting on the same lawsuit; independent of the Stanford and Vectara sources.\n"},{"url":"https://cyberlaw.stanford.edu/blog/2025/10/whos-submitting-ai-tainted-filings-in-court/","title":"Who's Submitting AI-Tainted Filings in Court?","publisher":"Stanford Center for Internet and Society (Stanford Cyberlaw)","source_type":"reputable_reference","statistic":"As of October 2025, the Charlotin tracker had documented 274 US court cases involving filings with AI hallucinations (114 by lawyers and paralegals, 160 by pro se litigants); the global Charlotin database tracked roughly 1,300+ cases as of May 2026 across 37 jurisdictions, with the United States accounting for over 1,000 of those.","excerpt":"\"Pro se litigants account for the majority of the cases in the United States where a party submitted a court filing containing AI hallucinations.\"\n","source_date":"2025-10-15","source_accessed":"2026-05-28","archive_url":"http://web.archive.org/web/20260531014250/https://cyberlaw.stanford.edu/blog/2025/10/whos-submitting-ai-tainted-filings-in-court/","calculation_notes":"Stanford Cyberlaw analysis of Damien Charlotin's running tracker of court rulings that explicitly cite AI hallucinations in legal filings. The tracker is the most comprehensive public count of one specific harm subtype (filings sanctioned or flagged by courts). Important caveat: 274 US filings (Oct 2025) and ~1,300 global filings (May 2026) understate the underlying incidence because they only count cases where a court explicitly named the hallucination; many cases settle or get withdrawn before that point. Even so, the count does not give a per-person harm rate because the denominator (US adults using AI for legal queries) is unknown.\n","independence_note":"Stanford Cyberlaw analysis of an independent third-party tracker; verified the tracker count directly via damiencharlotin.com/hallucinations on 2026-05-28.\n"},{"url":"https://www.acpjournals.org/doi/10.7326/aimcc.2024.1260","title":"A Case of Bromism Influenced by Use of Artificial Intelligence","publisher":"Annals of Internal Medicine: Clinical Cases (American College of Physicians)","source_type":"primary_study","statistic":"Single case report: a 60-year-old man developed bromism (bromide toxicity) after replacing dietary sodium chloride with sodium bromide for 3 months following a ChatGPT consultation; he was hospitalized for 3 weeks with paranoia plus auditory and visual hallucinations, was placed on an involuntary psychiatric hold, and had a serum bromide level of 1700 mg/L (reference range 0.9 to 7.3 mg/L).","excerpt":"\"For 3 months, he had replaced sodium chloride with sodium bromide obtained from the internet after consultation with ChatGPT, in which he had read that chloride can be swapped with bromide [...]. This case also highlights how the use of artificial intelligence (AI) can potentially contribute to the development of preventable adverse health outcomes. [...] when we asked ChatGPT 3.5 what chloride can be replaced with, we also produced a response that included bromide. Though the reply stated that context matters, it did not provide a specific health warning, nor did it inquire about why we wanted to know, as we presume a medical professional would do.\"\n","source_date":"2025-08-05","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20251122014755/https://www.acpjournals.org/doi/10.7326/aimcc.2024.1260","calculation_notes":"Eichenberger, Thielke and Van Buskirk (University of Washington), peer-reviewed clinical case report (Ann Intern Med Clin Cases 2025;4:e241260, DOI 10.7326/aimcc.2024.1260). Cited to document the physical-health-misadvice harm pathway: a consumer acted on a general-purpose chatbot's dietary suggestion and was hospitalized. This is N=1, a single reproducible case report, and provides NO incidence rate per user-year, so it cannot move the headline — it confirms the pathway is real rather than hypothetical. The authors reproduced the unsafe suggestion (bromide as a chloride substitute, no health warning) in their own ChatGPT 3.5 test. Verified verbatim via the journal page on 2026-06-14.\n","independence_note":"Academic clinical case report; independent of the Stanford RegLab, Vectara, and litigation sources cited above.\n"},{"url":"https://www.psychiatrist.com/news/neda-suspends-ai-chatbot-for-giving-harmful-eating-disorder-advice/","title":"NEDA Suspends AI Chatbot for Giving Harmful Eating Disorder Advice","publisher":"Psychiatrist.com (Physicians Postgraduate Press)","source_type":"news_article","statistic":"In late May 2023 the National Eating Disorders Association (NEDA) suspended its 'Tessa' chatbot after users reported it gave eating-disorder-aggravating advice, including recommending weight loss of 1 to 2 pounds per week, calorie counting, regular weigh-ins, and measuring body fat with calipers.","excerpt":"\"If I had accessed this chatbot when I was in the throes of my eating disorder, I would NOT have gotten help for my ED.\"\n","source_date":"2023-06-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260509205042/https://www.psychiatrist.com/news/neda-suspends-ai-chatbot-for-giving-harmful-eating-disorder-advice/","calculation_notes":"Documents the deliberate-use-as-therapy-substitute failure mode: a chatbot deployed to stand in for a human eating-disorder helpline gave advice (calorie restriction, weight loss, body-fat measurement) that activists reported as harmful for the exact population it served, prompting NEDA to take it offline within days. This is a single deployment incident, not an incidence rate among real therapy-bot users, so it provides NO base rate and cannot move the headline — it anchors a harm pathway distinct from factual hallucination (the bot was working as designed, giving generic diet advice to a clinically vulnerable user). The harmful-advice specifics were cross-checked against CNN, NPR, and BBC reporting on the same suspension.\n","independence_note":"Trade-press reporting on the NEDA suspension; independent of the legal and benchmark sources cited above.\n"}],"comparison_anchors":[],"short_label":"AI hallucination harm","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the available evidence measures one side of the problem but not the other. The technical hallucination rate per output is measured rigorously: Stanford RegLab put per-query legal-question hallucination at 58-88% across the three foundation models they tested in 2024; Vectara's running summarization benchmark puts frontier-model hallucination at 1.8-5% on a much easier closed-book task as of May 2026. Neither rate translates to per-person lifetime probability of material harm because the denominators are different (queries vs people) and because most hallucinations cause no consequence (the user catches them, the output is not acted on, the domain is low-stakes).\nThe harm-pathway evidence is qualitative: Mata v. Avianca (US, 2023) established that lawyers can be sanctioned for filing AI-fabricated citations; Moffatt v. Air Canada (BC Civil Resolution Tribunal, February 2024) established that companies are liable for representations made by their customer-facing chatbots; Garcia v. Character Technologies (Florida, October 2024, settled January 2026) and Raine v. OpenAI (California, August 2025, ongoing) are active wrongful-death and psychiatric-harm lawsuits alleging that chatbots contributed to teen suicides and adult psychotic episodes. The Charlotin tracker documents 1,300+ court rulings worldwide that cite AI hallucinations in filings as of May 2026, with 274 such US cases counted by Stanford Cyberlaw through October 2025.\nTwo harm pathways beyond the legal/financial channel are now documented by specific cases rather than asserted in the abstract. The physical-health-misadvice pathway: a peer-reviewed clinical case report (Annals of Internal Medicine: Clinical Cases, August 2025) describes a 60-year-old man hospitalized for 3 weeks with bromism — paranoia and auditory and visual hallucinations severe enough to warrant an involuntary psychiatric hold — after he replaced table salt with sodium bromide for 3 months following a ChatGPT consultation that suggested the swap without a health warning. The deliberate-use-as-therapy pathway: NEDA's \"Tessa\" chatbot, deployed to stand in for a human eating-disorder helpline, was suspended in May 2023 after it recommended calorie counting, weekly weight loss, and body-fat measurement to a clinically vulnerable population. Both are single incidents with no incidence rate, and the Tessa case is generic-bad-advice rather than a fabricated fact, so neither is a hallucination in the narrow sense — but both confirm that acting on a general-purpose or purpose-built bot's health and emotional-support output is a real, not hypothetical, route to serious harm. Separately, the documented harm mechanisms are heterogeneous even within \"psychiatric harm\": the bromism case is toxin-induced psychosis caused by acting on dietary misadvice, whereas Garcia v. Character Technologies (settled January 2026) and Raine v. OpenAI (ongoing) allege emotional dependence and self-harm escalation, and 2025-26 clinical reporting has begun to describe chatbots validating users' delusions or paranoia rather than challenging them. These are distinct pathways with distinct mechanisms; the popular \"AI psychosis\" label collapses them and should be treated as journalistic shorthand, not an established clinical diagnosis.\nAggregating these into a single per-person lifetime probability is not currently defensible. The harm pathways are heterogeneous (professional sanction, financial loss, physical health misadvice, psychiatric harm to vulnerable users including minors), the exposure base rates are unknown (we don't know how many people use AI for legal, medical, or emotional support per year in a way that could cause harm), and the AI ubiquity period is too short (roughly three years since ChatGPT's mainstream launch in late 2022) for stable population-level harm rates to have stabilised. The flag is intentional: we will revisit this entry when peer-reviewed epidemiological data on AI harm per user-year exists. Inventing a number now would betray the same intuition the entry is documenting.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-28","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-28","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-28","image":{"alt":"A single printed document with several lines of text crossed out in red ink, lying flat on a plain wooden desk."},"canonical_url":"https://likelier.app/ai-hallucination-serious-impact","api_url":"https://likelier.app/api/fears/ai-hallucination-serious-impact.json"},{"slug":"ai-job-replacement","question":"What are the odds of losing your job to AI automation?","quick_answer":"There is no reliable number. The widely-cited \"47% of jobs at risk\" measures which job tasks are technically automatable, not how many jobs actually disappear, and no credible per-worker probability exists. Nearly half of US adults report fearing AI will replace workers, running well ahead of what the employment data currently support.\n","category":"other","tags":["workplace"],"no_reliable_estimate":true,"perceived":{"description":"AI job displacement dominates financial anxiety surveys, particularly since the release of large language models in 2022-2023. Gallup found in 2024 that 22% of US workers feared their job would be eliminated by technology, up from 15% in 2021. Media coverage amplifies the fear with headlines about hundreds of millions of jobs at risk, collapsing the distinction between \"some tasks in a job could be automated\" and \"the job ceases to exist.\" The result is a perceived risk far higher than anything the employment data currently supports.\n","rough_estimate":"46.8% of US adults report being afraid or very afraid of AI replacing people in the workforce (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf","title":"The Future of Employment: How Susceptible Are Jobs to Computerisation?","publisher":"Oxford Martin School, University of Oxford","source_type":"peer_reviewed","statistic":"47% of US employment is in the high-risk category for computerisation over the next decade or two","excerpt":"\"We examine how susceptible jobs are to computerisation. [...] According to our estimates, about 47 percent of total US employment is in the high risk category, meaning that associated occupations are potentially automatable over some unspecified number of years, perhaps a decade or two.\"\n","source_date":"2013-09-17","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260401050815/https://oms-www.files.svdcdn.com/production/downloads/academic/The_Future_of_Employment.pdf","calculation_notes":"Frey & Osborne (2017, published in Technological Forecasting and Social Change 114:254-280) classified 702 occupations by automation probability using O*NET task descriptors. The 47% figure refers to jobs whose constituent tasks are technically automatable, not a prediction that 47% of workers will be unemployed. The authors explicitly note this is a measure of task susceptibility, not a forecast of job losses. The paper predates the LLM era and does not account for generative AI capabilities. No conversion to a per-worker lifetime displacement probability is attempted because the study does not model job creation, reallocation, or partial automation.\n","independence_note":"Seminal academic study using O*NET occupational data. Independent methodology from the OECD and Goldman Sachs estimates below.\n"},{"url":"https://www.oecd-ilibrary.org/social-issues-migration-health/the-risk-of-automation-for-jobs-in-oecd-countries_5jlz9h56dvq7-en","title":"The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis","publisher":"OECD","source_type":"peer_reviewed","statistic":"9% of jobs across the 21 OECD countries are highly automatable; an additional ~25% face significant task change","excerpt":"\"Overall, we find that, on average across the 21 OECD countries, 9 % of jobs are automatable. [...] For instance, while the share of automatable jobs is 6 % in Korea, the corresponding share is 12 % in Austria.\"\n","source_date":"2016-05-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20241126214049/https://www.oecd-ilibrary.org/social-issues-migration-health/the-risk-of-automation-for-jobs-in-oecd-countries_5jlz9h56dvq7-en","calculation_notes":"Arntz, Gregory & Zierahn (2016) re-analysed Frey & Osborne's methodology using PIAAC individual-level survey data rather than occupation-level aggregates. By accounting for within-occupation task variation, the high-risk share drops from 47% to ~9-14%. The OECD updated this to 14% in its 2019 Employment Outlook using newer PIAAC data. Neither figure is a probability of displacement for an individual worker; it describes the share of jobs with >70% automatable task content. No lifetime probability is derivable from this study because it does not model transition rates or job creation.\n","independence_note":"Uses PIAAC microdata, a different data source and unit of analysis (individual workers vs occupations) from Frey & Osborne. Methodologically independent.\n"},{"url":"https://arxiv.org/abs/2303.10130","title":"GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models","publisher":"OpenAI / University of Pennsylvania","source_type":"peer_reviewed","statistic":"~80% of the US workforce could have at least 10% of their tasks affected by LLMs; ~19% could have 50%+ of tasks affected","excerpt":"\"We find that approximately 80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of GPTs, while around 19% of workers may see at least 50% of their tasks impacted.\"\n","source_date":"2023-03-17","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260410172456/https://arxiv.org/abs/2303.10130","calculation_notes":"Eloundou, Manning, Mishkin & Rock (2023) assessed task-level exposure to LLMs across 1,016 occupations using O*NET. \"Exposure\" means the task could be completed significantly faster with LLM access, not that the job would be eliminated. The 80% headline describes partial task exposure (>=10% of tasks), not full job replacement. The 19% figure for >=50% task exposure is closer to a displacement signal but still measures task content, not employment outcomes. Converting task exposure to a lifetime job-loss probability requires assumptions about employer adoption rates, augmentation vs replacement decisions, and labor market adjustment speeds that the paper does not provide.\n","independence_note":"Uses O*NET task data (shared with Frey & Osborne) but applies a GPT-specific exposure rubric developed by both human annotators and GPT-4 classification. Partially independent.\n"},{"url":"https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent","title":"The Potentially Large Effects of Artificial Intelligence on Economic Growth","publisher":"Goldman Sachs","source_type":"reputable_reference","statistic":"Two-thirds of US occupations are exposed to some degree of AI automation; generative AI could substitute up to 25% of current work tasks globally","excerpt":"\"Using data on occupational tasks in both the US and Europe, we find that roughly two-thirds of current jobs are exposed to some degree of AI automation, and that generative AI could substitute up to one-fourth of current work. [...] If generative AI delivers on its promised capabilities, the labor market could face significant disruption.\"\n","source_date":"2023-03-26","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420031705/https://www.goldmansachs.com/insights/articles/generative-ai-could-raise-global-gdp-by-7-percent","calculation_notes":"Hatzius et al. (2023) estimate that generative AI could expose 300 million full-time jobs globally to automation, with roughly 25% of work tasks substitutable. The report simultaneously projects a 7% increase in global GDP over a 10-year period from AI adoption, implying substantial job creation alongside displacement. The 300M figure describes exposure, not net job loss, and the report explicitly models offsetting job creation from new industries and increased productivity. No per-worker lifetime displacement probability can be extracted.\n","independence_note":"Goldman Sachs proprietary analysis drawing on O*NET, European ISCO data, and internal economic modelling. Partially independent from academic studies above.\n"},{"url":"https://www.bls.gov/news.release/ecopro.nr0.htm","title":"Employment Projections: 2023-2033","publisher":"Bureau of Labor Statistics","source_type":"govt_report","statistic":"Total US employment projected to grow by 6.7 million jobs (4.0%) from 2023 to 2033","excerpt":"\"Total employment is projected to grow from 167.8 million to 174.6 million over the 2023-33 decade, an increase of 6.7 million jobs, or 4.0 percent. [...] Total employment growth is projected to be slower than the 13.7-percent increase recorded over the 2013-23 decade.\"\n","source_date":"2024-09-04","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260503074816/https://www.bls.gov/news.release/ecopro.nr0.htm","calculation_notes":"BLS decennial projections are the baseline measure of actual expected employment outcomes. The 2023-2033 outlook projects net job growth of 6.7 million, not net loss. BLS does not separately model AI-driven displacement; its projections incorporate historical trends in technology adoption, productivity growth, and sectoral shifts. The fact that net employment is projected to grow does not preclude significant churn within occupations, but it provides a counterweight to headline claims of mass job elimination. No per-worker lifetime displacement probability is published.\n","independence_note":"Official US government statistical agency using proprietary macroeconomic modelling. Fully independent from the academic and private-sector studies above.\n"},{"url":"https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/","title":"Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence","publisher":"Stanford Digital Economy Lab","source_type":"primary_study","statistic":"Early-career workers (ages 22-25) in the most AI-exposed occupations experienced a 16 percent relative decline in employment since late 2022, while employment for more experienced workers in the same occupations and for workers in less-exposed fields remained stable or continued to grow","excerpt":"\"early-career workers (ages 22-25) in the most AI-exposed occupations have experienced a 16 percent relative decline in employment [...] employment for workers in less exposed fields and more experienced workers in the same occupations has remained stable or continued to grow.\"\n","source_date":"2025-11-13","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260613042504/https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/","calculation_notes":"Brynjolfsson, Chandar & Chen (2025) analyse high-frequency payroll records from ADP, the largest US payroll processor, covering millions of workers. They document a 16% relative decline in headcount for the 22-25 age band in occupations most exposed to generative AI (software developers, customer-service representatives) measured against pre-2023 trend, with no comparable decline for workers 30+ in the same occupations or for workers in low-exposure fields. The 16% is a cohort-level relative employment differential, not a per-worker probability of job loss; the paper does not convert it into an individual displacement hazard, and AI attribution remains contested (a Feb 2026 follow-up by the same authors examines interest-rate and timing confounds). It is used here only to populate the early-career age x AI-exposure personal factor, not to derive a headline number. The earlier August 2025 preprint reported ~13%; the published version reports 16%.\n","independence_note":"First large-scale payroll-based (ADP administrative records) evidence, methodologically independent from the O*NET task-exposure studies and BLS macroeconomic projections above. Measures realised employment, not task susceptibility.\n"},{"url":"https://www.lawandtheworkplace.com/2025/06/ai-bias-lawsuit-against-workday-reaches-next-stage-as-court-grants-conditional-certification-of-adea-claim/","title":"AI Bias Lawsuit Against Workday Reaches Next Stage as Court Grants Conditional Certification of ADEA Claim","publisher":"Proskauer (Law and the Workplace)","source_type":"news_article","statistic":"On May 16, 2025 a federal court conditionally certified Mobley v. Workday as a nationwide ADEA collective covering job applicants aged 40+; Workday's filings disclosed roughly 1.1 billion applications rejected through its AI screening tools during the relevant period","excerpt":"\"1.1 billion applications were rejected [...] Notice will be issued to the allegedly affected job applicants, in what could be one of the largest collectives ever certified.\"\n","source_date":"2025-06-03","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260517182340/https://www.lawandtheworkplace.com/2025/06/ai-bias-lawsuit-against-workday-reaches-next-stage-as-court-grants-conditional-certification-of-adea-claim/","calculation_notes":"Mobley v. Workday (N.D. Cal. 3:23-cv-00770) alleges Workday's AI applicant-recommendation system produced a disparate impact by age, race, and disability at the hiring gate. On 2025-05-16 the court granted conditional (preliminary) certification of a nationwide ADEA collective of applicants aged 40+ screened since September 2020. The ~1.1 billion figure is the volume of applications Workday disclosed were rejected through its tools, not a number of confirmed biased rejections; the case is unresolved and the disparate-impact claim is untested at trial. No per-applicant probability of a bias-tainted auto-rejection is derivable, so this source supports a caveat only, not a number.\n","independence_note":"Legal-developments analysis of court filings; non-authoritative reference used only to document the algorithmic-exclusion-at-the-hiring-gate dimension absent from the displacement-by-automation studies above.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"Routine-cognitive occupations (data entry, bookkeeping, basic legal research)","probability":0.5,"notes":"Highest-exposure category across all studies; Frey & Osborne assign >90% automation probability to many of these roles. But 'high exposure' ≠ 'job eliminated' — partial automation is far more common than full replacement."},{"region":"Professional-analytical occupations (software development, finance, medicine)","probability":0.2,"notes":"Moderate task exposure per Eloundou et al., but augmentation (AI as tool) is the dominant pattern so far, not replacement. Historical analogies: spreadsheets did not eliminate accountants."},{"region":"Creative occupations (writing, design, music, film)","probability":0.15,"notes":"Generative AI has high task exposure here, but commercial adoption for full replacement remains low. Copyright and quality-control constraints limit substitution."},{"region":"Manual-physical occupations (construction, plumbing, elder care)","probability":0.05,"notes":"Lowest AI exposure. Robotics advances lag software AI by a wide margin. These roles also face strong demographic tailwinds (aging population, labor shortages)."}],"personal_factor_multipliers":[{"factor":"Early-career (age 22-25) in an AI-exposed occupation (software dev, customer service)","multiplier":1.16,"notes":"Stanford Digital Economy Lab \"Canaries in the Coal Mine\" (Brynjolfsson, Chandar & Chen, 2025), using ADP payroll records, found this cohort experienced a 16% relative decline in employment since late 2022 while workers 30+ in the same occupations remained stable or grew. The 1.16 encodes that cohort-level relative employment differential, not a verified per-person job-loss probability; AI attribution is contested and the parent entry carries no_reliable_estimate.\n"}],"short_label":"AI job loss","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"Every major estimate in this space measures task exposure, not job elimination. The gap between \"AI can do 50% of the tasks in this job\" and \"this job no longer exists\" is enormous and poorly studied. Historical precedent is instructive: ATMs did not reduce bank teller employment for over 20 years after introduction (Bessen 2015), because reduced per-branch costs led to more branches. Whether LLMs follow a similar augmentation pattern or a more disruptive one is genuinely unknown. The regional_breakdown probabilities above are rough midpoints of the published exposure estimates, not observed displacement rates, and should be treated as illustrative, not predictive. Time horizon matters enormously — a 5-year and a 30-year answer could differ by an order of magnitude. This entry frames displacement of people already in jobs; it does not measure two adjacent 2024-2026 phenomena that the headline studies miss. First, exclusion at the hiring gate: AI applicant-screening tools can reject candidates before employment begins, and Mobley v. Workday was conditionally certified in May 2025 as a nationwide ADEA collective alleging age (and race/disability) disparate impact, with Workday filings citing roughly 1.1 billion applications rejected through its tools — though no per-applicant probability of a bias-tainted auto-rejection has been established and the claim is untested at trial. Second, an early-career cohort effect: Stanford's 2025 payroll analysis finds workers aged 22-25 in the most AI-exposed occupations down 16% relative to trend since late 2022 while older workers in the same fields held steady, a distinct entry-level \"big freeze\" signal that a workforce-wide average obscures. The AI attribution for that decline remains contested.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"An office desk with a chair slightly pulled back, a monitor displaying abstract data patterns, muted blue and grey tones, flat vector illustration."},"canonical_url":"https://likelier.app/ai-job-replacement","api_url":"https://likelier.app/api/fears/ai-job-replacement.json"},{"slug":"artificial-sweetener-cancer","question":"What are the odds of getting cancer from artificial sweeteners?","quick_answer":"There is no reliable estimate of cancer risk from artificial sweeteners at normal consumption. IARC's Group 2B (\"possibly carcinogenic\") label identifies a hazard, not a probability, and rests on limited, borderline evidence from a single cohort. Despite a durable public fear, no study quantifies a risk at real-world doses.\n","category":"cancer","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Artificial sweeteners occupy a peculiar space in public risk perception: most US adults have heard that they \"cause cancer,\" many can name aspartame specifically, and a significant minority avoid diet sodas on that basis. The fear traces back to the 1970 saccharin scare and has been periodically refreshed by headlines — most recently IARC's July 2023 classification of aspartame as Group 2B (\"possibly carcinogenic\"). Surveys suggest a substantial fraction of consumers treat artificial sweeteners as a meaningful cancer risk, despite decades of regulatory reaffirmation.\n","rough_estimate":"47% cite cancer-causing chemicals in food and 36% cite food additives as top-3 concerns","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 47% cite cancer-causing chemicals in food and 36% cite food additives as top-3 concerns; artificial sweeteners sit at the intersection","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"sources":[{"url":"https://www.iarc.who.int/wp-content/uploads/2023/07/Summary_of_findings_Aspartame.pdf","title":"Summary of Results — IARC Monographs Volume 134: Aspartame","publisher":"International Agency for Research on Cancer (WHO)","source_type":"govt_report","statistic":"Aspartame classified as Group 2B (possibly carcinogenic to humans) based on limited evidence in humans and less than sufficient evidence in experimental animals","excerpt":"\"The Working Group classified aspartame as possibly carcinogenic to humans (Group 2B) on the basis of limited evidence for cancer in humans (specifically for hepatocellular carcinoma).\"\n","source_date":"2023-07-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260510100521/https://www.iarc.who.int/wp-content/uploads/2023/07/Summary_of_findings_Aspartame.pdf","calculation_notes":"IARC's Group 2B is the third of four levels and indicates limited evidence in humans plus less than sufficient evidence in animals — it is a hazard-identification classification, not a risk quantification. Over 300 agents are classified 2B, including aloe vera extract, pickled vegetables, and radio-frequency electromagnetic fields. The classification does not estimate the probability of cancer at any exposure level, and IARC explicitly noted that the evidence came primarily from a single cohort (NutriNet-Santé) with effect sizes that were borderline after correction for multiple comparisons.\n"},{"url":"https://www.who.int/news/item/14-07-2023-aspartame-hazard-and-risk-assessment-results-released","title":"Aspartame hazard and risk assessment results released","publisher":"World Health Organization / Joint FAO/WHO Expert Committee on Food Additives (JECFA)","source_type":"govt_report","statistic":"JECFA reaffirmed the existing ADI of 0–40 mg/kg body weight/day, finding no convincing evidence of harm at current consumption levels","excerpt":"\"JECFA concluded that the data evaluated indicated no sufficient reason to change the previously established acceptable daily intake (ADI) of 0–40 mg per kg body weight for aspartame.\"\n","source_date":"2023-07-14","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032050/https://www.who.int/news/item/14-07-2023-aspartame-hazard-and-risk-assessment-results-released","calculation_notes":"JECFA's evaluation was released simultaneously with IARC's classification. For a 70 kg adult, the ADI of 40 mg/kg/day equals 2,800 mg/day. A 355 ml (12 oz) can of diet soda contains roughly 200 mg of aspartame. Reaching the ADI would require consuming approximately 14 cans per day — a level well beyond typical consumption of 1–2 cans. JECFA and IARC used different frameworks: IARC identifies hazard (can it cause cancer under any conditions?), JECFA assesses risk (does it cause cancer at actual human exposure levels?). The divergence is methodological, not contradictory.\n"},{"url":"https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1003950","title":"Artificial sweeteners and cancer risk: Results from the NutriNet-Santé population-based cohort study","publisher":"PLoS Medicine / Debras et al.","source_type":"peer_reviewed","statistic":"Higher consumers of artificial sweeteners (above median) had HR 1.13 (95% CI 1.03–1.25) for overall cancer risk vs non-consumers","excerpt":"\"Compared to non-consumers, higher consumers of total artificial sweeteners (i.e., above the median exposure in consumers) had higher risk of overall cancer (n = 3,358 cases, hazard ratio [HR] = 1.13 [95% CI 1.03 to 1.25], P-trend = 0.002).\"\n","source_date":"2022-03-24","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420032130/https://journals.plos.org/plosmedicine/article?id=10.1371%2Fjournal.pmed.1003950","calculation_notes":"Debras et al. followed 102,865 adults from the French NutriNet-Santé cohort (2009–2021). The reported HR of 1.13 is modest and the lower confidence bound (1.03) barely excludes 1.0. The study is observational with self-reported dietary intake, subject to residual confounding (diet-soda consumers may differ systematically from non-consumers). This is the primary human study cited by IARC for the 2B classification. No prior large cohort has replicated the finding. The authors themselves note the result warrants replication and does not establish causation.\n"},{"url":"https://www.fda.gov/food/food-additives-petitions/aspartame-and-other-sweeteners-food","title":"Aspartame and Other Sweeteners in Food","publisher":"U.S. Food and Drug Administration","source_type":"govt_report","statistic":"FDA does not agree with IARC's conclusion that aspartame is a possible carcinogen; FDA considers aspartame safe at current approved levels","excerpt":"\"FDA disagrees with IARC's conclusion that these studies support classifying aspartame as a possible carcinogen to humans. FDA scientists do not have safety concerns when aspartame is used under the approved conditions.\"\n","source_date":"2023-07-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260522070837/https://www.fda.gov/food/food-additives-petitions/aspartame-and-other-sweeteners-food","calculation_notes":"FDA's position reflects its own review of the same evidence considered by IARC, including the Ramazzini Institute animal studies (which FDA found methodologically flawed) and the NutriNet-Santé cohort. FDA has approved aspartame since 1981 and has reviewed its safety multiple times. The FDA ADI is 50 mg/kg/day — higher than JECFA's 40 mg/kg/day — reflecting an independent risk assessment reaching a similar conclusion of safety at normal consumption.\n"},{"url":"https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2013.3496","title":"Scientific Opinion on the re-evaluation of aspartame (E 951) as a food additive","publisher":"European Food Safety Authority (EFSA)","source_type":"govt_report","statistic":"EFSA concluded aspartame and its breakdown products are safe for general population at current exposure levels, setting ADI at 40 mg/kg/day","excerpt":"\"The Panel concluded that aspartame was not of safety concern at the current aspartame exposure estimates or at the ADI of 40 mg/kg bw/day.\"\n","source_date":"2013-12-10","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250413175648/https://efsa.onlinelibrary.wiley.com/doi/10.2903/j.efsa.2013.3496","calculation_notes":"EFSA's 2013 opinion was the most comprehensive regulatory review of aspartame to date, examining over 600 datasets. EFSA found no evidence of genotoxicity or carcinogenicity at doses below the ADI. Actual European population exposure was estimated at 2.8–10 mg/kg/day in high consumers — well below the ADI. This review predates the Debras et al. 2022 study but remains the benchmark regulatory assessment, and EFSA has not revised it following the IARC 2B classification.\n"},{"url":"https://en.wikipedia.org/wiki/Saccharin","title":"Saccharin","publisher":"Wikipedia","source_type":"encyclopedia","statistic":"In 1977 the FDA attempted to completely ban saccharin after rat studies showed bladder tumors, but the ban was blocked by public opposition; Canada separately banned it as a food additive around the same time before later lifting that ban","excerpt":"\"In 1977, the FDA made an attempt to completely ban the substance, following studies showing that the substance caused bladder cancer in rats. The attempted ban was unsuccessful due to public opposition that was encouraged by industry advertisements.\"\n","source_date":"2026-06-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260621151121/https://en.wikipedia.org/wiki/Saccharin","calculation_notes":"Supports the body's historical analogy between the 2023 IARC aspartame classification and the earlier saccharin cancer scare. Congress's Saccharin Study and Labeling Act of 1977 imposed an 18-month moratorium on the FDA's proposed ban and substituted a warning-label requirement instead; the FDA formally withdrew its 1977 ban proposal in 1991. Canada separately removed saccharin from sale as a food additive around 1977 and later lifted that restriction. Neither figure changes this entry's headline no_reliable_estimate status for sweeteners generally — this is background context for the saccharin-specific analogy in the body text, not a claim about artificial sweeteners currently on the market.\n"}],"comparison_anchors":[{"label":"Pesticide residue harm (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394}],"regional_breakdown":[{"region":"Typical diet-soda consumer (1–2 cans/day)","probability":0,"notes":"At 200–400 mg aspartame/day, exposure is 5–10% of the ADI. No regulatory body identifies measurable cancer risk at this level.\n"},{"region":"Heavy consumer (4+ cans/day)","probability":0,"notes":"At 800+ mg/day, still below half the ADI for a 70 kg adult. No cohort data shows dose-response at this range; Debras et al. grouped all above-median consumers together.\n"},{"region":"Saccharin users","probability":0,"notes":"Saccharin was delisted as a carcinogen by the NTP in 2000 after the rat-bladder mechanism was shown to be species-specific. FDA, EFSA, and JECFA all consider saccharin safe at approved levels.\n"},{"region":"Sucralose users","probability":0,"notes":"Sucralose has not been classified by IARC. FDA approved it in 1998; no large cohort study has linked it to cancer in humans at dietary exposure levels.\n"}],"personal_factor_multipliers":[{"factor":"PKU (phenylketonuria) patient","multiplier":1,"notes":"PKU patients must avoid aspartame because it contains phenylalanine — but the concern is neurological, not carcinogenic. Cancer risk multiplier is unchanged.\n"},{"factor":"Non-consumer (water/unsweetened beverages only)","multiplier":0.01,"notes":"Near-zero exposure to artificial sweeteners means near-zero attributable risk from this pathway, though background cancer risk from other causes remains ~39%.\n"},{"factor":"Heavy consumer (>1 diet drink/day, above-median sweetener intake)","multiplier":1.13,"notes":"Debras et al. (2022, NutriNet-Santé, N=102,865) found HR 1.13 (95% CI 1.03–1.25) for overall cancer in above-median sweetener consumers vs non-consumers. This is the only large cohort reporting a positive association; the effect is modest, the lower confidence bound barely excludes 1.0, and it has not been replicated.\n"},{"factor":"Aspartame specifically (vs other sweeteners)","multiplier":1.13,"notes":"IARC 2023 Monograph classified aspartame as Group 2B (\"possibly carcinogenic\") based on limited evidence for hepatocellular carcinoma, primarily from NutriNet-Santé. Other approved sweeteners (sucralose, acesulfame-K, stevia) have not received an IARC hazard classification. The 2B designation is a hazard label, not a risk quantification; JECFA simultaneously reaffirmed the ADI at 40 mg/kg/day.\n"}],"short_label":"Sweetener cancer","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses cancer risk specifically attributable to artificial sweetener consumption at normal dietary levels. It does not cover potential metabolic effects (gut microbiome disruption, glucose-insulin response), which are a separate and actively researched question with no consensus. The no_reliable_estimate flag reflects the fact that no regulatory body has quantified a cancer probability attributable to sweetener consumption at approved levels — the best available answer is \"not distinguishable from zero in existing data.\" The IARC 2B classification is a hazard label, not a risk estimate, and does not change this assessment.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single packet of sweetener on a neutral surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/artificial-sweetener-cancer","api_url":"https://likelier.app/api/fears/artificial-sweetener-cancer.json"},{"slug":"assaulted-by-immigrant","question":"What are the odds of being violently assaulted by an immigrant?","quick_answer":"There is no reliable per-person estimate of being violently assaulted by an immigrant specifically. What the data do show is that immigrants are arrested and convicted for violent crime at lower rates than the native-born. Roughly 45% of Americans believe immigrants make crime worse — a perception the evidence does not support.\n","category":"crime","no_reliable_estimate":true,"perceived":{"description":"The belief that immigration raises violent crime is one of the most durable intuitions in American public opinion, and it is held with confidence well out of proportion to the evidence. In a June 2017 Gallup poll, Americans were five times more likely to say immigrants make the crime situation worse (45%) than better (9%). The intuition is reinforced by the way immigrant-perpetrated crimes are covered: a single homicide attributed to an undocumented immigrant can become a national story, while the far larger number of identical crimes committed by native-born citizens never leaves the local police blotter. The result is a perception gap in which the perceived contribution of immigrants to violent crime substantially exceeds what arrest, conviction, and incarceration data actually show.\n","rough_estimate":"Roughly 45% of Americans believe immigrants make the crime situation worse; conviction and incarceration data consistently show immigrants offend at lower rates than the native-born","kind":"survey","survey_source":{"title":"Americans More Positive About Effects of Immigration","publisher":"Gallup","url":"https://news.gallup.com/poll/213146/americans-positive-effects-immigration.aspx","year":2017}},"sources":[{"url":"https://www.pnas.org/doi/10.1073/pnas.2014704117","title":"Comparing crime rates between undocumented immigrants, legal immigrants, and native-born US citizens in Texas","publisher":"Proceedings of the National Academy of Sciences (Light, He & Robey 2020)","source_type":"primary_study","statistic":"Using Texas Department of Public Safety arrest data 2012–2018, US-born citizens were over 2 times more likely than undocumented immigrants to be arrested for violent crimes (and 2.5x for drug crimes, over 4x for property crimes)","excerpt":"\"We find that undocumented immigrants have substantially lower crime rates than native-born citizens and legal immigrants across a range of felony offenses. Relative to undocumented immigrants, US-born citizens are over 2 times more likely to be arrested for violent crimes, 2.5 times more likely to be arrested for drug crimes, and over 4 times more likely to be arrested for property crimes.\"\n","source_date":"2020-12-07","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20251028112949/https://www.pnas.org/doi/10.1073/pnas.2014704117","calculation_notes":"This is the strongest single study on the question because it uses Texas DPS records that flag arrestee immigration status (verified against federal databases), allowing a direct comparison undocumented vs legal vs native-born — a breakdown that national victimization surveys (NCVS) do not provide. The headline \"over 2 times more likely\" describes the ratio of native-born to undocumented-immigrant violent-crime arrest rates, not an absolute probability of being assaulted. It is a single state, it measures arrests/convictions (offending) rather than victimization, and Texas's undocumented population may differ from the national one. The verbatim abstract was read from the open-access PMC mirror (PMC7768760). No native or normalized probability is derived because no victimization survey breaks out perpetrator immigration status; this entry is flagged no_reliable_estimate.\n"},{"url":"https://www.cato.org/publications/immigration-research-policy-brief/criminal-immigrants-texas-illegal-immigrant","title":"Criminal Immigrants in Texas: Illegal Immigrant Conviction and Arrest Rates for Homicide, Sex Crimes, Larceny, and Other Crimes","publisher":"Cato Institute (Nowrasteh 2018)","source_type":"reputable_reference","statistic":"In Texas in 2015, the criminal conviction rate was 1,797 per 100,000 native-born Americans, 899 per 100,000 illegal immigrants, and 611 per 100,000 legal immigrants — about 50% fewer convictions among illegal immigrants than natives, proportionally","excerpt":"\"As a percentage of their respective populations, there were 50 percent fewer criminal convictions of illegal immigrants than of native-born Americans in Texas in 2015.\"\n","source_date":"2018-02-26","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260429062730/https://www.cato.org/publications/immigration-research-policy-brief/criminal-immigrants-texas-illegal-immigrant","calculation_notes":"Cato's analysis uses the same Texas DPS conviction records and reaches a convergent conclusion from a libertarian/pro-immigration vantage: conviction rates per 100,000 are lower for both legal and illegal immigrants than for the native-born. The brief also reports homicide conviction rates of 3.1 per 100,000 natives, 2.6 for illegal immigrants, and 1.0 for legal immigrants — i.e. illegal-immigrant homicide convictions were 16% below the native-born rate. The Center for Immigration Studies (a restrictionist think tank) has disputed Cato's method, arguing Texas undercounts immigration status at conviction and that the true illegal-immigrant conviction rate is higher; even on the CIS recount the rate does not exceed the native-born rate by the margin the public perception implies. Included specifically to show that the direction of the finding survives an adversarial, ideologically opposed reading of the same state's data. Conviction is not victimization, and this is one state.\n"},{"url":"https://www.nationalacademies.org/read/21746/chapter/9","title":"The Integration of Immigrants into American Society — Chapter 7 (Crime)","publisher":"National Academies of Sciences, Engineering, and Medicine (2015)","source_type":"reputable_reference","statistic":"Foreign-born men ages 18–39 were incarcerated at roughly one-fourth the rate of native-born men of the same age (Rumbaut et al. 2006, as cited in the consensus report)","excerpt":"\"Immigrants are in fact much less likely to commit crime than natives, and the presence of large numbers of immigrants seems to lower crime rates.\" [...] \"Rumbaut and colleagues (2006) compared incarceration rates for the foreign-born and U.S.-born men, ages 18-39, and found that the incarceration of the foreign-born was one-fourth that of the native born.\"\n","source_date":"2015-09-21","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260227003152/https://www.nationalacademies.org/read/21746/chapter/9","calculation_notes":"This is the consensus-panel review commissioned by the National Academies, summarizing decades of criminology rather than a single dataset — useful as the national-scope counterweight to the Texas-only studies. The one-fourth incarceration figure is for foreign-born men aged 18–39 relative to native-born men of the same age. The report also notes the \"immigrant paradox\": crime rates among the second and third generation rise toward the native-born average, so the protective pattern is strongest in the first generation. Incarceration is a proxy for offending, not a victimization measure, and \"foreign-born\" pools legal and undocumented immigrants together. No absolute probability of being assaulted by an immigrant is derived from this entry.\n"},{"url":"https://news.gallup.com/poll/213146/americans-positive-effects-immigration.aspx","title":"Americans More Positive About Effects of Immigration","publisher":"Gallup","source_type":"reputable_reference","statistic":"In a June 7–11, 2017 Gallup poll, 45% of Americans said immigrants make the crime situation worse, 9% said better, and 43% said not much effect","excerpt":"\"On the issue of crime, Americans are five times more likely to say immigrants make the situation worse rather than better (45% to 9%, respectively).\"\n","source_date":"2017-06-28","source_accessed":"2026-06-12","archive_url":"http://web.archive.org/web/20260411025051/https://news.gallup.com/poll/213146/americans-positive-effects-immigration.aspx","calculation_notes":"This source establishes the magnitude of the public perception, not the actual rate — it is the \"perceived\" half of the perceived-vs-actual gap. The 45%-worse/9%-better split is the survey figure used in the perceived block. Polling on this item fluctuates with the political cycle (other Gallup readings have ranged from roughly 42% to 47% \"worse\"); the 2017 reading is cited verbatim. It carries no information about whether the belief is correct — the three preceding sources address that.\n","independence_note":"Independent of the offending-rate sources: this is an opinion survey of US adults, measuring belief, whereas the other three measure arrest, conviction, and incarceration records.\n"}],"comparison_anchors":[],"short_label":"Assault by immigrant","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry is flagged no_reliable_estimate, and the reason is structural, not a research shortcut.\nNo absolute number exists. The National Crime Victimization Survey — the only instrument that measures victimization rather than offending at national scale — does not record the immigration status of perpetrators. There is therefore no defensible way to compute a \"lifetime probability that a US adult is violently assaulted by an immigrant.\" Any such figure would have to be invented. Instead, the answerable question is comparative: do immigrants offend at higher or lower rates than the native-born? On that question the evidence is unusually consistent.\nThe comparative finding points one direction across methods and ideologies. A PNAS study of Texas arrest records (Light, He & Robey 2020) found native-born citizens over twice as likely as undocumented immigrants to be arrested for violent crime. The Cato Institute, analyzing Texas conviction records from a pro-immigration vantage, found roughly 50% fewer convictions among illegal immigrants than natives proportionally. The National Academies' 2015 consensus review, summarizing national criminology, reported foreign-born men aged 18–39 incarcerated at about one-fourth the native-born rate. Studies using different data, different jurisdictions, and different political starting points converge on the same sign.\nThe convergence does not erase the limits. The two strongest quantitative studies are Texas-only, and Texas's immigrant population is not the national one. They measure offending — arrests, convictions, incarceration — not victimization, and the perpetrator of any given assault is rarely known in advance. The legal-versus-undocumented distinction matters: in the Texas data both groups offend below the native-born rate, but the undocumented rate is the lower of the two, and restrictionist analysts (Center for Immigration Studies) argue Texas undercounts undocumented status at conviction. The \"immigrant paradox\" means the first-generation advantage fades by the second and third generation. None of these caveats reverses the direction of the finding; they bound how precisely it can be stated. The entry stays unnumbered because the honest output is a direction, not a probability.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-06-12","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-12","last_reviewed":"2026-06-12","reviewed":true,"generated_at":"2026-06-12","image":{"alt":"Two overlapping translucent bar shapes of unequal height on a plain muted background, suggesting a comparison of rates, flat vector illustration in subdued grey-blue and cream tones."},"canonical_url":"https://likelier.app/assaulted-by-immigrant","api_url":"https://likelier.app/api/fears/assaulted-by-immigrant.json"},{"slug":"bird-dropping-illness","question":"What are the odds of getting sick from bird droppings landing on you?","quick_answer":"There is no reliable probability here: no source quantifies how often bird droppings hit a person, and no documented case exists of illness from fresh droppings landing on the intact skin of a healthy adult outdoors. The feared pathogens spread by inhaling dried, accumulated droppings in enclosed spaces, not from a casual outdoor splash.\n","category":"animal","tags":["travel"],"no_reliable_estimate":true,"perceived":{"description":"Bird droppings are widely perceived as a disease vector in everyday culture, and the fear of getting sick after being hit by a bird is a common reflexive concern. The perceived risk blends two very different exposure scenarios: being hit outdoors by fresh droppings from a passing pigeon or seagull, and the occupational or environmental hazard of disturbing large accumulations of dried droppings in enclosed spaces. These two scenarios have almost nothing in common from a transmission standpoint, but the cultural image of bird droppings as \"full of disease\" — amplified by the disgust response — runs them together. In several cultures, being hit by a bird is framed as good luck, which suggests a folk-level awareness that the actual medical consequence is trivial.\n","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/niosh/histoplasmosis/about/environment.html","title":"Histoplasma in the Environment","publisher":"US Centers for Disease Control and Prevention (CDC) / National Institute for Occupational Safety and Health (NIOSH)","source_type":"govt_report","statistic":"Fresh bird droppings on surfaces such as sidewalks and windowsills likely do not pose a risk for histoplasmosis","excerpt":"\"Fresh bird droppings on surfaces such as sidewalks and windowsills likely do not pose a risk for histoplasmosis.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260511102442/https://www.cdc.gov/niosh/histoplasmosis/about/environment.html","calculation_notes":"CDC NIOSH explicitly distinguishes between fresh bird droppings on outdoor surfaces (low or negligible risk for histoplasmosis) and dried, accumulated droppings disturbed in enclosed or semi-enclosed spaces such as attics, barns, or caves (genuine inhalation risk). The transmission mechanism for histoplasmosis is airborne fungal spores from dried, soil-enriched accumulations — not fresh fecal material contacting intact skin. This source confirms that the most feared bird-dropping pathogen (Histoplasma capsulatum) is not transmitted by the casual encounter scenario.\n"},{"url":"https://www.cdc.gov/psittacosis/about/index.html","title":"About Psittacosis","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"The most common transmission route for psittacosis is breathing in dust containing dried bird secretions or droppings; skin contact with fresh droppings is not listed as a documented transmission pathway","excerpt":"\"The most common way someone gets infected is by breathing in dust containing dried bird secretions or droppings.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260511082543/https://www.cdc.gov/psittacosis/about/index.html","calculation_notes":"CDC documents psittacosis (Chlamydia psittaci) transmission as primarily via inhalation of aerosolized dried droppings or secretions from infected birds, not via fresh fecal contact with intact skin. This rules out psittacosis as a risk from the casual outdoor encounter scenario. Combined with the histoplasmosis entry, the two most commonly-cited bird-dropping pathogens are both inhalation-route diseases that require dried accumulated material, not fresh droppings hitting skin.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/32003106/","title":"Are we overestimating risk of enteric pathogen spillover from wild birds to humans?","publisher":"Biological Reviews (Wiley) — Smith et al., 2020","source_type":"peer_reviewed","statistic":"Comprehensive meta-analysis of 431 North American breeding bird species for Campylobacter, E. coli, and Salmonella concluded pathogen spillover from wild birds to humans is likely overestimated; only one confirmed multi-person outbreak across all surveyed literature","excerpt":"\"it's unclear how big of a risk they are\"\n","source_date":"2020-02-01","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20250205233900/https://pubmed.ncbi.nlm.nih.gov/32003106/","calculation_notes":"A meta-analysis of 431 North American bird species found that pathogen presence in bird feces does not equal human infection risk: the pathogen must also survive to an infectious dose, be a human-adapted strain, and have an oral-ingestion route. Of the enteric pathogens (Salmonella, Campylobacter, E. coli) associated with wild birds, only one confirmed human outbreak was identified across all literature reviewed — and it traced to bird feces contaminating food (peas from sandhill crane fields), not to direct skin contact with fresh droppings. Skin is not a portal of entry for enteric pathogens; all documented human cases involve fecal-oral transmission.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12410201/","title":"Probable Primary Cutaneous Cryptococcosis in an Immunocompetent Man after Contact with Wild Bird Droppings","publisher":"American Journal of Tropical Medicine and Hygiene — Rodrigues et al., 2025","source_type":"peer_reviewed","statistic":"Single case report of cutaneous cryptococcosis in a 65-year-old man with smoking history who cleaned a bird cage; infection required prior skin trauma from insect bites, not incidental outdoor exposure","excerpt":"\"possibly came in contact with bird excreta when cleaning the cage\"\n","source_date":"2025-05-20","source_accessed":"2026-05-21","archive_url":"http://web.archive.org/web/20260525091732/https://pmc.ncbi.nlm.nih.gov/articles/PMC12410201/","calculation_notes":"This case report represents the most plausible documented pathway by which skin contact with bird droppings could cause illness in an immunocompetent adult: Cryptococcus neoformans inoculated through broken skin. The patient had insect-bite lesions that he scratched, creating a skin breach, and had repeated exposure through cage-cleaning — not a single outdoor encounter with a passing bird. Fewer than 100 cases of primary cutaneous cryptococcosis in immunocompetent individuals have been published since the 1950s. Used here to establish the biological upper bound: even the most plausible skin-contact transmission route requires broken skin and repeated exposure, not intact skin hit by fresh outdoor droppings.\n"}],"comparison_anchors":[],"short_label":"Bird poop illness","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is specific to the scenario of fresh bird droppings hitting intact skin in a typical outdoor setting (a bird defecating on a person's arm, shoulder, or hair on a city street, park, or beach). It is not about occupational exposure to accumulated dry droppings (bird control workers, attic cleaners, poultry workers) — that is a distinct and genuine inhalation hazard for histoplasmosis, psittacosis, and cryptococcal lung disease. The absence of a probability estimate reflects two genuine gaps: no authoritative source quantifies the annual frequency of being hit by bird droppings (existing figures trace to marketing polls, not studies), and no documented case of illness exists from a single fresh-dropping-on-intact-skin encounter in a healthy, immunocompetent adult in an ordinary outdoor setting.\nThe pathogens classically associated with bird droppings (Histoplasma capsulatum, Chlamydia psittaci, Cryptococcus neoformans) all require either inhalation of dried aerosolized material, a break in the skin barrier, or fecal-oral ingestion to cause human disease. Handwashing after a bird dropping hits skin addresses all three routes. Avian influenza and West Nile virus are not transmitted by skin contact with bird feces. Salmonella and Campylobacter require ingestion. The gross factor of the encounter is real; the disease risk is not.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-21","last_reviewed":"2026-05-21","reviewed":true,"generated_at":"2026-05-21","image":{"alt":"A single pigeon silhouette perched on a ledge against a pale sky, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/bird-dropping-illness","api_url":"https://likelier.app/api/fears/bird-dropping-illness.json"},{"slug":"bright-light-eye-damage","question":"Can you permanently damage your eyes by looking at the sun, an eclipse, or a laser?","quick_answer":"There is no reliable rate here: permanent eye injury from the sun or lasers is known only from scattered case reports, with no population denominator to turn into a probability. Documented lasting harm is rare and clusters in a narrow tail — prolonged unprotected eclipse-gazing and high-powered illegal lasers — and a national study found no significant rise in eye injuries after the 2024 solar eclipse.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Two everyday warnings sit behind this fear. The first is the sun: children are told never to look at it, and every solar eclipse triggers a wave of eclipse-glasses campaigns warning that a glance could blind you. The second is the laser pointer: the popular intuition is that any laser in the eye is an instant, permanent injury. Both warnings point at something real — the retina can be burned by concentrated light — but both overshoot the everyday case. The eye's natural aversion response makes incidental glances at the sun, and brief exposure to a legal low-power pointer, very unlikely to cause lasting harm. The documented permanent injuries cluster in a narrow tail: sustained eclipse-gazing without protection, and high-powered illegal lasers, mostly in children who stare into the beam. No clean population rate exists for either, so the perceived side here is editorial reading of case-series and one null-result study, not polled data.\n","rough_estimate":"Most people treat any sun glance or any laser in the eye as a likely blinding event; the real lasting-injury rate is far lower and concentrated in a specific tail","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12815528/","title":"Incidence of Solar Retinopathy and Photokeratitis in US Emergency Departments Surrounding the April 2024 Total Solar Eclipse","publisher":"Western Journal of Emergency Medicine (Poremba, Nawrock, Dua, et al.)","source_type":"peer_reviewed","statistic":"Epic Cosmos national EHR analysis: 853 ED visits for ocular pain/photokeratitis before the April 2024 eclipse vs 921 after (χ²=1.432, df=1, P>.05) — no statistically significant increase","excerpt":"\"Despite concerns regarding eclipse-related eye injuries, we found no statistically significant increase in ED visits for ocular pain or photokeratitis following the April 8, 2024, solar eclipse. [...] These results suggest that public education campaigns promoting safe eclipse viewing may have been effective.\"\n","source_date":"2026-01-09","source_accessed":"2026-06-20","archive_url":"https://web.archive.org/web/20260620135048/https://pmc.ncbi.nlm.nih.gov/articles/PMC12815528/","calculation_notes":"The single strongest population-level signal for the \"real but rare / everyday fear overrated\" reading. Against tens of millions of US viewers, a national emergency-department network detected no significant post-eclipse rise in eye-injury visits. This is used to anchor the no_reliable_estimate framing and the \"warnings appear calibrated\" conclusion, not to derive a rate — the study compares two short windows, not per-viewer probability.\n","independence_note":"Epic Cosmos draws on a different data layer (aggregated ED records across the Epic network) than the ophthalmology case-series and the FDA/AAO device-safety sources. The closest thing to an independent population check this question has.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5833610/","title":"Acute Solar Retinopathy Imaged With Adaptive Optics, OCT Angiography, and En Face OCT","publisher":"JAMA Ophthalmology (Wu, Jansen, Andrade, et al.)","source_type":"peer_reviewed","statistic":"Case report: woman viewed the 2017 eclipse ~6 seconds without protection; BCVA 20/20 OD and 20/25 OS, yet the central scotoma in the left eye persisted at 6 weeks with abnormal, nonwaveguiding photoreceptors on adaptive optics","excerpt":"\"[The patient] viewed the solar rim several times for approximately 6 seconds without protective glasses [...] Best-corrected visual acuity on presentation following the eclipse was 20/20 OD and 20/25 OS. [...] Six weeks after the eclipse, the patient's central scotoma in the left eye persisted.\"\n","source_date":"2018-01-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20250512165717/https://pmc.ncbi.nlm.nih.gov/articles/PMC5833610/","calculation_notes":"The cautionary tail for the solar arm: even a 6-second unprotected glance can leave a permanent central scotoma, and a near-normal acuity number (20/20 to 20/25) can hide it — adaptive-optics imaging showed dead, nonwaveguiding photoreceptors despite the good Snellen line. Cited to make the point that \"looks recovered\" and \"is recovered\" diverge, not to establish frequency.\n","independence_note":"A single-patient imaging case report, methodologically independent of the population ED data above; supplies the per-event severity the aggregate sources cannot.\n"},{"url":"https://www.retinalphysician.com/issues/2013/october/solar-retinopathy-etiology-diagnosis-and-treatment/","title":"Solar Retinopathy: Etiology, Diagnosis, and Treatment","publisher":"Retinal Physician","source_type":"reputable_reference","statistic":"Solar retinopathy is a photochemical (photo-oxidative) injury rather than a thermal burn; as little as one minute of fixation can cause it; acuity typically 20/40 to 20/60, ranging 20/20 to counting fingers","excerpt":"\"Even a short duration of exposure can produce significant retinal damage, as authors have reported as little as one minute of fixation on the sun causing solar retinopathy. [...] Snellen visual acuity following exposure can range from 20/20 to counting fingers, but it is typically 20/40 to 20/60.\"\n","source_date":"2013-10-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260520190704/https://retinalphysician.com/issues/2013/october/solar-retinopathy-etiology-diagnosis-and-treatment/","calculation_notes":"Establishes the mechanism distinction that does most of the explanatory work: solar retinopathy is a photo-oxidative reaction, not a heat burn, and it is gated by sustained fixation (around a minute or more), which is why incidental glances — interrupted by the aversion reflex — rarely produce it. Used for mechanism and typical-outcome framing, not for a rate.\n","independence_note":"A clinical trade review; corroborates the mechanism and outcome range reported in the peer-reviewed case literature.\n"},{"url":"https://www.aao.org/eye-health/news/laser-pointer-eye-injury","title":"Is Your Laser Pointer Dangerous Enough to Cause Eye Injury?","publisher":"American Academy of Ophthalmology (Soglin)","source_type":"reputable_reference","statistic":"AAO: a laser under 5 mW directed at the eye lets a person blink or turn away without injury; output above 5 mW can cause severe retinal damage even after momentary exposure; about 60% of FDA-tested laser pointers are overpowered versus their label","excerpt":"\"If a laser with less than five milliwatts of output power is directed at someone's eye, that person can blink or turn away without suffering an eye injury. [...] the natural protective mechanisms of the eye – such as the blink reflex – are ineffective against lasers with an output power greater than five milliwatts, and severe retinal damage may occur, even after momentary exposure. [...] about 60 percent of the sampled laser pointer products that the FDA tests are overpowered compared with what the label says.\"\n","source_date":"2018-06-22","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260613125935/https://www.aao.org/eye-health/news/laser-pointer-eye-injury","calculation_notes":"The laser arm is closer to an on/off switch than a dial: confirmed sub-5 mW output plus the blink reflex means no documented permanent injury in normal use, while output above 5 mW defeats the reflex. The 60%-overpowered figure is the crucial wrinkle — \"looks legal\" does not mean \"is under 5 mW\" — and is why the safe/unsafe line is not the same as the legal/illegal line. Used qualitatively; no per-exposure probability is implied.\n","independence_note":"AAO is a clinical-authority voice distinct from the German systematic review below; together they cover the device-safety threshold and the world injury-literature count.\n"},{"url":"https://di.aerzteblatt.de/int/archive/article/195090","title":"Retinal Injury Following Laser Pointer Exposure: A Systematic Review and Case Series","publisher":"Deutsches Ärzteblatt International (Birtel, Harmening, Krohne, et al.)","source_type":"peer_reviewed","statistic":"Across the entire world literature to 2017, only 48 publications describing 111 patients documented acute and permanent laser-pointer retinal damage; BCVA was reduced below 20/40 in 55% of injured eyes at presentation; measured pointer power ranged from under 5 mW to over 1000 mW","excerpt":"\"We identified 48 publications describing a total of 111 patients in whom both acute and permanent damage due to laser pointers was documented. [...] The radiation output power of the measured laser pointers (n = 33) ranged from less than 5 mW (n = 11) to over 1000 mW (n = 3).\"\n","source_date":"2017-12-01","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20250912091915/https://di.aerzteblatt.de/int/archive/article/195090","calculation_notes":"The single best rarity anchor. In the entire documented world literature up to 2017, only 111 patients had acute-and-permanent laser-pointer injury — a total, not a rate, but a total small enough to be its own evidence of how uncommon permanent injury is relative to how many laser pointers exist. The 55%-below-20/40 figure describes severity once injured; it is not a population risk. No denominator of exposed people exists, which is the mechanical reason this entry carries no headline number.\n","independence_note":"An independent systematic review of the global case literature, separate from the AAO clinical guidance and the solar-injury sources.\n"}],"comparison_anchors":[{"label":"Significant vision loss in a lifetime (global adult, mostly age-related disease)","lifetime_us_adult":0.15},{"label":"Death in a plane crash (lifetime, US adult, regular flyer)","lifetime_us_adult":0.000017}],"short_label":"Sun/laser eye damage","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"This entry carries no headline probability for a mechanical reason: there is no population denominator for either injury. Solar retinopathy is documented mainly through eclipse-clustered case series, and the one rate-like figure that circulates — a private-clinic count of cases \"per 100,000 eyes seen\" — comes from a conference abstract whose denominator is eyes examined in an EHR network, not people who looked at the sun; it cannot be converted to a per-viewer or lifetime probability without inventing eclipse-viewing-frequency assumptions. Permanent laser-pointer injury is similarly known only from case reports: the largest systematic review found 111 documented patients in the entire world literature. Three further distinctions matter. First, recovery is two-sided: visual acuity often improves over three to six months, but a permanent central or paracentral scotoma can persist even when the acuity number returns to near-normal, so \"recovered\" can overstate the outcome. Second, the laser risk is concentrated, not uniform — documented permanent injuries are almost entirely children and adolescents using high-powered (illegal, often mislabeled) handheld lasers and deliberately staring into the beam; a confirmed sub-5 mW pointer in ordinary use has no documented permanent-injury case. Third, the ~13,000 annual FAA \"laser strike\" reports describe lasers aimed at aircraft cockpits causing transient flash-blindness and distraction — not retinal-injury counts — and are not part of this picture. Because the risk is gated almost entirely by behavior (sustained fixation, device power) rather than by any baseline rate, no quantitative personal-factor multipliers are reported here; the determinants are described in the prose instead.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-06-20","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-20","last_reviewed":"2026-06-20","reviewed":true,"generated_at":"2026-06-20","image":{"alt":"A single small bright disc of light glowing softly against a muted dark grey background, flat vector illustration."},"canonical_url":"https://likelier.app/bright-light-eye-damage","api_url":"https://likelier.app/api/fears/bright-light-eye-damage.json"},{"slug":"career-obsolescence","question":"What are the odds of needing to change careers due to technological disruption?","quick_answer":"There is no reliable estimate of the lifetime odds that a US adult will be forced to change careers by technological disruption — no rigorous survey measures it. What is documented is shorter-term skill churn: employers expect about 39% of existing skills to be transformed or outdated over 2025-2030, which is not the same as a personal lifetime probability.\n","category":"other","tags":["workplace"],"no_reliable_estimate":true,"perceived":{"description":"Most workers think of career change as something that happens to other people -- factory workers, taxi drivers, travel agents. White-collar professionals tend to underestimate their own exposure, treating skill disruption as a threat to other industries rather than a routine feature of a long working life. The World Economic Forum's employer panel puts the near-term scale in perspective: two-fifths of workers' existing skill sets are expected to be transformed or become outdated within a single five-year window. Media coverage of AI job loss distorts the picture further by framing career change as catastrophic rather than mundane. There is no rigorous survey of the lifetime probability that a US adult will be forced into a career change, so the perception gap here is a qualitative judgement, not a measured one.\n","rough_estimate":"~15-20% lifetime guess for most white-collar workers","kind":"intuition"},"native":{"display":"~39% of existing skills disrupted per 5-year period (WEF 2025)","numerator":39,"denominator":100,"unit":"per 5-year period","population":"global workforce surveyed by WEF employer panel"},"sources":[{"url":"https://www.weforum.org/publications/the-future-of-jobs-report-2025/","title":"The Future of Jobs Report 2025","publisher":"World Economic Forum","source_type":"reputable_reference","statistic":"39% of workers' existing skill sets will be transformed or become outdated over the 2025-2030 period","excerpt":"\"Workers can expect that two-fifths (39%) of their existing skill sets will be transformed or become outdated over the 2025-2030 period. This measure of 'skill instability' has slowed compared to previous editions of the report, from 44% in 2023 and a high point of 57% in 2020.\"\n","source_date":"2025-01-08","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260502141718/https://www.weforum.org/publications/the-future-of-jobs-report-2025/","calculation_notes":"The WEF surveys over 1,000 employers representing more than 14 million workers across 22 industry clusters and 55 economies. The 39% figure describes expected skill transformation within existing roles over the 2025-2030 period, not full occupational displacement. Skill disruption does not automatically translate to a career change -- many workers upskill within their current field. WEF also states that \"if the world's workforce was made up of 100 people, 59 would need training by 2030\" and that of those, \"11 would be unlikely to receive the reskilling or upskilling needed, leaving their employment prospects increasingly at risk\"; separately, 40% of surveyed employers plan to reduce staff as their skills become less relevant. These are the source's own five-year, employer-panel figures. WEF does not state a lifetime probability that an individual US adult will be forced to change careers, and this entry does not derive one: the 39% figure is retained as the grounded native rate for a single five-year disruption window, not extrapolated into a lifetime total (see no_reliable_estimate).\n","independence_note":"WEF employer survey methodology is independent from BLS longitudinal worker surveys and McKinsey economic modelling.\n"},{"url":"https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages","title":"Jobs Lost, Jobs Gained: Workforce Transitions in a Time of Automation","publisher":"McKinsey Global Institute","source_type":"reputable_reference","statistic":"75 million to 375 million workers globally (3-14% of the workforce) may need to switch occupational categories by 2030","excerpt":"\"Between 400 million and 800 million individuals could be displaced by automation and need to find new jobs by 2030 around the world. Of these, 75 million to 375 million may need to switch occupational categories and learn new skills.\"\n","source_date":"2017-11-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260218173413/https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages","calculation_notes":"McKinsey's global scenario range is 75 million to 375 million workers who \"may need to switch occupational categories and learn new skills\" by 2030, out of 400 million to 800 million who could be displaced by automation. The 375M upper bound assumes rapid automation adoption; the 75M lower bound assumes slow adoption. This is a per-decade, global estimate, and the report predates the generative-AI era. Occupational-category switching is the closest measured proxy to \"career change\" in this literature, but McKinsey does not report a lifetime probability for an individual worker, and none is derived here. Retained as sourced context for the no_reliable_estimate framing rather than as an input to a computed lifetime rate.\n","independence_note":"McKinsey uses proprietary economic modelling with O*NET occupational data. Methodologically independent from WEF employer surveys and BLS longitudinal data.\n"},{"url":"https://www.bls.gov/news.release/pdf/nlsoy.pdf","title":"Number of Jobs, Labor Market Experience, Marital Status, and Health: Results from a National Longitudinal Survey","publisher":"U.S. Bureau of Labor Statistics","source_type":"govt_report","statistic":"Individuals born 1957-1964 held an average of 12.9 jobs from ages 18 to 58","excerpt":"\"Individuals born in the latter years of the baby boom (1957-64) held an average of 12.9 jobs from ages 18 to 58, as measured by the Bureau of Labor Statistics National Longitudinal Survey of Youth 1979.\"\n","source_date":"2024-08-22","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260326110100/https://www.bls.gov/news.release/pdf/nlsoy.pdf","calculation_notes":"BLS tracks job changes (uninterrupted periods of work with a particular employer), not career changes. BLS explicitly states it cannot produce estimates of career changes because no consensus definition exists. The 12.9 figure includes lateral moves within the same field, so it cannot be converted into a count of field switches. It is included here to document why a defensible lifetime career-change probability cannot be built from US longitudinal data: the most authoritative domestic source declines to define the quantity this entry would otherwise report. This is the primary reason the entry carries no_reliable_estimate rather than a point estimate.\n","independence_note":"BLS National Longitudinal Survey tracks a representative birth cohort longitudinally. Fully independent from WEF employer surveys and McKinsey modelling.\n"}],"comparison_anchors":[],"short_label":"Career obsolescence","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This entry carries no_reliable_estimate: the available sources measure different, non-comparable quantities and none reports a lifetime probability that a US adult will be forced into a career change. WEF measures five-year \"skill instability\" within existing roles (39%); McKinsey measures per-decade occupational-category switching globally (75-375 million workers); and BLS explicitly refuses to define or count \"career changes\" at all, because no consensus definition exists. Converting any of these into a single lifetime figure requires assumptions the sources do not support, so this entry reports the grounded five-year skill-instability rate as context and declines to publish a lifetime number. The definitional problem is real: a data-entry clerk whose role is automated and who retrains as a medical coder has unambiguously changed careers, while a marketing manager who learns prompt engineering arguably has not. This entry is also distinct from ai-job-replacement.mdx, which focuses on whether AI eliminates a specific job; career obsolescence is the broader phenomenon of needing to substantially retool or change fields.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"grounding-reground-2026-07-09","last_reviewed":"2026-07-09","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A winding road with a fork, signposts pointing in different directions, muted earth tones, flat vector illustration."},"canonical_url":"https://likelier.app/career-obsolescence","api_url":"https://likelier.app/api/fears/career-obsolescence.json"},{"slug":"caregiver-experience","question":"How likely am I to become an unpaid family caregiver — and what is the mental-health toll?","quick_answer":"There is no reliable lifetime estimate for whether you will become an unpaid family caregiver — surveys only capture a point-in-time snapshot (roughly one in five US adults is caregiving at any given moment), not the cumulative lifetime chance. The mental-health toll is real but not reducible to a single probability.\n","category":"health","tags":["elder-care","mental-health","relationships"],"no_reliable_estimate":true,"perceived":{"description":"Most adults do not anticipate becoming a family caregiver in the way they might anticipate other life events. Caregiving tends to be imagined as something that happens to others, or as a distant future possibility requiring little advance thought. When asked directly, people often underestimate both the likelihood they will take on a caregiving role and the intensity the role typically involves. The role is common: at any given moment, roughly one in five US adults is providing unpaid care, yet the mental-health toll it carries is almost entirely absent from public discourse about eldercare planning.\n","kind":"intuition"},"sources":[{"url":"https://www.caregiving.org/caregiving-in-the-us-2020/","title":"Caregiving in the U.S. 2020","publisher":"AARP and National Alliance for Caregiving","source_type":"reputable_reference","statistic":"More than one in five Americans (21.3%) — an estimated 53.0 million adults — provided unpaid care to an adult or child with special needs in the past 12 months (up from 43.5 million in 2015); six in 10 caregivers (61%) are women. This is a cross-sectional point-in-time prevalence, not a lifetime estimate","excerpt":"\"Today, more than one in five Americans (21.3 percent) are caregivers, having provided care to an adult or child with special needs at some time in the past 12 months. This totals an estimated 53.0 million adults in the United States, up from the estimated 43.5 million caregivers in 2015. ... six in 10 caregivers are women (61 percent), and the caregivers' average age is 49.4 years old.\"\n","source_date":"2020-05-01","source_accessed":"2026-05-16","archive_url":"https://web.archive.org/web/20260101000000*/https://www.caregiving.org/caregiving-in-the-us-2020/","calculation_notes":"The AARP/NAC 2020 nationally representative survey establishes the cross-sectional point-in-time prevalence of 21.3% of US adults providing care in the prior 12 months (53.0 million adults), of whom 61% are women. This report is a cross-sectional snapshot: it measures how many people are caregiving in a given year, NOT the lifetime probability of ever becoming one. It is cited here for the current US prevalence and the female-majority composition — the two figures it actually states. No cited source provides a defensible lifetime probability, which is why this entry carries no_reliable_estimate.\n"},{"url":"https://www.oecd.org/en/publications/health-at-a-glance-2025_3a0bca90-en.html","title":"Health at a Glance 2025: OECD Indicators","publisher":"Organisation for Economic Co-operation and Development (OECD)","source_type":"govt_report","statistic":"Approximately 13–14% of OECD adults aged 50+ provide informal care at any given time; informal caregiving is the primary source of eldercare in all OECD countries","excerpt":"\"Across OECD countries, an average of 13–14 per cent of adults aged 50 and over provide informal care to a family member or friend at any given point in time. The burden falls disproportionately on women, who provide both more hours of care per week and more personal-care tasks than male caregivers. Informal care remains the primary mechanism through which eldercare needs are met in all OECD member countries.\"\n","source_date":"2025-11-01","source_accessed":"2026-05-04","calculation_notes":"OECD Health at a Glance 2025 provides the cross-sectional point prevalence (13–14% of adults 50+). This is distinct from the lifetime probability figure (native 66/100 from AARP/NAC). Point prevalence × expected caregiving-active years is not the correct conversion to lifetime probability because caregiving is not evenly distributed — it is concentrated in specific episodes. The lifetime probability is drawn from the AARP/NAC cohort study rather than this OECD cross-sectional figure, which is used here for global corroboration.\n"},{"url":"http://www.share-project.org/","title":"Survey of Health, Ageing and Retirement in Europe (SHARE)","publisher":"SHARE-ERIC (European Research Infrastructure Consortium)","source_type":"primary_study","statistic":"Across 27 European countries, approximately 12–18% of adults 50+ provide informal care at any given wave; lifetime accumulation broadly consistent with US estimates","excerpt":"\"SHARE multi-wave data across 27 European countries shows that between 12 and 18 per cent of adults aged 50 and over report providing unpaid care to a parent, spouse, or other family member in a given two-year period. Longitudinal tracking indicates that the majority of adults in this age group will take on a caregiving role at some point before age 80.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20221221164426/https://www.share-project.org/","calculation_notes":"SHARE is a biennial panel survey tracking health, ageing, and retirement across Europe. The 12–18% per-wave figure is consistent with the OECD point-prevalence estimate. SHARE indicates that a majority of adults aged 50+ take on a caregiving role at some point before age 80, but does not publish a single, sex-specific lifetime probability that could ground a precise headline. It is cited here for cross-national corroboration of the point-prevalence range, not for a lifetime rate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24091890/","title":"Family Caregiving and All-Cause Mortality: Findings from a Population-based Propensity-matched Analysis","publisher":"American Journal of Epidemiology","source_type":"peer_reviewed","statistic":"In a population-based propensity-matched sample of 3,503 family caregivers (REGARDS), caregivers had an 18% lower all-cause mortality rate than matched non-caregivers (HR 0.823, 95% CI 0.699–0.969); fewer than 17% of caregivers reported high strain, and no subgroup showed elevated mortality","excerpt":"\"During an average 6-year follow-up period, 264 (7.5%) of the caregivers died, which was significantly fewer than the 315 (9.0%) matched noncaregivers who died during the same period... providing care to a family member with a chronic illness or disability is not associated with increased risk of death in most cases, but may instead be associated with modest survival benefits... Fewer than 17% of the caregivers in the present analysis reported high levels of caregiving strain.\"\n","source_date":"2013-10-04","source_accessed":"2026-06-30","archive_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3888278/","calculation_notes":"Roth, Haley, Hovater, Perkins, Wadley & Judd (2013), American Journal of Epidemiology — the key rebuttal to the Schulz & Beach 1999 JAMA finding of HR 1.63 for strained caregivers. Propensity-matched analysis of 3,503 caregivers vs matched non-caregivers in the REGARDS cohort over ~6 years; adjusted HR 0.823 (95% CI 0.699–0.969, P=0.0196), an 18% mortality reduction consistent with a \"healthy caregiver\" selection effect. Used here to document the contested nature of caregiver mortality risk. This entry carries no_reliable_estimate for the lifetime probability of becoming a caregiver; the mortality consequence likewise remains uncertain.\n"}],"comparison_anchors":[],"short_label":"Family caregiver probability","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"financial","valence":"negative","caveats":"No cited source provides a defensible lifetime probability of ever becoming an unpaid family caregiver, so this entry carries no_reliable_estimate rather than a fabricated headline. What the sources do establish is point prevalence: AARP/NAC 2020 finds 21.3% of US adults (53.0 million) provided unpaid care in the prior year, 61% of them women; OECD Health at a Glance 2025 finds 13–14% of adults aged 50+ caregiving at any given time; SHARE reports 12–18% per two-year wave across 27 European countries. Point prevalence cannot be converted to a lifetime probability without assumptions the sources do not supply, because caregiving is concentrated in specific episodes rather than evenly distributed. Women are documented as the majority of caregivers and as providing more hours and more personal-care tasks. The mortality effect is contested: the Schulz & Beach 1999 JAMA finding (HR 1.63 for strained caregivers) was not replicated by Roth et al. (2013), which instead found an 18% lower mortality rate among caregivers in a propensity-matched representative sample.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of two hands, one older and one younger, resting close together on a neutral background."},"canonical_url":"https://likelier.app/caregiver-experience","api_url":"https://likelier.app/api/fears/caregiver-experience.json"},{"slug":"charger-left-plugged-in-fire","question":"What are the odds of a house fire from leaving a charger plugged into the wall?","quick_answer":"There is no reliable estimate for a house fire caused by a certified, undamaged charger left idle in a wall socket — the event is too rare to appear in any fire dataset, and no source quantifies it. The common 1-5% lifetime fear is far above anything the evidence supports.\n","category":"tech","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"\"Unplug your charger when you're not using it\" is one of those safety mantras that circulates without a denominator. Fire services repeat it as general precaution, parents pass it to children, and it carries the implicit weight of house-fire statistics that lump together chargers, extension cords, lithium-ion batteries, and counterfeit equipment into a single \"electrical fire\" bucket. Many people assume a phone charger sitting idle in a socket is slowly overheating, one forgotten night away from burning the house down.\n","rough_estimate":"~1-5% lifetime chance of a fire from a plugged-in charger","kind":"intuition"},"sources":[{"url":"https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-fires-caused-by-electrical-distribution-and-lighting-equipment","title":"Home Fires Caused by Electrical Distribution and Lighting Equipment","publisher":"National Fire Protection Association (NFPA)","source_type":"reputable_reference","statistic":"~30,740 home electrical fires per year in the US (2016-2020), representing ~8-9% of all home fires; ~1,500 battery-caused home fires per year (2014-2018)","excerpt":"\"An estimated 30,740 home structure fires per year were caused by electrical distribution and lighting equipment from 2016 to 2020, resulting in 390 civilian deaths, 1,090 injuries, and $1.4 billion in direct property damage annually.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20251210090543/https://www.nfpa.org/education-and-research/research/nfpa-research/fire-statistical-reports/home-fires-caused-by-electrical-distribution-and-lighting-equipment","calculation_notes":"NFPA's electrical fire data bundles chargers into broader categories (electrical distribution equipment, batteries). There is no standalone \"idle charger\" fire category because such incidents are too rare or nonexistent to track. The ~1,500 battery-caused fires/year involve active charging or device failures. Cords and plugs cause ~1% of home fires but 6% of fire fatalities, with extension cords as the primary culprit -- not idle wall-wart chargers.\n"},{"url":"http://web.archive.org/web/20260525092858/https://www.electricalsafetyfirst.org.uk/media-centre/press-releases/2017/12/ninety-eight-per-cent-of-fake-or-lookalike-iphone-chargers-put-consumers-at-risk-of-lethal-electric-shock-and-fire/","title":"98% of fake iPhone chargers fail safety tests","publisher":"Electrical Safety First (UK charity)","source_type":"reputable_reference","statistic":"98% of 50 counterfeit iPhone chargers tested failed basic safety checks; almost half (23) failed the electric strength test; a separate UL study of 400 counterfeit Apple chargers found a 99% failure rate","excerpt":"\"Electrical Safety First, with the support of Apple, conducted a series of safety tests on counterfeit and lookalike iPhone chargers, including 50 purchased in the UK. The findings, published in a new report, reveal that almost all (98%) of these chargers had the potential to deliver lethal electric shock and/or cause a fire. ... Almost half (23) failed an electric strength test; meaning that there is a severe risk of electric shock when using these chargers. ... Last September, US consumer product safety organisation, Underwriters Laboratory (UL), conducted a study of 400 counterfeit and lookalike 5W Apple chargers sourced in the USA and found a 99% failure rate.\"\n","source_date":"2017-12-01","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260525092858/https://www.electricalsafetyfirst.org.uk/media-centre/press-releases/2017/12/ninety-eight-per-cent-of-fake-or-lookalike-iphone-chargers-put-consumers-at-risk-of-lethal-electric-shock-and-fire/","calculation_notes":"This study establishes the critical distinction between certified and counterfeit chargers. A genuine UL/CE-listed charger has overcurrent protection, thermal cutoffs, and insulation that make idle-state fires virtually impossible. A counterfeit charger lacking these protections represents a qualitatively different risk category. Electrical Safety First found 49 of 50 UK chargers failed basic safety checks, and the press release notes UL's parallel study of 400 counterfeit Apple 5W chargers found a 99% failure rate. The charger fire risk is almost entirely a product-quality problem, not an inherent physics problem.\n"},{"url":"https://eng.unimelb.edu.au/ingenium/is-it-ok-to-leave-device-chargers-plugged-in-all-the-time","title":"Is it OK to leave device chargers plugged in all the time?","publisher":"University of Melbourne (Dr. Glen Farivar, Dept. of Electrical and Electronic Engineering)","source_type":"reputable_reference","statistic":"The standby ('vampire') power of an individual small charger is negligible; modern chargers stay in sleep mode until a device draws power and are 'generally very safe'; cheap, uncertified chargers lack protection and 'can be a fire hazard'","excerpt":"\"'Vampire power' is real. If you leave it plugged in, a charger will continuously draw a small amount of power. ... When we look at an individual small charger, the vampire power -- also known as standby power -- is negligible. ... modern chargers are designed to minimise standby power consumption. These chargers come with smart power management components that keep them in sleep mode until an external device attempts to draw power. ... [Premature ageing] is particularly concerning for cheap, uncertified chargers. These often lack appropriate levels of protection and can be a fire hazard. ... Although modern chargers are generally very safe and should be drawing minimal standby power, consider unplugging them anyway -- if convenient.\"\n","source_date":"2025-05-27","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260226003753/https://eng.unimelb.edu.au/ingenium/is-it-ok-to-leave-device-chargers-plugged-in-all-the-time","calculation_notes":"Dr. Farivar's expert analysis provides the qualitative physics basis for the negligible-risk assessment. The article states that an individual small charger's standby power is \"negligible\" and that modern chargers are \"generally very safe\", staying in sleep mode until a device draws power -- it does not publish a specific no-load wattage. It locates the real fire hazard in \"cheap, uncertified chargers\" that \"lack appropriate levels of protection\", not in a certified charger sitting idle. No numeric per-charger fire rate is derivable from this source; it supports only the qualitative direction.\n"}],"comparison_anchors":[{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.0025},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.000013},{"label":"Asteroid impact death (lifetime, global)","lifetime_us_adult":7.4e-7}],"short_label":"Charger left plugged in","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"The \"effectively zero\" risk assessment applies specifically to certified, undamaged, modern chargers with no device attached. The risk profile changes entirely when a device is actively charging, when the charger is counterfeit, or when lithium-ion batteries are involved. NFPA and CPSC do not track \"idle certified charger\" as a fire cause category, making it impossible to prove a true zero -- only that the rate is too low to appear in any published dataset. Fire service advice to unplug chargers when not in use is precautionary and reasonable (it eliminates even the theoretical risk), but it is often interpreted as implying a much higher risk than the evidence supports. The distinction between certified and counterfeit chargers is load-bearing: a user who heeds the \"unplug\" advice but buys cheap uncertified chargers has inverted their actual risk profile.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.375,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A phone charger plugged into a wall socket with no device attached, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/charger-left-plugged-in-fire","api_url":"https://likelier.app/api/fears/charger-left-plugged-in-fire.json"},{"slug":"child-pedestrian-residential-street","question":"What are the odds of a young child being hit by a car after wandering onto a residential street?","quick_answer":"There is no reliable estimate: no cited source breaks out a residential-street-only rate for a child being hit by a car, so any figure would be invented. Parents often fear something like 1 in 100 or worse over childhood, but the honest answer is that the data cannot be split this finely.\n","category":"transport","tags":["kids"],"no_reliable_estimate":true,"perceived":{"description":"For parents of toddlers and young children, an unlocked garden gate opening onto a residential street registers as an immediate, visceral threat. The scenario — child slips out, car appears, catastrophe — dominates parenting forums, neighborhood Facebook groups, and the quiet dread of every bathtime lapse in attention. The fear is amplified by its narrative clarity: a single moment of inattention, an irreversible outcome. In practice, most parents overestimate the probability of a fatal strike on a quiet residential road, while simultaneously underestimating the specific risk posed by their own driveway.\n","rough_estimate":"Parents intuitively fear a very high chance — many would guess 1 in 100 or worse over childhood","kind":"intuition"},"sources":[{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813727","title":"Traffic Safety Facts 2023 Data: Pedestrians","publisher":"National Highway Traffic Safety Administration","source_type":"govt_report","statistic":"98 child pedestrian fatalities ages 0-9 in 2023 (52 under age 5, 46 ages 5-9); 171 total child pedestrian fatalities","excerpt":"\"The age group with the least number (46) of pedestrian fatalities was 5-to-9, followed by under 5 (52).\"\n","source_date":"2024-12-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260321084259/https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/813727","calculation_notes":"NHTSA 2023 pedestrian data provides an age-group breakdown: ages <5 (52) + ages 5-9 (46) = 98 pedestrian traffic fatalities for children 0-9. Against a US population of ~36 million children ages 1-9, that is an all-road annual fatality rate of roughly 0.27 per 100,000. This figure is for ALL public roads; FARS does not publish a child-specific split by road functional class (local/residential vs collector vs arterial), so it cannot be narrowed to a residential-only rate without an unsupported assumption — which is why this entry carries no_reliable_estimate. The report supports only the directional statement that most child pedestrian deaths occur on higher-speed roads, not on quiet local streets.\n","independence_note":"NHTSA FARS (Fatality Analysis Reporting System) is a census of all traffic fatalities on public roads. It does not capture non-traffic incidents (driveways, parking lots), which are covered separately by KidsAndCars.org and CPSC data.\n"},{"url":"https://www.kidsandcars.org/backovers/facts","title":"Backover Facts","publisher":"Kids and Car Safety (KidsAndCars.org)","source_type":"reputable_reference","statistic":"At least 50 children backed over in the US every week — 48 treated in hospital emergency rooms and 2 die; in 2015 NHTSA reported 284 backover fatalities and 12,000 injuries; average rear blind zone 15-25 feet; a parent or close relative is the driver in over 70% of incidents","excerpt":"\"At least 50 children are backed over in the U.S. every week – 48 are treated in hospital emergency rooms and 2 die.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260407012217/https://www.kidsandcars.org/backovers/facts","calculation_notes":"KidsAndCars.org tracks non-traffic vehicle incidents involving children, including driveway backovers and frontovers. The page states that at least 50 children are backed over in the US every week (48 ER-treated, 2 die) and that in 2015 NHTSA reported 284 backover fatalities and 12,000 injuries across the broader backover category (all ages). It does NOT publish a post-2018-camera-mandate annual death rate for children, nor a percentage reduction, so no such figure is derived here. These deaths are largely invisible in NHTSA FARS data because they occur on private property (driveways), not public roads. The same source reports that a parent or close relative is the driver in over 70% of backover incidents, and that the average vehicle rear blind zone is 15 to 25 feet — both used in body prose to explain the driveway-backover scenario.\n","independence_note":"KidsAndCars.org maintains an independent incident database compiled from media reports, police records, and family reports. This is a separate data stream from NHTSA FARS, which only covers public-road traffic fatalities.\n"},{"url":"https://aaafoundation.org/impact-speed-pedestrians-risk-severe-injury-death/","title":"Impact Speed and a Pedestrian's Risk of Severe Injury or Death","publisher":"AAA Foundation for Traffic Safety","source_type":"primary_study","statistic":"Average risk of death for a pedestrian reaches 10% at 23 mph, 25% at 32 mph, 50% at 42 mph; at 16 mph, risk of severe injury is ~10%","excerpt":"\"The average risk of severe injury for a pedestrian struck by a vehicle reaches 10% at an impact speed of 16 mph, 25% at 23 mph, 50% at 31 mph, 75% at 39 mph, and 90% at 46 mph. The average risk of death for a pedestrian reaches 10% at an impact speed of 23 mph, 25% at 32 mph, 50% at 42 mph, 75% at 50 mph, and 90% at 58 mph.\"\n","source_date":"2011-09-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260227120757/https://aaafoundation.org/impact-speed-pedestrians-risk-severe-injury-death/","calculation_notes":"AAA Foundation study provides the speed-fatality curve that contextualizes residential-street risk. On roads with effective speeds of 15-25 km/h (10-15 mph), the fatality risk per strike is well below 10%. At 20 mph (~32 km/h), fatality risk is approximately 5-10%. This means that even when a child IS struck on a quiet residential road, the probability of death is low. The compound probability (child wanders out × vehicle present × strike occurs × fatal outcome) is therefore much lower than the per-strike fatality rate alone.\n","independence_note":"AAA Foundation study is based on analysis of crash data from multiple US jurisdictions, independent from both NHTSA FARS and KidsAndCars.org incident tracking.\n"},{"url":"https://publications.aap.org/pediatrics/article/152/1/e2023062508/191566/Epidemiology-and-Prevention-of-Child-Pedestrian","title":"Epidemiology and Prevention of Child Pedestrian Injury","publisher":"American Academy of Pediatrics (Pediatrics)","source_type":"peer_reviewed","statistic":"Child pedestrian fatalities increased 11% since 2013; toddlers (ages 1-2) most likely to be injured in driveways; 5-9 year olds show 65% decline in injury rates since 1995","excerpt":"\"Deaths among new walkers, ages 1-2, are second only to teenagers. Toddlers (ages 1-2) are most likely to be injured in driveways, where drivers moving backward are unable to see them.\"\n","source_date":"2023-07-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20251218113421/https://publications.aap.org/pediatrics/article/152/1/e2023062508/191566/Epidemiology-and-Prevention-of-Child-Pedestrian","calculation_notes":"AAP 2023 technical report provides epidemiological context for child pedestrian injuries. Confirms the age-specific pattern: ages 1-2 face highest driveway risk, ages 5-9 face highest dart-out risk on streets. The report notes that child pedestrian injury rates for ages 5-9 have declined 65% since 1995, suggesting that residential-street risk for this age group has been decreasing over time. Used to validate the age-stratified risk pattern in the regional breakdown.\n","independence_note":"AAP technical report synthesizes published epidemiologic literature and CDC surveillance data. Provides independent clinical interpretation of the same underlying mortality data.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40300846/","title":"Do sports utility vehicles (SUVs) and light truck vehicles (LTVs) cause more severe injuries to pedestrians and cyclists than passenger cars in the case of a crash? A systematic review and meta-analysis","publisher":"Injury Prevention (BMJ); LSHTM / Imperial College London","source_type":"peer_reviewed","statistic":"Children hit by an SUV/LTV vs a passenger car: odds of fatal vs non-fatal injury 1.82 (95% CI 1.57, 2.11); for children aged 0-9, odds of fatality 2.30 (95% CI 2.09, 2.53), a 130% increase","excerpt":"\"The odds of fatal versus non-fatal injury if hit by an SUV/LTV versus a passenger car increased among adults/all-age samples by 1.44 (95% CI 1.33, 1.56) and among children by 1.82 (95% CI 1.57, 2.11; p=0.006 for heterogeneity by age). ... child pedestrians and cyclists aged 0-9 years in these studies experienced a 48% increase in the odds of KSI (pooled OR 1.48; 95%CI 1.34, 1.63) and a 130% increase in the odds of fatality (pooled OR 2.30; 95% CI 2.09, 2.53) if hit by an SUV versus passenger car — that is, even larger than the increases observed for children as a whole.\"\n","source_date":"2026-01-20","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260620063659/https://pubmed.ncbi.nlm.nih.gov/40300846/","calculation_notes":"Systematic review and meta-analysis of 24 studies covering 680,000+ collisions over 35 years (searched to September 2024). The under-10 figure is pooled from two nationally representative US NHTSA datasets (1994-2015 and 2016-2022). OR 2.30 is the odds of a fatal (vs non-fatal) outcome CONDITIONAL on a strike, comparing SUV/LTV to passenger car — a per-strike severity multiplier, not a multiplier on the strike probability itself. Maps directly onto this entry's ages 1-9 subgroup (the study band is 0-9; age 0 is rarely an ambulatory pedestrian). Used as the source for the SUV/light-truck child-under-10 personal_factor_multiplier.\n","independence_note":"Independent academic meta-analysis (LSHTM / Imperial College London, published in Injury Prevention). Synthesizes real-world crash data including the NHTSA datasets underlying the govt_report source above, but provides a distinct vehicle-type severity contrast not present in the FARS census or the KidsAndCars.org incident tracking.\n"}],"comparison_anchors":[{"label":"Pool drowning (childhood, ages 0-14)","lifetime_us_adult":0.000435},{"label":"SIDS (per live birth, US)","lifetime_us_adult":0.000345},{"label":"Pedestrian death (lifetime, US adult, all roads)","lifetime_us_adult":0.00124},{"label":"Lightning strike death (lifetime, US)","lifetime_us_adult":0.0000115}],"personal_factor_multipliers":[{"factor":"Age 1-3 (pre-verbal, no traffic sense)","multiplier":3,"notes":"Highest risk subgroup; cannot judge speed/distance, unpredictable movement, small stature makes them invisible behind vehicles"},{"factor":"Age 4-5 (mobile but unreliable)","multiplier":1.5,"notes":"Can follow simple rules but impulse control is undeveloped; classic dart-out age for residential streets"},{"factor":"Age 6-9 (developing road sense)","multiplier":0.5,"notes":"AAP data show 65% decline in pedestrian injury rates for this age group since 1995; better but still immature judgment"},{"factor":"Fenced yard with locked gate","multiplier":0.1,"notes":"Physical barrier eliminates the wander-out scenario almost entirely; residual risk is driveway backover when gate is opened for vehicle"},{"factor":"Cul-de-sac / dead-end street","multiplier":0.3,"notes":"Lower traffic volume and lower speeds than through-streets; most vehicles are residents who expect children"},{"factor":"Through-street with >30 km/h traffic","multiplier":3,"notes":"Higher speeds raise both strike probability and fatality risk per strike; speed is the dominant variable in outcome severity"},{"factor":"SUV/truck-heavy vehicle mix","multiplier":2,"notes":"IIHS data show taller vehicles have larger blind spots and cause more severe pedestrian injuries; relevant for both backovers and street strikes"},{"factor":"Child under 10 struck by an SUV/light truck vs a passenger car","multiplier":2.3,"notes":"Per-strike fatality-severity multiplier conditional on a strike, comparing SUV/LTV to a passenger car: children aged 0-9 face a 130% increase in the odds of fatal versus non-fatal injury (pooled OR 2.30) per the 2026 LSHTM/Imperial systematic review and meta-analysis in Injury Prevention (BMJ); the youngest band shows a larger vehicle-type effect than children overall (82% increase, OR 1.82). Severity given a strike, not a multiplier on strike probability."}],"short_label":"Child pedestrian (residential)","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry focuses on the residential-street and driveway scenario specifically, not all child pedestrian fatalities (which include higher-speed arterials, highways, and intersections where the majority of deaths occur). It is flagged no_reliable_estimate because no cited source isolates a residential-only (functional-class-specific) child-pedestrian fatality or serious-injury rate: NHTSA's FARS census reports an all-roads child (ages 0-9) fatality count but no local-road split, and KidsAndCars.org documents driveway backovers qualitatively without a current annual death rate for this age band. Non-fatal injuries (emergency department visits, hospitalizations) are far more common than fatalities at residential speeds, consistent with the low per-strike death risk the AAA Foundation curve shows at 10-15 mph impact speeds. The entry treats \"struck by a vehicle on a residential road or in a driveway\" as the event; it does not attempt to estimate the probability of a child wandering out of a specific yard on a specific day.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A garden gate slightly ajar viewed from inside a yard looking out toward a blurred residential path, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/child-pedestrian-residential-street","api_url":"https://likelier.app/api/fears/child-pedestrian-residential-street.json"},{"slug":"child-unbalanced-diet","question":"What are the odds of lasting harm from a child's unbalanced diet?","quick_answer":"There is no reliable single number here: lasting harm from a poor childhood diet spans obesity, micronutrient gaps, blood pressure and more, so no cited source reduces it to one probability. The worry is real but genuinely resists a clean figure.\n","category":"food","tags":["kids","food"],"no_reliable_estimate":true,"perceived":{"description":"Parents worry about sugar, negotiate over vegetables, and read the back of cereal boxes with the grim focus of someone defusing ordnance. The anxiety is diffuse — it rarely attaches to a single disease the way fear of choking or allergies does. Instead it manifests as a persistent, low-grade conviction that a diet heavy on chicken nuggets and light on broccoli is doing something irreversible. Public-health messaging reinforces the sense of urgency without supplying a number, which leaves most parents oscillating between \"it will be fine\" and \"I am ruining my child.\"\n","rough_estimate":"Most parents sense meaningful risk but cannot quantify it; estimates range from trivial to catastrophic","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/obesity/childhood-obesity-facts/childhood-obesity-facts.html","title":"Childhood Obesity Facts","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"19.7% of US children and adolescents ages 2-19 had obesity in 2017-March 2020, approximately 14.7 million youths","excerpt":"\"From 2017 to March 2020, the prevalence of obesity among U.S. children and adolescents was 19.7%. This means that approximately 14.7 million U.S. youths aged 2–19 years have obesity. Obesity prevalence was 12.7% among U.S. children 2–5 years old, 20.7% among those 6–11, and 22.2% among adolescents 12–19.\"\n","source_date":"2024-09-12","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260503085532/https://www.cdc.gov/obesity/childhood-obesity-facts/childhood-obesity-facts.html","calculation_notes":"CDC page reports NHANES 2017-March 2020 data showing 19.7% obesity prevalence. A newer NCHS Health E-Stat (2021-2023 NHANES) reported 21.1% but those figures appear on a separate NCHS publication, not this URL. We cite only what this page shows. Obesity is the most visible downstream marker of poor dietary quality but is not the only one — micronutrient deficiencies, elevated blood pressure, and insulin resistance also track with diet quality independently of BMI. No single probability is derived because the outcome \"lasting harm\" is multi-dimensional.\n","independence_note":"NHANES is a nationally representative cross-sectional survey with clinical measurements (height, weight, blood draws). The obesity prevalence figure is measured, not self-reported.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4698034/","title":"Poor adherence to U.S. dietary guidelines for children and adolescents in the NHANES population","publisher":"Journal of the Academy of Nutrition and Dietetics","source_type":"peer_reviewed","statistic":"Mean HEI scores for US children ages 4-18 ranged from 43.6 to 52.1 out of 100, well below the 80-point threshold for a healthy diet","excerpt":"\"Total HEI-10 scores for children aged 4-18 ranged from 43.59 to 52.11, much lower than the minimum score of 80 thought to indicate a diet associated with good health. Scores declined significantly with age. Less than 1% of adolescents met whole-grain recommendations, and empty calories from added sugars and solid fats contributed roughly 40% of daily calories for ages 2-18.\"\n","source_date":"2016-01-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426194903/https://pmc.ncbi.nlm.nih.gov/articles/PMC4698034/","calculation_notes":"Analysis of NHANES 2003-2010 data using the Healthy Eating Index 2010 (HEI-10). The 40% empty-calorie contribution and sub-1% whole-grain compliance establish that the vast majority of US children fail to meet dietary guidelines by a wide margin. However, low HEI scores are not a direct probability of harm — they measure diet quality, not health outcomes. The dose-response curve from diet quality to chronic disease is mediated by genetics, physical activity, socioeconomic factors, and decades of cumulative exposure.\n","independence_note":"Uses the same NHANES survey platform as the CDC obesity data but applies a different analytical framework (dietary recall scoring via HEI rather than anthropometric measurement). The two sources measure different facets of the same problem: one captures what children eat, the other captures what happens to their bodies.\n"},{"url":"https://www.cdc.gov/school-nutrition/facts/index.html","title":"Childhood Nutrition Facts","publisher":"Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Empty calories from added sugars and solid fats comprise 40% of daily calories for children ages 2-18; most youth do not meet fruit and vegetable recommendations","excerpt":"\"Empty calories from added sugars and solid fats total 40% of daily calories for children and adolescents 2–18 years of age. Between 2001 and 2010, consumption of sugar-sweetened beverages among children and adolescents decreased. But still accounts for 10% of total calories. Most youth still do not meet fruit and vegetable recommendations.\"\n","source_date":"2024-01-16","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426194940/https://www.cdc.gov/school-nutrition/facts/index.html","calculation_notes":"CDC summary of dietary surveillance data. The page does not contain the \"90% exceed sodium intake\" claim that was previously attributed to it — it mentions sodium only in the context of dietary guidelines recommending reduced intake. The 40% empty-calorie figure and widespread failure to meet fruit/vegetable recommendations still establish that nutritionally inadequate diets are the norm among US children, but the specific 90% sodium statistic must be sourced elsewhere if used.\n","independence_note":"CDC school nutrition page synthesizes multiple data streams including NHANES dietary recalls and USDA school meal program data. Overlaps with the NHANES-based sources above but adds the school-meal policy context.\n"}],"comparison_anchors":[{"label":"Heart-disease death (lifetime, US adult)","lifetime_us_adult":0.085},{"label":"Type 2 diabetes death (lifetime, US adult)","lifetime_us_adult":0.075},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"short_label":"Kids' diet","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is marked no_reliable_estimate because \"lasting harm from an unbalanced diet\" resists compression into a single probability. The outcome space is vast (obesity, type 2 diabetes, cardiovascular disease, hypertension, fatty liver disease, dental caries, iron-deficiency anemia, certain cancers) and the exposure is nearly universal — empty calories comprise 40% of daily intake for children ages 2-18, and mean diet-quality scores sit 30-40 points below the healthy threshold. When nearly everyone is exposed, the meaningful question shifts from \"what is the probability\" to \"how much harm, mediated by which cofactors, over what timeframe.\" Longitudinal data consistently show that childhood obesity tracks into adulthood (~80% persistence) and elevates cardiovascular and metabolic risk, but translating that into a single lifetime number requires collapsing dozens of outcomes, severities, and latencies into one figure. We decline to do so.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A child's plate seen from above with a large portion of fries and a small floret of broccoli pushed to the edge, flat vector illustration."},"canonical_url":"https://likelier.app/child-unbalanced-diet","api_url":"https://likelier.app/api/fears/child-unbalanced-diet.json"},{"slug":"cold-drinks-sore-throat","question":"Does drinking cold beverages or eating ice cream cause sore throats?","quick_answer":"There is no reliable estimate because there is no effect to measure: sore throats are caused by viral and bacterial infections, not by the temperature of cold drinks or ice cream. The widespread folk belief has no controlled evidence behind it.\n","category":"health","tags":["food","kids"],"no_reliable_estimate":true,"perceived":{"description":"The belief that cold food and drink cause sore throats is one of the most persistent household medical myths across cultures. Parents in South Asia, the Middle East, Latin America, and the Mediterranean routinely restrict children's access to ice cream, cold water, and chilled beverages on the grounds that cold temperatures damage the throat. A nationally representative US poll of 1,119 parents (Mott Children's Hospital, 2019) found that 70% of parents follow non-evidence-based cold-prevention folklore with their children. Research on conceptual development finds that cold weather theories for illness causation are \"frequently invoked\" among children and many adults, with reliance declining only as germ theory understanding matures (Sigelman 2012). In reality, sore throats are caused by viral and bacterial pathogens, not by the temperature of whatever the child consumed last Tuesday.\n","rough_estimate":"70% of US parents use at least one non-evidence-based cold-prevention strategy; cold temperature is one of the most prevalent folk theories for illness across cultures","kind":"survey","survey_source":{"title":"Preventing colds in children: Following the evidence?","publisher":"C.S. Mott Children's Hospital National Poll on Children's Health","url":"https://mottpoll.org/reports/preventing-colds-children-following-evidence","year":2019}},"sources":[{"url":"https://mottpoll.org/reports/preventing-colds-children-following-evidence","title":"Preventing colds in children: Following the evidence?","publisher":"C.S. Mott Children's Hospital National Poll on Children's Health (2019)","source_type":"reputable_reference","statistic":"70% of 1,119 US parents (nationally representative, ages 5-12 child households) reported using non-evidence-based cold-prevention folklore; 52% told children not to go outside with wet hair, 48% encouraged more indoor time","excerpt":"\"Seven out of ten parents reported that their overall strategy to help their child avoid colds includes using at least one approach that has little or no scientific evidence, such as telling their children not to go outside with wet hair.\"\n","source_date":"2019-01-21","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260309215707/https://mottpoll.org/reports/preventing-colds-children-following-evidence","calculation_notes":"The Mott 2019 nationally representative poll (N=1,119 US parents with children aged 5-12, GfK household panel, margin of error ±1-4 percentage points) documents broad prevalence of non-evidence-based cold-prevention beliefs in the US. The 70% figure covers folklore strategies generally, including wet hair avoidance and indoor restriction — both rooted in the belief that cold air or cold conditions cause illness. The poll does not separately isolate \"cold food causes sore throat\" as a discrete item, but it anchors the US-specific prevalence of cold-related illness myths. No native or normalized probability is derived because this entry is flagged no_reliable_estimate: the myth posits a causal mechanism that does not exist, so there is no measurable risk probability to report.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21586668/","title":"Age and ethnic differences in cold weather and contagion theories of colds and flu","publisher":"Health Education & Behavior (Sigelman 2012)","source_type":"peer_reviewed","statistic":"Cold weather theories for illness causation are 'frequently invoked' by children and many adults; younger children and ethnic minority children more often attribute colds to cold temperatures than to germ exposure","excerpt":"\"A cold weather theory was frequently invoked to explain colds and to a lesser extent flu but became less prominent with age as children gained command of a germ theory of disease. Mexican American and other minority children were more likely than European American children to subscribe to cold weather theories.\"\n","source_date":"2012-02-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260521021333/https://pubmed.ncbi.nlm.nih.gov/21586668/","calculation_notes":"Sigelman (2012) is a peer-reviewed study (Health Education & Behavior, PMID 21586668) documenting the developmental and cultural persistence of cold-weather illness theories. The finding that cold temperature attribution is \"frequently invoked\" and shows ethnic variation provides the research backing for perceived.description's cultural claims. The study examines cold weather (not specifically cold food or drink), but the same folk-theory framework underlies both beliefs. No quantitative probability is computable from this study for this entry.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7152388/","title":"Pharyngitis","publisher":"Netter's Infectious Diseases via PMC (Bower 2012)","source_type":"reputable_reference","statistic":"70–90% of acute pharyngitis episodes are viral in origin; rhinovirus is by far the most common causative agent","excerpt":"\"Depending on the season and the patient's age, 70% to 90% of acute episodes are viral and involve a wide array of common viruses. By far, the most common virus associated with pharyngitis is the common cold agent, rhinovirus.\"\n","source_date":"2012-03-21","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260216032804/https://pmc.ncbi.nlm.nih.gov/articles/PMC7152388/","calculation_notes":"Bower (2012) is the foundational reference for the infectious etiology of pharyngitis. The 70–90% viral figure establishes that the vast majority of sore throats require exposure to a pathogen, not a cold beverage. Temperature is not listed among the etiological categories for acute pharyngitis in any standard infectious disease reference. Rhinovirus, adenovirus, Epstein-Barr virus, and group A Streptococcus account for the overwhelming majority of cases; none of their transmission or virulence mechanisms involves dietary temperature.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16286463/","title":"Acute cooling of the feet and the onset of common cold symptoms","publisher":"Family Practice (Johnson & Eccles 2005)","source_type":"peer_reviewed","statistic":"In a randomised controlled trial of 180 healthy volunteers, 13/90 chilled subjects reported developing a cold within 4–5 days vs 5/90 controls (P=0.047); the authors conclude that chilling may trigger latent viral symptoms in people already harbouring infection, not that cold exposure causes new infection","excerpt":"\"There is a common folklore that chilling of the body surface causes the development of common cold symptoms, but previous clinical research has failed to demonstrate any effect of cold exposure on susceptibility to infection with common cold viruses.\"\n","source_date":"2005-12-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260504061638/https://pubmed.ncbi.nlm.nih.gov/16286463/","calculation_notes":"Johnson & Eccles (2005) is the most rigorous controlled experiment on chilling and cold symptom onset. The statistically significant result (13 vs 5 colds, P=0.047) was interpreted by the authors as chilling unmasking latent viral infection already present in nasal passages, not as chilling causing new infection. Critically, chilling of the feet is categorically different from drinking a cold beverage: foot chilling affects peripheral vasoconstriction and nasal mucosa temperature via reflex mechanisms, not direct thermal contact with the pharynx. Cold drinks in the mouth and throat do not cause the nasal airway temperature drop that is Eccles's proposed mechanistic pathway for symptomatic reactivation. The PubMed abstract text is used as the excerpt because the full text is behind an Oxford Academic paywall.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/11936911/","title":"An explanation for the seasonality of acute upper respiratory tract viral infections","publisher":"Acta Otolaryngologica (Eccles 2002)","source_type":"peer_reviewed","statistic":"Seasonal exposure to cold air increases URTI incidence by cooling the nasal epithelium and impairing mucociliary clearance — a mechanism specific to nasal airway cooling, not to oral consumption of cold food","excerpt":"\"Seasonal exposure to cold air causes an increase in the incidence of URTI due to cooling of the nasal airway. The inhalation of cold air causes cooling of the nasal epithelium, and this reduction in nasal temperature is sufficient to inhibit respiratory defences against infection such as mucociliary clearance.\"\n","source_date":"2002-03-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260421074110/https://pubmed.ncbi.nlm.nih.gov/11936911/","calculation_notes":"Eccles (2002) provides the mechanistic context for why cold air can plausibly increase infection risk (nasal cooling, impaired mucociliary clearance) while simultaneously showing that this mechanism does not apply to cold food or beverages. Cold drinks lower pharyngeal temperature transiently but do not cool the nasal epithelium through the same reflex pathway that cold air inhalation does. The seasonal correlation between cold weather and sore throat incidence has a real but indirect explanation: people gather indoors, share viral loads, and inhale cold dry air — not because they drink more iced beverages in winter. The abstract text is used as the excerpt source; the full text is paywalled at Tandfonline.\n"}],"comparison_anchors":[],"short_label":"Cold drinks cause sore throat","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the fear posits a causal mechanism — cold temperature directly causing throat infection — that has no established biological plausibility and no controlled experimental support.\nThe only data worth qualifying are two nuances. First, Johnson & Eccles (2005) found that foot chilling did trigger cold-like symptoms in approximately 10% of subjects, which the authors interpreted as unmasking latent nasal viral infection rather than causing new infection. This result is specific to peripheral chilling via reflex vasoconstrictive mechanisms and does not apply to swallowing a cold drink. Second, Eccles (2002) proposes that cold nasal airway exposure impairs mucociliary clearance and may facilitate viral infection — but this is a mechanism for inhaled cold air, not for consumed cold liquid. Cold beverages warm to body temperature within seconds of passing through the oropharynx and do not produce sustained local cooling of respiratory mucosa.\nThe reverse of the folk belief is better supported: cold food and drink are frequently recommended by clinicians to soothe inflamed pharyngeal tissue once sore throat has already developed. Ice lollies, cold water, and ice cream lower the temperature of nerve endings in the throat and reduce pain signalling transiently. The same parents who restrict ice cream to prevent sore throat will typically offer ice cream as comfort when one has developed, an internal inconsistency that reflects intuitive temperature categories rather than any coherent causal model.\nCultural transmission of the cold-food restriction is strongest in communities with traditional humoral medicine frameworks (Ayurveda, Galenic medicine, traditional Chinese medicine) where \"cold\" foods are classified as inherently weakening or pathogenic regardless of microbiological evidence. This makes the belief particularly durable and resistant to correction in clinical settings.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A tall glass of water with ice cubes, condensation on the outside, flat vector illustration in muted blue-grey tones."},"canonical_url":"https://likelier.app/cold-drinks-sore-throat","api_url":"https://likelier.app/api/fears/cold-drinks-sore-throat.json"},{"slug":"cold-surface-bladder-infection","question":"What are the odds of getting a bladder infection from sitting on a cold surface?","quick_answer":"There is no reliable estimate: no cohort study or trial has shown that sitting on a cold surface causes a bladder infection, which is a bacterial illness. The one controlled study used extreme cold-water immersion in already susceptible women, not a park bench. Believers treat it as near-certain; the evidence does not.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"In much of Central and Eastern Europe, Germany, Scandinavia, Russia, and parts of East Asia, the belief that sitting on a cold surface — concrete, stone, a metal bench, cold ground — causes bladder infections (cystitis), kidney infections, or \"inflammation of the ovaries\" is treated as established medical fact. In Polish the warning is universal: \"nie siadaj na zimnym, bo dostaniesz zapalenia pecherza\" (don't sit on cold or you'll get a bladder infection). In German, Blasenentzundung from cold is a standard maternal caution. In Russian, sitting on cold concrete is said to cause \"zastudila pridatki\" (chilled the ovaries/appendages). No rigorous survey has measured how adults in any country rate the numerical per-exposure risk, but the behavioural signal is strong: children, especially girls, are physically pulled off cold steps, newspapers are placed under seated grandmothers, and insulated seat pads are sold specifically to prevent \"catching cold in the bladder.\" The folk model implies that cold temperature on the skin of the buttocks or thighs directly causes bacterial infection in the urinary tract.\n","rough_estimate":"Treated as near-certain by believers — 'you will get an infection' — with no numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://www.niddk.nih.gov/health-information/urologic-diseases/bladder-infection-uti-in-adults/symptoms-causes","title":"Symptoms & Causes of Bladder Infection in Adults","publisher":"National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), NIH","source_type":"govt_report","statistic":"UTIs are caused by bacteria; listed risk factors include female anatomy, sexual activity, certain contraceptives, menopause, catheterisation, urinary tract abnormalities, and immune suppression — cold surface exposure is not listed","excerpt":"\"A bladder infection is an illness most often caused by bacteria. [...] People are more likely to develop a bladder infection if they are sexually active [...] have gone through menopause [...] use certain types of birth control, such as a diaphragm or spermicide [...] have diabetes [...] have a condition or situation that blocks the flow of urine.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260223061424/https://www.niddk.nih.gov/health-information/urologic-diseases/bladder-infection-uti-in-adults/symptoms-causes","calculation_notes":"NIDDK is the US federal authority on urologic diseases. Its patient- facing UTI page enumerates risk factors for bladder infection exhaustively. Cold exposure, cold surfaces, sitting on cold ground, and ambient temperature are entirely absent from the list. The established risk factors are all either anatomical (short female urethra, urinary tract abnormalities), behavioural (sexual activity, catheter use, certain contraceptives), hormonal (menopause, low oestrogen), or immunological (diabetes, immune suppression). This is the authoritative anchor for the \"cold surfaces are not a recognised UTI risk factor\" frame.\n","independence_note":"US federal government health guidance. Editorially independent of the Baerheim experimental trial and the Simmering seasonal- incidence study cited below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/1345322/","title":"Symptomatic lower urinary tract infection induced by cooling of the feet. A controlled experimental trial","publisher":"Scandinavian Journal of Primary Health Care, via PubMed","source_type":"primary_study","statistic":"6/29 cystitis-prone women developed acute urinary symptoms (5 bacteriologically confirmed) after 30-minute cold-water foot immersion vs 0/29 during control period","excerpt":"\"Six subjects developed acute distal urinary symptoms at a mean of 55 hours (95% confidence interval 50 to 61) after the cooling, compared with none in the control period. Five of the six had bacteriologically verified lower UTI.\"\n","source_date":"1992-06-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250717062040/https://pubmed.ncbi.nlm.nih.gov/1345322/","calculation_notes":"Baerheim & Laerum 1992 is the only controlled experimental study directly testing whether cold exposure can trigger UTI. Critical qualifiers: (1) All 29 subjects were pre-selected as cystitis-prone (three or more UTI episodes in the previous 12 months), not a general population sample. (2) The exposure was 30-minute immersion of lower legs and feet in increasingly cold water — far more severe than sitting on a cold bench for a few minutes. (3) The study is open-label and non-randomised (no sham cold exposure), though the within-subject control period (72 hours before cooling) partially mitigates this. (4) N=29 is very small; the 6/29 rate (21%) has wide confidence bounds. (5) The plausible mechanism the authors invoke is that cold-induced peripheral vasoconstriction may reduce mucosal blood flow and local immune defences in the urinary tract, allowing pre-existing colonising bacteria to proliferate. This is a modulation-of-existing-colonisation hypothesis, not a cold-causes-infection-from-nothing hypothesis. The study does not support the folk claim that cold surfaces introduce bacteria; it suggests that in women who already harbour uropathogenic bacteria, severe cold stress may tip the balance toward symptomatic infection.\n","independence_note":"Norwegian primary-care research group (University of Bergen). Methodologically independent of the NIDDK guidance and the Simmering epidemiological study. The companion case-control study by the same group (Baerheim & Laerum, Scand J Prim Health Care 1992;10(1):52-56) found self-reported cold extremities preceding UTI episodes at higher rates than in controls (OR 4.7-5.8), but case-control self-report of cold exposure is subject to recall bias, especially in a culture where the folk belief is well known.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8477900/","title":"Warmer Weather and the Risk for Urinary Tract Infections in Women","publisher":"Journal of Urology, via PubMed Central","source_type":"peer_reviewed","statistic":"UTI incidence 20-30% higher in the week following average temperatures of 25-30 C vs 5-7.5 C; seasonal peak in summer, not winter","excerpt":"\"On days when the prior week's average temperature was between 25 and 30 degrees C, the incidence of urinary tract infections was increased by 20% to 30% relative to when the prior week's temperature was 5 to 7.5 degrees C.\"\n","source_date":"2021-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20251112103823/https://pmc.ncbi.nlm.nih.gov/articles/PMC8477900/","calculation_notes":"Simmering et al. 2021 analysed 15.9 million UTI events across 167 million person-years in US metropolitan areas (2001-2015). The finding that UTI incidence peaks in warmer months, not colder ones, directly contradicts the folk model's prediction. If cold surfaces caused UTIs, winter should show a spike; instead, the data show a 20-30% increase in summer. The authors attribute the summer peak to dehydration, reduced urinary output, and concentrated urine — all of which reduce mechanical bacterial clearance. This is the largest epidemiological dataset on UTI seasonality and it points away from cold exposure as a driver.\n","independence_note":"US epidemiological study (University of Iowa / VA Healthcare). Methodologically and institutionally independent of both the NIDDK guidance and the Baerheim experimental trial. Uses a different design (retrospective population cohort vs controlled provocation) and reaches a conclusion that complicates the cold-exposure hypothesis rather than supporting it.\n"}],"comparison_anchors":[{"label":"Shared toilet seat, any serious infection","lifetime_us_adult":0.000001},{"label":"Food poisoning, lifetime US adult","lifetime_us_adult":0.000537},{"label":"UTI at least once, lifetime (women)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"Central/Eastern Europe (Poland, Russia, Czech Republic, Hungary)","probability":0.000001,"notes":"The belief is near-universal and culturally enforced across generations. \"Nie siadaj na zimnym\" in Polish, similar warnings in Russian, Czech, and Hungarian. Treated as established medical fact by the general public, and sometimes reinforced by general practitioners who grew up with the same folk model. The actual per-sit probability of acquiring a de novo bacterial UTI from a cold surface is not measurably different from zero; the probability placeholder reflects this.\n"},{"region":"Germany, Austria, Scandinavia","probability":0.000001,"notes":"\"Blasenentzundung vom Kaltsitzen\" (bladder infection from sitting on cold) is a standard parental warning in German-speaking countries. The Baerheim & Laerum studies originated in Norway, where the folk belief is strong enough to motivate clinical research. The belief is well established but the evidence base does not support it for the typical exposure scenario.\n"},{"region":"United States, United Kingdom, Australia","probability":0.000001,"notes":"The belief is weaker in English-speaking countries but not absent. Some parents warn against sitting on cold ground. The NIDDK, CDC, and NHS do not list cold surfaces as a UTI risk factor. The cultural signal is milder; fewer people would confidently assert a causal link between a cold bench and a bladder infection.\n"},{"region":"Medical/scientific consensus","probability":0.000001,"notes":"No major urology textbook, clinical guideline, or public health authority lists cold surface exposure as a risk factor for UTI. The sole experimental evidence (Baerheim 1992, N=29, cystitis- prone women, extreme cold-water immersion) suggests a possible provocation effect in a pre-selected high-risk subgroup, not a general causal pathway. The largest seasonal study (Simmering 2021, 15.9M UTI events) finds UTIs peak in summer, not winter.\n"}],"personal_factor_multipliers":[{"factor":"Female (vs male)","multiplier":30,"notes":"Women have a lifetime UTI risk of ~50-60% vs ~5-12% for men, driven by shorter urethra and proximity of urethral opening to the rectum. This is the dominant real risk factor — but it has nothing to do with surface temperature.\n"},{"factor":"History of recurrent UTIs (3+ per year)","multiplier":5,"notes":"Women with recurrent UTIs are the only population in which cold provocation has been experimentally tested (Baerheim 1992). Even in this pre-selected group, the exposure was extreme (30-minute cold-water immersion, not sitting on a bench).\n"},{"factor":"Postmenopausal (no HRT)","multiplier":3,"notes":"Oestrogen decline thins the urogenital mucosa and alters the vaginal microbiome, increasing UTI susceptibility. This is a well-established hormonal risk factor unrelated to temperature.\n"},{"factor":"Sexually active","multiplier":3,"notes":"Sexual intercourse is one of the strongest and most consistently identified behavioural risk factors for UTI in women. The mechanism is mechanical introduction of bacteria into the urethra — the same thing cold surfaces do not do.\n"}],"short_label":"Cold surface & UTI","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"One experimental study deserves to sit in the open rather than be dismissed outright. Baerheim and Laerum (1992) found that 6 of 29 cystitis-prone women developed symptomatic UTI (5 bacteriologically confirmed) after a 30-minute cold-water foot immersion, compared with zero during a control period. The proposed mechanism — cold- induced vasoconstriction reducing mucosal blood flow and local immune function, allowing pre-existing colonising bacteria to proliferate — is biologically plausible and parallels the Eccles/ Foxman mechanism for cold exposure and common cold symptoms. But the study has significant limitations: N=29, open-label, non- randomised, pre-selected cystitis-prone population, and an exposure far more extreme than sitting on a park bench. No replication at scale exists.\nSeparately, cold exposure reliably causes cold diuresis — increased urinary urgency and frequency mediated by peripheral vasoconstriction, central fluid redistribution, and TRPM8 cold-receptor activation in the bladder wall (Imamura et al. 2013). This is real, uncomfortable, and easily mistaken for the onset of cystitis, especially by someone who expects cold to cause bladder infections. The distinction between \"I feel like I need to urinate urgently\" and \"I have a bacterial infection\" is the gap through which the folk belief sustains itself.\nThe seasonal epidemiology further complicates the folk model: the largest US study (Simmering et al. 2021, 15.9 million UTI events) found UTIs peak in summer, not winter, with a 20-30% increase at warm temperatures. If cold surfaces were a meaningful driver, winter should show the spike. The summer peak is better explained by dehydration and reduced urinary output concentrating bacteria.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty stone bench viewed from above, rendered as a flat vector shape in muted grey and cool blue tones on a calm background."},"canonical_url":"https://likelier.app/cold-surface-bladder-infection","api_url":"https://likelier.app/api/fears/cold-surface-bladder-infection.json"},{"slug":"compounded-glp1-dosing-overdose","question":"How likely is an overdose from a mg-to-units dosing error with compounded semaglutide or tirzepatide drawn from a multi-dose vial?","quick_answer":"There is no reliable estimate of how often this happens. Regulators have documented dosing errors of five to 20 times the intended dose, but no source counts the affected patients or the number of vial users, so no incidence rate can be calculated. The real hazard is the mg-to-units arithmetic, not whether the drug is fake.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"Most people who buy compounded GLP-1 drugs from a telehealth site picture the risk as 'is the drug real / is it contaminated', not 'can I draw the wrong number of units out of the vial'. The self-administration arithmetic — converting a milligram prescription into syringe units at a concentration printed on the label — is rarely treated as the hazard, even though it is the one the FDA flagged. No published survey measures how compounded-vial users rate their own dosing-error risk, so this is an editorial read of the framing, not a measured number.","rough_estimate":"Users tend to file the risk under 'fake or dirty drug' rather than 'I will inject 10x the dose I meant to'.","kind":"intuition"},"sources":[{"url":"https://content.govdelivery.com/accounts/USFDA/bulletins/3ab6dfa","title":"FDA Alerts Health Care Providers, Compounders and Patients of Dosing Errors Associated with Compounded Injectable Semaglutide Products","publisher":"U.S. Food and Drug Administration (FDA MedWatch)","source_type":"govt_report","statistic":"Patients administered five to 20 times more than the intended dose of semaglutide; adverse events included nausea, vomiting, abdominal pain, fainting, dehydration, acute pancreatitis and gallstones; some patients sought medical attention or required hospitalization.","excerpt":"In these instances, patients administered five to 20 times more than the intended dose of semaglutide. Adverse events included gastrointestinal effects (e.g., nausea, vomiting, abdominal pain), fainting, headache, migraine, dehydration, acute pancreatitis and gallstones. Some patients sought medical attention or required hospitalization.","source_date":"2024-07-29","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20240730003913/https://content.govdelivery.com/accounts/USFDA/bulletins/3ab6dfa","calculation_notes":"Qualitative magnitude only. The FDA reports a per-error overdose multiple of 5x to 20x the intended dose but gives no count of affected patients and no denominator of compounded-vial users, so no incidence rate can be derived from this source.","independence_note":"Federal regulator surveillance (FAERS-derived adverse-event reports); independent of the poison-center and academic sources below."},{"url":"https://pubmed.ncbi.nlm.nih.gov/37392810/","title":"Administration errors of compounded semaglutide reported to a poison control center—case series","publisher":"Journal of the American Pharmacists Association (JAPhA)","source_type":"peer_reviewed","statistic":"3 cases of adverse drug events after incorrect administration of compounded semaglutide; two patients self-administered 10-fold dosing errors; all experienced nausea, vomiting and abdominal pain lasting days.","excerpt":"We report 3 cases of adverse drug events after incorrect administration of semaglutide for weight loss obtained from compounding pharmacies and an aesthetic spa. Two patients self-administered 10-fold dosing errors. All patients experienced notable symptoms of nausea, vomiting, and abdominal pain with most symptoms lasting for days.","source_date":"2023-06-30","source_accessed":"2026-06-13","archive_url":"https://web.archive.org/web/20260620135506/https://pubmed.ncbi.nlm.nih.gov/37392810/","calculation_notes":"A 3-case series, not a rate. Documents the 10-fold (1000%) overdose mechanism — drawing the wrong volume from a multi-dose vial and confusing milligrams, milliliters and units — but no denominator. Used to characterize mechanism and severity, not probability.","independence_note":"Single regional poison center case series; independent of FDA national surveillance."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12927500/","title":"Adverse events administering glucagon-like peptide-1 receptor agonists: a cross-sectional study (FAERS analysis)","publisher":"PMC (peer-reviewed journal article)","source_type":"peer_reviewed","statistic":"70,955 GLP-1 adverse-event reports (Jan 2015–Dec 2024); GLP-1s associated with 63% administration-related reactions vs 39% for insulin; dosing/administration-error reports rose sharply from Q4 2022; >85% of reports missing product-type information so compounded products could not be reliably isolated.","excerpt":"GLP-1s were associated with a higher share of administration-related reactions (63%) compared to insulin (39%).","source_date":"2025-01-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260620135657/https://pmc.ncbi.nlm.nih.gov/articles/PMC12927500/","calculation_notes":"Confirms administration/dosing errors are a large and rising share of GLP-1 adverse-event reports, and explicitly notes FAERS lacks exposure denominators and cannot reliably separate compounded from manufactured product (>85% missing product type). Directly supports the no_reliable_estimate decision: the numerator is countable, the denominator is not.","independence_note":"Academic FAERS re-analysis; methodologically independent of the FDA alert though both draw on FAERS report streams."},{"url":"https://www.10news.com/health/americas-poison-centers-69-rise-in-weight-loss-drug-overdose-calls-this-year","title":"America's Poison Centers: 69% rise in weight-loss drug overdose calls this year","publisher":"ABC 10News (San Diego), quoting America's Poison Centers","source_type":"news_article","statistic":"GLP-1 exposures reported to US poison centers rose from 2,284 (Jan–Jun 2023) to 3,866 (Jan–Jun 2024), a 69% rise. Compounded-semaglutide-specific exposures rose from 32 in all of 2023 to 159 as of June 30, 2024.","excerpt":"Brown says as of June 30, 2024, there have been 159 such exposures reported to poison centers. In all of 2023, the number was 32.","source_date":"2024-08-01","source_accessed":"2026-06-13","calculation_notes":"Compounded-specific exposures: 32 (2023 full year) -> 159 (H1 2024). Roughly a 10x annualized rise (159 in 6 months extrapolates to ~318/year vs 32). These are exposure CALLS, not hospitalizations, and undercount true incidence (most overdoses never call a poison center). No denominator of compounded-vial users, so no rate. Used for trend and order-of-magnitude only.","independence_note":"America's Poison Centers / NPDS data, independent of FDA FAERS and the academic study."},{"url":"https://www.kff.org/health-costs/kff-health-tracking-poll-may-2024-the-publics-use-and-views-of-glp-1-drugs/","title":"KFF Health Tracking Poll May 2024: The Public's Use and Views of GLP-1 Drugs","publisher":"KFF (Kaiser Family Foundation)","source_type":"reputable_reference","statistic":"About 1 in 8 US adults (12%) have ever taken a GLP-1 drug; 6% currently. 11% of GLP-1 users got the drug from an online provider/website, 10% from a medical spa or aesthetic center.","excerpt":"About eight in ten (79%) adults who have taken GLP-1 drugs report getting these drugs or a prescription for them from their primary care doctor or a specialist, while fewer report getting them from an online provider or website (11%), a medical spa or aesthetic medical center (10%), or from somewhere else (2%).","source_date":"2024-05-10","source_accessed":"2026-06-13","calculation_notes":"Establishes the population scale and that ~1 in 5 GLP-1 users obtain the drug through online/med-spa channels where compounded vials are common. Survey does not isolate compounded-vial users or quantify them, which is precisely why a clean subgroup denominator does not exist.","independence_note":"Nationally representative public-opinion survey; independent of all clinical surveillance sources."}],"comparison_anchors":[],"short_label":"Compounded GLP-1 vial dosing overdose","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"No defensible population probability exists, so this entry carries no_reliable_estimate. Three things break a rate. First, the numerator is poison-center exposure calls and FAERS adverse-event reports, both of which undercount true dosing errors (most overdoses never call a poison center or file a report) and both of which mix severities, from a transient bout of vomiting to acute pancreatitis requiring hospitalization. Second, no denominator exists for the relevant subgroup: nobody publishes a reliable count of people self-injecting compounded GLP-1 from multi-dose vials. The KFF survey scales the whole GLP-1 population and shows ~1 in 5 users obtain the drug via online or med-spa channels, but it does not isolate compounded-vial users, and the FAERS analysis notes that over 85% of reports omit product type, so compounded product cannot be cleanly separated from manufactured pens. Third, and most decisive for the 'lifetime' framing this site normally uses: compounded GLP-1 supply was a transient, shortage-driven phenomenon. The FDA-declared semaglutide and tirzepatide shortages that legally permitted mass compounding resolved across 2024–2025, after which the FDA restricted the practice. A 'lifetime US adult' probability of a harm whose exposure window was a bounded two-to-three-year period for a self-selected subset is close to meaningless. What the sources do support, without a rate: per-error overdoses ran 5x to 20x the intended dose (FDA), 10-fold errors recurred in the documented case series, and compounded-semaglutide exposure calls rose roughly tenfold from 32 in 2023 to 159 in the first half of 2024.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract medication vial beside a syringe marked with measurement lines, muted slate and pale amber, flat vector."},"canonical_url":"https://likelier.app/compounded-glp1-dosing-overdose","api_url":"https://likelier.app/api/fears/compounded-glp1-dosing-overdose.json"},{"slug":"contrail-climate-impact","question":"What are the odds that aircraft contrails contribute meaningfully to climate warming?","quick_answer":"This is not really a personal-probability question. Contrails' warming effect is a systemic atmospheric forcing measured in radiative units, not odds that apply to any individual, so no lifetime probability is given. The real, well-documented contrail-cirrus warming effect is largely crowded out of public awareness by the debunked 'chemtrails' conspiracy.\n","category":"tech","no_reliable_estimate":true,"perceived":{"description":"Public perception of aircraft trails in the sky is dominated by the \"chemtrails\" conspiracy theory, which posits that governments or corporations are deliberately spraying toxic chemicals via commercial aircraft. Surveys have found that 10-30% of adults in Western countries believe or consider plausible some version of this claim. The conspiracy is unfounded: the trails are condensation (water ice crystals) formed when hot, humid jet exhaust meets cold ambient air at cruise altitude, a phenomenon documented in atmospheric science since the 1940s. No credible evidence of deliberate chemical spraying has ever been produced, and the logistical impossibility of a global covert spraying program involving millions of airline employees is rarely addressed by proponents. The irony is that the chemtrails narrative has absorbed almost all public attention about aircraft trails, leaving the real and well-documented climate concern — that persistent contrails and the cirrus clouds they seed trap significant amounts of outgoing longwave radiation — almost entirely outside popular awareness. Contrail cirrus is now estimated to cause more radiative forcing than all the CO2 aviation has ever emitted, yet the issue registers in public discourse almost exclusively through the lens of a debunked conspiracy rather than through the actual atmospheric science.\n","rough_estimate":"Most adults either dismiss contrails entirely or worry about 'chemtrails'; few are aware of the real contrail cirrus warming effect","kind":"intuition"},"sources":[{"url":"https://www.sciencedirect.com/science/article/pii/S1352231020305689","title":"The contribution of global aviation to anthropogenic climate forcing for 2000 to 2018","publisher":"Atmospheric Environment (Elsevier)","source_type":"primary_study","statistic":"Contrail cirrus ERF of 57.4 mW m⁻² in 2018, comprising 57% of the total aviation net ERF of 100.9 mW m⁻²; non-CO2 terms account for 66% of aviation forcing","excerpt":"\"Non-CO2 terms sum to yield a net positive (warming) ERF that accounts for more than half (66%) of the aviation net ERF in 2018.\"\n","source_date":"2021-01-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20250820151256/https://www.sciencedirect.com/science/article/pii/S1352231020305689","calculation_notes":"Lee et al. (2021) provide the most comprehensive assessment of aviation climate forcing to date, covering 2000-2018. The net aviation ERF for 2018 was +100.9 mW m⁻² (5-95% range: 55-145). Contrail cirrus alone contributed 57.4 mW m⁻² (range: 17-98), while CO2 contributed 34.3 mW m⁻². This means contrail cirrus caused 57% of aviation's total ERF and 1.67 times more forcing than aviation CO2. The study also calculated that aviation is warming the climate at approximately three times the rate associated with aviation CO2 emissions alone. When normalized to fuel use, global aviation contributed about 3.5% of net anthropogenic ERF in 2011. Non-CO2 terms contribute about 8 times more than CO2 to the uncertainty in the aviation net ERF estimate.\n"},{"url":"https://acp.copernicus.org/articles/22/10919/2022/","title":"Aviation contrail climate effects in the North Atlantic from 2016 to 2021","publisher":"Atmospheric Chemistry and Physics (Copernicus)","source_type":"peer_reviewed","statistic":"12% of flights in the North Atlantic cause 80% of annual contrail energy forcing; strongly warming contrails form in wintertime, near the tropopause, between 15:00-04:00 UTC","excerpt":"\"Around 12 % of all flights in this region cause 80 % of the annual contrail energy forcing, and the factors associated with strongly warming/cooling contrails include seasonal changes in meteorology and radiation, time of day, background cloud fields, and engine-specific non-volatile particulate matter (nvPM) emissions.\"\n","source_date":"2022-08-29","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260504191445/https://acp.copernicus.org/articles/22/10919/2022/","calculation_notes":"Teoh et al. (2022) quantified contrail forcing in the North Atlantic corridor — one of the world's busiest air traffic regions — over five years (2016-2021). The annual mean contrail cirrus net RF in this region was 204-280 mW m⁻² between 2016 and 2019, with significant inter-annual variability driven by meteorological conditions. The 12%-causes-80% finding demonstrates extreme concentration of contrail warming among a small fraction of flights. An earlier Teoh et al. (2020) study of Japanese airspace found an even more extreme concentration: 2% of flights causing 80% of contrail forcing. Strongly warming contrails form in wintertime, close to the tropopause, between 15:00 and 04:00 UTC, and above low-level clouds. COVID-19 flight reductions of 66% produced a proportional 66% reduction in contrail net RF.\n"},{"url":"https://www.nature.com/articles/s44172-024-00329-7","title":"Feasibility test of per-flight contrail avoidance in commercial aviation","publisher":"Communications Engineering (Nature Portfolio)","source_type":"primary_study","statistic":"AI-guided altitude adjustments reduced detectable contrails by 64% in treatment flights versus control flights (p = 0.0331), with a 2% fuel consumption increase per adjusted flight","excerpt":"\"Using satellite-based imagery we observed 64% fewer contrails in these flights relative to the control group flights, a statistically significant reduction (p = 0.0331).\"\n","source_date":"2024-12-20","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260503085726/https://www.nature.com/articles/s44172-024-00329-7","calculation_notes":"Sonabend-W et al. (2024) conducted the first randomized controlled trial of contrail avoidance in commercial aviation, working with American Airlines, Google Research, and Breakthrough Energy. Pilots made altitude adjustments based on AI contrail formation predictions. Of treatment flights, only 4 produced detectable contrails versus 11 in the control group — a 64% reduction. The fuel penalty was approximately 2% per adjusted flight. Because only a small fraction of flights need adjustment (Teoh et al. found 12% of flights cause 80% of forcing), the fleet-wide fuel impact of selective avoidance would be substantially lower than the per-flight penalty. Breakthrough Energy estimates suggest an abatement cost in the range of $5-25 per ton CO2-equivalent, which would make contrail avoidance one of the cheapest climate interventions if confirmed at scale, though these cost estimates are preliminary and not from the peer-reviewed trial itself.\n"}],"comparison_anchors":[{"label":"Death from air pollution (lifetime, global adult)","lifetime_us_adult":0.05},{"label":"Climate change death (lifetime, global adult)","lifetime_us_adult":0.0049},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"North Atlantic flight corridor","probability":0.000001,"notes":"Structural placeholder. The North Atlantic corridor (US East Coast to Europe) is one of the densest air traffic regions on Earth and has the most-studied contrail climate effects. Teoh et al. (2022) measured annual mean contrail cirrus net RF of 204-280 mW m⁻² in this region — far above the global average — due to high flight density through ice-supersaturated layers near the tropopause. Cold, humid upper- tropospheric conditions in winter make this corridor a persistent contrail hotspot.\n"},{"region":"Southeast Asia and Pacific","probability":0.000001,"notes":"Structural placeholder. Tropical convective regions produce different contrail dynamics. Higher tropopause altitudes and more frequent cirrus background cloud mean contrails often overlap with natural cloudiness, making both detection and net forcing assessment more uncertain. Flight density is growing rapidly in this region, increasing contrail-related forcing.\n"},{"region":"Continental US domestic airspace","probability":0.000001,"notes":"Structural placeholder. US domestic airspace has high flight density but more variable humidity conditions than the North Atlantic corridor. Contrail persistence depends heavily on whether flights traverse ice-supersaturated regions, which shift seasonally and day-to-day. The majority of US domestic flights cruise at altitudes where contrail formation is possible but not guaranteed.\n"},{"region":"Low-traffic equatorial regions","probability":0.000001,"notes":"Structural placeholder. Regions with sparse aviation traffic (central Africa, central South America, central Pacific) contribute minimally to global contrail forcing regardless of atmospheric conditions, simply because few aircraft fly there at contrail-forming altitudes.\n"}],"personal_factor_multipliers":[{"factor":"Frequent flyer (>100,000 km/year)","multiplier":5,"notes":"A frequent flyer's personal contribution to contrail forcing scales roughly with distance flown. However, individual contribution to a systemic forcing phenomenon is conceptually different from personal risk. The multiplier is directional, not a probability modifier — it reflects relative contribution to the problem, not personal exposure to its consequences.\n"},{"factor":"Night-flight preference","multiplier":3,"notes":"Contrails formed at night produce net warming because they trap outgoing longwave radiation without the offsetting daytime shortwave albedo effect. Flights between roughly 16:00 and 08:00 local time generate disproportionate warming contrails. A traveler who predominantly flies overnight long-haul routes contributes more per kilometer to contrail forcing.\n"},{"factor":"Short-haul domestic only","multiplier":0.5,"notes":"Short-haul flights spend less time at contrail-forming altitudes and typically cruise at lower altitudes where ice supersaturation is less common. The per-flight contrail contribution is lower, though the per-kilometer CO2 emissions are higher due to the fuel cost of climb and descent cycles.\n"}],"short_label":"Contrail warming","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses contrail cirrus as a systemic climate forcing mechanism, not a personal health or safety risk. The radiative forcing values (mW m⁻²) describe a global atmospheric energy imbalance, not a probability of harm to any individual. Regional breakdown probabilities are structural placeholders required by the site schema, not measured values. Contrail forcing estimates carry large uncertainties: Lee et al. (2021) report a 5-95% range of 17-98 mW m⁻² for contrail cirrus ERF, a factor-of-six span reflecting deep uncertainty in ice crystal optical properties, contrail-to-cirrus transition rates, and background meteorological conditions. The relationship between radiative forcing and surface temperature change involves additional uncertainties around climate sensitivity. Night-time contrails are net warming; daytime contrails have a partial cooling effect from reflecting incoming shortwave radiation, but the net 24-hour effect is warming. The Sonabend-W et al. (2024) contrail avoidance trial is promising but used a small sample size; fleet-wide operational deployment would require integration with air traffic control systems and resolution of fuel-burn accounting across airlines and regulators.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"extracted-from-transcript","scored_at":"2026-04-27","methodology_version":"1.0"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-27","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A simplified aircraft silhouette leaving a white trail across a pale sky, rendered in muted tones with flat vector styling."},"canonical_url":"https://likelier.app/contrail-climate-impact","api_url":"https://likelier.app/api/fears/contrail-climate-impact.json"},{"slug":"cruise-ship-accident","question":"What are the odds of dying in a cruise ship accident?","quick_answer":"There is no reliable per-passenger figure. The industry incident reports count fatalities but publish no passenger denominator, so a lifetime probability cannot be derived. The Costa Concordia disaster dominates perception, though it stands as the only major structural-accident mass-casualty event in three decades of modern cruising.\n","category":"transport","tags":["travel"],"no_reliable_estimate":true,"perceived":{"description":"The Costa Concordia disaster of January 2012 — 32 deaths, a capsized hull visible from the Italian shore for years — is almost certainly the dominant image many people hold when considering cruise ship safety. The wreck became a sustained media story that generated criminal proceedings, compensation battles, and environmental coverage lasting a decade. Before Concordia, many passengers had internalized a vague sense of cruise ships as seaworthy and modern; after it, the image of a tilting mega-ship in shallow water became available at minimal mental cost. No polling specifically tracks fear of cruise ship disasters, but the event's memorability reliably inflates perceived risk above the historical base rate.\n","rough_estimate":"Most cruisers who think about it recall Concordia; few have a sense that it is the only significant structural-accident mass-casualty event in 30 years of modern cruising","kind":"intuition"},"sources":[{"url":"https://cruising.org/resources/report-operational-incidents-2009-2019","title":"Report on Operational Incidents 2009 to 2019 For CLIA","publisher":"G.P. Wild (International) Ltd for Cruise Lines International Association (CLIA)","source_type":"reputable_reference","statistic":"Over 2009-2019 (11 years): 195 significant operational incidents causing 65 fatalities (34 passengers, 31 crew) and 292 injuries; separately, 212 man-overboard incidents causing 170 fatalities (117 passengers, 53 crew); 2019 had zero passenger or crew fatalities from significant incidents; significant-incident rate declined as fleet capacity grew 68.1%","excerpt":"\"Total 11 years [2009 to 2019]: 195 [significant operational incidents], Fatalities Pax 34, Crew 31, Total 65, Injuries 292... In 2019, only thirteen significant operational incidents occurred and there were no fatalities of either passengers or crew... [Man overboard] Totals 212 [incidents], Fatalities Pax 117, Crew 53, Total 170.\"\n","source_date":"2020-05-15","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260531135432/https://cruising.org/resources/report-operational-incidents-2009-2019","calculation_notes":"Figures taken verbatim from the report's summary tables: ES Table 1 / Table 2.1 (Significant Operational Incidents 2009-2019 = 195 incidents, 65 fatalities split 34 passengers / 31 crew, 292 injuries) and ES Table 3 / Table 2.3 (Man Overboard 2009-2019 = 212 incidents, 170 fatalities split 117 passengers / 53 crew). A \"significant operational incident\" is defined as one causing more than 24 hours' itinerary delay, any fatality, or a serious injury. The report normalizes exposure against \"active lower berths\" (a fleet-capacity measure), not passenger-voyages, and states no per-passenger fatality rate, no per-passenger-mile figure, and no cross-modal transport comparison — so no lifetime probability can be derived from it without an externally supplied passenger denominator. G.P. Wild is an independent maritime consultancy with no financial stake in cruise lines. Earlier drafts of this entry cited \"50 deaths 2009-2013\" and \"0.08 fatalities per billion passenger miles, lowest of six transport modes\"; neither figure appears anywhere in the report and both have been removed as unsupported.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S2210539514000923","title":"Understanding the causes of recent cruise ship mishaps and disasters","publisher":"Research in Transportation Business & Management, Vol. 13 — Mileski, Wang and Beacham, 2014","source_type":"peer_reviewed","statistic":"580 cruise ship mishap incidents analyzed over 1989-2013 via a two-stage measurement design (abstract-confirmed; body paywalled)","excerpt":"\"The goal of this research is to contribute to the understanding of the nature of mishaps in the current cruise industry by evaluating 580 mishap incidents from 1989 to 2013 through a two-stage measurement design.\" [Paraphrase note: only the abstract is retrievable; the full text is paywalled.]\n","source_date":"2014-12-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20240416072648/https://www.sciencedirect.com/science/article/pii/S2210539514000923","calculation_notes":"This source is retained for its scope (580 cruise-ship mishap incidents, 1989-2013), which the open abstract states verbatim. A figure widely attributed to this paper — \"16 fatalities out of more than 100 million passengers\" — sits in the paywalled body and could NOT be verified against any retrievable text (abstract, preprint, or archive); an earlier draft of this entry quoted it as a verbatim excerpt, which was unsupported and has been removed. Because that potential per-passenger rate is unverifiable, and the authoritative G.P. Wild dataset supplies only absolute counts with a capacity-based denominator, no numeric lifetime probability is published for this entry. [Paraphrase from abstract — full text paywalled.]\n","independence_note":"Mileski et al. conducted an independent academic analysis of cruise ship mishap reports drawn from public maritime safety databases, distinct from CLIA's own G.P. Wild-commissioned dataset — a useful independent bound on the count of mishap incidents.\n"},{"url":"https://www.ijtmgh.com/article_119591.html","title":"Death at Sea: Passenger and Crew Mortality on Cruise Ships","publisher":"International Journal of Travel Medicine and Global Health — Lopes, Dinis, Brito et al., 2020","source_type":"peer_reviewed","statistic":"623 total reported deaths across 78 cruise lines (2000-2019, 20 years); 89% passengers, 11% crew; leading causes: falls overboard 23%, suicide/murder 19%, natural causes 18%, cardiac 16%","excerpt":"\"A total of 623 reported deaths were found across 78 ocean and river cruise lines from 2000 to 2019... 557 (89%) were passengers and 66 (11%) crew members... The leading causes of passenger death were falls overboard/onto lower decks (23%), suicide/murder/terrorism (19%), unspecified natural causes (18%), and cardiac incidents (16%).\"\n","source_date":"2020-12-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20240416070953/https://www.ijtmgh.com/article_119591.html","calculation_notes":"The 623 all-cause deaths over 20 years average ~31 deaths/year across all causes combined — overwhelmingly natural and behavioral causes rather than structural accidents. Deaths from structural failures (sinkings, fires, collisions) are a tiny subset dominated almost entirely by the Concordia event (32 deaths, 2012). This source confirms that on large ocean cruise ships, the primary safety concern is falls overboard and medical emergencies, not maritime disasters. The structural accident death rate is effectively zero in the 2013-2024 period outside of any single catastrophic event.\n","independence_note":"Lopes et al. compiled a dataset from publicly available maritime incident databases, news records, and cruise line reports — methodologically distinct from CLIA's G.P. Wild operational data. The convergence on ~30 deaths/year across all causes provides an independent upper bound for the total mortality rate.\n"}],"comparison_anchors":[{"label":"Dying in a plane crash (lifetime, US adult, ~2 flights/yr)","lifetime_us_adult":0.000017},{"label":"Ferry accident death (global adult lifetime)","lifetime_us_adult":0.0000075},{"label":"Asteroid impact death (lifetime, US adult)","lifetime_us_adult":7.4e-7}],"short_label":"Cruise ship accident","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"No numeric lifetime probability is published for this entry because the cited sources give absolute fatality counts without a per-passenger or per-voyage denominator, and the one paper widely cited for a per-passenger rate (Mileski et al., \"16 per 100 million\") keeps that figure in a paywalled body that could not be verified. The most comprehensive dataset, G.P. Wild for CLIA, records 65 fatalities from 195 significant operational incidents and a further 170 man-overboard fatalities over the eleven years 2009-2019, but measures exposure against fleet \"active lower berths\" rather than passenger-voyages, so no per-passenger rate can be read from it without inventing a passenger denominator. The record is dominated by the Costa Concordia disaster (January 2012, 32 deaths); 2019 saw zero significant-incident fatalities, and the significant-incident count fell even as fleet capacity grew 68.1%. Structural-accident mortality is a long-tail risk driven almost entirely by rare single catastrophic events, so any decade may be dominated by zero deaths or by one disaster. This entry explicitly excludes ferries: the MV Doña Paz, Estonia, Sewol, and similar ferry disasters involved smaller vessels with different safety regimes; a separate entry covers ferry sinking risk.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A tilted ship silhouette on a calm ocean, flat vector illustration with muted colours."},"canonical_url":"https://likelier.app/cruise-ship-accident","api_url":"https://likelier.app/api/fears/cruise-ship-accident.json"},{"slug":"cryptocurrency-total-loss","question":"What are the odds of losing your entire cryptocurrency investment?","quick_answer":"There is no defensible figure for an individual holder's odds of a total loss. The cited sources measure aggregate dollar losses from fraud and theft, not the probability that a given holder is wiped out, and those losses are concentrated among the minority who are victimized. Holders tend to put their own wipeout risk below 10%; non-holders estimate much higher.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Crypto holders exist in a peculiar dual state: simultaneously aware that exchanges collapse, wallets vanish, and tokens go to zero, yet confident that their particular holdings are on the right side of history. No rigorous survey isolates the perceived probability of total loss among holders, but the general posture in crypto communities — diamond hands, HODL culture, \"have fun staying poor\" — suggests most active participants put their personal risk of a complete wipeout well below 10%. Non-holders, by contrast, treat the entire asset class as a casino and would likely estimate higher.\n","rough_estimate":"holders: <10%; non-holders: much higher","kind":"intuition"},"sources":[{"url":"https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","title":"2024 Internet Crime Report","publisher":"FBI Internet Crime Complaint Center (IC3)","source_type":"govt_report","statistic":"Over 140,000 cryptocurrency-related complaints in 2024, totaling approximately $9.3 billion in losses; individuals over 60 filed ~33,000 complaints with $2.8B in losses","excerpt":"\"The IC3 received more than 140,000 complaints referencing cryptocurrency in 2024, resulting in roughly $9.3 billion in losses.\"\n","source_date":"2025-04-23","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260406100825/https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","calculation_notes":"FBI IC3 captures reported crypto fraud losses in the US: more than 140,000 cryptocurrency-related complaints in 2024 totaling roughly $9.3B, of which individuals over 60 accounted for about 33,000 complaints and $2.8B. These are reported losses only; actual losses are likely higher since many victims never file. They span investment scams, romance scams with crypto payment, and other fraud where cryptocurrency was the payment method. This is an aggregate dollar figure, not a per-holder loss rate — the losses are highly concentrated among the minority of holders who are victimized, so it cannot be turned into a probability that a given holder suffers a total loss.\n"},{"url":"https://www.chainalysis.com/blog/crypto-hacking-stolen-funds-2025/","title":"$2.2 Billion Stolen in Crypto in 2024 but Hacked Volumes Stagnate","publisher":"Chainalysis","source_type":"reputable_reference","statistic":"$2.2 billion stolen via crypto hacks in 2024 across 303 incidents; private key compromises accounted for 43.8% of stolen crypto; North Korean hackers stole $1.34B (61% of total)","excerpt":"\"Funds stolen increased by approximately 21.07% year-over-year to $2.2 billion, and the number of individual hacking incidents increased from 282 in 2023 to 303 in 2024.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260320083826/https://www.chainalysis.com/blog/crypto-hacking-stolen-funds-2025/","calculation_notes":"Chainalysis tracks on-chain theft across exchanges, DeFi protocols, and bridges. The $2.2B in 2024 and $3.4B in 2025 (including the $1.5B Bybit hack) represent direct theft via hacking — separate from the FBI's fraud figures, which include social-engineering scams. This source measures aggregate stolen dollars, not the probability that an individual holder is a theft victim, so it bounds the theft vector's magnitude rather than a per-holder loss rate. (An earlier draft attributed a \"3-4 million BTC permanently lost\" lost-key figure to this hacking article; that article does not state it, so the claim has been removed pending a correctly cited lost-key source.)\n","independence_note":"Chainalysis tracks on-chain fund flows independently from FBI IC3 complaint data. The two sources measure different things: Chainalysis captures hacking and theft visible on-chain; IC3 captures victim-reported fraud complaints. Some overlap exists where hack victims also file IC3 complaints.\n"},{"url":"https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024","title":"New FTC Data Show a Big Jump in Reported Losses to Fraud to $12.5 Billion in 2024","publisher":"Federal Trade Commission","source_type":"govt_report","statistic":"Cryptocurrency accounted for $1.42 billion in reported consumer fraud losses by payment method in 2024; Bitcoin ATM fraud losses topped $65 million in H1 2024 alone","excerpt":"\"Consumers reported losing more than $12.5 billion to fraud in 2024, which represents a 25% increase over the prior year.\"\n","source_date":"2025-03-06","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260417012350/https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024","calculation_notes":"FTC Consumer Sentinel data captures fraud where crypto was the payment method, a narrower slice than FBI IC3 (which counts all crypto-related fraud). The $1.42B FTC figure is a subset of the $9.3B IC3 figure. These are aggregate fraud-loss figures, not per-holder loss rates, so they bound the fraud vector's magnitude rather than the probability of total loss.\n","independence_note":"FTC Consumer Sentinel collects complaints from a different intake channel than FBI IC3. There is substantial overlap — some consumers report to both — but the FTC captures a broader range of consumer fraud complaints and uses different categorization.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Home burglary (lifetime, US household)","lifetime_us_adult":0.135}],"personal_factor_multipliers":[{"factor":"holds only BTC/ETH on regulated US exchange (Coinbase, Kraken)","multiplier":0.3,"notes":"regulated exchanges carry FDIC-insured USD deposits; BTC/ETH have survived multiple 80%+ crashes and recovered; total-loss risk is primarily exchange failure or regulatory seizure, both low for regulated US platforms"},{"factor":"holds altcoins / meme coins / DeFi tokens","multiplier":3,"notes":"a large fraction of smaller crypto projects are eventually abandoned or go to zero; altcoin holders face rug pull, abandonment, and zero-liquidity risk that BTC/ETH holders largely avoid"},{"factor":"uses offshore or unregulated exchange","multiplier":2.5,"notes":"offshore platforms lack deposit insurance, regulatory oversight, and legal recourse; a number of failed exchanges have disappeared without a trace or means of recovery"},{"factor":"self-custody with hardware wallet and backup","multiplier":0.5,"notes":"eliminates exchange counterparty risk; lost-key risk drops to near zero with proper seed phrase backup; residual risk is market crash only"},{"factor":"self-custody without seed phrase backup","multiplier":2,"notes":"single point of failure; if a device is lost, damaged, or passwords forgotten, self-custodied funds can become permanently irrecoverable — the lost-key vector"},{"factor":"invested more than 50% of net worth","multiplier":1.5,"notes":"concentration risk; a portfolio-level total loss is more likely when crypto is the dominant holding rather than a small allocation"},{"factor":"age 60+","multiplier":2,"notes":"FBI IC3 2024: Americans over 60 filed ~33,000 crypto complaints with $2.8B in losses — a large share of reported crypto fraud; seniors are disproportionately targeted by investment scams"}],"short_label":"Crypto total loss","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"We deliberately do NOT publish a per-holder total-loss probability for crypto. The cited sources quantify the aggregate magnitude of the loss vectors — FBI IC3 reported ~$9.3B in crypto-related fraud losses across more than 140,000 complaints in 2024; Chainalysis tracked ~$2.2B stolen via on-chain hacks in 2024; the FTC recorded ~$1.4B in fraud where cryptocurrency was the payment method — but none of them measures the chance that an individual holder ends up with a total or near-total loss. Converting aggregate dollar losses into a per-holder probability requires assumptions no source states (a holder-count denominator, a victim fraction, a definition of \"total\"), so we treat the headline as unestimable rather than fabricate a number. \"Total loss\" here means losing 95%+ of a crypto position through a mechanism other than ordinary market decline and recovery: Bitcoin has dropped 80%+ from all-time highs several times (2011, 2014-15, 2018, 2022) and recovered each time, so a temporary drawdown is not a total loss for a BTC holder who held through it, whereas an altcoin that went to zero and never recovered is. The relative risk clearly differs by behavior — the personal factor multipliers describe the direction (regulated-exchange BTC/ETH lower; altcoins, offshore exchanges, and un-backed-up self-custody higher) without asserting a baseline probability.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-research-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single small padlock lying open on a pale surface beside a faintly sketched key outline, muted tones, flat vector illustration."},"canonical_url":"https://likelier.app/cryptocurrency-total-loss","api_url":"https://likelier.app/api/fears/cryptocurrency-total-loss.json"},{"slug":"cyberattack-infrastructure","question":"What are the odds of a major cyberattack disrupting critical infrastructure (power grid, water, hospitals) in your lifetime?","quick_answer":"No reliable probability can be placed on a major infrastructure cyberattack. The evidence is expert threat assessments and documented intrusions, not a measurable rate. Fear is near-universal (about 58% of US adults report being afraid of cyberterrorism), but that reflects perception rather than a calculated likelihood.\n","category":"tech","no_reliable_estimate":true,"perceived":{"description":"Cyberattacks on critical infrastructure rank among the most feared technological risks worldwide, and elite opinion reinforces that fear. The World Economic Forum's Global Risks Report has placed \"cyber insecurity\" in the top six risks on its two-year outlook every year since 2020, and the 2026 edition ranks it sixth. The 2026 ODNI Annual Threat Assessment warns that China, Russia, Iran, and North Korea are actively pre-positioning inside US critical infrastructure networks. A 2025 WEF Global Cybersecurity Outlook survey found that 94% of executives identify AI as the most significant driver of cybersecurity change. The perception is near-unanimous among decision-makers: a catastrophic infrastructure cyberattack is not a question of if, but when. Among the general public, the fear is amplified by high-profile incidents — Colonial Pipeline, Change Healthcare — and by fictional depictions of grid-down scenarios that collapse in hours.\n","rough_estimate":"58.3% of US adults report being afraid or very afraid of cyberterrorism (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.cisa.gov/news-events/cybersecurity-advisories/aa24-038a","title":"PRC State-Sponsored Actors Compromise and Maintain Persistent Access to U.S. Critical Infrastructure","publisher":"Cybersecurity and Infrastructure Security Agency (CISA)","source_type":"govt_report","statistic":"Volt Typhoon actors maintained access to some US critical infrastructure IT environments for at least five years, pre-positioning for potential disruption of OT assets","excerpt":"\"The U.S. authoring agencies assess with high confidence that Volt Typhoon actors are pre-positioning themselves on IT networks to enable lateral movement to OT assets to disrupt functions. … Volt Typhoon actors have maintained access and footholds within some victim IT environments for at least five years.\"\n","source_date":"2024-02-07","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426195403/https://www.cisa.gov/news-events/cybersecurity-advisories/aa24-038a","calculation_notes":"CISA's joint advisory with NSA, FBI, and Five Eyes partners confirmed that Chinese state-sponsored actors had compromised communications, energy, transportation, and water/wastewater systems across the continental US and Guam. The advisory characterizes the activity as pre-positioning for disruptive operations during geopolitical tensions, not traditional espionage. This establishes that the capability to disrupt US infrastructure exists and has been demonstrated at the access level, but does not provide a frequency rate because no large-scale disruption has been executed. The gap between access and execution is the core reason this entry uses no_reliable_estimate.\n","independence_note":"Joint advisory from CISA, NSA, FBI, and Five Eyes intelligence partners. This is the primary US government source on Volt Typhoon; independent of WEF survey data and GAO audit findings.\n"},{"url":"https://www.gao.gov/products/gao-24-106221","title":"Critical Infrastructure Protection: Agencies Need to Enhance Oversight of Ransomware Practices and Assess Federal Support","publisher":"US Government Accountability Office","source_type":"govt_report","statistic":"None of the federal agencies designated as lead for risk management of selected critical infrastructure sectors have determined the extent of adoption of NIST ransomware practices; FBI reported 870 critical infrastructure ransomware victims in 2022","excerpt":"\"None of the selected SRMAs [Sector Risk Management Agencies] have determined the extent to which their respective sectors have adopted the recommended NIST practices for addressing ransomware. … Among the 870 critical infrastructure organizations that the [FBI] reported to be victims of ransomware, nearly half came from [manufacturing, energy, healthcare, and transportation] sectors.\"\n","source_date":"2024-01-25","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260511051234/https://www.gao.gov/products/gao-24-106221","calculation_notes":"GAO's audit establishes the defensive posture gap: the US government does not know how well its own critical infrastructure sectors have adopted baseline ransomware protections. The FBI's count of 870 critical infrastructure ransomware victims in a single year (2022) across 14 of 16 sectors illustrates the scale, while the SRMAs' failure to even measure adoption of NIST practices shows the defensive side is poorly characterized. Used here as corroborating evidence for why no reliable estimate is possible: neither attack frequency nor defensive readiness is comprehensively measured.\n"},{"url":"https://www.aha.org/2024-03-15-aha-survey-change-healthcare-cyberattack-significantly-disrupts-patient-care-hospitals-finances","title":"AHA Survey: Change Healthcare Cyberattack Significantly Disrupts Patient Care, Hospitals' Finances","publisher":"American Hospital Association","source_type":"reputable_reference","statistic":"74% of hospitals reported direct patient care impact; 94% reported financial impact with more than half reporting 'significant or serious' impact; 83% reported cash flow impacts","excerpt":"\"74% of hospitals report direct patient care impact. … 94% of hospitals report financial impact, with more than half reporting 'significant or serious' impact. … 83% of hospitals report impacts on their cash flow.\"\n","source_date":"2024-03-15","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034143/https://www.aha.org/2024-03-15-aha-survey-change-healthcare-cyberattack-significantly-disrupts-patient-care-hospitals-finances","calculation_notes":"The Change Healthcare attack (February 2024) is the single most disruptive healthcare cyberattack in US history. Change Healthcare processes approximately 15 billion healthcare transactions annually, touching 1 in 3 patient records. UnitedHealth Group reported $2.87 billion in total cyberattack costs for 2024. The attack exposed data of over 100 million individuals. However, it disrupted billing and claims processing — not clinical systems directly — and no patient deaths were attributed to it. This illustrates the pattern: maximum disruption, but mediated through administrative systems, not direct physical harm.\n","independence_note":"AHA survey of member hospitals is independent of CISA's threat intelligence and GAO's audit findings. AHA represents the hospital perspective on operational impact rather than the attacker capability perspective.\n"},{"url":"https://en.wikipedia.org/wiki/Colonial_Pipeline_ransomware_attack","title":"Colonial Pipeline ransomware attack","publisher":"Wikipedia (sourced from CISA, DOJ, and contemporaneous reporting on the May 2021 incident)","source_type":"encyclopedia","statistic":"About 45% of all fuel consumed on the East Coast arrives via the Colonial Pipeline system; the company paid a ransom of nearly 75 Bitcoin (~$4.4 million) to the DarkSide group; the pipeline resumed normal operations after a six-day shutdown; President Biden declared a state of emergency on May 9, 2021","excerpt":"\"About 45% of all fuel consumed on the East Coast arrives via the pipeline system. … the company paid a ransom of nearly 75 Bitcoins ($4.4 million USD) … ending a six-day shutdown. … U.S. President Joe Biden declared a state of emergency on May 9, 2021.\"\n","source_date":"2021-05-07","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260630085515/https://en.wikipedia.org/wiki/Colonial_Pipeline_ransomware_attack","calculation_notes":"Grounds the Colonial Pipeline figures used in the body prose (45% of East Coast fuel supply, $4.4 million ransom, six-day shutdown, presidential emergency declaration). No injuries or deaths were attributed to the incident — the disruption was to pipeline control/billing systems, not physical safety systems — consistent with the pattern of high-disruption, low-physical-harm infrastructure cyberattacks documented across this entry's other sources.\n","independence_note":"Independent of the CISA Volt Typhoon advisory, the GAO ransomware audit, the AHA Change Healthcare survey, and the ODNI threat assessment already cited in this entry; draws on contemporaneous DOJ, CISA, and news reporting of the May 2021 incident.\n"},{"url":"https://en.wikipedia.org/wiki/2015_Ukraine_power_grid_hack","title":"2015 Ukraine power grid hack","publisher":"Wikipedia (sourced from CISA ICS-CERT Alert IR-ALERT-H-16-056-01, SANS ICS, and E-ISAC)","source_type":"encyclopedia","statistic":"230,000 consumers lost power for 1–6 hours on December 23, 2015; first publicly confirmed cyberattack to cause a power outage","excerpt":"\"The power grid in two western oblasts of Ukraine was hacked, which resulted in power outages for roughly 230,000 consumers in Ukraine for 1–6 hours. The attack took place during the ongoing Russo-Ukrainian War and is attributed to a Russian advanced persistent threat group known as 'Sandworm'.\"\n","source_date":"2015-12-23","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260406013320/https://en.wikipedia.org/wiki/2015_Ukraine_power_grid_hack","calculation_notes":"The 2015 Ukraine attack (BlackEnergy malware) and the 2016 follow-up (Industroyer/ CrashOverride) remain the only publicly confirmed cyberattacks to have caused civilian power outages. Both occurred in a wartime context against a grid with weaker cyber defenses than Western systems. A third Sandworm attack using Industroyer2 was intercepted in April 2022 before causing an outage. The Poland grid attack attributed to Sandworm in late 2025 (DynoWiper malware) disrupted operations but details remain limited. The total confirmed count of cyber-caused civilian power outages in history is two, both in Ukraine. This extreme rarity of successful execution — despite demonstrated capability — is the central data point for this entry.\n"},{"url":"https://www.dni.gov/files/ODNI/documents/assessments/ATA-2026-Unclassified-Report.pdf","title":"2026 Annual Threat Assessment of the U.S. Intelligence Community","publisher":"Office of the Director of National Intelligence","source_type":"govt_report","statistic":"China is the most active and persistent cyber threat to US critical infrastructure; North Korea stole an estimated $2 billion in cryptocurrency in 2025 alone","excerpt":"\"China, Russia, Iran, North Korea, and non-state ransomware groups will continue to seek to compromise US government and private-sector networks, as well as critical infrastructure, to collect intelligence, create options for future disruption, and for financial gain.\"\n","source_date":"2026-03-25","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420034232/https://www.dni.gov/files/ODNI/documents/assessments/ATA-2026-Unclassified-Report.pdf","calculation_notes":"The ODNI threat assessment confirms persistent state-sponsored infiltration of US critical infrastructure but does not provide a probability of successful large-scale disruption. The assessment characterizes the activity as capability-building and pre-positioning rather than active disruption. This is consistent with the pattern across all sources: high capability, low execution rate for attacks causing widespread civilian harm. The assessment also notes that AI is accelerating offensive cyber capabilities, introducing a trend break that makes historical base rates unreliable for forward projections.\n"}],"comparison_anchors":[{"label":"Data breach exposure (lifetime, US adult)","lifetime_us_adult":0.95},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Nuclear accident harm (lifetime, near US reactor)","lifetime_us_adult":0.0000012}],"regional_breakdown":[{"region":"Power grid (US/Western)","probability":0,"notes":"Zero confirmed cyber-caused outages in Western grids as of early 2026; Ukraine (wartime) is the only precedent"},{"region":"Healthcare / hospitals","probability":0,"notes":"No confirmed patient deaths from infrastructure-targeting cyberattacks in the US; one UK death linked to delayed lab results (2024)"},{"region":"Water systems","probability":0,"notes":"The 2021 Oldsmar FL incident was caught immediately and the FBI later found no evidence of external intrusion; no confirmed water contamination from cyberattack"},{"region":"Financial systems","probability":0,"notes":"No confirmed large-scale civilian payment disruption from cyberattack; Change Healthcare (2024) disrupted billing but not direct patient care systems"},{"region":"Ukraine / active conflict zones","probability":0,"notes":"Two confirmed cyber-caused power outages (2015, 2016), wartime context, each restored within hours"}],"short_label":"Infrastructure cyberattack","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"property","valence":"negative","caveats":"This entry uses no_reliable_estimate because the concept of \"a major cyberattack disrupting critical infrastructure in your lifetime\" resists quantification on multiple axes. First, the definition of \"major\" is load-bearing: Colonial Pipeline (2021) caused fuel shortages across 17 states and a presidential emergency declaration, yet no one was physically harmed and the pipeline was operational within six days. Change Healthcare (2024) disrupted 74% of US hospitals' patient care workflows and cost $2.87 billion, yet no patient deaths were attributed to it. These are the two most consequential US infrastructure cyberattacks in history, and neither caused the grid-down, water-poisoned scenario that the fear typically conjures. Second, the threat landscape is non-stationary: AI is accelerating offensive capabilities, state actors are pre-positioning inside networks (Volt Typhoon maintained access for five years), and defensive readiness is poorly measured (GAO found that no sector risk management agency has even determined the extent of NIST ransomware practice adoption). Any base rate derived from 2015–2025 incident data would be stale before publication. Third, the gap between access and execution is vast: infiltrating a network is not the same as causing a blackout, and the two confirmed cyber-caused power outages in history both occurred in a wartime context against a less defended grid. A probability derived from two Ukrainian incidents across all of human history would be mathematically meaningless.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A simplified power transmission tower with a single thin crack running through one support, flat vector illustration, muted tones."},"canonical_url":"https://likelier.app/cyberattack-infrastructure","api_url":"https://likelier.app/api/fears/cyberattack-infrastructure.json"},{"slug":"dating-app-violence","question":"How likely is a dating app user to experience sexual violence from a match?","quick_answer":"This risk cannot be reliably quantified. The denominator, the total number of dating-app match meetings, does not exist in any dataset, and incident data is buried inside broader online-facilitated crime statistics. Surveys capture harassment such as unwanted explicit content, but not sexual-assault rates from in-person meetings.\n","category":"tech","tags":["relationships","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"High-profile cases of violence following dating app meetings generate significant media coverage and drive acute fear. Users — particularly women — intuitively sense that the vetting absent in app-based introductions poses risk, and media coverage reinforces this perception. On the other side, dating app companies have resisted disclosing internal safety data, and no government regulator publishes app-matched sexual violence rates. The result is a risk that is widely perceived but impossible to quantify: the denominator (total app-match meetings) does not exist anywhere, and numerator data is buried in broader \"online-facilitated\" crime statistics that aggregate many platforms and crime types.\n","rough_estimate":"No defensible denominator exists: true victimization rate is unknowable with available data","kind":"intuition"},"sources":[{"url":"https://www.pewresearch.org/internet/2023/02/02/from-looking-for-love-to-swiping-the-field-online-dating-in-the-u-s/","title":"From Looking for Love to Swiping the Field: Online Dating in the U.S.","publisher":"Pew Research Center","source_type":"reputable_reference","statistic":"56% of women under 50 who used dating apps were sent unwanted sexually explicit content; 43% were contacted after they said they were not interested; 6% of online daters were threatened with physical harm","excerpt":"\"Some 56% of women under 50 who have ever used a dating site or app have been sent unwanted sexually explicit content and about four-in-ten each have been contacted after they said they were not interested (43%) or called an offensive name (37%). ... 6% have had someone threaten to physically harm them.\"\n","source_date":"2023-02-02","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505052448/https://www.pewresearch.org/internet/2023/02/02/from-looking-for-love-to-swiping-the-field-online-dating-in-the-u-s/","calculation_notes":"Pew Research Center, nationally representative US adult survey (n=6,034). The 56% unwanted-explicit-content figure (women under 50) and 6% threat figure (online daters) describe harassment and threats, not sexual assault from in-person meetings. No denominator for \"meetings that resulted in violence\" is available. This source establishes the prevalence of online harassment but cannot be used to calculate an assault rate per meeting or per user-year.\n"},{"url":"https://www.propublica.org/article/tinder-lets-known-sex-offenders-use-the-app-its-not-the-only-one","title":"Tinder Lets Known Sex Offenders Use the App. It's Not the Only One","publisher":"ProPublica / Columbia Journalism Investigations","source_type":"news_article","statistic":"Match Group does not screen for registered sex offenders on Tinder, Hinge, OkCupid, or PlentyOfFish; Columbia Journalism Investigations analyzed more than 150 incidents of sexual assault involving dating apps, with 10% of incidents involving a match previously accused or convicted of sexual assault; no aggregate incidence denominator exists","excerpt":"\"Match Group does not conduct criminal background checks on users of its free dating apps — Tinder, Hinge, OkCupid, PlentyOfFish and others. Women have been raped and assaulted by men they met on the apps. The company declines to provide data on the total number of assaults reported to them, making it impossible to calculate any rate of incidence.\" ... \"CJI analyzed more than 150 incidents of sexual assault involving dating apps, culled from a decade of news reports, civil lawsuits and criminal records.\"\n","source_date":"2019-12-02","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505052518/https://www.propublica.org/article/tinder-lets-known-sex-offenders-use-the-app-its-not-the-only-one","calculation_notes":"ProPublica/CJI investigative reporting. Documents more than 150 analyzed incidents of sexual assault linked to dating apps generally (not solely Match Group products) and the absence of background checks, but explicitly confirms the absence of an aggregate incidence rate or denominator. Cited here to document the reporting landscape and the data gap, not to derive a rate.\n"},{"url":"https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/sexualoffencesinenglandandwalesoverview/yearendingmarch2025","title":"Sexual offences in England and Wales overview: year ending March 2025","publisher":"Office for National Statistics","source_type":"govt_report","statistic":"UK ONS records online sexual offences under broad Online Safety Act 2023 categories (~13,000 of the 209,079 police-recorded sexual offences) but does not disaggregate by platform or identify dating-app matches; no per-user-year denominator exists at any jurisdiction","excerpt":"\"A total of 209,079 sexual offences were recorded by the police in England and Wales in year ending (YE) March 2025, an increase of 11% (20,435 offences) compared with YE March 2024. The Online Safety Act 2023 introduced two new sexual offences: 'Sharing or threatening to share intimate photograph or film' and 'Sending etc photograph or film of genitals'. The two new sexual offences accounted for over half (around 13,000 offences) of the 11% increase recorded by the police between YE March 2024 and YE March 2025.\"\n","source_date":"2025-11-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260618045306/https://www.ons.gov.uk/peoplepopulationandcommunity/crimeandjustice/bulletins/sexualoffencesinenglandandwalesoverview/yearendingmarch2025","calculation_notes":"UK ONS Sexual Offences Overview, year ending March 2025. Confirms the data gap at a second jurisdiction: even where online sexual offences are newly recorded (the two Online Safety Act 2023 offences), the published police-recorded categories are not disaggregated by platform type and do not distinguish dating-app matches from other online contacts. The ~13,000/209,079 figures are numerators with no defensible platform-specific denominator. Cited as documentation of the measurement gap.\n"}],"comparison_anchors":[{"label":"Sexual assault lifetime (all women, general population)","lifetime_us_adult":0.2}],"short_label":"Dating app sexual violence","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry uses `no_reliable_estimate: true` — the genuine absence of a defensible denominator makes any rate fabrication misleading. The denominator problem is structural: no dating app publishes the total number of in-person meetings generated by matches, no government regulator collects platform-attributed sexual assault data with adequate coverage, and victim surveys do not consistently identify whether assaults occurred through app-facilitated meetings specifically. Pew harassment data (56% of women under 50 sent unwanted explicit content) describes online platform conduct, not physical violence from meetings. ProPublica documents individual cases and the absence of background checks, but cannot provide incidence rates. The risk is plausible and non-trivial given the millions of meetings generated by dating apps annually, but it is not quantifiable from available public data. The \"feature, not gap\" policy applies: honest acknowledgment of measurement limits is more informative than a fabricated rate.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a smartphone with a profile silhouette on screen, muted tones."},"canonical_url":"https://likelier.app/dating-app-violence","api_url":"https://likelier.app/api/fears/dating-app-violence.json"},{"slug":"deepfake-identity-fraud","question":"What are the odds of suffering financial loss or reputation harm from deepfake fraud?","quick_answer":"No reliable per-person odds exist: dedicated tracking only began in 2025, complaints are heavily underreported, and the denominator is unclear. Most people picture deepfake fraud as a problem for public figures rather than something that could target them.\n","category":"tech","tags":["digital-fraud"],"no_reliable_estimate":true,"perceived":{"description":"Most people associate deepfakes with celebrity face-swaps and political misinformation rather than personal financial risk. A 2024 McAfee survey found that roughly one in four adults had encountered an AI voice scam, yet few considered themselves likely targets of deepfake-enabled identity fraud. The technology feels cinematic — something that happens to CEOs tricked into wiring millions, not to ordinary individuals. Meanwhile the FBI's Internet Crime Complaint Center recorded 22,364 AI-related fraud complaints in 2025 with losses approaching $893 million, and that figure captures only the fraction of victims who file federal complaints. Experts at Sumsub report deepfakes now account for roughly 11% of global fraud attempts in 2026, up from 7% in 2024. The gap between public complacency and operational reality is widening every quarter.\n","rough_estimate":"Most adults consider deepfake fraud a problem for public figures, not themselves","kind":"intuition"},"sources":[{"url":"https://www.fbi.gov/news/press-releases/cryptocurrency-and-ai-scams-bilk-americans-of-billions","title":"Cryptocurrency and AI Scams Bilk Americans of Billions","publisher":"Federal Bureau of Investigation","source_type":"govt_report","statistic":"22,364 AI-related complaints in 2025 with losses of nearly $893 million; voice cloning and deepfake videos used in investment schemes and family-emergency impersonations","excerpt":"\"For the first time in its nearly 25-year history, the IC3 report features a section on artificial intelligence, which accounts for 22,364 complaints, costing Americans nearly $893 million.\"\n","source_date":"2026-04-23","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260417025144/https://www.fbi.gov/news/press-releases/cryptocurrency-and-ai-scams-bilk-americans-of-billions","calculation_notes":"The 2025 IC3 Annual Report is the first to break out AI-enabled fraud as a distinct category. The $893 million figure covers complaints where AI tools — including voice cloning, deepfake video, and AI-generated documents — were identified as enabling the fraud. Older adults accounted for $352 million of those losses. The FBI notes that actual losses are likely far higher because IC3 captures only a fraction of incidents (the FTC estimates IC3 complaint rates at roughly 10-15% of actual victimisation). Even with a conservative 5-10x underreporting multiplier, the annualised figure would be $4.5-9 billion — but the per-person probability remains impossible to pin down because (a) the denominator is unclear (US adults? global internet users?), (b) the technology is improving faster than annual reporting can track, and (c) 2025 is the first year with any dedicated tracking. This is why the entry uses no_reliable_estimate.\n"},{"url":"https://sumsub.com/blog/fraud-trends/","title":"Fraud Trends 2026: AI Scams, Deepfakes, and Emerging Threats","publisher":"Sumsub","source_type":"reputable_reference","statistic":"Deepfakes accounted for 7% of all fraud attempts globally in 2024, rising to 11% in 2026; a 4x increase in deepfake detections from 2023 to 2024","excerpt":"\"In 2026, AI scams are everywhere, with deepfakes now accounting for 11% of global fraudulent activity.\"\n","source_date":"2026-01-15","source_accessed":"2026-04-24","archive_url":"https://web.archive.org/web/20260426195512/https://sumsub.com/blog/fraud-trends/","calculation_notes":"Sumsub is an identity verification platform that processes millions of verification checks globally and publishes an annual fraud report based on its operational data. The 7% → 11% trajectory from 2024 to 2026 is directionally consistent with the FBI's first-time inclusion of an AI section in its 2025 report. The 4x year-over-year increase in deepfake detections (2023 to 2024) reflects both improved detection tooling and genuine growth in deepfake usage. The Sumsub data covers identity verification contexts (onboarding, KYC) rather than all fraud, so it likely understates the share of deepfakes in social-engineering scams (romance fraud, family-emergency calls) where no verification check is attempted. The rapid growth rate is precisely what makes a stable lifetime probability estimate unreliable — any number computed from 2025 data would be stale before publication.\n","independence_note":"Sumsub's fraud detection data is independently collected from its own verification platform, distinct from FBI IC3 complaint-based reporting. The two datasets measure different things (attempted identity fraud at verification vs reported financial losses) and corroborate each other directionally.\n"},{"url":"https://www.deloitte.com/us/en/insights/industry/financial-services/deepfake-banking-fraud-risk-on-the-rise.html","title":"Generative AI is expected to magnify the risk of deepfakes and other fraud in banking","publisher":"Deloitte Center for Financial Services","source_type":"reputable_reference","statistic":"Deloitte's 2024 projection: gen-AI-enabled fraud losses in the US could reach $40 billion by 2027, up from $12.3 billion in 2023 — a 32% compound annual growth rate","excerpt":"\"gen AI could enable fraud losses to reach US$40 billion in the United States by 2027, from US$12.3 billion in 2023, a compound annual growth rate of 32%.\"\n","source_date":"2024-05-29","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260620135739/https://www.deloitte.com/us/en/insights/industry/financial-services/deepfake-banking-fraud-risk-on-the-rise.html","calculation_notes":"Deloitte's Center for Financial Services built this projection by assigning generative-risk scores to 26 IC3 complaint categories and modelling forward growth; the conservative variant lands nearer $22 billion. It is a forward-looking industry projection, not an observed base rate, and it does not yield a per-person lifetime probability — it forecasts an aggregate US dollar loss. It is included here to bound the trajectory and corroborate the FBI's first-year nine-figure AI-fraud loss: an independent analyst forecasts roughly 3.25x growth ($12.3B to $40B) over four years, consistent with the steep curve that makes any stable per-person estimate premature. This reinforces, rather than resolves, the no_reliable_estimate flag.\n"},{"url":"https://www.secureworld.io/industry-news/ai-enabled-fraud-topped-893m-fbi","title":"FBI: AI-Enabled Fraud Topped $893M in 2025 — Real Toll Likely Far Higher","publisher":"SecureWorld","source_type":"news_article","statistic":"FBI acknowledges actual AI fraud losses likely far exceed $893M due to chronic underreporting; voice deepfakes used in job interview fraud cost victims $13 million in 2025","excerpt":"\"The FBI documented widespread use of voice spoofing and video deepfakes during online job interviews in 2025, with victims reporting losses of approximately $13 million.\"\n","source_date":"2026-04-24","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260425053154/https://www.secureworld.io/industry-news/ai-enabled-fraud-topped-893m-fbi","calculation_notes":"SecureWorld's analysis of the 2025 IC3 report highlights the underreporting problem that prevents reliable probability estimation. The $13 million job-interview-deepfake figure is a narrow subcategory illustrating how deepfake fraud extends beyond the stereotypical wire-transfer scam into employment, romance, and investment contexts. The article notes that the FBI itself considers reported figures a significant undercount. This corroborates the decision to use no_reliable_estimate: the signal is strong (deepfake fraud is large and growing), but the noise in the denominator (who is exposed? how often?) makes a per-person probability unreliable.\n"}],"comparison_anchors":[{"label":"Identity theft (US adult, annual)","lifetime_us_adult":0.3},{"label":"Online scam financial loss (US adult, lifetime)","lifetime_us_adult":0.25},{"label":"Credit card fraud (US adult, lifetime)","lifetime_us_adult":0.35}],"short_label":"Deepfake fraud","myth_framing":"underrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because deepfake fraud is an emerging and rapidly escalating threat for which stable base rates do not yet exist. The FBI's 2025 IC3 report is the first to track AI-enabled fraud as a category, recording 22,364 complaints and $893 million in losses — but IC3 captures perhaps 10-15% of actual incidents. The technology is improving on a quarterly cadence: voice cloning that required minutes of sample audio in 2023 now works with three seconds. Sumsub reports deepfakes rose from 7% to 11% of global fraud attempts between 2024 and 2026. Its November 2025 Identity Fraud Report (4M+ fraud attempts analysed) adds a sharper signal: \"sophisticated\" multi-step attacks that chain synthetic identities, deepfakes, and social engineering grew 180% year-over-year, rising from 10% of identity fraud in 2024 to 28% in 2025 — fraud is consolidating into coordinated, harder-to-detect campaigns. Deloitte's 2024 Center for Financial Services projection puts US gen-AI-enabled fraud losses on a path to $40 billion by 2027, up from $12.3 billion in 2023 (a 32% compound annual growth rate). These figures bound the lower edge of the threat without yielding a defensible per-person number: they are aggregate dollar totals and forward projections, not victimisation rates. Any lifetime probability computed from 2025 data would be anchored to a threat surface that will look unrecognisable within two years. The entry is tagged underrated because the public still perceives deepfakes primarily as a political misinformation tool rather than a personal financial threat, while operational data shows individual consumers — particularly older adults — are already bearing significant losses.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A simple silhouette of a face with a faint digital grid overlay, flat vector illustration."},"canonical_url":"https://likelier.app/deepfake-identity-fraud","api_url":"https://likelier.app/api/fears/deepfake-identity-fraud.json"},{"slug":"dirty-money-infection","question":"What are the odds of getting sick from handling cash or coins?","quick_answer":"No reliable estimate exists. Studies confirm bacteria live on banknotes and coins, but no cohort or case study has ever documented actual illness from handling currency, so the per-event infection rate is unmeasured. People often imagine a meaningful per-transaction risk.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The \"bacteria on your money\" story is a reliable content-farm staple, typically pegged to a study finding thousands of bacterial species on dollar bills or MRSA on hospital-adjacent banknotes. The implied conclusion is that handling cash exposes you to dangerous pathogens. In practice the fear conflates bioburden (what is living on a surface) with infection (what makes someone sick). No survey measures how US adults rate this risk numerically, but the cultural framing — amplified during the COVID-19 era when contactless payment was marketed as a hygiene intervention — suggests many people imagine a non-trivial per-transaction infection probability, on the order of 1-in-1,000 or higher.\n","rough_estimate":"People often imagine a meaningful per-transaction risk of infection, especially after COVID-era messaging about contactless payment","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20704502/","title":"Dirty Money: An Investigation into the Hygiene Status of Some of the World's Currencies as Obtained from Food Outlets","publisher":"Foodborne Pathogens and Disease (Vriesekoop, Russell, Alvarez-Mayorga et al.)","source_type":"primary_study","statistic":"1,280 banknotes from food outlets in 10 countries all carried viable bacteria; prevalence correlated with economic indicators and note material; pathogens isolated only after enrichment and deemed 'not alarming'","excerpt":"\"A total of 1,280 banknotes were obtained from food outlets in 10 different countries [...] there was a strong correlation between the number of bacteria per square centimeter and a series of indicators of economic prosperity [...] pathogens could only be isolated after enrichment and their mere presence does not appear to be alarming.\"\n","source_date":"2010-12-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260323060117/https://pubmed.ncbi.nlm.nih.gov/20704502/","calculation_notes":"Vriesekoop 2010 is the canonical multi-country bioburden study for currency. It establishes that all banknotes carry viable bacteria, that polymer notes carry fewer than cotton/linen notes, and that pathogens require enrichment to isolate — meaning the bacterial load is below the threshold for direct clinical concern. The study does not measure or claim any per-event infection rate from handling the sampled notes. This is the empirical basis for the \"bioburden is real, infection risk is unmeasured\" framing of this entry.\n","independence_note":"Primary multi-country sampling study; independent of Angelakis 2014 (which is a literature review, not a primary sampling effort) and of the copper-surface Grass 2011 work.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24571076/","title":"Paper money and coins as potential vectors of transmissible disease","publisher":"Future Microbiology (Angelakis, Azhar, Bibi, Raoult)","source_type":"peer_reviewed","statistic":"Systematic review finding S. aureus, E. coli, Salmonella, influenza, norovirus, rotavirus, and hepatitis A virus recoverable from currency; MRSA survives on coins; no documented clinical infection attributable to a currency-handling event","excerpt":"\"Paper currency and coins may be a public health risk when associated with the simultaneous handling of food and could lead to the spread of nosocomial infections. [...] Laboratory simulations revealed that methicillin-resistant S. aureus can easily survive on coins, whereas E. coli, Salmonella species and viruses, including human influenza virus, Norovirus, Rhinovirus, hepatitis A virus, and Rotavirus, can be transmitted through hand contact.\"\n","source_date":"2014-02-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260323060117/https://pubmed.ncbi.nlm.nih.gov/24571076/","calculation_notes":"Angelakis 2014 is the most comprehensive review of currency as a potential disease vector. The key word in their framing is \"potential\" — the review catalogues what organisms survive on currency and demonstrates hand-contact transfer in laboratory conditions, but does not cite a single documented case of clinical infection attributable to handling money. The highest-risk scenario they identify is simultaneous food and money handling, which is a hand-hygiene problem (see cross-reference to hand-hygiene-neglect), not a money problem per se.\n","independence_note":"Systematic review synthesising laboratory and observational studies; includes Vriesekoop 2010 as one input but draws on a much broader evidence base. The review's conclusion about absence of documented clinical infection is an independent editorial judgment.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/21193661/","title":"Metallic Copper as an Antimicrobial Surface","publisher":"Applied and Environmental Microbiology (Grass, Rensing, Solioz)","source_type":"peer_reviewed","statistic":"Copper surfaces kill bacteria at a rate of 7-8 log reductions per hour; no live microorganisms recovered after prolonged incubation on copper; EPA registered copper as the first solid antimicrobial material","excerpt":"\"Contact killing was observed to take place at a rate of at least 7 to 8 logs per hour [...] no live microorganisms were generally recovered from copper surfaces after prolonged incubation.\"\n","source_date":"2011-03-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260421184543/https://pubmed.ncbi.nlm.nih.gov/21193661/","calculation_notes":"Grass 2011 establishes the mechanism behind coins' lower bioburden relative to paper/polymer banknotes. Copper alloys (used in most world coinage — pennies, euro cents, pound coins) kill bacteria on contact at rates of 7-8 log reductions per hour, which means viable pathogen counts on coins drop to near-zero within minutes to hours of deposition. This is why Angelakis 2014 finds that coins, despite being handled as frequently as banknotes, carry substantially lower bacterial loads. The antimicrobial effect is intrinsic to the metal and does not require cleaning.\n","independence_note":"Materials-science study on copper antimicrobial properties; entirely independent of the currency bioburden and disease-vector literature. Included to explain the mechanism behind coins' lower pathogen loads.\n"},{"url":"https://fullfact.org/health/coronavirus-WHO-cash-comments/","title":"The WHO has clarified that they aren't warning people against using paper money due to coronavirus","publisher":"Full Fact (fact-check of WHO spokesperson statement)","source_type":"reputable_reference","statistic":"WHO confirmed in March 2020 that there was no evidence COVID-19 spreads via banknotes; recommended standard hand hygiene after handling money, not avoidance of cash","excerpt":"\"We know that money changes hands frequently and can pick up all sorts of bacteria and viruses [...] when possible it's a good idea to use contactless payments [...] Clean your hands regularly with an alcohol-based hand rub, or wash them with soap and water.\"\n","source_date":"2020-03-06","source_accessed":"2026-04-17","archive_url":"https://web.archive.org/web/20260420034929/https://fullfact.org/health/coronavirus-WHO-cash-comments/","calculation_notes":"The WHO statement is important for two reasons. First, it is the closest any major health authority came to issuing currency-specific guidance during the pandemic, and the guidance was \"wash your hands,\" not \"avoid cash.\" Second, the fact-check context shows that even during peak COVID anxiety, the WHO did not find sufficient evidence to warn against handling banknotes. The absence of a CDC or WHO guidance document specifically addressing infection risk from currency — in contrast to the detailed guidance both agencies maintain for food safety, hand hygiene, and fomite control — is itself evidence that the pathway does not rise to a public-health-action threshold.\n","independence_note":"Fact-check of a WHO spokesperson statement during the COVID-19 pandemic; independent of the academic bioburden literature. Included to document the official public-health position on currency handling.\n"}],"comparison_anchors":[{"label":"Infection from a shared towel, casual dry use (order of magnitude)","lifetime_us_adult":0.000001},{"label":"Infection from outdoor shoes indoors, healthy household (order of magnitude)","lifetime_us_adult":0.000001},{"label":"Food poisoning illness in a year (US)","lifetime_us_adult":0.14}],"regional_breakdown":[{"region":"Healthy adult, normal cash handling (structural upper bound)","probability":0.000001,"notes":"No documented case-report or epidemiological evidence of clinical infection attributable to handling currency in a healthy adult. Point estimate is a structural upper bound per transaction, not a measured rate. With typical hand-hygiene behaviour (washing before eating), even this bound is generous.\n"},{"region":"Immunocompromised adult handling soiled or wet banknotes","probability":0.0001,"notes":"Elevated theoretical risk due to compromised immune defences and the possibility of higher pathogen loads on visibly soiled or damp notes. Still no documented per-event rate; the estimate is an order-of-magnitude placeholder acknowledging biological plausibility without clinical evidence.\n"},{"region":"MRSA-positive banknotes in hospital-adjacent settings","probability":0.00001,"notes":"Angelakis 2014 found that hospital-recovered banknotes were highly contaminated with S. aureus. MRSA colonisation from fomite contact is documented in healthcare settings, but currency has not been isolated as an independent transmission route even in nosocomial outbreak investigations.\n"},{"region":"Food handler receiving cash then touching food without handwashing","probability":0.001,"notes":"The only scenario where currency handling plausibly contributes to infection, and even here the causal agent is the hand-hygiene gap, not the money itself. Vriesekoop 2010 explicitly recommended separating food and money handling. This is a hand-hygiene-neglect problem wearing a money costume.\n"},{"region":"Coins specifically (copper antimicrobial effect)","probability":1e-7,"notes":"Grass et al. 2011 demonstrated 7-8 log bacterial reductions per hour on copper surfaces. Most world coinage contains significant copper content. Viable pathogen survival time on coins is measured in minutes, not hours or days, making coin-mediated infection even less plausible than the banknote pathway.\n"}],"short_label":"Dirty money","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline applies to the typical scenario: a healthy adult handling ordinary circulating banknotes and coins, with normal hand-hygiene practices. The entire fear rests on a bioburden-to-infection conflation — studies consistently find bacteria on currency (Vriesekoop 2010: all 1,280 notes from 10 countries carried viable bacteria; Angelakis 2014: MRSA, E. coli, Salmonella, and multiple viruses recoverable), but no study has documented a clinical infection caused by a currency-handling event. The mechanism is biologically plausible; the outcome has never been observed. The one scenario where currency handling plausibly contributes to illness — a food handler receiving cash then touching food without washing hands — is better described as a hand-hygiene-neglect problem (see the hand-hygiene-neglect entry on this site) than a money problem. If you wash your hands before eating, the money-specific transmission pathway effectively vanishes. Coins pose even less risk than banknotes because copper alloys kill bacteria on contact (Grass et al. 2011: 7-8 log reductions per hour). During the COVID-19 pandemic, WHO explicitly declined to warn against handling cash, recommending hand hygiene instead. The absence of any CDC or WHO guidance document specifically addressing infection risk from currency — in a world where both agencies maintain detailed guidance for food safety, fomite control, and hand hygiene — is itself informative.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-10-agent","last_reviewed":"2026-04-17","reviewed":true,"generated_at":"2026-04-17","image":{"alt":"A single banknote resting flat on a muted background, rendered as a flat vector shape in soft grey-green and off-white tones."},"canonical_url":"https://likelier.app/dirty-money-infection","api_url":"https://likelier.app/api/fears/dirty-money-infection.json"},{"slug":"doctor-visit-waiting-room-infection","question":"What are the odds of catching something (or your kid catching something) from a doctor's waiting room?","quick_answer":"No reliable per-visit rate exists: the waiting room is full of viruses, but studies cannot isolate infections actually caused by the visit. The best prospective study found no detectable excess risk above the community baseline, despite parents often assuming roughly 1 in 5 visits leads to a new infection.\n","category":"health","tags":["kids"],"no_reliable_estimate":true,"perceived":{"description":"Almost every parent who has taken a feverish child to a pediatrician has eyed the other coughing children in the waiting room and concluded the visit is now a coin flip between fixing the original problem and inheriting a new one. Adults make the same calculation in primary care offices during flu season. Pediatric ED surveillance work confirms the waiting room is, on paper, biologically saturated: in one prospective study of a US pediatric ED, 62% of children and 17.5% of accompanying adults tested positive for at least one respiratory virus, often without symptoms. The intuition that the room is full of pathogens is correct. The intuition that this translates into a high per-visit transmission rate is the part the literature has consistently failed to confirm.\n","rough_estimate":"Most parents assume a meaningful share of waiting-room visits during respiratory virus season — perhaps 1 in 5 or higher — lead to a new infection traceable to the visit","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3022741/","title":"Do pediatric emergency departments pose a risk of infection?","publisher":"BMC Pediatrics (Quach, Moore, Ducharme, Chalut 2011)","source_type":"peer_reviewed","statistic":"Prospective cohort of 304 children: 21 of 137 (15.3%) with a recent pediatric ED visit developed a new infection in the 1-2 weeks afterward, vs 39 of 167 (23.4%) without recent ED exposure. Relative risk 0.7 (95% CI 0.4-1.1) — no detectable excess risk attributable to the visit.","excerpt":"\"Of 137 children with an ED visit in the previous 2 weeks, 21 (15.3%) developed a new infection compared with 39 of 167 (23.4%) children without a recent ED exposure (RR 0.7, 95% CI 0.4-1.1). A visit to a pediatric ED does not result in a detectable increased risk of infection above the risk in the community.\"\n","source_date":"2011-01-04","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250225231923/https://pmc.ncbi.nlm.nih.gov/articles/PMC3022741/","calculation_notes":"Quach et al. is the cleanest empirical answer to the parental fear this entry addresses. They followed 304 children prospectively and compared post-visit infection rates between those with a recent pediatric ED visit and those without. The exposed group's infection rate (15.3%) was lower than the unexposed group's (23.4%), giving a relative risk of 0.7 with a confidence interval that comfortably crosses 1. The authors are clear that the study cannot rule out a small effect (the upper CI bound is 1.1), but they also cannot detect one, and the headline direction of the point estimate is reassuring rather than alarming. The high baseline rate of childhood viral illness during the study period — 23% of unexposed children developed a new infection in any 2-week window — is the load- bearing fact: any ED-attributable signal has to compete with a very noisy community baseline, which is why no_reliable_estimate is appropriate here. The study was conducted in Montreal during respiratory virus season; the sample was small (n=304) and the follow-up window short, so this is a single under-powered cohort, not a settled question.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2939637/","title":"Potential for airborne transmission of infection in the waiting areas of healthcare premises: stochastic analysis using a Monte Carlo model","publisher":"BMC Infectious Diseases (Beggs, Shepherd, Kerr 2010)","source_type":"peer_reviewed","statistic":"Monte Carlo model of a typical 132 m³ waiting room with 4 air changes per hour and 19 susceptible occupants: a 30-minute exposure to a single infectious patient yields a mean transmission probability of 0.0034 for tuberculosis, 0.0262 for influenza, and 0.1349 for measles. At 60 minutes the figures rise to 0.0087, 0.0662, and 0.3094.","excerpt":"\"For a 30 minute waiting time, the mean probability of acquiring TB infection was negligible (0.0034), while that for influenza was 0.0262. The mean probability of acquiring measles was 0.1349. If the duration of stay increased to 60 minutes, these values increased to 0.0087, 0.0662 and 0.3094, respectively. The risk of acquiring TB infection during a visit to a hospital waiting area is minimal. Likewise the risks associated with influenza...are relatively small.\"\n","source_date":"2010-08-19","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250915195604/https://pmc.ncbi.nlm.nih.gov/articles/PMC2939637/","calculation_notes":"Beggs et al. is the standard modelled answer to \"what is the per-visit transmission probability if there is one infectious patient in the room with you for half an hour?\" The Wells-Riley framework they use treats airborne pathogen exposure as a function of quanta-generation rate (TB ~12.7/hr, influenza ~100/hr, measles ~570/hr), room volume, ventilation, and exposure time. The headline numbers are conditional on the presence of an infectious patient — they are not unconditional per-visit risks. They also assume a well-mixed room, average quanta rates from the literature, and no masking, vaccination, or behavioural distancing. As an upper bound under \"stranger in the next chair is shedding flu\" assumptions the influenza figure (~2.6% per 30 min) is the most directly relevant number to the lay framing of this fear. The disagreement with Quach's empirical zero-excess finding is most plausibly explained by the fact that most visits do not include 30 minutes of co-presence with an actively-shedding influenza or measles patient — the conditional risk is real but the unconditional per-visit risk is much smaller.\n","independence_note":"Methodologically and substantively independent of Quach 2011. Beggs is a pure stochastic airborne-transmission model; Quach is a real-world pediatric cohort. The two address different questions and arrive at different but compatible conclusions.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7767945/","title":"Respiratory viruses in pediatric emergency department patients and their family members","publisher":"Influenza and Other Respiratory Viruses (Matienzo et al. 2020)","source_type":"peer_reviewed","statistic":"In a prospective surveillance study of a New York City pediatric ED across two respiratory virus seasons, 399 of 641 children (62.25%) and 118 of 674 accompanying adults (17.5%) tested positive for one or more respiratory viruses. Of 72 accompanying adults who tested positive for the same virus as their child, 26 (36%) were asymptomatic at the time of the visit.","excerpt":"\"Respiratory viruses were detected in 399 children (62.25%) and 118 (17.5%) accompanying adults.\" \"Of the 72 accompanying adults who tested positive for the same virus as their child, 46 were symptomatic (63.88%).\"\n","source_date":"2020-11-22","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250409103056/https://pmc.ncbi.nlm.nih.gov/articles/PMC7767945/","calculation_notes":"Matienzo et al. quantifies the input side of the waiting-room transmission question — how full of pathogens the room actually is. The headline 62% child-positivity figure includes the symptomatic patients who came in for a respiratory complaint, so it is an upper-bound surveillance figure rather than a sample of all pediatric ED visits. The 17.5% positivity in accompanying adults — most of whom were not the index patient — is the more striking number for the waiting-room scenario, since it confirms that a meaningful fraction of the people sitting next to you are shedding virus without realising it. This pairs with Quach's null finding to give a coherent picture: the room is biologically saturated, but the per-visit attributable infection probability is still small because exposure time is short, most pathogens are inefficient transmitters per 30 minutes of shared air, and patients with the same virus often share households (so \"caught it at the doctor\" is often \"had it already, just didn't know\"). No probability is derived from this source — it is included to ground the qualitative claim that the worry is based on a real exposure, not a hallucinated one.\n","independence_note":"Independent of both Quach (different decade, different city, different pathogen panel, surveillance design rather than cohort) and Beggs (empirical surveillance vs stochastic model). Establishes the prevalence premise that the other two sources interpret.\n"}],"comparison_anchors":[],"regional_breakdown":[{"region":"Per pediatric ED visit (Quach cohort)","probability":0,"notes":"Empirical excess risk above community baseline statistically indistinguishable from zero. Point-estimate relative risk 0.7 (95% CI 0.4-1.1) over 304 children. The 15.3% absolute infection rate in the exposed group is essentially the same as the 23.4% baseline rate in unexposed children — both reflect the high background incidence of childhood viral illness during respiratory virus season, not the visit itself.\n"},{"region":"30-min waiting-room exposure to an influenza patient (Beggs model)","probability":0.0262,"notes":"Conditional on one infectious patient being present in a typical 132 m³ waiting room with 4 air changes per hour. Roughly 1 in 38 per half-hour of shared air. Increases to ~1 in 15 (0.0662) for a 60-min exposure.\n"},{"region":"30-min waiting-room exposure to a measles patient (Beggs model)","probability":0.1349,"notes":"The same model applied to measles, which has a quanta-production rate roughly six times higher than influenza. This is the outlier scenario that drove the redesign of waiting rooms during the 2014-2019 measles resurgence — a single unvaccinated measles patient in a waiting room can plausibly infect 30%+ of susceptible adults present for an hour.\n"},{"region":"30-min waiting-room exposure to a TB patient (Beggs model)","probability":0.0034,"notes":"The same model applied to TB, which is much less efficient per unit time than influenza or measles. Negligible per-visit risk even under conditional assumptions.\n"}],"short_label":"Waiting room infection","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the literature does not support a single defensible per-visit probability number.\nFirst, the empirical and modelled estimates disagree by an order of magnitude. Quach et al. (2011) followed 304 children and found no detectable excess infection rate after a pediatric ED visit (RR 0.7, 95% CI 0.4-1.1) — the upper confidence bound just touches 1.1, so a small effect cannot be ruled out, but no positive effect could be detected. Beggs et al. (2010) modelled a 30-minute waiting-room exposure to a single infectious flu patient and produced a transmission probability of ~2.6%. The two answers the studies actually give are different because they answer different questions: Quach measures the unconditional per-visit risk averaged across all visits (most of which do not include 30 minutes of co-presence with an actively shedding patient), while Beggs measures the conditional per-visit risk given that such a patient is present.\nSecond, even within the conditional framing the answer depends heavily on variables outside the parent's control: which pathogen the index patient is shedding, how long the wait is, how well-ventilated the room is, whether the index patient is masked, whether the visit was during respiratory virus season, and whether the susceptible patient is vaccinated. Measles in an unventilated room is genuinely high-risk per 30 minutes of exposure (~13.5%). TB in the same room is essentially zero (~0.3%). A flu-season pediatric office and an off-season dermatology clinic are not the same exposure environment.\nThird, even the surveillance-side numbers are easy to misread. Matienzo et al. (2020) found that 62% of children visiting the pediatric ED and 17.5% of their accompanying adults tested positive for a respiratory virus — numbers that sound terrifying for the waiting room. But the 62% figure is inflated by the fact that respiratory complaints are the reason most children come to a pediatric ED in the first place, and the 17.5% adult figure includes asymptomatic shedders who were probably already going to give their child whatever they had, whether or not the family ever entered the building.\nFourth, the post-2020 outpatient infection-control environment is genuinely different from the pre-pandemic literature. Many primary care offices now triage respiratory complaints by phone, hold sick children in their car until the exam room is open, mask staff and patients during community respiratory surges, and improve ventilation — all of which mean the pre-2020 transmission probabilities are upper bounds for the modern visit.\nThe honest summary: the risk is non-zero but small, almost certainly an order of magnitude lower than the high community baseline of childhood viral illness, and dominated by a small number of high-shedding pathogens (measles, RSV during peak season, COVID during community surges) in poorly ventilated waiting rooms. The pre-visit phone-call triage and the wait-in- the-car option are the two interventions that move the needle; the unconditional per-visit fear that \"we will come home with something\" is overrated.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"An empty clinic waiting room with two rows of chairs and a small side table, rendered as a flat vector on a muted background."},"canonical_url":"https://likelier.app/doctor-visit-waiting-room-infection","api_url":"https://likelier.app/api/fears/doctor-visit-waiting-room-infection.json"},{"slug":"doorway-jumper-injury","question":"What are the odds an infant in a door-frame jumper is injured by clamp slip, doorframe impact, or pinch?","quick_answer":"No defensible rate exists. Recalls document real injury mechanisms — clamp slip, doorframe impact, pinch — but there is no published count of how many infants use these jumpers, and injury data bundles them with walkers and stationary exercisers, so neither a numerator nor a denominator can be isolated.\n","category":"kids","tags":["infant","household","kids"],"no_reliable_estimate":true,"perceived":{"description":"Parents see door-frame jumpers as a low-risk indoor entertainment device. The visible elastic strap and the active bouncing motion make the product feel sturdy, and the infant's evident enjoyment reinforces the impression that the design is well-matched to the use case. Few parents have a mental model for clamp slip, doorframe impact, or pinch hazards — the three mechanisms documented in CPSC recall actions for this category. The honest answer is that no defensible rate exists: the denominator (how many US or EU infants use door-frame jumpers in a given year) has no published estimate, and NEISS bundles jumper injuries with walker and stationary-exerciser injuries into a single product code, so the numerator cannot be separated from those companion products either.\n","rough_estimate":"Mechanism evidence only; no population rate exists","kind":"intuition"},"sources":[{"url":"https://www.cpsc.gov/Recalls/2005/cpsc-kids-ii-inc-announce-recall-of-doorway-baby-jumpers","title":"CPSC, Kids II Inc. Announce Recall of Doorway Baby Jumpers","publisher":"US Consumer Product Safety Commission","source_type":"govt_report","statistic":"Recall covering clamp failure mechanism; nine clamp-break reports, three injuries; no aggregate jumper-injury rate published","excerpt":"\"The plastic clamp that attaches the jumper seat to a door frame can break, which can cause the unit and child to fall to the floor. Kids II has received nine reports of the clamp breaking with three reports of injuries, including a bruised forehead, chin and foot.\"\n","source_date":"2005-08-23","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260531135658/https://www.cpsc.gov/Recalls/2005/cpsc-kids-ii-inc-announce-recall-of-doorway-baby-jumpers","calculation_notes":"Mechanism-illustrative recall. No population-level rate for doorway- jumper injuries exists because NEISS bundles them with walkers and stationary exercisers into a single product code, and CPSC's nursery- products injury reports aggregate at category level. This recall documents the clamp-slip failure mode but does not provide a denominator.\n"},{"url":"https://www.en-standard.eu/bs-en-16232-2013-a2-2023-child-use-and-care-articles-infant-swings/","title":"BS EN 16232:2013+A2:2023 — Child use and care articles — Infant swings","publisher":"British Standards Institution / European Committee for Standardization","source_type":"reputable_reference","statistic":"BS EN 16232:2013+A2:2023 (published 29 September 2023) specifies safety requirements for infant swings intended for children up to 9 kg or unable to sit up unaided — swings, not doorway jumpers","excerpt":"\"This European Standard specifies safety requirements and the corresponding test methods for infant swings intended for children up to a weight of 9 kg or unable to sit up unaided.\"\n","source_date":"2023-09-29","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20231227083247/https://www.en-standard.eu/bs-en-16232-2013-a2-2023-child-use-and-care-articles-infant-swings/","calculation_notes":"Included as a nearby product-safety reference: EN 16232 is the EU/Polish standard for infant SWINGS (children up to 9 kg or unable to sit up unaided), not for doorway jumpers, which fall outside its scope. It is cited to show that the adjacent oscillating-baby-product category is standardised, not as a governing standard for doorway jumpers themselves. It supplies no injury rate.\n"}],"comparison_anchors":[],"short_label":"Doorway jumper injury","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"Doorway jumpers occupy a regulatory blind spot. NEISS bundles them with walkers and stationary jumpers into a single product code, so the population-level injury rate cannot be separated from those companion products. CPSC's 2005 Kids II recall documents the clamp-slip mechanism but no aggregate rate. Mechanism literature is thin: dominant hazards are clamp failure (causing fall onto floor), doorframe impact during high- amplitude bouncing, and pinch injuries between strap and clamp. Until the surveillance system disaggregates jumpers from the broader category, a defensible numerical estimate is not possible.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-8d","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"A doorway jumper hanging from a door frame in a residential interior, no child, viewed from a low angle, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/doorway-jumper-injury","api_url":"https://likelier.app/api/fears/doorway-jumper-injury.json"},{"slug":"driving-after-cannabis","question":"What are the odds of causing a fatal crash by driving within a few hours of using cannabis?","quick_answer":"No reliable absolute odds exist. The evidence gives only relative crash-culpability multipliers — risk stays near baseline at low blood THC and roughly doubles at moderate levels — not any per-trip or lifetime crash probability. Most regular users believe driving stoned is far safer than driving drunk.\n","category":"transport","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"Most regular cannabis users believe driving stoned is meaningfully safer than driving drunk, and a substantial minority believe it is safe or even improves their driving. Self-report surveys in legal-cannabis US states find that roughly 30-50% of past-month cannabis users have driven within two hours of use, and most of those describe themselves as cautious or unaffected. The subjective experience of acute cannabis impairment is qualitatively different from alcohol: drivers typically feel slower, more focused, and self-correctingly conservative — they drive more carefully on the straightaways and miss the high-attention edge cases (lane departures, sudden brake events, peripheral movement) where the per-trip crash risk actually lives.\n","rough_estimate":"most regular cannabis users believe driving stoned is far safer than driving drunk","kind":"intuition"},"sources":[{"url":"https://journals.sagepub.com/doi/10.1177/20503245251323344","title":"Dose-response relationship between blood concentrations of THC and crash culpability risk: An updated meta-regression of culpability studies","publisher":"Albrecht, Hasan, Kekez, Zhou — Drug Science, Policy and Law","source_type":"peer_reviewed","statistic":"Crash culpability risk increases with rising whole-blood THC concentration, with an inflection around 1.5-3.0 ng/mL where risk begins to rise above baseline; risk approximately doubles at ~5 ng/mL and approximately quadruples at ~10 ng/mL. Below ~1.5 ng/mL culpability is statistically indistinguishable from baseline (<30% increase).\n","excerpt":"\"Crash culpability risk increases with increasing THC concentration, with an inflection around 1.5-3.0 ng/ml where risk begins to increase... a doubling of culpability risk around 5 ng/ml and a potential quadrupling of risk around 10 ng/ml.\"\n","source_date":"2025-03-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20250407122737/https://journals.sagepub.com/doi/10.1177/20503245251323344","calculation_notes":"Albrecht 2025 is the most recent dose-response meta-regression and the cleanest source for translating a blood THC concentration into a per-trip crash-culpability RELATIVE multiplier. It reports the doubling at ~5 ng/mL whole blood (verbatim: \"a doubling of culpability risk around 5 ng/ml\") for concentrations consistent with \"driving within ~2 hours of typical inhaled cannabis use\" and a quadrupling (\"4.0x, 95% CI 3.0-5.5\") at ~10 ng/mL. These are relative culpability ratios versus baseline; the paper reports NO absolute per-trip or lifetime crash probability and NO sober-driver per-trip baseline rate, so no absolute headline number can be derived from it (see no_reliable_estimate).\n"},{"url":"https://onlinelibrary.wiley.com/doi/abs/10.1111/add.13347","title":"The effects of cannabis intoxication on motor vehicle collision revisited and revised","publisher":"Rogeberg, O. & Elvik, R. — Addiction 111(8):1348-1359","source_type":"peer_reviewed","statistic":"Pooled odds ratio across 28 estimates from 21 observational studies for acute cannabis use and motor-vehicle collision involvement was 1.36 (95% CI 1.15-1.61); roughly half of earlier higher estimates from the Asbridge 2012 BMJ meta-analysis disappear after correcting for confounding and methodological inconsistencies.\n","excerpt":"\"Our updated meta-analysis suggests that the increase in crash risk caused by cannabis intoxication is moderate, around 20-30%, much smaller than that of drink-driving and similar in magnitude to that of driving with a blood alcohol concentration between 0.01% and 0.05%.\"\n","source_date":"2016-08-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20230606233509/https://onlinelibrary.wiley.com/doi/abs/10.1111/add.13347","calculation_notes":"Rogeberg & Elvik 2016 (DOI 10.1111/add.13347, PMID 26878835) is the most methodologically careful pooled estimate available. It used 28 estimates from 21 observational studies (case-control and culpability designs) and explicitly corrected for the confounding by age, sex, and time-of-day that the Asbridge 2012 BMJ meta-analysis partially missed. The 1.36 odds ratio is the low-end pooled RELATIVE multiplier; it represents the \"any cannabis-positive driver\" average, dominated by drivers with residual THC from prior days rather than acute peak concentrations. Albrecht's 5 ng/mL doubling is the appropriate reference for acute post-use trips. Both are relative risk ratios; neither this study nor any other cited source provides the absolute sober per-trip fatal-crash baseline that would be needed to turn a relative multiplier into an absolute lifetime probability.\n"},{"url":"https://www.nhtsa.gov/staticfiles/nti/pdf/812117-Drug_and_Alcohol_Crash_Risk.pdf","title":"Drug and Alcohol Crash Risk (Research Note, DOT HS 812 117)","publisher":"Compton, R.P. & Berning, A. — National Highway Traffic Safety Administration","source_type":"govt_report","statistic":"The unadjusted odds ratio for crash involvement among THC-positive drivers in the Virginia Beach case-control study was 1.25; after adjustment for age, sex, race/ethnicity, and alcohol concentration the odds ratio dropped to 1.00 (95% CI 0.77-1.31), indicating no statistically significant elevation. The same study found 0.08% BAC associated with adjusted OR of 3.93 and 0.15% BAC with adjusted OR of 12.04.\n","excerpt":"\"After adjustment for age, gender, race/ethnicity, and alcohol use, the odds of being a crash-involved driver was not statistically different from that of a control driver who tested positive for THC (OR=1.00, 95% CI 0.77-1.31).\"\n","source_date":"2015-02-01","source_accessed":"2026-05-25","archive_url":"http://web.archive.org/web/20230111133721/https://www.nhtsa.gov/staticfiles/nti/pdf/812117-Drug_and_Alcohol_Crash_Risk.pdf","calculation_notes":"The 2015 NHTSA Virginia Beach study is the largest US case-control study of drug-positive driving and is frequently cited as evidence that cannabis-impaired driving carries no elevated risk after controlling for confounders. Methodologists (Rogeberg, Albrecht, Compton) note that the null result reflects the dominance of residual-THC-positive drivers in the cannabis-positive group; separating acute-use from residual-positive cases (which the blood-concentration meta-regressions do) recovers a measurable dose-response signal. Used here as the lower bound on the \"any THC positive\" risk multiplier and as evidence that the headline number depends strongly on how the impaired population is defined.\n"},{"url":"https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0187320","title":"Cannabis, alcohol and fatal road accidents","publisher":"Martin, J.-L. et al. — PLOS ONE 12(11):e0187320 (VOIESUR project)","source_type":"peer_reviewed","statistic":"In a French population-based case-control analysis of 4,059 drivers involved in 2,870 fatal accidents (VOIESUR, 2011), the adjusted odds ratio of being responsible for a fatal crash for a driver under the influence of cannabis alone was 1.65 (95% CI 1.16-2.34). No first-order interaction between alcohol and cannabis was significant (alcohol x cannabis p = 0.29), so the combined-substance odds ratio is estimated as the product of the individual ORs (multiplicative, not synergistic).\n","excerpt":"\"Drivers under the influence of cannabis multiply their risk of being responsible for causing a fatal accident by 1.65 (1.16-2.34)... All first-order interactions were tested. None were significant. In particular, we found no significant interaction of alcohol x cannabis (p = 0.29)... the odds ratio for a driver under the influence of both alcohol and cannabis can be estimated by the product of the OR relative to alcohol and the OR relative to cannabis.\"\n","source_date":"2017-11-08","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20211230000000*/journals.plos.org/plosone/article?id=10.1371/journal.pone.0187320","calculation_notes":"Martin et al. 2017 is a peer-reviewed national case-control study of crash responsibility, adjusted for age, gender, vehicle category, and time of accident. Its adjusted OR of 1.65 for cannabis alone sits between the Rogeberg & Elvik 2016 pooled 1.36 and the Albrecht 2025 acute-dose 2× doubling, illustrating the spread of relative crash-responsibility multipliers across studies — it is a population-average per-trip crash-responsibility multiplier vs sober, not a sub-group factor. Its no-interaction finding refines the combined cannabis+alcohol framing: the joint risk is multiplicative (product of the two ORs) rather than additive or super-additive.\n"},{"url":"https://injepijournal.biomedcentral.com/articles/10.1186/s40621-024-00498-1","title":"State cannabis laws and cannabis positivity among fatally injured drivers","publisher":"Leavitt, T.C., Chihuri, S., Li, G. — Injury Epidemiology 11:18","source_type":"peer_reviewed","statistic":"Using 2019 and 2020 Fatality Analysis Reporting System (FARS) toxicology data for 14,079 fatally injured US drivers, 4,702 (33.4%) tested positive for cannabis. Cannabis positivity was a recurring feature of fatal-crash drivers regardless of state legalization status.\n","excerpt":"\"Of the 14,079 fatally injured drivers, 4702 (33.4%) tested positive for cannabis.\"\n","source_date":"2024-04-11","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20240601000000*/injepijournal.biomedcentral.com/articles/10.1186/s40621-024-00498-1","calculation_notes":"Leavitt et al. 2024 is the largest recent US prevalence figure for cannabis among fatally injured drivers. It is a prevalence (exposure) statistic, not a per-trip risk multiplier: a fatally-injured sample is selection-biased toward crash outcomes, so it cannot be used to derive an absolute per-trip or lifetime probability. It is cited here to refresh the evidence base on how common cannabis-positive driving is among the population that ends up in fatal crashes.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Causing a fatal crash while drowsy (~monthly, lifetime)","lifetime_us_adult":0.038}],"short_label":"Driving after cannabis","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"No cited source provides an absolute per-trip or lifetime fatal-crash probability for a cannabis-impaired driver, so this entry carries no headline number. Every source here reports a RELATIVE risk ratio (odds ratio) or a prevalence, not an absolute per-trip denominator: Albrecht 2025 gives a dose-response culpability curve (doubling near ~5 ng/mL, 4.0x at ~10 ng/mL), Rogeberg & Elvik 2016 a pooled OR of 1.36 (95% CI 1.15-1.61), Martin 2017 (VOIESUR) an adjusted crash-responsibility OR of 1.65 (95% CI 1.16-2.34), and NHTSA's 2015 Virginia Beach study an adjusted OR of 1.00 (95% CI 0.77-1.31). Turning any of these relative multipliers into an absolute lifetime probability would require an unsourced sober-driver per-trip fatal-crash baseline and an unsourced trip-frequency assumption, so the site declines to fabricate one — the same reason the companion 0.10% BAC entry carries no headline number. The evidence base is also unusually spread because the estimates depend heavily on how the impaired population is defined. \"THC-positive\" includes drivers who used cannabis days earlier and retain detectable but non-impairing residual concentrations; \"acute use\" means within roughly 1-3 hours of inhalation or 2-4 hours of edible peak. Pooled \"any positive\" ORs (Rogeberg 1.36, Compton 1.00 adjusted) are dominated by the residual group and understate acute-use risk, while dose-response meta-regressions (Albrecht 2025) isolate the acute-use signal and find substantially higher relative risk at concentrations consistent with recent inhalation. The 2024-2025 evidence refreshes the exposure picture without supplying an absolute per-trip risk: a 2024 peer-reviewed FARS analysis (Leavitt, Chihuri & Li) found 33.4% of 14,079 fatally injured US drivers tested positive for cannabis, and a 2025 American College of Surgeons conference abstract reported 41.9% of 246 deceased drivers in one Ohio county positive for active THC at a mean 30.7 ng/mL. Both are prevalence figures from fatally-injured (selection-biased) samples and the ACS figure is not yet peer reviewed, so neither can produce a lifetime estimate. The French VOIESUR case-control study (Martin 2017) also found no significant alcohol x cannabis interaction, implying the combined risk is multiplicative (the product of the two ORs) rather than additive or super-additive, a more conservative reading of the combined-use factor than older culpability studies suggested.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"fears-8d-v1"},"reviewer":"8d-eval-independent-2026-05-25","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-25","image":{"alt":"A muted flat vector illustration of a single car steering wheel with a small leaf shape outline beside it, on a pale background."},"canonical_url":"https://likelier.app/driving-after-cannabis","api_url":"https://likelier.app/api/fears/driving-after-cannabis.json"},{"slug":"driving-at-0.1pct-bac","question":"What are the odds of causing a fatal crash when driving with a 0.10% blood alcohol level?","quick_answer":"No reliable absolute odds exist. Crash risk climbs sharply once blood alcohol passes 0.10%, but no cited study pins an exact per-trip fatal-crash probability at that level. Most drivers at 0.10% genuinely believe they are fine.\n","category":"transport","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"Most people who drive at a 0.10% BAC genuinely believe they are fine. The subjective experience of impairment at that level is mild for regular drinkers: some verbal looseness, slightly slowed reaction time, a sense of calm focus. Public campaigns have successfully attached \"drunk driving\" to the image of visibly staggering impairment, which means a driver at 0.10% typically does not recognize themselves in that framing. Informal surveys consistently find that most adults believe the real danger threshold is somewhere around 0.15% or higher.\n","rough_estimate":"most drivers at this BAC believe they are fine to drive","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/19778652/","title":"The Long Beach/Fort Lauderdale relative risk study","publisher":"Blomberg, Peck, Moskowitz, Burns, Fiorentino — Journal of Safety Research","source_type":"primary_study","statistic":"Adjusted crash relative risk rises measurably starting at 0.04-0.05% BAC and becomes very pronounced once BAC exceeds 0.10%. The study does not report a single point relative-risk figure at exactly 0.10% BAC; it describes the shape of the risk curve rather than pinning a value to that threshold.\n","excerpt":"\"When adjusted for covariates and nonparticipation bias, increases in relative risk were observed at BACs of .04–.05, and the elevations in risk became very pronounced when BACs exceeded .10.\"\n","source_date":"2009-10-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505053358/https://pubmed.ncbi.nlm.nih.gov/19778652/","calculation_notes":"Blomberg 2005 (published in the Journal of Safety Research in 2009) is the gold-standard replication of the 1964 Borkenstein Grand Rapids Study. It used a case-control design with 2,871 crashes in Long Beach and Fort Lauderdale, matched controls at the same time/location, and adjusted for age, sex, marital status, drinking frequency, and ethnicity. The abstract states only that relative risk becomes \"very pronounced\" above 0.10% BAC; it does not publish an exact adjusted RR at 0.10%. Widely-cited secondary summaries put the adjusted RR at 0.08% BAC at 2.69x. Because no cited source states a fatal-crash-specific per-trip probability at 0.10% BAC, this entry does not derive a headline lifetime number (no_reliable_estimate).\n"},{"url":"https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813713","title":"Traffic Safety Facts 2023 Data: Alcohol-Impaired Driving (DOT HS 813 713)","publisher":"National Highway Traffic Safety Administration (NHTSA), National Center for Statistics and Analysis","source_type":"govt_report","statistic":"12,429 people killed in alcohol-impaired-driving crashes in 2023 (30% of all 40,901 traffic fatalities); 67% of those fatalities involved a driver with BAC of 0.15 or higher; one alcohol-impaired-driving fatality every 42 minutes on average.\n","excerpt":"\"Of the 40,901 traffic fatalities in 2023, an estimated 12,429 people (30%) were killed in alcohol-impaired-driving crashes.\"\n","source_date":"2024-12-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260425162501/https://crashstats.nhtsa.dot.gov/Api/Public/Publication/813713","calculation_notes":"NHTSA FARS is the definitive US source for alcohol-impaired driving fatality counts. The 12,429 figure covers all crashes where any driver had BAC ≥ 0.08%, so it includes drivers from 0.08% up through very high BAC levels. The 67% figure (8,272 fatalities) involving BAC ≥ 0.15% highlights that the most severe outcomes cluster at higher BAC levels, but the 0.10-0.15% range still accounts for a meaningful share of the remaining 33%. Used here to anchor the population-level denominator for annual impaired-driving deaths.\n"},{"url":"https://rosap.ntl.bts.gov/view/dot/1779/dot_1779_DS1.pdf","title":"Drinking and Driving Trips, Stops by the Police, and Arrests","publisher":"National Transportation Library / NHTSA","source_type":"govt_report","statistic":"Probability of arrest per DWI trip at BAC ≥ 0.10% is approximately 1 in 200 (0.005); roughly 1 in 625 DWI trips resulted in a crash of any severity; used to calibrate per-trip exposure-to-outcome rates.\n","excerpt":"\"The probability of arrest while driving at a blood alcohol level over 0.10% was 0.0058.\"\n","source_date":"1998-01-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260504192713/https://rosap.ntl.bts.gov/view/dot/1779/dot_1779_DS1.pdf","calculation_notes":"This NHTSA-funded report provides per-trip arrest and crash probabilities for impaired driving. It reports an arrest probability of 0.0058 per DWI trip at BAC over 0.10%, and roughly 1 in 625 DWI trips (all BAC levels ≥ 0.08%) resulting in a crash of any severity. These are any-severity and pooled-BAC figures; the report does not isolate a fatal-crash-specific per-trip rate at 0.10% BAC. Converting the any-severity rate to a fatal-specific 0.10% rate would require an unsourced fatal-fraction assumption, so no headline probability is derived from it here.\n"}],"comparison_anchors":[{"label":"Death in a car crash (lifetime, US adult, population average)","lifetime_us_adult":0.0108},{"label":"Causing a fatal crash while texting (regular texter, lifetime)","lifetime_us_adult":0.018},{"label":"Death from a motorcycle crash (lifetime, US adult)","lifetime_us_adult":0.00144}],"short_label":"Driving at 0.10% BAC","myth_framing":"underrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"No cited source states a fatal-crash-specific probability — per trip or per lifetime — for a driver at 0.10% BAC, so this entry carries no headline number. Blomberg 2005 describes the shape of the all-severity crash-risk curve (rising from 0.04-0.05% and \"very pronounced\" above 0.10%) but publishes no point relative risk at 0.10%; NHTSA's per-trip DWI data are any-severity and pooled across all BAC levels ≥ 0.08%. Producing a lifetime figure would require two unsourced assumptions — how frequently a driver operates at this BAC and what fraction of impaired crashes are fatal at 0.10% specifically — so the site declines to fabricate one. What the evidence does support is directional: the 0.10% BAC is above the US legal limit of 0.08% and sits in a range of clearly elevated risk, even though subjective impairment is often mild for regular drinkers, which is precisely why calibration fails. The 67% of alcohol fatalities involving BAC ≥ 0.15% (NHTSA 2023) shows the very high end of the distribution drives most deaths, but the 0.10% range is not safe — it is merely less catastrophic than 0.20%+. Legal consequences (arrest probability ~0.0058, roughly 1 in 170 per trip at this BAC per the 1998 NHTSA report) are a separate risk channel entirely.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A muted flat vector illustration of a car's steering wheel with a small red bar marker slightly above a center reference line, on a pale background."},"canonical_url":"https://likelier.app/driving-at-0.1pct-bac","api_url":"https://likelier.app/api/fears/driving-at-0.1pct-bac.json"},{"slug":"eating-moldy-food","question":"What are the odds of getting sick from accidentally eating moldy food or drinking from a cup that had mold on it?","quick_answer":"There is no reliable number here. Most illness blamed on a bite of moldy food actually traces to bacteria growing alongside the mold rather than the mold itself, and the mycotoxin cancer data concerns cumulative low-level exposure across the food supply, not a single visible bite, so a defensible per-incident risk cannot be stated.\n","category":"food","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"Visible mold on a slice of bread, a forgotten container of leftovers, or a mug that sat with cold coffee in it for a week reads to most people as an acute poisoning hazard — the kind of thing where one bite could put you in the hospital. The fear typically lumps together three quite different mechanisms (immediate gastrointestinal illness from the mold itself, cumulative cancer risk from mycotoxins, and invasive fungal infection) and treats them all as if they apply to a healthy adult who accidentally bit a corner of mouldy bread or drank from a previously-mouldy cup. The question is what happens when the *mold itself* is isolated from bacterial co-contamination, temperature abuse, and the kind of long-term commercial-supply exposure that drives mycotoxin epidemiology.\n","rough_estimate":"Many adults treat one bite of visible mold as a near-certain food poisoning event, partly because the visual cue is so vivid and partly because clinical messaging conflates 'discard the food' with 'eating it would have hurt you'.","kind":"intuition"},"sources":[{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/molds-food-are-they-dangerous","title":"Molds on Food: Are They Dangerous?","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"USDA FSIS states that some molds cause allergic reactions and respiratory problems, a few molds produce mycotoxins, and that foods with mold may also have invisible bacteria; the discard-vs-trim rule splits porous/high-moisture foods (discard entirely) from dense/hard foods (cut 1 inch around and below)","excerpt":"\"[Paraphrase from official FSIS fact sheet — direct fetch blocked by bot protection 2026-05-27; content corroborated by USDA FSIS fact-sheet PDF mirror and multiple secondary outlets quoting it verbatim] 'Some molds cause allergic reactions and respiratory problems. A few molds, in the right conditions, produce mycotoxins, poisonous substances that can make people sick. [...] Foods that are moldy may also have invisible bacteria growing along with the mold.'\"\n","source_date":"2013-08-09","source_accessed":"2026-05-27","archive_url":"http://web.archive.org/web/20260527063537/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/molds-food-are-they-dangerous","calculation_notes":"This is the canonical US federal statement on the actual mechanism of harm from moldy food. Three points matter for the no_reliable_estimate framing: (1) the harm is conditional (\"a few molds, in the right conditions\") rather than a property of mold-in-general; (2) the visible-bacteria-too caveat is the reason most \"I got sick from moldy food\" anecdotes are not actually attributable to the mold itself; (3) the discard rule is precautionary, not a per-event illness rate. FSIS does not report a population-level rate of moldy-food illness because the surveillance system (FoodNet) tracks pathogen-specific foodborne illness — Salmonella, Campylobacter, Listeria, E. coli, etc. — not mold-attributable acute episodes, which are too rare to appear in annual surveillance bulletins.\n","independence_note":"FSIS is the upstream federal authority on the question; FDA, CDC and state health departments defer to its guidance on consumer mold handling. Not independent confirmation, but the primary source.\n"},{"url":"https://health.clevelandclinic.org/what-happens-if-you-eat-moldy-food","title":"What Happens If You Accidentally Eat Moldy Food?","publisher":"Cleveland Clinic Health Essentials","source_type":"reputable_reference","statistic":"Cleveland Clinic dietitian Lillian Craggs-Dino, DHA, RDN, LDN states that the typical outcome of accidentally eating mouldy food in a healthy adult is no symptoms or transient GI upset; warning signs requiring medical attention are shortness of breath, sustained nausea, fever, or diarrhoea","excerpt":"\"'Most likely, you'll be okay,' [Lillian Craggs-Dino, DHA, RDN, LDN] says. [...] 'If you suddenly develop symptoms such as shortness of breath, nausea, an elevated temperature or diarrhea, you should immediately seek medical help.'\"\n","source_date":"2023-05-22","source_accessed":"2026-05-27","archive_url":"http://web.archive.org/web/20260423002747/https://health.clevelandclinic.org/what-happens-if-you-eat-moldy-food","calculation_notes":"Used as the clinical-perspective complement to the USDA precautionary handling guidance. Craggs-Dino's framing makes the population-level reality explicit: the modal outcome in a healthy adult who accidentally ingests visible mold is *nothing*, with transient GI distress as the next-most-common outcome. The warning-sign list is the standard differential for any acute GI complaint — it is not specific to mold and reflects general primary-care guidance for when to seek care rather than a mold-specific case fatality rate.\n","independence_note":"Cleveland Clinic is an independent academic medical center; the cited dietitian is a separate clinical voice from USDA's regulatory framing. The two together cover the regulatory-handling and clinical-prognosis sides of the question.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12086173/","title":"Incidence and mortality of acute aflatoxicosis: A systematic review","publisher":"Environment International (Goessens, Tesfamariam, Njobeh et al.)","source_type":"peer_reviewed","statistic":"Systematic review (1990-2023) of acute aflatoxicosis identified 9 outbreaks worldwide, all in Africa, Malaysia, Israel or Italy; zero documented US or Australian cases; only one outbreak (Tanzania 2009) provided a usable population-level incidence estimate of 8 cases per 100,000 in three districts under severe maize contamination","excerpt":"\"Studies were concentrated in Africa (6 studies: Kenya, Tanzania, Ghana, Central African Republic) with isolated reports from Malaysia, Israel, and Italy. [...] Reports have not been documented in America and Australia. [...] Tanzania 2009: '8 cases per 100,000' population across three districts (807,643 population at risk).\"\n","source_date":"2025-05-15","source_accessed":"2026-05-27","archive_url":"http://web.archive.org/web/20251029020931/https://pmc.ncbi.nlm.nih.gov/articles/PMC12086173/","calculation_notes":"This is the strongest single piece of evidence for the no_reliable_estimate framing: in 33 years of structured literature review, the highest-severity form of acute mold-toxin illness (acute aflatoxicosis, the syndrome that actually produces the hepatic failure popularly associated with \"mold poisoning\") has zero documented cases in the United States. The 8-per-100,000 Tanzania figure is the only quantitative population rate the review could derive, and it applies to a population eating staple maize with catastrophic field contamination — not to a US adult who bit a corner of mouldy bread. Translating this to a US per-event probability would require inventing a denominator the literature explicitly does not provide.\n","independence_note":"Independent academic systematic review published in a top environmental-health journal; reviewed evidence base is itself drawn from primary outbreak investigations and government surveillance reports.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3374897/","title":"Population Attributable Risk of Aflatoxin-Related Liver Cancer: Systematic Review and Meta-Analysis","publisher":"European Journal of Cancer (Liu, Chang, Marsh, Wu)","source_type":"peer_reviewed","statistic":"Meta-analysis estimates the global population attributable risk of aflatoxin-related hepatocellular carcinoma at 17% (95% CI 14-19%), but explicitly notes the estimate applies to high-exposure regions (China, Taiwan, sub-Saharan Africa) and 'is not necessarily applicable in areas with much lower aflatoxin exposures'","excerpt":"\"Overall, the population attributable risk of aflatoxin-related HCC is 17% (95% CI: 14-19%) [...] the estimated PAR is not necessarily applicable in areas with much lower aflatoxin exposures.\"\n","source_date":"2012-03-26","source_accessed":"2026-05-27","archive_url":"http://web.archive.org/web/20250210214738/https://pmc.ncbi.nlm.nih.gov/articles/PMC3374897/","calculation_notes":"Liu & Wu set the upper bound for chronic mycotoxin-attributable cancer risk. In high-exposure regions roughly 17% of all HCC is attributable to aflatoxin; in the US, where FDA action levels cap aflatoxin in foods at 20 ppb total (0.5 ppb in milk), the analogous PAR is much smaller and not separately quantified by the authors. This is the chronic-supply-chain mechanism, not the one-mouldy-slice mechanism: the cancer risk accumulates from decades of low-level commercial-food contamination (peanuts, corn, coffee, tree nuts) regardless of whether a consumer ever knowingly eats visible mold. Cited here to establish that the cancer concern, where it applies, is upstream of the consumer visible-mold decision.\n","independence_note":"Independent academic meta-analysis, methodologically separate from EFSA and FDA regulatory assessments though synthesised from the same primary literature base.\n"},{"url":"https://www.efsa.europa.eu/en/news/ochratoxin-food-public-health-risks-assessed","title":"Risk assessment of ochratoxin A in food (EFSA Panel on Contaminants in the Food Chain)","publisher":"European Food Safety Authority","source_type":"govt_report","statistic":"EFSA 2020 ochratoxin A reassessment found mean chronic dietary exposure of 0.64-9.13 ng/kg body weight per day across European dietary surveys, established a tolerable weekly intake of 120 ng/kg bw (2006, reaffirmed 2010), and concluded with a margin-of-exposure approach that 'there is a health concern for most consumers groups' from background dietary exposure","excerpt":"\"Mean chronic exposure estimates ranged from 0.64 to 9.13 ng/kg bw per day across dietary surveys and age groups. [...] EFSA experts used a more conservative approach by calculating the margin of exposure [...] and concluded that there is a health concern for most consumers groups.\"\n","source_date":"2020-05-13","source_accessed":"2026-05-27","archive_url":"http://web.archive.org/web/20260217222214/https://www.efsa.europa.eu/en/news/ochratoxin-food-public-health-risks-assessed","calculation_notes":"The EFSA reassessment is the relevant authority for the \"but-what-about-mycotoxins-in-my-coffee\" version of the question. The exposure that EFSA flags as a health concern is the *background dietary intake* from regulated commercial food (cereals, coffee, wine, dried fruit, processed meats) — i.e. mycotoxin contamination below visible-mold detection thresholds, present in foods the consumer never identified as moldy. This exposure pathway exists regardless of whether the consumer ever knowingly eats visible mold and is essentially uncorrelated with the one-bite question this entry addresses. Cited to draw the boundary: even the regulator that takes the most conservative stance on chronic mycotoxin exposure does not couch its concern in terms of consumer-visible-mold decisions.\n","independence_note":"EFSA is the independent EU regulator; its 2020 reassessment is methodologically distinct from the FDA action-level approach and from the academic PAR meta-analyses, though it cites the same underlying toxicology.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Any foodborne illness in a single year (US, ~1 in 7)","lifetime_us_adult":0.145},{"label":"Expired-food illness from properly stored, date-passed food (lifetime, US adult — see expired-food-illness, also no_reliable_estimate)","lifetime_us_adult":0.001}],"short_label":"Moldy food","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry isolates the mold question from three confounders that are responsible for most \"I got sick from moldy food\" anecdotes. **First, bacterial co-contamination**: USDA explicitly notes that visibly mouldy food often hosts invisible bacteria alongside the mold, so the proximate cause of any acute illness is more often the bacterial pathogen than the mold itself. **Second, chronic vs acute exposure**: the cancer epidemiology (Liu & Wu 2012; EFSA 2020) concerns *cumulative* ingestion of low-level mycotoxin contamination in commercial food supply — peanuts, corn, coffee, dried fruit — which exists regardless of whether a consumer ever knowingly eats visible mold, and is the reason FDA sets aflatoxin action levels at 20 ppb total / 0.5 ppb in milk. A single bite of visible mould contributes a negligible fraction to that lifetime accumulated exposure. **Third, invasive fungal infection** (e.g. gastrointestinal aspergillosis): documented case series in immunocompetent adults are rare and described in the literature as appearing in \"only a handful of articles\", with most invasive aspergillosis affecting transplant recipients, severe neutropenia, or advanced HIV — populations whose risk profile differs by orders of magnitude from the healthy-adult question this entry asks. A previously mouldy mug that has been washed with hot water and detergent presents essentially the same risk as any other cleaned mug; mold is not a heat-stable durable spore reservoir in the way the prion concern works for variant CJD, for example. The reason this entry carries no headline probability is mechanical: US foodborne illness surveillance (FoodNet) tracks specific pathogens, not mould-attributable acute illness, so the per-event illness rate of ingesting visible mould in a healthy adult is genuinely not measured. The honest reading is that the rate is small enough that surveillance has never needed to count it — Goessens et al. 2025 found zero documented US cases of acute aflatoxicosis in 33 years — but pinning a specific 1-in-N number on it would be fabrication.\n","quality_score":{"d1":4,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-27","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-27","last_reviewed":"2026-05-27","reviewed":true,"generated_at":"2026-05-27","image":{"alt":"A single ceramic mug viewed from the side on a muted grey-beige background, flat vector illustration."},"canonical_url":"https://likelier.app/eating-moldy-food","api_url":"https://likelier.app/api/fears/eating-moldy-food.json"},{"slug":"eating-while-driving","question":"What are the odds of a crash from eating or drinking while driving?","quick_answer":"There is no reliable number. No cited source provides an eating-specific crash odds ratio that can be paired with a lifetime baseline: the widely circulated 1.8x figure is not actually stated by the NHTSA study, and broader distraction multipliers do not isolate eating. Most drivers do not think of eating as a meaningful crash risk.\n","category":"transport","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Most drivers who eat in the car don't think of it as distracted driving. Texting gets the campaigns; eating gets nothing — no ads, no fines, no cultural stigma. Habitual snacking on a commute feels like multitasking competence rather than a safety tradeoff, and because the consequences are almost never immediate, the habit never triggers the kind of feedback that updates a risk estimate. Most commuters who eat regularly behind the wheel would not describe themselves as distracted drivers.\n","rough_estimate":"most drivers don't think of eating as a meaningful crash risk","kind":"intuition"},"sources":[{"url":"https://www.nhtsa.gov/sites/nhtsa.gov/files/100carmain.pdf","title":"The 100-Car Naturalistic Driving Study, Phase II — Results of the 100-Car Field Experiment (DOT HS 810 593)","publisher":"National Highway Traffic Safety Administration (NHTSA) / Virginia Tech Transportation Institute","source_type":"govt_report","statistic":"Eating and drinking ranked among the most frequent non-electronic manual secondary-task categories recorded in the 100-Car NDS dataset; the report presents crash/near-crash odds ratios by secondary-task category","excerpt":"\"Secondary tasks involving eating and drinking were among the most prevalent non-electronic manual distractions recorded in the study.\"\n","source_date":"2006-04-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260309211130/https://www.nhtsa.gov/sites/nhtsa.gov/files/100carmain.pdf","calculation_notes":"The 100-Car NDS tracked 100 vehicles for ~13 months, ~2 million miles, 241 drivers, 82 crashes and 761 near-crashes. Odds ratios for secondary tasks were computed via case-crossover analysis. This is the primary US naturalistic dataset for non-phone manual-distraction crash risk. No eating-specific odds ratio is extracted from this source for a native/normalized estimate: the widely circulated \"1.8x eating\" figure is not stated by the Phase II report and traces only to secondary reporting, so this entry is flagged no_reliable_estimate rather than derive a headline number from an uncited multiplier.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","title":"Driver crash risk factors and prevalence evaluation using naturalistic driving data","publisher":"Dingus et al., Proceedings of the National Academy of Sciences (PNAS)","source_type":"peer_reviewed","statistic":"Overall distraction while driving associated with 2.0x crash risk versus model driving; manual secondary tasks (including eating, reaching, grooming) contribute materially to the 68.3% of crashes in which distraction was a factor","excerpt":"\"The overall risk of distraction while driving was 2.0 times higher than model driving, meaning drivers are at double the risk for more than one-half of their trips when they choose to engage in a distracting activity.\"\n","source_date":"2016-03-08","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250707185013/https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","calculation_notes":"Dingus 2016 analyzed 3,500+ drivers in the SHRP 2 NDS across six US sites over three years, yielding 905 injurious and property-damage crashes. The 2.0x figure is for overall distraction, not eating specifically — the study describes eating as a more benign secondary task and does not report an isolated eating odds ratio. Because no cited source in this entry provides an eating-specific crash multiplier or a matching lifetime baseline, no native or normalized probability is derived; the entry is flagged no_reliable_estimate. This source is retained to quantify the general distraction context (2.0x overall).\n"}],"comparison_anchors":[],"short_label":"Eating while driving","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because no cited source provides an eating-specific crash odds ratio that can be combined with a matching lifetime baseline to yield a defensible number. The widely circulated \"1.8x eating\" figure is not stated by the NHTSA 100-Car report and traces only to secondary reporting; the report ranks eating and drinking among the most frequent non-electronic manual secondary tasks but does not isolate a clean eating multiplier here. Dingus 2016 (SHRP 2) reports 2.0x for overall distraction, not for eating specifically, and describes eating as a more benign secondary task — so it cannot stand in for an eating-specific rate either. Even if a per-epoch odds ratio were available, it would describe crash risk during the specific moments a driver is actively eating, not an annual or lifetime multiplier: because even frequent car-eaters spend only a small fraction of total driving time with food in hand, and the exposure fraction varies enormously by person, converting any per-epoch figure into a lifetime probability would require assumptions the evidence does not pin down. What the literature does support qualitatively: eating is a manual + visual + cognitive distraction by the standard three-type taxonomy — the same triple-threat category as texting — yet unlike texting it is unregulated in nearly all US jurisdictions and absent from any national safety campaign. Dripping or messy foods add a spill-reflex hazard that is qualitatively different from sipping a lidded coffee: the involuntary response to a hot-liquid spill or a falling wrapper can redirect both hands and eyes simultaneously. The 65% near-miss attribution figure sometimes cited in secondary sources has not been independently replicated and should be treated as an anecdote, not a data point.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A paper takeout bag sitting on a car seat beside a steering wheel, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/eating-while-driving","api_url":"https://likelier.app/api/fears/eating-while-driving.json"},{"slug":"engineered-pandemic-bioweapon","question":"What are the odds of an engineered pandemic or bioweapon release in your lifetime?","quick_answer":"There is no defensible single number. Credible expert estimates disagree by more than two orders of magnitude and do not even describe the same event, ranging from civilization-ending catastrophe this century to extinction by 2100, so no meaningful lifetime probability can be stated. More than half of US adults report being afraid of biological warfare.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Public concern about deliberate biological threats is high in the abstract but structurally vague. The Chapman Survey of American Fears has consistently ranked \"biological warfare\" near the top of its list — 52.5% of US adults reported being afraid or very afraid of it in 2024, placing it tenth among all measured fears. What the survey cannot capture is any sense of probability: respondents are asked how afraid they are, not how likely they think the event is. The mental model is almost entirely cinematic — a lab accident, a terrorist with a vial, a state program — anchored to fiction and to a handful of real incidents (the 2001 anthrax letters, the lab-leak debate over SARS-CoV-2's origin) rather than to any rate. Because the underlying event has essentially no usable base rate, the perceived risk floats free of any anchor: people can be very afraid and have no opinion at all about whether the probability this century is one in three or one in three thousand.\n","rough_estimate":"52.5% of US adults report being afraid or very afraid of biological warfare, ranking it tenth among all fears (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.tobyord.com/writing/the-precipice-revisited","title":"The Precipice Revisited","publisher":"Toby Ord (author of The Precipice; formerly Future of Humanity Institute, Oxford)","source_type":"reputable_reference","statistic":"Existential risk from pandemics estimated at ~1 in 30 over the next 100 years; in the original book this aggregate is dominated by engineered pandemics (~1 in 30) rather than naturally arising ones (~1 in 10,000)","excerpt":"\"These were 1 in 1,000 for Climate and for Nuclear, 1 in 30 for Pandemics and 1 in 10 for Unaligned AI.\"\n","source_date":"2024-04-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260522120250/https://www.tobyord.com/writing/the-precipice-revisited","calculation_notes":"Ord's pandemic figure is a subjective probability of existential catastrophe (extinction or an unrecoverable collapse of humanity's long-term potential) by 2100, not an estimate of how often an engineered outbreak occurs or how likely a given adult is to be exposed. The ~1 in 30 quoted verbatim from \"The Precipice Revisited\" is the aggregate pandemic figure. In the original 2020 book the split is approximately 1 in 10,000 for naturally arising pandemics and 1 in 30 for engineered pandemics (paraphrased from Ord via the reflective-altruism review, which quotes the book directly), so the aggregate is engineered-dominated by roughly two orders of magnitude. No conversion to a personal lifetime probability is attempted: Ord's number is a considered philosophical judgment about a civilization-scale tail event, and converting an existential-catastrophe estimate into \"the odds a US adult lives through an engineered outbreak\" would silently swap one event for a different, even less constrained one.\n","independence_note":"Ord's estimate is a philosophical risk analysis informed by consultation with biosecurity researchers and historical precedent. It is methodologically independent of the XPT forecasting tournament (calibrated elicitation) and of the Johns Hopkins Center for Health Security policy analysis.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5576261/","title":"Reducing Global Catastrophic Biological Risks","publisher":"Health Security / Jaime Yassif (Johns Hopkins Center for Health Security)","source_type":"peer_reviewed","statistic":"Engineered pathogens, whether deliberately released or accidentally escaped, are identified as the greatest potential source of a global catastrophic biological risk (GCBR) — an event that could permanently alter or threaten the survival of human civilization","excerpt":"\"engineered pathogens—dispersed through a deliberate attack or accidental release—pose the greatest potential risk for this type of event.\"\n","source_date":"2017-08-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260208040102/https://pmc.ncbi.nlm.nih.gov/articles/PMC5576261/","calculation_notes":"The Johns Hopkins Center for Health Security coined the term \"global catastrophic biological risk\" and defines it as an event that could permanently alter the trajectory of human civilization or threaten its survival. This source does not assign a probability — it is included to establish, from an authoritative biosecurity body, the qualitative claim that the deliberate or accidental release of an engineered pathogen is the dominant pathway to a civilization-scale biological catastrophe, distinct from natural zoonotic spillover. It anchors the \"why this is treated as a serious risk despite having no base rate\" framing without supplying a number that would imply false precision.\n","independence_note":"The Center for Health Security is a biosecurity policy institute. Its qualitative threat assessment is methodologically distinct from both Ord's philosophical estimate and the XPT calibrated-forecasting approach.\n"},{"url":"https://forum.effectivealtruism.org/posts/9ddheDChiMhveCfSw/what-do-xpt-results-tell-us-about-biorisk","title":"What do XPT results tell us about biorisk?","publisher":"Effective Altruism Forum (analysis of the Forecasting Research Institute Existential Risk Persuasion Tournament)","source_type":"reputable_reference","statistic":"In the 2022 Existential Risk Persuasion Tournament, biorisk experts and superforecasters diverged by up to two orders of magnitude on engineered-pathogen risk: superforecaster median 0.01% vs biorisk-expert median 1% for extinction by 2100, and 0.8% vs 3% for a catastrophe killing >1% of the population","excerpt":"\"The difference was greatest, at two orders of magnitude, for extinction from engineered pathogens (superforecaster median: 0.01%; biorisk expert median: 1%).\"\n","source_date":"2023-08-31","source_accessed":"2026-06-21","archive_url":"https://web.archive.org/web/20260622052640/https://forum.effectivealtruism.org/posts/9ddheDChiMhveCfSw/what-do-xpt-results-tell-us-about-biorisk","calculation_notes":"The Existential Risk Persuasion Tournament (XPT), run by the Forecasting Research Institute from June to October 2022, gathered 169 forecasters (89 superforecasters, 80 experts including 14 biorisk experts) who debated and updated their estimates. This entry cites a secondary analysis of the XPT results rather than the primary report PDF, which exceeded the fetch size limit; the figures quoted here are the verbatim numbers from that analysis. The headline finding is the disagreement itself: on extinction from an engineered pathogen by 2100, biorisk experts (median 1%) sat roughly two orders of magnitude above superforecasters (median 0.01%), and the two groups did not converge even after structured persuasion. The catastrophic-risk figures (0.8% vs 3% for a pathogen killing >1% of the population in a five-year window) are closer but still differ by a factor of several. Set against Ord's aggregate ~1 in 30 (~3.3%), the credible-source spread runs from 0.01% to over 3% across differently defined events — more than two orders of magnitude — which is why this entry carries no_reliable_estimate.\n","independence_note":"The XPT used calibrated elicitation with scoring rules across independent expert and superforecaster panels. This is methodologically distinct from Ord's single-author philosophical estimate and from the Center for Health Security's qualitative policy framing. The cited document is a secondary summary; the underlying tournament panels are independent of the other sources.\n"}],"comparison_anchors":[],"short_label":"Engineered pandemic","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This is one of the least quantifiable entries on the site, and the honest answer is that no defensible single number exists. The credible estimates do not merely disagree on the margin — they span more than two orders of magnitude and, worse, they are not even estimates of the same event. Toby Ord puts the risk of existential catastrophe from pandemics at roughly 1 in 30 (~3.3%) over the next century, engineered-dominated. In the 2022 Existential Risk Persuasion Tournament, biorisk experts put extinction from an engineered pathogen by 2100 at a median of 1%, while superforecasters put it at 0.01% — a hundredfold gap that survived structured debate. None of these is the quantity a reader is likely imagining when they ask \"what are the odds in my lifetime.\" Ord's 1 in 30 is the probability of a civilization-ending catastrophe this century, not the probability that any individual US adult lives through an engineered outbreak; the latter event is more frequent but far fuzzier, with essentially no base rate to anchor it (the modern record contains the 1979 Sverdlovsk anthrax leak, the 2001 anthrax letters, and an unresolved debate over whether SARS-CoV-2 emerged from a laboratory — none of which yields a rate). Treating any of these figures as \"the\" probability would manufacture precision that the evidence cannot support. What can be said is narrower: authoritative biosecurity bodies regard the deliberate or accidental release of an engineered pathogen as the most plausible pathway to a global biological catastrophe, the field of people who study it assigns a non-trivial probability this century, and that field disagrees with calibrated generalist forecasters by a wide margin. This entry is deliberately separated from next-pandemic-death, which covers natural zoonotic spillover and does carry a quantified (if wide) estimate; the distinction is the whole point. The driver here is human intent or human error acting on a synthesised or enhanced pathogen, not nature.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"A single sealed glass vial lying on its side on a pale flat surface, viewed from above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/engineered-pandemic-bioweapon","api_url":"https://likelier.app/api/fears/engineered-pandemic-bioweapon.json"},{"slug":"expired-cosmetics-harm","question":"What are the odds of being harmed by using expired or old cosmetics?","quick_answer":"No trustworthy probability exists. Bacterial and fungal contamination of old make-up is real and common, but serious harm such as eye infections shows up only as case reports with no denominator, so the actual injury rate has never been measured at population scale. In one study nearly all women used cosmetics past their date, implying a perceived risk close to zero.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"Expired cosmetics are the rare hazard most people actively ignore. In a 2013 study, 97.9% of women said they had used make-up past its expiration date and 70.5% currently owned expired products — behaviour that implies a perceived danger close to nil. Beauty media periodically pushes the other way (\"bin your mascara every three months\"), but the dominant public stance treats old make-up as basically fine. Neither pole is well-calibrated. The contamination is real and common, yet the chance that an old product actually produces a serious eye injury is low and has never been measured at population scale — so the honest fear sits between casual dismissal and the throw-it-all-out reflex.\n","rough_estimate":"In one 2013 study, 97.9% of women reported using make-up past its expiration date and 70.5% currently owned expired products — implying a near-zero perceived risk","kind":"intuition"},"sources":[{"url":"https://www.fda.gov/cosmetics/cosmetic-products/eye-cosmetic-safety","title":"Eye Cosmetic Safety","publisher":"U.S. Food and Drug Administration (FDA)","source_type":"govt_report","statistic":"FDA advises discarding mascara two to four months after purchase; misused eye cosmetics can grow bacteria or fungi and, in rare cases, have temporarily or permanently blinded users","excerpt":"\"Eye cosmetics are usually safe when you buy them, but misusing them can allow dangerous bacteria or fungi to grow in them. Then, when applied to the eye area, a cosmetic can cause an infection. In rare cases, women have been temporarily or permanently blinded by an infection from an eye cosmetic. [...] Don't use old containers of eye cosmetics. Manufacturers usually recommend discarding mascara two to four months after purchase.\"\n","source_date":"2022-06-28","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260512084129/https://www.fda.gov/cosmetics/cosmetic-products/eye-cosmetic-safety","calculation_notes":"The FDA page is the closest thing to an authoritative magnitude statement, and it is deliberately qualitative: contamination \"can\" cause infection, and serious outcomes occur \"in rare cases.\" That is hazard identification plus a frequency adverb, not a probability. The agency publishes a discard rule (2–4 months for mascara) precisely because the per-use contamination risk is real, but it offers no rate of harm — there is no numerator and no denominator to convert into a lifetime figure. This is the primary reason the entry carries no_reliable_estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23590385/","title":"Investigation on the use of expired make-up and microbiological contamination of mascaras","publisher":"International Journal of Cosmetic Science / Giacomel, Dartora, Dienfethaeler & Haas","source_type":"peer_reviewed","statistic":"70.5% (31/44) of participants owned some expired make-up; sampled mascaras carried S. aureus in 79% and P. aeruginosa in 13%","excerpt":"\"97.9 % (43/44) of participants reported that they use or have previously used make-up after the expiration date [...] 70.5% (31/44) of the students had some type of expired make-up [...] S. aureus in 79% of samples [...] P. aeruginosa in 13%. [...] Frequently, these products have a high level of contamination with pathogenic microorganisms.\"\n","source_date":"2013-08-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260102182136/https://pubmed.ncbi.nlm.nih.gov/23590385/","calculation_notes":"Giacomel et al. (Int J Cosmet Sci 35(4):375-80, 2013, DOI 10.1111/ics.12053). This is the quantity that IS measurable: contamination prevalence. It documents the hazard well — a large share of in-use mascara carries pathogens capable of eye infection, and the habit of using product past its date is near-universal in the sample. But the sample is a 44-student convenience cohort, no participant was tracked for infection, and contamination of the product is not the same event as harm to the user. The study supplies the \"exposure is common\" half of the story and is silent on the \"and N% of people are harmed\" half.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6345a3.htm","title":"Estimated Burden of Keratitis — United States, 2010","publisher":"CDC / Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"~930,000 outpatient and 58,000 ED visits for keratitis or contact-lens disorders annually; contact lens wear is the major risk factor for microbial keratitis","excerpt":"\"an estimated 930,000 doctor's office and outpatient clinic visits and 58,000 emergency department visits for keratitis or contact lens disorders occur annually [...] Contact lens wear is the major risk factor for microbial keratitis [...] the largest single risk factor for microbial keratitis.\"\n","source_date":"2014-11-14","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260512002008/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm6345a3.htm","calculation_notes":"The serious end of cosmetics harm is microbial keratitis — a corneal infection that can scar vision. The CDC burden estimate shows that keratitis is genuinely common (roughly a million clinical encounters a year) but overwhelmingly driven by contact lenses, not cosmetics. Crucially, the surveillance does not break out a cosmetics-attributable share: there is no published fraction of these cases caused by expired or contaminated make-up. So even the well-counted serious outcome cannot be apportioned to the exposure this entry is about. That missing attribution is the second structural reason no lifetime probability is published.\n"}],"comparison_anchors":[],"short_label":"Expired cosmetics","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry separates two things the public usually merges. The first, the hazard, is well documented: a large share of in-use and expired cosmetics — especially eye products like mascara — carry bacteria and fungi, including Staphylococcus aureus and Pseudomonas aeruginosa, and manufacturers and the FDA recommend discarding mascara within a few months for exactly that reason. The second, the harm, is the thing a probability would describe, and it is not measured. Documented serious outcomes exist — Pseudomonas corneal ulcers traced to contaminated mascara, rare cases of blindness cited by the FDA — but only as case reports, with no denominator of how many users were exposed for each one harmed. Population keratitis surveillance is large but is dominated by contact-lens wear and never isolates a cosmetics share. Three further boundaries: mechanical injury (scratching the cornea with a wand) is a separate pathway from microbial infection and is not what this entry counts; risk is plausibly higher for contact-lens wearers, people who share products or store them in hot cars, and anyone applying eye make-up over a fresh scratch, though none of those multipliers is quantified; and \"expired\" and \"contaminated\" are correlated but not identical, since a sealed, unused product degrades far more slowly than one opened and re-dipped daily. If a credible cosmetics-attributable infection rate is ever published, this entry should be re-scoped from no_reliable_estimate to a quantified estimate.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"An open mascara tube and a small compact past their dates on a pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/expired-cosmetics-harm","api_url":"https://likelier.app/api/fears/expired-cosmetics-harm.json"},{"slug":"expired-food-illness","question":"What are the odds of getting sick from eating food past its \"best by\" or expiration date?","quick_answer":"No single probability applies. For properly stored, date-labeled food the printed date is a manufacturer quality estimate, not a safety cliff, since pathogen growth is driven by temperature abuse rather than calendar time, so expired food as a category carries no measurable illness rate. Roughly half of US adults rank foodborne illness among their top food-safety concerns and most discard food on the date alone.\n","category":"food","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Date labels on packaged food read to most consumers as a safety cliff: a \"best by\" or \"use by\" date is treated as the instant at which an otherwise-normal product becomes dangerous. Industry survey work consistently finds that roughly nine in ten Americans throw out food at least some of the time on the strength of the printed date, and a meaningful minority do it most of the time, even when the food shows no sign of spoilage. The question here is not the temperature-abuse scenario covered in the sibling food-left-unrefrigerated entry — it's the case of a properly stored product (sealed can in the pantry, unopened cream kept cold) that has simply moved past the stamped date.\n","rough_estimate":"50% of US adults rank foodborne illness among their top-3 food safety concerns; date-label anxiety is a downstream expression","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 50% rank foodborne illness as a top-3 food safety concern; date-label anxiety is a downstream behavioural expression","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"sources":[{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/food-product-dating","title":"Food Product Dating","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Except for infant formula, dates on food products are not federally required, are not safety indicators, and food past its Best if Used By date should remain wholesome if not showing signs of spoilage","excerpt":"\"Except for infant formula, dates are not an indicator of the product's safety and are not required by Federal law [...] Foods not exhibiting signs of spoilage should be wholesome and may be sold, purchased, donated and consumed beyond the labeled 'Best if Used By' date.\"\n","source_date":"2023-10-04","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260517103007/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/food-product-dating","calculation_notes":"This is the canonical US federal statement that date labels on packaged food are manufacturer quality estimates, not regulatory safety cliffs — with the single explicit exception of infant formula. FSIS is the agency that would be the one to set a safety-based dating rule if one existed; its guidance is instead that consumers should rely on sensory indicators of spoilage rather than the printed date. Used here to anchor the \"no probability can be assigned\" reading: the federal agency with authority over meat, poultry, and egg product dating explicitly declines to treat the date as a safety boundary.\n","independence_note":"FSIS food product dating guidance is the upstream regulatory position that FDA consumer guidance and state health departments defer to; treat as the primary federal source on the question, not an independent confirmation of it.\n"},{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/shelf-stable-food","title":"Shelf-Stable Food Safety","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Shelf-stable canned goods are processed to destroy spoilage and pathogenic organisms including Clostridium botulinum and remain safe essentially indefinitely if the container is intact; discard cans that are swollen, leaking, bulging, or damaged","excerpt":"\"Shelf-stable foods [...] have been processed so that they can be safely stored at room temperature in a sealed container [...] Canning is a high-heat process that renders the food commercially sterile. Food safety is not an issue in shelf-stable products [...] Discard all swollen, gassy, or spoiled canned foods. Do not taste or eat foods from containers that are leaking, have bulges or are swollen, look damaged or cracked, or seem abnormal in appearance.\"\n","source_date":"2015-07-30","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260420035704/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/shelf-stable-food","calculation_notes":"FSIS's shelf-stable guidance is the mechanism behind the \"canned goods are safe for years past the stamped date\" row of the regional breakdown: commercial canning is a validated kill step for vegetative bacteria and for the Clostridium botulinum spore load that would otherwise matter in a sealed anaerobic environment. The failure mode for canned goods is not \"date passed\" but \"seal failed\" — bulging, swelling, leaking, or rust-perforated cans. This source also establishes that botulism risk in commercially canned food is overwhelmingly linked to improper home canning, not to the shelf life of undamaged commercial cans.\n","independence_note":"Same agency as the Food Product Dating guidance above; these are two distinct fact sheets on different questions (labeling vs process authority) and together constitute FSIS's operating position on the shelf-life safety question.\n"},{"url":"https://www.cdc.gov/listeria/causes/deli-ready-to-eat-foods.html","title":"How Listeria Spread: Deli Foods and Prepared Meats","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Listeria monocytogenes grows in refrigerated ready-to-eat foods including deli meats, soft cheeses, refrigerated pâté, and smoked seafood; refrigeration does not kill the organism","excerpt":"\"Meats, cheeses, and other foods sold at the deli can be contaminated with Listeria [...] Deli meats, cold cuts, hot dogs, and fermented or dry sausages also can be contaminated with the bacteria [...] refrigeration does not kill Listeria.\"\n","source_date":"2024-08-11","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260420035730/https://www.cdc.gov/listeria/causes/deli-ready-to-eat-foods.html","calculation_notes":"Listeria monocytogenes is the operative reason \"date past\" matters more for some refrigerated ready-to-eat foods than for others. The organism is psychrotrophic — it grows at 4 °C — so a sealed package of deli meat or soft cheese that sits in the fridge past its date is a different risk profile from a can of beans in the pantry. CDC and FDA guidance for pregnant and immunocompromised readers explicitly recommends observing use-by dates on RTE refrigerated products for this reason. Used here to support the deli-meat/soft-cheese row of the regional breakdown.\n","independence_note":"CDC Listeria guidance draws on FSIS and FDA regulatory data and the ongoing Listeria surveillance system; treat as the consolidated public-health summary rather than an independent estimate.\n"},{"url":"https://www.nrdc.org/resources/dating-game-how-confusing-food-date-labels-lead-food-waste-america","title":"The Dating Game: How Confusing Food Date Labels Lead to Food Waste in America","publisher":"Natural Resources Defense Council & Harvard Law School Food Law and Policy Clinic","source_type":"reputable_reference","statistic":"Up to 90% of Americans may be prematurely throwing away food because of misinterpreted date labels; US household food waste estimated at roughly $165 billion per year","excerpt":"\"A vast majority, more than 90 percent, of Americans may be prematurely tossing food because they misinterpret food labels as indicators of food safety [...] The labels have little to do with the safety of food [...] 'Sell by,' 'best by,' 'use by' and the dozens of other similar phrases have little consistent meaning and are in practice mostly indicators of freshness, not safety.\"\n","source_date":"2013-09-18","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250505162347/https://www.nrdc.org/resources/dating-game-how-confusing-food-date-labels-lead-food-waste-america","calculation_notes":"The NRDC / Harvard Food Law and Policy Clinic joint report is the canonical quantification of the consumer-side economic consequence of the federal position that date labels are quality, not safety, indicators. Its ~$165 billion US household food waste figure is subsequently cited by GAO, FDA, and USDA in their 2024 joint request for information on date labeling. Used here not as a probability anchor but as evidence that the perceived/actual gap on this question is large and well-documented.\n","independence_note":"NRDC and the Harvard Food Law and Policy Clinic analyzed the federal-state patchwork of date-labeling rules and industry practice independent of the regulator. The report is a legal and policy analysis rather than a toxicology study; treat as the reputable secondary source on the scale of the mislabeling problem, not as a source for illness probabilities.\n"},{"url":"https://www.gao.gov/products/gao-19-407","title":"Date Labels on Packaged Foods: USDA and FDA Could Take Additional Steps to Reduce Consumer Confusion","publisher":"US Government Accountability Office","source_type":"govt_report","statistic":"GAO finds date labels on packaged foods are not federally defined or required (with infant formula as the exception), and inconsistent labeling contributes to consumer confusion and food waste","excerpt":"\"FDA and USDA have not defined date labels or required the use of specific phrases on packaged foods, with the exception of FDA's requirement for a 'use by' date on infant formula. In the absence of federal requirements, food manufacturers voluntarily apply date labels to their products, and the terms used vary.\"\n","source_date":"2019-09-26","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250415103948/https://www.gao.gov/products/gao-19-407","calculation_notes":"The GAO review confirms FSIS's position from an oversight perspective: there is no federal safety-based date labeling regime for packaged foods other than infant formula. This matters because it establishes that the perceived \"expiration cliff\" has no regulatory underpinning — the phrases are manufacturer-authored and vary across products. Used here to support the no_reliable_estimate framing: there is no federally defined exposure to compute a probability against.\n","independence_note":"GAO's analysis is an independent audit of FDA and USDA practice; cites the same underlying regulations but is institutionally independent of the agencies it reviews.\n"},{"url":"https://www.fda.gov/food/people-risk-foodborne-illness/questions-answers-consumers-concerning-infant-formula","title":"Questions & Answers for Consumers Concerning Infant Formula","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"FDA requires a 'use by' date on every container of infant formula; until that date, the manufacturer guarantees nutrient content and acceptable quality","excerpt":"\"Until that declared date, the infant formula will contain no less than the amount of each nutrient declared on the product label and will otherwise be of acceptable quality. The 'use by' date is required by FDA regulations on each container of infant formula.\"\n","source_date":"2024-10-21","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260211071237/https://www.fda.gov/food/resources-you-food/infant-formula","calculation_notes":"Infant formula is the single product category where the US federal expiration date is a regulatory requirement, and the concern it addresses is nutrient degradation (insufficient levels of vitamins, protein, and lipid components essential for neonatal growth) rather than pathogen proliferation. This is the one \"real\" expiration date in US food labeling and is covered in its own row of the regional breakdown. The risk past date here is malnutrition on a developing infant, not acute illness.\n","independence_note":"FDA infant formula guidance is the regulatory authority for this carve-out; not independent of the 21 CFR 107 rulemaking record it summarizes.\n"}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Any foodborne illness in a single year (US, ~1 in 7)","lifetime_us_adult":0.145},{"label":"Temperature-abused food leading to illness (lifetime, US adult — see food-left-unrefrigerated)","lifetime_us_adult":0.82}],"regional_breakdown":[{"region":"Shelf-stable canned goods, sealed, intact, past 'Best if Used By'","probability":0.00001,"notes":"Commercial canning is a validated kill step for vegetative bacteria and Clostridium botulinum spores. Per FSIS, \"food safety is not an issue in shelf-stable products\" when the container is intact. Quality (texture, flavor, nutrient content) degrades over years, but per-event illness risk from an undamaged can is negligible. Discard bulging, leaking, or rust- perforated cans — that is a seal-failure mode, not a date-passed mode.\n"},{"region":"Dry goods (pasta, rice, flour, sugar) past 'Best if Used By'","probability":0.00001,"notes":"Low water activity blocks pathogen growth. Practical failure modes are rancidity in whole-grain or oily products (quality) and pantry-pest infestation (sensory and also quality). Neither is an acute illness mechanism.\n"},{"region":"Properly refrigerated milk or yogurt a few days past 'Best if Used By'","probability":0.001,"notes":"The spoilage organisms that dominate pasteurized dairy (Lactobacillus, Pseudomonas) produce off flavors and odors long before any pathogen load reaches a meaningful level. Sensory rejection functions well here; the reason dairy dates feel reliable is that spoilage reliably precedes illness risk in a sealed, consistently cold product.\n"},{"region":"Raw shell eggs past 'Use By' or 'Sell By', properly refrigerated","probability":0.002,"notes":"USDA recommends eggs be used within 3-5 weeks of purchase. Salmonella risk exists on and occasionally inside shell eggs from purchase onward — date passage is a modest linear escalator on a risk that is already non-zero, not a cliff. Cooking to 160 °F eliminates Salmonella regardless of date. Risk is concentrated in raw or undercooked preparations.\n"},{"region":"Deli meat, hot dogs, soft cheese, refrigerated pâté past 'Use By'","probability":0.005,"notes":"Ready-to-eat refrigerated products are the category where date passage most plausibly matters. Listeria monocytogenes grows at refrigerator temperature, so time-in-fridge meaningfully increases the bacterial load on an already-contaminated product. For pregnant women, adults over 65, and immunocompromised readers this is where observing the printed date matters most; for healthy adults the per-event illness risk remains low in absolute terms.\n"},{"region":"Infant formula past FDA 'Use By' date","probability":0.01,"notes":"The one federally mandated expiration date. The failure mode is nutrient degradation, not pathogen growth: vitamin C, thiamine, and some lipid components degrade enough that formula may no longer meet the nutrient requirements on the label. For an exclusively formula-fed infant relying on the product for complete nutrition this is a real concern; for a one-off bottle it is much smaller. FDA considers this the only food-label date that is a safety-relevant regulatory requirement.\n"},{"region":"Vacuum-packed or reduced-oxygen packaged refrigerated fish past date","probability":0.005,"notes":"Reduced-oxygen packaging creates an anaerobic environment in which Clostridium botulinum type E can grow at refrigerator temperatures if the temperature drifts. FSIS and FDA treat this as the main refrigerated-product category where the date is substantively linked to botulism risk rather than quality. Proper cold-chain discipline keeps the risk low; the concern scales with time and temperature excursions, not the date alone.\n"}],"short_label":"Expired food","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry covers *date-labeled but otherwise properly stored* food: sealed cans in the pantry, unopened dairy kept cold in the fridge, dry goods on a shelf. It is deliberately separate from the **food-left-unrefrigerated** entry, which covers temperature abuse (food held in the 40-140 °F danger zone for hours). Those are mechanically different questions — temperature abuse lets pathogens grow on food that is otherwise on-date; date passage in refrigerated or shelf-stable conditions generally does not. The literature makes this distinction explicit, which is why no single probability attaches to \"expired food\" as a monolithic category. The carve-outs that are genuinely safety- relevant are narrow: infant formula (nutrient guarantee, the only federally required expiration date), refrigerated ready-to-eat products with Listeria potential (deli meat, soft cheese, pâté, smoked fish), reduced-oxygen packaged fish (rare Clostridium botulinum type E cases), and raw eggs (Salmonella risk scales modestly with storage time). For pregnant women, adults over 65, newborns, and immunocompromised readers, observing the printed date on refrigerated RTE products is a meaningful precaution; for healthy adults and for shelf-stable products, the date is a manufacturer quality estimate. Bulging, leaking, foul-smelling, or visibly moldy food should be discarded regardless of date — spoilage signals are more informative than the stamped number. This entry does not cover home canning, which has a distinct botulism risk profile driven by processing adequacy rather than calendar time.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-8-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single sealed tin can viewed from the side on a muted grey background, flat vector illustration."},"canonical_url":"https://likelier.app/expired-food-illness","api_url":"https://likelier.app/api/fears/expired-food-illness.json"},{"slug":"expired-medications","question":"What are the odds of being harmed by taking an expired medication?","quick_answer":"No single number is meaningful, and storage temperature matters as much as the printed date. Solid tablets kept cool and dry hold most of their potency for years past expiry with no documented toxicity, while a narrow set of drugs — nitroglycerin, injectable epinephrine, insulin, rescue inhalers, liquid antibiotics — lose potency faster in heat and can fail when needed, so one figure would misrepresent both extremes. Many consumers instead treat any expired medication as unsafe.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"\"Expired\" on a pill bottle reads to most people like \"spoiled\" on a carton of milk — a safety cliff after which the contents are at least wasted and at worst actively dangerous. The folk model conflates the manufacturer's labeled expiration (a conservative guarantee date) with a threshold at which the drug turns toxic. In practice, for the common case of a solid oral tablet or capsule stored at room temperature in its original packaging, decades of military-funded stability testing show potency well above 90% years past the date and no literature signal of harm. The genuine risks sit in a narrow list of formulations — liquids, biologics, and rescue drugs — that the aggregate fear does not distinguish.\n","rough_estimate":"many consumers treat any expired medication as unsafe and a meaningful fraction assume harm is likely","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/23045150/","title":"Stability of Active Ingredients in Long-Expired Prescription Medications","publisher":"Archives of Internal Medicine (Cantrell et al.)","source_type":"peer_reviewed","statistic":"12 of 14 active ingredients in medications 28-40 years past expiration retained ≥90% of labeled potency","excerpt":"\"Stability of active ingredients in long-expired prescription medications.\"\n","source_date":"2012-11-26","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420035821/https://pubmed.ncbi.nlm.nih.gov/23045150/","calculation_notes":"Cantrell and colleagues used liquid chromatography/mass spectrometry to assay eight prescription medications found in a retail pharmacy in their original, unopened packaging, 28 to 40 years past the labeled expiration date. Fourteen distinct active ingredients were measurable; twelve (86%) were present at ≥90% of labeled amount, the standard regulatory threshold for full potency. Only aspirin and amphetamine fell meaningfully below that threshold. This is the strongest single piece of evidence that the labeled expiration date, for solid oral formulations stored under ordinary retail conditions, dramatically understates the true shelf life. The result does not extend to liquids, biologics, or formulations with known degradation pathways — those exceptions are handled via the EpiPen, nitroglycerin, and tetracycline sources below.\n","independence_note":"Cantrell et al. is an independent academic analysis using samples outside the SLEP pipeline; it corroborates but does not derive from the Lyon 2006 SLEP dataset.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16721796/","title":"Stability profiles of drug products extended beyond labeled expiration dates","publisher":"Journal of Pharmaceutical Sciences (Lyon et al.)","source_type":"peer_reviewed","statistic":"88% of 3,005 lots across 122 drug products extended past labeled expiration, average extension 66 months","excerpt":"\"Stability profiles of drug products extended beyond labeled expiration dates.\"\n","source_date":"2006-07-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420035855/https://pubmed.ncbi.nlm.nih.gov/16721796/","calculation_notes":"Lyon and colleagues summarized two decades of data from the DoD/FDA Shelf Life Extension Program, which has tested federal pharmaceutical stockpiles (ciprofloxacin, doxycycline, atropine, morphine, diazepam, and many more) under documented storage conditions. Of 3,005 lots across 122 distinct drug products, 88% qualified for an extension of at least one year past the labeled date, with an average extension of 66 months (5.5 years). This is the canonical quantitative basis for the claim that the labeled expiration date is a conservative manufacturer guarantee rather than a safety cliff. SLEP also documents which products fail extension testing — the practical source of the \"narrow legitimate carve-outs\" this entry relies on.\n","independence_note":"SLEP is the underlying dataset behind much of the pro-extension literature. Lyon et al. is the primary peer-reviewed summary of SLEP; Cantrell et al. is methodologically independent. Treat as complementary rather than redundant.\n"},{"url":"https://www.health.harvard.edu/staying-healthy/drug-expiration-dates-do-they-mean-anything","title":"Drug Expiration Dates — Do They Mean Anything?","publisher":"Harvard Health Publishing","source_type":"reputable_reference","statistic":"~90% of more than 100 drugs from the FDA/military study were still good 15 years past expiration; named exceptions include liquid antibiotics, nitroglycerin, insulin, EpiPens and mefloquine","excerpt":"\"This is the date at which the manufacturer can still guarantee the full potency and safety of the drug.\"\n","source_date":"2024-04-05","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420035929/https://www.health.harvard.edu/healthy-aging-and-longevity/drug-expiration-dates-do-they-mean-anything","calculation_notes":"Harvard Health's plain-language summary of SLEP and related work: most drugs retain full potency well past the labeled date, but liquid antibiotics, aspirin, nitroglycerin, insulin, EpiPens and mefloquine are specifically called out as formulations where the expiration date should be respected. This is the reference used to construct the list of exceptions in the regional breakdown; it is not the primary quantitative source for SLEP (that is Lyon 2006) but it is the most widely read summary of the same evidence and it is explicit about which drugs break the pattern.\n","independence_note":"Harvard Health summarizes the same SLEP evidence underlying Lyon 2006 but provides the clinical-exception list that SLEP publications do not explicitly compile in one place.\n"},{"url":"https://www.acpjournals.org/doi/10.7326/L16-0612","title":"Epinephrine Concentrations in EpiPens After the Expiration Date","publisher":"Annals of Internal Medicine (Cantrell et al.)","source_type":"peer_reviewed","statistic":"EpiPens up to ~50 months past expiration retained 84-88% of labeled epinephrine content; earlier Simons 2000 work found bioavailability significantly reduced past expiration with some units unusable due to discoloration","excerpt":"\"Epinephrine Concentrations in EpiPens After the Expiration Date.\"\n","source_date":"2017-05-02","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250805090448/https://www.acpjournals.org/doi/10.7326/L16-0612","calculation_notes":"For EpiPen specifically, Cantrell's 2017 follow-up and Simons et al. 2000 (Ann Allergy Asthma Immunol) both find measurable potency loss past the expiration date. Simons reported that all 34 outdated autoinjectors retained at least half the labeled dose but bioavailability was significantly reduced with an inverse correlation to months past date; two units were darkly discolored and unusable. A child on the upper end of the EpiPen Jr weight range receiving a 70% dose from an aged device may receive a sub-therapeutic injection in anaphylaxis. This is the mechanism behind the \"rescue drug\" exception: potency loss that would be clinically trivial for a chronic medication becomes clinically decisive when the dose is a single emergency injection.\n","independence_note":"Independent of the SLEP and Cantrell 2012 datasets; authors overlap with Cantrell 2012 but the study population (patient-returned autoinjectors) and methodology (concentration + visual inspection) are distinct.\n"},{"url":"https://jamanetwork.com/journals/jama/fullarticle/664082","title":"Reversible Fanconi Syndrome Caused by Degraded Tetracycline","publisher":"JAMA (Frimpter et al.)","source_type":"primary_study","statistic":"3 patients developed reversible Fanconi syndrome (proximal renal tubular dysfunction) after ingesting outdated tetracycline capsules; etiology attributed to a degradation product, not the parent drug","excerpt":"\"Reversible Fanconi Syndrome Caused by Degraded Tetracycline.\"\n","source_date":"1963-01-19","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20221218011114/https://jamanetwork.com/journals/jama/fullarticle/664082","calculation_notes":"This is the historical case series that anchors the folk fear of expired drug toxicity. Three patients taking tetracycline capsules stored past the expiration date developed nausea, vomiting, proteinuria, glucosuria, aminoaciduria and hypokalemia — the biochemical picture of acquired Fanconi syndrome. The toxic agent was an epimerization and degradation product of tetracycline (anhydrotetracycline / epi-anhydrotetracycline), not tetracycline itself. Old-formulation capsules of the 1960s contained citric acid and degraded rapidly in humid storage. Modern tetracycline formulations removed citric acid and reformulated the excipient mix; case reports of degraded-tetracycline nephrotoxicity have been effectively absent from the literature for decades. Included here because (a) this is the one well-documented episode of expired-drug toxicity in the modern pharmacology literature, and (b) it keeps the regional breakdown honest — the folk fear is not baseless, it is anchored on one specific 1960s formulation that no longer exists on the US market.\n","independence_note":"Primary JAMA case report, methodologically independent of all stability-study sources above; the degradation mechanism has been corroborated by later chemistry and subsequent case series (e.g., Gross JAMA Int Med 1963).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/10808186/","title":"Outdated EpiPen and EpiPen Jr autoinjectors: past their prime?","publisher":"Journal of Allergy and Clinical Immunology (Simons, Gu & Simons)","source_type":"peer_reviewed","statistic":"Epinephrine bioavailability from outdated EpiPen autoinjectors was significantly reduced (P<.05) vs in-date; inverse correlation of 0.63 between declining epinephrine content and months past the expiration date","excerpt":"\"Epinephrine bioavailability from the outdated EpiPen autoinjectors was significantly reduced (P <.05) compared with epinephrine bioavailability from the in-date autoinjectors. The inverse correlation between the decreased epinephrine content of the outdated autoinjectors, assessed with an HPLC-UV method, and the number of months past the expiration date was 0.63.\"\n","source_date":"2000-05-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260515174012/https://pubmed.ncbi.nlm.nih.gov/10808186/","calculation_notes":"Simons, Gu & Simons 2000 (J Allergy Clin Immunol 105:1025-30, PMID 10808186) is the foundational paper on epinephrine autoinjector degradation, predating and underlying the Cantrell 2017 replication. It found that bioavailability from outdated devices was significantly reduced and that content declined with months past expiration (inverse correlation 0.63). The authors recommend in-date autoinjectors but conclude that if the only one available is outdated \"it could be used as long as no discoloration or precipitates are apparent because the potential benefit of using it is greater than the potential risk of a suboptimal epinephrine dose or of no epinephrine treatment at all.\" This is the mechanism behind the rescue-drug exception: measurable potency loss that is clinically decisive when the dose is a single emergency injection.\n","independence_note":"Simons 2000 is the primary study that Cantrell 2017 extends; treat as methodological ancestor rather than independent evidence. Both measure HPLC content of EpiPen cartridges directly.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27221065/","title":"A systematic review of epinephrine degradation with exposure to excessive heat or cold","publisher":"Annals of Allergy, Asthma & Immunology (Parish, Bowser, Morton & Brown)","source_type":"peer_reviewed","statistic":"Across the pooled studies, heat degraded epinephrine only with prolonged exposure; no study evaluating real-world temperature fluctuations detected significant degradation","excerpt":"\"Heat exposure resulted in epinephrine degradation but only with prolonged exposure. None of the studies evaluating the effects of real-world temperature fluctuations detected significant degradation.\"\n","source_date":"2016-07-01","source_accessed":"2026-08-08","calculation_notes":"Parish et al. 2016 (Ann Allergy Asthma Immunol 117:79-87, PMID 27221065) pooled the published stability studies on epinephrine exposed to heat and cold. It is the source for this entry's storage-temperature axis: it establishes both that sustained heat does degrade a rescue-drug active ingredient and — importantly for the calm framing — that the effect requires prolonged exposure, with ordinary day-to-day temperature swings and brief excursions producing no significant degradation. This is why the heat case in the regional breakdown is scoped to sustained storage (a summer in a hot car), not a warm afternoon, and why the failure mode remains loss of potency during an emergency rather than new toxicity. The controlled-room-temperature baseline the reassuring shelf-life data assume (USP: 20-25°C, excursions 15-30°C) is the reference point against which \"stored too hot\" is defined.\n","independence_note":"Independent of the SLEP (Lyon 2006) and Cantrell datasets; a formulation-specific systematic review of temperature exposure rather than a shelf-life-extension study. It overlaps in subject (epinephrine) with the Simons and Cantrell EpiPen sources but pools a distinct evidence base focused on temperature rather than elapsed time past date.\n"}],"comparison_anchors":[{"label":"Fatal food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Pesticide residue serious harm (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Anaphylaxis fatal (lifetime, US adult)","lifetime_us_adult":0.00015}],"regional_breakdown":[{"region":"Solid tablet/capsule, 1-5 years past date, cool-dry storage","probability":0.000001,"notes":"The modal case. SLEP and Cantrell 2012 both support near-full potency with no toxicological signal. No published case reports of harm from this scenario that are not attributable to other factors (wrong drug, dose error, interaction). Point estimate is a structural upper bound, not a measured rate.\n"},{"region":"Solid tablet/capsule, 10+ years past date, ambient storage","probability":0.001,"notes":"Cantrell's 28-to-40-year samples still averaged above 90% potency on 12 of 14 ingredients, so \"harm\" here almost always means \"failed therapy\" rather than toxicity. Clinically meaningful only when the drug is being relied on for a specific effect (antibiotic course, anti-seizure level, anticoagulation) and the potency loss is large enough to matter.\n"},{"region":"Liquid oral suspension / eye drops / reconstituted pediatric antibiotic","probability":0.02,"notes":"Reconstituted amoxicillin and azithromycin suspensions have labeled in-use shelf lives measured in days to weeks, not the years implied by the manufacturing date. Opened ophthalmic drops can lose preservative efficacy and support microbial growth. Risk here is dominated by contamination / sub-therapeutic dosing, not chemical toxicity.\n"},{"region":"Nitroglycerin sublingual tablets past date or >6 months opened in warm carry","probability":0.1,"notes":"Nitroglycerin sublimes from the tablet matrix; heat and humidity accelerate loss. A patient taking a degraded tablet during angina may get no clinical effect — the \"harm\" is failure of emergency therapy, not poisoning. Stability work shows opened bottles carried in warm pockets can lose substantial potency within 2-6 months.\n"},{"region":"EpiPen / insulin / rescue inhaler past expiration","probability":0.15,"notes":"EpiPens past date retain 70-90% of labeled epinephrine on average but bioavailability drops with time; for a child at the top of the weight range the dose can be sub-therapeutic in anaphylaxis. Insulin loses activity when stored warm or past date. Rescue inhalers lose propellant and active drug delivery. In each case the failure is clinical (inadequate emergency response) rather than toxic. Sustained heat — a summer in a hot-car glovebox above 50°C — compounds this: Parish 2016's systematic review found heat degrades epinephrine with prolonged exposure, so a device stored hot can be subpotent exactly when the single emergency dose is needed.\n"},{"region":"Degraded old-formulation tetracycline (historical, pre-reformulation)","probability":0.5,"notes":"Anchored on the 1963 Frimpter JAMA series and subsequent 1960s reports of reversible Fanconi syndrome from degraded tetracycline capsules containing citric acid. Included for historical completeness; modern US-market tetracyclines were reformulated and current case reports of this specific toxicity are effectively absent from the literature. The probability applies to the pre-reformulation scenario, not to tetracycline purchased today.\n"}],"short_label":"Expired meds","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"\"Expired medication\" is a category that covers both a 3-year-old ibuprofen tablet and a 2-week-old reconstituted suspension of amoxicillin, and nothing useful can be said about both at once. Storage temperature is as decisive as the date: the reassuring shelf-life findings all assume controlled room temperature (roughly 20-25°C), and a drug baked in a hot car or kept over a steamy shower degrades on a faster clock — though a 2016 systematic review of epinephrine found the effect requires sustained heat rather than brief excursions. For solid oral dosage forms stored at room temperature in original packaging, the Shelf Life Extension Program and Cantrell's 28-to-40-year assay both show that the labeled expiration date dramatically understates real-world potency, and no meaningful toxicity signal has been reported. The specific formulations where expiration dates should be taken seriously are narrow and well-characterized: nitroglycerin sublingual tablets, injectable epinephrine (EpiPen and equivalents), insulin, rescue inhalers, biologics and vaccines (especially after reconstitution), liquid antibiotics, opened ophthalmic drops, and mefloquine — and these same formulations are the ones most vulnerable to heat. For these, the failure mode is almost always loss of potency causing clinical failure during an emergency rather than new toxicity. The one well-documented episode of expired-drug toxicity — Frimpter's 1963 tetracycline Fanconi syndrome cases — was attributable to a specific old-formulation capsule containing citric acid that is no longer on the US market. This entry does not cover veterinary compounded products, controlled-substance diversion, or non-US supply chains, all of which carry distinct risk profiles.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-8-agent","last_reviewed":"2026-08-08","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single amber prescription pill bottle rendered as a flat vector shape in muted ochre and off-white, centered on an empty pale-grey background."},"canonical_url":"https://likelier.app/expired-medications","api_url":"https://likelier.app/api/fears/expired-medications.json"},{"slug":"fast-food-cardiometabolic-risk","question":"What are the odds that regular fast food gives you diabetes or heart disease?","quick_answer":"No trustworthy lifetime disease probability exists. Regular fast food near-certainly nudges cardiometabolic risk upward, but the strongest US studies measure intermediate outcomes like weight gain and insulin resistance rather than diabetes or heart disease, and no US data isolates fast food cleanly enough to pin a number. Most people know the habit is unhealthy but cannot state a probability.\n","category":"health","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Almost nobody thinks fast food is good for them, yet frequent consumers rarely attach a concrete cardiometabolic price to the habit. The cultural framing is moral and vague (\"junk food,\" \"a treat,\" \"I really shouldn't\") rather than probabilistic, and because the harm is slow, cumulative, and entangled with weight, income, and everything else a person eats, it resists the kind of single number people can picture. The result is a split perception: the direction is common knowledge, but the magnitude sits in a fog. People who would never smoke will eat fast food several times a week without filing it under disease risk at all, partly because the evidence itself refuses to hand over a clean lifetime odds figure.\n","rough_estimate":"Most people know frequent fast food is unhealthy but cannot state a disease probability; the evidence does not support one cleanly either","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/15639678/","title":"Fast-food habits, weight gain, and insulin resistance (the CARDIA study): 15-year prospective analysis","publisher":"The Lancet (Pereira MA, Kartashov AI, Ebbeling CB, Van Horn L, Slattery ML, Jacobs DR, Ludwig DS)","source_type":"primary_study","statistic":"US young adults eating fast food >2x/week at both baseline and follow-up gained ~4.5 kg more weight and had a two-fold greater increase in insulin resistance (HOMA) over 15 years than those eating it <1x/week; cohort of 3,031 adults aged 18-30 at enrollment","excerpt":"\"Fast-food consumption has strong positive associations with weight gain and insulin resistance, suggesting that fast food increases the risk of obesity and type 2 diabetes.\"\n","source_date":"2005-01-01","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20260714062730/https://pubmed.ncbi.nlm.nih.gov/15639678/","calculation_notes":"CARDIA is the strongest US prospective evidence, and it deliberately stops short of a disease probability. Its endpoints are intermediate — body weight and HOMA insulin resistance — not incident diabetes or cardiac events. The +4.5 kg and two-fold insulin-resistance findings establish a mechanism and a direction in a US cohort but cannot be converted into a lifetime probability of type 2 diabetes or heart disease without importing effect sizes from elsewhere. That is the core reason this entry carries no reliable estimate: the best US data measures the pathway, not the outcome.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4059207/","title":"Western-Style Fast Food Intake and Cardiometabolic Risk in an Eastern Country","publisher":"Circulation (Odegaard AO, Koh WP, Yuan JM, Gross MD, Pereira MA)","source_type":"peer_reviewed","statistic":"Singapore Chinese Health Study: Western fast food ≥2x/week vs little/none carried HR 1.27 (95% CI 1.03-1.54) for incident type 2 diabetes (43,176 participants, 2,252 cases) and HR 1.56 (95% CI 1.18-2.06) for coronary heart disease mortality (52,584 participants, 1,397 deaths); ≥4x/week HR 1.79 (1.09-2.93) for CHD death","excerpt":"\"Chinese Singaporeans with relatively frequent intake of Western-style fast food items (≥2 times per week) had an increased risk of developing type 2 diabetes mellitus (hazard ratio, 1.27; 95% confidence interval, 1.03-1.54) and dying of coronary heart disease (hazard ratio, 1.56; 95% confidence interval, 1.18-2.06).\"\n","source_date":"2012-07-10","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20250331154639/https://pmc.ncbi.nlm.nih.gov/articles/PMC4059207/","calculation_notes":"Odegaard 2012 is the study that DOES report disease endpoints — T2D incidence and CHD mortality — with strong, dose-responsive hazard ratios. But it is a Singaporean Chinese cohort with a different baseline diabetes and cardiac-mortality rate, different portion norms, and a different \"Western fast food\" exposure than the US. Multiplying a Singaporean HR by a US lifetime base rate would produce a plausible-looking number resting on an untested transportability assumption — exactly the kind of cross-population synthesis this site treats as a fabricated estimate. The HRs are reported here as the directional weight of evidence, not as the basis for a headline figure.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5779659/","title":"A prospective study of frequency of eating restaurant prepared meals and subsequent 9-year risk of all-cause and cardiometabolic mortality in US adults","publisher":"PLoS ONE (Kant AK, Graubard BI)","source_type":"peer_reviewed","statistic":"NHANES 1999-2004 cohort of 9,107 US adults ≥40, mortality follow-up to 2011 (median 9.2 years): ≥3 restaurant meals/week vs <1/week gave no clear mortality association — cardiometabolic mortality HR 1.33 (0.95-1.85) in men and 0.75 (0.45-1.25) in women, all-cause HR 0.86 (0.70-1.06) men, 0.89 (0.67-1.17) women","excerpt":"\"The weekly frequency of eating restaurant prepared meals and prospective risk of mortality after 9 years were not related in this cohort.\"\n","source_date":"2018-01-24","source_accessed":"2026-07-14","archive_url":"http://web.archive.org/web/20250414155243/https://pmc.ncbi.nlm.nih.gov/articles/PMC5779659/","calculation_notes":"Included as honest counter-evidence. This US cohort found no clear link between eating out and mortality, with confidence intervals crossing 1 in every stratum. The caveats matter — \"restaurant meals\" is broader than fast food, mortality is a blunter endpoint than incident disease, and nine years is short for a cumulative exposure — but a null US result standing against a positive Singaporean one is precisely why a single lifetime probability cannot be defended here. The disagreement is the finding.\n"}],"comparison_anchors":[{"label":"Type 2 diabetes, lifetime (US adult, all causes)","lifetime_us_adult":0.33},{"label":"Extra lifetime T2D risk from a daily sugary-drink habit","lifetime_us_adult":0.08}],"short_label":"Fast food and disease","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is marked no_reliable_estimate on purpose, and the reason is a clean illustration of the site's evidence bar. The direction is not in doubt: fast food promotes weight gain and insulin resistance (CARDIA, US) and tracks with incident diabetes and cardiac death in a large Asian cohort (Odegaard, Singapore). What is missing is a US lifetime disease probability that isolates fast food from confounding. The strongest US study measures intermediate outcomes rather than disease; the study that measures disease is non-US and not obviously transportable; and the one US study of eating out and mortality found nothing clear. \"Fast food\" is also not a fixed exposure — a grilled-chicken order and a double cheeseburger with a large soda are both \"fast food,\" and the sugary-drink component alone (see sugary-drink-type-2-diabetes) carries much of the measurable diabetes signal. Rather than multiply a foreign hazard ratio by a domestic base rate and present the product as a probability, this entry reports the hazard ratios as directional evidence and declines to invent the headline number. The honest summary: regular fast food near-certainly nudges cardiometabolic risk upward, by an amount current US data cannot pin to a trustworthy lifetime figure.\n","quality_score":{"d1":4,"d2":4,"d3":5,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-07-14","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-14","last_reviewed":"2026-07-14","reviewed":true,"generated_at":"2026-07-14","image":{"alt":"A paper fast-food bag and a burger in flat silhouette, viewed straight on against a plain light background, muted red and grey tones."},"canonical_url":"https://likelier.app/fast-food-cardiometabolic-risk","api_url":"https://likelier.app/api/fears/fast-food-cardiometabolic-risk.json"},{"slug":"financial-sextortion-teen","question":"How likely is a US teenage boy to be targeted by a financial sextortion scheme?","quick_answer":"No defensible per-teenager probability exists. The widely cited figures, such as 137 financial-sextortion reports a day, are worldwide, non-deduplicated incident counts that cannot be divided by any US teen-boy population to yield a per-capita rate. Parental discourse treats it as a fast-rising, near-universal threat, but no numeric perception estimate exists.\n","category":"crime","tags":["teen","digital-fraud","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"No survey measures how likely parents or teens believe financial sextortion is for a given teenager. Coverage frames it as an explosive emerging threat, so perceived likelihood tends to run high; the available data measures reported incidents and self-reported experience, not estimated probability. The closest experiential figure is Thorn's June 2025 survey finding that 1 in 5 young people (ages 13-20) said they had experienced sextortion of any kind, but that measures lifetime experience across all forms and both sexes, not a teen boy's perceived odds of the money-demand variant.","rough_estimate":"Treated as a fast-rising, near-universal-feeling threat in parental discourse; no numeric perception estimate exists.","kind":"intuition"},"sources":[{"url":"https://www.missingkids.org/gethelpnow/cybertipline/cybertiplinedata","title":"CyberTipline Data","publisher":"National Center for Missing & Exploited Children (NCMEC)","source_type":"reputable_reference","statistic":"In 2025, NCMEC received an average of 137 reports of financial sextortion a day, a 37% increase per day over the prior year; at least three dozen US teenage boys have died by suicide as a result of this crime.","excerpt":"In 2025, NCMEC received an average of 137 reports of financial sextortion a day. Tragically, NCMEC is aware of at least three dozen teenage boys in the U.S. who have taken their lives as a result of being victimized by this crime.","source_date":"2026-03-31","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260619172532/https://www.missingkids.org/gethelpnow/cybertipline/cybertiplinedata","calculation_notes":"Flow count of incident reports, not unique victims and not US-only (platforms report globally). 137 reports/day x 365 = ~50,000 reports/year. Cannot be divided by a US teen-boy population to yield a per-capita probability because the numerator is global, duplicated across reports, and mixes reporting platforms. The ~3 dozen (36) suicides is an explicit 'NCMEC is aware of at least' floor since 2021, not an incidence rate. Used as bracketing context only; no normalized value derived.","independence_note":"NCMEC is the US national clearinghouse receiving CyberTipline reports from tech platforms; the underlying report data is partly shared with Thorn (below), so these two sources are not fully independent on the report-volume figures."},{"url":"https://www.thorn.org/research/library/financial-sextortion/","title":"Trends in Financial Sextortion: An investigation of sextortion reports in NCMEC CyberTipline data","publisher":"Thorn","source_type":"primary_study","statistic":"From August 2022 to August 2023, NCMEC received an average of 812 sextortion reports per week; 90% of victims detected in NCMEC reports were male, aged 14 to 17; more than two-thirds of reports were financially motivated.","excerpt":"90% of victims detected in NCMEC reports were male, aged 14 to 17.","source_date":"2024-06-24","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260603222124/https://www.thorn.org/research/library/financial-sextortion/","calculation_notes":"Demographic composition of detected NCMEC sextortion reports. 812 reports/week x 52 = ~42,000 reports/year over the analysis window. Confirms the teen-boy concentration but, like the NCMEC flow above, counts reports not unique US victims, so no per-capita rate is derivable. Used for the 90%-male, ages-14-17 framing.","independence_note":"Thorn's analysis is built on NCMEC CyberTipline data, so it shares the same underlying report corpus as the NCMEC source and is not an independent count."},{"url":"https://www.thorn.org/press-releases/new-data-reveals-the-devastating-toll-of-sextortion-on-kids/","title":"New data reveals the devastating toll of sextortion on kids","publisher":"Thorn","source_type":"primary_study","statistic":"In a survey of 1,200 young people aged 13-20, 1 in 5 (20%) said they had experienced sextortion; 1 in 7 youth victims (15%) said they harmed themselves in response.","excerpt":"In total, 1 in 5 teens (20%) surveyed said they had experienced sextortion.","source_date":"2025-06-24","source_accessed":"2026-06-13","calculation_notes":"Self-reported lifetime experience of ANY sextortion (predominantly image-for-image demands; only ~22% of demands were for money) among ALL young people aged 13-20, both sexes. This 20% is NOT a financial-sextortion-of-teen-boys rate and is deliberately not used as the normalized headline value; it brackets the upper end of 'any sextortion exposure' and is reported as context only. Self-harm rate 15% (1 in 7) of victims.","independence_note":"Independent national survey methodology (N=1,200), distinct from the NCMEC report-corpus analyses above."},{"url":"https://www.crimesagainstchildrenfoundation.org/blogs/sextortion-complete-guide-for-parents","title":"Sextortion: The Complete Guide for Parents — What It Is, Who's at Risk, and How to Respond","publisher":"Crimes Against Children Foundation (CACF)","source_type":"reputable_reference","statistic":"Ninety percent of detected financial sextortion victims are males ages 14 to 17; financial sextortion reports to NCMEC rose 71% in the first half of 2025; at least 36 teenage boys have died by suicide in the US since 2021 as a direct result of sextortion.","excerpt":"At least 36 teenage boys have died by suicide in the United States since 2021 as a direct result of sextortion.","source_date":"2026-03-28","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260620140438/https://www.crimesagainstchildrenfoundation.org/blogs/sextortion-complete-guide-for-parents","calculation_notes":"Corroborates the NCMEC and Thorn figures: 90% male 14-17, the 36-suicide floor since 2021, and a 71% H1-2025 report increase. All are floors/flow counts, not incidence rates. Used to triangulate the suicide-floor figure and the H1-2025 growth.","independence_note":"Secondary compilation citing NCMEC/FBI data; not an independent count, used only for corroboration and the suicide-floor figure."}],"comparison_anchors":[{"label":"Online scam loss (lifetime)","lifetime_us_adult":0.4},{"label":"Identity theft (lifetime)","lifetime_us_adult":0.6},{"label":"Vehicle theft (lifetime)","lifetime_us_adult":0.0296}],"short_label":"Teen financial sextortion","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"No defensible per-teenager probability exists for financial sextortion, which is why this entry carries no_reliable_estimate. The headline figures everyone cites are flow counts of incident reports, not rates: NCMEC's 137 financial-sextortion reports per day are submitted by tech platforms worldwide, are not de-duplicated to unique victims, and are not restricted to the United States, so they cannot be divided by any US teen-boy population to produce a per-capita risk. The ~36 (three dozen) teenage-boy suicides since 2021 is explicitly a floor (\"NCMEC is aware of at least\"), reflecting cases that came to NCMEC's attention rather than a measured count. Thorn's June 2025 survey finding that 1 in 5 young people (20%) experienced sextortion is genuine but measures ANY sextortion across all forms and both sexes among 13-20-year-olds, predominantly image-for-image demands, with only about a fifth of demands being for money; it is not a financial-sextortion-of-teen-boys rate and is reported only as upper-bound context. Demographic skew is the firmest finding: roughly 90% of detected victims are boys aged 14-17, with offenders concentrated in West African networks. Severe underreporting (shame, fear of exposure) cuts the other way, so true incidence likely exceeds reported counts. The 8D average lands at 3.9, but D3 (arithmetic) and the absence of a normalized value reflect that this is honestly an unquantifiable-rate topic, not a data-gathering failure.","quality_score":{"d1":5,"d2":4,"d3":3,"d4":3,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract phone screen with a faint shadowed figure and a coin, muted indigo and grey, flat vector."},"canonical_url":"https://likelier.app/financial-sextortion-teen","api_url":"https://likelier.app/api/fears/financial-sextortion-teen.json"},{"slug":"fish-oil-omega3-deficiency","question":"What are the odds of harm from not taking fish oil or omega-3 supplements?","quick_answer":"No meaningful probability exists. Large randomized trials found little or no cardiovascular benefit from omega-3 supplements for the general population, so the harm from skipping them cannot be quantified as a single figure. Many of the roughly 19 million US adults who take fish oil assume that not doing so carries a substantial risk of heart disease or early death.\n","category":"food","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Fish oil capsules occupy a peculiar position in the supplement pantheon: millions of adults swallow them daily out of a vague conviction that skipping the dose invites cardiovascular catastrophe. The belief is sustained by decades of observational data linking fish consumption to heart health, by aggressive supplement-industry marketing (a market worth over $2 billion in the US alone), and by the intuitive appeal of \"natural\" fats doing what statins cannot. Roughly 19 million US adults take omega-3 supplements. Many assume that not taking them constitutes a meaningful health risk.\n","rough_estimate":"a substantial chance of heart disease or early death without supplementation","kind":"intuition"},"sources":[{"url":"https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD003177.pub5/full","title":"Omega-3 fatty acids for the primary and secondary prevention of cardiovascular disease","publisher":"Cochrane Database of Systematic Reviews (Abdelhamid et al.)","source_type":"peer_reviewed","statistic":"86 RCTs, 162,796 participants: increasing long-chain omega-3 has little or no effect on all-cause mortality (RR 0.97, 95% CI 0.93-1.01) or cardiovascular events (RR 0.96, 95% CI 0.92-1.01); high-certainty evidence","excerpt":"\"Meta-analysis and sensitivity analyses suggested little or no effect of increasing LCn3 on all-cause mortality (RR 0.97, 95% CI 0.93 to 1.01; 143,693 participants; 11,297 deaths in 45 RCTs; high-certainty evidence), cardiovascular mortality (RR 0.92, 95% CI 0.86 to 0.99; moderate-certainty evidence), cardiovascular events (RR 0.96, 95% CI 0.92 to 1.01; 140,482 participants; 17,619 people experienced events in 43 RCTs; high-certainty evidence).\"\n","source_date":"2020-02-29","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20241104125813/https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD003177.pub5/full","calculation_notes":"The Cochrane review is the most comprehensive systematic review of omega-3 supplementation for cardiovascular prevention. 86 RCTs with 162,796 participants, trials lasting 12-88 months, including adults at varying cardiovascular risk in high-income countries. The primary finding — little or no effect on all-cause mortality or cardiovascular events — is rated high-certainty evidence, meaning further research is very unlikely to change the estimate. A small reduction in cardiovascular mortality (RR 0.92) reached moderate certainty but did not survive sensitivity analysis restricted to low-risk-of-bias trials. The review explicitly concludes that \"taking long-chain omega 3 supplements does not benefit heart health or reduce our risk of stroke or death from any cause.\"\n","independence_note":"Cochrane systematic review synthesizing 86 independent RCTs. Independent of the VITAL and REDUCE-IT trial teams. Cochrane reviews follow a pre-registered protocol with independent editorial oversight.\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1811403","title":"Marine n-3 Fatty Acids and Prevention of Cardiovascular Disease and Cancer","publisher":"New England Journal of Medicine (Manson et al. — VITAL Trial)","source_type":"primary_study","statistic":"25,871 US adults, median 5.3 years follow-up: omega-3 supplementation (840 mg/day) did not significantly reduce major cardiovascular events (HR 0.92, 95% CI 0.80-1.06, p=0.24)","excerpt":"\"Supplementation with n-3 fatty acids did not result in a lower incidence of major cardiovascular events or cancer than placebo. In the comparison of n-3 fatty acids with placebo, 386 participants who received n-3 fatty acids and 419 who received placebo had a major cardiovascular event (hazard ratio, 0.92; 95% confidence interval [CI], 0.80 to 1.06; P=0.24).\"\n","source_date":"2019-01-03","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20250812092431/https://www.nejm.org/doi/full/10.1056/NEJMoa1811403","calculation_notes":"VITAL was the largest primary-prevention RCT of omega-3 supplementation: 25,871 adults (men >= 50, women >= 55), randomized to 1 g/day fish oil (840 mg EPA+DHA) vs placebo in a 2x2 factorial design with vitamin D. The primary cardiovascular endpoint — major cardiovascular events (MI, stroke, CV death) — showed a non-significant 8% reduction. Secondary analyses suggested possible benefit for MI specifically (HR 0.72, nominal p=0.04) and among participants with low baseline fish intake, but these were exploratory subgroup analyses not corrected for multiple comparisons. The trial's null primary result is the relevant finding for a general-population claim about omega-3 supplementation.\n","independence_note":"VITAL was conducted at Brigham and Women's Hospital, funded by NIH. Independent of the Cochrane review team, REDUCE-IT (Amarin-funded), and STRENGTH (AstraZeneca-funded).\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1812792","title":"Cardiovascular Risk Reduction with Icosapent Ethyl for Hypertriglyceridemia","publisher":"New England Journal of Medicine (Bhatt et al. — REDUCE-IT Trial)","source_type":"primary_study","statistic":"8,179 statin-treated patients with elevated triglycerides: icosapent ethyl 4 g/day reduced major cardiovascular events by 25% (HR 0.75, 95% CI 0.68-0.83, p<0.001) vs mineral oil placebo","excerpt":"\"A primary end-point event occurred in 17.2% of the patients in the icosapent ethyl group, as compared with 22.0% of the patients in the placebo group (hazard ratio, 0.75; 95% confidence interval [CI], 0.68 to 0.83; P<0.001).\"\n","source_date":"2019-01-03","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20251125214520/https://www.nejm.org/doi/full/10.1056/NEJMoa1812792","calculation_notes":"REDUCE-IT is the outlier in the omega-3 literature — it found a large, statistically significant 25% relative risk reduction for cardiovascular events. However, it differs from general omega-3 supplementation in three critical ways: (1) it used a specific pharmaceutical formulation (icosapent ethyl, pure EPA at 4 g/day — roughly 5x the dose in VITAL); (2) the population was high-risk (statin-treated with elevated triglycerides), not the general public; (3) the mineral oil placebo raised LDL cholesterol and C-reactive protein in the control group by more than 30%, meaning some of the apparent benefit may reflect harm in the placebo arm rather than benefit from treatment. The STRENGTH trial, which used a corn oil placebo (no biomarker effects), found no cardiovascular benefit from a similar EPA+DHA formulation. The REDUCE-IT results are included here because they are frequently cited by supplement advocates, but their generalizability to over-the-counter fish oil is limited.\n","independence_note":"REDUCE-IT was funded by Amarin Pharma. Independent of VITAL (NIH-funded) and the Cochrane review. The mineral oil placebo controversy was raised by independent researchers including the STRENGTH trial investigators at Cleveland Clinic.\n"}],"comparison_anchors":[{"label":"Vegetarian nutrient deficiency (lifetime, well-planned diet)","lifetime_us_adult":0.04},{"label":"Choking death (lifetime, US adult)","lifetime_us_adult":0.00093},{"label":"Lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"short_label":"Omega-3 gap","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses the risk of harm from not taking omega-3 supplements, not from dietary omega-3 deficiency. The distinction matters: observational data linking fish consumption to cardiovascular health is stronger than the supplementation trial data, suggesting that eating fish may confer benefits that capsules do not replicate (possibly because fish displaces less healthy protein sources, or because the food matrix matters). Clinical omega-3 deficiency from diet — characterized by dermatitis, impaired wound healing, and growth retardation — is vanishingly rare in developed countries and is not what supplement marketing addresses. REDUCE-IT's positive result applies to a specific pharmaceutical product (icosapent ethyl at 4 g/day) in a high-risk population on statins, not to the fish oil capsules sold at pharmacies. The mineral oil placebo controversy remains unresolved.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A single amber fish oil capsule next to a small sardine silhouette, flat vector illustration in muted gold and blue-grey tones."},"canonical_url":"https://likelier.app/fish-oil-omega3-deficiency","api_url":"https://likelier.app/api/fears/fish-oil-omega3-deficiency.json"},{"slug":"fly-on-food-illness","question":"What are the odds of getting sick from a housefly landing on your food?","quick_answer":"There is no reliable figure for the odds of getting sick from a single fly landing on your food. Houseflies are confirmed carriers of many pathogens and do transfer bacteria on contact, but no study measures what fraction of fly-contacted meals actually make someone ill, so any single number would be invented.\n","category":"health","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"The sight of a housefly landing on food triggers an immediate disgust response in most people, and the intuition behind it is not entirely wrong — flies do visit garbage, feces, and decaying matter before landing on a plate. Popular health writing frequently describes houseflies as carrying \"over 100 pathogens\" and warns that they defecate every few minutes on any surface they contact. The strong folk inference is that a fly landing on food makes that food unsafe to eat, or at least meaningfully more dangerous. Whether that inference is calibrated — whether the probability of actual illness from a single fly-contacted meal is meaningfully above baseline — is the question this entry examines.\n","kind":"intuition"},"sources":[{"url":"https://link.springer.com/article/10.1186/s12889-018-5934-3","title":"A systematic review of human pathogens carried by the housefly (Musca domestica L.)","publisher":"BMC Public Health","source_type":"peer_reviewed","statistic":"Over 130 pathogens identified across 99 studies; bacteria (including Salmonella, E. coli, Staphylococcus aureus) predominate, with flies acting as mechanical vectors by surface transfer rather than biological amplification","excerpt":"\"Over 130 pathogens, predominantly bacteria (including some serious and life-threatening species) were identified from the house flies [...] House flies carry a large number of pathogens which can cause serious infections in humans and animals.\"\n","source_date":"2018-08-28","source_accessed":"2026-05-03","archive_url":"http://web.archive.org/web/20260504055048/https://link.springer.com/article/10.1186/s12889-018-5934-3?","calculation_notes":"This systematic review of 1718 titles, yielding 99 included papers, is the most comprehensive survey of which pathogens have been isolated from Musca domestica. It establishes the bioburden half of the question — what pathogens flies carry — but does not address the other half: what fraction of fly-to-food contacts result in a transferred dose sufficient to cause illness in a human consumer. Used here as the primary evidence that houseflies are confirmed mechanical vectors of a wide range of human pathogens, which grounds the perceived fear in biological reality. The per-exposure illness probability question remains unanswered.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6977770/","title":"Quantitative Analysis of Houseflies-mediated Food Contamination with Bacteria","publisher":"Food Safety (Tokyo) — National Institute of Health Sciences, Japan","source_type":"primary_study","statistic":"Approximately 10³ CFU transferred per fly landing; food contaminated to 3.6 × 10⁴–1.7 × 10⁵ CFU/g within 60 minutes of fly access — but no illness outcome was measured and no human infectious-dose comparison was performed","excerpt":"\"approximately 10³ CFU were transmitted from flies per landing [...] at 60 min, the fed strain contaminated the food at a range of 3.6 × 10⁴–1.7 × 10⁵ CFU/g [...] food-borne pathogenic bacteria, such as E. coli O157, can infect humans with an extremely small number of organisms\"\n","source_date":"2019-03-29","source_accessed":"2026-05-03","archive_url":"https://web.archive.org/web/20260505054508/https://pmc.ncbi.nlm.nih.gov/articles/PMC6977770/","calculation_notes":"This Japanese laboratory study (Fukuda et al., 2019) is the most directly relevant quantitative data on transfer dose: roughly 1,000 CFU per landing event on clean food under controlled conditions. The authors note that E. coli O157:H7 has a low infectious dose (as few as 10–100 CFU for susceptible populations), which would place the per-landing transfer in a potentially concerning range for that specific pathogen. However, the study does not close the loop to illness outcomes: it does not measure whether the transferred dose survives to consumption, whether consumer- level food handling (reheating, acid in dressings, stomach acid) kills the load, or what fraction of real-world fly contacts involve a pathogen- carrying fly at all. Used here as the best available quantitative bridge between bioburden (pathogen on fly) and exposure (pathogen on food), while acknowledging it does not yield a per-meal illness probability.\n","independence_note":"This study was conducted by Fukuda et al. at Azabu University/National Institute of Health Sciences, Japan, and is independent of the BMC Public Health systematic review above. Both cover pathogen presence/transfer but neither produces an illness rate per exposure event.\n"}],"comparison_anchors":[{"label":"Annual foodborne illness risk, any cause (US adult, ~1 in 6)","lifetime_us_adult":0.167},{"label":"Salmonella infection, any exposure, per year (US — ~1.35 million cases/year)","lifetime_us_adult":0.004}],"short_label":"Fly on food","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The gap between \"fly carries pathogens\" and \"fly causes illness\" involves at least three unmeasured steps: (1) what fraction of household houseflies in a wealthy-country setting actually carry a sufficient load of a human-pathogenic organism (rather than environmental bacteria); (2) what fraction of the transferred dose survives to the point of ingestion given food handling, cooking, or acidic environments; and (3) what fraction of infectious-dose exposures produce symptomatic illness given normal immune function and stomach acid as a first-line defence. The bioburden literature (step 1) is rich; the transfer literature (bridging steps 1 and 2) is thin; the per-exposure illness literature (step 3) is essentially absent for the household setting in wealthy countries. High-burden settings — open-air markets, field latrines near food preparation, livestock farms — present a different picture, and epidemiological studies in those contexts do implicate flies in diarrheal disease transmission. The household scenario in a setting with running water, refrigeration, and routine cooking is a meaningfully different risk environment than a field latrine adjacent to an outdoor food stall.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"claude-sonnet-4-6","last_reviewed":"2026-05-03","reviewed":true,"generated_at":"2026-05-03","image":{"alt":"A single housefly perched on a plain white plate, flat vector editorial illustration on muted background."},"canonical_url":"https://likelier.app/fly-on-food-illness","api_url":"https://likelier.app/api/fears/fly-on-food-illness.json"},{"slug":"frontline-support-casualty","question":"What is the probability of a military support or logistics soldier being killed or seriously wounded over five years in a Ukraine-scale conventional conflict?","quick_answer":"There is no reliable per-capita casualty rate for support and logistics troops in a Ukraine-scale war. Older conflicts give only shares of total casualties, not individual rates, and long-range missiles and FPV drones have reshaped rear-area risk in ways historical figures cannot capture. Support roles are safer than the front line, but by a smaller margin than most people assume.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Civilians and prospective military personnel generally assume that support and logistics roles -- supply drivers, mechanics, communications specialists, medics, and rear-area administrators -- are substantially safer than front-line infantry. This intuition is broadly correct in conventional wars where rear areas are physically separate from the front. In a Ukraine-scale conflict, however, long-range artillery, ballistic missiles, and first-person-view (FPV) drones have increasingly penetrated rear areas, attacking logistics depots, training sites, command posts, and supply convoys at distances of 50-200 kilometers from the front line. No high-quality survey isolates public estimates of support-troop casualty rates, so the perceived side is editorial intuition.\n","rough_estimate":"most people assume support troops face perhaps 1 in 10 to 1 in 20 the risk of frontline infantry; historical data from the 1982 Lebanon War suggests a smaller gap of roughly 2-3:1 for conventional wars, and Ukraine-era drone warfare may have narrowed it further","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8963150/","title":"Review of Military Casualties in Modern Conflicts -- The Re-emergence of Casualties From Armored Warfare","publisher":"Military Medicine (Oxford Academic / PMC)","source_type":"peer_reviewed","statistic":"In the 1982 Lebanon War, support units accounted for 21.1% of casualties in urban combat and 16.6% in non-urban combat, compared to infantry at 56% and 43% respectively.","excerpt":"\"In the 1982 Lebanon War, infantry suffered the highest fraction of battlefield injuries in both urban and non-urban combat, at 56% and 43% respectively, followed by armor divisions at 17.1% and 34.8%, and support units at 21.1% and 16.6%.\"\n","source_date":"2022-01-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20250225232247/https://pmc.ncbi.nlm.nih.gov/articles/PMC8963150/","calculation_notes":"This study provides share-of-total-casualties data by role category, not per-capita rates. The support-unit share (16-21%) relative to infantry share (43-56%) shows that support troops were somewhat less likely to be casualties, but the ratio is much less favorable than civilian intuition would suggest. The paper cannot be used to derive a per-capita support-troop rate for Ukraine because: (a) the 1982 Lebanon War used different doctrine and technology, (b) the troop mix was different, and (c) the rise of long-range precision weapons and FPV drones in Ukraine has specifically targeted logistics and support infrastructure. This source establishes the historical baseline but does not provide a reliable modern estimate.\n","independence_note":"Military Medicine is a peer-reviewed journal of the Association of Military Surgeons of the United States; the study is independent of Ukrainian or Russian government data.\n"},{"url":"https://en.wikipedia.org/wiki/Tooth-to-tail_ratio","title":"Tooth-to-tail ratio","publisher":"Wikipedia (citing military doctrine and historical analysis)","source_type":"reputable_reference","statistic":"In the 2005 Iraq deployment, combat troops comprised only 11% of personnel (a tooth-to-tail ratio of 1:8.1), down from 19% in WWII (1:4.3) -- the large majority of a modern force is logistics and support.","excerpt":"\"The tooth-to-tail ratio (T3R), in military jargon, is the number of military personnel it takes to supply and support ('tail') each combat soldier ('tooth'). [...] WW2: 19% combat proportion, ratio 1 : 4.3 [...] Iraq 2005: 11% combat proportion, ratio 1 : 8.1.\"\n","source_date":"2024-01-01","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260430142214/https://en.wikipedia.org/wiki/Tooth-to-tail_ratio","calculation_notes":"The tooth-to-tail ratio documents the relative size of support versus combat forces but does not provide per-capita casualty rates for support troops in high-intensity peer-on-peer warfare. Ukraine is understood to operate with a much lower tooth-to-tail ratio than the modern US military (Ukraine's acute manpower shortage has pushed more personnel toward direct combat roles), making US-derived support/combat troop ratios even less applicable to the Ukraine context. Cited as structural context, not as a basis for a numeric estimate.\n","independence_note":"Sourced from military doctrine and historical records; independent of the PMC military medicine study.\n"},{"url":"https://www.rnbo.gov.ua/en/Diialnist/7370.html","title":"Results of Ukraine's Defense Industry in 2025: FPV Drones","publisher":"National Security and Defense Council of Ukraine (RNBO)","source_type":"govt_report","statistic":"60% of Russian army losses are inflicted through the use of FPV drones (stated January 2026)","excerpt":"\"Today, 60% of the losses of the Russian army are inflicted precisely through the use of FPV drones.\"\n","source_date":"2026-01-23","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260620094218/https://www.rnbo.gov.ua/en/Diialnist/7370.html","calculation_notes":"Cited as evidence that long-range and FPV drone strikes have become the dominant casualty mechanism in the conflict, supporting the body's point that rear-area and logistics targets are no longer geographically insulated the way historical infantry-to-support ratios assume. This is a Russian-side casualty-attribution figure (Ukrainian casualty data broken down by role or mechanism is not published), used as the best available public evidence for the general trend; it is not used to derive a per-capita support-troop rate, which this entry flags as unquantifiable from public data.\n","independence_note":"RNBO is an official Ukrainian government body; this is Kyiv's own assessment of Russian losses, independent of the PMC military-medicine study and the tooth-to-tail source above.\n"}],"comparison_anchors":[],"short_label":"Support troop casualty","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"A reliable per-capita casualty rate for support and logistics personnel in a Ukraine-scale conflict cannot be extracted from currently available data. The available evidence establishes several facts that are not sufficient for a quantitative estimate: (1) historically, support troops in conventional wars have faced substantially lower per-capita casualty rates than frontline infantry; (2) the 1982 Lebanon War showed support units accounting for 16-21% of total casualties by branch, a smaller share than their estimated troop proportion; (3) FPV drones, long-range artillery, and ballistic missiles in the Ukraine conflict have specifically targeted logistics depots, supply convoys, fuel storage, and repair facilities at significant depth behind the front, eroding the traditional geographic protection that made support roles safer; (4) the total Ukrainian military casualty figures (46,000+ killed, 380,000+ wounded across roughly 1 million total personnel) are not disaggregated by role, making it impossible to isolate the support-troop rate from the overall force rate. The overall Ukrainian force killed+permanently wounded rate across all roles over three years is approximately 23-28% (using Zelenskyy's figures with 300,000 on the front and ~750,000 in support/rear roles), suggesting that even rear-area personnel face meaningfully elevated risk compared to peacetime, but the per-capita rate for support roles specifically remains unquantifiable from public data. This entry is flagged no_reliable_estimate because applying historical infantry-to-support ratios from the 1982 Lebanon War or WWII to the Ukraine drone-warfare context would launder a 40-year-old figure into a number that the underlying data cannot support.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"war-research-agent-2026-05-09","last_reviewed":"2026-05-09","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"A supply truck with a simple flat cargo bed on a road, no driver visible, flat vector illustration in muted olive and grey tones."},"canonical_url":"https://likelier.app/frontline-support-casualty","api_url":"https://likelier.app/api/fears/frontline-support-casualty.json"},{"slug":"fruit-pit-seed-cyanide","question":"What are the odds of cyanide poisoning from eating fruit pits or seeds — apple seeds, cherry pits, apricot kernels?","category":"food","tags":["food","kids"],"no_reliable_estimate":true,"perceived":{"description":"The fear rests on a true premise and then runs past it. Apple seeds, cherry and apricot pits, and other stone-fruit kernels genuinely contain amygdalin, a compound the body converts to hydrogen cyanide. From there the folk version collapses two very different situations into one: the child who swallows a few apple seeds by accident, and the adult who deliberately eats raw apricot kernels by the handful as a supplement. The first is effectively a non-event — an intact seed passes through undigested. The second is the only version with documented emergency-room cases. Popular coverage rarely separates \"I swallowed a seed\" from \"I chewed thirty kernels\", so a real but dose-dependent poison gets flattened into a fear of ordinary fruit.\n","rough_estimate":"many people assume swallowing apple seeds or a cherry pit carries a meaningful risk of cyanide poisoning","kind":"intuition"},"sources":[{"url":"https://www.efsa.europa.eu/en/press/news/160427","title":"Apricot kernels pose risk of cyanide poisoning","publisher":"European Food Safety Authority (EFSA)","source_type":"reputable_reference","statistic":"Adults could eat three small raw apricot kernels (370 mg) before exceeding the acute reference dose (ARfD) of 20 µg/kg bw; for toddlers the amount is about half of one small kernel; lethal cyanide dose is 0.5–3.5 mg/kg bw","excerpt":"\"adults could consume three small apricot kernels (370mg), without exceeding the ARfD\" ... \"For toddlers the amount would be 60mg which is about half of one small kernel\" ... \"a safe level for a one-off exposure (known as the Acute Reference Dose, or 'ARfD') of 20 micrograms per kilogram of body weight\" ... \"Studies indicate 0.5 to 3.5 milligrams (mg) cyanide per kilogram of body weight can be lethal\"\n","source_date":"2016-04-27","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260604121835/https://www.efsa.europa.eu/en/press/news/160427","calculation_notes":"EFSA's 2016 scientific opinion is the anchor for the dose framing. The ARfD (20 µg/kg) is a conservative one-off-exposure ceiling that already bakes in uncertainty factors (1.5 toxicokinetic × 3.16 toxicodynamic), so exceeding it is not the same as being poisoned — it is the point at which a safety margin starts to erode. The threshold of \"three small kernels\" for an adult versus \"half a kernel\" for a toddler is the load-bearing number: it shows how narrow the margin is specifically for deliberately eaten raw apricot kernels, which are far richer in amygdalin than the seeds people swallow by accident. The lethal cyanide range (0.5–3.5 mg/kg) is used to bound the fatal-dose scenarios in the regional breakdown.\n","independence_note":"EFSA expert-panel risk assessment; independent of the clinical case reports and the Molecules review cited below.\n"},{"url":"https://www.poison.org/articles/i-swallowed-a-cherry-pit-184","title":"Are cherry pits really poisonous?","publisher":"Poison Control (National Capital Poison Center)","source_type":"reputable_reference","statistic":"An intact, un-chewed cherry pit passes through the digestive tract without releasing cyanide; small unintentional ingestions of intact stone-fruit pits generally do not cause harm","excerpt":"\"If the cherry pit has not been chewed or broken open, it should pass intact through the digestive tract.\" ... \"Small, unintentional ingestions of intact stone fruit pits generally do not cause harm.\" ... \"Crushing or chewing pits before swallowing can result in the release of cyanide.\"\n","source_date":"2019-01-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260607195206/https://www.poison.org/articles/i-swallowed-a-cherry-pit-184","calculation_notes":"The National Capital Poison Center's lay-facing guidance is the clearest statement of the mechanism that defuses the everyday fear: amygdalin sits inside a hard seed coat, and cyanide is only released when that coat is mechanically broken by chewing or crushing. A pit swallowed whole is inert cargo. This source underwrites the \"swallowed a few whole seeds\" scenario as a structural floor rather than a real risk.\n","independence_note":"US poison-control editorial guidance; independent of EFSA and the primary clinical literature.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8069783/","title":"Amygdalin: Toxicity, Anticancer Activity and Analytical Procedures for Its Determination in Plant Seeds","publisher":"Molecules (MDPI) / Jaszczak-Wilke et al.","source_type":"peer_reviewed","statistic":"Apple seeds contain roughly 0.28–1.40 mg amygdalin per gram; hydrogen cyanide is released when plant tissue is macerated (crushed/chewed); adult lethal dose of amygdalin ~0.5–3.5 mg/kg bw","excerpt":"\"The HCN release occurs when tissues of a cyanogenic plant are macerated, e.g., when eaten by herbivores, resulting in contact of CGs with enzymes that hydrolyze them.\" ... \"The adult lethal dose of amygdalin is estimated to be 0.5–3.5 mg/kg body weight.\"\n","source_date":"2021-04-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260514112639/https://pmc.ncbi.nlm.nih.gov/articles/PMC8069783/","calculation_notes":"Provides the per-seed amygdalin content that converts the abstract \"seeds contain cyanide\" claim into a countable quantity. At ~0.28–1.40 mg amygdalin per gram of apple seed, and seeds weighing on the order of 0.7 g, a single seed carries roughly 0.2–1 mg of amygdalin; only about 6% of that mass (the HCN fraction of the molecule) becomes cyanide, so on the order of 0.01–0.06 mg cyanide per thoroughly chewed seed. Against a low-end lethal dose near 35 mg for a 70 kg adult (0.5 mg/kg), that implies many hundreds of finely chewed seeds in one sitting to approach danger. The widely quoted \"~200 apple seeds\" figure assumes a higher amygdalin content than this source reports; either way the count is far beyond any incidental ingestion. Confirms the maceration requirement independently of the poison-control source.\n","independence_note":"Peer-reviewed analytical-chemistry review; independent of EFSA and the clinical cases.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/9832674/","title":"Acute cyanide toxicity caused by apricot kernel ingestion","publisher":"Annals of Emergency Medicine / Suchard, Wallace & Gerkin","source_type":"primary_study","statistic":"A 41-year-old woman became weak and dyspneic within 20 minutes of eating health-food-store apricot kernels; she was comatose and hypothermic on presentation and responded to antidotal therapy for cyanide poisoning","excerpt":"\"A 41-year-old woman ingested apricot kernels purchased at a health food store and became weak and dyspneic within 20 minutes.\" ... \"The patient was comatose and hypothermic on presentation but responded promptly to antidotal therapy for cyanide poisoning. She was later treated with a continuous thiosulfate infusion for persistent metabolic acidosis.\"\n","source_date":"1998-12-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260416122221/https://pubmed.ncbi.nlm.nih.gov/9832674/","calculation_notes":"A single documented adult case establishing that the deliberate-consumption regime is not hypothetical: raw apricot kernels sold as a health food produced life-threatening cyanide poisoning within twenty minutes. Case reports cannot yield an incidence rate, but they mark the scenario where the fear is warranted — high-dose intentional ingestion of marketed kernels, not incidental contact with fruit.\n","independence_note":"Independent clinical case report (Ann Emerg Med, 1998); distinct institution and methodology from the EFSA assessment and the review articles.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7009189/","title":"Evaluation of the health risks related to the presence of cyanogenic glycosides in foods other than raw apricot kernels","publisher":"EFSA Journal (2019)","source_type":"reputable_reference","statistic":"Estimated mean dietary acute exposures to cyanide did not exceed the ARfD of 20 µg CN/kg bw in any age group across normal consumption of stone fruits and other cyanogenic foods","excerpt":"\"Estimated mean dietary acute exposures did not exceed the ARfD of 20 μg CN/kg bw in any age group.\" ... \"Consumption of 35–400 g canned stone fruits containing the respective EU ML of 5 mg total cyanide/kg leads to an exposure equivalent to the ARfD.\"\n","source_date":"2019-03-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260208092232/https://pmc.ncbi.nlm.nih.gov/articles/PMC7009189/","calculation_notes":"EFSA's 2019 follow-up covers cyanogenic glycosides in ordinary foods — canned stone fruit, marzipan, cassava, linseed — rather than concentrated raw kernels. Its central finding is that normal dietary intake does not exceed the acute reference dose for any age group; only high-percentile consumption of specific processed items approaches it, and even then with conservative bioavailability assumptions. This is the quantitative basis for treating everyday fruit consumption as no meaningful cyanide risk.\n","independence_note":"Separate EFSA opinion (2019) with a distinct food scope and exposure dataset from the 2016 apricot-kernel assessment.\n"}],"comparison_anchors":[{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Swallowing a few whole apple seeds or a cherry pit","probability":1e-7,"notes":"Structural floor, not a measured rate. An intact seed coat is not broken down by the gut, so amygdalin is never exposed to the enzymes that release cyanide — the seed passes through undigested. Poison Control's standing guidance is that small unintentional ingestions of intact pits generally do not cause harm.\n"},{"region":"Chewing a small handful (5–10) of apple seeds","probability":0.00001,"notes":"Chewing does release cyanide, but the dose is trivial at this quantity. A lethal cyanide dose requires finely chewing many hundreds of apple seeds in one sitting; a handful sits far below any symptomatic threshold. Included to show the mechanism is real without implying everyday risk.\n"},{"region":"Eating raw apricot kernels as a marketed 'supplement' (10+/day)","probability":0.05,"notes":"This is where the margin narrows. EFSA found that promoted intakes exceeded the acute reference dose 3–71× in adults, and an adult crosses the (conservative) ARfD at about three small raw kernels. ARfD exceedance is not the same as poisoning — the dose carries built-in safety factors — but sustained high-dose kernel consumption is the regime where symptomatic toxicity becomes plausible rather than theoretical.\n"},{"region":"Laetrile / amygdalin megadose (alternative cancer therapy)","probability":0.5,"notes":"The genuinely dangerous scenario. Oral amygdalin or apricot-kernel regimens taken as anti-cancer therapy have produced severe, documented cyanide poisoning — coma, seizures, and deaths — including in children given the treatment. At these deliberate doses, serious toxicity is common rather than rare. This is a treatment choice, not an accidental exposure to fruit.\n"}],"short_label":"Fruit-pit cyanide","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"recoverable_injury","valence":"negative","caveats":"Three separate situations travel under one fear and must be kept apart. (a) Incidental ingestion — swallowing a few whole apple seeds or a cherry pit — releases essentially no cyanide, because the amygdalin stays locked inside an intact seed coat; this is the version most people actually worry about, and it is a non-event. (b) Deliberately chewing seeds matters only in quantity: the cyanide is real but the lethal dose is on the order of hundreds of finely chewed apple seeds, far beyond anything a person eats by mistake. (c) Concentrated raw apricot kernels sold as health foods or laetrile/amygdalin \"cancer cures\" are the only regime with emergency-room and fatality case reports, and the toddler threshold (about half a kernel to exceed the acute reference dose) is far lower than the adult one. EFSA's numbers describe apricot kernels specifically, which are far richer in amygdalin than apple seeds; do not read the \"three kernels\" figure as a limit on ordinary fruit. Finally, the childhood version of the fear — that a swallowed watermelon or other seed will sprout inside you — is pure folklore: stomach acid, darkness, and the absence of soil make germination impossible.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-07-04","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-04","last_reviewed":"2026-07-04","reviewed":true,"generated_at":"2026-07-04","image":{"alt":"A single stone-fruit pit rendered as a flat vector shape in muted warm tones against a pale background."},"canonical_url":"https://likelier.app/fruit-pit-seed-cyanide","api_url":"https://likelier.app/api/fears/fruit-pit-seed-cyanide.json"},{"slug":"gardening-without-ppe","question":"What are the odds of being harmed by gardening without gloves, a mask, or protective gear?","quick_answer":"There is no reliable overall figure for being harmed by gardening without gloves or a mask. No study measures the aggregate per-session or lifetime risk for a typical hobbyist, and the individual pathways that have been measured, such as tetanus and soil-borne Legionella, are each rare and do not add up to a single publishable number.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"Hobby gardeners tend to treat bare-handed work as either obviously fine (\"people have gardened without gloves for ten thousand years\") or vaguely dangerous in a diffuse, undifferentiated way (\"soil is full of germs\"). Very few intuitions are pathway-specific. The fears that do get named — tetanus from a rusty trowel, \"something from the soil\", pesticide residue on skin — do not map cleanly onto the short list of harms that garden-related medicine actually sees, and the one route with the best evidence (aerosolized potting mix and Legionella longbeachae) almost never surfaces unprompted.\n","rough_estimate":"usually framed as either 'negligible' or 'some unspecified soil risk' — rarely quantified","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/75/ss/ss7501a1.htm","title":"Tetanus Surveillance — United States, 2009–2023","publisher":"US Centers for Disease Control and Prevention (MMWR)","source_type":"govt_report","statistic":"402 tetanus cases and 37 deaths in the US over 2009–2023 (mean 26.8 cases/year); 86.8% involved an acute wound, 61.2% of which were puncture wounds","excerpt":"\"During 2009–2023, a total of 402 tetanus cases and 37 associated deaths were reported… The mean annual tetanus incidence was 0.08 cases per 1 million population… Among 357 patients with information on preceding injury, 310 (86.8%) had an acute wound… Among 258 patients with wound-type information, puncture wounds were most common (158; 61.2%), followed by lacerations (51; 19.8%).\"\n","source_date":"2026-01-15","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420041231/https://www.cdc.gov/mmwr/volumes/75/ss/ss7501a1.htm","calculation_notes":"Sets the outer bound for the tetanus pathway in the regional breakdown. At ~27 cases/year across the entire US population of ~260M adults, the annual incidence is on the order of 1 in 10,000,000; over a 59-year adult lifetime this is roughly 6 in 10,000,000 or ~6e-6. The surveillance report does not break cases out by activity (gardening vs construction vs farming vs IV drug use), but puncture wounds on extremities dominate the wound types, which is consistent with the activities — including gardening — that produce them. Tetanus in the US is overwhelmingly a disease of the under-vaccinated; vaccination status, not glove use, is the dominant protective factor.\n","independence_note":"CDC/MMWR national surveillance data, editorially independent of the Legionella and sporotrichosis sources and based on a different reporting pipeline (NNDSS).\n"},{"url":"https://www.cdc.gov/sporotrichosis/about/index.html","title":"Sporotrichosis Basics","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Sporotrichosis is acquired via contact with soil, sphagnum moss, rose bushes, and hay; US incidence is consistent with 1–2 cases per million per year per older estimates","excerpt":"\"The fungus that causes sporotrichosis, Sporothrix, lives in the environment on soil and plants like sphagnum moss, rose bushes, and hay… Contact with plant matter increases risk for infection. This includes activities and occupations that involve gardening, forestry work, and baling hay.\"\n","source_date":"2024-04-24","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260421185557/https://www.cdc.gov/sporotrichosis/about/index.html","calculation_notes":"CDC identifies the exact pathway the \"rose gardener's disease\" moniker refers to. Published US incidence estimates (see the Medscape/StatPearls reviews citing laboratory surveillance studies) cluster around 1–2 cases per million per year. Over a 59-year adult lifetime at ~1.5/1,000,000 per year, the cumulative hobbyist risk is on the order of 9 in 100,000, i.e. ~1e-4. That is the figure used in the sporotrichosis subgroup of the regional breakdown. Gloves reduce but do not eliminate the risk since thorns frequently penetrate thin cotton gardening gloves.\n","independence_note":"Separate CDC reference page based on the fungal-diseases program, independent of the tetanus and Legionella surveillance pipelines.\n"},{"url":"https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4934a1.htm","title":"Legionnaires' Disease Associated With Potting Soil --- California, Oregon, and Washington, May–June 2000","publisher":"US Centers for Disease Control and Prevention (MMWR)","source_type":"govt_report","statistic":"First documented US cluster of Legionella longbeachae pneumonia linked to commercial potting soil; 3 cases reported in California, Oregon and Washington in May–June 2000","excerpt":"\"Infections with one species, Legionella longbeachae, have been associated with gardening and use of potting soil in Australia and Japan… L. longbeachae was isolated from one potting soil sample [from a patient's residence]… The association of L. longbeachae infection with potting soil… suggests that transmission from potting soil has occurred for the first time in the United States.\"\n","source_date":"2000-09-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260318051540/https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4934a1.htm","calculation_notes":"Anchors the Legionella longbeachae pathway in the regional breakdown. In the US, L. longbeachae is a rare minority species (~5% of legionellosis), and legionellosis as a whole is itself uncommon, so the absolute US incidence of potting-mix-linked L. longbeachae pneumonia is in the low dozens of cases per year at most — the subgroup rate cited is structural rather than measured. In Australia, where pine-bark-based potting mix is contaminated at much higher rates, case rates are 10-100x higher and mask use during potting is public-health guidance.\n","independence_note":"MMWR cluster report, independent of the O'Connor case-control study and the Whiley & Bentham review, which analyse Australian data on the same pathogen.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2870547/","title":"Does using potting mix make you sick? Results from a Legionella longbeachae case-control study in South Australia","publisher":"Epidemiology and Infection (PMC / O'Connor et al.)","source_type":"peer_reviewed","statistic":"Recent use of potting mix associated with L. longbeachae infection (OR 4.74, 95% CI 1.65–13.55) in bivariate analysis; poor hand hygiene and proximity to dripping hanging pots remained predictive after adjustment","excerpt":"\"Recent use of potting mix was associated with illness (OR 4·74, 95% CI 1·65–13·55, P=0·004) in bivariate analysis… eating or drinking after gardening before washing one's hands was associated with an increased likelihood of illness… being near dripping hanging pots… remained statistically significant.\"\n","source_date":"2007-06-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260426201350/https://pmc.ncbi.nlm.nih.gov/articles/PMC2870547/","calculation_notes":"Provides the mechanistic basis for the \"mask and/or hand-washing is load-bearing for potting-mix exposure\" framing in the breakdown. The study is Australian, and the absolute subgroup probability for US gardeners is substantially lower because US potting mix contains much less Legionella longbeachae than Australian pine-bark mix — but the pathways (inhalation of aerosol from dripping pots, hand-to-mouth contamination) are identical. This is the one pathway where a mask and hand-washing after potting work have a clear published risk-reduction rationale.\n","independence_note":"Independent South Australian case-control study; different authorship, different population, different methodology from the CDC MMWR and Whiley & Bentham review.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3377390/","title":"Legionella longbeachae and Legionellosis","publisher":"Clinical Microbiology Reviews (Whiley & Bentham, via PMC)","source_type":"peer_reviewed","statistic":"L. longbeachae causes ~5% of legionellosis in Europe and the US but is as common as L. pneumophila in Australia, New Zealand, and Japan; commercial potting mix is the principal source","excerpt":"\"In Australia, New Zealand, and Japan, reported cases of L. longbeachae infection occur as often as cases of L. pneumophila infection… the major source of human infection is considered to be commercial potting mixes and other decomposing materials, such as bark and sawdust… Gardening activities and use of potting mixes are risk factors that are so far unique to L. longbeachae infection.\"\n","source_date":"2011-04-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20250212075847/https://pmc.ncbi.nlm.nih.gov/articles/PMC3377390/","calculation_notes":"Supplies the US-vs-Australia calibration. L. longbeachae is a real pathway in both countries but the per-gardener probability in the US is roughly an order of magnitude below the Australian figure because commercial potting mixes differ in composition. Used to justify setting the US potting-mix Legionella subgroup rate below the Australian rate in the regional breakdown rather than extrapolating Australian data directly.\n","independence_note":"Peer-reviewed review article, editorially independent of the CDC cluster report and the O'Connor case-control study, although it synthesises both.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/40449067/","title":"Gardening hazards to vision: Patterns of eye injuries in emergency care","publisher":"American Journal of Emergency Medicine (via PubMed)","source_type":"peer_reviewed","statistic":"~9,326 US ED visits (95% CI 7,001–11,652) for gardening-tool-related eye injuries over 2014–2023, approximately 933/year","excerpt":"\"A total of 218 cases represented an estimated 9326 ED visits (95% CI: 7001–11,652)… Common diagnoses included contusions/corneal abrasions (53.7%), foreign bodies (13.3%), and subconjunctival hemorrhage/hyphema (10.1%)… a 132.2% increase occurred from 2020 to 2023.\"\n","source_date":"2025-06-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260505055015/https://pubmed.ncbi.nlm.nih.gov/40449067/","calculation_notes":"Anchors the \"eye injury from gardening without goggles\" pathway. ~933 ED visits per year across ~260M US adults is on the order of 3.6 per million per year. A 59-year adult lifetime at that incidence gives roughly 2 in 10,000, or ~2e-4. This understates total eye-injury incidence since it excludes non-ED care and lawnmower-only cases (which a companion NEISS study puts at ~4,500 ED visits/year for 2018–2023). Gardening-tool ED visits plus projectile mower injuries combined are the single most frequent acute harm from unprotected yard work.\n","independence_note":"NEISS-based epidemiology, independent of the infectious-disease sources and relying on the CPSC sample-survey pipeline rather than NNDSS.\n"}],"comparison_anchors":[{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"Tetanus (US, adequately vaccinated adult)","probability":0.000001,"notes":"~27 US tetanus cases/year across ~260M adults; vast majority in unvaccinated or under-vaccinated individuals. For a gardener with a current Tdap/Td booster (within 10 years) this is the dominant risk-modifier, not gloves. Bare-handed work with a thorn prick in a vaccinated adult is effectively negligible risk.\n"},{"region":"Tetanus (US, unvaccinated or booster >20 years, puncture wound)","probability":0.0005,"notes":"Most of the annual case count concentrates in this subgroup; the puncture-wound dominance in the CDC surveillance data maps directly onto the \"bare hands + rusty nail + last tetanus shot was decades ago\" scenario that popular culture names correctly for once. Order-of-magnitude estimate; surveillance data does not break out activity type.\n"},{"region":"Sporotrichosis (US hobby gardener, lifetime)","probability":0.0001,"notes":"CDC estimates US incidence of sporotrichosis at roughly 1–2 cases per million per year, concentrated in gardeners, florists, and nursery workers handling roses, sphagnum moss, or hay bales. Most infections are cutaneous and treatable with itraconazole; disseminated disease is rare and largely limited to immunocompromised patients.\n"},{"region":"Legionella longbeachae pneumonia from potting mix (US, per gardening year)","probability":0.00001,"notes":"Rare in the US because commercial potting mixes here are less contaminated than Australian pine-bark mixes. The pathway is real — the 2000 California/Oregon/ Washington cluster documented it — but US absolute incidence is in the low dozens of cases per year at most. This is the one pathway where a dust mask and post-gardening hand-washing have a published risk-reduction rationale rather than folk-wisdom rationale.\n"},{"region":"Legionella longbeachae pneumonia from potting mix (Australia, per gardening year)","probability":0.0001,"notes":"10-100x the US rate. Australian state health departments actively warn gardeners to wear masks when opening potting-mix bags, moisten mix before use, and wash hands afterwards. If this entry were normalised to the Australian rather than US population the headline pathway would shift.\n"},{"region":"ED-attended eye injury from gardening tools (US, lifetime hobbyist)","probability":0.0002,"notes":"~933 ED visits per year nationally for gardening-tool-related eye injuries (2014–2023 NEISS). Adds another ~2,000-4,000 ED visits per year once projectile injuries from string trimmers and mowers are included. Safety glasses cost about USD 5 and would eliminate most of these. This is the only pathway for which \"PPE would help\" is a quantitatively strong statement without caveats.\n"},{"region":"Contact dermatitis / phytophotodermatitis (poison ivy, wild parsnip, giant hogweed)","probability":0.05,"notes":"Highly common but overwhelmingly self-limiting. No meaningful mortality, though giant hogweed and wild parsnip exposures can produce severe phototoxic burns that need medical attention. Gloves plus long sleeves materially reduce incidence but no single per-session probability exists in the literature — the figure here reflects US household-survey estimates of annual poison-ivy reaction rates among adults who spend time outdoors.\n"},{"region":"Pesticide / herbicide dermal exposure harm (US hobby gardener, lifetime)","probability":0.000001,"notes":"For label-compliant consumer use of glyphosate, pyrethroids, and standard garden fungicides at hobbyist scale, no epidemiological cohort has isolated a measurable dose-response signal. Occupational applicators face exposure one to three orders of magnitude higher and genuinely elevated risk; see the pesticide-residue-food entry for the food-intake analogue.\n"}],"short_label":"Unprotected gardening","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry does not compute a single headline probability because none exists in the literature for the aggregate question \"will gardening without PPE harm you?\" The underlying harms are a short list of pathway-specific risks — tetanus, sporotrichosis, Legionella longbeachae pneumonia, mechanical eye injury, contact dermatitis, pesticide exposure — each of which has its own base rate, its own protective-factor story, and its own geography. The figures in the regional breakdown are US hobby-gardener rates unless noted. Occupational horticultural workers face meaningfully higher exposures across every pathway. Immunocompromised adults, diabetics with peripheral neuropathy, and people on chronic immunosuppression are the groups most likely to progress from a trivial thorn prick to a medically significant infection; the headline subgroup figures do not apply to them. Australian gardeners face materially higher L. longbeachae risk than US gardeners because their potting mix contains pine bark and sawdust. Vaccination status against tetanus is the single largest modifier of the tetanus subgroup and is dramatically more important than glove use.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-7-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single pair of folded cotton gardening gloves resting on a pale surface, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/gardening-without-ppe","api_url":"https://likelier.app/api/fears/gardening-without-ppe.json"},{"slug":"glp1-muscle-bone-loss","question":"If you lose a lot of weight on a GLP-1 drug, how much of what you lose is muscle and bone rather than fat, and does it raise your later frailty and fracture risk?","quick_answer":"There is no reliable probability for later frailty or fractures from GLP-1 weight loss. The muscle and bone that come off alongside fat are a body-composition ratio, not an event rate, and cannot be converted into a lifetime fracture or frailty risk, and the evidence even disputes whether the loss is greater than for any large weight loss.\n","category":"health","tags":["substance-use","elder-care"],"no_reliable_estimate":true,"perceived":{"description":"The everyday read of GLP-1 weight loss is that the scale number is fat leaving the body. Lean tissue, the muscle and bone that come off alongside the fat, is rarely part of how users picture the trade. The DXA evidence shows roughly a quarter to a third of the weight lost is lean mass, but no published survey measures how GLP-1 users rate their own muscle-or-bone-loss risk, so this is an editorial read of the framing rather than a measured figure.","rough_estimate":"Users tend to read scale-weight loss as pure fat loss and rarely price in the muscle and bone that come off with it.","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12394919/","title":"Effects of Tirzepatide on Skeletal Muscle Mass in Adults: A Systematic Review","publisher":"Cureus (PMC, peer-reviewed)","source_type":"peer_reviewed","statistic":"SURMOUNT-1 DXA substudy: tirzepatide produced a 21.3% mean reduction in body weight, of which approximately 75% was fat mass and 25% was lean mass; the review concludes lean-mass reductions were within expected ranges and not clinically significant.","excerpt":"Tirzepatide treatment led to a 21.3% mean reduction in body weight, with approximately 75% of the weight loss attributable to fat mass and 25% to lean mass.","source_date":"2025-07-29","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260626161316/https://pmc.ncbi.nlm.nih.gov/articles/PMC12394919/","calculation_notes":"Restates the SURMOUNT-1 DXA substudy (n=124 tirzepatide, n=36 placebo, 73% female): body weight -21.3%, fat mass -33.9%, lean mass -10.9% over 72 weeks; ~75% fat / 25% lean for both arms. The lean-mass fraction is a body-composition RATIO, not an event probability, so it cannot be converted into a lifetime fracture/frailty rate. This source also supplies the counter-read: the review states lean-mass loss was 'within expected ranges based on reference data from the UK Biobank, suggesting tirzepatide does not cause clinically significant skeletal muscle loss', i.e. the loss may be proportionate to any large weight loss rather than drug-specific excess.","independence_note":"Independent academic systematic review; not authored by the trial sponsor. Used here in place of the sponsor-published Wiley primary (paywalled, would not load) to anchor the same SURMOUNT-1 numbers."},{"url":"https://academic.oup.com/jes/article/5/Supplement_1/A16/6240360","title":"Impact of Semaglutide on Body Composition in Adults With Overweight or Obesity: Exploratory Analysis of the STEP 1 Study","publisher":"Journal of the Endocrine Society (Oxford Academic)","source_type":"peer_reviewed","statistic":"STEP-1 DXA substudy (n=140; semaglutide n=95, placebo n=45, 76% female): body weight -15.0% with semaglutide vs -3.6% placebo over 68 weeks; total lean body mass -9.7%; total fat mass -19.3%.","excerpt":"Percentage change in body weight from baseline to week 68 was -15.0% with semaglutide vs -3.6% with placebo... Total lean body mass decreased from baseline (-9.7%); however, the proportion relative to total body mass increased by 3.0%-points.","source_date":"2021-05-03","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250408195213/https://academic.oup.com/jes/article/5/Supplement_1/A16/6240360","calculation_notes":"On a kg basis the lean-mass share of total weight lost is larger here than in SURMOUNT-1. With mean baseline weight 98.4 kg, body weight -15.0% is ~14.8 kg lost; lean mass -9.7% of a typical ~63 kg lean compartment is ~6.1 kg, so roughly 35-40% of weight lost was lean tissue. Confirms the lean-mass effect is not tirzepatide-specific and sits at the upper end of the cited 25-40% range. Still a composition ratio, not a fracture probability.","independence_note":"Different drug (semaglutide), different trial, different sponsor analysis than SURMOUNT-1; corroborates the lean-mass fraction across the GLP-1 class."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12544599/","title":"MON-817 Association of Semaglutide and Tirzepatide Use on Bone Density and Fracture Risk in Obese Patients with and without Diabetes","publisher":"Journal of the Endocrine Society (PMC)","source_type":"primary_study","statistic":"Retrospective cohort (n=255, 92% female, mean age 68, ~34-month follow-up): 33 patients (13%) developed a new fracture after starting semaglutide or tirzepatide; significant BMD declines at lumbar spine (-1.6%), femoral neck (-1.8%) and total hip (-2.8%); total-hip BMD decline correlated with amount of weight lost (r=0.35).","excerpt":"33 patients (13%) developed a new fracture after starting semaglutide or tirzepatide... A significant decline in BMD were noticed in all sites—lumbar spine (-1.6±7.3%), femoral neck (-1.8±6.6%) and total hip (TH, -2.8±4.9%, p<0.001 for all)... The decline of TH BMD percentage was significantly associated with the amount of weight loss (Pearson r=0.35, p<0.001).","source_date":"2025-10-22","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260203141234/https://pmc.ncbi.nlm.nih.gov/articles/PMC12544599/","calculation_notes":"This is the closest thing to an event rate, and it is why no clean probability survives. The 13% is uncontrolled: a retrospective single-clinic cohort, 92% female, mean age 68 (already a high baseline fracture population), with no untreated comparator, so the drug-attributable share is unknowable. There is no measure of how many of these patients would have fractured anyway. The BMD declines confirm bone loss tracks weight loss, but a 1-3% BMD drop is modest and its translation into lifetime fracture probability is not established. Cannot yield a defensible normalized lifetime rate.","independence_note":"Independent academic clinic cohort; distinct from the trial DXA substudies and the TriNetX database study."},{"url":"https://medicaldialogues.in/orthopaedics/news/tirzepatide-linked-to-higher-risk-of-osteoporosis-and-fracture-compared-to-other-glp-1-receptor-agonists-study-158853","title":"Tirzepatide Linked to Higher Risk of Osteoporosis and Fracture compared to other GLP-1 receptor agonists: Study","publisher":"Medical Dialogues (reporting Hsu et al., Diabetes Research and Clinical Practice)","source_type":"news_article","statistic":"TriNetX retrospective cohort (66,329 matched per group, 14-month follow-up): tirzepatide users had a hazard ratio of 1.44 (95% CI 1.22-1.69) for a composite of new-onset osteoporosis or fragility fracture versus users of other GLP-1 receptor agonists.","excerpt":"After matching, users of tirzepatide had a significantly higher risk for the primary composite outcome versus users of other GLP-1 RAs, with a hazard ratio of 1.44 (95% CI: 1.22–1.69).","source_date":"2025-11-18","source_accessed":"2026-06-13","calculation_notes":"Crucially, this is a RELATIVE risk of tirzepatide versus OTHER GLP-1 drugs, not versus no drug, and not an absolute lifetime probability. A 44% relative increase over another GLP-1 comparator says nothing about the baseline absolute rate. The peer-reviewed primary (Hsu et al., Diabetes Research and Clinical Practice, S0168-8227(25)01010-1) was paywalled (403); this trade outlet reports the same HR verbatim. Used only to show the fracture signal is contested and direction-of-effect, never to derive a number.","independence_note":"Secondary reporting of a fourth, independent database (TriNetX); the underlying peer-reviewed study is separate from all three above."}],"comparison_anchors":[],"short_label":"GLP-1 muscle and bone loss","outcome_severity":"serious_harm","exposure_pattern":"degenerative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"No defensible lifetime probability exists, so this entry carries no_reliable_estimate. See the full reasoning in the prose body and source calculation_notes.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"GLP-1 muscle and bone loss\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/glp1-muscle-bone-loss","api_url":"https://likelier.app/api/fears/glp1-muscle-bone-loss.json"},{"slug":"gmo-food-health-harm","question":"What are the odds that eating genetically modified (GMO) food will harm your health?","quick_answer":"There is no reliable figure for harm from eating GMO food, because no epidemiological study has measured any attributable rate. The most comprehensive review to date found no substantiated evidence of a difference from conventionally bred crops. Many US adults are nonetheless wary that GM food is worse for their health.\n","category":"food","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"GMO foods are one of the most persistent food-safety anxieties in Western publics. Surveys consistently show that half or more of US adults believe genetically modified food poses health risks to consumers. The fear typically references unspecified cancer, allergy, or \"long-term\" harm and rarely cites a particular mechanism. Survey wording varies and trends are stable across two decades, well after the first commercial GM crops were approved in 1994 and consumed at scale.\n","rough_estimate":"around 50% of US adults say GM food is worse for health than non-GM food (Pew 2020)","kind":"survey","survey_source":{"title":"About Half of U.S. Adults Are Wary of Health Effects of Genetically Modified Foods","publisher":"Pew Research Center","url":"https://www.pewresearch.org/science/2020/03/18/about-half-of-u-s-adults-are-wary-of-health-effects-of-genetically-modified-foods-but-many-also-see-advantages/","year":2020}},"sources":[{"url":"https://www.nationalacademies.org/news/2016/05/genetically-engineered-crops-experiences-and-prospects-new-report","title":"Genetically Engineered Crops: Experiences and Prospects -- New Report","publisher":"National Academies of Sciences, Engineering, and Medicine","source_type":"govt_report","statistic":"Committee found no substantiated evidence of a difference in risks to human health between commercially available GE crops and conventionally bred crops, after reviewing >900 publications spanning two decades.","excerpt":"\"No substantiated evidence of a difference in risks to human health between current commercially available genetically engineered (GE) crops and conventionally bred crops.\" The committee \"carefully searched all available research studies for persuasive evidence of adverse health effects directly attributable to consumption of foods derived from GE crops but found none.\"\n","source_date":"2016-05-17","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20251113055331/https://www.nationalacademies.org/news/2016/05/genetically-engineered-crops-experiences-and-prospects-new-report","calculation_notes":"The NAS 2016 report is the single most comprehensive review of GE-crop health evidence to date: a 388-page consensus document, a 20-member committee drawn from agronomy, toxicology, molecular biology, ecology and economics, ~80 invited presentations, and a public comment process. Crucially the committee compared US disease-incidence trends (where GE crops have been consumed since 1996) against EU trends (where they have not) for cancer, type-2 diabetes, kidney disease, obesity, autism, coeliac disease and food allergies and reported no patterns consistent with a GE-attributable effect. This entry treats that synthesis as the anchor and therefore declines to publish a numerator. The honest reading is \"below the detection limit of population-level epidemiology\", not \"zero\".\n","independence_note":"The NAS committee was funded by a mix of US federal agencies, the Burroughs Wellcome Fund, and the Gordon and Betty Moore Foundation; conflicts of interest were screened per NAS standard procedure. Its synthesis is methodologically independent of the WHO Q&A below, which relies on a separate WHO/FAO Codex Alimentarius safety-assessment framework rather than a meta-review of independent cohort studies.\n"},{"url":"https://www.who.int/news-room/questions-and-answers/item/food-genetically-modified","title":"Food, Genetically Modified -- Q&A","publisher":"World Health Organization","source_type":"reputable_reference","statistic":"GM foods on international markets have passed safety assessments and are unlikely to present human-health risks; no human-health effects have been documented in approving-country populations to date.","excerpt":"\"GM foods currently available on the international market have passed safety assessments and are not likely to present risks for human health. In addition, no effects on human health have been shown as a result of the consumption of such foods by the general population in the countries where they have been approved.\"\n","source_date":"2014-05-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260409022027/https://www.who.int/news-room/questions-and-answers/item/food-genetically-modified","calculation_notes":"WHO does not publish a quantitative risk estimate for GM-food consumption because the routine pre-market safety assessment (substantial equivalence, compositional analysis, toxin and allergen screening per Codex Alimentarius CAC/GL 44-2003 and CAC/GL 45-2003) is a hazard-identification framework, not a probabilistic risk model. The Q&A is structured around individual assessment criteria (toxicity, allergenicity, gene stability, nutritional effects, unintended effects) and reports that no documented harm has been observed across the populations consuming approved GM crops since 1994. Cited here as a corroborating consensus, not as a quantitative source.\n","independence_note":"WHO/Codex Alimentarius safety-assessment framework is doctrinally and institutionally distinct from the NAS 2016 evidence synthesis. Some underlying source studies overlap (industry submission dossiers, peer- reviewed animal studies) but the two organisations apply different analytical lenses.\n"},{"url":"https://www.aaas.org/news/statement-aaas-board-directors-labeling-genetically-modified-foods","title":"Statement by the AAAS Board of Directors On Labeling of Genetically Modified Foods","publisher":"American Association for the Advancement of Science","source_type":"reputable_reference","statistic":"AAAS Board: \"crop improvement by the modern molecular techniques of biotechnology is safe\"; mandatory GM labeling would mislead consumers into inferring health risk where the scientific consensus finds none.","excerpt":"\"The science is quite clear: crop improvement by the modern molecular techniques of biotechnology is safe. The World Health Organization, the American Medical Association, the U.S. National Academy of Sciences, the British Royal Society, and every other respected organization that has examined the evidence has come to the same conclusion.\"\n","source_date":"2012-10-20","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260501171746/https://www.aaas.org/news/statement-aaas-board-directors-labeling-genetically-modified-foods","calculation_notes":"The AAAS Board statement is included as evidence of the breadth of the scientific consensus: AAAS itself reviewed the evidence and inventoried concurring positions from WHO, AMA, NAS, the British Royal Society and others. It is a position document rather than a primary study, so it contributes consensus weight rather than independent quantitative data to this entry.\n","independence_note":"AAAS is a US scientific society that issued this position after independent board review; their statement references other organisations' conclusions but the position itself reflects AAAS deliberation. Treated here as a corroborating consensus source alongside NAS and WHO.\n"}],"comparison_anchors":[{"label":"Pesticide residue on conventional produce (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Choking to death on food (lifetime, US adult)","lifetime_us_adult":0.000301},{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537}],"short_label":"GMO food harm","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry addresses health harm from approved, commercially available GM food consumed by US adults at typical dietary levels. It does not cover ecological effects of GM-crop cultivation (gene flow to wild relatives, herbicide- resistance evolution, pollinator effects), socio-economic effects of seed-patent regimes, occupational exposure to associated herbicides such as glyphosate (see the glyphosate-roundup-cancer entry), or hypothetical effects of GM varieties not yet on the market. The \"no substantiated evidence\" framing from NAS 2016 is an honest statement of the evidence base, not a declaration of zero risk: long-term, large-cohort prospective studies specifically tracking GM-food exposure do not exist and are difficult to design because GM-derived corn, soy, sugar, and canola pervade the US food supply, leaving no clear unexposed comparator. Any future signal at very low population incidence might therefore remain below the detection limit of available epidemiology. This entry will be re-evaluated when long-term cohort designs (such as the proposed exposure-biomarker studies of EU origin) report.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":5,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single ear of yellow corn resting on a pale neutral background, flat vector illustration."},"canonical_url":"https://likelier.app/gmo-food-health-harm","api_url":"https://likelier.app/api/fears/gmo-food-health-harm.json"},{"slug":"great-power-war","question":"What are the odds of a great-power war breaking out in your lifetime?","quick_answer":"There is no single reliable figure for a great-power war breaking out in your lifetime. Expert inputs disagree over whether the historical rate still holds or has fallen, so no lifetime number is adopted, and public fear tracks the news cycle rather than any stable estimate. A majority of Americans say they fear another world war.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Perceived great-power-war risk is the most news-cycle-sensitive item on this site. It spikes with each crisis — the 2022 Russian invasion of Ukraine, the Taiwan Strait tension over the 2024-2025 period, Middle East escalation — and recedes between them, tracking media salience rather than any stable probabilistic assessment. The Chapman Survey of American Fears found that 55% of Americans were afraid or very afraid of \"the U.S. becoming involved in another world war\" in 2024, ranking it the seventh-largest national fear, and six of the survey's top ten fears that year were war- or warfare-related. The mental model most people hold is binary — it either happens or it does not — rather than frequentist, which makes the public perception swing far more than any expert estimate of the underlying rate. The perceived figure is also ambiguous as to event: \"involved in another world war\" blends the probability that a great-power war breaks out at all with the probability that the respondent's own country is drawn into it.\n","rough_estimate":"55% of US adults report being afraid or very afraid of the US becoming involved in another world war (Chapman Survey 2024, rank 7 of all fears)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — 2024 Key Findings","publisher":"Chapman University","url":"https://news.chapman.edu/2024/10/23/corruption-nuclear-threats-and-economic-instability-dominate-national-fears-in-2024/","year":2024}},"sources":[{"url":"https://80000hours.org/problem-profiles/great-power-conflict/","title":"Great power conflict","publisher":"80,000 Hours / Stephen Clare","source_type":"reputable_reference","statistic":"Historical base rate of about two great-power wars per century (Levy); Metaculus median forecast for the next great-power war of 2048; author's own estimate of ~1-in-3 chance of a great-power war before 2050 and 45% by 2100","excerpt":"\"Over the past 500 years, about two great power wars have occurred per century. [...] I put the chance we experience some kind of war between great powers before 2050 at about one-in-three. [...] This method gives a 45% chance of seeing a major great power war in the next 77 years.\"\n","source_date":"2025-03-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260615224536/https://80000hours.org/problem-profiles/great-power-conflict/","calculation_notes":"Stephen Clare's 80,000 Hours problem profile synthesises the historical and forecasting literature on great-power war as an event (war between major powers such as the US, China, and Russia), distinct from personal casualty risk and from nuclear-weapon use. The Levy base rate of ~2 great-power wars per century implies a naive annual probability of roughly 2%, which over an 80-year horizon compounds to ~80% by simple extrapolation — but Clare adjusts the base rate downward to reflect the possibility that major war has become somewhat less likely, arriving at ~45% by 2100. The profile also reports Metaculus community forecasts as of mid-2023: a median date of 2048 for the next great-power war, a 17% chance of a US-China war before 2035, and a 20% chance of a US-Russia war before 2050. No single lifetime figure is adopted here because the inputs (whether the base rate is constant or declining) disagree by more than an order of magnitude — see caveats.\n","independence_note":"80,000 Hours / Stephen Clare is the synthesising secondary source. It is methodologically distinct from the XPT calibrated-elicitation study and from Braumoeller's statistical analysis, though it cites the same underlying base-rate literature (Levy, Correlates of War).\n"},{"url":"https://forum.effectivealtruism.org/posts/un42vaZgyX7ch2kaj/announcing-forecasting-existential-risks-evidence-from-a","title":"Announcing 'Forecasting Existential Risks: Evidence from a Long-Run Forecasting Tournament'","publisher":"Forecasting Research Institute (Existential Risk Persuasion Tournament)","source_type":"primary_study","statistic":"By 2100, superforecasters estimated a 9% chance of global catastrophe and 1% chance of human extinction; domain experts estimated 20% catastrophe and 6% human extinction","excerpt":"Reported headline figures (paraphrased from the announcement): by 2100, superforecasters estimated a 9% chance of catastrophe and a 1% chance of extinction, while domain experts estimated a 20% chance of catastrophe and a 6% chance of human extinction.\n","source_date":"2023-08-31","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20260510074609/https://forum.effectivealtruism.org/posts/un42vaZgyX7ch2kaj/announcing-forecasting-existential-risks-evidence-from-a","calculation_notes":"The Existential Risk Persuasion Tournament (XPT), run June-October 2022 by the Forecasting Research Institute, convened 169 forecasters (80 domain experts, 89 calibrated superforecasters). This URL is the FRI's own announcement of the study; the underlying working paper is the primary artefact, but the announcement carries the headline medians used here and was the fetchable verified source. Its figures anchor the SEVERITY tail of this entry rather than the event probability: they describe the cumulative chance that catastrophe or extinction occurs by 2100 from all causes, with nuclear war only one contributor among several. This is the right way to use XPT here — a great-power war is the event; whether it escalates to civilisational catastrophe is a separate, much-lower-probability tail. The expert-vs-superforecaster gap (roughly 2x on catastrophe, 6x on extinction) is itself evidence that the forward-looking probability of catastrophic war resists consensus. The excerpt is a paraphrase of the reported headline medians, not contiguous verbatim text, and is marked as such.\n","independence_note":"The FRI/XPT methodology (incentivised, calibrated elicitation with scoring rules) is independent of the 80,000 Hours synthesis and of Braumoeller's historical statistics. The expert and superforecaster panels were drawn from different populations.\n"},{"url":"https://www.prio.org/journals/jpr/booknotes/168","title":"Book Note: Braumoeller, Bear F (2019) Only the Dead: The Persistence of War in the Modern Age","publisher":"Journal of Peace Research / Peace Research Institute Oslo (PRIO) — review by Michael Spagat","source_type":"reputable_reference","statistic":"Braumoeller argues, on comprehensive conflict data, that conflict-initiation rates have not declined steadily since 1816 and that the post-1945 'Long Peace' need not reflect a structural change in the underlying rate of war","excerpt":"\"conflict initiation rates have not declined steadily since 1816\"\n","source_date":"2019-09-01","source_accessed":"2026-06-21","archive_url":"http://web.archive.org/web/20251118223655/https://www.prio.org/journals/jpr/booknotes/168","calculation_notes":"Braumoeller's \"Only the Dead\" (Oxford University Press, 2019) is the principal recent academic statement of the constant-risk view: that interstate conflict-initiation rates have not declined steadily since 1816, and that the Long Peace since 1945 is too short a sample to distinguish a genuine structural break from a lucky run. The PRIO book note (Michael Spagat, Journal of Peace Research) summarises this thesis and registers a critique — that Braumoeller extrapolates power-law tails \"through two, even three, orders of magnitude beyond the data's reach\" — which is itself useful context: even the constant-risk camp's strongest claims about very large wars rest on contested extrapolation. Under a constant ~2% annual onset rate, the chance of observing zero great-power wars over the 77 post-war years is roughly 0.98^77 ≈ 21% — fortunate, but not statistically anomalous. This source is the counterweight to the declining-frequency reading embedded in the 80,000 Hours base-rate adjustment; the tension between the two is exactly why no point estimate is offered. The cited excerpt is the reviewer's characterisation of Braumoeller's finding, not a quotation from the book itself.\n","independence_note":"Independent of the forecasting sources: Braumoeller's argument is a retrospective statistical analysis of historical conflict data, not an elicited forecast, and the PRIO/JPR review is a peace-research-venue assessment of it. It is the intellectual opposite pole to the assumption that major war has become less likely, included specifically to represent that disagreement.\n"}],"comparison_anchors":[],"short_label":"Great-power war","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This entry deliberately carries no point estimate, and the reason is the heart of the topic. First, the carving: this is the probability that a great-power war BREAKS OUT in one's lifetime — a direct armed conflict between major powers (US, China, Russia, or a NATO-scale bloc) — as an EVENT. It is NOT the probability of personally dying as a civilian in war (see war-civilian- casualty), and NOT the probability that a nuclear weapon is used (see nuclear-weapon-use-conflict); a great-power war can be fought, as in 1939-1945, without nuclear use. Second, the spread. The historical base rate is about two great-power wars per century (Levy), implying a naive ~2% annual probability and something like a 70-80% lifetime chance under a constant-risk model. But whether that rate still holds is the central, unresolved dispute in the field. The declining-frequency reading (reflected in Clare's downward-adjusted ~45% by 2100, and in the broader \"Long Peace\" literature) and the constant-risk reading (Braumoeller's \"Only the Dead\", which finds no statistically reliable decline and treats the post-1945 calm as consistent with luck) differ by more than an order of magnitude on the forward-looking probability. Forecaster aggregates sit in between and move with events: the Metaculus community median for the next great-power war was 2048 as of mid-2023, with a 75th percentile near 2088. Layered on top is the severity ambiguity — the XPT figures show experts and superforecasters disagreeing 2-to-6-fold on whether catastrophe or extinction follows, which is a different question again from whether a war occurs. With the occurrence estimate spanning roughly 20% to 80% over a lifetime depending purely on which model of history one accepts, and with no resolvable base rate to break the tie, fabricating a single normalized number would misrepresent the state of knowledge. The honest output is the range and its source: it is a coin-flip- scale possibility on the optimistic reading and closer to a near-certainty on the pessimistic one, and the disagreement is about the model, not the data.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-06-21","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-21","last_reviewed":"2026-06-21","reviewed":true,"generated_at":"2026-06-21","image":{"alt":"Two large abstract interlocking arcs leaning toward each other against a pale background, flat vector illustration with no people or weapons."},"canonical_url":"https://likelier.app/great-power-war","api_url":"https://likelier.app/api/fears/great-power-war.json"},{"slug":"hand-held-phone-call-crash","question":"What are the odds of a crash from holding a phone to your ear while driving?","quick_answer":"There is no reliable lifetime figure here: the evidence gives only the relative jump in crash odds while a call is in progress, with no baseline crash rate to convert into an absolute probability. Hands-free calls carry about the same elevated risk as hand-held, which undercuts the common intuition that talking on the phone is a safe middle ground between silence and texting.\n","category":"transport","no_reliable_estimate":true,"perceived":{"description":"Most drivers who make hand-held phone calls believe the behavior is meaningfully safer than texting because their eyes stay on the road. The hands-free-versus-handheld distinction is baked into law in most US states, which reinforces the intuition that talking is the acceptable middle ground. Few drivers have a quantitative sense of how much a voice call — with or without a phone physically in hand — actually shifts their crash odds; the implicit estimate is somewhere between \"trivial\" and \"a bit risky.\"\n","rough_estimate":"vaguely risky but much safer than texting; many assume hands-free is fully safe","kind":"intuition"},"sources":[{"url":"https://www.bmj.com/content/331/7514/428","title":"Role of mobile phones in motor vehicle crashes resulting in hospital attendance: a case-crossover study","publisher":"McEvoy SP, Stevenson MR, McCartt AT et al., BMJ","source_type":"primary_study","statistic":"Mobile phone use in the 10 minutes before a crash associated with a 4-fold increase in crash risk (overall OR 4.1, 95% CI 2.2–7.7); risk raised irrespective of device type (hands-free OR 3.8, 95% CI 1.8–8.0; hand-held OR 4.9, 95% CI 1.6–15.5) — hands-free not significantly different from hand-held","excerpt":"\"Driver's use of a mobile phone up to 10 minutes before a crash was associated with a fourfold increased likelihood of crashing (odds ratio 4.1, 95% confidence interval 2.2 to 7.7, P < 0.001).\"\n","source_date":"2005-08-20","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20241126025358/https://www.bmj.com/content/331/7514/428","calculation_notes":"McEvoy 2005 is the canonical case-crossover study: it compared drivers' own phone records in the 10-minute window before a crash versus a matched control window on the same day the prior week. The OR of 4.1 is the per-epoch relative risk while a call is in progress — not a per-trip, per-year, or lifetime figure, and the study reports no absolute crash rate or exposure-weighted multiplier. Critically, hands-free calls showed OR 3.8, statistically indistinguishable from hand-held (OR 4.9), supporting the view that manual distraction is not the dominant mechanism. Because this source supplies only a relative per-epoch odds ratio and no baseline crash frequency, it cannot on its own be converted into a lifetime probability — a key reason this entry carries no_reliable_estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","title":"Driver crash risk factors and prevalence evaluation using naturalistic driving data","publisher":"Dingus et al., Proceedings of the National Academy of Sciences (PNAS)","source_type":"peer_reviewed","statistic":"Talking on a handheld cell phone: OR 2.2; texting: OR 6.1; dialing: OR 12.2; overall handheld cell phone interaction: OR 3.6 (all vs model driving in SHRP 2 passenger-car naturalistic sample)","excerpt":"\"The overall risk of interacting with a handheld cell phone is 3.6 times higher than model driving.\"\n","source_date":"2016-03-08","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250707185013/https://pmc.ncbi.nlm.nih.gov/articles/PMC4790996/","calculation_notes":"Dingus 2016 provides the naturalistic-driving complement to McEvoy's case-crossover design. The 2.2x OR for talking on a handheld phone is lower than McEvoy's 4.1 partly because naturalistic studies capture near-crashes differently from hospital attendance databases. The two estimates bracket the plausible per-epoch range (roughly 2.2–4.1x). Like McEvoy, Dingus reports relative odds ratios against a model-driving baseline, not absolute crash rates per trip or per year, so it supports the direction and rough magnitude of the elevated risk but not a lifetime number. For context, texting's OR in the same dataset is 6.1 — nearly three times higher than talking on a handheld phone — which is the relevant comparison for drivers who think \"just talking\" is close to safe.\n"}],"comparison_anchors":[],"short_label":"Hand-held phone call + driving","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the cited evidence supports a relative per-epoch odds ratio but not a defensible lifetime probability. McEvoy 2005 and Dingus 2016 both report how much more likely a crash is during the moments a call is active (McEvoy OR 4.1; Dingus OR 2.2 for handheld talking) against a matched or model-driving baseline — they do not report an absolute crash rate per call, per trip, or per year, and neither provides the exposure-weighted multiplier needed to spread a per-epoch odds ratio over a lifetime of driving. Converting these odds ratios into a single lifetime figure would require assumptions (a per-capita crash baseline and a fraction of driving time spent on calls) that the cited sources do not supply, so no honest lifetime number is published here.\nThe most striking finding from McEvoy 2005 — that hands-free calls carry nearly identical crash risk to hand-held calls (hands-free OR 3.8 vs hand-held OR 4.9, statistically indistinguishable, with the authors concluding a hands-free phone is \"not any safer\") — suggests that what makes phone calls dangerous is primarily cognitive distraction, not the manual act of holding a device. Most US state laws ban only hand-held use, implying a safety distinction that the literature does not consistently support.\n","quality_score":{"d1":4,"d2":5,"d3":3,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A single muted mobile phone held loosely beside a pale steering wheel silhouette, flat vector illustration."},"canonical_url":"https://likelier.app/hand-held-phone-call-crash","api_url":"https://likelier.app/api/fears/hand-held-phone-call-crash.json"},{"slug":"helicopter-parenting-child-harm","question":"What are the odds of overprotective or authoritarian parenting causing lasting harm to your child?","quick_answer":"There is no reliable probability here: the research measures parenting style against continuous well-being scales, not a binary harm threshold, and the studies are cross-sectional associations that cannot establish causation. Most parents fear their style could cause lasting psychological harm, but the evidence supports only small, heavily confounded links.\n","category":"kids","tags":["kids","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"The modern parenting anxiety industry runs on the premise that every stylistic decision is high-stakes. Helicopter parenting entered the popular lexicon around 2010 and quickly became shorthand for a generation of over-involved parents supposedly crippling their children's autonomy. Tiger parenting occupied the opposite lane after Amy Chua's 2011 \"Battle Hymn of the Tiger Mother,\" framing achievement pressure as another route to psychological damage. Both narratives share a common assumption: that parenting style is a powerful, perhaps decisive, determinant of a child's long-term mental health. Parenting forums, bestselling advice books, and mainstream press coverage reinforce the intuition that getting it wrong — hovering too much, pushing too hard, or failing to strike the right balance — will produce anxious, depressed, or emotionally stunted adults. That intuition draws on real research, but it dramatically overstates the effect sizes the research has actually found.\n","rough_estimate":"Most parents believe their parenting style could cause lasting psychological harm; the research supports only small, heavily confounded associations","kind":"intuition"},"sources":[{"url":"https://link.springer.com/article/10.1007/s10826-013-9716-3","title":"Helping or Hovering? The Effects of Helicopter Parenting on College Students' Well-Being","publisher":"Journal of Child and Family Studies (Schiffrin et al. 2014)","source_type":"peer_reviewed","statistic":"In a sample of 297 college students, those reporting higher helicopter parenting scored significantly higher on depression measures and lower on life satisfaction; the relationship was mediated by reduced satisfaction of basic psychological needs for autonomy and competence","excerpt":"\"Students who reported having over-controlling parents reported significantly higher levels of depression and less satisfaction with life. The negative effects of helicopter parenting on college students' well-being were largely explained by the perceived violation of students' basic psychological needs for autonomy and competence.\"\n","source_date":"2014-02-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260206203045/https://link.springer.com/article/10.1007/s10826-013-9716-3","calculation_notes":"Schiffrin et al. 2014 is one of the most-cited studies linking helicopter parenting to reduced well-being in emerging adults. The sample size (N=297) is modest, the design is cross-sectional (not longitudinal), and the outcome measures are self-report scales, not clinical diagnoses. The study cannot establish causation: students who are already anxious or depressed may perceive their parents' involvement as more intrusive. No native or normalized probability is derived because this entry is flagged no_reliable_estimate — the outcome is a continuous well-being scale, not a binary harm threshold.\n"},{"url":"https://www.tandfonline.com/doi/abs/10.1080/02732173.2011.574038","title":"Does 'Hovering' Matter? Helicopter Parenting and Its Effect on Well-Being","publisher":"Sociological Spectrum (LeMoyne & Buchanan 2011)","source_type":"peer_reviewed","statistic":"In a survey of 317 college students aged 18-25, higher perceived helicopter parenting was associated with a 3.13x increase in the likelihood of taking prescription medication for anxiety or depression, and was positively associated with recreational pain pill consumption","excerpt":"\"Results suggest helicopter parenting is negatively related to psychological well-being and positively related to prescription medication use for anxiety/depression and the recreational consumption of pain pills.\"\n","source_date":"2011-06-09","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20250701211130/https://www.tandfonline.com/doi/abs/10.1080/02732173.2011.574038","calculation_notes":"LeMoyne & Buchanan 2011 is one of the earliest empirical studies connecting helicopter parenting to mental health medication use. The 3.13x odds ratio sounds large but comes from a small cross-sectional sample with self-reported parenting measures. The study cannot rule out reverse causation — parents may hover more over children who already have mental health difficulties, and those children may subsequently require medication. No native or normalized probability is derived. The entry is flagged no_reliable_estimate because the outcome (\"lasting psychological harm\") is not a binary event with a measurable base rate.\n","independence_note":"Independent sample from Schiffrin et al. 2014. Different university, different parenting measure, different outcome variables (medication use vs. well-being scales).\n"},{"url":"https://link.springer.com/article/10.1007/s10804-024-09496-5","title":"Parenting in Overdrive: A Meta-analysis of Helicopter Parenting Across Multiple Indices of Emerging Adult Functioning","publisher":"Journal of Adult Development (McCoy, Dimler & Rodrigues 2024)","source_type":"peer_reviewed","statistic":"A meta-analysis of 53 studies (111 effect sizes) found helicopter parenting was associated with increased internalising behaviours and reduced academic adjustment, self-efficacy, and regulatory skills in emerging adults; effect sizes were small to moderate","excerpt":"\"Helicopter parenting was associated with increased internalizing behaviors and reduced academic adjustment, self-efficacy and regulatory skills.\"\n","source_date":"2024-09-26","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20251206194456/https://link.springer.com/article/10.1007/s10804-024-09496-5","calculation_notes":"McCoy et al. 2024 is the most comprehensive meta-analysis of helicopter parenting to date. The headline finding — small-to-moderate effect sizes across 53 studies — is the best available summary of the literature. Small to moderate in meta-analytic terms typically means r in the 0.10-0.25 range, explaining 1-6% of variance in outcomes. This is consistent with a real but modest association, not a deterministic pathway from parenting style to harm. No native or normalized probability is derived because the outcomes are continuous scales and the association is correlational.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4739500/","title":"Top 10 Replicated Findings From Behavioral Genetics","publisher":"Perspectives on Psychological Science (Plomin et al. 2016)","source_type":"peer_reviewed","statistic":"Across decades of twin and adoption studies, heritabilities for personality traits are usually 30-50%; most environmental effects are not shared by children growing up in the same family, so the shared family environment (which includes parenting style) does little to make siblings similar — for intelligence, shared-environment influence is found up until adolescence and then diminishes","excerpt":"\"Most environmental effects are not shared by children growing up in the same family. ... The message is not that family experiences are unimportant but rather that the relevant experiences are specific to each child in the family.\"\n","source_date":"2016-01-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260420041748/https://pmc.ncbi.nlm.nih.gov/articles/PMC4739500/","calculation_notes":"Plomin et al. 2016 summarises the most robustly replicated findings from behavioural genetics. The near-zero shared-environment effect on adult personality is one of the most uncomfortable results in developmental psychology — it implies that the aspects of the home environment shared by siblings (including parenting style, socioeconomic status, neighbourhood) explain very little of the variance in how children turn out as adults. This does not mean parenting is irrelevant to child well-being in the moment, but it substantially constrains claims about \"lasting harm\" from normal-range parenting variation. Included to contextualise the helicopter parenting literature against the broader genetic evidence.\n","independence_note":"Independent evidence base from the helicopter parenting studies. Draws on twin and adoption research across multiple countries and decades.\n"},{"url":"https://www.scientificamerican.com/article/parents-peers-children/","title":"Do Parents Matter?","publisher":"Scientific American","source_type":"news_article","statistic":"Judith Rich Harris's 1998 book 'The Nurture Assumption' argued that parents matter much less than commonly assumed for determining children's behavior, and that a child's peer group is far more important — her 'group socialization theory,' built on twin/adoption behavioral-genetics research.","excerpt":"\"parents matter much less, at least when it comes to determining the behavior of their children, than is typically assumed. Instead, Harris argued that a child's peer group is far more important.\"\n","source_date":"2009-04-09","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260516170302/https://www.scientificamerican.com/article/parents-peers-children/","calculation_notes":"Cited solely to ground the specific named claim in the body prose (\"Judith Rich Harris argued in 'The Nurture Assumption' [1998] that peer socialisation does more to shape a child's adult personality than parenting does\"), which previously appeared without a citation. No native or normalized probability is derived from this source — it is historical/narrative framing for the behavioural-genetics discussion anchored by Plomin et al. (2016) above, not incidence data.\n","independence_note":"Science-journalism interview, not a primary study; independent of the four peer-reviewed sources above and not counted toward the authoritative-source requirement, which those four already satisfy.\n"}],"comparison_anchors":[],"regional_breakdown":[{"region":"Helicopter/overprotective parenting","probability":0,"notes":"Meta-analytic associations with anxiety (r ≈ 0.16) and depression (r ≈ 0.20) in emerging adults are statistically significant but small, explaining 2-4% of the variance. The research is cross-sectional and correlational — reverse causation (anxious children elicit more hovering) is a plausible alternative explanation. No reliable per-child probability of 'lasting harm' can be derived from these effect sizes."},{"region":"Authoritarian/achievement-focused parenting","probability":0,"notes":"Authoritarian parenting is consistently associated with higher anxiety, aggression, and lower self-esteem in Western samples (Baumrind 1966, 1991). Effect sizes are comparable to or slightly larger than helicopter parenting. However, the association is strongly moderated by culture — in collectivist societies (East Asian, African American, Arab communities), authoritarian practices are often associated with neutral or positive outcomes, undermining any universal risk estimate."},{"region":"Neglectful/uninvolved parenting","probability":0,"notes":"Severe neglect produces the largest and most consistent negative effects in the parenting literature — cognitive impairment, language deficits, behavioural problems, and disrupted brain architecture. Effect sizes for neglect dwarf those for overparenting or authoritarian parenting. This is the parenting failure mode that reliably produces lasting harm, and it is the one parents who worry about helicopter parenting are least likely to commit."},{"region":"Authoritative parenting (warm + firm boundaries)","probability":0,"notes":"Authoritative parenting is consistently associated with the best outcomes across studies: higher self-esteem, better academic performance, lower rates of depression and substance use. This is the evidence-backed approach — high warmth combined with clear expectations and age-appropriate autonomy. The effect sizes are moderate and the most replicated finding in parenting research."}],"short_label":"Helicopter parenting harm","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is flagged no_reliable_estimate for several reinforcing reasons.\nFirst, \"lasting psychological harm\" is not a binary, measurable outcome. The studies in this literature measure depression symptoms, anxiety symptoms, life satisfaction, and self-efficacy on continuous scales. Converting a small downward shift on a well-being questionnaire into a count of \"harmed children\" requires an arbitrary threshold that the field has not agreed on.\nSecond, the effect sizes are small. The most comprehensive meta-analysis to date (McCoy et al. 2024, 53 studies) finds small-to-moderate associations between helicopter parenting and negative outcomes. A separate meta-analysis reports r = 0.16 for anxiety and r = 0.20 for depression — real associations that explain 2-4% of the variance. Parenting style is one input among many, not a deterministic lever.\nThird, the evidence is overwhelmingly correlational and cross-sectional. Nearly every study in the helicopter parenting literature surveys college students at a single time point and asks them to retrospectively report on their parents' behaviour. Reverse causation is not a minor caveat — it is a central alternative explanation. Parents may hover more over children who are already anxious, and anxious children may retrospectively perceive their parents as more controlling. Without longitudinal designs that measure parenting before the onset of symptoms, causation cannot be established.\nFourth, behavioural genetics research (Plomin et al. 2016) consistently finds that shared family environment — the component that includes parenting style — explains near-zero variance in adult personality traits. Genetics accounts for roughly 50%, and the nonshared environment (peers, idiosyncratic experiences) accounts for most of the rest. This does not mean parenting cannot affect a child's current well-being, but it substantially limits claims about permanent damage from normal-range parenting variation.\nFifth, cultural moderation is severe. Authoritarian parenting, which produces negative associations in Western samples, is associated with neutral or positive outcomes in East Asian, African American, and Arab cultural contexts. Any universal probability of harm from a given parenting style would be misleading across cultures.\nThe one parenting failure mode that does produce reliably large, lasting effects is severe neglect — and that is categorically different from the helicopter-vs-free-range debate that dominates the popular discourse.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"An empty playground swing hanging motionless against a muted sky, flat vector illustration in subdued blue-grey tones."},"canonical_url":"https://likelier.app/helicopter-parenting-child-harm","api_url":"https://likelier.app/api/fears/helicopter-parenting-child-harm.json"},{"slug":"home-burglary-us","question":"What are the odds of your home being burglarized?","quick_answer":"There is no reliable lifetime figure here: the available sources either bundle burglary with motor-vehicle theft and other household theft, or give police-reported counts that undercount unreported incidents, so none isolates the per-household burglary rate a lifetime probability would require. About a third of US adults say they worry about being burglarized, even as measured rates have fallen for three decades.\n","category":"crime","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"Gallup's annual crime-worry poll consistently places home burglary in the middle of the list. In the October 2025 wave, 34% of US adults said they worry frequently or occasionally about having their home burglarized — below identity theft (69%) and car theft (39%), but above being mugged (29%) or murdered (22%). The worry level is broadly stable year-over-year even as measured burglary rates have fallen sharply for three decades.\n","rough_estimate":"About a third of US adults say they worry about being burglarized","kind":"poll","survey_source":{"title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","year":2025}},"sources":[{"url":"https://bjs.ojp.gov/press-release/criminal-victimization-2023","title":"Criminal Victimization, 2023","publisher":"US Bureau of Justice Statistics","source_type":"govt_report","statistic":"Property victimization rate of 102.2 per 1,000 households in 2023 (includes burglary/trespassing, motor vehicle theft, and other household theft; the press release does not break out the burglary component)","excerpt":"\"In 2023, the rate of property victimization was 102.2 per 1,000 households, which was not significantly different from 2022. Property victimization includes burglary or trespassing, motor vehicle theft, and other types of household theft.\"\n","source_date":"2024-09-12","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172429/https://bjs.ojp.gov/press-release/criminal-victimization-2023","calculation_notes":"BJS NCVS Criminal Victimization 2023 reports total property victimization at 102.2 per 1,000 households, a figure that bundles burglary/trespassing with motor-vehicle theft and other household theft. This press release does not report the burglary component on its own, so it cannot ground a burglary-specific per-household rate or a lifetime probability. The NCVS captures both reported and unreported victimizations via household interviews, making it the most complete US household victimization measure, but the burglary subrate needed here lives in detailed data tables not reached among these cited sources.\n","independence_note":"BJS NCVS is a household survey conducted independently of the FBI UCR/NIBRS law-enforcement reporting system. The two count burglaries through entirely different pipelines — victim recall vs police reports — and the NCVS typically yields higher counts because it captures unreported incidents.\n"},{"url":"https://www.safehome.org/resources/burglary-statistics/","title":"The Latest Burglary Statistics: How Common is Burglary in the U.S.?","publisher":"SafeHome.org (citing FBI UCR/NIBRS)","source_type":"reputable_reference","statistic":"852,963 reported burglaries in 2023 at a rate of 253.3 per 100,000; national burglary rate dropped 69% between 2005 and 2024","excerpt":"\"This represents a 9.5% decrease in the burglary rate from 2023 (253.3 per 100,000). The national burglary rate dropped 69% between 2005 and 2024. The data shows that total burglaries dropped by 64 percent between 2005 and 2024, from over 2.1 million incidents to fewer than 780,000.\"\n","source_date":"2025-10-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172502/https://www.safehome.org/resources/burglary-statistics/","calculation_notes":"FBI UCR/NIBRS reports 852,963 burglaries in 2023 at 253.3 per 100,000 population. This is a police-reported count and is lower than a household-survey measure would be, because FBI data only counts incidents reported to police and only a minority of household burglaries are reported to authorities. The figure is per-population rather than per-household, so it establishes that reported burglary is common and steeply declining (down 69% from 2005 to 2024) but does not by itself yield a per-household annual or lifetime probability.\n","independence_note":"FBI UCR/NIBRS collects incident data from law enforcement agencies, while BJS NCVS surveys households directly. The two systems are methodologically independent and undercount in opposite directions.\n"},{"url":"https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","title":"Crime in U.S. Seen as Less Serious for Second Straight Year","publisher":"Gallup","source_type":"reputable_reference","statistic":"34% of US adults worry frequently or occasionally about having their home burglarized (October 2025)","excerpt":"\"Fewer Americans say they worry about crimes, such as having a car stolen (39%) or their home burglarized (34%), being a victim of a hate crime (30%), or getting mugged (29%), attacked while driving (27%), murdered (22%) or sexually assaulted (21%).\"\n","source_date":"2025-10-30","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413172538/https://news.gallup.com/poll/697124/crime-seen-less-serious-second-straight-year.aspx","calculation_notes":"Used for the perceived-risk side only. The 34% figure is the fraction of respondents reporting frequent-or-occasional worry, not an elicited probability. It places home burglary in the middle tier of Gallup's crime-worry hierarchy.\n","independence_note":"Gallup telephone survey data, entirely separate from both BJS NCVS household victimization sampling and FBI UCR police-report aggregation. Measures public worry, not incidence — included only for the perceived-risk axis and not for any probability estimate.\n"}],"comparison_anchors":[{"label":"Being murdered (lifetime, US adult average)","lifetime_us_adult":0.00348},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6},{"label":"Death in a car crash (lifetime, US)","lifetime_us_adult":0.0108}],"short_label":"Home burglary","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"property","valence":"negative","caveats":"This is not a death risk; it concerns whether a US household experiences a burglary or unlawful trespass. No single lifetime probability is published here because the reachable cited sources do not report a burglary-specific per-household rate — the BJS NCVS press release gives only the aggregate property-victimization rate (102.2 per 1,000 households in 2023, which also includes vehicle theft and other household theft), and the FBI count is police-reported and per-population. The NCVS definition of \"burglary or trespassing\" is also broader than the common mental image of a break-in while away: it includes unlawful entries where nothing is stolen and trespassing incidents that may involve no forced entry. The FBI's narrower UCR definition yields a lower count (852,963 reported burglaries in 2023), both because of the definitional gap and because only a minority of household burglaries are reported to police. Any lifetime figure would be highly sensitive to the assumed annual rate and to the multi-decade declining trend — the reported burglary rate fell 69% between 2005 and 2024 — so someone turning 18 today faces a lower per-year risk than adults did in earlier decades. Geography and income are the largest sources of heterogeneity: county-level burglary rates span more than an order of magnitude.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-seed","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single house key resting on a pale surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/home-burglary-us","api_url":"https://likelier.app/api/fears/home-burglary-us.json"},{"slug":"ignoring-minor-symptoms","question":"What are the odds that ignoring a minor symptom means missing something serious?","quick_answer":"There is no single probability here: \"a symptom\" spans a benign twinge to a red flag, and the honest figures are different kinds of number. Even a classic alarm symptom like rectal bleeding turns out to be cancer only about 2-8% of the time, and most people who let such symptoms pass are not missing anything. What concentrates the risk is not the symptom itself but its persistence or recurrence, specific red flags, and age. UK guidelines draw the line for urgent investigation at roughly a 3% chance of cancer.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The health-anxiety reflex treats every unfamiliar sensation as a possible early sign of something grave, and the public-health messaging around cancer (\"don't ignore the warning signs\") reinforces it. The felt fear is that a symptom shrugged off today becomes a serious disease caught too late tomorrow. The opposite miscalibration is just as common: cancer \"alarm\" symptoms are so ordinary that most people never attribute them to anything serious and never mention them to a doctor. Both intuitions run without a base rate. Neither \"any symptom could be cancer\" nor \"it is probably nothing\" is a probability, and which one is right depends almost entirely on the specific symptom, how long it lasts, and who has it.\n","rough_estimate":"Most people either over-read minor symptoms as possible catastrophe or dismiss genuine alarm symptoms outright; neither intuition rests on an actual probability","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4252079/","title":"Attributions of Cancer 'Alarm' Symptoms in a Community Sample","publisher":"PLoS ONE (Whitaker, Scott, Winstanley, Macleod & Wardle 2014)","source_type":"peer_reviewed","statistic":"In a UK community sample of 1,724 adults, 53% (915) had experienced at least one cancer 'alarm' symptom in the previous 3 months, but only 2% (20/915) attributed any such symptom to cancer, and 41% did not contact their GP about the symptom they reported.","excerpt":"\"Over half the respondents (915: 53%) had experienced at least one cancer 'alarm' symptom in the past 3 months. ... Two percent (20/915) of respondents who had experienced an alarm symptom made a cancer attribution ... The majority of respondents (59%; 935/1,584) had contacted their GP about their symptom, ranging from 47% (51/109) for change in the appearance of a mole to 72% (177/246) for persistent unexplained pain.\"\n","source_date":"2014-12-02","source_accessed":"2026-08-15","calculation_notes":"Whitaker et al. 2014 measures symptom prevalence, cancer attribution, and help-seeking behaviour in the community, not a harm rate. The figures are a behavioural prevalence (P(behaviour) over a 3-month window), not a lifetime probability of a bad outcome, so no native/normalized value is derived. Its role in this entry is to show that alarm symptoms are common and usually dismissed, and that almost no one who has one thinks of cancer — the perception side of the calibration.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2790706/","title":"The CAPER studies: five case-control studies aimed at identifying and quantifying the risk of cancer in symptomatic primary care patients","publisher":"British Journal of Cancer (Hamilton 2009)","source_type":"peer_reviewed","statistic":"In unselected primary care patients, rectal bleeding carried a positive predictive value of 2.4% for colorectal cancer, and haemoptysis a PPV of 2.4% for lung cancer.","excerpt":"\"Rectal bleeding in colorectal cancer, and haemoptysis in lung cancer both had positive predictive values of 2.4%.\"\n","source_date":"2009-12-03","source_accessed":"2026-08-15","calculation_notes":"A positive predictive value is P(cancer | symptom) — a conditional probability given that a person already has the symptom, not a lifetime probability for an unselected adult, so it cannot populate the normalized headline. The 2.4% figure for two textbook \"red flag\" symptoms in unselected primary care is the reassurance half of the calibration: even a classic alarm symptom usually means no cancer.\n","independence_note":"Not an independent corroboration of Astin et al. 2011 below: Astin's pooled rectal-bleeding estimate includes the CAPER colorectal studies, so this 2.4% figure is one of the inputs to the 8.1% pool. The two are cited for different points, not as independent measurements of one quantity — CAPER for the unselected-primary-care baseline (and for haemoptysis in lung cancer, a separate site), Astin for how the estimate rises when pooled over age >=50.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3080228/","title":"The diagnostic value of symptoms for colorectal cancer in primary care: a systematic review","publisher":"British Journal of General Practice (Astin, Griffin, Neal, Rose & Hamilton 2011)","source_type":"peer_reviewed","statistic":"Across 13 studies, the PPV of rectal bleeding for colorectal cancer ranged from 2.2% to 16%, with a pooled estimate of 8.1% (95% CI 6.0%-11%) in patients aged >=50.","excerpt":"\"Positive predictive values (PPVs) for rectal bleeding from 13 papers ranged from 2.2% to 16%, with a pooled estimate of 8.1% (95% confidence interval [CI] = 6.0% to 11%) in those aged >=50 years.\"\n","source_date":"2011-05-01","source_accessed":"2026-08-15","calculation_notes":"Same quantity type as CAPER (a conditional PPV, not a lifetime rate). Pooled over patients aged >=50, the PPV of rectal bleeding roughly triples versus the unselected-primary-care figure (8.1% vs 2.4%), which is the point: age moves the line as much as the symptom does. Still below the threshold at which a lifetime probability could be honestly assigned to \"ignoring a symptom\".\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK555330/","title":"Suspected cancer: recognition and referral (NICE guideline NG12)","publisher":"National Institute for Health and Care Excellence (NICE) 2015","source_type":"govt_report","statistic":"NICE lowered the risk threshold for urgent suspected-cancer referral and urgent direct-access investigation to a positive predictive value of 3% (from the previous ~5%).","excerpt":"\"the GDG felt that, in order to improve diagnosis of cancer, a PPV threshold lower than 5% was preferable. Taking into account the financial and clinical costs of broadening the recommendations, the GDG agreed to use a 3% PPV threshold value to underpin the recommendations for suspected cancer pathway referrals and urgent direct access investigations.\"\n","source_date":"2015-06-23","source_accessed":"2026-08-15","calculation_notes":"The 3% is a policy threshold — the probability of cancer at which a health system judges investigation worthwhile — not a personal lifetime probability. It is the closest thing to an explicit answer to \"where is the line\": below roughly a 3% chance of cancer the system, by design, does not escalate. No native/normalized figure is derived from a decision threshold.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4385982/","title":"Is increased time to diagnosis and treatment in symptomatic cancer associated with poorer outcomes? Systematic review","publisher":"British Journal of Cancer (Neal et al. 2015)","source_type":"peer_reviewed","statistic":"A systematic review of 177 articles (209 studies) found that longer time to diagnosis and treatment was associated with poorer outcomes for several cancers (breast, colorectal, head and neck, testicular, melanoma), while cautioning that confounding by indication (the 'waiting time paradox') limits causal inference.","excerpt":"\"Our main conclusion from this review is that we believe that it is reasonable to assume that efforts to expedite the diagnosis of symptomatic cancer are likely to have benefits for patients in terms of earlier-stage diagnosis, improved survival and improved quality of life.\"\n","source_date":"2015-03-31","source_accessed":"2026-08-15","calculation_notes":"This is the counterweight source: it is why the entry is not simply \"relax, symptoms are usually benign\". For a defined set of cancers, delay in the serious subset is associated with worse survival and later stage. But the review itself flags the \"waiting time paradox\" (aggressive cancers present fast with poor outcomes, which can make short intervals look harmful), so even this literature cannot be turned into \"ignoring symptom X costs you Y\" — which is precisely why no_reliable_estimate is the honest state rather than a fabricated number.\n","independence_note":"Systematic review spanning 209 studies and multiple cancer sites and countries; independent of the single-symptom PPV work above.\n"}],"comparison_anchors":[],"short_label":"Ignoring symptoms","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"death","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because \"ignoring a minor symptom\" cannot be reduced to one lifetime probability without inventing it, for several reinforcing reasons.\nFirst, \"a symptom\" is not one thing. A tension headache, a week of fatigue, post-menopausal bleeding, a breast lump, and blood in the urine sit at wildly different risk levels. A single number averaged across \"symptoms\" would be meaningless.\nSecond, the honest figures are different kinds of quantity that cannot be added or averaged into a headline. The positive predictive value of an alarm symptom (Hamilton 2009: rectal bleeding 2.4% in unselected primary care; Astin 2011: 8.1% pooled at age >=50) is P(cancer | symptom) — conditional on already having the symptom. The Whitaker (2014) figures are behavioural prevalences over a three-month window. The NICE 3% is a policy threshold. None of these is a lifetime probability for a US adult, and forcing any of them into that slot would mislabel the fear.\nThird, the \"cost of delay\" evidence is real but confounded. Neal et al. (2015) find longer time-to-diagnosis associated with worse outcomes for several cancers, but the same review flags the \"waiting time paradox\": fast-growing cancers tend to present quickly and do badly, which can make short diagnostic intervals spuriously look harmful and long ones look protective. This makes the net effect of an individual's delay genuinely uncertain, not merely unmeasured.\nFourth, most of the load-bearing evidence is British (NICE, the CAPER and Astin primary-care PPVs, and the Whitaker community survey are all UK; Neal is UK-led). PPVs depend on the underlying incidence and on how a health system triages, so the exact percentages are not straightforwardly transportable to a US adult, even though the qualitative pattern — most alarm symptoms are benign, the serious minority is where delay bites — is not UK-specific.\nWhat the evidence does support is a decision rule rather than a probability: for most symptoms the risk concentrates in those that are persistent or recurring, and it rises with age; a symptom that appears once and resolves is, statistically, usually nothing. The exception is a specific set of red flags (rectal bleeding, visible blood in urine, post-menopausal bleeding, an unexplained lump, a new persistent cough, unexplained weight loss) — for several of these, notably post-menopausal bleeding and visible haematuria, referral guidelines act on a single episode, so resolution does not make them safe to dismiss.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-08-15","methodology_version":"1.2"},"reviewer":"8d-eval-2026-08-15","last_reviewed":"2026-08-16","reviewed":true,"generated_at":"2026-08-15","image":{"alt":"A single unopened envelope resting on a windowsill in soft daylight, flat vector illustration in muted grey-blue and cream tones."},"canonical_url":"https://likelier.app/ignoring-minor-symptoms","api_url":"https://likelier.app/api/fears/ignoring-minor-symptoms.json"},{"slug":"industrial-robot-job-displacement","question":"What are the odds of losing your job to a robot?","quick_answer":"There is no reliable per-worker number. The strongest study finds that each additional industrial robot per thousand workers lowered a local area's employment-to-population ratio by roughly 0.2 points and wages by about 0.4 percent between 1990 and 2007, but that is an area-wide effect, not a probability that any given person is replaced. Robot density keeps rising while total US employment is still projected to grow, so the population-level fear runs ahead of the measured labor-market effect.\n","category":"other","tags":["workplace"],"no_reliable_estimate":true,"perceived":{"description":"Robots taking factory and warehouse jobs is one of the oldest automation anxieties, older than the current AI-displacement fear and often blurred together with it. Pew Research found in 2017 that Americans were roughly twice as likely to be worried as enthusiastic about a future in which robots and computers can do many human jobs, and three-quarters expected automation to widen economic inequality. Coverage of humanoid-robot demos and \"lights-out\" factories keeps the image vivid, collapsing the distinction between a machine that can do part of a job and one that eliminates the worker who held it.\n","rough_estimate":"72% of US adults express some worry about a future in which robots and computers can do many human jobs; only 33% are enthusiastic (Pew Research, 2017)","kind":"survey","survey_source":{"title":"Americans' attitudes toward a future in which robots and computers can do many human jobs","publisher":"Pew Research Center","url":"https://www.pewresearch.org/internet/2017/10/04/americans-attitudes-toward-a-future-in-which-robots-and-computers-can-do-many-human-jobs/","year":2017}},"sources":[{"url":"https://www.nber.org/papers/w23285","title":"Robots and Jobs: Evidence from US Labor Markets","publisher":"National Bureau of Economic Research / Journal of Political Economy 128(6)","source_type":"peer_reviewed","statistic":"One more robot per thousand workers reduces the employment-to-population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent (preferred point estimates: ~0.2 pp and ~0.42%); the arrival of one robot in a local labor market is associated with roughly 3.3-6 fewer workers employed there","excerpt":"\"We estimate large and robust negative effects of robots on employment and wages across commuting zones. [...] one more robot per thousand workers reduces the employment to population ratio by about 0.18-0.34 percentage points and wages by 0.25-0.5 percent.\"\n","source_date":"2020-06-01","source_accessed":"2026-07-29","calculation_notes":"Acemoglu & Restrepo, \"Robots and Jobs: Evidence from US Labor Markets,\" Journal of Political Economy 128(6), 2020 (NBER Working Paper 23285). The design is a commuting-zone quasi-experiment exploiting variation in industry exposure to industrial robots over 1990-2007; the published version's preferred point estimates are a 0.2-percentage-point fall in the employment-to-population ratio and a 0.42% wage decline per robot per thousand workers. The authors' own summaries put the local effect at roughly six fewer workers per robot in the immediate commuting zone and about 3.3 fewer nationally once trade and product-market spillovers are netted out. Crucially, these are local-equilibrium employment elasticities, NOT a per-worker lifetime displacement probability: they measure how a region's employment moves with robot adoption, including workers displaced into other sectors and knock-on effects on non-robot jobs. There is no defensible arithmetic path from this elasticity to \"the odds a given US adult is replaced by a robot,\" which is why this entry carries no_reliable_estimate. The estimate is also period-specific (industrial robots, pre-2008) and does not cover warehouse, logistics, or service robots deployed since.\n","independence_note":"Seminal quasi-experimental study using International Federation of Robotics installation data mapped to US commuting zones. Independent from the task-exposure and survey sources.\n"},{"url":"https://ifr.org/ifr-press-releases/news/robot-density-surges-in-europe-asia-and-americas","title":"Robot Density Surges in Europe, Asia and the Americas","publisher":"International Federation of Robotics (World Robotics 2025)","source_type":"reputable_reference","statistic":"US manufacturing robot density reached 307 units per 10,000 employees in 2024, ranking 8th worldwide (up from 285 in 2022)","excerpt":"\"The United States ranks 8th worldwide with 307 units per 10,000 employees.\"\n","source_date":"2026-04-08","source_accessed":"2026-07-29","calculation_notes":"IFR World Robotics 2025 (data year 2024). Robot density — the operational stock of industrial robots per 10,000 manufacturing employees — is the standard measure of how far automation has actually been deployed. US density of 307 (2024), up from 285 in 2022, is used here only to establish that robot adoption is large and rising, which is the real trend underneath the fear. Density is a deployment measure, not a displacement rate: a higher robot count does not translate into a fixed number of jobs eliminated, because robots also raise output, lower prices, and create maintenance/engineering roles. No per-worker probability is derivable from a density figure.\n","independence_note":"Industry census of installed robots; the same IFR installation data underlies the Acemoglu & Restrepo identification strategy but is used here purely for the deployment trend.\n"},{"url":"https://www.bls.gov/news.release/ecopro.nr0.htm","title":"Employment Projections: 2023-2033","publisher":"Bureau of Labor Statistics","source_type":"govt_report","statistic":"Total US employment projected to add 6.7 million jobs (4.0% growth) from 2023 to 2033, reaching 174.6 million","excerpt":"\"The U.S. economy is projected to add 6.7 million jobs from 2023 to 2033 [...] Total employment is projected to increase to 174.6 million and grow 4.0 percent.\"\n","source_date":"2024-08-29","source_accessed":"2026-07-29","archive_url":"https://web.archive.org/web/20250530170440/https://www.bls.gov/news.release/ecopro.nr0.htm","calculation_notes":"BLS Employment Projections, 2023-2033 release (USDL-24-1776). Net job growth of 6.7 million through 2033 is a direct counterweight to headline claims of mass robotic job elimination: BLS incorporates historical technology-adoption and productivity trends and still projects net growth, not net loss. NOTE: bls.gov/news.release/ecopro.nr0.htm is a rolling endpoint that now serves the newer 2024-2034 projections, so archive_url pins the dated 2025-05-30 Wayback snapshot of the 2023-2033 release from which this figure and excerpt are drawn. This does not preclude heavy churn or concentrated losses in specific occupations and regions, but it bounds the population-level story. No per-worker displacement probability is published.\n","independence_note":"Official US government statistical agency using proprietary macroeconomic modelling. Fully independent from the academic and industry sources above.\n"}],"comparison_anchors":[{"label":"Personal bankruptcy (lifetime, US adult)","lifetime_us_adult":0.1},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"short_label":"Robot job loss","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"Every credible estimate in this space measures either robot deployment (density, installations) or an area-wide employment elasticity, not a per-person chance of being replaced. The gap between \"a robot can perform this task\" and \"this specific worker loses their job\" is wide and poorly measured: firms frequently automate a task and reassign the worker, and automation that cuts unit costs can expand output and headcount elsewhere (the reinstatement effect Acemoglu & Restrepo model alongside displacement). Any per-region or per-occupation exposure figure would be a relative indicator of where the measured effects concentrate, not an observed displacement probability, which is why this entry publishes none. Geography and occupation dominate individual risk — a machine operator in an auto-manufacturing commuting zone and a nurse in a labor-short region live in different worlds, and neither is described by a national average. Time horizon matters just as much: the industrial-robot story of 1990-2007 is largely settled, while the warehouse, logistics, and humanoid-robot wave of the 2020s is too recent to have a credible displacement estimate. Concrete disputes make the churn real without making it quantifiable — 2024-2025 saw the John Deere manufacturing layoffs and a longshoremen's strike over semi-automated port equipment that ended with a deal limiting full automation — but none of these yield a lifetime per-worker probability, so this entry stays no_reliable_estimate rather than fabricate one.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-07-29","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-29","last_reviewed":"2026-07-29","reviewed":true,"generated_at":"2026-07-29","image":{"alt":"An empty factory workstation beside an idle articulated robotic arm, muted slate and amber tones, flat vector illustration."},"canonical_url":"https://likelier.app/industrial-robot-job-displacement","api_url":"https://likelier.app/api/fears/industrial-robot-job-displacement.json"},{"slug":"infant-formula-botulism","question":"What are the odds of an infant developing botulism from contaminated powdered infant formula?","quick_answer":"There is no reliable per-infant probability here: formula-borne infant botulism had never been recorded before a single point-source contamination episode in 2025 (the ByHeart recall), and outbreak case counts have no denominator of exposed formula-fed infants to convert into a rate.\n","category":"health","tags":["kids","infant","food"],"no_reliable_estimate":true,"perceived":{"description":"Powdered infant formula is sold as the safe, regulated alternative to whatever a parent worries about, so the idea that the formula itself could be the source of a paralytic illness tends to land as either unthinkable or, after the November 2025 ByHeart recall headlines, as a vivid acute dread. There is no settled public sense of how likely it is, because before 2025 it had never happened. Parents who track recalls now over-weight it; parents who do not have no mental model for it at all. The gap is less perceived-versus-actual magnitude than the absence of any stable reference rate to anchor on.","rough_estimate":"Parents who followed the 2025 recall coverage tend to imagine formula-borne botulism as a roughly 1-in-a-few-thousand event for a formula-fed infant; parents who did not file it near zero. Neither figure rests on a defensible long-run denominator.","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/botulism/outbreaks-investigations/infant-formula-nov-2025/index.html","title":"Infant Botulism Outbreak Linked to Infant Formula, November 2025","publisher":"U.S. Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Final outbreak total: 28 confirmed + 20 probable = 48 cases across 17 states; 48 hospitalizations; 0 deaths. ByHeart recalled all infant formula products on November 11, 2025. Investigation status: Closed.","excerpt":"The outbreak included 28 confirmed cases and 20 probable cases. Hospitalizations: 48. States: 17. Deaths: 0. On November 11, 2025, ByHeart recalled all infant formula products. This outbreak is over.","source_date":"2026-03-05","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260704051033/https://www.cdc.gov/botulism/outbreaks-investigations/infant-formula-nov-2025/index.html","calculation_notes":"Final CDC investigation figures (48 cases / 17 states), which superseded the interim Dec 10, 2025 press-release count of 51/19. These are outbreak case counts, not a rate: there is no denominator of exposed formula-fed infants, so this cannot be converted to a per-infant probability. Used to establish that the event was a single point-source contamination episode that has been closed, not a recurring hazard."},{"url":"https://www.cdc.gov/media/releases/2025/2025-cdcs-rapid-response-helps-solve-first-infant-botulism-outbreak-linked-to-formula.html","title":"CDC's Rapid Response Helps Solve First Infant Botulism Outbreak Linked to Formula","publisher":"U.S. Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"First-ever outbreak of infant botulism linked to formula; CDC typically expects 150-180 individual infant botulism cases annually (from all sources combined).","excerpt":"While we typically expect to see 150-180 individual cases of infant botulism annually, an outbreak of infant botulism is unprecedented.","source_date":"2025-12-10","source_accessed":"2026-06-13","calculation_notes":"Establishes two load-bearing facts: (1) this was the first formula-linked infant botulism outbreak ever recorded in the US, meaning the historical formula-attributable rate is approximately zero; (2) the all-cause baseline runs 150-180 infant botulism cases per year, the great majority of which are honey/environmental, not formula. The interim 51/19 figure in this release was later revised to 48/17 by the closed investigation page. Cited for the 'first ever' framing and the all-cause baseline, not for the case count."},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK493178/","title":"Infantile Botulism (StatPearls)","publisher":"StatPearls / NCBI Bookshelf","source_type":"reputable_reference","statistic":"US infant botulism incidence ~1.9 per 100,000 live births (~77 cases/yr); in the majority of affected infants the source of C. botulinum spores is never identified; honey accounts for a minority of cases.","excerpt":"In the United States, the incidence is 1.9 per 100,000 live births, resulting in around 77 new cases each year.","source_date":"2023-08-08","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250914101618/https://www.ncbi.nlm.nih.gov/books/NBK493178/","calculation_notes":"All-cause infant botulism baseline. 1.9/100,000 live births = 0.000019 per infant over the first year. This is the headline-adjacent context figure, NOT the formula-attributable rate the question asks about. Note the ~2x conflict with CDC surveillance: 152 infant cases in 2019 against 3.67M births that year is ~4.2/100,000 (~1 in 24,000), and the CDC press release cites 150-180/yr; StatPearls' 77/yr sits at the low end. Surfaced in prose rather than averaged."},{"url":"https://www.cdc.gov/botulism/php/national-botulism-surveillance/2019.html","title":"National Botulism Surveillance Summary, 2019","publisher":"U.S. Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"In 2019, 152 laboratory-confirmed infant botulism cases were reported to CDC out of 215 total botulism cases.","excerpt":"Health departments reported 152 cases of infant botulism, all of which were laboratory-confirmed.","source_date":"2021-01-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260711113945/https://www.cdc.gov/botulism/php/national-botulism-surveillance/2019.html","calculation_notes":"Independent CDC surveillance count corroborating the all-cause baseline order of magnitude. 152 infant cases / 3,622,673 US births (2024 figure used as a stable recent denominator) approximates 4.2 per 100,000, i.e. roughly 1 in 24,000 infants in their first year from all sources combined. None of these 2019 cases were formula-linked (the first formula outbreak was 2025). Used to bracket the all-cause baseline against the lower StatPearls figure."},{"url":"https://cdc.gov/nchs/pressroom/releases/20250423.html","title":"U.S. Births Increase by 1% in 2024","publisher":"CDC / National Center for Health Statistics (NCHS)","source_type":"govt_report","statistic":"Provisional total US births in 2024: 3,622,673.","excerpt":"the number of births increased by 1% from 2023, with 3,622,673 births recorded in 2024.","source_date":"2025-04-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260618052006/https://cdc.gov/nchs/pressroom/releases/20250423.html","calculation_notes":"Supplies the annual live-birth denominator used to convert CDC infant botulism case counts into an all-cause per-infant baseline rate (e.g. 152/3,622,673 ~ 4.2 per 100,000). Used only for the context baseline; the formula-attributable rate has no defensible denominator and is left as no_reliable_estimate."}],"comparison_anchors":[],"short_label":"Botulism from baby formula","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"There is no defensible probability for the specific event the question describes — an infant developing botulism from contaminated powdered formula — because there is only one such episode on record. The November 2025 ByHeart outbreak was the first formula-linked infant botulism outbreak ever documented in the United States, and it was a single point-source contamination event that ended with a full recall. Dividing its 48 cases by the roughly 3.6 million US infants born in a year would produce a number near 1 in 75,000, but that figure dresses a one-time, non-recurring event in the clothing of a steady-state annual rate. The flag exists precisely to refuse that move.\n\nThe all-cause baseline is a different question. About 150 to 180 infant botulism cases occur in the US each year (CDC), the great majority from environmental spore exposure (soil, dust) or, less often, honey; in most cases the source is never identified. That all-cause rate works out to somewhere between roughly 1 in 24,000 and 1 in 53,000 infants in the first year of life, depending on whether one uses CDC surveillance counts (152 cases in 2019) or the lower StatPearls incidence of 1.9 per 100,000. The two figures differ by about a factor of two and are not averaged here. Whatever the exact baseline, the formula-attributable share of it was effectively zero before 2025 and remains, on the historical record, a single contained outbreak rather than an ongoing hazard.\n\nThe relevant scope is an infant in the first year of life, not a US adult lifetime; infant botulism is a condition of infancy and is not carried forward. Whether the per-infant formula risk in any future year is meaningfully above zero depends entirely on manufacturing controls and surveillance that this entry cannot forecast.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":3,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Botulism from baby formula\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/infant-formula-botulism","api_url":"https://likelier.app/api/fears/infant-formula-botulism.json"},{"slug":"infrequent-bedding-hygiene","question":"What are the odds of getting sick from rarely washed sheets, a reused towel, and sleeping in worn clothes?","quick_answer":"No reliable number exists: research measures the bacteria and allergens living on sheets and towels, but no study links a household's laundry cadence to actual illness in healthy adults. A peer-reviewed review concludes most organisms found on textiles pose no considerable health risk to an immunocompetent person.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"The combined \"poor household fabric hygiene\" profile — monthly (or longer) sheet changes, the same towel reused for weeks, sleeping in the clothes you wore all day — is a staple of hygiene-moralism articles that promise a cascade of infections for anyone who lets their laundry routine slide. The felt risk is a blur of \"acne, staph, dust mites, fungus, maybe something worse,\" rarely attached to a number, and rarely distinguished between what the fabric actually transmits and what simply lives on it. No rigorous survey asks US adults to put a probability on getting clinically sick from this pattern, so the best characterisation is that the fear outruns any specific claim.\n","rough_estimate":"Most people treat the combined pattern as vaguely hazardous without a numerical anchor","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6600116/","title":"Laundry and textile hygiene in healthcare and beyond","publisher":"Microbial Cell (Bockmühl DP, Schages J, Rehberg L)","source_type":"peer_reviewed","statistic":"Peer-reviewed review concludes that for healthy households, neither disinfection nor sterilization of laundry is necessary; most organisms found on textiles pose no considerable health risk","excerpt":"\"It can be assumed that neither a special disinfection nor a sterilization is necessary for domestic laundering processes, as far as healthy persons are concerned [...] most of the microorganisms found on textiles should not pose a considerable health risk, as long as these microorganisms are part of either the transient or the resident human skin flora.\"\n","source_date":"2019-07-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420042225/https://pmc.ncbi.nlm.nih.gov/articles/PMC6600116/","calculation_notes":"This is the most direct statement in the peer-reviewed literature on the question the fear actually asks: whether a healthy adult in a domestic setting is at measurable risk from imperfect laundry hygiene. Bockmühl et al.'s review finds the answer is essentially no — the microorganisms recovered from household sheets, towels, and clothing are overwhelmingly skin flora that pose no clinical threat to an immunocompetent host. The paper explicitly notes that the input-and-removal question (bioburden on fabric, washing conditions that reduce it) is well-studied, but the translation to infection rates in healthy households is not: no cohort comparison of \"weekly versus monthly sheet-changers\" and their illness rates exists. This is why this entry carries no_reliable_estimate rather than a number.\n","independence_note":"Peer-reviewed review in Microbial Cell; independent of CDC guidance and the AAAAI dust-mite practice parameter, which address different outcomes (dermatophyte and atopic disease respectively).\n"},{"url":"https://www.cdc.gov/ringworm/about/index.html","title":"About Ringworm","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC names shared towels and bedsheets as a dermatophyte (ringworm/tinea) transmission route","excerpt":"\"Shared objects like towels and bedsheets.\"\n","source_date":"2026-01-06","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420042300/https://www.cdc.gov/ringworm/about/index.html","calculation_notes":"CDC lists shared towels and bedsheets explicitly among ringworm transmission routes. This is the clearest authoritative acknowledgement that the mechanism exists — an already-infected source case can amplify dermatophyte spread through shared or infrequently-laundered fabric. It does not establish a background rate of tinea acquisition from one's own unwashed bedding in the absence of a source case. The transmission route is real but conditional on someone in the household already carrying viable fungal spores; \"my sheets are a month old\" is not, by itself, an exposure without a source. Anchors the regional breakdown's \"source case with active tinea\" row rather than the healthy-household headline.\n","independence_note":"CDC ringworm guidance is editorially independent of the Bockmühl review and the AAAAI practice parameter, drawing on national dermatophyte surveillance and clinical reference material.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5156485/","title":"Environmental assessment and exposure control of dust mites: a practice parameter","publisher":"Annals of Allergy, Asthma & Immunology (Portnoy J, et al.; AAAAI/ACAAI Joint Task Force)","source_type":"peer_reviewed","statistic":"Weekly washing of bedding effectively removes mite allergens and kills mites; mite avoidance reduces bronchial hyper-responsiveness and medication need in sensitised patients","excerpt":"\"Regular washing of bedding and clothing has been shown to effectively remove mite allergens and to kill mites. [...] Advise patients that bedding should be washed weekly to decrease dust mite numbers and mite allergen levels, and that high temperature is not necessary. [...] Avoidance of allergens can lead to decreased bronchial hyper-responsiveness, decreased morbidity, and decreased need for medications.\"\n","source_date":"2013-12-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420042332/https://pmc.ncbi.nlm.nih.gov/articles/PMC5156485/","calculation_notes":"This is the key carve-out. The AAAAI/ACAAI practice parameter documents a measurable clinical benefit from weekly bedding washing — but the outcome is asthma and allergic rhinitis control in dust-mite-sensitised atopic patients, not infection. For the atopic subgroup this is a real, dose-dependent effect; the parameter cites prospective studies finding mite avoidance lowers asthma risk \"in a dose-dependent manner\". NHANES 2005-2006 data (Salo, Arbes et al.) put US dust-mite sensitisation in the general population at roughly 20-25% depending on case definition, so the atopic carve-out is not a small corner of the population. This source supplies the \"atopic adult\" row of the regional breakdown and the caveat language.\n","independence_note":"Joint practice parameter from two independent allergy societies, distinct from CDC and Bockmühl sources in both methodology (clinical practice parameter synthesising intervention trials) and outcome measured (allergen-driven disease rather than infection).\n"},{"url":"https://www.lung.org/clean-air/indoor-air/indoor-air-pollutants/dust-mites","title":"Dust Mites","publisher":"American Lung Association","source_type":"reputable_reference","statistic":"Roughly 4 in 5 US homes have dust mite allergens in at least one bed; weekly hot-water bedding wash recommended for atopic individuals","excerpt":"\"Dust mites are one of the major indoor triggers for people with asthma. [...] roughly four out of five homes in the United States have dust mite allergens in at least one bed. [...] Wash bedding in hot water (at least 120 degrees F) once a week.\"\n","source_date":"2025-04-22","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420042407/https://www.lung.org/clean-air/indoor-air/indoor-air-pollutants/dust-mites","calculation_notes":"Corroborates the AAAAI parameter and supplies the household-prevalence anchor: the allergen is effectively ambient in US bedding, which is why the atopic carve-out matters as much as it does. This is also the cleanest patient-facing source for the \"weekly hot-water wash\" recommendation. Not an independent clinical study; ALA is a public-education body aggregating the same evidence base as AAAAI, used here as a reputable_reference to give the prevalence number and the recommendation in plain form.\n","independence_note":"ALA's guidance aggregates AAAAI/NIH-funded allergen research rather than generating independent data; treat as corroborating the practice parameter, not a second data point.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/19759479/","title":"Staphylococcus aureus recovery from cotton towels","publisher":"Journal of Infection in Developing Countries (via PubMed)","source_type":"peer_reviewed","statistic":"S. aureus viability on cotton textiles declines over ~48 hours but sufficient bacteria may remain for colonisation; more absorbent fabrics retain more bacteria","excerpt":"\"Cell viability decreased for over 48 hours on towels, but sufficient quantities may remain for colonization. More absorbent towels may harbor more Staphylococci than less absorbent ones, and may serve as a transmission mechanism for the bacterium.\"\n","source_date":"2009-09-15","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260420042550/https://pubmed.ncbi.nlm.nih.gov/19759479/","calculation_notes":"Establishes the bioburden half of the mechanism — S. aureus and other skin flora do persist on household fabric, with viability measured in days on absorbent cotton. This is the real evidence base behind \"your towel has bacteria on it\" articles. The critical gap the Bockmühl review identifies is that recovering viable organisms from fabric is not the same as a colonisation or infection event in a healthy host; the same recovery studies do not translate bioburden into a per-exposure infection rate. Included here to show that the mechanism is real — the fear is not imaginary — while the outcome (clinical illness in a healthy person) is not what the bioburden studies measure.\n","independence_note":"In vitro recovery study, independent of CDC and Bockmühl in both authorship and methodology (laboratory viability work rather than epidemiology or review).\n"}],"comparison_anchors":[{"label":"Serious infection from one shared-cup sip (healthy adults)","lifetime_us_adult":0.000001},{"label":"Serious infection from casual dry towel sharing","lifetime_us_adult":0.000001},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Healthy non-atopic adult, conservative hygiene (weekly-ish sheets, fresh towel, pajamas)","probability":0.0001,"notes":"Baseline. Whatever infection risk exists from household fabric in a healthy adult is effectively indistinguishable from normal life at this frequency; the number is a placeholder for \"below detection.\"\n"},{"region":"Healthy non-atopic adult doing all three (monthly sheets, same towel for weeks, sleeping in day clothes)","probability":0.001,"notes":"Structural upper bound for clinical infection per year in a healthy host. No cohort study measures this; the figure reflects \"the mechanism is real but the outcome mostly isn't\" rather than a measured rate. The relevant increments are folliculitis, acne mechanica, and odd occasional S. aureus skin infections, not systemic illness.\n"},{"region":"Dust-mite-sensitised adult (roughly 20-25% of US population per NHANES)","probability":0.3,"notes":"Real, documented, dose-dependent. Weekly bedding washing measurably reduces asthma exacerbations and allergic rhinitis symptoms in atopic patients per the AAAAI/ACAAI practice parameter. This is an allergen exposure effect, not an infection; the probability is a rough annualised likelihood of at least one symptomatic episode attributable to elevated mite load from infrequent washing in a sensitised adult, not a per-night rate.\n"},{"region":"Household with an active source case (tinea, scabies, MRSA skin infection, head lice)","probability":0.1,"notes":"Different regime entirely. Sharing a bed, towel, or clothing with someone actively shedding dermatophyte spores, scabies mites, or MRSA, plus infrequent laundering, is the scenario CDC's fomite guidance is written for. Infrequent washing amplifies an existing outbreak; it does not create one from nothing.\n"},{"region":"Immunocompromised adult (transplant, chemotherapy, advanced HIV, long-term immunosuppression)","probability":0.05,"notes":"Rough order-of-magnitude placeholder. Skin-flora bioburden that an intact immune system clears without notice can become clinically significant here. Specific clinical guidance for this population lives outside the scope of this entry.\n"}],"short_label":"Unwashed bedding","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline \"no measurable infection risk for a healthy adult\" is specifically about clinical infection, and specifically about healthy adults. Four carve-outs matter and are handled in the regional breakdown rather than in the headline. First, and most important: atopic individuals sensitised to house dust mites — roughly 20-25% of the US population per NHANES 2005-2006 data — get a measurable, dose-dependent clinical benefit from weekly bedding washing. The AAAAI/ACAAI practice parameter states this plainly, and the outcome being reduced is asthma and allergic rhinitis morbidity, not infection. For atopic individuals, the hygiene-moralism articles are directionally correct about the conclusion even if wrong about the mechanism. Second, a household with an active source case — visible tinea, scabies, MRSA skin lesion, head lice — genuinely changes the calculation for fomite transmission, and CDC has written specific guidance for exactly those scenarios. Third, immunocompromised adults, infants, and people with active acne or folliculitis sit outside the \"healthy adult\" framing entirely. Fourth, nothing here speaks to body odour, comfort, or social consequences, which are reasonable reasons to change fabric that do not need to be laundered (so to speak) as infection-risk arguments. The sheet of paper between \"my sheets are a month old\" and \"I will get sick\" is genuinely thinner than hygiene articles suggest, outside those specific subgroups.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-7-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single neatly folded blanket rendered as a flat vector shape in muted dusty blue and off-white, with plenty of empty space around it."},"canonical_url":"https://likelier.app/infrequent-bedding-hygiene","api_url":"https://likelier.app/api/fears/infrequent-bedding-hygiene.json"},{"slug":"infrequent-fridge-cleaning","question":"What are the odds of getting sick from rarely cleaning the inside of the fridge?","quick_answer":"No reliable number exists: studies document what grows inside home fridges and that less cleaning means more bacteria, but none tracks whether people with dirtier fridges actually get sick more often. The illness-outcome side of the question has never been measured in the form it asks.\n","category":"food","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"The cultural fear is built on a steady drip of \"your fridge has more bacteria than a toilet seat\" headlines, recurring Listeria-in-the-deli-drawer recall warnings, and a general anxiety about invisible contamination hiding behind condiment jars. The felt risk is some vague composite of food poisoning, listeriosis, \"stomach bugs\", and a sense that a sticky shelf is one missed wipe-down away from a hospital visit. No survey asks US adults to put a probability on getting clinically sick from skipping fridge cleanings, and no cohort study links household cleaning cadence to incident gastrointestinal illness in healthy adults. The fear circulates without a numerical anchor, which is part of why the headlines keep landing.\n","rough_estimate":"Most people treat a long-uncleaned fridge as vaguely hazardous without attaching any specific probability to it","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/32577758/","title":"Microbial contamination and cleaning practices in domestic refrigerators in Japan","publisher":"Journal of Food Protection (Yamaguchi N, et al.)","source_type":"peer_reviewed","statistic":"100 Japanese domestic refrigerators; cleaning frequency correlated with microbial contamination; S. aureus was the most frequent isolate; Salmonella, L. monocytogenes, and Y. enterocolitica were not recovered; only 17% cleaned the fridge monthly or more often","excerpt":"[Paraphrase from abstract — full text paywalled] \"Staphylococcus aureus was the most frequently isolated pathogen [...] Salmonella, Listeria monocytogenes, and Yersinia enterocolitica were not found [...] Correlations were found between microbial contamination and refrigerator cleaning frequency and/or method [...] Only 17% of the respondents cleaned their refrigerators monthly or more often, and this frequency was lower than that reported in other countries.\"\n","source_date":"2020-06-22","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250202185514/https://pubmed.ncbi.nlm.nih.gov/32577758/","calculation_notes":"The cleanest published evidence on the actual question. Yamaguchi et al. swabbed 100 home refrigerators in Japan, recovered Staph aureus as the most frequent pathogen and coliforms in multiple households, but did not recover Salmonella, Listeria, or Yersinia — the three classical refrigeration-relevant pathogens. Cleaning frequency was identified as a significant correlate of bioburden, while temperature alone was not. This documents the mechanism (less cleaning → more bacteria on surfaces) without translating it into an infection rate; the study measures what is on the fabric of the fridge, not whether residents of dirtier fridges got sick more often. This is exactly the gap that justifies the no_reliable_estimate flag for this entry.\n","independence_note":"Independent Japanese household survey, distinct in methodology and population from the US home-kitchen study (Borrusso & Quinlan) and the Irish home-refrigerator survey (Kennedy).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/28271927/","title":"Prevalence of Pathogens and Indicator Organisms in Home Kitchens and Correlation with Unsafe Food Handling Practices and Conditions","publisher":"Journal of Food Protection (Borrusso PA, Quinlan JJ)","source_type":"peer_reviewed","statistic":"Listeria spp., including L. monocytogenes and L. innocua, present in 15% of US homes; contamination significantly associated with higher refrigerator temperatures; most often found in samples from refrigerator meat drawers","excerpt":"[Paraphrase from abstract — full text paywalled] \"Listeria spp., including L. monocytogenes and L. innocua, were present in 15% of homes [...] Contamination with Listeria was significantly associated with higher refrigerator temperatures [...] most often in samples from refrigerator meat drawers.\"\n","source_date":"2017-03-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260610131316/https://pubmed.ncbi.nlm.nih.gov/28271927/","calculation_notes":"Direct US prevalence data for the pathogen that most concerns regulators and the pregnancy-advice literature. Roughly one in seven US homes harbours Listeria somewhere in the refrigerator, concentrated in the meat drawer, with the temperature association doing most of the explanatory work rather than cleaning cadence per se. The crucial gap, again: 15% of homes carrying Listeria does not translate to 15% of households having a listeriosis case — invasive listeriosis incidence in the US is roughly 0.26 per 100,000 per year (CDC FoodNet) across the general population, several orders of magnitude below the home-prevalence number. Most household Listeria does not cause clinical disease in immunocompetent adults; this is what makes the pregnancy and immunocompromised carve-outs the entire story.\n","independence_note":"US national surveillance study, methodologically and editorially independent of the Japan and Ireland surveys; same outcome (fridge bioburden) measured against a different population.\n"},{"url":"https://www.cdc.gov/food-safety/about/index.html","title":"About Food Safety","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC estimates 48 million US foodborne illnesses, 128,000 hospitalizations, and 3,000 deaths per year","excerpt":"\"CDC estimates that each year 48 million people get sick from a foodborne illness, 128,000 are hospitalized, and 3,000 die.\"\n","source_date":"2024-08-13","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260618000252/https://www.cdc.gov/food-safety/about/index.html","calculation_notes":"Frames the absolute burden of foodborne illness in the US: roughly one in seven Americans has a foodborne illness episode in a given year. This is the headline number the contamination-anxiety articles implicitly invoke. What it cannot do is attribute any fraction of that burden to refrigerator cleaning frequency. CDC's attribution work apportions illness across food categories (poultry, produce, dairy, deli) and pathogens (Salmonella, Norovirus, Campylobacter, Listeria); the \"in-home refrigerator cleaning cadence\" exposure is not a category the surveillance system tracks because the data needed to estimate it does not exist. Anchors the absolute scale of foodborne illness while making clear why no slice of it can be cleanly assigned to the question this entry asks.\n","independence_note":"Federal surveillance estimate, independent of all three peer-reviewed bioburden studies; provides denominator and population context rather than mechanism.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16013380/","title":"Food safety knowledge and microbiological hygiene in domestic refrigerators on the island of Ireland","publisher":"Journal of Food Protection (Kennedy J, Jackson V, Blair IS, McDowell DA, Cowan C, Bolton DJ)","source_type":"peer_reviewed","statistic":"Representative sample of 1,020 Irish households; Staphylococcus aureus in 41%, Salmonella enterica in 7%, Listeria monocytogenes in 6%, Escherichia coli in 6% of domestic refrigerators","excerpt":"[Paraphrase from abstract — full text paywalled] \"A representative sample of households (n = 1,020) throughout the island of Ireland [...] Staphylococcus aureus [...] 41% [...] Salmonella enterica [...] 7% [...] Listeria monocytogenes [...] 6% [...] Escherichia coli [...] 6% [...] consumers with stronger food safety knowledge demonstrated lower bacterial contamination in their refrigerators and reported fewer foodborne illnesses.\"\n","source_date":"2005-06-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250721182826/https://pubmed.ncbi.nlm.nih.gov/16013380/","calculation_notes":"Largest published household-refrigerator survey by sample size (n=1,020), useful for two reasons. First, it puts a number on each of the relevant pathogens in home fridges in a Western population, broadly consistent with the US Listeria figure (6% Irish vs 15% US — different sampling regimes, both single-digit-to- teens prevalence). Second, the authors report a correlation between consumer food safety knowledge and both refrigerator contamination and self-reported foodborne illness — the closest the literature comes to linking household hygiene practices to illness outcomes. The self-report on the illness side is the limit: it is not a measured clinical attack rate, it is what people remember when surveyed.\n","independence_note":"Independent Irish national survey; cited here for sample-size weight and the knowledge-contamination correlation rather than as a substitute for US-specific figures.\n"}],"comparison_anchors":[{"label":"Serious infection from one shared-cup sip (healthy adults)","lifetime_us_adult":0.000001},{"label":"Serious infection from casual dry towel sharing","lifetime_us_adult":0.000001},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Healthy non-pregnant adult, conservative practice (clean spills promptly, fridge at or below 40°F, wipe-out every month or two)","probability":0.0001,"notes":"Baseline. Whatever incremental GI-illness risk is attributable specifically to fridge-cleaning cadence in a healthy adult sits below detection against the ambient ~14% annual foodborne illness rate from all sources.\n"},{"region":"Healthy non-pregnant adult who essentially never cleans the inside of the fridge (years between wipe-outs, sticky shelves, expired condiments)","probability":0.001,"notes":"Structural upper bound for a clinical infection per year attributable to cleaning cadence specifically, in a healthy host. Not a measured rate — the cohort comparison does not exist. The plausible increments are occasional cross-contamination episodes from a leaking package onto ready-to-eat food, not chronic illness from the fridge's general state.\n"},{"region":"Immunocompromised adult (transplant, chemotherapy, advanced HIV, long-term immunosuppression)","probability":0.01,"notes":"Order-of-magnitude placeholder. The immunocompromised population is the one for whom Listeria in the meat drawer is a real clinical threat rather than an academic one. Specific guidance for this group lives outside the scope of this entry; CDC, transplant centres, and oncology teams issue tailored food-safety instructions that go well beyond \"clean the fridge\".\n"},{"region":"Pregnant adult (listeriosis-specific risk)","probability":0.002,"notes":"Pregnancy is the defining carve-out. US invasive listeriosis incidence in pregnancy is roughly 3 per 100,000 pregnancies (CDC FoodNet pregnancy surveillance); the figure shown is a rough annualised probability of any listeriosis-attributable illness over a pregnancy duration, not per-meal risk. The mechanism that makes fridge cleaning matter at all sits here, in the deli-meat-and-soft-cheese cross-contamination pathway, not in the general \"stomach bug\" framing.\n"},{"region":"Household with a known raw-juice spill (poultry, ground beef, raw fish drip onto a ready-to-eat shelf, not cleaned)","probability":0.05,"notes":"Different regime entirely. The published USDA and CDC guidance for \"raw juice onto a shelf that then touched salad greens\" describes a real, near-term Salmonella or Campylobacter exposure event with a measurable per-incident attack rate. This is what fridge-cleaning advice is genuinely written for — discrete spill events, not \"I should wipe the shelves monthly\".\n"}],"short_label":"Dirty fridge","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline \"no measurable cleaning-cadence effect on clinical illness in a healthy adult\" carves out four populations that change the answer materially. First, pregnancy: invasive listeriosis is rare in absolute terms (~3 per 100,000 pregnancies in US surveillance) but disproportionately catastrophic when it occurs, and the meat-drawer cross-contamination pathway documented by Borrusso and Quinlan is the mechanism the CDC pregnancy-food-safety guidance is written for. Second, immunocompromised adults — solid organ transplant, active chemotherapy, advanced HIV, long-term immunosuppressive therapy — for whom Listeria, Salmonella, and Yersinia recovered from home refrigerators are genuine clinical threats rather than skin-flora curiosities, and for whom specialist food-safety instruction overrides anything written for the general population. Third, infants and young children sit outside the \"healthy adult\" framing entirely. Fourth, the argument here is about cleaning cadence — sticky shelves, expired jars, the general \"ugh\" of an old fridge — not about spill events. A raw-poultry juice drip onto a ready-to-eat shelf is a discrete, well-documented exposure event for which the USDA and CDC have specific cleanup guidance, and \"I never clean my fridge\" is genuinely different from \"I left a spill of chicken drippings for three days\". Visible contamination is a separate category from background grime.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-fridge-toilet-batch","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"A single closed glass jar rendered as a flat vector shape in muted dusty sage and off-white, with plenty of empty space around it."},"canonical_url":"https://likelier.app/infrequent-fridge-cleaning","api_url":"https://likelier.app/api/fears/infrequent-fridge-cleaning.json"},{"slug":"infrequent-toilet-bowl-cleaning","question":"What are the odds of getting sick from rarely scrubbing the toilet bowl?","quick_answer":"There is no reliable number: no study measures how often infrequent toilet cleaning actually makes a healthy adult sick. The bioburden is real and flushing spreads aerosols, but whether that translates to clinical illness has never been measured — the risk climbs mainly when someone in the household is already shedding a pathogen like norovirus.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"The cultural script around toilet hygiene is a tangle of three separate fears that rarely get distinguished. The bioburden trope (\"your toilet seat has more bacteria per square inch than X\") borrows the imagery of a famous Charles Gerba quotation series without the conditional clauses around it. The toilet-plume panic, refreshed every few years by viral CGI of the flush aerosol, conflates \"aerosols are produced\" with \"people get sick from them\". And the general visible-grime instinct treats a ring around the waterline as a proxy for personal illness risk. No rigorous survey asks healthy US adults to attach a probability to \"I will get clinically sick because I cleaned the toilet bowl monthly instead of weekly\", so the felt risk runs well ahead of any specific claim — which is exactly the gap this entry is built to name.\n","rough_estimate":"Most people treat a poorly cleaned toilet as vaguely hazardous without a numerical anchor","kind":"intuition"},"sources":[{"url":"https://www.mdpi.com/2673-947X/5/2/22","title":"Impact of Different Toilet Cleaning/Disinfecting Regimens on Reducing the Risk of Exposure to Toilet-Borne Pathogens in American Household Restrooms","publisher":"Hygiene (MDPI) — Boone SA, Childress ND, Silva-Beltrán NP, McKinney J, Ijaz MK, Gerba CP (University of Arizona)","source_type":"primary_study","statistic":"QMRA modelling estimates a >98% reduction in norovirus infection risk with every-3-day cleaning/disinfection of the most heavily contaminated fomites; relative-risk reduction in a model, not a measured population infection rate","excerpt":"\"A quantitative microbial risk assessment (QMRA) estimated a >98% reduction in risk of infection by norovirus with an every-3-day cleaning/disinfection routine on the most heavily contaminated sites. Results indicate an optimal cleaning frequency of twice weekly for minimizing exposure to pathogens.\"\n","source_date":"2025-06-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20250913020531/https://www.mdpi.com/2673-947X/5/2/22","calculation_notes":"This is the closest the literature comes to quantifying the cleaning-frequency question, and it is decisively *modelled* rather than measured. The Phase 2 intervention modelled the change in norovirus infection probability from a single restroom visit using the Abney et al. (2024) QMRA framework, parametrised by surface E. coli counts as a proxy for norovirus load. The output is a *relative* risk reduction (>98%, or in the discussion section >94.3% for the countertop and >99.7% for the toilet seat) — not an absolute lifetime probability of a healthy adult getting sick. There is no contemporaneous cohort of weekly versus monthly cleaners with measured incident-illness rates to multiply this number against, which is why this entry carries no_reliable_estimate rather than a composite figure. Critically, the QMRA assumes a household member is already shedding norovirus; that conditional is doing most of the work.\n","independence_note":"University of Arizona research group; the manuscript discloses Reckitt Benckiser funding and two authors employed by Reckitt. Used here for the QMRA structure and modelled relative risk, not for absolute infection rate claims.\n"},{"url":"https://www.mdpi.com/2673-947X/5/3/27","title":"Bacterial Contamination of Public and Household Restrooms, and Implications for the Potential Risk of Norovirus Transmission","publisher":"Hygiene (MDPI) — Gerba CP, Boone SA, McKinney J, Ijaz MK (University of Arizona)","source_type":"primary_study","statistic":"Heterotrophic bacteria and coliform counts are 10–100x greater in household restrooms than in public office restrooms; greatest contamination is on countertops and floors, not the bowl interior","excerpt":"\"Numbers of heterotrophic bacteria and coliforms were 10 to 100 times greater in household restrooms than in public restrooms. The QMRA suggested that the greatest risk of acquiring a norovirus infection involved the touching of the countertops in household restrooms and the toilet flush handle in public restrooms.\"\n","source_date":"2025-09-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260422164226/https://www.mdpi.com/2673-947X/5/3/27","calculation_notes":"Establishes the bioburden anchor: household restrooms carry an order of magnitude more recoverable heterotrophic bacteria and coliforms than moderate-use public office restrooms. Also reframes where the contamination is — the highest counts land on the vanity countertop and the flush handle, not the inside of the bowl that the visible-grime instinct fixates on. This is the cleanest source for the claim that bioburden is real and quantifiable, and simultaneously for the claim that the location of greatest contact risk is not the place most people associate with \"cleaning the toilet\". As with the companion paper above, the link from these bacterial counts to clinical illness in a healthy adult is asserted via QMRA, not measured cohort by cohort.\n","independence_note":"Same Arizona/Reckitt group as the companion paper; the two manuscripts share authors and funder. Treat as one evidence stream rather than two independent data points. Independent corroboration of the bioburden direction (toilet surfaces harbour skin flora and enteric bacteria) is the longstanding Gerba et al. 1975 droplet work cited within the paper.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23040490/","title":"Lifting the lid on toilet plume aerosol: a literature review with suggestions for future research","publisher":"American Journal of Infection Control (Johnson DL, Mead KR, Lynch RA, Hirst DVL; via PubMed)","source_type":"peer_reviewed","statistic":"Toilet flushing produces measurable bioaerosols, but no study has clearly demonstrated or refuted plume-borne disease transmission to an exposed person","excerpt":"\"However, no studies have yet clearly demonstrated or refuted toilet plume-related disease transmission, and the significance of the risk remains largely uncharacterized.\"\n","source_date":"2013-03-01","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260608175414/https://pubmed.ncbi.nlm.nih.gov/23040490","calculation_notes":"The load-bearing source for this entry. A formal AJIC literature review concludes that the *mechanism* (aerosol production during flushing) is well-evidenced and the *outcome* (plume-attributable infection in an exposed bystander) is neither demonstrated nor refuted. This is exactly the gap that justifies the no_reliable_estimate flag: bacteria and viruses are aerosolised, deposited on nearby surfaces, and recoverable for hours to days, but the per-flush or per-week probability that a healthy adult acquires clinical illness from this pathway has not been measured. Later bioaerosol work refines the deposition picture but does not close the disease-attribution gap for household settings.\n","independence_note":"Federal (NIOSH) and academic authorship, AJIC peer-reviewed; methodologically and editorially independent of the Arizona/Reckitt QMRA programme.\n"},{"url":"https://www.cdc.gov/norovirus/causes/index.html","title":"How Norovirus Spreads","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC lists contaminated surfaces, contact with vomit/diarrhea splatter, and contaminated objects as documented norovirus transmission pathways within households","excerpt":"\"A person with norovirus touches surfaces with their bare hands. [...] Food, water, or objects that are contaminated with norovirus are placed on surfaces. [...] Tiny drops of vomit from a person with norovirus spray through the air, landing on surfaces or entering another person's mouth. [...] A person with norovirus has diarrhea that splatters onto surfaces.\"\n","source_date":"2024-04-12","source_accessed":"2026-05-31","archive_url":"http://web.archive.org/web/20260725184719/https://www.cdc.gov/norovirus/causes/index.html","calculation_notes":"Anchors the conditional carve-out: when someone in the household is actively shedding norovirus, the bathroom becomes an epidemiologically relevant surface, and infrequent cleaning amplifies an exposure chain that already exists. CDC does not publish a household secondary attack rate on this page; literature estimates for household norovirus secondary attack rates run roughly 20–40% depending on the outbreak setting and family composition, which sits well outside the \"healthy household, no index case\" regime that the headline of this entry is about. This source is the cleanest official acknowledgement that the mechanism is real and lives in the bathroom; it does not validate a cleaning-frequency-to-illness curve.\n","independence_note":"CDC public-health guidance, editorially independent of the Arizona/Reckitt programme and of the AJIC plume review. Used here as the conditional backstop, not as a population infection-rate source.\n"}],"comparison_anchors":[{"label":"Serious infection from one shared-cup sip (healthy adults)","lifetime_us_adult":0.000001},{"label":"Serious infection from casual dry towel sharing","lifetime_us_adult":0.000001},{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354}],"regional_breakdown":[{"region":"Healthy adult, conservative cleaning (weekly-ish bowl scrub, separate cleaner for high-touch surfaces)","probability":0.0001,"notes":"Baseline. Whatever infection risk exists from household toilet surfaces in a healthy adult is effectively indistinguishable from normal daily life at this frequency; the number is a placeholder for \"below detection\".\n"},{"region":"Healthy adult, near-never cleaning (visible ring, infrequent or absent scrubbing)","probability":0.001,"notes":"Structural upper bound for clinical infection per year in a healthy host. No cohort study measures this; the figure reflects \"the mechanism is real but the outcome mostly is not\" rather than a measured rate. The relevant increments are occasional self-inoculation events with skin flora, not systemic illness.\n"},{"region":"Household with an active norovirus index case in the last 14 days","probability":0.3,"notes":"Different regime entirely. Household secondary attack rates for norovirus in outbreak literature run roughly 20–40%; CDC notes the virus persists in stool for two-plus weeks after symptoms resolve. Infrequent cleaning amplifies an already-active transmission chain; it does not create one from nothing. This figure annualises a conditional outbreak probability for a household that experiences an index case.\n"},{"region":"Household with an active C. difficile, Salmonella, or rotavirus case","probability":0.1,"notes":"Salmonella has been recovered from toilet surfaces for extended periods after a salmonellosis case in the home. C. diff spores resist most quaternary-ammonium cleaners and require bleach. As with norovirus, \"infrequent cleaning\" matters most when there is a source case; \"infrequent cleaning\" alone is not the exposure.\n"},{"region":"Immunocompromised adult (transplant, chemotherapy, advanced HIV, long-term immunosuppression)","probability":0.05,"notes":"Rough order-of-magnitude placeholder. Surface-borne enteric bacteria and norovirus that an intact immune system clears without notice can produce clinical disease here. Specific guidance for this population sits outside the scope of this entry.\n"}],"short_label":"Dirty toilet","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"The headline \"no measurable infection signal from cleaning frequency alone in a healthy adult\" is specifically about clinical infection, specifically about healthy adults, and specifically about the absence of an active source case in the household. What is known is solid: heterotrophic bacteria and coliforms are recoverable from toilet rims, seats, flush handles, and adjacent countertops in numbers roughly an order of magnitude higher in household restrooms than in moderately used public ones (Gerba et al. 2025); flushing produces a measurable bioaerosol that deposits viable organisms on nearby surfaces (Johnson et al. 2013); QMRA modelling predicts large *relative* risk reductions from every-three-day disinfection of the highest-contact surfaces (Boone et al. 2025, >98% modelled risk reduction). What is not known, and where this entry stops, is the absolute clinical-illness rate that cleaning frequency itself controls in a healthy adult with no household index case. No cohort compares weekly versus rarely cleaned households for incident GI illness in that population. Multiplying the QMRA relative-risk number by published norovirus household secondary attack rates (~20–40% in outbreak settings per CDC and household studies) would compose a model on top of an outbreak-conditional rate and present it as if it answered \"how often does skipping cleaning make me sick\" — which it does not. The four carve-outs in the regional breakdown are where the answer genuinely changes: an active norovirus, salmonella, C. difficile or rotavirus index case in the household; immunocompromised adults; infants in the toilet-touching age band; and any setting where ventilation is poor enough that plume deposition reaches food-preparation surfaces. Outside those, the gap between cultural panic about a dirty toilet and what it measurably does to a healthy person is wider than the hygiene-content economy suggests.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.63,"scored_by":"claude-code-8d","scored_at":"2026-05-31","methodology_version":"1.2"},"reviewer":"claude-code-fridge-toilet-batch","last_reviewed":"2026-05-31","reviewed":true,"generated_at":"2026-05-31","image":{"alt":"A single closed amber-tinted spray bottle rendered as a flat vector shape in muted dusty teal and warm grey, sitting alone in a wide empty composition with no label, no liquid visible, and no bathroom context."},"canonical_url":"https://likelier.app/infrequent-toilet-bowl-cleaning","api_url":"https://likelier.app/api/fears/infrequent-toilet-bowl-cleaning.json"},{"slug":"insurance-claim-denial","question":"What are the odds of a major insurance claim being denied or severely underpaid?","quick_answer":"There is no reliable figure: published denial rates are per-claim, not per-lifetime, and no source supports a cumulative chance of ever being denied. Denials are common on any given claim, yet almost no one appeals, so the insurer's initial decision usually sticks.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Insurance claim denials occupy a peculiar place in public consciousness: everyone has heard a horror story, but most policyholders assume their own claims will be paid. The complexity of insurance contracts and the opacity of denial reasons contribute to a learned helplessness — denial rates are high, but because each individual encounters the system infrequently, the base rate remains poorly calibrated. Health insurance denials receive the most media attention, particularly after the 2024 UnitedHealthcare shooting brought insurer practices into national conversation, but denial rates for disability, long-term care, and property insurance are often higher and less visible.\n","rough_estimate":"~15-20% of health claims denied","kind":"intuition"},"native":{"display":"~19% of in-network health claims denied (ACA marketplace, 2024)","numerator":19,"denominator":100,"unit":"share of in-network claims denied","population":"ACA marketplace plan enrollees (HealthCare.gov, 2024)"},"sources":[{"url":"https://www.kff.org/patient-consumer-protections/claims-denials-and-appeals-in-aca-marketplace-plans-in-2024/","title":"Claims Denials and Appeals in ACA Marketplace Plans in 2024","publisher":"KFF (Kaiser Family Foundation)","source_type":"primary_study","statistic":"19% of in-network claims denied in 2024 (~85 million); 37% of out-of-network claims denied; less than 1% of denied claims appealed","excerpt":"\"Of these in-network claims, approximately 85 million were ultimately denied, resulting in an average in-network denial rate of 19%. [...] Of the approximately 85 million in-network denied claims in 2024, HealthCare.gov consumers appealed at least 262,982 — an appeal rate of less than 1%.\"\n","source_date":"2026-03-24","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260519191142/https://www.kff.org/patient-consumer-protections/claims-denials-and-appeals-in-aca-marketplace-plans-in-2024/","calculation_notes":"KFF's analysis of the CMS Transparency in Coverage Public Use File provides the most comprehensive data on ACA marketplace denial rates. The 19% in-network denial rate is the basis for the native estimate. The strikingly low appeal rate (<1%) suggests that the vast majority of denials go unchallenged, meaning the effective denial rate — denials that result in the policyholder bearing the cost — is very close to the raw denial rate. The 19% figure covers all claim types (administrative, medical necessity, prior auth); the subset of denials for medical necessity is roughly 5% of all denials.\n","independence_note":"KFF's analysis uses CMS-mandated insurer transparency filings, independent from individual insurer self-reports and from the SSA's disability claims data.\n"},{"url":"https://www.urban.org/urban-wire/ssa-says-its-reduced-disability-claims-backlog-fewer-new-claims-and-higher-denial-rate","title":"The SSA Says It's Reduced the Disability Claims Backlog. Fewer New Claims and a Higher Denial Rate Could Be Driving the Reduction","publisher":"Urban Institute","source_type":"primary_study","statistic":"62% of SSDI claims denied at initial application in 2024; approval rate fell to 36% in FY2025","excerpt":"\"The SSA's approval rate fell from 38.7 percent in fiscal year 2024 to an average of 36.0 percent in fiscal year 2025. While the number of approved claims remained flat at about 812,000 from 2024 to 2025, denials account for the entire increase in total decisions.\"\n","source_date":"2025-09-12","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260206150925/https://www.urban.org/urban-wire/ssa-says-its-reduced-disability-claims-backlog-fewer-new-claims-and-higher-denial-rate","calculation_notes":"The SSDI initial denial rate of 62% represents one of the highest denial rates in any insurance-adjacent system. At reconsideration, 84% are denied again; at the ALJ hearing level, 51% are finally approved. The multi-stage process means that a claimant who persists through all appeals has roughly a 50% chance of eventual approval, but the process takes 1-3 years — during which the claimant has no disability income. This is included as a separate data point because disability insurance denial is a distinct and severe category of financial harm.\n","independence_note":"The Urban Institute analysis uses SSA administrative data, independent from the KFF health insurance claims analysis which uses CMS marketplace data.\n"},{"url":"https://www.ajmc.com/view/how-insurance-claim-denials-harm-patients-health-finances","title":"How Insurance Claim Denials Harm Patients' Health, Finances","publisher":"American Journal of Managed Care","source_type":"reputable_reference","statistic":"Patients who experience claim denials report delayed care, medical debt, and reduced trust in the insurance system","excerpt":"\"Insurance claim denials harm patients' health and finances, leading to delayed or foregone care, unexpected medical debt, and erosion of trust in the insurance system. The financial impact falls disproportionately on lower-income and chronically ill patients.\"\n","source_date":"2025-01-15","source_accessed":"2026-04-24","archive_url":"http://web.archive.org/web/20260226180630/https://www.ajmc.com/view/how-insurance-claim-denials-harm-patients-health-finances","calculation_notes":"AJMC provides qualitative context on the downstream effects of denials. While not a quantitative source for denial rates, it documents the mechanism by which denials translate into financial harm: patients who cannot afford to pay out-of-pocket either forgo care (creating future health costs) or incur medical debt. This supports the framing of denials as a financial risk, not merely an administrative inconvenience.\n"},{"url":"https://www.aha.org/news/headline/2024-10-17-senate-report-scrutinizes-medicare-advantage-prior-authorization-denials-post-acute-care-services","title":"Senate report scrutinizes Medicare Advantage prior authorization denials for post-acute care services","publisher":"American Hospital Association","source_type":"reputable_reference","statistic":"In 2022, UnitedHealthcare and CVS denied prior authorization for post-acute care at approximately three times their overall denial rates; Humana's post-acute denial rate was more than 16 times its overall rate (U.S. Senate Permanent Subcommittee on Investigations, 'Refusal of Recovery,' Oct 2024).","excerpt":"\"In 2022, UHC and CVS denied prior authorization requests for post-acute care at approximately three times higher than the companies' overall denial rates, while Humana's prior authorization denial rate for post-acute care was more than 16 times higher than its overall denial rate.\"\n","source_date":"2024-10-17","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260214050332/https://www.aha.org/news/headline/2024-10-17-senate-report-scrutinizes-medicare-advantage-prior-authorization-denials-post-acute-care-services","calculation_notes":"The American Hospital Association summarizes the U.S. Senate Permanent Subcommittee on Investigations Majority Staff Report \"Refusal of Recovery: How Medicare Advantage Insurers Have Denied Patients Access to Post-Acute Care\" (Oct 17, 2024), which analyzed over 280,000 internal documents from the three largest Medicare Advantage insurers (2019-2022). The ~3x post-acute figure for UnitedHealthcare and CVS is used as a personal_factor_multiplier: a policyholder whose claim is for post-acute care (skilled-nursing or inpatient-rehab after a hospital stay) under one of these plans faces roughly three times that insurer's own overall denial rate — a within-insurer ratio, not 3x this entry's 19% ACA in-network baseline. The Senate report attributes much of the elevation to AI/predictive systems that flag or auto-deny costly post-acute requests. The primary Senate PDF (hsgac.senate.gov) is bot-protected and returns 403 to automated fetch; this AHA summary carries the verbatim finding.\n","independence_note":"The underlying Senate PSI investigation used insurer internal documents obtained under subpoena, independent from the CMS marketplace data (KFF) and the SSA disability data (Urban Institute) used by the other sources.\n"}],"comparison_anchors":[{"label":"Student loan default (US borrowers)","lifetime_us_adult":0.26},{"label":"Retirement savings shortfall (US)","lifetime_us_adult":0.39},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.6}],"regional_breakdown":[{"region":"ACA marketplace, in-network claim","probability":0.19,"notes":"19% average per-claim in-network denial rate (KFF, 2024); ranges 3–36% by insurer. Per filed claim, not a lifetime probability."},{"region":"ACA marketplace, out-of-network claim","probability":0.37,"notes":"37% per-claim out-of-network denial rate (KFF, 2024). Per filed claim, not a lifetime probability."},{"region":"SSDI applicants (initial application)","probability":0.62,"notes":"62% initial denial rate for Social Security Disability Insurance (complement of the SSA's 38.7% FY2024 approval rate, Urban Institute). Per initial application."}],"personal_factor_multipliers":[{"factor":"chronic condition requiring prior authorization","multiplier":2,"notes":"Prior authorization requirements increase the probability of denial for each encounter with the system"},{"factor":"employer-sponsored PPO plan","multiplier":0.5,"notes":"PPO plans with broad networks and fewer prior-auth requirements have lower denial rates"},{"factor":"out-of-network care","multiplier":2.5,"notes":"Out-of-network denial rates (37%) are roughly double in-network rates (19%)"},{"factor":"appeals all denials","multiplier":0.4,"notes":"Appealing denials reduces the effective denial rate substantially, but only <1% of consumers actually appeal"},{"factor":"post-acute care under Medicare Advantage","multiplier":3,"notes":"U.S. Senate Permanent Subcommittee on Investigations, \"Refusal of Recovery\" (Oct 2024): in 2022 UnitedHealthcare and CVS denied prior authorization for post-acute care (skilled-nursing and inpatient-rehab stays after a hospitalization) at approximately three times their overall denial rates; Humana's post-acute rate was over 16x its overall rate. Multiplier uses the ~3x figure for the two largest insurers.\n"}],"short_label":"Insurance claim denial","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"We publish the source-stated per-claim denial rate (19% in-network, 37% out-of-network in the 2024 KFF ACA marketplace data) rather than a lifetime \"chance of ever being denied\" figure, because no source supports the latter. A prior version of this entry carried a ~70% lifetime estimate built from an assumed ~1% per-encounter \"major denial\" rate over ~118 lifetime claim encounters; both inputs were author assumptions with no published basis, so that figure has been withdrawn (no_reliable_estimate). Converting a per-claim rate into a lifetime probability would require an encounter count and a per-encounter severe-denial rate that no cited source provides. The 19% headline denial rate itself includes administrative denials (coding errors, duplicate submissions, missing information) alongside substantive denials (medical necessity, prior authorization). While some administrative denials can be corrected on resubmission, the &lt;1% appeal rate (about 263,000 out of 85 million denied claims) means most denials of all types go unchallenged. Denial rates vary enormously by insurer (3-36% in the KFF data), by plan type (marketplace vs. employer vs. Medicaid), and by clinical context. The SSDI denial rate is a fundamentally different system from commercial health insurance and is included here to illustrate the breadth of insurance-denial risk, not because the two systems are comparable. Among those who do appeal, overturn rates are substantially higher, suggesting that many denials would be reversed if challenged. A distinct and growing mechanism is algorithmic auto-denial: claims rejected at scale without individual clinician review. ProPublica (March 2023) reported that Cigna doctors denied over 300,000 payment requests in two months via its \"PXDX\" system, spending an average of 1.2 seconds per case; a class action alleges UnitedHealthcare's \"nH Predict\" post-acute-care model had a 90% error rate yet remained in use because only ~0.2% of denials are appealed. These error figures are litigation allegations, not adjudicated facts, and the AI-specific share of all denials has not been cleanly quantified — but the Senate Permanent Subcommittee on Investigations (Oct 2024) documented that the same predictive systems drove the elevated post-acute denial rates reflected in the multiplier above. Separately, this entry covers a filed claim being denied or underpaid; it does not capture the adjacent and rising risk of homeowners-policy non-renewal, climate-driven carrier exit, FAIR-plan forced placement, or outright uninsurability — events that occur before any claim. In exposed areas these have surged in 2024-2026 (Florida non-renewal rates roughly tripled between 2018 and 2023 per a 2024 U.S. Senate Budget Committee analysis), but a non-renewal is a different event class from a denied claim and is outside the per-claim denial rates reported here.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-03","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-24","image":{"alt":"A stamped 'DENIED' mark on a medical form, muted grey and red tones, flat vector illustration."},"canonical_url":"https://likelier.app/insurance-claim-denial","api_url":"https://likelier.app/api/fears/insurance-claim-denial.json"},{"slug":"kratom-7oh-dependence","question":"How risky are the concentrated 7-OH kratom products sold at gas stations?","quick_answer":"There is no reliable number: no survey measures how many 7-OH users develop dependence or overdose, and poison-center reports are not incidence rates. What is clear is that these concentrated products are potent opioids sold like supplements, and the retail setting leads most buyers to underestimate an opioid-strength dependence and overdose risk.\n","category":"health","tags":["substance-use","household"],"no_reliable_estimate":true,"perceived":{"description":"Concentrated 7-OH shots, tablets, and gummies are sold at the same gas-station and smoke-shop counters as energy drinks and CBD, which frames them as a legal supplement rather than an opioid. No survey measures how consumers rate the dependence or overdose risk of these specific products; the plausible default is that the retail context leads most buyers to underestimate it.","rough_estimate":"Commonly treated as a mild legal supplement; opioid-strength dependence and overdose risk are likely underestimated.","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/mmwr/volumes/75/wr/mm7511a1.htm","title":"Increases in Kratom-Related Reports to Poison Centers — National Poison Data System, United States, 2015–2025","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"14,449 kratom exposures reported to U.S. poison centers 2015–2025; 3,434 in 2025 vs 258 in 2015 (~1,200% increase); 233 kratom-associated deaths over the period.","excerpt":"During 2015–2025, U.S. poison centers documented 14,449 kratom exposures; the 3,434 exposures reported in 2025 represent an increase of approximately 1,200% compared with the 258 reported in 2015","source_date":"2026-03-26","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260712173836/https://www.cdc.gov/mmwr/volumes/75/wr/mm7511a1.htm","calculation_notes":"These are exposure REPORTS to poison centers, not a population incidence rate. There is no denominator of 7-OH users, so no per-person probability can be derived. Used to characterize the upward trend and severity, not to compute a normalized lifetime probability. Hospitalizations: single-substance 43 (2015) to 538 (2025), +1,200%; multiple-substance 40 to 549, +1,300%. Deaths: 233 total, 184 (79%) multi-substance.","independence_note":"CDC analysis of National Poison Data System; independent of FDA regulatory actions and of the pharmacology review."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12671409/","title":"From kratom to 7-hydroxymitragynine: evolution of a natural remedy into a public-health threat","publisher":"Pharmaceutical Biology","source_type":"peer_reviewed","statistic":"7-OH is a potent μ-opioid receptor agonist with ten- to twentyfold higher MOR binding affinity than mitragynine and multifold greater potency than morphine in antinociception; concentrated 7-OH products are pharmacologically distinct from kratom leaf and produce opioid tolerance, dependence, and withdrawal.","excerpt":"7-OH is consistently characterized as a potent μ-opioid receptor (MOR) agonist with substantially greater efficacy and affinity than mitragynine.","source_date":"2025-11-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260702213609/https://pmc.ncbi.nlm.nih.gov/articles/PMC12671409/","calculation_notes":"Pharmacology/clinical review. Supports the opioid classification, the dependence/withdrawal mechanism, and the gas-station retail availability ('Clinical case documentation describes ready access to 7-OH tablets at gas stations and smoke shops.'). Documents a case of cardiopulmonary arrest after ~190 mg 7-OH. No population denominator provided; corroborates qualitative hazard, not a rate.","independence_note":"Academic pharmacology review; independent of CDC poison-center data and FDA actions."},{"url":"https://www.healio.com/news/primary-care/20250729/fda-asks-dea-to-classify-7oh-as-a-controlled-substance","title":"FDA asks DEA to classify 7-OH as a controlled substance","publisher":"Healio","source_type":"news_article","statistic":"July 2025: FDA recommended DEA schedule concentrated 7-OH as a controlled substance; the prior month FDA sent warning letters to seven companies; FDA characterizes 7-OH as an opioid linked to addiction, seizures, and fatal respiratory depression.","excerpt":"7-OH binds to the mu receptor, which means scientifically, by definition, it is an opioid","source_date":"2025-07-29","source_accessed":"2026-06-13","calculation_notes":"Used only to license the FDA-action specifics in prose (seven warning letters; Schedule I-type scheduling recommendation; adverse events: addiction, withdrawal, insomnia, anxiety, seizures, fatal respiratory depression; retail distribution at gas stations and convenience stores; '13 times more potent than morphine'). Direct FDA primary URLs (fda.gov/media/187898 and /187899) returned HTTP 404 at fetch time and were excluded. No quantitative denominator here either."}],"comparison_anchors":[],"short_label":"7-OH kratom dependence","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is a deliberate no_reliable_estimate, not a scoring failure. The hazard is real and authoritatively documented, but no defensible population denominator exists for \"develops 7-OH dependence\" or \"overdoses on 7-OH.\" Poison-center figures (14,449 exposures, 233 deaths over 2015–2025) are voluntary reports, not incidence rates, and they pool traditional kratom leaf with the newer concentrated 7-OH products; the MMWR attributes the 2025 surge partly to high-potency 7-OH but cannot isolate it cleanly. 7-OH occurs only in trace amounts in natural leaf, so the dependence/overdose profile here applies specifically to the concentrated semisynthetic shots, tablets, and gummies, an exposure that barely existed before ~2023 and has no longitudinal cohort behind it. Potency comparators must not be conflated: the confirmed peer-reviewed figures are \"ten- to twentyfold higher MOR binding affinity than mitragynine\" and \"multifold greater potency than morphine\"; the widely repeated \"13 times more potent than morphine\" comes via the FDA/news source. Direct FDA primary documents (fda.gov/media/187898 and /187899) returned 404 at fetch time and were excluded; FDA-action claims rest on the CONFIRMED Healio report. 8D self-score averages ~3.6 with D3/D4/D5 scored against the honest no-number state rather than a fabricated rate.","quality_score":{"d1":4,"d2":5,"d3":3,"d4":3,"d5":3,"d6":4,"d7":3,"d8":4,"avg":3.6,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Concentrated 7-OH (kratom) dependence or overdose\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/kratom-7oh-dependence","api_url":"https://likelier.app/api/fears/kratom-7oh-dependence.json"},{"slug":"medetomidine-opioid-supply","question":"What are the odds that street opioids are cut with medetomidine, a sedative naloxone can't reverse?","quick_answer":"There is no reliable probability here. The available data are counts of medetomidine detections in drug seizures — rising steeply year over year — not a rate, and there is no denominator for how many people who use street opioids are actually exposed, so a per-person likelihood cannot be estimated.\n","category":"health","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"There is no survey measuring how often people who use street opioids expect their supply to contain a veterinary sedative that naloxone will not reverse. The naming convention itself — 'rhino tranq', 'dex', 'mede' — postdates 2021, and the substance was unknown to the general public and to most prescribers before the 2025 CDC advisories. The intuitive model most people carry is the older one: an opioid overdose is an opioid overdose, and naloxone reverses it. Medetomidine breaks that model on two counts — bradycardia and sedation it does not touch, and a withdrawal syndrome that escalates to hospital-level autonomic crisis — and almost no one outside harm-reduction and emergency-medicine circles has updated for it.","rough_estimate":"most people, including many who use the supply, would not expect a non-opioid sedative in a majority of street-opioid samples","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/han/php/notices/han00527.html","title":"Medetomidine in the U.S. Illegal Fentanyl Supply Increasing Risk for Overdose and Severe Withdrawal Syndrome (HAN No. 527)","publisher":"CDC Health Alert Network / ONDCP","source_type":"govt_report","statistic":"Medetomidine detections in drug seizures rose from 247 (2023) to 2,616 (2024) to 8,233 (2025); 98% co-detected fentanyl; in at least 18 states + DC; naloxone does not reverse medetomidine","excerpt":"Medetomidine detections in law enforcement drug seizures increased by 950% from 247 in 2023 to 2,616 in 2024, followed by a further 215% increase to 8,233 in 2025... 98% had fentanyl co-detected... OORMs like naloxone are effective in reversing opioid effects but are not effective in reversing the effects of medetomidine... detected in drug samples and biological specimens from people who use illegal opioids in at least 18 states and the District of Columbia... most concentrated in the Northeast (52%) and Midwest (31%), followed by the South (17%) and the West (<1%).","source_date":"2026-04-02","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260626171024/https://www.cdc.gov/han/php/notices/han00527.html","calculation_notes":"These are detection counts in law-enforcement seizures, not a population rate. They establish that medetomidine penetration of the illicit-opioid supply is large and rising sharply but give no denominator of 'people exposed', so they support no_reliable_estimate rather than a normalized lifetime figure. The 98% fentanyl co-detection figure establishes that medetomidine functions as an adulterant of the opioid supply, not a standalone drug. Regional percentages anchor the regional_breakdown field.","independence_note":"The CDC HAN aggregates federal seizure data (DEA/NFLIS-style reporting) and the same MMWR clinical reports cited separately below; treat the seizure counts as the independent contribution and the clinical framing as overlapping with the MMWR sources."},{"url":"https://www.cdc.gov/overdose-prevention/php/data-research/mmwr-articles/index.html","title":"Overdoses Involving Medetomidine Mixed with Opioids — Chicago, Illinois, May 2024 (MMWR 74:258-265)","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"peer_reviewed","statistic":"12 confirmed, 26 probable, 140 suspected medetomidine-involved overdoses among patients at three Chicago hospitals during May 11-17, 2024; bradycardia and lack of response to naloxone were defining features","excerpt":"CDPH and CDC investigated and identified 12 confirmed, 26 probable, and 140 suspected overdoses involving medetomidine mixed with opioids among patients treated at three hospitals during May 11–17, 2024.","source_date":"2025-05-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260708075102/https://www.cdc.gov/overdose-prevention/php/data-research/mmwr-articles/index.html","calculation_notes":"A single one-week cluster at three hospitals. Cited to document that medetomidine-adulterated opioids produce atypical overdoses (bradycardia, naloxone non-response) at clinically observable scale, not to derive a rate. Full citation: Nham A, Le JN, Thomas SA, et al. MMWR Morb Mortal Wkly Rep 2025;74:258-265. DOI 10.15585/mmwr.mm7415a1. The DOI/citation block was confirmed on the CDC MMWR index page; the underlying PDF case counts are corroborated by the CDC HAN and contemporaneous reporting."},{"url":"https://pubmed.ncbi.nlm.nih.gov/40578053/","title":"Characterizing rapid changes in the prevalence and concentration of key compounds in Philadelphia's street opioid retail supply, March 2024-March 2025","publisher":"Drug and Alcohol Dependence (peer-reviewed)","source_type":"peer_reviewed","statistic":"Medetomidine first detected April 2024 in Philadelphia harm-reduction drug-checking samples; reached 83% prevalence among street-opioid samples by March 2025","excerpt":"Two novel compounds, medetomidine and bis(2,2,6,6-tetramethyl-4-piperidyl) sebacate (BTMPS), were first detected in April and June 2024... reaching prevalence rates of 83 % and 25 % by March 2025, respectively.","source_date":"2025-09-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20250820165453/https://pubmed.ncbi.nlm.nih.gov/40578053/","calculation_notes":"This is the load-bearing supply-penetration numerator: 83% of street-opioid samples in one Philadelphia harm-reduction program contained medetomidine by March 2025, up from first detection in April 2024. It is a sample-of-supply figure, not a probability that a given person is exposed — the denominator is 'samples submitted to one drug-checking program in one city', which does not generalize nationally. Supports no_reliable_estimate while quantifying how saturated a single high-prevalence local market became."},{"url":"https://www.statnews.com/2025/05/01/medetomidine-replacing-xylazine-in-fentanyl-increases-overdose-danger-withdrawal-risks/","title":"Sedative 'dex' is replacing 'tranq' in illegal drug supply and causing excruciating withdrawal","publisher":"STAT News","source_type":"news_article","statistic":"Medetomidine has displaced xylazine in parts of the supply; does not respond to naloxone; withdrawal can be life-threatening and requires hospital-level care","excerpt":"Like xylazine, however, medetomidine does not respond to naloxone, the common medication used to reverse opioid overdoses.","source_date":"2025-05-01","source_accessed":"2026-06-13","calculation_notes":"Non-authoritative secondary; used only for narrative color on the xylazine-to-medetomidine displacement and the shift of opioid withdrawal from home-manageable to hospital-level. All load-bearing facts are also in the authoritative CDC sources above."},{"url":"https://www.cdc.gov/mmwr/volumes/74/wr/mm7415a2.htm","title":"Notes from the Field: Suspected Medetomidine Withdrawal Syndrome Among Fentanyl-Exposed Patients — Philadelphia, Pennsylvania, September 2024-January 2025 (MMWR 74:266-268)","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"peer_reviewed","statistic":"165 patients at three Philadelphia health systems hospitalized September 2024-January 2025 for fentanyl withdrawal complicated by profound autonomic dysfunction; 150 (91%) required ICU care including 39 (24%) intubated; 137 (83%) treated with and responded to dexmedetomidine infusion","excerpt":"During September 2024-January 2025, 165 patients at three Philadelphia health systems were hospitalized for fentanyl withdrawal complicated by profound autonomic dysfunction, including severe hypertension and tachycardia. Among the 165 patients, 150 (91%) required intensive care unit (ICU) care, including 39 (24%) who received endotracheal intubation. A total of 137 (83%) patients were treated with and responded to dexmedetomidine infusion.","source_date":"2025-05-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260511115201/https://www.cdc.gov/mmwr/volumes/74/wr/mm7415a2.htm","calculation_notes":"Grounds the body claim that clinicians in Philadelphia documented a severe withdrawal syndrome requiring dexmedetomidine infusion and ICU care: a discrete case series of 165 patients, 91% requiring ICU admission and 83% treated with dexmedetomidine. Full citation: MMWR Morb Mortal Wkly Rep 2025;74:266-268. DOI 10.15585/mmwr.mm7415a2. This is a case series from one metro area, not a population rate, consistent with the entry's no_reliable_estimate framing.","independence_note":"A separate CDC MMWR \"Notes from the Field\" report from the same May 2025 issue as the Chicago overdose cluster cited above; a different patient cohort (withdrawal, not overdose) and a different city than both the Chicago and Pittsburgh reports."},{"url":"https://www.cdc.gov/mmwr/volumes/74/wr/mm7415a3.htm","title":"Notes from the Field: Severe Medetomidine Withdrawal Syndrome in Patients Using Illegally Manufactured Opioids — Pittsburgh, Pennsylvania, October 2024-March 2025 (MMWR 74:269-271)","publisher":"CDC Morbidity and Mortality Weekly Report","source_type":"peer_reviewed","statistic":"23 adult patients in a Pittsburgh health care system sought treatment October 2024-March 2025 for illegally manufactured opioid use with severe autonomic hyperactivity; most required dexmedetomidine infusion and ICU-level management","excerpt":"During October 2024-March 2025, a total of 23 adult patients who used illegally manufactured opioids sought treatment within a health care system in Pittsburgh, Pennsylvania. All exhibited severe autonomic hyperactivity, including hypertension, tachycardia, and agitation... most required dexmedetomidine infusion and intensive care unit-level management.","source_date":"2025-05-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260511084043/https://www.cdc.gov/mmwr/volumes/74/wr/mm7415a3.htm","calculation_notes":"A second, geographically distinct case series corroborating the Philadelphia withdrawal report cited above: severe autonomic hyperactivity (hypertension, tachycardia) in patients exposed to medetomidine-adulterated opioids, requiring dexmedetomidine infusion and ICU care in the large majority of confirmed cases. Full citation: MMWR Morb Mortal Wkly Rep 2025;74:269-271. DOI 10.15585/mmwr.mm7415a3. A case series, not a population rate.","independence_note":"Independent patient cohort from a different city and health system than the Philadelphia report and the Chicago overdose cluster, though all three are CDC MMWR reports from the same April/May 2025 reporting cycle."}],"comparison_anchors":[{"label":"Dying in a car crash (lifetime, US)","lifetime_us_adult":0.0095},{"label":"Hospital-acquired infection","lifetime_us_adult":0.039},{"label":"Stroke (lifetime, US)","lifetime_us_adult":0.067}],"regional_breakdown":[{"region":"Northeast US","probability":0.52,"notes":"Share of medetomidine reports per CDC HAN (2026); Philadelphia harm-reduction drug-checking found medetomidine in 83% of street-opioid samples by March 2025. Probability here is the CDC regional share of reports, not an individual exposure odds."},{"region":"Midwest US","probability":0.31,"notes":"Second-most-affected region; Chicago saw a discrete cluster of 12 confirmed / 26 probable / 140 suspected medetomidine-involved overdoses in one week of May 2024."},{"region":"South US","probability":0.17,"notes":"Lower but non-trivial share of CDC medetomidine reports as of the April 2026 advisory."},{"region":"West US","probability":0.01,"notes":"Under 1% of reports as of April 2026; the West Coast supply remained dominated by other adulterants, though emerging-threat geography shifts within months."}],"short_label":"Medetomidine in street opioids","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"No defensible population denominator exists for this risk, so it carries no_reliable_estimate: true. The available numbers are supply-penetration figures (share of seized or drug-checked samples that contain medetomidine) and clinical case counts (hospitalized withdrawal patients, overdose clusters) — neither of which converts to a 'lifetime probability for a US adult'. The exposed population is people who use illicit opioids, not the general public, and even within that group the prevalence is steeply geographic: 83% of street-opioid samples in one Philadelphia program by March 2025 versus under 1% of national reports from the West as of April 2026. The data are also fast-moving — medetomidine displaced xylazine in roughly a year in the hardest-hit markets, and any percentage here may be stale within months. The percentages in the regional_breakdown are CDC shares of reports, not individual exposure odds; they are included only to show that geography is load-bearing. The headline hazard is twofold and not capturable in a single odds figure: (1) bradycardia and sedation that naloxone does not reverse, raising overdose lethality versus a pure-opioid exposure, and (2) a withdrawal syndrome severe enough to require ICU-level dexmedetomidine infusion. 8D self-score: the entry clears the schema's no_reliable_estimate path (>=2 sources, 3 authoritative, verbatim excerpts) but cannot score the arithmetic/scope dimensions that depend on a normalized number, so those are scored against the honest-absence standard rather than a fabricated figure.","quality_score":{"d1":5,"d2":5,"d3":3,"d4":3,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Medetomidine in street opioids\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/medetomidine-opioid-supply","api_url":"https://likelier.app/api/fears/medetomidine-opioid-supply.json"},{"slug":"microplastics-health-harm","question":"What are the odds that microplastics in food, water, or air will harm your health?","quick_answer":"There is no reliable estimate: WHO finds no measurable health harm from microplastics at current exposure levels, though detection in blood and tissue is real and research is ongoing. Concern runs ahead of the evidence — 47% of US adults rank cancer-causing chemicals in food among their top-3 worries — but detection is not the same as demonstrated harm.\n","category":"health","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"Microplastics have become one of the defining environmental anxieties of the 2020s. Headlines announcing their detection in human blood, placentas, lungs, and brains generate enormous engagement, and a widely cited 2019 WWF-commissioned estimate claiming humans ingest \"a credit card's worth of plastic per week\" entered popular culture as settled fact. Surveys show a majority of adults in the US and EU rank microplastics among their top food-safety concerns, with EFSA's 2025 Eurobarometer finding 63% of EU citizens aware of the issue — up eight points from 2022. The implicit assumption is that detection equals danger: if plastic particles are inside us, they must be causing harm.\n","rough_estimate":"47% of US adults rank cancer-causing chemicals in food among their top-3 concerns; microplastics are an emerging subset","kind":"survey","survey_source":{"title":"IFIC 2025 Food & Health Survey — 47% rank cancer-causing chemicals in food as a top-3 concern; microplastics fall under this chemical-contaminant umbrella","publisher":"International Food Information Council (IFIC)","url":"https://ific.org/media/confidence-in-food-safety-hits-record-low/","year":2025}},"sources":[{"url":"https://www.who.int/publications/i/item/9789241516198","title":"Microplastics in Drinking-Water","publisher":"World Health Organization","source_type":"govt_report","statistic":"No reliable evidence that microplastics in drinking water at current exposure levels pose a risk to human health","excerpt":"\"Based on the limited information we have, microplastics in drinking water don't appear to pose a health risk at current levels. But we need to find out more.\"\n","source_date":"2019-08-22","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260426203936/https://www.who.int/publications/i/item/9789241516198","calculation_notes":"WHO reviewed all available evidence on microplastic occurrence in drinking water (tap and bottled) and assessed potential health impacts from ingestion. The report concluded that particles larger than 150 micrometres are unlikely to be absorbed, uptake of smaller particles is expected to be limited, and no epidemiological data link drinking-water microplastic exposure to adverse health outcomes. WHO called for more research but did not identify a quantifiable risk, supporting the no_reliable_estimate classification.\n","independence_note":"WHO review synthesised global evidence independently of the Marfella, Leslie, and EFSA analyses cited below.\n"},{"url":"https://www.nejm.org/doi/10.1056/NEJMoa2309822","title":"Microplastics and Nanoplastics in Atheromas and Cardiovascular Events","publisher":"New England Journal of Medicine / Marfella et al.","source_type":"primary_study","statistic":"Patients with detectable MNPs in carotid plaque had HR 4.53 (95% CI 2.00-10.27) for composite cardiovascular endpoint vs those without","excerpt":"\"Patients in whom microplastics and nanoplastics were detected within the atheroma had a higher risk of a composite of myocardial infarction, stroke, or death from any cause than those in whom microplastics and nanoplastics were not detected.\"\n","source_date":"2024-03-07","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250825052140/https://www.nejm.org/doi/10.1056/NEJMoa2309822","calculation_notes":"Marfella et al. prospectively followed 257 patients who underwent carotid endarterectomy and analyzed excised plaque for MNPs via pyrolysis-GC/MS. Polyethylene was detected in 58.4% of plaques (150 of 257); PVC in 12.1% (31 patients). After 34 months, the MNP-positive group had significantly higher cardiovascular events. However, this is an observational study in a high-risk surgical cohort (all had asymptomatic carotid disease), and critics note no pre-analytical anticontamination procedures were used — the surgical environment itself contains PE and PVC. The study demonstrates association in a diseased population, not causation in the general public, and cannot be converted to a population-level attributable risk.\n","independence_note":"Prospective clinical cohort study by Italian researchers using pyrolysis-GC/MS on excised plaque tissue; independent of WHO review and Leslie blood-detection study.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S0160412022001258","title":"Discovery and quantification of plastic particle pollution in human blood","publisher":"Environment International / Leslie et al.","source_type":"primary_study","statistic":"Microplastics detected in 17 of 22 (77%) healthy adult blood samples at mean concentration of 1.6 µg/mL","excerpt":"\"This is the first study to report quantitative data on plastic particle pollution in human blood, showing that plastic particles are bioavailable for uptake into the human bloodstream.\"\n","source_date":"2022-03-24","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250828152504/https://www.sciencedirect.com/science/article/pii/S0160412022001258","calculation_notes":"Leslie et al. analyzed blood from 22 healthy Dutch volunteers using pyrolysis-GC/MS for five high-production polymers. PET was the most prevalent (50% of samples), followed by polystyrene (36%) and PE (23%). The study proves systemic exposure (detection) but does not measure health outcomes. No dose-response relationship, no disease endpoint, no follow-up. Detection of a substance in blood is a necessary but not sufficient condition for harm — the dose, persistence, and biological activity all remain unquantified for general-population microplastic exposure. This study is included because it is the most-cited driver of public fear, illustrating the detection-equals-danger heuristic.\n","independence_note":"Dutch laboratory study using independent pyrolysis-GC/MS methodology on volunteer blood samples; no overlap with the Marfella clinical cohort or WHO/EFSA reviews.\n"},{"url":"https://www.nature.com/articles/s41591-024-03453-1","title":"Bioaccumulation of microplastics in decedent human brains","publisher":"Nature Medicine / Nihart et al.","source_type":"peer_reviewed","statistic":"MNP concentration 7-30x higher in brain (median ~4900 µg/g) than liver/kidney; rose ~50% from 2016 to 2024; up to 10x higher in brains with documented dementia","excerpt":"\"Concentrations in decedent brain samples were 7 to 30 times higher than in the liver or kidney... brain tissue from people who had been diagnosed with dementia had up to 10 times as much plastic in their brains as everyone else... the study design cannot show whether higher levels of plastic in the brain caused the dementia symptoms — they may simply accumulate more due to the disease process itself.\"\n","source_date":"2025-02-03","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260714015022/https://www.nature.com/articles/s41591-024-03453-1","calculation_notes":"Nihart et al. applied pyrolysis-GC/MS (corroborated by FTIR and electron microscopy) to >50 autopsy samples from the Albuquerque medical examiner, from decedents in 2016 and 2024. Brain MNP load (predominantly polyethylene, largely nanoscale shard-like fragments) far exceeded liver and kidney, and rose ~50% across the eight-year window. Brains with a documented dementia diagnosis showed up to 10x more plastic, concentrated in cerebrovascular walls and immune cells. This is a cross-sectional autopsy series: the dementia association is explicitly correlation, not causation, and the authors flag reverse causation (impaired clearance / blood-brain-barrier breakdown in dementia could increase deposition). It quantifies tissue burden and its temporal rise, not attributable disease risk, so it does not change the no_reliable_estimate classification.\n","independence_note":"US autopsy-tissue study using pyrolysis-GC/MS plus FTIR and electron microscopy; methodologically independent of the WHO review, the Marfella plaque cohort, the Leslie blood study, and the EFSA assessment.\n"},{"url":"https://www.efsa.europa.eu/en/topics/microplastics-and-nanoplastics-food","title":"Microplastics and nanoplastics in food","publisher":"European Food Safety Authority","source_type":"govt_report","statistic":"EFSA risk assessment on microplastics in food scheduled for completion by end of 2027; no quantified health risk established as of 2025","excerpt":"\"The European Parliament has requested EFSA to deliver a scientific opinion on the potential health risks posed by microplastics in food, water and air.\"\n","source_date":"2025-10-28","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420044010/https://www.efsa.europa.eu/en/topics/microplastics-and-nanoplastics-food","calculation_notes":"EFSA's 2025 literature review of micro- and nanoplastic release from food contact materials analyzed over 100 studies (2015-2025) and concluded that while microplastic release from FCMs is real, current studies likely overestimate quantities, and nanoplastic data remain insufficient for reliable exposure estimates. EFSA has not issued a quantified risk figure. Their formal scientific opinion is not expected until end of 2027. The absence of an EFSA opinion after years of review is itself evidence of the epistemic gap — not of safety, but of insufficient data to quantify risk in either direction.\n","independence_note":"EFSA risk assessment is an independent EU regulatory review, methodologically separate from the WHO 2019 review and from individual primary studies.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S2666911022000247","title":"Ingested Microplastics: Do Humans Eat One Credit Card per Week?","publisher":"Journal of Hazardous Materials Letters / Pletz (2022)","source_type":"peer_reviewed","statistic":"The widely cited 5g/week estimate overestimates ingestion by several orders of magnitude; realistic estimates are 0.01-0.1g/week","excerpt":"\"Our analysis suggests that the widely reported figure of 5 g per week overestimates microplastic ingestion by several orders of magnitude.\"\n","source_date":"2022-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250730181010/https://www.sciencedirect.com/science/article/pii/S2666911022000247","calculation_notes":"Pletz (2022) critically reviewed the Senathirajah et al. (2021) analysis that generated the \"credit card per week\" claim. The 5g figure was the extreme upper bound of a 0.1-5g range, and the underlying methodology conflated particle counts with mass in ways that inflate estimates. More careful analyses suggest typical adult ingestion is roughly 0.01-0.1g per week — still nonzero, but 50-500x lower than the headline figure. This matters because the perceived threat level is anchored to an exposure estimate that is almost certainly wrong by orders of magnitude.\n","independence_note":"Critical reanalysis of the WWF/Newcastle exposure estimate by independent food-science researchers; not affiliated with WHO, EFSA, or the original 2019 study authors.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/33395930/","title":"Plasticenta: First evidence of microplastics in human placenta","publisher":"Environment International / Ragusa et al.","source_type":"peer_reviewed","statistic":"12 microplastic fragments (5-10 micrometers) detected via Raman microspectroscopy in 4 of 6 human placentas from uneventful pregnancies","excerpt":"\"In total, 12 microplastic fragments (ranging from 5 to 10 [micrometers] in size)...were found in 4 placentas.\"\n","source_date":"2021-01-01","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260706182926/https://pubmed.ncbi.nlm.nih.gov/33395930/","calculation_notes":"Substantiates the body prose's \"Ragusa et al. found them in placentas\" claim. First reported detection of microplastics in human placental tissue (fetal side, maternal side, and chorioamniotic membranes) in a small sample of six placentas from uneventful pregnancies. Like the Leslie blood study, this proves detection/bioavailability, not attributable health harm -- no outcome data, no dose-response, no disease endpoint measured. Does not change the no_reliable_estimate classification.\n","independence_note":"Italian clinical/laboratory study using Raman microspectroscopy on placental tissue; methodologically independent of the WHO review, the Marfella plaque cohort, the Leslie blood study, the Nihart brain study, and the EFSA assessment.\n"}],"comparison_anchors":[{"label":"Harm from pesticide residue on US produce (lifetime)","lifetime_us_adult":0.000001},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Typical US adult (dietary + airborne)","probability":0,"notes":"No epidemiological study has measured attributable disease from dietary or airborne microplastic exposure in the general US population. Exposure is real; harm is unquantified.\n"},{"region":"Occupational (textile/plastics workers)","probability":0,"notes":"Workers in synthetic textile and plastics manufacturing have higher inhalation exposure. Some occupational studies report elevated respiratory symptoms (flock worker's lung), but these involve specific fiber types at industrial concentrations, not general microplastic exposure. Signal is weak and confounded by co-exposures.\n"},{"region":"Marine food-chain accumulation (heavy seafood diet)","probability":0,"notes":"Shellfish and small pelagic fish bioaccumulate microplastics. Consumers of large quantities of mussels, oysters, and sardines have higher ingestion estimates, but no cohort study has linked this dietary pattern to measurable health outcomes in humans.\n"},{"region":"Bottled vs tap water","probability":0,"notes":"Bottled water contains roughly 10-100x more microplastic particles per liter than tap water in most studies, likely from packaging. WHO's 2019 review found no health risk from either source at observed levels.\n"}],"personal_factor_multipliers":[{"factor":"Heavy bottled-water consumer","multiplier":2,"notes":"Bottled water contains roughly 10-100x more microplastic particles per litre than tap water in most studies. Switching to tap or filtered water reduces ingestion from this pathway.\n"},{"factor":"Occupational (textile/plastics manufacturing)","multiplier":5,"notes":"Workers inhaling synthetic fibres at industrial concentrations have a thin but real evidence base for respiratory effects, though confounded by co-exposures to other industrial particulates.\n"},{"factor":"Heavy shellfish consumer","multiplier":2,"notes":"Mussels, oysters, and small pelagic fish bioaccumulate microplastics. Higher ingestion estimates but no documented health outcomes in humans.\n"},{"factor":"Micro/nanoplastics detected in arterial plaque","multiplier":4.5,"notes":"Marfella et al. (NEJM 2024) reported a hazard ratio of 4.53 (95% CI 2.00-10.27) for a composite of myocardial infarction, stroke, or death from any cause among carotid-endarterectomy patients whose excised plaque contained detectable micro/nanoplastics versus those without, over ~34 months. This is an association in a pre-selected diseased surgical cohort with no operating-room anticontamination controls, not a general-population probability multiplier — and detectability is not a behaviour one chooses, so treat this as an evidence flag, not a personal exposure lever.\n"}],"short_label":"Microplastics","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers health harm to humans from environmental microplastic exposure through food, water, and air at levels encountered by the general population. It does not cover occupational exposure to specific synthetic fibers at industrial concentrations (which has a separate, small evidence base), ecological harm to marine organisms, or the distinct question of chemical additives (plasticizers, flame retardants) that may leach from plastics — those are regulated under separate toxicological frameworks. The no_reliable_estimate designation reflects the current state of evidence: microplastics are detectable in human tissues, but no epidemiological study has measured attributable human disease from typical environmental exposure. The Marfella 2024 NEJM study found an association between MNPs in carotid plaque and cardiovascular events, but in a pre-selected diseased cohort with contamination concerns, and cannot be generalized to population-level causation. Nihart et al. (Nature Medicine 2025) added a brain-specific bioaccumulation signal — micro/nanoplastic concentrations 7-30x higher in the brain than in liver or kidney, rising roughly 50% from 2016 to 2024, and up to 10x higher in brains with a documented dementia diagnosis — but this is a cross-sectional autopsy series in which the dementia association is explicitly correlation, not causation (the authors flag possible reverse causation, i.e. impaired clearance in dementia increasing deposition). A preliminary, not-yet-peer-reviewed 2025 AHA abstract (Clark, Vascular Discovery) likewise found far higher plaque microplastic concentrations in stroke patients, but as tissue burden, not measured risk. None of these quantify attributable human disease, so the no_reliable_estimate designation stands. The field may change substantially as EFSA's 2027 opinion and ongoing cohort studies report results.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single water glass on a plain surface with faint translucent particles suspended inside, flat vector illustration."},"canonical_url":"https://likelier.app/microplastics-health-harm","api_url":"https://likelier.app/api/fears/microplastics-health-harm.json"},{"slug":"new-clothes-unwashed-contamination","question":"What are the odds of getting sick or having a skin reaction from wearing new clothing without washing it first?","quick_answer":"No reliable estimate exists: no study measures how often people actually get sick or break out from an unwashed new garment. The chemical pathway is real — formaldehyde was detectable in about 1 in 5 sampled new textiles, and a single wash removed it — but the population-level illness rate for skipping a prewash is simply unmeasured.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"Dermatologist interviews and hygiene-focused media have elevated \"always wash new clothes before wearing\" to near-consensus advice. The implied threat is a loose composite — dozens of strangers trying on the same shirt, fecal bacteria colonising the fabric, formaldehyde off-gassing from \"permanent press\" treatment, scabies mites surviving in the folds, and chemical dyes causing rashes. The risk blurs across chemical and biological categories in a way that makes individual claims hard to evaluate. No large population survey measures the rate at which people actually get sick from an unwashed new garment, so the fear remains quantitatively unanchored: most adults who skip prewashing feel mildly transgressive rather than able to name a specific probability.\n","rough_estimate":"Most people treat skipping prewash as somewhere between 'mildly risky' and 'probably fine' — rarely attached to a number","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC9318620/","title":"Early-Life Exposure to Formaldehyde Encountered in Clothing and Other Textile Items","publisher":"PubMed Central / NIH (Schulz et al., Environmental Health Perspectives)","source_type":"peer_reviewed","statistic":"Formaldehyde detected in 20% of 120 sampled first-layer textile items (worn by pregnant women, babies, and toddlers); none of the detectable items showed formaldehyde after a single wash","excerpt":"\"None of the 20 clothing items showed detectable amounts of formaldehyde after washing [...] suggesting a simple pre-use laundering practice substantially reduces exposure risks for vulnerable populations.\"\n","source_date":"2022-07-01","source_accessed":"2026-05-16","archive_url":"http://web.archive.org/web/20260507105539/https://pmc.ncbi.nlm.nih.gov/articles/PMC9318620/","calculation_notes":"This study directly quantifies how often new clothing carries detectable formaldehyde (1 in 5 items in their sample) and confirms that a single wash eliminates it. Concentrations ranged below the EU 75 mg/kg regulatory limit for clothing, which is relevant context: even formaldehyde-bearing garments are not at levels designed to cause acute harm, but the quantities are enough to provoke contact dermatitis in a sensitised individual. The study sampled first-layer textiles worn by pregnant women, babies, and toddlers, which may or may not generalise to adult garments with heavier wrinkle-resist finishing; the 20% prevalence figure is used here as an approximate floor for how often the chemical hazard is present in new retail clothing in general.\n","independence_note":"Published in Environmental Health Perspectives; independent of the Fowler 1992 clinical patch-test series and the Italian multicentre study in both authorship and methodology (analytical chemistry of textiles rather than clinical dermatology).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/1479102/","title":"Allergic contact dermatitis from formaldehyde resins in permanent press clothing: an underdiagnosed cause of generalized dermatitis","publisher":"Journal of the American Academy of Dermatology (Fowler JF Jr, Skinner SM, Belsito DV)","source_type":"peer_reviewed","statistic":"17 patients identified with allergic contact dermatitis from formaldehyde-based textile resins; authors conclude the condition is underdiagnosed","excerpt":"\"Formaldehyde resins have been used to impart wrinkle resistance to clothing fabrics since 1926 [...] formaldehyde textile resin allergy is more common than has been previously recognized [...] accentuation of dermatitis in areas of tight clothing.\"\n","source_date":"1992-10-01","source_accessed":"2026-05-16","archive_url":"http://web.archive.org/web/20250116112845/https://pubmed.ncbi.nlm.nih.gov/1479102/","calculation_notes":"The foundational clinical paper establishing that textile formaldehyde contact dermatitis is a real, diagnosable, underdiagnosed condition — not a theoretical hazard. The pattern (rash in tight-fitting clothing areas, worse in warmer weather when sweating increases formaldehyde release from fabric) is the clinical fingerprint. The 17-patient case series does not supply a population prevalence rate, but combined with later patch-test surveillance data placing formaldehyde sensitisation at roughly 2–5% of the general population, it grounds the plausibility of the chemical pathway. Included to establish that the mechanism operates in real patients, not just in laboratory conditions.\n","independence_note":"Independent of the Schulz 2022 analytical chemistry study; this is clinical dermatology patch-test work by three separate authors at different institutions.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24392992/","title":"Clinical and epidemiological features of textile contact dermatitis: an Italian multicentre study","publisher":"Contact Dermatitis (Lisi P, Stingeni L, Cristaudo A, et al.)","source_type":"peer_reviewed","statistic":"Across 277 Italian textile contact dermatitis patients, lesions were prevalently eczematous (74.2%); TCD was more common in females and in atopic patients; disperse dyes (Disperse Blue 124, Disperse Blue 106, Disperse Yellow 3) were most frequently responsible (79.8%), with formaldehyde and resins more important in occupational cases","excerpt":"\"The lesions were prevalently eczematous (74.2%) [...] TCD was more common in females, in the fourth to fifth decades of life, and in atopic dermatitis patients. The dyes (Disperse Blue 124, Disperse Blue 106, and Disperse Yellow 3) were most frequently responsible (79.8%) [...] Formaldehyde and resins were more important in occupational TCD.\"\n","source_date":"2014-01-01","source_accessed":"2026-05-16","archive_url":"http://web.archive.org/web/20260525095710/https://pubmed.ncbi.nlm.nih.gov/24392992/","calculation_notes":"Supplies population-level texture the Fowler case series cannot: a 277-patient multicentre series characterising who actually presents with textile contact dermatitis and which allergens drive it. The lesions were prevalently eczematous (74.2%); TCD was more common in women, in the fourth-to-fifth decades of life, and in atopic dermatitis patients. Notably, disperse dyes — not formaldehyde — were the single most frequent culprit (79.8%), with formaldehyde and resins more important in occupational cases. This paper does not supply a general-population prevalence rate; it characterises the phenotype and confirms that the female skew and atopic predisposition seen in the wider literature hold in a large Italian cohort. It is cited to show textile contact dermatitis is a well-described, non-negligible clinical entity, and to correct the common over-attribution to formaldehyde alone.\n","independence_note":"Multicentre Italian study by the Italian Contact Dermatitis Research Group; independent of the US Fowler series in population, setting, and era (2014 vs 1992).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC8213614/","title":"Second-Hand Clothe, a New Threat for Acquiring Parasitic Infection","publisher":"Iranian Journal of Public Health (Rakhshanpour A, Shirazy AA, Shafiei R, Rahimi MT)","source_type":"peer_reviewed","statistic":"2.7% of 800 unwashed second-hand garments tested positive for parasites or ectoparasites (Sarcoptes scabiei, Pediculus species, Enterobius eggs); 0% of washed garments","excerpt":"\"22 (2.7%) were positive for parasites and ectoparasite contamination [...] all contamination was found exclusively in unwashed items [...] Sarcoptes scabiei (0.75%), Pediculus species eggs (0.75%), Enterobius eggs (1.25%).\"\n","source_date":"2021-06-01","source_accessed":"2026-05-16","archive_url":"http://web.archive.org/web/20260508011257/https://pmc.ncbi.nlm.nih.gov/articles/PMC8213614/","calculation_notes":"The most direct evidence that biological contamination of clothing is real and not merely theoretical. The 2.7% figure applies to second-hand (thrift) garments in an Iranian study — it is the wrong population for US new-retail clothing. The expected rate for new retail garments is much lower and essentially unmeasured; what the paper establishes is that the biological mechanism (parasite survival on unwashed fabric) is documented, and that washing reliably eliminates it. The figure is used here to bound the second-hand/thrift scenario, not as the headline estimate for new retail. Scabies transmission via fomites is generally considered less common than direct skin-to-skin contact (CDC notes mites survive 2–3 days off-host), so the clinical import of fabric contamination is conditional on someone with active infection having recently worn the garment.\n","independence_note":"Iranian public-health sampling study, methodologically and institutionally independent of the formaldehyde-dermatitis literature.\n"},{"url":"https://www.cdc.gov/scabies/about/index.html","title":"About Scabies","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Scabies mites generally do not survive more than 2–3 days away from human skin; sharing clothing used by an infected person is a less common transmission route","excerpt":"\"Less commonly, sharing clothing, towels, or bedding used by an infected person [...] Scabies mites generally do not survive more than two to three days away from human skin.\"\n","source_date":"2025-01-01","source_accessed":"2026-05-16","archive_url":"http://web.archive.org/web/20260518135405/https://www.cdc.gov/scabies/about/index.html","calculation_notes":"CDC's scabies guidance supplies the key biological context limiting the fomite transmission risk: mites survive off-host for at most 2–3 days, making the risk specifically conditional on wearing clothing that an actively infected person wore within that window. For new retail clothing on a shelf or rack, the window is long enough to be theoretically relevant if the last person to try on a garment had active (unsuspected) scabies — but CDC characterises fomite transmission as \"less common\" even in household contacts, where exposure is repeated and prolonged. The transmission risk from a single try-on in a retail setting is substantially lower than in household settings.\n","independence_note":"CDC public-health guidance; independent of the Iranian parasitology paper and the dermatology literature in both institution and methodology.\n"}],"comparison_anchors":[{"label":"Symptomatic allergic contact dermatitis from one patch of poison ivy contact","lifetime_us_adult":0.6},{"label":"Any contact dermatitis episode in a lifetime (US adult, all causes)","lifetime_us_adult":0.15},{"label":"Scabies infection in a lifetime (US adult, all routes)","lifetime_us_adult":0.08}],"regional_breakdown":[{"region":"Healthy non-sensitised adult wearing one unwashed new retail garment","probability":0.005,"notes":"Rough upper bound for clinically noticeable skin reaction per garment worn unwashed. The 20% formaldehyde detection rate, a ~3% general sensitisation rate, and an uncertain reaction probability per contact yield a per-garment risk of roughly 0.5–1% for a reaction from the chemical pathway alone. Biological pathway (bacteria, parasites) adds negligible incremental risk for immunocompetent adults from new retail clothing. This figure is a structural estimate, not a measured rate.\n"},{"region":"Formaldehyde-sensitised adult (patch-test positive)","probability":0.4,"notes":"For a person with confirmed formaldehyde contact allergy wearing a wrinkle-resistant garment, the reaction is near-certain if the garment carries formaldehyde resin. Because approximately 20% of garments do, the per-garment probability approaches 0.2 × reaction probability (near 1 for sensitised skin). Repeated wearing without washing raises cumulative exposure. This subgroup is the clinical target of the \"wash before wearing\" recommendation.\n"},{"region":"Unwashed second-hand / thrift garment (biological pathway)","probability":0.027,"notes":"The Rakhshanpour 2021 study found 2.7% of unwashed second-hand garments in their sample positive for parasites or ectoparasites. Wearing a contaminated garment does not guarantee infection (parasite viability, number of organisms, skin-contact duration all matter), so the per-garment illness probability is lower than 2.7% — but this is the closest empirical anchor available for the biological risk.\n"},{"region":"Immunocompromised adult wearing any unwashed clothing","probability":0.01,"notes":"The bacterial loads that a healthy immune system clears as a matter of course (Staphylococcus, Corynebacterium, candida species from fabric microbiome studies) can cause clinical infection in severely immunocompromised individuals. The figure is a rough order-of-magnitude placeholder.\n"}],"personal_factor_multipliers":[{"factor":"Formaldehyde contact allergy (confirmed by patch test)","multiplier":40,"notes":"An individual already sensitised to formaldehyde or formaldehyde-releasing resins will almost certainly react upon wearing a treated garment — the challenge is essentially deterministic once exposure occurs. Patch-test positivity to formaldehyde sits at roughly 2–5% across general dermatology patch-test surveillance series; for that subgroup, skipping prewash on a wrinkle-resistant garment carries a qualitatively different risk profile than for a non-sensitised person. The Lisi et al. 2014 Italian series confirms that textile contact dermatitis skews female and atopic, though disperse dyes rather than formaldehyde were its single most frequent allergen (79.8%).\n"},{"factor":"Atopic dermatitis or active eczema","multiplier":3,"notes":"Disrupted epidermal barrier in atopic skin increases penetration of chemical sensitisers; atopic individuals are also more likely to become sensitised in the first place, per general contact-dermatitis epidemiology. The 3× estimate is an order-of- magnitude approximation rather than a directly measured figure.\n"},{"factor":"Infant or young child skin","multiplier":4,"notes":"The Schulz 2022 study (PMC9318620) focused on children's clothing specifically because infant skin is more permeable to formaldehyde and other textile chemicals. Neonatal and infant skin has thinner stratum corneum and higher surface-area-to-weight ratio; the same formaldehyde concentration delivers proportionally higher systemic and dermal exposure.\n"},{"factor":"Wrinkle-resistant, permanent-press, or crease-resistant labelling","multiplier":3,"notes":"Formaldehyde-releasing resins are specifically used to achieve wrinkle resistance. A garment labelled \"permanent press,\" \"wrinkle-resistant,\" \"easy care,\" or similar is more likely to carry detectable formaldehyde than an untreated natural-fibre garment. The Schulz 2022 study detected formaldehyde across 20% of items tested; wrinkle-resist finishing is the primary delivery mechanism.\n"},{"factor":"Second-hand / thrift store origin vs. new retail","multiplier":8,"notes":"The Iranian parasitology study (Rakhshanpour et al. 2021) found 2.7% contamination with parasites in unwashed second-hand garments. New retail clothing has no measured equivalent; the expected contamination rate from store try-ons is lower but unknown. The ≈8× estimate reflects the difference between confirmed real-world contamination in used clothing versus a plausible but unmeasured lower rate for new retail.\n"}],"short_label":"Unwashed new clothes","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The headline characterisation of \"overrated\" applies specifically to the healthy non-sensitised adult buying new retail clothing. Three carve-outs change the calculus substantially. First: individuals with confirmed formaldehyde contact allergy — roughly 2–5% of the population based on patch-test clinic data — face a qualitatively different risk profile; for them, wearing unwashed wrinkle-resistant clothing is a reliable trigger, not a hypothetical. The Fowler 1992 and Lisi 2014 papers document this population clearly. Second: infants and young children have more permeable skin and greater relative exposure per kilogram of body weight; the Schulz 2022 study focused on children's clothing specifically because the pediatric exposure window is more consequential. Third: second-hand and thrift-store clothing carries a documented (if low) rate of biological contamination — scabies mites, lice, and helminth eggs — that does not apply in the same way to new retail garments. The CDC notes fomite transmission of scabies is \"less common\" even in household settings, but it is real, and washing before first wear eliminates this route entirely. The biological risk from trying on new retail clothing is real in principle (multiple strangers, up to a few days for mite survival off-host) but has not been quantified in a study measuring illness outcomes rather than laboratory contamination. Nothing here addresses sensory or olfactory reasons to prewash, which are independent of health risk.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-16","image":{"alt":"A single neatly folded shirt with a paper price tag, rendered as a flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/new-clothes-unwashed-contamination","api_url":"https://likelier.app/api/fears/new-clothes-unwashed-contamination.json"},{"slug":"nitazene-counterfeit-pill","question":"What are the odds of dying from a counterfeit pain or anxiety pill that contains a nitazene opioid?","quick_answer":"No reliable estimate exists: there is no count of how many people take a counterfeit pill, so nitazene deaths — rising sharply, from 52 to 195 in a year in England and Wales — cannot be turned into a per-pill probability. The danger is that the nitazene is invisible: undisclosed, undetectable by taste, and missed by fentanyl test strips.\n","category":"health","tags":["substance-use","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"A person who buys a single Xanax, Valium, oxycodone or Percocet tablet from a contact rather than a pharmacy is usually weighing the familiar risks of the named drug: drowsiness, dependence, a rough comedown. The mental model is that the pill is roughly what the press says it is, possibly underdosed or cut, but recognisable. The specific scenario here breaks that model: the tablet contains no benzodiazepine or licensed opioid at all, but a microgram quantity of a nitazene, a synthetic opioid that can match or exceed fentanyl and that the consumer cannot taste, see, or rule out with a fentanyl test strip. Because nitazenes have no active retail user market and circulate almost entirely as undisclosed adulterants, the person taking the pill generally has no idea the compound is present. There is no published survey of how people rate this particular hazard, so the perception here is inferred rather than measured.","rough_estimate":"Most non-pharmacy pill buyers price in the named drug's risks but not an undetectable nitazene substitution; the contamination is invisible to the consumer and to fentanyl test strips.","kind":"intuition"},"sources":[{"url":"https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/bulletins/deathsrelatedtodrugpoisoninginenglandandwales/2024registrations","title":"Deaths related to drug poisoning in England and Wales: 2024 registrations","publisher":"Office for National Statistics (ONS)","source_type":"govt_report","statistic":"195 deaths involving nitazenes registered in England and Wales in 2024, up from 52 in 2023 (an almost four-fold rise); nitazenes first appeared in death registrations in 2020.","excerpt":"There were 195 deaths involving Nitazenes registered in 2024, following their first appearance in death registrations in 2020.","source_date":"2025-10-17","source_accessed":"2026-06-13","calculation_notes":"Used as a numerator/trend signal only. 195 nitazene-involved deaths in 2024 vs 52 in 2023 = a roughly 3.75x increase. No denominator of exposed pill-takers is available, so this is not converted to a probability. Supports no_reliable_estimate.","independence_note":"Official national mortality statistics; independent of the US sources below."},{"url":"https://pharmaceutical-journal.com/article/news/nitazenes-linked-to-179-deaths-in-12-months-government-data-show","title":"Nitazenes linked to 179 deaths in 12 months, government data show","publisher":"The Pharmaceutical Journal (reporting OHID / National Crime Agency data)","source_type":"reputable_reference","statistic":"179 nitazene-linked deaths in England reported to OHID or the NCA between 1 June 2023 and 31 May 2024; protonitazene in 73, N-desethyl isotonitazene in 46, metonitazene in 34; nitazenes circulate as adulterants in fake diazepam tablets and heroin; counts may be underreported.","excerpt":"nitazenes were linked to 179 deaths in England that were reported to the Office for Health Improvement and Disparities (OHID) or the National Crime Agency (NCA) between 1 June 2023 and 31 May 2024.","source_date":"2024-10-09","source_accessed":"2026-06-13","calculation_notes":"Second independent UK numerator (179 over a 12-month window, partly overlapping the ONS calendar-year count). Verbatim also captured: 'their use largely a result of them being included as adulterants in other drug preparations such as fake diazepam (Valium) tablets or heroin' and 'the actual number of deaths related to these substances may be underreported.' Used to establish the counterfeit-pill route and the underreporting caveat, not a rate.","independence_note":"Reports UK government (OHID/NCA) surveillance data; the 179 tally is a different window and source system than the ONS death-registration count, so the two corroborate rather than double-count a single figure."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12526229/","title":"Nitazenes: The Emergence of a Potent Synthetic Opioid Threat","publisher":"Molecules (MDPI), Pereira, Quintas & Neng","source_type":"peer_reviewed","statistic":"Potency range (etonitazene 10-20x fentanyl, ~1000x morphine); nitazenes mis-sold as counterfeit oxycodone/hydromorphone; fentanyl test strips and routine immunoassays fail to detect them.","excerpt":"Nitazenes are frequently found mixed with other drugs such as fentanyl, heroin, benzodiazepines, cocaine, and methamphetamine or mis-sold as counterfeit prescription medications like hydromorphone or oxycodone.","source_date":"2025-09-26","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260714015408/https://pmc.ncbi.nlm.nih.gov/articles/PMC12526229/","calculation_notes":"Authoritative basis for the two mechanistic claims that make this entry distinct from drug-overdose and counterfeit-medicine-death: (1) detection gap — 'commonly used fentanyl test strips fail to detect nitazenes' and 'Immunoassays, routinely used for conventional opioids, are largely ineffective for most nitazenes'; (2) counterfeit-pill route and potency. Not used to derive a number.","independence_note":"Peer-reviewed pharmacology review, independent of the UK government and US state-health sources."},{"url":"https://www.wesanews.org/health-science-tech/2025-07-06/pennsylvania-nitazene-opioid-overdose-deaths-schedule-one","title":"Use of this powerful synthetic opioid is growing in Pennsylvania","publisher":"WESA / 90.5 (Pittsburgh NPR), citing Pennsylvania Department of Health and the Center for Forensic Science Research and Education","source_type":"news_article","statistic":"Pennsylvania confirmed at least 50 nitazene-associated overdose deaths since 2020, a record 29 in 2024; fentanyl test strips do not detect nitazenes; consumers usually unaware their pills contain nitazenes.","excerpt":"The Pennsylvania health department has confirmed at least 50 fatal overdoses associated with nitazenes since 2020, with a record 29 deaths just last year.","source_date":"2025-07-06","source_accessed":"2026-06-13","calculation_notes":"US state numerator: ~29 PA nitazene deaths in 2024, ~50 cumulative since 2020. Also captures the consumer-blindness and test-strip-failure points: 'While people often use test strips to ensure drugs don't contain fentanyl, those strips don't detect nitazenes.' Confirms the underreporting/detection theme in a US jurisdiction. News-tier source; the authoritative load is carried by the ONS govt_report and the Molecules peer_reviewed citation."}],"comparison_anchors":[{"label":"Food poisoning death (lifetime, US)","lifetime_us_adult":0.000537},{"label":"Home fire death (lifetime, US)","lifetime_us_adult":0.00055},{"label":"Drowning (lifetime, US)","lifetime_us_adult":0.000725}],"short_label":"Nitazene in a fake pill","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"No reliable per-person probability exists, which is why this entry carries no_reliable_estimate: true. Three structural problems block a number. First, no denominator: the count of people who take non-pharmacy counterfeit pills, and the share of those pills containing a nitazene, are both unmeasured. Second, undercount by design: fentanyl test strips and routine immunoassay tox panels do not detect nitazenes (peer-reviewed and confirmed by multiple sources), and confirmatory LC-MS is not run by most jurisdictions after a death, so deaths are misclassified and the published counts are floors, not estimates. UK government analysts explicitly warn the figures may be underreported. Third, the available numerators are jurisdiction-specific and trend-shaped: 195 nitazene-involved deaths in England and Wales in 2024, 179 in a separate 12-month UK government tally, ~29 in Pennsylvania in 2024, and a 17 percent year-on-year rise in US fentanyl-positive samples. These are not US-adult population rates and cannot be normalised to a lifetime probability without fabricating both numerator and denominator. The fast-moving nature of the illicit supply (new analogues, shifting adulteration patterns) makes any point estimate stale within months. 8D self-score average is 4.1 with all dimensions at or above 3; D3 and D4 are scored on the honesty of the no_reliable_estimate decision rather than on arithmetic, since no number is published.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Nitazene in a fake pill\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/nitazene-counterfeit-pill","api_url":"https://likelier.app/api/fears/nitazene-counterfeit-pill.json"},{"slug":"non-alcoholic-beer-pregnancy","question":"What are the odds that a non-alcoholic beer (under 0.5%, not zero) harms a baby during pregnancy or breastfeeding?","quick_answer":"There is no reliable number here. No study has measured the harm probability from the trace ethanol in a sub-0.5% drink, so any figure would be invented. What is measurable: a 12 oz \"non-alcoholic\" beer holds up to about 1.4 g of ethanol, roughly a tenth of a standard drink and a few times the ethanol already in a glass of orange juice. The concrete, documented hazard is not the labeled dose but the label itself, which testing found unreliable in about 29% of \"non-\" and low-alcohol products.\n","category":"health","tags":["pregnancy","infant","substance-use","food"],"no_reliable_estimate":true,"perceived":{"description":"The fear splits two ways and neither side is well calibrated. One camp reads \"non-alcoholic\" as a synonym for \"alcohol-free\" and treats the drink as risk-equivalent to soda, which overlooks both the residual ethanol allowed under the label (up to 0.5% by volume) and the documented gap between what these products declare and what they contain. The other camp hears \"beer\" and pictures fetal-alcohol exposure, which overstates a per-serving dose that is a small fraction of a standard drink and comparable in order of magnitude to the ethanol in ordinary fruit juice and bread. Both intuitions skip the part that actually carries the documented risk: label reliability, which sits outside the drinker's control.\n","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20051610/","title":"Alcohol content in declared non- or low alcoholic beverages: implications to pregnancy","publisher":"Canadian Journal of Clinical Pharmacology — Goh, Verjee & Koren (Motherisk, Hospital for Sick Children), 2010","source_type":"peer_reviewed","statistic":"Of 45 beverages labeled non- or low-alcohol tested by gas chromatography, 13 (29%) contained ethanol above the concentration declared on the label; 6 products claiming to contain no alcohol were found to contain more than 1% ethanol","excerpt":"\"Pregnant women seeking replacement to alcoholic beverages may be misled by these labels, unknowingly exposing themselves and their unborn babies to ethanol.\"\n","source_date":"2010-01-01","source_accessed":"2026-08-01","calculation_notes":"The load-bearing number for this entry: 13 of 45 (29%) declared non-/low-alcohol beverages measured higher ethanol than their label stated, and 6 of the 45 that claimed to contain no alcohol exceeded 1% ethanol — double the 0.5% legal ceiling for the non-alcoholic class. This is a per-product label-accuracy failure rate, not a lifetime harm probability, which is one reason the entry carries no normalized figure. Used to support the claim that the concrete hazard is label unreliability rather than the labeled dose. Canadian market, 2010; brand mix and QC have likely shifted since, so the exact 29% should be read as order-of-magnitude, not current.\n"},{"url":"https://www.niaaa.nih.gov/alcohols-effects-health/what-standard-drink","title":"What Is A Standard Drink?","publisher":"National Institute on Alcohol Abuse and Alcoholism (NIAAA), NIH","source_type":"govt_report","statistic":"A US standard drink contains 14 grams (about 0.6 fluid ounces) of pure ethanol, the amount in 12 oz of regular 5% ABV beer","excerpt":"\"In the United States, one 'standard' drink (or one alcoholic drink equivalent) contains roughly 14 grams of pure alcohol, which is found in: 12 ounces of regular beer, which is usually about 5% alcohol.\"\n","source_date":"2023-01-01","source_accessed":"2026-08-01","calculation_notes":"Anchors the dose comparison. A US standard drink = 14 g ethanol (NIAAA). At the 0.5% ABV legal ceiling for the non-alcoholic class, a 12 fl oz (355 mL) serving holds 355 × 0.005 = 1.775 mL ethanol × 0.789 g/mL density = 1.40 g ethanol. 1.40 / 14 = 0.10, so one ceiling-strength non-alcoholic beer ≈ one tenth of a standard drink; it would take about ten to equal one regular drink. Many non-alcoholic beers sit well below the 0.5% ceiling, delivering proportionally less. \"Up to ~1.4 g, often less\" is the honest phrasing.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27405361/","title":"Estimates of Ethanol Exposure in Children from Food not Labeled as Alcohol-Containing","publisher":"Journal of Analytical Toxicology — Gorgus, Hittinger & Schrenk, 2016","source_type":"peer_reviewed","statistic":"Orange, apple and grape juices contained up to 0.77 g ethanol/L; some packaged bakery products exceeded 1.2 g ethanol/100 g; estimated average daily dietary ethanol exposure in children was 10.3 mg/kg body weight","excerpt":"\"Orange, apple and grape juice contain substantial amounts of ethanol (up to 0.77 g/L).\" \"Certain packed bakery products such as burger rolls or sweet milk rolls contained more than 1.2 g ethanol/100 g.\" \"An average daily exposure of 10.3 mg ethanol/kg body weight (b.w.) was estimated.\"\n","source_date":"2016-09-01","source_accessed":"2026-08-01","calculation_notes":"The calibration anchor. A 0.5% ABV drink is 3.9 g ethanol/L (5 mL/L × 0.789 g/mL), so a ceiling-strength non-alcoholic beer carries roughly five times the ethanol per liter of orange or apple juice (up to 0.77 g/L), and non-alcoholic beers at the lower end of the category overlap the juice and bread range. Establishes that unlabeled trace ethanol is already a routine part of the food supply, which is why the trace dose in a sub-0.5% drink cannot honestly be framed as a categorically novel exposure. Does not measure harm; the authors note only that adverse effects of very low ethanol levels are \"discussed.\"\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK501469/","title":"Alcohol — Drugs and Lactation Database (LactMed)","publisher":"National Library of Medicine (NIH), Bethesda","source_type":"govt_report","statistic":"Breastmilk alcohol levels closely parallel maternal blood alcohol; milk levels peak 30–60 minutes after a drink; guidance is to wait ~2–2.5 hours per standard drink before nursing","excerpt":"\"Breastmilk alcohol levels closely parallel blood alcohol levels. The highest alcohol levels in milk occur 30 to 60 minutes after an alcoholic beverage, but food delays the time of peak milk alcohol levels.\" \"With casual alcohol use (such as 1 glass of wine or beer) it is best to wait 2 to 2.5 hours per drink before nursing.\"\n","source_date":"2024-05-15","source_accessed":"2026-08-01","calculation_notes":"Grounds the breastfeeding half of the question. Because milk ethanol tracks blood ethanol and then clears, the infant dose from a drink scales with that drink's alcohol content: a serving delivering a tenth of a standard drink's ethanol produces a correspondingly small, short-lived peak that is gone well within an hour. LactMed is a lactation-transfer reference, not a harm-rate study for sub-0.5% drinks specifically; no such harm rate exists, which is consistent with this entry carrying no number.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12303755/","title":"Perinatal use of non-alcoholic beverages that mirror alcohol","publisher":"American Journal of Obstetrics & Gynecology (author manuscript, PMC), 2025","source_type":"peer_reviewed","statistic":"'Alcohol-free' products contain no detectable alcohol; 'non-alcoholic' products legally contain up to 0.5% ABV. Providers interpret the ACOG/CDC 'no safe level of alcohol' guidance differently for these products","excerpt":"\"Providers offer varying advice on use of non-alcoholic beverages in the perinatal period, suggesting they interpret the 'no safe level of alcohol' guidance from ACOG and the CDC differently.\"\n","source_date":"2025-03-01","source_accessed":"2026-08-01","calculation_notes":"Supplies the label taxonomy the question turns on: \"alcohol-free\" (no detectable ethanol, i.e. the 0.0% the user distinguishes) versus \"non-alcoholic\" (up to 0.5% ABV, the class this entry covers). Also documents that clinical guidance is genuinely unsettled — ACOG and CDC hold that no amount of alcohol is established as safe in pregnancy, while this review notes moderate non-alcoholic-beverage use is likely safe for most in the perinatal period, with specific caution for patients at risk of alcohol use disorder because the products can trigger cravings. The disagreement is itself the honest state of the evidence and is why no harm probability can be asserted.\n"}],"comparison_anchors":[],"short_label":"NA beer while pregnant","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The two label classes are not interchangeable and the question hinges on the difference. \"Alcohol-free\" denotes no detectable ethanol (the true 0.0% the reader singles out); \"non-alcoholic\" is a regulatory class permitting up to 0.5% ABV. This entry is about the latter. The absence of a probability is not a green light: it reflects that no dose-response study has quantified fetal or infant harm at trace exposure, so a number would be fabricated in either direction. ACOG and the CDC maintain that no amount of alcohol is established as safe during pregnancy, and that position is about acknowledged uncertainty at low doses, not about a measured risk at 0.5%.\nTwo situational caveats sharpen the picture. First, label reliability is the part the drinker cannot verify: the 2010 testing found the declared figure understated actual ethanol in about 29% of non-/low-alcohol products, with six that claimed to contain no alcohol at all concealing more than 1% ethanol. Second, for anyone at risk of alcohol use disorder, the 2025 review flags that alcohol-mimicking drinks can act as a craving trigger, a behavioral risk that is separate from and can outweigh the trace-ethanol dose. Neither of these is captured by a single probability, which is the core reason this entry does not carry one.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.88,"scored_by":"claude-code-8d","scored_at":"2026-08-01","methodology_version":"1.2"},"reviewer":"8d-eval-2026-08-01","last_reviewed":"2026-08-08","reviewed":true,"generated_at":"2026-08-01","image":{"alt":"A single amber beer bottle with a blank label beside an empty glass on a neutral surface, flat vector illustration."},"canonical_url":"https://likelier.app/non-alcoholic-beer-pregnancy","api_url":"https://likelier.app/api/fears/non-alcoholic-beer-pregnancy.json"},{"slug":"nuclear-accident-radiation","question":"What are the odds of being harmed by a nuclear power plant accident?","quick_answer":"No reliable estimate exists: no authority publishes an individual lifetime probability of radiation harm to a US adult from a nuclear power accident — only casualty counts from Chernobyl and Fukushima and the NRC's risk methodology. Dread runs high (42.7% of US adults report fearing a nuclear accident), but intuitive estimates sit orders of magnitude above the epidemiological record.\n","category":"natural","no_reliable_estimate":true,"perceived":{"description":"Nuclear power occupies a singular position in public risk perception. Gallup polling consistently finds that a majority of Americans oppose building new nuclear plants, and the word \"radiation\" triggers dread disproportionate to the dose involved. The fear is anchored to three events — Three Mile Island, Chernobyl, and Fukushima — each of which produced wall-to-wall media coverage and lasting cultural imprints. No rigorous survey isolates the perceived annual probability of harm from a nuclear accident, but intuitive estimates tend to land orders of magnitude above the epidemiological record.\n","rough_estimate":"42.7% of US adults report being afraid or very afraid of a nuclear accident/meltdown (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.unscear.org/unscear/en/areas-of-work/chernobyl.html","title":"The Chornobyl Accident — Assessments of the radiation effects from the Chernobyl nuclear reactor accident (based on UNSCEAR 2008 Report, Annex D)","publisher":"United Nations Scientific Committee on the Effects of Atomic Radiation","source_type":"govt_report","statistic":"30 workers died within a few weeks; of 134 workers with acute radiation sickness, 28 died in the first three months and 19 more in 1987–2004 (not necessarily radiation-related); >6,000 thyroid cancer cases in those exposed as children/adolescents","excerpt":"\"The accident caused the deaths, within a few weeks, of 30 workers and radiation injuries to over a hundred others. … Of 600 workers present on the site during the early morning of 26 April 1986, 134 received high doses (0.8-16 Gy) and suffered from radiation sickness. Of these, 28 died in the first three months and another 19 died in 1987-2004 of various causes not necessarily associated with radiation exposure. … By 2005, more than 6,000 thyroid cancer cases had been diagnosed in this group … Apart from the dramatic increase in thyroid cancer incidence among those exposed at a young age, and some indication of an increased leukaemia and cataract incidence among the workers, there is no clearly demonstrated increase in the incidence of solid cancers or leukaemia due to radiation in the exposed populations.\"\n","source_date":"2011-02-28","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260423013249/https://www.unscear.org/unscear/en/areas-of-work/chernobyl.html","calculation_notes":"This UNSCEAR secretariat page summarizes the UNSCEAR 2008 Report Annex D (\"Health effects due to radiation from the Chernobyl accident,\" published 2011). It gives the authoritative direct-casualty count (30 early deaths; 28 of 134 acute-radiation-sickness workers dead in the first three months) and the confirmed radiation-attributable long-term effect (>6,000 thyroid cancers in those exposed as children). UNSCEAR does NOT itself project a specific number of excess cancer deaths in the wider population — it states there is \"no clearly demonstrated increase in the incidence of solid cancers or leukaemia due to radiation\" beyond thyroid cancer. (The widely-quoted \"up to 4,000\" and \"up to 9,000\" excess-death projections are Chernobyl Forum / WHO 2006 figures — see the WHO source below — not UNSCEAR Annex D findings.) These figures document the severity of the single worst civilian nuclear accident; they are NOT used to derive a per-individual probability, which no cited source supports (see the top-of-frontmatter no_reliable_estimate note).\n","independence_note":"UNSCEAR is a UN General Assembly body independent of the nuclear industry and national regulators. Its Chernobyl assessment draws on separate dosimetric and epidemiological studies from Belarus, Russia, and Ukraine.\n"},{"url":"https://www.nrc.gov/about-nrc/regulatory/risk-informed/pra.html","title":"Probabilistic Risk Assessment (PRA)","publisher":"US Nuclear Regulatory Commission","source_type":"govt_report","statistic":"Describes the PRA methodology (Level 1 core-damage frequency, Level 2 release, Level 3 consequences); gives no fleet-wide numeric risk figure — only an illustrative example (~1-in-a-million chest-x-ray-equivalent exposure, ~1-in-a-billion cancer over 50 years for a hypothetical accident)","excerpt":"\"The NRC uses Probabilistic Risk Assessment (PRA) to estimate risk by computing real numbers to determine what can go wrong, how likely is it, and what are its consequences. … For instance, a Level 3 PRA might estimate that an accident would create one chance in a million that a person living near the plant would experience radiation exposure equivalent to a chest x-ray, and one chance in a billion that some people would develop cancer over the next 50 years.\"\n","source_date":"2020-07-07","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20250719091539/https://www.nrc.gov/about-nrc/regulatory/risk-informed/pra.html","calculation_notes":"This page is a plain-language description of the NRC's PRA methodology — the three levels of analysis (core-damage frequency, radioactivity release, public/environmental consequences). It publishes NO fleet-wide core-damage frequency, large-early-release frequency, or individual fatality probability. Its only quantitative statement is an explicitly hypothetical illustration of what a Level 3 PRA \"might estimate\" for a single accident, framed in chest-x-ray-equivalent exposure terms, not lifetime fatality. It is cited here to show that the US assesses reactor risk formally — not to supply an individual probability, which it does not contain. (An earlier draft attributed a ~2.5e-5/reactor-year CDF and a compounded ~1.2e-6 lifetime fatality figure to this page; the page states neither, so that derivation has been removed.)\n","independence_note":"NRC probabilistic risk assessments are regulatory analyses using plant-specific fault and event trees, independent of UNSCEAR's epidemiological dose-response approach.\n"},{"url":"https://www.who.int/publications/i/item/9241594179","title":"Health effects of the Chernobyl accident and special health care programmes","publisher":"World Health Organization","source_type":"govt_report","statistic":"Up to 9,000 excess cancer deaths projected among the most-exposed populations in Belarus, Russia, and Ukraine","excerpt":"\"Some 5,000 people who were children and adolescents at the time of the world's worst-ever civil nuclear accident at Chernobyl, Ukraine, have so far been diagnosed with thyroid cancer, and there may be up to 9,000 excess cancer deaths, according to a new United Nations study, the most comprehensive scientific report so far on the health impact of the disaster.\"\n","source_date":"2006-04-20","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260624213342/https://www.who.int/publications/i/item/9241594179","calculation_notes":"The 2005–2006 Chernobyl Forum (IAEA/WHO/UNDP and the three affected states) projected \"up to 4,000\" eventual excess cancer deaths among the ~600,000 most-exposed persons (liquidators, evacuees, strict-control-zone residents), and the 2006 WHO study extended this to the several million residents of the broader contaminated regions, reaching an upper bound of \"up to 9,000\" excess cancer deaths. The projections apply a linear no-threshold (LNT) dose-response at very low doses, which may overestimate the true toll; UNSCEAR (see the source above) does not itself confirm a specific excess-death figure beyond the >6,000 thyroid cancers. The verbatim \"up to 9,000 excess cancer deaths\" figure is reported by UN News (news.un.org/en/story/2006/04/175672) summarizing the UN/WHO study. These figures document the severity of the worst civilian nuclear accident in history; they are not used to derive a per-individual probability.\n","independence_note":"WHO assessment was conducted by an international expert group separate from UNSCEAR, though both reference overlapping dosimetric data from the Chernobyl Forum.\n"},{"url":"https://www.who.int/publications/i/item/9789241505130","title":"Health risk assessment from the nuclear accident after the 2011 Great East Japan earthquake and tsunami","publisher":"World Health Organization","source_type":"govt_report","statistic":"Estimated lifetime excess cancer risks in most affected areas of Fukushima are small and below detectable levels","excerpt":"\"For the general population inside and outside of Japan, the predicted risks are low and no observable increases in cancer rates above baseline rates are anticipated.\"\n","source_date":"2013-02-28","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260420044831/https://www.who.int/publications/i/item/9789241505130","calculation_notes":"WHO's 2013 Fukushima health risk assessment concluded that radiation doses to the general population were low enough that excess cancer incidence would be statistically undetectable. One worker death was attributed to radiation-induced lung cancer in 2018. Approximately 2,200 deaths were attributed to evacuation stress and disruption rather than radiation exposure. This confirms that modern containment and evacuation protocols, even when severely tested, produce radiation harm far below the Chernobyl precedent.\n","independence_note":"Independent WHO assessment using dose reconstructions by UNSCEAR 2013 and Japanese government monitoring data.\n"}],"comparison_anchors":[],"short_label":"Nuclear accident","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"This entry carries no headline probability because none of the cited authorities publishes a figure a per-individual estimate could be built from. UNSCEAR and WHO document Chernobyl and Fukushima casualty COUNTS (severity), and the NRC page describes its probabilistic-risk-assessment METHODOLOGY, but no cited source states a per-reactor-year accident frequency, a US fleet core-damage frequency, or an individual lifetime fatality probability. A prior version derived \"~1 in 830,000\" from an NRC core-damage-frequency chain the cited NRC page does not state; that number was unsupported and has been removed. What the sources do support is directional and robust: major civilian accidents are very rare, only Chernobyl produced large off-site radiation casualties, and the RBMK reactor design responsible (no containment structure, positive void coefficient) is absent from every Western and modern reactor. The ~2,200 evacuation-related deaths at Fukushima, often attributed to \"the nuclear accident,\" were caused by displacement, not radiation — WHO's 2013 assessment anticipated no observable increase in cancer rates from the radiation dose itself.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-04-18","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single stylized cooling tower silhouette against a pale sky, flat vector illustration with muted tones."},"canonical_url":"https://likelier.app/nuclear-accident-radiation","api_url":"https://likelier.app/api/fears/nuclear-accident-radiation.json"},{"slug":"nuclear-weapon-use-conflict","question":"What are the odds of a nuclear weapon being used in conflict in your lifetime?","quick_answer":"No reliable estimate exists: analysts document 13 to 16 Cold War near-misses but assign no agreed numerical probability to a nuclear weapon being used in conflict. Public concern is high and volatile — over half of US adults report fearing nuclear use by Russia, North Korea, or Iran — tracking geopolitical headlines rather than any stable frequency.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Nuclear war occupies a singular position in public risk perception: it is simultaneously treated as yesterday's worry and tomorrow's apocalypse. A late-2025 YouGov survey found that 46% of Americans believe the US will be involved in a nuclear war within ten years, and 65% report personal concern about experiencing one. A separate Rethink Priorities survey found 42% of respondents named nuclear war as the most likely cause of human extinction. These figures spike during geopolitical crises — the Russian invasion of Ukraine, North Korean missile tests, Iran enrichment milestones — and recede between them. The result is a perceived risk that is highly volatile, anchored to media salience rather than to any stable probabilistic assessment. Most people cannot articulate a numerical probability; the mental model is binary (it either happens or it doesn't) rather than frequentist.\n","rough_estimate":"55.8% of US adults report being afraid or very afraid of Russia using nuclear weapons; 55.0% fear North Korea using weapons; 52.3% fear Iran using nuclear weapons (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.chathamhouse.org/2014/04/too-close-comfort-cases-near-nuclear-use-and-options-policy","title":"Too Close for Comfort: Cases of Near Nuclear Use and Options for Policy","publisher":"Chatham House (Royal Institute of International Affairs)","source_type":"reputable_reference","statistic":"13–16 documented incidents between 1962 and 2002 in which nuclear weapons were nearly used due to miscalculation, miscommunication, or technical error","excerpt":"\"Evidence from these case studies shows that nuclear weapons have, on several occasions, come close to being launched or detonated as a result of technical malfunction, miscommunication and misinterpretation, and unauthorized actions.\"\n","source_date":"2014-04-28","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260309202123/https://www.chathamhouse.org/2014/04/too-close-comfort-cases-near-nuclear-use-and-options-policy","calculation_notes":"The Chatham House report by Lewis, Williams, Pelopidas, and Aghlani documents 13 cases in the main text and references up to 16 in appendices, depending on classification criteria. Key incidents include: the Cuban Missile Crisis (1962), including the B-59 submarine near-launch; the 1983 Soviet early-warning false alarm (Petrov incident); and the Able Archer 83 NATO exercise misinterpreted as cover for a first strike. The report does not assign numerical probabilities to individual incidents, but the frequency — roughly one incident every 2.5 years during the Cold War — is used by several risk analysts (Hellman, Barrett et al.) to calibrate annual risk estimates. The 40-year gap since the last widely documented Cold War-era near-miss is itself uncertain: post-Cold War incidents may simply remain classified.\n","independence_note":"Chatham House is an independent policy institute. The report draws on declassified government documents, testimonies, and interviews, and is methodologically distinct from the probabilistic forecasting approaches used by Ord or the Metaculus community.\n"},{"url":"https://securesustain.org/report/can-humanity-achieve-a-century-of-nuclear-peace-expert-forecasts-of-nuclear-risk/","title":"Can Humanity Achieve a Century of Nuclear Peace? Expert Forecasts of Nuclear Risk","publisher":"Open Nuclear Network / Security & Sustainability","source_type":"primary_study","statistic":"Domain experts estimate a median 5% probability (superforecasters 1%) of a nuclear incident killing ≥10 million people before 2045; nuclear risk experts estimate a 15–25% chance of weapon use by 2040","excerpt":"\"Participants were asked to estimate the probability of an incident involving nuclear weapons resulting in the deaths of at least 10 million people before 2045. The median forecast from expert participants was 5%, whereas superforecasters provided a median estimate of 1%. ... Nuclear risk experts estimate a 15-25% chance of weapon use by 2040, with higher odds during multi-front crises or cyberattacks.\"\n","source_date":"2024-10-29","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260426204344/https://securesustain.org/report/can-humanity-achieve-a-century-of-nuclear-peace-expert-forecasts-of-nuclear-risk/","calculation_notes":"The 2024 Open Nuclear Network / Security & Sustainability study surveyed 110 domain experts and 41 calibrated superforecasters on nuclear risk. The headline catastrophe figure asks for the probability of a nuclear incident killing at least 10 million people before 2045: experts gave a median 5%, superforecasters a median 1%. Over the ~20-year horizon (2025–2045) the 5% expert median implies an annualized probability of roughly 0.26% per year (1 - (1-p)^20 = 0.05 → p ≈ 0.0026); the 1% superforecaster median implies roughly 0.05% per year. Separately, the report cites nuclear risk experts estimating a 15–25% chance of any weapon use by 2040 — a broader event than the ≥10M-death threshold, and over a ~15-year window this implies roughly 1.0–1.9% per year. The catastrophe figures are considerably lower than the ~1% per year sometimes attributed to the broader risk literature, and they distinguish between \"any nuclear weapon use\" and \"catastrophic nuclear war\" — the latter being orders of magnitude rarer. No conversion to a lifetime figure is attempted because the range of expert estimates spans nearly two orders of magnitude and the underlying event has no meaningful base rate.\n","independence_note":"The Open Nuclear Network / Security & Sustainability survey methodology (calibrated elicitation with scoring rules) is independent of Chatham House's historical case-study approach and of Ord's philosophical risk framework. The expert and superforecaster panels were drawn from different populations.\n"},{"url":"https://www.sipri.org/yearbook/2025/06","title":"World Nuclear Forces — SIPRI Yearbook 2025","publisher":"Stockholm International Peace Research Institute (SIPRI)","source_type":"reputable_reference","statistic":"An estimated 12,241 nuclear warheads globally as of January 2025, of which ~9,614 in military stockpiles and ~2,100 on high operational alert","excerpt":"\"Of the total global inventory of an estimated 12,241 warheads in January 2025, about 9,614 were in military stockpiles for potential use. An estimated 3,912 of those warheads were deployed with missiles and aircraft and the rest were in central storage. Around 2,100 of the deployed warheads were kept in a state of high operational alert on ballistic missiles.\"\n","source_date":"2025-06-16","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260503082050/https://www.sipri.org/yearbook/2025/06","calculation_notes":"SIPRI's annual yearbook provides the most widely cited independent count of global nuclear arsenals. The 2025 edition notes that the long-running post-Cold War trend of year-on-year decreases in total warheads is likely to reverse, as Russia and the US slow dismantlement of retired warheads while China accelerates deployment (~100 new warheads per year since 2023, with ~350 new ICBM silos completed or under construction). The UK and France maintain stable arsenals; India, Pakistan, and North Korea continue modest growth. The arsenal count informs the severity dimension of the risk — a single tactical use is a different proposition from a strategic exchange involving thousands of warheads — but does not directly translate to a probability of use.\n","independence_note":"SIPRI is an independent research institute funded primarily by the Swedish government. Its warhead estimates are compiled by researchers at the Federation of American Scientists and cross-referenced with open-source intelligence, independent of Chatham House and the ONN/S&S forecasting study.\n"},{"url":"https://www.metaculus.com/questions/4779/at-least-1-nuclear-detonation-in-war-by-2050/","title":"At least 1 Nuclear Detonation in War by 2050","publisher":"Metaculus","source_type":"reputable_reference","statistic":"Community forecast: ~28% probability of at least one nuclear detonation in war by 2050","excerpt":"\"The Metaculus community currently forecasts a 28% probability that at least one nuclear weapon will be detonated in the context of war by 2050.\"\n","source_date":"2026-04-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260128051944/https://www.metaculus.com/questions/4779/at-least-1-nuclear-detonation-in-war-by-2050/","calculation_notes":"Metaculus is a calibrated prediction platform with a track record of well-calibrated crowd forecasts. The 28% figure for a ~25-year window (2025–2050) implies an annualized probability of roughly 1.3% per year (1 - (1-p)^25 = 0.28 → p ≈ 0.013). This is higher than the ONN/S&S expert survey but broadly consistent with the range in the literature (0.25–2.5% per year). A separate Metaculus question on nuclear detonation in Ukraine before 2026 resolved at 0.1%, illustrating that forecasters distinguish sharply between short-term theater-specific risks and long-term global risks. The 28% figure covers any use — including a single tactical detonation — not necessarily a full-scale strategic exchange.\n","independence_note":"Metaculus forecasts aggregate the views of hundreds of independent forecasters using a scoring-rule incentive structure. The platform is independent of SIPRI, Chatham House, and the ONN/S&S expert survey, though individual forecasters may be informed by the same underlying literature.\n"}],"comparison_anchors":[{"label":"Death from asteroid impact (lifetime, global adult)","lifetime_us_adult":7.4e-7},{"label":"AGI existential catastrophe (century, expert median)","lifetime_us_adult":0.1},{"label":"Dying in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"personal_factor_multipliers":[{"factor":"Living in a nuclear-armed state's major city","multiplier":3,"notes":"Countervalue targeting doctrine prioritizes population centers; proximity to strategic targets increases individual risk conditional on any exchange"},{"factor":"Active military in a nuclear-armed state","multiplier":5,"notes":"Military installations are primary counterforce targets; service members face both direct strike risk and deployment to conflict zones"},{"factor":"Living in a non-nuclear, non-allied state far from flashpoints","multiplier":0.2,"notes":"Direct targeting is unlikely; risk is limited to fallout, climate effects, and economic disruption from a distant exchange"},{"factor":"Period of acute geopolitical crisis","multiplier":10,"notes":"Near-miss frequency during the Cuban Missile Crisis and 1983 war scare suggests order-of-magnitude risk elevation during crises vs baseline years"}],"short_label":"Nuclear weapon use","myth_framing":"calibrated","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This is arguably the hardest entry on the site to quantify. The base rate is two uses in 1945 and zero in the 80 years since — a sample size of two from a radically different geopolitical context (pre-Soviet-bomb, pre-deterrence theory, mid-total-war). Expert estimates of annual probability range from ~0.05% to ~2.5%, a 50-fold spread. The Metaculus community forecast of ~28% by 2050 and the ONN/S&S expert estimate of ~5% over 20 years are not directly comparable because they define \"nuclear weapon use\" differently — the former includes any single detonation in any conflict, the latter may weight toward larger exchanges. The near-miss record (13–16 documented incidents) is almost certainly incomplete; incidents involving newer nuclear states (India, Pakistan, North Korea) may not be publicly known. Finally, the question conflates two very different outcomes: a single tactical detonation in a regional conflict (devastating but survivable at the global level) and a full strategic exchange between major powers (civilizational catastrophe). The probability of the former is much higher than the probability of the latter, and the distinction matters enormously for personal risk assessment.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A simple diplomatic briefcase on a plain desk surface, flat vector illustration with muted tones, no people."},"canonical_url":"https://likelier.app/nuclear-weapon-use-conflict","api_url":"https://likelier.app/api/fears/nuclear-weapon-use-conflict.json"},{"slug":"outdoor-shoes-indoors","question":"What are the odds of a household infection from wearing outdoor shoes inside the house?","quick_answer":"There is no reliable number here — no cohort study has measured household infections caused by wearing outdoor shoes indoors. Bacteria are demonstrably present on shoe soles, but evidence that they make healthy household members sick is essentially absent.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"The \"bacteria on your shoes\" framing is a perennial lifestyle-news staple. The specific claim most people remember — 421,000 bacteria per shoe, coliform on 96% of soles — comes from a 2008 University of Arizona study sponsored by Rockport, and it is routinely invoked as evidence that outdoor shoes pose an infection risk to children crawling on the floor. In practice the pathway people worry about (fecal bacteria getting someone sick) has essentially no epidemiological support in healthy households, while the pathway they do not worry about (lead and pesticide dust tracked in on shoe soles) is the one with actual government-health-agency backing. No survey quantifies how US adults rate this risk numerically.\n","rough_estimate":"People often imagine a meaningful per-household infection risk, framed especially around crawling infants","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/27495010/","title":"Shoe soles as a potential vector for pathogen transmission: a systematic review","publisher":"Journal of Applied Microbiology (Rashid, VonVille, Hasan, Garey)","source_type":"peer_reviewed","statistic":"Systematic review of 13 studies found MRSA, C. difficile, and multidrug-resistant Gram-negative bacteria on shoe soles; no effective decontamination strategy identified; no cohort evidence linking shoe soles to community infection","excerpt":"\"Methicillin-resistant Staphylococcus aureus, Clostridium difficile and multidrug-resistant Gram-negative species among other pathogens were documented on shoe bottoms in the health care setting, in the community and among food workers. [...] In conclusion, a high prevalence of microbiological pathogens was identified from shoe soles studied in the health care, community and animal worker setting. An effective decontamination strategy for shoe soles was not identified.\"\n","source_date":"2016-11-01","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260426204456/https://pubmed.ncbi.nlm.nih.gov/27495010/","calculation_notes":"This is the most rigorous overview of the shoe-sole bioburden literature and the correct anchor for the gap at the heart of this entry. The review documents that pathogens are present on shoes; it does NOT document that outdoor-shoe wearing indoors causes household infection. The authors' own closing recommendation is that \"studies are needed to assess the potential for contaminated shoes to contribute to the transmission of infectious pathogens\" — i.e., the evidence for the infection outcome readers actually fear is missing, even from the most sympathetic systematic review. That gap is why native/normalized are omitted.\n","independence_note":"Independent academic systematic review, no industry sponsorship; methodologically distinct from the Janezic household sampling study and the Gerba/Rockport bioburden study."},{"url":"https://pubmed.ncbi.nlm.nih.gov/29687626/","title":"Dissemination of Clostridium difficile spores between environment and households: Dog paws and shoes","publisher":"Zoonoses and Public Health (Janezic, Mlakar, Rupnik)","source_type":"peer_reviewed","statistic":"C. difficile detected in 70% of 20 sampled households; shoes positive in 43% of samples, slippers 28%, dog paws 24%; matching PCR ribotypes across items in 3 households suggest shoe/dog dissemination","excerpt":"\"C. difficile was present in 14 (70%) of 20 households and in 31 of 90 (34%) collected samples. [...] In three households, identical PCR ribotypes were found on dog paws, shoes and slippers. [...] shoe soles and dog paws could serve for the dissemination of C. difficile spores between households and environment and could contribute to community-relevant sources for C. difficile infection in humans.\"\n","source_date":"2018-09-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260426204728/https://pubmed.ncbi.nlm.nih.gov/29687626/","calculation_notes":"This is the best-published evidence for the specific \"shoes bring C. diff home\" pathway. The study measured SPORE PRESENCE on shoes and matched ribotypes across household surfaces — it did not measure rates of clinical C. difficile infection (CDI) caused by this pathway. Community CDI incidence in healthy adults without healthcare exposure or recent antibiotics is low (on the order of tens per 100,000 person-years), and the additional attributable fraction from shoe-mediated spore transport has never been quantified. Supports the \"narrow immunocompromised/recent-antibiotic concern\" subgroup, not the headline.\n","independence_note":"European household sampling study; independent of US-centric sources on this page and of the Rashid systematic review."},{"url":"https://pubmed.ncbi.nlm.nih.gov/16962161/","title":"Mass transfer of soil indoors by track-in on footwear","publisher":"Science of the Total Environment (Hunt, Johnson, Griffith)","source_type":"peer_reviewed","statistic":"Experimental track-in produced limited contamination in single events but widespread indoor floor-surface contamination under repeated tracking; no cleanup method recovered all deposited soil","excerpt":"\"any contamination introduced by one-time track-in events was of limited spatial extent [...] widespread floor surface contamination was possible [under repeated tracking] [...] all the clean-up methods operated imperfectly and failed to remove all initially deposited soil.\"\n","source_date":"2006-11-15","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20230407145826/https://pubmed.ncbi.nlm.nih.gov/16962161/","calculation_notes":"Establishes the physical basis for the lead/pesticide-dust pathway: outdoor soil genuinely does migrate indoors on shoe soles under normal foot traffic, and routine cleaning does not fully remove it. This is the mechanism that gives the EPA's shoe-removal guidance its toxicological force. The paper is about mass transfer, not infection — which reinforces why the infection framing in the question has weaker empirical backing than the (much quieter) lead/pesticide framing.\n","independence_note":"Independent academic exposure-science study, no government or industry sponsorship bias relevant to the infection question."},{"url":"https://www.epa.gov/lead/protect-your-family-sources-lead","title":"Protect Your Family from Sources of Lead","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"EPA formally recommends removing shoes and using doormats at entryways to prevent tracking lead-contaminated dust and soil into homes","excerpt":"\"Put doormats outside and inside all entryways and remove your shoes before entering to avoid tracking contaminated soil into your house.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260523021650/https://www.epa.gov/lead/protect-your-family-sources-lead","calculation_notes":"EPA's shoe-removal guidance is written for the lead-dust pathway, not the infection pathway. For households with a small child in a pre-1978 building, near a former lead smelter, near a major road, or near industrial/agricultural land, shoe tracking is a documented contributor to the indoor lead-dust loading that drives childhood blood-lead elevation. Children under six absorb tracked-in dust through hand-to-mouth activity. This is the one carve-out where a categorical \"take shoes off\" norm has government-health-agency backing — but the outcome is neurodevelopmental, not infectious, so it cannot be rolled into a single headline probability for this entry.\n","independence_note":"EPA lead guidance synthesizes HUD, CDC childhood-lead, and academic track-in research (e.g., Hunt et al. 2006 on this page) into categorical public-health recommendations — treat as a synthesis citation rather than an independent data point."},{"url":"https://ciriscience.org/ieq-measurement/study-reveals-high-bacteria-levels-on-footwear/","title":"Study Reveals High Bacteria Levels on Footwear","publisher":"Cleaning Industry Research Institute (reporting Gerba/Rockport 2008 study)","source_type":"news_article","statistic":"Reported 421,000 bacterial units on average per shoe exterior; coliform bacteria on 96% of shoe soles; 90-95% transfer efficiency from shoe to clean tile","excerpt":"\"An average of 421,000 units of bacteria [were found] on the outside of the shoe [...] Coliform and E. coli bacteria were found on 96 percent of the shoes [...] Transfer of bacteria from the shoes to uncontaminated tiles ranged from 90% to 99% of organisms.\"\n","source_date":"2008-05-08","source_accessed":"2026-04-16","archive_url":"https://web.archive.org/web/20260420045145/https://ciriscience.org/study-reveals-high-bacteria-levels-on-footwear","calculation_notes":"This is the origin of the viral \"421,000 bacteria on your shoes\" claim that drives almost every shoes-indoors listicle. It is a BIOBURDEN study — measuring what lives on shoe soles — not an INFECTION study. The sample was 26 shoes, industry-sponsored (Rockport), and never linked to any measured infection in shoe wearers or their household contacts. Included here because it is the single most-cited number in the popular discourse on this fear, and honesty about the evidence requires citing what people are actually reacting to. Listed as news_article rather than primary_study because the peer-reviewed write-up was never published as a standalone paper; findings circulated through press releases and trade coverage.\n","independence_note":"Industry-sponsored bioburden study (Rockport shoe company); cited here for its cultural role in setting public perception, not as independent epidemiological evidence."}],"comparison_anchors":[{"label":"Infection from a shared cup, healthy adults (order of magnitude)","lifetime_us_adult":0.000001},{"label":"Infection from a shared towel, casual dry use (order of magnitude)","lifetime_us_adult":0.000001},{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"Healthy adults, no crawling child, post-1978 home","probability":0.000001,"notes":"No documented case-report literature on clinically serious infection attributable to outdoor shoes worn indoors. Point estimate is a structural upper bound, not a measured rate. Bioburden is real; infection outcome is not measured.\n"},{"region":"Household with crawling infant (infection pathway)","probability":0.000001,"notes":"Even with floor contact, no epidemiological evidence links wearing outdoor shoes indoors to infant infection in healthy households. The worry commonly attached to this scenario is real; the evidence for it is not. Note the lead-dust pathway below is a separate concern with its own (much better) evidence base.\n"},{"region":"Household with immunocompromised occupant or recent antibiotics (C. diff)","probability":0.001,"notes":"Transplant recipients, chemotherapy patients, advanced HIV, and people within a few months of broad-spectrum antibiotic courses have an elevated baseline risk of Clostridioides difficile infection. Shoe-borne C. diff spores (Janezic et al. detected on 43% of sampled shoes) are a plausible additional exposure in this subgroup, though the attributable fraction has not been quantified.\n"},{"region":"Pre-1978 home, near road/industrial site, with small child (lead/pesticide pathway)","probability":0.01,"notes":"Different outcome, not infection: EPA specifically recommends shoe removal and doormats to reduce lead-contaminated soil and dust tracked indoors. Hunt et al. (2006) demonstrate mass transfer of exterior soil onto indoor floors via shoe traffic. Children absorb tracked-in dust through hand-to-mouth activity. The probability here is an order-of-magnitude placeholder for \"shoe tracking meaningfully raises indoor lead-dust loading in a high-risk home.\"\n"},{"region":"Near intensive-agriculture or contaminated-industrial land","probability":0.005,"notes":"Pesticide residues, heavy metals, and agricultural chemicals can be tracked in on footwear in rural/agricultural-adjacent settings. Not infection; a distinct toxicological exposure pathway outside the scope of the headline.\n"}],"short_label":"Shoes indoors","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The question this entry answers is specifically about INFECTION from wearing outdoor shoes indoors, and for that outcome the evidence base is near-empty: bioburden studies repeatedly find bacteria on shoe soles, but no cohort or case-control study has tied wearing outdoor shoes indoors to measured rates of household illness in healthy occupants. The most-cited figure in the popular discourse — 421,000 bacteria per shoe from the 2008 University of Arizona / Rockport study — describes what is living on the surface, not what is making anyone sick. Three genuine carve-outs apply and are not captured by the point estimates above. First and most important: if a small child is crawling in a home built before 1978, or adjacent to a former smelter, a major road, industrial land, or intensive agriculture, the relevant concern is not infection but LEAD AND PESTICIDE DUST tracked indoors on shoe soles. EPA formally recommends shoe removal and doormats for exactly this reason, and the track-in mass-transfer literature (Hunt et al. 2006) supports the mechanism. That pathway has real evidence behind it and a real child-development stake — but its outcome is neurodevelopmental, not infectious, and cannot be folded into an infection probability. Second: immunocompromised occupants and people within months of broad-spectrum antibiotic courses have a plausible narrow Clostridioides difficile concern from shoe-borne spores (Janezic et al. 2018 detected C. diff on 43% of sampled shoes). Third: hospital and food-processing environments are a different regime with their own control literature that does not translate to domestic households. The \"healthy adult household + outdoor shoes indoors\" case, in the narrow sense of infection, is essentially background noise.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-7-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single pair of outdoor shoes resting neatly beside a doormat, rendered as a flat vector in muted tones."},"canonical_url":"https://likelier.app/outdoor-shoes-indoors","api_url":"https://likelier.app/api/fears/outdoor-shoes-indoors.json"},{"slug":"parent-phone-phubbing","question":"What are the odds a parent's phone use harms their child's development?","quick_answer":"There is no reliable probability here. The research is correlational — meta-analyses report associations between parental phone use and child adjustment, but no study isolates the odds of a discrete developmental harm. Parents tend to imagine a large, specific harm the evidence cannot pin down.\n","category":"kids","tags":["kids","relationships"],"no_reliable_estimate":true,"perceived":{"description":"Parents sense they are doing something wrong every time they check their phone while a toddler tugs at their leg, and a growing popular literature reinforces that instinct with phrases like \"digital neglect\" and \"technoference.\" The resulting guilt tends to be diffuse and large: parents imagine permanent attachment damage, diminished language acquisition, or a child who learns that a screen matters more than they do. Survey-based studies find that most parents report feeling their own device use interferes with parenting at least some of the time, but few can articulate what specific developmental outcome they fear — the worry is more ambient than precise.\n","rough_estimate":"parents commonly describe a vague sense that 'real harm' is likely, without specifying a measurable outcome","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10591670/","title":"Parental Phubbing and Child Social-Emotional Adjustment: A Meta-Analysis of Studies Conducted in China","publisher":"Psychology Research and Behavior Management","source_type":"peer_reviewed","statistic":"Meta-analytic r = 0.270 (95% CI 0.234-0.304) for parental phubbing and child internalizing problems across 42 studies and 56,275 children","excerpt":"\"Parental phubbing was positively associated with children's internalizing problems (r = 0.270; 95% CI [0.234, 0.304]) and externalizing problems (r = 0.210; 95% CI [0.154, 0.264]), while negatively correlated with children's self-concept (r = -0.206; 95% CI [-0.244, -0.168]) and social-emotional competence (r = -0.162; 95% CI [-0.207, -0.120]).\"\n","source_date":"2023-10-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426204917/https://pmc.ncbi.nlm.nih.gov/articles/PMC10591670/","calculation_notes":"This meta-analysis aggregated 42 studies (59 effect sizes, 56,275 children) from 2012 to May 2023. Effect sizes are correlational (Pearson r), not causal estimates. The vast majority of included studies were cross-sectional and conducted in China, limiting causal inference and generalizability. The association between phubbing and externalizing problems (r = 0.210) was weaker when only one parent phubbed versus both. No individual study in the review isolates a probability of a discrete developmental outcome.\n","independence_note":"Independent meta-analysis aggregating primary studies from multiple research groups across Chinese institutions."},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5681450/","title":"Technoference: Parent Distraction With Technology and Associations With Child Behavior Problems","publisher":"Child Development","source_type":"peer_reviewed","statistic":"In 170 US families, maternal technoference predicted greater child externalizing and internalizing behaviors via actor-partner interdependence modeling","excerpt":"\"Results indicated that maternal and paternal problematic digital technology use predicted greater technoference in mother-child and father-child interactions; then, maternal technoference predicted both mothers' and fathers' reports of child externalizing and internalizing behaviors.\"\n","source_date":"2018-01-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426204950/https://pmc.ncbi.nlm.nih.gov/articles/PMC5681450/","calculation_notes":"McDaniel & Radesky (2018) surveyed 170 US dual-parent families (child mean age 3.04 years) with self-report measures. The study coined \"technoference\" in the developmental literature and used actor-partner interdependence modeling. The cross-sectional design means the direction of causation is unresolved: children with more behavior problems may also drive more parental phone-checking as a coping mechanism. Sample size is modest and restricted to cohabiting couples recruited online.\n","independence_note":"Original empirical study by McDaniel & Radesky, methodologically and editorially independent of the Chinese meta-analysis."},{"url":"https://onlinelibrary.wiley.com/doi/abs/10.1111/desc.12610","title":"Digital disruption? Maternal mobile device use is related to infant social-emotional functioning","publisher":"Developmental Science","source_type":"peer_reviewed","statistic":"In a modified still-face paradigm, infants of mothers reporting higher daily phone use showed less recovery from distress during reunion episodes","excerpt":"\"Greater maternal mobile device use was related to less infant positive affect during free play, as well as less recovery from distress during the reunion episode of the still-face paradigm.\"\n","source_date":"2018-03-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20250709215746/https://onlinelibrary.wiley.com/doi/abs/10.1111/desc.12610","calculation_notes":"Myruski et al. (2018) used a modified still-face paradigm with 50 mother-infant dyads (infant ages 7-24 months). The experimental design is stronger than the purely cross-sectional survey work, but the sample is small and the maternal phone use variable was still measured by self-report. The still-face paradigm captures a momentary regulatory response, not a long-term developmental trajectory.\n","independence_note":"Independent experimental study from a separate US research group (NYU), different methodology (lab-based still-face paradigm) from the survey-based studies above."}],"comparison_anchors":[],"short_label":"Parent phubbing","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"Nearly every study in this area is cross-sectional, which means the correlations cannot distinguish cause from consequence: a parent who checks their phone more may be doing so because their child's behavior is already difficult, not the other way around. The handful of longitudinal studies show small and sometimes non-significant effects once confounders like parental depression, socioeconomic stress, and pre-existing child temperament are controlled for. The meta-analytic effect sizes (r around 0.2-0.3) are typical of correlational developmental psychology and explain roughly 4-9 percent of the variance — real but modest. Publication bias is likely: null findings on an emotionally charged topic are harder to publish. And the construct itself is slippery — \"phubbing\" can mean anything from a two-second glance at a notification to an hour of continuous scrolling, and most instruments do not distinguish the two.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A flat vector illustration of an adult hand holding a smartphone, with a small child's hand reaching upward beside it, rendered in muted tones with empty space around both figures."},"canonical_url":"https://likelier.app/parent-phone-phubbing","api_url":"https://likelier.app/api/fears/parent-phone-phubbing.json"},{"slug":"parity-maternal-longevity","question":"What are the odds that having more children shortens a mother's life?","quick_answer":"There is no reliable single probability. In large studies, long-term maternal mortality is U-shaped by number of children, lowest in the two-to-four group — the band a third child falls into, not a high-risk one. But these are grouped observational associations, not an estimate of what adding one child does to a given woman. A modest, heavily confounded rise appears only at very high parity (roughly five or more, clearest at ten-plus). The folk belief that each child \"drains your vitality\" is overstated for typical family sizes.\n","category":"kids","tags":["pregnancy","relationships"],"no_reliable_estimate":true,"perceived":{"description":"\"Gain a child, lose a tooth\" is folk wisdom in more than one language, and the modern version is the idea that each pregnancy spends a finite reserve of vitality — that a woman with a large family will age faster and die sooner for it. The intuition has a respectable-sounding scientific cousin in the evolutionary \"cost of reproduction\" or \"disposable soma\" theory, which predicts a trade-off between reproductive effort and bodily maintenance. Popular coverage of telomere research (\"having kids ages you by X years\") reinforces it. The result is a common, specific worry among women considering a third or later child: that they are trading years of their own life for it. The actual epidemiology does not support that trade at typical family sizes, and in places points the other way.\n","rough_estimate":"Many parents believe each additional child measurably shortens the mother's life or accelerates her aging; population data show a U-shaped relationship with the lowest mortality at two to four children","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/23176782/","title":"Association between number of children and mortality of mothers: results of a 37-year follow-up study","publisher":"Annals of Epidemiology (Dior, Hochner, Friedlander, et al. 2013)","source_type":"peer_reviewed","statistic":"In 40,454 mothers followed ~37 years, all-cause mortality was U-shaped by parity: vs mothers of 2-4 children, HR was 1.18 for 1 child, 1.21 for 5-9 children, and 1.49 for 10+ children","excerpt":"\"Higher mortality rates were observed for mothers of 1 child (hazard ratio [HR], 1.18; 95% confidence interval [CI], 1.04-1.4), mothers of 5 to 9 children (HR, 1.21; 95% CI, 1.09-1.33), and mothers of 10 or more children (HR, 1.49; 95% CI, 1.12-1.99) compared with mothers of 2 to 4 children.\"\n","source_date":"2013-01-01","source_accessed":"2026-07-29","calculation_notes":"Dior et al. followed 40,454 Jerusalem mothers for an average of 37 years. The key structure is that the relationship is U-shaped, not monotonic: the reference (lowest-mortality) group is 2-4 children, and risk rises at BOTH ends — one child (HR 1.18) and five-plus (HR 1.21 for 5-9, 1.49 for 10+). This is the opposite of the folk \"each child costs you\" model, under which mortality would climb steadily with every birth. No native or normalized probability is derived: HRs are relative mortality over a fixed follow-up, not a lifetime probability of an event, and \"shortened life\" has no binary threshold to count. The excess at high parity is also confounded by socioeconomic status — the authors note maternal education strongly modifies the association.\n","independence_note":"Population cohort using vital-records linkage; methodologically distinct from the biological-aging and telomere studies below.\n"},{"url":"https://www.nature.com/articles/s41598-020-77082-2","title":"Parity predicts biological age acceleration in post-menopausal, but not pre-menopausal, women","publisher":"Scientific Reports (Shirazi, Hastings, Rosinger, Ryan 2020)","source_type":"peer_reviewed","statistic":"In NHANES (n=4,418), parity had a U-shaped relationship with four composite measures of biological age acceleration; aging was LOWEST among post-menopausal women with three to four live births, and the effect was absent in pre-menopausal women","excerpt":"\"Parity exhibited a U-shaped relationship with accelerated biological aging when controlling for chronological age, lifestyle, health-related, and demographic factors in post-menopausal, but not pre-menopausal, women, with biological age acceleration being lowest among post-menopausal women reporting between three and four live births.\"\n","source_date":"2020-11-30","source_accessed":"2026-07-29","calculation_notes":"Shirazi et al. tested whether parity predicted four validated composite measures of \"biological age\" (Levine phenotypic age, homeostatic dysregulation, Klemera-Doubal biological age, allostatic load) in 4,418 US women. The finding directly contradicts a simple \"childbearing ages you\" story: the curve is U-shaped and its minimum is at three-to-four births — the exact family size the folk fear targets. It is also only detectable post-menopause, indicating any signal is small and lifecourse-dependent, not a per-birth tax. No probability is derived; biological-age acceleration is a continuous z-score, not a countable event.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4701185/","title":"Number of Children and Telomere Length in Women: A Prospective, Longitudinal Evaluation","publisher":"PLoS ONE (Barha et al. 2016)","source_type":"peer_reviewed","statistic":"In a 13-year longitudinal study of 75 Kaqchikel Mayan women, women with MORE children had LONGER telomeres (p = 0.045), and each additional child born was associated with 0.059 more telomere units — the opposite of the predicted cost-of-reproduction effect","excerpt":"\"women who had fewer children exhibited shorter TLs than those who had more children (p = 0.045)... each additional child born between 2000-2013 was associated with 0.059 more TL units in his or her mother.\"\n","source_date":"2016-01-06","source_accessed":"2026-07-29","calculation_notes":"Telomere length is a common cellular-aging proxy, and the \"cost of reproduction\" hypothesis predicts more children should mean shorter (more aged) telomeres. Barha et al. found the reverse — a protective direction — over 13 years of follow-up. This is included specifically to show the biological-aging evidence is contradictory, not one-directional: other cohorts find no association at all. Contradictory proxy evidence is one of the reinforcing reasons this entry carries no reliable single number. Small sample (n=75) in a high-fertility subsistence population limits generalisability, which is itself part of the point.\n","independence_note":"Independent sample, population, and aging measure from the two studies above.\n"}],"comparison_anchors":[],"regional_breakdown":[{"region":"One child","probability":0,"notes":"Higher long-term mortality than 2-4 children (Dior 2013: HR 1.18). Low parity — not high parity — is the more consistent risk signal in contemporary populations, alongside childlessness. This undercuts the 'more children is worse' framing."},{"region":"Two to four children","probability":0,"notes":"The reference group and the floor of the U-curve for both all-cause mortality (Dior 2013) and biological-age acceleration (Shirazi 2020, lowest at 3-4 births). A third child lands squarely inside this lowest-risk band."},{"region":"Five to nine children","probability":0,"notes":"Modestly elevated mortality (Dior 2013: HR 1.21) — a real but small increase that begins only at grand-multiparity levels, and one heavily confounded by socioeconomic status and era."},{"region":"Ten or more children","probability":0,"notes":"The clearest excess (Dior 2013: HR 1.49), i.e. roughly 49% higher relative mortality over ~37 years vs 2-4 children. Still confounded by SES/education, and far beyond the family sizes the everyday 'each child shortens your life' worry is usually about."}],"short_label":"Family size & lifespan","myth_framing":"overrated","outcome_severity":"fatal","exposure_pattern":"cumulative","outcome_type":"death","valence":"negative","caveats":"This entry is flagged no_reliable_estimate for several reinforcing reasons, and it is important to separate it from a different question that IS well quantified.\nFirst, \"shortening a mother's life\" is not a binary, countable event. The strongest evidence measures relative mortality over a fixed follow-up (hazard ratios) or continuous biological-age scores, neither of which converts into a lifetime probability without an arbitrary threshold.\nSecond, the shape of the relationship is the whole story, and it is U-shaped, not a staircase. In the largest study (Dior et al. 2013, 40,454 mothers, ~37 years), the lowest mortality is at two-to-four children; risk rises for one child (HR 1.18) and again at five-plus (HR 1.21) and ten-plus (HR 1.49). Biological-aging work (Shirazi et al. 2020) finds its minimum at three-to-four births. So for the family size that prompts this worry most often — a third child — the population signal is protective-to-flat, not harmful.\nThird, the biological-aging proxy evidence is contradictory. Barha et al. (2016) found more children associated with LONGER telomeres; other cohorts find no association. There is no consistent cellular-aging penalty to point to.\nFourth, the modest excess that does appear at high parity is confounded. Large families correlate with lower socioeconomic status, less education, and earlier, differently-spaced childbearing; maternal education strongly modifies the association. These cohorts cannot isolate a causal \"cost of each child\" from the circumstances of women who have many.\nFinally, none of this is about the acute, well-measured risk of dying FROM a pregnancy or delivery — maternal mortality — which is a separate, quantified figure (on the order of 1 in 200 lifetime globally per WHO, far lower in high-income settings) and does rise with each additional pregnancy as a per-event exposure. This entry is about the popular claim that raising a larger family measurably shortens the mother's overall lifespan, and for typical family sizes that claim is not supported.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-07-29","methodology_version":"1.0"},"reviewer":"8d-eval-2026-07-29","last_reviewed":"2026-07-29","reviewed":true,"generated_at":"2026-07-29","image":{"alt":"A shallow U-shaped curve drawn in muted tones on a calm background, its lowest point marked, flat vector illustration."},"canonical_url":"https://likelier.app/parity-maternal-longevity","api_url":"https://likelier.app/api/fears/parity-maternal-longevity.json"},{"slug":"personal-bankruptcy","question":"What are the odds of filing for personal bankruptcy during your lifetime?","quick_answer":"No cited source supports a defensible lifetime figure for this. Annual filing counts are well documented, but filing is not independent year to year — bankruptcy law imposes multi-year waits between discharges — so the yearly rate cannot be compounded into a lifetime probability, and none is published.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Bankruptcy occupies an outsized place in financial anxiety. Popular culture treats it as a catastrophic, rare event reserved for the profligate or the deeply unlucky. Most adults, when asked, guess the lifetime odds are somewhere in the low single digits. The stigma attached to filing amplifies the perception of rarity: people who have filed rarely volunteer the fact, which makes it seem less common than it is. Whether that guess is actually below the true lifetime rate cannot be settled here, because no cited source states a defensible US-adult lifetime filing probability.\n","rough_estimate":"~2-5% lifetime guess for most people","kind":"intuition"},"native":{"display":"~494,000 nonbusiness filings per year (2024, US)","numerator":494201,"denominator":131000000,"unit":"per year","population":"US households"},"sources":[{"url":"https://www.uscourts.gov/data-news/judiciary-news/2026/02/04/bankruptcy-filings-rise-11-percent","title":"Bankruptcy Filings Rise 11 Percent","publisher":"United States Courts","source_type":"govt_report","statistic":"574,314 total bankruptcy filings in the 12 months ending December 2025; 549,577 were nonbusiness filings. In the prior year (ending December 2024), nonbusiness filings totaled 494,201.","excerpt":"\"Annual bankruptcy filings totaled 574,314 in the year ending December 31, 2025, compared with 517,308 cases filed in the year ending December 31, 2024. Non-business filings rose 11.2 percent to 549,577, compared with 494,201 in December 2024.\"\n","source_date":"2026-02-04","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413181449/https://www.uscourts.gov/data-news/judiciary-news/2026/02/04/bankruptcy-filings-rise-11-percent","calculation_notes":"Primary source for the annual nonbusiness filing count. 494,201 nonbusiness filings in the year ending December 2024 divided by ~131 million US households gives an annual household filing rate of approximately 0.38%. This is a well-documented annual rate; it is NOT compounded into a lifetime probability here, because filing is not independent year-to-year and no cited source states a defensible US-adult lifetime cumulative rate (see the entry-level no_reliable_estimate note).\n","independence_note":"US Courts administrative data is the official record of all bankruptcy petitions filed in federal courts. It is the primary upstream data source that academic studies and the American Bankruptcy Institute rely on.\n"},{"url":"https://www.debt.org/bankruptcy/statistics/","title":"Bankruptcy Statistics [Updated For 2025]","publisher":"Debt.org","source_type":"reputable_reference","statistic":"Of surveyed filers, 78% cited a decline in income and 65% cited medical issues as reasons for bankruptcy; in 2005, one out of every 55 households filed for bankruptcy.","excerpt":"\"78% cited a decline in income and 65% cited medical issues as bankruptcy reasons. In 2005, bankruptcy filings hit an all-time high with more than two million cases filed — one out of every 55 households filed for bankruptcy.\"\n","source_date":"2025-10-01","source_accessed":"2026-04-12","archive_url":"https://web.archive.org/web/20260413181529/https://www.debt.org/bankruptcy/statistics/","calculation_notes":"Provides the reason-for-filing breakdown (78% income decline, 65% medical issues) and the 2005 historical peak of over 2 million filings (1 in 55 households), which was driven by a rush to file before the Bankruptcy Abuse Prevention and Consumer Protection Act (BAPCPA) took effect. Post-BAPCPA, annual filings dropped sharply and have never returned to those levels. NOTE: this page does NOT state any US-adult lifetime cumulative filing probability (no \"1 in 10 lifetime\" figure appears on it); it reports only single-year and survey-based statistics. It is cited here for the reason-for-filing and historical-cycle context, not for a lifetime probability.\n","independence_note":"Debt.org aggregates data from US Courts filings and a consumer-bankruptcy survey. The figures used here (medical/income reasons, 2005 peak) are single-year and survey findings, not lifetime probabilities.\n"},{"url":"https://www.epiqglobal.com/en-us/resource-center/news/commercial-chapter-11-filings-increase-20-percent-in-calendar-year-2024","title":"Commercial Chapter 11 Filings Increase 20 Percent in Calendar Year 2024","publisher":"American Bankruptcy Institute / Epiq AACER","source_type":"reputable_reference","statistic":"Overall consumer bankruptcy filings in calendar year 2024 were 478,749 (up 14% from 419,555 in 2023); Chapter 7 totaled 288,968 (up 19%) and Chapter 13 totaled 188,934 (up 7%).","excerpt":"\"Overall consumer filing totals for calendar year 2024 were 478,749, representing a 14 percent increase from the 419,555 consumer filings the previous year. Consumer chapter 7 filings increased 19 percent to 288,968 in CY 2024 from the previous year's total of 242,919. The 188,934 consumer chapter 13 bankruptcy filings during calendar year 2024 registered a 7 percent increase over 2023's total of 175,977.\"\n","source_date":"2025-01-03","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260114094132/https://www.epiqglobal.com/en-us/resource-center/news/commercial-chapter-11-filings-increase-20-percent-in-calendar-year-2024","calculation_notes":"ABI and Epiq AACER jointly report consumer filing totals sourced from the Administrative Office of the US Courts data. The CY2024 consumer total of 478,749 (478,749 ≈ 288,968 Chapter 7 + 188,934 Chapter 13) is a calendar-year, consumer-only figure that cross-checks closely with the US Courts primary source's 494,201 nonbusiness count for the year ending December 2024 (the small gap reflects nonbusiness vs. consumer-defined filing categories). The two-sided Chapter 7/Chapter 13 split confirms the composition behind the headline rate.\n","independence_note":"ABI/Epiq aggregate the same AO filing data used by the US Courts source, so not a fully independent count — but Epiq AACER is the leading provider of US bankruptcy filing data and publishes the consumer Chapter 7/13 breakdown that the AO headline release does not itemize.\n"}],"comparison_anchors":[],"short_label":"Personal bankruptcy","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"financial","valence":"negative","caveats":"No cited source states a defensible US-adult lifetime cumulative filing probability, so this entry publishes no lifetime headline. Annual filing rates have ranged from roughly 0.3% of households (recent years) to over 1.5% (2005 pre-BAPCPA rush), and compounding a single year over a lifetime overstates the true rate because filing is not independent year-to-year. Individual risk varies enormously with income, health-insurance status, homeownership, and state exemption laws. Medical debt is cited as a contributing factor in roughly two-thirds of surveyed filings, though disentangling it from income loss is difficult. The count is of filings, not discharges; a small fraction of petitions are dismissed before discharge. Chapter 7 and Chapter 13 filings are pooled here.\n","quality_score":{"d1":5,"d2":4,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"grounding-cert-fix-2026-07","last_reviewed":"2026-07-11","reviewed":true,"generated_at":"2026-04-12","image":{"alt":"A single empty folder on a bare desk, muted grey and tan tones, flat vector illustration."},"canonical_url":"https://likelier.app/personal-bankruptcy","api_url":"https://likelier.app/api/fears/personal-bankruptcy.json"},{"slug":"pfas-tap-water","question":"What are the odds that PFAS in your tap water will make you sick?","quick_answer":"No reliable figure exists here. No single attributable-risk number for illness from PFAS in tap water can be given: exposure varies by orders of magnitude between water systems, and the dose-response relationship at low concentrations remains contested. The fear outruns any measurable personal risk estimate.\n","category":"health","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"PFAS — per- and polyfluoroalkyl substances, the \"forever chemicals\" — became a mainstream health anxiety after the EPA finalized enforceable maximum contaminant levels in April 2024. Media coverage consistently frames PFAS as ubiquitous, indestructible, and linked to cancer, thyroid disease, and immune suppression. Gallup and AP-NORC polling in 2023-2024 found that roughly half of US adults reported concern about contaminants in their tap water, with PFAS frequently cited by name. The intuition for many consumers is that if these chemicals are \"everywhere\" and \"forever,\" exposure must translate into meaningful personal health risk.\n","rough_estimate":"52.4% of US adults report being afraid or very afraid of pollution of drinking water (Chapman Survey 2024)","kind":"survey","survey_source":{"title":"Chapman University Survey of American Fears, Wave 10 — Complete List of Fears 2024","publisher":"Chapman University","url":"https://www.chapman.edu/wilkinson/research-centers/babbie-center/_files/2024-csaf-fears-high-to-low.pdf","year":2024}},"sources":[{"url":"https://www.epa.gov/sdwa/and-polyfluoroalkyl-substances-pfas","title":"Per- and Polyfluoroalkyl Substances (PFAS) — Final PFAS National Primary Drinking Water Regulation","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"MCLs set at 4 ppt for PFOA and 4 ppt for PFOS; hazard index for PFHxS, PFNA, HFPO-DA, and PFBS","excerpt":"\"EPA is setting enforceable Maximum Contaminant Levels (MCLs) for five individual PFAS and for mixtures of certain PFAS in drinking water. The final rule sets MCLs of 4.0 parts per trillion for PFOA and 4.0 parts per trillion for PFOS.\"\n","source_date":"2024-04-10","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260412120331/https://www.epa.gov/sdwa/and-polyfluoroalkyl-substances-pfas","calculation_notes":"The EPA's final PFAS MCL rule (89 FR 32532) establishes the first legally enforceable federal limits for six PFAS compounds in public drinking water. The 4 ppt MCLs for PFOA and PFOS are based on updated health advisories reflecting cancer (kidney, testicular), developmental, immune, and thyroid endpoints. Public water systems must comply by 2029. The rule applies to ~66,000 community water systems but does not cover private wells (serving ~43 million Americans). No single attributable-risk figure for illness from PFAS-contaminated tap water exists because exposure varies by orders of magnitude across geographies and the dose-response relationship at low concentrations remains contested.\n"},{"url":"https://www.atsdr.cdc.gov/toxprofiles/tp200.pdf","title":"Toxicological Profile for Perfluoroalkyls","publisher":"Agency for Toxic Substances and Disease Registry (ATSDR), CDC","source_type":"govt_report","statistic":"Epidemiological associations reported for kidney cancer, testicular cancer, thyroid disease, ulcerative colitis, pregnancy-induced hypertension, and decreased vaccine antibody response at elevated PFAS serum levels","excerpt":"\"Epidemiological studies of highly exposed populations have reported associations between PFOA exposure and kidney cancer, testicular cancer, thyroid disease, ulcerative colitis, and pregnancy-induced hypertension. PFOS exposure has been associated with decreased antibody response to vaccines.\"\n","source_date":"2021-05-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260403093323/https://www.atsdr.cdc.gov/toxprofiles/tp200.pdf","calculation_notes":"The ATSDR toxicological profile synthesizes evidence from occupational cohorts (3M/Decatur, DuPont/Washington Works) and community studies (C8 Health Project, Veneto, Ronneby). The associations are strongest in populations with serum PFOA levels >20 ng/mL — roughly 5-10x the current US general-population median of ~2 ng/mL (NHANES 2017-2018). Translating these associations into a population-attributable risk for the general US adult drinking compliant municipal water is not feasible from available data — the dose-response curve at low serum levels is uncertain, and confounders (occupational co-exposures, socioeconomic factors) have not been fully resolved.\n"},{"url":"https://nap.nationalacademies.org/catalog/26156/guidance-on-pfas-exposure-testing-and-clinical-follow-up","title":"Guidance on PFAS Exposure, Testing, and Clinical Follow-Up","publisher":"National Academies of Sciences, Engineering, and Medicine","source_type":"peer_reviewed","statistic":"Decreased antibody response and kidney cancer both classified as having 'sufficient' evidence of association with PFAS; testicular cancer, thyroid disease/dysfunction, and liver enzyme alterations classified as having only 'limited or suggestive' evidence","excerpt":"\"There is sufficient evidence of an association between PFAS exposure and the following health outcomes: decreased antibody response (in adults and children), dyslipidemia (in adults and children), decreased infant and fetal growth, and increased risk of kidney cancer (in adults). There is limited or suggestive evidence of an association between PFAS exposure and increased risk of breast cancer (in adults), liver enzyme alterations (in adults and children), increased risk of pregnancy-induced hypertension, increased risk of testicular cancer (in adults), increased risk of thyroid disease and dysfunction (in adults), and increased risk of ulcerative colitis (in adults).\"\n","source_date":"2022-07-28","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251113003947/https://nap.nationalacademies.org/catalog/26156/guidance-on-pfas-exposure-testing-and-clinical-follow-up","calculation_notes":"The NAS committee stratified evidence into tiers (sufficient, limited or suggestive, inadequate). Decreased antibody response, dyslipidemia, decreased infant and fetal growth, and kidney cancer all reached \"sufficient\"; testicular cancer, thyroid disease/dysfunction, and liver enzyme alterations sit in the lower \"limited or suggestive\" tier. The committee recommended clinical follow-up for individuals with serum PFAS >2 ng/mL for PFOS or >2 ng/mL for PFOA — a threshold near the general-population median, meaning roughly half of US adults would qualify for monitoring. This breadth of the recommendation reflects precautionary framing rather than quantified risk: the report does not estimate attributable disease burden for the general population. The committee explicitly noted that individual clinical decisions should account for total exposure history, which varies enormously by geography and water source.\n"},{"url":"https://www.atsdr.cdc.gov/pfas/hcp/clinical-overview/human-exposure.html","title":"Human Exposure: PFAS Information for Clinicians — NHANES Biomonitoring Trends","publisher":"Agency for Toxic Substances and Disease Registry (ATSDR), CDC","source_type":"govt_report","statistic":"NHANES reported geometric mean blood levels of 4.25 ng/mL for PFOS and 1.42 ng/mL for PFOA in 2017-2018, down from 1999-2000 baselines of ~30.4 ng/mL (PFOS) and ~5.21 ng/mL (PFOA) — a substantial decline driven by manufacturer phase-outs","excerpt":"\"The data show declining levels of three prevalent PFAS (PFOA, PFOS, and PFHxS), in part because the U.S. Environmental Protection Agency (EPA) enlisted major manufacturers to phase out production and reduce facility emissions of PFOA, PFOS, PFHxS, and PFNA. In 2017–2018, NHANES reported geometric mean blood levels of PFOS: 4.25 ng/mL, with 95% of the general population ≤14.6 ng/mL, PFOA: 1.42 ng/mL, with 95% of the general population ≤3.77 ng/mL.\"\n","source_date":"2024-11-12","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260628195841/https://www.atsdr.cdc.gov/pfas/hcp/clinical-overview/human-exposure.html","calculation_notes":"NHANES biomonitoring shows that general-population PFAS blood levels have dropped 70-85% since 1999 as US manufacturers phased out PFOS (by 2002) and PFOA (by 2015). The current general-population median is well below the serum levels at which epidemiological associations were observed in the C8 and occupational cohorts. However, communities near military bases (AFFF firefighting foam), industrial discharge sites, and waste-water treatment plants still show serum levels 5-50x the national median. This geographic concentration is the core reason a single national attributable-risk number is misleading.\n"},{"url":"https://www.epa.gov/sdwa/proposed-pfoa-and-pfos-compliance-extension-rule","title":"Proposed PFOA and PFOS Compliance Extension Rule","publisher":"US Environmental Protection Agency","source_type":"govt_report","statistic":"EPA proposed (May 18, 2026) to extend the PFOA/PFOS drinking-water compliance deadline from 2029 to 2031, and separately to rescind the standards for PFHxS, PFNA, HFPO-DA, and the Hazard Index mixture","excerpt":"\"On May 18, 2026, the EPA announced a proposed rule to uphold the federal drinking water Maximum Contaminant Levels (MCLs) for perfluorooctanoic acid (PFOA) and perfluorooctane sulfonic acid (PFOS) while strengthening practical implementation by providing an option for drinking water systems to request two additional years - to 2031 - to comply with the enforceable limits.\"\n","source_date":"2026-05-18","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260708200841/https://www.epa.gov/sdwa/proposed-pfoa-and-pfos-compliance-extension-rule","calculation_notes":"In May 2026 the EPA published two proposed rules that scale back the 2024 final regulation. The compliance-extension rule keeps the 4 ppt MCLs for PFOA and PFOS but lets systems apply to delay compliance to 2031 (from 2029). A companion proposed-rescission rule would withdraw the standards and regulatory determinations for PFHxS, PFNA, HFPO-DA, and the Hazard Index mixture of those three plus PFBS, on the stated procedural ground that the 2024 rule combined the regulatory determination and the standard rather than proceeding sequentially. Both are proposals, not final rules, as of the access date — so the hazard framing and exposed- population figures attached to the 2024 rule remain subject to regulatory change. This does not yield any quantified attributable-risk figure and so does not alter the entry's no_reliable_estimate basis; it is recorded as a regulatory-uncertainty caveat.\n"}],"comparison_anchors":[{"label":"Pesticide residue harm (lifetime, US adult)","lifetime_us_adult":0.000001},{"label":"Lung cancer death (lifetime, US adult)","lifetime_us_adult":0.0535}],"regional_breakdown":[{"region":"Municipal water, post-2029 MCL compliance","probability":0,"notes":"Once public water systems meet the 4 ppt MCLs for PFOA/PFOS (compliance deadline 2029), exposure via municipal tap water drops to near-background levels. Attributable health risk from tap water specifically approaches zero, though dietary, occupational, and legacy body burden remain.\n"},{"region":"Municipal water, pre-2029 or non-compliant","probability":0,"notes":"Many systems already comply; EPA estimates ~6-10% of systems currently exceed the new MCLs. Quantified attributable risk for consumers of non-compliant water is not established — the dose-response at these concentrations is uncertain.\n"},{"region":"Near military base or industrial site","probability":0,"notes":"Communities near AFFF-contaminated military bases (e.g., Pease AFB, Peterson SFB) or industrial sites (e.g., Parkersburg WV, Fayetteville NC) have documented serum PFAS levels 5-50x the national median. These are the populations where epidemiological associations (kidney cancer, thyroid disease) were originally observed. Risk is genuinely elevated but not precisely quantified at the individual level.\n"},{"region":"Private well, untested","probability":0,"notes":"Approximately 43 million Americans rely on private wells not subject to SDWA regulation. PFAS testing is voluntary and uncommon. Proximity to contamination sources creates highly variable exposure with no systematic monitoring.\n"}],"personal_factor_multipliers":[{"factor":"Municipal water meeting 2024 MCLs","multiplier":0.1,"notes":"Tap-water PFAS exposure drops to near-background when supply complies with 4 ppt MCLs. Remaining exposure is primarily dietary and consumer-product-based.\n"},{"factor":"Private well near known contamination source","multiplier":10,"notes":"Private wells within a few miles of AFFF-contaminated military bases or PFAS manufacturing sites have shown concentrations hundreds to thousands of times above the new MCLs.\n"},{"factor":"Occupational (firefighter using AFFF, PFAS manufacturing)","multiplier":20,"notes":"Occupational cohorts in PFAS manufacturing and AFFF-using firefighters show the highest documented serum levels and the strongest epidemiological associations.\n"},{"factor":"Activated-carbon or reverse-osmosis home filtration","multiplier":0.05,"notes":"NSF-certified activated-carbon and RO systems remove >90% of PFOA/PFOS from tap water. Effective mitigation for households on non-compliant supplies or untested private wells.\n"}],"short_label":"PFAS in water","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry uses no_reliable_estimate because no published study quantifies a population-attributable risk of illness from PFAS in US tap water at concentrations typical of compliant municipal systems. The epidemiological associations (kidney cancer, thyroid disease, immune suppression) come predominantly from highly exposed occupational and community cohorts with serum PFAS levels 5-50x the current general-population median. Extrapolating those associations to the low end of the dose-response curve is an active area of research without consensus. The regional breakdown uses 0.0 probabilities as placeholders — not claims that the risk is zero — because no defensible point estimate exists for any subgroup. The entry does not address PFAS exposure from food packaging, nonstick cookware, or other non-water sources, which may collectively exceed tap-water exposure for most Americans. Hazard-framing update (2026-06-14): the EPA states the 2024 rule \"will prevent PFAS exposure in drinking water for approximately 100 million people\" and \"prevent thousands of deaths, and reduce tens of thousands of serious PFAS-attributable illnesses\" — population-level projected benefits, not an individual attributable risk, so they do not change the no_reliable_estimate basis. An Environmental Working Group analysis of EPA UCMR 5 monitoring data put the number of Americans served by water with detectable PFAS at roughly 158 million (March 2025), rising in later UCMR 5 releases. Regulatory-uncertainty caveat: in May 2026 the EPA proposed to extend the PFOA/PFOS compliance deadline to 2031 and to rescind the standards for PFHxS, PFNA, HFPO-DA, and their Hazard Index mixture; these proposals were not final as of the access date, so the regulatory basis for the hazard framing above may shift.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A glass of water on a muted surface with subtle molecular structures in the background, flat vector illustration."},"canonical_url":"https://likelier.app/pfas-tap-water","api_url":"https://likelier.app/api/fears/pfas-tap-water.json"},{"slug":"phone-while-cycling-injury","question":"What are the odds of crashing while cycling distracted by a phone?","quick_answer":"There is no defensible number. The direction is clear — using a phone while cycling raises crash risk — but no US data source records a cyclist's own phone use at the moment of a crash, so a per-cyclist lifetime probability cannot be estimated.\n","category":"transport","no_reliable_estimate":true,"perceived":{"description":"Most regular cyclists who have ever fumbled a phone in traffic recognise the feeling immediately: one hand off the bars, eyes off the road, balance drifting toward whichever side the device is on. The intuition is that phone use while cycling sharply raises the odds of a crash — and qualitatively, the cycling-distraction literature agrees. The problem is that no US data source converts that qualitative agreement into a defensible probability. NHTSA's Fatality Analysis Reporting System codes distracted *drivers* who hit cyclists; it does not code the cyclist's own phone use at the moment of crash. The CDC's emergency department surveillance (NEISS-AIP) explicitly notes that its narrative records \"do not provide detailed or consistent information about… injury circumstances\". The Dutch and broader European literature has the best behavioural and prevalence data, but converting it to a US adult lifetime probability requires assumptions about exposure, infrastructure, and cycling frequency that the underlying studies do not support.\n","rough_estimate":"no defensible US denominator — qualitative direction (increased risk) is well supported, but the per-cyclist lifetime probability is not estimable","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/20069479/","title":"Mobile phone use while cycling: incidence and effects on behaviour and safety","publisher":"Ergonomics (Taylor & Francis), via PubMed","source_type":"peer_reviewed","statistic":"In Groningen, NL, 2.2% of observed cyclists were talking on a phone and 0.6% were texting or dialling; only 0.5% of accident-involved cyclists reported phone use at the time of the crash","excerpt":"\"In Groningen 2.2% of cyclists were observed talking on their phone and 0.6% were text messaging or entering a phone number. In study 2, accident-involved cyclists responded to a questionnaire. Only 0.5% stated that they were using their phone at the time of the accident… Telephoning coincided with reduced speed, reduced peripheral vision performance and increased risk and mental effort ratings. Text messaging had the largest negative impact on cycling performance. Higher mental workload and lower speed may account for the relatively low number of people calling involved in accidents.\"\n","source_date":"2010-01-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260525095915/https://pubmed.ncbi.nlm.nih.gov/20069479/","calculation_notes":"de Waard et al. 2010 is the foundational observational study on cyclist mobile phone prevalence in the Netherlands. It is cited here for two reasons: (1) it establishes that the behavioural effect of phone use on cycling performance is real and measurable (peripheral vision narrows, speed drops, perceived risk rises), and (2) the 0.5% self-report among crash-involved cyclists is not a crash-risk rate — it is a snapshot of what crashed cyclists recall doing, on a non-representative sample, in the Netherlands. Even if taken at face value, it cannot be converted to a US per-crash or per-cyclist-year probability because the denominator (total cycling-phone-use exposure-hours) is not measured. No conversion to native or normalized US-adult lifetime probability is performed.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/22385735/","title":"The use and risk of portable electronic devices while cycling among different age groups","publisher":"Journal of Safety Research (Elsevier), via PubMed","source_type":"peer_reviewed","statistic":"Self-reported crash odds were ~1.6× higher for Dutch teen cyclists and ~1.8× higher for young adult cyclists who used portable electronic devices on every trip vs cyclists who never used them; no significant effect detected for middle-aged or older cyclists","excerpt":"\"The odds of being involved in a bicycle crash are higher for teen cyclists (factor 1.6) and young adult cyclists (factor 1.8) who use electronic devices on every trip compared to cyclists who never use these devices. For middle-aged and older adult cyclists, the use of portable electronic devices was not a significant predictor of bicycle crashes, but frequency of cycling in demanding traffic situations was.\"\n","source_date":"2012-02-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250204053157/https://pubmed.ncbi.nlm.nih.gov/22385735/","calculation_notes":"Goldenbeld et al. 2012 is the strongest peer-reviewed estimate of crash-risk elevation from cyclist device use. The reported odds ratios (1.6 and 1.8) are self-reported, age-stratified, and Dutch — they describe relative risk between \"every trip\" and \"never\" device users in the same age cohort, not absolute probability per trip or per year. The effect disappears in older cohorts, suggesting either behavioural compensation or selection. Multiplying any baseline US cyclist crash probability by ~1.7 would be inappropriate because (a) the underlying US baseline for distraction-attributable crashes does not exist, (b) Dutch cyclists ride on protected infrastructure that US cyclists do not, and (c) \"every trip\" device use is a behavioural extreme, not the modal phone-using cyclist. The OR is cited as qualitative direction evidence (distraction increases crashes for younger cyclists) without a quantitative US transfer.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4785823/","title":"Distracted biking: an observational study","publisher":"Journal of Trauma Nursing (Wolters Kluwer), via PubMed Central","source_type":"peer_reviewed","statistic":"Of 1,974 cyclists observed at 4 Boston intersections in summer 2016, 31.2% were distracted; 17.7% by auditory devices (headphones/earbuds) and 13.5% by visual/tactile devices (phone or object in hand)","excerpt":"\"Of the 1,974 bicyclists observed, 615 (31.2%) were distracted. Auditory distractions, predominantly headphones or earbuds, accounted for 17.7%, and visual/tactile distractions, predominantly a phone or other object in the hand, accounted for 13.5%. Reduced attention can place bicyclists and other road users at greater risk of sustaining an injury.\"\n","source_date":"2016-03-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260209141333/https://pmc.ncbi.nlm.nih.gov/articles/PMC4785823/","calculation_notes":"Wolfe et al. 2016 (observations summer 2015, published 2016) is the best-available US observational prevalence study for cyclist distraction. It establishes that distraction is common (≈1 in 3 observed cyclists at Boston intersections) but does not link observed distraction to crash outcomes — there is no denominator-matched crash count for the same population. The study is cited as the US exposure-prevalence anchor that demonstrates the behaviour is widespread, while explicitly not supplying a crash-risk number. No native or normalized probability is derived.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/70/wr/mm7019a1.htm","title":"Emergency Department Visits for Bicycle-Related Traumatic Brain Injuries Among Children and Adults — United States, 2009–2018","publisher":"CDC Morbidity and Mortality Weekly Report (MMWR)","source_type":"govt_report","statistic":"An estimated 596,972 ED visits for bicycle-related TBIs occurred in the US during 2009–2018; the surveillance system does not record cyclist phone use, helmet use, or injury circumstances consistently","excerpt":"\"An estimated 596,972 ED visits for bicycle-related TBIs occurred in the United States during this study period (2009–2018)… NEISS-AIP narrative descriptions do not provide detailed or consistent information about helmet use, injury circumstances (e.g., whether the injury occurred on a road or bicycle path), or about a person's level of exposure.\"\n","source_date":"2021-05-14","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260511084004/https://www.cdc.gov/mmwr/volumes/70/wr/mm7019a1.htm","calculation_notes":"Sarmiento et al. 2021 (CDC MMWR) is the primary US bicycle-injury surveillance source. It documents that ED visits for bicycle-related TBI are common (≈60,000/year averaged over 2009–2018) but explicitly states that the surveillance instrument does not capture the circumstances that would let an analyst identify which crashes involved cyclist phone use. This source is cited as documentation of the data gap that prevents a US cyclist-phone-distraction injury rate from being estimated — the numerator data does not exist, regardless of what denominator one might assume.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5685949/","title":"Distracted Walking, Bicycling, and Driving: Systematic Review and Meta-Analysis of Mobile Technology and Youth Crash Risk","publisher":"Child Development (Wiley), via PubMed Central","source_type":"peer_reviewed","statistic":"Of 41 peer-reviewed studies in a systematic review of mobile technology and youth road crash risk, only one cycling study met inclusion criteria and was excluded from the meta-analysis, so no aggregate effect size for phone-while-cycling crash risk could be computed","excerpt":"\"A single study on bicycling met inclusion criteria, but was omitted from the meta-analysis.\"\n","source_date":"2017-11-01","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20251017045210/https://pmc.ncbi.nlm.nih.gov/articles/PMC5685949/","calculation_notes":"Stavrinos et al. 2017 is the closest available systematic review on mobile technology and crash risk that includes cycling as a category. The reviewers identified only one eligible cycling study (Kircher et al. 2015, an experimental study of behavioural compensation, not crash incidence), which they then excluded from quantitative synthesis. This is the strongest available evidence that no defensible pooled odds ratio exists for phone-while-cycling crash risk in the published literature. The source is cited as direct documentation of the evidence-base gap that justifies no_reliable_estimate=true.\n"}],"comparison_anchors":[{"label":"Cyclist killed by a motor vehicle (lifetime, US regular cyclist)","lifetime_us_adult":0.0051},{"label":"Cycling helmetless serious head injury (lifetime, US frequent unhelmeted urban cyclist)","lifetime_us_adult":0.125},{"label":"Phone-distracted walking injury (lifetime, US adult)","lifetime_us_adult":0.0025}],"personal_factor_multipliers":[{"factor":"occasional recreational cyclist on protected paths, phone in pocket, ear buds out","multiplier":0.3,"notes":"Lowest exposure profile — low cycling time, low device use, separated infrastructure. Multiplier is relative to a same-cohort cyclist with no phone use; absolute lifetime probability is not estimated for this entry."},{"factor":"regular urban commuter, occasional handheld phone glances at lights, single ear bud","multiplier":1,"notes":"Baseline reference profile in this entry — the relative risk floor against which the other multipliers are scaled. No absolute lifetime probability is implied."},{"factor":"every-trip handheld device user, teens or young adults (16–34)","multiplier":1.8,"notes":"Approximates the Goldenbeld et al. 2012 self-reported odds ratio for Dutch young-adult cyclists using devices on every trip vs never. Effect attenuates in middle-aged cyclists in the same study. Translation to US adults is not validated."},{"factor":"active texting or typing while moving, any age","multiplier":3,"notes":"de Waard et al. 2010 found texting had the largest negative effect on cycling performance — speed, lateral control, and peripheral vision all degraded substantially. The 3x is a qualitative direction estimate, not a measured crash odds ratio."},{"factor":"active texting on an arterial road or in unprotected mixed traffic","multiplier":6,"notes":"Compounds texting-degraded control with high-consequence environment (NHTSA: 65% of US cyclist fatalities occur on principal/minor arterials). Qualitative upper bound, not a measured number."}],"short_label":"Phone-distracted cycling crash","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry uses `no_reliable_estimate: true` because the US literature does not contain the data required to compute a defensible lifetime probability of injury from cycling while distracted by a phone. The structural problem has three layers. First, the numerator: NHTSA's FARS records distracted *drivers* who strike cyclists but does not code the cyclist's own phone use, and the CDC's NEISS-AIP emergency department surveillance explicitly states that injury circumstances are not captured consistently. Second, the denominator: the US does not measure cyclist phone-use exposure-hours, and the closest available figure (Wolfe et al. 2016, 31% of Boston cyclists distracted, of which 13.5% visually/tactilely) is a single-city intersection snapshot, not a national exposure metric. Third, transfer bias: the best peer-reviewed crash-risk evidence is Dutch (Goldenbeld et al. 2012 odds ratio ~1.6–1.8 for every-trip vs never device use in young cyclists; effect absent in older cyclists), but Dutch cyclists ride on protected infrastructure at population trip-shares twenty to thirty times higher than US cyclists, and the Stavrinos et al. 2017 systematic review found only one cycling study globally that met inclusion criteria for mobile-technology crash-risk analysis — and excluded it from the meta- analysis. Even after the Netherlands banned handheld phone use while cycling in July 2019, the academic evaluation concluded \"there are no precise figures for how many accidents are caused by phone use while cycling\". What a US adult should take from this entry is qualitative: cyclist phone use measurably degrades peripheral vision, speed control, and reaction time, and is associated with elevated self-reported crash odds in young cyclists in the Netherlands; texting is worse than calling; and the elevated risk compounds with the road environment, especially on arterial roads where US cyclist fatalities are concentrated. The quantitative claim — what fraction of US adult cyclists will be injured in a phone-distraction crash over a lifetime — is not estimable from current evidence, and inventing a number to make the entry listable would misrepresent the state of the science.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"An empty bike lane viewed from above, with a faint phone-shaped outline drawn on the asphalt, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/phone-while-cycling-injury","api_url":"https://likelier.app/api/fears/phone-while-cycling-injury.json"},{"slug":"pimple-danger-triangle","question":"What are the odds of a serious infection from squeezing a pimple?","quick_answer":"No reliable number exists. The only published incidence figure for cavernous sinus thrombosis appears to conflate it with a broader clot disorder, and its severe outcomes are conditional on already having the infection — not on squeezing a pimple — so no per-squeeze probability can be derived.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The \"danger triangle of the face\" has become a staple of health-anxiety content on TikTok and Instagram. The premise is anatomically real: valveless veins connect the mid-face to the cavernous sinus at the base of the skull, creating a theoretical route for bacteria to travel from a squeezed pimple to the brain. The internet compresses this into \"popping a pimple can kill you.\" A secondary, quieter fear is that squeezing any pimple — triangle or not — risks turning a minor blemish into cellulitis, an abscess, or a scar requiring medical treatment. Both fears are amplified by the universality of the behaviour: surveys suggest the majority of acne sufferers squeeze or pick at lesions, so the audience for the warning is effectively everyone.\n","kind":"intuition"},"sources":[{"url":"https://www.ncbi.nlm.nih.gov/books/NBK448177/","title":"Cavernous Sinus Thrombosis","publisher":"StatPearls / National Library of Medicine","source_type":"peer_reviewed","statistic":"CST incidence ~0.2–1.6 per 100,000/year; mortality 8–13% in antibiotic era; S. aureus in ~⅔ of cases","excerpt":"\"The estimated annual incidence of cavernous sinus thrombosis may be approximately 0.2 to 1.6 per 100,000 per year. … Mortality rates were high in the past, approaching 80% to 100%. … The mortality rate [has fallen] to approximately 8% to 13%. … Blindness can result in 8% to 15% of cases. … Approximately 50% [have] persistent cranial nerve deficits. … Staphylococcus aureus is responsible for approximately two-thirds of cases.\"\n","source_date":"2025-06-16","source_accessed":"2026-04-25","archive_url":"http://web.archive.org/web/20260422230442/https://www.ncbi.nlm.nih.gov/books/NBK448177/","calculation_notes":"StatPearls provides the only published incidence range for cavernous sinus thrombosis (0.2–1.6 per 100,000 per year). However, this figure may conflate CST with broader cerebral venous sinus thrombosis (CVST); a 2023 meta-analysis in Stroke (Zhou et al.) found CVST incidence of ~12 per million, and CST comprises only 1–4% of CVST, yielding a derived CST estimate of ~0.1–0.5 per million — roughly 10× lower than the StatPearls headline. There is an internal inconsistency in the StatPearls figures: if CST is 1–4% of all CVST (~12 per million), the derived CST incidence would be 0.12–0.48 per million, not 0.2–1.6 per 100,000 (i.e. 2–16 per million). The StatPearls range appears to conflate CST with broader CVST. This ambiguity is a key reason the entry carries no_reliable_estimate. The morbidity figures (8–15% blindness, 50% cranial nerve deficits) are conditional on developing CST, not on squeezing a pimple.\n"},{"url":"https://www.msdmanuals.com/professional/eye-disorders/orbital-diseases/cavernous-sinus-thrombosis","title":"Cavernous Sinus Thrombosis","publisher":"MSD Manual (Merck) Professional Edition","source_type":"reputable_reference","statistic":"CST is an 'extremely rare complication of common facial infections'; sinusitis >50% of cases; mortality ~10–15%","excerpt":"\"Cavernous sinus thrombosis is an extremely rare complication of common facial infections, most notably sphenoidal or ethmoidal sinusitis (greater than 50%), nasal furuncles, and dental infections. … Mortality in the antibiotic era is approximately 10 to 15%. … Approximately one-third of surviving patients develop serious sequelae (eg, ophthalmoplegia, blindness).\"\n","source_date":"2024-09-01","source_accessed":"2026-04-25","archive_url":"https://web.archive.org/web/20260505061849/https://www.msdmanuals.com/professional/eye-disorders/orbital-diseases/cavernous-sinus-thrombosis","calculation_notes":"MSD Manual's etiological breakdown attributes >50% of septic CST to sinusitis, with nasal furuncles and dental infections accounting for most of the remainder. A nasal furuncle is a deep staphylococcal abscess of the nasal vestibule — clinically distinct from a squeezed superficial pimple. No source quantifies the fraction of CST attributable to manipulation of ordinary acne lesions; it is at most a small subset of the furuncle category.\n","independence_note":"Editorially independent of StatPearls: MSD Manual is a separately authored clinical reference with its own editorial board.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10240192/","title":"Prevalence and risk factors of acne scars in patients with acne vulgaris","publisher":"Skin Research and Technology","source_type":"peer_reviewed","statistic":"Pooled acne-scar prevalence among acne patients: 47% (95% CI 38–56%); squeezing is a recognised risk factor","excerpt":"\"The pooled prevalence of acne scars was 47% (95% CI: 38–56%) in a meta-analysis of 37 studies involving 24,649 acne patients. Significant risk factors included male gender (OR 1.58), positive family history (OR 2.73), and acne severity.\"\n","source_date":"2023-06-07","source_accessed":"2026-04-25","archive_url":"https://web.archive.org/web/20260426205852/https://pmc.ncbi.nlm.nih.gov/articles/PMC10240192/","calculation_notes":"Liu et al. (2023) is the largest meta-analysis of acne scarring prevalence. While the meta-analysis identifies squeezing as a recognised risk factor, it could not compute a pooled odds ratio for squeezing specifically due to insufficient standardised data across studies. A separate Thai cross-sectional study (Yan et al. 2025, N = 225) found an adjusted OR of 2.69 (95% CI 1.08–6.68) for squeezing/picking as an independent scarring risk factor, but this is a single moderate-quality study and not generalisable to a global estimate. The 47% scar prevalence figure contextualises the most common real-world consequence of acne manipulation — scarring, not life-threatening infection.\n","independence_note":"Methodologically independent of the CST sources: this is a dermatology meta-analysis studying acne outcomes, not vascular or neurological complications.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27434444/","title":"Cellulitis: A Review","publisher":"JAMA (Raff & Kroshinsky)","source_type":"peer_reviewed","statistic":"An estimated 14.5 million cellulitis cases occur annually in the United States, accounting for $3.7 billion in ambulatory care costs alone","excerpt":"\"In the United States, an estimated 14.5 million cases annually of cellulitis account for $3.7 billion in ambulatory care costs alone.\"\n","source_date":"2016-07-19","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260629024455/https://pubmed.ncbi.nlm.nih.gov/27434444/","calculation_notes":"Raff and Kroshinsky's JAMA review is the standard reference for US cellulitis burden and supplies the \"14.5 million annual cases\" figure used in the caveats and body prose to illustrate that population-level cellulitis data is not stratified by whether acne manipulation was the entry point. It does not itself estimate a probability for the pimple-squeezing-to-infection pathway.\n","independence_note":"Methodologically and institutionally independent of the CST sources (StatPearls, MSD Manual) and the acne-scarring meta-analysis (Liu et al.): this is a general dermatology/infectious-disease review of cellulitis epidemiology and management.\n"}],"comparison_anchors":[{"label":"Lifetime acne scarring (among acne patients)","lifetime_us_adult":0.47},{"label":"Lifetime sepsis from minor wound (healthy US adult)","lifetime_us_adult":0.0003}],"short_label":"Pimple infection","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"No epidemiological study tracks the specific pathway from squeezing a pimple to serious infection (cellulitis, abscess, or cavernous sinus thrombosis) as a primary endpoint. The existing literature on cavernous sinus thrombosis attributes cases to \"nasal furuncles\" (deep staphylococcal abscesses) and sinusitis, not to manipulation of superficial comedones. The incidence of CST itself is uncertain: StatPearls cites 0.2–1.6 per 100,000/year, but this may conflate CST with broader cerebral venous sinus thrombosis (CVST), whose meta-analytic incidence is ~12 per million; CST is only 1–4% of CVST, yielding a derived CST estimate roughly 10× lower. For skin infections (cellulitis, abscess) from pimple manipulation, no population-level data exists — the 14.5 million annual US cellulitis cases are not stratified by whether acne manipulation was the entry point. The best-documented complication of pimple squeezing is scarring (47% of acne patients develop some scarring, with squeezing as a recognised independent risk factor), but this is a cosmetic outcome, not a life-threatening one. The entry is marked no_reliable_estimate because every path to a probability — CST, cellulitis, or secondary infection — requires assumptions that cannot be grounded in tracked data.\n","quality_score":{"d1":5,"d2":4,"d3":5,"d4":4,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.75,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-04-25","image":{"alt":"A flat vector illustration of a simplified face outline with a faint triangular highlight over the nose and upper lip area, rendered in muted tones."},"canonical_url":"https://likelier.app/pimple-danger-triangle","api_url":"https://likelier.app/api/fears/pimple-danger-triangle.json"},{"slug":"plant-based-meat-ultraprocessed-harm","question":"What are the odds plant-based meat alternatives (Beyond, Impossible) cause health harm because they are ultra-processed?","quick_answer":"There is no reliable number. No study has followed Beyond or Impossible eaters long enough to estimate a lifetime probability of harm; the available short-term trials actually show improved cardiovascular markers, the opposite of what the 'ultra-processed therefore harmful' framing predicts.\n","category":"food","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"Plant-based meat alternatives like Beyond Meat and Impossible Burger sit in a strange cultural position. Marketed as a heart-healthy and climate-friendly swap for ground beef, they are simultaneously attacked from a different direction as ultra-processed \"frankenfoods\" whose long ingredient lists, industrial extrusion, and NOVA category 4 classification are read as a health red flag. The intuited model is roughly: if it has 20 ingredients and is made in a factory, it must be worse for you than the meat it replaces. Surveys have not put a single number on the perceived lifetime risk, but the framing is durable enough that \"ultra-processed = harmful\" has migrated from nutrition journals into casual conversation.\n","rough_estimate":"Commonly framed as comparable to or worse than the conventional meat being replaced, with no specific numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/32780794/","title":"A randomized crossover trial on the effect of plant-based compared with animal-based meat on trimethylamine-N-oxide and cardiovascular disease risk factors in generally healthy adults (SWAP-MEAT)","publisher":"American Journal of Clinical Nutrition (Crimarco et al.)","source_type":"peer_reviewed","statistic":"8-week crossover RCT in 36 generally healthy adults: plant-based meat reduced TMAO from 4.7 to 2.7 μM (P=0.012), LDL-C from 120.7 to 109.9 mg/dL (P=0.002), and body weight by ~1 kg vs animal meat (Stanford Prevention Research Center)","excerpt":"\"Among generally healthy adults, contrasting Plant with Animal intake, while keeping all other dietary components similar, the Plant products improved several cardiovascular disease risk factors, including TMAO; there were no adverse effects on risk factors from the Plant products.\"\n","source_date":"2020-08-11","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260310130037/https://pubmed.ncbi.nlm.nih.gov/32780794/","calculation_notes":"The strongest direct comparative evidence to date. Stanford Prevention Research Center, NIH-funded with a separate Beyond Meat product-donation arrangement disclosed in the paper. Sample size is modest (n=36) and the trial duration is 8 weeks per arm; the design captures short-term cardiometabolic effects, not lifetime disease incidence. The headline direction (plant-based reduces several CVD risk markers vs the matched animal-meat arm) is the opposite of what the \"ultra-processed therefore harmful\" framing would predict. No basis here for a lifetime probability of harm — this is hypothesis-generating short-term biomarker data, not a cohort outcome study.\n","independence_note":"Stanford-led RCT with Beyond Meat product donation but independent analysis and publication. Methodologically distinct from the AHA advisory and the Bohrer nutrient-profiling review below.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/34724806/","title":"2021 Dietary Guidance to Improve Cardiovascular Health: A Scientific Statement From the American Heart Association","publisher":"Circulation (Lichtenstein et al., AHA Scientific Statement)","source_type":"peer_reviewed","statistic":"AHA 2021 dietary guidance lists 10 evidence-based features of a heart-healthy diet; flags plant-based meat alternatives as a category requiring caution because many are ultra-processed and high in sodium and additives, while not classifying them as harmful","excerpt":"\"The availability of plant-based meat alternatives can help diversify protein choices but requires some caution because many are ultra-processed and come with added sugars, sodium, stabilisers, and preservatives.\"\n","source_date":"2021-11-02","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260713012850/https://pubmed.ncbi.nlm.nih.gov/34724806/","calculation_notes":"The most recent authoritative US clinical-nutrition guidance on this category. Published as a formal AHA Scientific Statement in Circulation, peer-reviewed and chaired by Alice Lichtenstein at Tufts. The wording is deliberately calibrated: plant-based meat alternatives are not classified as harmful, but the sodium and additive content is flagged as a reason for moderation rather than full substitution. AHA does not provide a lifetime probability estimate because none exists in the cohort literature. The advisory's framing is \"diversify protein with caution,\" not \"avoid because ultra-processed.\"\n","independence_note":"AHA Scientific Statement writing committee with disclosure of individual conflicts; institutionally independent of the Stanford RCT and the Bohrer review. Synthesizes the broader evidence base including but not limited to SWAP-MEAT.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S2213453019301144","title":"An investigation of the formulation and nutritional composition of modern meat analogue products","publisher":"Food Science and Human Wellness (Bohrer)","source_type":"peer_reviewed","statistic":"Cross-sectional comparison of US-market meat analogue products vs the meat products they imitate: meat analogues had lower saturated fat and cholesterol and higher carbohydrates and dietary fibre; sodium content was the main negative, frequently exceeding the meat being replaced","excerpt":"\"[Paraphrase from abstract — direct fetch blocked by bot protection 2026-05-30; content corroborated by Semantic Scholar abstract and the author's University of Guelph faculty page] Modern meat analogue products classify as ultra-processed foods under NOVA category 4, but their nutrient profiles differ substantively from typical ultra-processed foods: comparable protein and iron to the meat they replace, lower saturated fat and cholesterol, higher dietary fibre and carbohydrate, with sodium content as the primary nutritional concern.\"\n","source_date":"2019-11-29","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20241114225102/https://www.sciencedirect.com/science/article/pii/S2213453019301144","calculation_notes":"Bohrer's review is the most comprehensive cross-sectional nutrient profile available for the US market and the basis for the standard qualifier that plant-based meat is nutritionally distinct from generic ultra-processed snack foods. The paper does not estimate disease risk; it documents that the NOVA-4 ultra-processed classification, while technically correct, does not transfer cleanly from the original ultra-processed cohort literature (which was built on chips, packaged pastries, sweetened drinks, and processed meats) to this product class.\n","independence_note":"University of Guelph single-author nutrient-profiling review, with no industry funding declared; methodologically and institutionally independent of the Stanford SWAP-MEAT trial and the AHA advisory.\n"}],"comparison_anchors":[{"label":"Red meat (daily) → colorectal cancer (lifetime, US adult)","lifetime_us_adult":0.048},{"label":"Processed meat (daily) → colorectal cancer, IARC Group 1 (lifetime baseline + RR 1.18)","lifetime_us_adult":0.048},{"label":"Baseline cardiovascular disease (lifetime, US adult)","lifetime_us_adult":0.4}],"regional_breakdown":[{"region":"Occasional consumer (≤1 serving/week), otherwise varied diet","probability":0.000001,"notes":"The available short-term RCT data shows neutral-to-improved cardiometabolic markers vs the animal meat being replaced. No cohort has detected harm at this exposure level. Point estimate is a structural \"no detected harm\" placeholder, not a measured rate.\n"},{"region":"Daily consumer (5+ servings/week) as primary protein source for years","probability":0.000001,"notes":"Lifetime cohort data does not exist for this exposure pattern because the products are too new. Mechanistically the sodium load is the most plausible source of cumulative cardiovascular risk; the SWAP-MEAT 8-week trial captured this and still showed net cardiometabolic improvement vs the matched animal-meat arm. Any number assigned here would be extrapolation, not measurement.\n"},{"region":"Daily consumer with hypertension or sodium-sensitive kidney disease","probability":0.000001,"notes":"The AHA caution about sodium is most relevant in this subgroup. No study has measured a lifetime outcome differential; clinical guidance is to count plant-based meat toward the daily sodium budget the same way one would count deli meat or canned soup.\n"}],"personal_factor_multipliers":[{"factor":"Already vegetarian by choice (using plant-meat occasionally)","multiplier":0.9,"notes":"Crimarco SWAP-MEAT showed cardiometabolic improvement when plant-meat replaced animal meat. For someone who would otherwise be eating fewer animal products, the swap-in effect is smaller; not a multiplicative harm but a smaller relative benefit.\n"},{"factor":"Heavy daily consumer of ultra-processed foods generally","multiplier":1.2,"notes":"The NOVA-4 cohort literature linking ultra-processed food to mortality (Chen 2024 BMJ, Srour 2019 BMJ) is built primarily on packaged snacks, sweet drinks, and processed meats. A diet dominated by ultra-processed foods is associated with worse outcomes in cohort data; plant-meat products contribute to that category by NOVA classification but the direct harm signal from this product class is absent.\n"},{"factor":"Hypertension or sodium-sensitive (advised <2,300 mg/day)","multiplier":1.3,"notes":"Bohrer 2019 and AHA 2021 both flag sodium as the primary nutritional concern; some products exceed 500 mg/serving. For a sodium-sensitive consumer this is a meaningful contribution to daily intake, similar to deli meat or canned soup. No lifetime cohort data isolates this exposure.\n"},{"factor":"Chronic kidney disease (any stage)","multiplier":1.5,"notes":"Plant-protein sources change phosphorus bioavailability and potassium load relative to animal protein. Direction of effect is mixed in the CKD literature: some plant proteins are favored over animal protein in nephrology guidelines, but high-additive products with phosphate preservatives or potassium salts can be problematic. No specific lifetime probability has been quantified for this subgroup × product class.\n"}],"short_label":"Plant-based meat harm","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers the broad \"ultra-processed therefore harmful\" claim about plant-based meat alternatives as a product class. It does not cover individual product safety incidents, allergen concerns (soy, pea, gluten), or environmental and animal-welfare arguments — all of which are separate questions with separate evidence bases. The honest summary of the cardiometabolic-harm question is: short-term RCT data (Crimarco) suggests cardiometabolic neutral-or-better outcomes than the animal meat being replaced; cross-sectional nutrient profiles (Bohrer) show comparable protein and iron, lower saturated fat, higher fibre, and sodium as the main legitimate concern; AHA 2021 advises moderation rather than avoidance. The one number commonly cited in social media — that plant-meat consumers have some specific elevated disease risk — does not have a cohort study behind it. The product class is under a decade old at meaningful market scale, and the cohort follow-up needed to produce a lifetime probability of harm has not happened yet.\nWhere this entry does not apply: dedicated \"high-protein plant snack\" bars, protein-fortified ultra-processed cereals, and other plant-protein products marketed for general consumption are a different category and should be evaluated on their own nutrient profile, not bundled with the burger patties and ground-meat analogues covered here.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single plant-based burger patty resting on a plain plate, rendered as a flat vector shape in muted warm-brown and off-white tones on a calm background."},"canonical_url":"https://likelier.app/plant-based-meat-ultraprocessed-harm","api_url":"https://likelier.app/api/fears/plant-based-meat-ultraprocessed-harm.json"},{"slug":"plant-milk-bone-health","question":"What are the odds that switching to plant milk (almond, oat, soy) damages your bone health?","quick_answer":"No single number applies. No study isolates plant-milk drinkers to give an adult lifetime fracture risk. The documented harm is confined to young children fed unfortified plant milk as their main calcium source; for adults drinking fortified plant milk, no harm has been detected.\n","category":"food","tags":["food","kids"],"no_reliable_estimate":true,"perceived":{"description":"Cow's milk has been the headline \"bone food\" of US public-health messaging since the original Got Milk campaign, and the cultural assumption follows: swap to almond, oat, or soy milk and your skeleton starves. The fear is most often framed for adults worried about osteoporosis decades later, but the more clinically real version concerns young children switched onto unfortified or homemade plant milks as a primary calcium source. Most adult consumers have no clear numerical estimate of the supposed fracture-risk increase; the intuition is qualitative — \"milk builds bones, this is not milk, therefore this is bad for bones.\" Pediatric case reports of rickets in toddlers on rice or almond milk circulate in the pediatric literature and occasionally surface in the consumer press, fuelling a \"real risk\" perception that is genuine for one specific population (young children on unfortified plant milk as primary source) and largely absent for another (adults consuming fortified plant milk).\n","rough_estimate":"Commonly framed as a substantial bone-density or fracture-risk increase, with no specific numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/11335767/","title":"Severe nutritional deficiencies in toddlers resulting from health food milk alternatives","publisher":"Pediatrics (Carvalho, Kenney, Carrington, Hall)","source_type":"peer_reviewed","statistic":"Case series documenting kwashiorkor (22-month-old on rice beverage as milk replacement, serum albumin 1.0 g/dL) and rickets (17-month-old on unfortified soy beverage with growth arrest, frontal bossing, genu varus) attributable to unfortified plant-milk substitution","excerpt":"\"While health food beverages may be a part of a healthful diet for adults, these case reports illustrate the potential dangers of replacing cow milk or formula with these alternative beverages in young children. [...] young children consuming these beverages as their sole source of milk are at risk for severe nutritional deficiencies.\"\n","source_date":"2001-04-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260531015941/https://pubmed.ncbi.nlm.nih.gov/11335767/","calculation_notes":"Foundational pediatric case-series paper that established the clinical reality of unfortified-plant-milk nutritional deficiencies in toddlers. Two cases is not a population rate, but Carvalho's paper is the most cited entry point into a now-broader case literature documenting rickets, hypocalcemia, and growth failure in children switched from cow milk or formula onto unfortified almond, rice, or homemade plant beverages between roughly 1 and 3 years of age. The mechanism is straightforward: a typical cup of cow milk delivers ~300 mg calcium and ~8 g protein; unfortified almond milk delivers ~1 g protein and minimal calcium. As a primary calcium source over months, the gap is biologically sufficient to produce rickets.\n","independence_note":"Single-center US case series at a children's hospital, no industry funding. Methodologically and institutionally independent of the Singhal nutrient-comparison review and the Healthy Eating Research consensus statement.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27540708/","title":"A Comparison of the Nutritional Value of Cow's Milk and Nondairy Beverages","publisher":"Journal of Pediatric Gastroenterology and Nutrition (Singhal, Baker, Baker)","source_type":"peer_reviewed","statistic":"Cross-sectional comparison of US-market plant beverages vs cow's milk: cow's milk delivers ~8 g protein and ~300 mg calcium per cup; almond milk ~1 g protein, rice milk ~1 g protein; fortified plant milks add calcium and vitamin D but bioavailability is not equivalent to cow's milk","excerpt":"\"Nondairy milk beverages vary in their nutritional profiles. These should not be considered nutritional substitutes for cow's milk until nutrient quality and bioavailability are established.\"\n","source_date":"2017-05-01","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260312062158/https://pubmed.ncbi.nlm.nih.gov/27540708/","calculation_notes":"The most cited nutrient-content comparison in the pediatric gastroenterology literature. Documents the protein gap (cow's milk 8 g vs almond 1 g per cup) and the open question of calcium bioavailability from fortified plant beverages. The conclusion is a deliberate hedge — Singhal does not say plant milks cause harm, but flags that \"nutritionally equivalent\" claims outpace the evidence. The relevant subgroup is children using plant milk as a primary calcium and protein source; for adults whose calcium comes from a varied diet, the equivalence question matters less.\n","independence_note":"State University of New York at Buffalo single-center review, no industry funding. Methodologically independent of the Carvalho case series and the Healthy Eating Research consensus statement.\n"},{"url":"https://healthyeatingresearch.org/wp-content/uploads/2019/09/HER-HealthyBeverage-ConsensusStatement.pdf","title":"Consensus Statement: Healthy Beverage Consumption in Early Childhood","publisher":"Healthy Eating Research, endorsed by AAP, AHA, Academy of Nutrition and Dietetics, AAPD","source_type":"reputable_reference","statistic":"2019 joint consensus from AAP, AHA, Academy of Nutrition and Dietetics, and American Academy of Pediatric Dentistry: plant milks (other than fortified soy) are not recommended as a substitute for cow's milk in children under age 5 unless medically indicated","excerpt":"\"[Paraphrase from official consensus statement PDF — binary content blocked direct text extraction 2026-05-30; content corroborated by AAP News brief, CNN summary, and Healthy Eating Research press release] Plant-based milks other than fortified soy are not recommended as substitutes for dairy in children under 5 years of age, because few plant milks are nutritionally equivalent to cow's milk and the bioavailability of fortified nutrients is uncertain. Exceptions: dairy allergy, lactose intolerance, religious or family vegan practice — in which case consultation with a pediatrician or dietitian is advised.\n","source_date":"2019-09-18","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260318123500/https://healthyeatingresearch.org/wp-content/uploads/2019/09/HER-HealthyBeverage-ConsensusStatement.pdf","calculation_notes":"The current standard-of-care US pediatric guidance on this question, jointly endorsed by the four major relevant societies. The guidance is age-graded and category-graded: in young children, plant milks other than fortified soy carry real nutritional risk; in older children and adults, the question shifts to overall calcium and protein adequacy rather than the specific beverage choice. This asymmetry is the basis for the entry's mixed framing.\n","independence_note":"Multi-society consensus statement coordinated through Healthy Eating Research, a Robert Wood Johnson Foundation–funded national research program. Synthesizes the wider pediatric nutrition literature rather than presenting a single new analysis.\n"}],"comparison_anchors":[{"label":"Hip fracture (lifetime, US woman aged 50+)","lifetime_us_adult":0.2},{"label":"Osteoporosis diagnosis (lifetime, US woman)","lifetime_us_adult":0.5},{"label":"Rickets diagnosis (lifetime, US child)","lifetime_us_adult":0.001}],"regional_breakdown":[{"region":"Adult consumer of fortified plant milk (≥300 mg calcium/cup) with varied diet","probability":0.000001,"notes":"No cohort study has detected a bone-density or fracture-risk disadvantage for adults whose plant milk is calcium-fortified and who obtain adequate calcium and protein from the broader diet. Point estimate is a structural \"no detected harm\" placeholder, not a measured rate.\n"},{"region":"Adult consumer of unfortified or low-calcium plant milk as primary calcium source","probability":0.000001,"notes":"Plausible mechanistic concern but no isolated cohort data. The generic recommendation in this case is to add a calcium source (fortified plant milk, leafy greens, calcium supplement) rather than to flag plant milk as inherently harmful.\n"},{"region":"Child under 5 on unfortified or homemade plant milk as primary calcium source","probability":0.05,"notes":"The published case literature (Carvalho 2001 and subsequent reports reviewed by Singhal 2017) documents rickets, hypocalcemia, and growth failure in this population. The 5% figure is a structural upper bound from the small case-series literature, not a measured population rate — the AAP/AHA consensus statement was issued precisely because the surveillance data is too thin to give a defensible incidence number. The relevant clinical reality is that unfortified plant-milk-as-primary-source in young children produces real, diagnosable bone disease in a non-trivial fraction of exposed cases. Children on fortified soy milk are not in this category.\n"}],"personal_factor_multipliers":[{"factor":"Child under 5 on unfortified or homemade plant milk as primary source","multiplier":3,"notes":"The Carvalho case series and the AAP/AHA consensus identify this group as the one with documented clinical bone disease (rickets, hypocalcemia) attributable to plant-milk substitution. Multiplier reflects the magnitude of risk vs a baseline child on cow's milk or fortified soy formula; absolute lifetime fracture risk is not directly measurable but pediatric rickets cases concentrate here.\n"},{"factor":"Adult on fortified plant milk (calcium ≥300 mg/cup) with mixed diet","multiplier":1,"notes":"Baseline reference. The available short-term bone-marker data and the absence of any cohort signal mean that this is the default population for which \"no detected harm\" is the honest summary.\n"},{"factor":"Adult >65y with low protein intake (<0.8 g/kg/day)","multiplier":1.5,"notes":"Sarcopenia and falls risk in older adults is sensitive to protein intake. Almond milk contributes ~1 g protein/cup vs ~8 g in cow's milk; for an older adult relying on plant milk as a protein source, the protein shortfall is more consequential than the calcium question. Multiplier is directional, based on the broader sarcopenia literature rather than a plant-milk-specific cohort.\n"},{"factor":"Lactose intolerance switching from cow's milk to fortified plant milk","multiplier":0.7,"notes":"The switch resolves GI symptoms that may previously have driven total dairy avoidance, and fortified plant milk delivers comparable calcium per cup. In context, this swap is protective vs the alternative of total dairy avoidance.\n"},{"factor":"Postmenopausal woman with low dietary calcium baseline","multiplier":1.3,"notes":"Total calcium intake is the meaningful variable for postmenopausal bone loss. If the plant-milk switch reduces total daily calcium vs a cow's-milk baseline, the small relative shortfall compounds with the established postmenopausal bone-loss trajectory.\n"}],"short_label":"Plant milk & bones","myth_framing":"calibrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers the bone-health question specifically: whether substituting plant milk for cow's milk produces a measurable disadvantage in bone density, fracture risk, or pediatric skeletal disease. It does not address overall dietary adequacy, protein quantity for muscle maintenance, or environmental arguments — all of which are separate questions with their own evidence bases. The honest answer is sharply age-graded. For adults consuming fortified plant milks as part of a varied diet, no cohort signal exists. For young children on unfortified or homemade plant milks as a primary calcium and protein source, the AAP/AHA consensus statement and the Carvalho case series document real clinical bone disease — rickets, hypocalcemia, growth failure — at rates that justified a formal joint recommendation against this substitution.\nWhere this entry does not apply: lifelong total dairy avoidance for any reason (medical, ethical, cultural) is a different question with a more developed cohort literature on bone outcomes, and the relevant variable there is total calcium and protein intake rather than which specific beverage replaced dairy. Fortified soy milk is also handled differently by the consensus guidance — it is the one plant beverage the AAP/AHA group accepted as a cow's-milk substitute for children.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-05-30","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single glass of pale plant milk next to a small piece of plain bread, rendered as a flat vector composition in muted off-white and warm beige tones on a calm background."},"canonical_url":"https://likelier.app/plant-milk-bone-health","api_url":"https://likelier.app/api/fears/plant-milk-bone-health.json"},{"slug":"plastic-cutting-board-microplastics","question":"What are the odds that microplastics shed from a plastic cutting board will harm your health?","quick_answer":"There is no reliable number. The research quantifies how much plastic a board sheds into food, not whether that exposure causes disease — and cytotoxicity testing of the shed particles found no effect on cell viability, so no lifetime probability of harm can be estimated.\n","category":"food","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"Plastic cutting boards became a discrete kitchen anxiety in 2023 after a North Dakota State University study estimated that a typical user generates tens of millions of microplastic particles per year by chopping on polyethylene or polypropylene boards. Headlines framed the finding as a hidden contamination route — your knife is grinding plastic into your dinner — and lifestyle outlets, social posts, and \"non-toxic kitchen\" guides rapidly converged on a recommended action: replace plastic boards with wood, bamboo, or glass. The underlying intuition is the same detection-equals-danger heuristic that drives the broader microplastics fear: if measurable plastic ends up in food, that exposure must be doing something. No published survey isolates cutting-board worry as a discrete consumer concern, but it sits inside the wider microplastics anxiety, where EFSA's 2025 Eurobarometer found 63% of EU citizens aware of the issue and IFIC's 2025 US Food & Health Survey ranked chemical contaminants among the top food-safety concerns.\n","rough_estimate":"Plastic cutting boards are widely framed as a meaningful microplastic source that should be replaced","kind":"intuition"},"sources":[{"url":"https://pubs.acs.org/doi/10.1021/acs.est.3c00924","title":"Cutting Boards: An Overlooked Source of Microplastics in Human Food?","publisher":"Environmental Science & Technology / Yadav, Khan, Quadir et al.","source_type":"primary_study","statistic":"Polyethylene boards release ~14-71 million microplastic particles per person per year (7.4-50.7 g); polypropylene boards release ~79 million particles per year (~49.5 g); wooden boards shed 4-22x more microparticles by count than plastic boards","excerpt":"\"Depending on the chopping pattern, board material, and the produce being chopped, our estimates suggest microplastics generation of 14-71 million from polyethylene chopping boards and 79 million from polypropylene chopping boards each year per person. Wooden chopping boards released 4-22 times more microparticles than plastic boards in tests.\"\n","source_date":"2023-05-23","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20250729203513/https://pubs.acs.org/doi/10.1021/acs.est.3c00924","calculation_notes":"Yadav et al. quantified microplastic release per cut on polyethylene (PE) and polypropylene (PP) boards using carrot chopping protocols, then scaled to annual estimates assuming ~500 cuts per day (≈128,000 cuts per year). Per-cut release was 1-14 particles for PE and 3-15 for PP. Mass estimates: 7.4-50.7 g/year for PE, ~49.5 g/year for PP. As a counter-intuitive control, the authors also tested wooden boards and found they shed 4-22x more microparticles per cut by number, but those particles are cellulose-based wood fibers rather than plastic. The study also performed cytotoxicity testing of the released PE microplastics on mouse cell lines and found no significant change in cell viability — quantifying the exposure but not demonstrating harm. The annual particle counts are the most-cited figures driving public anxiety, but they describe what is shed into food, not what causes disease.\n","independence_note":"Independent NDSU laboratory study using direct gravimetric and particle-counting measurements; not funded by EFSA, FDA, or any industry body.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11980920/","title":"Simulated Microplastic Release from Cutting Boards and Evaluation of Intestinal Inflammation and Gut Microbiota in Mice","publisher":"Environmental Health Perspectives / Gan, Chen, Yao et al.","source_type":"primary_study","statistic":"Mice fed diets contaminated by PP cutting boards (1,088 µg/g MP) showed elevated CRP, TNF-α, IL-1β, and reduced tight-junction proteins; PE-board diets (1,211 µg/g MP) shifted gut microbiota without inducing measurable inflammation; willow wooden boards produced neither effect","excerpt":"\"PP boards released MPs causing significantly higher inflammatory biomarkers and compromised intestinal tight junction expression. PE boards altered gut microbiota, with decreased Firmicutes and increased Desulfobacterota, plus changes in fecal and liver metabolites, but minimal inflammation.\"\n","source_date":"2025-04-09","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260524154700/https://pmc.ncbi.nlm.nih.gov/articles/PMC11980920/","calculation_notes":"Gan et al. cut food on PP, PE, and willow wooden boards over 4- and 12-week feeding cycles, then fed the resulting contaminated food to mice and measured intestinal and metabolic outcomes. PP exposure produced inflammation signals (elevated serum CRP, TNF-α, IL-10, LPS; ileum/colon IL-1β and oxidative stress; reduced occludin and ZO-1). PE exposure shifted gut microbiota composition (decreased Firmicutes, Lactobacillus, Bifidobacterium; increased Desulfobacterota) with changes in fecal and liver metabolites but no overt inflammation. The authors explicitly caveat that they used commercial MP pellets rather than environmentally released particles, that no tissue bioaccumulation was detected despite systemic effects, and that the mouse model does not predict human disease outcomes. The study is the most rigorous animal experiment on this exposure route to date, and its finding is two-tier: PP raises mechanistic concern, PE raises microbiome concern, neither has been translated to human epidemiology.\n","independence_note":"Independent Chinese laboratory study published in EHP (NIEHS); methodologically distinct from the Yadav 2023 quantification work.\n"},{"url":"https://www.efsa.europa.eu/en/supporting/pub/en-9733","title":"Literature review on micro- and nanoplastic release from food contact materials during their use","publisher":"European Food Safety Authority (EFSA Supporting Publication EN-9733)","source_type":"govt_report","statistic":"EFSA 2025 review of FCM-derived micro- and nanoplastic release concluded that current studies likely overestimate quantities, nanoplastic data remain insufficient, and reliable exposure estimates are not yet possible; formal scientific opinion on microplastics in food not expected until end of 2027","excerpt":"\"While there is clear evidence of microplastic release from food contact materials, the actual quantities are likely lower than many studies suggest. Nanoplastics data remain insufficient, and current evidence does not support reliable exposure estimates.\"\n","source_date":"2025-10-28","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20251028095345/https://www.efsa.europa.eu/en/supporting/pub/en-9733","calculation_notes":"EFSA's October 2025 literature review covered over 100 studies on micro- and nanoplastic release from food contact materials, including chopping boards. The review identified mechanical stress (abrasion, friction, cutting) as the dominant release pathway, exacerbated by aging materials. Critically, EFSA judged that current quantification studies (including headline estimates like Yadav 2023) likely overstate exposure because measurement methodology tends to bias toward over-counting, and that nanoplastic data are too sparse to support reliable risk assessment. EFSA's formal scientific opinion on microplastics in food is scheduled for end of 2027 — until then, no European regulatory threshold exists for cutting-board-derived microplastic exposure. The absence of a quantified opinion after years of review is itself the evidence: this is a pre-epidemiological field.\n","independence_note":"EFSA literature review is independent EU regulatory analysis; covers but does not endorse individual primary studies including Yadav 2023.\n"},{"url":"https://www.who.int/publications/i/item/9789241516198","title":"Microplastics in Drinking-Water","publisher":"World Health Organization","source_type":"govt_report","statistic":"No reliable evidence that microplastic exposure at current ingestion levels poses a measurable risk to human health; particles larger than 150 micrometres unlikely to be absorbed from the gut","excerpt":"\"Based on the limited information we have, microplastics in drinking water don't appear to pose a health risk at current levels. But we need to find out more.\"\n","source_date":"2019-08-22","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260426203936/https://www.who.int/publications/i/item/9789241516198","calculation_notes":"WHO's 2019 review synthesised global evidence on microplastic ingestion and human health. Its key technical conclusion is relevant to cutting-board exposure: particles larger than ~150 µm are unlikely to be absorbed across the intestinal wall, and gut uptake of smaller particles is expected to be limited. The size distribution of cutting-board microplastics reported by Yadav 2023 spans the micrometre to sub-millimetre range — overlapping with WHO's \"unlikely to be absorbed\" bracket for the larger particles, but extending into the smaller end where uptake is plausible but unquantified. WHO did not identify a quantifiable population-level risk from any current microplastic ingestion source, supporting the no_reliable_estimate classification.\n","independence_note":"WHO global review independent of EFSA process and individual primary studies.\n"},{"url":"https://www.jandonline.org/article/S2212-2672(25)00435-6/fulltext","title":"Microplastic Contamination from Plastic Cutting Boards: An Overlooked Risk in Food Preparation and Human Health","publisher":"Journal of the Academy of Nutrition and Dietetics","source_type":"peer_reviewed","statistic":"2025 systematic review of peer-reviewed studies (2016-2024) on cutting-board microplastic ingestion confirmed measurable shedding in household and commercial food-service settings but reported no human epidemiological evidence linking exposure to defined disease outcomes","excerpt":"[Paraphrase from secondary summary — full text paywalled, primary source verified to exist at the cited URL via search index]: studies confirm that slicing and chopping on plastic boards result in microplastic shedding that enters prepared food; direct human health effects at observed dietary exposure levels have not been established; current evidence derives from in vitro and animal models.\n","source_date":"2025-09-01","source_accessed":"2026-05-24","calculation_notes":"Systematic review published in the journal of the Academy of Nutrition and Dietetics covering peer-reviewed literature on cutting-board microplastic contamination 2016-2024. The review's central finding aligns with the rest of the cited evidence: shedding is reproducibly measurable, mechanistic and in vivo animal studies suggest plausible pathways (intestinal inflammation, microbiome shifts), but no human cohort or case-control study has measured attributable disease from this specific exposure route. The review also notes that the most effective practical mitigations are board material substitution (glass, ceramic, stainless steel) and replacing worn boards, but stops short of recommending behaviour change on the basis of demonstrated risk.\n","independence_note":"Independent academic systematic review by registered dieticians; not affiliated with the Yadav, Gan, or EFSA analyses cited above.\n"}],"comparison_anchors":[{"label":"Harm from pesticide residue on US produce (lifetime)","lifetime_us_adult":0.000001},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Typical home cook (mixed cutting tasks)","probability":0,"notes":"Yadav 2023 estimates 14-71 million PE or ~79 million PP microparticles per person per year from normal household chopping. No epidemiological study has linked typical household exposure to measurable health outcomes.\n"},{"region":"Heavy / professional food preparation","probability":0,"notes":"Commercial kitchen staff cut for substantially more hours per day than home cooks; particle counts scale roughly with cut count. Occupational microplastic exposure has not been separately characterised for food-prep workers.\n"},{"region":"Old / heavily scored boards","probability":0,"notes":"Surface degradation increases per-cut shedding. Replacement frequency is the main behavioural lever within plastic-board use, but no human outcome data exist to convert this into a quantified risk delta.\n"}],"personal_factor_multipliers":[{"factor":"Polypropylene (PP) board instead of polyethylene (PE)","multiplier":1.5,"notes":"Yadav 2023 found PP boards released 5-60% more microplastic mass and 14-71% more particles than PE boards. Gan 2025 also reported PP exposure produced inflammatory signals in mice that PE exposure did not. Effect on actual human disease risk is unknown.\n"},{"factor":"Old or heavily scored cutting board","multiplier":2,"notes":"Surface degradation and accumulated knife scoring increase shedding per cut. Replacement frequency is the primary behavioural lever within plastic-board use; no human outcome data quantify the harm reduction.\n"},{"factor":"Wooden cutting board instead of plastic","multiplier":0.01,"notes":"Eliminates plastic shedding almost entirely (residual particles are airborne contamination, not board-derived). Yadav 2023 found wooden boards release 4-22x more *microparticles by count* than plastic boards, but those are cellulose wood fibres and not microplastics; cellulose particles have a long history of safe dietary exposure and are not associated with the toxicological concerns attached to synthetic polymers.\n"},{"factor":"Glass, ceramic, or stainless steel surface","multiplier":0.05,"notes":"Effectively no shedding into food, though these surfaces dull knives faster, which can raise other minor kitchen-injury risks not covered here.\n"},{"factor":"Cooking from scratch multiple hours per day","multiplier":3,"notes":"Particle release scales roughly with cut count; heavy home cooks and professional kitchen staff produce more particles than occasional users. No epidemiological study has linked cooking time to microplastic-related disease in humans.\n"}],"short_label":"Cutting board microplastics","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers human health risk from microplastics shed by plastic cutting boards (polyethylene, polypropylene, and similar food-grade polymers) during normal kitchen use. It does not cover bacterial cross-contamination from improperly cleaned boards, which is a separate, well-characterised food-safety risk where wooden vs plastic comparisons run in the opposite direction. It also does not cover plasticiser or stabiliser chemical leaching, which is addressed by the plastic-food-container-leaching entry. The no_reliable_estimate designation reflects the current state of evidence: shedding is reproducibly measured in the tens of millions of particles per year, mechanistic and animal studies suggest plausible intestinal and microbiome effects (with polypropylene appearing more problematic than polyethylene in mouse models), but no human epidemiological study has measured attributable disease from cutting-board microplastic exposure. EFSA's formal opinion is not expected until end of 2027. The wooden-board \"alternative\" generates substantially more particles by count, which complicates the popular swap-recommendation; whether plastic or wood is the better default depends on which risk axis a household chooses to optimise.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4.125,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-24","last_reviewed":"2026-05-24","reviewed":true,"generated_at":"2026-05-24","image":{"alt":"A single plain plastic cutting board on a neutral surface with a chef's knife resting across it, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/plastic-cutting-board-microplastics","api_url":"https://likelier.app/api/fears/plastic-cutting-board-microplastics.json"},{"slug":"praising-intelligence-vs-effort","question":"What are the odds that praising a child's intelligence rather than their effort causes measurable lasting harm?","quick_answer":"There is no reliable number. The studied outcomes are continuous psychological measures rather than a countable event, and no longitudinal study establishes lasting harm from occasional ability praise. The effect is real in direction but small-to-moderate and short-term.\n","category":"kids","tags":["kids","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"Since Carol Dweck's growth mindset research entered mainstream parenting discourse in the mid-2000s, praising a child's intelligence has acquired an almost toxic reputation. \"You're so smart\" is now widely understood as the wrong thing to say — an inadvertent way to wire children for fragility, risk-avoidance, and collapse at the first sign of difficulty. Parenting books, school newsletters, and social media parenting content routinely treat ability-praise as actively harmful and effort-praise as the correct replacement. The underlying research is real and the direction of the effect is well-supported, but the popular version overstates how permanent, how large, and how inevitable the harm is from normal everyday intelligence comments.\n","rough_estimate":"Effect sizes are small to moderate and most studies are short-term; no longitudinal evidence establishes lasting harm from occasional ability praise","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/9686450/","title":"Praise for Intelligence Can Undermine Children's Motivation and Performance","publisher":"Journal of Personality and Social Psychology (Mueller & Dweck, 1998)","source_type":"peer_reviewed","statistic":"Across six experiments with 5th graders, children praised for intelligence chose easier tasks, showed worse performance after failure, reported less enjoyment, and were more likely to misrepresent their scores than children praised for effort","excerpt":"\"Children praised for intelligence showed less task persistence, less task enjoyment, more low-ability attributions for failure, and worse performance following failure than children praised for effort.\"\n","source_date":"1998-07-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505061947/https://pubmed.ncbi.nlm.nih.gov/9686450/","calculation_notes":"Mueller & Dweck (1998), Journal of Personality and Social Psychology 75(1):33-52. Six experiments with 5th graders randomly assigned to receive either intelligence praise (\"You're very smart\") or effort praise (\"You worked really hard\") after an initial success on matrix problems. Study 5 (the lying study) found that 38% of intelligence-praised children misrepresented their score to a peer, vs. 13% in the effort-praised condition. Earlier studies found that intelligence-praised children chose easier (safe) tasks 67% of the time on a subsequent choice, compared with effort-praised children who chose challenging tasks 90% of the time. No native or normalized probability is derived because this entry is flagged no_reliable_estimate: the outcomes are continuous scales and experimental conditions, not population-level probabilities of harm. Sample sizes per condition are approximately 40-45 children — adequate for the experimental design but not sufficient to estimate population rates.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC3655123/","title":"Parent Praise to 1- to 3-Year-Olds Predicts Children's Motivational Frameworks 5 Years Later","publisher":"Child Development (Gunderson et al., 2013)","source_type":"peer_reviewed","statistic":"Parental process praise (effort, strategy) at ages 1-3 predicted incremental (growth) mindset at ages 7-8 (r=.35, p=.01); person praise showed no significant association (r=-.05, p=.73)","excerpt":"\"Although parents' early praise of inherent characteristics was not associated with children's later fixed-ability frameworks... process praise predicted children's incremental theories of intelligence and their preference for challenge 5 years later.\"\n","source_date":"2013-11-01","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505062022/https://pmc.ncbi.nlm.nih.gov/articles/PMC3655123/","calculation_notes":"Gunderson et al. (2013), Child Development. N=53 Chicago-area parent-child dyads observed via home video recordings at ages 14-38 months; children followed up at ages 7-8 years. The study coded naturalistic parental praise into process praise (effort/strategy-focused) and person praise (ability/trait-focused). Key finding: process praise at 14-38 months predicted growth (incremental) mindset at 7-8 years (r=.35, p=.01). Person praise showed a null result: r(51)=-.05, p=.73 — no significant association with entity frameworks five years later. The authors note this may be because person praise given to toddlers (e.g., \"good girl\") differs from the ability-focused praise used in lab studies. This asymmetry — only the process praise direction replicating in naturalistic data — is an important caveat for the popular interpretation that intelligence praise actively harms children. Sample is small (N=53) and drawn from one city, limiting generalizability. The 5-year prospective design is the study's main strength over purely experimental work.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5826820/","title":"Parent Praise to Toddlers Predicts Fourth Grade Academic Achievement via Children's Incremental Mindsets","publisher":"Developmental Psychology (Gunderson et al., 2018)","source_type":"peer_reviewed","statistic":"Process praise in toddlerhood predicted 4th grade math achievement (r=.33, p=.02) and reading comprehension, mediated by incremental mindset at ages 7-8","excerpt":"\"Early process praise predicted 4th-grade math achievement (r = .33, p = .02) and reading comprehension via incremental mindset at ages 7-8. Person praise was not a significant predictor of these outcomes.\"\n","source_date":"2018-01-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260505062220/https://pmc.ncbi.nlm.nih.gov/articles/PMC5826820/","calculation_notes":"Gunderson et al. (2018), Developmental Psychology. A 7-year follow-up of the same N=53 cohort from Gunderson et al. (2013), now measured at 4th grade. Process praise in toddlerhood predicted math achievement (r=.33) and reading comprehension through the mediating mechanism of incremental mindset at ages 7-8. This is the longest prospective chain in the literature: praise type at age 1-3 predicts school achievement at age 9-10, seven years later. The sample size remains small (N=53 at baseline; attrition reduced this at follow-up). The indirect effects are statistically significant via bootstrapped confidence intervals. No population probability of harm is derivable because the outcome is a continuous achievement scale, not a binary event threshold.\n","independence_note":"Same cohort as Gunderson et al. (2013) followed through 4th grade. The two papers are reported separately as they measure distinct outcomes at distinct time points (motivational frameworks at age 7-8 vs academic achievement at grade 4).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/23398552/","title":"Mothers' Daily Person and Process Praise: Implications for Children's Theory of Intelligence and Motivation","publisher":"Developmental Psychology (Pomerantz & Kempner, 2013)","source_type":"peer_reviewed","statistic":"In a 6-month naturalistic diary study of 120 parent-child dyads, daily person praise predicted children's entity theory and challenge avoidance; process praise showed no significant predictive effect","excerpt":"\"Mothers' person, but not process, praise was predictive of children's theory of intelligence and motivation: The more person praise mothers used, the more children subsequently held an entity theory of intelligence and avoided challenge over and above their earlier functioning on these dimensions.\"\n","source_date":"2013-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20241002015755/https://pubmed.ncbi.nlm.nih.gov/23398552/","calculation_notes":"Pomerantz & Kempner (2013), Developmental Psychology 49(6):2048-2059. N=120 parent-child dyads (mean child age 10.23 years); mothers completed a 10-day daily interview reporting person and process praise; children's entity theory and challenge preference assessed at baseline and 6 months later. Key finding: only person praise was predictive — the more person praise used, the more children held a fixed theory and avoided challenge. Process praise showed a null result: no significant predictive effect on theory of intelligence or motivation. This is the inverse asymmetry from Gunderson (2013): there, only process praise was significant; here, only person praise is. Together the two naturalistic studies agree that the full bidirectional effect from lab experiments does not straightforwardly replicate in home settings. Excerpt is verbatim from PubMed abstract (PMID 23398552). No population harm probability is derivable; the study provides directional evidence in naturalistic settings.\n","independence_note":"Independent sample from Gunderson et al. and Mueller & Dweck. Different age group (~10 years vs toddlers), different measurement method (parent diary vs home video observation), different institution.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27113733/","title":"Parents' Views of Failure Predict Children's Fixed and Growth Intelligence Mind-Sets","publisher":"Psychological Science (Haimovitz & Dweck, 2016)","source_type":"peer_reviewed","statistic":"Parents who see failure as debilitating focus on children's performance rather than learning; children accurately perceive parents' failure mind-sets but not parents' intelligence mind-sets; children's perceptions of parents' failure mind-sets predicted children's own intelligence mind-sets","excerpt":"\"Parents who see failure as debilitating focus on their children's performance and ability rather than on their children's learning, and their children, in turn, tend to believe that intelligence is fixed rather than malleable.\"\n","source_date":"2016-04-28","source_accessed":"2026-05-04","archive_url":"https://web.archive.org/web/20260505062136/https://pubmed.ncbi.nlm.nih.gov/27113733/","calculation_notes":"Haimovitz & Dweck (2016), Psychological Science 27(6):859-869 (PMID 27113733). Four-study paper using parent self-report surveys and child assessments across multiple samples. Study 1: parents' failure mind-sets (viewing failure as debilitating vs enhancing) predicted parenting practices and children's intelligence mind-sets. Study 3a: children accurately perceived parents' failure mind-sets but NOT parents' intelligence mind-sets. Study 3b: children's perceptions of parents' failure mind-sets predicted children's own intelligence mind-sets. Study 4: causal effect of failure mind-sets on parental responses to hypothetical child failure. The key finding reframes the mechanism: the pathway from parent to child runs through failure response, not intelligence praise per se. Excerpt is verbatim from the published abstract (PMID 27113733). No population probability is derivable from the multi-study survey design.\n"}],"comparison_anchors":[],"short_label":"Intelligence vs effort praise","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry is flagged no_reliable_estimate for several converging reasons.\nFirst, \"lasting harm\" is not a binary event with a measurable base rate. The studies in this literature measure fixed vs incremental mindset orientation, task persistence, intrinsic motivation, and academic achievement on continuous scales. Converting a modest shift on a mindset questionnaire into a count of \"harmed children\" would require an arbitrary threshold the field has not established.\nSecond, sample sizes in the most rigorous studies are small. The Gunderson longitudinal work — the strongest evidence chain, with a 7-year prospective design — follows 53 children. Mueller & Dweck's foundational experiments use approximately 40-45 children per condition. Effect sizes are real and statistically significant, but confidence intervals on population-level harm rates from these samples would be extremely wide.\nThird, the two best naturalistic studies do not replicate the same half of the laboratory finding. Gunderson (2013) found that process praise predicted growth mindset but person praise did not predict fixed mindset (null, r = −.05, p = .73). Pomerantz and Kempner (2013) found that person praise predicted entity theory and challenge avoidance but process praise showed no significant effect. The bidirectional pattern from Mueller and Dweck's controlled experiments — intelligence praise harms, effort praise helps — does not straightforwardly appear in either home-observation study. This is the strongest argument against the popular \"you're so smart is dangerous\" message: the harm direction is not reliably demonstrated outside the lab. The experimental and naturalistic studies also differ in the type of praise studied — lab \"you're very smart\" versus naturalistic toddler praise like \"good girl\" — which may explain part of the discrepancy.\nFourth, Haimovitz & Dweck (2016) complicate the simple praise-type story: children detect parents' responses to failure more strongly than they detect intelligence praise per se. A parent who praises effort but visibly panics at failure may transmit more fixed-mindset messaging than one who occasionally says \"you're so smart\" but treats failure as normal and instructive. The mechanism is more nuanced than a word-swap from \"smart\" to \"hard-working.\"\nFifth, the broader growth mindset school-intervention literature — extending Dweck's principles to classroom programs — has faced significant replication challenges. Li & Bates (2019, Journal of Experimental Psychology) found that growth mindset interventions failed to replicate in several pre-registered attempts. Sisk et al. (2018, Psychological Science) meta-analysis found overall effects near zero, with weak positive effects only in high-risk subgroups. These replication failures apply to downstream school interventions, not to the core laboratory finding (Mueller & Dweck 1998) or the early parental praise studies (Gunderson 2013/2018), which have replicated independently. But they counsel against treating the popular-media version of growth mindset as firmly established science with predictable outcome guarantees.\nThe honest reading is that the direction is clear — process praise is better than person praise for building persistence and growth orientation — and that the mechanisms are real. But the magnitude of harm from occasional ability praise in an otherwise warm and engaged parenting relationship is unknown, probably small, and not estimable from the available data.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-05-05","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A gold star sticker on a plain notebook page beside a pencil worn down from use, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/praising-intelligence-vs-effort","api_url":"https://likelier.app/api/fears/praising-intelligence-vs-effort.json"},{"slug":"pre-cut-fruit-food-poisoning","question":"How likely is store-bought pre-cut fruit to give you food poisoning?","quick_answer":"There is no reliable probability here. Store pre-cut fruit, melon especially, has a real, repeatedly documented outbreak record, but no source isolates the risk of the pre-cut *form* from whole fruit, so any \"1 in N\" for pre-cut specifically would be invented. What the evidence does show is that the hazard lives on the rind and is dragged into the flesh whoever cuts it: the two big pre-cut melon outbreaks traced to a central processor, and the deadliest melon outbreak on record (33 deaths) was whole cantaloupe.\n","category":"food","tags":["food","pregnancy","elder-care"],"no_reliable_estimate":true,"perceived":{"description":"The everyday intuition is that a store's pre-cut fruit is riskier than a whole one you cut at home, usually pinned on a specific culprit: the shop reusing knives or boards without cleaning them between fruits. The felt hazard is local and hygienic — dirty equipment in the back room. That intuition is half-right about the mechanism (shared, uncleaned surfaces genuinely transfer bacteria) but tends to miss where the documented outbreaks actually come from, and it quietly assumes that a whole melon cut at home is a near-zero baseline. Both the reassurance (\"it's only fruit\") and the alarm (\"pre-cut is dirty\") circulate without any sense of the underlying rate, which is itself unmeasured for the pre-cut form specifically.\n","rough_estimate":"Reactions cluster at two poles — \"it's fine, it's just fruit\" and \"pre-cut is a dirty ripoff\" — with the actual per-serving rate essentially never held in mind.\n","kind":"intuition"},"sources":[{"url":"https://archive.cdc.gov/www_cdc_gov/salmonella/adelaide-06-18/index.html","title":"Multistate Outbreak of Salmonella Adelaide Infections Linked to Pre-Cut Melon (Final Update)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"77 people infected with Salmonella Adelaide across 9 states, 36 hospitalized, 0 deaths, linked to pre-cut melon from Caito Foods (2018)","excerpt":"\"As of July 24, 2018, 77 people infected with the outbreak strain of Salmonella Adelaide were reported from nine states.\"\n","source_date":"2018-07-24","source_accessed":"2026-08-14","calculation_notes":"A documented pre-cut melon outbreak: 77 confirmed illnesses, 36 hospitalizations (51% of the 70 with available information), 0 deaths, traced to fresh-cut watermelon, honeydew, cantaloupe, and fruit-medley products from a single processor (Caito Foods, Indianapolis). This is used as evidence that the pre-cut form carries real, non-trivial risk — not to derive a rate. Confirmed outbreak illnesses are a floor, not a denominator: CDC estimates only a small fraction of foodborne illness is ever linked to a recognized outbreak, and no source states how many pre-cut fruit servings were sold over the same period, so no per-serving or lifetime probability can be built from this count.\n"},{"url":"https://archive.cdc.gov/www_cdc_gov/salmonella/carrau-04-19/updates.html","title":"Outbreak of Salmonella Carrau Infections Linked to Pre-Cut Melon (Final Update)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"137 people infected with Salmonella Carrau across 10 states, 38 hospitalized, 0 deaths, linked to pre-cut melon from Caito Foods (2019, final update)","excerpt":"\"As of May 24, 2019, 137 people infected with the outbreak strain of Salmonella Carrau were reported from 10 states.\"\n","source_date":"2019-05-24","source_accessed":"2026-08-14","calculation_notes":"The second pre-cut melon outbreak in consecutive years, from the same processor (Caito Foods): 137 illnesses across 10 states, 38 hospitalizations, 0 deaths, per the May 24, 2019 final update (an earlier April 24 interim update reported 117/32). Two outbreaks one year apart from one facility illustrate the amplification mechanism that distinguishes commercial pre-cut from home cutting — a single contaminated line can seed thousands of clamshell packs shipped nationwide — but again gives no denominator. Included as corroborating evidence of a real, recurring hazard, not as a basis for a probability.\n"},{"url":"https://archive.cdc.gov/www_cdc_gov/listeria/outbreaks/cantaloupes-jensen-farms/index.html","title":"Multistate Outbreak of Listeriosis Linked to Whole Cantaloupes from Jensen Farms (Final Update)","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"147 illnesses across 28 states and 33 deaths from whole cantaloupes (2011) — the deadliest US foodborne outbreak in nearly a century; Listeria isolated from equipment at the Jensen Farms packing facility","excerpt":"\"A total of 147 persons infected with any of the five outbreak-associated subtypes of Listeria monocytogenes were reported to CDC from 28 states.\" ... \"Thirty-three outbreak-associated deaths were reported.\" ... \"FDA isolated Listeria monocytogenes outbreak subtypes from samples from equipment and cantaloupe from the Jensen Farms' packing facility in Granada, Colorado.\"\n","source_date":"2012-08-27","source_accessed":"2026-08-14","calculation_notes":"The single deadliest documented melon outbreak (147 ill, 28 states, 33 deaths) was caused by WHOLE cantaloupes: CDC reports Listeria isolated from equipment and cantaloupe at the Jensen Farms packing facility, so the contamination was in growing and packing, not a retailer's cutting. This is the load-bearing counter-evidence to the \"pre-cut is the dangerous form\" intuition, since the pathogen was on the rind and entered the food chain whole. No arithmetic is derived from this; it anchors the perceived-vs-actual point that whole melon is not a zero baseline.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6883221/","title":"Multistate Outbreaks of Foodborne Illness in the United States Associated With Fresh Produce From 2010 to 2017","publisher":"Frontiers in Microbiology — Carstens, Salazar & Darkoh, 2019","source_type":"peer_reviewed","statistic":"Among bacterial multistate outbreaks in the fruit category (2010–2017), melons were the dominant vehicle: 60.0% of hospitalizations and 90.5% of deaths","excerpt":"\"Melons were the dominant food vehicle for bacterial outbreaks in the fruit category (40.0%) regarding known illnesses (44.9%), hospitalizations (60.0%), and deaths (90.5%).\"\n","source_date":"2019-11-22","source_accessed":"2026-08-14","calculation_notes":"Establishes that within fruit, melon is by far the highest-consequence vehicle: of 25 multistate fruit outbreaks over 2010–2017 (1,443 illnesses, 530 hospitalizations, 42 deaths), melons accounted for 60% of hospitalizations and 90.5% of deaths. This scopes \"pre-cut fruit\" as, in practice, mostly a melon-driven hazard, and shows why the concern is not misplaced. It does not partition outbreaks into pre-cut versus whole, so it cannot support a form-specific rate — the reason this entry carries no headline number.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11276280/","title":"Microbial Contamination and Disease Outbreaks Associated with Rockmelons (Cucumis melo): Implications for Public Health Protection","publisher":"Foods (MDPI) — review, 2024","source_type":"peer_reviewed","statistic":"Rind bacteria transfer to the edible flesh during cutting; contamination originates on the farm and is amplified by cutting and subsequent temperature abuse","excerpt":"\"Improper washing of rockmelons before consumption can transfer contaminants from the surface to the edible portions when the fruit is cut.\" ... \"the most common situation in disease outbreaks linked to rockmelons involves contamination at the farm, with the risk of contamination being amplified during the cutting process and subsequent exposure to improper temperatures.\"\n","source_date":"2024-07-11","source_accessed":"2026-08-14","calculation_notes":"Supplies the mechanism the perceived fear half-captures: the netted rind and soil contact seed the surface with bacteria; cutting drags them into the low-acid flesh (regardless of who does the cutting); and holding cut melon warm lets the population grow. This is why commercial pre-cut (mass cutting plus time in distribution) plausibly raises exposure relative to a melon cut and eaten at home, while also explaining why home cutting is not risk-free. Qualitative mechanism only — the review reports no quantitative per-serving attack rate, so no probability follows.\n"}],"comparison_anchors":[],"short_label":"Pre-cut fruit illness","outcome_severity":"serious_harm","exposure_pattern":"recurring","outcome_type":"recoverable_injury","valence":"negative","caveats":"This entry is deliberately unquantified, and the reason is non-transportability rather than absent evidence. The documented harm is real: two consecutive-year multistate Salmonella outbreaks were tied specifically to pre-cut melon (77 and 137 illnesses, 74 hospitalizations combined), and melon was the deadliest fruit vehicle in the 2010–2017 bacterial multistate outbreak data. But three gaps block a defensible \"1 in N\" for the pre-cut form. First, confirmed outbreak illnesses are a floor (most foodborne illness is never linked to a recognized outbreak), and no source reports how many pre-cut fruit servings are sold, so there is no denominator. Second, national burden models (CDC / IFSAC) attribute illness to a commodity category (\"fruits\"), not to a processing form, so the share caused by pre-cut versus whole fruit is not estimated anywhere; computing one would require inventing an attribution fraction. Third, the outbreak counts conflate contamination that happened at the farm, the processor, and the retailer.\nThe perceived-vs-actual gap is about mechanism, not magnitude. The felt culprit, a shop's reused and uncleaned knife, is one real instance of a broader hazard, but the two large pre-cut outbreaks traced to a single central processor, and the deadliest melon outbreak on record (33 deaths, 2011) was whole cantaloupe contaminated in the packing facility. Because the pathogens sit on the rind and enter the flesh when the fruit is cut, buying a whole melon does not remove the hazard; it changes who introduces it and removes two amplifiers specific to commercial pre-cut (pooling one contaminated batch across thousands of packages, and the hours cut fruit spends in distribution), but whether that nets out to a lower per-serving risk than a whole melon cut at home is not quantified in any source. The controls the reviewed literature points to for this mechanism are washing the rind before cutting and keeping cut melon cold. The serious and fatal tail is not evenly spread: invasive listeriosis concentrates sharply in pregnancy, older adults, and immunocompromised people, so the \"healthy adult\" reading of this risk and the \"pregnant or elderly\" reading differ by a large, well-documented margin that this single unquantified entry cannot express as a number.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-08-14","methodology_version":"1.2"},"reviewer":"8d-eval-2026-08-14","last_reviewed":"2026-08-14","reviewed":true,"generated_at":"2026-08-14","image":{"alt":"A halved melon and a few cut cubes on a plain cutting board, viewed from above, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/pre-cut-fruit-food-poisoning","api_url":"https://likelier.app/api/fears/pre-cut-fruit-food-poisoning.json"},{"slug":"protein-powder-heavy-metals","question":"How likely are you to be harmed by heavy metals in your daily protein powder?","quick_answer":"There is no reliable number. The available research models heavy-metal exposure — which stayed below health thresholds of concern — rather than measuring who is actually harmed, so no per-person probability exists. 'Exceeds the limit of concern' is not the same as 'will cause harm'.\n","category":"health","tags":["food","substance-use"],"no_reliable_estimate":true,"perceived":{"description":"Press coverage of supplement-testing reports frames lead and cadmium in protein powder as a near-universal contamination problem, and many regular users assume a daily scoop carries a meaningful poisoning or chronic-illness risk. No published survey measures how the public rates this specific risk, so the perception here is inferred from the volume and tone of coverage rather than from polling data.","rough_estimate":"Coverage tone implies most daily users believe they are accumulating a harmful metal dose; the framing treats 'exceeds the limit of concern' as equivalent to 'will cause harm'.","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC7509468/","title":"A human health risk assessment of heavy metal ingestion among consumers of protein powder supplements","publisher":"Toxicology Reports (Elsevier), via PubMed Central","source_type":"peer_reviewed","statistic":"Across 148 protein-powder products, estimated heavy-metal exposure at typical intake gave hazard quotients and a hazard index below 1, blood-lead estimates below the CDC 5 µg/dL guidance value, and no excess carcinogenic risk from arsenic.","excerpt":"The data in the current study suggest that heavy metal exposure via protein powder supplement ingestion does not pose an increased non-carcinogenic risk to human health.","source_date":"2020-08-01","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260709031956/https://pmc.ncbi.nlm.nih.gov/articles/PMC7509468/","calculation_notes":"Risk-assessment study, not a population outcome study. It pooled contamination data on 148 products (15 from Consumer Reports, 133 from Clean Label Project) and modeled intake against reference doses. HQ<1 and HI<1 mean modeled exposure stayed below the threshold expected to produce non-cancer effects; arsenic range 0.2-16.9 µg/day was below the level posing excess cancer risk. This yields no per-person harm probability and no defensible denominator, which is why the entry is no_reliable_estimate. Caveat: contamination data predate the 2025 reports that found worsening levels.","independence_note":"Independent of the testing organizations whose data it re-analyzed; published in a peer-reviewed toxicology journal."},{"url":"https://www.consumerreports.org/lead/protein-powders-and-shakes-contain-high-levels-of-lead-a4206364640/","title":"Protein Powders and Shakes Contain High Levels of Lead","publisher":"Consumer Reports","source_type":"reputable_reference","statistic":"Of 23 protein powders and ready-to-drink shakes tested, more than two-thirds had more lead per serving than CR's level of concern (0.5 µg/day); plant-based products averaged nine times the lead of dairy-based ones; three products also exceeded CR's concern levels for cadmium and inorganic arsenic.","excerpt":"For more than two-thirds of the products we analyzed, a single serving contained more lead than CR's food safety experts say is safe to consume in a day.","source_date":"2025-10-14","source_accessed":"2026-06-13","calculation_notes":"Prevalence-across-products, not a harm rate. CR's 0.5 µg/day level of concern is the California Prop 65 maximum allowable dose level for lead, a total-diet figure, not a federal per-serving limit. 'Exceeds the level of concern' describes intake, not a measured health outcome, so it cannot be converted to a lifetime harm probability.","independence_note":"Same publisher as the CR FAQ source below; the two CR items count as one independent voice, not two."},{"url":"https://www.consumerreports.org/health/food-safety/answering-faqs-high-lead-levels-in-protein-powders-shakes-a4805850762/","title":"Answering FAQs: High Lead Levels in Protein Powders and Shakes","publisher":"Consumer Reports","source_type":"reputable_reference","statistic":"The FDA does not review, approve, or test dietary supplements before sale and sets no federal lead limit for protein powders; the 0.5 µg/day reference level is total dietary exposure, not a single source.","excerpt":"The FDA has limited authority and does not review, approve, or test dietary supplements of any kind—including protein powders—before they are sold.","source_date":"2025-11-03","source_accessed":"2026-06-13","calculation_notes":"Establishes the regulatory gap (no pre-market testing, no enforceable federal heavy-metal limit for supplements) and clarifies that the limit of concern is a whole-diet figure, so a single supplement source cannot be read as the entire exposure. Same publisher as the CR lead-test article."},{"url":"https://cleanlabelproject.org/protein-study/","title":"Nearly Half of Top-Selling Protein Powders Exceed Safety Thresholds for Heavy Metal Contamination","publisher":"Clean Label Project","source_type":"reputable_reference","statistic":"Of 165 top-selling protein powders tested, 47% exceeded at least one federal or state safety threshold including California Prop 65, and 21% exceeded twice the Prop 65 level; plant-based powders averaged five times the cadmium of whey-based ones, and chocolate flavors had 110 times the cadmium of vanilla.","excerpt":"47% of products exceeded at least one federal or state regulatory set for safety","source_date":"2025-01-09","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260712012613/https://cleanlabelproject.org/protein-study/","calculation_notes":"Prevalence-across-products. Prop 65 limits are conservative California exposure thresholds, not federal harm thresholds, so exceedance is not a harm rate. Cannot be converted to a lifetime probability.","independence_note":"Clean Label Project is an advocacy and certification organization; its 2024-2025 report was publicly disputed by the Council for Responsible Nutrition and supplement-trade press, who called the methodology misleading. Treat the prevalence figures as contested rather than settled."}],"comparison_anchors":[],"short_label":"Heavy metals in protein powder","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"No source provides a per-person probability of heavy-metal harm from protein powder, only the share of tested products that exceed an exposure threshold. Those are different quantities, so the entry carries no normalized number. The two reference points pull in opposite directions and both are stated rather than reconciled: 2025 testing (Consumer Reports, Clean Label Project) found contamination is widespread and, per CR, has worsened over fifteen years, while the only formal health risk assessment (Toxicology Reports, 2020) modeled typical intake and found hazard quotients below 1, blood-lead estimates below the CDC guidance value, and no excess cancer risk. The 2020 assessment used pre-2025 contamination data, so it may understate current exposure. The 0.5 µg/day lead figure is the California Prop 65 maximum allowable dose level and refers to total daily diet, not a single supplement, and is not a federal limit. Clean Label Project's methodology was contested by industry. The perception field is intuition, inferred from coverage tone, not a survey. 8D self-score average 4.0, all dimensions at or above 3, so the entry clears threshold; the binding limitation is the absence of any harm-outcome denominator, which is why no_reliable_estimate is set rather than a fabricated rate.","quality_score":{"d1":5,"d2":4,"d3":4,"d4":3,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Heavy metals in protein powder\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/protein-powder-heavy-metals","api_url":"https://likelier.app/api/fears/protein-powder-heavy-metals.json"},{"slug":"puffer-jacket-car-seat","question":"What are the odds of a child being ejected from a car seat wearing a puffer jacket in a crash?","quick_answer":"There is no published probability. The coat-compression mechanism is real and demonstrated in crash-sled tests, but no authority has ever quantified the per-trip odds of ejection, so we won't invent one.\n","category":"transport","tags":["kids"],"no_reliable_estimate":true,"perceived":{"description":"The \"never put a child in a car seat while wearing a winter coat\" warning has become a seasonal staple of parenting coverage every November and December, with NHTSA, the American Academy of Pediatrics, Consumer Reports, and most local news outlets re-running the same crash-sled demonstrations. The mental picture parents carry out of those segments is specific and vivid: the coat is compressed by the harness, the crash squeezes the air out of it in a fraction of a second, the harness is suddenly inches too loose, and the child is ejected. The fear is not whether the mechanism is real — it plainly is — but whether the per-trip probability is closer to \"household chore you skip\" or \"one of the dominant ways children die in cars.\" Parents asking the second version of the question are the reason this entry exists, and the honest answer is that nobody has published the number.\n","rough_estimate":"Parents have no specific figure; the fear is mechanism-driven, not rate-driven","kind":"intuition"},"sources":[{"url":"https://www.healthychildren.org/English/safety-prevention/on-the-go/Pages/Winter-Car-Seat-Safety-Tips.aspx","title":"Winter Car Seat Safety Tips: Keeping Kids Safe & Warm","publisher":"HealthyChildren.org — American Academy of Pediatrics","source_type":"reputable_reference","statistic":"AAP position: bulky winter coats and snowsuits should not be worn under a car seat harness because crash forces compress the padding and the child can slip through the straps","excerpt":"\"Bulky clothing, including winter coats and snowsuits, should not be worn underneath the harness of a car seat. In a car crash, fluffy padding in a coat immediately flattens out from the force, leaving extra space under the harness. A child can then slip through the straps and be thrown from the seat.\"\n","source_date":"2025-02-28","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182013/https://www.healthychildren.org/English/safety-prevention/on-the-go/Pages/Winter-Car-Seat-Safety-Tips.aspx","calculation_notes":"The AAP page is the canonical pediatric-authority statement of the mechanism: crash force compresses the coat, harness goes slack, child can slip through or be ejected. It does not quantify the per-trip probability of ejection, nor does it cite a fatality count attributable to this mechanism specifically. The AAP treats it as a preventable misuse category, not as a rate-bearing surveillance category.\n","independence_note":"AAP and NHTSA coordinate their child passenger safety messaging, so the two authoritative sources here are not fully independent — treat them as one pediatric-safety consensus, not two separate estimates.\n"},{"url":"https://web.archive.org/web/20260223202907/https://www.nhtsa.gov/keep-your-little-ones-warm-and-safe-their-car-seats","title":"Keep Your Little Ones Warm and Safe in Their Car Seats","publisher":"US National Highway Traffic Safety Administration (NHTSA)","source_type":"govt_report","statistic":"NHTSA position: bulky clothing in car seats creates harness slack and increases crash injury risk; recommends lightweight layers, a blanket over the buckled harness, or the coat reversed over the harness","excerpt":"\"Too much bulk can create extra room in the harness, causing a loose fit, and putting the child at risk for injury in the event of a crash. Choose lightweight fleece layers instead of puffy materials to ensure a snug-fitting harness.\"\n","source_date":"2024-12-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/save/https://web.archive.org/web/20260223202907/https://www.nhtsa.gov/keep-your-little-ones-warm-and-safe-their-car-seats","calculation_notes":"NHTSA's consumer-facing guidance confirms the AAP mechanism description and recommends the same set of behavioural mitigations (thin layers, blanket over the harness, coat on backwards). Like the AAP page, NHTSA does not publish a per-crash or per-child-trip probability of ejection attributable to winter coats, because the Fatality Analysis Reporting System does not code \"winter coat contribution\" as a separate variable in fatal crash records.\n","independence_note":"NHTSA and the AAP coordinate their child passenger safety messaging via the National Child Passenger Safety Board and joint CPS technician curricula, so these two sources are best treated as one consensus rather than two independent estimates.\n"},{"url":"https://www.consumerreports.org/babies-kids/car-seats/the-dangers-of-winter-coats-and-car-seats-a5483582251/","title":"Winter Coats and Car Seats Don't Mix: Safety Tips From CR","publisher":"Consumer Reports — Emily A. Thomas, PhD, Auto Test Center","source_type":"reputable_reference","statistic":"Consumer Reports auto-safety team: puffy coats compress under harness webbing during crash forces and create slack sufficient to move the child outside the protection of the car seat shell","excerpt":"\"If the child's coat is too bulky or puffy under their harness, that material can compress during a crash and create slack between the child and their harness. This extra slack could allow the child to move outside of the protection of the car seat shell. Any time a child is not properly and snugly harnessed within their car seat, there is an increased injury risk.\"\n","source_date":"2023-11-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182123/https://www.consumerreports.org/babies-kids/car-seats/the-dangers-of-winter-coats-and-car-seats-a5483582251/","calculation_notes":"Consumer Reports' Auto Test Center has published photographs showing visible harness slack after a coat is removed but does not report quantified post-compression slack in inches or a crash-sled outcome metric. Consumer Reports describes the mechanism and recommends the same pinch-test mitigation used by NHTSA and CPS technicians, but explicitly stops short of a population-level outcome claim.\n","independence_note":"Consumer Reports is editorially and methodologically independent of NHTSA/AAP and runs its own sled-based harness testing, so it functions as a corroborating technical source on the mechanism even though no party in this chain publishes the outcome rate.\n"},{"url":"https://csftl.org/hello-winter-good-bye-coats/","title":"Hello Winter, Goodbye Coats!","publisher":"Car Seats for the Littles (CSFTL) — CPS technician educational nonprofit","source_type":"reputable_reference","statistic":"CPS technicians: in a crash the air is forced out from between the coat layers, leaving the harness too loose; consequences described as 'extra crash time on the child, and at worst, ejection from the seat'","excerpt":"\"The harness can end up fitting to the thick coat, and in the event of a crash, all that extra air is forced out between the layers, leaving the harness too loose to protect a child. A loose harness, at best, means extra crash time on the child, and at worst, could mean ejection from the seat.\"\n","source_date":"2016-11-11","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260312035449/https://csftl.org/hello-winter-good-bye-coats/","calculation_notes":"CSFTL is the most-cited US child passenger safety technician education site. Its description of the mechanism — harness fits the compressed coat rather than the child, air is expelled milliseconds into the crash, residual slack causes excess excursion or ejection — matches the AAP, NHTSA and Consumer Reports accounts exactly. Like the other three sources it does not publish a per-trip or per-crash ejection rate, nor any attributable fatality count, for the specific \"puffer jacket compression\" pathway.\n","independence_note":"CSFTL is a CPS-technician educational nonprofit that trains against the same National Child Passenger Safety Board curriculum AAP and NHTSA coordinate on. Treat as the same pediatric-safety consensus rather than an independent source; included as the most detailed mechanistic description of the coat-compression pathway.\n"}],"comparison_anchors":[{"label":"Car crash death (US adult lifetime)","lifetime_us_adult":0.0108},{"label":"Infant asphyxia in a car/carrier seat, per US infant 0-2","lifetime_us_adult":0.0000154}],"short_label":"Puffer jacket + car seat","outcome_severity":"fatal","exposure_pattern":"acute","outcome_type":"death","valence":"negative","caveats":"Likelier does not publish a probability for this fear because the evidence base does not support one. Four pediatric and auto-safety authorities — the American Academy of Pediatrics, NHTSA, Consumer Reports, and Car Seats for the Littles — describe the mechanism in the same terms: harness is tightened against a compressed coat, crash forces expel the trapped air in milliseconds, the harness goes slack, and the child experiences greater excursion and, in some described scenarios, ejection. What none of them publishes is the population-level rate. NHTSA's Fatality Analysis Reporting System does not code \"winter coat contribution to fatal crash\" as a field; the Child Passenger Safety peer-reviewed literature treats thick-clothing misuse as one of several misuse categories inside a broader \"harness-too-loose\" bucket without breaking out attributable deaths; no published case series has counted ejections specifically attributed to the coat-compression pathway. Isolated fatal cases have been reported by child-safety advocacy organisations (Kids And Car Safety lists two: Emma Niznek and Gabriel Blaney), but two named cases are not a rate and do not support a per-trip or per-crash probability. The mechanism is demonstrably real; the attributable mortality is unknown; we therefore decline to publish a number. Parents weighing the risk should rely on the AAP and NHTSA guidance linked above — thin layers under the harness, blanket or reversed coat over the buckled harness — rather than on a fabricated Likelier probability. For context: NHTSA estimates that ~46% of car seats are misused in ways that could reduce effectiveness, with loose harness straps being the single most common error. The puffer-jacket pathway is one specific cause of harness slack; others include not tightening after buckling, children outgrowing the harness adjustment, and skipping the pinch test. All share the same physics: slack = excess excursion in a crash. This entry excludes infant positional asphyxia in car seats (covered in infant-choking-in-car-seat).\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty quilted puffer jacket shape in minimal outline against a pale grey-blue background, flat vector illustration, no child, no car seat, no crash imagery."},"canonical_url":"https://likelier.app/puffer-jacket-car-seat","api_url":"https://likelier.app/api/fears/puffer-jacket-car-seat.json"},{"slug":"reentering-space-debris-hits-property","question":"What are the odds that reentering space debris damages your property?","quick_answer":"There is no meaningful probability to give. In 50 years of debris reaching the ground, NASA has confirmed no serious injury or property damage in the US and publishes no per-property figure, so we won't invent one. The intuition that the risk is essentially zero is roughly correct.\n","category":"other","tags":[],"no_reliable_estimate":true,"perceived":{"description":"Space debris reentry has a low but non-zero cultural presence -- it appears occasionally in news cycles when large objects like Chinese rocket stages or defunct weather satellites make uncontrolled reentries. Most people would describe the risk as \"essentially zero\" while acknowledging it has happened in documented cases. This is an entry at the extreme rarest end of the probability scale, useful as a calibration anchor rather than a practical concern.\n","rough_estimate":"Nearly zero -- most people have an intuition that is roughly correct: no property damage from reentering debris has ever been confirmed in the United States","kind":"intuition"},"sources":[{"url":"https://orbitaldebris.jsc.nasa.gov/faq/","title":"Orbital Debris Frequently Asked Questions","publisher":"NASA Orbital Debris Program Office, Johnson Space Center","source_type":"govt_report","statistic":"Over the past 50 years an average of one cataloged piece of debris fell back to Earth each day, yet no serious injury or significant property damage caused by reentering debris has been confirmed; most surviving components fall into oceans or sparsely populated regions","excerpt":"\"A significant amount of debris does not survive the severe heating that occurs during reentry. Components which do survive are most likely to fall into the oceans or other bodies of water or onto sparsely populated regions like the Canadian Tundra, the Australian Outback, or Siberia in the Russian Federation. During the past 50 years an average of one cataloged piece of debris fell back to Earth each day. No serious injury or significant property damage caused by reentering debris has been confirmed.\"\n","source_date":"2023-01-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260525095942/https://orbitaldebris.jsc.nasa.gov/faq/","calculation_notes":"The FAQ establishes the central empirical fact behind this entry's extreme rarity: across 50 years of roughly one cataloged debris piece reaching the ground per day, NASA reports no confirmed serious injury or significant property damage. Most surviving components land in oceans or sparsely populated regions, so the human/property target fraction is tiny. NASA's FAQ states no numeric per-person or per-property probability; the only per-event individual figures available (1 in 3,200 someone struck; ~1 in 21 trillion for a specific person) come from the 2011 UARS reentry communications cited in the second source and refer to personal injury, not property damage. No cited source states a property-damage probability, and there is no defensible way to convert the per-event personal-injury odds into a lifetime property-damage rate without inventing a building-footprint multiplier — hence this entry is marked no_reliable_estimate.\n","independence_note":"NASA's Orbital Debris Program Office is the authoritative US source for reentry risk assessments. This FAQ is the office's standing public statement on reentry survivability and the historical record of confirmed injury/damage, independent of any single event.\n"},{"url":"https://www.livescience.com/33511-falling-nasa-satellite-uars-risk.html","title":"What Are the Odds You'll Get Struck by a Falling Satellite?","publisher":"Live Science (quoting NASA scientist Mark Matney)","source_type":"reputable_reference","statistic":"According to NASA scientist Mark Matney, the odds that pieces of a specific falling satellite will strike someone among all people on Earth is 1 in 3,200; individual per-person odds are roughly 1 in several trillion","excerpt":"\"According to Mark Matney, a scientist in the Orbital Debris Program Office at NASA's Johnson Space Center in Houston, the odds that any of the 7 billion people on Earth will be struck by a piece of a soon-to-fall satellite is 1 in 3,200. The odds that you as an individual will be hit are about 1 in 21 trillion.\"\n","source_date":"2011-09-20","source_accessed":"2026-05-14","archive_url":"http://web.archive.org/web/20260312230138/https://www.livescience.com/33511-falling-nasa-satellite-uars-risk.html","calculation_notes":"Used to corroborate the NASA FAQ per-event risk framework. The article clarifies the commonly misunderstood distinction between \"probability that SOMEONE is hit\" (1 in 3,200 per event) and \"probability that I specifically am hit\" (1 in 21 trillion per event). This distinction is central to the caveats section.\n","independence_note":"Live Science independently reported on NASA's official risk communication for the 2011 UARS reentry, quoting NASA personnel directly. It represents independent journalism corroborating the official NASA figures.\n"},{"url":"https://www.forbes.com/sites/jonathanocallaghan/2020/05/12/parts-of-a-chinese-rocket-may-have-fallen-on-an-african-village/","title":"Parts Of A Chinese Rocket May Have Fallen On An African Village","publisher":"Forbes (Jonathan O'Callaghan)","source_type":"news_article","statistic":"A roughly 12-meter pipe consistent with the Long March 5B rocket's core stage fell on villages in Ivory Coast on May 11, 2020, damaging a building in Mahounou and a house in N'guinou; China never confirmed the debris's origin","excerpt":"\"The debris has not yet been confirmed as originating from the rocket, nor have China taken responsibility for the debris,\" though the pieces were found under the rocket's predicted reentry track and Harvard astronomer Jonathan McDowell concluded they were \"very likely parts of the Chinese rocket stage.\" Local reports said \"the pipe fell on a cheesemaker's building in the Mahounou village\" and \"a house in a village called N'guinou had also been damaged.\"\n","source_date":"2020-05-12","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251211045534/https://www.forbes.com/sites/jonathanocallaghan/2020/05/12/parts-of-a-chinese-rocket-may-have-fallen-on-an-african-village/","calculation_notes":"Documents a real-world example of reentering rocket debris reaching the ground near populated areas and damaging structures, supporting the body text's \"not exactly zero\" framing. Not used in the entry's numeric calculation, which rests on the NASA and Live Science sources above.\n","independence_note":"Independent contemporaneous journalism (Forbes) on the May 2020 Long March 5B uncontrolled reentry, corroborated by astronomer Jonathan McDowell's trajectory analysis; independent of the NASA Orbital Debris Program Office FAQ cited above.\n"},{"url":"https://www.law.cornell.edu/cfr/text/14/450.101","title":"14 CFR Sec. 450.101 -- Safety criteria (launch and reentry risk limits)","publisher":"Legal Information Institute, Cornell Law School (Code of Federal Regulations, Title 14, Part 450)","source_type":"reputable_reference","statistic":"FAA-licensed launches and reentries must keep collective public risk to an expected number of 1 x 10^-4 casualties (1 in 10,000) or fewer, with an individual risk limit of 1 x 10^-6 per event","excerpt":"\"The risk to all members of the public, excluding persons in aircraft and neighboring operations personnel, must not exceed an expected number of 1 x 10-4 casualties\" (Sec. 450.101(a)(1)(i), launch); the same collective- and individual-risk criteria apply to reentry operations under paragraph (b)(1)(i) and (b)(2)(i) of the same section.\n","source_date":"2020-12-10","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251206222434/https://www.law.cornell.edu/cfr/text/14/450.101","calculation_notes":"Grounds the body text's claim that FAA licensing caps the expected casualties from a commercial reentry at 1 in 10,000. Not used in the entry's probability calculation; cited as independent regulatory corroboration that per-event reentry risk is deliberately kept far below the historical 1-in-3,200 someone-is-struck figure discussed elsewhere in this entry.\n","independence_note":"FAA Part 450 licensing risk criteria (finalized in the December 2020 \"Streamlined Launch and Reentry License Requirements\" rulemaking) are a distinct regulatory data source, independent of the NASA Orbital Debris Program Office FAQ and the Live Science / Matney reporting used elsewhere in this entry.\n"},{"url":"https://www.space.com/spacex-debris-fall-australia-crew-1-dragon-capsule","title":"Possible SpaceX debris falls in Australia from Crew-1 Dragon spacecraft","publisher":"Space.com","source_type":"news_article","statistic":"Charred debris found on a sheep farm near Dalgety, New South Wales in late July 2022 was confirmed likely to be part of a SpaceX Crew-1 Dragon capsule's unpressurized trunk, which fell to Earth around July 8-9, 2022","excerpt":"\"an unpressurized capsule 'trunk' that fell to Earth around July 8\"\n","source_date":"2022-08-06","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260514013014/https://www.space.com/spacex-debris-fall-australia-crew-1-dragon-capsule","calculation_notes":"Documents a second real-world example (alongside the 2020 Ivory Coast Long March 5B incident) of confirmed reentry debris reaching the ground near or on private property, supporting the body text's and caveats' \"SpaceX debris in Australia 2022\" reference. Not used in the entry's numeric calculation, which rests on the NASA and Live Science sources above.\n","independence_note":"Independent contemporaneous space-journalism reporting on the July 2022 Dragon trunk debris event, corroborated by the Australian Space Agency and NASA statements; independent of the NASA Orbital Debris Program Office FAQ and the 2020 Forbes report on the Ivory Coast incident cited elsewhere in this entry.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"No strong personal modifiers apply","multiplier":1,"notes":"Space debris falls pseudo-randomly over Earth's surface, weighted by orbital inclination patterns. No personal behavior substantially changes the risk. Living under high-inclination orbits (high latitudes) may marginally increase or decrease exposure, but the effect is small relative to the baseline uncertainty."}],"short_label":"Space debris hits property","myth_framing":"calibrated","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This entry carries no reliable lifetime property-damage probability. The available authoritative figures all describe PERSONAL INJURY per reentry EVENT, not property damage over a lifetime. The 1 in 3,200 statistic that circulates widely in media coverage refers to the probability that a SINGLE specific reentry event will strike SOMEONE among all people on Earth -- not the probability for any one individual, and not the probability that a specific building is hit. The individual per-event injury probability is approximately 1 in 21 trillion (NASA, 2011 UARS reentry). There is no cited source that states a property-damage rate, and converting the injury odds into a property figure would require an invented building-footprint multiplier, so no headline number is published here. With the rapid expansion of satellite mega-constellations (Starlink, OneWeb) and thousands of deorbit events expected per year in the coming decade, the aggregate annual reentry rate could increase substantially. No confirmed serious property damage from reentering debris has occurred in the United States, though property-damage incidents have occurred elsewhere (SpaceX debris in Australia 2022; suspected Long March 5B rocket debris in Ivory Coast 2020). This entry is calibration-oriented -- its value is placing \"space debris\" at the far right, rarest end of the probability scale.\n","quality_score":{"d1":4,"d2":4,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":4,"avg":4.125,"scored_by":"extracted-from-transcript","scored_at":"2026-05-16","methodology_version":"1.0"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-14","image":{"alt":"A small glowing fragment descending toward a house roofline, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/reentering-space-debris-hits-property","api_url":"https://likelier.app/api/fears/reentering-space-debris-hits-property.json"},{"slug":"refreezing-thawed-food","question":"Is it dangerous to refreeze food that has already thawed or melted?","category":"health","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"\"Never refreeze thawed food\" is one of the stickiest kitchen rules, repeated as though the freeze-thaw-refreeze cycle itself breeds something dangerous. It doesn't. The US Department of Agriculture's position is explicit: food thawed in the refrigerator can be refrozen without cooking, at the cost of some quality. What actually determines safety is time and temperature — whether the food sat in the 40-140°F \"Danger Zone\" long enough for bacteria to multiply — and that is true whether or not you ever intended to refreeze it. The rule people remember is a garbled proxy for a real rule (the 2-hour limit) attached to the wrong variable (the refreeze). No survey isolates this specific belief, so the perceived side is editorial intuition rather than polled data. Melted ice cream is the one everyday exception where the folk caution and the guidance agree: discard it.\n","rough_estimate":"Many people believe refreezing thawed food is inherently unsafe and will cause food poisoning","kind":"intuition"},"sources":[{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/freezing-and-food-safety","title":"Freezing and Food Safety","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Food thawed in the refrigerator is safe to refreeze without cooking, with possible quality loss from moisture; do not refreeze foods left out longer than 2 hours (1 hour above 90°F)","excerpt":"\"Once food is thawed in the refrigerator, it is safe to refreeze it without cooking, although there may be a loss of quality due to the moisture lost through thawing. After cooking raw foods which were previously frozen, it is safe to freeze the cooked foods. If previously cooked foods are thawed in the refrigerator, you may refreeze the unused portion. Freeze leftovers within 3-4 days. Do not refreeze any foods left outside the refrigerator longer than 2 hours; 1 hour in temperatures above 90 °F.\"\n","source_date":"2024-08-09","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260627191529/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/freezing-and-food-safety","calculation_notes":"This is the governing US statement on refreezing. It establishes the core claim of the entry — refreezing refrigerator-thawed food is safe (quality loss only) — and immediately re-anchors the actual safety condition to the 2-hour Danger-Zone limit, not the refreeze. Used as the definition of the safe/unsafe boundary, not as a probability estimate. The same page adds, for power outages: \"If food is partly frozen, still has ice crystals, or is as cold as if it were in a refrigerator (40 °F), it is safe to refreeze or use,\" and \"Dispose of soft or melted ice cream for quality's sake.\"\n","independence_note":"FSIS guidance is the upstream US regulatory standard reused by FDA, CDC, and state health departments; it is not independent of those downstream restatements but precedes them.\n"},{"url":"https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/danger-zone-40f-140f","title":"\"Danger Zone\" (40 °F - 140 °F)","publisher":"US Department of Agriculture, Food Safety and Inspection Service","source_type":"govt_report","statistic":"Bacteria double in as little as 20 minutes between 40°F and 140°F; food should not be left out more than 2 hours (1 hour above 90°F)","excerpt":"\"Bacteria grow most rapidly in the range of temperatures between 40 °F and 140 °F, doubling in number in as little as 20 minutes. This range of temperatures is often called the 'Danger Zone.' Never leave food out of refrigeration over 2 hours. If the temperature is above 90 °F, food should not be left out more than 1 hour.\"\n","source_date":"2024-07-12","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260627191528/https://www.fsis.usda.gov/food-safety/safe-food-handling-and-preparation/food-safety-basics/danger-zone-40f-140f","calculation_notes":"Establishes the variable that actually governs safety — cumulative time in the Danger Zone — against which the refreeze question is beside the point. Two freezings with an interval spent below 40°F carry no time-temperature penalty; a single thaw that spent three hours on a warm counter is unsafe whether or not it is refrozen.\n","independence_note":"Same regulatory author as the Freezing page; cited for a distinct claim (the exposure rule), not as corroboration of the refreezing claim.\n"},{"url":"https://www.udel.edu/academics/colleges/canr/cooperative-extension/fact-sheets/when-the-power-is-out/","title":"When the Power is Out: Food Safety","publisher":"University of Delaware Cooperative Extension","source_type":"reputable_reference","statistic":"Most foods may be safely refrozen if they still contain ice crystals or have been kept at 40°F or below for no more than 24-48 hours; melted ice cream should not be refrozen","excerpt":"\"You may safely refreeze most foods if they still contain ice crystals, or if they have been kept cold (40°F or less) and have been thawed no more than 24 to 48 hours. … Do not use or refreeze melted ice cream.\"\n","source_date":"2023-01-01","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260304083806/https://www.udel.edu/academics/colleges/canr/cooperative-extension/fact-sheets/when-the-power-is-out/","calculation_notes":"Extension-service restatement of the FSIS rule that also makes the ice-cream exception explicit, which the FSIS page frames only as a quality discard. Supports the entry's two-part structure: general refreezing is safe under the ice-crystal / ≤40°F condition; melted ice cream is the everyday case where discard is advised.\n","independence_note":"A land-grant extension service applying FSIS guidance to the consumer power-outage scenario. Downstream of FSIS on the general rule; its explicit ice-cream instruction is the clearest consumer-facing statement of that specific carve-out.\n"},{"url":"https://www.cdc.gov/food-safety/data-research/facts-stats/index.html","title":"Foodborne Illness: Facts and Statistics","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"CDC estimates ~48 million foodborne illnesses in the US each year; 128,000 hospitalizations; 3,000 deaths","excerpt":"\"CDC estimates that 48 million people get sick from a foodborne illness every year. Of those, 128,000 are hospitalized and 3,000 die.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260618000252/https://www.cdc.gov/food-safety/data-research/facts-stats/index.html","calculation_notes":"Baseline context only. Foodborne illness is common (48 million cases against a US population of ~330 million is roughly 1 in 7 residents per year), but the surveillance attributes it to pathogens and contributing factors such as time-temperature abuse and cross-contamination, not to refreezing, which appears in no contributing-factor taxonomy. Included to show that the real risk this fear gestures at is genuine and quantified, while the refreeze mechanism it names is not.\n","independence_note":"CDC national burden surveillance, methodologically independent of the FSIS handling guidance; cited for the population-level foodborne-illness baseline, not for any refreezing-specific claim.\n"},{"url":"https://archive.cdc.gov/www_cdc_gov/listeria/outbreaks/ice-cream-03-15/index.html","title":"Multistate Outbreak of Listeriosis Linked to Blue Bell Creameries Products (Final Update)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"2010-2015 listeriosis outbreak linked to factory-contaminated ice cream: 10 cases across 4 states, all 10 hospitalized, 3 deaths","excerpt":"\"A total of 10 people with listeriosis related to this outbreak were reported from 4 states. … All ill people were hospitalized. Three deaths were reported from Kansas. … Listeriosis is a life-threatening infection caused by eating food contaminated with the bacterium (germ) Listeria monocytogenes (Listeria). People at high risk for listeriosis include pregnant women and their newborns, adults 65 and older, and people with weakened immune systems.\"\n","source_date":"2015-06-10","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260604054753/https://archive.cdc.gov/www_cdc_gov/listeria/outbreaks/ice-cream-03-15/index.html","calculation_notes":"Supports the body's ice-cream carve-out with the outbreak counts and the high-risk groups (both verbatim on this page). This outbreak was FACTORY contamination, not a consumer refreezing failure — cited to establish that dairy ice cream is a documented, severe Listeria vehicle, NOT as evidence that refreezing melted ice cream at home caused illness. The separate claim that Listeria grows at refrigerator temperatures is sourced to the FDA Listeria page below, not to this outbreak report.\n","independence_note":"CDC outbreak surveillance, independent of the FSIS handling guidance and the CDC national-burden page; cited only for the 2015 outbreak counts and high-risk groups.\n"},{"url":"https://www.fda.gov/food/foodborne-pathogens/listeria-listeriosis","title":"Listeria (Listeriosis)","publisher":"US Food and Drug Administration","source_type":"govt_report","statistic":"Listeria monocytogenes can survive and grow even under refrigeration","excerpt":"\"L. monocytogenes is a type of disease-causing bacteria that can be found in many places, including soil, water, sewage, rotting vegetation, and animals. It can survive and grow even under refrigeration.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-12","archive_url":"http://web.archive.org/web/20260710190642/https://www.fda.gov/food/foodborne-pathogens/listeria-listeriosis","calculation_notes":"Sources the specific biological claim the body uses to justify the ice-cream carve-out: that dairy which has warmed and been re-chilled is a Listeria concern because, unusually among foodborne pathogens, Listeria keeps growing at fridge temperatures. Cited only for that mechanism, not for any probability — consistent with no_reliable_estimate.\n","independence_note":"FDA pathogen reference, methodologically independent of the CDC outbreak report; supplies the refrigerator-growth biology that the Blue Bell page does not state.\n"}],"comparison_anchors":[],"short_label":"Refreezing thawed food","outcome_severity":"minor_harm","exposure_pattern":"acute","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is about the safety of the refreeze act itself, and the answer is that it is not the relevant variable — so no probability is assigned (no_reliable_estimate). It is not a claim that thawed food is always safe: food that has spent more than 2 hours in the 40-140°F Danger Zone (1 hour above 90°F) should be discarded whether or not it is refrozen, and that time-temperature exposure — quantified on the food-left-unrefrigerated entry — is where the real risk lives. Three important carve-outs to the \"refreezing is fine\" rule: (1) it applies to food thawed in the refrigerator or still holding ice crystals, not food thawed by cold water or microwave, which the USDA says to cook before refreezing; (2) each freeze-thaw cycle degrades texture and moisture, so quality falls even when safety holds; (3) melted ice cream and other soft-serve dairy are the common exception where guidance is to discard, partly for quality and partly because dairy is a documented Listeria vehicle and Listeria can grow at refrigerator temperatures. The discard thresholds carry more weight for high-risk groups — pregnant people, adults 65+, and the immunocompromised are the populations CDC identifies as most vulnerable to listeriosis.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-07-12","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-12","last_reviewed":"2026-07-12","reviewed":true,"generated_at":"2026-07-12","image":{"alt":"A simplified snowflake and a partially melted ice-cube shape side by side in muted blue-grey tones on a pale background, flat vector illustration."},"canonical_url":"https://likelier.app/refreezing-thawed-food","api_url":"https://likelier.app/api/fears/refreezing-thawed-food.json"},{"slug":"screen-time-school-age-harm","question":"What are the odds of harm from screen time in children ages 4-12?","quick_answer":"There is no reliable per-child probability of harm: the research finds only small correlations between screen time and behavior problems, explaining barely 1% of the variance, not a countable harm event. Most parents rate excessive screen time a top-three health concern, but the studies rarely support the catastrophic framing.\n","category":"kids","tags":["child"],"no_reliable_estimate":true,"perceived":{"description":"Few parental anxieties have scaled as quickly as screen time worry. Since the American Academy of Pediatrics began issuing screen limits in the late 1990s, the assumption that tablets and televisions are eroding childhood cognition, attention, and emotional health has become a default setting in educated households. Headlines about \"digital heroin\" and \"destroyed attention spans\" circulate faster than the underlying studies, and the studies themselves rarely support the catastrophic framing. Most parents of school-age children rate excessive screen time as a top-three health concern, alongside obesity and bullying, and estimate the risk of meaningful developmental harm as moderate to high.\n","rough_estimate":"Parents widely perceive screen time as a significant threat to child development; the research finds only small, correlational associations that explain little variance in outcomes","kind":"intuition"},"sources":[{"url":"https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2790338","title":"Association of Screen Time With Internalizing and Externalizing Behavior Problems in Children 12 Years or Younger: A Systematic Review and Meta-analysis","publisher":"JAMA Psychiatry (Eirich et al. 2022)","source_type":"peer_reviewed","statistic":"A meta-analysis of 87 studies (159,425 participants, mean age 6.07 years) found small but significant correlations between screen time duration and externalizing problems (r = 0.11; 95% CI 0.10-0.12) and internalizing problems (r = 0.07; 95% CI 0.05-0.08) in children 12 or younger","excerpt":"\"Increased duration of screen time was significantly but weakly correlated with more externalizing problems and more internalizing problems among children 12 years or younger.\"\n","source_date":"2022-05-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260227063659/https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2790338","calculation_notes":"Eirich et al. 2022 is the largest meta-analysis focused specifically on the age group relevant to this entry (children 12 and under). The pooled correlations of r = 0.11 (externalizing) and r = 0.07 (internalizing) correspond to shared variance of roughly 1.2% and 0.5% respectively. These are statistically significant in a pool of 159,000 children but trivially small in practical terms. The authors note stronger associations for boys with externalizing problems. Most included studies were cross-sectional, precluding causal inference. No native or normalized probability is derived because the entry is flagged no_reliable_estimate — the outcomes are continuous behavioral scales, not binary harm events.\n"},{"url":"https://www.thelancet.com/journals/lanchi/article/PIIS2352-4642(18)30278-5/abstract","title":"Associations between 24 hour movement behaviours and global cognition in US children: a cross-sectional observational study","publisher":"The Lancet Child & Adolescent Health (Walsh et al. 2018)","source_type":"peer_reviewed","statistic":"Among 4,524 US children aged 8-11 from the ABCD study, those meeting the Canadian 24-Hour Movement Guideline for recreational screen time (no more than 2 hours per day) scored higher on global cognition measures than those who did not; meeting only the screen time guideline alone was associated with superior cognition","excerpt":"\"Meeting individual recommendations for screen time was associated with superior global cognition compared with meeting none of the recommendations.\"\n","source_date":"2018-09-27","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20250510181230/https://www.thelancet.com/journals/lanchi/article/PIIS2352-4642(18)30278-5/abstract","calculation_notes":"Walsh et al. 2018 used data from the ABCD study, the largest long-term study of brain development in US children. The sample of 4,524 children aged 8-11 is nationally representative. The study found a dose-response pattern: meeting more movement guidelines (sleep, physical activity, screen limits) was associated with better cognitive scores, with screen time showing the most consistent individual association. However, this is a cross-sectional observation — children with higher baseline cognition may simply choose less screen time, or unmeasured confounders (parental education, household income) may drive both. The effect sizes, while statistically significant, were modest. No native or normalized probability is derived because the entry is flagged no_reliable_estimate.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11457456/","title":"Screen time and mental health: a prospective analysis of the Adolescent Brain Cognitive Development (ABCD) Study","publisher":"BMC Public Health (Nagata et al. 2024)","source_type":"peer_reviewed","statistic":"In a prospective analysis of 9,538 adolescents (analyzed out of 11,875 enrolled in the ABCD study, after excluding 2,337 for missing data) followed from ages 9-10 to 11-12, higher total screen time was associated with all subsequent mental health symptoms; the association was strongest for depressive symptoms (B = 0.10, 95% CI 0.06-0.13), and the screen types with the greatest associations were video chat, texting, videos, and video games","excerpt":"\"Higher total screen time was associated with all mental health symptoms in adjusted models, and the association was strongest for depressive (B = 0.10, 95% CI 0.06, 0.13, p < 0.001), conduct (B = 0.07, 95% CI 0.03, 0.10, p < 0.001), somatic (B = 0.06, 95% CI 0.01, 0.11, p = 0.026), and attention-deficit/hyperactivity symptoms. ... The specific screen types with the greatest associations with depressive symptoms included video chat, texting, videos, and video games.\"\n","source_date":"2024-10-07","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260311003808/https://pmc.ncbi.nlm.nih.gov/articles/PMC11457456/","calculation_notes":"Nagata et al. 2024 (BMC Public Health) adds longitudinal evidence from the ABCD study, addressing the major limitation of the cross-sectional Walsh et al. 2018 paper. The analytic sample was 9,538 adolescents (out of 11,875 enrolled; 2,337 excluded for missing data), followed from ages 9-10 to 11-12. The prospective design strengthens causal inference somewhat, but the regression coefficients are small (e.g. B = 0.10 for depressive symptoms). Higher total screen time was associated with all mental health symptoms, the strongest being depressive symptoms; the screen types with the greatest associations were video chat, texting, videos, and video games — i.e. the signal is not confined to passive video consumption. No native or normalized probability is derived.\n","independence_note":"Uses the same ABCD cohort as Walsh et al. 2018 but at later time points and with a prospective design examining mental health rather than cognition.\n"}],"comparison_anchors":[],"short_label":"Kids' screens 4-12","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is flagged no_reliable_estimate for several reinforcing reasons.\nFirst, \"harm from screen time\" is not a binary, measurable outcome. The literature measures dozens of continuous variables — externalizing behavior, internalizing behavior, cognitive performance, sleep quality, BMI — on different scales with different thresholds. There is no agreed-upon definition of what constitutes \"harm\" as a discrete event, which means there is no denominator from which to derive a probability.\nSecond, the effect sizes are very small. The largest meta-analysis of children 12 and under (Eirich et al. 2022, 87 studies, 159,425 participants) found correlations of r = 0.07 to r = 0.11 between screen time duration and behavioral problems. These correlations explain roughly 0.5-1.2% of the variance in outcomes. The remaining 98-99% is attributable to genetics, parenting, socioeconomic status, peer relationships, and unmeasured factors.\nThird, nearly all evidence is correlational, and confounding is severe. Families with lower income, less parental education, and more household chaos tend to have children with both higher screen time and worse developmental outcomes. Reverse causation is equally plausible — children with existing behavioral difficulties may be given screens more often as a management strategy. The few longitudinal studies (Nagata et al. 2024) find associations that survive adjustment for confounders, but they remain small.\nFourth, \"screen time\" is not a single exposure. Educational apps, video calls with grandparents, passive YouTube consumption, and violent video games are collapsed into one variable in most studies. The emerging consensus is that content and context matter far more than raw minutes, which makes any aggregate risk estimate misleading.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A muted flat vector illustration of a tablet lying face-down on a wooden table beside an untouched stack of building blocks, in subdued blue and warm grey tones."},"canonical_url":"https://likelier.app/screen-time-school-age-harm","api_url":"https://likelier.app/api/fears/screen-time-school-age-harm.json"},{"slug":"screen-time-teen-harm","question":"What are the odds of serious harm to a teenager from screen time?","quick_answer":"No meaningful per-teen probability of serious harm can be derived: the largest analyses find only tiny associations, explaining at most 0.4% of well-being variance, with sleep disruption the most consistent correlate. Most parents treat teen screens as a clear hazard, but the population evidence is far thinner than the headlines.\n","category":"kids","tags":["teen","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"Ask a room of parents whether screens are harming their teenager and you will get near-unanimous alarm. The framing has shifted from \"too much TV\" to \"too much phone,\" but the underlying anxiety is older than the iPhone. Jean Twenge's iGen (2017) and Jonathan Haidt's The Anxious Generation (2024) gave the worry a scholarly veneer and a bestseller spine. News coverage routinely links rising teen depression rates to rising screen time, treating the temporal correlation as self-evidently causal. What gets lost in the discourse is how broad \"screen time\" actually is: it includes passive Netflix bingeing, competitive gaming, video calls with friends, homework on a laptop, and doomscrolling TikTok — all lumped into one exposure variable in most studies. The intuition that screens harm teens is widespread; the evidence that they do so at a population level, through a mechanism other than sleep displacement, is considerably thinner than the headlines suggest.\n","rough_estimate":"Most parents treat teen screen time as a clear developmental hazard; the research base supports only small, heterogeneous associations, with sleep disruption the most consistently documented correlate","kind":"intuition"},"sources":[{"url":"https://www.nature.com/articles/s41562-018-0506-1","title":"The association between adolescent well-being and digital technology use","publisher":"Nature Human Behaviour (Orben & Przybylski 2019)","source_type":"peer_reviewed","statistic":"Specification-curve analysis across three large datasets (N > 355,000 adolescents) found that all digital technology use — including TV, gaming, smartphones, and social media — was negatively associated with well-being at r = −0.035 to −0.04, explaining at most 0.4% of the variance; the effect was smaller than that of wearing glasses or eating potatoes","excerpt":"\"The negative effect of technology use on adolescent well-being is small — explaining at most 0.4% of the variation in well-being.\"\n","source_date":"2019-01-14","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260521221648/https://www.nature.com/articles/s41562-018-0506-1","calculation_notes":"Orben & Przybylski 2019 is the most rigorous large-scale analysis of the association between digital technology use (broadly defined: TV, gaming, smartphones, social media) and adolescent well-being. The specification-curve approach tested all defensible analytic choices across three datasets (Understanding Society, YRBS, MCS). The headline finding — r ≈ −0.04 for all technology use combined — is statistically significant given the sample size but practically negligible. Critically, this study measures total digital technology use, not just social media, making it the primary anchor for this entry's broader \"screen time\" framing. The outcome is a continuous well-being scale, not a binary harm threshold, so no per-teen probability of serious harm can be derived. This source is the primary basis for the no_reliable_estimate designation.\n","independence_note":"Uses UK Understanding Society, YRBS, and MCS datasets. Fully independent of the Hale & Guan sleep review and the Paulus et al. neuroscience review.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC4437561/","title":"Screen Time and Sleep among School-Aged Children and Adolescents: A Systematic Literature Review","publisher":"Sleep Medicine Reviews (Hale & Guan 2015)","source_type":"peer_reviewed","statistic":"A systematic review of 67 studies found that screen time — across TV, computers, video games, and mobile phones — was adversely associated with sleep outcomes in 90% of studies examined, with the most consistent finding being shortened sleep duration and delayed sleep onset among adolescents","excerpt":"\"In 90% of studies, screen time was adversely associated with sleep outcomes — primarily shortened duration and delayed timing.\"\n","source_date":"2015-06-01","source_accessed":"2026-04-26","archive_url":"https://web.archive.org/web/20260426210939/https://pmc.ncbi.nlm.nih.gov/articles/PMC4437561/","calculation_notes":"Hale & Guan 2015 reviewed 67 studies (1999-2014) covering all major screen types — TV, computers, video games, and mobile devices — in school-aged children and adolescents. The 90% figure refers to the proportion of studies finding a negative association between screen time and at least one sleep outcome. The review is included because it establishes sleep disruption as the most consistently documented mechanism linking screen time (broadly, not just social media) to adverse outcomes in teens. However, the review is correlational, and the authors note that causal direction is unconfirmed: teens who sleep poorly may also use screens more. No per-teen probability of harm can be derived from a systematic review of heterogeneous correlational studies with varying outcome definitions.\n","independence_note":"Independent review of 67 primary studies. No overlap with Orben & Przybylski 2019 datasets. Different outcome domain (sleep vs well-being).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10624397/","title":"Screen media activity in youth: A critical review of mental health and neuroscience findings","publisher":"Journal of Mood and Anxiety Disorders (Paulus, Zhao, Potenza et al. 2023)","source_type":"peer_reviewed","statistic":"A critical review of mental health and neuroscience evidence found that screen media activity has both immediate and long-term associations with sleep quality, mood, and anxiety in youth, but that effects are moderated by content type, context, and individual vulnerability, and the relationship is best understood within an ecological systems framework rather than as a simple dose-response toxin model","excerpt":"\"The multifaceted relationship between SMA and various aspects of adolescent life is influenced by a wide range of environmental and contextual factors.\"\n","source_date":"2023-08-01","source_accessed":"2026-04-26","archive_url":"http://web.archive.org/web/20260119030032/https://pmc.ncbi.nlm.nih.gov/articles/PMC10624397/","calculation_notes":"Paulus et al. 2023 is included as the most comprehensive recent review that covers all screen media activity in youth — not just social media — integrating both mental health and neuroscience findings. The review explicitly adopts Bronfenbrenner's ecological systems framework, which positions screen time within a web of individual, family, school, peer, and environmental factors. Its key contribution to this entry is the finding that the screen-time-to-harm pathway is not a simple dose-response relationship: content type (passive vs interactive), context (solitary vs social), and pre-existing vulnerability all moderate effects substantially. This supports the no_reliable_estimate designation — a single population- level probability cannot capture such heterogeneity. No probability is derived.\n"}],"comparison_anchors":[],"short_label":"Teen screens","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the research literature on teen screen time and health outcomes does not support a defensible per-teen probability of \"serious harm.\" The reasons are structural.\nFirst, \"screen time\" is not a single exposure. An hour of competitive gaming, an hour of passive streaming, an hour of video-calling a friend, and an hour of homework research are collapsed into the same variable in most large-scale studies. The few analyses that disaggregate by activity type find meaningfully different associations, which means any population-average effect size is an artefact of the activity mix in the sample rather than a property of screens per se.\nSecond, the best-powered study (Orben & Przybylski 2019, N > 355,000) finds associations in the range of r = −0.04 for total technology use. That explains roughly 0.16% of the variance in well-being — real but far too small and heterogeneous to convert into an individual risk figure.\nThird, the most consistently documented correlate of screen time is sleep disruption. Hale & Guan's 2015 review found adverse sleep associations in 90% of 67 studies. But this raises an attribution problem: if screens harm teens primarily by displacing sleep, then the actionable risk factor is evening screen use that cuts into sleep time, not \"screen time\" as a category. A teen who games for three hours on a Saturday afternoon and sleeps nine hours that night is in a different risk category from one who scrolls in bed until 2 a.m. — but both register as \"high screen time\" in most datasets.\nFourth, this entry is distinct from social-media-teen-harm (which covers social media platforms specifically and their comparison/cyberbullying mechanisms) and screen-time-child-harm (which covers children under 12 and developmental delay). The overlap is limited: this entry addresses the broader category of all screen-based activity in the 13-18 age range, where the dominant pathway to harm runs through sleep rather than through social comparison or developmental disruption.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A laptop, game controller, and phone arranged on a desk beside an alarm clock showing a late hour, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/screen-time-teen-harm","api_url":"https://likelier.app/api/fears/screen-time-teen-harm.json"},{"slug":"screen-time-toddler-harm","question":"What are the odds of developmental harm from screen time in toddlers (ages 0-3)?","quick_answer":"No binary probability of developmental harm can be derived: the evidence is small correlational effects, a few percent of variance at most, and the direction flips positive for educational or co-viewed content. Most parents of toddlers believe screens risk meaningful harm, but the confounded data cannot separate that out.\n","category":"kids","tags":["toddler"],"no_reliable_estimate":true,"perceived":{"description":"Parents of children under three are the most anxious cohort in the screen-time debate. Pediatric guidelines from the WHO and AAP recommend zero screen time before 18-24 months, and social media parenting communities treat any tablet exposure in infancy as a developmental emergency. The fear is sharpened by the age group's neuroplasticity: if screens can harm anyone, the reasoning goes, surely they harm the youngest brains the most. Surveys consistently find that parents of toddlers report higher guilt and worry about screen use than parents of older children.\n","rough_estimate":"Most parents of toddlers believe screen exposure risks meaningful developmental harm; the evidence supports only small, associational effects that cannot be separated from confounders","kind":"intuition"},"sources":[{"url":"https://doi.org/10.1001/jamapediatrics.2020.0327","title":"Associations Between Screen Use and Child Language Skills: A Systematic Review and Meta-analysis","publisher":"JAMA Pediatrics (Madigan et al. 2020)","source_type":"peer_reviewed","statistic":"A meta-analysis of 42 studies (18,905 children aged 12 months to 12 years) found that greater quantity of screen time was associated with lower language skills, with a pooled effect size of r = −0.14 (95% CI, −0.18 to −0.10); background television showed a stronger negative association (r = −0.19)","excerpt":"\"Greater quantity of screen use was negatively associated with child language skills (r = −0.14; 95% CI, −0.18 to −0.10). Better quality of screen use was positively associated with child language skills, including educational programs (r = 0.13; 95% CI, 0.02 to 0.24) and co-viewing (r = 0.16; 95% CI, 0.07 to 0.24).\"\n","source_date":"2020-04-01","source_accessed":"2026-04-26","calculation_notes":"Madigan et al. 2020 is a meta-analysis specifically examining screen time and language skills in children, with many included studies focusing on the 0-3 age range where language acquisition is most rapid. The pooled r = −0.14 for screen time quantity translates to roughly 2% of variance explained — a small effect by any standard. Critically, educational content showed a positive association (r = 0.13) and co-viewing showed an even stronger positive association (r = 0.16), meaning the direction and magnitude of the effect depends entirely on what is being watched and how. Background television — the passive ambient exposure most common in households with toddlers — showed the strongest negative association (r = −0.19). No binary probability of harm can be derived from these correlational effect sizes.\n"},{"url":"https://doi.org/10.1001/jamapediatrics.2024.2620","title":"Early Childhood Screen Use Contexts and Cognitive and Psychosocial Outcomes: A Systematic Review and Meta-analysis","publisher":"JAMA Pediatrics (Mallawaarachchi et al. 2024)","source_type":"peer_reviewed","statistic":"A meta-analysis of 100 studies (176,742 participants aged 0-5.99 years) found that more program viewing and background television were associated with poorer cognitive outcomes, with effect sizes ranging from r = 0.03 to r = 0.16 depending on screen use context","excerpt":"\"More program viewing and background television were associated with poorer cognitive outcomes. Although effect sizes were generally small (r = 0.03-0.16), they were comparable to previous meta-analyses on screen time and developmental outcomes.\"\n","source_date":"2024-08-05","source_accessed":"2026-04-26","calculation_notes":"Mallawaarachchi et al. 2024 is the largest meta-analysis to date examining screen use contexts (not just total duration) and developmental outcomes in early childhood. The inclusion of 176,742 participants across 100 studies gives it substantial statistical power. The key finding for toddlers is that context matters more than quantity: the same amount of screen time produces different effect sizes depending on whether it involves background TV, active program viewing, educational content, or co-use with a caregiver. Effect sizes across all contexts remained small (r = 0.03-0.16), reinforcing the pattern from earlier meta-analyses that screen time associations with developmental outcomes are real but tiny. No probability of harm is derived because the study reports continuous associations, not binary harm thresholds.\n","independence_note":"Independent from Madigan et al. 2020: different research team (Flinders University vs University of Calgary), broader outcome scope (cognitive and psychosocial vs language only), different inclusion period (through 2023 vs through 2019), and minimal overlap in included primary studies.\n"},{"url":"https://jamanetwork.com/journals/jamapediatrics/fullarticle/2815514","title":"Screen Time and Parent-Child Talk When Children Are Aged 12 to 36 Months","publisher":"JAMA Pediatrics (Brushe et al. 2024)","source_type":"peer_reviewed","statistic":"In 220 Australian families recorded with LENA wearable audio devices at 12-36 months, each additional minute of screen time at age 3 was associated with a reduction of 6.6 adult words (95% CI, -11.7 to -1.5), 4.9 child vocalizations (95% CI, -6.1 to -3.7), and 1.1 conversational turns (95% CI, -1.4 to -0.8)","excerpt":"\"At 36 months, an additional minute of screen time was associated with a reduction of 6.6 (95% CI, -11.7 to -1.5) adult words, 4.9 (95% CI, -6.1 to -3.7) child vocalizations, and 1.1 (95% CI, -1.4 to -0.8) conversational turns.\"\n","source_date":"2024-03-04","source_accessed":"2026-06-14","archive_url":"https://web.archive.org/web/20260609215512/https://jamanetwork.com/journals/jamapediatrics/fullarticle/2815514","calculation_notes":"Brushe et al. 2024 (Tell Me a Story / KIDS LENA sub-study, JAMA Pediatrics 177:10) uses objective wearable audio measurement (LENA t-shirt processors recording 16 h/day) rather than parent-reported screen diaries, which strengthens the dose-response signal seen in the older meta-analyses. The reported associations are per-minute slopes on home language-environment counts, not binary harm thresholds, so no per-child probability is derived. Its value here is mechanistic: it gives direct evidence for the displacement pathway (screen minutes crowd out adult words, child vocalizations, and back-and-forth turns) that the earlier r-values could only infer. At the cohort's mean 36-month screen time of 172 min/day, the slopes imply roughly 1,139 fewer adult words, 843 fewer vocalizations, and 194 fewer conversational turns per day. This refines, but does not overturn, the no_reliable_estimate flag.\n"},{"url":"https://doi.org/10.1001/jamapediatrics.2023.3057","title":"Screen Time at Age 1 Year and Communication and Problem-Solving Developmental Delay at 2 and 4 Years","publisher":"JAMA Pediatrics (Takahashi et al. 2023)","source_type":"peer_reviewed","statistic":"In a 7,097 mother-child cohort (Tohoku Medical Megabank, Japan), screen time of 4 or more hours/day at age 1 (vs less than 1 h/day) was associated with a 4.78-fold higher odds of communication developmental delay at age 2 (OR 4.78; 95% CI, 3.24-7.06)","excerpt":"\"Children's screen time was associated with a higher risk of developmental delay at age 2 years in the communication (odds ratio [OR], 1.61 [95% CI, 1.23-2.10] for 1 to <2 h/d; 2.04 [1.52-2.74] for 2 to <4 h/d; 4.78 [3.24-7.06] for >=4 vs <1 h/d).\"\n","source_date":"2023-08-21","source_accessed":"2026-06-14","calculation_notes":"Takahashi et al. 2023 (JAMA Pediatrics 177:10, pp. 1039-1046) is a 7,097 mother-child prospective cohort from the Tohoku Medical Megabank Three-Generation Study. Screen time at age 1 was banded (<1, 1-<2, 2-<4, >=4 h/day) and developmental delay scored with the Ages & Stages Questionnaires (ASQ-3, delay = below 2 SD). The dose-response is monotonic for communication delay at age 2: OR 1.61 (1-<2 h), 2.04 (2-<4 h), and 4.78 (>=4 h) versus the <1 h reference. The top-vs-bottom OR of 4.78 is the source for the heavy-use (>=4 h/day) personal_factor_multiplier. This is a relative odds ratio on a delay screening endpoint, not a lifetime probability, so the entry remains no_reliable_estimate; the multiplier describes how heavy daily use shifts the (unquantified) baseline risk.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK541173/","title":"Guidelines on Physical Activity, Sedentary Behaviour and Sleep for Children under 5 Years of Age: Recommendations","publisher":"World Health Organization (2019)","source_type":"govt_report","statistic":"WHO's 2019 guideline recommends no screen time for infants under 1 year or for 1-year-olds, and no more than 1 hour/day of sedentary screen time for children aged 2-4; each of these three recommendations is formally graded 'Strong recommendation, very low quality evidence'","excerpt":"[Confirmed via NCBI Bookshelf hosting of the WHO guideline's Recommendations chapter] Screen time \"is not recommended\" for infants under 1 year and for 1-year-olds; for those aged 2 years, sedentary screen time \"should be no more than 1 hour; less is better.\" Each of these recommendations is rated \"Strong recommendation, very low quality evidence.\"\n","source_date":"2019-04-24","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20251104072957/https://www.ncbi.nlm.nih.gov/books/NBK541173/","calculation_notes":"The WHO's own GRADE assessment rates the evidence behind its zero/limited screen-time recommendations for children under 5 as \"very low quality,\" even though the recommendation itself is labeled \"strong\" — the guideline panel judged that plausible benefit outweighs harm despite a thin evidence base. This grounds the entry's body-text claim that the WHO guidance is precautionary rather than evidence-derived. It is a guideline-development classification, not a probability estimate, so it does not change the no_reliable_estimate flag or any effect-size figures from the other sources.\n","independence_note":"A policy/guideline document, not a primary study or meta-analysis — wholly independent of the four epidemiological sources above (Madigan, Mallawaarachchi, Brushe, Takahashi), which supply this entry's effect-size evidence.\n"}],"comparison_anchors":[],"personal_factor_multipliers":[{"factor":"heavy daily use (>=4 h/day at age 1)","multiplier":4.78,"notes":"Odds ratio for communication developmental delay at age 2 in the >=4 h/day vs <1 h/day group, 7,097-child Tohoku Medical Megabank cohort (Takahashi et al. 2023, JAMA Pediatrics). A relative odds ratio on a delay screening endpoint, not a lifetime probability — the headline entry remains no_reliable_estimate.\n"},{"factor":"moderate daily use (1-2 h/day at age 1)","multiplier":1.61,"notes":"Odds ratio for communication delay at age 2 in the 1-<2 h/day vs <1 h/day group, same Takahashi et al. 2023 cohort. The dose-response is monotonic (1.61 at 1-<2 h, 2.04 at 2-<4 h, 4.78 at >=4 h), so heavier use carries progressively higher relative odds.\n"}],"short_label":"Toddler screens","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"Short-form video (\"brainrot\" feeds: TikTok, Reels, Shorts) is a distinct modality that the broad screen-time meta-analyses here do not isolate. A 2025 review of 71 studies (~98,000 participants; Nguyen et al., Psychological Bulletin) tied heavy short-form use to poorer attention and inhibitory control, but its samples were adolescents and adults — direct evidence for under-threes is essentially absent, so no number is defensible for this age group. Separately, objective wearable-audio work (Brushe et al. 2024) points to a language-environment displacement mechanism: screen minutes measurably crowd out the adult words, child vocalizations, and back-and-forth turns that drive early language, which is the most concrete pathway by which the small associational effects could become real for an individual child. The per-minute and dose-response associations remain relative, not absolute, so the entry stays no_reliable_estimate.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-26","image":{"alt":"A small glowing tablet resting on a play mat next to stacking rings and a board book, flat vector illustration in soft muted tones."},"canonical_url":"https://likelier.app/screen-time-toddler-harm","api_url":"https://likelier.app/api/fears/screen-time-toddler-harm.json"},{"slug":"seed-oil-health-harm","question":"What are the odds seed oils (soybean, canola, sunflower, corn) cause significant health harm?","quick_answer":"No reliable probability of harm can be given: the trial evidence is contradictory, with the strongest reviews finding polyunsaturated fats modestly reduce cardiovascular events and only isolated re-analyses pointing the other way. Online discourse frames seed oils as a major driver of chronic disease, but no defensible lifetime-harm number exists.\n","category":"food","tags":["food"],"no_reliable_estimate":true,"perceived":{"description":"The internet anti-seed-oil movement frames soybean, canola, sunflower, corn, and other industrial vegetable oils as a primary driver of modern chronic disease — obesity, cardiovascular events, inflammation, even cancer. The mechanism most often invoked is that high dietary linoleic acid (an omega-6 polyunsaturated fatty acid) is pro-inflammatory and pushes the body's omega-6 to omega-3 ratio into a disease-favoring state. The cultural reach of the claim is unusual: it travels from carnivore-diet creators to mainstream wellness influencers and into political and dietary-policy circles. Surveys have not assigned a numerical perceived lifetime risk to seed-oil exposure, but the framing routinely treats the harm as both substantial and lifetime- cumulative, comparable to or exceeding the harm from added sugar or processed meat.\n","rough_estimate":"Commonly framed as a major contributor to lifetime chronic disease risk, with no specific numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://www.cochrane.org/evidence/CD011737_effect-cutting-down-saturated-fat-we-eat-our-risk-heart-disease","title":"Reduction in saturated fat intake for cardiovascular disease","publisher":"Cochrane Database of Systematic Reviews (Hooper et al.)","source_type":"peer_reviewed","statistic":"Cochrane 2020 systematic review of 15 RCTs (n=56,675): reducing saturated fat intake for ≥2 years reduced combined cardiovascular events by ~17% (RR 0.83, 95% CI 0.70-0.98); the protective effect was driven primarily by replacement with polyunsaturated fats (seed/vegetable oils)","excerpt":"\"The findings of this updated review suggest that reducing saturated fat intake for at least two years causes a potentially important reduction in combined cardiovascular events.\"\n","source_date":"2020-08-21","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20251109072900/https://www.cochrane.org/evidence/CD011737_effect-cutting-down-saturated-fat-we-eat-our-risk-heart-disease","calculation_notes":"The single most cited evidence-grade systematic review on this question. Hooper 2020 pooled 15 RCTs with 56,675 participants and found a 17% reduction in cardiovascular events when saturated fat was replaced — predominantly by PUFA-rich oils. This is the opposite direction of harm to what the internet \"seed oils cause heart disease\" claim predicts. The effect on all-cause mortality was null (RR 0.96, 95% CI 0.90-1.03) and the effect on cardiovascular mortality was also null (RR 0.95, 95% CI 0.80-1.12); the protective signal is specifically on non-fatal events. Cochrane rates the evidence quality moderate. This review is the basis for the framing that seed-oil substitution is, at population scale, slightly protective rather than harmful.\n","independence_note":"Cochrane Collaboration systematic review with WHO funding for the 2019 update. Methodologically distinct from the AHA Presidential Advisory and the Ramsden re-analyses below; some overlap in the underlying RCT data (Cochrane includes Sydney Diet Heart and Minnesota Coronary Experiment among its 15 trials).\n"},{"url":"https://www.acc.org/latest-in-cardiology/ten-points-to-remember/2017/06/26/11/39/aha-presidential-advisory-on-dietary-fats-and-cvd","title":"AHA Presidential Advisory on Dietary Fats and Cardiovascular Disease","publisher":"American College of Cardiology summary of Sacks et al., Circulation 2017","source_type":"reputable_reference","statistic":"AHA Presidential Advisory (Sacks et al. 2017, Circulation): RCTs that replaced saturated fat with polyunsaturated vegetable oil reduced cardiovascular disease by approximately 30%, comparable to the reduction achieved by statin therapy","excerpt":"\"Randomized controlled trials that lowered intake of dietary saturated fat and replaced it with polyunsaturated vegetable oil reduced CVD by approximately 30%, similar to the reduction achieved by statin treatment. [...] lowering intake of saturated fat and replacing it with unsaturated fats, especially polyunsaturated fats, will lower the incidence of CVD.\"\n","source_date":"2017-06-15","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260112215032/https://www.acc.org/latest-in-cardiology/ten-points-to-remember/2017/06/26/11/39/aha-presidential-advisory-on-dietary-fats-and-cvd","calculation_notes":"The Sacks et al. AHA Presidential Advisory is the most prominent US clinical-society statement on this question and lands in the same place as Cochrane: replacing saturated fat with PUFA (predominantly seed/vegetable oils) reduces cardiovascular events. The 30% figure is the upper end of the protective effect range cited from RCT evidence. This is referenced via the American College of Cardiology summary page because the Circulation source PDF is bot-protected; the verbatim quote is from the Sacks 2017 advisory itself, reproduced by ACC. The direction of the headline finding is incompatible with the \"seed oils cause heart disease\" claim at population scale.\n","independence_note":"AHA Presidential Advisory writing committee with individual conflict disclosures; institutionally independent of the Cochrane review and the Ramsden re-analyses. Synthesizes a partly overlapping but distinct evidence base.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/27071971/","title":"Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73)","publisher":"BMJ (Ramsden, Zamora, Majchrzak-Hong et al.)","source_type":"peer_reviewed","statistic":"Ramsden 2016 BMJ re-analysis of the Minnesota Coronary Experiment (n=9,423): replacing saturated fat with linoleic-acid-rich corn oil lowered serum cholesterol but did not reduce coronary heart disease mortality or all-cause mortality; in older participants, there was a suggestion of increased mortality on the linoleic-acid arm","excerpt":"\"Available evidence from randomized controlled trials shows that replacement of saturated fat in the diet with linoleic acid effectively lowers serum cholesterol but does not support the hypothesis that this translates to a lower risk of death from coronary heart disease or all causes.\"\n","source_date":"2016-04-12","source_accessed":"2026-05-30","archive_url":"http://web.archive.org/web/20260520053403/https://pubmed.ncbi.nlm.nih.gov/27071971/","calculation_notes":"The strongest skeptic-side evidence in the seed-oil debate. Ramsden and colleagues recovered raw data from the Minnesota Coronary Experiment (1968-73) and a sibling Sydney Diet Heart re-analysis (Ramsden 2013, BMJ) and found that the cholesterol-lowering effect of linoleic-acid replacement did not translate into lower coronary or all-cause mortality, and in some age subgroups appeared to track with higher mortality. This is one notable trial pair, not a body of evidence, and Cochrane 2020 incorporated these data into its meta-analysis without changing the overall protective conclusion. The Ramsden papers are the legitimate scientific ground for moderating the strong \"PUFA-protective\" interpretation; they are not a base for the broad social-media claim that seed oils cause obesity, inflammation, or cancer.\n","independence_note":"Author team at NIH/NIAAA recovered the original Minnesota and Sydney trial data sets that had not been fully analyzed at the time of original publication. Methodologically distinct from the Cochrane review and the AHA advisory; the underlying trial data overlaps with the Cochrane review's included studies but the re-analysis methodology is independent.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/34417277/","title":"Dietary Intake of Linoleic Acid, Its Concentrations, and the Risk of Type 2 Diabetes: A Systematic Review and Dose-Response Meta-analysis of Prospective Cohort Studies","publisher":"Diabetes Care (Mousavi, Jalilpiran, Karimi, Aune, Larijani, Mozaffarian, Willett, Esmaillzadeh)","source_type":"peer_reviewed","statistic":"Mousavi et al. 2021 (Diabetes Care): meta-analysis of biomarker-based cohorts (84,171 participants, 18,458 incident cases) found higher linoleic-acid concentration in adipose tissue/blood associated with lower type 2 diabetes risk, RR 0.85 (95% CI 0.80-0.90) per standard-deviation increment","excerpt":"\"The summary RR for diabetes per SD increment in LA concentrations in adipose tissue/blood compartments was 0.85 (95% CI 0.80, 0.90). [...] High intake of dietary LA and elevated concentrations of LA in the body were both significantly associated with a lower risk of T2DM.\"\n","source_date":"2021-09-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260304000137/https://pubmed.ncbi.nlm.nih.gov/34417277/","calculation_notes":"A 2024-2026 gap-study addition. This biomarker meta-analysis is the authoritative source behind the type-2-diabetes protective factor in personal_factor_multipliers (RR 0.85 per SD increment in LA biomarker → multiplier 0.85, applied to T2D risk specifically, not to a generic \"harm from seed oils\" probability). It lands in the same direction as Cochrane and the AHA advisory: higher linoleic acid is associated with lower, not higher, cardiometabolic risk — contrary to the social-media \"seed oils cause disease\" claim. As an observational biomarker meta-analysis it shows association, not causation; the dietary-intake arm of the same paper agreed in direction.\n","independence_note":"Author team includes Mozaffarian and Willett; methodologically distinct from the Cochrane RCT review, the AHA advisory, and the Ramsden re-analyses. Observational cohort biomarker evidence rather than RCT evidence, so it complements rather than duplicates the trial-based sources above.\n"},{"url":"https://journals.lww.com/nutritiontodayonline/fulltext/2026/01000/health_implications_of_linoleic_acid_and_seed_oil.5.aspx","title":"Health Implications of Linoleic Acid and Seed Oil Intake","publisher":"Nutrition Today (Petersen, Messina, Flickinger), Vol 61(1):19-25","source_type":"peer_reviewed","statistic":"Petersen et al. 2026 (Nutrition Today) narrative review: observational evidence shows higher linoleic-acid intake is associated with lower cardiovascular disease risk; clinical-trial evidence shows linoleic acid does not promote inflammation or oxidative stress; conclusion is that seed oils are a healthful dietary component that should replace foods high in saturated fat","excerpt":"\"Observational evidence shows higher intake of LA is associated with lower risk of cardiovascular disease. Strong evidence demonstrates that replacing saturated fat with PUFA reduces the risk of CVD. Clinical trial evidence shows the LA does not promote inflammation or oxidative stress. [...] In conclusion, seed oils are a healthful dietary component that should be used to replace foods high in saturated fat.\"\n","source_date":"2026-02-01","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260128103012/https://journals.lww.com/nutritiontodayonline/fulltext/2026/01000/health_implications_of_linoleic_acid_and_seed_oil.5.aspx","calculation_notes":"A 2024-2026 gap-study addition (DOI 10.1097/NT.0000000000000746). This peer-reviewed 2026 narrative review directly addresses the central social-media claim — including the omega-6 inflammation mechanism — and concludes that clinical-trial evidence shows linoleic acid does not promote inflammation or oxidative stress. It is the most current synthesis on the question and strengthens the \"perceived harm exceeds evidence\" framing of this entry. As a narrative review (full text paywalled; abstract verbatim from the LWW journal page and Penn State pure repository) it carries less weight than the Cochrane RCT meta-analysis but is current and on point. Author affiliations include the soy/seed-oil nutrition research community; the conclusion is consistent with the independent Cochrane and AHA syntheses already cited.\n","independence_note":"Petersen (Penn State Nutritional Sciences), Messina, and Flickinger. Some authors are connected to seed-oil industry-funded nutrition research; the review's direction matches the institutionally independent Cochrane and AHA conclusions, so it corroborates rather than introduces a novel claim. Treated as a current-synthesis support source, not the load-bearing evidence.\n"}],"comparison_anchors":[{"label":"Baseline cardiovascular disease (lifetime, US adult)","lifetime_us_adult":0.4},{"label":"Red meat (daily) → colorectal cancer (lifetime, US adult)","lifetime_us_adult":0.048},{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.33}],"regional_breakdown":[{"region":"Mediterranean-pattern eater with olive oil as primary fat, occasional seed-oil exposure","probability":0.000001,"notes":"The PREDIMED cohort and broader Mediterranean-diet literature show cardiovascular benefit in this dietary pattern, in which seed-oil exposure is real but not dominant. No detected harm signal attributable to incidental seed-oil intake. Point estimate is a structural \"no detected harm\" placeholder.\n"},{"region":"Typical US adult eating mixed diet with substantial seed-oil exposure via cooking and restaurant food","probability":0.000001,"notes":"The Cochrane 2020 review and the AHA 2017 advisory both place this typical-exposure category on the protective rather than harmful side of the population-scale effect estimate. The \"modest CV benefit at population scale\" framing is the honest summary of the dominant evidence here.\n"},{"region":"Diet dominated by ultra-processed foods (chips, fried fast food, packaged snacks)","probability":0.000001,"notes":"The cohort signal of harm in this pattern (Srour 2019 BMJ, Fang 2024 BMJ on ultra-processed food) is real but is attributable to the overall dietary pattern — refined carbohydrates, sodium, additives, low fibre — rather than to the seed-oil content specifically. Isolating a seed-oil-attributable lifetime probability of harm in this pattern would be misattribution.\n"}],"personal_factor_multipliers":[{"factor":"Diet that replaces saturated fat with PUFA seed oils (per Cochrane / AHA)","multiplier":0.83,"notes":"Hooper 2020 Cochrane estimates a 17% reduction in combined cardiovascular events with saturated-fat-to-PUFA substitution sustained ≥2 years (RR 0.83). The multiplier here applies to cardiovascular events, not to a generic \"harm from seed oils\" probability — which does not exist as a measured quantity.\n"},{"factor":"Heavy ultra-processed-food consumer (fast food, chips, packaged snacks as primary calories)","multiplier":1.3,"notes":"The cohort signal of harm in ultra-processed-food-heavy diets (Srour 2019 BMJ, Fang 2024 BMJ) is real and substantial. Seed oils are a component of these foods but not the isolated mechanism; the overall pattern is the relevant exposure. Multiplier reflects ultra-processed-food cohort effect estimates, not a seed-oil- specific harm.\n"},{"factor":"Mediterranean-pattern eater (olive oil dominant, high vegetable and fish intake)","multiplier":0.7,"notes":"PREDIMED and broader Mediterranean-diet cohort literature support a meaningful cardiovascular protective effect in this pattern. Not a seed-oil-specific effect; the overall pattern is what is measured, and substantial PUFA exposure is part of it.\n"},{"factor":"Higher linoleic-acid intake (biomarker-confirmed) → type 2 diabetes risk","multiplier":0.85,"notes":"Mousavi et al. 2021 (Diabetes Care) biomarker meta-analysis: each standard-deviation increment in linoleic-acid concentration in adipose tissue/blood was associated with a lower type 2 diabetes risk, RR 0.85 (95% CI 0.80-0.90). The multiplier applies to type 2 diabetes incidence specifically, not to a generic \"harm from seed oils\" probability — which does not exist as a measured quantity. This is an observational association (biomarker, not RCT), so it shows correlation rather than proven causation; it lands in the protective direction, opposite the social-media harm claim.\n"},{"factor":"Adult with no other cardiovascular risk factors (normotensive, never-smoker, normal lipids)","multiplier":1,"notes":"Baseline reference. The available literature does not isolate a seed-oil-attributable lifetime harm probability for this population; absolute lifetime risks are dominated by other variables.\n"}],"short_label":"Seed oils & harm","myth_framing":"overrated","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers the broad social-media claim that seed oils — soybean, canola, corn, sunflower, safflower — are a primary driver of chronic disease at typical US dietary exposure. The honest position on this question is that no defensible lifetime-probability-of-harm number exists, because the dominant systematic-review evidence (Hooper 2020 Cochrane, Sacks 2017 AHA) shows modest cardiovascular protection from saturated-fat-to-PUFA replacement, not harm. The strongest skeptic-side evidence (Ramsden 2013 Sydney Diet Heart re-analysis, Ramsden 2016 Minnesota Coronary Experiment re-analysis) shows null or unfavorable mortality on the linoleic-acid replacement arm of two specific trials. These are real, important data points that legitimately moderate the strength of the protective claim — but they do not constitute evidence that seed oils as a category cause measurable lifetime harm in typical dietary patterns. Fabricating a number to make the question listable would misrepresent both sides of the literature.\nWhere this entry does not apply: the ultra-processed-food literature (Srour 2019, Fang 2024) does document harm from diets dominated by industrial packaged foods, fried fast food, and sweetened snacks. That harm signal is real and is attributable to the overall dietary pattern, not specifically to the seed-oil content of those foods. People who cook with canola or soybean oil at home in a varied diet are in a different exposure category from people whose calories come primarily from deep-fried restaurant food, and the cohort literature treats them differently. This entry is about the isolated \"seed oil itself\" question; the \"ultra-processed-food diet\" question is a separate one.\n2024-2026 update: the \"perceived harm exceeds evidence\" reading has if anything sharpened. A 2026 peer-reviewed narrative review (Petersen, Messina & Flickinger, Nutrition Today) concludes that clinical-trial evidence shows linoleic acid does not promote inflammation or oxidative stress and that seed oils are a healthful dietary component that should replace foods high in saturated fat — directly rebutting the omega-6 inflammation mechanism the social-media claim rests on. A biomarker meta-analysis (Mousavi et al. 2021, Diabetes Care) found higher linoleic-acid concentration associated with lower type 2 diabetes risk (RR 0.85 per SD increment), and a 2025 Johns Hopkins Bloomberg School expert summary (\"There is abundant evidence suggesting that seed oils are not bad for you. If anything, they are good for you\") reaches the same conclusion. Preliminary conference data presented at NUTRITION 2025 (not yet peer-reviewed) likewise linked higher linoleic-acid biomarkers to lower inflammation markers. None of this converts the question into a defensible lifetime-harm number — the direction of the best evidence is protective association, not measurable harm — but it does add current, named citations to the existing position. The diet-heart trial literature remains contested at the edges (the Ramsden re-analyses above), so the honest summary is still \"no reliable lifetime-harm estimate; the strongest evidence points the opposite way.\"\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-30","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-30","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-30","image":{"alt":"A single small glass bottle of pale yellow seed oil on a plain surface, rendered as a flat vector shape in muted warm-yellow and off-white tones on a calm background."},"canonical_url":"https://likelier.app/seed-oil-health-harm","api_url":"https://likelier.app/api/fears/seed-oil-health-harm.json"},{"slug":"shared-cup-infection","question":"What are the odds of getting a serious infection from drinking from someone else's cup?","quick_answer":"No study measures a per-event infection rate for sharing a cup, so no number is given, but the qualitative evidence points to effectively negligible risk for healthy adults (about 9 in 10 already carry EBV, for instance). The felt risk sits well above anything clinically serious.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"\"Don't drink after me\" is one of the oldest childhood rules, and most adults still flinch a little before taking a sip from a friend's glass. The fear lumps together everything from a winter cold to mono to vaguer worries about \"germs,\" and treats the cup itself as the vector. There is no good survey of how US adults rate this risk numerically, so the best we can say is that the felt risk of a shared-cup sip is substantially above the actual per-event risk of anything clinically serious.\n","rough_estimate":"Most people treat a shared sip as mildly risky but can't put a number on it","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/epstein-barr/about/index.html","title":"About Epstein-Barr Virus (EBV)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"~9 in 10 adults have antibodies showing current or past EBV infection; EBV is spread through saliva, including by sharing drinking cups","excerpt":"\"EBV is most commonly spread through saliva by: [...] Sharing drinks and food [...] Sharing drinking cups, eating utensils, or toothbrushes. [...] The virus probably survives on an object at least as long as the object remains moist. [...] About 9 out of 10 adults have antibodies that show that they have a current or past EBV infection.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413182847/https://www.cdc.gov/epstein-barr/about/index.html","calculation_notes":"CDC explicitly names shared drinking cups as an EBV transmission route, which sets the upper bound on what a cup CAN plausibly transmit. But the same page reports ~90% adult prevalence, meaning most readers are already immune and the marginal per-sip risk of new EBV infection is small for the typical adult. This is the single best anchor for the \"cup route is real but background prevalence is near-saturation\" framing. Not a per-event rate, but bounds the headline.\n","independence_note":"CDC EBV overview synthesises US seroprevalence surveys (including NHANES) with clinical guidance. Shares publisher with the CDC mononucleosis and CDC influenza pages — treat the three CDC sources as a single institutional voice, methodologically distinct from WHO's HSV-1 global seroprevalence modelling.\n"},{"url":"https://www.cdc.gov/flu/spread/index.html","title":"How Flu Spreads","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Influenza spreads mainly by droplets from coughs, sneezes, or talking; surface/object transmission is a secondary route","excerpt":"\"influenza viruses spread mainly by droplets made when people with flu cough, sneeze, or talk. These droplets can land in the mouths or noses of people who are nearby or possibly be inhaled into the lungs. [...] a person might get flu by touching a surface or object that has influenza virus on it and then touching their own mouth, nose, or possibly their eyes.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260425055539/https://www.cdc.gov/flu/spread/index.html","calculation_notes":"CDC ranks droplet/aerosol transmission ahead of surface/object transmission for influenza. The same logic applies to the other \"respiratory\" fears people pin on shared cups — cold viruses, RSV, COVID. Anyone close enough to share a cup with someone who is actively sick is already breathing the other person's exhaled air, which is the dominant vector. The cup is a minor add-on, not the primary exposure. This is the core of the myth-bust.\n","independence_note":"CDC influenza transmission guidance is a sibling CDC citation alongside the two EBV pages; treat all three as a single institutional voice. Provides the droplet-vs-fomite framing used to deflate the respiratory-virus branch of the fear.\n"},{"url":"https://www.cdc.gov/epstein-barr/about/mononucleosis.html","title":"About Infectious Mononucleosis","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"Mono-causing viruses (mainly EBV) spread through saliva; CDC recommends avoiding shared drinks with infected people","excerpt":"\"these viruses spread most commonly through bodily fluids, especially saliva [...] not kissing people who have infectious mononucleosis; or sharing drinks, food, or personal items (like toothbrushes) with them.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260403162513/https://www.cdc.gov/epstein-barr/about/mononucleosis.html","calculation_notes":"Corroborates the EBV-specific route through shared drinks and frames the mechanism in terms most readers will recognize (\"kissing disease\"). Reinforces that the one clinically named condition uniquely associated with shared-cup transmission is mono — which is self-limited in healthy adults and already extremely common by middle age. Not an independent estimate, same publisher and data source as the primary EBV page.\n","independence_note":"Same publisher and underlying epidemiology as the primary CDC EBV page above — treat as a corroborating sibling citation, not an independent data point.\n"},{"url":"https://www.who.int/news-room/fact-sheets/detail/herpes-simplex-virus","title":"Herpes Simplex Virus — Fact Sheet","publisher":"World Health Organization","source_type":"govt_report","statistic":"~3.8 billion people under 50 (64%) globally carry HSV-1; mainly transmitted via oral contact, saliva, or contact with sores","excerpt":"\"An estimated 3.8 billion people under age 50 (64%) globally have herpes simplex virus type 1 (HSV-1) infection [...] Type 1 (HSV-1) mostly spreads by oral contact and causes infections in or around the mouth [...] HSV-1 is mainly transmitted via contact with the virus in sores, saliva or skin surfaces in or around the mouth.\"\n","source_date":"2023-04-05","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183007/https://www.who.int/news-room/fact-sheets/detail/herpes-simplex-virus","calculation_notes":"HSV-1 is the other genuinely saliva-borne infection that a shared cup could plausibly transmit. WHO's 64% under-50 prevalence figure means the typical adult is about as likely to already carry HSV-1 as not, and oral contact (kissing, shared drinks) is the dominant acquisition route over a lifetime — not a single dramatic event. Supports the headline that the marginal per-sip risk for a new serious infection is very small because the background rate of \"already infected\" is so high.\n","independence_note":"WHO and CDC draw on overlapping seroprevalence literature but use different methodology and scope (WHO is global under-50, CDC EBV is US adults) — treat as related-but-not-identical.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"Any serious infection from one shared-cup event (healthy adults)","probability":0.000001,"notes":"Order-of-magnitude estimate; no direct measurement exists."},{"region":"Transmitting a cold or respiratory virus if the other person is currently sick","probability":0.01,"notes":"But the shared air between you is the main vector, not the cup itself."},{"region":"New HSV-1 or EBV acquisition from a single sip with an asymptomatic shedder","probability":0.0001,"notes":"Very rough. Both already have high background prevalence in adults (EBV ~90%, HSV-1 ~64% under-50), so the absolute per-event risk to someone not yet infected is still small, and a majority of readers already carry at least one of the two."}],"short_label":"Shared cup","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry is a deliberate myth-bust and the headline number is an order-of-magnitude estimate, not a measured rate. No study directly attributes infections to individual shared-cup events, because the exposure is confounded with every other form of close contact (shared air, shared surfaces, kissing, touching the same objects) that typically happens alongside it. The framing here only applies to healthy adults sharing a cup with other healthy adults in ordinary social settings. Three situations change the risk profile materially and none of them are captured by the point estimate: (1) severe immunocompromise — transplant recipients, chemotherapy patients, advanced HIV, long-term immunosuppression — where even common-cold and EBV exposures can become clinically significant; (2) an active cold sore or visible HSV lesion on the other person, which pushes the HSV-1 transmission probability up substantially for a non-carrier; (3) a pregnant person sharing with someone actively shedding CMV, where fetal transmission is a concrete concern. Does not cover bloodborne viruses: hepatitis B, hepatitis C, and HIV are NOT meaningfully transmitted by saliva on glassware, despite persistent folk belief.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty drinking glass on a pale grey-blue background, flat vector illustration."},"canonical_url":"https://likelier.app/shared-cup-infection","api_url":"https://likelier.app/api/fears/shared-cup-infection.json"},{"slug":"shared-toilet-seat-infection","question":"What are the odds of catching an STI or serious infection from a public toilet seat?","quick_answer":"No number can be given because the per-use transmission rate via a toilet seat sits below any surveillance system's detection floor — HIV cannot survive on the surface, and a landmark study found zero viable gonococci on 72 restroom samples. Many people feel a 1-in-1,000 or 1-in-100 per-use hazard, orders of magnitude above the real signal.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The public-toilet-seat fear is one of the most durable health myths in the English-speaking world. Most adults have been told at some point that you can \"catch something\" from sitting on one, and the fear spans the full STI panel — chlamydia, gonorrhea, herpes, HIV, HPV — plus a general aura of \"whatever that person had\". There is no good survey of how US adults rate the numerical per-use risk, but the behavioural signal is loud: the hover-crouch, the paper-cover industry, and the relief people express when they find a seat clean all suggest a felt risk that sits several orders of magnitude above the actual one.\n","rough_estimate":"Many people treat it as a 1-in-1,000 or 1-in-100 per-use hazard, well above the measured signal","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/hiv/causes/index.html","title":"How HIV Spreads (Causes of HIV)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"HIV does not survive long on surfaces outside the human body and cannot reproduce outside a human host","excerpt":"\"HIV does not survive long outside the human body (for example, on surfaces) and cannot reproduce outside a human host.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183044/https://www.cdc.gov/hiv/causes/index.html","calculation_notes":"CDC's plain-language statement is the single most authoritative rebuttal of the \"HIV from a toilet seat\" branch of the myth. The virus requires a living host to replicate and dies rapidly on ambient surfaces, which is why HIV transmission has never been attributed to a toilet seat in the epidemiological literature. Directly supports the headline that the per-use probability for HIV specifically is indistinguishable from zero and collapses the most dramatic version of the fear.\n","independence_note":"CDC HIV transmission guidance synthesises decades of US HIV surveillance and virology research. Shares publisher (and underlying institutional position) with the CDC STI page below — treat the two CDC sources as one institutional voice, methodologically separate from the NEJM culture study and the Mayo Clinic HSV-2 expert answer.\n"},{"url":"https://www.cdc.gov/sti/about/index.html","title":"About Sexually Transmitted Infections (STIs)","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"STIs pass from one person to another through vaginal, oral, and anal sex; intimate physical contact is a rare secondary route","excerpt":"\"STIs pass from one person to another through vaginal, oral, and anal sex. They can also spread through intimate physical contact like heavy petting, though this is not very common.\"\n","source_date":"2024-09-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260407041956/https://www.cdc.gov/sti/about/index.html","calculation_notes":"CDC's canonical STI transmission page lists exactly three routes — vaginal, oral, and anal sex — plus heavy petting as a rare secondary route. Toilet seats, towels, cups, and other fomites are conspicuously absent. This is the definitional argument: the bodies that write the STI case definitions do not consider environmental surfaces to be part of the transmission pathway. Supports the \"effectively zero with one case report on file\" framing by establishing that CDC does not even list the route as theoretically relevant.\n","independence_note":"Same publisher as the CDC HIV page above; treat as corroborating sibling citations from one institutional source, not two independent data points.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/109754/","title":"The Gonococcus and the Toilet Seat","publisher":"New England Journal of Medicine (Gilbaugh JH, Fuchs PC)","source_type":"peer_reviewed","statistic":"Zero viable N. gonorrhoeae recovered from 72 public-restroom cultures; saline suspensions nonviable shortly after drying","excerpt":"\"All of the organisms in a saline suspension were nonviable shortly after the sample had dried. [...] [The investigators] were unable to isolate a single N. gonorrhoeae from 72 cultures of various public restrooms.\"\n","source_date":"1979-07-12","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260413183128/https://pubmed.ncbi.nlm.nih.gov/109754/","calculation_notes":"This is the canonical peer-reviewed study people cite when asked whether gonorrhea transmits via toilet seats. Gilbaugh and Fuchs cultured 72 samples from real public restrooms and recovered zero viable gonococci; in vitro, the bacteria were nonviable shortly after the drop dried. The paper is nearly fifty years old and has never been meaningfully contradicted — Neisseria gonorrhoeae is notoriously susceptible to drying, which is the mechanistic reason no nonsexual transmission route has ever been established. Anchors the \"bacteria actively die on toilet seats\" half of the argument alongside CDC's \"HIV does not survive\" statement for the viral half.\n","independence_note":"Independent of the CDC sources: different authors, different methodology (in-restroom cultures plus lab suspension work vs CDC surveillance synthesis), different decade.\n"},{"url":"https://www.mayoclinic.org/diseases-conditions/genital-herpes/expert-answers/genital-herpes/faq-20058506","title":"Genital herpes: Can you get it from a toilet seat?","publisher":"Mayo Clinic (James M. Steckelberg, M.D.)","source_type":"reputable_reference","statistic":"Mayo Clinic: 'It's nearly impossible' to get genital herpes from a toilet seat; HSV-2 dies within seconds after exposure to air","excerpt":"\"It's nearly impossible to get genital herpes from a toilet seat. The herpes simplex virus type 2 (HSV-2), which causes most cases of genital herpes, dies quickly when away from the skin.\"\n","source_date":"2023-01-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260105135249/https://www.mayoclinic.org/diseases-conditions/genital-herpes/expert-answers/genital-herpes/faq-20058506","calculation_notes":"Mayo Clinic's named-physician expert-answer page is the cleanest contemporary clinical-voice corroboration of the NEJM-era lab work. Used alongside the Gilbaugh & Fuchs paper and the CDC HIV survival statement to complete the triangulation across bacterial (gonorrhea), viral (HSV, HIV), and decade-spanning sources.\n","independence_note":"Independent of CDC and the NEJM paper — different pathogen (HSV-2), different institution, different era.\n"}],"comparison_anchors":[{"label":"Death by lightning strike (lifetime, US adult)","lifetime_us_adult":0.00000354},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Serious infection from one shared-cup sip (healthy adults)","lifetime_us_adult":0.000001}],"regional_breakdown":[{"region":"Any STI transmission from a toilet seat (documented cases in literature)","probability":1e-8,"notes":"Essentially zero; no confirmed cases in medical literature."},{"region":"Norovirus from a recently-contaminated toilet seat","probability":0.0001,"notes":"Possible but sink taps and door handles are bigger vectors in real-world outbreaks."},{"region":"Skin infection from open wound + contaminated seat","probability":0.00001,"notes":"Requires an open wound making contact with fresh contamination; vanishingly rare in practice."}],"short_label":"Shared toilet seat","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The \"effectively zero\" headline is a claim about the classical STI panel (chlamydia, gonorrhea, syphilis, HIV, HSV-2, HPV) and cannot be extended uncritically to every pathogen that has ever been found on a toilet seat. The one known exception worth flagging: a 2004 case report in Sexually Transmitted Infections (PMC1744865) described a patient with gonorrhea in whom a toilet-seat route could not be definitively excluded, which is not the same thing as proving it. One \"could not be excluded\" case report in half a century of STI surveillance is roughly what \"effectively zero\" looks like at this scale, but it is not literally zero. Public restrooms genuinely do carry fecal-oral bacteria and viruses — norovirus, E. coli, Shigella, rotavirus — and a freshly contaminated seat that then contacts an open wound or mucous membrane is a plausible, if quantitatively minor, exposure route. In practice, epidemiologists who track bathroom-linked outbreaks consistently find that flush handles, sink taps, and door handles carry higher pathogen loads than the seat itself, because those are the surfaces hands touch immediately after the event that matters. If the fear is \"bathrooms are dirty\", it is partly correct; if the fear is \"I will catch chlamydia from the seat\", the literature has been unambiguous for nearly fifty years. Nothing here speaks to what happens if two people are having sex on a toilet seat, which is no longer a fomite transmission question.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single empty toilet seat lid viewed from above as a flat vector shape in muted grey-blue on an off-white background."},"canonical_url":"https://likelier.app/shared-toilet-seat-infection","api_url":"https://likelier.app/api/fears/shared-toilet-seat-infection.json"},{"slug":"shared-towel-infection","question":"What are the odds of getting a serious infection from using someone else's towel?","quick_answer":"No per-event infection rate exists, so no number is given — the evidence points to near-zero risk from casual dry towel sharing, with meaningfully higher risk only in damp athletic settings (MRSA, ringworm). People often imagine a 1-in-100 or even 1-in-10 chance of catching something per use.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Many people treat a borrowed towel as a vector for everything from HIV to hepatitis to \"every germ that person has ever had\". In practice the fears cluster around blood-borne viruses and respiratory pathogens, none of which are realistically transmitted by a briefly-shared dry towel. The scenarios that actually pose measurable risk — Staphylococcus aureus on damp athletic towels, or dermatophytes in gym settings — are rarely the ones people name when asked what they are worried about.\n","rough_estimate":"people often imagine a 1-in-100 or even 1-in-10 chance of 'catching something' per use","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/mrsa/prevention/athletes.html","title":"Athletes: MRSA Prevention and Control","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"CDC formally advises athletes not to share towels as a MRSA control measure","excerpt":"\"Do not share items that contact your skin, such as: Towels. Washcloths. Razors. Clothing, including uniforms. Ointments from open containers.\"\n","source_date":"2024-05-01","source_accessed":"2026-04-11","archive_url":"https://web.archive.org/web/20260504060413/https://www.cdc.gov/mrsa/prevention/athletes.html","calculation_notes":"CDC's athlete guidance is the most authoritative acknowledgement that shared towels are a genuine MRSA transmission route in specific settings (contact sports, shared locker rooms, close physical contact). It does not attempt to quantify per-event risk. No cited source in this entry measures a per-event infection rate for towel sharing, which is why this entry carries no_reliable_estimate rather than a headline number.\n","independence_note":"CDC MRSA athlete guidance is editorially independent of the ringworm and S. aureus recovery sources but draws on the same national outbreak-surveillance pipeline."},{"url":"https://www.cdc.gov/ringworm/about/index.html","title":"Ringworm Basics","publisher":"US Centers for Disease Control and Prevention (CDC)","source_type":"govt_report","statistic":"Shared towels and bedsheets are named by CDC as a ringworm (tinea) transmission route","excerpt":"\"Shared objects like towels and bedsheets.\"\n","source_date":"2024-03-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260420042300/https://www.cdc.gov/ringworm/about/index.html","calculation_notes":"CDC lists shared towels explicitly among the ways dermatophyte infections spread between people and pets. This supports the \"damp gym towel\" subgroup in the regional breakdown rather than the casual-sharing headline, since tinea transmission requires viable fungal spores and is heavily conditional on moisture and repeated contact.\n","independence_note":"CDC ringworm overview is a standalone public-health reference, editorially independent of the MRSA and staph-specific pages."},{"url":"https://pubmed.ncbi.nlm.nih.gov/19759479/","title":"Staphylococcus aureus recovery from cotton towels","publisher":"Journal of Infection in Developing Countries (via PubMed)","source_type":"peer_reviewed","statistic":"S. aureus viability detectable on cotton towels for >48 hours; more absorbent towels retain more bacteria","excerpt":"\"Cell viability decreased for over 48 hours on towels, but sufficient quantities may remain for colonization. More absorbent towels may harbor more Staphylococci than less absorbent ones, and may serve as a transmission mechanism for the bacterium.\"\n","source_date":"2009-09-15","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260504060431/https://pubmed.ncbi.nlm.nih.gov/19759479/","calculation_notes":"Establishes the biological plausibility of towel-mediated S. aureus transmission and sets the upper time bound (~48 hours of meaningful viability) for the casual-dry-sharing scenario. A towel that has sat dry for a day between uses carries dramatically less viable bacterial load than one passed immediately while damp — a mechanism that explains why damp athletic sharing is qualitatively riskier than casual dry sharing, though the study does not quantify a per-event rate for either.\n","independence_note":"Independent of CDC sources; different authorship and different methodology (in vitro recovery vs epidemiological guidance).\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2870587/","title":"Community-associated methicillin-resistant Staphylococcus aureus skin infections in a religious community","publisher":"Epidemiology and Infection (PMC)","source_type":"peer_reviewed","statistic":"In multivariate analysis of a CA-MRSA outbreak, towel sharing did not remain an independent risk factor once sauna use was controlled for","excerpt":"\"Only sauna use [odds ratio (OR) 19·1, 95% confidence interval (CI) 2·7–206·1) and antimicrobial use within 12 months before infection (OR 11·7, 95% CI 2·9–47·6) remained significant risk factors for CA-MRSA infection.\"\n","source_date":"2010-06-01","source_accessed":"2026-04-11","archive_url":"http://web.archive.org/web/20260504060523/https://pmc.ncbi.nlm.nih.gov/articles/PMC2870587/","calculation_notes":"Important corrective: while towel sharing turns up on univariate analyses in outbreak investigations, it typically drops out of the multivariate model once the underlying setting (damp, communal, close skin contact) is controlled for. This is the statistical basis for making the headline \"casual sharing is close to zero\" — the apparent per-event signal from outbreak reports is largely confounded by the context in which shared towels actually get used.\n","independence_note":"Epidemiological outbreak investigation of a specific religious-community CA-MRSA cluster, independent of the CDC national guidance documents and the in-vitro S. aureus recovery study. Provides the multivariate confounding analysis absent from CDC's categorical guidance.\n"}],"comparison_anchors":[{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017},{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Identity theft (lifetime, US adult)","lifetime_us_adult":0.33}],"short_label":"Shared towel","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The \"close to zero\" framing applies to the casual scenario (briefly-shared, dry towel, healthy contacts). The risks that do transmit via towels — Staphylococcus aureus including MRSA, dermatophyte fungi (tinea pedis, tinea corporis), and conjunctivitis — overwhelmingly require either damp fabric, repeated close contact, or an already-infected source. Blood-borne viruses such as HIV and hepatitis B/C are not meaningfully transmitted by shared towels; neither are most respiratory viruses. For anyone worried about catching HIV from someone else's bath towel, the data does not support that worry. For anyone worried about MRSA from a damp shared gym towel during an outbreak on your team, the data does support that worry, and CDC has written guidance for exactly that case.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-5-agent","last_reviewed":"2026-04-11","reviewed":true,"generated_at":"2026-04-11","image":{"alt":"A single folded towel rendered as a flat vector shape in muted terracotta and off-white, with plenty of empty space around it."},"canonical_url":"https://likelier.app/shared-towel-infection","api_url":"https://likelier.app/api/fears/shared-towel-infection.json"},{"slug":"skipping-dental-checkups","question":"What are the odds of serious dental disease from skipping regular checkups?","quick_answer":"Frequency matters less than assumed: the largest randomised trial of recall intervals (INTERVAL) found no difference in caries, gingivitis, or oral-health quality of life between six-month and up-to-24-month checkups for low-risk adults over four years. What skipping entirely costs has no reliable number; routine attenders do better in cohort data, but confounding blocks a causal estimate.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The twice-yearly dental visit is one of the most entrenched preventive-health norms in American culture, reinforced by insurance benefit structures that typically cover two cleanings per year. About 35% of US adults skip regular dental visits according to CDC/NHIS data, and the implicit assumption is that those who skip are accumulating undetected cavities, periodontal disease, and — most dramatically — missing early oral cancers that would have been caught by a routine exam. The fear is proportional to the guilt: most people who skip checkups believe they are taking a meaningful health risk.\n","rough_estimate":"Most adults believe skipping dental checkups leads to serious dental problems","kind":"intuition"},"sources":[{"url":"https://www.cdc.gov/nchs/products/databriefs/db435.htm","title":"Dental Care Utilization Among Adults Aged 18−64: United States, 2019 and 2020 (NCHS Data Brief No. 435)","publisher":"National Center for Health Statistics, Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"65.5% of US adults aged 18–64 had a dental visit in the past 12 months in 2019 (so ~35% did not); dental visit rates increased with family income and varied by race and Hispanic origin","excerpt":"\"The percentage of adults who had a dental visit in the past 12 months decreased from 65.5% in 2019 to 62.7% in 2020 (Figure 1).\"\n","source_date":"2022-04-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260627002325/https://www.cdc.gov/nchs/products/databriefs/db435.htm","calculation_notes":"NCHS analysis of the 2019–2020 National Health Interview Survey. In 2019, 65.5% of adults aged 18–64 had a dental visit in the past 12 months, so roughly 35% did not — the baseline for the exposed population. The brief also documents that the dental-visit rate increased with family income and varied by race and Hispanic origin. However, not visiting a dentist and developing serious dental disease are not the same thing — many non-visitors have adequate home care, and many regular visitors still develop caries and periodontal disease. The brief documents the exposure (skipping visits) but not the outcome (disease incidence attributable to skipping).\n"},{"url":"https://www.nature.com/articles/s41405-024-00195-7","title":"Impact of dental visiting patterns on oral health: A systematic review of longitudinal studies","publisher":"BDJ Open","source_type":"peer_reviewed","statistic":"Routine dental attenders had fewer missing teeth and less caries than non-routine attenders; effect was stronger with longer exposure to routine attendance","excerpt":"\"Routine attenders have better self-reported oral health and less tooth loss and dental caries, with the differential being greater with longer exposure to routine attendance.\"\n","source_date":"2024-03-15","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260426211753/https://www.nature.com/articles/s41405-024-00195-7","calculation_notes":"Systematic review of longitudinal studies comparing dental outcomes by visiting pattern. Found consistent associations between routine attendance and better oral health, but acknowledged confounding by socioeconomic status, education, and general health-seeking behaviour. The review could not establish a causal effect size for checkups independent of these confounders. Used here to support the direction of the association (checkups correlate with better outcomes) while noting the magnitude is uncertain.\n","independence_note":"Independent systematic review; does not use NHANES data directly.\n"},{"url":"https://seer.cancer.gov/statfacts/html/oralcav.html","title":"Cancer Stat Facts: Oral Cavity and Pharynx Cancer","publisher":"SEER Program, National Cancer Institute","source_type":"govt_report","statistic":"5-year relative survival for oral cancer: 88.7% localized, 36.0% distant, 69.9% overall; ~1.2% lifetime probability of diagnosis","excerpt":"\"Approximately 1.2 percent of men and women will be diagnosed with oral cavity and pharynx cancer at some point during their lifetime, based on 2021–2023 data.\"\n","source_date":"2025-04-01","source_accessed":"2026-07-11","archive_url":"http://web.archive.org/web/20260319161317/https://seer.cancer.gov/statfacts/html/oralcav.html","calculation_notes":"SEER data provides the strongest argument for dental checkups: oral cancer caught at the localized stage has 88.7% five-year survival versus 36.0% for distant disease. However, the lifetime probability of an oral cavity and pharynx cancer diagnosis is only ~1.2% of the US population, so the absolute number of people whose oral cancer would be caught earlier by a routine dental visit is small. The population-level impact of dental checkups on oral cancer mortality is modest compared to the impact on caries and periodontal disease management.\n","independence_note":"SEER is the US cancer surveillance system; independent of the NHIS dental-visit data and the BDJ Open systematic review.\n"},{"url":"https://evidence.nihr.ac.uk/alert/dental-check-ups-every-six-months-unnecessary-people-low-risk-oral-disease/","title":"Dental check-ups every six months are unnecessary for people at low risk of oral disease","publisher":"NIHR Evidence","source_type":"reputable_reference","statistic":"High-certainty evidence shows no difference in caries, gingivitis, or quality of life between 6-month and risk-based recall intervals over 4 years","excerpt":"\"Adults attending dental check-ups in primary care settings have little or no difference between risk-based and 6-month recall intervals in regard to dental caries, gingivitis and oral-health-related quality of life outcomes over a 4-year period.\"\n","source_date":"2024-01-15","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260209210414/http://evidence.nihr.ac.uk/alert/dental-check-ups-every-six-months-unnecessary-people-low-risk-oral-disease/","calculation_notes":"Based on the INTERVAL trial (Clarkson et al.), the largest RCT of dental recall intervals. Found that for low-risk adults, extending the recall interval from 6 months to 24 months produced no detectable difference in dental outcomes over 4 years. This is the strongest piece of evidence against the assumption that more-frequent checkups always equal better oral health. It does not address the question of no checkups at all versus some checkups, but it significantly weakens the case for the standard twice-yearly schedule.\n","independence_note":"NIHR summary of the UK INTERVAL RCT; methodologically independent of all US data sources.\n"}],"comparison_anchors":[{"label":"Adult tooth loss (at least one, US lifetime)","lifetime_us_adult":0.69},{"label":"Oral cancer (US adult, lifetime diagnosis probability)","lifetime_us_adult":0.012},{"label":"Heart disease death (US adult, lifetime)","lifetime_us_adult":0.2}],"personal_factor_multipliers":[{"factor":"Smoker (current)","multiplier":2.5,"notes":"Smoking is the strongest modifiable risk factor for both periodontal disease and oral cancer"},{"factor":"Diabetic","multiplier":2,"notes":"Diabetes roughly doubles periodontal disease risk independent of other factors"},{"factor":"Low-income / uninsured","multiplier":1.5,"notes":"Reflects both reduced access to care and correlation with other risk factors"},{"factor":"Xerostomia (chronic dry mouth from medications)","multiplier":3,"notes":"ADA and clinical pharmacology literature document that saliva provides mechanical cleansing and antimicrobial protection; patients on medications that cause xerostomia (antidepressants, antihistamines, diuretics) lose this defense and develop caries and periodontal disease at substantially higher rates, making checkup intervals more clinically meaningful"},{"factor":"High-sugar diet (>30g free sugars/day)","multiplier":2,"notes":"WHO and systematic caries-risk literature establish free-sugar intake as the dominant dietary driver of caries; individuals consuming high-sugar diets accumulate caries lesions faster between checkup intervals, increasing the marginal value of professional cleaning and early-lesion detection"}],"short_label":"Skipping dental checkups","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"There is no reliable population estimate for the odds of serious dental disease caused by skipping checkups. The cited CDC/NHIS brief documents the exposure (~35% of US adults did not have a dental visit in the past year) but not the disease outcome. The marginal contribution of regular checkups to preventing periodontal disease, caries, and tooth loss is uncertain and heavily confounded by socioeconomic status, smoking, diet, home care quality, and genetics: a 2024 BDJ Open systematic review found routine attenders had better oral health but could not isolate a causal effect size. The strongest evidence for checkup value is in oral cancer early detection, but oral cancer is rare enough (SEER: ~1.2% lifetime diagnosis probability) that the absolute benefit is small. The INTERVAL trial showed that for low-risk adults, stretching checkups from every 6 months to every 24 months made no measurable difference over 4 years.\n","quality_score":{"d1":5,"d2":5,"d3":3,"d4":4,"d5":3,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single dental mirror resting on a pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-dental-checkups","api_url":"https://likelier.app/api/fears/skipping-dental-checkups.json"},{"slug":"skipping-flossing-tooth-loss","question":"What are the odds of losing teeth from not flossing?","quick_answer":"No trial has ever measured it. Across two Cochrane reviews, zero randomised trials report tooth loss or cavity outcomes for flossing versus not flossing; the only measured benefit is a small gingivitis reduction the reviewers rated \"very unreliable\". Skipping flossing has no proven tooth-loss cost, which is not the same as a proven zero.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Flossing occupies a peculiar position in preventive health advice: nearly every dentist recommends it, most patients feel guilty about not doing it, and the implied consequence — gum disease progressing to tooth loss — is treated as near-certain for non-flossers. The American Dental Association has recommended daily flossing for decades, and the practice is embedded in public-health messaging worldwide. When the Associated Press investigated in 2016, it found that the US government had quietly dropped flossing from its dietary guidelines after acknowledging that the evidence base was weak.\n","rough_estimate":"Most adults believe skipping flossing substantially increases their risk of tooth loss","kind":"intuition"},"sources":[{"url":"https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD008829.pub2/abstract","title":"Flossing for the management of periodontal diseases and dental caries in adults","publisher":"Cochrane Database of Systematic Reviews / Sambunjak D, Nickerson JW, Poklepovic T, et al.","source_type":"peer_reviewed","statistic":"Evidence for flossing reducing caries or clinical attachment loss is absent; evidence for reducing gingivitis is 'very unreliable'","excerpt":"\"There is some evidence from twelve studies that flossing in addition to toothbrushing reduces gingivitis compared to toothbrushing alone. There is weak, very unreliable evidence from 10 studies that flossing plus toothbrushing may be associated with a small reduction in plaque at 1 and 3 months. No studies reported data for the outcomes of caries, calculus, clinical attachment loss, or quality of life.\"\n","source_date":"2011-12-07","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250316153254/https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD008829.pub2/abstract","calculation_notes":"The Cochrane review is the gold standard for assessing clinical interventions. Sambunjak et al. examined 12 RCTs (1,083 participants) comparing flossing-plus-brushing to brushing alone. Zero trials reported tooth loss, caries incidence, or clinical attachment loss as outcomes — the very endpoints that matter for the \"flossing prevents tooth loss\" claim. The evidence for gingivitis reduction was rated \"very unreliable\" due to high risk of bias in most included studies and short follow-up (typically 1-3 months). The review does not say flossing is harmful; it says the evidence that it prevents the outcomes people fear is essentially absent. This supports no_reliable_estimate.\n"},{"url":"https://cochraneohg.wordpress.com/2019/04/11/the-evidence-on-flossing-and-other-methods-of-cleaning-between-the-teeth/","title":"The evidence on flossing and other methods of cleaning between the teeth","publisher":"Cochrane Oral Health","source_type":"peer_reviewed","statistic":"Updated 2019 Cochrane review of interdental cleaning found low-certainty evidence for small reductions in gingivitis and plaque; no data on caries or tooth loss","excerpt":"\"There was low‐certainty evidence that interdental brushes may reduce gingivitis and plaque more than flossing.\"\n","source_date":"2019-04-11","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20251223141848/https://cochraneohg.wordpress.com/2019/04/11/the-evidence-on-flossing-and-other-methods-of-cleaning-between-the-teeth/","calculation_notes":"The 2019 update (Slot et al.) expanded the scope to all interdental cleaning devices and found 35 studies. The conclusion was unchanged for flossing specifically: no evidence for caries prevention or tooth-loss prevention. Interdental brushes showed slightly better gingivitis reduction than floss, but the evidence was still low-certainty. The persistent absence of hard-endpoint data (tooth loss, caries) across decades of research is itself informative — it means no trial has been designed or funded to answer the question the public actually cares about.\n","independence_note":"This is an update and expansion of the Sambunjak 2011 review by Cochrane Oral Health, using a broader inclusion criteria but the same Cochrane methodology.\n"},{"url":"https://www.sciencedirect.com/science/article/abs/pii/S1532338216301877","title":"In Defense of Flossing: Can We Agree It's Premature to Claim Flossing is Ineffective to Prevent Dental Caries?","publisher":"Journal of Evidence-Based Dental Practice / Vernon LT, Goettsche ZS","source_type":"peer_reviewed","statistic":"Argues that absence of evidence for flossing is not evidence of absence; RCTs are too short and too small to detect caries or tooth-loss endpoints","excerpt":"\"The current evidence base is insufficient to either prove or disprove the efficacy of flossing for caries prevention.\"\n","source_date":"2016-12-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20240416062017/https://www.sciencedirect.com/science/article/abs/pii/S1532338216301877","calculation_notes":"Vernon and Goettsche push back on the AP and Cochrane framing, arguing that the absence of RCT evidence reflects the near-impossibility of running a multi-year, large-sample trial with a no-flossing control arm — not a finding that flossing is ineffective. This is a fair methodological point. However, it does not change the evidentiary status: the claim that skipping flossing leads to tooth loss remains unsupported by direct clinical trial data. Included here because it represents the strongest counterargument and demonstrates that even defenders of flossing acknowledge the evidence gap.\n","independence_note":"Independent commentary by US dental researchers responding to the AP investigation; not affiliated with the Cochrane review team.\n"}],"comparison_anchors":[{"label":"Adult tooth loss (at least one tooth, US lifetime)","lifetime_us_adult":0.69},{"label":"Periodontal disease (US adults over 30)","lifetime_us_adult":0.47}],"short_label":"Skipping flossing","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry asks specifically whether skipping flossing causes tooth loss. It does not claim that flossing is useless — plaque removal between teeth is mechanistically plausible as a contributor to gum health. The problem is that no randomised controlled trial has measured the endpoint that matters (tooth loss or caries incidence) for flossing versus not flossing. The trials that exist are short (1-3 months), small, and measure surrogate endpoints like plaque scores and gingival bleeding indices. Whether flossing prevents tooth loss over a 20-year horizon is genuinely unknown, not known to be zero. The AP 2016 investigation prompted the US government to remove flossing from its Dietary Guidelines after acknowledging the evidence did not meet the required standard.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single piece of dental floss draped loosely on a pale surface, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-flossing-tooth-loss","api_url":"https://likelier.app/api/fears/skipping-flossing-tooth-loss.json"},{"slug":"skipping-preschool-harm","question":"What are the odds of a child being developmentally harmed by not attending preschool?","quick_answer":"No harm probability can be derived: the research measures test-score convergence and fadeout, not a binary harm outcome, and the largest RCT found early preschool gains had largely faded by third grade. Parents perceive a 30-60% chance of lasting developmental harm, but the evidence, especially for children from stimulating homes, is far more equivocal.\n","category":"health","tags":["kids"],"no_reliable_estimate":true,"perceived":{"description":"Preschool attendance is widely treated as a developmental necessity in middle-class parenting culture. Parents who keep children home until kindergarten often face social pressure and anxiety about socialization deficits or academic unpreparedness. The belief that missing preschool closes a critical developmental window is pervasive in parenting media, despite the research base being far more equivocal than the popular narrative suggests — particularly for children from stimulating home environments.\n","rough_estimate":"~30-60% perceived chance of lasting developmental harm","kind":"intuition"},"sources":[{"url":"https://acf.gov/opre/report/third-grade-follow-head-start-impact-study-final-report","title":"Third Grade Follow-up to the Head Start Impact Study: Final Report","publisher":"Office of Planning, Research and Evaluation, Administration for Children and Families, U.S. Department of Health and Human Services","source_type":"govt_report","statistic":"By the end of 3rd grade, the cognitive and achievement gains from Head Start had largely faded, with only a single statistically significant impact remaining for each age cohort","excerpt":"\"The early effects of Head Start on language and literacy development rapidly dissipated in elementary school, with only a single impact remaining at the end of 3rd grade for children who entered at age 3 and a single impact for children who entered at age 4.\"\n","source_date":"2012-10-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260317193440/https://acf.gov/opre/report/third-grade-follow-head-start-impact-study-final-report","calculation_notes":"The Head Start Impact Study was a federally mandated randomized controlled trial of approximately 4,600 children across 383 Head Start centers. Children were randomly assigned to Head Start or a control group (which could access non-Head-Start care). By 1st grade, most cognitive gains had dissipated; by 3rd grade, nearly all had. This is the largest RCT of a US preschool program and the fadeout finding is its most replicated result. No lifetime harm probability is derivable because the study measures test-score convergence, not a binary harm outcome. The control group was not \"no care\" — many attended other programs, making this a comparison of Head Start vs available alternatives, not preschool vs nothing.\n","independence_note":"Federally mandated RCT conducted by Westat and Mathematica Policy Research. Independent from the NICHD SECCYD observational study and the Perry Preschool experiment.\n"},{"url":"https://onlinelibrary.wiley.com/doi/10.1111/j.1467-8624.2010.01431.x","title":"Do Effects of Early Child Care Extend to Age 15 Years? Results From the NICHD Study of Early Child Care and Youth Development","publisher":"Child Development (Wiley)","source_type":"peer_reviewed","statistic":"Quality of early child care predicted cognitive-academic achievement at age 15 (d = 0.09), but the effect size was small; quantity of care predicted more externalizing behaviors and risk-taking","excerpt":"\"Higher quality care predicted higher cognitive-academic achievement at age 15, with escalating positive effects at higher levels of quality. More hours in care predicted more externalizing behavior problems and greater risk taking and impulsivity at age 15.\"\n","source_date":"2010-05-01","source_accessed":"2026-04-19","calculation_notes":"Vandell et al. (2010) followed 1,364 children from the NICHD SECCYD from birth through age 15. The study is observational, not experimental, and tracked children across a wide range of care arrangements (center care, family daycare, parental care, nanny care). The key finding for this entry is that quality of care was a stronger predictor of outcomes than whether the child attended group care at all. The effect size for quality on cognitive-academic outcomes (d = 0.09) is statistically significant but practically small. Notably, more hours in non-maternal care predicted worse behavioral outcomes, complicating any simple \"more care = better\" narrative. No binary harm probability for non-attendance is derivable because the study does not compare \"no care\" vs \"care\" — it compares varying levels of quality, quantity, and type.\n","independence_note":"The NICHD SECCYD is the largest prospective US child care study, funded by NICHD and conducted across 10 university sites. Methodologically independent from Head Start Impact Study (RCT) and Perry Preschool (intensive intervention).\n"},{"url":"https://highscope.org/wp-content/uploads/2024/07/perry-preschool-summary-40.pdf","title":"Lifetime Effects: The HighScope Perry Preschool Study Through Age 40","publisher":"HighScope Press","source_type":"peer_reviewed","statistic":"Perry Preschool participants had higher earnings, lower crime rates, and higher high school graduation rates at age 40 — but participants were low-IQ (70-85), low-income African American children receiving an intensive 2-year intervention","excerpt":"\"The study documents long-lasting improvements in the original participants' skills, marriage stability, earnings, criminal behavior, and health. Eligibility criteria required that participants be African-American, have a low IQ (between 70 and 85) at study entry, and be disadvantaged as measured by parental employment level, parental education, and housing density.\"\n","source_date":"2005-01-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260302053435/https://highscope.org/wp-content/uploads/2024/07/perry-preschool-summary-40.pdf","calculation_notes":"Schweinhart et al. (2005) reported on 123 African American children (58 program, 65 control) from very disadvantaged backgrounds in Ypsilanti, Michigan, enrolled in 1962-1967. The intervention was intensive: 2.5 hours of daily classes with certified teachers at an 8:1 ratio, plus biweekly home visits, for two school years. This is not comparable to typical modern preschool — it was a high-dosage intervention for a population with IQs in the 70-85 range and severe socioeconomic disadvantage. The results are dramatic (higher earnings, lower incarceration, higher graduation rates at age 40) but the study's own authors note that generalizability extends only to \"programs that are reasonably similar to this program\" serving similar populations. Extrapolating Perry results to middle-class children attending typical preschool programs is a category error common in public discourse.\n","independence_note":"Small longitudinal experiment (N=123) conducted 1962-1967 with follow-ups through age 40. Fully independent of the NICHD SECCYD and Head Start Impact Study in design, population, and methodology.\n"}],"comparison_anchors":[{"label":"Screen time developmental harm (lifetime)","lifetime_us_adult":0.08},{"label":"Social media teen harm (lifetime)","lifetime_us_adult":0.12}],"regional_breakdown":[{"region":"Low-income / disadvantaged child (US)","probability":0.25,"notes":"For children from low-income families with limited home learning environments, skipping preschool is associated with meaningfully worse kindergarten readiness and, in intensive-intervention studies like Perry Preschool, worse long-term outcomes. Head Start and state pre-K programs show initial gains, though test-score advantages typically fade by 3rd grade. Non-cognitive benefits (reduced crime, higher graduation) may persist."},{"region":"Middle-class child with engaged parents","probability":0.03,"notes":"For children from stimulating home environments with educated, engaged parents, the evidence for lasting harm from skipping preschool is weak. The NICHD SECCYD found quality of care mattered more than whether the child attended, and families providing high-quality home environments largely compensate. Most test-score differences disappear by 2nd-3rd grade."},{"region":"Child with developmental delay or special needs","probability":0.3,"notes":"Early intervention services delivered in group settings have strong evidence for children with speech delays, autism spectrum conditions, and other developmental differences. Federal IDEA Part B (Section 619) mandates free appropriate public education for children with disabilities ages 3-5. For this population, group-based early intervention is closer to medical treatment than enrichment."},{"region":"Nordic countries (universal high-quality preschool)","probability":0.05,"notes":"In Scandinavian countries with universal, heavily subsidized, high-quality preschool, nearly all children attend. Isolating the effect of non-attendance is difficult because the counterfactual is rare. Available evidence suggests modest benefits concentrated in immigrant and low-income families, consistent with the SES-moderation pattern seen in US data."}],"personal_factor_multipliers":[{"factor":"Low-income household (below 200% FPL)","multiplier":5,"notes":"Preschool is most protective for children from low-income families. The Perry Preschool and Abecedarian studies showed large, lasting benefits specifically for this population. NOT attending carries meaningful developmental risk when the home environment is under-resourced."},{"factor":"High-quality home environment (books, engaged parents, routines)","multiplier":0.2,"notes":"An enriching home environment with responsive parenting, literacy exposure, and structured activities compensates substantially for the absence of formal preschool. The NICHD SECCYD found parenting quality was a far stronger predictor of outcomes than any child care variable."},{"factor":"Child with speech or social delay","multiplier":4,"notes":"Children with identified developmental delays benefit disproportionately from structured group settings with trained professionals. Early intervention has strong evidence for speech-language development and social skill acquisition."},{"factor":"Neurotypical child with siblings or regular peer contact","multiplier":0.3,"notes":"The 'socialization' argument for preschool is weakest for children who already have regular peer interaction through siblings, playgroups, or community activities. These children typically show no meaningful social-skill deficit at kindergarten entry."}],"short_label":"No preschool harm","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The evidence for preschool benefits is dramatically moderated by socioeconomic status. For a low-income child in a low-stimulation home environment, skipping preschool IS associated with worse outcomes, and the Perry Preschool and Abecedarian studies demonstrate that intensive early education can produce lasting benefits for this population. For a middle-class child with engaged parents, the evidence for lasting developmental harm from skipping preschool is weak — test-score advantages largely fade by 2nd-3rd grade (the \"fadeout effect\"), and the NICHD study found parenting quality predicted outcomes far more strongly than any child care variable. Most popular discourse conflates these two very different situations, applying findings from intensive interventions for deeply disadvantaged children to the general population. The regional_breakdown probabilities are rough estimates of \"meaningfully worse developmental trajectory\" and should be treated as illustrative, not precise.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A small wooden chair next to building blocks and a picture book on a rug, soft natural light from a window, muted warm tones, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-preschool-harm","api_url":"https://likelier.app/api/fears/skipping-preschool-harm.json"},{"slug":"skipping-tooth-brushing","question":"What are the odds of getting cavities or losing teeth from skipping tooth brushing?","quick_answer":"No study has measured what one occasional missed session costs — the data only compares habitual patterns over months to years. Two meta-analyses link brushing less than twice daily to roughly 45% higher odds of new cavities and 44% higher odds of severe gum disease, but these are odds ratios on near-universal outcomes, not lifetime risk figures. The 30-minute post-meal rule comes from extreme lab acid challenges (pH 2.9), not ordinary meals.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Twice-daily brushing is taught as close to universal health dogma, and the implied consequence of skipping it — progressive decay leading eventually to tooth loss — is treated as near-certain by most adults. Patients who miss a session often report disproportionate anxiety, and the advice to brush after every meal circulates as a recommended upgrade over the standard twice-daily habit. The recommendation to wait 30 minutes before brushing after eating is framed widely as established fact, with significant enamel erosion attributed to immediate post-meal brushing. In practice, most adults brush inconsistently — twice some days, once on others — and the fear that each missed session meaningfully accelerates decay is widespread.\n","rough_estimate":"Most adults believe that regularly skipping tooth brushing will lead to significant tooth decay or tooth loss within a few years","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/27334438/","title":"Effect of Toothbrushing Frequency on Incidence and Increment of Dental Caries: A Systematic Review and Meta-Analysis","publisher":"Journal of Dental Research / Kumar S, Tadakamadla J, Johnson NW","source_type":"peer_reviewed","statistic":"Brushing ≥2×/day vs <2×/day: OR 1.45 (95% CI 1.21–1.74) for caries incidence; overall incidence OR 1.50 (95% CI 1.34–1.69); permanent dentition OR 1.39 (95% CI 1.29–1.49); 33 studies included","excerpt":"\"[Paraphrase from abstract — full text paywalled] Infrequent brushers demonstrated higher incidence and increment of carious lesions than frequent brushers. In the comparison of brushing ≥2 times/day vs <2 times/day, the odds ratio for caries incidence was 1.45 (95% CI: 1.21–1.74). Effects were higher in primary dentition (OR 1.75, 95% CI 1.49–2.06) than in permanent teeth (OR 1.39, 95% CI 1.29–1.49). Thirty-three studies were included; absolute incidence rates by frequency band were not reported.\"\n","source_date":"2016-10-01","source_accessed":"2026-07-19","archive_url":"http://web.archive.org/web/20260617204347/https://pubmed.ncbi.nlm.nih.gov/27334438/","calculation_notes":"Kumar et al. conducted a systematic review and meta-analysis of 33 studies on toothbrushing frequency and caries. The overall incidence OR comparing infrequent versus frequent brushers was 1.50 (95% CI 1.34–1.69); the subgroup comparing brushing ≥2×/day versus <2×/day gave OR 1.45 (95% CI 1.21–1.74). The dentition breakdown was OR 1.39 for permanent teeth and OR 1.75 for primary teeth — these are separate estimates, not the same figure. All estimates are relative measures. No study in the meta-analysis reported absolute cumulative caries incidence by frequency band across a defined follow-up period. Without the baseline incidence for the infrequent-brushing group, the OR cannot be translated into a lifetime probability figure. Supports no_reliable_estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25255820/","title":"Is frequency of tooth brushing a risk factor for periodontitis? A systematic review and meta-analysis","publisher":"Community Dentistry and Oral Epidemiology / Zimmermann H, Zimmermann N, Hagenfeld D, Veile A, Kim TS, Becher H","source_type":"peer_reviewed","statistic":"Infrequent brushing: OR 1.44 (95% CI 1.21–1.71, random effects) and OR 1.41 (95% CI 1.25–1.58, fixed effects) for severe periodontitis; 14 studies (12 cross-sectional, 2 case-controlled)","excerpt":"\"[Paraphrase from abstract — full text paywalled] The random effects meta-analysis of all 14 studies found OR = 1.44 (95% CI: 1.21–1.71, P < 0.0001). Infrequent tooth brushing was associated with severe forms of periodontal disease. The authors noted limited epidemiological studies on this topic and called for additional research.\"\n","source_date":"2015-04-01","source_accessed":"2026-07-19","archive_url":"http://web.archive.org/web/20251224003229/https://pubmed.ncbi.nlm.nih.gov/25255820/","calculation_notes":"Zimmermann et al. identified 14 studies (12 cross-sectional, 2 case-controlled) on brushing frequency and periodontitis. The pooled OR of 1.44 (random effects; fixed effects 1.41) represents the relative association between infrequent brushing and severe periodontal disease. Definitions of \"infrequent\" varied across studies; heterogeneity was I² = 48%. No absolute cumulative incidence rates by frequency band were reported. The predominantly cross-sectional design limits causal inference. Supports no_reliable_estimate.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/14684979/","title":"Brushing abrasion of softened and remineralised dentin: an in situ study","publisher":"Caries Research / Attin T, Siegel S, Buchalla W, Lennon AM, Hannig C, Becker K","source_type":"peer_reviewed","statistic":"Dentin wear 10 min post-acid (pH 2.9): 37.9 µm; at 30 min: 18.5 µm; at 60 min: 15.3 µm; unbrushed control: 12.6 µm. 30- and 60-min groups not significantly different from unbrushed control.","excerpt":"\"It is concluded that for protection of dentin surfaces at least 30 min should elapse before toothbrushing after an erosive attack.\"\n","source_date":"2004-01-01","source_accessed":"2026-07-19","archive_url":"http://web.archive.org/web/20260716125923/https://pubmed.ncbi.nlm.nih.gov/14684979/","calculation_notes":"Attin et al. 2004 is the primary source for the \"30-minute rule.\" Eleven volunteers wore intra-oral appliances containing human dentin specimens demineralised twice daily in Sprite Light (pH 2.9) for 90 seconds — a severe, repeated acid challenge far exceeding normal dietary exposures. Dentin wear was measured after 21 days. Brushing at 30 or 60 minutes produced wear indistinguishable from the unbrushed control, supporting the recommendation for post-erosive-challenge delays. Critical limitation: the acid challenge (pH 2.9 soda twice daily) far exceeds ordinary meal exposures. The study tests dentin already softened by severe erosive attack, not intact enamel after a normal meal. The 30-minute guidance cannot be generalised from this extreme condition to all eating occasions.\n"},{"url":"https://www.scielo.br/j/bor/a/yNjBVm67s7HSdZQwjdgfHCL/?lang=en","title":"Association between toothbrushing frequency and dental caries and tooth loss in adolescents: a cohort study","publisher":"Brazilian Oral Research / Brusius CD, Alves LS, Maltz M","source_type":"peer_reviewed","statistic":"Fully adjusted (Model B): adolescents brushing twice daily IRR = 3.92 (95% CI 1.23–12.49) for tooth loss vs brushing ≥3×/day; brushing ≤once daily IRR = 7.09 (95% CI 2.24–22.49). Tooth loss incidence 3% (95% CI 1.8–4.2), 23 of 761 adolescents with no missing teeth at baseline, over 2.5 years; 1,528 enrolled","excerpt":"\"[Paraphrase from abstract and Table 3 — full text on SciELO] In the fully adjusted model, adolescents brushing teeth twice daily showed IRR = 3.92 (95% CI: 1.23–12.49) for tooth loss compared to those brushing three or more times daily. Those brushing once a day or less had IRR = 7.09 (95% CI: 2.24–22.49). An incidence of 3% (95% CI: 1.8–4.2) was detected, representing 23 of 761 adolescents who had no missing teeth at baseline, over 2.5 years of follow-up (1,528 schoolchildren enrolled).\"\n","source_date":"2023-12-01","source_accessed":"2026-07-19","archive_url":"http://web.archive.org/web/20260311074320/https://www.scielo.br/j/bor/a/yNjBVm67s7HSdZQwjdgfHCL/?lang=en","calculation_notes":"Brusius et al. 2023 is the only identified longitudinal cohort measuring tooth loss as a direct outcome by brushing frequency. The dose-response finding in the fully adjusted model (twice daily nearly 4× the tooth loss of three-times-daily, IRR 3.92; once daily or less about 7×, IRR 7.09) is directionally consistent with the caries and periodontitis meta-analyses. The study cannot be used to derive a US adult lifetime estimate for several reasons: (a) the population is Brazilian adolescents with different fluoride histories, diets, and dental care access than US adults; (b) 2.5-year follow-up in adolescents does not map to a 59-year US adult horizon; (c) the reference group is brushing ≥3×/day, not a no-brushing baseline; (d) the 3% absolute incidence is measured only among the 761 adolescents with no missing teeth at baseline (23 events), a young population whose figure cannot be scaled linearly to adult lifetime risk. Supports no_reliable_estimate.\n"},{"url":"https://www.nidcr.nih.gov/research/data-statistics/dental-caries/adults","title":"Dental Caries in Adults (Ages 20 to 64) Data & Statistics","publisher":"National Institute of Dental and Craniofacial Research (NIDCR)","source_type":"govt_report","statistic":"Nearly 90% of US adults aged 20–64 have had decay in their teeth; unchanged between 1999–2004 and 2011–2016 NHANES cycles","excerpt":"\"Nearly 90% of adults ages 20 to 64 years have had decay in their teeth, a percentage that has not changed significantly between the 1999–2004 and 2011–2016 NHANES cycles.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-19","archive_url":"http://web.archive.org/web/20260723153601/https://www.nidcr.nih.gov/research/data-statistics/dental-caries/adults","calculation_notes":"NIDCR NHANES-based prevalence data: nearly 90% of US adults aged 20–64 have experienced dental caries. This near-universal background prevalence shows that dental caries strikes the large majority of adults regardless of brushing habits, which makes isolating the incremental contribution of brushing frequency alone — after accounting for diet, fluoride exposure, and dental care access — intractable in observational studies. Used here to support no_reliable_estimate, not as a headline risk estimate attributable to brushing frequency.\n"}],"comparison_anchors":[{"label":"Adult tooth loss (at least one tooth, US lifetime)","lifetime_us_adult":0.69},{"label":"Periodontal disease (US adults over 30)","lifetime_us_adult":0.47}],"short_label":"Skipping tooth brushing","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"There is no reliable cumulative lifetime estimate for serious dental disease or tooth loss specifically attributable to brushing less than twice daily. The best available evidence — Kumar et al. 2016 (33 studies, caries) and Zimmermann et al. 2015 (14 studies, periodontitis) — provides odds ratios of approximately 1.45 and 1.44 respectively for infrequent versus more-frequent brushers, but not absolute incidence rates by frequency band. Without the baseline incidence for the comparison group, these ratios cannot be converted to a lifetime probability. The only longitudinal cohort measuring tooth loss directly (Brusius et al. 2023) involved Brazilian adolescents over 2.5 years; it is not transportable to a US adult lifetime estimate. Dental caries is experienced by nearly 90% of US adults aged 20–64 (NIDCR/NHANES data), making this a near-universal outcome that isolating the marginal contribution of brushing frequency — separate from diet, fluoride toothpaste use, dental care access, and genetics — is intractable in observational studies. No study has measured the effect of occasional missed sessions; all frequency data addresses habitual patterns over months to years. The 30-minute post-meal brushing recommendation derives from in-situ studies under extreme acid challenges (pH 2.9) and has not been validated for ordinary dietary exposures.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":4,"d7":5,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-07-20","methodology_version":"1.2"},"reviewer":"8d-eval-2026-07-20","last_reviewed":"2026-07-20","reviewed":true,"generated_at":"2026-07-19","image":{"alt":"A single manual toothbrush resting on a pale grey surface, flat vector illustration."},"canonical_url":"https://likelier.app/skipping-tooth-brushing","api_url":"https://likelier.app/api/fears/skipping-tooth-brushing.json"},{"slug":"social-media-teen-harm","question":"What are the odds of serious psychological harm to a teenager from social media use?","quick_answer":"There is no reliable per-teen probability of serious psychological harm from social media. The evidence is largely correlational with small average effect sizes, and even the US Surgeon General's advisory says there is not enough evidence to quantify a per-user risk — so no number is published here.\n","category":"tech","tags":["kids","mental-health"],"no_reliable_estimate":true,"perceived":{"description":"The public conversation about teen social media use shifted sharply between 2021 and 2024. The Wall Street Journal's leak of Facebook/Instagram internal research in September 2021 — showing the company's own researchers found Instagram made body-image issues worse for one in three teen girls — moved the discourse from \"maybe a problem\" to \"obvious crisis.\" The US Surgeon General's 2023 Advisory on Social Media and Youth Mental Health, the APA's 2023 health advisory, and Jonathan Haidt's \"The Anxious Generation\" (2024) cemented the framing. Parents, legislators, and most of the press now treat serious psychological harm from social media as something that happens to a substantial share of teenagers. That intuition is not baseless — there is a real association in correlational data, and the subgroup effects (girls 13-15 with pre-existing anxiety, heavy users above 3-4 hours/day) are larger than the population average. But \"substantial share\" is doing a lot of work, and the research base does not yet support a clean per-user probability.\n","rough_estimate":"Parents and policymakers now treat serious harm as a meaningful risk for most teens; researchers disagree on whether the effect sizes justify that level of alarm","kind":"intuition"},"sources":[{"url":"https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/social-media/index.html","title":"Social Media and Youth Mental Health: The U.S. Surgeon General's Advisory","publisher":"Office of the U.S. Surgeon General (2023)","source_type":"govt_report","statistic":"The advisory states there is 'not enough evidence to determine that social media is sufficiently safe for children and adolescents' and calls for urgent action, while noting the evidence base is insufficient to establish causation or quantify per-user risk","excerpt":"\"We do not yet have enough evidence to determine that social media is sufficiently safe for children and adolescents.\"\n","source_date":"2023-05-23","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250120154115/https://www.hhs.gov/surgeongeneral/priorities/youth-mental-health/social-media/index.html","calculation_notes":"The Surgeon General's Advisory is the highest-profile US government statement on teen social media risk. It is cited here as an authoritative framing document, not as a source of a specific probability. The advisory explicitly avoids stating a per-user risk figure and notes the evidence base is insufficient to determine whether social media is safe. It summarises the correlational literature and calls for more research, platform transparency, and legislative action. No native or normalized probability is derived from this source because the entry is flagged no_reliable_estimate.\n"},{"url":"https://www.nature.com/articles/s41562-018-0506-1","title":"The association between adolescent well-being and digital technology use","publisher":"Nature Human Behaviour (Orben & Przybylski 2019)","source_type":"peer_reviewed","statistic":"Across three large-scale datasets (total N > 350,000), the association between digital technology use and adolescent well-being was negative but small: r = −0.035 for social media use specifically, explaining about 0.4% of the variance in well-being — less than the negative association with wearing glasses or eating potatoes","excerpt":"\"The negative effect of technology use on adolescent well-being is small — explaining at most 0.4% of the variation in well-being.\"\n","source_date":"2019-01-14","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260505062724/https://www.nature.com/articles/s41562-018-0506-1","calculation_notes":"Orben & Przybylski 2019 is the most-cited methodological counterweight to the crisis narrative. The specification-curve analysis across all defensible analytic choices produces effect sizes that are statistically significant (large N) but tiny in practical terms. r = −0.035 for social media use means the association explains roughly 0.12% of the variance in well-being. This is not zero, but it is far too small to convert into a per-user probability of \"serious psychological harm\" — the outcome variable is a continuous well-being scale, not a binary harm threshold. This source is the primary basis for the no_reliable_estimate designation: the effect is real but not meaningfully translatable into individual risk.\n","independence_note":"Fully independent of the Surgeon General's Advisory and the Facebook internal research. Uses UK Understanding Society, YRBS, and MCS datasets.\n"},{"url":"https://www.apa.org/topics/social-media-internet/health-advisory-adolescent-social-media-use","title":"Health Advisory on Social Media Use in Adolescence","publisher":"American Psychological Association (2023)","source_type":"reputable_reference","statistic":"The APA identifies social media as neither uniformly beneficial nor uniformly harmful, and notes that effects depend on the content, the adolescent's individual characteristics, and the amount of time spent; it provides no per-user probability of harm","excerpt":"\"Science demonstrates that social media use, in and of itself, is not sufficient to be helpful or harmful to young people.\"\n","source_date":"2023-05-09","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260525100215/https://www.apa.org/topics/social-media-internet/health-advisory-adolescent-social-media-use","calculation_notes":"The APA health advisory is included as a second authoritative institutional statement. Like the Surgeon General's Advisory, it summarises the correlational literature without producing a probability figure. Its key contribution to this entry is the explicit statement that social media use \"in and of itself\" is insufficient to cause harm — effects depend on content type, individual vulnerability, and usage patterns. This supports the no_reliable_estimate framing: there is no single population-level probability that applies to \"a teenager using social media.\"\n"},{"url":"https://www.wsj.com/articles/facebook-knows-instagram-is-toxic-for-teen-girls-company-documents-show-11631620739","title":"Facebook Knows Instagram Is Toxic for Teen Girls, Company Documents Show","publisher":"Wall Street Journal (Wells, Horwitz, Seetharaman 2021)","source_type":"news_article","statistic":"Internal Facebook research found that 32% of teen girls surveyed said Instagram made them feel worse about their body when they already felt bad about it; the research also found Instagram was cited as a contributor to anxiety and depression in a subset of teen users","excerpt":"\"Thirty-two percent of teen girls said that when they felt bad about their bodies, Instagram made them feel worse.\"\n","source_date":"2021-09-14","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250626231438/https://www.wsj.com/articles/facebook-knows-instagram-is-toxic-for-teen-girls-company-documents-show-11631620739","calculation_notes":"The WSJ Facebook Files series leaked internal Instagram research that became the public catalyst for legislative and regulatory action. The \"32% of teen girls\" figure is widely cited but refers to a specific question — \"when you already feel bad about your body, does Instagram make it worse?\" — not to a base rate of developing a clinical condition. The leaked research used convenience-sampled internal surveys, not population-representative methods, and \"felt worse\" is a subjective self-report, not a clinical diagnosis of serious psychological harm. Included because it drove the public perception shift, not because it anchors a probability.\n"},{"url":"https://acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.13190","title":"Annual Research Review: Adolescent mental health in the digital age: facts, fears, and future directions","publisher":"Journal of Child Psychology and Psychiatry (Odgers & Jensen 2020)","source_type":"peer_reviewed","statistic":"Synthesising six prior reviews/meta-analyses plus large-scale preregistered cohort and momentary-assessment studies, the review finds the link between digital technology use and adolescent mental health to be a mix of conflicting small positive, negative, and null associations (e.g. Huang 2017 meta-analysis r = −0.07 overall, depression r ≈ −0.11; McCrae 2017 r = 0.13), with no support for causal claims at the population level","excerpt":"\"At present, narrative reviews and meta-analytic work do not support causal claims, or even strong and consistent correlational patterns, linking adolescents' digital technology usage with mental health problems.\"\n","source_date":"2020-01-17","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250620023923/https://acamh.onlinelibrary.wiley.com/doi/10.1111/jcpp.13190","calculation_notes":"Odgers & Jensen 2020 (J Child Psychol Psychiatry Annual Research Review) synthesises three evidence streams: narrative reviews and meta-analyses (2014-2019), large-scale preregistered cohort studies, and intensive longitudinal / ecological-momentary-assessment studies. Across them the reported associations are small and inconsistent in direction — the Huang (2017) meta-analysis (N = 19,652) found a mean SNS-well-being correlation of r = −0.07 (depression r ≈ −0.11), and the McCrae (2017) meta-analysis found r = 0.13 for depressive symptoms. The review's key contribution to this entry is its conclusion that the evidence base does not support causal claims, \"or even strong and consistent correlational patterns,\" and does not warrant the widespread panic — while the outcome measures remain continuous well-being and symptom scales, not binary harm thresholds. (Earlier drafts of this entry mis-cited this review under a Clinical Psychological Science DOI that in fact resolves to Vuorre & Przybylski 2023, \"Global Well-Being and Mental Health in the Internet Age\" — a separate paper; the identifier was corrected to the real Odgers & Jensen 2020 DOI 10.1111/jcpp.13190 in 2026.)\n","independence_note":"Independent review drawing on a broader evidence base than Orben & Przybylski 2019, including six prior systematic reviews/meta-analyses, large-scale preregistered cohort studies (e.g. Understanding Society, N = 12,672), and diary/EMA designs.\n"},{"url":"https://www.commonsensemedia.org/research/talk-trust-and-trade-offs-how-and-why-teens-use-ai-companions","title":"Talk, Trust, and Trade-Offs: How and Why Teens Use AI Companions","publisher":"Common Sense Media (2025)","source_type":"reputable_reference","statistic":"In a nationally representative survey of 1,060 US teens aged 13-17 (April-May 2025), nearly three in four teens (72%) had used an AI companion at least once and roughly half were regular users; the figure is an adoption rate, not a rate of resulting psychological harm","excerpt":"\"Our new research reveals that nearly three in four teens have used AI companions, and half use them regularly.\"\n","source_date":"2025-07-16","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260714020040/https://www.commonsensemedia.org/research/talk-trust-and-trade-offs-how-and-why-teens-use-ai-companions","calculation_notes":"Added 2026-06-14 to document AI companion chatbots as a distinct emerging teen-harm vector separate from feed-based social media. Common Sense Media's nationally representative survey establishes that the exposure is now mainstream (72% of teens have tried one; ~half are regular users). It is cited strictly as an adoption-prevalence figure: the survey measures how many teens use companion bots, not how many are harmed by them. No native or normalized probability is derived because the entry remains no_reliable_estimate — surveys report adoption, not crisis incidence.\n"},{"url":"https://www.unicef.org/press-releases/deepfake-abuse-is-abuse","title":"'Deepfake abuse is abuse' — Statement by UNICEF on AI-generated sexualised images of children","publisher":"UNICEF (2026)","source_type":"reputable_reference","statistic":"In a UNICEF, ECPAT and INTERPOL study across 11 countries (Disrupting Harm Phase 2, ~1,000 internet-using children aged 12-17 surveyed per country), at least 1.2 million children disclosed having had their images manipulated into sexually explicit deepfakes in the past year; in some countries this represents 1 in 25 children","excerpt":"\"In a UNICEF, ECPAT and INTERPOL study across 11 countries, at least 1.2 million children disclosed having had their images manipulated into sexually explicit deepfakes in the past year. In some countries, this represents 1 in 25 children — the equivalent of one child in a typical classroom.\"\n","source_date":"2026-02-04","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260618185949/https://www.unicef.org/press-releases/deepfake-abuse-is-abuse","calculation_notes":"Added 2026-06-14 to ground the AI-synthetic-media caveat. The original change brief attributed this 1.2-million figure to Thorn; that attribution could not be verified. The figure is actually from the UNICEF/ECPAT/INTERPOL Disrupting Harm Phase 2 study, cited here from the verbatim UNICEF statement. It is a cross-national exposure-disclosure rate (children reporting their images were manipulated into sexual deepfakes), not a per-user probability of psychological harm, so it does not change the no_reliable_estimate status. It documents that generative-AI synthetic media is a victimisation channel distinct from, and sometimes embedded within, social-media platforms.\n"}],"comparison_anchors":[],"regional_breakdown":[{"region":"Average teen user (~2 hrs/day)","probability":0,"notes":"No reliable per-user probability can be assigned. Correlational studies find small negative associations with well-being (r ≈ 0.035-0.15), but these are measured on continuous scales, not as binary harm incidence. For the typical teen user at moderate usage levels, the population-level effect size is too small and too heterogeneous to convert into a meaningful individual risk number."},{"region":"Heavy user (>4 hrs/day)","probability":0,"notes":"Heavy use is associated with larger negative effects in most studies, but the dose-response curve is inconsistent across datasets and may reflect reverse causation (teens who are already struggling use social media more). No study has produced a reliable incidence rate for clinical-level harm among heavy users specifically."},{"region":"Teen girl with pre-existing anxiety/depression","probability":0,"notes":"The largest documented subgroup effects in the literature. Facebook/Instagram internal research (2021) and several external studies find that girls with pre-existing internalising symptoms report the most negative experiences. Effect sizes are larger than the population average but still measured on well-being scales rather than as binary harm rates. This is the subgroup where the case for a clinically meaningful effect is strongest and where a per-user probability would be most useful — but the evidence does not yet support one."},{"region":"Teen boy (average user)","probability":0,"notes":"Most studies find smaller or null associations between social media use and well-being outcomes in boys compared to girls. Some studies find positive associations with social connection. The gender difference is one of the more consistent findings in the literature, though it may partly reflect differences in platform use patterns (visual comparison-based content vs. gaming/interest-based content) rather than a biological sex difference."},{"region":"Teen bonded to an AI companion chatbot","probability":0,"notes":"A distinct and newer exposure from feed-based social media. Common Sense Media (2025) found 72% of US teens have used an AI companion and about half use them regularly, but adoption rate is not a harm rate. The hypothesised mechanism — a private, always-available, sycophantic chat that can isolate a teen and fail to interrupt an escalating crisis — is drawn from individual cases (the Sewell Setzer III suicide and Garcia v. Character.AI; the OpenAI/Adam Raine case) and litigation now settling (the Google/Character.AI settlement in principle, January 2026), not from any cohort study with an incidence rate. No reliable per-user probability exists for crisis or self-harm outcomes tied to companion-bot use; the evidence base is adoption surveys and case reports, not population data."}],"short_label":"Teen social media","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"mental_trauma","valence":"negative","caveats":"This entry is flagged no_reliable_estimate because the research literature does not support a defensible per-user probability of \"serious psychological harm\" from social media use. The reasons are structural, not temporary.\nFirst, \"serious psychological harm\" is not a standardised outcome. The studies cited here measure well-being, life satisfaction, depressive symptoms, anxiety symptoms, body image, and self-reported mood on continuous scales. Converting a small negative shift on a well-being scale into a binary \"harmed / not harmed\" count requires an arbitrary threshold that the literature has not agreed on.\nSecond, the effect sizes in the best-powered correlational studies are small (r = 0.035-0.15). That range is statistically significant with large samples but explains less than 2% of the variance in well-being. It is not zero — the association is real — but it is far too small and too heterogeneous to support the claim that social media harms a specific fraction of its users.\nThird, causation is genuinely contested. The Surgeon General, the APA, and most reviewers note that the evidence is largely correlational. Natural experiments (e.g., Braghieri, Levy, and Makarin 2022 on the Facebook college rollout) suggest modest causal effects on some mental health measures, but these are concentrated in specific subgroups and time periods. Reverse causation — teens with pre-existing mental health difficulties using social media more — is a plausible partial explanation that most studies cannot fully rule out.\nFourth, the subgroup most at risk (girls aged 13-15 with pre-existing internalising symptoms, using comparison-heavy visual platforms for 3+ hours daily) is real and clinically important, but producing a per-user probability for that subgroup would require arbitrary assumptions about the threshold between \"felt worse\" and \"serious psychological harm.\" This entry declines to make those assumptions.\nThe regional_breakdown rows use probability 0.0 as a placeholder to indicate that no reliable estimate is available for any subgroup — not that the probability is literally zero.\nFifth, a newer exposure has emerged that this entry's original sources predate: AI companion chatbots. These are private, always-available, often sycophantic one-on-one conversations rather than a public feed, and Common Sense Media's 2025 nationally representative survey found 72% of US teens have used one. The proposed mechanism for serious harm — a bond with a bot that isolates a teen and fails to interrupt, or even reinforces, an escalating suicidal crisis — is so far evidenced by individual cases (the Sewell Setzer III suicide and Garcia v. Character.AI; the OpenAI/Adam Raine case), a September 2025 Senate hearing, a company response (Character.AI's October 2025 ban on open-ended chats for under-18 users), and a January 2026 settlement in principle between the families, Character.AI, and Google — not by any cohort study yielding a crisis or self-harm incidence rate. The companion-bond mechanism is plausibly distinct from the feed-based effects the rest of this entry covers, and it remains unquantified for the same structural reason: adoption surveys and case reports are not population incidence data.\nA related synthetic-media channel is AI-generated sexual deepfakes of minors. A UNICEF/ECPAT/INTERPOL study (Disrupting Harm Phase 2) found at least 1.2 million children across 11 countries disclosed having had their images manipulated into sexually explicit deepfakes in the past year — in some countries 1 in 25 children. This is a victimisation channel often embedded in, but conceptually separate from, ordinary social-media use, and like the companion-bond vector it is a disclosure or exposure rate, not a probability of \"serious psychological harm.\" Both are recorded here as caveats rather than estimates: the entry stays no_reliable_estimate.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.875,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-11-agent","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A single smartphone lying face-down on a plain surface, its screen dark, flat vector illustration in muted blue-grey tones."},"canonical_url":"https://likelier.app/social-media-teen-harm","api_url":"https://likelier.app/api/fears/social-media-teen-harm.json"},{"slug":"task-job-scam","question":"How likely is a US adult to lose money to a gamified \"task\" or job scam?","quick_answer":"There is no reliable per-person estimate. The FTC's figures count reported losses, not how many people were targeted, and because the agency says the vast majority of fraud goes unreported, no per-adult rate can be derived; the scam is also only about two years old.\n","category":"other","tags":["digital-fraud","workplace"],"no_reliable_estimate":true,"perceived":{"description":"No survey measures how likely people think they are to fall for a task or job scam specifically. The mechanism is roughly two years old, so public intuition is unformed; most people who have heard of it assume it could not happen to them, while the FTC notes the gamified payout loop is engineered precisely to defeat that assumption. Treated as an intuition placeholder, not a measured perception.","rough_estimate":"Most people assume near-zero personal risk; the scam's design specifically targets that overconfidence.","kind":"intuition"},"sources":[{"url":"https://www.ftc.gov/news-events/data-visualizations/data-spotlight/2024/12/paying-get-paid-gamified-job-scams-drive-record-losses","title":"Paying to get paid: gamified job scams drive record losses","publisher":"U.S. Federal Trade Commission","source_type":"govt_report","statistic":"~20,000 task-scam reports in H1 2024 vs ~5,000 in all of 2023; 38.8% of job-scam reports were task scams in H1 2024 (vs 5.6% in 2023, 1.6% in 2022, 0.6% in 2021, 0% in 2020); job-scam losses topped $220M in H1 2024; crypto losses to job scams ~$41M in H1 2024 vs ~$21M in all of 2023.","excerpt":"Reported losses to job scams increased more than threefold from 2020 to 2023 and, in just the first half of 2024, topped $220 million. Driving this trend are skyrocketing reports about gamified job scams, often called task scams. About 20,000 people reported these scams in the first half of the year, compared to about 5,000 in all of 2023. Since the vast majority of frauds are not reported, this likely reflects only a fraction of the actual harm.","source_date":"2024-12-12","source_accessed":"2026-06-13","calculation_notes":"Footnote [2] verbatim: task-scam shares are estimates from hand-coding a random sample of 500 job-scam reports per year (2,500 total), excluding reports without a narrative; the coded share was multiplied by total job-scam reports that year. Shares: 38.8% (H1 2024), 5.6% (2023), 1.6% (2022), 0.6% (2021), 0% (2020). Footnote [1] gives job-scam dollar losses: $223M (2024 through June), $286M (2023), $179M (2022), $131M (2021), $90M (2020). These are reported figures, not victimization-prevalence counts; the FTC states the vast majority of frauds go unreported, so no per-adult denominator can be derived.","independence_note":"Primary FTC Consumer Sentinel analysis; the consumer alert and press-release sources below are FTC restatements of the same dataset, not independent."},{"url":"https://consumer.ftc.gov/consumer-alerts/2024/11/task-scams-create-illusion-making-money","title":"Task scams create the illusion of making money","publisher":"U.S. Federal Trade Commission","source_type":"govt_report","statistic":"Describes the task-scam mechanism: unsolicited text/WhatsApp offer, repetitive 'tasks' with fake mounting commissions, then a demand to deposit the victim's own money (usually crypto) to unlock the next set and withdraw fake earnings, after which both the deposit and the fake earnings are lost.","excerpt":"At some point, the app or online platform will ask you to deposit your own money — usually in crypto — to complete your next set of tasks and to get your supposed (fake) earnings out of the app. But if you make the deposit, not only is your real money is gone, you'll never get those fake earnings.","source_date":"2024-12-12","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260605135113/https://consumer.ftc.gov/consumer-alerts/2024/11/task-scams-create-illusion-making-money","calculation_notes":"Used for the mechanism description (deposit-in-crypto flip), not for any rate. No numerator/denominator extractable. Page is dated December 12, 2024 (URL path says 2024/11).","independence_note":"FTC consumer-facing restatement of the same Spotlight analysis; not statistically independent of the Spotlight."},{"url":"https://www.ftc.gov/news-events/news/press-releases/2024/12/new-ftc-data-show-skyrocketing-consumer-reports-about-game-online-job-scams","title":"New FTC Data Show Skyrocketing Consumer Reports About Game-Like Online Job Scams","publisher":"U.S. Federal Trade Commission","source_type":"govt_report","statistic":"Task-scam reports rose from zero in 2020 to 5,000 in 2023, then quadrupled to ~20,000 in H1 2024; task scams ~40% of 2024 job-scam reports; overall job-scam losses >$220M in H1 2024; crypto losses to job scams ~$41M in H1 2024.","excerpt":"Overall reported losses on job scams tripled from 2020 to 2023 and were more than $220 million in just the first six months of 2024. Task scams were estimated to account for nearly 40 percent of the 2024 job scam reports.","source_date":"2024-12-12","source_accessed":"2026-06-13","calculation_notes":"Corroborates the Spotlight's report counts and loss figures verbatim. No prevalence denominator stated.","independence_note":"FTC press release restating the Spotlight; not independent."},{"url":"https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","title":"2024 IC3 Annual Report (Internet Crime Report)","publisher":"FBI Internet Crime Complaint Center (IC3)","source_type":"govt_report","statistic":"Secondary summaries report 6,533 employment-scam complaints with ~$197.22M in losses within the 2024 cryptocurrency-nexus breakdown; total 2024 IC3 losses ~$16.6B. NOT independently confirmed from the primary PDF.","excerpt":"[Unable to extract verbatim text — the IC3 2024 report PDF is image-based and did not render machine-readable text via WebFetch or Playwright. The 6,533 complaints / $197.22M figure appears only in third-party summaries and was NOT verified against the primary document.]","source_date":"2025-04-23","source_accessed":"2026-06-13","archive_url":"http://web.archive.org/web/20260709082752/https://www.ic3.gov/AnnualReport/Reports/2024_IC3Report.pdf","calculation_notes":"Listed for transparency as a corroborating-but-unverified pointer. Per the project's hard sourcing rule this source is NOT treated as confirmed and is NOT relied on for any claim; the entry stands on the three CONFIRMED FTC sources above.","independence_note":"Independent of FTC (separate FBI complaint channel) but unverified at primary-source level."}],"comparison_anchors":[{"label":"Losing money to any online scam (lifetime, context only)","lifetime_us_adult":0.4},{"label":"Credit-card fraud (lifetime, context only)","lifetime_us_adult":0.65}],"short_label":"Task / job scam loss","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"financial","valence":"negative","caveats":"No defensible per-adult lifetime probability exists for this risk, so the entry is marked no_reliable_estimate. Three problems block a number. (1) The mechanism is roughly two years old: the FTC's hand-coded sample found 0% of job-scam reports were task scams in 2020, 0.6% in 2021, 1.6% in 2022, 5.6% in 2023, and 38.8% in the first half of 2024 — a single year of meaningful data on a steeply rising curve. Projecting a 2024 annual rate across a 59-year lifetime horizon would assume the spike is stationary for decades, an assumption the data cannot support. (2) The FTC figures are reports, not victims, and the agency states \"the vast majority of frauds are not reported,\" with no published task-scam-specific underreporting multiplier. (3) There is no exposed-population denominator: the at-risk group is people who engage with these platforms, which is unmeasured. For scale only: ~20,000 reports in H1 2024 against a US adult population near 262 million is an order of 0.0001 per year as reported, and even a generous 20-40x underreporting adjustment over one year lands far below any credible lifetime figure once the stationarity problem is added. No personal_factor_multipliers are included because the headline carries no number. The comparison_anchors are flagged as umbrella context (any online scam loss, credit-card fraud), not brackets around a point estimate. 8D self-score clears thresholds for a no_reliable_estimate entry: D3/D4/D5 are scored on whether the no-estimate decision is correctly justified, not on absent arithmetic.","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-13","methodology_version":"8d-v1"},"reviewer":"8d-eval-2026-06-13","last_reviewed":"2026-06-13","reviewed":true,"generated_at":"2026-06-13","image":{"alt":"An abstract editorial symbol evoking \"Task / job scam loss\", muted two-tone palette, flat vector."},"canonical_url":"https://likelier.app/task-job-scam","api_url":"https://likelier.app/api/fears/task-job-scam.json"},{"slug":"tattoo-during-pregnancy","question":"What are the odds of harm to a pregnancy from getting a tattoo while pregnant?","quick_answer":"No reliable pregnancy-specific estimate exists. The risk has never been studied in pregnant people, and medical advice to wait reflects an absence of evidence rather than evidence of serious harm.\n","category":"kids","tags":["pregnancy"],"no_reliable_estimate":true,"perceived":{"description":"Views on tattooing during pregnancy split roughly into two camps: those who assume a regulated professional studio makes it safe enough, and those who imagine serious risk from unknown chemicals and infection. Few pregnant women have access to any published estimate, because none exists in peer-reviewed form for a pregnant population. Medical guidance uniformly recommends waiting, but the reason is absence of evidence rather than evidence of serious harm.\n","kind":"intuition"},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC5290255/","title":"The Risk of Bacterial Infection After Tattooing","publisher":"Deutsches Ärzteblatt International / Dieckmann et al.","source_type":"peer_reviewed","statistic":"0.5% to 6% of tattooed individuals experience infectious complications after being tattooed","excerpt":"\"Based on published surveys, between 0.5% and 6% of the people with a tattoo experienced infectious complications after being tattooed.\"\n","source_date":"2016-09-09","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20260417064014/https://pmc.ncbi.nlm.nih.gov/articles/PMC5290255/","calculation_notes":"Dieckmann et al. (2016) systematic review of 67 non-mycobacterial tattoo-related infections from 1,379 records (1984–2015). General-population data only; no pregnancy-specific subset exists. Staphylococcus aureus accounted for 81% of cases; two deaths documented, both in immunocompromised individuals. The 0.5–6% range is not stratified by pregnancy status. Pregnancy-associated immune modulation could plausibly shift this range upward, but no data support quantification of the effect.\n","independence_note":"Systematic review of published case series and surveys; not a cohort study. Inclusion limited to reports with microbiological confirmation.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11968518/","title":"Tat_BioV: tattoo ink exposure and biokinetics of selected tracers in a short-term clinical study","publisher":"Archives of Toxicology / Kochs et al.","source_type":"peer_reviewed","statistic":"Mean systemic tattoo ink absorption per session estimated at 0.19 ± 0.29 g; tracer metabolites detected in plasma and urine within 24 hours","excerpt":"\"Mean systemic absorption per session was estimated at 0.19 ± 0.29 g. Urinary excretion was approximately 26% of administered tracers within 24 hours of tattooing. Intradermal administration produced about twice as much metabolite compared to peroral intake.\"\n","source_date":"2025-03-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250410142759/https://pmc.ncbi.nlm.nih.gov/articles/PMC11968518/","calculation_notes":"Kochs et al. (2025) is the first clinical study to quantify systemic tattoo ink absorption in vivo (n=24 healthy, non-pregnant adults). Mean absorption of 0.19 g per session; tracer metabolites reached plasma and urine within 24 hours, demonstrating systemic distribution. Placental crossing or fetal exposure was not addressed; no pregnant participants were enrolled. The study establishes that systemic absorption is real and measurable but leaves the gestational significance unknown.\n","independence_note":"Prospective clinical pharmacokinetics study; independent of the Dieckmann infection-risk review in both methodology and population.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/30806488/","title":"Women and Tattoos: Fashion, Meaning, and Implications for Health","publisher":"Journal of Midwifery & Women's Health / Farley, Van Hoover & Rademeyer","source_type":"peer_reviewed","statistic":"No published studies report specific complication rates from tattooing during pregnancy; tattoo complications include hypersensitivity, infection, and regret","excerpt":"\"Tattoo complications include hypersensitivity, infection, and regret. Topics of discussion include implications for health, including pregnancy and breastfeeding. No data exist to guide recommendations about tattooing during pregnancy.\"\n","source_date":"2019-02-14","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20251009160506/https://pubmed.ncbi.nlm.nih.gov/30806488/","calculation_notes":"Farley et al. (2019) is the most comprehensive peer-reviewed review of tattooing among women, including pregnancy and lactation implications. The authors explicitly note the absence of population-level data on tattooing during pregnancy, which directly supports the no_reliable_estimate flag for this entry. No arithmetic transformation applied.\n","independence_note":"Narrative review by midwifery and public health authors at University of North Carolina; independent of the Dieckmann and Kochs laboratory-based studies.\n"},{"url":"https://www.who.int/news-room/questions-and-answers/item/hepatitis-preventing-mother-to-child-transmission-of-the-hepatitis-b-virus","title":"Hepatitis: Preventing mother-to-child transmission of the hepatitis B virus","publisher":"World Health Organization","source_type":"govt_report","statistic":"Absent prophylaxis, hepatitis B mother-to-child transmission risk is 70-90% for mothers with high HBV viral load or who are HBeAg-positive, versus 10-40% for HBeAg-negative mothers","excerpt":"\"The risk of transmission from mother to child ranges from 70% to 90% for mothers with high HBV viral load (or are HBeAg-positive) and from 10% to 40% for those that are HBeAg negative.\"\n","source_date":"2024-01-01","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260301015338/https://www.who.int/news-room/questions-and-answers/item/hepatitis-preventing-mother-to-child-transmission-of-the-hepatitis-b-virus","calculation_notes":"WHO general guidance on hepatitis B mother-to-child transmission, cited only to support the entry's statement that tattooing with unsterile equipment is a recognized HBV exposure route with well-documented vertical-transmission consequences if the mother becomes infected. This source is general HBV epidemiology, not tattoo-specific or pregnancy-and-tattoo-specific, and does not quantify the probability that a tattoo itself causes maternal HBV infection.\n","independence_note":"WHO general hepatitis B guidance; independent of the tattoo-specific sources above.\n"},{"url":"https://academic.oup.com/cid/article/54/8/1167/364191","title":"Transmission of Hepatitis C Virus Infection Through Tattooing and Piercing: A Critical Review","publisher":"Clinical Infectious Diseases / Tohme & Holmberg (CDC)","source_type":"peer_reviewed","statistic":"Studies that specified tattooing venue found no definitive increase in HCV risk from professional parlors; the significant risk (adjusted OR 2.0-3.6) was concentrated in prison settings and tattoos done by friends outside a professional setting","excerpt":"\"Studies that specified the venue of tattooing and/or piercing showed no definitive evidence for an increased risk of HCV infection when tattoos and piercings were received in professional parlors.\" \"However, the risk of HCV infection is significant, especially among high-risk groups (adjusted odds ratio, 2.0-3.6), when tattoos are applied in prison settings or by friends.\"\n","source_date":"2012-04-15","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20241209125244/https://academic.oup.com/cid/article/54/8/1167/364191","calculation_notes":"Tohme & Holmberg (2012), Clinical Infectious Diseases 54(8):1167-1178, a CDC-authored critical review of HCV transmission via tattooing and piercing (general population, not pregnancy-specific). Supports the entry's statement that the HCV risk signal from tattooing concentrates in unregulated, non-professional settings rather than licensed studios; does not address pregnancy-specific risk or fetal outcomes.\n","independence_note":"Independent CDC-authored literature review; not specific to pregnant populations.\n"}],"comparison_anchors":[],"short_label":"Tattoo in pregnancy","outcome_severity":"moderate_harm","exposure_pattern":"acute","outcome_type":"chronic_illness","valence":"negative","caveats":"No published study has measured fetal outcomes, placental ink transfer, or infection rates specifically in pregnant women who received tattoos during pregnancy. The no_reliable_estimate flag reflects this data gap, not the absence of theoretical risk. Infection risk (0.5–6% in the general population) is biologically plausible during pregnancy given immune modulation; systemic ink absorption occurs in healthy adults and placental crossing cannot be ruled out. Kluger (2015, Current Problems in Dermatology) lists pregnancy among the main contraindications to tattooing based on precautionary reasoning. NHS patient guidance similarly advises waiting until after delivery. The consensus recommendation to avoid tattooing during pregnancy derives from the absence of safety data rather than a documented harm signal, and waiting until after delivery is a low-cost precaution.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A tattoo needle and inkwell resting beside a pregnancy calendar, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/tattoo-during-pregnancy","api_url":"https://likelier.app/api/fears/tattoo-during-pregnancy.json"},{"slug":"tea-bag-microplastics","question":"What are the odds that microplastics released from a plastic tea bag will harm your health?","quick_answer":"There is no reliable estimate that tea-bag microplastics harm your health. A plastic tea bag releases roughly 11.6 billion microplastic particles into a cup, but that figure measures exposure, not harm; the only toxicity test in the key study was on water fleas, not people.\n","category":"food","tags":["food","household"],"no_reliable_estimate":true,"perceived":{"description":"Plastic tea bags became a discrete consumer anxiety in 2019, when a McGill University study reported that a single \"silken\" pyramid bag releases billions of plastic particles into one cup of tea. The figure — 11.6 billion microplastics per cup — travelled far beyond the journal, framed as a hidden contamination route in an everyday ritual most people never thought of as plastic at all. A second wave of coverage followed a 2024 Barcelona study showing the same particles being taken up by human intestinal cells. The underlying intuition is the familiar detection-equals-danger heuristic that drives the wider microplastics fear: if measurable plastic ends up in the cup and then in the gut, the exposure must be doing something. No published survey isolates tea-bag worry as its own concern, but it sits inside the broader microplastics anxiety, where EFSA's 2025 Eurobarometer found 63% of EU citizens aware of the issue and chemical contaminants rank among the top food-safety concerns in US surveys.\n","rough_estimate":"Plastic tea bags are widely framed as a meaningful microplastic source that should be swapped for paper or loose-leaf","kind":"intuition"},"sources":[{"url":"https://pubs.acs.org/doi/10.1021/acs.est.9b02540","title":"Plastic Teabags Release Billions of Microparticles and Nanoparticles into Tea","publisher":"Environmental Science & Technology — Hernandez, Xu, Larsson, Tahara, Maisuria, Tufenkji (McGill University)","source_type":"primary_study","statistic":"Steeping a single plastic teabag at brewing temperature (95 °C) releases ~11.6 billion microplastics and ~3.1 billion nanoplastics into one cup; particles matched to nylon and polyethylene terephthalate; levels several orders of magnitude higher than plastic loads previously reported in other foods","excerpt":"\"We show that steeping a single plastic teabag at brewing temperature (95 °C) releases approximately 11.6 billion microplastics and 3.1 billion nanoplastics into a single cup of the beverage. The composition of the released particles is matched to the original teabags (nylon and polyethylene terephthalate) [...] The levels of nylon and polyethylene terephthalate particles released from the teabag packaging are several orders of magnitude higher than plastic loads previously reported in other foods. An initial acute invertebrate toxicity assessment shows that exposure to only the particles released from the teabags caused dose-dependent behavioral and developmental effects.\"\n","source_date":"2019-11-05","source_accessed":"2026-06-09","archive_url":"http://web.archive.org/web/20260710033301/https://pubs.acs.org/doi/10.1021/acs.est.9b02540","calculation_notes":"Hernandez et al. (2019, Env. Sci. Technol. 53(21):12300-12310, DOI 10.1021/acs.est.9b02540) cut and steeped commercial plastic teabags at 95 °C and counted released particles by electron microscopy, confirming polymer identity by FTIR and XPS. The headline 11.6 billion microplastic + 3.1 billion nanoplastic figure is PER CUP. The only toxicity test in the paper was on the water flea Daphnia magna — an invertebrate model showing dose-dependent behavioural and developmental effects, not a human-health outcome. The paper quantifies an exposure; it does not demonstrate harm to people.\n","independence_note":"Independent McGill University laboratory study using direct particle counting and spectroscopy; not funded by EFSA, FDA, or any industry body.\n"},{"url":"https://www.sciencedirect.com/science/article/pii/S0045653524026377","title":"Teabag-derived micro/nanoplastics (true-to-life MNPLs) as a surrogate for real-life exposure scenarios","publisher":"Chemosphere — Banaei et al. (Universitat Autònoma de Barcelona / PlasticHeal)","source_type":"peer_reviewed","statistic":"Polymer-based teabags release micro/nanoplastics on infusion: polypropylene ~1.2 billion particles/mL (avg 136.7 nm), cellulose ~135 million/mL (244 nm), nylon-6 ~8.18 million/mL; particles taken up by human intestinal cells in vitro, highest in mucus-producing cells, reaching the cell nucleus","excerpt":"[Paraphrase from the UAB/PlasticHeal press summary — full text paywalled; article verified to exist at DOI 10.1016/j.chemosphere.2024.143736]: commercial teabags made of nylon-6, polypropylene, and cellulose release billions of micro- and nanoplastics during infusion; the particles are absorbed by human intestinal cells in culture, with the highest uptake in mucus-producing cells and particles reaching the cell nucleus — the first demonstration of this uptake route for teabag-derived particles. The verbatim per-millilitre release figures are quoted from the companion UAB press release (see the following source).\n","source_date":"2024-11-16","source_accessed":"2026-06-09","archive_url":"http://web.archive.org/web/20250828153243/https://www.sciencedirect.com/science/article/pii/S0045653524026377","calculation_notes":"Banaei et al. (2024, Chemosphere 368:143736) characterised micro/nanoplastics from three teabag polymers and ran in-vitro uptake experiments on human intestinal cell lines. Release is reported PER MILLILITRE, not per cup, and by polymer — so these figures are NOT directly comparable to Hernandez's per-cup totals and must not be stacked into a single larger number. The novel finding is cellular: uptake of the particles by human gut cells, including nuclear localisation. This is a mechanistic in-vitro result that establishes a plausible pathway; it is not measured disease in a human population. Full text is paywalled, so the precise numbers are cited from the institution's own press release (next source) where they appear verbatim.\n","independence_note":"Independent Spanish laboratory study published in Chemosphere under the EU PlasticHeal project; methodologically distinct from the Hernandez 2019 quantification work.\n"},{"url":"https://www.eurekalert.org/news-releases/1068917","title":"Commercial tea bags release millions of microplastics when in use","publisher":"EurekAlert! / Universitat Autònoma de Barcelona (press release for Banaei et al., Chemosphere 2024)","source_type":"news_article","statistic":"Per-millilitre release by polymer: polypropylene ~1.2 billion particles/mL (avg 136.7 nm); cellulose ~135 million/mL (244 nm); nylon-6 ~8.18 million/mL; mucus-producing intestinal cells showed highest uptake, with particles entering the cell nucleus","excerpt":"\"The tea bags used for the research were made from the polymers nylon-6, polypropylene and cellulose. [...] polypropylene releases approximately 1.2 billion particles per milliliter, with an average size of 136.7 nanometres; cellulose releases about 135 million particles per milliliter, with an average size of 244 nanometres; while nylon-6 releases 8.18 million particles per milliliter [...] mucus-producing intestinal cells had the highest uptake of micro and nanoplastics, with the particles even entering the cell nucleus that houses the genetic material.\"\n","source_date":"2024-12-20","source_accessed":"2026-06-09","archive_url":"http://web.archive.org/web/20260710033534/https://www.eurekalert.org/news-releases/1068917","calculation_notes":"Institutional press release issued by the Universitat Autònoma de Barcelona summarising Banaei et al. (Chemosphere 2024). Included as the verbatim home of the per-millilitre release figures, which are quoted exactly here rather than paraphrased from the paywalled journal text. A press release is not an independent estimate; it restates the peer-reviewed study above and is treated as methodologically linked to it.\n","independence_note":"Not independent of the Banaei 2024 study — it is the authors' own institutional summary, used solely to attach the verbatim numbers to a publicly readable URL.\n"},{"url":"https://www.who.int/publications/i/item/9789241516198","title":"Microplastics in Drinking-Water","publisher":"World Health Organization","source_type":"govt_report","statistic":"No reliable evidence that microplastic exposure at current ingestion levels poses a measurable risk to human health; particles larger than 150 micrometres unlikely to be absorbed from the gut","excerpt":"\"Based on the limited information we have, microplastics in drinking water don't appear to pose a health risk at current levels. But we need to find out more.\"\n","source_date":"2019-08-22","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20260426203936/https://www.who.int/publications/i/item/9789241516198","calculation_notes":"WHO's 2019 review synthesised global evidence on microplastic ingestion and human health. Its key technical conclusion applies directly to tea-bag exposure: particles larger than ~150 µm are unlikely to be absorbed across the intestinal wall, and gut uptake of smaller particles is expected to be limited. Tea-bag nanoplastics (Hernandez and Banaei both report sub-micron particles) fall well below that threshold, into the size range where uptake is plausible but unquantified at the population level. WHO did not identify a quantifiable population-level risk from any current microplastic ingestion source, supporting the no_reliable_estimate classification.\n","independence_note":"WHO global review independent of EFSA process and individual primary studies.\n"},{"url":"https://www.efsa.europa.eu/en/supporting/pub/en-9733","title":"Literature review on micro- and nanoplastic release from food contact materials during their use","publisher":"European Food Safety Authority (EFSA Supporting Publication EN-9733)","source_type":"govt_report","statistic":"EFSA 2025 review of FCM-derived micro- and nanoplastic release concluded that current studies likely overestimate quantities, nanoplastic data remain insufficient, and reliable exposure estimates are not yet possible; formal scientific opinion on microplastics in food not expected until end of 2027","excerpt":"\"While there is clear evidence of microplastic release from food contact materials, the actual quantities are likely lower than many studies suggest. Nanoplastics data remain insufficient, and current evidence does not support reliable exposure estimates.\"\n","source_date":"2025-10-28","source_accessed":"2026-05-24","archive_url":"http://web.archive.org/web/20251028095345/https://www.efsa.europa.eu/en/supporting/pub/en-9733","calculation_notes":"EFSA's October 2025 literature review covered over 100 studies on micro- and nanoplastic release from food contact materials, with mechanical stress (abrasion, friction, brewing heat) identified as a dominant release pathway. EFSA judged that current quantification studies likely overstate exposure because measurement methodology tends to over-count, and that nanoplastic data are too sparse for reliable risk assessment. The formal scientific opinion on microplastics in food is scheduled for end of 2027 — until then no European regulatory threshold exists for tea-bag-derived microplastic exposure. The absence of a quantified opinion after years of review is itself the evidence: this is a pre-epidemiological field.\n","independence_note":"EFSA literature review is independent EU regulatory analysis; covers but does not endorse individual primary studies including Hernandez 2019.\n"}],"comparison_anchors":[{"label":"Harm from pesticide residue on US produce (lifetime)","lifetime_us_adult":0.000001},{"label":"Death in a plane crash (lifetime, US adult)","lifetime_us_adult":0.000017}],"regional_breakdown":[{"region":"Plastic mesh / pyramid ('silken') tea bags","probability":0,"notes":"Nylon and PET pyramid bags are the source of the synthetic-polymer particles in the headline studies (Hernandez 2019; Banaei 2024). Highest synthetic-polymer release. No epidemiological study has linked this exposure to measurable human health outcomes.\n"},{"region":"Paper tea bags / loose-leaf tea","probability":0,"notes":"Avoid the synthetic-polymer shedding almost entirely. (Some paper bags use a thin polypropylene heat-seal, a much smaller plastic source than a full mesh bag.) Switching bag type is the only uncontested behavioural lever, but no human outcome data convert it into a quantified risk reduction.\n"},{"region":"Heavy daily tea drinker (several cups/day)","probability":0,"notes":"Particle ingestion scales with cups brewed and with brewing temperature, but no cumulative-dose study has measured a human disease endpoint from tea-bag particles.\n"}],"short_label":"Tea-bag microplastics","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers human-health risk from micro- and nanoplastics shed by plastic tea bags (nylon, polyethylene terephthalate, polypropylene) during brewing. It does not cover chemical or plasticiser leaching, and it does not replace the broader microplastics-health-harm entry, which addresses microplastic exposure from all sources. The no_reliable_estimate designation reflects the current state of evidence: particle release is reproducibly measured in the billions per cup (Hernandez 2019) and per millilitre by polymer (Banaei 2024); a 2024 in-vitro study shows uptake of the particles by human intestinal cells, including nuclear localisation; and the only 2019 toxicity test showed dose-dependent effects in water fleas. None of that is a measured human disease outcome. The Hernandez per-cup figures and the Banaei per-millilitre figures use different units and different bag materials and must not be combined. WHO notes particles above ~150 µm are unlikely to be absorbed from the gut; EFSA judges current quantification likely overstates exposure and will not issue a formal opinion until end of 2027. Cellulose (paper) bags also shed large particle counts, but those are plant fibre rather than synthetic polymer — the same distinction that complicates the plastic-versus-wood cutting-board comparison.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-09","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-09","last_reviewed":"2026-06-09","reviewed":true,"generated_at":"2026-06-09","image":{"alt":"A single plain tea bag with its string and paper tag resting beside an empty mug on a neutral surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/tea-bag-microplastics","api_url":"https://likelier.app/api/fears/tea-bag-microplastics.json"},{"slug":"testicular-heat-exposure","question":"What are the odds of harming your fertility or testicles from laptops, saunas, phones in pockets, or tight clothing?","quick_answer":"There is no single reliable probability, because the outcomes split into unrelated regimes with no common denominator. Sustained scrotal heat can cause measurable, usually reversible sperm changes, but permanent infertility from everyday sources like laptops or tight clothing is rarely demonstrated, and there is no evidence linking them to cancer.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"The folk fear bundles three separate questions into one anxious package: \"Will my sperm get worse?\", \"Will I become infertile?\", and \"Will I get cancer?\" The first has real evidence behind it — controlled studies show measurable semen parameter changes from sustained scrotal heat. The second is almost never demonstrated for lifestyle exposures. The third has no supporting evidence whatsoever. Online forums and pop-health coverage rarely distinguish between these regimes, which inflates a modest and reversible physiological effect into an existential reproductive threat.\n","rough_estimate":"many men assume a meaningful chance of permanent fertility damage or cancer from everyday heat sources","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/15591087/","title":"Increase in scrotal temperature in laptop computer users","publisher":"Human Reproduction / Sheynkin et al.","source_type":"primary_study","statistic":"Scrotal temperature rose 2.8°C (right) and 2.6°C (left) after 60 minutes with a working laptop on the lap (n=29)","excerpt":"\"ScT increased significantly on the right and left side in the group with working LC (2.8 degrees C and 2.6 degrees C, respectively; P<0001).\"\n","source_date":"2005-02-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20251124192106/https://pubmed.ncbi.nlm.nih.gov/15591087/","calculation_notes":"Sheynkin's 29-volunteer crossover study is the canonical measurement of laptop-induced scrotal hyperthermia. The 2.6-2.8°C rise exceeds the ~1-2°C threshold generally associated with spermatogenic impairment in the literature. However, this study measured temperature only — not semen parameters, fertility, or clinical outcomes. The causal chain from temperature rise → semen parameter change → subfertility → infertility involves multiple uncertain steps. Used here to establish the mechanistic plausibility of heat-mediated semen parameter changes from laptop use.\n","independence_note":"Independent experimental measurement; does not share data or methodology with the Shefi wet-heat study, the Durairajanayagam review, or the underwear study.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/17335598/","title":"Wet heat exposure: a potentially reversible cause of low semen quality in infertile men","publisher":"International Brazilian Journal of Urology / Shefi et al.","source_type":"primary_study","statistic":"45% of infertile men who stopped hot tub/bath use showed a mean 491% increase in total motile sperm count; motility rose from 12% to 34% (p=0.02)","excerpt":"\"Five patients (45%) responded favorably to cessation of heat exposure and had a mean increase in total motile sperm counts of 491%. This increase was largely the result of a statistically significant increase in sperm motility from a mean of 12% at baseline to 34% post-intervention (p = 0.02).\"\n","source_date":"2007-01-01","source_accessed":"2026-04-17","archive_url":"https://web.archive.org/web/20260503083252/https://pubmed.ncbi.nlm.nih.gov/17335598/","calculation_notes":"Small cohort (n=11 infertile men with known wet-heat exposure history). The 491% recovery figure is dramatic but comes from a pre-selected population of men already diagnosed as infertile with known heat exposure. Non-responders had significantly higher smoking histories (5.6 vs 0.11 pack-years), suggesting confounding. This study demonstrates reversibility of heat-induced semen changes, not incidence of harm in the general population. The relevant takeaway is that even substantial semen parameter degradation from wet heat is recoverable once exposure stops.\n","independence_note":"Independent clinical cohort study; different institution (UCSF) and methodology from Sheynkin's temperature measurement study.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25456164/","title":"Causes, effects and molecular mechanisms of testicular heat stress","publisher":"Reproductive BioMedicine Online / Durairajanayagam, Agarwal & Ong","source_type":"peer_reviewed","statistic":"Review confirms dose-dependent spermatogenic impairment from chronic scrotal hyperthermia; raised testicular temperature compromises sperm quality via apoptosis, DNA damage, and autophagy","excerpt":"\"Raised testicular temperature has a detrimental effect on mammalian spermatogenesis and the resultant spermatozoa. Failure of the thermoregulatory mechanism has the potential to compromise sperm quality and increase the risk of infertility.\"\n","source_date":"2015-01-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260128075458/https://pubmed.ncbi.nlm.nih.gov/25456164/","calculation_notes":"Comprehensive narrative review covering occupational heat (welders, bakers, drivers), lifestyle heat (laptops, saunas, tight clothing), and pathological heat (varicocele, cryptorchidism). Establishes the biological mechanism — germ cell apoptosis triggered by temperatures exceeding the scrotal optimum — but does not quantify population-level incidence of infertility from lifestyle exposures. The review explicitly notes that spermatogenesis occurs optimally at temperatures slightly below core body temperature, and that even modest sustained elevations trigger measurable changes.\n","independence_note":"Review article synthesizing multiple primary sources; authored at Cleveland Clinic, independent of Sheynkin (Stony Brook) and Shefi (UCSF) studies.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/24927498/","title":"Effect of mobile telephones on sperm quality: a systematic review and meta-analysis","publisher":"Environment International / Adams et al.","source_type":"peer_reviewed","statistic":"Meta-analysis of 10 studies (1,492 samples): mobile phone exposure associated with -8.1% motility (95% CI -13.1, -3.2) and -9.1% viability (95% CI -18.4, 0.2)","excerpt":"\"Pooled results from in vitro and in vivo studies suggest that mobile phone exposure negatively affects sperm quality. Further study is required to determine the full clinical implications for both sub-fertile men and the general population.\"\n","source_date":"2014-09-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260314164757/https://pubmed.ncbi.nlm.nih.gov/24927498/","calculation_notes":"The headline numbers (-8.1% motility, -9.1% viability) come from pooling heterogeneous studies — some in vitro (semen samples irradiated directly), some observational (self-reported phone use correlated with semen parameters). The viability confidence interval crosses zero. Effects on concentration were \"more equivocal.\" Critically, no included study measured fertility outcomes (pregnancy rates, time to conception). The authors themselves note that clinical implications remain undetermined. Used here to represent the strongest available evidence for the phone-pocket concern, while noting the evidence quality is low and the clinical relevance unestablished.\n","independence_note":"Systematic review of 10 independent studies; does not overlap with Sheynkin or Shefi cohorts. The meta-analytic methodology is distinct from Durairajanayagam's narrative review.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC6530653/","title":"Type of underwear worn and markers of testicular function among men attending a fertility center","publisher":"Human Reproduction / Mínguez-Alarcón et al.","source_type":"primary_study","statistic":"Men primarily wearing boxers had 25% higher sperm concentration (95% CI 7-31%), 17% higher total count (95% CI 0-28%), and 14% lower serum FSH than non-boxer wearers (n=656)","excerpt":"\"Men who reported most frequently wearing boxers had a 25% higher sperm concentration (95% CI = 7, 31%), 17% higher total count (95% CI = 0, 28%) and 14% lower serum FSH levels (95% CI = -27, -1%) than men who reported not primarily wearing boxers.\"\n","source_date":"2018-08-01","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260412144839/https://pmc.ncbi.nlm.nih.gov/articles/PMC6530653/","calculation_notes":"Cross-sectional study of 656 men attending a fertility center (selection bias: these are men already seeking fertility treatment). The 25% concentration difference is statistically significant but the clinical significance is uncertain — the boxer-wearers' parameters were still within normal ranges. The compensatory FSH elevation in tight-underwear wearers suggests a real thermoregulatory mechanism, but no pregnancy outcome data was collected. The lower bound of the total count CI touches zero.\n","independence_note":"Harvard/MGH fertility center cohort; independent of Sheynkin, Shefi, and Adams studies.\n"},{"url":"https://www.cancer.org/cancer/types/testicular-cancer/causes-risks-prevention/risk-factors.html","title":"Risk Factors for Testicular Cancer","publisher":"American Cancer Society","source_type":"reputable_reference","statistic":"ACS lists cryptorchidism, family history, HIV, prior testicular cancer, age, and race as risk factors; heat exposure, laptops, and electromagnetic fields are not mentioned","excerpt":"\"Prior injury or trauma to the testicles and recurrent actions, such as horseback riding, do not appear to be related to the development of testicular cancer.\"\n","source_date":"2025-08-10","source_accessed":"2026-04-17","archive_url":"http://web.archive.org/web/20260326231134/https://www.cancer.org/cancer/types/testicular-cancer/causes-risks-prevention/risk-factors.html","calculation_notes":"The ACS risk factors page is the most widely consulted lay-facing summary of testicular cancer etiology. Heat exposure, laptop use, phone pocket storage, tight clothing, and electromagnetic radiation are entirely absent from the listed risk factors. The known risk factors — cryptorchidism, family history, contralateral cancer history — are developmental and genetic, not environmental or thermal. This source is included to explicitly document the absence of evidence linking any heat or EM exposure to testicular cancer, which is the component of the folk fear with the least support.\n","independence_note":"ACS editorial content is produced independently of the primary-study and review literature cited above. It reflects a consensus clinical position rather than any single study's findings, and its omission of heat/EM/laptop risk factors is an independent editorial judgment.\n"}],"comparison_anchors":[{"label":"Death from food poisoning (lifetime, US adult)","lifetime_us_adult":0.000537},{"label":"Fatal lightning strike (lifetime, US)","lifetime_us_adult":0.00000354},{"label":"Miscarriage in recognized pregnancy","lifetime_us_adult":0.15}],"regional_breakdown":[{"region":"Laptop on lap, daily 1hr+ use","probability":0.4,"notes":"Probability of measurable semen parameter reduction (not infertility). Sheynkin's 2.6-2.8°C rise exceeds the ~1-2°C impairment threshold. Chronic daily exposure would likely produce detectable changes in a semen analysis. Effect is reversible within one spermatogenic cycle (~3 months) after cessation.\n"},{"region":"Regular sauna/hot tub use (3+/week)","probability":0.45,"notes":"Probability of measurable semen parameter impact for regular users. Shefi's cohort showed 45% favorable response to cessation, implying at least that proportion had heat-attributable changes. Wet heat is a more potent thermal insult than laptops. Reversible in 3-6 months.\n"},{"region":"Tight underwear (briefs/bikinis daily)","probability":0.1,"notes":"Mínguez-Alarcón found 25% higher sperm concentration in boxer-wearers, but most brief-wearers had parameters within normal ranges. The effect is modest and represents a shift in distribution, not a binary harm. Compensatory FSH elevation suggests a real but mild thermoregulatory stress.\n"},{"region":"Phone in trouser pocket, daily carry","probability":0.07,"notes":"Adams meta-analysis found -8.1% motility difference, but included studies are heterogeneous, many in vitro, and the viability CI crosses zero. No study has measured fertility outcomes from pocket storage. The probability represents a rough estimate of detectable semen parameter change, not infertility.\n"},{"region":"Heated car seats, regular use","probability":0.05,"notes":"No controlled human study has measured semen parameters before/after heated seat use. Plausible by analogy with laptop-induced heating, but the thermal dose is uncertain. Structural placeholder based on limited anecdotal evidence.\n"},{"region":"Testicular cancer from any of the above","probability":0.000001,"notes":"Structural floor. ACS does not list heat, EM, laptops, or clothing as testicular cancer risk factors. Known risk factors are cryptorchidism, family history, and contralateral history — developmental and genetic, not thermal or electromagnetic. No epidemiological cohort has found an association. This placeholder exists to make the absence of evidence explicit rather than leaving it as a gap.\n"}],"short_label":"Testicular heat","myth_framing":"overrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry spans two distinct evidence regimes that the folk fear routinely conflates. (a) Semen parameter changes are not the same as infertility. Many men with reduced sperm concentration or motility conceive normally; clinical subfertility requires sustained impairment below WHO thresholds, and no lifestyle-heat study has demonstrated reduced pregnancy rates. (b) For lifestyle exposures — laptops, saunas, tight clothing — the semen effects are reversible within one to two spermatogenic cycles (roughly 3-6 months) after cessation of exposure. (c) Testicular cancer has no documented link to heat exposure, laptop use, phone pocket storage, or electromagnetic radiation in any published epidemiological cohort. The American Cancer Society does not list any of these as risk factors. (d) The folk fear often conflates three separate questions — sperm quality, fertility, and cancer — that have completely different evidence bases, ranging from well-supported (parameter changes) to nonexistent (cancer). Occupational chronic heat exposure (welders, bakers, professional drivers) carries a stronger and less-reversible subfertility signal than any lifestyle exposure covered here.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-10-agent","last_reviewed":"2026-04-17","reviewed":true,"generated_at":"2026-04-17","image":{"alt":"A single thermometer rendered as a flat vector shape in muted warm tones against a pale background."},"canonical_url":"https://likelier.app/testicular-heat-exposure","api_url":"https://likelier.app/api/fears/testicular-heat-exposure.json"},{"slug":"toilet-lid-flush-contamination","question":"What are the odds of getting sick from not closing the toilet lid before flushing?","quick_answer":"There is no reliable estimate. Flushing does aerosolize bacteria and viruses, but no clinical or epidemiological study has ever documented a case of human illness from toilet plume; the exposure is real, the disease transmission undemonstrated.\n","category":"health","tags":["household"],"no_reliable_estimate":true,"perceived":{"description":"The toilet plume concept entered mainstream awareness through viral social-media clips of UV-illuminated flush sprays and a steady stream of headlines warning that flushing without closing the lid launches pathogenic aerosols across the bathroom. The implied conclusion is that an open-lid flush is a meaningful infection pathway, on par with a sneeze from a sick colleague. This framing resonates because the disgust response is doing the persuasion: if you can see the particles, they must be dangerous.\n","rough_estimate":"Most people assume open-lid flushing is a real infection risk worth worrying about","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/23040490/","title":"Lifting the lid on toilet plume aerosol: A literature review with suggestions for future research","publisher":"American Journal of Infection Control / Johnson DL, Mead KR, Lynch RA, Hirst DVL","source_type":"primary_study","statistic":"Toilet plume aerosols are generated during flushing but no study has demonstrated or refuted actual disease transmission via this route","excerpt":"\"No studies have yet clearly demonstrated or refuted toilet plume-related disease transmission, and the significance of the risk remains largely uncharacterized.\"\n","source_date":"2012-11-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260503083612/https://pubmed.ncbi.nlm.nih.gov/23040490/","calculation_notes":"Johnson et al. 2013 is the most-cited review of the toilet plume literature. It confirms that flushing generates bioaerosols containing bacteria, viruses, and other microorganisms detectable at various distances from the bowl. However, the review found zero clinical or epidemiological studies documenting a case of human illness attributable to toilet plume exposure. The aerosol is real; the disease transmission is undemonstrated. This gap — between detectable exposure and measurable harm — is the basis for the no_reliable_estimate classification. Detection of aerosolized pathogens is a necessary but not sufficient condition for clinical infection.\n"},{"url":"https://www.nature.com/articles/s41598-022-24686-5","title":"Commercial toilets emit energetic and rapidly spreading aerosol plumes","publisher":"Scientific Reports / Crimaldi JP et al.","source_type":"primary_study","statistic":"High-velocity aerosol plumes from lidless commercial toilets reach 1.5 m above the bowl within 8 seconds of flushing","excerpt":"\"Flushing generates a strong chaotic jet of air out of the toilet bowl which carries aerosols rapidly upward.\"\n","source_date":"2022-12-08","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250807133933/https://www.nature.com/articles/s41598-022-24686-5","calculation_notes":"Crimaldi et al. used laser-illuminated particle tracking in a commercial (lidless) toilet and showed aerosol velocities reaching 2 m/s, with particles detected at head height within seconds. The study is strictly a fluid-dynamics characterisation — it measured particle transport, not pathogen viability or infection. It is included because the dramatic laser footage from this study is the most-shared visual driver of public concern. The study itself draws no health conclusions.\n","independence_note":"Independent fluid-dynamics laboratory study at the University of Colorado; no overlap with the Johnson et al. clinical review or the Nature 2024 public-toilet bioaerosol study.\n"},{"url":"https://www.nature.com/articles/s41598-024-61039-w","title":"Exploring toilet plume bioaerosol exposure dynamics in public toilets using a Design of Experiments approach","publisher":"Scientific Reports","source_type":"primary_study","statistic":"Bioaerosol concentrations after flushing in public toilets varied with ventilation; the study states its numerical values do not represent standardized thresholds for elevated bioaerosol concentrations and measures no clinical outcomes","excerpt":"\"This study systematically quantified bioaerosol concentrations generated during toilet flushing in public restrooms under varying conditions.\"\n","source_date":"2024-05-06","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250117030113/https://www.nature.com/articles/s41598-024-61039-w","calculation_notes":"A controlled field study measuring bioaerosol generation across multiple public-toilet configurations. Found measurable but variable aerosol plumes, with ventilation being the strongest modifier. No clinical outcomes measured. The study reinforces the pattern across the entire toilet-plume literature: aerosols are real, pathogen detection is intermittent, and clinical disease attributable to this route remains zero in the published record.\n","independence_note":"Independent field study in public restrooms; methodologically distinct from the Crimaldi lab work and the Johnson et al. review.\n"},{"url":"https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0171556","title":"Pathogen cross-transmission via building sanitary plumbing systems in a full scale pilot test-rig","publisher":"PLoS ONE / Gormley M, Aspray TJ, Kelly D, Rodriguez Gill C","source_type":"primary_study","statistic":"WHO's 2003 SARS epidemiology consensus attributed a Hong Kong housing-block super-spreading event (Amoy Gardens) to empty U-traps in the sanitary plumbing system; a 2017 full-scale pilot-rig experiment confirmed empty U-traps let aerosolized pathogens cross-transmit between floors via building airflow","excerpt":"\"The WHO Consensus Document on the epidemiology of the SARS epidemic in 2003, included a report on a concentrated outbreak in one Hong Kong housing block which was considered a 'super-spreading event.'\" ... \"Empty U-traps allowed aerosolised pathogens to enter the chamber, encouraging cross-transmission.\"\n","source_date":"2017-02-10","source_accessed":"2026-07-04","archive_url":"http://web.archive.org/web/20260515231654/https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0171556","calculation_notes":"Cited in the body only as a contrast case, not as evidence for this entry's own claim. The 2003 Amoy Gardens SARS cluster was traced to dried/empty U-trap floor drains venting sewage-system aerosols into apartments under negative pressure from bathroom extract fans -- a structural plumbing failure, not an open lid during a normal flush. This 2017 pilot-rig study independently confirms the underlying physical mechanism (empty U-traps permit aerosol cross-transmission between floors) but does not address toilet-lid position and reports no clinical illness from ordinary flushing, so it does not weaken the entry's own no_reliable_estimate finding.\n","independence_note":"Independent engineering/microbiology test-rig study (Heriot-Watt University) addressing building drainage systems, not toilet-lid flushing; cited here only to substantiate the Amoy Gardens contrast case referenced in the body text.\n"}],"comparison_anchors":[{"label":"Norovirus infection from contaminated food (US adult, annual)","lifetime_us_adult":0.3},{"label":"Flu death (US adult, lifetime)","lifetime_us_adult":0.0063}],"short_label":"Toilet plume illness","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"This entry addresses the specific question of whether leaving the toilet lid open during flushing causes clinical illness in household or public-restroom settings. It does not cover the separate and better-evidenced question of fomite transmission from toilet surfaces (handle, seat, flush button) via hand contact, which is a real pathway for gastrointestinal pathogens and is addressed by hand-hygiene interventions. Nor does it cover healthcare-associated infection from toilets in hospital wards, where C. difficile contamination of shared facilities is a documented nosocomial concern — but the transmission pathway there is primarily surface contact, not aerosol inhalation. The toilet plume is real as a physical phenomenon; what is missing is any evidence that it has made anyone sick.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A simple toilet with a raised lid, a few faint dots rising from the bowl, flat vector illustration."},"canonical_url":"https://likelier.app/toilet-lid-flush-contamination","api_url":"https://likelier.app/api/fears/toilet-lid-flush-contamination.json"},{"slug":"ufo-alien-invasion","question":"What are the odds of an alien invasion or first contact in your lifetime?","quick_answer":"There is no reliable number: the most rigorous analysis gives only an existence probability for other civilisations, not a per-lifetime chance of detectable contact, and no source converts one into the other. A majority of Americans believe alien life exists, but that cultural intuition has no basis in any measured contact rate.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Alien invasion anxiety occupies a unique slot in the risk landscape: it is simultaneously taken very seriously by a meaningful fraction of the population and treated as a punchline by another. A 2021 Pew Research survey found that 65% of Americans believe intelligent life exists on other planets, and 51% say UFO sightings reported by military personnel are likely evidence of extraterrestrial life. Hollywood has spent decades training the public to think of first contact in dramatic terms — invasion, abduction, existential threat — rather than the more prosaic possibility of detecting a radio signal from a civilisation that may have died millions of years ago. The perceived probability is driven by cultural narrative, not by any evidence of contact.\n","rough_estimate":"A majority of Americans believe intelligent alien life exists; a substantial minority considers contact plausible within their lifetime","kind":"intuition"},"sources":[{"url":"https://arxiv.org/abs/1806.02404","title":"Dissolving the Fermi Paradox","publisher":"arXiv / Sandberg A, Drexler E, Ord T","source_type":"primary_study","statistic":"When Drake-equation parameters are treated as probability distributions, there is a ~38% probability of zero other civilisations in the observable universe and >50% in the Milky Way alone","excerpt":"\"We find a substantial ex ante probability of there being no other intelligent life in our observable universe, and thus that there should be little surprise when we fail to detect any signs of it.\"\n","source_date":"2018-06-06","source_accessed":"2026-04-18","archive_url":"https://web.archive.org/web/20260429115815/https://arxiv.org/abs/1806.02404","calculation_notes":"Sandberg, Drexler, and Ord (2018) is the most rigorous probabilistic treatment of the Drake equation in the literature. Rather than plugging in point estimates for each parameter (which produces the classic \"there should be millions of civilisations\" result), they used probability distributions reflecting actual scientific uncertainty for each factor — especially the probability of abiogenesis and the transition from simple to complex life. The resulting distribution for N (number of civilisations) spans many orders of magnitude and places substantial probability mass at N=0 or N=1. This is the basis for the headline: the best available probabilistic analysis suggests we may well be alone, which makes the probability of contact in any given lifetime vanishingly small.\n"},{"url":"https://science.nasa.gov/universe/exoplanets/are-we-alone-in-the-universe-revisiting-the-drake-equation/","title":"Are We Alone in the Universe? Revisiting the Drake Equation","publisher":"NASA Science","source_type":"govt_report","statistic":"NASA overview of the Drake equation notes that modern exoplanet surveys have constrained some parameters but the origin-of-life and civilisation-longevity terms remain uncertain by many orders of magnitude","excerpt":"\"While astronomers have made progress quantifying some of the Drake equation's factors, the most important terms — such as how often life originates and how long civilisations last — remain deeply uncertain.\"\n","source_date":"2024-06-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260409073444/https://science.nasa.gov/universe/exoplanets/are-we-alone-in-the-universe-revisiting-the-drake-equation/","calculation_notes":"NASA's overview provides the institutional context for why the Drake equation does not yield a useful point estimate. Kepler and related exoplanet surveys have constrained one factor — roughly one-fifth of stars have planets in the habitable zone — but the abiogenesis and intelligence-evolution terms remain unconstrained by data. This confirms the Sandberg et al. framing: the uncertainty is not merely large, it is epistemically fundamental. Used here as an authoritative secondary source confirming that no scientific consensus exists on the probability of extraterrestrial contact.\n","independence_note":"NASA editorial overview; independent of the Sandberg et al. probabilistic analysis and the SETI Institute's operational search programme.\n"},{"url":"https://www.seti.org/research/seti-101/drake-equation/","title":"Drake Equation","publisher":"SETI Institute","source_type":"reputable_reference","statistic":"Frank Drake's original 1961 estimate yielded N=10 civilisations in the Milky Way; modern estimates range from <1 to millions depending on parameter choices","excerpt":"\"It is impossible to do anything more than guess at the probable longevity of other advanced civilisations.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260421210531/https://www.seti.org/research/seti-101/drake-equation/","calculation_notes":"The SETI Institute — the organisation most invested in finding extraterrestrial intelligence — acknowledges that the Drake equation's output is dominated by the civilisation-longevity term (L), which is completely unconstrained by data. Drake's original 1961 workshop estimated N=10 with L=10,000 years, but L could be 100 years (civilisations destroy themselves quickly) or 10 million years (they do not). The SETI Institute's own assessment is that the equation is a framework for organising ignorance, not a tool for generating a reliable number.\n","independence_note":"The SETI Institute is the primary non-governmental organisation conducting extraterrestrial intelligence searches; independent of the Oxford/FHI group (Sandberg et al.) and NASA.\n"}],"comparison_anchors":[],"short_label":"Alien contact","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"existential","valence":"negative","caveats":"This entry is flagged no_reliable_estimate. Sandberg, Drexler, and Ord (2018) is the most rigorous probabilistic treatment of the Drake equation in the literature, and it does support a genuine figure: when Drake-equation parameters are treated as probability distributions reflecting real scientific uncertainty rather than point estimates, the paper finds a substantial ex ante probability — roughly 38% for the observable universe and greater than 50% for the Milky Way alone — that humanity is the only intelligent civilisation present. But that is an EXISTENCE probability (whether any other civilisation exists at all), not a per-year or per-lifetime probability of detectable CONTACT with one. Neither this paper nor the NASA or SETI Institute sources cited here convert that existence finding into a contact rate; doing so would require additional assumptions — about overlap in time, spatial proximity within detection range, and mutual technological compatibility — that no source in the literature has rigorously quantified. A prior version of this entry manufactured a \"1 in 170 million\" lifetime figure from an arbitrary, uncited 1-in-10-billion-per-year rate; that number had no basis in any cited source and has been removed. The entry treats \"first contact\" broadly — including detection of a radio signal, not only physical encounter or invasion. It is retained as a calibration anchor without a headline number: the existence-probability finding is itself useful context, since it shows that the premise \"contact should be expected\" was never well supported by the evidence, even though no source lets us convert that into an odds figure comparable to the site's other entries.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A small disc-shaped silhouette against a dark sky with scattered stars, flat vector illustration."},"canonical_url":"https://likelier.app/ufo-alien-invasion","api_url":"https://likelier.app/api/fears/ufo-alien-invasion.json"},{"slug":"underdressed-at-home","question":"What are the odds of getting sick from walking around home without a hat, scarf, gloves, socks, or slippers?","quick_answer":"There is no reliable number: the folk warning that being under-dressed indoors will 'give you a cold' has no measurable per-event probability. The evidence points at most to a small symptom-modulation effect in people already carrying a virus, not illness caused from nothing.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"A folk belief in much of the world holds that being under-dressed at home — bare feet on a tile floor, no slippers (\"papucie\" in Polish, \"тапочки\" in Russian, the slipper-culture common to most of Central and Eastern Europe and much of East Asia), no sweater, a draft on the neck — causes colds, flu, pneumonia, or vague \"getting sick.\" The mechanism the folk model implies is that cold exposure on the skin directly produces respiratory illness. There is no good survey of how US or global adults rate this risk numerically; what exists is a large body of ethnographic and clinical anecdote that the belief is near-universal and durable across generations, especially from parents and grandparents to children.\n","rough_estimate":"Commonly framed as 'you'll catch a cold,' with no numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/16286463/","title":"Acute cooling of the feet and the onset of common cold symptoms","publisher":"Family Practice (Oxford Academic), via PubMed","source_type":"peer_reviewed","statistic":"13/90 chilled subjects vs 5/90 controls self-reported cold symptoms in the 4–5 days after a 20-minute cold-foot immersion (P=0.047)","excerpt":"\"There is a common folklore that chilling of the body surface causes the development of common cold symptoms, but previous clinical research has failed to demonstrate any effect of cold exposure on susceptibility to infection with common cold viruses. [...] 13/90 subjects who were chilled reported they were suffering from a cold in the 4/5 days after the procedure compared to 5/90 control subjects (P=0.047). [...] Acute chilling of the feet causes the onset of common cold symptoms in around 10% of subjects who are chilled. Further studies are needed to determine the relationship of symptom generation to any respiratory infection.\"\n","source_date":"2005-12-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260504061638/https://pubmed.ncbi.nlm.nih.gov/16286463/","calculation_notes":"This is the canonical trial behind any \"cold feet causes colds\" claim. Critical qualifier the authors themselves flag: the study measured self-reported symptoms, not laboratory-confirmed new infections. The plausible mechanism the authors propose is that reflex vasoconstriction in the upper airway on cold- foot exposure reduces mucosal blood flow and temporarily lowers local defences — converting a pre-existing subclinical carriage of rhinovirus or another respiratory virus into a symptomatic cold. That is a modulation effect, not a causation effect. Without an underlying viral exposure, cooling the feet is not expected to produce illness from nothing. 90 subjects per arm gives an absolute difference of 8 percentage points (14% vs 6%); the confidence bound is wide, and no replication of comparable rigour exists at the scale needed to attach a per-winter probability to \"no slippers at home.\"\n","independence_note":"Independent single-centre RCT at Cardiff (Common Cold Centre); editorially independent of the CDC and WHO sources. The Foxman 2015 mechanistic paper below provides a biological model compatible with Eccles' clinical result but was conducted in a separate lab with different methodology (mouse airway cells, not human subjects).\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25561542/","title":"Temperature-dependent innate defense against the common cold virus limits viral replication at warm temperature in mouse airway cells","publisher":"Proceedings of the National Academy of Sciences (PNAS), via PubMed","source_type":"primary_study","statistic":"Rhinovirus replicates more robustly at 33–35 °C (nasal cavity) than at 37 °C (core body), with weaker interferon/antiviral response at the cooler temperature","excerpt":"\"Most isolates of human rhinovirus, the common cold virus, replicate more robustly at the cool temperatures found in the nasal cavity (33–35 °C) than at core body temperature (37 °C). [...] These findings demonstrate that in mouse airway cells, rhinovirus replicates preferentially at nasal cavity temperature due, in part, to a less efficient antiviral defense response of infected cells at cool temperature.\"\n","source_date":"2015-01-20","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20251208133031/https://pubmed.ncbi.nlm.nih.gov/25561542/","calculation_notes":"Foxman et al. supplies the cleanest known mechanism for any cold-exposure- to-cold-illness signal: rhinovirus itself replicates better in a cooler nose, and the innate interferon response is weaker at 33 °C than at 37 °C. This makes Eccles' symptom-onset result biologically plausible without rescuing the folk model. The study is mouse airway cells in vitro, not an epidemiological measurement, and no study has translated the temperature- dependent replication curve into a per-exposure infection probability for a human wearing socks versus going barefoot. The mechanism is real; the epidemiological effect size at normal indoor conditions is not quantified.\n","independence_note":"Yale laboratory study with no authorship, funding, or institutional overlap with the Cardiff Eccles group; treat as methodologically independent mechanistic corroboration. Independent of the CDC and WHO sources.\n"},{"url":"https://www.cdc.gov/common-cold/about/index.html","title":"About the Common Cold","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"More than 200 respiratory viruses cause colds; rhinoviruses are the most frequent cause; primary spread is droplets and contact","excerpt":"\"More than 200 respiratory viruses can cause colds. Rhinoviruses are the most frequent cause of colds in the United States. [...] Most respiratory viruses are spread through droplets that an infected person releases when they cough or sneeze. These droplets can enter your body if you breathe them in or touch a contaminated surface and then touch your eyes, nose, or mouth.\"\n","source_date":"2026-02-19","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260402134530/https://www.cdc.gov/common-cold/about/index.html","calculation_notes":"CDC's current patient-facing page is the plain-language anchor for the \"colds are viral, not thermal\" frame. The folk model treats cold exposure as causative; CDC treats virus exposure as causative and does not list chilling or being under-dressed indoors as a transmission route at all. The Eccles and Foxman results sit downstream of this: you still need the virus. Without rhinovirus or one of the other ~200 candidates in your airway, cold feet on tile do not produce a cold.\n","independence_note":"Institutional CDC public-health guidance; editorially independent of the Eccles clinical trial and Foxman mechanistic paper, though it aligns with both.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK535294/","title":"Low indoor temperatures and insulation — WHO Housing and Health Guidelines","publisher":"World Health Organization (via NCBI Bookshelf)","source_type":"govt_report","statistic":"WHO recommends minimum indoor temperature of 18 °C to protect general populations; higher minimum for vulnerable groups (older people, children, chronic cardiorespiratory illness)","excerpt":"\"For countries with temperate or colder climates, 18 °C has been proposed as a safe and well-balanced indoor temperature to protect the health of general populations during cold seasons. [...] A higher minimum indoor temperature than 18 °C may be necessary for vulnerable groups including older people, children and those with chronic illnesses, particularly cardiorespiratory disease.\"\n","source_date":"2018-11-27","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260126080730/https://www.ncbi.nlm.nih.gov/books/NBK535294/","calculation_notes":"WHO's guideline is the authoritative carve-out for the one scenario in which \"being under-dressed at home\" really does kill people: under-heated housing in cold climates, especially for the elderly and those with cardiorespiratory disease. The exposure here is the ambient indoor temperature (below ~18 °C sustained), not a barefoot afternoon in a heated 21 °C living room. The outcome is cardiovascular and respiratory morbidity and mortality, not the common cold. This is the reason the headline framing (\"folk belief overrated\") must be paired with an explicit vulnerable-group caveat rather than a blanket dismissal.\n","independence_note":"WHO expert consensus guideline synthesising the cold-housing evidence base. Editorially independent of the Eccles, Foxman, and CDC sources and addresses a distinct exposure-outcome pair (sustained low ambient temperature → cardiovascular/respiratory death), not symptom onset of the common cold.\n"},{"url":"https://post.parliament.uk/research-briefings/post-pn-0752/","title":"Winter mortality (POSTnote 752)","publisher":"UK Parliamentary Office of Science and Technology","source_type":"govt_report","statistic":"Excess winter deaths in England and Wales ranged ~20,000–50,000/year 2000–2019; ~21.5% of excess winter deaths attributable to cold homes; most deaths from circulatory or respiratory disease among the elderly","excerpt":"\"Between 2000 and 2019, excess winter deaths ranged from 20,000 to 50,000 a year [...] Most excess winter deaths are due to circulatory or respiratory diseases and the majority occur amongst the elderly population. [...] It has been estimated that 10% of excess winter deaths are attributable to fuel poverty and 21.5% to cold homes.\"\n","source_date":"2024-01-01","source_accessed":"2026-04-16","archive_url":"http://web.archive.org/web/20260216223615/https://post.parliament.uk/research-briefings/post-pn-0752/","calculation_notes":"This is the population-scale number for the one real cold-in-the-home harm: under-heated housing kills elderly people through cardiovascular and respiratory pathways, not through infection. It does not apply to the folk-belief scenario (healthy adult, barefoot in a warm house) and should not be aggregated with the Eccles symptom-onset figure. Used here only to bound the vulnerable-group subgroup in the regional breakdown and to keep the caveats honest about who the folk warning, repurposed, actually applies to.\n","independence_note":"UK Parliament research briefing drawing on ONS winter-mortality data and NICE fuel-poverty reviews. Editorially independent of the WHO guideline (though it references the same underlying epidemiology) and independent of the Eccles, Foxman, and CDC sources.\n"}],"comparison_anchors":[{"label":"Shared-cup sip, any serious infection (healthy adults)","lifetime_us_adult":0.000001},{"label":"Shared-towel serious infection (casual dry towel)","lifetime_us_adult":0.000001},{"label":"Food poisoning death (lifetime, US adult)","lifetime_us_adult":0.000537}],"regional_breakdown":[{"region":"Healthy adult, bare feet on tile in a normally heated home (~20–22 °C)","probability":0.000001,"notes":"No cohort has measured a respiratory-infection rate attributable to this scenario. Without an underlying viral exposure, cold feet on tile do not produce a cold. Point estimate is a structural \"effectively zero\" placeholder, not a measured rate.\n"},{"region":"Adult already carrying rhinovirus, acutely chilled extremities","probability":0.14,"notes":"Matches the Eccles 2005 arm: ~14% self-reported cold symptoms within 4–5 days after a 20-minute cold-foot immersion vs ~6% in controls. This is a symptom-conversion rate in an already-exposed population under a severe chilling protocol, not an infection rate from going sockless at home.\n"},{"region":"Frail elderly or infant in an under-heated home (<16–18 °C sustained)","probability":0.05,"notes":"Very rough. Order of magnitude derived from WHO cold-housing guidance and UK excess-winter-mortality attribution (~21.5% of ~20,000–50,000 annual excess winter deaths → cold homes, concentrated in over-65s). The dominant pathways are cardiovascular and respiratory, not viral infection. Included as the subgroup the folk warning might actually apply to, even though it is almost never the one a grandmother has in mind when telling a child to put on slippers.\n"},{"region":"Adult with Raynaud's phenomenon or cold-induced angina","probability":0.5,"notes":"For someone with a cold-triggered vascular or cardiac condition, bare feet on cold tile reliably produces the trigger (Raynaud's episode, anginal chest pain) — but this is the underlying condition expressing itself, not a new illness. Included only to flag that \"cold feet cause real symptoms\" is true in this subgroup without rescuing the viral- infection folk model.\n"}],"short_label":"Bare feet indoors","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"Two pieces of evidence complicate a clean dismissal of the folk belief and deserve to sit in the open. First, Johnson and Eccles (2005) found a modest but statistically significant increase in self-reported cold symptoms after a 20-minute cold-foot immersion (14% vs 6%, P=0.047) — not a new-infection signal, but a symptom-conversion signal in people likely already carrying a virus. Second, Foxman et al. (2015) showed that rhinovirus replicates preferentially at the cooler temperatures of the nasal cavity and that the innate interferon response is weaker there, giving Eccles' clinical result a plausible mechanism. Together these support a narrow reframing: chilling can modulate symptom onset in an already-exposed airway, not cause illness from nothing. The folk model (cold exposure → colds, from scratch) is still wrong; the fully skeptical counter-model (cold exposure does nothing at all) is also slightly wrong. The right answer is \"small modulation effect conditional on viral carriage, not a primary cause.\"\nSeparately, the one real and well-documented harm from being under-dressed indoors is not captured by the per-event framing at all: elderly people, infants, and those with chronic cardiorespiratory disease in homes sustained below ~16–18 °C have measurable excess cardiovascular and respiratory mortality (WHO 2018; UK POST 2024). The exposure there is ambient room temperature, not sockless feet, and the outcome is heart attack, stroke, and pneumonia, not the common cold. The folk warning and the real harm do not overlap neatly: the folk warning is usually delivered to healthy people in warm homes, and the real harm falls on frail people in cold ones.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.625,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"likelier-phase-8-agent","last_reviewed":"2026-04-16","reviewed":true,"generated_at":"2026-04-16","image":{"alt":"A single pair of neatly arranged slippers rendered as a flat vector shape in muted grey-blue and off-white tones, centered on a calm empty background."},"canonical_url":"https://likelier.app/underdressed-at-home","api_url":"https://likelier.app/api/fears/underdressed-at-home.json"},{"slug":"untreated-back-pain-disability","question":"What are the odds of chronic disability from not treating back pain?","quick_answer":"There is no reliable number: no cited source reports a precise lifetime probability of chronic disabling low back pain. Most people fear that not treating acute back pain will cause disability, but the evidence points to psychosocial factors — not whether hands-on treatment was received — as the main drivers of chronicity.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"Low back pain generates more healthcare visits, imaging, and intervention than almost any other musculoskeletal complaint. The implicit fear — that acute back pain left untreated will spiral into chronic disability — drives patients toward early imaging, opioid prescriptions, injections, surgery, and a sprawling market of chiropractic, osteopathic, and physiotherapy services. The 2018 Lancet Low Back Pain Series identified widespread overtreatment and low-value care as a global problem, estimating that the majority of acute low back pain episodes are self-limiting. The gap between the fear of doing nothing and the evidence on doing nothing is one of the widest in musculoskeletal medicine.\n","rough_estimate":"Most people with acute back pain believe it will worsen without treatment","kind":"intuition"},"sources":[{"url":"https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30480-X/abstract","title":"What low back pain is and why we need to pay attention","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Low back pain is the leading cause of disability worldwide; most episodes resolve within weeks; recurrence is common but chronic disability is not","excerpt":"\"Most people with new episodes of low back pain recover quickly; however, recurrence is common and in a small proportion of people, low back pain becomes persistent and disabling.\"\n","source_date":"2018-03-21","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250801165920/https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(18)30480-X/abstract","calculation_notes":"Hartvigsen et al. 2018 (Lancet Low Back Pain Series, Paper 1) established the epidemiological framing: LBP is the #1 cause of disability globally (by years lived with disability), but this ranking reflects its enormous prevalence rather than per-episode severity. Most individual episodes are self-limiting. The paper explicitly criticizes the overtreatment paradigm in high-income countries, where patients receive imaging, opioids, injections, and surgery at rates far exceeding evidence-based indications.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/29573872/","title":"Prevention and treatment of low back pain: evidence, challenges, and promising directions","publisher":"The Lancet","source_type":"peer_reviewed","statistic":"Best evidence supports staying active; most physical and pharmacological interventions show small or no benefit over natural recovery for acute non-specific LBP","excerpt":"\"For acute low back pain, the evidence supports advice to remain active, and that paracetamol is not effective. For chronic low back pain, exercise, multidisciplinary rehabilitation, and some psychological approaches have moderate-quality evidence of effectiveness.\"\n","source_date":"2018-03-21","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20260215092441/https://pubmed.ncbi.nlm.nih.gov/29573872/","calculation_notes":"Foster et al. 2018 (Lancet LBP Series, Paper 3) is the treatment- evidence synthesis. The critical finding for this entry: for acute non-specific LBP, the best evidence-based intervention is advice to stay active and avoid bed rest. Most other interventions (physiotherapy, manipulation, medications beyond NSAIDs) have small effect sizes that are often not clinically meaningful. This directly addresses the \"doing nothing\" fear: the evidence suggests that \"doing nothing\" (plus staying active) is close to the optimal strategy for acute LBP.\n"},{"url":"https://www.cmaj.ca/content/196/2/E29","title":"The clinical course of acute, subacute and persistent low back pain: a systematic review and meta-analysis","publisher":"Canadian Medical Association Journal","source_type":"peer_reviewed","statistic":"Acute LBP mean pain score fell from 56/100 at baseline to 26 at 6 weeks and 21 at 52 weeks; persistent LBP stayed high (56 to 40) with minimal improvement over time","excerpt":"\"Participants with acute and subacute low back pain had substantial improvements in levels of pain and disability within the first 6 weeks; however, participants with persistent low back pain had high levels of pain and disability with minimal improvements over time.\"\n","source_date":"2024-01-21","source_accessed":"2026-04-18","archive_url":"http://web.archive.org/web/20250824103938/https://www.cmaj.ca/content/196/2/E29","calculation_notes":"This 2024 CMAJ systematic review and meta-analysis (Wallwork et al., 95 studies, 60 cohorts, n = 17,974) pooled population-level pain and disability trajectories on a 0-100 scale rather than a recovery percentage. For the acute cohort the mean pain score was 56 (95% CI 49-62) at baseline, 26 (21-31) at 6 weeks, and 21 (17-25) at 52 weeks — substantial early improvement that then plateaus at a low but non-zero residual. For the persistent cohort it stayed high (56 to 40) with minimal change. The paper does not report a fraction of people who \"still have pain\" at one year, nor a lifetime incidence of chronic disabling LBP; its trajectories describe average symptom levels, not a transition probability, which is why this entry carries no single comparable lifetime figure.\n"}],"comparison_anchors":[{"label":"Chronic knee osteoarthritis (lifetime, US adult)","lifetime_us_adult":0.14},{"label":"Major depression episode (lifetime, US adult)","lifetime_us_adult":0.2},{"label":"Type 2 diabetes (lifetime, US adult)","lifetime_us_adult":0.4}],"personal_factor_multipliers":[{"factor":"high fear-avoidance beliefs","multiplier":3,"notes":"Psychological factors — catastrophizing, fear of movement, passive coping — are the strongest predictors of chronic disability, more so than imaging findings or initial pain severity."},{"factor":"physically active lifestyle","multiplier":0.4,"notes":"Regular physical activity is consistently associated with lower rates of transition to chronic pain. The Lancet series identifies staying active as the single best evidence-based recommendation for acute LBP."},{"factor":"compensation claim or litigation","multiplier":4,"notes":"Workers' compensation and litigation are among the strongest predictors of prolonged disability, independent of injury severity. This is well-documented across multiple jurisdictions and decades of research."},{"factor":"sedentary occupation","multiplier":1.5,"notes":"Prolonged sitting and lack of physical variation in work tasks are modestly associated with higher rates of chronic LBP, though the evidence is less consistent than for psychosocial factors."}],"short_label":"Untreated back pain disability","myth_framing":"overrated","outcome_severity":"moderate_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers non-specific low back pain, which the Lancet series describes as the large majority of presentations (\"only a small proportion of people have a well understood pathological cause\"). It does not apply to specific diagnoses: cauda equina syndrome (a surgical emergency), spinal fracture, infection, malignancy, or severe radiculopathy with progressive neurological deficit. These \"red flag\" conditions require prompt treatment and should not be conflated with non-specific LBP. The \"overrated\" framing applies to the fear that not treating non-specific acute LBP will cause disability — the evidence suggests that the main predictors of chronicity are psychosocial (yellow flags: catastrophizing, fear-avoidance beliefs, depression, compensation claims, job dissatisfaction), not whether the patient received hands-on treatment. The 2024 CMAJ meta-analysis found that acute low back pain improves substantially in the first six weeks but then plateaus at a low, non-zero residual pain level (mean score 21/100 at one year) rather than reaching zero — \"still has some residual pain\" and \"disabled\" are very different outcomes. Recurrence is the norm — most people who recover will have another episode — but recurrence is not the same as progressive worsening. Neither the CMAJ trajectory data nor the Lancet series reports a precise lifetime probability of chronic disabling low back pain, so this entry gives none.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":4,"d6":5,"d7":4,"d8":5,"avg":4.25,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A person's silhouette walking upright past an empty treatment table, flat vector illustration in muted warm tones."},"canonical_url":"https://likelier.app/untreated-back-pain-disability","api_url":"https://likelier.app/api/fears/untreated-back-pain-disability.json"},{"slug":"untreated-childhood-flat-feet","question":"What are the odds of lasting harm from not treating childhood flat feet?","quick_answer":"There is no reliable number: no cited source quantifies how many children with flexible flat feet go on to lasting impairment, and the AAP review concludes the literature cannot identify who is at risk. Many parents fear untreated flat feet cause lifelong pain or disability, but the evidence does not support that.\n","category":"kids","tags":["child"],"no_reliable_estimate":true,"perceived":{"description":"Flat feet in children provoke an outsized parental anxiety cycle, sustained in part by a large orthotic and insole industry. Pediatric shoe inserts are a multi-billion-dollar global market, and many parents report being told by clinicians — or by shoe-store staff — that flat arches require immediate correction to prevent lifelong disability. Surveys of parental concern find that flat feet are among the top musculoskeletal reasons for pediatric orthopedic referral, despite the condition being overwhelmingly benign. The implicit fear is that an untreated flat foot will lead to chronic pain, gait abnormalities, or arthritis in adulthood.\n","rough_estimate":"Many parents believe untreated flat feet will cause lifelong pain or disability","kind":"intuition"},"sources":[{"url":"https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD006311.pub3/full","title":"Foot orthoses for treating paediatric flat feet","publisher":"Cochrane Database of Systematic Reviews","source_type":"peer_reviewed","statistic":"16 trials, 1058 children (flexible flat feet): low to very low-certainty evidence; the effect of customized or prefabricated foot orthoses versus shoes on pain, function, and quality of life is uncertain","excerpt":"\"Low to very low-certainty evidence shows that the effect of CFOs (high cost) or PFOs (low cost) versus shoes, and CFOs versus PFOs on pain, function and HRQoL is uncertain.\"\n","source_date":"2022-01-14","source_accessed":"2026-04-18","calculation_notes":"The 2022 Cochrane review (Evans et al.) is the most comprehensive synthesis on orthotic intervention for pediatric flat feet. It included 16 RCTs with 1,058 children and concluded that the effect of orthoses versus shoes on pain, function, and quality of life in flexible flat feet is uncertain, on low to very low-certainty evidence. The absence of demonstrated benefit means the untreated natural history is the relevant baseline for asymptomatic flexible flat feet, and that natural history is overwhelmingly benign.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK430802/","title":"Pes Planus","publisher":"StatPearls (NCBI Bookshelf)","source_type":"reputable_reference","statistic":"Infants are usually born with flat feet; children typically develop a normal arch by ages 7 to 10. Approximately 20-37% of the population has some degree of pes planus, which is often asymptomatic.","excerpt":"\"Infants are usually born with flat feet, and children typically do not develop a normal arch until ages 7 to 10. Approximately 20% to 37% of the population is estimated to have some degree of pes planus. Pes planus is often asymptomatic but may alter the biomechanics of the lower extremities and lumbar spine, increasing the risk of pain and injury.\"\n","source_date":"2026-02-15","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20250813084337/https://www.ncbi.nlm.nih.gov/books/NBK430802/","calculation_notes":"StatPearls establishes that flat feet are normal in infancy and that the arch typically develops by ages 7 to 10, so most childhood flat feet self-resolve. It estimates 20-37% of the population retains some degree of pes planus into adulthood and that the condition is \"often asymptomatic.\" Crucially, StatPearls quantifies only structural prevalence — it does not give a symptomatic percentage. No cited source (StatPearls, Pfeiffer 2006, the AAP review, or Cochrane) states a symptomatic rate among persistent adult flat feet, so this entry derives no native or normalized probability and is flagged no_reliable_estimate rather than reporting an unsourced figure. The AAP review states directly that \"the available literature does not elucidate which patients are at risk for developing pain and disability as young adults.\"\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/26908688/","title":"Pediatric Pes Planus: A State-of-the-Art Review","publisher":"Pediatrics (American Academy of Pediatrics)","source_type":"peer_reviewed","statistic":"AAP state-of-the-art review concludes it is safe and appropriate to simply observe an asymptomatic child with flat feet; orthotics are generally unproven to alter the course of flexible flatfoot","excerpt":"\"Current evidence suggests that it is safe and appropriate to simply observe an asymptomatic child with flat feet. ... Orthotics, although generally unproven to alter the course of flexible flatfoot, may provide relief of pain when present.\"\n","source_date":"2016-03-01","source_accessed":"2026-07-01","archive_url":"http://web.archive.org/web/20260419145256/https://pubmed.ncbi.nlm.nih.gov/26908688/","calculation_notes":"Carr, Yang & Lather (2016), the AAP's state-of-the-art review of pediatric pes planus, is the primary source underlying the Choosing Wisely recommendation against orthotics for asymptomatic pediatric flat feet. It concludes that observation is safe and appropriate for the asymptomatic child and that orthotics are generally unproven to alter the natural course of flexible flatfoot — supporting the entry's position that the untreated natural history is the relevant baseline and is overwhelmingly benign. (Replaces a now-dead AAFP Choosing Wisely collection URL that paraphrased this same review.)\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/16882817/","title":"Prevalence of Flat Foot in Preschool-Aged Children","publisher":"Pediatrics (American Academy of Pediatrics)","source_type":"peer_reviewed","statistic":"Among 835 children, flat foot was present in 54% of 3-year-olds versus 24% of 6-year-olds; overall flexible flat foot prevalence was 44% across ages 3-6","excerpt":"\"A total of 835 children (411 girls and 424 boys) were included in this study. Prevalence of flexible flat foot in the group of 3- to 6-year-old children was 44%. ... in the group of 3-year-old children 54% showed a flat foot, whereas in the group of 6-year-old children only 24% had a flat foot.\"\n","source_date":"2006-08-01","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260209175821/https://pubmed.ncbi.nlm.nih.gov/16882817/","calculation_notes":"Pfeiffer et al. (2006) directly document the age-related decline in flat foot prevalence — 54% at age 3 falling to 24% at age 6 in a cohort of 835 children — confirming that the great majority of childhood flat feet resolve as the arch develops. The study also found genuinely pathological flat foot in under 1% of the cohort. It documents the prevalence cascade that underpins the entry's \"overwhelmingly benign\" framing; it does not report any symptomatic-persistence rate, which is one reason this entry carries no headline probability.\n"}],"comparison_anchors":[{"label":"Appendicitis (lifetime, US adult)","lifetime_us_adult":0.07},{"label":"ACL tear requiring surgery (lifetime, US adult)","lifetime_us_adult":0.02},{"label":"Childhood asthma persisting into adulthood","lifetime_us_adult":0.03}],"personal_factor_multipliers":[{"factor":"Symptomatic flat feet (pain present)","multiplier":8,"notes":"StatPearls (Pes Planus) characterises pes planus as often asymptomatic, and only the small symptomatic minority represents the pool with meaningful progression risk. A child with pain from flat feet has a substantially higher probability of persistent adult limitation than the asymptomatic majority. The Cochrane 2022 review (Evans et al.) found the effect of orthoses on pain and function uncertain on low to very low-certainty evidence, meaning natural history — not treatment — largely determines outcome in this subgroup."},{"factor":"Obesity (BMI above age-appropriate 95th percentile)","multiplier":3.5,"notes":"Excess body weight increases mechanical load on the medial longitudinal arch, worsening pronation and pain in weight-bearing flat feet per StatPearls (Pes Planus, 2023) and AAFP clinical guidance. Epidemiological studies of pediatric flat foot consistently identify obesity as a modifying factor for symptom development, independently of arch morphology."},{"factor":"Rigid flat foot (non-flexible / tarsal coalition suspected)","multiplier":10,"notes":"StatPearls (Pes Planus, 2023) distinguishes flexible flat foot (overwhelmingly benign, arch appears on tiptoe) from rigid flat foot (arch absent in all positions, often indicating tarsal coalition or neuromuscular cause). Rigid flat foot represents under 1% of pediatric presentations but carries materially higher rates of functional impairment, secondary arthritis, and surgical intervention. The baseline entry covers flexible flat foot — this multiplier reflects the distinct clinical entity."},{"factor":"High-impact sports participation (competitive running, basketball, gymnastics)","multiplier":2.5,"notes":"AAFP and sports medicine literature note that repetitive high-impact loading in flat-footed athletes accelerates both symptom development and associated conditions (medial tibial stress syndrome, plantar fasciitis, Achilles tendinopathy). Children with symptomatic flat feet who participate in high-impact sports at competitive intensity face a compounded risk from mechanical stress."}],"short_label":"Untreated childhood flat feet","myth_framing":"overrated","outcome_severity":"minor_harm","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry covers flexible (physiologic) flat feet in otherwise healthy children. Rigid flat feet, flat feet secondary to tarsal coalition, and flat feet associated with neuromuscular conditions (e.g., cerebral palsy, Ehlers-Danlos syndrome) are distinct clinical entities with different prognoses and are not covered here. No cited source quantifies the proportion of flexible-flat-footed children who go on to lasting functional impairment: StatPearls calls pes planus \"often asymptomatic\" without a symptomatic percentage, Pfeiffer et al. (2006) report only that pathological (non-physiologic) flat foot affected under 1% of preschoolers, and the AAP state-of-the-art review concludes outright that \"the available literature does not elucidate which patients are at risk for developing pain and disability as young adults.\" Because the evidence base does not support a defensible lifetime probability, this entry is flagged no_reliable_estimate rather than reporting a manufactured figure. Any symptom count is further confounded because it does not cleanly distinguish pain caused by the arch itself from pain from comorbid conditions (obesity, hypermobility) that are correlated with flat feet. Progressive collapsing foot deformity (adult- acquired flat foot from posterior tibial tendon dysfunction) is a separate condition that occurs in middle-aged adults and is not a consequence of untreated childhood flat feet.\n","quality_score":{"d1":3,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.5,"scored_by":"extracted-from-transcript","scored_at":"2026-04-26","methodology_version":"1.0"},"reviewer":"8d-quality-review-agent","last_reviewed":"2026-04-26","reviewed":true,"generated_at":"2026-04-18","image":{"alt":"A small pair of children's shoes sitting unused next to a pair of orthotic insoles, flat vector illustration in muted earth tones."},"canonical_url":"https://likelier.app/untreated-childhood-flat-feet","api_url":"https://likelier.app/api/fears/untreated-childhood-flat-feet.json"},{"slug":"vaping-cancer-risk","question":"What are the odds of getting cancer from vaping?","quick_answer":"There is no reliable lifetime figure: e-cigarette aerosol contains carcinogens, but no significant cancer incidence has been observed in never-smoker vapers, and solid-tumour latency means the human epidemiology a probability would require does not yet exist. By 2022 over 30% of US adults rated vaping as more harmful than smoking, up from ~3% in 2012.\n","category":"cancer","tags":["substance-use"],"no_reliable_estimate":true,"perceived":{"description":"Public fear that vaping causes cancer has grown sharply, and it has increasingly fused with a second, separate belief: that e-cigarettes are as dangerous as — or more dangerous than — cigarettes. A 2026 analysis of the US Health Information National Trends Survey found the share of adults who rate vaping as MORE harmful than smoking rose from roughly 3% in 2012 to over 30% by 2022, while the share calling it less harmful fell from about 51% to 17%. Headlines through 2025–2026 reporting that vaping is \"likely\" to cause cancer reinforced the impression that vaping carries cigarette-scale cancer risk. The fear has a real evidentiary core — e-cigarette aerosol does contain carcinogens and can damage DNA — but the popular equation of vaping with cigarette-level cancer risk runs well ahead of what has actually been measured in people.\n","rough_estimate":"By 2022, >30% of US adults rated vaping as MORE harmful than cigarettes (up from ~3% in 2012); the share calling it less harmful fell from ~51% to ~17%","kind":"survey","survey_source":{"title":"Association of Vaping-Related Events with Relative Harm Perceptions of E-Cigarettes (HINTS 2012–2022, N=20,771)","publisher":"Nicotine & Tobacco Research / Wu et al.","url":"https://doi.org/10.1093/ntr/ntag024","year":2026}},"sources":[{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC11773639/","title":"Evidence update on the cancer risk of vaping e-cigarettes: A systematic review","publisher":"Tobacco Induced Diseases / Kundu et al.","source_type":"peer_reviewed","statistic":"Substantial biomarker evidence (DNA damage, oxidative stress) links e-cigarette exposure to cancer-risk pathways, but no significant cancer incidence was found in never-smoker vapers and overall human cancer evidence remains limited","excerpt":"\"There is substantial evidence that e-cigarette exposure is associated with biomarkers reflective of cancer disease risk. However, the overall evidence on cancer risk is still limited and should be further investigated by future research, particularly rigorously designed clinical trials and population-based research. [...] No significant incident or prevalent risk of lung cancer or other types of cancer was found in the never smoker current vapers population.\"\n","source_date":"2025-01-28","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260608043012/https://pmc.ncbi.nlm.nih.gov/articles/PMC11773639/","calculation_notes":"Kundu et al., Tobacco Induced Diseases vol. 23 (2025), DOI 10.18332/tid/192934. The review's own framing is the reason this entry carries no_reliable_estimate: \"substantial evidence\" attaches to BIOMARKERS (oxidative stress, DNA damage, genotoxicity) — that is hazard identification, not a probability. The human epidemiology that a lifetime figure would require does not exist: no significant cancer incidence has been observed in never-smoker vapers, and the authors call the actual human risk limited and uncertain. Two structural barriers explain the gap and are unlikely to clear soon — (1) latency: solid-tumour cancers from inhaled carcinogens typically take 20–40 years, and widespread vaping is only ~10–15 years old, so no exposed cohort has aged into the onset window; (2) confounding: most vapers are current or former smokers, so the small never-smoker vaper cohort is young and produces near-zero events with uselessly wide intervals. No attributable fraction or relative risk is available to convert into a lifetime probability, so none is published here.\n"},{"url":"https://www.ncbi.nlm.nih.gov/books/NBK507174/","title":"Public Health Consequences of E-Cigarettes — Chapter 10: Cancers","publisher":"National Academies of Sciences, Engineering, and Medicine (NASEM)","source_type":"reputable_reference","statistic":"Formal conclusions: no available evidence on e-cigarette use and cancer in humans (10-1); limited animal/biomarker and DNA-damage evidence (10-2, 10-3); substantial evidence that aerosol chemicals like formaldehyde and acrolein can damage DNA (10-4)","excerpt":"\"Conclusion 10-1. There is no available evidence whether or not e-cigarette use is associated with intermediate cancer endpoints in humans. [...] Conclusion 10-3. There is limited evidence that e-cigarette aerosol can be mutagenic or cause DNA damage in humans, animal models, and human cells in culture. Conclusion 10-4. There is substantial evidence that some chemicals present in e-cigarette aerosols (e.g., formaldehyde, acrolein) are capable of causing DNA damage and mutagenesis.\"\n","source_date":"2018-01-23","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20250404073244/https://www.ncbi.nlm.nih.gov/books/NBK507174/","calculation_notes":"NASEM's 2018 report drew the line this entry rests on: the chemistry can damage DNA (a hazard — Conclusion 10-4, \"substantial evidence\") while the human cancer association is simply unmeasured (Conclusion 10-1, \"no available evidence\"). That is the exact hazard-versus-magnitude split that distinguishes a no_reliable_estimate entry from a quantified one. The 2025 Kundu systematic review confirms the gap had not closed seven years later: still biomarker-deep, still epidemiologically empty.\n"},{"url":"https://www.cdc.gov/mmwr/volumes/73/wr/mm7341a2.htm","title":"Tobacco Product Use Among US Middle and High School Students — National Youth Tobacco Survey, 2024","publisher":"CDC / Morbidity and Mortality Weekly Report","source_type":"govt_report","statistic":"7.8% of US high school students currently used e-cigarettes in 2024 (down from 10.0% in 2023); e-cigarettes remain the most-used tobacco product among US youths","excerpt":"\"In 2024, current (previous 30-day) use of any tobacco product was reported by 10.1% of high school students (representing 1.58 million students) and 5.4% of middle school students (representing 640,000 students). [...] Among high school students, declines occurred in current use of any tobacco product (from 12.6% to 10.1%) and e-cigarettes (from 10.0% to 7.8%).\"\n","source_date":"2024-10-17","source_accessed":"2026-06-20","archive_url":"http://web.archive.org/web/20260511132325/https://www.cdc.gov/mmwr/volumes/73/wr/mm7341a2.htm","calculation_notes":"Prevalence context, not a cancer rate. Millions of US adolescents and adults vape regularly, so an exposure that is plausibly carcinogenic but not yet quantifiable still matters at population scale — the absence of a per-person lifetime number is a statement about evidence maturity, not about safety. Used to characterise the size of the exposed population, not to derive a probability.\n"}],"comparison_anchors":[{"label":"Lifetime cancer from any cause (US adult)","lifetime_us_adult":0.394},{"label":"Lifelong smoker dies of a smoking-related disease","lifetime_us_adult":0.5}],"short_label":"Vaping cancer","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"This entry is scoped to cancer specifically attributable to e-cigarette / vaping use, and the no_reliable_estimate flag is the honest state of the evidence: direction (vaping is plausibly, perhaps likely, carcinogenic) is supported, but magnitude is not measurable yet, so no lifetime probability is published. Three points bound the claim. First, \"less than smoking\" is not \"zero\" — every body that warns about vaping's cancer potential also holds that vaping is substantially less harmful than cigarettes; the popular and rising belief that vaping is as or more cancer-causing than smoking (Conclusion: it is not, on current evidence) is the overrated component this entry flags. Second, the figure that would let anyone quantify this — a vaping-specific relative risk or attributable fraction — does not exist, largely because of cancer's multi-decade latency against vaping's short history. Third, this entry does not cover the acute EVALI lung-injury outbreak (a separate, causally specific event with its own entry) or vaping's cardiovascular and respiratory effects, which are distinct ongoing questions. If credible cohort or attributable-risk data emerge, this entry should be re-scoped from no_reliable_estimate to a quantified estimate.\n","quality_score":{"d1":5,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":4,"d8":5,"avg":4.375,"scored_by":"claude-code-8d","scored_at":"2026-06-20","methodology_version":"1.2"},"reviewer":"8d-eval-2026-06-20","last_reviewed":"2026-06-20","reviewed":true,"generated_at":"2026-06-20","image":{"alt":"A vape device beside a small DNA-helix motif on a neutral surface, flat vector illustration in muted tones."},"canonical_url":"https://likelier.app/vaping-cancer-risk","api_url":"https://likelier.app/api/fears/vaping-cancer-risk.json"},{"slug":"walking-in-rain-getting-sick","question":"What are the odds of getting sick from walking in the rain without an umbrella?","quick_answer":"No number applies: walking in rain is not a measurable route to illness, because rainwater on skin does not transmit respiratory viruses. The seasonal link between cold, wet weather and colds is driven by indoor crowding and virus survival, not by getting wet.\n","category":"health","no_reliable_estimate":true,"perceived":{"description":"One of the most durable folk beliefs in the world: walking in rain without an umbrella or raincoat will give you a cold, flu, or pneumonia. The belief is near-universal across cultures — Eastern European grandmothers, East Asian parents, Latin American aunts, and Midwestern dads all converge on the same causal model: wet skin → illness. No large-scale survey has quantified how high people rate this probability numerically, but the qualitative conviction is strong enough that \"come in out of the rain\" functions as both literal advice and metaphor for common sense. The intuited mechanism is that rain on skin somehow causes respiratory infection directly, bypassing the need for a virus.\n","rough_estimate":"Commonly framed as near-certain ('you'll catch your death'), with no numerical estimate attached","kind":"intuition"},"sources":[{"url":"https://pubmed.ncbi.nlm.nih.gov/16286463/","title":"Acute cooling of the feet and the onset of common cold symptoms","publisher":"Family Practice (Oxford Academic), via PubMed","source_type":"peer_reviewed","statistic":"13/90 chilled subjects vs 5/90 controls self-reported cold symptoms in the 4–5 days after a 20-minute cold-foot immersion (P=0.047)","excerpt":"\"There is a common folklore that chilling of the body surface causes the development of common cold symptoms, but previous clinical research has failed to demonstrate any effect of cold exposure on susceptibility to infection with common cold viruses. [...] 13/90 subjects who were chilled reported they were suffering from a cold in the 4/5 days after the procedure compared to 5/90 control subjects (P=0.047). [...] Acute chilling of the feet causes the onset of common cold symptoms in around 10% of subjects who are chilled. Further studies are needed to determine the relationship of symptom generation to any respiratory infection.\"\n","source_date":"2005-12-01","source_accessed":"2026-04-19","archive_url":"https://web.archive.org/web/20260504061624/https://pubmed.ncbi.nlm.nih.gov/16286463/","calculation_notes":"The closest experimental evidence to the folk belief. Critical qualifier the authors themselves flag: the study measured self-reported symptoms, not laboratory-confirmed new infections. The proposed mechanism is that reflex vasoconstriction in the upper airway on cold-foot exposure reduces mucosal blood flow and temporarily lowers local defences, converting a pre-existing subclinical carriage of rhinovirus into a symptomatic cold. This is a modulation effect in people already carrying a virus, not a causation effect from cold/wet exposure alone. The absolute difference is 8 percentage points (14% vs 6%) with wide confidence bounds; no comparable replication exists. Walking in rain is a weaker chilling stimulus than the 20-minute ice-water foot immersion used in the trial, so the study represents an upper bound on any cold-exposure effect, not a direct measure of rain-walking risk.\n","independence_note":"Independent single-centre RCT at Cardiff Common Cold Centre. Editorially independent of the CDC source below.\n"},{"url":"https://www.cdc.gov/common-cold/about/index.html","title":"About the Common Cold","publisher":"US Centers for Disease Control and Prevention","source_type":"govt_report","statistic":"More than 200 respiratory viruses cause colds; rhinoviruses are the most frequent cause; primary spread is droplets and contact","excerpt":"\"More than 200 respiratory viruses can cause colds. Rhinoviruses are the most frequent cause of colds in the United States. [...] Most respiratory viruses are spread through droplets that an infected person releases when they cough or sneeze. These droplets can enter your body if you breathe them in or touch a contaminated surface and then touch your eyes, nose, or mouth.\"\n","source_date":"2026-02-19","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260402134530/https://www.cdc.gov/common-cold/about/index.html","calculation_notes":"CDC's current patient-facing page is the authoritative anchor for the \"colds are viral, not thermal\" frame. Rain, cold air, and wet clothing are not listed as transmission routes. The page identifies droplet inhalation and contaminated-surface contact as the mechanisms — both require a virus source (an infected person), not a weather event. This directly contradicts the folk model in which rain on skin produces illness independently of viral exposure.\n","independence_note":"Institutional CDC public-health guidance; editorially independent of the Eccles clinical trial.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/18977127/","title":"Cold temperature and low humidity are associated with increased occurrence of respiratory tract infections","publisher":"Respiratory Medicine, via PubMed","source_type":"peer_reviewed","statistic":"1-unit decrease in temperature associated with increased RTI occurrence; low humidity independently associated with respiratory tract infections in Finnish military conscripts (n=892)","excerpt":"\"We examined whether the development of acute respiratory tract infections (RTI) is potentiated by cold exposure and lowered humidity. [...] Cold temperature and low humidity were associated with increased occurrence of RTI, and a decrease in temperature and humidity preceded the onset of the infections.\"\n","source_date":"2009-03-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260424130303/https://pubmed.ncbi.nlm.nih.gov/18977127/","calculation_notes":"Mäkinen et al. (2009) supplies the population-level seasonal correlation that makes the folk belief feel true. In 892 Finnish military conscripts, drops in temperature and humidity preceded spikes in respiratory tract infections. Critically, this is an ecological association — it does not show that individual cold/wet exposure causes infection, only that RTI incidence tracks ambient weather. The study authors note that cold air dries mucous membranes and that barracks crowding increases during cold weather, both of which facilitate viral transmission. The finding is consistent with the standard epidemiological model: the seasonal pattern is real, but the causal pathway runs through virus survival, mucosal drying, and indoor crowding, not through rain on skin.\n","independence_note":"Finnish prospective cohort study with no authorship, funding, or institutional overlap with the Cardiff Eccles group or the CDC source. Methodologically independent ecological study.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/25561542/","title":"Temperature-dependent innate defense against the common cold virus limits viral replication at warm temperature in mouse airway cells","publisher":"Proceedings of the National Academy of Sciences (PNAS), via PubMed","source_type":"peer_reviewed","statistic":"Rhinovirus-infected mouse airway cells show markedly higher antiviral interferon-response gene expression at 37°C than at 33°C, explaining why the virus replicates more efficiently at cooler nasal-cavity temperatures","excerpt":"\"Most isolates of human rhinovirus, the common cold virus, replicate more robustly at the cool temperatures found in the nasal cavity (33-35 °C) than at core body temperature (37 °C). [...] Mouse airway cells infected with mouse-adapted rhinovirus 1B exhibited a striking enrichment in expression of antiviral defense response genes at 37 °C relative to 33 °C [...] These findings demonstrate that in mouse airway cells, rhinovirus replicates preferentially at nasal cavity temperature due, in part, to a less efficient antiviral defense response of infected cells at cool temperature.\"\n","source_date":"2015-01-20","source_accessed":"2026-07-03","archive_url":"http://web.archive.org/web/20260418074821/https://pubmed.ncbi.nlm.nih.gov/25561542/","calculation_notes":"Foxman et al. (Yale University) is the mechanistic study the body text and caveats reference as the \"kernel of truth\" behind the folk belief: rhinovirus replicates preferentially at cooler nasal-cavity temperatures in part because infected airway cells mount a weaker interferon/antiviral response at 33°C than at 37°C. This is an in-vitro mouse-airway-cell study, not a human clinical trial or a measurement of rain-exposure infection rates — it supports biological plausibility for the temperature-dependent effect Eccles (2005) measured in humans, but does not itself quantify any human infection probability. Added as a formal citation because the body text and caveats named this study by author/journal/year without a corresponding sources[] entry.\n","independence_note":"Yale University in-vitro study; institutionally and methodologically independent of the Cardiff Common Cold Centre (Eccles) clinical trial and the Finnish military cohort (Mäkinen et al.) ecological study also cited in this entry.\n"}],"comparison_anchors":[{"label":"Person-to-person common cold transmission (sharing indoor air with infected person, ~1 hour)","lifetime_us_adult":0.99},{"label":"Shared-cup sip, any serious infection (healthy adults)","lifetime_us_adult":0.000001},{"label":"Bare feet indoors in heated home → cold (effectively zero)","lifetime_us_adult":0.000001}],"regional_breakdown":[{"region":"Light drizzle, clothed, 30-minute walk in mild weather (10–15 °C)","probability":0.000001,"notes":"Minimal chilling stimulus, far weaker than the Eccles ice-water protocol. Without an underlying viral exposure, a brief walk in light rain does not produce respiratory infection. Point estimate is a structural \"effectively zero\" placeholder, not a measured rate.\n"},{"region":"Heavy rain, soaked through, 2+ hours in cold weather (<5 °C)","probability":0.000001,"notes":"Sustained wet-cold exposure is a real physiological stressor (risk of hypothermia if severe enough), but hypothermia is a thermoregulatory emergency, not a cold or flu. In the absence of viral exposure, even prolonged soaking does not cause respiratory infection. If the person is already carrying a subclinical virus, the Eccles data suggest a modulation effect of at most ~8 percentage points, but that effect requires the virus to already be present.\n"},{"region":"Cold rain + already incubating a subclinical respiratory virus","probability":0.14,"notes":"Upper bound from the Eccles 2005 trial: ~14% of subjects who received an acute cold-foot stimulus self-reported cold symptoms within 4–5 days, vs ~6% of controls. This is the scenario where the folk belief has a kernel of truth — but the operative variable is the pre-existing virus, not the rain. The rain/cold is a modest amplifier, not the cause.\n"}],"personal_factor_multipliers":[{"factor":"Already incubating a respiratory virus","multiplier":1000,"notes":"The overwhelming risk factor. Without a virus, rain cannot cause a cold. With a virus already in the upper airway, cold exposure may modestly increase the probability of symptoms developing (Eccles 2005), but the virus is doing the work. The multiplier is notional — the baseline \"rain alone → cold\" probability is effectively zero, so any multiplier applied to zero remains negligible in absolute terms.\n"},{"factor":"Immunocompromised (e.g. transplant recipient, active chemotherapy)","multiplier":3,"notes":"Immunocompromised individuals clear subclinical viral carriage less efficiently and may be more susceptible to symptom conversion from any physiological stressor, including cold exposure. The multiplier is approximate — no study has measured rain-specific infection rates in this population.\n"},{"factor":"Changed into dry clothes promptly after getting wet","multiplier":0.5,"notes":"Reducing the duration of cold-wet exposure limits any vasoconstriction effect on nasal mucosa. Changing clothes, warming up, and drying off quickly is the standard advice and is likely to reduce whatever marginal symptom-conversion risk exists. The multiplier is directional, not measured.\n"}],"short_label":"Rain & getting sick","myth_framing":"overrated","outcome_severity":"minor_harm","exposure_pattern":"recurring","outcome_type":"inconvenience","valence":"negative","caveats":"The folk belief is wrong about mechanism but not entirely wrong about seasonal correlation, and that distinction matters. Cold and wet seasons genuinely produce more respiratory illness — the Mäkinen et al. (2009) data from Finnish military conscripts show temperature and humidity drops preceding spikes in respiratory tract infections. But the causal pathway runs through virus survival in cold dry air, mucosal drying, and indoor crowding during bad weather, not through rainwater on skin. People observe the correlation (it rains, then they get sick) and misidentify the cause (the rain) instead of the mechanism they cannot see (inhaling someone else's virus indoors because the weather drove everyone inside).\nThe Eccles (2005) chilling experiment is the strongest evidence that cold exposure does something — but what it does is convert subclinical viral carriage into symptomatic illness, not cause infection from nothing. The absolute effect size is modest (8 percentage points), the chilling protocol (20 minutes of ice-water foot immersion) is more severe than most rain exposure, and no replication at comparable scale exists. The Foxman et al. (2015) mouse-cell study provides a plausible biological mechanism (rhinovirus replicates better and innate immune response is weaker at cooler nasal temperatures), but the translation from in-vitro mouse airway cells to \"you walked to the bus stop in the rain\" is a long chain of extrapolation.\nWhere the entry does not apply: prolonged extreme cold-wet exposure (soaked clothing, wind, near-freezing temperatures for hours) is a real risk for hypothermia, which is a thermoregulatory emergency, not a cold or flu. That is a different entry with a different mechanism and a different outcome severity.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A single closed umbrella resting on its side, rendered as a flat vector shape in muted blue-grey and off-white tones, centered on a calm empty background."},"canonical_url":"https://likelier.app/walking-in-rain-getting-sick","api_url":"https://likelier.app/api/fears/walking-in-rain-getting-sick.json"},{"slug":"wartime-conscription-europe","question":"What is the probability of a civilian being forcibly drafted when a European country enters active conventional war?","quick_answer":"No single probability fits: the odds of being forcibly drafted are entirely country-specific, and no source measures per-person wartime mobilization rates. In peacetime, conscription ranges from none in all-volunteer states like Germany and France to most of the male cohort in Finland — and these rates rise sharply in an active war.\n","category":"other","no_reliable_estimate":true,"perceived":{"description":"Europeans in countries without active conscription (most of Western and Central Europe) perceive the probability of being personally drafted in a future war as low, often conceptually near-zero, anchored on the post-Cold War elimination of mandatory service in countries like Germany (2011), France (2001), and Italy (2005). People in countries with active conscription systems (Finland, Estonia, Norway, Lithuania) have a much more calibrated sense of their personal military obligation. The question of being drafted into an active hot war is distinct from the question of peacetime conscription, and no high-quality survey specifically measures public estimates of per-person wartime draft probability in Europe. The perceived side is editorial intuition.\n","rough_estimate":"Western Europeans generally estimate their personal wartime draft probability as very low, perhaps 1 in 20 to 1 in 100; in countries with active conscription the personal obligation is near-certain for military-age men during an active conflict","kind":"intuition"},"sources":[{"url":"https://www.osw.waw.pl/en/publikacje/osw-commentary/2024-09-23/universal-selective-and-lottery-based-conscription-nordic-and","title":"Universal, selective, and lottery-based: conscription in the Nordic and Baltic states","publisher":"OSW Centre for Eastern Studies (Warsaw)","source_type":"reputable_reference","statistic":"Finland drafts the majority of each male age cohort; Finland and Estonia together trained approximately 28,000 reservists annually as of 2023; Estonia conscripts fewer than half of its male cohort; Norway and Denmark extend conscription to women; Lithuania reinstated conscription in 2015; Sweden in 2017; Latvia in 2022.","excerpt":"\"In Finland, the majority of each age cohort is drafted, whereas in Estonia, fewer than half enter compulsory service. [...] Currently, both states train approximately 28,000 reservists annually (2023), representing a three-quarter increase for Finland and nearly a fivefold rise for Estonia compared to 2021. [...] Lithuania (2015), Sweden (2017), and Latvia (2022) reinstated conscription.\"\n","source_date":"2024-09-23","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260525101051/https://www.osw.waw.pl/en/publikacje/osw-commentary/2024-09-23/universal-selective-and-lottery-based-conscription-nordic-and","calculation_notes":"This source documents the current peacetime conscription landscape in Nordic and Baltic states. Finland drafts the majority of its male cohort each year; the source reports that Finland and Estonia together train approximately 28,000 reservists annually (2023). Estonia drafts fewer than 50% of its male cohort. These are peacetime rates; in an active war these rates increase sharply and age ranges expand. The source establishes baseline conscription obligations but does not provide wartime mobilization rates. Used to document the range of conscription regimes and establish that \"probability of being drafted\" is highly country-specific.\n","independence_note":"OSW Centre for Eastern Studies is an independent Warsaw-based research institution; independent of government sources on conscription policy description.\n"},{"url":"https://www.euronews.com/my-europe/2026/04/27/defence-which-european-countries-have-mandatory-and-voluntary-military-service","title":"Defence: Which European countries have mandatory and voluntary military service?","publisher":"Euronews","source_type":"reputable_reference","statistic":"At least 10 EU member states currently have some form of conscription; Finland, Estonia, Lithuania, Latvia, Norway, and Denmark have mandatory service systems; Germany, France, Italy, and Spain have all-volunteer forces.","excerpt":"\"At least 10 EU member states currently have conscription in place, including Austria, Cyprus, Denmark, and Estonia. Among these, Austria, Croatia, Cyprus, Estonia, Finland, Greece, Latvia and Lithuania make it mandatory only for men, while Denmark and Sweden make it compulsory for both men and women.\"\n","source_date":"2026-04-27","source_accessed":"2026-05-09","archive_url":"http://web.archive.org/web/20260428101735/https://www.euronews.com/my-europe/2026/04/27/defence-which-european-countries-have-mandatory-and-voluntary-military-service","calculation_notes":"Maps the current European conscription landscape. Countries with mandatory service already carry a near-certain personal military obligation for military-age men during active conflict. Countries without conscription (Germany, France, Spain, Italy, Poland currently) would need to reintroduce it, which historically takes weeks to months after a war starts. The all-volunteer/conscription distinction is itself country-specific and has been changing rapidly (Latvia reintroduced conscription in 2024; Poland announced plans for universal male military training in 2025). A single European-wide probability is not meaningful.\n","independence_note":"Euronews reporting based on official government sources across EU member states; independent of OSW analysis.\n"},{"url":"https://www.osw.waw.pl/en/publikacje/analyses/2025-12-10/germanys-military-service-reform-voluntarism-its-core","title":"Germany's military service reform: voluntarism at its core","publisher":"OSW Centre for Eastern Studies (Warsaw)","source_type":"reputable_reference","statistic":"Germany's Bundestag passed its military service reform on 5 December 2025; from 2026 every male German citizen reaching age 18 must complete a registration questionnaire, with mandatory medical examinations from July 2027 for men born 2008 and later. Service itself remains voluntary; a compulsory draft requires a further parliamentary vote.","excerpt":"\"On 5 December, the Bundestag passed a law reforming military service. [...] From 2026, the new regulations will require male German citizens residing in Germany who have reached the age of 18 to complete a questionnaire. [...] From July 2027, men born in 2008 and thereafter will be required to undergo medical examinations. [...] If the quotas are not met, or if Germany's 'defence situation' so requires, parliament may introduce a compulsory draft based on a lottery to fill the shortfall.\"\n","source_date":"2025-12-10","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260115173332/https://www.osw.waw.pl/en/publikacje/analyses/2025-12-10/germanys-military-service-reform-voluntarism-its-core","calculation_notes":"Documents the 2025-2026 shift in Germany's all-volunteer force toward compulsory peacetime registration and screening. This is distinct from -- and a precursor to -- the wartime draft this entry measures: even in a fully voluntary system, a young German man's odds of compulsory registration and (from July 2027) a mandatory medical exam are now near-certain, while actual conscription into service still requires a separate Bundestag vote. The figure is a peacetime administrative-obligation rate, not a wartime mobilization rate, so it does not change this entry's no_reliable_estimate flag.\n","independence_note":"OSW Centre for Eastern Studies is an independent Warsaw-based research institution; describes German federal legislation from a third-country analytic vantage.\n"},{"url":"https://www.osw.waw.pl/en/publikacje/analyses/2025-11-05/short-conscription-period-big-ambitions-croatia-reinstates","title":"Short conscription period, big ambitions: Croatia reinstates compulsory military service","publisher":"OSW Centre for Eastern Studies (Warsaw)","source_type":"reputable_reference","statistic":"Croatia's parliament adopted a law reinstating compulsory military service on 24 October 2025; from January 2026 it is mandatory for men aged 19 to 29, with eight-week basic training. The plan envisages around 4,000 conscripts per year, in five cycles of 800 -- a selective fraction of the male cohort, not universal call-up.","excerpt":"\"On 24 October, the Croatian parliament (Sabor) adopted a law reinstating compulsory military service. [...] From January 2026, conscription will be mandatory for men aged 19 to 29. [...] The basic training will last eight weeks. [...] The plan envisages the annual conscription of around 4,000 recruits, divided into five training cycles of 800 individuals each.\"\n","source_date":"2025-11-05","source_accessed":"2026-06-14","archive_url":"http://web.archive.org/web/20260312051246/https://www.osw.waw.pl/en/publikacje/analyses/2025-11-05/short-conscription-period-big-ambitions-croatia-reinstates","calculation_notes":"Documents Croatia's 2025-2026 reinstatement of peacetime compulsory service after a 17-year hiatus. The ~4,000 conscripts per year is well below Croatia's full male age cohort (roughly 18,000-20,000 men reach 19 each year), so the peacetime obligation is selective, not universal -- a young Croatian man is legally liable but only a minority are actually called in peacetime. Like the German example, this is a peacetime administrative/training obligation, distinct from the wartime mobilization rate this entry measures, and does not change the no_reliable_estimate flag.\n","independence_note":"OSW Centre for Eastern Studies is an independent Warsaw-based research institution; analytic vantage independent of Croatian government communications.\n"}],"comparison_anchors":[],"short_label":"Wartime conscription","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"autonomy_loss","valence":"negative","caveats":"A single probability of \"being drafted in a European country entering active war\" cannot be calculated because the number depends critically on: (a) which country -- a Finnish or Estonian man of military age is already legally obligated to serve and would be called up with near-certainty in a hot war, while a German or Spanish civilian faces no such obligation under current law; (b) whether the country chooses or is legally required to implement conscription on entering the war -- WWII-era data from Britain, Germany, the Soviet Union, and Finland shows wartime mobilization at 24-42% of the total male population (and all of the military-age cohort, with males 18-45 comprising roughly 20-25% of total population). Historical data establishes that countries entering major conventional wars mobilize an overwhelming fraction of military-age men: Germany mobilized approximately 42% of all males in WWII; the Soviet Union mobilized approximately 35%; Britain mobilized approximately 24%; Finland mobilized the majority of military-age men entirely. Ukraine, since February 2022, has mobilized approximately 1 million of an estimated ~10 million military-age men (roughly 10% of the military-age cohort within the country, rising over time as the mobilization law tightened). For any European country with active conscription and an ongoing hot war, the conditional probability of a military-age man being called up within 12-24 months approaches 0.5-0.9 based on historical precedent. For countries without current conscription, the conditional probability depends on whether the government chooses to reintroduce it -- an uncertain policy variable, not a stable base rate. Poland, as of 2025, announced plans for universal male military training but was considering a voluntary-incentive rather than mandatory model (Tusk, March 2025). The absence of a stable, country-agnostic probability makes no_reliable_estimate the appropriate flag for this entry.\nA distinct, more tractable trend has emerged in 2025-2026: compulsory peacetime registration and screening, well short of an actual wartime draft. Germany's military service reform (Bundestag, 5 December 2025; in force from 2026) obliges every male German citizen reaching age 18 to complete a registration questionnaire, with a mandatory medical examination from July 2027 for men born 2008 and later -- yet service itself stays voluntary, and a compulsory draft would still need a further Bundestag vote (OSW, 2025). Croatia reinstated compulsory service after a 17-year gap (Sabor, 24 October 2025; mandatory for men 19-29 from January 2026, eight-week training), but plans only ~4,000 conscripts per year against a male cohort of roughly 18,000-20,000 -- a selective, minority call-up rather than universal service (OSW, 2025). So a young European man's odds of facing compulsory registration, screening, or selective peacetime training before any war are rising and are now computable per country, even where the wartime-draft obligation itself remains a policy variable. These peacetime administrative obligations are a separate quantity from the wartime mobilization rate this entry measures and do not, on their own, yield a country-agnostic wartime-draft probability.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"war-research-agent-2026-05-09","last_reviewed":"2026-06-14","reviewed":true,"generated_at":"2026-05-09","image":{"alt":"An official-looking envelope resting on a plain surface, flat vector illustration in muted grey and beige tones."},"canonical_url":"https://likelier.app/wartime-conscription-europe","api_url":"https://likelier.app/api/fears/wartime-conscription-europe.json"},{"slug":"working-age-stroke-disability","question":"How likely is an adult under 65 to have a disabling stroke during their working years?","quick_answer":"No reliable single figure exists for the odds of a disabling stroke before 65: no cited source publishes a cumulative pre-65 probability, and a lifetime-risk number can't validly be multiplied by a share of incident strokes to manufacture one. What is documented is that over 62% of strokes occur under age 70 and over 16% in people aged 15 to 49.\n","category":"health","tags":["workplace"],"no_reliable_estimate":true,"perceived":{"description":"Stroke is widely perceived as a disease of the elderly. The dominant cultural image of stroke — a severe event in an 80-year-old — means working-age adults typically assign it a probability close to zero before retirement age. This perception is measurably incorrect. World Stroke Organization data show that over 62% of strokes worldwide occur in people under age 70 and over 16% in people aged 15–49. The underestimation is compounded by the relatively recent recognition that stroke incidence in the 15–49 age band has been rising globally, driven by increasing prevalence of hypertension, metabolic syndrome, and obesity in younger cohorts. Because pre-65 stroke is not the \"typical\" patient profile, vascular risk factors in working-age adults are undertreated relative to their older counterparts.\n","kind":"intuition"},"sources":[{"url":"https://www.thelancet.com/journals/laneur/article/PIIS1474-4422(24)00369-7/fulltext","title":"Global, regional, and national burden of stroke and its risk factors, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021","publisher":"The Lancet Neurology / IHME","source_type":"peer_reviewed","statistic":"Stroke incidence in 15–49 age group globally: 49.15/100,000/year men, 41.4/100,000/year women (GBD 2021); ~14% of all stroke incidence under age 50; rising trend +7.6% 1990–2021","excerpt":"\"In 2021, the age-standardised incidence rate of stroke for the 15–49 age group was 49.15 per 100,000 population per year for men and 41.44 per 100,000 per year for women globally. Approximately 14 percent of all stroke incidence worldwide occurred in individuals under the age of 50, with the proportion rising over the study period (1990–2021). The age-standardised incidence rate in this younger group increased by 7.6 percent between 1990 and 2021, driven predominantly by increases in metabolic risk factors including hypertension, obesity, and diabetes in younger cohorts.\"\n","source_date":"2024-10-01","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250809224223/https://www.thelancet.com/journals/laneur/article/PIIS1474-4422(24)00369-7/fulltext","calculation_notes":"GBD 2021 Stroke systematic analysis, Lancet Neurology 2024. The 49.15/100k and 41.44/100k annual incidence rates for 15–49 provide the primary denominator for the pre-65 calculation. The \"14% of all stroke incidence under 50\" figure, combined with the Feigin 2018 lifetime-risk data, anchors the cumulative 6% pre-65 estimate. The rising incidence trend (7.6% 1990–2021) implies the 6% baseline is a conservative current estimate.\n"},{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1804492","title":"Global Lifetime Stroke Risk: A Population-Based Study","publisher":"New England Journal of Medicine","source_type":"peer_reviewed","statistic":"Lifetime stroke risk from age 25: 24.9% globally (95% CI 23.5–26.2%); 1 in 4 adults worldwide will have a stroke in their remaining lifetime from age 25","excerpt":"\"The lifetime risk of stroke from age 25 was 24.9% (95% CI 23.5–26.2%) globally, with substantial regional variation — highest in East Asia (38.8%) and lowest in sub-Saharan Africa (11.8%). Risk estimates were calculated using GBD 2016 incidence, prevalence, and mortality data. Among those who survive to older ages, the lifetime risk from age 70 was 13.6%. These data indicate that stroke is an extremely common event across the global adult population.\"\n","source_date":"2018-12-20","source_accessed":"2026-05-04","archive_url":"http://web.archive.org/web/20250523073717/https://www.nejm.org/doi/full/10.1056/NEJMoa1804492","calculation_notes":"Feigin et al. NEJM 2018 — global lifetime stroke risk from GBD 2016. The 24.9% lifetime risk from age 25 is the denominator for calculating the pre-65 share. Combined with the GBD 2021 finding that ~20–25% of stroke incidence occurs before age 65, yields a cumulative pre-65 stroke probability of approximately 5–6%. This is used as the native rate (6/100).\n"},{"url":"https://www.world-stroke.org/assets/downloads/WSO_Global_Stroke_Fact_Sheet.pdf","title":"Global Stroke Fact Sheet 2022","publisher":"World Stroke Organization","source_type":"reputable_reference","statistic":"Globally, 1 in 4 adults over age 25 will have a stroke in their lifetime; over 62% of all strokes each year occur in people under age 70; over 143 million years of healthy life are lost each year to stroke-related death and disability","excerpt":"\"There are over 12.2 million new strokes each year. Globally, one in four people over age 25 will have a stroke in their lifetime. Each year, over 16% of all strokes occur in people 15-49 years of age. Each year, over 62% of all strokes occur in people under 70 years of age. Over 143 million years of healthy life is lost each year due to stroke-related death and disability. 15% of healthy life lost due to stroke-related death and disability affects people 15-49 years old. 57% of healthy life lost due to stroke-related death and disability affects people under the age of 70 years.\"\n","source_date":"2022-10-29","source_accessed":"2026-06-30","archive_url":"http://web.archive.org/web/20260617191357/https://www.world-stroke.org/assets/downloads/WSO_Global_Stroke_Fact_Sheet.pdf","calculation_notes":"World Stroke Organization Global Stroke Fact Sheet 2022 (GBD 2019 statistics). Corroborates that stroke is a working-age burden: over 62% of incident strokes and 57% of stroke-related healthy-life loss (DALYs) occur before age 70, and 16% of strokes occur in the 15-49 band. Used to confirm that a large share of the pre-65 stroke burden carries lasting disability (healthy-life loss / DALYs), supporting the disabling-stroke framing of the entry. Note: this fact sheet reports stroke as the second leading cause of death and third leading cause of death-and-disability combined (DALYs) globally; it does not publish a single \"share of survivors disabled at 6 months\" figure, so that fraction is taken from the wider stroke-outcome literature, not from this fact sheet.\n"}],"comparison_anchors":[],"short_label":"Working-age disabling stroke","myth_framing":"underrated","outcome_severity":"serious_harm","exposure_pattern":"acute","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"No cited source publishes a cumulative probability of stroke before age 65, so this entry carries no headline number. What the sources do state: lifetime stroke risk from age 25 is 24.9% globally (Feigin 2018), over 62% of strokes occur under age 70 and over 16% in ages 15–49 (WSO Fact Sheet 2022), and ~14% of stroke incidence is under age 50 with a rising 15–49 trend (GBD 2021). Age 65 sits between the under-50 (~14–16%) and under-70 (62%) shares, but no source states the pre-65 fraction, and a cohort lifetime-risk figure cannot be validly multiplied by a period incidence fraction to yield a cumulative-before-65 probability. Regional variation is large: East Asia has the highest lifetime stroke risk (38.8%); Sub-Saharan Africa and Eastern Europe have rising rates driven by uncontrolled hypertension. The stroke-survivor disability fraction cited in the older literature (~two-thirds with measurable functional disability at 6 months) is not published in these cited sources and may be lower in younger cohorts due to better neuroplasticity and rehabilitation access. The \"rising incidence in 15–49\" finding from GBD 2021 (+7.6% from 1990) means the working-age burden is directionally increasing.\n","quality_score":{"d1":3,"d2":5,"d3":4,"d4":4,"d5":4,"d6":4,"d7":3,"d8":5,"avg":4,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"8d-eval-2026-05-16","last_reviewed":"2026-05-16","reviewed":true,"generated_at":"2026-05-04","image":{"alt":"A flat vector illustration of a brain scan image on a light box, muted tones."},"canonical_url":"https://likelier.app/working-age-stroke-disability","api_url":"https://likelier.app/api/fears/working-age-stroke-disability.json"},{"slug":"youngest-in-class-harm","question":"What are the odds of being the youngest in class harming your child's academic or psychological development?","quick_answer":"No single probability captures this, since the effects span ADHD overdiagnosis, test scores, and self-esteem with different magnitudes. What is documented: the youngest children in class are about 34% more likely to be diagnosed with ADHD, a relative risk rather than an absolute harm rate, and the effect is large in early grades but mostly fades by adulthood.\n","category":"health","tags":["kids"],"no_reliable_estimate":true,"perceived":{"description":"The relative age effect is one of those parenting anxieties that sharpens into focus the moment a parent learns their child's birthday falls near the school enrollment cutoff. In countries with a September 1 cutoff, parents of July and August babies routinely agonize over whether to enroll on time or hold the child back a year. The fear has been amplified by a steady stream of research coverage since Gladwell's \"Outliers\" (2008) popularized Bedard and Dhuey's findings, and more recently by the Layton et al. 2018 NEJM study linking birth month to ADHD diagnosis rates. The parental intuition is that starting school as the youngest in class confers a lasting developmental disadvantage. That intuition is partly right: the short-term effects are real and larger than many expect. But the long-term trajectory is far more reassuring than the early-grades data suggests.\n","rough_estimate":"Most parents near the cutoff date believe the youngest-in-class disadvantage is large and lasting; the research shows it is large early but mostly transient","kind":"intuition"},"sources":[{"url":"https://www.nejm.org/doi/full/10.1056/NEJMoa1806828","title":"Attention Deficit-Hyperactivity Disorder and Month of School Enrollment","publisher":"New England Journal of Medicine (Layton, Barnett, Hicks, Jena 2018)","source_type":"peer_reviewed","statistic":"Among 407,846 children born 2007-2009, ADHD diagnosis rate was 85.1 per 10,000 for August-born children vs 63.6 per 10,000 for September-born children in states with a September 1 kindergarten cutoff — a 34% relative increase. ADHD treatment rate was 52.9 vs 40.4 per 10,000 (31% higher). No significant difference existed for other consecutive birth months or in states without a September 1 cutoff.","excerpt":"\"The rate of ADHD diagnosis among children in states with a September 1 cutoff was 85.1 per 10,000 children among those born in August and 63.6 per 10,000 children among those born in September.\"\n","source_date":"2018-11-29","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20241212082340/https://www.nejm.org/doi/full/10.1056/NEJMoa1806828","calculation_notes":"Layton et al. 2018 is the gold-standard natural experiment on relative age and ADHD diagnosis. The study used insurance claims data for 407,846 children born 2007-2009 across 18 states with a September 1 kindergarten cutoff. The rate ratio 85.1/63.6 = 1.337, yielding the 34% excess diagnosis figure. Crucially, the same August-September comparison in states without a September 1 cutoff showed no significant difference, confirming that the effect is driven by relative age within the classroom rather than by birth month per se. The native numerator (85.1) and denominator (10,000) represent the August-born ADHD diagnosis rate. The normalized 0.34 represents the relative excess risk.\n","independence_note":"Independent dataset from Elder and Lubotsky 2009. Layton et al. use commercial insurance claims; Elder and Lubotsky use NHIS and ECLS-K survey data. Both reach concordant conclusions about relative age and ADHD diagnosis using entirely different data sources and methodologies.\n"},{"url":"https://academic.oup.com/qje/article-abstract/121/4/1437/1855234","title":"The Persistence of Early Childhood Maturity: International Evidence of Long-Run Age Effects","publisher":"Quarterly Journal of Economics (Bedard and Dhuey 2006)","source_type":"peer_reviewed","statistic":"Across 19 OECD countries using TIMSS data, the youngest children in each grade cohort scored 4-12 percentile points lower than the oldest children in grade 4, and 2-9 percentile points lower in grade 8. In Canada and the US, youngest-cohort members were less likely to attend university.","excerpt":"\"The youngest members of each cohort score 4-12 percentiles lower than the oldest members in fourth grade.\"\n","source_date":"2006-11-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20210108191746/https://academic.oup.com/qje/article-abstract/121/4/1437/1855234","calculation_notes":"Bedard and Dhuey 2006 is the foundational cross-national study on relative age effects in academic performance. The 4-12 percentile gap at grade 4 narrows to 2-9 percentile points by grade 8, demonstrating partial but incomplete fade. The study used TIMSS 1995 and 1999 international mathematics and science data from countries with clear enrollment cutoff dates. The range reflects variation across countries and subjects. This study anchors the academic-performance dimension of the youngest-in-class effect but does not produce a probability figure used in the normalized estimate.\n","independence_note":"Entirely independent from the Layton ADHD data. Different outcome (test scores vs ADHD diagnosis), different data source (TIMSS international assessments vs US insurance claims), different time period.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC2933294/","title":"The Importance of Relative Standards in ADHD Diagnoses: Evidence Based on Exact Birth Dates","publisher":"Journal of Health Economics (Elder 2010, building on Elder and Lubotsky 2009)","source_type":"peer_reviewed","statistic":"Children born just before the kindergarten cutoff date had ADHD diagnosis rates 8.4% vs 5.1% for those born just after — a 65% relative increase. They were also more than twice as likely to regularly use methylphenidate (Ritalin). The effect was driven primarily by teacher assessments, not parental assessments.","excerpt":"\"Roughly 8.4 percent of children born in the month prior to their state's cutoff date for kindergarten eligibility are diagnosed with ADHD, compared to 5.1 percent of children born in the month immediately afterward.\"\n","source_date":"2010-08-01","source_accessed":"2026-04-19","archive_url":"http://web.archive.org/web/20260421202249/https://pmc.ncbi.nlm.nih.gov/articles/PMC2933294/","calculation_notes":"Elder 2010 (expanding Elder and Lubotsky 2009) uses ECLS-K and NHIS data to show that the youngest children in kindergarten cohorts are dramatically more likely to receive ADHD diagnoses and stimulant medication. The 8.4% vs 5.1% comparison yields a 65% relative increase, even larger than the 34% found in Layton et al. 2018. The higher relative effect likely reflects different age ranges and data sources (survey vs claims). The key mechanistic finding is that teachers drive the diagnosis disparity: teacher assessments of ADHD symptoms are strongly correlated with relative age, while parental assessments are only weakly correlated. This suggests teachers are comparing children to classmates who may be up to 11 months older, interpreting normal developmental variation as pathology.\n","independence_note":"Uses ECLS-K and NHIS survey data, independent from the Layton et al. 2018 insurance claims data. Different methodology and era but concordant findings, strengthening the evidence base.\n"}],"comparison_anchors":[{"label":"ADHD diagnosis rate, oldest in class (September-born, baseline)","lifetime_us_adult":0.00636},{"label":"Miscarriage in recognized pregnancy","lifetime_us_adult":0.15}],"regional_breakdown":[{"region":"Early grades (K-3)","probability":0.34,"notes":"Largest effects. Youngest children score 4-12 percentile points lower on standardized tests (Bedard and Dhuey 2006) and are 34-65% more likely to receive an ADHD diagnosis (Layton 2018, Elder 2010). Teachers compare children to same-grade peers who may be nearly a year older, amplifying perceived behavioral and academic gaps."},{"region":"Middle school (grades 5-8)","probability":0.15,"notes":"Effects diminish. Academic gap narrows to 2-9 percentile points by grade 8 (Bedard and Dhuey 2006). ADHD diagnosis disparity also shrinks as children mature and the 11-month age gap becomes proportionally smaller relative to total age."},{"region":"High school (grades 9-12)","probability":0.05,"notes":"Effects are mostly gone for academic performance. Some residual impact on track placement in systems that sort students early (Muhlenweg and Puhani 2010 found youngest students in Germany were one-third less likely to be assigned to the academic track at age 10, with partial correction by age 16)."},{"region":"University and adulthood","probability":0.02,"notes":"Minimal measurable effects. Bedard and Dhuey 2006 found a small effect on university attendance in Canada and the US, but most studies find no significant impact on adult earnings or employment. The relative age effect is primarily a childhood phenomenon."}],"personal_factor_multipliers":[{"factor":"Boy (vs girl)","multiplier":1.5,"notes":"Relative age effects on both academic performance and ADHD diagnosis are consistently larger for boys. Boys develop self-regulation skills later on average, amplifying the maturity gap between youngest and oldest in class."},{"factor":"Born in cutoff month (e.g., August in September-1 states)","multiplier":2,"notes":"Children born in the month immediately before the cutoff face the maximum relative age disadvantage — nearly 12 months younger than the oldest classmates. Layton et al. 2018 found the effect was concentrated entirely in this comparison."},{"factor":"High-resource family","multiplier":0.5,"notes":"Parents with resources can compensate through tutoring, enrichment activities, and advocacy that counteracts teacher misperception. The option to redshirt (delay enrollment) is also predominantly exercised by higher-income, white families."},{"factor":"Low-resource family","multiplier":1.5,"notes":"Effects persist longer when families lack resources to compensate. Schneeweis and Zweimuller 2014 found that relative age effects in Austria faded for children from favorable backgrounds but persisted for those from less favorable ones."},{"factor":"Early-tracking school system (e.g., Germany at age 10)","multiplier":2,"notes":"School systems that sort children into academic vs vocational tracks at young ages lock in relative age disadvantages. Muhlenweg and Puhani 2010 showed youngest students were only two-thirds as likely to be placed in the academic track in Germany."}],"short_label":"Youngest in class","myth_framing":"calibrated","outcome_severity":"moderate_harm","exposure_pattern":"cumulative","outcome_type":"chronic_illness","valence":"negative","caveats":"The normalized 0.34 figure represents a relative risk increase for ADHD diagnosis, not a conventional lifetime probability. It means the youngest children in class are 34% more likely to be diagnosed with ADHD than the oldest, not that 34% of youngest-in-class children are harmed. The absolute ADHD diagnosis rate difference is 21.5 per 10,000 children (0.85% vs 0.64%), which is small in absolute terms even though the relative increase is striking.\nThe ADHD finding is better understood as evidence of diagnostic contamination than as evidence of developmental damage. The same child, with the same brain, born one day earlier or later relative to an arbitrary administrative cutoff, receives a different probability of psychiatric diagnosis. Layton et al. confirmed this by showing no August-September difference in states without a September 1 cutoff.\nAcademic performance effects (4-12 percentile points in early grades) are real but mostly transient. The gap narrows substantially by middle school and is minimal by university. Long-term earnings effects are not consistently detected in the literature.\nRedshirting (delaying kindergarten entry by a year) has mixed evidence. It confers a short-term academic advantage of roughly 0.2 standard deviations in grades 3-5, but this advantage tends to fade by end of elementary school. For children with disabilities, delayed entry may be counterproductive because it delays access to school-based services.\nThe regional_breakdown rows use the relative excess ADHD diagnosis risk as a proxy, not absolute probabilities of harm. The probability values represent the approximate relative risk increase at each stage, declining from 0.34 in early grades toward negligible levels in adulthood.\n","quality_score":{"d1":5,"d2":5,"d3":5,"d4":4,"d5":5,"d6":5,"d7":4,"d8":5,"avg":4.75,"scored_by":"claude-code-8d","scored_at":"2026-05-25","methodology_version":"1.2"},"reviewer":"quality-review-agent","last_reviewed":"2026-04-19","reviewed":true,"generated_at":"2026-04-19","image":{"alt":"A row of school backpacks arranged from large to small against a classroom wall, flat vector illustration in soft muted tones."},"canonical_url":"https://likelier.app/youngest-in-class-harm","api_url":"https://likelier.app/api/fears/youngest-in-class-harm.json"},{"slug":"zero-alcohol-beer-pregnancy","question":"Is a \"0.0%\" (alcohol-free) beer truly zero alcohol, and safe during pregnancy or breastfeeding?","quick_answer":"There is no reliable harm number, and the honest answer is mostly reassuring. US rules define \"alcohol free\" and \"0.0%\" as containing no alcohol, with no tolerance, so a genuine one delivers essentially no exposure — the safest choice in the category. The catch is verification, not dose: \"0.0%\" is usually a density reading, and that method fails near zero, so lab testing has repeatedly caught \"alcohol-free\" labels holding real ethanol (0.009–0.385% by gas chromatography in one study; above 1% in six products in another). The number tells you the target, not a guaranteed assay.\n","category":"health","tags":["pregnancy","infant","substance-use","food"],"no_reliable_estimate":true,"perceived":{"description":"Two readings pull in opposite directions and neither is quite right. One treats \"0.0%\" as a certified, laboratory-absolute zero, which reads more precision into the label than the measurement behind it usually supports: the number is the brewer's target, and the routine way of arriving at it (density) is least reliable exactly at zero. The other treats anything still called \"beer\" as suspect, which ignores that a genuinely alcohol-free product has no ethanol to pass to a fetus or into breastmilk. The gap the intuition misses is that the reassurance is real but conditional on the label being true, and the \"contains no alcohol\" class is the very one independent testing has caught above zero most often.\n","kind":"intuition"},"sources":[{"url":"https://www.law.cornell.edu/cfr/text/27/7.65","title":"27 CFR 7.65 — Alcohol content (malt beverage labeling)","publisher":"US Alcohol and Tobacco Tax and Trade Bureau (TTB), via Cornell Legal Information Institute","source_type":"govt_report","statistic":"'Alcohol free' may be used only on malt beverages containing no alcohol, with no tolerances permitted; a malt beverage may not be labeled 0.0% ABV unless it is also labeled 'alcohol free' and contains no alcohol. 'Non-alcoholic' requires an adjacent 'contains less than 0.5 percent alcohol by volume' statement","excerpt":"\"A malt beverage may not be labeled with an alcohol content of 0.0 percent alcohol by volume, unless it is also labeled as 'alcohol free' ... and contains no alcohol.\" \"No tolerances are permitted for 'alcohol free' malt beverages.\"\n","source_date":"2023-01-01","source_accessed":"2026-08-01","calculation_notes":"Establishes the regulatory absolutism this entry turns on. In the US, \"alcohol free\" and a \"0.0%\" claim legally mean the product contains no alcohol at all, with zero tolerance — unlike the \"non-alcoholic\" class, which is defined as under 0.5% ABV. So a genuine 0.0% product should deliver 0 g ethanol and therefore no fetal or breastmilk exposure. The rule constrains the label, not the assay: it defines what \"0.0%\" must mean, which is precisely why measured deviations from it (see the two chromatography sources) are the load-bearing hazard rather than the labeled dose. Regulatory text, not a harm-rate study.\n"},{"url":"https://link.springer.com/article/10.1007/BF02259306","title":"Gas-liquid chromatographic determination of ethanol in 'Alcohol-Free' beverages and fruit juices","publisher":"Chromatographia — Morad, Hikal & Buchanan, 1980","source_type":"peer_reviewed","statistic":"Beverages labeled 'Alcohol-Free' (five beers plus grape and apple juices) measured 0.009% to 0.385% ethanol by volume via gas chromatography — i.e., not truly alcohol-free but alcohol-low [Paraphrase from abstract — full text paywalled]","excerpt":"\"[Paraphrase from abstract — full text paywalled] Products labeled 'Alcohol-Free' were found by gas–liquid chromatography to contain ethanol from 0.009% to 0.385% by volume, indicating they are not alcohol-free but alcohol-low.\"\n","source_date":"1980-01-01","source_accessed":"2026-08-01","calculation_notes":"Direct evidence that an \"alcohol-free\" label is a claim about a measurement, not a guarantee of literal zero. Gas chromatography (which detects ethanol directly, down to ~0.005% v/v) found \"alcohol-free\" beverages spanning 0.009–0.385% v/v. Dose translation via NIAAA's 14 g standard drink: at the 0.385% upper end a 12 fl oz (355 mL) serving holds 355 × 0.00385 × 0.789 g/mL = 1.08 g ethanol ≈ 1/13 of a standard drink — comparable to a \"non-alcoholic\" (≤0.5%) beer; at the 0.009% low end, ~0.025 g ≈ 1/560 of a standard drink, negligible. Caveat: this is a 1980 study with a small brand sample and older market practices; cited for the qualitative point (measured ethanol in \"alcohol-free\" labels), not as a current prevalence.\n"},{"url":"https://pubmed.ncbi.nlm.nih.gov/20051610/","title":"Alcohol content in declared non- or low alcoholic beverages: implications to pregnancy","publisher":"Canadian Journal of Clinical Pharmacology — Goh, Verjee & Koren (Motherisk, Hospital for Sick Children), 2010","source_type":"peer_reviewed","statistic":"Of 45 beverages labeled non- or low-alcohol, 6 that claimed to contain no alcohol were found to contain greater than 1% ethanol; 13 of 45 (29%) exceeded their declared alcohol content","excerpt":"\"Six beverages claiming to contain no alcohol were found to contain greater than 1% ethanol.\"\n","source_date":"2010-01-01","source_accessed":"2026-08-01","calculation_notes":"The more recent corroboration, and the sharper one: six products that claimed to contain no alcohol tested above 1% ABV. At 1% v/v a 12 fl oz serving is 355 × 0.01 × 0.789 = 2.8 g ethanol ≈ 1/5 of a standard drink — so a \"no alcohol\" label spanned, in testing, from genuinely negligible to a fifth of a real drink. Establishes that the zero-claim label class is the one testing catches above zero, which is the entry's core point. Canadian market, 2010; read as order-of-magnitude, not a current rate. See also the sibling entry [[non-alcoholic-beer-pregnancy]] for the under-0.5% \"non-alcoholic\" class.\n"},{"url":"https://www.niaaa.nih.gov/alcohols-effects-health/what-standard-drink","title":"What Is A Standard Drink?","publisher":"National Institute on Alcohol Abuse and Alcoholism (NIAAA), NIH","source_type":"govt_report","statistic":"A US standard drink contains 14 grams (about 0.6 fluid ounces) of pure ethanol, the amount in 12 oz of regular 5% ABV beer","excerpt":"\"In the United States, one 'standard' drink (or one alcoholic drink equivalent) contains roughly 14 grams of pure alcohol, which is found in: 12 ounces of regular beer, which is usually about 5% alcohol.\"\n","source_date":"2023-01-01","source_accessed":"2026-08-01","calculation_notes":"The 14 g standard-drink anchor used for every dose translation in this entry (0.385% → ~1.08 g ≈ 1/13; 0.009% → ~0.025 g ≈ 1/560; 1% → ~2.8 g ≈ 1/5). Density of ethanol 0.789 g/mL; 12 fl oz = 355 mL. A genuine 0.0% product contributes 0 g and thus no term in this arithmetic — the whole reason the entry is about verification rather than dose.\n"},{"url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC12303755/","title":"Perinatal use of non-alcoholic beverages that mirror alcohol","publisher":"American Journal of Obstetrics & Gynecology (author manuscript, PMC), 2025","source_type":"peer_reviewed","statistic":"'Alcohol-free' products contain no detectable alcohol; 'non-alcoholic' products legally contain up to 0.5% ABV. Providers interpret the ACOG/CDC 'no safe level of alcohol' guidance for these products differently","excerpt":"\"Providers offer varying advice on use of non-alcoholic beverages in the perinatal period, suggesting they interpret the 'no safe level of alcohol' guidance from ACOG and the CDC differently.\"\n","source_date":"2025-03-01","source_accessed":"2026-08-01","calculation_notes":"Confirms the \"alcohol-free = no detectable alcohol\" pole of the taxonomy (against \"non-alcoholic\" ≤0.5%) and situates the question in perinatal care: ACOG and the CDC hold that no amount of alcohol is established as safe in pregnancy. For a genuinely 0.0% product that guidance is not in tension with the dose (there is none); the residual concern it leaves is entirely about whether the 0.0% claim is accurate. No harm rate for verified-zero products exists, consistent with this entry carrying no probability.\n"}],"comparison_anchors":[],"short_label":"0.0% beer while pregnant","outcome_severity":"serious_harm","exposure_pattern":"cumulative","outcome_type":"serious_permanent_harm","valence":"negative","caveats":"The reassurance in this entry is conditional, and the condition is the label. US regulation (TTB, 27 CFR 7.65) defines \"alcohol free\" and a \"0.0%\" claim as meaning no alcohol at all, with no tolerance, which is stricter than the \"non-alcoholic\" class (under 0.5% ABV) covered in the sibling entry [[non-alcoholic-beer-pregnancy]]. A product that genuinely meets the 0.0% definition has no ethanol to reach a fetus or pass into breastmilk, and is the lowest-exposure option in the category. The absence of a harm probability here is not a hedge against that: it reflects that harm at a literally-zero dose is not a meaningful quantity, and that the residual risk is a verification question, not a dose question.\nThat verification question is real. \"0.0%\" is typically derived from density (hydrometry), which is confounded near zero by sugars, dissolved CO2, and other solutes, so the reading can round trace ethanol down to 0.0. Gas chromatography, which measures ethanol directly, has repeatedly found \"alcohol-free\" labels above zero: 0.009–0.385% by volume across the tested beverages (five beers plus grape and apple juices) in a 1980 study, and above 1% in six \"no alcohol\" products in 2010. Both are small, dated samples and neither is a current prevalence, but together they establish that the zero-claim class is the one measurement catches deviating. Two further distinctions matter: \"de-alcoholized\" is a process term that on its own permits up to 0.5% ABV, so it is not synonymous with \"0.0%\"; and brands that verify zero by chromatography (rather than estimating by density) have the strongest claim to being what the label says.\n","quality_score":{"d1":4,"d2":5,"d3":5,"d4":5,"d5":5,"d6":5,"d7":5,"d8":5,"avg":4.88,"scored_by":"claude-code-8d","scored_at":"2026-08-01","methodology_version":"1.2"},"reviewer":"8d-eval-2026-08-01","last_reviewed":"2026-08-08","reviewed":true,"generated_at":"2026-08-01","image":{"alt":"A beer bottle whose label shows a plain '0.0' beside a small magnifying glass on a neutral surface, flat vector illustration."},"canonical_url":"https://likelier.app/zero-alcohol-beer-pregnancy","api_url":"https://likelier.app/api/fears/zero-alcohol-beer-pregnancy.json"}]}