12.7% of Crashes Have a Medical Cause. Only 6 States Ask Doctors to Say Anything.
According to the toxicology reports, there aren't any. When a 74-year-old with uncontrolled diabetes blacks out behind the wheel and crosses the median at 55 mph, nobody runs a tox screen because there's nothing to screen for. NHTSA built its entire crash investigation system around what you drank. It has no protocol for what your body did.
FARS logged 1,513 driver deaths per year tied to medical events behind the wheel from 1995 to 1997.[1] Cardiac episodes: 83%. Alcohol-related crash deaths that same period: 13,436 per year. NHTSA built a billion-dollar enforcement machine around alcohol and the medical column got nothing, even though 1,513 is the floor. Police reports attribute 1 to 2% of crashes to medical causes, but when researchers actually interviewed crash-involved drivers, 12.7% had a direct medical cause.[2] Six-fold underreporting, because nobody asks the right questions at the scene. Drivers over 70 were 11.4% of the study population but 30% of its medically-caused crashes.
The regulatory response: a 2024 JAMA Network Open study found six states mandate physician reporting of medically impaired drivers: California, Delaware, Nevada, New Jersey, Oregon, and Pennsylvania.[3] Four of those only cover "lapses of consciousness." Progressive dementia, uncontrolled arrhythmia, insulin-dependent diabetes with a history of hypoglycemic blackouts? In 44 states, the physician can shrug and refill the prescription. A third of state DMV websites lack instructions for voluntary reporting, only three states accept anonymous reports, and a quarter offer physicians zero legal immunity, so reporting a dangerous driver can trigger a lawsuit from the patient.
When Ontario, Canada, tried mandatory reporting, crashes dropped 41%, saving an estimated $200 million per year.[4] But patients showed a 25% spike in depression visits and a 30% drop in return visits to the reporting physician. A separate 2024 study found mandatory dementia-reporting laws were associated with physicians simply not making the diagnosis.[5] That is a real tradeoff, and it doesn't make the current system any more defensible.
A University of Utah study found drivers with three or more medical conditions crashed at 93 incidents per million miles versus 29 for all drivers.[6] A 3.2x multiplier. A BAC of 0.08 gives you roughly 4x, and we built an entire legal system around it. Three comorbidities get you to 3.2x and we built a suggestion box.
What you can do: If you or a family member has a condition that causes syncope, seizures, or blackouts, check your state's DMV medical reporting page. In the six mandatory states your physician must notify the DMV; in the other 44, that conversation probably hasn't happened, and you should have it. If your state's DMV website has no reporting instructions, that is the problem this article is about. Courts have held physicians liable for crashes their unreported patients caused,[7] so the liability cuts both ways.
Sources & References
- Sheth SG, Krauss G, Seizure-related car crashes, Epilepsia 2001;42(suppl 7):215-216. Analysis of FARS 1995–1997 data. medscape.com
- Vézina C, Crash Rates of Quebec Drivers with Medical Conditions, PMC 2013, citing Lindsey T, 2011 (secondary citation). pmc.ncbi.nlm.nih.gov
- Tran EM, Lee JE, Reporting Requirements, Confidentiality, and Legal Immunity for Physicians Who Report Medically Impaired Drivers, JAMA Network Open 2024. pmc.ncbi.nlm.nih.gov
- Redelmeier DA et al., Ontario physician warning study, 2012. medicalxpress.com
- Mattke S et al., Mandatory dementia reporting and diagnostic patterns, 2024. medicalxpress.com
- Thiese MS et al., Multiple Conditions Increase Preventable Crash Risks among Truck Drivers, J Occup Environ Med 2015. pmc.ncbi.nlm.nih.gov
- NHTSA, Countermeasures That Work: Older Drivers — Licensing Agency Referrals. nhtsa.gov
Source: FARS 1995–1997 via Sheth & Krauss (2001); JAMA Network Open (2024); Lindsey (2011) via Vézina (2013). FARS medical-cause data has not been systematically updated since the late 1990s. The 12.7% figure comes from one rigorous study and may not generalize nationally. Commercial driver risk data (Thiese) may not apply to non-commercial drivers. The 1,513 annual fatality count reflects 1990s-era driving volumes; current figures could differ. See methodology for caveats.