Critical reexamination reveals inflated estimates of cannabis traffic fatalities

This research commentary critically examines recent estimates claiming that cannabis use causes about 5,000 traffic fatalities annually in the US. The authors identify and correct two significant mathematical errors in the original calculations, demonstrating how estimation methodology directly impacts public health conclusions. When these errors are corrected, the estimated cannabis-attributed fatalities drop substantially to approximately 3,550 deaths per year, highlighting the sensitivity of traffic safety estimates to methodological accuracy.

Beyond mathematical corrections, the authors argue that previous estimates failed to properly account for alcohol co-use, which is critical since drivers using both cannabis and alcohol represent a different population with potentially different risk profiles. When more realistic and defensible values are incorporated—particularly accounting for the prevalence of dual substance use—the estimated cannabis-attributed traffic fatalities falls to below 2,000 annually, and the authors suggest the true number could plausibly be much closer to zero. This commentary underscores an important principle: public health claims about cannabis safety require rigorous scrutiny of underlying assumptions and statistical methods.

The implications are significant for cannabis policy debates, where traffic safety is frequently cited as a major concern. The authors argue that while cannabis use should not be dismissed as completely safe in driving contexts, the current scientific evidence does not reliably support claims of 5,000+ annual deaths. This work emphasizes the need for more rigorous epidemiological research that carefully disentangles the independent effects of cannabis from confounding factors like concurrent alcohol use, driver behavior, and other risk factors.