Odds of culpability associated with use of impairing drugs in injured drivers in Victoria, Australia

•Culpability analysis was conducted on 5000 injured drivers in whom drug testing was conducted.•Alcohol gave large increases in the odds of culpability that was concentration-dependent.•Methylamphetamine also showed a large increase in the odds of culpability.•THC gave a modest increase in the odds of culpability•Benzodiazepines also gave a modest increase in the odds of culpability. Culpability analysis was conducted on 5000 drivers injured as a result of a vehicular collision and in whom comprehensive toxicology testing in blood was conducted. The sample included 1000 drivers for each of 5 years from approximately 5000–6000 drivers injured and taken to hospital in the State of Victoria. Logistic regression was used to investigate differences in the odds of culpability associated with alcohol and drug use and other selected crash attributes using the drug-free driver as the reference group. Adjusted odds ratios were obtained from multivariable logistic regression models in which other potentially explanatory driver and crash attributes were included. Drivers with alcohol present showed large increases in the odds of culpability similar to that seen in other studies investigating associations between blood alcohol concentration and crash risk. Methylamphetamine also showed a large increase in the odds of culpability (OR 19) compared to the reference group at both below and above 0.1 mg/L, whereas those drivers with Δ9-tetrahydrocannabinol (THC) present showed only modest increase in odds when all concentrations were assessed (OR 1.9, 95 %CI 1.2–3.1). Benzodiazepines in drivers also gave an increase in odds (3.2, 95 %CI 1.6–6.1), but not other medicinal drugs such as antidepressants, antipsychotics and opioids. Drivers that had combinations of impairing drugs generally gave a large increase in odds, particularly combinations of alcohol with THC or benzodiazepines, and those drivers using both THC and methamphetamine.