Impact of psychoactive substances and attention-related impairments on driving performance: Sex differences and road...

Crashes are a global public health concern, with psychoactive substance use and attention impairments potentially increasing risk and affecting male and female drivers differently. This study aims to identify factors related to psychoactive substance use and attention impairments that predict road crash involvement by sex, offering insights for targeted prevention effort.

This observational study examined individuals assessed for driving license reissuance after driving under the influence of alcohol and/or drugs. Data collected included personal and sociodemographic details, results from a Continuous performance test (CPT-3) (a neuropsychological test assessing inattentiveness, impulsivity, sustained attention, and vigilance), and toxicological analyses of biological samples. Statistical analyses (chi-square, logistic regression, and ROC curve) aimed at identifing sex-specific predictors of road crash involvement.

The study included 169 participants (82.8 % males, 17.2 % females). Initial univariate analyses indicated sex differences that were not confirmed after Bonferroni correction for multiple comparisons. Logistic regression found blood alcohol concentration predictive of crashes in males (p = 0.005, OR = 2.52), while inattentiveness was a major factor for females (p = 0.045, OR = 12.0). In the combined model, sex itself was not an independent predictor after adjusting for these factors. The model showed moderate accuracy for males (ROC-AUC = 0.695) and higher accuracy for females (ROC-AUC = 0.816), suggesting that inattentiveness may play a particularly important role in predicting crashes in the female subgroup.

Inattentiveness and alcohol use emerged as key predictors of crash involvement, with inattentiveness showing particular relevance in the female subgroup. Prevention should focus on modifiable risk factors rather than sex. The model's higher accuracy in females suggests attentional measures may help identify at-risk drivers, though the small female sample limits generalizability.