Naturalistic Driving Study on Cannabis Use in Washington and Virginia

Abstract

This study examined the consumption behavior of cannabis users and its influence on driving performance through a comprehensive naturalistic driving study (NDS) conducted in Washington and Virginia. The study aimed to address gaps in research by leveraging real-world data to evaluate how cannabis consumption impacts driver behavior, safety-critical events, and crash risk. The study’s objectives included assessing the prevalence of driving under the influence of cannabis (DUIC), examining variations in impairment across different consumption methods and doses, and exploring the relationships between self-reported intoxication levels and objective performance metrics. Participants were selected based on their regular cannabis use and self-reported DUIC history. Data was collected via in-vehicle instrumentation, a smartphone-based journal app, breathalyzer readings, and oral fluid tests. The study offered key insights into the impact of self-reported cannabis consumption on driving behavior, with cannabis trips occurring alongside sober trips with similar frequency and temporal distributions. Self-reported substance use data revealed that cannabis consumption methods differed between regions, with dabs being the preferred form in Washington and smoking cannabis flower (e.g., joints, pipes, bowls) dominating in Virginia. Polysubstance use, particularly with alcohol, was prevalent, with 13.7% of Washington and 20.1% of Virginia journal entries involving multiple substances. Breathalyzer data showed that 20% of Washington’s and 14.5% of Virginia’s alcohol-positive trips exceeded the 0.08% blood alcohol concentration (BAC) limit. Quantisal oral fluid tests highlighted variations in tetrahydrocannabinol (THC) levels, with mean delta-9 THC levels significantly higher in Washington (1,662 ng/ml) compared to Virginia (260.9 ng/ml). While 85% of Quantisal tests were successfully submitted, challenges such as outlier THC readings due to participant noncompliance with testing protocols were noted, indicating the complexity of linking subjective impairment levels to objective performance metrics. The findings highlight the complexity of DUIC and the need for further research to inform public policy, law enforcement practices, and safety guidelines. The dataset provides a valuable resource for understanding cannabis-related driving risks and developing targeted interventions. Further analyses should explore the nuanced effects of cannabis potency, user tolerance, and polysubstance interactions on driving performance. Enhanced data collection techniques could improve the reliability of future studies.