Risk of motor vehicle collision associated with cannabis and alcohol use among patients presenting for emergency care
Esther K. Choo,Stacy A. Trent,Daniel K. Nishijima,Angela Eichelberger,Steve Kazmierczak,Yu Ye,Karen J. Brasel,Ariane Audett,Cheryl J Cherpitel
DOI: https://doi.org/10.1016/j.aap.2024.107459
IF: 5.9
2024-01-27
Accident Analysis & Prevention
Abstract:Background The objective of this study was to examine the relationship between cannabis and alcohol use and occurrence of motor vehicle collision (MVC) among patients in the emergency department (ED). Methods This was a cross-sectional study of visits to EDs in Denver, CO, Portland, OR, and Sacramento, CA by drivers who were involved in MVCs and presented with injuries (cases) and non-injured drivers (controls) who presented for medical care. We obtained blood samples and measured delta-9-THC and its metabolites. Alcohol levels were determined by breathalyzer or samples taken in the course of clinical care. Participants completed a research-assistant-administered interview consisting of questions about drug and alcohol use prior to their visit, context of use, and past-year drug and alcohol use. Multiple logistic regression was used to estimate the association between MVC and cannabis/alcohol use, adjusted for demographic characteristics. We then stratified participants based on levels of cannabis use and calculated the odds of MVC across these levels, first using self-report and then using blood levels for delta-9-THC in separate models. We conducted a case-crossover analysis, using 7-day look-back data to allow each participant to serve as their own control. Sensitivity analyses examined the influence of usual use patterns and driving in a closed (car, truck, van) versus open (motorcycle, motorbike, all-terrain vehicle) vehicle. Results Cannabis alone was not associated with higher odds of MVC, while acute alcohol use alone, and combined use of alcohol and cannabis were both independently associated with higher odds of MVC. Stratifying by level of self-reported or measured cannabis use, higher levels were not associated with higher odds for MVC, with or without co-use of alcohol; in fact, high self-reported acute cannabis use was associated with lower odds of MVC (odds ratio [OR] 0.18, 95% confidence interval [CI] 0.05–0.65). In the case-crossover analysis, alcohol use alone or in combination with cannabis was associated with higher odds of MVC, while cannabis use alone was again associated with decreased odds of MVC. Conclusions Alcohol use alone or in conjunction with cannabis was consistently associated with higer odds for MVC. However, the relationship between measured levels of cannabis and MVC was not as clear. Emphasis on actual driving behaviors and clinical signs of intoxication to determine driving under the influence has the strongest rationale.
public, environmental & occupational health,transportation,ergonomics,social sciences, interdisciplinary