Systematic review of the effects of cannabis retail outlets on traffic collisions, fatalities and other traffic-related outcomes

Abstract Aims In recent years, many jurisdictions have legalized medicinal and/or recreational cannabis. Yet, limited information is available on the effects of cannabis commercialization (provision of legal supply of cannabis to adults through licensed retail outlets) on adverse outcomes, such as traffic fatalities. The purpose of this study was to systematically review the literature on the effects of cannabis retail outlets on traffic collisions, fatalities and other traffic-related outcomes. Methods Eight electronic bibliographic databases were searched (MEDLINE, PUBMED, PsychINFO, CINAHL, EMBASE, Scopus, Dissertation Abstracts, Google Scholar) to identify documents of original research on effects of cannabis retail outlets on traffic-related outcomes, including hand searches and a canvass of all members of the International Council on Alcohol, Drugs and Traffic Safety. The guidelines and rating criteria by the Scottish Intercollegiate Guidelines Network informed the review. Results 190 documents were identified, of which nine met all inclusion/exclusion criteria. All nine studies were conducted in the US. Six studies showed some increase in adverse traffic-related outcomes for states with legalized retail outlets compared to states with no legalized cannabis outlets. Three studies showed no overall effect, although one study showed a significant reduction in collision fatalities for those 15–44 years of age. Studies using Difference-in-Difference designs, total collision fatalities rates by billion vehicle miles travelled or by cannabis positive results found positive outcomes. Conclusions This systematic review indicates that two thirds of reviewed studies found some positive association between cannabis commercialization and fatality outcomes. However, the few studies available and the quality grading for the studies suggest only a moderate probability of causal attribution. Future studies need move beyond examining only fatalities or assessing mere presence of cannabis; reliable measures indicating driver impairment at time of collision are needed to establish better causal attribution and to provide greater utility for policy development and implementation.