PT - JOURNAL ARTICLE AU - Morrison, Christopher AU - Jacoby, Sara AU - Dong, Beidi AU - Delgado, M Kit AU - Wiebe, Douglas TI - 36 An interrupted time-series analysis of ridesharing and motor vehicle crashes in us cities AID - 10.1136/injuryprev-2017-042560.36 DP - 2017 Sep 01 TA - Injury Prevention PG - A14--A14 VI - 23 IP - Suppl 1 4099 - http://injuryprevention.bmj.com/content/23/Suppl_1/A14.1.short 4100 - http://injuryprevention.bmj.com/content/23/Suppl_1/A14.1.full SO - Inj Prev2017 Sep 01; 23 AB - Statement of purpose Ridesharing is a novel technology that connects contracted owner-operator drivers with prospective passengers through a mobile application. The worlds largest ridesharing company, Uber, reports it has booked over 2 billion journeys since operations began in 2010. However, the effect on motor vehicle crashes is unclear. Because the lower convenience and financial costs of ridesharing compared to drunk driving could reduce the incidence of alcohol-involved crashes, we hypothesised that alcohol-involved crash rates would be lower in cities with ridesharing.Methods/approach We identified four US cities where Uber launched, ceased, and then resumed operations (Las Vegas, Portland, Reno, and San Antonio). Using crash data from state Departments of Transportation, we constructed two weekly time-series (2013 to 2016) for each city: counts of all injury crashes, and the proportion of injury crashes that were alcohol involved. Using interrupted time series models, we tested Ubers launch, cessation, and resumption as interruptions.Results The majority of changes to Uber operations were unrelated to all injury crashes or alcohol-involved injury crashes, though some interruptions supported our hypotheses. For example, in Portland, the resumption of Uber operations coincided with a 61.8% decrease in alcohol-involved crashes, equating to 3.5 fewer alcohol-involved crashes per week. We did not detect a concomitant change in injury crashes in Portland at that time.Conclusions Ridesharing may reduce alcohol-involved crashes in some cities, but may not affect overall injury crash incidence. Relationships differ between geographic locations and within geographic locations over time, and may depend on specific local characteristics.Significance and Contributions to Injury and Violence Prevention Science Motor vehicle crashes are the leading cause of death for people aged 13 to 25, and alcohol-involved crashes account for around one third of all motor vehicle deaths. Ridesharing may be an effective intervention to reduce alcohol-involved crashes in some cities.