Evaluating the impact of bicycle infrastructure on bicycle-motor vehicle crashes in Iowa

Background/Purpose: Each year bicycle-motor vehicle crashes result in more than 600 deaths and 50,000 injuries, demonstrating the public health importance. The purpose of this study is to determine whether bicycle-specific pavement markings (bicycle lanes and shared lane arrows) and signage are protective against crashes and if there are differences between these facility types. Methods: We used a novel case-site control-site study design that combines current traffic volume data with retrospective crash data from the Iowa DOT, 2007-2010. We randomly selected 147 crash-intersections and 147 control-intersections. Control-sites were matched to case-sites on neighborhood and road classification. Traffic volume was collected during 30 minute observation periods on the same day of week and time of day the crash occurred. Descriptive statistics and conditional regression analyses were conducted to predict the impact of bicycle facilities on crashes. Results/Outcomes: 11.56% of case-sites and 14.96% of controls had a bicycle facility present. Case intersections had higher bicycle volume (3.52 vs 3.34 per 30min) and motor vehicle volume (248.77 vs 205.76 per 30min) than controls. After adjusting for bicycle volume, motor vehicle volume, street width, and sidewalks, results suggest that compared to no bicycle facility, pavement markings were more protective against crashes (OR=0.42, 95%CI:0.10-1.83) than signage (OR= 0.75, 95% CI: 0.18-3.13). Conclusions: These results can be used by city traffic engineers and planners to support the installation of on-road bicycle facilities. Future examination of a larger sample would provide the power needed to further evaluate differences of impact between the various types of on-road bicycle facilities.