Fatal-crashes involving cyclists (1985-2009): Crash-, cyclist-, and driver-related factors associated with crash responsibility.

Author(s): Erica Sawula, Sacha Dubois, Bruce Weaver, Michel Bédard

Slidedeck Presentation Only (no paper submitted):

5C - Dubois


Background/Aims: Cycling is a healthy, environmentally friendly, mode of transportation. However, in 2012, 726 cyclists were killed and an additional 49,000 cyclists were injured in motor vehicle traffic crashes in the United States (National Highway Traffic Safety Administration, 2014). Given the risk of cyclist harm, we identified crash- and cyclist-related factors associated with the odds of contributing to cyclist-related crashes.
Methods: United States fatal crashes (1985-2009) involving one cyclist and one car were selected from the Fatality Analysis Reporting System (FARS). Utilizing unsafe driver actions (UDAs), we classified bicycles and drivers by crash responsibility scenario (both, cyclist, driver, or neither responsible). We examined the odds of key crash- and cyclist/driver-related factors by responsibility scenario using logistic regression. Age was measured in decades and centered on two decades. Blood alcohol content (BAC) was scaled such that .05 mg/L was equal to one unit.
Results: Between 1985-2009 there were 5510, 2872, 944, 4669 paired-crashes where the cyclist, driver, both, or neither committed an UDA (i.e., were responsible). For crashes where both cyclist and driver were responsible, the most common UDAs committed by the cyclist were: failure to yield right-of-way (34.5%), operating without required equipment (19.9%), and failure to obey traffic signs (12.5%); and for drivers were: speeding (51.4%), erratic operation (14.8%), and failure to keep in the proper lane (9.6%). Drivers who were younger (Age OR: 0.66; 95%CI: 0.58, 0.74), and male (OR: 1.44; 95%CI: 1.21,1.70), increased the odds of both driver and cyclist being considered at fault compared to the neither responsible scenario. It should be noted that for drivers the effect of age was curvilinear (Age2 OR: 1.05; 95%CI: 1.02, 1.07). Further, wet surfaces (OR: 1.54; 95%CI: 1.02, 2.33) and rural areas (OR: 1.24; 95%CI: 1.04, 1.48) increased the odds of the both driver and cyclist being responsible for the crash relative to neither being responsible. Cyclist sex, age, and whether the crash occurred on either a curved road, intersection, on the road-way, on a high-speed limit roadway, and during daylight did not significantly result in differing odds between the both- and neither-responsible scenarios. In a sub-analysis of those crashes where both car and cyclist were both tested for drugs and alcohol, driver BAC increased the odds (BAC OR: 2.01; 95% CI: 1.36, 2.97) of a crash where both driver and cyclist were considered at fault. However this effect diminished as BAC increased (BAC2 OR: 0.85; 95% CI: 0.77, 0.94). Results for when the cyclist or driver was solely responsible compared to when neither was responsible will also be presented.
Discussion/Conclusions: Cyclist, driver, and environmental conditions contributed to increased odds of crashes where both cyclist and driver or either were considered responsible. Therefore, efforts to prevent cyclist-related crashes should focus on measures to improve the cycling environment and also those measures aimed at controlling cyclist- and driver- related factors (including information, legislation, and education).