Active living in a dangerous world: Characteristics of the built environment surrounding pedestrian-vehicle collisions

Background: Autocentric design has created communities with limited walkability. Simultaneously, pedestrian fatalities account for a disproportionate number of overall traffic fatalities, comprising approximately 12.1% of all traffic fatalities. Motivations to reduce pedestrian injury are typically concerned with the acute injury inflicted. However, concerns about pedestrian injury can influence walking behavior. Since physical inactivity is associated with numerous chronic health conditions, pedestrian injury may have long-term health consequences. Researching how the built environment can minimize pedestrian injury is essential to creating safe communities and promoting active lifestyles. Study Purpose: To identify characteristics of the built environment associated with pedestrian-vehicle collision (PVC) locations along primary arterials, and to assess whether there are significant differences in area walkability between collision sites and non-collision sites. Methods: Data describing Clark County, Washington PVCs and the built environment were compiled from various data sources using GIS software. Built environment characteristics surrounding PVC locations (n=60) were compared to non-collision locations (n=120) using unique retrospective case-control methods and logistic regression. Results: Walkability of the surrounding environment was correlated with PVC locations, but was not significant in the multivariate model. For every additional 10,000 passengers/month, locations with higher-volume transit stops had 24% greater odds of having a nearby collision (multivariate OR=1.24, 95% CI: 1.06-1.43). Areas with sidewalks, high-traffic volume intersections, and supermarkets were also significantly associated with collision locations. Conclusions: Locations with lower walkability did not have significantly more PVCs after adjustment for other factors. Our findings support previous research indicating that locations near high-volume transit stops have greater odds of PVC, underscoring the importance of maximizing safety surrounding transit stops. Future research should assess these relationships by type of collision, since different factors of the built environment may be associated with different types of collisions. Additionally, future studies should consider adjusting the spatial buffer used to calculate walkability.