Pedestrian Injury Severity Levels in the Halifax Regional Municipality, Nova Scotia, Canada

Abstract

Pedestrians are particularly vulnerable road users within the urban environment. Many studies have examined the factors that contribute to the frequency and severity of collisions, but limited research has examined the influence of the built environment on pedestrian injury severity. This study used the Halifax regional municipality in Nova Scotia, Canada, as a case study to examine the effect of the built environment on the injury severity of pedestrians. Two ordered response models were used: a conventionally ordered probit model and a hierarchical ordered probit (HOPIT) model, which accommodated unobserved heterogeneity because the thresholds could vary across observations. In the HOPIT model fit in this study, the threshold covariates varied with whether the collision occurred at an intersection and with the number of walking commuters in the neighborhood. Built environment contributing factors, including a variety of street pattern classifications, land use types, transit supply, and demographic characteristics, were examined with other variables (e.g., pedestrian and driver characteristics, collision characteristics, environmental conditions). Nova Scotia Collision Record Database data were used for the years 2007 to 2011 to develop the ordered response models of the injury severity of pedestrians. The study found personal and collision characteristics to be significant to explain the injury severity outcomes of pedestrians. In addition, environmental characteristics (e.g., land use type, presence of activity centers, demographic attributes) were found to influence injury severity outcomes. The study results may help inform policy development to improve pedestrian safety in Nova Scotia.