Research Papers

Modeling Associations between Pedestrian Injury Severity and Road Infrastructure

Version 1
Date added June 18, 2019
Downloaded 0 times/fois
Category 2019 CARSP XXIX Calgary
Tags Research and Evaluation, Session 6C
Author/Auteur Pascua, Dubois, Pernia, Bedard
Stream/Volet Research and Evaluation

Slidedeck Presentation Only:



Background/Context: Approaches to promote transportation safety have been increasingly gaining popularity over the past couple of decades. As such, numerous North American transportation agencies have been prioritizing safety of vulnerable road users such as pedestrians to mitigate traffic-related injury frequency and severity. This relatively recent shift in road safety culture has brought about many significant changes to the built environment, specifically to the transportation infrastructure used by many. Understanding the fundamental relationship between infrastructure and pedestrian injury severity is key to accomplishing such a task.

Aims/Objectives: On the basis of pedestrian crash frequencies, this study investigates the relationship between pedestrian injury severity and transportation infrastructure, among other potential contributors, at intersections and midblock segments on a nationwide scale.

Methods/Targets: Due to the intrinsic ordered nature of injury severity levels, an ordinal probit modeling approach was proposed to examine factor significance regarding pedestrian injury severity. Pedestrian injury data from 2011 to 2015 were extracted from the United States' General Estimates System (GES), which contains nationally-representative estimates of traffic safety data. Pedestrian injury severity was trichotomized into the following categories: no/possible injuries, minor injuries and severe injuries. Several explanatory variables were derived from GES literature and categorized into one of four groups: roadway/environmental factors, pedestrian characteristics, driver maneuver and vehicle type.

Results/Activities: To examine potential causative variable differences in different types of roadway environments, the GES pedestrian dataset was partitioned into two subsets, each focusing on crashes that have occurred on or near an intersection or a midblock setting, respectively. Marginal effects for significant explanatory variables will be computed to quantitatively estimate their influence on pedestrian injury severity. One model output per roadway environment (i.e., intersections and midblock segments) will also be displayed. Results and discussion will be emphasized towards roadway infrastructure and pedestrian characteristics.

Discussion/Deliverables: The results from the two model outputs will be interpreted with an emphasis toward roadway infrastructure and pedestrian characteristics. Several high-level countermeasures for each roadway feature will be proposed. In addition, several research limitations will be highlighted.

Conclusions: Identifying critical infrastructural features at a variety of roadway environments that contribute to increased likelihood of severe pedestrian injuries is an essential step in developing countermeasures to ultimately reduce traffic injury rates, as per Vision Zero doctrine. While substantial research efforts have been made to understand the relationship between various forms of road infrastructure and crash risk, much of these efforts have focused on motorist safety. The findings of this research are intended to assist transportation professionals with determining locations requiring safety interventions.