Social inequalities in child pedestrian collisions in Toronto, Canada. The role of the built environment

Author(s): Schwartz, Howard, Mitra, Saunders, Fuselli, Macpherson, Cloutier, Rothman

Slidedeck Presentation:

1A Schwartz



In Canada, road injury is a leading cause of death among children and youth. Important inequalities in child pedestrian-motor vehicle collisions (CPMVC) have been observed, with higher rates in low socioeconomic status neighbourhoods. Social inequalities in collision rates in other dimensions such as visible minorities and new immigrant populations remain under-explored in the Canadian context. Reasons for inequalities in CPMVC are not well understood, including whether disparities result from inequitable roadway-built environments, such as fewer traffic interventions and more high-speed roads in socially marginalized areas.


To examine social inequalities in CPMVC and the role of the built environment in Toronto, Canada.


We conducted a cross-sectional study to examine the relationship between area-level social indicators (material deprivation, ethnic concentration, proportion recent immigrants, proportion visible minority) with CPMVC by census tract (CT) (n=530) in Toronto, Canada. Police-reported pedestrian-motor vehicle collisions among children, age 1-17, from 2010-2018 were mapped by CT. Social indicators were derived from the Ontario Marginalization Index and the 2016 Canadian census. Built environment features, derived from the 2016 census and City of Toronto built environment data, were categorized into 3 groups; i) road interventions (crosswalks, crossing guards, speed humps, traffic calming, one-way streets, bike lanes, per road kilometer) ii) road features (proportion living along arterial roads, intersection density, traffic signal density), and iii) land-use features (proportion of multi-dwelling homes (apartments, duplexes), density of old homes, percent residential, commercial, or park land-use). Multivariate negative binomial regression models are being conducted to estimate associations between social indicators and CPMVC per road kilometer, controlling for number of children and built environment features.


A total of 2028 CPMVCs occurred within the study area between 2010-18, with an average of 3.8 collisions per CT (range 0-20). Variation was seen across social indicators, with deprivation and ethnic concentration z-scores (i.e., distance from the mean) ranging from -2.3 - 3.65 and -1.67 - 3.14, respectively, percent new immigrants (mean 6%) ranging from 0-24%, and percent visible minority (mean 48%), ranging from 9-98%. Variability in density of road interventions, road features (e.g., proportion arterial roads), and land uses was seen across Toronto CTs. In univariate analyses, all social indicators were significantly associated with CPMVC, including material deprivation (Incidence Rate Ratio (IRR): 1.37, 95% Confidence Interval (CI): 1.28-1.47), ethnic concentration (IRR: 1.34, 95%CI 1.25-1.44), proportion recent immigrants (IRR: 2.06, 95%CI: 1.74-2.44), and proportion visible minority (IRR: 1.10, 95%CI: 1.07-1.13).


Social inequalities in CPMVC exist across Toronto, particularly in areas with higher proportions of recent immigrants. Subsequent analyses will examine whether accounting for built environment differences attenuate the effects of these area-based inequalities. We found that Toronto includes important socioeconomic and ethnic diversity, and diversity in built environments by CT, providing an ideal setting for analysing inequalities in the built environment and CPMVC.


This study will provide further insights into the relationship between social indicators and CPMVC, and the influence of the built environment. Results from this study can be used to support goals of equitability in achieving Vision Zero goals, promoting road safety for all road users.