Driving behaviours near elementary schools and child pedestrian-motor vehicle collisions

Author(s): Linda Rothman, Alison MacPherson

Slidedeck Presentation Only (no paper submitted):

2C - MacPherson


The promotion of walking to school has become a priority in many urban centres in order to improve health, foster community relationships and to lessen traffic congestion. Dangerous driving behaviours near schools have not been well described and there has been no investigation into the relationship with child pedestrian-motor vehicle collision (PMVC) outcomes.
AIMS - To determine the incidence of dangerous driving behaviours observed during morning school drop off and their association with police-reported child pedestrian-motor vehicle collision (PMVC) rates near elementary schools in Toronto, Canada.

Police-reported child pedestrian collisions (ages 4-12) from 2000-2011 during school travel times were mapped within 200 m of 118 schools along with features of the built environment. Dangerous driving behaviours during morning drop off were measured by trained observers and included; double parking, cars waiting in front of school blocking the vision of other motorists and pedestrians and cars dropping off children on the opposite side of the road to the school. Data was collected regarding the number of children observed walking to school, speed limits in front of the school, jaywalking, traffic congestion and the presence of school crossing guards. A composite measure of children’s socioeconomic status (SES) attending the school was obtained from the Toronto District School Board. A multivariate poisson regression was used to model the rates of PMVCs per number of children walking, adjusting for the built environment, SES and traffic conditions.

There were 45 child PMVCs with a mean age of 8.3 (SD 2.02). Sixteen children (36%) had minimal injuries and 29 (64%) had minor injuries resulting in an emergency department visit. The greatest proportion of collisions occurred on collector roads (18, 40%) and local roads (16, 36%), with 10 collisions on major or minor arterials (22%). The mean collision rate was .31/1000/year (SD = 7.0) and the mean number of children walking to school was 137 (SD =82.7). Dangerous driving behaviours were observed in 104 schools (88%), with two or more dangerous driving behaviours observed in 75 schools (63%). In the multivariate analysis, each additional dangerous driving behaviour was associated with a 42% increase in collision rates (IRR = 1.42, 95% CI 1.01, 1.99) and each additional 100 meters of major or minor arterial roads were associated with a 30% increase in collision rates (IRR = 1.20, CI 1.16, 1.46).

Dangerous driving behaviours were associated with non-fatal child PMVC rates during school travel times in the area immediate surrounding elementary schools after controlling for the presence of higher speed roads. The association between lower socioeconomic status and PMVCS was no longer significant after adjusting for driver behaviour and higher speed roadways.

Other dangerous driving behaviours near schools and their relationship to child PMVCs need to be examined such as failure to stop at stop signs. Limitations included the small number of collisions, however; the results emphasize the importance of effective interventions to address dangerous driving behaviour and the need to lower traffic speeds near schools.