Major urban road characteristics and injured pedestrians: A representative survey of intersections in Montréal, Quebec.

OBJECTIVES In urban settings, pedestrian fatalities and injuries are concentrated on major roads. This study aims to describe urban intersections with major roads (arterials and collector roads) and explore the association between intersection characteristics and injured pedestrians. METHODS From a stratified random sampling in Montréal, Quebec, 512 intersections were selected and their characteristics collected. The number of injured pedestrians from 1999 to 2008 was obtained from ambulance services. Binomial negative regression models (including IRR: incidence rate ratios) were calculated to determine associations between intersection characteristics and injured pedestrians: i) at all intersections; ii) at intersections with multi-lane roads and iii) at signalized intersections with available vehicle and pedestrian counts. RESULTS Major intersections had more traffic lanes (3.8 vs. 1.7, p < 0.01) and longer pedestrian crossings (18.8 m vs. 12.7 m, p < 0.01) than minor intersections. Bus stops were also more frequent at these intersections (75% vs. 6%, p < 0.01). Overall, each additional traffic lane was associated with a 75% increase in the number of injured pedestrians (IRR = 1.75; 95% confidence interval [CI] = 1.41-2.18). At intersections with multi-lane roads, a fourth branch (IRR = 2.40; 95% CI = 1.53-3.77), vehicles parked within 5 m of the intersection (IRR = 2.40; 95% CI = 1.64-3.51), and marked crosswalks (IRR = 1.79; 95% CI = 1.08-2.95) significantly increased the number of injured pedestrians. Raised medians had no significant protective effect. CONCLUSIONS The results show that besides traffic and pedestrian volumes, intersection characteristics contribute to pedestrian injuries. The reduction of traffic lanes, parking prohibition near intersections and implementation of appropriate pedestrian refuge areas would improve pedestrian safety.