Examining the impact of bicycle lanes on cyclist-motor vehicle collisions in the city of Toronto: a quasi-experimental study.

Author(s): Deepit Bhatia, Sarah Richmond, Jennifer Loo, Linda Rothman, Colin Macarthur, Andrew Howard

Slidedeck Presentation Only (no paper submitted):

5C - Bhatia


Urban bicycling has been increasing in popularity in Canada over the last two decades, leading to a corresponding increase in bicycling infrastructure, such as painted cycle lanes and cycle tracks. Despite the construction of over 300km of cycle lanes in the City of Toronto, Canada in the last 20 years, there is a paucity of evidence regarding the safety of these lanes. Unlike separated cycle tracks, painted cycle lanes are not physically separated from motor-vehicle traffic.

The primary objective of this study was to examine the effect of the installation of painted cycle lanes on the incidence rate of cyclist motor vehicle collisions (CMVC) in high-volume downtown cycle lanes in Toronto. A secondary objective was to examine changes in injury severity with the installation of cycle lanes.

A quasi-experimental design was used to evaluate the frequency of CMVC collisions pre- and post-installation of 7 cycle lanes in Toronto. Study data were obtained from Toronto Police Service reports for collisions occurring between 1991 and 2010. A zero-inflated Poisson regression model was used to estimate the rate of collisions. Changes in collision frequency pre- to post-installation of cycle lanes were reported using incidence rate ratios (IRR) with 95% confidence intervals (CI)

Over the study period (January 1, 1991 to December 31, 2010), a total of 23,959 CMVC were reported in Toronto. Of these, 329 occurred on the 7 lane segments included in the analysis. There was a non-significant 19% reduction in the rate of collisions per segment-month (IRR = 0.81, 95% CI: 0.65, 1.01). Point estimates suggest a decrease in the frequency of collisions that resulted in minimal/minor injuries (IRR = 0.84, 95% CI: 0.59, 1.20), and in major/fatal injuries (IRR = 0.72, 95% CI: 0.51, 1.01).

The implementation of painted cycle lanes in Toronto demonstrated a modest effect in reducing the frequency of CMVC. The primary limitation was the lack of available volume data for the lane segments, but census and City of Toronto data suggest an overall 30% increase in the number of commuter cyclists in Toronto between 1996 and 2001. Thus, our estimates, and the protective effect of cycle lanes were conservative.

Cycle lanes lead to modest decreases in CMVC in high-volume, downtown streets. Further research is required, incorporating cycling volumes in order to calculate a more accurate effect of cycle lanes. Further research is also needed to examine changes in how CMVCs occurred (for example, changes in right hooks or rear-ending collisions).

Deepit Bhatia1, 4, Sarah Richmond1, Jennifer Loo3, Linda Rothman2, Devon Williams1, Colin Macarthur5, Andrew Howard1
1: Child Health Evaluative Sciences, The Hospital for Sick Children, Toronto.
2: York University, Toronto.
3: Public Health and Preventive Medicine, University of Toronto, Toronto.
4: Dalla Lana School of Public Health, University of Toronto, Toronto.
5. Department of Pediatrics, University of Toronto, Toronto