Safety effects of cycle track implementation on cyclist-motor vehicle collisions in Toronto, Canada
Author(s): Ling, Rothman, Cloutier, Howard
Slidedeck Presentation Only:
Abstract:
Background/Context: There has been a rise in cycling and cycling infrastructure in Toronto, Canada. A previous study found that there was a non-statistically significant reduction in cyclist-motor vehicle collisions following the implementation of painted bike lanes in Toronto. Cycle tracks, which are painted bike lanes with a physical separation from traffic, may offer additional protection for cyclists; however, the evidence on the safety of cycle tracks are limited and mixed. Furthermore, area-wide safety effects in streets surrounding cycle tracks have not yet been studied.
Aims/Objectives: To examine the safety effects of cycle tracks on the rate of cyclist-motor vehicle collisions on the tracks as well as the area-wide safety effects on peripheral streets surrounding the tracks in Toronto.
Methods/Targets: Cycle track locations and implementation dates were obtained from the City of Toronto. All police-reported cyclist-motor vehicle collisions from 2000-2016 were provided from the Toronto Police Service. Analyses were restricted to 2 years pre- and post-track implementation. Collision rates per segment-month were calculated for each track. Streets between distances of 26m to 550 m in 100m increments of the tracks were evaluated for area-wide effects. Collision rates per segment-month for each area were calculated. Zero-inflated Poisson models were used to evaluate the safety effects of cycle tracks and at varied distances on the peripheral streets, controlling for season of collision and baseline collision rates.
Results/Activities: We examined six cycle tracks implemented in 2013-2014, spanning 9.9 km in total length. Among 190 collisions observed on tracks, the rate of collision significantly increased more than 2-fold following their implementation (Incidence Rate Ratio (IRR) = 2.06, 95% CI: 1.51-2.81, p<0.001). There were no significant effects of cycle tracks among the 191 collisions on streets within 26-150m. However, significant differences before and after track implementation were detected on streets between 151-250m, 251-350m, 351-450m and 451-550m to the tracks. Analyses were subsequently pooled for all 602 collisions on streets at 151-550m distances and following track implementation, a significant reduction of 35% in collision rates was observed (IRR= 0.65, 95% CI: 0.54 -0.76, p<0.0001).
Discussion/Deliverables: Although collision rates have increased on cycle tracks following its implementation, this may be attributed to increased cycling volume over the years and changes in cycling routes, as cyclists' may prefer to ride on cycle tracks versus painted bike lanes or street. Cycling volume data is not available on all tracks, so it is difficult to estimate changes in exposure. For area-wide effects, there was no change seen on streets closest to cycle tracks since these may be frequent access/exit points from the tracks. However, a safety halo effect appeared on streets at further distances of cycle tracks which may be a result of changes in cycling routes and shifting preferences to ride on newly implemented cycle tracks.
Conclusions: Overall, benefits for cycle tracks can be seen in surrounding areas of the cycle tracks. However, safety enhancements on cycle tracks itself are needed to protect and accommodate increase numbers of cyclists. These findings will help to inform current research on cycling infrastructure in Toronto and cyclist safety.