Pedaling towards safer streets: evaluating the impact of cycling infrastructure on road safety
研究了加拿大三个城市自行车道（分离式车道和划线车道）对骑行伤亡率的影响，发现分离式车道能降低伤亡风险，而划线车道效果不一。
• COVID-19 boosted cycling and infrastructure in Canada, but long-term trends remain unclear. • Cyclist volumes rose with new cycling infrastructure, especially on cycle tracks. • Impact of cycling infrastructure on killed or seriously injured (KSI) per 10km varied by city and infrastructure type. • Vancouver’s painted lanes were associated with higher cyclist KSI; Calgary’s cycle tracks with fewer KSI. • KSIs did not rise despite more riders, especially on cycle tracks, suggesting reduced injury risk per cyclist. Cycling provides health and environmental benefits but poses safety risks. Over the past decade, Canadian cities have expanded cycling infrastructure, such as cycle tracks (physically separated from motor vehicles) and painted lanes, to support safer active transportation. This study evaluates the impact of cycling infrastructure implemented in Toronto, Calgary, and Vancouver. We used infrastructure data (2011–2022) from municipal reports, validated with street view imagery. The primary outcome was killed or seriously injured (KSI) victims per 10 km of road, based on police-reported data (2009–2023). A difference-in-differences design, adapted for gradual infrastructure implementation, was applied using road segments as the unit of analysis. Subgroup analyses examined cyclist KSI and infrastructure type. Crude data on seasonal ridership from intersection counters on a subset of road segments (9 %) were used to estimate post-installation changes. Results varied by city and infrastructure type. In Toronto, no clear associations were found. In Calgary, new cycle tracks were associated with 2.1 fewer cyclist KSIs per 10 km (95 % CI: −4.5 to 0.3). In Vancouver, painted lanes were associated with 4.7 more KSIs per 10 km (95 % CI: 0 to 9.3), and one additional cyclist KSI per 10 km (95 % CI: 0 to 2.1). After implementation, cycle tracks showed substantial increases in cyclist volumes in Toronto (76 %), Calgary (207 %), and Vancouver (732 %), whereas painted lanes exhibited more modest changes. Although low KSI counts introduced uncertainty, cycle tracks were associated with no change or reductions in KSIs, while painted lanes showed no change or slight increases.