Korte weergave conclusies:
• Cyclist crashes (including non-injury crashes) are more common than would be suggested by more conventional analyses of police or insurance data and travel surveys
• Crash rates decreased with increasing age category, increased with higher BMI category and were higher for men compared to women.
• Crash rates tended to be highest for participants who disagreed that cycling for travel was comfortable, well regarded in their neighbourhood or common.
• Participants who lived in neighbourhoods with a higher density of bike lanes, higher NDVI or lower building density tended to have lower crash rates.
Zwaktes van studie:
• PASTA participants are more educated and younger than the general population (Gaupp-Berghausen et al., 2019), although recruitment specifically oversampled cyclists, so there may be better representation of the cycling populations.
• Spatially resolved exposure data would allow for further important analyses, such as risks associated with specific route characteristics, but at the beginning of PASTA large scale collection of spatially resolved route data from participants was not feasible due to limitations of available tracking apps at the time. Passive detection of cycling routes through mobile tracking apps (Geurs et al., 2015) could enable the widespread collection of spatially resolved exposure data and more detailed investigation of policy relevant risk factors in future studies.
Een nieuwe toevoeging aan studie over fietsongevallen:
• A new contribution of this research is an inquiry on the association between differences in social environment and crash risk. We found that individual perceptions of social norms around cycling were associated with crash risk, where those who agreed that cycling was a well-regarded mode of transport in their neighborhood were at lower risk for a crash than those that were neutral (1.16 times higher) or disagreed (1.28 times higher risk). The perception question was asked at baseline, so preceded any reported crashes. We suggest this variable may be in part capturing different built environment conditions, where those participants who think cycling is well regarded may live and travel in safer areas for cyclists, within their respective cities. A part of this may also be the safety in numbers effect, where a higher level agreement corresponds to an area with more cyclists due to the presumably more supportive social environment for cycling.