Population-level estimate of bicycle use and fatality risk in a data-poor setting.

Abstract

Lack of data on exposure for walking and cycling poses a significant barrier to understanding the injury risk of these road users. Though this data paucity is most prevalent across low-and-middle-income countries (LMICs), it remains a challenge in many high-income countries as well. A new and simple method has been proposed to estimate population-level cycling distance travelled, with New Delhi, India as a case study. I used two independent estimates to calculate this distance. First, a ratio of motorcycle volume counts to that of cycle volume counts across major roads, and second, the total annual distance travelled by motorcycles. I validate this method using data from London, where cycling distance estimates are available from city-wide traffic volume counts as well as household travel survey. Combining the distance estimates with annual fatalities of corresponding road users, I found that cyclists have about 2 times greater fatality risk per kilometre than motorcycle occupants and about 40 times greater risk than car occupants. To encourage greater use of cycling, there is an urgent need to narrow this gap between the safety of cyclists and that of car occupants. The proposed method can be used to monitor cycling usage and its risk for many settings where traffic surveillance systems do not exist.