Citation

Myhrmann MS, Janstrup KH, Møller M, Mabit SE. Accid. Anal. Prev. 2020; 149: e105875.

Copyright

(Copyright © 2020, Elsevier Publishing)

DOI

10.1016/j.aap.2020.105875

PMID

33242711

Abstract

The majority of research on bicyclist injury severity relates to bicycle-motor vehicle crashes, even though single-bicycle crashes make up more than half of bicycle crashes. This study explores the factors related to the injury severity outcome of single-bicycle crashes. We use single-bicycle crash data obtained from medical records collected in the period 2010-2015 combined with road maintenance data. The data includes three injury severity categories: 'severe injury', 'slight injury', 'no injury'. The relation between the factors surrounding single-bicycle crashes and the resulting injury severity is estimated using a latent class ordered probit model. The model estimation identifies three latent classes where the likelihood of cyclist membership depends on the bicyclist's age and gender. Furthermore, several factors appear to affect the likelihood of injuries in single-bicycle crashes. These are the road geometry (i.e. if the crash occurred on a bicycle lane or a road section), maintenance level, and the interaction between road geometry and maintenance level. The findings suggest that single-bicycle crashes on road sections result in more severe injuries than single-bicycle crashes on bicycle lanes. The largest effect is seen when a single-bicycle crash occurs on a road section with a poorly maintained bicycle lane being available. Crashes on low volume roads with few bicyclists are also related to an increased probability of severe injury as well as crashes occurring after dark.

Language: en

Keywords

Injury severity; Latent class ordered probit; Medical records; Road safety; Single-bicycle crashes