Determinants and patterns of truck-involved crash injury severities in Hong Kong: Insights from latent class and multinomial logit models
Trucks are crucial to the supply chain and logistics efficiency of Hong Kong, so understanding the injury outcomes of truck-involved crashes matters for both economic activity and road safety. Using 11,045 casualty records describing 6969 truck-involved crashes in Hong Kong from 2022 to 2023, this study applies the latent class logit model to examine the determinants influencing injury outcomes in truck collisions and identify their heterogeneous crash patterns. Driver, vehicle, environmental, road and casualty characteristics are examined as determinants of minor, serious and fatal injury outcomes, and the multinomial logit, random parameters logit and latent class logit models are estimated on identical data. The two-class latent class model was interpreted jointly with the multinomial logit benchmark. In both models, pedestrian collisions and casualties over 50 were associated with higher serious- and fatal-injury probabilities, darkness and multi-vehicle movement with a higher probability of fatal injury, and middle-aged drivers with a lower one; light goods vehicles with a higher probability of minor injury, and junctions with a lower probability of serious but not fatal injury. The findings could serve as an evidence-based reference for transportation administrators and industries to formulate effective safety strategies to better protect road users from truck-involved crashes.
Authors
- Qiaoqiao Ren (ORCID: https://orcid.org/0000-0002-2482-2856)
- Yuhe Zhang
- Min Xu
Institutions
- Hong Kong Polytechnic University (HK)
- Zhuhai College of Science and Technology (CN)
Publication Details
- Journal
- Journal of Transportation Safety & Security
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1080/19439962.2026.2761867
- Primary Topic
- Traffic and Road Safety
- Type
- article
- Field-Weighted Citation Impact
- 0.00