Determinants of active accident involvement among motorcycle couriers in Türkiye: a cross-sectional study

BACKGROUND: Motorcycle couriers are increasingly central to urban delivery systems, yet they experience disproportionately high rates of traffic-related injuries. Existing research often attributes crash involvement primarily to individual riding behaviours, overlooking the potential influence of employment models, environmental exposure, and physical strain. Evidence distinguishing rider-related errors from systemic working conditions, particularly for active accidents in which the rider is at fault, remains limited. METHODS: This cross-sectional study was conducted between June and December 2025 among 372 motorcycle couriers actively working in Ankara, Türkiye. Participants were recruited using a snowball sampling strategy through face-to-face and online networks. Data were collected via an electronic questionnaire capturing sociodemographic characteristics, employment model, working conditions, accident history, psychosocial status (Depression Anxiety Stress Scale-21), risky riding behaviours (Motorcycle risk-taking behaviour), and musculoskeletal symptoms. Self-reported active accident involvement (defined as at-fault crashes) within the previous 12 months was the primary outcome. Associations were examined using bivariate analyses and multivariable binary logistic regression. RESULTS: Overall, 79.6% of participants reported at least one active accident in the previous 12 months. Compared with freelance couriers, company-employed (OR = 7.066; 95% CI: 2.104 to 23.727) and self-employed couriers (OR = 2.783; 95% CI: 1.109 to 6.984) had significantly higher odds of active accident involvement. Non-collision falls showed the strongest association with active accident involvement, with four or more falls associated with markedly higher odds of active accident involvement (OR = 12.887; 95% CI: 4.623 to 35.923). Bad weather conditions (OR = 2.049; 95% CI: 1.029 to 4.083) and motorcycle ownership (OR = 5.540; 95% CI: 2.250 to 13.644) also remained significantly associated with active accident involvement, whereas heavy traffic exposure and risky riding behaviour scores did not. CONCLUSIONS: In this sample, active accident involvement was more strongly associated with employment model, non-collision falls, bad weather conditions, and motorcycle ownership than with self-reported risky riding behaviour. These findings should be interpreted cautiously in light of the cross-sectional design, self-reported outcome definition, and non-probability sampling approach.