Modeling and analyzing crash severity determinants across severity levels for evidence based road safety strategies on the Debre Markos to Bahir Dar highway

Road crash severity is a multifactorial outcome shaped by the complex interplay of human, vehicular, infrastructural, and temporal determinants. This study employs a Multivariate Regression (SUR) model to holistically analyze these determinants on Ethiopia’s Debre Markos–Bahir Dar Highway, capturing critical interdependencies among fatal, serious, slight, and property-damage-only (PDO) crash outcomes. The analysis reveals a distinct and policy-relevant hierarchy of determinant sensitivity (Fatal > PDO > Serious > Slight), fundamentally guided by empirical strength, societal cost, and strategic preventive rationale. Key findings identify male, middle-aged, and inexperienced drivers as high-risk cohorts, with vehicle type (notably buses and passenger car), specific collision dynamics (side/rear-end/run-off), temporal patterns (afternoon/night), and road geometry-including the counterintuitive high-risk of straight sections and downgrades-as principal factors elevating severity. The significant sensitivity of PDO crashes to factors like unlicensed driving and over-speeding underscores their role as critical precursors to severe injuries, necessitating their prioritization in proactive safety management. The implications demand an integrated, evidence-based policy framework. This study concludes by recommending targeted interventions: graduated licensing and demographically-tailored training for high-risk drivers; strategic infrastructure investment in pedestrian protection, speed controlling, and sight-distance improvements; enhanced vehicle regulation against overloading; and temporal enforcement leveraging telematics and automated systems. Supported by robust model fit (R2 = 0.472–0.678; p < 0.001), this research provides a multivariate foundation for scenario-informed safety strategies. The findings are vital for advancing Ethiopia’s Safe System goals and offer a transferable model for regions with similar traffic and infrastructural profiles, enabling sustained reductions across all crash severities through prioritized, cost-effective intervention. Key factors affecting crash severity were identified on the Debre Markos to Bahir Dar highway. Drivers, pedestrians, and passengers, as well as case specific factors show different levels of risk in road crashes. Results support targeted road safety actions to reduce crash severities.