Identifying policy-relevant traffic crash risk factors in Cheongju, South Korea using logistic regression and explainable machine learning

Clinical Snapshot

Key Findings

• P Value: Not reported in abstract; variables described as 'significant predictors' without threshold specification

• Effect Size: Not reported in abstract; qualitative identification of traffic signal violations, failure to comply with safe driving obligations, and absence of median barrier on undivided roads as consistent significant predictors across both analytical approaches

• Primary Outcome: Traffic crash severity (ordinal: property damage only / injury / fatal) in Cheongju, South Korea

• Nnt Or Sensitivity: Not applicable (observational severity study); ML model performance metrics (AUC, sensitivity, specificity) not reported in abstract

• Confidence Interval: Not reported in abstract

Clinical Application

Findings are operationally feasible for policy translation: signal enforcement and median barrier installation are established road safety interventions. However, direct clinical application (e.g., injury prevention protocols) requires linkage to trauma registry data not present in this study Limited direct applicability to Australian practice. Australian road safety is governed by the National Road Safety Strategy 2021–2030 and state-based road authorities (e.g., Transport for NSW, VicRoads, MRWA). The Safe System approach underpinning Australian policy already incorporates median barrier installation on high-speed roads (consistent with this study's findings) and automated speed/signal enforcement. The RACGP has no direct role in traffic crash prevention policy, though GPs contribute to driver fitness assessments under Austroads guidelines. The specific enforcement and infrastructure gaps identified in Cheongju may not reflect Australian conditions. Korean road law and enforcement culture differ substantially from Australian jurisdictions. Road safety policymakers, traffic engineers, and public health authorities in mid-sized urban centres with comparable road infrastructure and traffic enforcement systems

Abstract

Rapid urbanization and increasing traffic volumes have made the occurrence of traffic crashes and the resulting harm a major public safety concern. This study analyzes traffic crash data from Cheongju, a mid-sized city in Chungcheongbuk-do, to identify key determinants of crash severity and provide evidence-based policy recommendations. Our approach is novel in that it integrates statistical modeling and machine learning methodologies; this dual approach not only overcomes the limitations inherent in using either technique alone but also allows for the identification of consistent risk factors influencing traffic crash severity that may have gone unrecognized otherwise. Marginal effects of explanatory variables were interpreted using ordinal logistic regression, while feature importance in machine learning models-including Support Vector Machine, Random Forest, XGBoost, and LightGBM-was evaluated using SHAP (SHapley Additive exPlanations) values. Both analytical approaches consistently identified traffic signal violations, failure to comply with safe driving obligations, and the absence of a median barrier on undivided roads as significant predictors of crash severity. By leveraging empirical data specific to Cheongju, our research provides regionally tailored insights that distinguish our work from prior studies with broader or less localized focus. These findings highlight the need for stricter enforcement of traffic regulations and structural improvements in roadway infrastructure and can inform policymakers in formulating effective, context-specific measures to enhance road safety.

References

1. 1.Lee, E.-J., Kim, S., Lee, H.-J., & Jhong, J.-H. (2026). Identifying policy-relevant traffic crash risk factors in Cheongju, South Korea using logistic regression and explainable machine learning. PLoS ONE. https://doi.org/10.1371/journal.pone.0350616