Understanding Fatality Risk in Traffic Accidents: RTS and Individual Characteristics in Near-Zero PTD Cases

Abstract

This study aims to evaluate fatality risk in traffic accidents from a multifactorial perspective,
incorporating not only injury severity but also physiological signs (RTS) and individual characteristics (age and medical history). Particular attention is given to the “near-zero PTD group,” in which deaths occurred despite non-lethal injuries, to explore non-trauma-related fatal factors.
To assess the contribution of explanatory variables, several logistic regression models were constructed. The model using only ISS showed limited predictive power (Recall = 0.147, Pseudo R² = 0.335). By adding RTS and an age-based score, the model improved (Recall = 0.573, Pseudo R² = 0.557). Further enhancement was achieved by introducing an interaction term (age² × number of medical conditions²), leading to a substantial improvement (Recall = 0.874, Pseudo R² = 0.640), indicating that elderly individuals with multiple medical histories are at higher risk of death.
Additionally, RTS deterioration was found to be associated with psychological and environmental factors such as impact direction, nighttime occurrence, adverse weather, and alcohol use. These results suggest that RTS may reflect not only physiological trauma severity but also internal stress reactions triggered during the accident.
The findings indicate that future traffic safety strategies should address not only injury reduction, but also stress-related factors and provide targeted interventions for high-risk individuals, such as the elderly with preexisting conditions.