Effect of emergency medical service response time on fatality risk of freeway crashes: Bayesian random parameters spatial logistic approach

Peng Huang, Sheng Ouyang, Han Yan, Xiaofei Wang, Jaeyoung Lee, Qiang ZengFrontiers in Public Health 2024 · DOI

Abstract

It is crucial to reduce Emergency Medical Service (EMS) response time to decrease the fatality likelihood of freeway crashes. This study utilizes a Bayesian random parameters spatial logistic approach to quantify the impact of response delays on crash outcomes, providing a rigorous statistical deep dive into the "golden hour" of emergency response.

Introduction

Traffic safety research typically focuses on pre-crash factors (why accidents happen) and crash-event factors (vehicle safety ratings, seatbelt use). Post-crash factors, specifically the efficiency of the medical response, are often overlooked due to the difficulty of obtaining reliable, high-resolution data. Freeway crashes present a unique challenge: they often occur far from hospitals, involve high speeds, and result in severe trauma. This study seeks to bridge the gap by determining the precise relationship between EMS response time and fatality risk while accounting for "unobserved heterogeneity"—hidden factors like driver behavior or specific mechanical failures that standard models often miss.

Methodology

The researchers developed a Bayesian random parameters spatial logistic model to move beyond the limitations of standard regression.

  • Random Parameters: This model allows the effect of variables (like "truck involvement") to vary across observations, capturing unobserved differences between individual accidents.
  • Spatial Correlation: Using a Leroux CAR prior, the model accounts for the fact that crashes occurring near one another often share similar unmeasured traits, such as roadway lighting or micro-climates.
  • Bayesian Inference: The authors estimated posterior distributions of parameters using Markov Chain Monte Carlo (MCMC) simulations in WinBUGS, providing a more nuanced view of risk than frequentist p-values.
Figure 2: History plot of the parameter for EMS response time in the random parameters model, demonstrating MCMC convergence.

Data and Variables

The study utilized a dataset of 1,414 crash records from the Kaiyang Freeway in Guangdong, China, spanning 2014 and 2015. The primary variable was EMS response time, measured from notification to arrival. Control variables included vehicle types, crash configurations (rear-end vs. angle), time of day, roadway geometry (horizontal curvature and vertical grade), and real-time weather data matched to the hour of the crash.

Figure 1: The distribution of EMS response time across the study dataset.

Key Results and Insights

The study found a clear, positive correlation between longer wait times and higher fatality risks. Specifically, a 1-minute increase in EMS response time results in a 2.6% increase in the odds of a fatality. A 10-minute reduction in response time is estimated to prevent 0.23 fatalities per 100 crashes.

Other significant findings include:

  • Road Geometry: A 1% increase in vertical grade (steeper hills) increases fatality odds by 197%, likely due to reduced sight distances. Conversely, tighter horizontal curves were associated with lower fatality risk, suggesting drivers compensate by reducing speed and increasing caution.
  • Crash Type: Angle crashes and rear-end collisions were 6.75 and 4.35 times more likely to be fatal than single-vehicle crashes, respectively.
  • Unfamiliarity: Vehicles registered outside the local province were linked to significantly higher fatality risks, possibly due to driver unfamiliarity with the specific roadway geometry.
  • Truck Heterogeneity: The random parameters approach revealed that truck involvement increased fatality risk in 66% of cases, a nuance that standard models would miss by averaging the effect.

Conclusion

Optimizing EMS facility locations and enforcing emergency lane usage are critical life-saving necessities. While the study is limited by its focus on a single freeway and the absence of certain driver-specific attributes, it proves that accounting for spatial and unobserved factors is critical for accuracy in traffic safety modeling. Future work may involve integrating 5G-enabled real-time traffic data to further optimize rescue routes across broader networks.

Summary updated

Abstract

It is crucial to reduce EMS response time to decrease the fatality likelihood of freeway crashes. Some countermeasures have been proposed to shorten EMS response time.
BibTeX
@inproceedings{huang2024effect,
  author = {Peng Huang and Sheng Ouyang and Han Yan and Xiaofei Wang and Jaeyoung Lee and Qiang Zeng},
  title = {{Effect of emergency medical service response time on fatality risk of freeway crashes: Bayesian random parameters spatial logistic approach}},
  booktitle = {Frontiers in Public Health},
  year = {2024},
  doi = {10.3389/fpubh.2024.1453788},
}
Figure