Examining Risk Factors Influencing Fatal Pedestrian Crashes on Rural Highways: Matched Case-Control Study Design

Abstract

Pedestrian fatalities on high-speed rural roads are a growing concern in low- and middle-income countries (LMICs), where infrastructure often prioritizes vehicular mobility over pedestrian safety. This study investigates the effects of factors contributing to fatal pedestrian crashes on high-speed rural roads in India using a matched case-control (C-C) study design. Six high-speed road stretches were selected, and data related to crashes, road geometry, built environment, pedestrian exposure, and traffic characteristics were collected. Four models using conditional logistic regression on four datasets with varying matching ratios (1:1 to 1:4) and one model using binary logistic regression on unmatched data were developed to estimate odds ratios for potential risk factors. Key predictors of fatal pedestrian crashes included the presence of junctions, authorized median gaps, service roads, bus stops, schools, roadside eateries, canals/bridges/culverts, petrol pumps, and segments with high pedestrian population exposure. Model 4 (1:4 matched) was the best fit, demonstrating that matching improved estimate precision and controls for confounding. Segments with medians showed protective effects, while the unauthorized median gap produced a counterintuitive result. This study highlights the impact of geometric and environmental factors on pedestrian safety in the rural areas of LMICs. It also demonstrates the feasibility and effectiveness of the matched C-C design in data-constrained settings. The findings provide valuable insights to guide targeted interventions, including infrastructural enhancements, speed management, and strategic planning of bus stop locations and pedestrian crossing facilities within rural settlements along highways.