Identification of the Determinants of Severity for Pedestrian Injuries on Brazilian Federal Highways between 2017 and 2019: A Multinomial Logistic Regression Analysis

Abstract

Walking is one of the main forms of mobility among the Brazilian population, representing 41% of the trips made in the country in 2016. However, the sharing of space between pedestrians and motor vehicles in locations lacking adequate infrastructure has created conflict situations, increasing the risk of injuries and fatalities among pedestrians. This study analysed the determinants of injury severity among pedestrians on Brazilian federal highways between 2017 and 2019. Official data from the Brazilian Federal Highway Police were used to create a multinomial logistic regression model that classified the injury severity of pedestrians involved in highway crashes based on a set of explanatory variables. The results indicate that pedestrians struck by large vehicles, such as trucks, buses, and utility vehicles, have higher odds of death; periods of the day with lower natural light, such as night, dawn, and dusk, are associated with a higher risk of severe injuries and fatalities; urban road segments are less prone to fatalities than those located in rural areas; male pedestrians have higher odds of severe injury or death; and multi-carriageway roads and weekends are associated with higher odds of pedestrian death or severe injury. Finally, an annual upward trend was identified in the odds of severe injury (9.9% per year) and death (3.3% per year) among Brazilian pedestrians, alongside a higher risk of pedestrian injuries in the North, Northeast, and Central-West regions.