Identification of the determinants of severity for pedestrian injuries at Brazilian federal highways between 2017 and 2019: A multinomial logistic regression-based analysis

Abstract

Walkability is a fundamental means of transportation among the Brazilian society, representing 41% of the travels inside the country in 2016. However, the absence of adequate infrastructure has caused conflict in places where people and vehicles share space, leading to injuries and fatalities among pedestrians. This study analyses the determinants of the severity of pedestrian injuries that occurred at Brazilian Federal Highways between 2017 and 2019. We used official data from the Federal Highway Police to build a multinomial logistic regression model, which described the physical condition of pedestrians based on the explanatory variables available in the database. The results indicate that pedestrian-vehicle crashes involving heavy vehicles, such as trucks, buses, and SUVs, caused higher odds of pedestrian-death. Furthermore, dark daytimes such as night, dawn, and dusk, are risky of both severe injuries and pedestrian-deaths; urban zones have a lower risk of pedestrian-death compared to rural areas; male pedestrians are more threatened by death and severe injuries than female pedestrians; and multiple corridor highways and weekends hold higher odds of pedestrian death or severe injuries. Finally, we identified an annual rise in the odds of severe injuries (9,9% per year) and deaths (3,3% per year) among Brazilian pedestrians, as well as a higher risk of physical damage for northern, northeastern, and middle-western pedestrians.