Modeling Injury Outcomes of Crashes Involving Heavy Vehicles in Rural and Urban Settings in Texas

摘要

Concern related to crashes involving large trucks in rural and urban settings has increased due to the potential level of sustained injuries and associated socioeconomic impacts. However, detailed studies on crashes involving large trucks in rural and urban areas have not been conducted for the Texas interstate system. This study analyzed the factors contributing to injury severity by utilizing Texas crash data, based on discrete outcome models, which account for possible unobserved effects or heterogeneity related to drivers, vehicles, and roadway environment. The study estimated random parameter logit (i.e., mixed logit) models specific to rural and urban areas separately to predict the likelihood of five severity levels used in the Crash Records Information System (CRIS) in Texas—fatality, incapacitating, non-incapacitating, possible, and no-injury or property-damage-only. The estimated models indicate that a complex interaction of factors lead to injury severities with a flexibility of the effects to vary across the observations that is very likely to occur in the conventional crash reporting system. Separate rural and urban models unveiled the risk factors associated with driver demographics, driving behavior, traffic characteristics, temporal characteristics, roadway geometrics, and environmental characteristics. Critical factors in rural crashes resulting in severe injuries include hitting roadside fixed objects, unprotected median and median width, traffic flow, wet road surface, summer months, day of week, driver demographics (race), and unsafe driving speed. Likewise, critical factors in urban crashes resulting in severe injuries are shoulder width, median width, time of day, light condition (dark but lighted), and driver demographics (age group).