Analysis of truck occupant injury severities in fatal crashes involving trucks under different light conditions on rural roadways.

INTRODUCTION: Truck-involved fatal crashes on rural roadways often result in severe safety consequences for truck occupants, and the effects of contributing factors may vary across lighting conditions. This study aimed to examine truck occupant injury severity under daylight, dark-not-lighted, and dark-lighted conditions. METHODS: Using 9,332 rural fatal truck-involved crashes from the 2012-2016 Trucks Involved in Fatal Accidents/Crashes database, we defined the dependent variable as the maximum injury severity among truck occupants, grouped into no injury, minor injury, and major injury based on the KABCO scale. Generalized ordered logit models were estimated separately for the three lighting conditions, and likelihood-ratio tests were used to justify the separate model structure. RESULTS: The generalized ordered logit models showed better relative in-sample fit than the ordered logit models, with pseudo R-squared values of 0.19, 0.24, and 0.44 for the daylight, dark-not-lighted, and dark-lighted subsets, respectively. Driver-behavior factors - including fatigued driving, careless or inattentive driving, drug or alcohol involvement, and failure to keep in the proper lane - were associated with increased probabilities of major truck-occupant injury. Roadway-context variables, such as roadway grade, two-lane segments, adverse weather, and segments with higher posted speed limits, showed condition-specific associations with injury severity, particularly under nighttime conditions. DISCUSSION: The findings suggest that rural lighting improvement and nighttime truck safety management should be prioritized for high-risk roadway and driver-behavior scenarios. These results underscore the need for targeted countermeasures that account for different lighting environments to mitigate severe injuries in rural truck crashes.