Injury Severity in Multivehicle Rear-End Crashes

Abstract

Rear-end crashes constitute a substantial portion of the total crashes in the United States. They are also amenable to reduction through emerging intelligent transportation systems technologies such as crash warning systems. A specific objective was to analyze the effect of information and vehicle technology on injury severity in rear-end crashes, while controlling for the effects of driver, vehicle, and roadway factors. The study is based on real-life data from the Highway Safety Information System for North Carolina access-controlled roadways. The results show that in two-vehicle crashes the leading driver is more severely injured, whereas in three-vehicle crashes the driver in the middle is more severely injured. To analyze injury severity on the KABCO scale, three separate ordered probit models were estimated for Drivers 1 (leading), 2 (striking), and 3 (striking, in a three-vehicle crash). A vehicle age variable was used in the model specification to capture the effect of vehicle age and to serve as a proxy for safety improvements, in particular the center high-mounted stoplight (CHMSL). The modeling results show that being in a newer vehicle protects the driver in rear-end collisions. Similarly, being in a newer vehicle protects Driver 2. Interestingly, striking a newer Vehicle 1 can reduce the chance of both Driver 2 and Driver 3 injuries, partly as a result of CHMSL on Vehicle 1. Also examined is whether drivers of vans, pickup trucks, and station wagon cars/trucks sustain less-severe injuries because of their larger vehicle mass or more-severe injuries because of their information-blocking effect. The results show that technological improvements have a quantifiable beneficial effect on safety.