Seat belt design and risk of injury in rollover crashes: a comparison of conventional and seat-integrated restraints

Background: Passenger vehicles in the United States currently utilize two seat belt designs: frame-mounted ("conventional restraints"; most common) and seat-mounted ("seat-integrated restraints.") The relative risk of injury in relation to seat belt configuration is unknown. Methods: The National Automotive Sampling System (NASS) Crashworthiness Data System (CDS) provided detailed information for a representative sampling of motor vehicle crashes in the United States. Modified Poisson regression was used to compare the risk of moderate or severe injury (Abbreviated Injury Scale (AIS) ≥2) for seat-integrated restraints relative to conventional restraints. Results: After adjustment for vehicle model year and vehicle type, the relative risk of moderate or severe injury associated with seat-integrated restraints was 1.47 (95% Confidence Interval (CI) 0.95 to 2.26). The relative risk of moderate or severe head or neck injuries associated with seat-integrated restraints was 2.09 (95% CI 1.05 to 4.15). Discussion: Seat-integrated restraints were associated with a statistically significantly higher risk of moderate or severe head or neck injury, and a non-significant yet elevated risk of moderate or severe overall injury. In light of known performance limitations of seat-integrated restraints in other types of crashes, lack of evidence for a protective effect of seat-integrated restraints may be cause to re-evaluate their utility.