Rear compared to front seat restraint system effectiveness in preventing fatalities: L. Evans. Society of Automotive Engineers, Paper #870485, Detroit, February 1987. SAE Publication SP-691, Detroit, MI
- Publisher
- Elsevier Science
- Year
- 1987
- Tongue
- English
- Weight
- 93 KB
- Volume
- 19
- Category
- Article
- ISSN
- 0001-4575
No coin nor oath required. For personal study only.
✦ Synopsis
Each of these publications has similar objectives, but uses very different approaches to achieve them. Each examines the effectiveness of occupant restraints-primarily in the rear seat-in preventing or reducing injuries in traffic accidents. Given the controversy surrounding the conclusions of the National Transportation Safety Board in their recent study of rear seat belts [1986], the publication of these reports is particularly apt.
Campbell uses data from towaway cars involved in accidents in North Carolina between 1979 and 1985. These cars have been classified by severity of deformation, type of accident, seating position of the occupants, their age and belt use, etc. The proportion of rear seat occupants with severe or fatal injuries is usually higher among unbelted occupants than among those belted, after controlling for type and severity of collision and age and seat position of the occupant. Exceptions occur when the severity of impact and subsequent deformation is greatest, when the risk of serious or fatal injury is high whether the occupant is restrained or not.
Campbell also compares the 26 cases described by the National Transportation Safety Board [1986] with cases from the North Carolina data which appear to satisfy the same selection criteria, and comments on each of the conclusions of that report.
The remaining two reports use data from the Fatal Accident Reporting System (FARS) and attempt to summarize risks across different levels of severity. The first report presents some schematic representations of the relationship between severity of impact and the risks of fatality in restrained and unrestrained car occupants, to come up with the number killed, in theory, in each situation. The second paper presents a numerical example to compare the effectiveness of restraint systems for front and rear seat occupants. The analysis of these data is not entirely conventional; in particular, adding an arbitrary amount to the variance of the odds ratio used in the comparison seems unusually conservative.