Citation

Donelson A, Ramachandran K, Zhao K, Kalinowski A. Transp. Res. Rec. 1999; 1665: 109-117.

Copyright

(Copyright © 1999, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/1665-15

PMID

unavailable

Abstract

Fatality-based rates--for example, the numbers of occupant deaths per 10,000 registered vehicle years--are often used to compare the "realworld performance" of passenger vehicles. One assumption, often implicit, is that such rates reflect more the design or characteristics of vehicles and less the human and environmental factors that can increase the risk of fatality. The purpose of this study was to examine the importance of selected fatality-risk factors for the magnitude and meaning of fatality-based rates. The objectives were to quantify the influence of fatality-risk factors, to adjust fatality-based rates for that influence, and to assess how well adjusted rates measured differences for various groupings of vehicles. The focus of this study was on rollovers in single-vehicle crashes involving light-duty trucks. Statistical models of fatality risk were developed with multivariate logistic regression applied to data on single-vehicle rollovers of any severity. Raw counts of occupant fatalities based on data from the fatality analysis reporting system were then adjusted. Finally, the statistical reliability of differences in adjusted rates was estimated.

RESULTS indicated that adjusted rates were a very small fraction of crude rates. Differences in rates for light trucks decreased greatly and, in general, became statistically insignificant. Studies comparing fatality-based rates among vehicles need careful, statistical control of factors that increase the risk of fatal injury. Fatal crashes are rare and have special attributes. Rollover research specific to vehicles would do well to concentrate on crashes of any severity, which, for vehicles grouped by make and model, demands reference to large-volume files maintained by states.

Language: en