The objective of the study presented in this scientific poster was to determine the relative involvement of drivers from four ethnic groups (Whites, Blacks, Hispanics, and Asians) in alcohol-related fatal crashes using two normalising methods: (1) rates per vehicle mile of travel; and (2) the involvement rate of drinking compared to nondrinking drivers in fatal crashes. Fatally injured driver records in the Fatality Analysis Reporting Systems (FARS) file were linked with the Hyde Cause of Death (HCOD) file. Matched data were available for the years from 1990 to 1996, and yielded 160,313 fatally injured drivers with known race/ethnicity. Measured blood alcohol contents (BACs) were available for 116,145 of those fatally injured drivers. It is argued that the best measure of the relative significance of alcohol involvement in fatal crashes is provided by the ratio of drinking to nondrinking drivers in fatal crashes. This odds ratio (OR) measures the probability that a fatally injured driver in any age, gender, or ethnic category will have been drinking. Comparisons across racial/ethnic groups showed that drivers ranging from 21 to 50 years old had the highest probability of being BAC-positive. Using this measure of alcohol involvement (drivers with positive BACs/drivers with zero BACs), Hispanics had the highest OR, followed by Whites, Blacks, and Asians. The patterns for female drivers were similar to, but lower than, the male drivers. The results of a logistic regression analysis suggest that a closer look is merited among the patterns within ethnic groups, and how the age and gender effects on alcohol involvement differ from one racial/ethnic group to another.
Abstract