Estimating alcohol involvement in fatal crashes in light of increases in restraint use

Abstract

In 1994, 40.8% of traffic fatalities occurred in crashes involving alcohol. Reducing the incidence of alcohol involvement in traffic crashes is one of the National Highway Traffic Safety Administration's (NHTSA's) highest priority programs. Another of NHTSA's highest priorities is increasing the use of occupant protection systems - safety belts, child safety seats, and motorcycle helmets. While these two safety programs complement one another, it is worth investigating their interaction with regard to measuring and assessing program progress over time. The use of occupant restraints saves lives, and a strong inverse relationship between restraint use and alcohol involvement has been observed in the fatal crash data. The primary database used to study the role of alcohol involvement in serious crashes is the Fatal Accident Reporting System (FARS), which shows that drivers in fatal crashes who have been drinking are less likely to be restrained than are sober drivers. In addition, the increase in restraint use is greater among non-drinking drivers. As restraint use increases, more lives are saved. Since restraint use is higher among sober drivers, the true trend in alcohol involvement in fatal crashes may be masked by the differential increases in occupant restraint use. Those that are most likely to use restraints are less likely to be involved in an alcohol-related crash. Nonalcohol-involved occupants will be at a higher rate than those in alcohol-related crashes, are less likely to become a crash fatality, and therefore will not appear in the FARS database. As these restrained, nonalcohol-involved occupants are removed from the pool of fatalities, the percentage of alcohol involved fatalities becomes comparably larger. To compensate for the disproportionate rates at which occupants at different levels of alcohol involvement are being saved by restraints, the percentage of fatalities at each blood alcohol content (BAC) level have been recalculated, taking into account restraint effectiveness. From this is derived a more representative estimate of alcohol involvement, the rate of alcohol involvement in potentially fatal crashes, or the potential rate of alcohol involvement. Potentially fatal crashes are those crashes which would have been fatal to an unrestrained occupant. In determining potential fatalities, specific restraint type (manual or automatic safety belt, child seat, motorcycle helmet) as well as vehicle type and seating position were accounted for. In 1994, the actual rate of alcohol involvement for fatalities was 40.8%, while the potential rate of alcohol involvement was 38.1%. As the rates of restraint use increase between alcohol- and nonalcohol-involved fatalities, this difference will continue to grow.