A Controlled study of fatal automobile accidents in New York City

Abstract

Drivers of non-commercial automobiles fatally injured in accidents in New York City were compared with non-involved drivers passing the accident sites at the same times of day and on the same days of week. The greatest difference between the two groups was in the prior use of alcohol. Among drivers rated as probably responsible for their accidents 73 per cent had been drinking to some extent whereas only 26 per cent of the similarly exposed, but non-involved drivers had been drinking. Forty-six per cent of the accident-responsible group had blood alcohol concentrations in the very high, 250 mg % and over, range. In contrast, not a single one of the drivers in the large control group had a concentration in this range. Also represented in the fatally-injured group, but not in the control group, were drivers of fleeing stolen cars, one driver killed in a drag race and a small group of drivers, all in the fifth decade of life, whose accidents resulted from the prior, medical incapacitation. It is pointed out that all of the drivers in these latter categories were sober, and that the occurrence of accidents in the medical group should not per se be used in justification of programs of medical license restrictions on the basis of present evidence. Those fatally-injured were significantly closer to home than were the similarly exposed, but non-involved drivers, and almost none of those fatally-injured lived outside the city. The case group was composed entirely of males. Although few women were driving at the times and places of the predominantly nighttime and early morning accidents, males were nonetheless significantly overrepresented in the case group. The fatally-injured were also significantly less often married, and in the entire case-control group those not married had significantly higher alcohol concentrations than those married. The fatally-injured drivers were not significantly different in age or socioeconomic status from the similarly exposed but non-involved controls, and no association between accident involvement and vehicle age were found. Finally, it is suggested that alcoholism rather than merely social drinking was involved in the case of the drivers with very high alcohol concentrations.