ALCOHOL AND HIGHWAY SAFETY: BEHAVIORAL AND MEDICAL ASPECTS

Drivers involved in fatal and serious injury highway crashes throughout Vermont were compared with drivers: (1) using the same roads at similar times, but without crash involvement; (2) with recent drunken driving arrests; (3) with arrests for other serious traffic violations; and (4) with no crashes or citations in the previous five years. Variables studied included presence of alcohol, drinking patterns, driving patterns, and social problems, especially those related to alcohol use. Among driver fatalities, 54% had alcohol, and 42% had 100 mg% or greater. Driver fatalities with alcohol tended to be young to middle-aged males with histories of medium to heavy drinking of both beer and liquor and with fatty degeneration of the liver. Those without alcohol tended to be older and to be light drinkers. Regulated drugs other than alcohol were almost never found. Twenty-three percent of fatalities died of probably survivable injuries due to problems throughout the emergency care system. Among roadblock comparison drivers, 14% had alcohol and 2% had 100 mg% or greater. Twelve percent were classified as heavy drinkers (5+ drinks per sitting) and they were over-represented among persons who had alcohol when stopped, and who had prior crashes or citations. Most drivers below legal age for drinking reported that they do drink, many heavily. Drinking, and heavy drinking, were more common among men than women. At a blood alcohol concentration of 100 mg%, risk of being involved in and responsible for a fatal crash was 7 times that without alcohol, and at 150 mg%, it was 25 times greater than the risk with no alcohol. Among clear-record drivers (no crashes or citations in the previous five years), 2% had alcohol when tested at roadblocks and 0% had 100 mg% or greater. Almost all drunken drivers were males, and most (60%) were heavy drinkers with excessive numbers of prior crashes and citations. Many were classified in the lower occupational level, and many were divorced, widowed, or separated. A discriminant function analysis succeeded in correctly classifying 95% of clear-record drivers and 87% of convicted drunken drivers on the basis of four significant variables: number of lifetime citations, occupational level, frequency of beer consumption, and quantity of liquor consumption. Three types of induced-intoxication experiments were conducted to study the influences of alcohol upon driving-related behavior: small-group studies, laboratory experiments, and a closed-course pilot study using an instrumented car. It was concluded that doses of alcohol which result in presumptive legal impairment may be associated with: (1) reductions in performance on both auditory and visual attention tasks which require the monitoring of multi-channel inputs; (2) decreases in responsiveness to stimulation of the retinal periphery; (3) alterations of visual perception in ambiguous situations; (4) increases in the likelihood of risky behavior in gaming or chance-taking situations; (5) differential mood and performance effects with respect to personality; and (6) reductions in driving accuracy and changes in automobile control-use patterns.