Drinking status and fatal crashes: which drinkers contribute most to the problem?

• Published in: Journal of studies on alcohol Vol. 67; no. 5; p. 722
• Main Authors: Voas, Robert B, Romano, Eduardo, Tippetts, A Scott, Furr-Holden, C Debra M
• Format: Journal Article
• Language: English
• Published: United States 01.09.2006
• Subjects: Accidents, Traffic - mortality; Adolescent; Adult; Alcohol Drinking - epidemiology; Automobile Driving - statistics & numerical data; Female; Humans; Male; Risk Factors
• ISSN: 0096-882X
• DOI: 10.15288/jsa.2006.67.722

The object of this study was to estimate the relative contribution of various classes of drinkers (including those with alcohol-use disorders) to alcohol-related fatal motor vehicle crashes. Using the National Epidemiologic Survey on Alcohol and Related Conditions conducted in 2000, the percentage of state residents falling into six nonoverlapping alcohol-user categories-dependent drinkers, abusive drinkers, dependent and abusive drinkers, heavy episodic drinkers, current normative drinkers, and current nondrinkers- was determined based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, classifications. The percentage of residents in each state in each of these user categories and their relationships to the number of drinking drivers involved in fatal crashes in that state were determined through regression analysis using data from the Fatality Analysis Reporting System. The proportion of drinkers in a state in each of the six consumption categories was positively related to the number of drinking drivers in fatal crashes in that state. Conversely, the percentage of the state's population who were current nondrinkers was negatively related to the number of drinking road users in crashes. Although alcohol abusive and heavy episodic drinkers had substantially higher associations with impaired drivers in fatal crashes, half of such drivers were associated with the percentage of current nominative drinkers in the state. Despite the relevancy of these findings, they must be qualified by statistical limitations associated with the use of state as the unit of analysis.