Development of a composite measure for assessing alcohol treatment outcome: operationalization and validation.

This article operationally describes and empirically validates a composite outcome measure developed for use in a multisite alcohol treatment matching study. Using empirically based clinical guidelines to establish alcohol consumption and alcohol-related problems criteria, 1,726 subjects were classified as abstinent, moderate drinking without problems, heavy drinking or problems, or heavy drinking and problems at intake and 3, 6, 9, 12, and 15 months postintake. Subjects with poorer composite outcome also had poorer outcomes related to quantity and frequency of alcohol consumption, alcohol-related problems, serum gamma-glutamyltranspeptidase and other nonalcohol-related measures assessing psychiatric dysfunction, psychosocial functioning, and purpose or meaning in life. Differences in the present composite measure relative to other categorical measures used in research to date and the potential for incorporating nonalcohol-specific variables into composite measures are discussed.