Concurrent and prospective screening for problem drinking among college students

Abstract

Previous studies suggest that combining the CAGE questionnaire with the Perceived Benefit of Drinking Scale (PBDS), information about an adolescent's use of tobacco, and best friend's drinking pattern is a useful composite screening measure for problem drinking. The present study was undertaken to evaluate this composite screening measure prospectively as a predictor of subsequent problem drinking among late adolescents across 3 years of college. A random sample of 452 college freshmen entered a longitudinal study of alcohol use at the beginning of their freshman year. A total of 184 (58%) completed follow-up measures of alcohol use 32 months later. Outcome measures included the quantity and frequency of alcohol use and a composite measure of specific alcohol-related problems. CAGE scores, PBDS scores, tobacco use, and best friend's drinking patterns as reported at college entry together explained 33% of the variance in the quantity/frequency measure and 37% of the variance in the alcohol-related problems measure from the end of the junior year. These same variables as reported at the end of the junior year explained 50% of the variance in the quantity/frequency measure and 61% of the variance in the alcohol-related problems measure. The composite screening measure as reported at college entry had a sensitivity of 73%, specificity of 70%, positive predictive value (PPV) of 63%, and negative predictive value (NPV) of 78% for students at high risk for problem drinking at the end of the junior year. A similar concurrent composite screening measure consisting of the same variables reported at the end of the junior year had a sensitivity of 88%, specificity of 56%, PPV of 60%, and NPV of 83% for high-risk drinkers. A total of 70-73% of students could be correctly categorized by each composite screening measure. These composite screening tests had significantly better test characteristics than the CAGE or PBDS alone. College students' responses to the CAGE, PBDS, tobacco use, and their friends' drinking remain consistent over 3 years and correlate with concurrent and future risk for problem drinking. These variables explain significant variance in drinking and alcohol-related problems and may constitute a useful screening measure for current and future problem drinking.