Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120
Presentation
Screening for At-Risk College Alcohol Use: How Much is Too Much?
Introduction: Excessive alcohol consumption puts college students at increased risk for alcohol-related injury. To utilize intervention resources cost-effectively, it is necessary to identify students at greatest risk for the most severe consequences of excessive drinking. Current widely used screening tools for at-risk college alcohol use measure binge drinking, defined as 5+drinks per occasion for men and 4+drinks per occasion for women. This study evaluates alternative alcohol use cut-offs for risk of alcohol harms by analyzing data from the College Health Intervention Project Study (CHIPS), a randomized clinical trial of brief alcohol intervention delivered in health care settings at four US and one Canadian university. Study Sample: 2,090 college students who exceeded threshold alcohol use levels on a general Health Screening Survey and participated in a face-to-face baseline assessment for the CHIPS study. Measures: Alcohol consumption in the past 28 days is assessed with the Timeline Follow-back (TLFB) method in which participants are guided by a researcher to report number of standard alcohol drinks consumed using personally memorable events to enhance recall. Subjects self-report alcohol-related injuries in the past 6 months. Methods: Logistic regression log likelihoods are used to choose the best predictor of alcohol-related injury. The optimal cut-off point is chosen as the daily threshold which maximizes the log likelihood test statistic. Decision tree analysis simulates intervention results to determine the economic impact of screening and intervention at the proposed optimal cut-off point compared to the standard 5+(male)/4+(female) drinks cut-off for risk of alcohol-related physical harm. Results: The number of heavy drinking days (HDD), defined as 8+drinks per day for men and 5+drinks per day for women, maximizes the log likelihood test statistic for alcohol-related injury risk compared to other daily alcohol cut-off points. 1,559 subjects (75%) exceeded the 8+/5+ HDD cut-off at least once in the past month. 1,826 subjects (87%) exceeded the 5+(male)/4+(female) drinks per day cut-off at least once in the past month. Exceeding the 8+/5+ HDD cut-off identifies 91% of the subjects in the sample who experienced alcohol-related injury, while exceeding the 5+(male)/4+(female) cut-off identifies 97% of the subjects who experienced alcohol-related injury. Use of the HDD measure for screening for at-risk alcohol use compared to a standard 5+/4+ measure results in 267 fewer alcohol interventions, with 11 of those subjects suffering a recent alcohol-related injury. In simulated decision tree analysis, if the cost savings due to injury reduction are less than 24 times the cost of screening and intervention, then the use of the 8+/5+ HDD screening measure is a cost-effective alternative to screening with the 5+/4+ drinking measure. Conclusions: This study suggests that the number of days of drinking 8+drinks per day for men or 5+drinks per day for women maximizes likelihood of alcohol-related injury. Screening and intervention for alcohol-related physical harms may need to utilize higher daily drinking thresholds than the current standard in order to make the most cost-effective use of resources.