A Meta-Analysis of Computer-Delivered Drinking Interventions for College Students: A Comprehensive Review of Studies From 2010 to 2016

摘要

Objective: Computer-delivered drinking interventions (CDIs) are administered to tens of thousands of college students each year, yet recent evidence for their efficacy has not been summarized. This meta-analysis extends the work of past reviews and investigates the efficacy of CDIs in reducing college students' alcohol consumption and alcohol-related harms. Method: Following the systematic review standards set by PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses), the literature was searched for published and unpublished data available from 2010 to 2016. We reviewed 35 randomized controlled trials (64 CDIs, N = 20,068, 57% female) that compared CDIs with control conditions for college students and calculated between- and within-groups weighted mean effect sizes. We analyzed effects at three follow-up assessment points (short term = <6 weeks, intermediate term = 7-26 weeks, long term = >27 weeks). Results: Within-group effect sizes showed that CDI participants did make reductions in drinking over time; however, between-groups effect sizes revealed that these effects rarely differed from those of control participants. CDIs were associated with very small but statistically significant reductions in quantity (d = 0.06, 95% CI [0.02, 0.10]) and frequency (d = 0.07, 95% CI [0.02, 0.12]) of alcohol consumption when compared with controls at short-term follow-up. However, at intermediate (d = -0.07, 95% CI [-0.11, -0.02]) and long-term follow-ups (d= -0.06, 95% CI [-0.12, -0.01]), CDIs were associated with statistically significantly more alcohol-related problems than controls. Conclusions: Computer-delivered drinking interventions result in small reductions in college students' alcohol consumption over time. However, these interventions rarely reduce drinking more than controls and may be associated with some increased risk of harm over time.