Brief Substance Use Interventions in General Healthcare Settings: Understanding Variability in Effects

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT Although there is some consensus that brief interventions (BI) in general medical settings can reduce self-reported alcohol consumption among patients who drink risky amounts, effect sizes vary widely, most patients continue to drink risky amounts, and there are many null studies. Despite the lack of understanding of why BI effects can vary so much across settings, screening and brief intervention are still widely recommended for implementation in general health settings. Knowing what works and what does not, how much is necessary to achieve desired effects, and in what settings and circumstances BIs work more or less, has become critical. The proposed study will therefore use meta-analytic techniques to analyze variability in the effects of BIs for drugs and alcohol delivered in general healthcare settings. The specific aims of the study are to 1) assess the overall effects and variability in effects of BIs for drugs and alcohol delivered in general healthcare settings, 2) examine variability in effects of BIs according to active intervention components, participant characteristics, provider and setting characteristics, and study methodology, and 3) create predicted profiles of the contexts and settings in which BIs may be most or least effective in reducing drug and alcohol use. The proposed study will compile a meta-analytic database from the corpus of experimental studies examining the effects of BIs in general healthcare settings. Studies eligible for inclusion will be those involving a BI delivered in a general healthcare setting, intended to reduce drug or alcohol use among patients (adults and youth) screened for heavy substance use. Randomized controlled trials conducted in any country and reported in any language will be eligible for inclusion. Detailed information and aggregate data will be extracted from study reports. Individual participant data will also be collected for all eligible studies, to permit more in- depth exploration of variability in intervention effects across different types of patients. Random-effects meta- analysis and mixed-effects meta-regression models with robust variance estimates will be used to synthesize findings across studies and create the predicted profiles identifying the contexts in which BIs are most or least effective. An expert panel of advisors will be used to guide the study, providing feedback on data collection and data analysis strategies, facilitating collection of individual participant data, and ensuring the study addresses clinically meaningful patient profiles for BI research. The results from the proposed study will significantly advance the field of BI research by documenting and exploring the variability in BI effects. Inconsistent findings across primary trials currently present a challenge for researchers and practitioners interested in implementing BIs in healthcare settings. This study will identify the intervention contexts and patient profiles for which BIs may be most or least effective, providing timely and valuable guidance for future research on and implementation of this form of drug prevention program.