Medication development for AUD: A systematic review of clinical trial methodology.

ABSTRACT

Refining clinical trial methodology has become increasingly important as study design is shown to influence treatment efficacy. To maximize the efficiency of randomized clinical trials (RCTs), researchers aim to establish standardized practices. The goal of this systematic review is to describe methodological practices of clinical trials for alcohol use disorder (AUD) over the past 40 years. To achieve this goal, a PubMed search was conducted in April 2023 for RCTs on AUD medications published between July 2018 through April 2023. Resulting studies were combined with a previous search from 1985 through 2018. Inclusion criteria for the RCT studies were (1) a randomized controlled trial, (2) double or single blinded, (3) placebo or active control condition, (4) alcohol use as the primary endpoint, (5) 4 or more weeks of treatment, and (6) 12 or more weeks of follow-up. In total, methodological data from 139 RCTs representing 19 medications and spanning the past four decades were summarized. Results indicated that the most common medications tested were naltrexone (k = 42), acamprosate (k = 24), and baclofen (k = 11). On average, participants were 74% male and consumed 226 drinks per month pre-randomization. The median length of treatment was 12 weeks (IQR = 12-16; min = 4 max = 52) and the median follow-up duration was 12.5 weeks (IQR 12-26; min = 7 max = 104). There were two broad domains of outcomes (i.e., abstinence and heavy drinking), with most studies featuring outcomes from both domains (k = 87; 63%). Reporting practices were summarized by decade, revealing an increased enrollment of females, better reporting of race and ethnicity data, and less studies requiring pre-trial abstinence. This review summarizes the current state of the literature on randomized clinical trials for AUD including effect sizes for individual studies and summaries of key methodological features across this representative set of clinical trials.