Background: Unhealthy alcohol use is the third leading preventable cause of death in the United States, accounting for more than 140,000 deaths annually. Only 0.9 percent of Americans who reported having alcohol use disorder (AUD) in the past year indicated they received medication-assisted AUD treatment.
Methods: We updated a 2014 Agency for Healthcare Research and Quality (AHRQ) report on pharmacotherapy for AUD treatment, following AHRQ Evidence-based Practice Center Guidance. We assessed efficacy and comparative effectiveness of specific medications for improving consumption outcomes (Key Question [KQ] 1) and health outcomes (KQ 2). We assessed harms (KQ 3) and sought to identify evidence for the use of pharmacotherapy to treat AUD in primary care (KQ 4) and among subgroups (KQ 5). When possible, we conducted quantitative analyses using random-effects models to estimate pooled effects. When quantitative analyses could not be conducted, we used qualitative approaches.
Results: We included 118 studies (156 articles) in our review, which included 81 studies (106 articles) from the 2014 review and 37 studies (50 articles) published since then. Studies generally included counseling co-interventions in all study groups, and the benefits observed reflect the added benefit of medications beyond those of counseling and placebo. Oral naltrexone at the 50 mg dosage had moderate strength of evidence (SOE) for reducing return to any drinking, return to heavy drinking, percent drinking days, and percent heavy drinking days. The addition of a new randomized controlled trial of injectable naltrexone conducted in a population experiencing homelessness resulted in positive outcomes for a reduction in drinking days and heavy drinking days with low SOE. Acamprosate had moderate SOE for a significant reduction in return to any drinking and reduction in drinking days. Topiramate had moderate SOE for several outcomes as well, but with greater side effects. Two other medications demonstrated low SOE for benefit in at least one consumption outcome—baclofen (reduced return to any drinking) and gabapentin (reduced return to drinking and to heavy drinking). With no new studies on disulfiram, there remains inadequate evidence for efficacy compared to placebo for preventing return to any drinking or for other alcohol consumption outcomes. No new eligible studies provided head-to-head comparisons.
Conclusions: Oral naltrexone at the 50 mg dose had moderate strength of evidence across multiple outcomes and relative ease of use as a once-daily oral medication. Acamprosate and topiramate also have moderate evidence of benefit with a less desirable side effect profile (topiramate) and a higher pill burden (acamprosate). Clinicians and patients may want to consider which treatment outcomes are most important when choosing among the medications. Current data are largely insufficient for understanding health outcomes. Finally, there is relatively little research to assess the use of medications for AUD among subgroups (9 studies) or in primary care settings (1 study).