Alcohol medications: what the research supports

What the review found

McPheeters and colleagues’ 2023 review included 118 trials and 20,976 participants. It supported oral naltrexone and acamprosate for improving alcohol-consumption outcomes alongside psychosocial care.

The estimated number needed to treat to prevent a return to any drinking was 11 for acamprosate (95% confidence interval 1–32) and 18 for oral naltrexone (4–32). For preventing return to heavy drinking, the estimate for oral naltrexone was 11 (5–41). These are pooled results for specific outcomes and study conditions, not a guarantee for an individual.

What “number needed to treat” means

A number needed to treat describes the estimated additional benefit relative to the comparator over the study period. It does not mean only one person in that group improves; people in both groups may improve. It also does not tell us that the same result persists indefinitely.

A smaller number can suggest a larger absolute benefit only when the outcome, comparator, follow-up and population are comparable. The figures above concern different endpoints and should not be turned into a simple medication ranking. Confidence intervals show that the estimate has uncertainty.

Medication and withdrawal are separate decisions

NHS alcohol guidance distinguishes support for reducing or stopping alcohol from the management of dependence and withdrawal. Someone who may be physically dependent should seek medical advice before abruptly stopping. Severe confusion, hallucinations or seizures require emergency help.

This article concerns evidence to discuss with a clinician, not instructions for starting, stopping or dosing medicine. A prescription decision also needs a review of contraindications, interactions, monitoring and the person’s treatment goals.

Questions that make an appointment more useful

• Which outcome are we aiming for, and how will we track it?
• Which options fit my health history and other medication?
• What side effects or warning signs should prompt contact?
• What practical support will help me follow the plan?
• When will we review benefit and decide whether to continue or change treatment?

Write down the agreed review date and who to contact if the plan becomes difficult. A treatment option can have a strong research basis while still needing adaptation to the individual.

Putting emerging findings in perspective

Research on newer medicines should be evaluated separately from established options. A small trial of a new intervention does not automatically overturn a larger body of evidence or answer questions about long-term safety. Our GLP-1 research explainer compares two recent semaglutide trials and the populations they studied.

For a broader introduction, read the alcohol guide and withdrawal safety guide. The research should make a clinical conversation better informed, not replace it.