Minozzi 2025 (Cochrane): medicine plus counselling for alcohol use disorder

What they tested

Cochrane reviewers searched databases and trial registers in November 2023 for randomized trials comparing medicine plus psychosocial treatment (talk-based help such as cognitive behavioural therapy) with either one alone, or with no treatment or treatment as usual, in adults with alcohol use disorder.

Naltrexone was the most studied medicine, used in 81.0% of the trials, and cognitive behavioural therapy the most studied psychosocial treatment, in 66.7%. Most participants were men (74%), aged about 44, without other mental health or substance use disorders, and 15 trials detoxified people before treatment.

What they found

Compared with psychosocial treatment alone, the combination probably reduced the number of heavy drinkers (absolute difference -10%; RR 0.86; 8 studies, 1,609 participants; moderate certainty) and may increase the number who stayed continuously abstinent (absolute difference 5%; RR 1.17; low certainty).

It probably had little to no effect on the rate of abstinent days, serious adverse events or dropping out of treatment, judged against thresholds the authors set for a meaningful difference. Dropouts because of side effects were higher with the combination (RR 1.91), but the absolute difference of 2% fell below the authors' 5% threshold. Effects on heavy-drinking days and drinks per drinking day may have been small or absent.

Compared with medicine alone, adding psychosocial treatment may have little to no effect on abstinent days or dropouts (low certainty), and its effects on other outcomes were uncertain (very low certainty). Comparisons with treatment as usual were also very uncertain.

Only 28.5% of the studies were at low risk of bias across the main areas assessed, and the certainty of evidence ranged from moderate to very low.

Limits

• Certainty was moderate at best, downgraded mainly for risk of bias and imprecision.
• Few trials compared the combination with medicine alone or with treatment as usual.
• Most participants were men without other mental health or substance use problems, which limits how far the results apply to women and people with other conditions.

What it means if you're considering naltrexone

This review suggests that adding medicine, most often naltrexone, to counselling helps more people stop drinking heavily than counselling alone. Whether counselling adds much on top of medicine is less clear from the trials so far.