Two pills for alcohol use disorder have solid evidence. Most people aren't offered them.

Alcohol use disorder is common and treatable, and the treatments with the best evidence are two inexpensive, decades-old medications that most people who could benefit are never offered. A comprehensive meta-analysis lays out how well they work, and a national screening recommendation explains where the treatment path is supposed to begin.

The scale of the problem

Alcohol use disorder affects more than 28.3 million people in the United States and drives elevated rates of illness and death [s1]. Excessive alcohol use is among the most common causes of premature death in the country: the US Preventive Services Task Force cited an estimated 88,000 alcohol-attributable deaths a year over 2006 to 2010, from both acute causes such as crashes and chronic ones such as liver disease [s2].

What the medications do

A meta-analysis in JAMA pooled data from 118 clinical trials and 20,976 participants to compare treatments for alcohol use disorder [s1]. Two oral medications came out with at least moderate strength of evidence for benefit, expressed as the number of people who need to be treated for one to avoid a bad outcome — the "number needed to treat" [s1].

To prevent one person from returning to any drinking, the number needed to treat was 11 for acamprosate and 18 for oral naltrexone at 50 mg a day [s1]. For the specific outcome of returning to heavy drinking, oral naltrexone at 50 mg a day had a number needed to treat of 11 [s1]. Injectable naltrexone was associated with fewer drinking days over a 30-day treatment period — a weighted mean difference of 4.99 fewer days [s1]. Numbers needed to treat in that single-digit-to-teens range are comparable to many routinely prescribed medications, which is the point clinicians make when they call these drugs underused [s1].

They are not free of side effects

The same analysis quantified the harms, which are real but generally not dangerous. Compared with placebo, acamprosate raised the risk of diarrhoea (risk ratio 1.58), and naltrexone raised the risk of nausea (risk ratio 1.73) and vomiting (risk ratio 1.53) [s1]. These gastrointestinal effects are the main reason people stop the medications, and they are part of an honest account of what treatment involves [s1].

Note what is not on the list of best-supported options. Disulfiram, the oldest alcohol medication, which makes drinking acutely unpleasant, did not emerge from this analysis as a drug with moderate or strong evidence for benefit on the primary outcomes — its effect depends heavily on supervised adherence, and the controlled-trial evidence is weaker than for acamprosate and naltrexone [s1].

Where treatment is meant to start

Medication sits at one end of a path that is supposed to begin with detection. The US Preventive Services Task Force recommends screening adults for unhealthy alcohol use in primary care and providing brief behavioural counselling to those engaged in risky drinking [s2]. "Unhealthy alcohol use" there spans a spectrum from risky drinking to alcohol use disorder, and the recommendation reflects a judgement that screening plus brief counselling produces a net benefit [s2].

The gap between that recommendation and reality is the recurring theme. The medications work, screening is recommended, and yet a small minority of people with alcohol use disorder receive either [s1][s2]. The barrier is not primarily the evidence; it is that alcohol treatment is under-delivered in ordinary medical settings.

The limits

The meta-analysis is a synthesis of trials that varied in quality and duration, and the confidence intervals around some estimates are wide — the number needed to treat for acamprosate, for instance, ranged from 1 to 32 [s1]. Averages across trials do not predict how any individual will respond, and these medications work best as part of a broader treatment plan rather than alone [s1]. The screening recommendation applies to primary-care populations and does not by itself treat established dependence [s2].

What the evidence supports is straightforward: effective, low-cost medications for alcohol use disorder exist and are underused, and the first step — screening — is a recommended part of routine care that often does not happen [s1][s2]. This article describes what the evidence shows. It is not medical advice, and decisions about medication or treatment for alcohol use belong with a qualified clinician.

Sources

• Pharmacotherapy for Alcohol Use Disorder: A Systematic Review and Meta-Analysis — JAMA, 2023-11-01
• Screening and Behavioral Counseling Interventions to Reduce Unhealthy Alcohol Use: US Preventive Services Task Force Recommendation Statement — JAMA, 2018-11-01