Quit Smoking, Vaping, Alcohol

Alcohol Treatment: What the Evidence Says Actually Works

· the team behind the app

Ask most people what the options are for stopping drinking and you’ll get two answers: willpower, or AA. That’s it. That’s the whole menu, as far as public knowledge goes.

Both halves of that picture are wrong in interesting ways. There are prescription medicines with solid randomised-trial evidence behind them that hardly anyone is ever offered. And AA — routinely dismissed as unscientific — turns out to have better evidence than several of the therapies people assume are more rigorous.

This is the clinical companion to our guide on how to quit drinking, which covers twelve practical, day-to-day strategies: quit dates, triggers, urge surfing, replacing rituals. That article is about what you do. This one is about what treatment does.

Read this before anything else

If you drink heavily every day, do not stop abruptly on your own.

Alcohol is one of the few drugs where withdrawal itself can kill you. The NHS is blunt about it: it can be very dangerous to stop drinking suddenly if you’re dependent on alcohol. Symptoms typically start 6 to 12 hours after the last drink and last three to seven days, and can include tremor, sweating, a racing heart, confusion, hallucinations, and seizures. Severe withdrawal — delirium tremens — is a medical emergency.

Warning signs that you may be physically dependent include morning shakes, sweating or nausea that eases once you drink, and drinking to hold withdrawal off. If any of that is familiar, talk to a doctor before you stop. Medically supervised withdrawal is routine, safe, and nothing to be embarrassed about. In the US, the SAMHSA National Helpline (1-800-662-4357) is free, confidential and open 24/7.

Everything below assumes you’ve handled that part first.

The gap: effective treatment exists, and almost nobody gets it

Fewer than 10% of people with alcohol use disorder receive any help at all, according to NIAAA — despite the existence of evidence-based medicines and therapies. And less than 10% of patients with the condition are treated with medication.

Three medications are approved by the FDA for alcohol use disorder: acamprosate, naltrexone, and disulfiram. Two of them have genuinely good evidence. Most people who could benefit have never heard of them, and many will never have a doctor raise the subject.

First, a number worth learning: NNT

Research on medicines is usually reported as “number needed to treat,” or NNT. It sounds technical; it’s actually the most honest way to describe a treatment.

NNT is simply: how many people do you have to treat for one extra person to get the good outcome? An NNT of 15 means that if you treat fifteen people, one ends up better off than they would have been. The other fourteen either would have done fine anyway or don’t respond.

That framing kills the fantasy of a magic pill, which is good — but it also correctly sizes the benefit. Plenty of medicines doctors prescribe without hesitation have NNTs in this range, and unlike a lottery ticket, you get to combine this one with everything else you’re doing.

For alcohol, acamprosate and oral naltrexone have the best evidence for improving drinking outcomes, and they’re roughly equally effective at preventing a return to any drinking, with an NNT of about 12 to 20. Four head-to-head trials found no statistically significant difference between them.

Naltrexone or acamprosate depends on what you’re actually trying to do

Here’s where it gets useful. A meta-analysis by Maisel and colleagues, published in Addiction in 2013, found the two drugs pull in slightly different directions:

That distinction matters because not everyone wants the same thing, and it’s fine that they don’t. The NHS puts it plainly: the usual aim is to help you stop completely, especially if you’re dependent — but if you’re not ready to stop yet, or you just want to cut down, you’ll be supported to reduce your drinking to a safer level.

So the useful preparation isn’t memorising a drug name. It’s going in knowing what outcome you want, because your goal genuinely changes which medicine makes more sense.

Side effects, plainly

Neither drug is a sedative and neither gets you high. Both have real side effects worth knowing about.

No dosing information appears in this article on purpose. These are prescription medicines, the right choice depends on your kidneys, liver, other medications, and your goal, and that assessment is a doctor’s job.

Availability also differs by country. In the UK, NICE recommends acamprosate or oral naltrexone for alcohol dependence. Disulfiram — which makes drinking actively unpleasant rather than reducing the desire — is also approved, though the evidence for its effectiveness is inconsistent, and NICE positions it as an option when the other two aren’t suitable. What’s licensed, funded, and easy to get varies a lot between health systems.

The AA finding people don’t expect

In 2020, Cochrane published a review by Kelly, Humphreys and Ferri covering 27 studies and 10,565 participants on AA and Twelve-Step Facilitation (TSF) — the structured, clinician-delivered version designed to get people into AA and keep them there.

The headline: there is high-certainty evidence that manualised AA/TSF is more effective than other established treatments, including CBT, for achieving continuous abstinence. At 12 months, the risk ratio was 1.21 (95% CI 1.03 to 1.42) — that particular estimate comes from 2 of the studies, covering 1,936 people — and the effect held at 24 and 36 months. The review also found AA/TSF performs at least as well as other treatments for drinking intensity, alcohol-related consequences and addiction severity, and probably produces substantial healthcare cost savings.

For a fellowship that is free, has no professional staff, and is frequently written off as folk wisdom, that is a remarkable result.

Three caveats, because they matter:

  1. This is specifically about manualised TSF — a standardised programme, delivered properly. Non-manualised versions performed about as well as other established treatments, not better. The review’s own explanation is that TSF works largely by getting people to keep attending AA after formal treatment ends.
  2. “Better on average” is not “better for you.” These are group averages across thousands of people, some proportion of whom hated the rooms and left.
  3. If the spiritual framing doesn’t fit, that’s a real reason to look elsewhere — SMART Recovery and other secular options exist. The takeaway isn’t that you should do AA. It’s that “AA is unevidenced” is no longer a defensible thing to say.

Medication and support are designed to go together

Guidelines are consistent here. A joint SAMHSA/NIAAA consensus panel recommends pharmacotherapy alongside behavioural interventions. NICE recommends acamprosate or oral naltrexone in combination with an individual psychological intervention — not on its own.

The COMBINE trial (Anton et al., JAMA, 2006) tested this across 1,383 people, and the results are more interesting than the slogan. Those receiving naltrexone plus “medical management” — structured, fairly brief check-ins with a health professional — did better than placebo plus medical management. But the trial also found something genuinely surprising: placebo pills plus regular meetings with a clinician outperformed intensive behavioural therapy delivered without that contact.

Two honest caveats. COMBINE found no additional benefit from stacking naltrexone and intensive therapy on top of medical management, so “more treatment” isn’t automatically better. And acamprosate showed no effect in that particular trial, which sits awkwardly beside the broader meta-analytic evidence drawn largely from European studies.

The practical read: medication works best inside a structure where someone is checking on you regularly. Not pills instead of support — pills plus someone who notices.

Where counting days fits into any of this

None of the above covers most of your actual life. Treatment happens in appointments. The rest is you, a Tuesday evening, and a decision.

Counting days is not a treatment, and no trial has shown that a day counter cures anything. What it does is cheaper and more specific than that: it makes an invisible achievement visible, and gives it a value that grows. Day 3 is easy to throw away. Day 74 is not. The NHS’s own self-care advice for cutting down starts with keeping a daily record of your drinking — a diary or a tracking app — precisely because self-monitoring changes behaviour.

It also reframes the question at the moment it matters. A craving asks whether one drink would be nice. A counter reading 74 asks whether one drink is worth going back to zero. That’s a much easier question to answer. Our 30-day page walks through what that first month actually feels like, and the sobriety calculator will show you where a given date puts you.

Count your days, not your slip-ups. The Quit Smoking, Vaping, Alcohol app is a free, private day counter for quitting drinking, smoking, vaping, or any habit — milestones, money saved, and health progress, one day at a time. Download on the App Store · Get it on Google Play

Frequently asked questions

Do I have to want to quit completely to get treatment?

No. The NHS states that if you’re not ready to stop, you’ll be supported to reduce your drinking to a safer level instead. That goal also affects which medication fits: acamprosate showed the larger effect on maintaining abstinence, naltrexone on reducing heavy drinking and craving.

How do I actually get one of these medications?

Through a doctor — a GP, a specialist alcohol service, or an addiction clinician. They’re prescription-only everywhere they’re licensed. Go in prepared: how much you drink, how often, what you’ve already tried, and whether your goal is stopping or cutting down. In the US, the SAMHSA National Helpline (1-800-662-4357) can point you toward local treatment options at any budget.

If AA works so well in the research, should I just go to AA?

Try it, by all means — it’s free and it’s everywhere. But the Cochrane finding is specifically about structured Twelve-Step Facilitation, and it’s an average across thousands of people, not a prediction about you. Plenty of people get sober through SMART Recovery, therapy, medication, or some combination. The best programme is the one you’ll still be doing in six months. For what those months feel like physically, see what happens when you stop drinking.

Sources and further reading

This article is for information only and isn’t medical advice. It deliberately contains no dosing or prescribing guidance. If you’re worried about your drinking, or thinking about stopping, talk to a healthcare professional first.