You can see our post there to the right. That was the image we shared. Here’s a sampling of how people responded:
- I find this to be a dangerous overgeneralization.
- Be careful with this. Lives are at stake.
- This is just wrong!!
- Dangerous, irresponsible statement!
Not only do we stand by the quote, the finding itself isn’t especially controversial in the world of research on treatment for substance use disorders. It’s mostly controversial among professionals who don’t want it to be true.
The specific finding comes from a large-scale review of the literature in Hester and Miller’s Handbook of alcoholism treatment approaches: Effective alternatives. They ranked 48 alcohol treatments by effectiveness, and found that the most effective treatments were brief interventions, motivational enhancement, and medication. Attending 12-step groups (like Alcoholics Anonymous) ranked 37th out of the 48 studied modalities. The no-treatment-at-all approach, which they labeled “self-monitoring,” came in several spots higher, in 30th.
Like any meta-analysis, this study has meaningful methodological imperfections. And each of the studies that went into the meta-analysis has imperfections of its own. But the relative ineffectiveness of 12-step approaches is well-known and has been for some time.
AA has worked for thousands – but just a small percentage of those who have tried it
The second edition of the Big Book for Alcoholics Anonymous reported that “of alcoholics who came to A.A. and really tried, 50% got sober at once and remained that way; 25% sobered up after some relapses.” But the group has been studied many times in many ways over the years, and the numbers offered by AA appear wildly out of step with what researchers more typically find. To be sure, the clause “and really tried” in the Big Book muddies the issue, since it isn’t clear what it takes to be considered “really trying.” It also, notably, shifts the blame for failure onto the individual alcohol user, and away from AA.
A good, objective review of the available literature on AA’s effectiveness can be found on Wikipedia. It includes some studies and meta-analyses that support AA, and others that either do not support AA or that show a negative impact of attendance.
There are a number of challenges involved in evaluating any voluntary, anonymous program like AA. Most notably, unless a study involves random assignment to various forms of treatment, it is difficult to separate correlation from causation. If someone stays sober while attending AA, they might be experiencing a benefit of AA — or they might be well-motivated to stay sober, with that motivation driving both the AA attendance and the sobriety. If it’s the latter, then AA isn’t actually doing anything. Notably, the few experiments that have been done with AA that have involved random assignment have generally shown no positive effect, with at least one showing a negative effect of AA. To AA proponents, this simply demonstrates that AA attendance needs to be voluntary for it to work.
There is no doubt that AA has helped many thousands of people. It also is not especially effective overall. It almost certainly helps a lower percentage of patients than other modern methods. And AA remains a useful referral for many clients struggling with alcohol use issues. Despite the protests of some of our colleagues, none of these statements are contradictions of any of the others.
The reason AA can be both helpful to many thousands and also not especially effective is that the program is so large. AA groups can be found all over the world. Millions have attended at least one meeting in an effort to seek help for their drinking.
Even though it is likely less effective than other forms of treatment, AA still can be a useful referral. Remember that research gives us large-scale averages, not individual-level predictions (or at least, not very good ones). If your particular client is struggling with alcohol use and needs social support, AA offers precisely that. If the message of personal responsibility and faith in a higher power is a good match for the client’s worldview, they may feel very at home in AA groups. And if they can’t afford other forms of treatment but do not appear likely to succeed on their own, AA can provide useful accountability.
We’ve got no beef with AA, or any other 12-step program. We just want to help mental health professionals practice from the best available evidence — something that we’re all ethically required to do.
Based on the current evidence, AA should not be an automatic referral for all clients with alcohol issues. Of course, you can and should refer to AA if there is a compelling rationale to do so. You should just be prepared to explain the referral if challenged on it. AA is not especially likely to be effective. There are better frontline referrals out there.