Discussion about this post

User's avatar
Alcibiades's avatar

Excellent analysis. I wanted to comment on some of these studies that show, for example, 45% of alcoholics can quit without treatment. I'm a recovered alcoholic, and spent a lot of time talking about these types of studies in rehab with the counsellors. They are endlessly annoyed by the very simple errors made, which make these studies extremely misleading.

Every single person in AA (aside from a few who were court-ordered) are there because they have already tried abstinence and failed. They've likely tried it dozens of times on their own. Those trying various medications have also attempted abstinence and failed. This should be pretty obvious; You only go to your doctor or a meeting once you've accepted that you can't fix it on your own.

But what these studies do is take a group of people and they find that abstinence works for 45% of them. That's great! But the second group of people in treatment who are being compared have already self selected out of this 45%. This second group of people has a 0% success rate with abstinence.

So if a program works more than 0% of the time, it's better than abstinence! Because we already know abstinence worked 0% of the time for these people.

Sadly, there are essentially no studies that properly compare abstinence with treatment. Remember when people tried to study porn usage in men, but found it very hard because they couldn't find men who had never watched porn? There's a similar issue here. You can't find alcoholics ready to go through treatment who haven't already tried and failed with abstinence.

And maybe an unpopular opinion: If you can just decide one day give to up alcohol, and successfully do so.... were you really an alcoholic? Is that who we care about here? Many people struggle for decades trying to quit. They lose all their friends and family and job along the way. It's pretty silly to treat these people similarly.

Expand full comment
Jake Smith's avatar

On the Antabuse not showing a significant effect in blinded studies, you write: "The only explanation that works here is that people were so scared of the negative effect of Antabuse that even those on the placebo decided to abstain from alcohol entirely."

I would guess this is exactly what's happening. I don't know any of this for sure, but I would guess that a lot of the effectiveness from Antabuse comes from people not even trying to have a drink because they don't want the misery, rather than trying to have a drink and experiencing the misery first-hand. So if much of the effectiveness is just coming from that anticipation, it really won't matter whether you're given a placebo if you think there's a decent chance of it being the real thing.

Expand full comment
6 more comments...

No posts