Speaking of meta-things, what sorta bothers me is that the more we talk about antidepressants being ineffective, the more ineffective they are going to become. I understand all the scientific and ethical quandaries of prescribing placebos, but just for a second, consider the possibility that SSRIs are all a bunch of sugar pills. (harp music starts here) Folks come in to treatment with depression, you give them a pill, they get hopeful, they get activated, they agree to come to therapy (which we believe is pretty helpful) because you told them they really needed to or the "meds wouldn't work as well" (and they wouldn't have come in at all unless they thought you were going to pick up your pen and write a script). They come in 6 months later with some symptoms coming back, and you bump up the prozac from 20mg to 40mg. They get hopeful again that they are going to feel better, and lo and behold, they do. They even start feeling better faster than they would have with simple regression to the mean, because they expected to feel better. They expected to do so because the seemingly caring psychiatrist spent 30 minutes talking to them once a month, telling them they have a good chance of feeling better with a little bit of work, and making sure they went to therapy (which they half do just to make you happy, and just to prove "they aren't dependent on pills," which of course they aren't). Without the antidepressant, they never come see you. No insurance company will pay them to see you, at least not regularly if you're not their active therapist. And maybe they never go to therapy, because if there's one thing we know about folks with depression, is that they are good at all-or-nothing thinking. And they stay depressed.
We don't talk about hypertension or diabetes or COPD as biopsychosocial diseases. The placebo effect directly targets the psycho part of that model. It's simply more important in psychiatry, and I'm not sure it's a valid comparator. Every psych study with placebo should also have a wait list comparison. Because in the real world, THAT'S the comparison. If Prozac helps 65% of the time, placebo helps 55% of the time, and the wait list helps 25% of the time, then gosh darn it, giving someone an antidepressant and getting them into treatment is pretty effective. I don't have to treat 10 people to help one person get better, I have to treat about 3 people. You can't apply the same logic to diabetes to the same degree. Does that make our specialty a non-medical, snake oil specialty? I don't give a **** what Tom Cruise thinks. I am happy when my patient comes back and is feeling better and wants to keep working with me to feel better.
Now, I don't believe antidepressants are sugar pills, and we have plenty of very good studies that show how helpful they can be, and there are plenty of lousy studies that show they don't, and only some overlap of the two study types and outcomes. But I am pretty sure that the more the antimedicaton crowd tells people they don't work, the more they are going to not work. And that makes me sad. I guess I should go take my Wellbutrin.
