There’s a fascinating study in today’s Journal of the American Medical Association. It’s a
meta-analysis of randomized controlled trials comparing antidepressants vs. placebo. And it showed that the placebo effect is so strong in depression that placebos work as well as paroxetine (Paxil) and imipramine (Tofranil) for all patients except those with major depressive disorder that’s classified as “very severe.” Placebo tied active medication for “mild,” “moderate,” and even “severe” depression.
Why didn’t we know this before? The study’s authors explained that there are two reasons. One is that the vast majority of the thousands of clinical trials on depression enroll only patients with very severe depression, even though the vast majority of people taking antidepressants have less severe forms of the disorder.
The second reason is that many studies include what’s called a “placebo washout” period, in which all patients being considered for the trial are given a placebo for a few days up to a week. Those who show more than 20% improvement in their depression are dropped. But removing placebo responders clearly biases the study in favor of finding significant medication effects.
It’s a demonstration that the “gold standard” of medical research–the randomized, double-blind, placebo-controlled trial–can be misleading, especially when we bow to the temptation to generalize results from a highly selected group of test subjects to the wider population.