I always thought the whole point of science that nothing is ever “finally put to bed,” but never mind that for now. What did the study actually find?
The researchers examined data from 522 randomized controlled trials, averaging eight weeks apiece, of 21 different antidepressant drugs, involving 116,477 patients. The purpose was to determine the drugs’ efficacy, which was defined as a 50 percent or more reduction in a patient’s depression score on a particular scale. Every one of the antidepressants was found to be more effective than a placebo, with odds ratios, which ranged from 1.37 to 2.13, suggesting that exposure to any of the drugs was more likely to lead to the 50 percent reduction.
Unfortunately, these findings give us only the relative difference in response rates, meaning the scale of the difference between giving someone a drug or placebo. They do not give us the absolute — or actual — difference. They don’t tell us what we really want to know: How many patients do we have to expose to antidepressants to get one to respond?
Not one of the blizzard of laudatory news articles I saw asked this question, but Andrea Cirpiani, associate professor of psychiatry at Oxford and the lead author of the paper, kindly informed me via email that between 35-40 percent of patients responded to the placebo, while 55-60 percent responded to the active drug. In other words, for every five patients who took the drug, approximately one additional patient responded to the drug than placebo.
So are we justified in exposing patients to the hazards of antidepressants, which include insomnia, reduced blood clotting, suicidal thinking and loss of libido — especially when we know that most of them will not respond any better than they would to a placebo? The Lancet meta-analysis offers us no help at all in answering this question, since the authors did not look at serious adverse events.
And all of this further raises the question of whether the results obtained in short-term clinical trials, often performed on a group of professional patients who in no way resemble the folks who will be taking the drug in actual clinical practice, bear any relation to real-world outcomes. A “significant difference” between drug and placebo just means that the patients’ scores on a questionnaire changed. The difference can be tiny, as long as it is judged unlikely to have arisen by accident. These trials do not assess whether the people taking these drugs were more likely to go back to work, make friends, get married or enjoy life.
But we do know that as the consumption of these drugs has skyrocketed, so has the incidence of depression, which now is the leading cause of disability worldwide. This is not what happens when treatments work.
The headline in Newsweek proclaimed “Antidepressants do work, and many more people should take them: major international study” — going preposterously beyond the available evidence, since the paper never considered the question of whether antidepressants are superior to non-drug interventions, such as cognitive behavioral therapy.
David Healy is a professor of psychiatry at Bangor University and the author of “Pharmageddon.” “The key issue here is access to data,” he told me when I asked him about the Lancet paper.
“It’s a paper based on a series of ghostwritten articles, company publications, company reports of what the trials showed. Given what we know about the ghostwriting process, the ghostwriters don’t have access to the data behind the clinical trials either. You have to figure that we’ve got a garbage-in garbage-out kind of situation.
“It’s very clear to me that we’re giving [antidepressants] to far more people than we should be giving them to, and the reason we are doing that is people are hooked on them.”
So what to do about the burgeoning global burden of depression?
In his book “Ordinarily Well,” psychiatrist Peter Kramer gives us a look at a world most of us probably never suspected existed — the world of professional patients who spend their days going from one drug trial to another. And what do they get in return? They get income, something to organize their days around, a sense of purpose, a feeling of belonging, a modicum of concern for their well-being.
Didn’t people used to get these things from work, family, church, community? Perhaps therein lies our answer.