Pfizer, Zoloft And The Vexing Placebo Effect
By Ed Silverman
February 1st, 2013
In an unusual gambit, a California woman has filed a lawsuit claiming that the Zoloft antidepressant sold by Pfizer is no more effective than a placebo and that the drugmaker conspired to hide this fact from the medical community and the public in hopes of generating as many sales as possible. For its part, Pfizer contends that studies have shown that Zoloft is effective and maintains the lawsuit is frivolous. The lawsuit was filed by Laura Plumlee, who maintains that Zoloft failed to help her during three years of treatment. She argues that most clinical trials that were designed to substantiate effectiveness demonstrated that Zoloft was not superior to placebo and that the difference between Zoloft and placebo in “the only two studies” that did show the pill was more effective than a placebo had been obtained through “data manipulation” [here is the lawsuit].
The complaint relies, in part, on research by Irving Kirsch, a director of the Placebo Studies Program at Harvard Medical School, who caused a ruckus last year when ’60 Minutes’ featured his work in a segment about antidepressants and the placebo effect. He filed Freedom of Information Act requests to obtain unpublished trial data and contended that, when combining results with published data, various drugs were no better than dummy pills. “The only place where you get a clinically meaningful difference is at these very extreme levels of depression,” Kirsch, who is serving as an expert in the case, told 60 Minutes. “…The reason they get better is not the chemical in the drug. The difference between drug and placebo is very, very small, and in half of the studies, non-existent… You can get the same benefit without drugs” [see the clip here].His conclusions generated considerable controversy. For instance, American Psychiatric Association president Jeff Lieberman, who is also psychiatry department chair at Columbia University, told Medscape that Kirsch “is badly and potentially dangerously misleading people. Using his criteria, he included a relatively small number of the total number of antidepressant studies that have been done and he looked at the mean effects of the different treatments. Clearly, there are some people who received medication who improved tremendously, but there are some who did not improve”…
I never say too much about lawsuits because I am such a law dunce, but this suit is about exactly what those last two posts addressed [gone missing…, dodginess…]. So even if it dies in the legal whirlpool, it makes a solid point. In 1977, I left a psychiatric residency [pre-DSM-III and pre-SSRI] seeing the existing antidepressants [MAOIs and TCAs] as drugs for Melancholic and other severe depressions just like Kirsch said [“The only place where you get a clinically meaningful difference is at these very extreme levels of depression”], and since my own practice was not with those patients, my experience after training was nil – even in the post-Prozac Era. But as a volunteer in a charity clinic, I’ve written more and had some successes. Most of the patients I see there had been on [and are on] a compendium of medications and my job often involves slowly bringing the list down to size. Slowly because of withdrawal syndromes, but even more because people get attached to their medications, making it hard to discriminate between medication as therapeutic and medication as symbolic. But overall, I’m not so negative as Kirsch, because I’ve seen some convincing responses. But I’m close.