In the BMJ this week there is a debate about the antidepressants. On the “Yes, The antidepressants are overprescribed” side is Des Spence. This is hardly a new debate and Des Spence makes a good case for the overuse of the antidepressants, but what caught our eye was the response by Adrian Preda, and his discussion about the findings of Irving Kirsch. In his essay Preda never mentions a specific problem with Kirsch’s data. Instead, his main argument is that, when it comes to understanding the efficacy of the SSRIs, the main problem is the disconnect between the uncritical media and the uninformed general public on one side, and the well-informed experts on the other side.
He sees a biased and “uncritical media,” guilty of “selection bias,” who are “spinning”, “misusing language”, and “trumpeting” a story for a general public that has “its own opinion of mental illness” all of which has resulted in a “snow-balling effect that further inflames an already uninformative discourse.” On the other side are the well-informed scientists, who understand that the “placebo literature can be confusing,” but who carefully read the literature and understand the more “nuanced” questions, and realize that the Kirsch study was “not big news to the scientific community.” Surprisingly, Preda never provides a real world example of faulty reporting.
Let’s take a specific example of recent reporting. Earlier this year, 60 Minutes ran a piece on the Kirsch studies, and along with Kirsch they interviewed several other professionals, including several critics. The FDA and MHRA representatives largely agreed with him, and even his critics acknowledged that the antidepressant-placebo difference is small. Did 60 Minutes do something wrong here? From reading Preda it seems that even he agrees with Kirsch’s findings. We have written before how the pharmaceutical companies have published advertisements agreeing with Kirsch. So where is the problem?
Preda says that science had little to do with how the Kirsch study became the poster study of the antidepressant movement because it simply reconfirmed a well-known antidepressant effect. Several times he says that it is fine for the press to report the “non-spinned conclusion of Kirsch” which simply confirmed the well-known difference between drug and placebo. This is true, but why should the press not inform the general public that the difference between drug and placebo is minimal, or that according to guidelines set by NICE the difference is not significant?
In an earlier blog posting, as proof of the media’s inability to understand the science, Preda cited two studies which supposedly contradict Kirsch, including one study by Fountoulakis and Moller, and another by Gibbons and colleagues. Since Preda published his blog, but not since he has published his response in the BMJ, both of these studies have been called into question. Kirsch and his colleagues have pointed out a discrepancy in the methodology that Fountoulakis and Moller used in their reanalysis, and demonstrated that their original analysis was in fact correct. And David Healy and other physicians have raised serious questions about the data and statistical methods that Gibbons has used in several of his papers.
When Kirsch’s first meta-analysis was published In Prevention and Treatment, the entire issue was devoted to an analysis of his results by other professionals, many of whom were critics. While some of his critics argued about the details of his findings, not a single author disagreed with his main thesis. In fact, Steven Hollon, a clinical trial researcher commented: “Many have long been unimpressed by the magnitude of the differences observed between treatments and controls, what some of our colleagues refer to as the ‘dirty little secret’ in the pharmaceutical literature” (Hollon et al. 2002). All the media attention now being paid to this story clearly has something to do with the fact that the secret is out. Most surprisingly, from Preda’s response it seems that he thinks the story should still be kept secret.
Preda proposes that in the future, if the media wants to be more ethical, they need to emphasize the ambiguity of the scientific discovery.There is always ambiguity in science, but when the drugs first came on the market were patients told about the ambiguity of the six-week trials that were sponsored by the pharmaceutical companies? Or were they simply told that the drugs were extremely successful? When a medication first comes on the market this would seem to be the time for ambiguity. If we are now at a point where Kirsch has found nothing new and has simply confirmed what scientists knew all along, as Preda believes, then why should the press talk about ambiguity now? Confirmation leads to less ambiguity.
Preda’s essay is just one example of how the proponents of antidepressant medications appear to have undergone an abrupt shift as a result of recent media attention being paid to the minimal benefits of these drugs. Having remained silent for decades while the media uncritically promoted antidepressant efficacy and unsubstantiated chemical imbalance claims, antidepressant apologists now complain that the small antidepressant-placebo difference demonstrated by Kirsch and others is old news and of little interest to the general public. Rather than assume any responsibility for ten-years of problematic advertising, the main concern seems to be to shift the blame to others.
Over the past decade, we are not aware of any official professional or governmental organization publicly attempting to set the record straight, or to try and correct the public’s misunderstanding of the SSRIs, at least with anything near the effort that they are now putting forward in their rush to defend the SSRIs. Rather than blame everyone else for misunderstanding the true efficacy of the antidepressants wouldn’t the medical community be better served by even just a bit of self-reflection into their role in how the public failed to understand the true efficacy of the antidepressants?
Note: In addition to Jeff and Jon, Brett Deacon is one of the co-authors of this blog. Brett is an Associate Professor at the University of Wyoming and was one of the co-authors of the Kirsch paper in PLoS Medicine.