Reader Comments

Post a new comment on this article

Bias in Scientific Reporting

Posted by stefan on 10 Apr 2009 at 18:34 GMT

In their article, Serotonin and Depression: A Disconnect between the Advertisements and the Scientific Literature, authors Lacasse and Leo pose a question at the end of their preface: “The pertinent question is: are the claims made in SSRI advertising congruent with the scientific evidence?” However, the authors failed to respond in a fair and balanced manner to their own question. Instead, they chose to do what they accuse pharmaceutical manufacturers of doing, providing a limited scientific evidentiary basis to suit their needs. Unfortunately, the question that they chose to answer was: “Can we find flaws in the serotonin or monoamine hypothesis of depression sufficient to discredit those who would argue it as a basis for explaining the mechanism of drug action or a treatment for depression?” That is something altogether different. In their response to that question, they succeeded.

This is especially disconcerting to my associates and me who continue to make small attempts to illuminate, in the peer-reviewed medical literature, authors and their published conclusions whose works---often pharmaceutically-funded clinical research--- may be more marketing than science.

I want to provide a simple example, often used before, to illustrate my point. A glass of milk that was sitting on a table in an otherwise empty room has just been spilled. The only person who was in the room before the milk was spilled and who exited the room after the milk was spilled was a ten-year old boy, Johnny. Johnny’s mother looks in the room after Johnny exits and exclaims, “Johnny, you spilled the milk.” Despite Johnny’s insistence that he didn’t do it, his mother is convinced that he did. “Johnny, the glass of milk was upright before you went in, no one else has been in the room, after only two minutes in the room, you came out and the milk was already spilled. You even have milk on your hands! You did it, all the evidence points to you. You are grounded.” Does that seem plausible? Perhaps. Trouble is, the family cat was in the room all along, is still in the room, likes milk and has milk in his belly from recent ingestion.

Analogously, finding evidence that exclusively supports an argument against sertraline or the monoamine hypothesis of depression by ignoring the contrarian evidence that argues in the opposite direction is bad science… because it is not science. It is just polemic that serves a very different purpose.

That’s exactly what Lacasse and Leo did and what PLoS validated by publishing their one-sided work. There are articles that do demonstrate a link between serotonin and depression that were left out of the Lacasse and Leo article. Oquendo, as one example, reviews many of the articles that support the serotonin hypothesis in an article published in Clinical Neuroscience Research. (1)

Both extremes in publishing and media are bad: When pharmaceutical companies publish advertisements and media that showcase limited and incomplete findings without providing a fair and balanced approach, they mislead doctors and healthcare consumers. Likewise, when anti-pharmaceutical writers marginalize fact-gathering and science to support conclusionary premises (“pharmaceutical companies don’t tell us the truth about drugs”), they perform the same disservice to doctors and healthcare consumers by offering their own biased, incomplete and misleading conclusions.

Stefan P. Kruszewski, MD
Steven G. Klotz, MD

1. Oquendo M, Neuroimaging findings in major depression, suicidal behavior and aggression. Clinical Neuroscience Research 2001 Nov; 1(5): 377-380.

Competing interests declared: From Dr. Kruszewski: Between the years 1984 and 2001, I spoke for many major pharmaceutical companies whose products included neuropsychopharmaceuticals. Since 2003, I have been published in many leading medical journals, including but not limited to: New England Journal of Medicine, American Journal of Psychiatry, Neurology, British Medical Journal and Archives of General Psychiatry about conflicts of interest in medicine, accountability in clinical research findings and academic misrepresentations. Since 2004, I have been hired by numerous legal groups as an expert in neuropsychopharmacology and worked on behalf of private, public and governmental regulatory bodies to improve accountability in psychiatric research. My resume and complete disclosure is available on my website, My reasons for commenting on this article from 2005 from Leo and Lacasse is that I became aware of it recently, through the dialogue going on between JAMA and the authors; I believe that an essential commentary on their article has not yet been articulated. Dr Steven G. Klotz is currently a speaker for Eli Lily and Pfizer. He has had previous contractual relationships with Wyeth, AstraZeneca and Bristol-Myers Squibb (speaker and consultant) and Pfizer (as a consultant). He earns more than $5,000.00 per year from the pharmaceutical industry.