A group of researchers reported in the New England Journal of Medicine Jan 17 (N Engl J Med 2008:358:252-60) results from FDA-reported studies of 12 antidepressant agents involving over 12,000 patients. They then searched the medical literature to identify matching publications and compared the published outcomes with the outcomes of the FDA-reported studies.
According to the published literature 94% of the studies appeared to have a positive outcomes. By contrast, the FDA analysis showed that 51% were positive. Among 74 FDA-registered studies, 31% were not published. 22 studies viewed by the FDA “as having negative or questionable results were either…not published or published in a way that, in our opinion, conveyed a positive outcome (11 studies).”
In summary, it appears there is extreme selection bias toward positive results in reporting these clinical trials. Although the cause might have been due to “failure to submit manuscripts…or decisions by journal editors and reviewers not to publish,” this would seem in my view, to be extremely unlikely. In any event, the authors of this paper concluded that “Selective reporting of clinical trial results may have adverse consequences…” (To put it mildly).
I hope this shocking article receives the media coverage it deserves.