Pharmaceutical Companies Often Do Not Publish Negative Results From Clinical Trials of Antidepressants, Study Finds

Pharmaceutical companies often do not publish negative results from clinical trials that they conduct to obtain approval for antidepressants, a practice that can mislead physicians and consumers about the effectiveness of the medications, according to a study published on Thursday in the New England Journal of Medicine, the New York Times reports (Carey, New York Times, 1/17).

For the study -- led by Erick Turner, a psychiatrist at Oregon Health & Science University and a former reviewer of data on psychotropic medications at FDA -- researchers used results from 74 trials registered with FDA between 1987 and 2004 that involved 12 antidepressants and 12,564 patients (Armstrong/Winstein, Wall Street Journal, 1/17). The study found that pharmaceutical companies published 37 of 38 trials that FDA considered to have positive results, compared with 14 of 36 trials that the agency considered to have negative results. In addition, 11 of those 14 trials "conveyed a positive outcome" not justified by reviews conducted by FDA, the study found.

According to the Times, the study likely will "inflame a continuing debate about how drug trial data is reported" (New York Times, 1/17). Pharmaceutical companies do not have to publish the results of trials that they submit to FDA, but many companies promise to disclose all results. An FDA spokesperson said that the agency does not withhold documents related to reviews of medications (Wall Street Journal, 1/17).

Critics also have said that medical journals publish only positive results from trials and that researchers revise their hypotheses after trials conclude to make the results appear more positive (Desmon, Baltimore Sun, 1/17).

NEJM Editor-in-Chief Jeffrey Drazen said that the journal since 2002 has "made a conscious effort" to publish more negative results from trials and that an analysis by the staff found one-third of trials published in the journal in 2003 had negative results (Bloomberg/Long Island Newsday, 1/17). In addition, 12 journals in recent years have agreed to only accept trials registered before they begin to ensure that researchers do not revise their hypotheses. FDA in December 2007 mandated registration for all federally funded trials before they begin (Baltimore Sun, 1/17).

An abstract of the study is available online.

NPR's "Morning Edition" on Thursday reported on the study. The segment includes comments from Turner and Drazen (Silberner, "Morning Edition," NPR, 1/17). Audio of the segment is available online.