Journal Reports Not Consistent with FDA Studies Database

A Yale University School of Medicine review of studies is raising questions about the accuracy of trial results published in top journals, according to a research letter published in the Journal of American Medical Association.

Investigators analyzed 96 studies published between July 1, 2010 and June 30, 2011 in 19 high-impact journals such as the New England Journal of Medicine (23 studies), Lancet (18 studies), and JAMA (11 studies). Then, the studies were compared to ClinicalTrials.gov, which the US Food and Drug Administration (FDA) requires all studies to be posted on within 12 months of their completion. Analyzers found 93 of the 96 studies had at least 1 discrepancy between the journal report and the ClinicalTrials.gov results.

The most common incongruity fell under the completion rate and trial intervention category, such as different descriptions of dosages, frequencies, and durations of intervention.

Overall, 81 of the 156 (52%) primary efficacy end points listed were described accurately in both sources. For 6 listings, the difference among the results altered the trial interpretation. Only 338 of 2089 (16%) of secondary efficacy end points were described accurately in both sources.

The researchers speculate that possible causes for the inconsistencies include reporting or typographical errors, changes made during the peer review process, journal space limitations, and publications’ intentional dissemination of more favorable end points and results. Researchers note that although their study analyzed less than 100 studies, they represented high-impact journals and therefore, likely the best-case editing scenarios.