Discrepancies are common between results reported on the ClinicalTrials.gov results database and those published in peer-reviewed medical journals, according to a study published in Annals of Internal Medicine.
Clinical trial results may be selectively reported in medical journals. ClinicalTrials.gov is a clinical trial registry developed to provide the public with a Web-based, searchable source of information about trials but validity of posted results is unclear. Researchers looked at a random sample of 110 phase 3 or 4 trials completed before January 2009 to assess the consistency of results reported on the registry and those published in matching peer-reviewed medical journals. Of the trials studied, 20 percent inconsistently reported the primary outcome result. However, only a few of those inconsistencies were potentially meaningful. Descriptions of primary outcomes were inconsistent 15 percent of the time, while 80 percent of trials contained a secondary outcome reporting discrepancy. Adverse events were reported inconsistently in more than one third of trials with fewer serious adverse events being reported in published articles than on ClinicalTrials.gov.
The authors write that data entry errors may be to blame for some inconsistencies. Other inconsistencies may take shape during the peer review process where modifications in how results are analyzed or reported could contrast with data submitted to the ClinicalTrials.gov results registry. The authors suggest that until there is a gold standard clinical trial reporting source, discrepancies will have to be clarified by the investigator.