- Clinical Trials
- Research News
- Industry Trends
- Agency Actions
- Drug Safety Issues
- Approvals, Launches, & New Indications
- Health Care Reform
Researchers Find Bias and Selective Reporting in Surgical Clinical Studies
Surgical journals more at fault than medical journals (Mar. 14)
According to a recent Reuters report, drug trials aren’t the only studies that can be clouded by selective reporting of positive results. Researchers in the Netherlands have found that trials of surgical treatments can be even more biased. Their study was published in the Annals of Surgery.
The authors’ objective was to compare the primary outcomes specified in trial registries with those reported in published papers to determine whether outcome-reporting bias favored significant outcomes. Outcome-reporting bias was defined as the selective reporting of a subset of original registered outcome measures based on their results.
In a PubMed search, the investigators identified 327 surgical randomized controlled trials (RCTs) indexed between 2007 and 2012 in 10 surgical journals and 10 general medical journals. They found that registration was lacking for 109 (33%) of these papers — 107 (98%) in surgical journals and two (2%) in medical journals.
Twelve trials (6%) were still recruiting patients at the time of publication, and 48 (22%) were registered after the study was completed. Only 152 studies (46%) were registered before the end of the trial, with the primary outcome clearly specified. Of these papers, the authors found that 75 (49%) showed evidence of discrepancies between registered and published outcomes, most often related to the omission of a stated primary outcome or the introduction of a new one. These discrepancies favored statistically significant results in 28% of the papers.
The authors concluded that the quality of registration of surgical trials published in surgical journals was inferior to those published in general medical journals. Moreover, surgical journals were more likely to feature selective outcome reporting.
According to the Reuters report, incomplete or modified publication of clinical trial results can have serious consequences, including the promotion of ineffective or harmful treatments, or the use of expensive therapies when cheaper, more effective alternatives are available.