Comparison of Registered and Published Primary Outcomes in Randomized Controlled Trials of Orthopaedic Surgical Interventions.

BACKGROUND

The selective reporting of a subset of the outcomes that had been originally reported to a registry is a potential threat to the validity of evidence-based medicine. The extent of selective reporting has not been described for randomized controlled trials (RCTs) assessing the effectiveness of orthopaedic surgical interventions. The objective of this study was (1) to determine the percentage of orthopaedic surgical RCTs published in high-impact orthopaedic journals that were reported to have been registered, (2) to evaluate the consistency between the primary outcome measures recorded in the registry and those reported in the article, and (3) to evaluate whether selective reporting favored statistically significant outcomes.

METHODS

We searched PubMed for articles on RCTs assessing orthopaedic surgical interventions indexed from January 2010 through December 2014 and published in the ten orthopaedic journals with the highest impact factors. For every article in which the authors reported registration of the RCT, we extracted the number and nature of the outcome measures from the article and the corresponding information from the registry. We then evaluated the consistency between the primary outcome measures reported in the registry and those reported in the published article. Moreover, we evaluated whether selective reporting favored statistically significant outcomes.

RESULTS

Of the 362 articles on orthopaedic surgical RCTs, ninety (24.9%) reported that the RCT had been registered and thirty-four (37.8%) of the ninety had been registered adequately (registered before the study end with a clear description of the primary outcome measure and its time frame, with no substantial change after the study end). Twenty-six reports were eligible for our evaluation of the consistency between the registered primary outcome measures and those reported in the published article. This analysis identified one or multiple major discrepancies for fourteen articles, eight of which favored statistically significant results.

CONCLUSIONS

Few articles on orthopaedic surgical RCTs reported registration of the trial, and even fewer of these trials were registered adequately. Inconsistencies between registered primary outcome measures and those reported in the published articles, as well as selective outcome reporting favoring statistically significant outcomes, were prevalent.

CLINICAL RELEVANCE

Although trial registration is now the rule, it is currently far from optimal for orthopaedic surgical RCTs, and selective outcome reporting is prevalent. Full involvement of authors, editors, and reviewers is necessary to ensure publication of quality, unbiased results.