Related Experiment Videos
Completion, publication, and outcome reporting of registered randomized trials in abdominal surgery: registry-based
Klaudia Narbekovas1, Christoph Kuemmerli2, Sandro Huegli3
1Department of General Surgery, Division of Visceral Surgery, Medical University of Vienna, Vienna, Austria.
Background:
Abdominal surgery poses unique challenges to the conduct of randomized clinical trials (RCTs). However, trial discontinuation, non-publication, and selective reporting may result in an incomplete picture of the available evidence base. This study assessed the trajectories of RCTs in abdominal surgery from trial registration to the publication and reporting of results.
Methods:
RCTs in adult abdominal surgery were identified through a systematic search of ClinicalTrials.gov, the EU Clinical Trials Register, and the International Clinical Trials Registry Platform from database inception until 31 January 2018. Corresponding trial publications were identified by searching PubMed, Google Scholar, and Google. Selective reporting was evaluated by examining consistency of primary endpoints and reported P values in published trials.
Results:
After screening 2654 registry records, 295 trials were deemed eligible for inclusion. Among these, 160 (54%) were reported as completed in trial registries. An elective non-cancer trial population was associated with lower odds of non-completion (odds ratio (OR) 0.46; 95% c.i. 0.27 to 0.78), whereas trials with no reported funding had higher odds of non-completion compared with funded trials (OR 2.05; 95% c.i. 1.16 to 3.66). Among completed trials, a corresponding publication was identified for 119 trials (74%). Completed superiority trials with added or modified primary endpoints were more likely to report P values < 0.050 than primary endpoint-unmodified trials (OR 4.28; 95% c.i. 2.23 to 8.22).
Conclusion:
Nearly half of registered trials are not reported as completed, and a significant proportion of completed trials are not published. Among published trials, discrepancies between registered and published primary outcomes are common. This is consistent with possible selective outcome reporting and evidentiary bias in abdominal surgery, with implications for both guidelines and meta-analytical work.