Objective: To determine the extent to which systematic reviews published in surgery journals reported a clinical trial registry search as part of their search strategy and whether systematic reviews that omitted such searches would have located additional trials for inclusion.
Background: Systematic reviews are used by clinicians to guide clinical decision making. When conducting systematic reviews, the comprehensive search strategy is particularly critical to identify all studies-whether published or not-for producing an overall summary effect. Inclusion of only published studies may lead to overestimated and inaccurate summary effects; thus, it is important to consider unpublished studies. Here, we investigate the extent of clinical trial registry searches performed in surgical systematic reviews because trial registries may be the most viable approach to locate unpublished trial data.
Methods: We retrieved systematic reviews from the top surgery journals and the Cochrane Collaboration. Each was reviewed to determine which bibliographic databases were used and which, if any, trial registries were searched.
Results: Of 996 total systematic reviews, 252 (25.3%) reported having included a clinical trial registry search, with systematic reviews published in journals reporting searches of unpublished research at a rate of 6.4% (47/737). Reviews published by the Cochrane Collaboration included searches of unpublished research 79.2% of the time (205/259).
Conclusions: Many systematic reviews published in surgery journals include only published research, which may contribute to publication bias. We recommend that authors maximize available information by using unpublished trial data found in clinical trial registries.
Keywords: Clinical trial registry; Clinicaltrials.gov; Meta-analysis; Meta-review; Publication bias; Systematic reviews.
Copyright © 2019 Elsevier Inc. All rights reserved.