Clinical trials registries are under-utilized in the conduct of systematic reviews: a cross-sectional analysis

BACKGROUND: Publication bias is a major threat to the validity of systematic reviews. Searches of clinical trials registries can help to identify unpublished trials, though little is known about how often these resources are utilized. We assessed the usage and results of registry searches reported in systematic reviews published in major general medical journals. METHODS: This cross-sectional analysis includes data from systematic reviews assessing medical interventions which were published in one of six major general medical journals between July 2012 and June 2013. Two authors independently examined each published systematic review and all available supplementary materials to determine whether at least one clinical trials registry was searched. RESULTS: Of the 117 included systematic reviews, 41 (35%) reported searching a trials registry. Of the 29 reviews which also provided detailed registry search results, 15 (52%) identified at least one completed trial and 18 (62%) identified at least one ongoing trial. CONCLUSIONS: Clinical trials registry searches are not routinely included in systematic reviews published in major medical journals. Routine examination of registry databases may allow a more accurate characterization of publication and outcome reporting biases and improve the validity of estimated effects of medical treatments.