Prevalence of trial registration varies by study characteristics and risk of bias

AbstractObjectivesThe objective of this study was to determine the prevalence of trial registration in health research, whether trial registration status and timing vary depending on trial characteristics, and the relationship between trial registration status and risk of bias. Study Design and SettingWe systematically reviewed all clinical trials published from January to June 2017 in 28 high- and low-impact factor general and specialty medicine journals. ResultsWe identified 370 trials and assessed risk of bias in 183 trials. Trial registration rates were high; 95% of trials were registered prospectively or retrospectively before enrollment completion. Larger sample size, multiple recruitment countries, and primary industry funding were all predictors of earlier trial registration. Prospectively registered trials had a significantly lower risk of bias compared to unregistered trials across all domains. Prospectively registered trials had a similar risk of bias compared to retrospectively registered trials across four out of six domains, and a lower risk of bias across the remaining two domains. ConclusionTrial registration is an imperfect proxy for risk of bias. Systematic reviewers should assess risk of bias on a case-by-case basis and conduct sensitivity analyses excluding high risk of bias studies. In the longer term, mechanisms should be implemented to facilitate prospective registration of all trials.