Marginal Improvements in Risk of Bias of Nephrology Randomized Controlled Trials Over Time: A Meta-research Study of Cochrane Reviews of Randomized Controlled Trials.

Rationale & Objective: The rigor of randomized controlled trials (RCTs) is essential for evaluating treatment efficacy, but their validity depends on methodological rigor. In nephrology, a field historically underrepresented in RCTs, concerns about trial quality are particularly important. This study evaluates time trends in the conduct and reporting of nephrology RCTs to assess improvements in methodological quality. Study Design: This meta-research study analyzed systematic reviews of kidney disease interventions from the Cochrane Database of Systematic Reviews. Setting & Participants: We included all RCTs and quasi-RCTs of from Cochrane reviews with 7 Cochrane risk of bias domains. Predictors: We used the year 2001 as a cut-point to evaluate the impact of increased awareness of trial methodology following the revised the Consolidated Standards of Reporting Trials (CONSORT) statement. Outcomes: The primary outcome was the change in risk of bias domains over time. Analytical Approach: Piecewise linear regression assessed change in bias domains before and after 2001. Results: From 153 Cochrane reviews, 3083 trials were identified. High bias was most common in detection bias (47%) and least in sequence generation (3%). Between 2001 and 2019, there were increases in low bias across selection bias (1.6% per year; Limitations: Findings are based on Cochrane reviews, which may limit generalizability. Bias assessment has evolved, potentially affecting findings over time, and our analyses was unadjusted and did not incorporate potential confounding variables. Conclusions: The conduct of clinical trials in nephrology has improved modestly largely because of better reporting. Ongoing efforts are needed to enhance trial design and ensure RCTs in nephrology are trustworthy and useful to clinical decision makers.