Risk of bias and the reporting of surgeons' experience in randomized controlled trials of total hip and total knee arthroplasty: A systematic review

Rationale, aims, and objectives: The potential bias introduced by surgeons' lack of comparable, relevant experience when performing the procedures in different arms of randomized controlled trials (RCTs) is arguably not well-managed or reported. The aim of this work was to review the frequency and nature with which surgeons' relevant experience is reported in RCTs of total hip (THA) and total knee arthroplasty (TKA), and to relate this to other risk of bias domains for this study design. Methods: A systematic review of RCTs comparing different minimally invasive procedures for TKA and comparisons of THA and hemiarthroplasty (HA). We searched MEDLINE, EMBASE, Science Citation Index, The Cochrane Library, Conference Proceedings Citation Index-Science (CPCI-S), Current Controlled Trials, and Clinical Trials.gov. Results: Seventy-five relevant RCTs were identified, 65 RCTs comparing minimally invasive with standard or other minimally invasive approaches to TKA, and 10 for THA compared with HA. Risk of bias based on the reported details of surgeons' relevant experience was categorized as low, high, or unclear. There was a clear distinction before and after 2009, with a substantial decrease in trials at high or unclear risk of bias after this date. There were no strong associations between this domain and other, standard risk of bias domains for RCTs. Conclusion: The surgeons' relevant experience in an evaluated procedure is often poorly reported but has improved since 2009. The variable is not adequately captured by any other risk of bias domain. Future work should concentrate on conducting research on a much larger sample of studies and in procedures other than knee and hip arthroplasty.