Departures from community equipoise may lead to incorrect inference in randomized trials

Area of Science:

• Orthopedics
• Clinical Trial Methodology
• Biostatistics

Background:

• Randomized controlled trials (RCTs) are essential for evaluating treatment effectiveness.
• Patient enrollment criteria can influence trial outcomes.
• The impact of selective enrollment on the efficacy of arthroscopic partial meniscectomy for meniscal tears with osteoarthritis (OA) requires investigation.

Purpose of the Study:

• To assess how selective enrollment affects the results of randomized controlled trials (RCTs).
• To evaluate the impact of OA severity on the perceived efficacy of arthroscopic partial meniscectomy.

Main Methods:

• Simulated an RCT comparing arthroscopic partial meniscectomy to nonoperative therapy for meniscal tear and OA.
• Estimated efficacy using risk ratio (RR) under four enrollment scenarios: nonselective, mild OA bias, severe OA bias, and high severe OA bias.
• Simulated 100 trials per scenario with a sample size of 340.

Main Results:

• Nonselective enrollment yielded results consistent with the underlying population (mean RR: 1.87).
• Selective enrollment favoring severe OA cases reduced estimated surgical efficacy by 28% (mean RR: 1.34).
• In scenarios with selective enrollment, the 95% confidence intervals contained the true efficacy in only 25% of trials, with 44% empirical power.

Conclusions:

• Selective enrollment based on factors influencing efficacy can significantly alter trial results.
• Biased enrollment can lead to inaccurate conclusions regarding treatment effectiveness.
• Ensuring representative patient populations in RCTs is critical for valid scientific evidence.