Real-time Adaptive Randomization of Clinical Trials

Abstract

Listen

Translate article

ObjectiveTo evaluate real-time (day-to-day) adaptation of randomized controlled clinical trials (RCTs) with delayed endpoints – a “forward-looking optimal-experimentation” form of response-adaptive randomization (RAR). To identify the implied tradeoffs between lowered mortality, confidence intervals, statistical power, potential arm misidentification, and endpoint-rate change during the trial. Study Design and SettingUsing data from RCTs in acute myocardial infarction (30,732 patients in GUSTO-1) and coronary heart disease (12,218 patients in EUROPA), we resample treatment-arm assignments and expected endpoints to simulate (1) real-time assignment, (2) forward-looking assignments adapted after observing a fixed number of patients (“blocks”), and (3) a variant that balances RCT and real-time assignments. Blinded RTARs adjust day-to-day arm assignments by optimizing the tradeoff between assigning the (likely) best treatment and learning about endpoint rates for future assignments. ResultsDespite delays in endpoints, real-time assignment quickly learns which arm is superior. In the simulations, by the end of the trials, real-time assignment allocated more patients to the superior arm and fewer patients to the inferior arm(s) resulting in fewer mortalities over the course of the trial. Endpoint rates and odds ratios were well within (resampling) confidence intervals of the RCTs, but with tighter confidence intervals on the superior arm and less-tight confidence intervals on the inferior arm(s) and the odds ratios. The variant and patient-block-based adaptation each provide intermediate levels of benefits and costs. When endpoint rates change within a trial, real-time assignment improves estimation of the end-of-trial superior-arm endpoint rates, but exaggerates differences relative to inferior arms. Unlike most RARs, real-time assignment automatically adjusts to reduce biases when real changes are larger. ConclusionReal-time assignment improves patient outcomes within the trial and narrows the confidence interval for the superior arm. Benefits are balanced with wider confidence intervals on inferior arms and odds ratios. Forward-looking variants provide intermediate benefits and costs. In no simulations, was an inferior arm identified as statistically superior.