Combining Covariate Adjustment with Group Sequential and Information Adaptive Designs to Improve Randomized Trial Efficiency
share
In clinical trials, there is potential to improve precision and reduce the required sample size by appropriately adjusting for baseline variables in the statistical analysis. This is called covariate adjustment. Despite recommendations by the U.S. Food and Drug Administration and the European Medicines Agency in favor of covariate adjustment, it remains underutilized leading to inefficient trials. We address two obstacles that make it challenging to use covariate adjustment. A first obstacle is the incompatibility of many covariate adjusted estimators with commonly used stopping boundaries in group sequential designs (GSDs). A second obstacle is the uncertainty at the design stage about how much precision gain will result from covariate adjustment; an incorrect projection of a covariate's prognostic value risks an over- or underpowered trial. To address these obstacles, we extend the theory of information-monitoring in GSDs to handle covariate adjusted estimators. In particular, we propose a new statistical method that modifies the original estimator so that it becomes compatible with GSDs, while increasing or leaving unchanged the estimator's precision. This is needed since many covariate adjusted estimators don't satisfy the key property (i.e., independent information increments) needed to apply commonly used stopping boundaries in GSDs. Our approach allows the use of any asymptotically linear estimator, which covers many estimators used in randomized trials. Building on this, we propose using an information adaptive design, that is, continuing the trial until the required information level is achieved. Such a design adapts to the amount of precision gain due to covariate adjustment, resulting in trials that are correctly powered and that fully leverage prognostic baseline variables; this can lead to faster, more efficient trials, without sacrificing validity or power.
READ FULL TEXT
