Accounting for Inconsistent Use of Covariate Adjustment in Group Sequential Trials

Group sequential designs are common in clinical trials, as they allow for interim efficacy and futility monitoring. Adjustment for baseline covariates can increase power and precision of estimated effects in clinical trials. Statistical methods for both group sequential trials and covariate adjustment are well-developed, including when covariate adjustment is used in a group sequential trial. What is less studied is how to perform interim monitoring, estimation, and inference in group sequential trials where covariate adjustment is applied inconsistently throughout the trial (e.g., if there is a delay in covariate data being made available). We show how to bridge this gap by studying the asymptotic behavior of test statistics obtained from a group sequential trial using ANOVA versus ANCOVA. We focus on two-arm trials with simple, balanced randomization and continuous outcomes. We study the performance of our boundary, estimation, and inference adjustments in simulation studies. We end with recommendations about the application of covariate adjustment in group sequential designs.