Inference on the average treatment effect under minimization and other covariate-adaptive randomization methods

Summary Covariate-adaptive randomization schemes such as minimization and stratified permuted blocks are often applied in clinical trials to balance treatment assignments across prognostic factors. The existing theory for inference after covariate-adaptive is mostly limited situations where a correct model between the response covariates can be specified or method has well-understood properties. Based on stratification with covariate levels utilized further adjustment not used randomization, we propose several model-free estimators of average effect. We establish asymptotic normality proposed under all popular schemes, including method, show that distributions invariant respect methods. Consistent variance constructed inference. Asymptotic relative efficiencies finite-sample properties also studied. recommend using one our valid randomization.