ABSTRACT
Department of Public Health Sciences, Medical University of South Carolina
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 5.2 Allocation Randomness and Concealment . . . . . . . . . . . . . . . . . . . . . . . 118 5.3 Restricted Randomization Designs for Stratified Randomization 119 5.4 The Minimal Sufficient Balance Randomization . . . . . . . . . . . . . . . . . 121
5.4.1 The Concept of the Minimal Sufficient Balance . . . . . . . . 122 5.4.2 The Implementation of the Minimal Sufficient Balance 123 5.4.3 An Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
5.5 Balancing versus Adjustment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Covariate-adaptive randomization has been widely used in clinical trials in order to minimize biases due to covariate imbalances [17], and to increase the efficiency of the trial [15]. It is important to realize that within a treatment allocation algorithm, benefits in baseline covariate balancing are always associated with costs in allocation randomness [30]. As briefly described in Chapter 1 Section 1.3 of the current volume, there are three major types of covariate-adaptive randomization designs: stratified randomization [27], the minimization method [16, 24], and the model-based optimal design procedures [3]. Balancing covariates is the primary, if not the sole, objective for the minimization method and the model-based optimal design procedure. It is arguable to classify the minimization and the model-based optimal design as covariate-adaptive randomization procedures, because they do not include the consideration of treatment allocation randomness.
