A SAS Macro for Covariate-Constrained Randomization of General Cluster-Randomized and Unstratified Designs.

摘要: Ivers et al. (2012) have recently stressed the importance to both statistical power and face validity of balancing allocations study arms on relevant covariates. While several techniques exist (e.g., minimization, pair-matching, stratification), covariate-constrained randomization (CCR) approach proposed by Moulton (2004) is favored when clusters can be recruited prior randomization. CCRA V1.0, a macro published Chaudhary (2006), provides SAS implementation CCR for particular subset possible designs (those with two arms, small numbers strata clusters, an equal number within each stratum, constraints that expressed as absolute mean differences between arms). This paper presents more comprehensive macro, CCR, applicable across wider variety statistics describing range meeting in addition performing actual random assignment.