Many cluster randomized trials have relatively few numbers of clusters to be randomized. When baseline cluster-level covariates are available prior to randomization, the set of potential allocations can be restricted so as to ensure balance across study arms. This article discusses why and how restrictions can be made, and the ramifications of so doing. The Fisher–Bailey validity is explained, and examples are given regarding the tradeoff between balance and validity.
RoyallRM. Current advances in sampling theory: implications for human observational studies. Am J Epidemiol1976; 104(4): 463–474.
3.
BaileyRA. Restricted randomization: a practical example. J Am Stat Assoc1987; 82: 712–719.
4.
BaileyRARowleyCA. Valid randomization. Proc R Soc Lon Ser-A1987; 410: 105–124.
5.
LiFTurnerELHeagertyPJ, et al. An evaluation of constrained randomization for the design and analysis of group-randomized trials with binary outcomes. Stat Med2017; 36(24): 3791–3806.
ChaudharyMAMoultonLH. A SAS macro for constrained randomization of group-randomized designs. Comput Methods Programs Biomed2006; 83(3): 205–210.
8.
GreeneEJ. A SAS macro for covariate-constrained randomization of general cluster-randomized and unstratified designs. J Stat Softw2017; 77(CS1): 1–20.
9.
LorenzEGabryschS. Covariate-constrained randomization routine for achieving baseline balance in cluster-randomized trials. Stata J2017; 17(2): 503–510.
10.
YuHLiFGallisJA, et al. Cvcrand: a package for covariate-constrained randomization and the clustered permutation test for cluster randomized trials. R Journal2017; 11(2): 191–204.
11.
RaabGMButcherI. Balance in cluster randomized trials. Stat Med2001; 20(3): 351–365.
12.
MorganKLRubinDB. Rerandomization to improve covariate balance in experiments. Ann Stat2012; 40: 1263–1282.
BertsimasDJohnsonMKallusN. The power of optimization over randomization in designing experiments involving small samples. Oper Res2015; 63(4): 868–876.
15.
KapelnerAKriegerAMSklarM, et al. Harmonizing optimized designs with classic randomization in experiments. Am Stat2021; 75(2): 195–206.
16.
KasyM. Why experimenters might not always want to randomize, and what they could do instead. Polit Anal2016; 24(3): 324–338.
17.
De HoopETeerenstraSvan GaalBG, et al. The “best balance” allocation led to optimal balance in cluster-controlled trials. J Clin Epidemiol2012; 65(2): 132–137.
18.
PerryMFaesMReelickMF, et al. Studywise minimization: a treatment allocation method that improves balance among treatment groups and makes allocation unpredictable. Clin Epidemiol2010; 63(10): 1118–1122.
19.
MoultonLHGolubJEDurovniB, et al. Statistical design of THRio: a phased implementation clinic-randomized study of a tuberculosis preventive therapy intervention. Clin Trials2007; 4(2): 190–199.
