Partial randomization design in a support group intervention study

Randomized clinical trial (RCT) designs, in which participants are randomly assigned to treatment and control groups, are considered the gold standard of experimental design. However, if potential participants have a strong preference for a particular study group and/or when participatory effort must be sustained over a lengthy study time frame, they may deem randomization unacceptable, refuse to be in the study, and hence prevent the purpose of randomization from being realized. This article describes several alternative designs and the decision-making process prior to selecting a partially randomized preference trial (PRPT) design to accommodate preferences of women with breast cancer for participation in a support group intervention pilot study. Recruiting women into the two randomized arms of the PRPT was problematic because most potential participants expressed a preference for a particular study group. Possible reasons for failure to randomize more of the potential participants are discussed and suggestions for future research proposed.