| BackgroundMinimization is a kind of dynamic randomization method for clinical trials. The advantages of minimization include "the balanced groups achieved with respect to both the numbers in each treatment arm and the characteristics of each group" and "the ability to incorporate more prognostic factors than for stratified randomization, which is particularly valuable in smaller sample trials". Much recommended as it is by researchers, there has been little report on the application of minimization in randomized trials probably due to "controversy surrounding the validity of conventional analyses following minimization", "selection bias due to the fact that next assignment can be predicted in some situations" and "the additional organizational complexity with the potential to harm recruitment and increase costs". However, both the statistical and clinical validities of minimization have been demonstrated in recent studies. Some random allocation software package or system using minimization methods had been developed to facilitate its application,... |
