Block randomization in clinical trials

Abstract

Randomization is the key to ensure the balance of confounders between the comparison groups in clinical trials, and it is the statistical basis for making the study results comparable. A simple randomization in a clinical trial with large sample-size allows for a balanced comparison of the number of subjects and confounding factors between groups, but in a clinical trial with small sample-size, it is necessary to use a restricted randomization method (the blocked randomization). Block randomization ensures that the number of subjects between groups is basically equal, maximize the effectiveness of clinical trials as the standard error of the treatment-effect estimate is decreased, which affords big rewards in scientific accuracy and credibility.