Methods and Applications of Statistics in Clinical Trials, Volume 1: Concepts, Principles, Trials, and Designs

Chapter 71

Simple Randomization

Inna T. Perevozskaya

71.1 Introduction

The objective of any clinical trial is to provide an unbiased estimate of treatment effect, usually measured by comparison of the two treatment groups (e.g., experimental treatment versus conventional treatment) with respect to the endpoint of interest. Although there are several potential sources of bias for such an estimate, this articles focuses on one in particular, namely, lack of comparability of the treatment groups with respect to known or unknown prognostics factors (covariates). Randomization mitigates the risk of such an imbalance with respect to important covariates and thus addresses the source of bias arising from patient selection. However, randomization alone is not sufficient to completely eliminate bias. To claim that a study’s results are unbiased, additional measures may be required to alleviate concerns about other potential sources of bias such as methods of outcome assessment and missing data. (See the article on bias for more details on these additional sources.)

71.2 Concept of Randomization

Randomization is a method based on chance by which clinical trial participants are assigned to study treatment groups. Randomization minimizes the differences among groups by balancing the distribution of patients with particular characteristics (prognostic factors) among all the trial arms. In case of a two-arm trial with equal allocation to both treatment groups, each study participant has a fair ...