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

Chapter 11

Block Randomization

Damian McEntegart

11.1 Introduction

Randomization techniques can be classified as static, in that allocation is made from a randomization list that is generated before the start of the trial, or they can be dynamic, in that the allocation depends on the current balance of allocated treatments either overall or within particular sub-groups. This article deals with static randomizations from a list composed of one of more blocks. Blocked randomization is the most commonly used randomization technique as judged by reports in medical journals [1] and my own experience (80% of trials on our database of over 1500 randomized trials use blocked randomization).

Most theory and examples presented will relate to the case of randomization into two treatment groups with a desired equal allocation. This is for simplicity of exposition and everything can be generalized to cover multiple treatments and/or unequal randomization ratios.

11.2 Simple Randomization

To understand the rationale for blocked randomization and what it entails, it is first necessary to define simple randomization as a basis for comparison. Simple randomization (alternatively called complete or unconstrained randomization) occurs when each treatment assignment is made at random, and the allocation is completely in-dependent of previous allocations. It has the attractive property of complete unpredictability. One method of constructing a list in a trial of two treatments would be to toss a coin ...