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

Chapter 21

Composite Endpoints in Clinical Trials

Peter C. O’Brien, Peter J. Dyck, and Barbara C. Tilley

21.1 Introduction

Composite endpoints have become increasingly useful as a measure of a patient’s disease status. They have been found useful in both clinical practice and medical research. Here, we shall examine the usefulness of composite endpoints as a primary endpoint in clinical trials. The basic rationale for their use is discussed and illustrated with examples, followed by a variety of procedures for forming composites including how they should be interpreted. In the section on global assessment variables, alternate approaches to conducting a primary analysis with multiple endpoints are considered and these methods are compared with the use of composite endpoints.

21.2 The Rationale for Composite Endpoints

We consider the outcome measures used in the typical phase III clinical trial for efficacy, for simplicity assuming that an experimental therapy is being compared with a placebo. Most commonly, a protocol for such a study specifies a single primary endpoint (see the discussion of primary endpoints), various secondary endpoints (see the discussion of secondary endpoints), and often additional tertiary endpoints. The goal of categorizing study endpoints in this manner is to control the type I error rate (see the discussion of type I error). Thus, the study is judged to have achieved statistically significant evidence of efficacy and to be a positive trial only if the ...