The History of Surrogate Endpoint Evaluation: Single-Trial Methods


The seminal papers of Prentice (1989), Freedman et al. (1992), and Buyse and Molenberghs (1998) have set the scene for a large research line into surrogate marker evaluation methods. The methods that were proposed by these authors were based on the premise that information regarding the surrogate and the true endpoints is available from a single clinical trial. Hence, these methods are collectively referred to as single-trial surrogate evaluation methods. In this chapter, these methods are briefly reviewed.

It is important to emphasize that all single-trial methods are hampered by some fundamental theoretical and/or applied problems (see Sections below). This chapter merely reviews the single-trial methods to set the scene for the statistically more informative multiple-trial surrogate evaluation methods, which assume that the data of multiple trials or other relevant clustering units are available (see Part II of this book). The use of these single-trial methods is not recommended in practice, for reasons outlined in Section 3.5.

< Prev   CONTENTS   Source   Next >