Success Probability Estimation with Applications to Clinical Trials

PART I

SUCCESS PROBABILITY ESTIMATION IN PLANNING AND ANALYZING CLINICAL TRIALS

The first part of the book presents the concepts related to success probability estimation and their usefulness in clinical trials. Part I is devoted to both statisticians and non-statisticians: each Section of the first four Chapters begins with a boxed text explaining, in non-technical and non-statistical language, the concepts contained in the Section.

Chapter 1 illustrates the basic statistical tools and introduces the success probability (SP). In Chapter 2 the SP of a new trial with identical settings of the one previously performed is studied; this probability assumes the meaning of reproducibility probability (RP). The evaluation of the stability of the results of statistical tests is presented, together with the important new tool of RP-testing, which allows statistical testing only on the basis of the estimate of the RP. In Chapter 3 the use of pilot data for estimating SP and then the sample size of a subsequent trial is studied. In other words, the problem of planning clinical trials on the basis of pilot samples is discussed. The important concepts of launching criteria are introduced, and the variability of pilot data is taken into account, leading to the so-called conservative sample size estimation (CSSE). Several CSSE strategies are introduced and their performances are compared. In Chapter 4 the robustness of CSSE strategies is evaluated, and some correction techniques are presented ...