Statistical Issues in the Design and Analysis of Clinical Trials

Chapters 1-4 concern statistical methods in designing and analyzing data for survival clinical trials, and predicting trial duration. Chapters 5-7 concern statistical methods in clinical trials with sequential parallel designs, which have been proposed for trials with high placebo response rates which can lead to a higher failure rate in drug development. Utilizing Box and Muller (1958), one of the most popular methods of generating standard normal random variable using two independent uniform (0, 1) deviates, a new method is proposed in Chapter 8 to combine two p-values from two disjoint samples for designing a trial with two stages.