This book is unique in that it presents the mathematical theory and SAS programming examples side by side. Since SAS is popular software for data analysis in pharmaceutical industry, this book is a useful reference for people using SAS to report clinical trials data.
There are significant changes that are implemented in this second edition. A new feature includes the topics based on the stages (early or late) of clinical trials, which is beneficial to statisticians working in clinical trials for regulatory submissions. The topics in this edition are expanded to cover new approaches addressing statistical problems that were introduced in the first edition. These new approaches are the recent developmental research in clinical trials.
There are 7 chapters in this second edition which are grouped into three parts:
The first part (Chapters 1 and 2) presents the general statistical methods used at all stages of drug development. It contains the same content as the first edition with additional revisions.
The second part (Chapters 3 and 4) contains information not found in the first edition. It focuses on statistical methods in dose-escalation in Phase I (Chapter 3) and dose-finding in Phase II clinical trials (Chapter 4). The most attentions in does-escalation designs are given to continual reassessment method with examples from oncology trials. For dose-finding methods, this book confers the pairwise contrast-based and multi-contrast tests for dose-finding algorithms based on the MCP-Mod procedure.
The third part (Chapters 5-7) describes the statistical methods in late phase clinical trials. It is an extension of the three chapters (multiplicity adjustment method, interim data analysis, and incomplete data) from the first edition. More specifically, the extension of these three chapters include new macro to support gatekeeping procedures (Chapter 5), SAS procedures (PROC SEQDESIGN and PROC SEQTEST) that support a broad class of group-sequential designs (Chapter 6), PROC GEE to support weighted generalized estimating equations analyses, and PROC MI to run MNAR assumption for missing data (Chapter 7).
One thing to note is that the complicated methods used in this second edition rely heavily on SAS macros, which cannot be run through standard SAS procedures directly. Instead, you will need to download the macros through the designated websites.
Overall, this book is well balanced since it contains not only the SAS applications but also the theories behind statistical methods. It is a notable addition to the growing collection of Design and Analysis of Clinical Trials books already published.
Annpey Pong, PhD.
Merck Research Laboratories,
Rahway, New Jersey, USA