This is the second edition of a classic book on survival analysis. The first edition appeared in 1982, see the review Zbl 0541.62081. The field of survival analysis has developed enormously since then, and this fact is partly reflected in, and a motivation for, this second edition. The first edition was less inclined to medical applications than contemporary books in the field, and the author is following up on this line in the present edition. However, this does not mean that medical applications are absent, the author’s aim is to cover lifetime data analysis without concentrating exclusively on any field of applications. He succeeds quite well, but most examples and applications come from the engineering and biomedical fields. Another characteristic of this book, compared to other books on survival analysis, is the generous space given to fully parametric models.
A summary of the contents would be as follows: After an introductory chapter, where some basic concepts and some life time distributions are presented, the next two chapters discuss censoring and truncation, and standard nonparametric graphical methods. The connection to standard life table techniques is discussed. Then, in three chapters, the theory for parametric survival analysis is given, including a chapter on the fundamental families of log-location-scale distributions. Two chapters deal with semi-parametric survival analysis, essentially Cox’s regression model. The book concludes with some specialized topics: multiple modes of failure (essentially the same as competing risks), goodness-of fit tests, and multivariate survival analysis. Each chapter ends with a section of exercises. Compared to the first edition, this book is less focused on computational details and more on using available software for illustration of the methods by examples.
There are no less than seven appendices to the book. Six of them give more background details for interested readers, and one contains data sets, some old and some newer.