Chapter 7 extends survival analysis methods to a class of survival models, called parametric models, in which the distribution of the outcome (i.e., the time to event) is specified in terms of unknown parameters. Many such parametric models are acceleration failure time models, which provide an alternative measure to the hazard ratio called the "acceleration factor". The general form of the likelihood for a parametric model that allows for left, right, or interval censored data is also described. The chapter concludes with an introduction to frailty models.

Chapter 8 considers survival events that may occur more than once over the follow-up time for a given subject. Such events are called "recurrent events". Analysis of such data can be carried out using a Cox PH model with the data layout augmented so that each subject has a line of data for each recurrent event. A variation of this approach uses a stratified Cox PH model, which stratifies on the order in which recurrent events occur. The use of "robust variance estimates" are recommended to adjust the variances of estimated model coefficients for correlation among recurrent events on the same subject.

Chapter 9 considers survival data in which each subject can experience only one of several different types of events ("competing risks") over follow-up. Modeling such data can be carried out using a Cox model, a parametric survival model or a model which uses cumulative incidence (rather than survival).

The Computer Appendix in the first edition of this text has now been revised and extended to provide step-by-step instructions for using the computer packages STATA (version 7.0), SAS (version 8.2), and SPSS (version 11.5) to carry out the survival analyses presented in the main text. These computer packages are described in separate self- contained sections of the Computer Appendix, with the analysis of the same datasets illustrated in each section. The SPIDA package used in the first edition is no longer active and has therefore been omitted from the appendix and computer output in the main text.

In addition to the above new material, the original six chapters have been modified slightly to correct for errata in the first edition, to clarify certain issues, and to add theoretical background, particularly regarding the formulation of the (partial) likelihood functions for the Cox PH (Chapter 3) and extended Cox (Chapter 6) models.