- Shopping Bag ( 0 items )
Doody's Review ServiceReviewer: Sharon M. Homan, PhD (Kansas Health Institute)
Description: Since the publication of the first edition of this book in 1999, there have been significant advances in statistical methods and software applications to analyzing survival data in medical, epidemiological, and other health-related research. The approach taken in this second edition is similar to that in the first edition, but it expands coverage of interactions and covariate-adjusted survival functions, variable selection, time-dependent covariates, parametric regression models, competing risk models, and missing data methods.
Purpose: The authors emphasize practical and contemporary applications of survival data regression modeling. Their goal is to "provide a focused text on regression modeling for the time-to-event data typically encountered in health related studies." Researchers and analysts increasingly work with time-to-event data, and have many more software options to perform analyses. Choosing among various modeling and software options is critical to developing a good analysis plan and conducting optimal analyses. This book is a great resource, with its clear and accessible presentation of modeling techniques, case studies, and practical applications.
Audience: It is intended as a textbook for graduate courses in biostatistics, statistics, and epidemiologic methods, and as a reference for practitioners and researchers in health-related fields who have a foundation in linear and logistic regression methods. The authors are highly regarded biostatisticians, with enormous expertise in the development, application, and teaching of survival data analysis methods.
Features: The book begins by building a mathematically sound and practical foundation of survival data, survival time functions, rand egression modeling. Next, the proportional hazards model is developed - estimation, model fitting, and model adequacy. Extensions of the proportional hazards model include stratified models, time-varying covariates, truncation and censoring. The final chapters cover parametric modeling, recurrent event and frailty models, and competing risk models. One of the best features of this book is the presentation of well-developed applications and excellent examples and illustrations of analyses using Stata 9.0 and SAS 9.1. Challenging exercises end each chapter, but notes and/or solutions are not provided. Data sets can be downloaded from the publisher's web site. Adding the programming code for the examples and chapter exercises in an appendix on the web site would be even better.
Assessment: This is a superb resource - a practical guide with up-to-date applications. The authors are excellent teachers of the mathematics and application of survival data regression modeling including how to handle complexities such as time-varying covariates and correlated observations. The second edition is a significant revision, incorporating the new capabilities of Stata, easy-to-use software with good graphical and statistical analysis capabilities, and introducing contemporary applications of regression modeling of time-to-event data.