The Statistical Evaluation of Medical Tests for Classification and Prediction (Oxford Statistical Science Series) [Hardcover]

Margaret Sullivan Pepe (Author)

Book Description

ISBN-10: 0198509847 | ISBN-13: 978-0198509844 | Publication Date: May 29, 2003 | Edition: 1

This book describes statistical concepts and techniques for evaluating medical diagnostic tests and biomarkers for detecting disease. More generally, the techniques pertain to the statistical classification problem for predicting a dichotomous outcome. Measures for quantifying test accuracy are described including sensitivity, specificity, predictive values, diagnostic likelihood ratios and the Receiver Operating Characteristic Curve that is commonly used for continuous and ordinal valued tests. Statistical procedures are presented for estimating and comparing them. Regression frameworks for assessing factors that influence test accuracy and for comparing tests while adjusting for such factors are presented. This book presents many worked examples of real data and should be of interest to practicing statisticians or quantitative researchers involved in the development of tests for classification or prediction in medicine.

Editorial Reviews

Review

It is a clear and authorative work which will be invaluable for anyone trying to establish a new diagnostic test. Derek Robinson, University of Sussex, Brighton Statistical results are given a methodical treatment. In each chapter statistical results are motivated, stated and usually proven, and then illustrated on a variety of datasets based on actual trials. Many of the results were developed by Pepe and her colleagues, who have advanced the field in seminal ways. Chapters end with a concluding remarks section, a set of exercises, and proofs of more involved theoretical results, when needed. Concluding remarks sections nicely summarize results and discuss open research questions. References to key papers are given throughout. This structure allows the book to serve as both a classroom text and an excellent reference to the current literature. Journal of Biopharmaceutical Studies I very much recommend this book to a range of audiences, from those wanting an introduction to diagnostic test concepts and methods to active researchers in the area. This book is likely to stimulate considerable progress in development of new statistical methods for diagnostic tests, an area that relative to therapeutics has received little attention from biostatisticians. Journal of Biopharmaceutical Studies

About the Author

Margaret Sullivan Pepe is a Professor of Biostatistics, University of Washington and Fred Hutchinson Cancer Research Center, Washington, USA.

Product Details

• Hardcover: 320 pages
• Publisher: Oxford University Press, USA; 1 edition (May 29, 2003)
• Language: English
• ISBN-10: 0198509847
• ISBN-13: 978-0198509844
• Product Dimensions: 5.6 x 5.3 x 0.4 inches
• Shipping Weight: 1.2 pounds
• Average Customer Review: 4.5 out of 5 stars  See all reviews (2 customer reviews)
• Amazon Best Sellers Rank: #2,436,802 in Books (See Top 100 in Books)

More About the Author

Discover books, learn about writers, read author blogs, and more.

Customer Reviews

Most Helpful Customer Reviews

34 of 34 people found the following review helpful:

5.0 out of 5 stars well written book on an important topic, June 8, 2008

By 

Michael R. Chernick "statman31147" (Holland PA) - See all my reviews
(REAL NAME)   

This review is from: The Statistical Evaluation of Medical Tests for Classification and Prediction (Oxford Statistical Science Series) (Paperback)

With the increase in cancer trials both in medical research and the pharmaceutical industry, medical diagnostic tests including medical imaging evaluations are being used more and more. There is a need for adjudication when radiologists disagree on a diagnosis and some standard statistical measures from other fields are finding new application.

This book provides a thorough background on the subject, the methodology and the applications. It is very clearly written and not overly technical. The methodology is the classical frequentist approach to statistical inference. Recently, Lyle Broemeling has published a book on this topic that takes exclusively the Bayesian approach and explains the Bayesian approach for those who are not acquainted with it. When dealing with predictions based on evolving data the Bayesian approach might be more natural. This would be the case in a clinical trial where the data is reviewed sequentially.

A particularly important part of diagnostic accuracy analysis is the area under the ROC curve. Pepe does an excellent job of covering that topic. This text is referenced often in Broemeling's book since it provides an excellent explanation of the same topics using the frequentist approach that he covers with the Bayesian approach.

7 of 8 people found the following review helpful:

4.0 out of 5 stars Review, August 27, 2007

By 

M. Inguanez (MALTA, Europe) - See all my reviews
(REAL NAME)   

This review is from: The Statistical Evaluation of Medical Tests for Classification and Prediction (Oxford Statistical Science Series) (Paperback)

An excellent book on the evaluation of medical tests. Includes a lot of information on techniques used for evaluating diagnostic tests. Easily read even by persons with no medical background.