Design and Analysis of Group-Randomized Trials

Abstract

Design and Analysis of Group-Randomized Trials Author: MURRAY, David M., PhD (Univ of Minnesota) Bibliographic Data: (ISBN: 0-19-512036-1, Oxford University Press Inc, 1998, $65.00) Series Title: Monographs in Epidemiology and Biostatistics, v. 27, 10 chapters, 467 pages, hard cover Audience 1: Biostatisticians Audience 2: Epidemiologists Other Info: The book contains black-and-white illustrations. DESCRIPTION: This book introduces the purpose and history of group randomized trials, then presents design considerations and analytic approaches. It also presents applications of analytic approaches in previously conducted trials. PURPOSE: The stated purpose is to help those involved with group-randomized trials do a better job of planning, reviewing, funding, conducting, analyzing, and interpreting their studies. This objective is met. AUDIENCE: Given the purpose stated, this book is intended for a wide audience, primarily researchers in both academic and non-academic centers, including analysts, graduate students, and evaluation staff involved with the trials. I believe it meets the needs of the middle group well, but may fall short, not in quality, but in detail, for those conducting analyses. Those who make funding or interpretation decisions may find the detail overwhelming. FEATURES: This book flows smoothly and both the table of the contents and the index are helpful in finding topics of specific interest. The analysis section includes information and SAS approaches to analyses being discussed, and the index includes references to SAS procedures and syntax in a distinctive font. Another useful feature is information on the strengths and weaknesses of alternative analytic approaches. ASSESSMENT: This book is an important addition to any public health or medical library. It is well-written and much needed. Too often, researchers conducting group-randomized trials are forced to mix and match information from clinical trials and social sciences literature. Several reviews cited point out the inherent difficulties, listing weakness in design and analysis in many published group-randomization trials. This book provides comprehensive information on the increasingly common and important method of group-randomization. Analysts will need to supplement this book, but it will be a valuable resource for anyone interested in or involved with group-randomized trials. SCORE: Weighted Numerical Score: 92 - **** Reviewed by: Seema Sonnad, PhD (University of Michigan Medical Center)