This book is a revised and substantially expanded successor to Causal Relationships in Medicine: A Practical System for Critical Appraisal, published by Oxford in 1988. This new edition has been thoroughly rewritten and new sections on meta-analysis, evidence-based medicine and misclassification have been added. The appraisal system proposed here is applicable to studies involving randomized clinical trials, and epidemiologic studies of the survey, cohort, and case control design. It covers all major study designs used in clinical medicine and public health to assess the etiology of conditions, the efficacy of treatment, and the outcomes of health management.
The book presents a logical system of critical appraisal, from the description of the nature of the study and its results to the consideration of observational bias, the effects of extraneous factors, chance variation, the types of positive evidence expected where causal relationships exist and the issues of generalization of the results. Each chapter concludes with a summary and a set of self-test questions, with answer key. The book has two unique features. One is the presentation of six important studies illustrating randomized trials, prospective and retrospective cohort designs and case control designs, each with a critical appraisal. The other is the summary of major statistical methods used in the analysis of each study design in a series of logical tables with worked examples.
Reviewer:James C. Torner, MS, PhD(University of Iowa College of Public Health)
Description:This book provides a framework for epidemiological studies and clinical trials and then describes how studies are combined for synthesis and policy integration. The last section has several examples of published articles with a detailed evaluation of each based upon the prior chapters of the book. The previous edition was published in 1998.
Purpose:The book establishes a knowledge base of the principles of epidemiological studies and clinical trials and the design and methods. The organization begins with the concept of causation and how study design assesses it. Designs and methods are used to illustrate the contribution of each. Useful sections on bias and confounding help readers understand the caveats of observational and experimental studies. Lastly, the book establishes criteria and provides examples to use the criteria in evaluation. The premise of the book is laudable in that critical thinking and evaluation of peer-reviewed literature is imperative in the training of public health and clinical researchers. In general, the book lays a foundation, establishes the issues of bias and bias control, and provides a framework for evaluation with examples. Other authors may have different criteria, but the criteria in this book are useful.
Audience:The first eight chapters cover basic concept, design, and methods used in observational research. For someone new to the area this is a useful background. Chapters 9-15 are the real substance of the critical appraisal. Hence, for those familiar with design and methods, these chapters will be most useful. This book will be more useful to healthcare practitioners and policy makers who need to evaluate the quality of the literature. Health science students may also benefit from understanding the critical thinking needed to assess the evidence. The author addresses several concepts and methods issues and provides his own assessment of critical evaluation.
Features:The book begins with a discussion of causation, followed by epidemiologic measures and then design. Study designs are woven throughout and then specific chapters on design are described. Chapters on bias, confounding, and control of confounding provide for a logical assessment of the topics. Statistical methods are discussed, then some key methods of comparison. However, the depth should not scare off the non-statistician. Useful chapters on synthesis, including systematic reviews and meta-analysis are included. The key to the book is chapter 9 where the author details the 20 questions he uses for evaluation. The remaining chapters give examples of published studies by study design. The book is full of examples and generally they are useful in supporting the concept or method. The organization is somewhat repetitive and it is not clear why certain sections are included. The progression of topics is not traditional in epidemiological education, but would be suitable for those with a clinical or health policy background.
Assessment:This book provides a useful discussion of critical appraisal. There are a few books that directly address this issue. In general, this would be a useful adjunct in health science education. Its utility is really in chapters 9-15. This edition does provide a new and updated view of critical appraisal with useful examples.