The international call to practice evidence-based medicine was first published in JAMA as part of the “Rational Clinical Examination Series”.¹ Almost 10 years and a World Wide Web later, we now have McGee's book Evidence-Based Physical Diagnosis. The book's intent is “to explore the origins, pathophysiology, and diagnostic accuracy of many of the physical signs used today in adult patients.” The book presents an amalgam of expert opinion and peer-reviewed evidence for what appears to be a serendipitous collage of various topics on the clinical examination. Because the current state of knowledge about the clinical examination lacks the breadth of well-done trials found in treatment research and therefore results in this eclectic mix, few individuals have the temerity to take on such a project by themselves.

McGee S.Evidence-Based Physical Diagnosis. Philadelphia: Saunders, 2001.

Although not peer reviewed, the text seems to be similar in quality to those manuscripts published as part of the “Rational Clinical Examination Series” and will appeal to the same audience, including generalist physicians, clinical diagnosis instructors, resident physicians, and medical students. The method for compiling the necessary resources was not explicitly stated in this book. It is implicitly apparent that collecting the references cited required years of work. I have the impression that the topics covered in this book were first put together from a wise physician's “files of great articles” collected over many years, supplemented by present-day computer database searches for additional evidence. To be included in the evidence tables in this book, studies of adult patients had to meet 4 criteria: no asymptomatic controls so that the study population included patients for whom the condition was being considered; definitions for the physical signs; independent comparisons to a diagnostic standard; and data that allowed construction of 2 × 2 tables with likelihood ratios. Wherever possible, McGee presents summary measures for diagnostic accuracy as likelihood ratios to permit the reader to quickly categorise physical findings as “useful” versus “useless” tests. For those of us who bemoan the difficulties in finding such data, the evidence tables are reason enough to peruse the text.

The coverage of topics is interesting and provides as much insight into the interests of the author as it does the physical examination. Some of the chapters are full of such historical items as an entire page on Sister Mary Joseph's (née Julia Dempsey) nodule that make for interesting reading but not much evidence. A lengthy chapter on gait assessment contains much more information than is available in standard physical examination texts, but little evidence beyond expert opinion despite the 53 included references. If anything, the book may be “over-referenced” because few of the cited articles actually meet the criteria for inclusion in the evidence tables. The eclectic coverage of topics makes the book unsuitable as a single physical examination book for medical students or physicians. I believe it is strictly a supplementary text that will enrich the more standard, comprehensive texts used in most medical schools. Readers looking for quantitative evidence about the physical examination can concentrate on the references included in the tables and summarised in the final chapter.

As a resource that could be used in real time on a medical ward or in a clinic to rapidly find specific answers, the book is difficult to use because, after all, it is only a book. The information is currently not available in electronic format, so those desiring the data will either need the actual text or will need to extract the information from the last chapter for their personal digital assistants, or “peripheral brains”, carried in their coat pocket. Because of this limitation, I think it will be difficult to use this book as a frequent reference for answering questions about diagnosis on individual patients; however, while perusing this book for 2 months, I found myself often reviewing it to pull out snippets of information that later proved useful. It is a good read.