Reviews and Notes: Differential Diagnosis

Differential Diagnosis

Jeremiah A. Barondess and Charles C. J. Carpenter; eds. 993 pages. Philadelphia: Lea & Febiger; 1994. $75.00.

This multiauthor text begins with an introductory chapter on the basic importance of accurate and detailed clinical data collection during the medical interview and physical examination. Concern is raised over the excessive use of laboratory and imaging procedures as a substitute for bedside examination and reasoning, and several methods for organizing complex patient data into a few diagnostic possibilities are outlined. Dr. Harold Sox describes several mathematical aids for diagnostic decision making and explains them so that they can be understood and used even by those who shy away from mathematical methods in medicine. Sox includes concepts such as odds, probability, pretest odds, use of clinical prediction rules, receiver operating characteristic curves, and the useful odds ratio form of Bayes theorem. Because of the introductory nature of this chapter, the problems related to obtaining accurate data for Bayesian analysis are not addressed.

The remaining 18 chapters deal with the differential diagnosis of individual symptoms (such as chest pain); symptom clusters (congestive heart failure and embolic syndromes, for example); signs (such as fever, lymphadenopathy, and splenomegaly); organ system dysfunctions, including liver disease and renal disease, and laboratory abnormalities (such as anemia and thrombocytopenia). The initial portion of each chapter lists and discusses both common and rare causes of symptoms and signs related to the topic of that chapter. Difference sets between listed disorders are not always explicitly presented. Each diagnostic entity is described in terms of symptoms, signs, and laboratory and imaging abnormalities. In some cases, tables listing diagnostic possibilities are included.

A unique feature is the re-publication of several cases from The New England Journal of Medicine at the end of each chapter. The clear and concise discussion of pathophysiology and diagnostic possibilities in each chapter fortifies the reader to tackle the usually complex cases that follow, and the excellent discussions provided on differential diagnosis in the case reports of the journal allow readers to test as well as to extend their knowledge base.

I had expected more use of diagnostic criteria, prediction rules, and likelihood ratios used to differentiate one entity from another, but these techniques were only sporadically discussed throughout the book. Such quantitative approaches are absent in the case reports as well and may reflect limited and unreliable data in the literature. I found the writing clear and easy to read, and I believe the organization provides an excellent way to teach medicine to students, residents, and even practitioners who want to update their knowledge of a specific topic. The text could also be used to generate a list of possible causes of a symptom, sign, or laboratory abnormality, but it is not organized for rapid implementation of a strategy for a specific differential diagnosis unless the reader is already familiar with the relevant chapter. While The New England Journal of Medicine case reports provide outstanding discussions of differential diagnosis, the clinical descriptions of the patient illnesses are often sketchy, and the data provided are weighted heavily with imaging and laboratory results, a major concern voiced in the introductory chapter. All negatives aside, this book is full of valuable information on common problems in internal medicine and is likely to markedly expand the knowledge base and diagnostic skills of the reader at every level of clinical training and practice.

Stanley L. Wiener, MD