Foreword

I remember my introduction to the medical history and clinical examination as the most exciting moments of my early career. As each item in the history and physical examination was explained and given meaning and significance, I believed that after the long preclinical years I had at last reached the threshold of becoming a physician. I could begin to hold more than a comforting conversation with a patient. I could use my ears, eyes, and hands to disclose the patient's problem and so begin to be of actual use to a real patient. As I polished my skills, it did not occur to me that the divination of all those signs and symptoms was anything but an art: the epitome of the art of medicine.

But, with time, I realized that many of the so-called pathognomonic symptoms and signs were so merely because someone, often the person whose name was attached to them, had declared that they were. Doubt started to overtake accepted wisdom as it became clear to me that little worthwhile evidence supported the artist's tools I thought I had mastered.

Towards the end of the 1980s, my friend David Sackett, then chief of medicine and clinical epidemiology and biostatistics at McMaster University, showed me a new way of thinking about all this. He equated items in the history and the physical examination with traditional diagnostic laboratory tests, each susceptible to evidentiary testing. So he and I began planning 2 series of articles on evidence-based medicine to appear in JAMA. One of these, the Users' Guides to the Medical Literature, was soon placed into the capable hands of Gordon Guyatt, also of McMaster University, and articles began to appear in JAMA in 1993. By 2002, they were printed in updated form in 2 books, an Essentials and a fuller Manual,^1,2 both of which have been so successful that second editions^3,4 have just been published.

The other series consisted of The Rational Clinical Examination articles and started appearing in 1992. With the first article, Sackett and I published an editorial.⁵ We reminded our readers of studies that showed that primary care providers usually establish the correct diagnosis at the end of a brief history and some subroutine of the physical examination. So on practical grounds alone, it made sense to improve our understanding of the parts of the history and examination that were useful, or useless, in pinning down, usually at an early stage of the disease, one diagnosis and ruling out others. We contrasted symptoms and signs with laboratory tests, which were subjected to rigorous testing before adoption, but which might have far less ability to narrow the diagnostic possibilities. As an example, we observed the overwhelming probability of coronary stenosis in a 65-year-old man who has smoked all his life when he tells you that he gets central chest tightness regularly on exertion, which forces him to stop and which disappears when he rests.^6,7

Perhaps most important, by encouraging research into the history and physical examination, we wanted to restore respectability to a part of medicine that seemed to have been eroding as academic and financial rewards went to those who most resembled scientists relying on expensive diagnostic tests and least behaved as physicians relating to patients.

It is no coincidence that both Sackett and I, authors of the editorial launching the series, have served roles in the Cochrane Collaboration, an initiative that has had a massive effect on the way we see evidence and a profound influence on the methods and popularity of systematic review and metaanalysis. These sciences, as well as that of decision making, had grown up and spread to medicine during the 1970s and 1980s. Without them, both the Cochrane Collaboration and The Rational Clinical Examination series would have been impossible undertakings; indeed, the entire evidence-based movement would have grown far more slowly.

At the same time, because of the unfamiliarity of these techniques and the revolutionary approach we were taking, namely, a scientific examination of what most clinicians considered to be an ineffable art not susceptible to dissection, we published a primer on the precision and accuracy of the clinical examination. This laid out the approach to be taken and took the reader through the terms, methods, and calculations underpinning clinical diagnosis.⁸

Although each article's purpose could be worked out from its title, the full meaning of the concepts took time to sink in, as I discovered from comments sent in by many of the expert specialty peer reviewers to whom I sent the manuscripts as they came in to JAMA. Indeed, it was unfamiliar even to some prospective authors. David Sackett had a firm belief that the reviews would be done best by generalist physicians who had learned basic critical appraisal skills. As the editor, I learned that these generalist physicians were often speaking a different language from our specialist reviewers. Sackett was clearly correct, and it remains commonplace for specialty reviewers to ask that specialists be added to the writing team because, well, they are specialists. What has happened in our process is that both authors and reviewers learn from the editorial review process, with specialty reviewers ensuring that authors interpret the data in the proper context. In return, the specialists often learn that much of what they took for granted has no basis in evidence.

The Rational Clinical Examination book should not replace books on clinical diagnosis. But, somewhat as the Cochrane Database of Systematic Reviews provides a systematic evaluation of all studies on a particular intervention without becoming prescriptive, so articles in The Rational Clinical Examination series are careful systematic efforts to assess the accuracy of items from the patient's medical history and the clinical examination. In this sense, they are a revolutionary departure from what we have regarded as books on physical diagnosis, which, until the first had never taken that approach. Since then, however, such books have already started using the evidence as summarized in articles in the series. In his preface to articles in The Rational Clinical Examination series appeared, had never taken that approach. Since then, however, such books have already started using the evidence as summarized in articles in the series.

In his preface to the eighth edition of DeGowin's Diagnostic Examination, Richard LeBlond writes:

References to articles from the medical literature are included in the body of the text. We have chosen articles which provide useful clinical information including excellent descriptions of disease and syndromes and, in some cases, photographs illustrating key findings. Evidence-based articles on the utility of the physical exam are included, mostly from The Rational Clinical Examination series published over the last decade in the Journal of the American Medical Association. They are included with the caveat that they evaluate the physical exam as a hypothesis-testing tool, not as a hypothesis generating task. …⁹

Our series is indeed about testing tests (symptoms, signs) to separate the useful from the useless and so is about testing hypotheses. Books on physical diagnosis are hypothesis generating in that they are a compendium of instructions on how to elicit all symptoms and signs, typically presented in the absence of any certain disease consideration or context, typically organized by organ system (eg, "the cardiovascular examination"). In contrast, our articles are usually organized by a certain condition (eg, "Does this patient have systolic dysfunction?"). And, although there are a few articles in which the authors take a more hypothesis-generating tack (eg, those on splenomegaly and hepatomegaly), we always frame them in a clinical context.

An issue all along has been whether, and how much, to integrate the evidence on symptoms and signs with that provided by diagnostic tests. In general, we have had so much material to deal with, and there are so many good texts on diagnostic tests, that we have limited our approach as much as common sense would allow. Some articles do include assessments of a few basic laboratory and radiologic studies that are commonly available to the clinician and that can be interpreted only by the physician in the clinical context (eg, the sedimentation rate for temporal arthritis or vascular congestion on a chest radiograph for systolic dysfunction). Recently, we expanded the series to include "rational clinical procedures," because many procedures are actually part of the clinical examination and tightly linked to the presence of the history and physical examination findings.¹⁰

David Simel of Duke University had been immediately excited by the concept and was a coauthor of the first article in the series, "Does This Patient Have Ascites? How to Divine Fluid in the Abdomen."¹¹ At that time, 1992, Simel made it clear that he intended to devote his research career to investigating this crucial area of medicine, and soon after he took over as primary editor of the series. Since then, he has stimulated large numbers of authors to complete these systematic reviews. His personal involvement with authors has brought us many more articles than we could otherwise have expected and ensured a uniform presentation. He also made certain that every manuscript had been through review before submission to JAMA, where I put each manuscript through rigorous external peer review, just as with all original submissions to JAMA.

Each review is a considerable undertaking, often requiring more than a year of unpaid and often unappreciated work, which explains why it has taken 15 years to produce what is now more than 70 articles in JAMA. As news of the series spread, volunteer authors suggested their own topics of interest. The appearance of fully fledged review articles depended on the skills and persistence of the authors and on the persuasive powers and analytic assistance of David Simel. Even then, more than a fifth of the proposed topics failed to result in publishable manuscripts, usually because the authors found insufficient evidence. It is for that reason that Simel and I published in 1995 a plea for support for a wide research agenda and the formation of collaborations to ensure that the wide gaps in our knowledge were filled.¹²

With the publication of this book, Simel has updated the first 51 published articles either alone or with the original authors. In addition, he has updated the primer⁸-essential for all readers of this book. David Simel's contributions to this series, and the transformation he has wrought in how we think about the clinical examination, have been immense, and working with him has been a privilege and a delight.

This is the first book in The Rational Clinical Examination series. Our plan is to keep soliciting and publishing in JAMA articles on fresh Rational Clinical Examination topics. We welcome volunteers with good ideas who are prepared to undertake the work. We will accumulate these articles, keeping them current with updates, and publish them as new chapters online and in succeeding editions of The Rational Clinical Examination book. The Rational Clinical Examination will be published online with a set of teaching/learning slides for each chapter and will be integrated with the Users' Guides to the Medical Literature and other online-only content and features in an extensive evidence-based medicine Web site called, JAMAevidence (http://www.JAMAevidence.com).

David Simel and I welcome Sheri Keitz (recently of the Durham Veterans Affairs Medical Center and Duke University, who has now moved to the University of Miami) as editor of The Rational Clinical Examination Education Guides. Sheri has many talents, including a fine critical eye. She has prepared or supervised development of all the teaching slides, and she has reviewed most of the Updates to the original manuscripts.

The series started with the encouragement of George Lundberg, then editor-in-chief of JAMA and the Archives journals. His successor, Cathy DeAngelis, has consistently and very strongly supported us, helping negotiate the complex path to publication. Annette Flanagin has been a tireless worker in this, as in so many other JAMA causes. This book would not have been possible without her.

We are grateful to Barry Bowlus for directing the publishing of this book and to Richard Newman for his advice and support. We are also grateful for the expertise of Jim Shanahan, Robert Pancotti, Helen Parr, and others at McGraw-Hill, as well as Peter Compitello at NewGen, and Holly Auten and her colleagues at Silverchair.

Publishing, like medicine, moves forward. During the last few years, the illustrations in JAMA have come under the care of Ronna Siegel and 2 medical illustrators, Cassio Lynm and Alison Burke. The series articles have benefited from their extraordinary skills, and improvements continue with the introduction of video images, as well as teaching clips. We also thank Cara Wallace and Angela Grayson for their expert editing and support.

The response to the articles published in JAMA tells us that this book will be useful. We also hope that readers will be stimulated to conduct research on aspects of the clinical examination. Perhaps readers will contact us if they believe they can undertake the sort of review that could constitute future articles in JAMA and chapters in the next book.

—Drummond Rennie, MD, FRCP, MACP