Dr. Ryle, a practicing primary care physician, has written a book directed at other practicing physicians and the general public. His goal is to explicate the process of diagnostic reasoning in medicine, particularly in primary care settings, in order to encourage clinicians to reflect on that process and reduce errors and biases as well as to help patients understand the inevitable uncertainties in the diagnostic process. He starts by stressing the central role of accurate diagnosis in effective medical care. He argues that the basic principles of effective diagnostic reasoning are an application of the general mechanisms of using human reasoning. He draws on work in cognitive science, particularly the now well-established positing of two systems of reasoning, one rapid and intuitive (System 1) the other slow and analytical (System 2). Although his characterization of this process is somewhat simplistic and omits discussion of the many controversies surrounding this model, it provides an effective heuristic for presenting his arguments to his target audience. It’s likely that this approach to cognitive functioning will be new to many practicing physicians and the public at large. He also attempts to tie this to a global theory of neurological functioning, emphasizing the distinct roles of the two hemispheres. I think this attempt to integrate a neurological framework is unnecessary for his argument (as the author himself acknowledges), and given the gross oversimplification involved, actually weakens the power of his presentation.

Before launching into his discussion of diagnostic reasoning in general, he has a chapter discussing diagnostic terms and classifications. While stressing the need for a diagnostic classification system in medicine, he quite astutely notes its limitations. He recognizes that some diagnostic terms carry significant precision with clear implications for the pathophysiology of disorders, while others are significantly less precise and carry the risk of implying a degree of certainty and knowledge that is not present. His brief discussion of diagnosis in psychiatry is particularly balanced and astute for a non-psychiatric physician. He notes the arbitrary nature of many psychiatric diagnoses and the lack of clear separation of psychopathology from normal functioning in many situations. He also comments on the fact that many of the phenomena relevant to psychiatric diagnoses are not precisely measurable and quite correctly observes that scoring instruments in psychiatry, by attaching a number to a judgment, often imply a degree of precision that is not actually there.

The discussion then turns to a characterization of the diagnostic reasoning process itself. He stresses that human knowledge in any field results in the construction of mental models of how the world works that in turn influence the interpretation of new information and observations. On the one hand we cannot operate without such models, while at the same time they can lead to premature closure and biased interpretations of new observations. He then discusses the complementary roles of the rapid intuitive approach of System 1 versus the slower more analytic approach of System 2. He argues for the critical role of System 1 in rapidly generating diagnostic hypotheses given the time sensitive nature of much medical evaluation and treatment. He noted however that this system is heavily prone to a wide range of cognitive biases which he reviews in a brief and useful way. He maintains that System 2 can correct some of these biases, especially if one is aware of them, but that this is a much slower and time-consuming process, and can never be completed with total certainty of objective and valid results. Optimal clinical reasoning, then, involves achieving the best balance between these two approaches. The possibility for error however is always present. The author does an excellent job of showing the intrinsic risk of error in medical practice even for the most conscientious clinician. One must balance between the risk of diagnostic error and premature closure on the one hand, and excessive delay or a failure to reach any diagnostic conclusion at all on the other. Both kinds of errors can have catastrophic consequences. It is argued however that awareness of the diagnostic reasoning process itself, and particularly the risk of cognitive biases, can optimize the clinician’s performance.

After this discussion of diagnostic reasoning the author presents an extremely engaging and useful series of case studies from his own clinical experience that illustrate the successes and the pitfalls in diagnostic reasoning. These cases put real flesh on the theoretical skeleton and add immensely to the book. Clinicians will identify strongly with these experiences. At the same time non-clinicians reading the book will gain a considerable understanding of the actual challenges facing the clinician in the real- world setting.

Overall this book does a superb job in meeting the goals that the author sets for himself. It offers much to the practicing clinician and to patients who want to understand better the medical reasoning process. The argument of the book does not deal directly with strictly philosophical issues likely to be of interest to the philosopher of medicine. The implicit philosophical assumptions are at times rather simplistic and questionable. Specifically, biases are presented as intrinsic to intuitive thinking with the implication that analytic thinking is relatively bias free. Analytic thinking is at times equated with science. When this is done, there appears to be an assumption of a rather outmoded notion of what constitutes scientific method, namely a hypothetical deductive approach rooted in logical positivism, with a tip of the hat to Popper’s notion of refutability. This approach has been heavily challenged in recent decades in the philosophy of science. Having said this however, I would still recommend this book to philosophers of medicine, especially if they have had limited direct experience in observing actual clinical work. It provides an excellent sense of the challenges of clinical practice with cases that raise interesting philosophical questions. Again, in all fairness to the author, his target audience is not philosophers or cognitive scientists. His somewhat simplified theoretical framework serves its purpose as a useful heuristic for the average physician reader for whom much of this material will be quite new.

© 2019 Douglas Heinrichs

Douglas Heinrichs, MD, Ellicott City, MD

Categories: Philosophical

Keywords: Risk, Reasoning