Medical Reasoning
THE NATURE AND USE OF MEDICAL KNOWLEDGE
Erwin B. Montgomery, Jr., MD
MEDICAL DIRECTOR
GREENVILLE NEUROMODULATION CENTER
GREENVILLE, PENSYLVANIA, UNITED STATES
PROFESSOR OF NEUROLOGY
DEPARTMENT OF MEDICINE
MICHAEL G. DEGROOTE SCHOOL OF MEDICINE AT MCMASTER UNIVERSITY
HAMILTON, ONTARIO, CANADA
Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America.
© Oxford University Press 2019
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above.
You must not circulate this work in any other form and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data
Names: Montgomery, Erwin B., Jr., author.
Title: Medical reasoning : the nature and use of medical knowledge / Erwin B. Montgomery, Jr.
Description: New York, NY : Oxford University Press, [2019] | Includes bibliographical references and index.
Identifiers: LCCN 2018018343 (print) | LCCN 2018019011 (ebook) | ISBN 9780190912932 (online content) | ISBN 9780190912949 (updf) | ISBN 9780190912956 (epub) | ISBN 9780190912925 (cloth : alk. paper)
Subjects: | MESH: Logic | Philosophy, Medical | Decision Making | Thinking | Metaphysics
Classification: LCC R723 (ebook) | LCC R723 (print) | NLM W 61 | DDC 610.1—dc23
LC record available at https://lccn.loc.gov/2018018343
This material is not intended to be, and should not be considered, a substitute for medical or other professional advice. Treatment for the conditions described in this material is highly dependent on the individual circumstances. And, while this material is designed to offer accurate information with respect to the subject matter covered and to be current as of the time it was written, research and knowledge about medical and health issues is constantly evolving and dose schedules for medications are being revised continually, with new side effects recognized and accounted for regularly. Readers must therefore always check the product information and clinical procedures with the most up-to-date published product information and data sheets provided by the manufacturers and the most recent codes of conduct and safety regulation. The publisher and the authors make no representations or warranties to readers, express or implied, as to the accuracy or completeness of this material. Without limiting the foregoing, the publisher and the authors make no representations or warranties as to the accuracy or efficacy of the drug dosages mentioned in the material. The authors and the publisher do not accept, and expressly disclaim, any responsibility for any liability, loss or risk that may be claimed or incurred as a consequence of the use and/or application of any of the contents of this material.
9 8 7 6 5 4 3 2 1
Printed by Sheridan Books, Inc., United States of America
For Lyn Turkstra, who saved my life in so many ways and for those clinicians who did their utmost when it was not forced or popular.
CONTENTS
Preface xiii
Glossary of Concepts xxi
Companion Website xxxix
1. Introduction: Places to Explore the Ramifications of Uncertainty 1
Deduction and Its Derivatives 2
Probability and Statistics 6
Extra-Logical Considerations 10
Induction 12
The Discipline of Logic 15
Origins of Ideas as Hypotheses 16
Rationalist/Allopathic Medicine Versus the Empirics 17
Science and Scientism 19
Perspective 20
2. What Are We to Make of Reasoning in Modern Medicine? 21
Certainty, Knowledge, and Understanding 22
Track Record on Beneficence 25
Reproducibility in Biomedical Science 26
Approach in this Text and an Emphasis on Logic in Its Widest Connotation 27
Evolutionary Epistemology and Logic 27
Use and Misuses of Logical Errors 29
Applied Epistemology 29
Descriptive, How Things Are, Versus Normative, How Things Should Be 30
The Dichotomization of Medicine Between Science and Art 30
Logic and Its Extensions Versus Science 32
Medicine Beyond the Realm of Science but Within the Realm of Epistemology, Logic, and Logic’s Extensions 34
3. Epistemic Challenges and the Necessary Epistemic Responses 36
The False Choice Between Universal Scientific Judgment and Particular Common Sense Judgment 37
The Myth of the Inevitability of Certainty 40
Clinicians’ Obligation to Reason 40
The Evolution of Medical Reasoning and Misreasoning 41
The Importance of Philosophical Analyses 43
Future Areas of Study in Logic 44
4. Medical Epistemology: The Issues 45
Case 45
Deduction 46
The Limits of Deduction 47
The Patient Imperative 48
Dichotomization Based on Statistical Significance 49
What Physician A and What Physician B Thought 50
Comparing Apples and Oranges 52
Facts Are Insufficient 54
5. Deduction, Induction, and Abduction: The Basics 56
Case 56
Logical Argumentation 57
Consequence of the Failure to Uphold the Principle of the Excluded Middle 59
The Problematic Nature of the Throat Culture 61
Sensitivity, Specificity, and Positive and Negative Predictive Values as Logical Forms 63
Deduction 65
Fallacy of Confirming the Consequence 66
Rescuing the Hypothetico-Deductive Method 67
Abduction 69
Principle of Transitivity and the Fallacy of Pseudotransitivity 69
Induction 71
Induction and the Scientific Method 73
6. Evolution of Medical Reasoning 75
Ascendency of Allopathic (Scientific) Medicine 76
The Inverse Problem 77
Historical Approaches to Diagnosis 78
Medicine and Science 79
Diagnosis and Treatment 80
Evolution of Medical Science 81
The Persistence of Galenic Ideas with the Advancement of Medical Science 84
Dominance of Mechanistic Theories of Physiology and Pathophysiology 87
History of Medical Abduction 87
Institutionalization of Scientific Medicine and Further Reinforcement of Abduction 89
A Different Notion of Science 91
7. Variability Versus Diversity in Variety: The Epistemic Conundrum and Responses 93
Variability Versus Diversity 93
The Human Epistemic Condition 96
8. The Meaninglessness of the Mean 98
Medians and Quartiles 102
Cumulative Percentage (Probability) Function 104
The Metaphysical Notion of the Mean (and Median) 105
Different Types of Means 107
9. The Value of Statistical and Logical Thinking 108
Wired for Intuition? 109
Statistical and Epistemic Thinking 111 Are Randomized Controlled Trials Privileged? 113
Undersampled and Biased Experience and Induction 114
Non–Evidence-Based Medicine Methods of Medical Reasoning 116
10. The Centrality and Origins of Hypotheses 118
The Importance of Hypotheses 119
Origins of Hypotheses 120
Hypotheses and Postmodernism 121
Hypothesis Generation in Medical Decisions 123
Origin of Hypotheses from Medical Science 124
Role of Presuppositions or Metaphors in the Origin of Hypotheses 125
Diagnostic and Statistical Manual V (DSM-5): Modern-Day Battle of the Allopaths and Empirics 126
11. Necessary Presuppositions: The Metaphysics 128
The Centrality of Metaphysics (Properly Defined) 128
Allopathic Medicine and Science 130
Role of Theory in Shaping Observation 131
Aristotle’s Notion of the Contraries 132
The Triumph of Allopathic Medicine (the Modern Doctor of Medicine and Doctor of Osteopathic Medicine) and Its Presuppositions 133
Rationalism (Reductionism), Empiricism, and Dogmatism and Their Presuppositions 133
Dealing with Variety and Its Consequent Uncertainty 135
Allopathic Medicine and Reductionism 136
Cell Theory, the Rise of Pathology (Particularly Histopathology), and Pathological–Clinical Correlations of William Osler 138
Mereological Fallacy 141
The Conundrum Is the Result of the Inverse Problem 142
Neurology and the Neuron Doctrine 143
A Case Study 144
12. The False Notion of Intention, Choice, and Inhibition 150
One-Dimensional Push–Pull Dynamics 150
Methodological Reduction Creating New Ontology 152
Misperception of Inhibition 154
Potentiality 157
13. The Role of Metaphor 160
Metaphors in Medical and Scientific Knowledge and Understanding 161
Metaphors Leading to Errors in Medical and Scientific Knowledge and Understanding 162
Metaphors and Metonymies as Structuring Observation 163
Nonlinguistic Metaphors 164
Reductionism as a Process Metaphor 166
14. Dynamics 167
The Necessity of Addressing Dynamics, but How? 168
Continuing Concerns for Telos 170
One-Dimensional Push–Pull Dynamics 171
Understanding Dynamics Through Metaphor 173
Chaos and Complexity 174
15. Medical Science Versus Medical Technology 178
Medical Science, Medical Technology, or Both? 179
Lessons from the Past 180
A Special Case of Technology Looking for a Question 181
Mathematics as a Technology 182
Science for Its Own Sake, Technology for the Sake of Others 183
Experimentalism and Science as Technology 183
16. Irreproducibility in Biomedical Science 186
The Magnitude of the Problem 186
Scope of the Issue 188
Information Loss and Irreproducibility 189
Areas to Explore 191
Reconstruction: The Other Half of Reductionism and Relevance to Reproducibility 194
17. Medical Solipsism 195
Knowledge in Medicine 195
Nature of Solipsism 197
The Use of Consequence to Resolve the Solipsist’s Advantage 199
The Solipsism of Evidence-Based Medicine 200
The Solipsism of US Food and Drug Administration Approval 202
The Solipsism of Balkanized Medicine 203
18. Critique of Practical and Clinical Medical Reasoning 204 Distinctions Without Distinctions? 205
Clinical Intuition 209
Fundamental Limits of Science and Scientific Reasoning Necessitating
Practical Reasoning 213
Clinical Meaningfulness 215
Clinical Assessment 217
The Diagnostic/Therapeutic Scheme 219 Difference Between What Can Be Done and What Should Be Done 220
19. A Calling to Be Better than Ourselves 222
Malpractice and Accountability 223
Peer Review 224 Too Few to Fail 224
Ethical Principles and Moral Theory 225
References 227 Index 237
PREFACE
Writing and publishing this text entail risks because a large receptive audience cannot be assumed a priori. The topic is one seldom addressed, and therein lies the challenges. It is not as though this text addresses a widely recognized and greatly appreciated need. It does not offer an answer to the cure for any disease for which there is not one already, nor for new ways to relieve incurable diseases. It offers no new laboratory tests or imaging studies. So, what does it offer?
The attempt here is to exercise and thereby sharpen the most important tool any clinician or scientist can wield—their brains. Some may dismiss such an effort, finding it unnecessary or even demeaning. Yet clinicians are human and as prone to errors as anyone else. This can be embarrassing: witness the 83% of radiologists who did not recognize the image of a gorilla in a computed tomography (CT) scan of the chest despite looking directly at it (Drew et al., 2013), and, as such, clinicians have a duty to themselves, the profession, and patients to be cognizant of the errors to which the human flesh is err as the prerequisite to avoiding them.
It is likely that every clinician and scientist is concerned about the sensitivities, specificities, positive and negative predictive values, and consequences of failure to recognize false positives and false negatives in any laboratory or imaging test. Every clinicians’ and biomedical scientists’ mind may not be considered a tool, but the consequences of the mind’s use for patients are just the same as the application of any tool or device. The consequences of a false positive or false negative resulting in a wrong diagnosis or treatment is the same whether the false positive or false negative was the result of a blood test or the clinician’s mind. Does not the similarity in consequence, if not in operation, between the use of tools versus the clinician’s mind warrant the same degree of concern and thus scrutiny? The affirmative answer motivates this book.
This text is a critique of medical reasoning, but the term “critique” is taken in its philosophical connotation, which means “rigorous analysis.” It is not a value judgment in any ethical or moral sense. Rather, terms of “productive” or “counterproductive” to the intended purpose are used. Whether productive medical reasoning is demanded or counterproductive reasoning countenanced is an ethical question resolved in the context of the controlling moral theory. Thus, good or bad medical reasoning depends on ethics. However, this is not as relative as some may think because there is a general principle—common
morality (Beauchamp and Childress, 2013)—induced from the consideration of reasonable persons, that provides an operationalization of good and bad.
Throughout the text, I will present examples of what generally would be considered bad medical reasoning. These are raised solely to demonstrate the importance of a critical analysis of medical reasoning and to highlight that complacency is not an option. The very large majority of clinicians that I have encountered in more than 40 years of practice are very smart and talented, yet misadventures in medical reasoning are not rare. So, whence the bad medical reasoning?
Caring for humans is difficult because humans are complex; consequently, manifestations of health and disease present with great variety. The variety presents an epistemic conundrum in gaining new knowledge. In response, specific epistemic choices must be made. Each choice renders knowledge problematic and therefore uncertain. The needs of patients force actions even when knowledge is uncertain. Certainty in medical knowledge depends on valid deductive logic with true premises and valid arguments. However, gaining medical knowledge requires the use of logical fallacies. Furthermore, probability, and thus statistics, is founded on logical fallacies, and statistical errors often derive from the misuse of these inherent logical fallacies. The judicious use of logical fallacies leads to new knowledge and optimal care; injudicious use leads to failures and errors. For example, it will be demonstrated that randomized clinical trials, virtually the sine qua non of evidence-based medicine, rest on the logical Fallacy of Four Terms. The hypothetico-deductive approach to diagnosis and treatment rests on the logical Fallacy of Confirming the Consequence, which risks confirmation bias. However, no reasonable clinician or scientist would jettison randomized controlled trials or the hypothetico-deductive approach, and that is not the argument in this book. Rather, this book attempts to understand the dangers of misusing the Fallacy of Four Terms in order to strengthen the utility of randomized controlled trials and the hypotheticodeductive approach.
Rarely in my experience are there discussions of logic in medicine or biomedical science. To be sure, numerous publications relate to the logic of medicine, but the term “logic” is more or less generic in the sense of a method or algorithm, and it is used more often in a descriptive (how things appear to work) rather than a in normative (how things should work) sense. Thus, many animals can be described as having logic but not likely the type of deductive logic first championed by Aristotle. Indeed, the type of logic common to most animals (including humans) can be held synonymous, roughly, with cognitive function, and human medical reasoning can be understood in terms of cognitive functions and errors in terms of cognitive biases (see, e.g., the work of Norman and colleagues, 2006). Yet these, too, can be understood as variations of deductive and inductive logic. Some have contrasted analogical reasoning as being somehow different from deduction (and its variants) and
induction. When understood as the Fallacy of Pseudotransitivity derived from the Principle of Transitivity in deductive logic, then analogical reasoning is an extension of deductive logic. Understanding inductive and deductive logic and the extensions to logical fallacies, probability, and statistics is critical to nearly every aspect of medical reasoning.
Some authors have attempted to formulate medical reasoning in deductive terms, and attempts have been made to extend the utility of deductive logic to the complex and often imprecise realm of medicine (for example, through the use of fuzzy logic). However, I have not encountered discussions of a deductive logic approach to medical reasoning where logical fallacies are embraced rather than avoided, the necessity to embrace will be demonstrated. If true that it is necessary to embrace logical fallacies, the relative novelty of the concept, at the very least, offers a different perspective that perhaps could be helpful.
This notion of logic, expanded to the judicious use of fallacies, is evolutionary and organic. Because these fallacies arise from the need to treat individual, specific patients, they are, hence, organic. The epistemic nature of medical care centers on the need for certainty but also utility. The absolute certainty of deductive logic is very limited in utility. Deductive logic derives its certainty from the severe constraints placed on possibilities entailed by the Principle of the Excluded Middle. Yet questions confronting clinicians seldom have only true or false answers.
Confronted by great variety in the expressions of health, disease, and disorder, the epistemic question is whether the variety derives from variability or diversity. Variability suggests variations around some economical set of canonical forms. Medical knowledge and decision-making can be made more economical by addressing the canonical forms rather than individual instantiations. Reductionism is one example. Diversity, by contrast, argues that no such economization is possible and that each patient is a new and independent phenomenon—taken as de novo. The epistemic conundrum resulting from the variety of manifestations has driven different approaches to medicine since the ancient Greeks and continues to do so to this day.
Mechanistically, humans are incredibly complex, far more complex than would be inferred from their observable behaviors. This notion is inherent in the concept of medical syndromes as distinct from medical diagnoses. Congestive heart failure is a syndrome of specific symptoms and signs that follow from any number of different diagnoses. This means that, at the level of the patient’s direct manifestations, one cannot determine which of the diagnoses is responsible for the individual patient’s syndrome. This lack of one-to-one correspondence between the symptoms and signs to a single diagnosis presents the epistemic conundrum of the Inverse Problem, meaning one cannot infer the specific diagnosis from the manifestations of the syndrome.
The analogue to the Inverse Problem in logic is the Duhem–Quine thesis, which states that, should an argument result in a false conclusion, one cannot
know which of the component premises were false or which proposition was invalid. The instantiation in probability theory is the Gambler’s Fallacy. These issues are explored in detail throughout the book.
Induction, an inference to general principles based on a set of observations, is another source of knowledge, whether of medical science or a patient’s unique condition. Yet induction presents epistemic conundrums. These include the Fallacy of Induction and the A Priori Problem of Induction, discussed in detail in later chapters.
Some authors, such as Groopman (2007), fail to appreciate the nature of logic and do a disservice by creating straw-man arguments (which is casting one’s opponent as obviously flawed or failed, independent of the argument, in order to advance one’s own position). Groopman describes a brilliant pediatric cardiologist who thought he was being logical, with the result being a near catastrophe. Basically, the conclusion in Groopman’s work was not to trust logic, but, as will be seen, logic—when used properly—is the only source of confidence in decision-making.
Groopman’s case involved a pediatric patient with severe narrowing of the mitral valve of the heart and a hole in the wall between the atrium of the two halves of the heart. The result was that blood in the left half of the heart would flow to the right side and not go through into the left ventricle and then on to the aorta to supply the body because higher blood pressures were required to push the blood through the narrow mitral valve and out to the body. The cardiologist reasoned that if one closed the hole in the wall, there would be higher blood pressure in the left side of the heart because the blood pressure would not be reduced by the blood going into the right side of the heart. The cardiologist was quoted as saying “It has to be right, correct? It is very sound logic. But it’s wrong.” The child got worse when the hole in the wall of the heart was closed. The explicit implication was to not trust logic.
The reasoning here was illogical, as will be discussed in Chapter 5. The failure of the argument has nothing to do with logic, but is instead due to a false proposition (closing the hole in the wall would increase the blood pressure in the upper left half of the heart), for which the cardiologist is at fault, not logic. Even the soundest logic does not guarantee true conclusions if the premises are false. If someone chooses to use a hammer to saw a piece of wood, it is not the fault of the hammer that the piece of wood is shattered. Thus, attributing the cardiologist’s argument to logic is faulty and, if anything, demonstrates the clear need for clinicians to understand and apply logic appropriately.
Rapid advances in biological sciences, particularly since the early 1800s, have had and continue to have profound impacts on medical reasoning but perhaps not in a manner commonly thought of today. As will be seen, it was not an achievement of biological science in improving medical care that led to the current dominance of allopathic (modern) medicine. Rather, it was allopathic medicine’s adoption of scientism, following on the heels of advancing
biomedical science, that drove the process. “Scientism” here refers to the metaphysical stance (perhaps best described as faith) that we can “science” our way to certainty. This history is reviewed in subsequent chapters.
Fortunately, there is a long history of concern with efforts to understand reasoning going back to the ancient Greek philosophers. Their knowledge and experience can help clinicians and biomedical scientists, and, correspondingly, philosophical terminology is used. Many philosophical terms are convenient shorthand references to important and complex concepts and are no different from the shorthand of medical and scientific technology. Understanding the general form of logic, its extensions into judicious fallacies, and probability and statistics allows one to spot potential trouble in many contexts. In any medical discussion, when a decision is recognized as a specific instantiation of the general logical form if a implies b is true, b is true, therefore a is true , then that medical decision is immediately suspect because it represents the Fallacy of Confirming the Consequence. It does not matter exactly what a or b is.
Philosophers (and others whose professions involve the analysis of reasoning) can make important contributions to understanding and teaching medical reasoning to clinicians and biomedical scientists. To do so, it is important that each appreciates and converses in each other’s language. Clinicians may become philosophers in their own right. As the philosopher Bertrand Russell said, “To teach how to live without certainty, and yet without being paralyzed by hesitation, is perhaps the chief thing that philosophy, in our age, can still do for those who study it” (Russell, 1940).
The fear is that this text may be greeted like so much unsolicited advice. Such advice often is disturbing because it presupposes a problem that the advice is intended to rectify. As the receiver of the advice never asked for it in the first place, the unsolicited advice is perceived as an insult, not only because there is a problem for which the receiver is responsible, but also because the receiver did not have the sense or presence of mind to recognize the problem. The majority of clinicians will tolerate being told that their diagnoses or treatment recommendations are a mistake, but it is a rare individual who tolerates being told that the way he or she thinks led to the mistake. Yet the ubiquity of medical errors and the lack of reproducibility in biomedical research suggest that something is amiss and that it is important to at least consider the possibility of misreasoning (Chapter 2).
As will be addressed, medicine is not synonymous with science, and conflating the two produces potential sources of errors in medical reasoning. Indeed, the very methods, such as statistical inference and reductionism used in science, result in an irretrievable information loss of precisely the type needed to apply science to the care of individual patients. It is a matter of physics (based on information theory and the Second Law of Thermodynamics), as will be discussed.
Ultimately, as will be seen, uncertainty can only be managed—it cannot be eliminated. Scientific and statistical methods only convey an impression of reducing uncertainty. A reduction in apparent uncertainty in one context only surfaces as uncertainty elsewhere. Managing uncertainty properly requires a full and clear understanding of the nature of uncertainty and the nature of knowledge, a concern tackled by the philosophical discipline of epistemology.
The reader may note some degree of redundancy in the text. This is necessary and parallels experience in biomedical ethics. One can know the ethical principles and moral theories, but how these combine to resolve ethical questions requires discussion in the specifics. The theme here is that logic and its necessary extensions of judicious fallacies, probability, and statistics form the economical set of principles that can be recombined to understand specific cases. Importantly, it is only in the context of many specific cases that the full dimensions of logic and its extensions can be appreciated.
It is worth drawing attention to two features that may help readers not familiar with logic, epistemology, logic, probability, and statistics. First, a Glossary of Concepts (presented next) provides a brief introduction to a variety of concepts addressed in the book. Some readers may find it helpful to read the Glossary of Concepts in preparation for reading the main text. Also, two online appendices provide brief introductions to logic and to probability and statistics; they are available online at www.oup.com/us/medicalreasoning.
From what vantage do I come to recognize and understand the errors of medical reasoning? It certainly did not come from my undergraduate studies in biochemistry, nor from medical school, residency, or the fellowship that followed. As an assistant professor in the Department of Neurology at Washington University in St. Louis, I took graduate courses in philosophy, particularly in epistemology. What I learned from that experience is that, in philosophy, nothing was above or beyond debate. Through practice in graduate school and subsequently, I acquired some skill to unpack arguments, which is deconstructing arguments to their fundamental premises, particularly assumptions, presuppositions (implicit assumptions), propositions, and logical structures. I had the distinct advantage of prior efforts extending back through thousands of years by a continuous tradition of intellectual rigor in philosophy. It is to my professors of philosophy from 1981 to 1990 at Washington University in St. Louis that I owe a very large debt of gratitude.
What drove this writing are my many years of attending on the teaching wards and seeing the bafflement on the faces of very intelligent women and men during their later training in medical school when confronted with errors in medical reasoning that they had seen as standard practice elsewhere. Some students battled the confusion and sought to understand; unfortunately, there was so little time, with so much wrong thinking to be undone and new intellectual rigor to instill. Also, there was the tyranny of trying to be a success at my
mainstream academic career, which provided little time for efforts in the epistemology of medical reasoning.
In July 2014, I was offered a unique opportunity to become the medical director of the Greenville Neuromodulation Center and the Greenville Neuromodulation Scholar in Neuroscience and Philosophy at Thiel College in Greenville, Pennsylvania. The fact of being rather senior with nothing to prove to anyone and more time to pursue my latent philosophical interest allowed me to take up this effort to understand medical reasoning, as it is and how it might be. Teaching students in the small liberal arts college provided a valuable challenge and sounding board. Also, I received an unimaginable stroke of good fortune to meet and interact with Dr. Arthur “Buddy” White, a true philosopher and chair of the Department of Philosophy at Thiel College. His passion for philosophy was contagious and would heat to flame even the dampest kindling. I am indebted to the leadership of Greenville Neuromodulation Center for the opportunities and support for this effort.
In the efforts that follow, whether I am right or wrong in my analyses and recommendations is of little consequence. If this text sparks a discussion among others, one that continues to evolve and perhaps one day change the way clinicians and biomedical scientists think, then that would be a satisfaction that no one can deserve but nonetheless is a blessing. Finally, thanks to Erwin B. Montgomery III, PhD and Melissa Revell who translated from my poor prose.
Erwin B. Montgomery, Jr., MD
GLOSSARY OF CONCEPTS
This glossary is intended to be a ready reference for novel concepts or novel variations on established concepts as they are encountered while reading the book. Alternatively, the reader may first wish to review these concepts to become familiar with them prior to encountering them in the main text of the book. This glossary presumes some familiarity with logic and probability. Those wishing to refresh this knowledge might first review Appendices A and B, available at www.oup.com/us/medicalreasoning.
Abduction a form of reasoning that resembles deduction, but, in reality, it is the logical Fallacy of Confirming the Consequence, which is of the form if a implies b is true and b is true then a is true (see Fallacy of Confirming the Consequence). For example, one might reason that if patient A had strep throat then the patient should have a sore throat, fever, and an exudate in the throat; the patient has a sore throat, fever, and an exudate; therefore, the patient has strep throat. However, the patient could have a sore throat from any number of other causes. The hypothetico-deductive approach is an example of abduction in allopathic medicine (see Hypothetico-Deductive Reasoning and the Scientific Method).
Actor/Action Distinction the problem of inferring the mechanisms of action from observations of the actor. Originally derived from an analysis of meaning, the question was posed whether an expert actor, such as a Shakespearian actor, flawlessly reciting Shakespeare’s lines, actually understood what Shakespeare meant or intended. By extension, observation of the actor alone cannot convey an understanding of what Shakespeare meant or intended. Thus, observations of expert clinicians may not reveal the actual modes of reason, particularly as the expert clinician may be unaware of the reasoning modes used.
Allopathic Medicine also called rational or regular medicine. This program or school of medicine constructs diagnosis and treatment on the basis of an economical set of principles from which an explication of the individual patient is reconstructed. As such, allopathic medicine was well poised to take advantage of the emergence of modern medical science with emergence of the cell theory, the germ theory, histopathology, and microbiology. Allopathic medicine stands in contrast to the Empirics, who held that the health or disease of individual persons could not be explained on
an economical set of fundamental principles but rather that each patient has to be taken on as an epistemic problem in his or her own right, de novo.
Analogical Reasoning a mode of reasoning based on analogies; for example, treatment of patient A with disease X is based on the similarity of patient B, thought to have disease X, and her response to treatment Y. Use of analogy rather than synonymy allows for accommodation of the great variability among patients, their health, and the manifestation of their disease. A clear example is in the off-label use of medications in the United States. Similarity of patient A to patient B for which treatment Y has been approved or recommended becomes justification for the use of treatment Y for patient A. Some argue that analogical reasoning is of a different kind than logic. However, from the perspective of evolutionary logic (see Evolutionary Logic), analogies are an extension of deductive logic, particularly the concept of the Fallacy of Pseudotransitivity (see Fallacy of Pseudotransitivity), which derives from the logical Principle of Transitivity.
A Priori Problem of Induction induction typically involves generalizing a principle from a set of observations; for example, every raven seen is black, therefore all ravens are black. The A Priori Problem of Induction asks, “What are the rules, implying prior knowledge, by which a specific set of birds is selected to form the set of ravens from which the inference based on commonality—that of being black—is derived?” Note that if blackness is a rule, then the induction becomes a tautology conveying no new knowledge. Nor can the rules be such that crows or black swans would be included. Thus, induction requires some prior knowledge or circumstance. The A Priori Problem of Induction affects and limits pattern recognition as a means of medical decision-making (see Pattern Recognition).
Baconian Science a version of science that promulgates itself as being strictly empiric and arrived at by experimentation. The motto of the Royal Society Nullius in Verba (Take No One’s Word for It) typifies this approach. Taken to its extreme, no claim can be demonstrated by derivation or extrapolation from prior principles and initial conditions. For example, mathematical proofs would not be acceptable as evidence in support of a scientific claim. Baconian science tends to typify the version of science claimed by medical science and dominating medical reasoning.
Bayes’ Theorem see “Hoof Beats and Zebras” (Bayes’ Theorem)
Cartesian Science an alternative version to Baconian science (see Baconian Science) that emphasizes rationalism where new “scientific” knowledge is synthesized from fundamental principles. For example, one does not need to demonstrate Newton’s laws of motion empirically (inductively) to determine an astronomical anomaly that subsequently proved to be the planet Pluto. By the same token, one does not need to demonstrate empirically the therapeutic effect of penicillin in the event that the eventual use of penicillin may be considered.
Causal Syllogism a derivation from the practical syllogism, which is a variant of the deductive syllogism where the state-of-being linking verb is replaced by the action verb, in this case cause. For example, the typical syllogism is of the form all a’s are b’s, and c is an a, therefore c is a b. One can see the certainty and utility if a is disease A (sample of such patients), b is effective treatment B, and one’s patient is c. However, not all circumstances are amendable to syllogistic deduction. For example, one might want a to be as symptoms of a disease, b, and c is a patient with symptoms a. Here the relation is one of cause, and the syllogism becomes all a’s are caused by disease b, one’s patient (c) has symptom a, therefore one’s patient has disease b. This is an example of a causal syllogism that does not have the certainty of a syllogistic deduction.
Causational Synonym, Principle of see Principle of Causational Synonymy.
Chaos and Complexity originally demonstrated in mathematical systems and now with growing evidence in physical systems, chaos and complexity refer to systems that cannot be defined definitively, and hence any future behavior cannot be predicted. Indeed, it is the unpredictability of chaotic and complex systems that is their main defining feature. Additionally, such systems are highly dependent on the initial conditions from which any system would start. Chaotic systems are relatively simple systems, but their interactions are highly nonlinear. For example, the state equation y = x is linear in that a doubling of x leads to a doubling of y. However, equations such as y = −xx are nonlinear. Complex systems may have simpler dynamics, but their sheer number of interactions makes the outcomes difficult to predict. These systems present a problem for most current statistical approaches to make “sense” or inferences. To the degree that human biology is chaotic or complex, the application of traditional statistical analysis likely is unproductive, at the least, and misleading, at the worst.
Clinical Judgment see Practical Judgment.
Complexity and Chaos see Chaos and Complexity.
Contraries originating from Aristotle’s Physics (2001), it is a method used to simplify understanding of the interaction between phenomena (dynamics) and the phenomenon itself. Simplification is achieved by reducing different dynamics to a dynamic that is one-dimensional. For example, consider a gray scale that goes from white to black. The question is, “How many shades of gray are there?” The answer depends on the resolution of the device used to measure “grayness,” but, potentially, there are an infinite number of grays. However, such a position would obligate one to an infinite number of elements (ontologies)—shades of gray—such that managing them in any science of the gray scale would become very complex. Alternatively, one can take the position that there are just two shades of gray—black and white—that are the extremes of the continuum. Thus, the potential infinity of ontologies reduces to just two. The dichotomization
of a phenomenon alone as a single dimension, often characterized by a push–pull dynamic, is ubiquitous in medicine and may well be the source of error.
Cookbook Medicine a term used to describe the practice of medicine that directs patient diagnosis and management to a predefined algorithm. However, in typical use, it refers to an economical set of algorithms that are expected to be optimal for every patient. This is analogous to the dogmatic school of medicine originating with the ancient Greeks (see Dogmatics). Those using the term “cookbook medicine” in the usual pejorative sense often do so based on skepticism. The problem is often that skepticism reflects an unfounded bias, perhaps because of a specific tradition or, in a polemical sense, that if clinicians can be considered equivalent to algorithms, then a reduction in the clinician’s autonomy is threatened.
Deduction a method of argumentation subsumed in the discipline of logic. The value of deductive logic is that, given true premises and valid propositions, then the conclusions must be true. This provides the greatest level of certainty possible. However, the degree of certainty is related reciprocally to the utility of such arguments. Deductions typically resolve down to tautologies (see Appendix A) and thus do not lead to new knowledge. Deduction is “truth preserving” but not “truth generating.” A significant problem in medicine is that reasoning often is described as deductive and thus trades on the confidence that can be derived from true deductive arguments. But they are not strictly deductive and, in reality, do not have the certainty of deduction with true premises and valid propositions. Examples include the hypothetico-deductive approach to diagnosis and the scientific method, which embody the logical Fallacy of Confirming the Consequence (see Fallacy of Confirming the Consequence).
Differential Diagnosis a list of possible diagnoses causal to the patient’s phenomenological syndrome. Discussion of the differential diagnosis is made to contrast it with the actual diagnosis. The distinction is critical because the differential diagnosis is the starting point of medical diagnostic decision-making, and the diagnosis is a consequence. Proceeding directly to a diagnosis without first considering a differential diagnosis risks misdiagnosis, which stems from the Fallacy of Limited Alternatives (see Fallacy of Limited Alternatives) in the context of the Inverse Problem (see Inverse Problem).
Dogmatics a school of medicine at least since the ancient Greeks that held that patients should be managed by direct referral to accepted medical texts—possibly a form of cookbook medicine (see Cookbook Medicine).
“Don’t Change Too Many Things at Once Lest You Get Confused” related to admonition in medicine often raised as an example of clinical or practical reasoning. Sometimes the use of clinical or practical reasoning is cited as evidence that medicine cannot be scientific or logical and that extra-logical
reasoning is necessary. This practical admonition reflects the epistemic condition of the Inverse Problem (see Inverse Problem).
Duhem–Quine Thesis a thesis related to finding the cause of a failed argument. In any deduction, a false conclusion necessarily means that one or more premises are false or one or more propositions are invalid. The problem is that, within the providence of the argument, there is no way to tell which premise is false or which proposition is invalid.
Economical in the context of this book, economical does not refer to anything monetary (except as specific designated exceptions). Rather, it is used in a quantitative sense and relates to the ratio of different phenomena that must be explicated and the number of underlying principles necessary for each explanation.
Empirics also called irregular medicine an approach or school of medicine that stands in contrast to allopathic medicine (see Allopathic Medicine). The fundamental premise of the Empirics is that an explication of any individual patient on the basis of an economical set of principles, such as those related to physiology and pathophysiology, is not possible. Thus, the only means to treatment is to analyze the specific individual symptoms and signs manifest by each patient and direct specific treatments to the specific symptoms and signs. In homeopathy, this approach is reflected in the saying “like cures like.”
Epistemic Condition this condition relates to the fundamental dilemma that confronts the acquisition and use of human knowledge, particularly as it relates to medical decision-making. The greatest certainty in knowledge comes from logical deduction (see Deduction), yet deduction does not yield new knowledge, which is crucial to any empiric discipline such as medicine. Rather, logical fallacies must be employed in order to gain new knowledge but at the loss of certainty. The incorporation of judiciously used logical fallacies defines evolutionary epistemology and evolutionary logic. Methods and approaches have been developed to minimize the loss of certainty, such as probability and statistics. The risk for uncertainty can be appreciated from epistemic risk (see Epistemic Risk).
Epistemic Degrees of Freedom a component of epistemic risk (see Epistemic Risk) that relates to the degrees of freedom necessary to make the conceptual linkage between hypotheses and predictions, as in the hypotheticodeductive approach (see Hypothetico-Deductive Reasoning) or between the source and target domains used in the analogical approach (see Analogical Reasoning). The degrees of freedom relate to translations between linked arguments; for example, the number of instrumental assumptions and presuppositions. For example, hypotheses relating cognitive functions to neurometabolic image changes require a number of intervening translations, such as changes in deoxygenated hemoglobin to neuronal activities and then patterns of neuronal activities to cognitive functions.
