Diagnosis is arguably the most important job of a physician. But what does it actually mean to make a diagnosis? In this episode, we’ll explore this question by tracking the development of the “classical” model of diagnosis and pathological anatomy and discussing three cases over three hundred years. Along the way, we’ll ponder the concept of the lesion, iatromechanistic theories of the human machine, the birth of the International Classification of Diseases, and the rise and decline of the autopsy.
You can sign up for the iMED Digital Education conference at cmeregistration.hms.harvard.edu/digitaleducation.
- Hooper R, The Physician’s Vade-Mecum: Containing the Symptoms, Causes, Diagnosis, Prognosis and Treatment of Diseases. 1812.
- Holdman L et al, The Value of the Autopsy in Three Medical Eras. N Engl J Med 1983; 308:1000-1005.
- Cabot RC et al. CASE RECORDS OF THE MASSACHUSETTS GENERAL HOSPITALANTE-MORTEM AND POST-MORTEM RECORDS AS USED IN WEEKLY CLINICO-PATHOLOGICAL EXERCISES. Case 9431. Boston Med Surg J 1923; 189:595-599.
- Shojania KG and Burton EC, The Vanishing Nonforensic Autopsy. N Engl J Med 2008; 358:873-875
- Morgagni GB. The seats and causes of diseases investigated by anatomy in five books, containing a great variety of dissections, with remarks. To which are added … copious indexes. 1769. Retrieved online: https://openlibrary.org/works/OL15821973W/The_seats_and_causes_of_diseases_investigated_by_anatomy
- Castiglioni A, GB Morgagni and the Anatomico-pathological Conception of the Clinical. Proceedings of the Royal Society of Medicine, November 7, 1934.
- Thiene G, Padua University: The role it has played in the History of Medicine and Cardiology and its position today. European Heart Journal, Volume 30, Issue 6, March 2009, Pages 629–635.
- Zampieri F et al, Origin and development of modern medicine at the University of Padua and the role of the “Serenissima” Republic of Venice. Glob Cardiol Sci Pract. 2013; 2013(2): 149–162.
- Conner, Annastasia (2017) “Galen’s Analogy: Animal Experimentation and Anatomy in the Second Century C.E.,” Anthós: Vol. 8: Iss. 1, Article 9.
- Zampieri F et al. The clinico-pathological conference, based upon Giovanni Battista Morgagni’s legacy, remains of fundamental importance even in the era of the vanishing autopsy. Virchows Arch. 2015 Sep;467(3):249-54.
- Ghosh SK, Giovanni Battista Morgagni (1682-1771): father of pathologic anatomy and pioneer of modern medicine. Anat Sci Int. 2017 Jun;92(3):305-312.
- O’Neal JC, Auenbrugger, Corvisart, and the Perception of Disease. Eighteenth-Century Studies Vol. 31, No. 4, The Mind/Body Problem (Summer, 1998), pp. 473-489
- Brown TM, THE COLLEGE OF PHYSICIANS AND THE ACCEPTANCE OF IATROMECHANISM IN ENGLAND, 1665-1695. Bulletin of the History of Medicine Vol. 44, No. 1 (JANUARY-FEBRUARY 1970), pp. 12-30
- Roos AM, Luminaries in Medicine: Richard Mead, James Gibbs, and Solar and Lunar Effects on the Human Body in Early Modern England. Bulletin of the History of Medicine Vol. 74, No. 3 (Fall 2000), pp. 433-457 (25 pages)
- Frith J, History of Tuberculosis. Part 1 – Phthisis, consumption and the White Plague. JMVH. 2014; 22(2).
- World Health Organization, History of the development of the ICD. https://www.who.int/classifications/icd/en/HistoryOfICD.pdf
- Laennec RTH, A treatise on diseases of the chest and on mediate auscultation. 1834. Retrieved from: https://archive.org/details/b21987002.
- Marinker M, Why make people patients? J Med Ethics. 1975 Jul; 1(2): 81–84.
- King LS, Medical Thinking: A Historical Preface. Princeton University Press, 2016.
Please note that this transcript was made for editorial purposes, and therefore may not be the same as the ultimate episode.
This is Adam Rodman, and you’re listening to Bedside Rounds, a monthly podcast on the weird, wonderful, and intensely human stories that have shaped modern medicine, brought to you in partnership with the American College of Physicians. This episode is called “Cry of the Suffering Organs,” and it’s the first in a two parter that is going to examine the history of medical diagnosis. In that sense, this is sort of a sequel to Episode 15: Innumeracy, which I made all the way back in 2016. That was a very meaningful episode to me — it was basically a retelling of my senior talk that I did in residency (so you can see, I’ve been repurposing Grand Rounds since the beginning!), and it was a transition point to the Bedside Rounds you know today — the first really to focus on epistemology, “how we know things,” and to look at this from a historical perspective. It was also the first episode I produced as an attending — well, sort of, I guess, since I wrote it as a resident, and it was the transition to my first modern episode about Phineas Gage that really engaged in narrative storytelling as a didactic structure. In case you haven’t memorized every episode I’ve ever put out, it was generally about how doctors do very poorly translating test characteristics to actual patients, especially with screening tests, and advocated using Gerd Gigerenzer’s approach of natural frequency to make these discussions easier, especially with patients in the setting of cancer screening. There is nothing seemingly controversial there. But there were a lot of assumptions unstated. So in this two part series, I’m going to explore what it means to make a diagnosis from a historical perspective. First, I’m going to talk about the classical view of diagnosis — pathologic anatomy, or the cry of the suffering organs, which is an absolutely amazing name that I kinda wish we had gone with instead, and in the second I’m going to talk about how diagnostic tests challenged this model, leading to the development of test characteristics like sensitivity and specificity, and the integration of Bayesian probability.
This is also going to be the most “clinical” episode to date, because I’m not just going to tell you how modern diagnosis developed, I’m going to show you, by presenting three different cases over three hundred years of the same disease — starting with Morgagni’s anatomical dissections in Padua in the 18th century, Laennec and pathologic anatomy in Paris in the 19th, and Cabot and the clinicopathologic conference in Boston in the 20th. Along the way, we’ll ponder the origins of nosology, iatromechanistic theories of breakdowns in the human machine, the birth of the International Classifications of the Disease, and the rise and decline of the autopsy. This is a rather ambitious episode, so let’s get started!
We’re going to start with some very basic medical anthropology 101 to discuss some terminology. I want to start with a common framework, first proposed by Marshall Markiner in a fantastic essay back in 1975 called “Why make people patients?” It is in the shownotes, it’s short, and I highly recommend it. He describes three “modes of unhealth”. The first is what I am mostly interested in this episode — disease, which he defines as a physical, pathological process that causes deviance from a biological norm called health. Disease is the major focus of modern medicine — so much so that in the most cynical Foucauldian telling, the patient is the “accident of the disease.” The second category is illness, which is a patient’s personal feeling of unhealth. Illness often accompanies disease, but of course does not always — we talk about “silent killers” like hypertension, but honestly more troublesome there is a cornucopia of terms we use to label patients who have “illnesses” but not identifiable diseases — the neurasthenia of the 19th century is the “functional” disorder of the 21st. And then finally we have a third category, sickness. Whereas illness is an internal state of being, sickness is social and external. Disease, illness, and sickness all have complex interplays with each other. For example, we can think of breast cancer as the invasive, disordered growth of the milk production cells of the breast, but also the lump, the pain of lobectomy, the night sweats of taking a SERM, the fear of surveillance imaging. And it’s also a social phenomenon — the role of being a cancer patient in 2020 and all that entails.
In this episode, I am most interested in the concept of diagnosis, so I am exclusively going to be talking about diseases. And not just any disease — I’m going to be sticking throughout both these episodes to diseases of the chest — pneumonia and tuberculosis in particular. There are two reasons — the first of course is that, reasonably enough, I want to keep the cases as similar as possible so we can clearly see how conceptions of disease and diagnosis have changed. But there’s a major historical reason, and that’s that the major paradigm shifts in diagnosis that I’m describing in both this episode and the next were largely pushed by attempts to diagnose tuberculosis.
So what is a disease? You would think we would start medical school by defining what is basically the raison d’etre of a physician. But, of course, we don’t even really have a good working definition of what health is, let alone the types of unhealth. The WHO famously defined health in 1946 as a state of complete physical, mental and social well-being, not merely the absence of disease or infirmity, which on the one hand is an lovely utopian goal of a medical system, and on the other hand, can you think of the last time you had “complete physical, mental, and social well-being”? Especially in 2020? And diseases, whatever they are, change frequently. The most dramatic example in living memory is probably homosexuality, which over the course of a century went from a sexual act to a disease state with specific treatments, finally to being a normal state of being, when the American Psychiatric Association voted democratically in 1974 to remove it from the Diagnostic and Statistical Manual.
You can probably see me spinning up for a postmodern deconstruction of the entire conception of a disease and disability. But that will have to wait for another day (and in typical me fashion, I’ve been slowly working on an ambitious series on this for the past two years). But for now, I’m going to go with the definition of disease that largely defines our current episteme, outside of the field of psychiatry — a movement that historically was called pathologic anatomy. 21st century physicians love to use the word “lesion” so much that it seems almost trite. But the concept of a lesion was in fact one of the most revolutionary ideas of the past two centuries. Lesion technically means an abnormal change in the tissues. But the implications of this are huge — with the medicine of the lesion, disease lives SOMEWHERE in the body.
I’m going to trace the development of the medicine of the lesion, but first, we have to think about where disease “lives” in other nosological models. So the classic model of Western medicine, of course, is the humoral model — that health and disease are made by different balances of the four constituent body fluids, black bile, yellow bile, blood, and phlegm, as well as pneuma — the vital source of all life, coming from inspired air. If you have phthisis, for example, it may be from a surfeit of black bile. So where does this disease “live”? The answer, of course, is nowhere — or everywhere. A disease takes hold of the body, until it lets loose again. And this has major implications for treatment as well — because if depression and phthisis are both caused by an excess of black bile, the treatment for them should probably be the same. Taken to its logical conclusion, there are just four treatments in a humoral model. This reductio ad absurdum was described by the naturalist and philosopher JG Zimmermann in the 1770s, who wrote scathingly of humoralism: “I knew a certain Esculapius who has 50 or 60 patients every morning in his ante-chamber. He just listens a moment to the complaints of each, and then arranges them in four divisions — to the first, he prescribes blood-letting, to the second, a purge, to the third a clyster, and to the fourth, a change of air.” A quick aside here — Esculapius is a synonym of “Aesclepiad” or a follower of Aeusclepius, the Greek god of medicine, which was an acceptable synonym for “physician”, which of course means a practitioner of the physick spelled PHYSICK, which traces back to Latin and Greek via old French with two related meanings, the first “the healing arts” and the second “the study of nature” which is why physician and physicist are so different despite the same roots. Tangent alert: physick was also used in its old meaning (a medicine, or more specifically a laxative) until the early 20th century, and a few months ago, a patient told me she had taken “a physick, as my grandmother would have called it,” meaning senna, and it just made my day.
But Zimmermann was throwing shade because he was an advocate of a more botanical model of disease categorization — nosology. Zimmermann was a naturalist, so it’s no surprise that he would prefer a more botanical description of disease. Just as you would classify, say, a flower that you would find by, say, shape of the stamen or the number of petals, you would classify a disease by taking a careful history and fitting it into a pre-existing diagnosis. So to give you an example of Cullen’s nosology, which was very popular in the US — a fever might be in the genera of intermittent, which was further subdivided into every-day fevers, every three-days, or every four-days. Or you might have a flux, which was further subdivided into where the flux was coming from — catarrhus, the nose, or dysenteria, diarrhea
You can see why a nosological model is attractive. Theoretically, if you could precisely qualify the specific disease, you would be able to prescribe a specific treatment. So, for example, you might treat a quotidian fever with cool baths and blisters, but a tertian fever with cinchona. But where does the disease itself live? And this is the key point — disease is defined completely by symptomatology, so if the symptoms switch — say, you have a non-bloody diarrhea, but then develop blood — the diagnosis itself changes. Despite having a larger number of diagnoses and potentially specific treatments, disease still equally “takes hold” of the human body. There is no specific “place” that a disease lives.
The lesion does not exist in either a humoral or nosological model. It would take a completely different mental model to develop. And it should not be particularly surprising that this came from the anatomists, and their closely related fellows, the physiologists, largely centered around the medical school in Padua. The man credited for this is Giovanni Battista Morgagni. Now, I want to be very careful with my words here, because there has been a significant amount of hagiography surrounding both the early anatomists and Morgagni in particular, largely dating to the beginning of scientific medicine in the late 19th century. Its founders saw in Morgagni the antecedent of all their work. Virchow, for example, wrote:
“And thus we can say that, beginning with Morgagni and resulting from his work, the dogmatism of the old schools was completely shattered, and that with him the new medicine begins.”
Or from the head of the Scottish Medical Society: “All more eminent moderns, and even the men of our own time, although they work with new means and appliances, amid a flood of new light from physiology and histological anatomy, and amid a science of organic chemistry … are all of them successors and legitimate heirs of Morgagni’s labors and method.”
But the story of Morgagni contains considerably more nuance than that. In fact, just like talking about Hopkins and medical education in the US, it’s necessary to consider the entire university where Morgagni worked — the University of Padua. The University itself was founded in 1222 — one of the oldest extant universities in the world, founded when a group of scholars left the University of Bologna for more academic freedom. In 1404, Padua was subsumed into the growing Republic of Venice. The Venetian Senate decided that the already prospering University of Padua would become the educational center of their Republic, and gave both money and administrators to hire the best professors and guarantee academic freedom (in fact, Paduan and Venetian citizens were prevented from teaching in usual circumstances in order to prevent nepotism). After the Reformation, Padua was the only Catholic university open to Protestants. The organization of the medical school was copied all over Europe. There were five chairs — theoretical medicine, which meant studying the Hippocratic Corpus and Galen; practical medicine, largely focusing on Arabic commentaries, so Avicenna and Rhazes; anatomy and surgery; botany; and semiotics. Semiotics, the study of signs, is the closest thing to modern medical education, and was practiced in the town hospital, where students actually examined living patients.
Initially, anatomy at Padua was taught on a classical basis, but at the start of the 16th century, anatomists started to increasingly deviate from classical descriptions and describe what they saw in the human body. Vesalius is the most famous example, and has been held up as the first “modern” anatomist. But again, these anatomists did not see themselves as fundamentally breaking with the past. Vesalius himself wrote:
“Indeed, those who are now dedicated to the ancient study of medicine […] are beginning to learn to their satisfaction how little and how feebly men have laboured in the field of Anatomy to this day from the times of Galen, who, although easily chief of the master, nevertheless did not dissect the human body; and the fact is now evident that he described (not to say imposed upon us) the fabric of the ape’s body, although the latter differs from the former in many respects” (Vesalius 1543, p. 12). 7
Vesalius’ words are telling; he viewed his work as commentary on Galen, rather than refuting it. And in fact, in the Arab world, Galenic ideas had already been superseded, most notably Ibn Nafis, who accurately described pulmonary circulation three centuries before Harvey.
Regardless of the historicity of the story of Vesalius’s break with Galen, over the next two hundred years Padua hosted a series of anatomists and physiologists who slowly refined our image of the human body — Harvey, of course, but there are also a list of names that any first year medical student will recognize from all their eponyms: Colombo, Casserio, Wesling, Falloppia, Santoria, Wirsung, Valsalva — and of course, Galileo, who was a professor of mathematics while Harvey was working on his ideas about circulation. My point here is not to go over a history of anatomy — but to place Morgagni in a context not as a revolutionary genius, but as working in the context of two centuries of pioneering anatomists.
So the man himself. Giovanni Battista Morgagni was a talented anatomist and student of Valsalva, who became the first chair of anatomy at Padua at quite a young age — the same position Vesalius had held. He was a talented anatomist in his own right, and in his first major publication Adversaria Anatomica he describes a number of new anatomical findings that still bear his name. Unlike other prominent anatomists, Morgagni had been a practicing physician, and under Valsalva he had treated many of the patients that he later dissected. Like many anatomists before him, he noted that bodies showed pathologic changes when they were dissected — fluid in the abdomen, atrophied organs, purulence, clotted blood. But it was never truly clear what this had caused this. Morgagni, however, was an advocate of a new way of thinking about the body.
Morgagni was an iatromechanist, from iatromechanism sometimes also called iatrophysics, and closely related to iatrochemistry. That’s a lot of iatros. I haven’t really talked about this on the show before, because iatromechanism was only briefly popular in the 17th and 18th century. The dawning of the modern era had brought new challenges to the humoral model of medicine. The most prominent was from Palacelsus, who felt that the practice of medicine should integrate the latest theoretical science, in this case, alchemy. The details aren’t terribly important, but he felt that the function of the human body could be explained by chemical reactions, and that the mainstay of therapy therefore would be chemicals, most notably mercury. He called this “iatrochemistry”, iatros being Greek for doctor. Ultimately, this field would not last more than a generation or two after Paracelsus’ death, though the use of heavy metals in medicine persists arguably to this day. But in the 17th century, a related movement gained steam. Increased understanding of the laws of physics suggested that the human body might similarly run like a machine. Essentially, just as the mysteries of the heavens had been redefined in terms of gravity and orbits, iatromechanists felt physics and mechanics could accurately describe the functioning of the human body.
This is all a bit esoteric, so I want to give a concrete example. Thomas Willis, for whom the circle of Willis is named after, and who invented the word “neurology,” was one of the most prominent iatromechanists. In his masterpiece the Diatribae duae, he tackles the pathophysiology of fevers. After noting that wine spoils when it ferments, he reasoned that blood too must undergo a similar process. This fermentation leads to the blood “boiling over” just as in a boiling pot — hence causing a fever. This excess fluid explains why bloodletting works so well. So we have both a mechanistic explanation of fever — and conveniently a justification for traditional therapies.
Willis was certainly not a quack, and many of the people we now renown as early physiologists operated from a model of the human body as a Cartesian machine. Harvey, Boerhave, von Haller — all viewed the body as a series of hydraulic pumps, and when they broke, disease was caused. You can see here, in iatromechanistic thinking, the antecedents to the lesion — a defect exists in a specific location, though it’s not entirely clear how that links to specific diseases and symptoms noted during life.
Morgagni had this iatromechanistic machine in mind when he started his dissections. So just where did the human machine break down? Based on his dissections of the patients he had treated in life, he had an idea — the organs. As he would later (and possibly apocryphally) write, manifestations of disease — symptoms — are merely the “cry of the suffering organs.” Morgagni spent the second half of his career investigating this, and he published his life’s work in 1761 — at the age of 79, The Seats and Causes of Disease Investigated by Anatomy in Five Books, best known by its Latin shorthand De Sedibus. In it, Morgagni references 646 cases of patients he had seen in life and later dissected, attempting to correlate their symptoms while alive with changes in the organs after death.
And like I promised, I’m going to walk you through Morgagni’s thinking. We are going to start with a case from Letter 22, in which he describes pulmonary phthisis, or consumption in English, what we would call tuberculosis.
“A man of fifty years of age, having suffered great fatigues in journeying, and at length returning home, complained very much of a pain in his chest and a cough, by means of which he spat up a very little quantity of matter. He lay down most commodiously with his head bent forward. His belly and feet were tumid (swollen); he thirsted much; he breathed with a kind of panting motion, and was extremely oppressed after food. To these symptoms a spitting of blood was added. But this was appeased, the others, however, continued, except that almost a whole month before his death he was quite free from a pain of the breath. Yet on the three or four last days, he was not able to lie down on the right side. At length, a great quantity of blood burst forth from the lungs, he was suffocated.
After this very brief walkthrough of the case, Morgagni describes the man’s autopsy.
The belly was universally filled with a yellowish water, not unlike the serum we see swimming round the blood, after it has been taken away from a vein, and is coagulated. …
In the right cavity of the thorax, the lungs were found, except that they were very red in the inferior part. But in the left cavity the whole lobe was pallid, contracted, and hard, and contained a sanguinous matter. Besides it was so connected with the surrounding parts that it seemed to form one body, as it were, with them, nor could it be separated without applying violence.
Throughout the case, he references previous work by Wirsung, Ingrassia, and Willis, who similarly found “red hardness” in the lungs of patients who died of hemoptysis. And it’s in this observation that he makes this revolutionary connection — there was a connection between the clotted blood seen in the bottoms of the lungs, and his disordered breathing and “spitting” of blood, as well as his fatal hemoptysis. You could argue that this was a lesion, but there is still a problem — while Morgagni made these observations after death, he made no attempt to make diagnoses of these patients while they were still alive. If a patient came to him, complaining of swelling, early satiety, and chest pain, could Morgagni be confident that this “lesion” existed in his lungs? His only recourse would be to dissect the man’s body — and that is hardly an effective way to practice medicine.
So despite this “proto-lesion”, this is still not the conception of a disease that we understand today — from both a practical perspective, since there’s no way to diagnose the living, and also from a theoretical perspective, since the human body is not just a series of tubes with organs in the middle to become diseased. For that to come to fruition, we need to wait a half century after, in Paris, generally clustered around Corvisart in the Paris Clinical School. I have spoken so much about the Paris Clinical School on Bedside Rounds, so I won’t belabor the point to much — but seriously, if you want it to be belabored, please listen to Episode 43, 41, 38, 37, 34… well, you get the picture. There’s a reason people make fun of me for being obsessed with the Paris Clinical School. Very briefly, in Paris in the early 1800s, two parallel innovations would create modern medicine as we know it. The first was the continued focus of autopsy — from Morgagni, Bichat made the realization that it was not organs that were the seat of disease, but the tissues, of which he identified 21 different types. With this method, for example, Bayle was able to make the observation that the defining feature of tuberculosis was the tubercle — giving the disease its modern name — but importantly was able to show that tubercles were in other parts of the body than the lung, that tuberculosis was truly a systemic disease, not just a dysfunction of the organ. And separately, a generation of diagnosticians realized that they could detect these pathologic changes in the body with diagnostic skills — the physical exam. Initially Corvisart used immediate auscultation – that is, the ear on the chest — and Auenbrugger’s percussion, tapping the finger on cavities, but quickly the School introduced new techniques, most notably the stethoscope, invented by Rene Laennec in 1816. The melding of these two ideas — that disease existed in tissues, and could be identified by exam — was known as pathologic anatomy.
What did this medicine look like in practice? Fortunately, Corvisart, Laennec, and the other members of the Paris Clinical School kept detailed notes on their patients, and I want to read you a case description that Laennec published in his seminal “On Mediate Auscultation” where he first describes the use of a stethoscope.
On May 30th, 1818, a 42 year-old man with a 20 year history of various complaints of “fever, severe head-ache, and severe pains between the shoulders” presented to the Necker Hospital in Paris. The month before he had tried a “quack medicine,” but he started coughing up “extremely fetid expectoration” which brought him to the hospital. You know he must have been desperate, because at this time, hospitals were refuges for the desperate poor. Laennec took a detailed history of his present illness; the man complained of a frequent cough, often in paroxysms, and coughing up copious amounts of sputum.
Laennec performed an exam of his patient. There was good respiration on the right side, but a mucous rattle on the left with dullness to percussion in the axilla and bottom of the left lung. From this he made a working diagnosis — slight chronic peripneumony occupying the center of the left lung — peripneumony essentially means bronchopneumonia. After admission, the sounds in the patient’s left lower lung decreased markedly, and Laennec adapted his diagnosis — a “seropurulent effusion on the left side,” and over several days in the month of June he noted this effusion grow larger and larger, to the point dullness was up to the left clavicle. The poor man died on the 31st of July, two months after admission.
After his death, Laennec performed an autopsy immediately — I’m just going to read directly from the book, because it’s so dramatic: “on perforating the scalpel the left chest, a considerable quantity of gas, having the fetor of gangrene, made its escape. The left lung was compressed towards the spine and mediastinum but was united at its anterior edge to the cartilages of the ribs by means of a false membrane… The substance of the lung was flabby, fleshy, and contained little blood. It was firmer, almost hepatic, to the distance of half an inch around the excavation.” And it goes on like this for some time — in fact, most of the case is the description of the man’s body after death.
This, I think you will recognize, is our modern idea of diagnosis, and of the lesion. Laennec gathers information from the patient much in the way that I would — not just cataloguing symptoms, but to try and figure out where his disease lies, where the lesion is. So now the history helps narrow down the type of disease — you can imagine that Laennec was already thinking an infection of the chest from the minute the patient walked in — and he uses diagnostic tools, auscultation with the stethoscope as well as percussion with his fingers, to firmly localize the disease during life. The autopsy, then, is not just making associations between symptoms during life and findings after death, but confirming (or disproving) Laennec’s diagnosis. And it is an iterative process as well — with each detailed history and exam on the living, coupled with an autopsy when they died (and they often did), Laennec builds his diagnostic acumen.
While it’s tempting to think of the medicine of the Paris Clinical School as fundamentally the same as our own, we have to temper our enthusiasm. Different members had considerably divergent views on what actually caused these changes — Corvisart considered himself a “Hippocratic” and cared little for theory, while Laennec and Bayle had vitalist notions of a “life force,” similar to the ancient pneuma, or the Chinese qi. And they had no truly effective treatments — just as the nosologists and iatromechanists before them, they largely relied on old standards like bloodletting, purgatives, and medicinal herbs. But their system was useful, in that it could make predictions about patients — and more importantly, these predictions could be confirmed on the autopsy table when patients inevitably died. Old diseases were redefined in terms of “lesions,” and over the 19th century, scores of new diseases were added using this method — from specific types of cancers defined by mutated cells, to the new field of neurology, now defined by specific lesions in the brain and nervous system, and even in the late 19th century to infectious diseases.
Inderlying all of this new nosology was the autopsy. By the middle of the 19th century, the nosology of symptoms had died out, and there were several similar and overlapping classifications of disease, all based on pathological findings. In 1855, the first International Statistical Congress tasked William Farr with preparing a comprehensive classification of all diseases and causes of death that could be used in any country in the world. Farr’s system used five groupings — epidemic diseases, constitutional diseases, local diseases according to anatomical location, and then diseases that are caused by violence. This basic system — classifying diseases by their anatomical site, lasted with modification until 1889, though it was never universally accepted. Then in 1891, Jacques Bertillon, the chief statistician for the city of Paris, formalized an International List of Causes of Death, which was soon adopted around the world as a standardized way to classify deaths for vital statistics registrars. By the early 20th century, this had morphed into the International Lists of Diseases and Causes of Death, adopted by the League of Nations, and after WWII, was renamed the International Classification of Disease by the World Health Organization. The ICD, outside of the field of psychiatry, is the de facto nosology of medicine in the year 2020, now working on its 11th revision — in fact, I can’t finish out a patient encounter without listing a code from it — and the basis of this classification is the lesion, that disease lives in a particular anatomical site. If you peruse the ICD today — a dull task I know, you’ll still see this legacy. In fact, the first 20 codes, from 1A, all the way through LD95, all refer to pathologic anatomy, with the only real exceptions being the most controversial addition to the ICD-11 — Chinese traditional medicine codes — and a supplemental section for assessing functioning.
I know a lot of my listeners are medical trainees of various types, so it should be no surprise to you that pathological anatomy was enshrined as the bedrock of medical education. I don’t think there’s a finer example than the rise of clinico-pathological conferences, or CPCs, which were started by Walter Cannon and Richard Cabot here at Harvard in 1906, influenced by a recent innovation in legal scholarship — the case system. They envisioned a system where detailed case histories would be discussed by both resident physicians as well as expert discussants, including the primary treating physicians as well as the subspecialists and surgeons involved in their care. A clinical diagnosis is given, and then after the patient dies the “true” pathologic diagnosis is given and discussed by the pathologist and remainder of the medical team.
So let’s walk through the first published CPC, from 1923, in NEJM, which is a 24 year-old “American” student who presented to the Massachusetts General Hospital complaining of pain and dyspnea. The patient awoke on March 23rd, 1923 with a backache and fever and chills. This progressed with a severe pain immediately below his sternum, which brought him into the hospital. On exam, he was toxic-appearing and shivering. There was better lung expansion on the left than the right with diminished breath sounds in the lower third of the right lobe, with no wheezes or extra sounds. Two blood cultures were performed and were negative. An x-ray — a photofluorogram, since radiographs had not become standard yet in chest imaging, and you can see that the colors are revered in the facsimile, was performed and showed “findings are due to a process in the lung parenchyma most marked in the left base; shadows in the remainder of the lung fields are suggestive of a scattered infectious condition throughout both lungs probably off the same nature as the denser area noted.” The patient was admitted, and on hospital day two was noted to have dullness to percussion on the left side of the chest, as well as tactile fremitus and loud bronchial breath sounds from the scapula down. Similarly, he had worsened purulent sputum, which grew staphylococcus aureus. Unfortunately, he died later the next day.
After hearing the case delivered by his residents, Dr. Cabot discusses these findings and develops a differential diagnosis — in particular, he briefly considers both a malignant process as well as tuberculosis, before determining that the time course did not fit — this poor young man became sick far too quickly. His final clinical diagnosis is bronchopneumonia — the same diagnosis that his house officers had decided while treating him in the hospital
Then the pathologist who performed the was invited to discuss the findings. There were large abscesses abutting the pleura on the left, the biggest 7 cm in diameter. There was pus throughout the pleural space as well — all consistent with an empyema starting as a bronchopneumonia. Post-mortem cultures of this purulence grew both staph aureus and Pfeiffer’s bacillus. There was a lack of other systemic findings of septicemia — notably no petechia, and no thrombosed blood vessels. Based on this, he gave his pathologic diagnosis — pneumonic influenza, leading to bronchopneumonia with abscess formation and hemorrhages.
What’s interesting in this case, I think, is you can see how the addition of more sophisticated diagnostic tools didn’t fundamentally shift the medicine of the lesion, or of pathologic anatomy. Just like Laennec, Cabot takes a history of the patient, and builds a list of possibilities — a differential diagnosis. He then uses his diagnostic tools to help narrow down where the lesion might be — and the chest x-ray and sputum culture cognitively work in the same fashion as Laennec’s auscultation and percussion. And ultimately, the diagnosis is still confirmed with an autopsy, which has now integrated better histopathological and microbiological techniques, but still fundamentally proceeds in the same fashion.
We’ve now gone over three cases from three different eras in the development of the medicine of the lesion. So what does it mean to make a diagnosis? Well, the initial steps aren’t terribly different than what happened in a humoral, iatrophysical, or nosological model. Da Costa, who wrote the famous 19th century book on diagnosis called, well, Medical Diagnosis: Guide to the Knowledge and Disrimination of Diseases, in his introduction provides a step-by-step walkthrough.
One: discover the facts of the disease. In modern parlance, this means performing a history, physical, and appropriate diagnostic tests.
Two: Appreciate the significance of the facts, meaning recognize what of which you’ve found are “signs,” or in modern parlance, develop a differential diagnosis.
And finally, third: Use your “reasoning faculties” to fit what you’ve found to a final diagnosis.
This diagnostic model continues today, though with increasing recognition that medical diagnosis is an abductive process where the first two steps are iterative until a final diagnosis is settled upon
But a final assumption goes unstated, and that’s that the diagnosis can theoretically be confirmed on autopsy of the patient; that if the physician fails to identify and treat the disease during life, it’s sitting there in the body to be found.
Given this, iit should be no surprise that autopsies were incredibly popular and common throughout the late nineteenth and 20th centuries. A systematic review of autopsy rates at major medical centers in the United States suggests almost 80% of in-hospital deaths received an autopsy around World War II. And they appeared to make an actual impact in confirming and changing diagnoses. A random review of 100 autopsies from the 1960s, 70s, and 80s, at Brigham and Women’s Hospital showed that in all three eras there was a missed major diagnosis that might have changed management 10 percent of the time — and in an additional 12% of the time, there were major unknown diagnoses that would not have changed management. The authors cut right to the point: “We conclude that advances in diagnostic technology have not reduced the value of the autopsy, and that a goal-directed autopsy remains a vital component in the assurance of good medical care.”
From the tone of that conclusion, you can imagine that autopsy rates have plummeted in the US — they were about 40% when that article was written in the early 80s, and now are less than 10%. There is a whole body of literature out there about why this has happened — decreasing reimbursement, lack of pathologist time given how much the job description has expanded, family disinterest (and medical ethics now requiring consent for autopsy, which was not the case in the past), physician disinterest as imaging and other diagnostic technologies have decreased the mystery of what an autopsy might show.
But I would also argue that an intellectual shift in the nature of diagnosis was another cause. Because just at the height of the triumph of pathologic anatomy and the autopsy in the middle of the 20th century, it started to become apparent that the classic diagnostic model — and by extension, pathologic anatomy — had some significant problems. In this episode I’ve walked you through a “traditional” model of diagnosis — how we developed taking a history and applying diagnostics, recognizing signs and building a differential, and settling on a final diagnosis, all the while knowing that a “gold standard” existed inside the patient’s body, discoverable (at least theoretically, since we did get a lot better at treating disease) on autopsy. But in the sequel to this episode — and you’ll have to wait a few months for this, though I’m giving a version of this next episode in several grand rounds in the next few months — I’m going to talk about how physicians first formalized deficiencies in diagnostic tests, which we now call test characteristics, especially sensitivity and specificity, and how this has led to a re-thinking of the diagnostic process and even the nature of disease. So until then, that’s it for the show!
I hope you enjoyed the episode! Like so much that I end up doing on Bedside Rounds, this started as a small introduction to a lecture that I gave at the Boston University History of Medicine Society and then quickly grew into a behemoth that requires multiple episodes. But I also had a great time doing it — I actually read Laennec’s On Mediate Auscultation (or really, the section of pulmonary disease) years ago and still use some of the cases for teaching, so it was fun to revisit, and despite reading a lot ABOUT Morgagni, I’d never tackled De Sedibus before (and to be clear, I only read the section on diseases of the chest — the whole thing in enormous). I have links to English translations of both in the shownotes, if you want to peruse them yourselves. As long as I’m talking sources, I have to put a plug in for the spiritual source of this episode, which is the book Medical Thinking: A Historical Preface by Lester Snow King. It was published in 1982 (though thankfully remains in print), but the whole book is basically Bedside Rounds for an earlier generation. I’d say it’s my career’s goal to write a book like that, but I suppose this podcast in the 21st century version. If you can get your hands on a copy of this book, you will thoroughly enjoy it. And finally, some personal news — almost a month ago, my second child was born! I’m actually technically on paternity leave right now — but of course, I get antsy pretty easily, even with sleep deprivation, hence the slate of new episodes coming out in the upcoming months.
CME is available for this episode if you’re a member of the American College of Physicians at www.acponline.org/BedsideRounds. All of the episodes are online at www.bedsiderounds.org, or on Apple Podcasts, Spotify, Google Podcasts, or the podcast retrieval method of your choice. The facebook page is at /BedsideRounds. The show’s Twitter account is @BedsideRounds. I personally am @AdamRodmanMD on Twitter, and I think I’ve been pretty well-behaved lately.
All of the sources are in the shownotes, and a transcript is available on the website.
And finally, while I am actually a doctor and I don’t just play one on the internet, this podcast is intended to be purely for entertainment and informational purposes, and should not be construed as medical advice. If you have any medical concerns, please see your primary care practitioner.