Three Sciences of Medicine
Certainly we all agree that modern medicine takes its power from science. One kind of science concerns how we can do things. Let me call it ‘applied science’ for want of a name. The other kind concerns how nature does things. This is often called ‘basic science’ or ‘natural science’. Drug development and drug trials are obvious examples of the first. Mechanisms of disease are examples of the second.
These statements are so obvious I hesitate to make them, yet in such phrases as ‘evidence based medicine’, ‘basic vs. applied science’, ‘personalized medicine’, and ‘translational research’ the distinction, itself perfectly clear, can blur.
If basic science seeks to learn how nature does things, empirical science seeks to learn what things nature does. It is perhaps the oldest science. Were the star gazers first, or perhaps those who first enumerated the edible plants or named the large and the small cats and the wolves and elephants? Pure observation, measurement, and putting in order. Without it, nothing.
The benign presence looking down on this post is Karl Popper. His opus magnum, The Logic of Scientific Discovery, is one of two compelling influences on my views of science and medicine expressed here.
Process of Applied Science
The process seeks to fulfill a need. That need can be as various as all of medicine. Being a need for better tests, treatments, or ways to practice, it can be recognized by anyone: patients, doctors, even society as a whole. Furthermore, it need not be within the common boundaries of medicine proper. For example, the electronic medical record has come to be in response to a perception that such a record is needed, but that perception might have arisen not only from physicians but insurance payers, government, epidemiologists, and public intellectuals – as examples.
Kidney stone disease offers some remarkably clear instances of perceived needs that led to successful applied science. Think about shock wave lithotripsy and modern flexible ureteroscopes that were developed in response to evident problems of surgical stone removal. Similarly, the three extant 3 year trials of thiazide for stone prevention arose out of a need for evidence that such drug therapy would prevent stones.
The Imaginative Component
All science arises in part from imagination, and in our two examples that part is apparent. Shock waves have long been known as a way of transferring energy through fluid filled systems such as human bodies, and also as a way in which hard objects can be fractured. By adapting well established engineering methods and techniques to the problem of fracturing stones in kidneys manufacturers were able to create a workable instrument.
That thiazide diuretics lower urine calcium was noted as an observation as early as 1960 and individual physicians reported on the potential clinical benefits. Because kidney stones are often made of calcium crystals, and kidney stone patients often excrete more calcium in their urine daily than non stone forming people – a topic we have not as yet taken up on this site, it was a simple leap of imagination to see where thiazide could prevent calcium stones.
The Deductive Components
It is one thing to have in hand an instrument or a pill fashioned to do something needed – fracture stones or prevent their formation, and another to test the imaginative leap embodied in the instrument or pill against the realities of clinical medicine.
Both were known to have their actions – stone fracturing and urine calcium lowering – at the beginning of the process, but whether they could satisfy the needs – freeing kidneys of stones or preventing new stones – was unknown.
The logic of science was, in both cases, simple.
Many stones in kidneys are manifestly too large to pass. If you can fracture such stones the stones will change from objects of one size into a larger number of objects of smaller sizes, and these smaller objects will have a higher probability of traversing the ureter than the original stone before fracture. Shock waves can fracture stones, our instrument can fracture stones inside of models made to resemble kidneys inside humans, therefore our machine will improve the probability that large stones can be passed instead of surgically removed and will improve patient care provided the shock waves do not produce other unwanted effects – such a kidney injury – which reduce the value of the procedure.
Being composed of calcium crystals, stones must obey the laws of solution chemistry that pertain to all crystals. If crystals form at any particular level of urine calcium excretion, then lowering urine calcium excretion will lower urine calcium molarity for any given urine flow rate. Therefore, for any random array of flow rates, thiazide will lower urine calcium molarity and the formation of calcium crystals and stones if thiazide does indeed reliably lower urine calcium excretion over the time scales needed, of many years and do not produce other changes in the urine that foster stones.
The Importance of the Deductive Components
The mind, being agile, easily recognizes these laborious exercises in deduction as unnecessary – but that is illusion. In our two cases, the deductive links are indeed sound, and lead one to undertake the onerous business of trying the machines or the pills to see if they can satisfy the needs that gave rise to their original discovery. However, careful inspection often reveals errors or breaks, and when present these abrogate testing and force on onto another path. Or might do so. In the worst cases, a break in logic is found after failed testing.
The Experimental Component
Often called ‘trials; they are indeed a trial to judge the truth of the deductions made from the proposed answer to a perceived need. The trial can be conducted only for patients who pose the need which the new treatment has been created to satisfy. One must make sure that the need is present and specify the setting well enough to delimit the range within which the treatment can be assured to work. Then what one means by ‘work’ must be specified so as to compare the final result with some set of criteria that are relevant, and fixed so others may use them in choosing the treatment for new patients.
This can sound stuffier than it really is. Consider refrigerators, and Consumer’s Guide. People need refrigerators that are not overly expensive, do not waste electricity, run without excessive repairs, and, above all, keep food cool but not too cool. Manufacturers have made huge numbers of models to compete for the attention of buyers, but buyers want, if possible, to buy those that most fully meet their needs. The setup of a refrigerator trial and of the trials for shock waves or thiazide do not pose logically different problems of interpretation or design. They differ in material details, but only in that way.
The Meaning of a Positive Result
Being about how we can fulfill a need, a result tells us yes or no, not as a simple binary but a graded outcome consequence. Shock waves will fracture stones and let some pass, and over time it has become apparent that for some kinds of stones it has a reasonable stone free outcome and reasonable safety under certain circumstances. For thiazide diuretics as a way of preventing stones, we have 3 three year prospective controlled trials which indicate a reduction from about 60% to about 20% of patients forming a new stone over 3 years.
Although time and experience will always modify the implications of such trial findings, and new and better ways of meeting needs will general supplant older ways, the results will always remain ‘true’ in a sense. The ways of meeting the needs worked as specified, and nothing that comes later will radically alter that statement.
Process of Basic or Natural Science
The process seeks to satisfy a desire. The objective world, the world we live in, present itself to our senses, but the causes of what our senses perceive are not presented. Once again kidney stones provide wonderful examples.
The common urinary stone is obvious, and likewise that it passes, causes pain and bleeding and obstruction, and must be removed in some cases.
But how the stone came to be is not obvious, and even the phrase is filled with ambiguities.
We, for example, have found out new details of ‘how the stone came to be’ as an overgrowth on interstitial hydroxyapatite plaque. But like some unending hall of mirrors, or a set of Chinese boxes one inside the other, each revelation presents new desires: In this example, how the plaque forms, how the urothelium gives way so a stone can grow over the plaque, and so on into an endless series of open questions which are essentially about how Nature does things.
Even the stone itself, once formed and presented as an object, presents mysteries. Stones are filled with organic molecules of urine origin. What are they doing? What did they do? Even, being so numerous, what are they named?
One may say these issues of cause have practical meaning. If we knew, then we could do something to fulfill some need. But in reality, scientists are not so pure – or perhaps are so pure – and work in thrall to desire, curiosity, and the thrill of the chase.
What, then, are the objectives of basic or natural science?
They are nouns, often multiple, modified in some cases by verbs indicating some interactions between them. ‘How plaque forms’ is not an objective so much as a general or overall desire to know. ‘What causes plaque to form’ is to indicate an objective which can be satisfied by nouns – hypercalciuria, for example, oxidative stress, osteogenesis. Likewise, ‘which urine proteins predominate in calcium kidney stones’ is an objective which can be satisfied by nouns – calgranulin 2, for example. But these two differ: The first concerns how nature works, the second only concerns details of what is already available to our senses – aided by machines. ‘What organic molecules in stones do’ is an example of how nature works, and is utterly not available to our senses.
The Imaginative Component
The problem with objectives in natural science is their manifest entrapment within the known and consequent incapacity to forward discovery. They are desires that cannot progress matters. This is exactly what objectives in applied research abundantly do: They move science forward along lines desired by some portion of society whose needs are known and whose needs are justified.
To say ‘what causes plaque to form’ is to state a desire but not a direction that can be followed. To say ‘find a way to fracture stones in kidneys’ is to state a desire and a direction that can be followed: Toward any and all known technologies that are known to fracture hard objects.
Whereas in the latter case, a way to fracture stones, imagination ranges over the known and selects some potentially useful way, in the former cases, what causes plaque to form, or what organic molecules in stones do, imagination must come up with a proposition. Such a proposition is entirely imaginative, by its very nature, and such imaginings are the only known way to progress from what our senses can grapple with directly to that vast region of nature concerned with causes which our senses cannot ever reach.
Science attaches this name to the imagining or guess, if you like. It also offers two alternative names, conjecture and speculation, as contrasts. The first of these two arises from the idea of throwing together – in this case the desire and the dream. The second arises from the latin word for a watch tower, and implies looking out at – in this case the desire and by implication having some idea about what you are looking at. Hypothesis arises from Greek and means what I put under, in this case the base for a proposition about cause.
The distinctions have real meaning when you want to proceed beyond naming, because an hypothesis is meant to be taken as true for the purposes of proceeding, and the manner of proceeding is to assume it is true and deduce from that assumption some features that must be present in the real world. This link to deduction is what makes vision more than speculation or conjecture.
The Deductive Components
The profound difference between imagining a way to fracture stones or prevent them by employing what we have to work with and imagining – in our examples – what causes plaque to form or what organic molecules in stones do places remarkable burdens on the deductive component of natural science compared with applied or practical science.
Unlike applied science where one must only envision a workable and convincing test of a way of doing something, here one must find some way of making manifest in the real world the truth of what is, at heart, merely an imagined vision of how nature works.
The approach is particular, specific, and the very heart of science as it has evolved in the West since the Renaissance. One must deduce predictions about what one must find in the real world if the imagined vision were true. As an example, if plaque formed because of osteogenesis then one must find, in regions of plaque formation, key bone genes such as Runx -2. Failure to find such genes defeats the vision. As another, one might envision that the organic molecules in stones ‘glue’ tiny crystals together so as to make up the bulk and size of the stone. If this were true – an example from real life – then in phosphate stones whose carbon atoms would all come from the organic component one must find very short and specific inter-atomic distances between calcium and carbon atoms; using NMR this prediction has indeed been tested.
The Importance of the Deductive Components
In applied research, the deduction ends with the idea of ‘if it works’. In basic, or natural science, the deduction ends in a specific finding that is necessary if what is in fact a dream is a true dream. The very word ‘true’ means, simply, that the dream can predict via one or more logical statements some aspect of the real world that is verified by experiments designed specifically to test that prediction.
But that kind of truth is not the same as the truth of whether shock wave lithotripsy works. The latter is absolute: It does or it does not, and that is that. If from a dream one can deduce something that can be verified by experiment the dream is not necessarily true as a statement about how nature works. It is simply useful as an assumption which has predictive value.
Dramatic examples are better found outside the shallow waters of stone research. Consider peptic ulcer and gastric acid. In states of high acid production ulcers are prominent; in the absence of acid they do not occur. Drugs that reduce gastric acid cure peptic ulcers. Acid is certainly erosive. The picture – acid erosion of the gastric lining leads to ulcer – could not be more convincing, more powerful as an image and a source of action, and could not be more wrong. We now know it was completely wrong. You need bacteria for gastric ulcers. So a century of correct predictions did not make the dream true, yet it did lead to useful drugs.
The Experimental Component
This is not different in kind from that of practical science. Something is predicted to happen or be found under specified circumstances of an experiment – a contrived experience – and professional scientists know how to contrive the experience and be sure if it has or has not happened. It is true that imagination and deduction are the soul of science, but the profession is about this difficult and crucial business of measurement without which we have nothing. I find nothing different between a drug trial and some exotic NMR study concerned with inter-atomic distances even though the two experiments look utterly different. They are and they are not. The confusion is merely superficial. Some scientists are trained to do one kind of measurements, others another kind. That is all the difference.
The Meaning of a Positive Result
I have already alluded to the fundamental weakness of science, that positive predictions do not establish truth of a vision whereas they certainly do establish the truth – value – of a way of doing something. But there are subtle meanings here that bear special remarks. For example, high urine calcium excretion is so inviting a ’cause’ of calcium kidney stones. The links – higher excretion, higher supersaturation, crystallization – are seemingly tight and in fact incontrovertible. Lowering urine calcium excretion with thiazide does reduce new stones. Even epidemiology agrees; people who form new stones over decades of observation have higher urine calcium losses than those who do not.
Is this not just like peptic ulcer and gastric acid? Are we not the same satisfied generations of scientists in possession of a certain basis for action and new science as they were who pushed forward acid suppression drugs? Nature seems forever bemusing, but nature is indifferent so far as one can tell. So it must be ourselves who are fond of being bemused, who like to fall in love with our dreams.
The word mean the entire collection of what is known in an area, but it is specialized to work in basic science. The hypotheses – dreams – which have been tested and found robust become the accepted truths in a field until they are overthrown, as will always be the case. For drugs that have worked, we do not need the idea of theory – they work and will be superseded, but that they work will not be overthrown. Gastric acid as the cause of peptic ulcer worked and was superseded and proven wrong utterly and discarded as a truth. Theory is like the Olympian gods and, if I might say so, most deans and department chairs: powerful and doomed.
Empiricism: The Third Science
One example, the organic molecules in kidney stones, slipped by without objection, I believe, although perhaps some readers were brought up short. These molecules are there and can be measured and named. But to do this naming is neither to get a new way to do something or to know how nature does things. Likewise, the manner in which stones form in the kidneys of people can be observed directly during stone surgery, and has been. It is indeed various.
Naming the organic molecules and describing the manners in which stones form in kidneys do not tell us how nature brings about what we can find in the real world. Nor does such describing tell us how we can do what we desire to do. These kinds of activity simply define the real world with a greater detail in the area one wants to study.
What do we call this defining in a greater detail? Is it part of science or just looking around?
It is both. For science begins with desire that arises from contemplation of the real world. One cannot seek to understand the origins or causes of an imagined world; that would be a distraction and a fool’s waste of time. But often the real world is more like the organic molecules in kidney stones than a gastric ulcer. The fine details of the stone, or of the kidneys that produce and harbor the stones, or of the seemingly coherent clinical phenotypes of stone formers can be critical to having a reality from which productive visions can be had.
This kind of science is called empirical, and science. Empirical because ‘a source of knowledge acquired by means of observation or experimentation.’ Scientific because the acquisition is organized so as to produce reliable knowledge and performed using the skills and instruments that belong to professional scientists. Like practical science, empirical science is true to the extent it was done right. One might say that all of the experiments done to test the dreams of basic scientists add to our knowledge of the real world, being observations made in the real world by skilled professionals, so that an offshoot of even failed dreams is some increase of knowledge.
The Extreme Centrality of Mandarin Scholarship
One hesitates to mention this issue, but one massive reason for this very site is to bring forward what is known in the service of practical and basic research as well as medical practice – something we have not as yet alluded to here. One may discover what is already known, but that is a distraction and a foolishness which will destroy any scientific career if it occurs very often.
To find a way of doing something one begins with a full knowledge of all that has already been made available, and all that might be re-processed for some new task. To guess how nature might have produced some aspect of the real world is to know that portion of the real world as fully as is known, to know all of the dreams that have been brought forward – those brought down by their failures to predict and those not yet brought down – and all of the tests mounted in pursuit of their predictions. Naturally, the relevant methods of science are the very mark and make of a professional scientist, who must know them as a professional, and all their uses.
One might say science begins with need and desire, scientists with a perfected knowledge.
This Site and the Three Sciences of Medicine
Perhaps my examples have already made the point, but I doubt they are sufficient.
The first science, concerning how to do things, relates most uniquely to patients, whose concerns are rightly that stones stop forming and those present somehow make the most graceful forms of exit. But, how things got to be as they are cannot be of no interest to them.
The second science, concerning how nature does things, relates most uniquely to scientists, whose concerns are rightly the advancement of new knowledge about the world. But, being practical, few would disdain to find a new way of doing what patients might benefit from. I do not believe that scientists are ever only one way – concerned with nature or with the human condition.
I believe, although I offer no proof at all, that the two sciences combine most in the lives and thoughts of physicians. On the one hand, science concerning how to do things is the very foundation of medical care – the trials of treatments and tests – including prognostic guides. When evidenced based writers speak about evidence based medicine they mean precisely this.
On the other, the scientific study of how diseases come to be, the so called mechanisms of disease, cannot be totally absent from the physician’s mind. For example, even if only one molecule of the urine proteome proves – someday – to have prognostic power concerning recurrence, the other molecules that might permit stones through their abnormal characters would remain of interest.
For these reasons, the site contains and will proliferate and broadcast both sciences. And, as an ongoing theme, consider exactly what purposes the second science, of how nature does things, serves physicians and their patients.
The third science is everywhere, for practical and basic science begin with how the world is now. For physicians, empirical science concerns such matters as how groups of patients present themselves, coherent syndromes and phenotypes. The site is already filled with examples, once you stop to think about them.