Which is the role of the unknown in contemporary Medicine? Can Philosophy provide some help in dealing with it? To address these issues, we publish an interview with Dien Ho, PhD, Associate Professor of Philosophy and Healthcare Ethics, Director of the Centre for Health Humanities at Massachusetts College of Pharmacy and Health Sciences (MCPHS) in Boston. His A Philosopher Goes to the Doctor: A Critical Look at Philosophical Assumptions in Medicine has been recently published.
Q. In this Evidence-Based Medicine world, what is the role of the unknown?
DH. Evidence-Based Medicine (EBM) typically ranks the quality of evidence from clinical trials. Randomized clinical trials or meta-analyses containing them usually stand atop with anecdotal evidence or bio-plausibility further below. Although there are serious epistemic questions one can ask about EBM (e.g. what happens why one treatment has limited RCT evidence while another has extensive anecdotal evidence? How do we balance quality with quantity?), many of the challenges with unknown diseases are conceptually prior to EBM.
In order to identify the aetiology of disease and potential treatments for them, we have first to determine that the physiological presentations are in fact pathological. In other words, whether some human condition constitutes a disease or not is a question we must answer prior to exploring how EBM can help us tackle them. The definition of disease (and relatedly health) is not a simple biological matter. There are plenty of “normal” biological functions that we pathologize and treat. The front-loading of calcium absorption (in our life span) might make sense from an evolutionary point of view. Prioritizing our ability to repair our bones at a younger fertile age at the expense of more brittle bones at an older age maximizes our genes’ chances to replicate. In a world where most humans died before 45, we never had to pay for the front-loading. However, for those who live well beyond our “natural” expiration, fragile bones become a real health risk. It is hard to imagine that anyone would suggest that osteoporosis is not pathology because there it is the result of evolution.
The categorization of a condition as pathology must be determined extra-biologically. No amount of empirical research in clinical trials can possibly answer the question for us. In this sense, before EBM becomes relevant, we must ask some basic social and philosophical questions regarding the pathologization of a condition. This is particularly germane since the medicalization of human conditions has been a growing phenomenon since the advent of modern medicine. Our increasing ability to control physiology can easily tempt us to the conclusion that human conditions must be controlled. It behoves us to keep in mind that pathologization has historically been often guided by social norms, bigotry, and financial incentives.
Q. Doctor struggle with the unknown, also because patients would like to receive certainties. Can Philosophy provide some aids to study, explore and cope with the unknown?
DH. One of the recurring themes that I have encountered in my research on the philosophy of medicine is the ubiquitous assumption that medical sciences can map the taxonomy of diseases onto some real objective categories in the world. Or to put it more bluntly, medical sciences can get things right in that it can sort diseases into real etiologies.
At the same time, even the most cursory look at disease taxonomy would tell you that we often lump diseases together by their symptoms and presentations. These functional diseases have no obvious underlying common aetiology; the only thing linking them together is that they look similar. Melancholy, for instance, covered everything from depression to the anxiety of being marginalized. Or, take focal segmental glomerulosclerosis (FSGS)—a disease categorized by the scarring of kidney tissues that can lead to kidney failures in adults and nephrotic syndrome in children. The causes of FSGS include obesity, lupus, heroin usage, and HIV. Primary FSGS is categorized as an idiopathic condition with unknown causes. The only thing that links all these conditions under the same FSGS heading is their surface presentations. Indeed, whenever a clinician encounters a diagnosis that seems to lack the common underlying aetiology, it is typically characterized as an idiopathic version of the disease. Functional presentations seem to play a key role in how we taxonomize diseases.
In order to claim that medicine can map disease taxonomy onto some real objective causal features of the world, we need to assume that there is something to get right about. This is obviously a deep philosophical question. It is, in fact, a variation of the classic debate of scientific realism. More importantly, this question of whether medicine can get things right also forces us to think about the ultimate aim of medicine both as a research discipline and as a therapeutic modality. If we can successfully alleviate all the unwanted symptoms (including impending mortality!), does it matter if medicine gives up on the task of getting it right? If clinical medicine can successfully improve our lives, does it matter that we know nothing of the underlying aetiology (if there is something to be known)? In this respect, we begin to blur the line between unknown diseases and diseases that we are incapable of treating. In philosophy, to collapse metaphysics and epistemology is a grave sin; in medicine, its service to human well-being might make the conflation less sinful.
Q. Undiagnosed diseases represent a challenge: do you think that only the genetic study may give a solution? (Personally, I think that if every DNA mutation would represent a disease, we all will be affected by millions of daily disease).
DH. Genetic advances have clearly given us tools to make monumental changes to our phenotypical expressions. There is also little doubt in my mind that many dreadful human conditions are the results of genotypical deviances. Still, understanding of how our genes work does not answer questions as to whether a genetic variation is pathological. As with the definition of diseases, how we delineate pathological genotypes from their healthy counterparts is not an empirical question. I often ask my students this question: what physiological or genotypical discovery must occur in order for you to conclude that homosexuality is a disease? For me, the answer is none. Suppose it turns out to be the case that homosexuality can be traced to some genetic variance that causes an above-average secretion of a certain hormone in utero. Would one conclude that it is thus a disease? We consider the removal of homosexuality as a sexual disorder from DSM as significant progress in psychology. I suspect that the change (which took place over decades) came about as the result of accepting homosexuality as an equally legitimate form of human sexuality. The science, in this case, follows the ethics. Indeed, my sense is that science always follows ethics; it is just that we often are blinded to how values and norms implicitly and explicitly guide our scientific practice.
There is also another issue that is worth keeping in mind. Within the human genome, the vast majority of our genes have either no phenotypical expressions or have decidedly bad ones. Many of these genes are “legacy” in that they have stayed in our DNA from our ancestors. From an evolutionary point of view, they get passed on generation after generation because they do not impose any reproductive cost to us; they are free-riders, so to speak. But there are also genes that benefit themselves at the expense of the host organism. A particular segment of genes found in the genome of common fruit flies shortens their lifespans by 15%. It would appear that these genes should not survive since their host are less likely to reproduce in virtue of dying younger. Typically, any gene has a 50% chance of passing onto the next generation. But these very special genes have somehow managed to tip the scale in their favour such that they have a 60+% chance of passing on. This slight reproductive advantage gives the genes a survival edge even though they are sabotaging the very organism whose genome they are a part of. The moral of the story is that thinking of our genes as one unified entity with roughly the same reproductive interest is incorrect. Not only is it the case that what is in the interests of our genes might not be in our interests qua people, within the same genome, not all segments have the same interests. I like to think of our genes as a giant train station. Just as it makes little sense to talk about THE destination that everyone wants to go, it makes little sense to talk about OUR genetic makeup. We are a messy anarchy of competing genetic agendas that happens to settle on a détente.
When genetic research ventures into the domain of what is pathological and what is not or what is human nature and what is not, they are making the mistake of thinking there is a simple and obvious way to delineate what segments of our genome constitute US and what does not. There is no more a single genetic identity as there is a single American culture.
Q. You have recently given a lecture at the MIT on the unknown placebo effect: what is the main message that we can learn from a placebo?
DH. Implicit within much of the placebo research is the idea that, somehow, their working and nature are mysterious and perhaps even unscientific. This, of course, presupposes that non-placebic interactions are clearly understood and fall under some clean causal mechanism – two assumptions that are highly questionable.
Placebos are fascinating for me because they seem to challenge the way we practice clinical medicine. Ultimately, there is really nothing mysterious about placebos. Whenever someone says placebos are about mind over body, I point out that your mind is also your body (Cartesian dualism notwithstanding): placebos are really about body over the body. To be sure, placebo effects are likely all in your head but then, again, so is your brain.
The existence of placebo effects strikes me as rather well established, especially in terms of their analgesic power. A common criticism of placebos is that they are only effective for “subjective” conditions such as pain, anxiety, and fatigue (outcomes that depend largely on patients’ reports) but that they are ineffective for “objective” conditions. A few things to keep in mind. Firstly, there are some studies that show placebos can have remarkable effects on “objective” conditions such as T-cell count, mobility, and blood pressure. One of the most incredible cases I read involves a young man who, in an attempt to commit suicide, ingested all the anti-depressants that were given to him in a clinical trial. His blood pressure dropped to 80/40, and he needed fluid infusion in order to maintain normal pressure. The attending in the emergency room contacted the principal investigator of the study and learned that the young man was in fact, in the placebo arm of the study. Instead of ingesting anti-depressants, he had ingested 29 placebo pills. Within 15 minutes of learning of the news, his blood pressure rose back to 120/80. The case review describes the young man as having overdosed on placebos. Similar “objective” results have been noted in cases of sham surgeries in which clearly indicated fractures had not been treated, but patients regained mobility.
Secondly, the dismissal of placebo as being entirely subjective stems from modern medicine’s myopic focus on “measurables”. Not only do we ask patients how they feel as a way to assess the effectiveness of a treatment, but it also shifts clinical attention away from alleviating patients’ discomfort to obsessing over numbers. The bias to treat and the bias to medicalize, I suspect, come partly from our emphasis on numbers.
Finally, there is a basic assumption that we can draw a relatively clear line between subjective and objective measurements. Philosophers of science have longed known that even the most precise measuring device ultimately rests on a plethora of values and norms. Briefly, so long as conceptual taxonomy (blood types, disease grouping, or sex) depends on values, any device that aims to measure these taxonomies would obviously inherit the same values. The idea that objective measurements are somehow more immune from subjective biases and values is a pipedream leftover from the days of positivism–a view that has largely been abandoned by those in the philosophy of science.
One thing that I learned from my research on placebo is how much of the therapeutic benefits of modern medicine depends on certain performative. From how we label a drug to the colour of a pill, from the belief in the therapeutic benefits of medicine to a clinician’s outfit, subtle features of the healing environment contribute to therapeutic effectiveness. Indeed, rather than thinking there is a distinction between the healing environment (a doctor’s office, her demeanour, one’s trust in medicine, etc.) and the intervention itself (chemotherapy, antibiotics, physical therapy, etc.), it is better to think of the entire healing experience as one large unit with no significant distinction between the context and intervention. Placebos have typically been relegated to the contextual part of the therapy. If the picture I am proposing is correct than the line between placebos and standard treatments does not exist. At the end of the day, what matters is improving patients’ well-being; how much of it is due to the context and how much of it is due to the intervention is not only unimportant, it is in principle impossible to discover (as I have argued elsewhere).
Q. Naming the disease: most of the rare disease have the name of their discovery. Very different from what happened to the Parkinson’s disease, that once discovered by Parkinson was given the name of “paralysis agitans” by him… Much better name to explain what was happening in the body. Do you think that language should help in giving the name to the new (and old disease) replacing confusing names used more as patronymic names?
DH. I recall a similar story for Alzheimer. Apparently, patients were understandably distraught to learn that they have been diagnosed with “the gradual deterioration of the mind.” “Alzheimer” was introduced in order to mask the severity of the condition. In my clinical experience, I have seen oncologists refrained from saying “cancer;” instead, softer terms such as “mass” and “unusual growth” took the place of the c-word. I have some qualms about intentionally misleading patients with terms that one hopes would avoid the blowbacks from difficulty diagnoses.
A wiser approach is to emphasize the lack of certainty in our diagnoses and treatments. As a well-known physician once said to me, “We know about a third of what we do is probably useless. The tricky part is figuring out which third.” Medicine, both as a research and therapy, contains far more uncertainty that it projects. A bit of humility not only eases some of the ethical problems with paternalism; it also invites patients to participate as equals.
On a flip side, however, I want to offer an alternative viewpoint. Perhaps the air of confidence is necessary for the therapeutic effectiveness of medicine. If patients discover that doctors know far less than they say they do, it might actually make routine treatments less effective. A recent case in Australia illustrates the complexity of the problem. A pharmaceutical company began to market brand ibuprofen in different boxes, each highlighting the special purpose of the drug. There was one for headache, one for backpain, and so on. However, all the different types contain the exact same drug of the exact same dosage. Yet, they were sold at different price points. After public outcry, the pharmaceutical company pulled the drugs off the shelves. As much as this seems like a great example of the need for transparency, I am not entirely certain that it was good for medicine and patients. A drug works better if we are told that it specifically treats a certain condition. “Backpain” drugs work better for backpain than a “headache” drug even when they contain the exact same ingredients. By “lifting the curtain,” we might very well demystify the performative that constitutes a part of the therapeutic magic. It is often said to those interested in politics that one does not want to see how a sausage is made, and one does not want to see how a bill is passed. Perhaps we do not want to see how medicine works. Seeing, it turns out, might be unbelieving.
