Blog Post

Medical Research and the Myth of Scientific Truth

I.

In "Lies, Damned Lies, and Medical Science" (Atlantic Monthly, October 2010), David H. Freedman takes jabs at what he describes as the scientific fallacy behind most medical research. Covering the pioneering work of Dr John Ioannidis, a meta-researcher who has made a career scrutinizing thousands of published studies, he concludes that "as much as 90 percent of the published medical information that doctors rely on is flawed. ” However appalling this statement (true or not) might sound, the real problem is that we expect it to be otherwise: the trouble is not (or not only) that medical science fails to deliver, it is that we, as a culture, have a mythical image of what science and medicine can do. We believe in medicine as the purveyor of objective knowledge, but we fail to recognize this attitude for what it is: a belief, which should be examined as such in its relation to a more general belief system, what we call a culture.

David H. Freedman's fascinating, if scary, article reminds us that medicine has only recently joined the ranks of what we now call “hard-sciences.” As Freedman details the numerous ways that randomized trials fail to deliver consistent, or even simply applicable results, we start to wonder if this recent promotion was not a hasty one. For centuries, medicine has been envisaged as magic or religious ritual, as a techne or craft for its practical sides (surgery, obstetrics), or as a literary discipline relying on the careful reading and glossing of canonical authors from antiquity (Galen and Hippocrates) for its theoretical part (anatomy, pharmacopea).

Now that medical literature is secularized and based on facts and observation, one would assume that medicine would apply the most rigorous scientific methods and provide evidence-based knowledge of ailments and conditions to guide doctors. But as Dr. John Ioannidis’ work and Freedman's article show, sound methodology and selective publications criteria are not the norm. Nor are they in themselves enough to deliver the kind of hard-core truth that patients expect. Both authors point out that randomized trials purports to prove things that often yield little (if any) therapeutic value or to support claims for which there is just no sound methodological protocol because the reasoning behind them is flawed. This points out to a more fundamental issue: our culture’s mythical representation of what medicine and science (rather than scientists) can do. Scientists only respond to the demands and expectations of the culture and economy that produce and fund them. According to the article, the medical community was rather relieved to be exposed in Dr. John Ioannidis’ publication, as if they could finally stop pretending to be doing what they knew all along was not their task: provide bulletproof "truth" about ways to master our bodies.

Today’s faith in medicine as science (and in science as the quest for absolute truth, and in technology as the solution for everything) is a dominant misconception that does not make the need for medical research and the tremendous service that doctors do to humanity less real. Yet, this misconception is costly financially, humanly, and intellectually: the blind belief in drugs or discoveries leads to at best ineffective treatments, at worst to potentially dangerous ones. It substitutes the hype for the tried-and-true and undermines our collective ability to face our all too human and inescapable vulnerability. We can only be grateful to David Freeman, Dr. John Ioannidis and his team of scholars for debunking the myths behind medical “news”: this act of sanitary questioning forces us to think about the body as something more complex and precious than a machine to be fixed. The lessons to be drawn from this investigation are certainly not to cut funds for medical research, but to acknowledge that our expectations as much as our methodologies need to be examined.

II.

I like to think that research, in sciences and in the Humanities, is as much about the process as it is about the results: it is a quest spurred by intellectual curiosity that seeks to make sense of things for the sake of thinking them through. And it should be a pursuit guided by an ethics of knowledge. With that in mind, the admission of our shortcomings in understanding and mastering our bodies is in itself a source of knowledge — maybe not a practical knowledge that leads to revolutionary drugs (but sometimes withholding potentially harmful treatment is as positive as it gets) but an intellectual knowledge that could serve us morally, philosophically, emotionally, and politically, in the ways we distribute resources, help patients, and understand ourselves as mortal beings. The cultural framework within which we approach pain and illness is as important for the healing process as the "objective" treatment plan offered. Health is never just a question of body mechanics: if the mind participates in significant ways through cognitive processes and homeostatic responses, then cultural assumptions learned and unlearned play a part too. Mapping out what we don’t know and why is as important as making unsustainable claims about our superpowers.

The underlying ideology of man’s power over the living world is indeed a treacherous and pervasive one: as David H. Freedman’s points out, patients are the first to demand drugs and are the strongest believers in their effectiveness. This, in itself, is only human: who would not want to try anything when in pain? And it could be put to good use: the placebo effect, which accounts for up to 30% of positive results of any kind of medical intervention, real or sham, could be recruited to improve patients’ morale and medical outcomes. But it is deemed unethical to offer sugar pills instead of “real” drugs, even though the latter might be more dangerous than effective. Since most drugs, as Dr. John Ioannidis shows, already seem to either rely on absent or flimsy evidence or to cause significant side-effects, why not try to devise a system that would exploit human nature’s suggestibility and drugs’ aura, without the cost and consequences of downing massively invasive medicines? In his "Essays," Montaigne famously recalled the anecdote of a friend who confessed to him that he feared becoming impotent the night of his wedding. Montaigne gave him a “magic” medal to wear in a particular fashion at the fatidic moment: it worked, and his friend thanked him profusely. Montaigne, who had just invented this “treatment” with the first object he could find, comments philosophically on the power of words and rituals to impress the imagination enough to cure the body. That producing mere signs of medical certainty brings relief was not news to him, nor is it today: he called it the imagination, we call it the placebo effect. Yet our understanding of how the human psyche tricks us is currently used as an argument against alternative medicines, rather than being used as a leverage point with potential therapeutic benefits.

More fundamentally, experiments and results can rarely if at all be classified according to the gold standard of “certified knowledge” that the public demands: even the most uncontestable study needs to be interpreted before it can matter intellectually and therapeutically. Medical “news” become such only when presented as potentially revolutionary or, more modestly, useful. To extrapolate from the clinical environment to real-life effectiveness always requires a leap of faith—a welcome one when it alleviates pains and save lives, but one that can never be itself grounded in absolute scientific certainty. In other words, in medicine as in any other field in the humanities, interpretation is what turns data into meaning.

Often, it is the way that most of the “fit to print” medical news is stated that is flawed, rather than the findings themselves which in most cases don't tell much in terms of real-life application. Statistical findings (valid or not) have no prognostic value. The odds of rolling a double six with two dice are always one in thirty-six, regardless of the age, gender and medical history of the person that throws the dice. Not so with the probability that you will get improvements from drug XYZ because 40% of the test group did in clinical studies. We are not inanimate objects; we are not interchangeable: each body responds differently and comes to the examination table with its own soup of genes, hormones, blood count, fluid imbalance, aged cells, vulnerabilities, temperament and medical beliefs and expectations (the latter have been shown repeatedly to influence the outcome of any treatment). Those variables cannot be eliminated: even the best conceived clinical trial can only point out to the likelihood, in certain conditions, to see such and such changes. But for each specific case, the odds can only be individually analyzed, since it is the whole medical picture and the cultural identity of the patient that will influence the results. Very few drugs are targeted drugs that address the cause of a disease or illness: most address symptoms, sometimes quiet obliquely (who has not been prescribed a drug for illness X because experience had shown that it could alleviate symptoms of Z?).

Studies that point to the medical benefits of certain nutrients or vitamins are not left of the hook. It may be true that one, or two, or a dozen studies showed that people receiving high doses of the omega-3's showed lower risks of, say, heart attacks and colon cancer then a test group. It does not mean that if you take high doses of fish oils, you will be protected against those diseases or even reduce your own risks of contracting them.

III.

I was unwittingly given to experience first-hand how little is actually known of the way treatments work (or don’t) by the medical community (the best doctors were usually those who were upfront about it). As a patient showing “atypical” but severe symptoms of carpal tunnel syndrome and unexpected responses to routine treatments, my case was alternatively misdiagnosed or dismissed (a doctor once told me that “the literature” showed I should have gotten complete remission from a shot. I had to apologize that it did not.)

I will admit that I was the first to beg for treatments, the first to be infuriated and depressed that no one could provide a fast, tested and proved cure. My belief system was that all I needed was adequate care: doctors would know what I had and what to do. End of the story. Not so.

Not only did traditional (as well as alternative) treatments fail to help, when they did not made things worst, but I was amazed at how inconsistent the criteria to select one rather than another course of action were. A long-time hand therapist once commented philosophically that one doctor “liked” to do neurolyses (the surgical freeing of a nerve from inflammatory adhesions) rather than the canonical carpal tunnel release procedure. When pressed further, she admitted that in her experience (taking care of rehabilitation for post-operative patients for close to 30 years) the outcomes tended to be worst. “So why does he do it?” I asked, nervously. “He prefers it” was her laconic answer.

I was lucky to be followed by a conservative surgeon. He loved to discuss the meta-data with me. Yet, at the end of each visit, he would consistently prescribe what I now regard in hindsight as the best possible treatment in my specific case: “Let’s wait. I’ll see you in six weeks.”

Cecile Alduy's picture
« Je ne puis tenir registre de ma vie par mes actions: fortune les met trop bas; je le tiens par mes fantasies. » Montaigne, Essais, III, 9, 945 A prescient definition of blogging, no? Cécile Alduy, Ph.D., is Associate Professor of French Studies at Stanford University. A former student from the Ecole Normale Supérieure in Paris, she teaches French literature and film, with an emphasis on gender and ethnic studies.  Her research interests include Renaissance literature and culture, the history of the body, poetry, cognitive theory, and more generally how we make sense of the world.