Stats Can’t Explain Everything; The Anecdote Remains Relevant

Lately, I’ve been thinking about the difference between stories and stats — those hard and fast numbers that give us “objective” information about everything from the body politic to the human body.

Social scientists like data, perhaps because it makes social science seem more “scientific.” They like to square things off and measure them. They like to count:  How many? How much? What do the polls say?  Percentages are impressive.

Try to tell a story, however, and a purist will remind you that “the plural of data is not anecdote.”

But what some social scientists (and some physicians) forget is that statistics measure only what can be counted. Many of the things that are most important, in medicine as in life, are immeasurable. Stories are valuable because they can capture some of the messiness of reality, including the ambiguities and contradictions that make both human experience and the human mind/body just beyond comprehension. (Since we have only the mind with which to understand the mind, ultimately investigation must end in a stand-off.)

I began thinking about the difference between stories and statistics this week end, while reading Dr. Chris Johnson’s blog. A pediatrician and former head of pediatric critical care at the Mayo Clinic, Johnson confides that he sees medicine as “complicated mish-mash of science, near-science, intuition, guesswork, and blind luck.”

And that, he explains, is why he has been thinking about “the enduring power of the anecdote in how we humans understand things. These days physicians are exhorted to use only the hardest of hard evidence to make decision,” Johnson observes, and he has no objection to this — when there is sufficient data to lead to a clear diagnosis. But often there isn’t. And this is why physicians “also use anecdotes — stories we have heard or things we have seen,” Johnson writes. Those stories feed a doctor’s intuition. “We should always use the best science we can,” he concludes, “but somewhere in the mix there is a place for the anecdote, the story.”

Johnson then goes on to provide a link to what Rafael Campo, the award-winning physician and poet, had to say on the subject in a 2006 essay titled “Anecdotal Evidence: Why Narratives Matter to Medical Practice.”

Campo begins simply: “I want to tell you a story.

“After a lecture I gave recently at a well-known medical school on the possible utility of narrative to clinical practice, from the back of the auditorium came the first question of the traditional question and answer portion of the program: ‘Don’t you feel, Dr. Campo, that what you seem to regard as the arrogant biomedical science model of medicine is already sufficiently under attack these days?’

“As the lights came up, I could make out a tall, bearded man in a long white coat, standing as if at attention near the end of one of the aisles. ‘We have creationists trying to teach “intelligent design” in our children’s science classes, [he declared], and even closer to home, nurses and optometrists being given the right to prescribe medications.’ Their applause having ceased, my audience now grew hushed as he went on, his voice steadily rising.

“‘Do you really expect physicians to accept the notion that what any ignorant patient tells us about his disease should carry a weight equal to what our years of training and expertise reveals to us about complex pathophysiology?’ Then came what was clearly meant to be his coup de grace, delivered in an almost derisive tone. ‘Really, sir, do you have anything more than the anecdotal evidence you shared to support your thesis?’

“Of course, like any physician trained in the past several decades, I too had learned to view the anecdote with the greatest amount of skepticism, if not outright disdain,” Camp acknowledges. “The anecdote, though beguiling in its familiar engagement of our human sensibilities, is, we are all taught, the enemy of objective, dispassionate observation.

“The anecdote is rife with such difficulties as openness to interpretation, and the biases of faulty memory and foolish optimism; it is just as likely to be explained by fickle chance as by anything truly under the clinician’s control. It is colored by the inflections in our voices and shaped by our gestures and facial expressions. The case report counts not for academic promotion, while the randomized controlled trial of thousands of anonymous subjects has become the lingua franca of our profession, and for good reason, as rigorous epidemiologic studies have replaced mere conjecture with sound, evidence-based understanding of the causes of countless diseases and effective treatments for them. Yet to offer an anecdote these days is almost to admit the insufficiency of one’s knowledge, and so we do so, at least to our fellow physicians, very apologetically.”

Nevertheless, Campo insists: “Whether we choose to admit it or not, the anecdote continues to be an important engine of novel ideas in medicine…

“The inscrutably enduring power of the anecdote itself is what incites all our most fearsome defenses,” Campo continues. “So furious are we in our rejection of the merely anecdotal one cannot help but begin to wonder at it.”

Why are so we defensive? Because stories are potent. Rick Diamond and Mishra Moezzif, scientists at the Berkeley National Laboratory, elaborate on Campo’s point:

“Whether we formally recognize it or not anecdotes can have powerful effects in challenging assumptions, although they may  be ignored when they confront popular preconceptions of what is true,” Diamond and Moezzif observe, citing Thomas Kuhn’s classic The Structure of Scientific Revolutions. Yet “anecdotes are told because they illustrate or crystallize important concepts. Accordingly, anecdotes can provide extremely valuable information, if their significance is actually understood.”

In the same vein, Campo writes: “Whether we choose to admit it or not, the anecdote continues to be an important engine of novel ideas in medicine. No matter how wide the perceived rift between science and the humanities, and no matter what new technologies may deliver unto us in terms of more precise tests and life-prolonging therapies, the work of doctors will always necessarily take place at the intersection of science and language. [my emphasis] How many of us have first felt inspired to dig deeper into a question that first took shape in the form of ‘a couple of interesting cases’— the beginnings of a case series, in epidemiological parlance — shared by a colleague over a cup of bad doctors’ lounge coffee?”

The Patient’s History

In many instances the most important story in medicine — and the one that is most difficult to grasp — is the patient’s history. Yet rather than listening carefully, today’s rushed clinicians often rely on a battery of tests to tell them what they need to know about the patient. Meanwhile, the clue to the diagnosis remains hidden in a half-heard anecdote.

“Our patients’ stories too, if only we could listen to them less critically and cynically, might similarly inspire us to the more practically important discoveries of what truly ails them,” Campo acknowledges. “Yes, we must always be wary of the ways in which the interlocutor may lead us astray; the possibility of violation of the narrative contract, that implicit agreement between us that the story being told is truthful and offered in the service of best care, is a real one. A patient in distress may speak to us across a chasm so vast that what we can hear is terribly distorted — by our professional distance, by our own most unprofessional fears and misapprehensions, and by society’s attitudes which inescapably contextualize our every action.

“One common clinical scenario has become so familiar as to be regarded as paradigmatic of our distaste for the subjective. The patient, we frequently suspect, is exaggerating her pain to obtain more narcotics, so we check to see if she is tachycardic, or whether she perspires or writhes in her sheets, ever on the lookout for more reliable objective signs of what her suspiciously anecdotal description fails to convey. Yet even in the face of language’s shortcomings and betrayals, understanding narrative ultimately helps us. If we can recognize a breakdown in our communication with a suffering patient, we can begin the crucial process of repair—usually by explicitly re-establishing the ground rules of empathetic mutual trust upon which any exchange of language must be based.

“Perhaps,” he adds, “it is our own mistrust of the anecdotal that has engendered the backlash against science to which my interrogator at that recent lecture alluded. We seem to be of two minds when it comes to science as it relates to our ever defiantly human bodies. While we look to medicine to offer us the fruits of its inquiry into our innermost life-giving processes, at the same time we refuse to be entirely explicated. We want answers, but not all the answers. We want Tamiflu [a bird flu drug] as well as talismans to protect us from avian influenza.”

Uncertainties, Doubts, and Mysteries

There remain many medical miracles that we cannot fully explicate. For example, how do we explain the “placebo effect?” Are depressed patients who take medication really helped, or are the drugs, as some have recently suggested, simply placebos? How could we possibly know?  Our only evidence consists of the stories that the patient tells himself about himself and how he feels — the stories that, taken together, create his reality.

Campo considers the problem: “A daughter of a patient of mine wrote a poem about a flamingo,” Campo tells us, “so the birds won’t get mad and make us sick.’ At the bottom of the page blazed a hot pink stick figure of a bird, as if she had drawn fever itself. Might her fervent belief in the power of her own words somehow stimulate her immune system to fend off an unlucky exposure to a bird-borne virus?”

“In all the millions of epidemiologic studies we have published in thousands of medical journals,” Campo observes, “we have yet to prove the mechanism behind a phenomenon evident in nearly all of them: the placebo effect. Perhaps there remain ideas about ourselves and our bodies that can never be summarily studied?

“ ‘I want to tell you a story,’ another patient of mine said to me a few weeks later, back home in Boston, in the quieter theater of daily life. She was dying of multiple myeloma that afternoon. No more melphalan and prednisone, which had caused diabetes, nor more thalidomide, which had given her neuropathy; instead, she received only morphine now, because all that was left to treat was her pain. Rain fell relentlessly outside, streaking the windows in a way that made me think inanimate objects might somehow feel sadness.

“One of her daughters clutched my hand. I looked into her mother’s watery, deep brown eyes, which at that moment seemed a well of stories so absorbing and so numerous that they might unspool forever. ‘I want to tell you a story,’ she said again. Perhaps she was going to God, a notion that consoled us all; perhaps nothing was left of her but the fading impulse generated by the brain’s physiology, whose final expression would be these last words. But before she could go on, her breathing stopped — leaving it all at once plainly obvious, and yet utterly incomprehensible.”

Indeed, a test might provide the data needed to confirm the obvious fact that she had died, but even the most precise tests couldn’t “comprehend,” in the sense of fully grasping the mystery of her death. Only Campo’s story could begin to do that.

Maggie Mahar is an award winning journalist and author. A frequent contributor to THCB, her work has appeared in the New York Times, Barron’s and Institutional Investor. She is the author of  “Money-Driven Medicine: The Real Reason Why Healthcare Costs So Much,” an examination of the economic forces driving the health care system. A fellow at the Century Foundation, Maggie is also the author the increasingly influential HealthBeat blog, one of our favorite health care reads, where this piece first appeared.

Categories: Uncategorized

Tagged as:

3 replies »

  1. There is real science and then there is medical science. The two are so different that medical science depends heavily on posturing and manipulating to make what they do look like real science.
    Contrast what a real scientist does in an experiment compared to a medical scientist. A real scientist tries to create a closed system, eliminate as many variables as possible, strives to modify only one variable, and measure all the rest. Even then, they are cautious about explaining their observations. Was the system completely closed? Was only one variable modified? Are there other variables influencing results that are not being measured.
    Medical science is nothing like this. A human is an open system, it’s impossible to modify only one variable, the measures are egregiously imprecise… on and on. So, medical scientists are left pretending what they do is science.
    Thy use “anecdote” is used as a pejorative, when it could as easily be termed an “observation”. With so many mysteries around health and medical care, doctors should use every scrap of input they can come across, instead of trying to train themselves to ignore everything that is not a randomized, double-blind clinical trial so they can pretend they are scientists.
    The placebo effect is extremely limited in what it can do to improve a subject’s health. The amount of time spent blinding studies and trying to create realistic placebos, which subjects generally figure out right away anyway, is incredibly wasteful. Doctors should stop pretending to be scientists and be doctors.

  2. An anecdote is a story of something happening to someone in a certain way at a certain time. Another word for it in medicine is the case history. It is said that anecdotal evidence is the weakest level of evidence. It is said that there are four stronger levels of evidence. The highest being the randomized clinical trial (remember Vioxx?).
    However, we are all unique. What works in one person may not work in another. What works in a subset of patients with a certain disease may not work in another. The idea that approving drugs based on “population” studies has its limits. What may or may not work for the average “population” may not apply to the “individual.”
    In cancer medicine, the selection of chemothrapy for cancer patients has been traditionally based on results from phase III comparative trials (population studies) that define the most active drugs and drug combinations. Unfortunately, few patients with advanced disease are helped using currently available regimens.
    The new paradigm of requiring a companion diagnostic as a condition for approval of new “targeted” therapies, tries to improve the selection process for individual patients (not average populations), various types of in vitro tests that assess the activity of standard and targeted drugs on a patient’s tumor have been developed.
    Significant predictive correlations between in vitro drug response assays, and cancer patient response and survival have been demonstrated. The tumors of different patients have different responses to chemotherapy. It requires individualized treatment based on testing the individual properties of each patients’ cancer, not populations.

  3. My parents are physicians, and both believe that this is very true. Medicine is a concerted effort on the part of the patient and the patient’s friends, families, doctor, nurse, and all those related to the patient’s well-being. Treating the individual requires remembering the community.