In Praise of Uncertainty

Harriet Hall

There is an old joke about a patient who had high praise for her new doctor. She had headaches for years, and no one was ever able to explain why. The cause remained uncertain. But the uncertainty ended when her wonderful new doctor promptly diagnosed her with cephalalgia. The joke is that cephalalgia is not a diagnosis; it’s just the medical synonym for headache. She already knew she had headaches—but putting a medical label on it validated it for her.

We humans crave certainty, and that craving is a handicap imposed on us by the evolution of our brains. But uncertainty is far more realistic, and we would do well to train ourselves to live with uncertainty and tolerate ambiguity. This is especially true in medicine, where it’s better to have no answer than to have the wrong answer. The wrong answer could kill you.

Certainty is overrated. Research on the Dunning-Kruger effect shows that the less people understand something, the more certain they are.1 Robert A. Burton, a neurologist, wrote the excellent book On Being Certain: Believing You Are Right Even When You’re Not. He showed that your level of certainty that you are right has nothing to do with how right you are.

When I was practicing medicine, I sometimes had to resort to the diagnosis of IBTHOOM, which stands for It Beats The Hell Out Of Me! It described patients whose symptoms remained unexplained after an appropriately extensive workup. They wanted an answer and thought if we just did more tests, the answer would become evident. I had to explain to them that we had done all the indicated tests and that further indiscriminate testing would only muddy the waters. How would we even know what tests to choose? Should we x-ray their toes? Do blood tests for polio or chickenpox? Do an exploratory operation to cut their belly open and look around inside? Do a brain biopsy? None of the tests I could think of were indicated or likely to find a legitimate answer. When the probability of a disease is very low, false positive results are far more likely than true positives. If we did enough tests, we would surely find abnormal results on one or more of them—results that were meaningless and would only lead to further fruitless diagnostic efforts, including possibly dangerous invasive procedures. I pointed out the good news: we had already done all the appropriate tests, and while we didn’t know what they had, we knew what they didn’t have. No life-threatening problems had been identified; we had confirmed that they were not in any danger.

At this point, one of three things could happen: the symptoms could improve, get worse, or stay the same. If they improved or went away, it really wouldn’t matter what had caused them. If they got worse, or if new symptoms developed, we could always reevaluate and do more tests. Serious medical conditions are likely to progress and reveal themselves over time, so progression or new symptoms might give us some valuable clues about what to test. If the symptoms stayed the same, we would be reassured that nothing deadly was threatening and could do what we were already doing—trying to help the patient cope with the symptoms and maintain quality of life. My patients seemed to understand, accept, and appreciate my explanations. Of course, I retired before the heyday of alternative medicine; today they might reject my advice, seek care elsewhere, and end up in the clutches of a quack who could offer certainty divorced from reality.

Craving certainty is one of the drivers behind the use of alternative medicine. The mainstream medical doctor tells you that what you have is the common garden-variety low back pain that almost everyone gets at some time during their life, that we don’t know what causes it, and that it will subside on its own (50 percent of cases resolve in two weeks, 80 percent in six weeks). All he has to offer is back care advice, pills, and other treatments to keep you a bit more comfortable while you wait for it to run its course. Alternative medicine providers tell you they know exactly what caused it, how to fix it, and how to prevent it from recurring. That certainty is very satisfying to patients. But it’s a false certainty. The acupuncturist says it’s due to a disturbance of qi that he can fix with needles; the chiropractor says it’s due to a subluxation that he can fix with manual adjustments; the naturopath offers dietary advice, supplements, stress reduction advice, etc. There is no agreement or consensus on any of it. They don’t even critique each other. They don’t offer objective truth, only a variety of belief systems.

The craving for certainty persuades patients to accept bogus diagnoses, such as chronic Lyme disease, adrenal fatigue, and systemic yeast infections. These pseudo-diagnoses provide a convenient explanation for any number of vague symptoms. After doctor-shopping for years, patients are delighted to find a health care provider who is certain they have “the answer”—never mind that it’s a wrong answer. The provider spends a lot of time with them, has clear treatment plans backed by a bogus but sciencey-sounding rationale, and offers certainty and hope for a cure. They may submit to his quackery for years until the problem spontaneously resolves or until they finally realize they are not getting better under this care. Unfortunately, they may then seek out another quack who is equally certain about another bogus diagnosis.

Science doesn’t claim certainty but instead probability. When a scientist says vaccines don’t cause autism, the statement is only a convenient shorthand for a more nuanced reality that other scientists understand but laymen don’t. There is no way to absolutely prove that vaccines don’t cause autism, but studies have consistently shown no link between vaccines and autism—not even a correlation, much less causation. The negative evidence has accumulated to the point that it would be perverse to continue to believe vaccines cause autism, yet some perverse people, including some doctors, reject the scientific evidence and continue to trust ideology and anecdote instead of science. “My child was diagnosed with autism right after he got the MMR vaccine” creates a certainty in the parents’ minds that controlled scientific studies can’t compete with.

Nothing engenders certainty more than personal experience. “It worked for me” sounds very impressive, but it may not be true. Maybe it worked, but maybe it didn’t. There are a lot of other possible explanations. What you experienced was only this: you had a symptom, you tried a remedy, and your symptom went away. It might have gone away without any treatment, it might have gone away because of the treatment, or it might have gone away despite the treatment—which might have been somewhat harmful. Assuming it “worked” is an example of the post hoc ergo propter hoc logical fallacy, i.e., assuming that because Y followed X, X must have caused Y. If you applied that kind of thinking to the rooster who crows just before sunrise, you would conclude that the crowing was what made the sun come up! Things do get better without treatment. Before antibiotics, there were people who survived pneumonia. Spontaneous remissions occur, even in advanced cancers. Mammography studies have estimated that as many as 22 percent of breast cancers undergo spontaneous remission.2

The certainty of belief triumphs over the uncertainty and provisional nature of science. Science does not pretend to have final answers, absolute truth, or absolute proof. Its conclusions are always provisional, especially in medicine. They are always open to revision when better evidence comes along. That leads some people to distrust science, because it seems to be always changing its mind. But that’s a virtue, not a defect. As it responds to better evidence, it corresponds more and more to reality.

Certainty interferes with science. “Integrative medicine” guru Andrew Weil set up tests of osteopathic manipulation for ear infections, and when the experiments showed no effect, he said, “I’m sure there’s an effect there. We couldn’t capture it in the way we set up the experiment.” He couldn’t accept that there might not be anything to capture.

I have often thought that certainty is the root of all evil, or at least the biggest problem facing humanity. A few years ago, two women came to my door. I didn’t talk to them long enough to find out which religion they were proselytizing for, but I was intrigued by their approach. They asked if I knew what the biggest problem facing the world today was, and I answered that it was certainty, people who were absolutely sure they were right about something. They agreed with me, saying, “I know just what you mean.” They didn’t realize I was talking about them.

Unreasonable certainties have led to conflicts in medicine, politics, religion, and every other sphere of human interaction. What if we could all embrace uncertainty and ambiguity? What if religious believers were less dogmatically certain that God exists and that theirs is the one true religion? Might we see an end to religious persecution, religious conflicts, and dogmatic enforcement of dietary and clothing rules? Might acceptance of uncertainty facilitate workable compromises between opponents such as Democrats and Republicans, Jews and Arabs? Might racial and sexual prejudices and hate crimes diminish? What if everyone rejected dogmatic certainties around issues such as abortion, same-sex marriage, gun control, LGBTQ rights, evolution, climate change, and vaccination? Wouldn’t the world be a better place?

Richard Feynman said it best: “I can live with doubt and uncertainty and not knowing. I have approximate answers and possible beliefs and different degrees of certainty about different things … . It doesn’t frighten me.”

Medicine is more uncertain than most sciences. It is an applied science that often has to rely on insufficient information, probabilities, and considerable guesswork. There are degrees of certainty. In medicine, I am uncertain about which diet is healthiest, but I am reasonably certain that the following is good, evidence-based advice: exercise more, eat less, and don’t smoke.

Uncertainty is the most realistic stance in medicine as well as in politics and societal problems. Can we learn to live with uncertainty and ambiguity? I’d like to think so, but I’m not certain.




Harriet Hall

Harriet Hall, MD, a retired Air Force physician and flight surgeon, writes and educates about pseudoscientific and so-called alternative medicine. She is a contributing editor and frequent contributor to the Skeptical Inquirer and contributes to the blog Science-Based Medicine. She is author of Women Aren’t Supposed to Fly: Memoirs of a Female Flight Surgeon and coauthor of the 2012 textbook Consumer Health: A Guide to Intelligent Decisions.

There is an old joke about a patient who had high praise for her new doctor. She had headaches for years, and no one was ever able to explain why. The cause remained uncertain. But the uncertainty ended when her wonderful new doctor promptly diagnosed her with cephalalgia. The joke is that cephalalgia is not …

This article is available to subscribers only.
Subscribe now or log in to read this article.