Living With Probabilities
Scientific uncertainty leads some to ignore the probable
By Bob Bernstein, MD
We are at an interesting point in time in the modern world. The scientific method has brought us computers and the internet, and has eradicated diseases like polio and smallpox. Yet there is a growing distrust of science, ranging from climate change deniers to the anti-vaccine movement. Science is thought of as just another belief system, on an equal footing with religion. And so we let children suffer from preventable illness, and the earth's environment degrade to the detriment of the health and survivability of all of us. It seems as if the Age of Enlightenment has come and gone and we are again entering another period like the Dark Ages.
How can this be? We have come so far, and yet it seems that we are walking backwards.
There is an old tradition in medical education, thankfully not practiced much anymore, of "one-upmanship." A hapless student groping for answers to a gotcha question from a distinguished professor would be quoted the latest article from the New England Journal of Medicine. I vowed I'd never be that kind of teacher. But I'm about to break that vow. In the latest issue of the New England Journal there is an article entitled "Communicating Uncertainty - Ebola, Public Health, and the Scientific Process" (New England Journal of Medicine, vol. 372; 7-9, 2015).
The nub of the argument is that science is inherently uncertain. We never know "absolute truths." We build a model of the universe as it fits the known facts and observations, and we call it a theory—a semantic artifice which is a source of more confusion—and we can say that it is true until someone makes a new observation that does not fit with the theory. Sometimes the theory is modified (e.g.: Evolution, which remains substantially unchanged by new observations), and other times replaced (e.g.: Newton's theory of gravity—by Einstein's theory of relativity).
The same holds true in medicine. Answers to scientific questions beget more questions. We never truly know anything as a certainty. Guidelines, the bastion of "evidence-based medicine" change over time, sometimes radically. Remember hormone replacement? In the 70s and 80s, family doctors were criticized for not prescribing hormone replacement to every menopausal woman. Now hormone replacement is a dirty word, so much so that women who would benefit from it shy away because of perceived risks. How can we trust the medical profession if it is collectively always changing its mind?
Back to the New England Journal article. The author describes a weather prediction that said with scientific certainty that the water in the Red River in North Dakota would crest at 49 feet. When the river rose to 54 feet the resultant flood was devastating. The truth is that the river height prediction had a margin of error; it was not an absolute prediction, but a probability. But the forecasters were afraid that communicating uncertainty about it would betray public confidence! Maybe, but the actual flood betrayed public confidence even more.
Same thing happens in medicine. Remember swine flu when Gerald Ford was the American president? An ill-conceived vaccination campaign against swine flu caused some 500 cases of a rare neurological condition called Guillain-Barre syndrome, leading to about 25 deaths. The publicity put people off from getting influenza shots, yet the numbers of dead from flu because people were not immunized far outweighs the small numbers of G-B syndrome cases. Yet many of my patients even today refuse flu shots because of fear of side effects.
Hormone replacement has some probability of benefit and some probability of harm. The decision to be made is whether, for a specific person (you, for example), the benefit outweighs the harm. This kind of risk-benefit analysis goes on for all medical decisions. Sometimes it is blazingly obvious and not much thought about. Other times the risk-reward calculation is very difficult and statistics about what happens to a group of people do not necessarily apply to an individual. Many cancer treatments are like this. What is the best way to treat a low-grade prostate cancer? Is it safe to watch and wait, is it harmful to overtreat? What is likely to happen to me if I take a wait-and-see approach? What will happen to me if I have my prostate removed? No one can tell for sure; there is only probability. I tell my patients that it is like managing a baseball team. If the opposing team sends in a left-handed batter to face your right-handed pitcher, a good statistical move for you to make is to pull your pitcher in favour of a leftie. On average, it will work out, but no-one can predict what will happen this time.
So we are left with uncertainty as a matter of course in medicine. This uncertainty is at the root of why many people do not trust the medical establishment. We need to do a better job of communicating uncertainty to our patients so that they understand how to make decisions for themselves. We need to stop proclaiming that we know the answers when in fact we only know probabilities. Worse, we need to stop saying to the patient, in effect, "It's up to you, your choice," without any guidance from us. With an honest explanation of uncertainty, and navigating with our professional judgement, knowledge, and experience, we docs can help guide patients to make the best informed choices about their treatments in an atmosphere of trust and respect.
Robert Bernstein, MD, is a family practice physician and is part of the Bridgepoint Family Health Team.