Uncertainty in medicine: Clich or oblivion

Bernard W K Lau 劉偉楷

Human beings are, to some degree or other, afraid of uncertainty; and if possible we always work towards seeking certainty. There seems in our culture to be an expectation that doctors always have the answers and they are presumed to be experts in knowing, but unfortunately uncertainty is a way of life for doctors. In reality, there is at least a grain of truth that medical knowledge is often engulfed and infiltrated by a certain degree of uncertainty,¹ which is defined as the inability to determine the meaning of events, in a situation where the decision-maker is unable to predict outcomes accurately.² That is, given current knowledge, more than one event is often possible. That is why a clinical condition may have several, if not many, differential diagnoses and in the same vein a patient's problem may indeed have more than one determinant cause. In medicine, one wins a diagnostic gamble by believing and acting on a correct diagnosis or by rejecting an incorrect diagnosis. Or one loses by believing and acting on an incorrect diagnosis or by rejecting a correct diagnosis. For a doctor trained in the Mechanistic Paradigm's model of causality, he will prefer to work from the heuristic principle:

"Find the cause, then treat". Without a known cause there seems to be no clear basis for treatment, and the doctor is often uneasy with whatever treatment he undertakes.³

For all concerned in health care, the passion for certainty will keep costs high. It leads the doctor to use whatever technology is available to obtain a less equivocal diagnosis, and it also leads patients to demand such diagnostic overkill. Even when doctors themselves do not see a need to perform all the prescribed procedures, they may do so anyway to protect themselves against malpractice suits in which the courts apply medicine's own standards of certainty. By those standards a doctor must do something necessary to be certain as is possible before acting.³

In most cases uncertainty is easily tolerated because the consequences of getting it wrong are relatively unimportant (for example, is the patient's sore throat due to virus or bacterium?). Uncertainty is more stressful when misdiagnosis may have dire consequences, and most doctors would go as far as possible to reduce uncertainty when they believe the probability of serious illness is high.⁴

While patients might be interested to know the diagnosis, sometimes this means very little to them and it is prognosis that is more important to them as individuals.⁵ Questions about prognosis frequently raised by patients and their families take varying forms:

When will I get better? Will I be completely well? What are the chances of the disease recurring? No doubt, one of the most difficult tasks in medicine is predicting how long someone with a terminal illness may live. No accurate method is usually available, largely because of the multiple variables that influence when a patient will die.⁶ When considering prognosis, the doctor must make a complete diagnosis relating the pathology present to the individual concerned. Even when an accurate diagnosis can be made, it can be notoriously difficult to predict the course of events in a given instance of a disease. In general practice the diagnosis is often in doubt, at least in the early or undifferentiated stages, and this further complicates the task.^5,8

In so far as prognosis implies a forecast of the probable course and result of an illness, particularly with regard to the prospect of recovery,⁷ such a prediction always contains an element of irreducible uncertainty. Uncertainty in prediction simply means that, given current knowledge, there are multiple possible future states of nature. Moreover, any prediction involving people tends to be misleading, precisely because the numbers needed to make accurate predictions are so huge. The result is that predictions become meaningless - because they do not apply to any individual person.⁹ This is where the beauty of evidence-based medicine comes in.

Essentially evidence-based medicine is a process of having clinical practice based on validated information.¹⁰ This approach suggests that a doctor should use the best available evidence when making a decision to use a diagnostic test or choose a treatment.¹¹ However, it must be acknowledged that evidence-based practice is no panacea.⁴ Even its proponents make no pretence that it can provide all the answers. The main criticism against it is that it may foster the rigidity of "cookbook medicine" based on the biased opinions of western scientific medicine and statistics, at the expense of ignoring the art of medicine and individual patient and physician and patient variation in experience and judgement.¹² The contention is that there has always been a place in medicine for clinical judgement that is based on personal experience and intuition and that takes into account intangible factors such as physician's knowledge of a particular patient or family. Traditionally this kind of skill has been thought of as "the art of medicine".³

By all means good medicine would allow probability estimates to be derived not only from tables of figures, but also from the subjective judgement of the doctor, patient, and family. It means that a doctor's knowledge of a particular patient's history may be more useful than statistical tables in assessing the likelihood that the patient has a particular disease or will respond to a particular treatment. And it means that there may be times when the patient or the patient's family are better judges of these probabilities than the doctor, especially a doctor who does not have prior knowledge of the patient.3

There is a saying in medicine that every patient is unique and should therefore deserve to be treated judiciously as a different individual from others. In any event, what appropriate evidence-based medicine should be in any locality in the world is again a matter of judgement.

It is possible that shared decision-making,^13,14 a particular type of doctor-patient interaction whereby the patient brings his/her individual preferences and the doctor contributes the medical expertise, is a partial solution to the current state of medical uncertainty.¹⁵ Moreover, shared decision-making has been shown to lead to higher rates of satisfaction and better treatment results. This might be an area worthy of more research.

Hence, is the problem one of uncertain diagnosis or uncertain management? At the end of the day, it may well be the uncertain doctor, as ultimately it is the doctor, who plays the key role in managing uncertainty.⁸ It would seem that as long as medical uncertainty is part of reality in life, as veracious as the Heisenberg Uncertainty Principle, we may have to put up with it, after all. It then begs the question: is it good enough to be sure beyond reasonable doubt, not beyond all doubt?

Editors' Note

This editorial was written by Bernard shortly before his sudden and unexpected death on June 13^th.
Bernard was a regular contributor to the College journal. A thoughtful, kind and considerate person who was never afraid to offer his opinion nor to challenge the accepted paradigm.
We struggled with the decision to publish this piece, but in the end decided that Bernards' own words should speak for him. He will be sadly missed.