The importance of uncertainty

Iona Heath

HK Pract 2017;39:23-28

Introduction

The meanings of uncertainty include ‘the state of not being known or perfectly clear’ and a ‘liability to chance or accident’. Anyone who works in clinical practise will recognise these states. There is almost no clinical situation in which everything is known and perfectly clear, and which is not liable to chance or accident, and I suspect that the same is true for health policy.

Speaking in the British House of Commons in 1803, the great Irish-British playwright and parliamentarian Richard Brinsley Sheridan said:

The glorious uncertainty of the law was a thing well known and complained of by all ignorant people, but all learned gentlemen considered it as its greatest excellency.

Just like Sheridan’s view of the law, my view of the uncertainty within medicine is that it too is glorious, despite being much complained of.

The inevitability of uncertainty within clinical practise

In medical school, students are taught the mantra of history, examination, diagnosis and treatment. Each is supposed to follow the other in a rational and linear sequence with a single diagnosis suggesting appropriate treatment. Yet in practise, it is almost never like this. Consider a single patient with hypertension, but she also has bronchiectasis after her many years of smoking. She has a long history of mental illness following on from a violent and abusive childhood. She has just been diagnosed with cancer of the oesophagus. She has a child with severe learning difficulties and she is fearful, not only for herself but also about what will happen to her child. She is still married and but the relationship is difficult. Her sister has lung cancer and is already very ill. The sister’s children have problems and children of their own and this is only the beginning of a story which became richer and richer the longer I knew this patient and her family. It is a story with multiple components each of which interacts with the others unpredictably. Each of the components has a history which affects the interaction and each has the capacity to affect my patient’s blood pressure and to support or undermine the treatment that I am prescribing for her. Nothing is certain or clear or fully known and all of it feels susceptible to chance and accident.

In stark contrast, most biomedicine has been developed on the principles of Newtonian mechanics using a metaphor of the body as a machine and every effect having an identifiable cause. Yet anyone who has worked with the uncertainties of clinical practise is painfully aware of the limitations of this model. We need to examine our uncertainties carefully so that we can look at old and intractable situations in a new way and find new ways of resisting the simplistic and deterministic reductionism which informs much biomedical science and most of the organisation and evaluation of healthcare.¹

The sixteenth century French humanist scholar Michel de Montaigne had the ceiling of his library inscribed with quotations including one in Latin from the Roman scholar Pliny the Elder. The English translation reads:

Only one thing is certain: that nothing is certain. And nothing is more wretched or arrogant than man.

Nothing is certain. In science, we have tried, for centuries, to reduce our uncertainties by producing models and maps to try and capture our experience and understanding of ourselves and of the world around us.

The Polish-American philosopher and scientist, Alfred Korzybski, is famous for his description of the gap between the map and the territory. In 1933 he wrote “A map is not the territory it represents”. The experience of looking at a map, however detailed, is nothing like the experience of walking through a landscape. Similarly the map of medical science is nothing like the landscape of human suffering. The map provides a guide but it does not even begin to capture the reality of experience. Uncertainty pervades this gap between the territory of human suffering and the map of biomedical science. The task of making the medical map useful to those trapped within the territory of suffering is, and will always be, fraught with uncertainty because of the vast extent and infinite variation of the territory and because of the comparatively rudimentary nature of the map. But the uncertainty and doubt that clinicians experience every day are also what make new knowledge and understanding possible. We have to doubt existing explanations if we are ever to discover better ones. And because of the gap we need always to be worried by simple explanations of situations that we know to be complex.

As British philosopher Mary Midgley puts it:

… For any particular problem, we need a solution that sorts out the particular complications that puzzle us, not one that ignores them because they are untidy.²

Many of the algorithms and guidelines used in medicine today do just this: they ignore the particular complications that affect our individual patients. And they ignore a truth that was well known to the ancient Greeks:

It is a part of probability that many improbable things will happen.³

A key component of the medical model or map is the concept of diagnosis: the labelling system of medicine. The Harvard medical historian Charles Rosenberg pinpoints the underlying problem:

This modern history of diagnosis is inextricably related to disease specificity, to the notion that diseases can and should be thought of as entities existing outside the unique manifestations of illness in particular men and women.⁴

Experience is fluid and continuous, while diagnoses are discrete and dichotomise the normal from the abnormal in a way which has proved useful, but which is totally artificial. Notions of disease have become detached from the human experience of illness and, today, they are more and more defined on the basis of deviant numbers. Yet these numbers, always and forever, are surrounded by the uncertain and the unmeasurable, those great givens of human existence and of medicine.

The relevance of uncertainty to decision-making in medicine

In his book A Fortunate Man, an account of the working life of a country doctor in England in the 1960s, John Berger writes –

Good general diagnosticians are rare, not because most doctors lack medical knowledge, but because most are incapable of taking in all the possibly relevant facts – emotional, historical, environmental as well as physical.⁵

Doctors miss these relevant facts because they are all too often blinded by the imperatives of public health utilitarianism on the macro scale and those of biomedical science on the micro scale: both of which obscure the view of the unique personality in front of them. Population-based public health objectives, with centralised control and a strong emphasis on costeffectiveness, damage and detract from the individual focus of patient-centred care. Patients’ needs extend far beyond the biomedical and are easily marginalised if the agenda of the consultation is dictated by forces outside it.

It is the essentially unknowable individual human being who suffers and it is the need to respond to human suffering that is the origin of medicine. Medicine is unique in aspiring to be a science whose object is also a subject. Within the consultation both doctor and patient oscillate between perceiving the body as an object and as a subject. When the body is perceived as an object, the gaze of biomedical science sees only what the particular patient has in common with other patients. When the body is perceived as a subject, we see what is unique about this person – their life context, its story and the meanings that are found in both. The perspective involves two unique subjects: the patient and the doctor.

Sadly medicine and medical research have a strong bias towards the body as object and tend to marginalise subjective experience. Yet, the more we discover and understand about the extent to which our life stories affect the workings of our bodies, the more we discover about the ways in which violence, abuse and the stresses of poverty affect physiology, the more we understand the biology of biography, the more important our understanding of the detail of the lives of our patients becomes.

Adverse experiences early in childhood include physical, emotional, or sexual abuse; witnessing domestic violence; and growing up with household substance abuse, mental illness, parental divorce, or a household member in prison. These experiences have been shown to predict future premature mortality to the extent that people with six or more adverse childhood experiences died nearly 20 years earlier on average than those without any.⁶ In the context of a damaging childhood, the developing brain is exposed to repeated stress responses which lead to impairment in multiple brain structures and functions.

The Adverse Childhood Experiences (ACE) Study pyramid summarises the current state of understanding.⁷ Adverse child experiences lead to disrupted neurodevelopment, then to social, emotional and cognitive impairment, which in turn lead to the early adoption of risky health behaviours. These play out in disease, disability and social problems, and culminate in early death.

Adverse child experiences markedly increase the risk of smoking, alcohol and substance misuse, obesity, depression, self-harming and suicide, and multiple sexual partners. As Norwegian Professor of General Practise Anna Luise Kirkengen puts it:

Children who, more than anything, have learned that they are worthless and not appreciated turn into young people who know neither their right to be protected nor how to protect themselves.⁸

The effects are real: biography affects biology in ways we only just begin to understand:

Slow growth to the age of 7 years has been associated with an increased risk of unemployment in young men regardless of their adult stature, thus producing a complex pattern of continuity between biological and psychosocial disadvantage. ⁹

This makes all the more important our –

- growing understanding of the biological pathways that link stress, or chronic anxiety, to the observed physiological changes that put health at risk.¹⁰

Yet, because of the current emphasis on individual risk factors, we have ended up attempting to treat the health effects of socioeconomic inequality with pharmaceuticals.

If we, as doctors, are to help our patients make useful decisions in the context of the uncertain and the unmeasurable, we need foster certain great qualities. We will need courage to get close enough to the patient to see and to hear; courage not to rely completely on the attenuated truth of scientific medicine; courage to trust the patient’s account; courage, in the world of protocols and directives, to help patients to make their own decisions which accommodate their own values. We also need curiosity, wonder and imagination and doubt. The existence of the gap between the map of medicine and the territory of suffering implies the necessity of doubt. Doubt is defined as a subjective sense of uncertainty with regard to the truth or reality of anything and it is the source of both wisdom and freedom

Without these great qualities of courage, curiosity, imagination and doubt, we will continue to try to solve the big problems of human existence – suffering, loss, ageing and death – by using futile biotechnical means.

One of the most fundamental problems that we are all facing was captured by the Norwegian medical ethicist Bjørn Morten Hofmann:

Our abilities to produce and use technologies appear to outrun our abilities to reflect on their application.¹¹

We risk becoming technological giants but ethical midgets – technologically brilliant but insufficiently thoughtful, with a dangerous tendency to do before we really think. Every day, in advanced medical settings, we demonstrate the truth of the saying that to a man with a hammer everything looks like a nail and to a specialist with an Magnetic resonance imaging (MRI) or Computed tomography (CT) scanner every symptom needs a scan just to be sure and safe.

There is enormous variation in the number of MRI and CT scanners per million population in countries across Europe. 12 And in Asia Pacific there is even greater variation with some countries clearly having too few but with Korea, Australia and Japan having more CT scanners than even the most equipped European countries. Japan’s figure is 4 times the European average. 13 Yet each CT scan results in excessive exposure to radiation.

As contemporary Dutch philosopher Annemarie Mol writes:

Technologies always have unexpected effects: they generate forms of pain and pleasure that nobody predicted. … Do not just pay attention to what technologies are supposed to do, but also to what they happen to do, even if this is unexpected.¹⁴

With the rise of concern about overdiagnosis across the globe, we seem to be finally beginning to pay the necessary attention to the unexpected effects of new diagnostic technologies. It is increasingly clear that diseases diagnosed through the use of our ever more sophisticated technology are very different from those identified as a result of the combination of a patient’s account of his or her symptoms, the clinician’s clinical skill and the use of basic and relatively crude technologies. As Hoffman and Cooper write:

- we must recognize the enormous difference between a disease that presents clinically and “the same” disease that is found only because we have decided to search for it, in the absence of compelling clinical concern.¹⁵

And they point to the:

- multiple examples of overdiagnosis that arise when technology, rather than clinical findings, are the catalyst for finding disease.

The use of ultrasound to screen for thyroid cancer in South Korea provides just one example. This screening led to a rapidly rising rate of diagnosis associated with a static mortality rate: a pattern which is now recognised as characteristic of overdiagnosis. More people are diagnosed but the same number die. And here is the underlying uncertainty: it has become clear that many cellular malfunctions can correct themselves and overdiagnosis occurs whenever a condition is diagnosed that would not have caused the patient harm if left undiagnosed. Yet people are being labelled and needless fears are created and exacerbated. And meanwhile we understand more and more about how fear, anxiety and stress can, in themselves, undermine health. The harms are multiplying. As a direct result of this rapidly rising rate of diagnosis, thousands of people have undergone thyroidectomy despite this being an operation with significant potential complications and patients also have to take replacement thyroid hormone for the rest of their lives.

There is no screening programme in the United States but there is still a rising incidence, although it is much less striking, and this again seems to be due to sophisticated technology with small tumours being found as what are called ‘incidentalomas’ on ultrasounds or CT and MRI scans done for other reasons. Breast cancer screening and lung cancer screening seem to be following the same pattern.

It is time we really acknowledged the extent of scientific uncertainty and doubted more of the accepted explanations of causation and progression of disease states.

Physicians need to insist that the body is NOT a machine and that we cannot solve the problems of healthcare with algorithms. When human beings are sick, there are very few easy yes/no answers despite the hopes, aspirations and imperatives of politicians and policy-makers, and, underlying these, those of the medical industrial complex.

The laws of mechanics and machines are deterministic because they imply that anything that happens at any future time is completely determined by what happens now and moreover that everything now was completely determined by what happened in the past so that, for any given system, the same initial conditions will always produce the same outcome. In medicine, we can never extrapolate either backwards or forwards in time with any degree of certainty. Two individuals with the same diagnostic label and given the same recommended treatment can have quite different outcomes.

There are limits to what can be changed - but these limits are not obvious at the beginning. It is difficult to predict what may work and what will fail. … Try, be attentive to what happens, adapt this, that or the other, and try again.¹⁶

And this sort of caring attentiveness is precisely what we need to negotiate all the uncertainties, particularly towards the end of life

The particular relevance of uncertainty to the care of those towards the end of life

In her 1876 novel Daniel Deronda, the English novelist George Eliot asked:

- how to make sure that snatching from death was rescue?¹⁷

In a 2010 article in the New York Times, writer Katy Butler describes the ruination of her elderly parents’ lives by the insertion of an ill-considered pacemaker.

I watched them lose control of their lives to a set of perverse financial incentives — for cardiologists, hospitals and especially the manufacturers of advanced medical devices — skewed to promote maximum treatment. At a point hard to precisely define, they stopped being beneficiaries of the war on sudden death and became its victims.¹⁸

At the age of 79, her father was suddenly severely disabled by a stroke which robbed him of most of his language, his mobility and his ability to care for himself. A year later he suffered a strangulated inguinal hernia and the hospital refused to operate on him unless he agreed to have a pacemaker inserted because he had longstanding bradycardia and “might die during or shortly after the operation”. He had previously declined such a pacemaker when he was competent to do so and no-one told his family about the option of temporary pacing. The pacemaker kept him alive for six more terrible years of worsening dementia – exemplifying American bioethicist Dan Callahan’s description of the Difficult Child of Medical Progress:

- the 1 percent of patients who consume some 21 percent of health care costs, usually succumbing gradually from multi-organ failure, illustrate the progress problem. Fifty years ago they would have died faster and, in many cases, with less suffering. We have traded off shorter lives and faster deaths for just the opposite, longer lives and slower death.¹⁹

In her subsequent book, Katy Butler made this very perceptive comment:

A sense of urgency, combined with the assumption that the treatment offered has no alternatives and no downsides, are common ingredients in medical decisions, later regretted, involving the frail elderly.²⁰

This is the harm caused when we donot acknowledge the degree of uncertainty that surrounds all treatments towards the end of life and it is in these situations that, most of all, we need to take decisions slowly – really thinking them through in conversation with the patient and the family

A study published in 2008 attempted to measure statin use during the last 6 months of life and to determine if statin prescribing varies according to the presence of a recognisable, life-limiting condition. The researchers studied a group of patients who had died in 2004 and who were taking statins within 6 months of death. They compared those who were known to have a life-limiting condition with controls matched on number of comorbidities, age, and socioeconomic status. They found that there was no significant difference in the time off statins between cases and controls even though the study was sufficiently powered to detect one and concluded that there had been a missed opportunity to reduce the therapeutic burden upon dying patients and to limit health care spending.²¹

Another paper published in July 2015 studied the use of chemotherapy at the end of life. The conclusion was that “Although palliative chemotherapy is used to improve Quality Of Life for patients with end-stage cancer, its use did not improve QO life near Death for patients with moderate or poor performance status and worsened QOD for patients with good performance status. The QOD in patients with end-stage cancer is not improved, and can be harmed, by chemotherapy use near death, even in patients with good performance status.”²²

These sorts of harms are predicated on a false certainty that has become all too prevalent within contemporary medicine. Yet, the great comfort is that none of us knows exactly what will happen to us tomorrow. We know a lot about probability but probability is a long way from certainty. People do not always get the result predicted by their lifestyle. Not everyone who smokes or is obese dies prematurely. Conversely, a good diet and regular exercise does not provide complete protection from random disaster. Nonetheless, when death or disease occurs prematurely and unpredictably, the rhetoric of preventive medicine suggests that someone somewhere must somehow be at fault. The situation is immensely more complex. We are only just beginning to understand the power of the human mind to influence the functioning of the body - both happiness and despair affect prognosis.

Uncertainty becomes the basis of intellectual freedom and political resistance.

Zygmunt Bauman, the emeritus Professor of Sociology at the University of Leeds in the United Kingdom, clearly demonstrates the false certainty of some medical predictions by still being alive at the age of 90 despite his love of tobacco pipe-smoking. Bauman writes:

We understand now that uncertainty is not a temporary nuisance, which can be chased away through learning the rules, or surrendering to expert advice, or just doing what others do. Instead it is a permanent condition of life.

To be responsible does not mean to follow the rules; it may often require us to disregard the rules or to act in a way the rules do not warrant.²³

Conclusion

This breaking of the rules involves asserting that change does not equate with progress and that the present is more important than the future. My hope is that we can learn to nurture and treasure the glorious uncertainty of medicine and not be afraid of it.

1. Goodwin JS. Chaos, and the limits of modern medicine. JAMA 1997;278(17):1400.
2. Midgley M. Science and Poetry. Abingdon: Routledge, 2001.
3. Agathon 448-400 BCE. Quoted by Aristotle in Poetics, 335BCE.
4. Rosenberg CE. The Tyranny of Diagnosis: specific entities and individual experience. The Milbank Quarterly 2002;80(2):237-260.
5. Berger J, Mohr J. A Fortunate Man. Harmondsworth: Allen Lane The Penguin Press, 1967.
6. Brown DW, Anda RF, Tiemeier H, Felitti VJ, Edwards VJ, Croft JB, et al. Adverse childhood experiences and the risk of premature mortality. Am J Prev Med 2009;37:389-396.
7. https://www.cdc.gov/violenceprevention/acestudy/
8. Kirkengen AL. The Lived Experience of Violation: how abused children become unhealthy adults. Bucharest: Zeta Books, 2010.
9. Bartley M, Blane D, Montgomery S. Health and the life course: why safety nets matter. BMJ 1997;314:1194-1196.
10. Wilkinson R, Pickett K. The Spirit Level: Why Equality is Better for Everyone. London: Allen Lane, 2009.
11. Hofmann BM. Too much technology. BMJ 2015;350:h705.
12. OECD. Health at a Glance: Europe 2014.
13. OECD. Health at a Glance: Asia Pacific 2014.
14. Mol A. The Logic of Care. Health and the problem of patient choice. Abingdon: Routledge, 2006.
15. Hoffman JR, Cooper RJ. Overdiagnosis of Disease: A Modern Epidemic. Arch Intern Med. 2012;172(15):1123-1124. doi:10.1001/ archinternmed.2012.3319.
16. Mol A. The Logic of Care. Health and the problem of patient choice. Abingdon: Routledge, 2006.
17. Eliot G. Daniel Deronda, 1876.
18. Butler K. What Broke My Father’s Heart. The New York Times, June 18, 2010.
19. Callahan D. The Difficult Child of Medical Progress, Bioethics Forum, 2012.
20. Butler K. Knocking on Heaven’s Door. New York: Scribner, 2013.
21. Silveira MJ, Kazanis AS, Shevrin MP. Statins in the last six months of life: a recognizable, life-limiting condition does not decrease their use. J Palliat Med 2008;11:685-693.
22. Prigerson HG et al. Chemotherapy use, performance status, and quality of life at the end of life. JAMA Oncol. 2015 Sep;1(6):778-84. doi: 10.1001/ jamaoncol.2015.2378.
23. Bauman Z. Alone Again: Ethics After Uncertainty. London: Demos, 1994.