It’s tough to make predictions*

by Dominic Wilkinson, @Neonatalethics

The Court of Protection is due to review very soon the case of a teenager with a relapsed brain tumour. The young man had been diagnosed with the tumour as a baby, but it has apparently come back and spread so that according to his neurosurgeon he has been “going in and out of a coma”. In February, the court heard from medical specialists that he was expected to die within two weeks, and authorized doctors to withhold chemotherapy, neurosurgery and other invasive treatments, against the wishes of the boy’s parents.

However, three months after that ruling, the teenager is still alive, and so the court has been asked to review its decision. What should we make of this case? Were doctors and the court wrong?

The first point to note is that prognosis (medical prediction of outcome) is almost always riven by uncertainty. That is because of the nature of what we are attempting to do – predict the future. Clinicians don’t have a crystal ball. All we can know with absolute certainty is what has happened in the past. Doctors look back on past cases, and published scientific studies. Based on that experience it is possible to estimate what will happen in the future to a group of patients with similar features. However, it is much harder to say what will happen for the single patient in front of us.

Next, the fact that in this case, the patient has done better than predicted does not mean that the initial prediction was wrong. Imagine that a doctor estimates that there is a 99% chance that a patient will die within a short period of time. The patient returns some time later – still very much alive, and says to the doctor – ‘you were wrong’. But that is not necessarily correct. Based on the doctor’s prediction, we would expect one out of 100 patients to survive for longer. Even if there is only a one in a million chance of prolonged survival, there will still be very occasional patients who live longer. Those patients have not ‘defied predictions’ – they are exactly what was predicted.

Media reports often report such exceptional cases. Almost every week there are stories in the newspapers of miracle survivors. This can give the impression that doctors are wrong in their predictions frequently. But that does not follow either. Cases where the patient died as predicted are not newsworthy. Stories of patients dying within a short space of time don’t get reported.

There is more robust evidence about how well doctors are able to predict. A systematic review, published in the British Medical Journal in 2003, reviewed 17 studies (including a total of more than 1500 patients) that had assessed how accurately doctors had predicted survival duration in terminally ill patients with  cancer. That study found that predicted survival was strongly related to actual survival. Patients who were predicted to live for a longer time, survived longer than those predicted to die quickly. However, and this is important, doctors mostly overestimated survival. Doctors were correct to within a week in only 1/4 of cases. They overestimated survival by at least four weeks in another 1/4 of cases. The recent report by the parliamentary and health services ombudsman highlighted that many patients do not receive the end of life care that they need – in part because of this problem with overestimating survival

One interesting possibility about this specific case is that the patient lived for longer than expected because he did not receive the chemotherapy that had been requested by his parents. We could call this the “palliative paradox”. People think that making a decision to have palliative care or to go into a hospice will lead to the patient dying sooner. But the opposite is at least sometimes the case. An important study  published in 2010, found that patients with advanced lung cancer who saw palliative care specialists lived 25% longer than those patients who had usual care from oncologists. Other studies have found that hospice care extended the life of patients with pancreatic cancer by 3 weeks, and those with heart failure, by 3 months. US surgeon Atul Gawande has written: “The lesson seems almost Zen: you live longer only when you stop trying to live longer”.

Finally, media reports suggest that in this specific case, the courts are revisiting whether their previous orders should be revised – given that the patient has survived for longer than expected. This is entirely appropriate. Families, patients and doctors make decisions together to withhold some medical treatments because they are not working, or because they seem to risk doing more harm than good. But that may change over time. More information may become available, or, as in this case, the patient may live longer than expected. It may or may not be in his best interests now to receive chemotherapy or other potentially life-prolonging treatments. That question should be reviewed.

People sometimes think about decisions about life prolonging treatment as a fork in the road.*** They see palliative care as a one-way track, and worry that once they have left the main road they will never be able to return. But palliative care and life-prolonging treatments are not in conflict and they are not mutually exclusive. Conventional medical treatment is more akin to a river, while palliative care is the path running alongside. Sometimes it is appropriate to climb out of the water, and to walk more slowly along the tow path. Often, once people have left the water, they have no desire to return to the stream. But others will find, after some time, that they are ready to plunge back in. This may be for a short dip, or for a long swim. But the palliative path will be there at the side when they need it.

 

A/Prof Dominic Wilkinson is Director of Medical Ethics at the Oxford Uehiro Centre for Practical Ethics, and Consultant Neonatologist in Oxford

 

 

*”It’s tough to make predictions, especially about the future” Yogi Berra**

 

** This prediction is often attributed to Berra, though, ironically, its actual provenance is unclear. (It’s tough to make attributions?)

 

***One of Berra’s other (in)famous quotes: “If you come to a fork in the road, take it!”

  • Facebook
  • Twitter
  • Reddit

One Response to It’s tough to make predictions*

  • Anthony Drinkwater says:

    Thank you for this excellent overview . I hope that it also gets published to a wider audience.

Authors

Subscribe Via Email

Affiliations