A costly separation between withdrawing and withholding treatment

by Dominic Wilkinson

Television child star Gary Coleman died recently following a severe brain haemorrhage. He was taken to an intensive care unit, but the next day was taken off life support because of the severity of his brain injury.

Decisions like the one made by Gary Coleman’s doctors are common in intensive care. Many deaths follow decisions to stop intensive treatment because it is believed to have no chance of succeeding or because of the burden of illness even if the treatment does work. One question raised about cases like these is about the importance (or risks) of living wills. A separate question (and one that was raised during the critical care grand round earlier today) is about the difference between stopping treatment and failing to start treatment.

Is it more morally serious to stop intensive care after it has been started (ie to withdraw treatment) than to decide not to start intensive care (ie to withhold treatment)?

Most doctors and nurses in the US and UK believe that withholding and withdrawing treatment are not the same. They tend to worry far more about, and be far more reluctant to make decisions to withdraw treatment than to withhold treatment.

But ethicists, professional guidelines and the courts say that withholding and withdrawing treatment are equivalent.

“Where it has been decided that a treatment is not in the best interests of the patient, there is no ethical or legal obligation to provide it and therefore no need to make a distinction between not starting the treatment and withdrawing it.” GMC

One problem about making a distinction between withdrawing and withholding treatment is that this can lead people to decide not to start treatment that might be effective because they are worried that they will not be able to stop it.

However, here are some reasons why withdrawing (WD) and withholding (WH) treatment are not treated alike.

The first is that many cases of WH treatment are different from cases of WD treatment in terms of prognosis or the burdens of treatment – for example a patient might have a <5% chance of CPR being successful if they are resuscitated after their heart stops. However a patient who has had a cardiac arrest and is on life support in intensive care may have a much higher than 5% chance of survival – simply by virtue of the fact that they have made it to intensive care.

But secondly, there is one practical difference between WH and WD treatment that doesn’t appear in most descriptions of the debate. Decisions about admitting a patient to intensive care are often influenced by resources. It is a sad fact of life that there are not enough intensive care beds available to treat every patient with acute organ failure. Doctors have to prioritise intensive care beds to those who will benefit from them the most. However, there is a general belief that resource considerations should never play a role in decisions to stop life support.

If that is correct, that resources play a role in treatment starting decisions but never in treatment stopping decisions, then we would have a good reason for distinguishing between the two. Distributive justice demands that we make difficult decisions about limited resources. But, if we can only make such decisions at the door of intensive care, and not once people are through the door, then we necessarily have to treat those decisions differently.

There are several possible responses to this account of the difference between WD and WH treatment. The first is that maybe doctors shouldn’t be allowing resources to influence their decisions about starting intensive treatment. Perhaps we should make more health care funds available, increase intensive care units’ capacity etc. However, in the meantime decisions have to be made on the basis of limited resources.

A second response is to worry about the consequences of making this distinction – because of the greater uncertainty early in illness it may not be possible to distinguish between patients who will and will not respond to intensive care. If we have a higher threshold for starting intensive care treatment (because of resources) some patients will be excluded from intensive care and die who would have responded to treatment and survived.

A third response is to suggest that perhaps the reason why doctors are willing to take resources into account in treatment WH decisions but not treatment WD decisions is because they believe (erroneously) that WH decisions are less morally serious. Then the difference in practice cannot be used as evidence to substantiate a genuine difference between stopping and not-starting.

However, if treatment withholding and withdrawal are genuinely ethically and legally equivalent, perhaps we should revisit the principle that in practice separates them. Perhaps resources should sometimes be taken into account in decisions to withdraw intensive care. 

  • Facebook
  • Twitter
  • Reddit

5 Responses to A costly separation between withdrawing and withholding treatment

  • G Owen Schaefer says:

    You mention divergence of opinion between medical professionals and ethicists/lawyers on the WH/WD distinction. Still, there’s another group whose opinions are equally, if not more relevant: the patients and/or their proxies. Indeed, I would argue a survey of those individuals would be more valuable than one of docs or ethicists or lawyers, when it comes to determining policy. After all, it is the patients whose interests are at stake here.

    In this vein, I think it’s important to note that it was Gary Coleman’s ex-wife, not his doctors, who appeared to have made the decision to withdraw life support: http://www.cnn.com/2010/SHOWBIZ/TV/06/14/gary.coleman.death.decision/?hpt=T2

    Coleman’s living will (which, notably, his ex-wife appeared to contravene) did not appear to directly address the distinction between withholding and withdrawing treatment; still, we would want to respect (within reason) any distinction he and/or his duly appointed proxy made between withholding and withdrawing treatment, regardless of others’ philosophical opinions about that distinction. Moreover, if the general patient population makes such a distinction (as doctors appear to), perhaps the default policy should be to respect that apparent distinction. This is not to contravene the GMC guidelines, which actually seem geared at providing treatment which is in the patient’s interest, even if later withdrawal would be emotionally difficult. Rather, in other sorts of cases where the WH/WD distinction is relevant (like the resource-allocation you mention), perhaps we should err on the side of recognizing it specifically because patients (whose interests are at stake) view it as valid. Or so I presume – again, a survey of patient views on this matter would be of use.

  • Dominic Wilkinson says:

    Thanks for these helpful comments Owen

    You are right that patients’ views may be relevant in individual cases. Most patients don’t differentiate in their advance directives between withholding and withdrawing treatment, but if they did that would give us a reason to treat them differently for that patient.

    You are also right that our policy in a democracy is going to reflect the values and beliefs of the general public (and not just those of ethicists). I haven’t found a good survey of public attitudes to the WH/WD distinction, but there is good evidence that public views about intensive care decisions are different from the views of doctors (http://www.springerlink.com/content/74hfnc2ycf8rnvjf/. However, if there is a good reason to think that certain beliefs by the public are mistaken we may not want to give as much weight to those beliefs, or we may wish to try to actively correct them. I didn’t mention it in the post, but one likely contributor to the perception that there is a difference between WH and WD treatment is the status quo bias, an unreasonable preference for the status quo over a change.
    http://www.ncbi.nlm.nih.gov/pubmed/16162749

  • Marco Antonio Oliveira de Azevedo says:

    Dear Dominic. You’ve said that “resources play a role in treatment starting decisions but never in treatment stopping decisions”; that would explain the difference between WH and WD treatments. I think you’re right, but the question is: why? Why resources can play a role in Treatment Starting Decisions (STARTING) but NEVER in Treatment Stopping Decisions (STOPPING)? My suggestion is that people think they’ve obtained a right in cases of treatments in progress. And patients can have personal reasons for the continuation of a treatment, even in case of (posterior) considerations of futility. Suppose a doctor has decided to initiate some treatment in a patient of some grave disease but concludes, after some days, that the treatment is futile. He figures it out that if he had that opinion previously he simply wouldn’t have prescribed the treatment. But the fact is that he actually did it, and the patient consented (even tacitly). Hence, formerly he had a privilege to make the prescription; but he doesn’t have now any privilege of withdrawing the treatment without his patient’s consent. Formerly, if he hadn’t prescribed, he wouldn’t be in error, for the patient has a right only to appropriated treatments (and futile treatments are not appropriated “by definition”). But now the fact is that he actually prescribed the treatment, and since the treatment is not “physiologically futile” (see Jonsen, Singer & Winslade, Clinical Ethics, for a distinction between “futile” and “physiologically futile” treatments), the patient could have some reasons for the continuation. Hence the patient can claim the continuation and the doctor must talk with him. Here we are in front of a great difference between WD and WH treatments, even if those treatments aren’t technically appropriated to the case any longer. If I’m right, by the same token, considerations of just distributions of resources don’t matter in STOPPING, only in STARTING (this is a kind of TROLLEY – see Judith Jarvis Thomson).
    Marco Azevedo.

  • Dominic Wilkinson says:

    Hi Marco Antonio,

    thanks for your comment. If I understand you correctly, you suggest that the patient has a ‘right’ to treatment that has been started, but that they do not have any such right for treatments that have not yet commenced. Furthermore, you suggest that treatments can only be withdrawn with the patient’s consent, but that this need to consult with the patient is not necessary if the doctor decides not to start the treatment.
    I agree with you that this is the way that doctors perceive such decisions. But it is far from clear that this asymmetry is justified. If the treatment would benefit the patient, why does the fact that the patient has or hasn’t yet received it make a difference to whether or not it should be provided? You are right that the doctor may have to justify a decision to withdraw treatment – to the patient or to the family. If they are just withholding the treatment they might get away without mentioning it. But again, this seems irrelevant to the question of whether or not it ought to be provided.
    Perhaps though, the necessity of talking to the patient/surrogate in the stopping case, is behind the reason for not taking resources into account. Denying patients treatment because of inadequate resources is a very hard sell. It is controversial, difficult and likely to cause the patient or their family some distress. Since resource decisions cannot be avoided it wouldn’t be surprising if doctors preferred to make them in situations where they can get away without having to broach the subject openly.

  • sue says:

    I wonder what it is called when they give only one medicine that is not to be used with the illness that the patient has. It actually makes the illness worse. Is that extremely severe neglect or maybe just murder????

Authors

Subscribe Via Email

Affiliations