Ethics of withholding life sustaining therapy

In contrast to the chapter on futile and non-beneficial treatments, this article concerns the beneficial effect of withholding such treatments, and how that may be the most appropriate step. The other issues to consider are matters of responsibility (i.e. who is responsible for making the decision). Consensus of medical experts is one of the preconditions which must be met, and even though doctors are not obliged to offer pointless therapy there should also be some agreement from the patient and their family. If there is disagreement, conflict resolution is a staged process which escalates from gentle social worker mediated meetings to the gradual involvement of more specialists and high levels of hospital administration, and ultimately to the appointment of a public guardian by the Guardianship Board.

This comes up in real life a ot more than it comes up in the exam:

  • Question 30 from the first paper of 2006 discussed the ethics of withdrawing treatment following a suicide attempt in a chronically ill patient
  • Question 1d from the first paper of 2004 asked candidates to manage the scenario of a cancer patient in multi- system failure, with treatment limitations the only sensible option.

An answer to SAQs which ask for an approach to a clearly unsalvageable situation might carry on in the following manner:

  • Establish the utility of therapy. As CICM themselves put it, "there is no obligation to initiate therapy known to be ineffective, nor to continue therapy that has become ineffective".
  • Where possible, manage acute issues. If the patient is acutely unwell and no decision regarding their treatment limitations has been made, then it may be appropriate to rescue them in the acute setting. This might mean going ahead with such burdensome therapies as intubation.
    The argument for this course of action is:
    • To disregard uncertainty regarding prognosis and patient wishes may result in "false positive" situations where potentially beneficial treatments are withheld inappropriately
    • Resuscitation may offer some relief of symptoms (eg. decreased effort of breathing following intubation, and the relief of pain by the associated anaesthesia)
  • Establish medical consensus. Agreement should be sought between the medical teams and the ICU team regarding what is appropriate in the management of the patient. The ethical concepts of beneficience, non-maleficence and autonomy need to be considered.
  • Establish competence. The patient's right to self-determination is not degraded by mental illness or anaesthesia, but their ability to make reasoned judgments about their future certainly is. If the patient is not competent, family members may act as surrogate decisionmakers.
  • Establish responsibility. Typically the patient themselves is responsible for making decisions regarding their health. If the patient is incompetent, one needs to establishing who is the "person responsible". This may be the nominated enduring guardian or "medical power of attorney".
  • Establish broad consensus. There should ideally be an agreement between the medical consensus and the views of the patient and persons responsible. Important features of this process are:
    • Adequate time should be allowed
    • Adequate information (eg. factors taken into account in reaching the medical consensus) should be fully offered
    • Responsibility for decisonmaking does not rest solely with the next-of-kin - this should be abundantly clear
    • Issues should be discussed in the forum of a family conference, with input from a multidisciplinary care team involving nurses and social workers.
  • Document the consensus decision. Important components need to include:
    • Basis for the decision needs to be clearly documented
    • The document should identify those amongst whom the consensus has been reached
    • A completed Advanced Care Directive in writing
    • A plan for escalation of therapy with clearly defined limits
    • A clear plan of symptom control
  • Provide high-quality end of life care. This is characterised by:
    • Adequate pain and symptom management
    • Avoiding the inappropriate prolongation of dying
    • A sense of control over the process
    • Strengthening relationships with loved ones
    • Relieving the burden on loved ones
      (Singer et al, 1999)

For required reading, one cannot go past the CICM Document IC-14, "ANZICS Statement on withholding and withdrawing treatment". This probably represents some sort of minimum standard. Another good reference for this sort of thing is a 2010 editorial from CC&R, relevant for many reasons not the least of which is the fact that it was co-authored by the current president of CICM.

"Withdrawal", "limitation", euthanasia and assisted suicide

In summary:

  • Limitation of therapy is the decision not to offer some sort of therapy even if it is indicated. When you say you're going to "limit" therapy, it is generally expected to mean that you are going to offer some treatments but not others. This might mean that dialysis, intubation or CPR will not be offered to the patient even if the indications for these therapies are met, and the patient will be offered "comfort care" instead.
  • Withdrawal of therapy is the active cessation of some therapeutic or life-sustaining process. This might be something immediately fatal (an extreme example, the discontinuation of VA ECMO) or something associated with a more gradual decline, like dialysis. In any case, the life-prolonging therapy is being withdrawn and is replaced by some sort of palliative measures.
  • Euthanasia is "a deliberate act undertaken by one person with the intention of ending the life of another person to relieve that person’s suffering". This practice enjoys variable support, and there is even disagreement as to whether it can even be considered a medical procedure.
  • Assisted suicide is "the act of intentionally killing oneself with the assistance of another who deliberately provides the knowledge, means, or both". Most authors do not make any ethical distinction between asissted suicide and euthanasia.

At least from a purely philosophical standpoint, these can be viewed as morally equivalent. Moreover, there is no moral hierarchy of treatments: turning off the ventilator is equivalent to stopping regular dialysis in terms of its lethality (they differ only by timeframe).

It can be said that if a person has the right to self-determination, they have the right to decide whether or not they receive life-sustaining treatment. From that, it follows that the person then has some right to decide as to how and when they die. Ergo, right to use various means to accelerate or precipitate the process of dying can be seen as the continuation of those rights.

On the other hand, when a clinician assists suicide or performs euthanasia, their objective is the patient's death - which seems morally wrong. In withholding or limiting therapy, death is caused by the (fatal, end stage) disease and the clinician's objective is purely to relieve suffering (which seems somehow more palatable to some people).

The legal foundation

Most legal systems make a firm distinction between acts intended to cause death and acts which passively allow natural death to occur. The former are almost always illegal. The latter are illegal only when there is some sort of "duty of care" whereby one is legally obliged to prevent death. In some countries, a competent patient may request a clinician to assist their suicide or end their suffering by euthanasia. In most situations the clinician is not obliged to agree (for instance in Oregon, only one in six explicit requests are honored). In places where there is no specific legislation regarding euthanasia, the clinician is not protected from homicide charges even by the fact that the patient had consented to the procedure.

Decisionmakers at the end of life

In their answer to Question 30 from the first paper of 2006 , the college mention "power of attorney" multiple times, mentioning that it in imperative to determine who that person is and whether they have jurisdiction over health or financial matters. This is probably a local Australian thing. Insofar as the author can establish, in NSW this is a purely financial position. Making somebody your attorney does not invest them with the immediate right to make life-or-death decisions about you; it only gives them the legal authority to look after your financial affairs.

Judging by the tone of the college answer, what they probably meant was Medical Enduring Power of Attorney, which is the Victorian equivalent of the NSW "Enduring Guardian". That is a person you nominate who "you appoint to make lifestyle, health and medical decisions for you when you are not capable of doing this for yourself". The excellent LITFL page on this topic also refers to it (Chris Nickson being from Melbourne).

Generally speaking, if you (like most people) have not made a special effort to nominate an enduring guardian, the decision defaults to the "person responsible". To quote, the Guardianship Act 1987 (NSW) defines the hierarchy of person responsible as follows:

  1. An appointed guardian (including enduring guardian) with the function of consenting to medical and dental treatment. If there is no-one in this category:
  2. A spouse or de facto spouse who has a close and continuing relationship with the person. If there is no-one in this category:
  3. The carer or person who arranges care regularly or did so before the person went into residential care, and who is unpaid (note: the carer’s pension does not count as payment). If there is no-one in this category:
  4. A close friend or relative.

Self-determination in writing: the advance directive

People obsess about having some written document which defines the patient's wishes at the end of life. As an expression of the patient's autonomy, such a document has ethical validity, but probably little legal weight. Consider that it is impossible to determine whether the patient was capable of making reasoned decisions about these matters at the time that document was completed. The situation is made worse by the crude informality of this documentation. "There is no prescribed format or form for an Advance Care Directive", writes the NSW Trustee and Guardian website. "Should you want to make an Advance Care Directive you can simply write down your wishes". Clearly this is not ideal. One can envision Advanced Care Directives presenting in the form of drink coasters, tattoos or murals.

Of slightly greater use is the physician-completed advance directive. In short, when you sit down with the patient or their family, you take it upon yourself as a clinician to determine whether or not they are capable of participating in the decisionmaking. The result is a more formal document, which also bears your doctorly signature.

Conflict resolution at the end of life

Let's consider a (purely hypothetical) scenario where the medical consensus varies substantially from the opinions of he patient and their family. The ensuing family conference is therefore characterised by raw emotion which threatens to spill over into lawsuits and violence. How does one approach this?

An excellent NSW Health document (CRELS, 2010) is a goldmine of information. For the stress-crazed exam candidate its 56-page body is a daunting obstacle, and generally speaking this whole "communication and ethics" section ends up beign left until the very end; ergo it is unreasonable to expect anybody to read it. Instead, the following summary is offered:

  • Allow time.
  • Have multiple repeat discussions
  • Accumulate evidence about the patient’s poor prognosis and present it in the same way each time
  • Avoid daily discussions and constant repetition of the prognosis (it will be seen as coercive)
  • Have continuity of participants (same family members, same doctors)
  • Health professionals should not outnumber the family members
  • Express certainty about poor outcome (avoid expressing uncertainty)
  • Obtain a second opinion from another medical specialist
  • Involve an impartial third party mediator (who doe snot have the authority to decide the outcome)

If the conflict escalates to direct threats, lawyers or the media:

  • Involve senior hospital and Area Health Service administration
  • Involve the Department of Health, including Legal Branch and/or Media Unit.
  • Apply to the Guardianship Tribunal for guardianship orders.

The Tribunal stage is a dire situation, and in a number of ways represents the abandonment of the principle of patient autonomy. The Tribunal can consent to specific medical treatment including palliative care. If the family are particularly difficult and there is no consistency or consensus in their decisionmaking, the clinician may apply to the Tribunal for a guardian, for example the Public Guardian, to be appointed. The appointed guardian is obliged to consult family members about their opinions, but is not obliged to be guided by them. The appointed guardian can then make a decision about the patient even if that decision is not supported by the family.

High quality end-of-life care

A much-quoted paper by Singer et al (1999) describes the findings of a series of interviews held with chronically ill or terminal patients. The following major domains were identified:

  • Adequate pain and symptom management
  • Avoiding the inappropriate prolongation of dying
  • A sense of control over the process
  • Strengthening relationships with loved ones
  • Relieving the burden on loved ones


Myburgh, John, et al. "End-of-life care in the intensive care unit: Report from the Task Force of World Federation of Societies of Intensive and Critical Care Medicine.Journal of critical care 34 (2016): 125-130.

Palda, Valerie A., et al. "“Futile” care: Do we provide it? Why? A semistructured, Canada-wide survey of intensive care unit doctors and nurses." Journal of critical care 20.3 (2005): 207-213.

Sibbald, Robert, James Downar, and Laura Hawryluck. "Perceptions of “futile care” among caregivers in intensive care units." Canadian Medical Association Journal 177.10 (2007): 1201-1208.

Danbury, C. M., and C. S. Waldmann. "Ethics and law in the intensive care unit." Best Practice & Research Clinical Anaesthesiology 20.4 (2006): 589-603.

Myburgh, John, et al. "End-of-life care in the intensive care unit: Report from the Task Force of World Federation of Societies of Intensive and Critical Care Medicine." Journal of critical care 34 (2016): 125-130.

Schneiderman, Lawrence J. "Defining medical futility and improving medical care." Journal of bioethical inquiry 8.2 (2011): 123.

Ardagh, Michael. "Futility has no utility in resuscitation medicine." Journal of medical ethics 26.5 (2000): 396-399.

Schneiderman, Lawrence J., Kathy Faber-Langendoen, and Nancy S. Jecker. "Beyond futility to an ethic of care." The American journal of medicine 96.2 (1994): 110-114.

Waisel, David B., and Robert D. Truog. "The cardiopulmonary resuscitation-not-indicated order: futility revisited." Annals of internal medicine 122.4 (1995): 304-308.

Corke, Charlie, William Silvester, and Rinaldo Bellomo. "Avoiding nosocomial dysthanasia and promoting eleothanasia." Critical Care and Resuscitation 12.4 (2010): 221.

Dickens, Bernard M., Joseph M. Boyle Jr, and Linda Ganzini. "Euthanasia and assisted suicide." The Cambridge textbook of bioethics (2008): 72.

Singer, Peter A., Douglas K. Martin, and Merrijoy Kelner. "Quality end-of-life care: patients' perspectives." Jama 281.2 (1999): 163-168.

CICM Document IC-14, "ANZICS Statement on withholding and withdrawing treatment"