Skip to content

Physician Assisted Suicide and the Conflict Between Autonomy and Non-Maleficence

In the run-up to the 6th November ballot in Massachussets concerning the prescription of medication to end life, two prominent US physicians, Dr. Ronald Pies and Dr. John Grohol, have been debating the pros and cons of physician assisted suicide (PAS). In his article, Dr Ronald Pies argues against the legalisation of PAS, claiming in the title of his essay that “medical ethics must sometimes trump patient choice”. Although Pies’ acknowledges that respect for autonomy is one of the four cardinal principles in medical ethics, he goes on to claim that “. . . medical ethics must set limits on a patient’s autonomous requests” in the light of another of these cardinal principles, namely the principle of non-maleficence. What are we to make of this claim?

In considering this question, it is instructive to return to one of the original treatments of the four principles of medical ethics, provided by Beauchamp and Childress. According to Beauchamp and Childress, the four principles of medical ethics, (the principle of beneficence, the principle of non-maleficence, the principle of justice, and the principle of respect for autonomy) are to be understood as prima facie duties; [1] they are not to be understood as absolute inviolable principles, but rather as principles that can be overridden by other prima facie duties. This of course leads to something of a difficulty, since the principles can often come into conflict. PAS seems to be a case in point; here the principle of non-maleficence (which enjoins physicians from harming their patients intentionally) may be regarded as being in conflict with respecting the autonomy of the patient who wants to end their own life. This is problematic because when two prima facie principles conflict, it is often not clear how we are to determine which principle should take priority in a given situation.

In light of this, why should we accept Pies’ claim that the principle of non-maleficence should take priority over the principle of autonomy in cases of PAS? In one of his arguments, Pies’ appeals to our intuitions concerning (what he claims is) a case which is analogous to PAS in which we believe that non-maleficence ought to trump the patient’s autonomy. In the article referenced at the beginning of this post, he writes:

. . . a patient may feel completely justified and “empowered” in requesting that the dose of her pain medication be doubled, even though that would be extremely dangerous to her health.

Governed by the principle of nonmalfeasance, the physician will wisely decline the patient’s request, and explain why that is necessary. So, too, in my view, with a dying patient’s request for lethal medication or “assisted suicide.”

Pies’ point is that in the pain medication case, we intuitively believe that the principle of non-maleficence should trump what he claims to be the patient’s autonomous choice; surely we don’t think that we ought to allow a patient to take a drug which she doesn’t know will harm her just so that we can respect her autonomy. Surely then, we might think, non-maleficence must be the most important consideration in this case and in PAS?

However, Pies argument here is flawed because it is misleading to treat the pain medication case as analogous to the case of PAS.  Our reaction to the pain medication case is not an instance in which we really believe that non-maleficence should trump patient autonomy. This is because Pies is incorrect to imply that respecting the patient’s autonomy in this case requires complying with her request.

Let me explain why I think this is the case. When we consider the patient’s autonomy in the pain medication case, we might first ask, ‘in asking to have her medication doubled, what is the end that the patient is motivated to achieve, and is she autonomous with respect to that motive?’ At the most fundamental level, it seems that the answer to this question is most likely to be that the patient wishes to have her pain alleviated; assuming that she is competent and in pain, we have no reason to suppose that she is not autonomous with respect to this fundamental motive.

However, although the patient may be autonomous with respect to her motive to alleviate her pain, there is a good case for claiming that doubling her medication (as she has requested) would not in fact serve her autonomy. This is because, unbeknownst to the patient, doubling her medication will not serve as the means to the end that she is autonomously motivated to achieve; rather than merely alleviating her pain, doubling her medication would be extremely bad for her health. This seems crucial; indeed, the physician would surely be morally blameworthy to double the patient’s medication without informing her that doing so would be extremely dangerous.

The reason that the patient’s lack of information here seems crucial is that, as Suzy Killmister has recently argued,[2] in order to be autonomous, it is not sufficient for agents to merely be autonomous with respect to their motives; rather, autonomous agents must also have true beliefs regarding the relationship between the ends that they are motivated to achieve, and the actions which they undertake in order to achieve them. In lacking these sorts of true beliefs, agents seem to lack something which is necessary for acting effectively in pursuit of their ends, and thereby their autonomy.

Having recognised the importance of certain true beliefs to autonomous agency, we are now in a position to see the disanalogy between the pain medication case and PAS. In  the pain medication case, we are not deciding between (a) respecting the patient’s autonomy by doubling her medication and (b) adhering to the principle of non-maleficence by wisely declining the patient’s request, and explaining why that is necessary (to paraphrase Pies). Rather, the principle of respect for autonomy also demands that we decline the patient’s request and explain to them that doubling the medication would be extremely dangerous. This is because autonomous agency requires that agents have true beliefs about the nature of their actions and how they relate to the ends that they are motivated to achieve. Respecting autonomy does not require the physician to comply with this patient’s ill-informed choice.

Of course, it might be the case that even armed with information about the dangers of doubling her medication, the patient might re-iterate her request; then the principle of autonomy would demand that the doctor doubles the medication, if we assume that the patient is fully competent. However, having altered the case in this way, it now seems far less obvious that the doctor ought to prioritise the principle of non-maleficence, as was our intuition in the original case.

The pain medication case as Pies presents it sets up an illusionary conflict between the principle of autonomy and the principle of non-maleficence, a conflict in which only the most ardent liberal would deny that the latter principle should take precedence. We don’t think that we ought to allow a patient to take a drug which she doesn’t know will harm her just so that we can respect her autonomy. But this is not because non-maleficence ought to trump the patient’s autonomy. Rather it is because respecting autonomy does not require physicians to comply with such ill-informed choices; respecting autonomy requires that we allow people to act in accordance with their fully informed choices.

In Pies’ pain medication case, the principles of non-maleficence and autonomy are not, I have claimed, in conflict. However, in the case of PAS, the principles of non-maleficence and autonomy are in a conflict that is both very real, and extremely important. There may yet be good arguments for why non-maleficence should trump patient’s autonomy, and we should consider these carefully. However, our judgment concerning which prima facie principle ought to take precedence in PAS ought not to be clouded by an analogy which serves only to distort our appreciation of the importance of autonomy by misconstruing its nature.

[1]See Beauchamp and Childress, The Principles of Biomedical Ethics, OUP, 1977, pp. 45-7, and p. 89.

[2]  Killmister, “Autonomy and False Beliefs”, Philosophical Studies, (forthcoming).

Share on

3 Comment on this post

  1. Can the pharmaceutical industry be behind the suppression of autonomy and against PAS? Expensive medical equipment and drugs, along with doctors’ fees, will drastically be reduced if PAS becomes a choice for those who simply cannot endure the suffering caused by a disease or a condition.

  2. Re Khalid’s reply it is correct that the health industry in totum will lose with PAS, but on a broader basis (perhaps a bit cynical or cold hearted) it is the community as a whole that benefits by having expenses drastically reduced: one quarter or more of Medicare expenses are incurred on the last six month’s of life. However on a broader sense would it be “less maleficient” i.e. doing less harm, inflicting less pain, for the physician to attend to the patient’s needs? What will s/he be empathetically speaking?

  3. We all die. The questions are when and how?

    With the word suicide inserted into the jargon of PAS (Physician Assisted Suicide), this topic will always raise great emotion and prejudice (political and religious). The physician is caught in the middle and current social values prevent him from advocating the best clinical advice and course of action for the dying patient.

    The arguments will always be polarised should PAS be the description used. RFS (relief from suffering) as a concept and a criteria for the physician’s role with an individual person’s life would set a less prejudiced background to this important debate.

    The dilemma for physicians’ is that they have no clinical freedom to make a professional judgement that is in the best interest of the patient and accords with that patient’s wish as to their own inevitable destiny. Rather the physician faces the threat of criminal prosecution should they even comtemplate the inevitable for the patient and are denied by society the right to exercise professional clinical judgement.

Comments are closed.