Abstract
Like many behaviours whose categorisation has varied through history – for example as possession, as mortal sin, as crime, as illness – suicide appears to have now settled into its ‘natural’ place as a final sign of certain psychiatric disorders. Suicide has been medicalised. While suicide does not inevitably result from conditions like depression, personality disorder or schizophrenia, these psychiatric illnesses are typically and generally accepted as necessary motivators of all its instances. People who are mentally well do not commit suicide, and suicide is the sign, par excellence (albeit a retrospective one), of mental illness. It is the business of psychiatrists, clinical psychologists and some other health professionals to prevent suicide, because their jurisdiction is mental health and disorder, and because taking one’s own life is surely the worst thing that a person could do to themselves.
If these conventions apply to suicide, logic might suggest they should apply as well to assisted suicide. Assisting someone to commit suicide is to assist them to take their own life; and if that is the worst thing that people could do to themselves, then mental health workers should arguably attempt to prevent this from occurring too. How then might it be possible to argue the opposite case, i.e. to agree with the thoughts in the previous paragraph, but question their application to certain cases of assisted suicide? We know that at least some psychiatrists who believe that suicide generally ought to be prevented also believe that some cases of assisted suicide should be permitted (Ganzini et al., 1996; Schwartz et al., 1999; Shah et al., 1998). The most likely explanation of the apparent inconsistency is that, as Wittgenstein claimed, our intelligence can be bewitched by language (Wittgenstein, 1953/2001); in this case the evaluative meanings and implications of the same word can vary according to circumstances.
Suicide is a word that carries much evaluative baggage, in addition to indicating the bald physical fact of taking one’s own life. Suicide is a terrible thing, just because it is a matter of someone responding to their acute psychological pain in a highly effective but disastrous way, by acting in a way to which they saw no alternative, even though in the vast majority of cases, alternatives exist. It is in this sense that we usually describe suicide as irrational, because the existence of psychological pain that can be treated is generally agreed not to be a good reason to kill oneself. Assisted suicide, at least to those who believe that there are legitimate cases, and therefore that not all cases should be prevented, differs in the sense that the person who competently asks for such assistance, in circumstances that we can sympathise with, is not always motivated by psychological pain for which there is an acceptable alternative. At least some people, in circumstances of terminal illness, arguably make a rational decision to die, and to ask for assistance to die. 1 The meaning and value that we attribute to the same word is thus not necessarily the same as in the typical case of someone who commits suicide in the context of a mental disorder.
The claim that any decision about or request for assisted suicide can be rational is disputed, at least in relation to people with no faith in an after-life, who believe that after death comes nothing. How can we rationally decide that non-existence is better than existence? Since nothing can be known about non-existence, because there is nothing to be known about it, it cannot be rational to prefer it, even to an existence being experienced as very bad (Cowley, 2006; Devine, 1978). This claim in its turn has been disputed. There does not appear to be a problem of rationality, so this line of thinking goes, in the case where a competent person, unaffected by mental disorder and with adequate information, decides that their interests in continuing living have ended (Graber, 1981), even if what follows death is unknown. Moreover, we generally accept that people can rationally request the removal of life-sustaining treatment, where complying with the request has the same result as assisting suicide, and where the same opacity about non-existence exists. If it be contested that this is a rational request for the removal of burdensome treatment, it is still the case that the nature of the consequence of death is unknown, but it is also the case that there is no in-principle, morally relevant difference between the burdensomeness of a certain treatment and the burdensomeness of a particular condition. It is the burdensomeness, not its source, that should be seen as the relevant factor, on pain of discriminating against a subset of burdened people on the basis of the religiously based acts-omissions distinction.
Nevertheless, the claim that all requests for assistance to die are grounded in psychopathology has been advanced. This goes hand in hand with the consideration that assisted suicide, like ‘ordinary’ suicide, will always be a bad thing, and should always be prevented. In a previous issue of the journal, I analysed the proposal for the new psychiatric diagnosis of demoralisation syndrome (Clarke and Kissane, 2002; Kissane et al., 2001) as being, at least in part, motivated by an undeclared moral position concerning requests for assisted death, masquerading as a medical diagnosis concerning such requests (Parker, 2004a). Demoralisation syndrome defined any request for assistance to die as a cry for help, requiring therapeutic intervention, as it could never be rational. The problem for its proponents is that within its putative symptomatology, including hopelessness, helplessness, powerlessness, meaninglessness, impotence, and loss of a sense of control and security, is entwined a welter of evaluative language, concepts and concerns that are pertinent to end-of-life considerations. This is not to require of all psychiatric diagnoses some hard-nosed scientific objectivity: it is now widely accepted that many, if not most, psychiatric diagnoses combine descriptive (factual) and evaluative elements (Sadler, 2002). 2 But it calls into question the medicalisation of all requests for assistance to die on the basis of there being alternative responses that will be acceptable to patients. Whether or not demoralisation syndrome is included in DSM-5 (American Psychiatric Association, 2011), the proposed diagnosis suffers from at least this paternalistic feature of superimposing physician values on patient rationality.
Neither philosophy (Stanford Encyclopedia of Philosophy, 2011) nor the law (Re T, 1993) equate rationality with decision-making capacity, because different people’s rationalities involve different substantive beliefs and commitments, whereas decision-making capacity is a formal process that is not dependent on the nature of those beliefs and commitments. However, the same risks concerning superimposing physician values on patient rationality are run when the issue is the assessment of capacity. Readers may recall that when the Northern Territory became the first jurisdiction in the world to legalise voluntary euthanasia, one of the safeguards against mistakes was the inclusion of the requirement that people requesting assistance to die had to be judged by a psychiatrist not to be suffering from a treatable clinical depression in respect of the illness (Rights of the Terminally Ill Act, 1995). Now while depression does not necessarily extinguish competence (Ganzini et al., 1994; Hooper et al., 1996; Lee and Ganzini, 1994), this subsection was clearly designed to exclude decisions that could be judged as incompetent. For reasons of medical demography, the prominent Sydney psychiatrist John Ellard became a fly-in/fly-out consultant called to exclude depression from the first person to utilise the legislation (Alcorn, 1996).
There was no reason to question Ellard’s judgment about Bob Dent, but while we might intuitively choose psychiatrists as the best judges of patient competence, due to their familiarity with cognition and emotion, they are not immune from ignorance concerning the legal assumption and definition of capacity (PVM, 2000), from considerable variability in capacity assessment (Kim et al., 2011; Markson et al., 1994; Shah and Mukherjee, 2003), or from the influence of the gravity of clinical situations on their judgments of capacity (McKinnon et al., 1989).
A related matter at issue here is the need to disrupt the false conceptual connection often made between the level of risk involved in a decision and the ‘level’ of competence required to make it. This mistake has been made by philosophers (Brock, 1991; Buchanan, 2004; Buchanan and Brock, 1989) and judges (Re T, 1993), as well as clinicians. Given (a) the frequent linking of assisted suicide with suicide per se, (b) the fact that psychiatrists have an important role in the prevention of suicide, and (c) the general moral imperative of medicine towards beneficence, it should be no surprise that doctors, including psychiatrists, would consider that when the risks are high, a higher ‘level’ of decision-making capacity must be sought and established. Once this belief is in place, however, patient autonomy is immediately at risk, because demanding what is thought to be a higher level of capacity than in less serious clinical situations is likely to result in a number of judgments of incapacity that are inconsistent with the legal definition (Parker, 2004b; Wicclair, 1991).
The reason to think that this is the case is that the concept of different levels of capacity is incoherent. As long as I understand what my illness and the proposed treatment or refusal of treatment, including respective risks, involve, and as long as I have weighed these considerations in the light of my particular values and have somehow communicated them, then I am competent (Re C, 1994; Stewart and Biegler, 2004). This applies to minor illnesses, surgical procedures, life-threatening conditions and other grave situations alike. If these are the criteria common to all decisions, what would it be to have a higher ‘level’ of any of them? What is crucial in the more serious matters is that judgments of competence are made on the basis of good evidence of such understanding, because a mistaken attribution of competence can have such dire consequences. But attributing competence on the basis of poor evidence is a completely different matter than insisting on a higher ‘level’ of competence (Devereux and Parker, 2006; Kerridge et al., 2009). Such insistence, motivated by perceived seriousness of outcome, risks effacing the legitimate exercise of autonomy.
Conceptual clarity is a crucial requirement in the debate over assisted suicide. The possibilities for confusion by language, conflation of scientific and evaluative concepts, and the failure to adequately expose and test intuitions, abound. Psychiatrists especially need to avoid the puzzles and tricks that can bewitch them. There are two great elephants in the waiting room of psychiatry’s approach to assisted death. They are related, and they are seeking clarity and honesty from the subsequent consultation. The first is the puzzle of why it is only in the setting of active assistance, rather than across the spectrum of all decisions that bring about death, that the clangour about rationality and capacity appears, and why expertise, including psychiatric expertise, in assessing capacity is raised as a crucial requirement (Ryan, 1996). This suggests that it is killing per se that is not to be countenanced. So the second elephant waiting for clarification (and then neutralisation) is the lingering presence of the historical, religious allegiance to the sanctity of life doctrine in medicine (Kuhse, 1987). This allegiance no longer steps out from the shadows to explicitly speak its name in the assisted suicide debate, because simply to state that killing is wrong because God says so, or because it is self-evidently wrong, will no longer wash in a pluralistic society. But it continues to confuse, conflate and obscure, by its endurance in the culture.
There may be good reasons not to permit assisted suicide, but those canvassed here in the context of psychiatrists’ responsibilities and roles are not amongst them.