Abstract
In rejecting active voluntary euthanasia, many medical practitioners and their organisations continue to rely upon the distinction between acts which hasten death and omissions/withdrawals which lead to death, despite considerable challenge to the moral relevance of the distinction over the last three decades [1–4]. As support for euthanasia has grown, mandatory assessment by psychiatrists of people requesting active assistance to die has been urged on the grounds that protection against abuses and oversights is required [5],[6]. Mandatory psychiatric assessment was written into the short-lived Rights of the Terminally III (ROTI) Act (1995) of the Northern Territory [7].
One's position on mandatory psychiatric assessment is not fixed by that on the ethical and legislative acceptability of euthanasia. Nevertheless, the acts-omissions distinction is often relied on, somewhat unwittingly, in advocating such assessment. In this paper, I challenge the proposition that psychiatrists should formally assess terminally ill patients' requests for active assistance to die through active voluntary euthanasia or physician-assisted suicide, by considering some links between the act-omission distinction and a number of historical, cultural and institutional features of medicine and psychiatry.
Acts and omissions
Killing another human being is widely regarded as morally wrong. Some people take it to be self-evidently true that killing is intrinsically and always wrong. Alternatively, killing might be considered morally wrong because it usually ends the life of a person who desired that his life and all its experiences, promises, connections and hopes continue. The acts and omissions doctrine amounts to the claim that there is a moral difference between an act and an omission even if the two events have the same consequence: the act of killing a person is morally worse than the act of allowing a person to die. Opponents of the acts and omissions doctrine claim that, everything else being the same, killing a person is morally equivalent to allowing him to die. They accept that killing is in most cases seriously morally wrong (for the reasons mentioned above), but it will be morally acceptable when a person no longer has the relevant desires. Consider two terminally ill patients, one who is respirator-dependent. According to the acts and omissions doctrine, withdrawing ventilatory support at the competent and informed request of the dependent patient, with that patient's death as the foreseen outcome, is morally acceptable, whereas actively bringing about the death of the second patient at her request is not. According to the moral equivalence thesis, there is no moral difference between the two cases: if the withdrawal is seen as acceptable, so too should be active assistance to die. Supporters of the acts and omissions doctrine often base the claimed moral difference on a causal claim: in the case of treatment withdrawal, the patient's death is caused by the disease, whereas in the second case, death is caused directly by a human agent.
Those committed to the view that killing is intrinsically wrong will have no interest in debating the moral equivalence question or investigating the conditions under which euthanasia might be acceptable, since for them there will be no such conditions. Conversely, anyone who advocates the importance of assessing patients who request active assistance to die, to exclude depression or ensure the rationality of their requests, is thereby committed to accept that, in some circumstances at least, the moral equivalence thesis holds. This is because the assessment aims at distinguishing those whose request is competent, rational and uncoerced from those where such factors will cause us to refuse the request. There would be no point making such assessments if one viewed active assistance to die as never acceptable (i.e. as intrinsically wrong).
Medicine and death
Medicine's association with death and its prevention is obvious, but is a historically recent development [8]. While death was often perceived as bad in past eras, it was also generally seen as natural and inevitable. Medicine's power to prevent or postpone death has increasingly made death something which our culture considers ought to be avoided [9]. The modern physician's duty to save life derives from that power in the context of a strong allegiance to the sanctity of human life. However, in recent years, we have witnessed the deconstruction of the unquestioned imperative to prolong life, underpinned by the loss of absolute allegiance to the sanctity of life through the secularisation of society. This has led to a greater emphasis on patients' rights to refuse treatment [10–12], a clearer distinction between therapeutic and comfort measures elaborated within the palliative care/hospice movements, and subsequent accommodations within the medical profession itself towards a ‘re-naturalisation’ of death and dying [13]. The euthanasia movement, too, is motivated by these de-medicalising imperatives, since euthanasia arguably would deliver certain individuals from a socially imposed prolongation of life, in which the institution of medicine plays a crucial normative role. While some perceive euthanasia as a pinnacle of medicalisation, the crucial issue in the balance between medical power and patient autonomy is whether patients' settled and competent wishes are met, not the identity of the person who assists in their dying, whether by active assistance or a treatment withdrawal. Similarly, while palliative care is argued as reducing curative medicine's power over the terminal phase of life, some terminally ill patients come to perceive palliative care as inappropriate, and as unacceptable medicalisation, the more it is normalised as the only possible kind of ‘good death’ [14].
As terms such as ‘natural’, ‘artificial’, ‘ordinary’, ‘extraordinary’, ‘medicalisation’ and ‘demedicalisation’ have increasingly been perceived to function evaluatively as well as descriptively according to context, movements to weaken the grip of medical practice on dying have become stronger. While doctors often claim that the law has intruded too far into the doctor-patient relationship, an alternate perception is that it has gone some way to reducing modern medicine's powerful, but unarticulated, dominion over death and dying.
Medicine, euthanasia and treatment withdrawal
‘Euthanasia is the single most serious challenge to the Hippocratic tradition that has nurtured doctors for well over two thousand years’ [15] and organised medicine continues to reject it, adhering to the sanctity-of-life principle and the moral distinction between acts and omissions [16],[17]. Medical participation in active euthanasia is also seen as violating the goals and ideals of a healing profession [18],[19]. However, while acceptance of active euthanasia is lower in the medical profession than it is in the community, the level of medical acceptance is significant [20]. At least some practitioners consider that ideals can sometimes blind us to the real needs of suffering patients [21]. Empirical studies demonstrate that a significant number of individual doctors flout official ethics and the law to actively assist patients to die [22],[23].
Organised medicine has had far less difficulty accepting the legal formalisation of the right to refuse unwanted treatment than involvement in active euthanasia, due largely to the perception that when treatment is refused, it is the disease and not the physician which causes death [9]. This intuition makes it easier to think that requests for active assistance, but not those for withdrawal of treatment, require interpretation and exploration. Recent research has been undertaken into factors which may influence refusals of treatment, including clinical depression [24–27]. That such research is now occurring arguably results, at least in part, from a new, but unstated (perhaps unconscious), recognition that the need to explore and interpret requests for withdrawals is no different from the need in respect to requests for active assistance to die. If so, the acts-omissions distinction does not mark the moral boundary that the profession has long supported, but the fact that the research is recent also demonstrates how official doctrine influences both research agendas and clinical practices, both of which have important social implications.
The recognition of the need to explore requests for treatment withdrawals also weakens the claim that active euthanasia should not be legalised because we have not yet had sufficient enquiry into the many factors which can motivate euthanasia requests [28]. There are no new calls for a moratorium on treatment withdrawals pending adequate enquiry into these decisions.
The de-medicalisation movements share concerns about personal autonomy and quality of life in the face of the harmful possibilities of medical power. They focus on issues of decision-making competence and voluntariness, the meaning of a person's dying, interpretation of requests, burdensomeness and coercion as these influence end-of-life decision-making, and these issues must be faced in the contexts of both active euthanasia and treatment withdrawal. They are therefore not factors to which we may indefinitely appeal to reject assisted death while remaining neutral about them in situations of treatment withdrawal. They demand reasoned distinctions and judgements now that active euthanasia is widely socially accepted, as indicated by both public opinion polls and academic research [20].
Psychiatrists are clinically involved in assessing competence and interpreting mental symptoms, and with suicide. Because active euthanasia, but not refusal of life-sustaining treatment, is conventionally conceptualised as suicide, there is a strong assumption in favour of mandatory psychiatric assessment of euthanasia requests. If such assessments are supported, however, there appears to be as much reason to require mandatory psychiatric review of treatment withdrawal decisions.
Suicide, society and psychiatry
Suicide has been variously conceptualised across historical periods and cultural groups as heroic, sinful, criminal or rational [29]. Suicidal thoughts are now conceived as symptoms of mental illness, as are suicide attempts, which were, until recently, punishable choices. There are strong correlations between suicide and mental illness [30]. Psychiatrists occupy a frontline position in the fight against the problem of suicide, and the perceived culpability for failure to prevent it adds strength to the generalisation of a medical, preventable conception [31]. The medical-isation of suicide and the consequent relief from moral responsibility for the act [30] may also have been strengthened by the relatively high rate of suicide among psychiatrists themselves [32].
Psychiatry still searches for acceptability within the medical, and (to a lesser extent), the wider communities [8],[33]. Its subject matter includes phenomena accepted as causally determined and actions which most people would describe as the exercise of free will. This distinction has forensic implications: psychiatric advice determines offender status. Diagnostic systems such as the DSMs of the American Psychiatric Association have been analysed as inevitably evaluative as well as descriptive classifications [34]. Despite its enormous influence, Freud's project purportedly failed to provide a true science of persons [35], partly because it failed Popper's claim that a scientific theory must in principle be falsifiable [36], and there was no method of refuting psycho-analysis. Because it deals necessarily with illnesses of the mind, because the mind-body distinction remains the enigma that it does, and because our recalcitrant concepts concerning freedom cannot be accommodated within reductionist science, psychiatry remains at the fringe of scientific respectability.
Thus conceived, psychiatry could be expected to try to earn the respect of its sister disciplines by being as scientific as it can. Science aims to discover objective truth; this requires causal explanations which show why a particular event could not but occur, given just those conditions existing at the time. As a seeker of causes and explanations (e.g. of newly noted epidemiological patterns, such as increases in suicide rates among young men), scientific medicine tends to reduce reasons for action to causes, narrowing the possibilities of conceiving various actions as rationally free. The psychiatrisation of suicide has arguably narrowed the conceptual possibilities of the phenomenon, including that of rational suicide and its entailed notion of free choice.
One of the ways in which psychiatry has attempted to establish a respectable scientific basis is by presenting a ‘biopsychosocial’ model of disease [37]. However, while this approach importantly extends the purely biomedical treatment of illness, it is not an adequate explanatory model [38]. An adequate model for psychiatry would provide a satisfactory unitary conception of the mind: no small thing. But the assumption of the biopsychosocial approach as a complete, descriptive model encourages exhaustive exploration of each of the three dimensions to reveal a cause of the phenomenon being studied. Finding a cause promises a remedy, since a consequence of explaining something is the possibility of better controlling it. Requests for assistance to die invite the search for such remediable causes. The duty of care demands a therapeutic response, by tending to rule out the possibility of a rational judgement that a life has ceased to be worth living.
This would have an interesting and paradoxical result. In Nazi Germany, doctors and others defined certain lives as not worth living on the basis of morally irrelevant criteria, including race. If modern doctors disallow the possibility that some people's judgements that their lives are not worth continuing can be rational, they will be formally repeating this gate-keeper function. While we might agree that having your life prolonged against your will would usually be preferable to having your life terminated against your will, defining the value of others' lives is the same paternalistic process in both cases.
Perhaps another reason we have not, until recently, questioned the rationality of treatment refusals, is that we perceive such refusals as motivationally removed from death, in contrast to requests for active assistance to die. The directness of death from euthanasia encourages us to think that the patient's core motivation is a desire to be dead. What we conventionally associate with treatment refusal is a desire to be free of unwanted treatment, and this desire is more easily construed as prima facie rational than a desire to be dead. But, as for those who seek withdrawal of life-sustaining treatment, there is as little reason to suppose that those who request assistance to die wish primarily to be dead. Rather, they too have reached a point where their lives are of no further benefit to them, and desire release from burdens which are no longer acceptable. This leaves both kinds of decisions which can lead to death as presumptively rational. But it also leaves intact the general medical obligation to ensure that treatment refusals and requests for assistance to die do not emanate from remediable causes.
Psychiatry, psychiatrists and rational suicide
I submit that a request for assistance to die would be rational if there is good reason to believe that the quality of one's life renders it not worth living, and that this situation will not improve [32]. Now, while such a belief is a subjective state, its rationality depends on a person's being justified in losing hope, and this justification is obtained through the deliberations of members of a community [39]. The concern that subjective rationality, unconstrained by community consensus, would underwrite unreasonable cases of assisted dying, should also be applied to cases of treatment withdrawal. But laws which now provide for treatment refusal do not attempt to define suffering, nor do they specifically attempt to exclude undue influence on refusal decisions [40]. The safeguards contained in the Rights of the Terminally Ill Act were significantly more protective of a consensus on hopelessness and suffering than any Australian death with dignity guidelines or treatment refusal legislation [40]. Laws pertaining to all types of end-of-life decisions should ensure that decisions are free, rational and proceed from states of suffering which, while all reasonable people might not deem them hopeless from their own point of view, all could view as hopeless in the light of the individual's particular situation and values.
That psychiatrists are not required to evaluate treatment refusals reflects a presumption in these cases in favour of patient competence, a cornerstone of law in western jurisdictions, whereas the requirement for psychiatric assessment of requests for assistance to die reflects a presumption of incompetence for which there is neither empirical evidence nor moral justification. Sometimes hopelessness will be a psychiatric symptom, and sometimes a request for assistance to die will represent a psychodynamic process calling for exploration [41]. But if rational suicide is possible [42], the need to explore requests should be tied to the possibility that rational responses to objectively hopeless situations may sometimes mimic depressive or other symptoms which in other circumstances would demand treatment. Although some studies show that treatment of depression can sometimes influence treatment refusals, treatment does not always have this effect [25],[27] and there is no obvious reason to think that the same would not apply to requests for assistance to die. There is a conceptual and a practical distinction between the diagnosis of depression and the assessment of decision-making competence [26]. The failure to recognise and respect this distinction can lead to the overriding of autonomous decisions of some people who happen to be depressed.
The connections between assessments of decision-making competence, seriousness of decisions and the values of the physicians who assess competence are also relevant. There has been recent debate over the appropriateness of risk-related standards as against fixed thresholds in defining decision-making competence [43–45]. Risk-related models require higher levels of competence for decisions with more serious consequences. Fixed threshold models apply the same level to all decisions. If a request for assistance to die is perceived to be motivationally closer to a desire for death than a request for withdrawal of life-sustaining treatment, there will be a greater tendency to utilise risk-related assessment. But tying competency to seriousness of outcome in this way trivialises the notion of self-determination, if we accept that adult autonomy does not require that all our decisions will be in our objective interest [45]. Employing the risk-related model increases the like-lihood that others' values concerning the seriousness of outcome will infiltrate the decision-making process [45],[46]. While such biases are clearly not restricted to psychiatrists, their very alertness to the possibilities of depression and other motivational states in the setting of patients' assertions of suffering, together with their own motivation towards treating depression and preventing suicide, may move some to utilise a risk-related model and discount the validity of patients' assertions. The veto power given to psychiatrists under a mandatory assessment regime may tend to both discount the rational validity of some patients' requests for assistance to die, and illegitimately medicalise the solution of a difficult social issue which instead requires decisions by individuals, communities and polities [46].
Psychiatric gate-keeping for euthanasia
Despite these possibilities, we might think that, in view of the strong connections between suicide and mental illness, and since psychiatrists are involved more frequently than other practitioners in assessments of competence, psychiatrists are still better able to gauge this element. Further, given that psychiatrists presumably have greater insight than other practitioners into the mechanisms of introjection and projection which occur in relationships with their patients, they should be better equipped to deal with the possible distortions which may arise in emotionally charged situations and which may influence assessments of competence and interpretations of requests. However, these strengths of theoretical insight have not rendered psychiatrists any more immune to errors of judgement than other practitioners, for example, those who rationalise harmful sexual relations with patients [47]. If there have been failures to translate theoretical abilities to discern one's own and one's patient's motivations, needs, projections and introjections into practices which protect patients' interests by significant numbers of the very group which could be expected to do so, there is no reason to suppose that in other areas where practitioners' motivations are so strong they will be less prone to exploiting patients under cover of what they perceive to be the provision of beneficent care.
Kelly and Varghese have argued the reverse of this point: through the provision of assisted death, doctors may disguise their own feelings of futility and failure and regain feelings of control, and these countertransference responses may influence patients' perceptions and decisions [48]. Kelly and Varghese regard requests for assisted death as analogous to requests by a psychiatric patient for a sexual relationship; such requests cannot be autonomous but must be examined in the context of the relationship; such examination requires the framework provided by the prohibition. But, again, this is to mandate that requests for assistance in dying are always reducible to some therapeutic explanation that will enable the reversal of the initial request. To make the prohibition of assisted death a necessary condition of the therapeutic endeavour is surely to codify beneficence according to a particular moral view of the acts-omissions distinction.
Responding to requests for assistance to die or for treatment withdrawal which will lead to death involves the conflict between the duties to preserve life and those to relieve suffering. The Dutch have allowed this conflict to be resolved in certain cases through the concept of euthanasia as a lawful exception to a retained prohibition on active killing [49]. The less satisfactory way of resolving the conflict is through a process of moral splitting, whereby the choices involved in the conflict are attributed as exclusively good or bad according to the acts-omissions distinction, so that the duty to relieve suffering is rendered impossible to fulfil through its association with killing. Such splits avoid the complexities of these situations, but can impose on the patient an undesired prolongation of life.
It is also claimed that depression is under-diagnosed by non-psychiatrists [6], and that these practitioners may prematurely acquiesce to requests for assistance to die which may otherwise have been retracted. While the under-diagnosis of depression by non-psychiatrists should certainly be of concern, this should not rule out the possibility that (especially in a field where diagnostic categories are often problematic due to their syndromal nature and evaluative components), depression may actually be over-diagnosed by psychiatrists and other practitioners influenced by psychiatry as part of the more general process of medicalisation of some adjustment problems and other vicissitudes of life [50].
The model of gate-keeping followed in the ROTI Act mandated only one psychiatric assessment. Both opponents and supporters of euthanasia agree that this is not sufficient to make an adequate assessment of competence and to interpret requests for assistance to die, although their motivations for rejecting the model will be different. Supporters of euthanasia who oppose mandatory psychiatric review will claim that this shifts responsibility and accountability away from where that very process of exploration and support has occurred and should continue: between patient, family and usual caregivers, including the patient's doctor. On this view, psychiatrists should be involved optionally as part of a healing team looking to augment autonomy, not merely to check that it is present. The Oregon Physician-assisted Suicide Act adopts this ‘encouragement’ model, in contrast to mandated psychiatric assessment [51]. Opponents of euthanasia may raise the same objection, but if this is on the basis that one mandated assessment would displace a more exploratory and therapeutic relationship with the patient, with different goals from that of euthanasia [52], this again rules out, a priori, euthanasia as a possibly reasonable choice.
Conclusions
I have explored the question of mandating psychiatric assessment of patients who request assistance to die in terms of historical trends in the relationship between the institution of medicine and the society it serves, and in terms of the influences of and upon the discipline of psychiatry. Mandatory assessment is an inappropriate medicalisation of decisions which, while they must be informed by medical expertise, are essentially non-medical. Psychiatrists particularly should be aware of the relationships between beliefs, politics, power and individual freedom [53], and of the historical abuses of their discipline in pursuit of a number of political goals. They should consider the implications of the institutional and cultural forces which they may unwittingly instantiate. Such forces can threaten the freedom of those facing the end of life, surely a time when freedom ought to be protected and promoted.
Footnotes
Acknowledgements
The author acknowledges review of earlier drafts of this paper by Colleen Cartwright.