Abstract
The Royal Dutch Medical Association Special Committee on the Acceptability of Termination of Life has recently issued guidelines concerning euthanasia for mentally incapacitated patients [1]. This startling development in the Netherlands has the potential to profoundly influence the practice of psychiatry and warrants careful consideration by our profession. It comes shortly after the experience of legalised euthanasia in the Northern Territory of Australia, and at a time when, in Oregon, USA, there is new legislation permitting physician-assisted suicide (PAS) of the terminally ill. The euthanasia debate seems unlikely to go away. In our view, therefore, psychiatrists should be well informed about relevant clinical and ethical issues.
In this paper, we present our perspective on these issues in the hope of enhancing greater understanding about the clinical complexity and generating further informed discussion. We begin by reviewing surveys that inform us of psychiatrists' views regarding euthanasia and PAS. Then we consider how the desire for death presents clinically. What are the important diagnostic issues if a patient is not depressed? To what extent does interest in life-sustaining treatment change once depression or demoralisation are effectively treated in the medically ill? When is the depressed or demoralised patient no longer competent in their decision-making with respect to suicide? What is the role of the psychiatrist in the assessment of death wishes and what are the concerns generated by the Dutch guidelines for euthanasia of psychiatric patients?
Surveys of psychiatrists' views about PAS and euthanasia
A recent survey in the UK of 322 psychiatrists, whose mean age was 47, revealed that 86%% believed in rational suicide, while some 40%% favoured euthanasia and PAS [2]. Groenewoud and colleagues [3] estimated that 64%% of Dutch psychiatrists endorse PAS for intractable mental disorders, even in the absence of terminal physical illness. This view is striking given our knowledge from psychological autopsies that 80%% of persons with cancer who complete suicide have depressive syndromes, the same proportion as people without cancer who commit suicide [4].
Consultation-liaison psychiatrists surveyed in the USA showed strong opposition to the practice of PAS [5]. It seemed that the precarious nature of informed consent in vulnerable and seriously ill patients was well understood by this group of practitioners. Furthermore, when Oregon psychiatrists were asked about the value of a single assessment session to appraise a patient seeking PAS, only 6%% believed that such a session could generate a competent assessment [6]. However, Ganzini and Lee [7] suggested that beliefs about what is permissible expand as a result of acculturation to change in social policy, thus offering explanation for the more liberal view found among Dutch psychiatrists.
A sizeable survey of American specialists likely to be engaged in the care of the dying revealed that some 11%% were willing to assist with PAS, 7%% euthanasia [8]. Interestingly, only 18%% of these physicians had received a request for assistance with dying. Within a small group of some 5%% who acknowledged that they had assisted patients with euthanasia, the majority attributed death to opioids; only 17%% used potassium chloride. As increasing recognition is given to the safety of morphine in the care of the dying [9], it seems likely that many of those deaths attributed to morphine actually died from their disease. Despite widespread interest in managed death in the developed world, the majority of practitioners engaged in caring for the dying appeared unwilling to become involved with euthanasia. Respiratory physicians were noted to be more likely to accede to such requests.
Depression and demoralisation in the medically ill
There is strong empirical evidence demonstrating the association between desire for death and depression [10–13]. Recent studies of patients with motor neurone disease that explored their attitudes toward PAS have provided evidence that hopelessness and depression are not synonymous [14],[15]. Depressed patients did express hopelessness, but other patients were noted to be hopeless without being depressed. Indeed, Beck and colleagues [16] argued that hopelessness was a better predictor of suicidal ideation than depression, a view supported by a range of similar studies [17–23], particularly in younger adults [24]. In contrast, a study of suicidality in older adults found that hopelessness was not effective at predicting suicidal ideation at low levels of depressive symptoms. Once higher depressive symptomatology occurred, hopelessness accounted for greater variance in suicidal ideation [25]. In a terminally ill cohort in Canada, Chochinov's group showed that hopelessness contributed uniquely to prediction of suicidal ideation when depression was controlled for [26]. This differentiation of hopelessness from depression was further explored through latent trait analysis of a comprehensive symptom list from an Australian cohort [27], which showed that anhedonic depression, demoralisation and grief were separate dimensions in the medically ill. Indeed, demoralisation may be a comorbid, clinical syndrome in which the cognitions of hopelessness, existential despair and pessimism exist without a formal diagnosis of depression being present [25]. Such cognitive attitudes may be trait phenomena that persist in individuals across many years. Historically, they were reflected in constructs such as the depressive personality [28]. Some might argue that the diagnosis of adjustment disorder following the stress of illness covers demoralisation. Alas, clinicians often fail to recognise the maladaptiveness of cognitions that occur in demoralisation, which points to the benefit of delineating demoralisation as a specific clinical syndrome of the medically ill. This is not to pathologise normal reactions to life adversity, such as when a person is disheartened, for the intensity of distress and the maladaptiveness of response is excessive in the demoralised.
It is also important to differentiate desire for death due to demoralisation from acceptance of dying, which may be found as a common attitude in the elderly. Individuals who carry a high level of life satisfaction may be noted in their senior years to acknowledge a readiness for their life to come to a close [29]. Such acceptance of dying does not lead to an urgent desire for death. Freud, who battled cancer across 22 years, was an example of one man who appeared quite accepting of his death, having had a vitally strong life force. Irrespective of whether his physician, Dr Schur, gave Freud 20 or 30 mg of morphine and whether the dose was repeated once (as stated by Schur in Freud, living or dying [30]) or twice (as stated by Peter Gay in Freud. A Life for Our Time [31]), 36 h unfolded before Freud died. Given this time that elapsed, palliative care physicians would have no doubt that Freud actually died from his disease [9].
It is the cognitive rather than somatic features of depressive syndromes that are important in the medically ill. These include loss of interest and pleasure, depressed mood, loss of concentration and inattentiveness, indecisiveness, social withdrawal, guilt, worthlessness, hopelessness and helplessness, loss of value in continued life and the development of a suicidal urge and active suicidal ideation. Clearly, features of psychomotor retardation, melancholia and psychotic behaviour are also critical, but less common in the medically ill.
Anhedonia and loss of interest with depressed mood are the cardinal features of a major depressive episode. In contrast, demoralisation syndrome has hopelessness as its core construct. Its other key features include cognitive attitudes such as pessimism, stoicism and fatalism; existential despair with loss of purpose, sense of failure or meaninglessness; and the development of social isolation and alienation, the latter being associated with a poor social support network or dysfunctional family. Negative automatic thoughts with thinking errors such as all or nothing thinking, selective attention to the negative, labelling oneself negatively and emotional reasoning (i.e. the connection is illogical) may be particularly prominent. Others demonstrate hypochondriasis and a tendency to somatise, often associated with a long history of poor health. The effects of such cognitions on pain [32], evaluation of the future, erosion of hope, self-esteem and self-worth, and on determination of the risks or benefits of any treatment may be substantial [7],[33], yet often unrecognised. When we look back on Chochinov's Canadian study of palliative care patients [12], 60%% of those desiring death suffered from a depressive disorder. We should ask whether the other patients were rationally choosing suicide (lack of family support and pain were other identified predictors), or did they suffer from a demoralisation syndrome? More research is needed to empirically evaluate this newly proposed syndrome of demoralisation.
Importantly, subtle effects of the demoralised perspective can remain unchallenged by those around the patient, including the doctor, who may too closely identify with and share this demoralisation [34],[35]. Key tasks in adaptation to progressive illness that require recognition by the doctor include promoting change in the perceived need for independence and fear of loss of autonomy or control, rather than colluding with the notion that a life lived without such independence is ultimately unbearable [36]. Collusion in this manner would confirm the patient's negative self-precepts and escalate demoralisation.
Examples of patients with demoralisation syndrome were recorded among those who sought euthanasia in Darwin [37]. For example, a single 56-year-old woman had a 20-year history of carcinoid tumour and a more recent 5-year period involving surgery for recurrence, with complications spoiling her quality of life (see case 7 [37]). She withdrew socially, stopped reading and letter writing and gradually lost purpose in her existence. She thought increasingly about euthanasia over a number of months and completed the necessary documentation as was required by the Rights of the Terminally Ill Act, Northern Territory, 1995 [38], but did not yet want to die. She was judged to be not depressed at that stage by the reviewing psychiatrist and continued to receive care from her general practitioner and nursing friends. Her completion of the euthanasia documentation could be argued to correspond with a period of hopelessness, without fully developed features of a depressive syndrome. Later, as her disease progressed and she became more distressed, active suicidal ideation occurred with more clear-cut features of a major depressive episode. Other patients cited the pointlessness of their lives as their reason for requesting euthanasia (see cases 2 and 5 [37]). This loss of meaning suggests demoralisation.
Treatment of both demoralisation syndrome and depression
Demoralisation syndrome could be a harbinger of major depression, but a clinician engaged in the care of the above patients should not wait until major depression developed before seeking to intervene. Clearly, we could work psychotherapeutically and socially to improve the quality of life of patients who develop a demoralisation syndrome. A supportive psychotherapist would aim to provide encouragement with a focus on present day activities, while also promoting mourning and facilitating other expression of feelings. Perhaps the crux of this issue concerns the capacity of those caring for the dying patient to be able to bear the patient's suffering, listen to their experience and tolerate the distress. This ability to contain helplessness challenges clinicians at a personal level, confronting them with death anxiety and related existential concern [39]. Regrettably, it is in these tasks of communication with patients that doctors perceive their greatest deficiencies, feel considerable strain and yet continue to receive less than optimal training [40],[41]. Clinicians often struggle with identifying realistic hope, differentiating adaptive minimisation from unhelpful forms of denial, discerning an inner hopelessness from the apparently ‘hopeless’ prognosis of terminal illness, and challenging ‘negative’ attitudes about daily activities when prospects for recovery appear absent.
Cognitive-behaviour therapy would promote activity scheduling and reframe negative cognitions to counter pessimism and sustain a realistic attitude, focusing on the here and now, generativity and continuing with worthwhile activities that promote meaning in any remaining life. Social approaches would capitalise on the support of a volunteer program, make use of a variety of complementary therapies and promote interactions with family and friends to sustain genuine social encounter [42]. Alongside psychotherapy, pharmacological therapies are efficacious in treatment of depression in the medically ill, with modern agents causing fewer side-effects [43].
A key question is whether interest in life-sustaining treatment increases once depression or demoralisation are treated in the medically ill. Successful treatment of depression is indeed accompanied by an increased desire for life-sustaining interventions [44–46]. However, some patients may remain hopeless because of a persistent cognitive style, beyond improvement in mood and pleasure. These should be particularly targeted with psychotherapeutic management alongside pharmacotherapy.
Competence in association with depression
It is often argued that mild depression will not mar a patient's informed clinical decision. Undoubtedly, competence is lost with severe depression. However, the transition phase through which competence is sustained before a depressed patient loses their previously held positive world-view is unclear. Pain and other physical symptoms, demoralisation, depression or coercion from another may gradually confound the autonomy of any decision. The clinician is thus challenged to identify and actively treat such states, while also exploring systemic aspects with staff and family.
It has long been recognised that when patients have taken a firm decision to suicide and completed in their mind a plan of such action, they may appear calm and composed. They will trap the unwary clinician that does not continue to explore their cognitive stance.
Regrettably, poor clinical decision-making does occur and this was well exemplified by case 4 in the Darwin euthanasia series [37], in which a patient was diagnosed with a major depressive illness, but did not receive conventional treatment because of the doctor's reported fear of side-effects. She suffered from mycosis fungoides, was troubled by pruritis and left on doxepin 150 mg, with the doctor giving no consideration to prescribing selective serotonergic reuptake inhibitors or a less anticholinergic tricyclic and psychotherapeutic treatment of her depression. Her psychiatrist's opinion reflected a lack of experience in the field of consultation-liaison (C–L) psychiatry.
The psychiatrist's role in assessing the medically ill with a desire to die
Debate exists about the wisdom of a psychiatrist becoming involved in the assessment of patients seeking euthanasia. Orr and Bishop [47] argued that while suicide prevention remains seminal for psychiatrists, their ability to accurately appraise a patient's motivation to die, or their competence and autonomy in decision-making is doubtful. Orr and Bishop cautioned against a purely bureaucratic psychiatric involvement, a view that contrasted with that formed by the Australian and New Zealand College of Psychiatry's Working Party for the Section of Consultation-Liaison Psychiatry. The latter group saw the request for euthanasia as a plea for help, needing intervention and hence necessitating an active and ongoing psychiatric role [48].
One challenge for the psychiatrist is in the development of a therapeutic alliance. The experience of euthanasia in the Northern Territory provided evidence of patients withholding key information, because they saw the psychiatric assessment as a legislative hurdle that needed to be overcome [37]. The success of any psychiatric intervention is dependent upon the establishment of trust and assumes a comprehensive assessment prior to the development of management plans. It was clear that the Rights of the Terminally Ill legislation [38] defined a narrow gatekeeping role for a psychiatrist that proved unworkable [37]. The Oregon legislation has called for a more therapeutic role, but there is no evidence yet of whether this attempt will succeed [49].
Another clinical problem is the patient with treatment-resistant depression. In the setting of terminal illness, one might realistically ask when such a patient's suffering becomes too great. There is an emerging body of work that supports the efficacy of treatment of depression in the elderly and the medically ill [50], but the ‘hard case’ of treatment resistance may cause concern. The problem for a psychiatrist, who is sympathetic to the plight of such an anguished person and tempted to support their suicide, is the issue of a boundary violation. The framework of the doctor-patient relationship requires a safe space contained by a boundary, within which both can understand the patient's wishes, fears and experiences. This becomes an important source of orientation amid the complex emotions that can arise for the doctor [51]. Such an orienting ‘landmark’ or boundary is critical in the care of a dying patient [52]. The boundaries of the therapeutic relationship might easily be transgressed if one does not recognise the similarity between the acting out of erotic feeling and the suicidal urge [53]. The psychiatrist's therapeutic role is to be a container of anguished, despairing and hopeless emotions of a demoralised or depressed patient. As such, he or she waits for the window of opportunity to instil hope and encouragement back into such a person. Significant transference and countertransference phenomena exist in such a relationship, predicated on power, inequality and charging the psychiatrist with considerable professional responsibility. There are serious ethical issues for the psychiatrist who shifts position from encouraging hope to admitting hopelessness through PAS. In the latter, the patient may be precipitated into a seemingly quick resolution through suicide. Countertransference enactment and projective identification from a patient who feels worthless and despairing, or hateful and angry, are important elements of the doctor-patient relationship when a patient is suicidal [54]. If psychiatry as a discipline has an understanding of the psychology of such relationships, including conscious and unconscious dimensions and the impact of illness on each player, how can this be reconciled with a response that supports the narrow, legislated position of psychiatric gatekeeping for assisted suicide? What role is the psychiatrist cast in when, from the patient's perspective, he or she is not only the doctor from whom the patient seeks treatment, but also potentially the one who may be able to assist their suicide? The report from the New York State Task Force on Life and the Law [55] emphasised that medicalising assisted suicide fundamentally violates the values of medical practice.
Such complexity highlights the need for the psychiatrist to have considerable experience in management of the medically ill. As we move to an era of increased specialisation within psychiatry, we argue that the psychiatric care of patients dying from cancer or AIDS should be the domain of the C–L psychiatrist.
Research has confirmed high levels of psychiatric morbidity in the terminally ill. Structural barriers preventing access to care to address this morbidity must be removed [56]. Often, dying patients are cared for in settings remote from ready access to C–L psychiatrists and the stigma of psychiatric disorder may discourage general practitioners, families and patients from seeking psychiatric treatment [56].
The Dutch experience of euthanasia of the mentally ill
There have been two reported cases of psychiatrists assisting in the suicide of the mentally ill in the Netherlands [53]. In the city of Assen, Chabot, a psychiatrist, was reprimanded for failing to provide necessary treatment to his patient, a 50-year-old woman with major depression following the death of her second child from cancer. Her first son had previously suicided some years earlier, both children dying at the age of 20. Chabot met this woman through the Voluntary Euthanasia Society. Hendin, who interviewed Chabot extensively about his 30 h spent with this woman, concluded that the patient had pressured him into his decision to assist in her suicide and that his role took on that of identification with the aggressor [36]. She was a patient seeking her executioner and both she and Chabot experienced closeness in her death. Chabot appeared to have entered into her fantasy of death as a reunion with her boys. She chose to die in what had been her younger son's room and on his bed. She played the Bach flute sonata that was played at both her sons' funerals and kissed a photograph of her two boys following ingestion of her suicide-inducing capsules [36]. This case of a psychiatrist assisting in the suicide of a mentally ill patient in 1993 lead to the eventual promulgation of the Royal Dutch Medical Association's guidelines on the acceptability of termination of life in mentally incapacitated patients [1].
Subsequently, in 1997, a second case of a psychiatrist assisting in suicide was reported. This time it involved a 48-year-old woman who had had chronic anorexia nervosa and severe depression since age 17. She expressed her wish after both her husband and mother had died. There was no prosecution in this case [53]. Groenewoud et al. [3] estimated that 320 requests for PAS for mental illness were made annually to psychiatrists in the Netherlands, with between 2 and 5 patients being assisted each year. Depressive disorder, comorbid with a terminal somatic disease, was thought to be the most frequent psychiatric condition of those requesting PAS. Furthermore, in the more usual practice of euthanasia for terminal illness in the Netherlands, psychiatric consultation was only documented in 3%% of all requests [3].
Dutch guidelines for termination of life of the mentally ill
There are four explicit components to the Dutch guidelines for the termination of life of mentally incapacitated patients [1]. First, the patient's wish to die must be enduring, voluntary and well considered. Second, when there is a reasonable chance of recovery within a surveyable period of time, attempt at treatment is imperative. Third, suffering must be considered unacceptable to the patient and incurable from a prognostic perspective. Finally, an independent psychiatrist must personally evaluate and agree.
A range of problems emerges as one considers the application of these guidelines. Prognostic predictions prove repeatedly flawed in the cancer setting; how much more subjective and variable must they be in the psychiatric context? Whether a psychiatrist tends to be more psychotherapeutically or pharmacologically oriented may influence their interpretation of whether pessimistic and negative cognitions are fixed, as seen in patients with a demoralisation syndrome. What length of hospitalisation and period of intensive psychotherapy must be considered sufficient before the patient's wish is acted upon to compassionately relieve their suffering? How does psychiatry deal with the countertransference enactment inherent in such practice? What does this convey about the profession's view of the worth of a person with chronic mental illness? To what extent does the provision of assisted suicide represent the response of psychiatry to a challenge to its therapeutic efficacy, sense of professional esteem and level of negative regard for patients whose illness is long-standing? These Dutch guidelines have the potential to seriously change the nature of psychiatric practice. They must surely be considered a social experiment within a specific community in the world and, as such, should be carefully studied to clarify the benefits or perils of such behaviour.
The New York State Task Force [55] highlighted the risks of provision of assisted suicide for those whose autonomy and wellbeing have already been compromised by poverty, lack of access to good medical care, advanced age or membership of a stigmatised group. The responsibilities of psychiatrists as advocates for the mentally ill and/or physically disabled are clearly evident, and hard to reconcile with the provision of assisted suicide or tacit agreement in the presence of such legislation.
What of age, burden, dignity and dementia?
The practice of euthanasia and PAS is clearly related to age. Two-thirds of cases in the Netherlands were older than 60 years; cases in the age brackets of 70 + and 80 + increased across a recent decade [57]. Euthanasia undertaken for cancer and cerebro-vascular accident was significantly associated with increasing age, while euthanasia for AIDS and multiple sclerosis occurred in younger age groups.
Fear about being a burden to one's children and loss of dignity was evident within the Australian cases of euthanasia [37]. In palliative care, parents often express concern about being a burden to their children. This can be especially evident in the nuclear family where both partners work and struggle with commitments to their own, younger children, as well as their ill parent. The sense of duty held by offspring for an incapacitated parent, seen previously in the traditional family, is less evident in modern society, but the use of community volunteers can do much to enhance support and ease the burden of care. Open discussion of carer issues often facilitates expression by a family of their willingness to care for their parent.
Fear about loss of dignity is common in the medically ill and usually involves prior witness of another person living through illness perceived with disgust, revulsion or shame. Typically, such onlookers rate quality of life as less than the patient in reality does. This can be particularly hazardous when doctors assume they know what sort of life their patients want [34]. For example, when a healthy person witnesses incontinence, he or she commonly perceives greater loss of dignity than is necessarily experienced by a frail patient. Identification mechanisms underpin the development in onlookers of fear about loss of dignity, while they may carry guilt over feeling revolted by bodily decay. Considerable time can be spent helping patients and their families understand the process of dying, empowering them to prepare more adequately for this experience.
The ‘right to die’ notion has developed predominantly from concern about unwanted life-prolonging treatment, but has achieved a status of mantra for end-of-life decision-making. Reliance on and reverence for individual autonomy as the dominant and over-riding ethical principle in medical decision-making is limited in this context. The goal of autonomous choice may be unachievable and misleading given the realities of severe illness and disability, and the patient's need to regularly alter goals, level of independence and hopes for their future [58]. The patient's individual autonomy as the determining principle for medical decision-making is also limited in its application across cultures [59].
An understanding of the psychology of autonomy and dependency, and how these are altered in relationships involving chronically ill or dying patients is important in this debate. It has been argued that our community denies dependence on others, which is a reality across all stages of life [58]. Doctors can experience particular difficulty in coping with the increased dependency of chronically ill or dying patients, and experience frustration and resentment towards the patient as a response to their need for care [60]. These factors may contribute to the physician's perception of indignity in the dying.
Dementia is a feared clinical state because of the perception of loss of dignity. However, competence fluctuates with cognitive and emotional functioning and the rate of progression to cognitive disability is unpredictable in dementia [61]. The patient seeking pre-emptive suicide may be sacrificing many comfortable and relatively capable years. The American Geriatric Society's amicus curiae (‘friend of the court’) brief to the US Supreme Court 1996 hearing about a constitutional right to PAS argued that the elderly are often frail, disabled and economically disadvantaged, thus being at risk of abuse in a society that supports euthanasia [61]. Dying persons can be physically weak, prone to strong emotion and vulnerable to suggestion, expectation and guidance. The image of an independent, capable and voluntary person is an ideal not easily found in the context of the usual care of the dying person. The potential for undue influence or coercion is worrisome. Furthermore, voluntariness is problematic in the face of untreated symptoms and financial pressures. A number of uncertainties exist about the standard of medical care delivered to the elderly and the dying, their prognoses, assessments of competence and depression, which make legislative constraints around PAS problematic. Moreover, health care at the end of life is undergoing a major change; pain and other symptoms can be relieved. Most people can now die quietly [62]. However, the standard of care lags behind what is possible to achieve (see The SUPPORT study [63]). We are challenged to work towards substantial improvement in care provision and resourcing.
What role for our profession?
It is vitally important that psychiatrists are well informed, as there is little likelihood that clinicians can bypass the ethical issues involved. Psychiatrists for a long time have been the custodians of the suicidal. The battle to prevent suicide continues on many fronts. Further empirical studies are needed to clarify the experience of demoralisation and depression in the medically ill. The C–L psychiatrist has a particular role in the treatment of these patients. We need research and education concerning psychiatric morbidity in the terminally ill and its treatment, and particularly about the needs of doctors, nurses and other health professionals in providing care for dying patients [56].
Psychiatry can provide medicine with perspectives on psychosocial development, the psychology of doctor-patient relationships, clinical issues in the care of the dying and knowledge of the complex determinants of suicidality; yet such perspectives are curiously put aside in this debate, as if they cannot be applied in understanding the issues at stake. The difficulties doctors experience in caring for dying patients, comorbid psychiatric illness, the clinician's emotional distress, deficiencies in psychosocial care, and problems of access to appropriate care, including adequate analgesia and specialist palliative care, are all features of the clinical context in which the issue of assisted suicide emerges. Such problems can be used to better understand what underlies the wish for assisted suicide. The euthanasia debate has highlighted the difficulties that medicine, including psychiatry, has in dealing with death and suffering [64],[65].
Our profession has a responsibility to inform the broader community of the complexity of these issues, so that debate is less emotive and based upon stronger empirical evidence. Psychiatrists need to take up advocacy for the vulnerable, disabled and infirm. The Dutch guidelines for assisted suicide of the mentally ill cause grave concern. It is surprising that they have occurred in a Europe still recovering from the effects of the German euthanasia program (1900–45) which targeted the mentally ill, the physically disabled and those stigmatised as ‘unworthy of life’ [66]. Nonetheless, the new guidelines represent a social experiment being conducted in an increasingly liberalised society that promotes the autonomy of the individual. A balance must exist between individual rights and the safety of the broader community, including a state interest in preserving life. While social experiments such as the recently issued Dutch guidelines do occur as an expression of cultural change, not all social experiments prove worthy of continuation. In our view, these Dutch guidelines are so dangerous that they will eventually need to be rescinded.