In Word, or Sigh, or Tear: Depression and End-of-Life Choices

Chapter 15
In Word, or Sigh, or Tear:1 Depression and End-of-Life Choices²

Colin Gavaghan

Introduction

Although debates about the (legal and moral) status of voluntary euthanasia (VE) and assisted suicide (AS) seem as intractable as ever, it may be that the terrain on which they are played out has recently shifted. While earlier exchanges were often characterised by ‘first order’ arguments between the relative importance of (typically) autonomy and the relief of suffering, on the one hand, and the value of life or importance of not killing, on the other, much of the recent discourse has been characterised by what may be seen as ‘second order’ concerns. That is to say, both sides agree about which ethical principles matter, but disagree about how these principles are promoted, or contravened, by the practices in question. Opponents of reform, then, may agree with supporters that autonomy and beneficence are important ethical values, while disagreeing that they will be promoted by allowing the practices of voluntary euthanasia and assisted dying.

Often, the points of divergence concern putatively ‘vulnerable’ people. Thus, again, while we might agree that autonomy is important, we might disagree that people in this particular category are capable of acting autonomously, or at least fear that their autonomy will be seriously compromised by their circumstances. Equally, the concern might be that their best interests and future happiness will not be best served by allowing them to act on their current preferences.

This shift of focus can be seen in the context of the recently-revived New Zealand debate on assisted dying, where John Kleinsman, Director of the Nathaniel Centre – the New Zealand Catholic Bioethics Centre – was at pains to point out that: ‘I am, in all honesty, not interested in imposing my religious views on anyone’; but rather, his argument against assisted dying ‘… centres on safety and protection of those who are vulnerable’.³ In a recent polemic against proposed law reform in Scotland, journalist Kevin McKenna has issued a similar warning: ‘Many people who take their own lives do so because of depression. Society has not yet begun to consider urging those suffering from depression to kill themselves and have done with it’.⁴

In this chapter, I intend to focus on this last category of people: those whose autonomy and safety are often thought to be threatened by legalised VE/AS because they have been diagnosed with, or at least suspected of being affected by, a depressive disorder. Although there has been considerable recent disagreement about the precise diagnostic boundaries of such conditions, there is substantially greater consensus that they present a particular concern for end-of-life decision-making.

At the same time, controversy exists around the role of depression as an excluding condition with regard to competent choices. The presence, actual or suspected, of such disorders is often thought to invalidate, or at least cast doubt upon, competence to make autonomous choices. Even the generally pro-reform Falconer Commission did ‘… not consider that a person with depression, whose judgement might be significantly impaired as a result of this depression, should be permitted to take such a momentous decision as ending their own life’.5

This chapter explores the relationship between depressive disorder and end-of-life decisions in both of these contexts. Drawing on and developing themes touched on by Sheila McLean in 2007,⁶ it also examines this issue against the backdrop of what some commentators see as an expanding definition of ‘depression’ that threatens to pathologise a range of thoughts and wishes common among terminally or chronically ill people. While it is beyond my remit (and probably my expertise) to take an informed position on the desirability of this ‘diagnostic inflation’, it seems important to consider its likely implications for decisions as to legal capacity to consent to, and eligibility to be provided with, VE or AS.

A (Brief) History of Depression

Though use of the term ‘depression’ in this context is relatively recent, its roots can – with some reservations⁷ – be traced to early writings about ‘melancholia’. ‘If fear and sadness continue long’, noted Hippocrates in his Aphorisms, ‘the person is melancholic’.⁸ The term derives from the belief of ancient Greek physicians that the condition was caused by an imbalance of bodily fluids, or humours: the term derives from the Greek words melas (black) and kholé (bile).

A slight detour in the Middle Ages saw the concept become entangled with the Christian Church’s notion of accidie, and become associated with the sin of sloth.⁹ Otherwise, two aspects of the Hippocratic approach persisted until the Renaissance: the attribution of melancholia to humoral imbalance,¹⁰ and the ‘clustering of traits around the poles of sadness and fear’.11 References to ‘depression’ in this context can be traced to the seventeenth and eighteenth centuries; Samuel Johnson and George Eliot certainly used it in much the way we do today.¹²

During the nineteenth century, the term ‘depression’ came increasingly into use, though often to describe a symptom rather than a condition; ‘the basic diagnostic term was still usually melancholia or melancholy’.¹³ By the beginning of the twentieth century, however, the terms ‘depression’ and ‘melancholia’ were coming to be used more or less synonymously,¹⁴ with the former term gaining increasing favour among medical professionals.¹⁵ Emil Kraepelin’s work was highly influential in the formative years of ‘scientific’ psychiatry, and this was no less true regarding his approach to depression. An important facet of Kraepelin’s approach to mental illness was a distinction between ‘endogenous’ and ‘exogenous’ disorders¹⁶ – that is, those with an identifiable external cause, and those that, with no such identifiable cause, were deemed to originate internally.¹⁷

This binary approach became the dominant one for the early part of the twentieth century, where exogenous aetiology developed into the concept of ‘reactivity’,¹⁸ but it came to be challenged in more recent times by those who saw both ‘exogenous/melancholic’ and ‘endogenous/neurotic’ depression as lying along a continuum.¹⁹ This, in turn, led to a major shift in how depression was perceived and diagnosed; the dominant approach, as explained below, now defines it by reference to symptoms rather than causes. This move was controversial at the time and, as I discuss later in this chapter, remains so today.

It is not clear whether depression or melancholia were always regarded as threats to autonomy. According to Charles F. McKhann:

[T]he concept has evolved throughout the nineteenth century and into the twentieth that suicide is a manifestation of mental disturbance and that the individual is not responsible for his or her action. … This principle grew out of the observation that many suicides occurred in patients who had severe manic depressive (biphasic) disorders, those who were under severe emotional stress, and those who had suffered psychological trauma.²⁰

The standard definitions today are to be found in the Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-V), and the World Health Organization’s International Classification of Diseases (ICD-10). These differ in certain respects, but common to both are what we might call the ‘core’ symptoms of lowered or ‘depressed’ mood,21 and diminished capacity for enjoyment or pleasure. Other symptoms are commonly, but not invariably, associated with the condition, such as tiredness or reduced energy, disturbed sleep patterns, loss of appetite, weight loss, reduction in libido, feelings of worthlessness or low self-esteem, and ideas of ‘excessive’ guilt.

ICD-10 subdivides depressive episodes into ‘mild’, ‘moderate’ and ‘severe’.²² A ‘mild’ episode is one in which ‘two or three’ of the generic symptoms are present, with four or more indicating a ‘moderate’ episode. A ‘severe’ episode is one in which:

… several of the above symptoms are marked and distressing, typically loss of self-esteem and ideas of worthlessness or guilt. Of particular relevance for present purposes, suicidal thoughts and acts are common and a number of ‘somatic’ symptoms are usually present.²³

DSM-V requires that at least five of the typical symptoms be present, including at least one of the core symptoms.²⁴ Depressive episodes may or may not be accompanied by psychotic symptoms, such as ‘… hallucinations, delusions, psychomotor retardation, or stupor so severe that ordinary social activities are impossible’.²⁵

Definitional Controversy

Debates around the ‘pathologisation’ of normal human experience are not new. For over half a century, Thomas Szasz, Ivan Illich and their successors have railed against ‘the medicalization of life’²⁶ and the ‘tyranny of pharmacracy’.²⁷ The practice of psychiatry and the ‘myth of mental illness’²⁸ has been a particular focus for these impassioned polemics. While the echoes of the general ‘anti-psychiatry’ critique can still be heard from time to time, most of today’s critiques are more focused on particular examples of putative medical, psychiatric and pharmacological overreach. Many psychiatric conditions are argued to be dramatically over-diagnosed and over-medicated – either at the behest of a greedy pharmaceutical industry, or by a state bent on greater social control29 – if not invented outright for the purposes of manufacturing demand for ‘treatments’.³⁰

‘Clinical depression’ has not been insulated from these sorts of controversies. 2007 saw the publication of two prominent critiques of perceived over-diagnosis. In the British Medical Journal, professor of psychiatry Gordon Parker argued that: ‘A low threshold for diagnosing clinical depression, however, risks treating normal emotional states as illness, challenging the model’s credibility and risking inappropriate management’.³¹

In similar terms, Allan Horwitz and Jerome Wakefield argued:

… that the recent explosion of putative depressive disorder, in fact, does not stem primarily from a real rise in this condition. Instead, it is largely a product of conflating the two conceptually distinct categories of normal sadness and depressive disorder and thus classifying many instances of normal sadness as mental disorders.³²

Such critiques gained momentum and publicity when the American Psychiatric Association (APA) published its draft reforms of the DSM, reforms which soon attracted allegations of diagnostic inflation. In particular, the proposal to extend a diagnosis of major depression to situations of bereavement elicited strong criticism. The British Psychological Society had this to say:

[W]e are concerned that clients and the general public are negatively affected by the continued and continuous medicalisation of their natural and normal responses to their experiences; responses which undoubtedly have distressing consequences which demand helping responses, but which do not reflect illnesses so much as normal individual variation.³³

Allen J. Francis, a psychiatrist and Chair of the committee which produced the previous edition of the DSM, spoke out in even starker terms, claiming: ‘Making grief a mental disorder will be a bonanza for drug companies, but a disaster for grievers. The decision is also self destructive for DSM 5 and further undermines the credibility of APA. Psychiatry should not be mislabeling the normal’.³⁴

Although it is not my intention in this chapter to take sides in the debate over putative diagnostic inflation, this discussion of the controversy around it is relevant for my purposes. For one thing, it illustrates that ‘clinical depression’ is a contested diagnosis, with many fuzzy borders. If a diagnosis of ‘depression’ is to serve either as a qualifying or an excluding criterion for assisted dying, then eligibility for such assistance will often depend on where the psychiatric profession chooses – and it is, unquestionably, a choice – to set those borders.

Whether depression will satisfy the qualifying criteria will depend on the jurisdiction in question. In jurisdictions such as the Netherlands which require only ‘lasting and unbearable’ suffering,35 the question has arisen as to whether this suffering must be somatic³⁶ in origin, or whether psychiatric suffering³⁷ could suffice. In 1994, the Dutch Supreme Court decided that non-somatic suffering could form the basis for a defence for performing euthanasia. In the widely discussed Chabot decision, it held:

A claim of necessity can … not be excluded simply on the ground that the patient’s unbearable suffering, without prospect of improvement, does not have a somatic cause and that the patient is not in the terminal phase. … the cause of the suffering does not detract from the extent to which suffering is experienced.³⁸

However, when the suffering was non-somatic in origin, additional safeguards were put in place, specifically the requirements:

• that the applicant must have been examined by an independent colleague; and

• that the applicant cannot have refused any ‘realistic’ therapeutic alternative.

Whether these additional safeguards are sufficient,39 or indeed necessary,⁴⁰ has been a ground of some contention. In the wake of the judgment, it was also widely predicted that it was only a matter of time before the courts were called upon to decide whether severe unhappiness that did not satisfy the criteria for ‘clinical depression’ would qualify.⁴¹

In 2002, the Supreme Court was faced with precisely this question. Unlike the patient in Chabot’s case, GP Philip Sutorius had not diagnosed his patient as suffering from depression or any other ‘medically classified disease or disorder’. Rather, Edward Brongersma was affected by loneliness, ‘feelings of senselessness, physical deterioration, and a long-standing wish to die not associated with depression’.⁴² For the Dutch Supreme Court, this was a step too far. The defence of necessity was rejected; merely being ‘tired of life’ is not sufficient for the purposes of Dutch law.⁴³

The law in Switzerland has recently become the focus of considerable judicial attention with regard to non-somatic suffering and the putative ‘right to die’, with two cases reaching the European Court of Human Rights.⁴⁴ In theory, Swiss law is amongst the more liberal with regard to assisted dying, in that inciting or assisting suicide are only offences when carried out ‘for selfish motives’.⁴⁵

In practice, the most humane and effective means of suicide are considerably more tightly regulated. Under Swiss federal medicine laws, the lethal substances most suitable for these purposes are available only in accordance with a medical prescription. And both case law⁴⁶ and the professional medical guidelines by which Swiss doctors are bound allow them only to issue such prescriptions where:

(1) the patient’s illness makes it likely that death is close, (2) other options for support have been discussed and, where appropriate, put into place, and (3) the patient is capable of discernment, his or her wish appears to be carefully considered, is not the result of external pressure and is to be regarded as final, which must be verified by an independent third party who need not necessarily be a doctor. (Federal Court, translation from ECtHR, para. 16)

It appeared, then, that it would only be those near the end of life who would be able to avail themselves of physician-assisted suicide (PAS); and no other route is readily available for those who want to access lethal drugs. The reality, as Isra Black has explained, may have been somewhat different:

… empirical evidence from the Swiss right-to-die organisations before and after the Zurich Case has revealed a relatively low yet significant prevalence of PAS among individuals with non-fatal conditions, without corresponding criminal and/or professional proceedings.47

The applicant in the Haas case⁴⁸ had suffered from ‘… a serious bipolar affective disorder for about 20 years. During this period he has twice attempted suicide and has stayed in psychiatric hospitals on several occasions’.⁴⁹ Starting in 2004, he attempted to obtain assistance in ending his life, and approached several psychiatrists with a view to receiving a prescription of a lethal substance, specifically sodium pentobarbital.⁵⁰ When none of the psychiatrists was willing to provide such a prescription, Haas turned first to the Swiss courts, claiming that his right to control the timing and manner of his death (under Article 8 of the European Convention on Human Rights) was being infringed by this requirement.

Having been unsuccessful at first instance, Mr Haas appealed in 2006 to the Federal Court. The appeal was rejected. It is interesting to consider in detail the Court’s ruling. First, it held that – contrary to what the earlier case law seemed to require – mental illness and psychological suffering could provide a legal basis for assistance with dying:

It must not be forgotten that a serious, incurable and chronic mental illness may, in the same way as a somatic illness, cause suffering such that, over time, the patient concludes that his or her life is no longer worth living. The most recent ethical, legal and medical opinions indicate that in such cases also the prescription of sodium pentobarbital is not necessarily precluded or to be excluded on the ground that it would represent a breach of the doctor’s duty of care ….⁵¹

Second, it acknowledged that ‘the greatest restraint must be exercised’ when providing assistance in such cases. In particular:

[I]t is necessary to distinguish between a desire to die as the expression of a psychological disorder which can and must be treated, and a wish to die that is based on the considered and sustained decision of a person capable of discernment (‘pre-suicide assessment’), which must be respected as applicable.⁵²

The key consideration was whether ‘the wish to die is based on an autonomous and all-embracing decision’.53 Since the ‘question of whether the conditions have been met in a given case cannot be examined without recourse to specialised medical – and particularly psychiatric – knowledge’, ‘a thorough psychiatric examination thus becomes necessary’.⁵⁴

Although this ruling seemed to have widened the criteria for assisted suicide (in this case, with the help of a physician), it did not satisfy Mr Haas, who regarded the requirement to undergo medical examination as both an interference with his Art 8 right, and a practically insurmountable obstacle to achieving assistance,⁵⁵ and he appealed this decision to the European Court of Human Rights. The Court, while expressing sympathy for Haas’s ‘wish to commit suicide in a safe and dignified manner and without unnecessary pain and suffering’,⁵⁶ found in favour of the Swiss Government. In particular, it held:

[T]he regulations put in place by the Swiss authorities, namely the requirement to obtain a medical prescription, pursue, inter alia, the legitimate aims of protecting everybody from hasty decisions and preventing abuse, and, in particular, ensuring that a patient lacking discernment does not obtain a lethal dose of sodium pentobarbital.⁵⁷

The more recent Gross decision⁵⁸ concerned a situation that was, on the face of it, very similar to that in Haas. Aldo Gross had also attempted to take her own life, and had also received psychiatric in-patient care in the aftermath. Like Mr Haas, she sought a quick and humane means to end her own life, and approached various psychiatrists seeking a prescription for sodium pentobarbital. Gross’s case differed in that she was not suffering from any recognised illness at all, but only from a persistent desire to die. Unsurprisingly, she had been unable to persuade the physicians she consulted to issue her with a prescription for lethal drugs. Although satisfied as to her decision-making capacity, they regarded themselves as constrained by the fact that she was not suffering from any illness.

As with Mr Haas, Ms Gross was unsuccessful before the domestic courts, and ultimately brought her case to the European Court of Human Rights. By a narrow majority, the Court held against the Swiss Government, on the grounds that: ‘Swiss law, while providing the possibility of obtaining a lethal dose of sodium pentobarbital on medical prescription, does not provide sufficient guidelines ensuring clarity as to the extent of this right’.⁵⁹ In a ruling somewhat redolent of that of the House of Lords in Debbie Purdy’s case,⁶⁰ the European Court of Human Rights required the Swiss Government:

… to issue comprehensive and clear guidelines on whether and under which circumstances an individual in the applicant’s situation – that is, someone not suffering from a terminal illness – should be granted the ability to acquire a lethal dose of medication allowing them to end their life.⁶¹

In doing so, however, the Court was careful to stress that the substantive content of those guidelines was largely a matter for the Swiss Government to decide.62 As Isra Black has observed, this could see Ms Gross’s ‘victory’ transpiring to be a pyrrhic one, because:

… the Government might respond with clear guidance that PAS for non-fatal conditions (with the exception of mental disorders meeting the Haas conditions) or, more narrowly, for existential suffering, is not responsible medical practice.⁶³

The ‘fuzzy edges’ that previously seem to have allowed Swiss physicians to ‘provide … PAS [physician assisted suicide] to elderly individuals with apparent existential suffering, ostensibly without censure’ will now have to be made clearer. It is at least possible that this will be done in such a way as to render the situation more, rather than less, restrictive for those whose wish for PAS is borne of existential suffering.

The two jurisdictions best known for allowing euthanasia or assisted suicide, then, either already have, or are probably about to adopt, positions that accept ‘clinical’ depression as a qualifying criterion, but draw the line at ‘mere’ existential suffering. This attempt to draw clear distinctions between ‘psychiatric’ and ‘existential’ non-somatic suffering has elicited critical responses, from at least two directions. For opponents of euthanasia, its provision to ‘clinically depressed’ people is problematic. Herbert Hendin described the Chabot case as ‘especially troubling, as it marked Dutch acceptance of depressed suicidal patients as eligible for assisted suicide or euthanasia. Depressed patients’, he warned, ‘with or without medical conditions who did not respond quickly to treatment could be vulnerable’.⁶⁴

Opponents have also expressed fears that accepting non-somatic suffering as a ground for euthanasia will lead invariably to a situation wherein euthanasia will be provided on demand to anyone who is merely unhappy.⁶⁵ The requirement that the applicant be ‘medically ill’, we might fear, will be readily circumvented, and indeed may provide an incentive for sympathetic medics to provide pseudo-diagnoses. As well as circumventing the spirit of the law, this might risk contributing to the ‘diagnostic inflation’ of mental illness.

For some supporters of AS/VE, on the other hand, the distinction is at best arbitrary, and at worst irrational and inhumane. Regardless of which of the two most frequently advanced arguments for legalisation – autonomy or beneficence – is followed, nothing of ethical significance hangs on the presence of absence of a recognised ‘medical illness’. Thus, Griffiths and Weyers have argued that: ‘By trying to draw a line between suffering that is medically classifiable and other forms of suffering the court seems to choose the route of repression. It seems doubtful whether the demarcation will survive in practice’.⁶⁶

Although in profound disagreement as to how the situation should be improved, both sides seem to be in agreement in doubting that the distinction between ‘clinical’ depression and ‘mere’ unhappiness is particularly clear or likely to withstand challenge. One of the more overt exponents of this view is Julian Savulescu, who has argued that Edward Brongersma ‘… was in fact suffering from the commonest terminal disease: ageing. … We do not consider it to be a disease because of the way we have arbitrarily defined the concept of disease’.67

As I discussed above, the view that the boundaries of disease – and in particular, depressive disease – have been defined arbitrarily is by no means an uncommon one. Even among those who do not consider the definition wholly arbitrary, it is widely recognised that the boundary between ‘clinical’ depression and ‘mere’ unhappiness can be somewhat fuzzy. If there is merit in this critique, then it seems that an approach to euthanasia that accepts ‘depressive illness’ as a qualifying condition, but not non-pathological ‘existential suffering’, is likely to be problematic.

Furthermore, for those concerned about the capacity of euthanasia applicants to make autonomous decisions, it may be seen as somewhat anomalous that those diagnosed with psychiatric illness will be permitted to choose to die with assistance, while those with no such diagnosis will not.

In 2010, Dutch media reported on the formation of an organisation – ‘Out of Free Will’ – dedicated to campaigning for the right to assisted suicide for ‘All Dutch people over 70 years of age who feel tired of life’. Although widely depicted as a radical initiative to expand the availability of euthanasia, it is interesting to note that the group seeks to limit such a right to situations where: ‘The suicide assistants must make sure the death wish is more than a rash impulse, the product of depression, or the symptom of another illness, and that the patient has considered the consequences of his actions for those who will survive him’.⁶⁸

If the presence of depression is at least to serve as a red flag for problems with mental capacity, then the ‘Out of Free Will’ proposal may be seen as having certain advantages over the current position of Dutch law. Can we, though, assume that such a red flag response would be justified? As I discuss in the next section, the precise relationship between depressive illness and decision-making capacity is not straightforward.

Depression as an Excluding Criterion

Concerns about the influence of depression on decision-making capacity (henceforth simply ‘capacity’) lie at the heart of the second role depression can play in discussions and proposals about euthanasia. The moderate form of the argument here is that any proposals to allow euthanasia must be drafted so as to protect depressed (or at least seriously depressed) people from harmful or non-autonomous choices. In practice, this may mean excluding any depressed people from being able to avail themselves of VE/AS.

The extreme form of the argument is that the threat to depressed people is of such a nature and magnitude that it counts as a strong reason against allowing VE/AS at all. What both the moderate and extreme versions rely on, then, is an assumption that allowing (at least some) depressed people to access euthanasia would be a bad outcome.

The moderate version can be seen in the report of the generally pro-reform Falconer Commission:

Although the distinction between ‘appropriate sadness’ and depression in the context of terminal illness is complex, the Commission does not consider that a person with depression, whose judgement might be significantly impaired as a result of this depression, should be permitted to take such a momentous decision as ending their own life.69

The phrasing of this finding is somewhat ambiguous as to whether the phrase ‘whose judgement might be significantly impaired’ is a restrictive or non-restrictive clause. That is, is it being assumed that people with depression have suspect judgment merely by virtue of being depressed, or is the report alluding to a subset of depressed people of whom this might be said, and towards whom more care should be shown?

In either event, the draft legislation to which the report gave rise spoke only of ‘recognising and taking account of the effects of depression or other psychological disorders that may impair a person’s decision-making’.⁷⁰ This suggests an intention that a diagnosis of depression will function as a warning flag rather than a Stop sign; as an indicator of a need for particular scrutiny of an applicant’s decision-making capacity, rather than an assumption that it is lacking. Whether it would be interpreted in that way in practice is, of course, another question.

Another more recent legislative proposal – Margo MacDonald’s Assisted Suicide (Scotland) Bill,⁷¹ introduced in November 2013 – appeared to take a harder line on the issue. In the section headed Capacity, the Bill laid out standard criteria for decision-making capacity (which are discussed in the next section), but supplemented this with a requirement that the person ‘is not suffering from any mental disorder (within the meaning of section 328 of the Mental Health (Care and Treatment) (Scotland) Act 2003 …) which might affect the making of the request’.⁷²