Article Text

Download PDFPDF

Advance decisions in dementia: when the past conflicts with the present
  1. George Gillett
  1. Correspondence to George Gillett, Medical Sciences Division, University of Oxford, Oxford OX1 2JD, UK; george.gillett{at}live.co.uk

Abstract

As the prevalence of dementia increases across the Western world, there is a growing interest in advance care planning, by which patients may make decisions on behalf of their future selves. Under which ethical principles is this practice justified? I assess the justification for advance care planning put forward by the philosopher Ronald Dworkin, which he rationalises through an integrity-based conception of autonomy. I suggest his judgement is misguided by arguing in favour of two claims. First, that patients with dementia qualify for some right to contemporary autonomy conceptualised under the ‘sense of liberty’ it provides. Second, that respecting precedent autonomy, such as an advance care plan, is not essential to Dworkin’s integrity-based account of autonomy. Together, my claims problematise the practice of using advance decisions in the context of dementia.

  • autonomy
  • dementia
  • elderly and terminally ill
  • living wills/advance directives
  • psychiatry

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

In recent years, Western countries have experienced a proliferation of the number of patients living with neurodegenerative disease. In 2013, 815 827 people in the UK had dementia, with this figure expected to rise to 1 142 677 by 2025.1 This trend makes pertinent the ethical discussions around advance care planning in dementia.

Advance decisions (or living wills) are statements about how one wishes to be treated after they have lost legal decision-making capacity. The Mental Capacity Act introduced living wills into law in England and Wales, where clinicians are obliged to honour them, given a number of criteria are fulfilled.2

The legitimacy of advance decisions might be questioned in cases where the apparent wishes of a patient living with dementia conflict with wishes outlined in a prior advance decision. Oft-cited examples include that of a lifelong vegetarian patient who begins to crave meat or a monogamous happily married heterosexual patient who exhibits same-sex behaviour.3–5

Although patients with dementia often retain legal capacity to amend advance decisions some time after their diagnosis, the majority will eventually lose this capacity.6 7 Legal capacity in such circumstances is commonly defined as the ability to understand, retain and weigh up the information required to make a decision and to communicate that decision.8 It is patients who have lost this capacity that this paper refers to. I will assess the legitimacy of advance care planning and argue that in cases where the apparent wishes of a patient living dementia conflict with a prior advance decision, contemporary wishes should be respected over a conflicting advance care plan in order to safeguard a patient’s sense of liberty. Although this novel argument reaches similar conclusions to an argument proposed by Agnieszka Jaworska,9 my paper challenges Dworkin’s approach for different reasons. Unlike Jaworska, I do not rely on a conceptualisation of autonomy hinged on a capacity to value which remains contentious and fails to accommodate many patients with dementia.10

This paper will first outline a relevant case study before discussing the mainstay of ethical debate. I will then challenge a popular consensus in a series of claims that question the validity of advance decisions in dementia.

Case study

Sally is a 75-year-old retired teacher. She is married with two daughters and describes herself as a ‘family person’. Shortly after her 75th birthday, she receives a diagnosis of Alzheimer’s disease. In a consultation with her general practitioner, Sally recalls caring for her father when he had dementia. She remembers it being traumatic and discloses a strong wish to ‘not become a burden’ to her own family. In her words, she ‘doesn’t want them to see me in a bad way’. Consequently, Sally writes an advance decision asking to be moved into a care home once her dementia progresses.

Sally’s condition deteriorates quicker than expected. As her dementia worsens, she appears increasingly confused and struggles to have basic conversations. Like most dementia patients, she lacks the legal capacity to amend her advance decision. However, she seems most at ease and happiest in her own home, becoming aggressive and agitated when encouraged to leave the house, even for a short walk. It appears Sally’s wishes have changed. She is now reluctant to leave her house and no longer has inhibitions about being dependent on her family. They are willing to care for her, and there is no clinical reason necessitating admission to a care home. Should Sally’s advance decision be honoured, or her apparent contemporary wishes?

Ethical consensus

A standard ethical justification for advance decisions comes from Ronald Dworkin.11 He addresses the issue by asking why we value autonomy and suggests there are two possible interpretations: the ‘evidentiary view’ and the ‘integrity view’. He argues in favour of the ‘integrity view’ of autonomy. Both interpretations are outlined below.

First is what Dworkin describes as the ‘evidentiary view’, which states that the value of autonomy lies in allowing patients to make decisions in their own best interests. According to this view, autonomy is important ‘because as a general matter each person knows what is in his best interests better than anyone else’.12 As an account of autonomy, Dworkin dismisses this view as inadequate, making the epistemological argument that there are often cases where health professionals are better informed about disease processes or treatment options than their patients. The fact that we still respect autonomy in cases where patients, knowingly or otherwise, make decisions contrary to their best interests suggests that our understanding of the value of autonomy must entail something more than outlined in this ‘evidentiary view’.

Instead, Dworkin supports an ‘integrity view’ of autonomy. He argues that ‘the value of autonomy lies in the scheme of responsibility it creates: autonomy makes each of us responsible for shaping his own life according to some coherent and distinctive sense of character, conviction, and interest’.11 Under this view, autonomy is not important in allowing patients to achieve their best interests but in ‘allowing us to lead our own lives, so that each of us can be what he has made himself’.11

How we judge a patient’s capacity to make valid decisions depends on how we conceive the value of autonomy. Under his integrity-based approach to autonomy, Dworkin argues that capacity for autonomy should depend on a patient’s ability ‘to evaluate decisions in the structured context of an overall life organised around a coherent conception of character’.11 The nature of dementia makes this difficult: patients with advanced dementia often lack the cognitive capabilities necessary to evaluate decisions in such a way.7 Therefore, Dworkin argues that patients with advanced dementia lack a contemporary right to autonomy. In our case study, Sally lacks capacity as understood by the integrity-based approach; she is unable to follow basic conversations and unable ‘to evaluate decisions in the structured context of an overall life’ as Dworkin suggests.

Given this, we are left with two options: to administer care in line with a best interests judgement justified by beneficence, or to follow a patient’s advance decision justified by a right to precedent autonomy: the right to make decisions about one’s future. Dworkin’s response provides the mainstay of ethical justification for advance care planning. He argues that respecting advance decisions, or as he terms these wishes, ‘prudential concerns’, is essential to an integrity-based approach to autonomy. If we are to value someone’s ability to determine the course of their own life, he argues, then we must honour their prior wishes when they lack capacity for contemporary autonomy. Applied to our case study, Dworkin’s approach suggests that clinicians honour Sally’s advance decision over her contemporary wishes.

My argument

Since Dworkin’s argument is often viewed as the standard account, it is his that I will primarily focus on. I will make the following claims: (A) that Dworkin is too eager to dismiss alternative understandings of the value of autonomy and that patients with dementia qualify for some right to contemporary autonomy and (B) precedent autonomy is not essential to respecting integrity-based autonomy. Given this, I conclude there is no reason why dementia patients should be deprived of contemporary wishes in favour of a previous wish.i

Claim A: patients with advanced dementia qualify for a right to contemporary autonomy

Dworkin’s assessment of the value of autonomy using the evidentiary-based and integrity-based views is appealing. I agree that patients with advanced dementia lack capacity for autonomy on an integrity-based account. His broader argument, however, that dementia patients fail to qualify for any capacity to autonomy relies on an unjustified conceptual assumption: that there must be a single, universal reason to value autonomy which should be generalisable to all patients and circumstances.

Instead, a more appropriate approach would conceptualise autonomy as a principle that patients may derive value from for different reasons. This would be in keeping with our understanding of other moral principles, such as beneficence. I may value my well-being for purely hedonistic reasons, while my neighbour may value his well-being because it equips him to better contribute to society. Importantly, the pluralistic reasoning of why we value well-being does not detract from the moral status of beneficence. Rather, what is important is that each justification for valuing beneficence is valid in the context in which it is applied. The same is true when understanding the value of autonomy; we need not find a single reason to justify respecting the autonomy of every patient, but rather find a coherent and valid reason for the patient’s circumstances we are debating.

Why is this important? Dworkin dismisses the evidentiary view using examples where we value patients’ decisions even when their choices are not in their best interests. He claims this may arise from ‘weakness of the will’, such as when smoking a cigarette, or from deeply held beliefs, such as when refusing life-prolonging treatment because of religious views. Dworkin is right to conclude that the evidentiary account is insufficient to explain why we value autonomy in these cases. However, two conclusions can be drawn from his case studies. Either (1) the evidentiary approach might sometimes be a justification for autonomy, but in these circumstances it is not valid, or (2) that these case studies are evidence that the evidentiary view cannot be used as a justification for respecting autonomy in any circumstance. Dworkin reaches the second, bolder conclusion without justification.

Imagine we instead follow argument 1. Where would this leave our understanding of the value of autonomy? Unlike Dworkin, I suggest that both the integrity and evidentiary accounts can sometimes provide reason to value a patient’s autonomy. Imagine a patient with cancer who is faced with a choice between surgery and chemotherapy. Although a doctor may advise which treatment is most efficacious, it is often the patient who is best placed to judge how each treatment will affect their life and well-being, and consequently which treatment aligns with their best interests. The evidentiary account seems like a valid justification for respecting a patient’s autonomy in such circumstances.

Nonetheless, it is unlikely that a patient with advanced dementia would retain capacity for autonomy under an evidentiary view. However, given that it is possible to hold multiple justifications for respecting autonomy, I wish to articulate a third reason a patient’s autonomy might be respected, conceptualised around protecting a ‘sense of liberty’.

This ‘sense of liberty’ should be understood broadly as the awareness of freedom that acting out one’s will offers, which I argue is integral to a settled experience of personhood and our environment. Conceptualising a value of autonomy around this justification neither suggests that patients with dementia are expert judges of their best interests (the evidentiary view) nor that their autonomy is important in building and acting out a settled sense of character (the integrity view). I outline these distinctions in a thought experiment.

Imagine a machine is created that can make decisions on our behalf. The machine has been proven to deliver our best interests with 100% accuracy. However, it is not advisory; one must allow the machine to become a surrogate decision maker while retaining awareness of this externally controlled volition. Similar to Nozick’s13 Experience Machine, I suspect that most people would reject such an offer, preferring instead to safeguard their autonomy. Moreover, I believe this to be true even if the machine allowed us to first dictate a sense of character, values and beliefs to customise the machine’s judgements. This suggests that we would not solely reject the machine for reasons understood by an integrity account of autonomy. Nor would we reject the machine for reasons understood by the evidentiary account; the machine promises to deliver our best interests. Instead, I believe most people would reject the machine on the basis of it limiting our sense of liberty, which is intimately connected to our immediate experience of personhood.

Importantly, my ‘sense of liberty’ argument is not a justification for respecting patients’ wishes in order to maximise their well-being but is in itself an explanation of the value of autonomy. If my argument were merely a variant of beneficence, we would find little reason to reject the machine in the thought experiment; it offers consistent judgements as to what actions are in our best interests, just as doctors might offer a patient with dementia.

Therefore, my ‘sense of liberty’ argument is not about respecting someone’s right to exist in a state of well-being but to exist as a person at all. Under this conception of autonomy, capacity for autonomy would not depend on whether patients can ‘evaluate decisions in the structured context of an overall life’ but on a much simpler judgement, whether they are present and whether their innate sense of liberty is as important to them as it is to anybody else. As anyone who has worked with patients with dementia can testify, the desire to act freely is perhaps the sole desire to be consistently held by all patients, at all times, in all environments. This is unsurprising; my thought experiment demonstrates that we derive value from our decision-making processes, however disordered or inaccurate they may be, because they are inherent to a settled sense of liberty and personhood. Hence, even patients with severe dementia would qualify for a right to autonomy conceptualised under the ‘sense of liberty’ approach.

Despite having intuitive value to us, I believe the ‘sense of liberty’ justification for autonomy to be fairly specific to dementia. The ‘sense of liberty’ justification takes on a greater significance in the context of dementia compared with adults of sound mind, those with transient mental disorders and children.

The first reason for this is that the mind of a dementia patient is necessarily contemporary and impulsive. The nature of the disease limits patients’ ability to plan for the future or form deeply held values. In contrast, adults in sound mind often voluntarily contravene their contemporary wishes in order to benefit from delayed gratification or to pursue moral principles. This ability of patients in sound mind to rationalise through values and forward planning suggests the importance of their immediate ‘sense of liberty’ is less significant than for a patient whose existence is necessarily contemporary and impulsive. Therefore, the ‘sense of liberty’ justification might make for a less convincing argument when applied to a healthy adult’s right to autonomy.

The second characteristic of dementia that qualifies patients to a right to autonomy under the ‘sense of liberty’ account is the irreversible nature of their contemporary and impulsive existence. If we are to assign such significance to a patient’s ‘sense of liberty’, then we should be confident that this contemporary existence is a fair reflection of their capacity to experience the world more generally and that all attempts have been made to enhance their critical faculties to a state qualifying them for an integrity-based autonomy. Therefore, in the context of reversible mental disorders, it might be justified to restrict a patient’s ‘sense of liberty’ to help them achieve an integrity-based autonomy at a later date.

Likewise, it might be justifiable to violate a child’s ‘sense of liberty’ because of their potential to develop an existence that is not contemporary and impulsive but that involves a state of mind better understood by the integrity account of autonomy. In contrast, terminally ill children may be better qualified for a ‘sense of liberty’ right to autonomy, due to the irreversible state of their contemporary existence. This chimes with intuition, further suggesting the irreversible nature of one’s contemporary and impulsive existence should be a factor in judging capacity for a ‘sense of liberty’ based autonomy.

Thereby, I have articulated a new justification for the value of autonomy, which applies to patients with dementia, therefore reaching the conclusion of my first claim: patients with advanced dementia qualify for a right to contemporary autonomy.

It is important to note that this argument is distinct from the normal contemporary ‘change of mind’, which a healthy person may experience. The patient with dementia does not change their mind in the sense that they understand the rationale behind a prior wish, reconsider it and revise that wish. If the patient were capable of doing so, then there would be no apparent conflict; the patient would have the capacity to simply amend their advance decision as understood by an integrity-based conception of autonomy.

Instead, as figure 1 illustrates, my argument is that patients with dementia lack capacity to amend their decisions as understood by an integrity-based account of autonomy but that they qualify for a right to contemporary autonomy as understood by the ‘sense of liberty’ account. This leaves two competing wishes: a prior advance decision guided by the integrity-based account of autonomy and a contemporary wish guided by a ‘sense of liberty’ account of autonomy. Obviously, these wishes must come into some practical conflict for an ethical dilemma to exist, but there need not be a considered ‘change of mind’ for this conflict to be created.

Figure 1

Conceiving the right to contemporary autonomy. As the case study illustrates, the argument outlined in claim A applies to scenarios taking place in the time period represented by the shaded box A. Here, patients lack the capacity to make contemporary decisions under the integrity-based approach, but they may qualify for a right to contemporary autonomy under a ‘sense of liberty’ approach. However, this claim does not relate to whether advance directives can be used as a proxy for contemporary decisions, which is addressed in claim B.

Claim B: precedent autonomy is not essential to respecting autonomy under the integrity view

Having argued that patients with dementia qualify for some right to contemporary autonomy, there are now two competing claims for a potential course of action. Either the patient’s apparent contemporary wishes (defended by a ‘sense of liberty’ account) can be honoured or their prior wishes outlined in an advance decision (defended by an integrity-based approach) can be followed. My second claim confronts Dworkin’s defence of this latter conception of autonomy, which he rationalises through a patient’s right to make decisions about their future self, known as precedent autonomy. Dworkin’s argument consists of two parts:

  1. Autonomy, as understood by the integrity view, means that patients have a right to express prudential concerns about their future, justified by our right to ‘shape our own life’.

  2. The need to fulfil these prudential concerns justifies restricting the wishes of our future self. In other words, the concept of precedent autonomy is essential to respecting autonomy as understood by the integrity view.

I have argued that the integrity view of autonomy is valid, and I find no problem with Dworkin’s first claim. However, in his second claim, I argue that Dworkin overlooks a distinction between our right to enact prudential concerns and our right to impose a retrograde concern on our future self. I illustrate this distinction with a case study:

Imagine two 25-year-old obese patients: A & B. Both wish to lose weight before their 30th birthday. Patient A commences a diet and undergoes bariatric surgery at his earliest convenience, reaching a healthy body mass index (BMI) before his 30th birthday. Patient B, however, decides to sign a contract committing him to have a liposuction operation the day before his 30th birthday.

Both of these patients are expressing a prudential concern for their future selves: to have a healthy BMI on their 30th birthday. However, in the case of patient B, there is an additional consideration: he wishes to commit his future self to a retrograde concern.

This distinction is important. The integrity view of autonomy allows us to make prudential choices that will shape our future character and the reality in which we live. This is qualitatively different to limiting our choices within that future reality. In short, a prudential concern shapes the choice architecture available to our future self, whereas a retrograde concern commits our future self to choices within that architecture. While an integrity view of autonomy that allows people to shape their life by enacting prudential concerns for their future is sensible, I see no justification why the ability to enact retrograde concerns is essential to an integrity view of autonomy. Since advance decisions would be understood as retrograde concerns, it is therefore possible to honour a patient’s integrity-based autonomy without honouring their advance decisions.

Conclusion

I have argued that patients with dementia qualify for some right to autonomy, related to their immediate sense of liberty. Second, I have disputed Dworkin’s argument that we have a right to impose retrograde concerns on our future selves and have argued that this overstates integrity-based autonomy. Together, these arguments suggest there is no reason for dementia patients to be deprived of a contemporary wish in favour of a previous wish. This problematises advance decisions and, in our case study, argues in favour of Sally’s contemporary wishes.

Finally, it is worth noting that my argument entertains a different notion of autonomy to previous approaches. Specifically, claim A’s emphasis on a patient’s ‘sense of liberty’ rather than informed choice might justify deceptive strategies in dementia care. Although such strategies are common, such as in covert medicine administration, they are usually justified on the grounds that patients lack capacity rather than by conceptualising a new understanding of autonomy.14 Therefore, my arguments may have implications for those debates. Despite this caveat, I believe my argument is practically significant to much of dementia care, as it is difficult to imagine long-term deceptive care strategies that do not involve ethically unjustified infringements, even if only short lived, on a patient’s ‘sense of liberty’ at some point.

The sense of liberty argument may also problematise other areas of care. If it is to be accepted, we might find it difficult to justify administering life-prolonging antibiotics to a dementia patient who is refusing them, for instance. Currently, doctors may justify forceful strategies by appealing to a patient’s best interests, but if we are to defend a sense of liberty account of autonomy, this would leave such actions on ethically problematic grounds. This requires further discussion about how far sense of liberty should be protected. If we find it too problematic to honour a patient’s apparent contemporary refusal of life-prolonging treatment, for instance, it might be prudent to return to wider questions about the tension between autonomy and beneficence in clinical ethics. These questions go beyond the scope of this paper but are nonetheless important to consider if we are to entertain a sense of liberty account of autonomy.

Acknowledgments

I would like to thank Camillia Kong, PhD (Department of Psychiatry, University of Oxford) for her supervision and feedback on earlier drafts of this paper.

References

Footnotes

  • i One consideration which this paper will not address is personal identity. Like Dworkin’s argument, the arguments presented here rest on the assumption that the patient with dementia is the same person they were before developing the disease. If this is wrong, and arguments made by Derek Parfit and Rebecca Dresser are correct, this paper would prove largely irrelevant.15–18

  • Contributors I take responsibility for the content of this manuscript, wrote it and decided to submit it for publication. The manuscript represents original work undertaken entirely by the proposed author.

  • Funding The author has not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.