By George Gillett

Last month, the Guardian columnist Polly Toynbee wrote an article in support of assisted dying. She wrote about Katherine Whitehorn, her former colleague at the Observer. Describing Whitehorn, Toynbee writes:

She is not herself. Her old self would not recognise herself in this other being who sits in the care home. What or who she has become is a difficult philosophical question, but she is no longer Katharine Whitehorn as was.

Whitehorn has Alzheimer’s disease. An increasing number of us will develop the condition as we age and over 1.1 million people will be diagnosed with dementia in the UK by 2025.

Despite this, we are unprepared to deal with the ethical issues the condition raises. One question is especially persistent; who should decide how patients are treated once their dementia has progressed? Is it ethical for us to outline the details of care for our future selves, even if our future self later appears to reject it?

This dilemma is not only important in the context of assisted dying. Commonly cited examples include vegetarian patients who begin to crave meat or happily-married monogamous patients who exhibit same-sex behaviour as their dementia progresses. How should we deal with these changes of mind and behaviour which seem so easily attributable to pathological processes?

As Toynbee’s quote hints, one way to approach these questions is to consider theories of personal identity, inspired by the work of Derek Parfit. A simplification of such arguments might suggest that dementia sufficiently alters one’s identity such that the patient with dementia is not the same person as they were before they developed the condition. Hence, there may be little reason why the wishes of the earlier self should be applied to the later self. The two are ultimately different persons and have no right to make decisions about the other’s care.

This argument may be persuasive, but also has short-comings. A true endorsement of such an approach to personal identity would see us reject many modern societal norms. We would have little claim over our property from one minute to the next and there’d be no reason for any of our decisions – medical or otherwise – to be applied to our future selves. It is difficult to justify adopting such an approach in the domain of medical ethics without first considering its implications in wider society.

My paper puts personal identity arguments to one side. Let’s assume the person with dementia is the same person as they were before their disease progressed. Might there still be reason to question the ethical status of advance care planning?

Current consensus would suggest not. Popular conceptions of autonomy appeal to an integrity-based approach developed by Ronald Dworkin. He suggests autonomy is important because it allows each of us “to lead our own lives… according to some coherent and distinctive sense of character, conviction, and interest”.

Dworkin’s analysis correctly argues that patients with dementia do not qualify for a right to autonomy under such an account, since they lack the cognitive capacities required to make decisions of sufficient value. He also argues that if we are to truly respect a patient’s autonomy as understood by such an integrity-based account, we must fulfil the requests patients have made in the past. If we ignore their prior decisions, he suggests, then we do not respect their autonomy at all.

My paper argues that Dworkin overlooks an important point. I argue that the value which is found in autonomy may be different for different groups of patients, and that patients with dementia may qualify for some right to contemporary autonomy. Secondly, I argue that honouring advance decisions is not required to respect autonomy as understood by Dworkin’s integrity-based account and that Dworkin overstates the reach of this right. Together, these arguments problematise advance decisions.

My arguments have no particular implications on the assisted dying debate, provided that patients are able to consent to the given procedure when it takes place. However, it does oppose the claim that an assisted death – or any other aspect of care – should be carried out simply because a patient has previously outlined a wish for it to be so.

Advance decisions are in vogue, and they no doubt give patients a sense of control. This empowerment can be an important tool to help patients navigate difficult territory when they are first diagnosed with such conditions. However, we should be wary of empowering newly-diagnosed patients while neglecting our ethical duties to the voiceless patients who are bound to advance decisions they have no memory or awareness of.

Therefore my paper argues that, even putting aside questions of personal identity, there may still be reason to oppose our recent drive towards advance care planning.

Author: George Gillett

Affiliations: Medical Sciences Division, University of Oxford.

Paper: Advance decisions in dementia: when the past conflicts with the present

Competing interests: None declared.