[I am very grateful to George Szmukler, Professor of Psychiatry and Society at the Institute of Psychiatry, KCL, for this detailed comment responding to Lucy Series’ characteristically thoughtful comment on my post on the proposed Mental Capacity Bill in Northern Ireland which seeks to combine mental capacity and mental health legislation. Given its length, this merits a separate post, but I would urge readers first to read the original post on the Bill and Lucy’s comment. This is exactly the sort of debate I want to get up and running!]
The Northern Ireland Bill is a ground-breaking step in the right direction. My comments in this post are mainly about law applied to mental health. As is recognised in the Bill, conventional mental health legislation discriminates unfairly against persons with a mental illness. It is time for a radical change.
I think it is worth rehearsing how mental health legislation discriminates before looking at some of the significant issues raised by Lucy Series and others. Here I will examine, in particular, the nature of the assessment of decision-making capacity. Discussions often fail to do justice to its depth and breadth.
How does conventional mental health legislation discriminate?
First, let’s consider a person’s health interests. We accept that patients with ‘physical’ disorders provided they have decision-making capacity (DMC), can make decisions that may be seriously detrimental to their health or safety. In contrast to the necessary role of impaired decision-making in patients with ‘physical disorder’, under conventional mental health legislation capacity plays little or no role in decisions to initiate psychiatric treatment against a patient’s wishes. The criteria for the involuntary treatment of ‘mental disorders’ thus fail to respect the autonomy of the patient. The patient’s reasons for rejecting the treatment are not explored (as they must be in determining DMC), nor is the question of whether treatment is in the ‘best interests’ (BI) of the patient – from the perspective of the patient, not the clinical team. The key considerations are the presence of a mental disorder and risks to the patient’s health or safety. For persons with physical disorders with DMC, their personal values are given dominion. Those with mental disorder are not accorded this privilege.
There seems to be an underlying assumption in legislation of this kind that ‘mental disorder’ necessarily entails mental incapacity, so the question does not need to be asked, and that the wishes and preferences of a person with a disordered mind are not to be taken seriously in determining where their best interests lie. Research has shown that even among the most ill patients, those admitted to acute psychiatric wards, 40-60% retain capacity (Owen et al, BMJ 337 (2008): a448). By failing to respect the mentally disordered patient’s autonomy; by not presuming capacity unless there is reason for doubt; by assuming that mental disorder entails incapacity; and by enshrining these prejudices in legislation that applies uniquely to those with ‘mental disorder’, current mental health legislation discriminates against those with mental disorders.
Treatment for the protection of others
People with mental disorders are unusual in being liable to detention (usually in hospital) because they are assessed as presenting a risk of harm to others, but before they have actually committed an offence. This constitutes a form of preventive detention that is selective.
Imagine two overlapping populations of people – first, all people with a mental disorder and second, all people who present a risk to others. Some people with a mental disorder will also fall into the group who present a risk. Let us imagine that we can accurately assess the risk presented by an individual. Now take everyone in the community who presents a level of risk above an unacceptable threshold of risk. A small proportion of these will be people with a mental disorder. The vast majority will be in the non-mental disorder group. Conventional mental health law allows the detention only of those with mental disorder on the basis of risk alone, not the much larger remainder of the group (who are not mentally ill) who present an equal level of risk. How can this be justified? There is no evidence that risk is easier to assess in those with mental disorder, or that violence is more predictable in this group. In the absence of antisocial personality disorder or drug abuse, violence is at most only modestly more frequent in people suffering from a psychosis. Even if this were not so, selective detention on the basis of risk would not be justified. But might treatability be a justification; if risky people with mental disorders are treatable perhaps that is a good reason for involuntary treatment? Lengthy hospitalisations, lasting many years if not decades, in secure hospitals suggest this is not the case. And it is possible, probably likely, that psychosocial interventions such as anger management or controlled drinking programmes for non-mentally disordered risky persons could prove more effective in reducing the total number of violent incidents in the community than treating the relatively small number of patients with mental disorder?
One cannot escape the conclusion that the ‘protection of others’ criterion in mental health legislation is discriminatory. If preventive detention is to be allowed for the mentally disordered solely on account of their risk to others, if we are to avoid discrimination, so should it be for all of us. Fairness demands that all those presenting an equal level of risk to others should be equally liable to detention. This of course amounts to a generic dangerousness or preventive detention provision which many will find unacceptable. But the principle of non-discrimination requires that either we have generic legislation applicable to all of us; or that we have no preventive detention for anyone, including those with mental disorder.
This does not mean that a person’s dangerousness is unimportant. If it is reliably linked to an individual’s mental disorder for which there is treatment, then, as discussed below, if the person lacks DMC, involuntary treatment may be justified if it is in their best interests. What if the patient has capacity? If involuntary ‘treatment’ is to be imposed under these circumstances, no health interest is necessarily served – the person with capacity is usually regarded as the best judge of what is in their best interests. Protection of the public becomes the sole interest. Why then is health legislation appropriate?
Persons with mental disorder do not receive the protections from preventive detention that the rest of us do. Mental health legislation denies such protection, thus reinforcing an underlying stereotype that the mentally ill are inherently dangerous.
The Fusion Law proposal
The solution is a ‘fusion’ law (Dawson, J, and Szmukler, G. British Journal of Psychiatry 188 (2006): 504-09) such as is proposed for Northern Ireland. Involuntary treatment would only be permitted, firstly, when the person has an impairment of DMC and, secondly, if treatment would be in the person’s BI. What is especially important is that the impaired DMC could be of any cause, in any person– a head injury, post-epileptic confusion, schizophrenia, confusion due to an adverse drug reaction or infection, Alzheimer’s Disease, and so on. Involuntary treatment would cease when DMC has been re-established.
Why is this approach termed, ‘fusion’? The proposal builds on the strengths of the two existing regimes. Existing capacity-based legislation’s strength in giving due weight to autonomy is nearly always counterbalanced by a number of weaknesses. These lie in the lack of sufficient attention to emergency treatment, non-consensual treatment and detention in hospital. But these are exactly those areas in which civil commitment schemes are strong; detention and the use of force are clearly authorised and regulated. The lack of clarity in these areas in capacity-based legislation may pose a problem for the treatment of patients with ‘physical disorders’ who object to treatment, as it would for those with mental disorders. Patients who object to treatment but who do not suffer from a ‘mental disorder’ may be treated under civil commitment – inappropriately – because of a reluctance of clinicians to use force unless they can rely on clear statutory authority. Or perhaps they are not treated at all even though treatment would have been in their best interests – again because of the lack of clarity about its authorisation.
Some problems with a capacity-based regime
Lucy Series’ comment rightly points to some important problems that might arise in a capacity-based regime.
There are problems with the baggage associated with the terms ‘decision-making capacity’ and ‘best interests’, but for the moment let’s stick with those terms.
DMC and best interests in the Fusion proposal carry a heavy burden in justifying coercive interventions. The concepts thus need to be clear enough to meet this use. In medicine, one expects investigations to have a high degree of objectivity. Many slip into expecting the same precision in a test of DMC. However, the evaluation is of a different kind. A standardised ‘objective’ test of cognitive functioning such as the Mini Mental State has only a modest correlation with a person’s ability to make a specific decision at a particular time. Assessing DMC may involve, as we shall see, an ‘interpretation’ of complex phenomena such as beliefs and values in a particular context. Nevertheless, the research evidence on the inter-rater reliability of DMC assessments points to excellent agreement, especially when made by clinicians with a similar training (e.g. Cairns et al, British Journal of Psychiatry 187 (2005): 372-78). Nonetheless, there will always be difficult cases.
The kind of assessment that is described in the MCA 2005 has been criticised as being too ‘cognitive’, thus failing to take account of the role of values and emotions in decision-making. While a large proportion of persons, even those with a psychosis, have a clear impairment in understanding or retention – that is, they are unable to repeat in their own words the information they have just been given – or they are unable to articulate or arrive at a settled choice (Owen et al, British Journal of Psychiatry 203 (2013): 461-67), there are many instances where the assessment is more problematic. In particular, the person’s beliefs and values call for some kind of consideration. The problem here is that the assessment of DMC should respect a person’s beliefs and values and not reflect the beliefs and values of the assessor. Thus what has been aimed for, explicitly in the thinking behind the MCA 2005, is that the test of DMC should be ‘procedural’, tracking the thinking processes used by the person to arrive at a decision. It restricts itself to how the person manipulates the relevant information in the light of the beliefs and values the person holds, which themselves, are not to be subverted by those of the assessor. Such a procedural test has a laudable aim as it prevents clinicians from deeming a person to lack DMC because they make a seemingly imprudent decision. Thus if the decision-making process is intact, the person is entitled to freely decide.
Unfortunately, a procedural view does not take us as far as our intuitions may lead us. For example, a person may refuse a medication because of a belief that they are the subject of an experiment intended to control them, or a person with anorexia nervosa may accept the risks of dying through self-starvation claiming that being thin is more important to them than life itself. Their behavior is ‘rational’ if one accepts their guiding beliefs or values. In these cases, to understand whether the person’s capacity to decide is in some way impaired, it appears that substantive judgments about the content of these beliefs or values can’t be avoided. However, diagnoses and treatments can be contested; disagreeing with a medical recommendation is not evidence of a failure to appreciate significant facts about one’s condition, but may be the result of reflecting critically on the information offered. So, when assessing DMC, how might evaluative judgments about a person’s beliefs and values be made, and what degree of ‘objectivity’ can be achieved?
There is clearly a relationship between DMC and autonomy in some senses; varying accounts suggest that autonomy may be foundational for capacity or that having capacity is a necessary condition for autonomy (Freyenhagen & O’Shea, International Journal of Law in Context 9, no. 01 (2013): 53-70). However, the question for the assessor in the situation I am describing is relatively circumscribed – how far does the person’s specific decision at a particular time have the property of being ‘self-determined’ or of being ‘owned’ by the person. Buchanan and Brock (“Deciding for Others: The Ethics of Surrogate Decision Making.” Cambridge: University Press, 1990) pointing to what this might mean stated that “a competent decision-maker also requires a set of values or conception of what is good that is at least minimally consistent, stable and affirmed as his or her own.” Clearly required then is that an evaluation of an unusual or apparently unwise treatment decision must go beyond a simple examination of the content of the decision and the immediate explanation offered by the person. A more complex judgment, or ‘interpretation’, may be required.
Decision-making capability and the crucial role of ‘interpretation’
Natalie Banner, a philosopher, has drawn on Donald Davidson’s theory of ‘radical interpretation’ in thinking about DMC (“Judging By a Different Standard? Examining the Role of Rationality in Assessments of Mental Capacity.” University of Central Lancashire 2010 (pdf at http://www.clok.uclan.ac.uk)). In my view, this kind of approach offers helpful guidance for understanding what an assessment of DMC involves in the more problematic cases (Banner and Szmukler, Journal of Applied Philosophy 30 (2014): 379-74.) Davidson examined the necessary conditions for interpretation of a ‘radically unfamiliar’ speaker. This would seem to resonate with the clinical problem we are considering; for example, where a person expresses seemingly unusual ideas that lie behind an apparently unwise treatment decision. If such conditions for interpretation can be defined, they could shed light on normative standards governing how we go about making sense of a person’s unusual decision.
Davidson generated an idealised method for understanding a radically unfamiliar speaker. He argued that the interpreter is licensed to make certain assumptions about the speaker’s beliefs and values, because there are intrinsic, normatively governed relationships between the meanings of the speaker’s utterances, the beliefs he holds and the actions he performs, on the basis of which the interpreter can frame his attempts to understand the speaker’s behaviour. Without these normative structures it would, after all, not be possible to understand any other person.
Davidson‘s account of these norms is known as the ‘Principle of Charity’. It states that we must assume that a speaker is by and large consistent and correct in his beliefs. This is a critical starting point (which echoes the presumption of DMC). These conditions of consistency and correctness are termed ‘Coherence’ and ‘Correspondence’, respectively. A person’s beliefs, intentions and actions hang together broadly as a more or less coherent whole, and they largely reflect what is true. An interpreter is thus guided by these norms as a first step in trying to understand the speaker. The caveat of ‘by and large’ reflects the fact that an interpretation of a behaviour is imprecise, may be subject to revision, and need not be strictly contained by the demands for correspondence and coherence. Our mentality is inherently messy; we form beliefs on partial, perhaps conflicting, evidence, we may act against our avowed interests, and we make decisions based on emotional rather than rational thought processes. The theory allows significant latitude for errors. What the standards of Charity do imply is that there must necessarily be a broad background of true and coherent beliefs in order for a speaker to be interpretable at all. Indeed, it is only against this background of largely true and coherent beliefs, shared by an interpreter, that incoherence can be recognized.
A significant consequence of this view is that the ‘mental’ is holistically constituted. Rather than simply focussing on the content of a belief for identifying a problem with DMC, we must examine the relations it bears to the person’s other beliefs, evaluative commitments and actions. Exploring the context in which such beliefs arise, their biographical evolution, and their connectedness with other beliefs, values and behaviours will identify whether they broadly cohere and are capable of supporting, rather than undermining, the capacity to make a particular decision. This is presumably why we accept a Jehovah’s Witness’s refusal of a blood transfusion.
This level of assessment will thus take substantive content into account. However, this need not lead to unwarranted paternalism as the judgement is not arbitrary but normatively governed – interpretation is based on the normative, relational structure of a person’s beliefs, values and decisions. Taking the broader holistic context into account, points towards the kinds of questions an assessor conducting an evaluation of DMC ought to bear in mind. For example, one might examine the consistency of the person’s decision-influencing beliefs with his broader worldview and other values, the support such beliefs have, whether they are amenable to revision or argument, their stability over time, their evolution, the extent of their self-endorsement, past commitments they have engendered, and their cultural meaningfulness and rationality, to name but a few possible considerations. It is only if we acknowledge the context surrounding each decision and the relational nature of the beliefs and values impacting on decisions that we are in a position to identify where anomalies occur that might indicate a problem with DMC. Whilst a clear-cut prescription for how DMC judgements should work cannot be offered, there are nonetheless strong grounds for employing a wider scope of enquiry than the ‘procedural’ conception of capacity permits. This assessment is most pertinent to the ability to ‘use or weigh’ component of capacity in the Mental Capacity Act 2005.
Davidson recognised an inherent ‘indeterminacy’ in interpretation. Perhaps an account thus far not coherent, would not remain so with further probing of the person’s beliefs and values. This is certainly a consideration, but in a real-world encounter, where a decision needs to be made whether a person lacks DMC for a decision carrying important consequences, the aim is to reach a point in the dialogue (where it should be noted there is also a critical role for support) when it is reasonable to conclude whether there is a significant-enough breach in the relational structure of the person’s beliefs and values. A degree of ‘objectivity’ appropriate for this purpose is the goal.
The assessment will be further tested by asking others (for example, those who know the person well, or through an appeal to a tribunal or court) whether the judgement is sufficiently convincing. Freyenhagen and O’Shea (International Journal of Law in Context 9, no. 01 (2013): 53-70) suggest the possibility of a process of institutional oversight involving among others those who have been subject to non-consensual interventions. They propose that those previously affected should have a voice in deciding how evaluations should be interpreted.
If the assessment of DMC, especially in difficult cases, involves more than a reductionist procedural approach, there are some practical implications. Understanding the approach outlined, although based on shared norms of interpretation, probably requires training that goes beyond simple interviewing skills. A DMC assessment may require substantial time. Such procedures may not fit easily with time pressures imposed by emergency departments, for example. Nor do they comport well with trends in clinical practice to devise straightforward algorithms, checklists or ‘decision-aids’ to assist, standardise and speed-up decision-making.
Others have written accounts of ‘interpretation’ that have similarities to what I have described; for example, Jonathan Glover (“Towards Humanism in Psychiatry” in Tanner Lectures on Human Values, Volume 24: University of Utah Press, 2004; “Alien Landscapes: Interpreting Disordered Minds” Harvard University Press 2014 – in press). I am not suggesting that Davidson’s approach is the only one that we might draw on. The task is to develop a framework that is clinically applicable – but one that does justice to the complexity of the task.
A change in language that might clarify our thinking
It is probably a step too far at this time for a government to adopt the terms promoted in the UN Convention on the Rights of Persons with Disabilities – ‘will and preferences’, but thinking in those terms may help to make clearer what DMC and BI might mean. I suggest that the process of ‘interpretation’ described above is essentially aimed at ascertaining what a person’s ‘will and preferences’ are. There may be a problem in the person’s ability to express these at a time when an important decision needs to be made. Or perhaps the person is expressing their contemporaneous will and preferences very clearly, but there may be a serious doubt about the ‘authenticity’ of these when set against the person’s previously expressed will and preferences, particularly when those previous will and preferences have been enduring, stable, self-endorsed, evidenced a strong commitment, and have shaped the person’s significant life choices. Determining the answer to such a question is the task of interpretation. DMC would then become a question of whether an apparently imprudent decision is in accord with the person’s ‘true’ or ‘authentic’ will and preferences. If it is not, then acting in the ‘best interests’ of the person means acting to facilitate and give expression to the person’s now ascertained ‘true’ will and preferences, or the best possible interpretation of what these are.
I recognise that there will be some instances where this kind of assessment will be complex and difficult. However, I don’t think it can be avoided, and it is clearly much more respectful of the person’s self-determination and autonomy than the test applied under conventional mental health legislation. Fortunately, research indicates that the difficult cases are likely to be relatively uncommon (Owen et al, British Journal of Psychiatry 203, (2013): 461-67). Though the difficult cases present problems, they are the right problems that need to be addressed.
Institute of Psychiatry, King’s College London