Main findings

For most patients who are able to express a choice but in whom understanding was rated to be poor, DMC was absent, and this was unrelated to the type of hospital they were in. Among patients able to understand information and express a choice, a higher proportion lacked DMC in the psychiatric hospital compared with the medical hospital (39% and 13% respectively). In these patients, the medical patient sample typically had poor reasoning associated with lack of DMC, whereas the psychiatric patients typically had poor appreciation.

Strengths and limitations

This study has limitations. Formal diagnoses of dementia or delirium were not made in the medical sample and psychiatric comorbidity (such as depression) was not analysed. However, the prevalence of cognitive impairment in this medical sample was high at 25%, with a strong association to loss of DMC. ^{Reference Raymont, Bingley, Buchanan, David, Hayward and Wessely13,Reference Kim20}

The assessments conducted in these two studies were made by different research clinicians at different times. This introduces the possibility of observer biases. However, there was high interrater reliability ^{Reference Raymont, Buchanan, David, Hayward, Wessely and Hotopf16,Reference Cairns, Maddock, Buchanan, David, Hayward and Richardson17} and the judgements of DMC were checked by the same panel of experienced clinicians. We think it unlikely that the interpretation of DMC changed significantly in the period between the two clinical researchers' work in the medical and psychiatric hospital. We used psychiatrists to assess DMC - which reflects usual clinical practice, and although DMC is a legal, social and ethical concept the courts have looked to clinical expertise in their rulings on DMC. Using doctors with knowledge of psychopathology is therefore a strength of the study. However, some might argue that anthropologists or other social scientists may be better attuned than doctors to understand the relationships between patients' explanations and their cultural or religious background.

The MacCAT-T has often been used to collect scores without judgement of DMC and without clinical assessment, so a limitation to this study may be the way the tool is incorporated into clinical contexts and shaped by them with loss of objectivity. However, use of earlier versions of the MacCAT-T in clinical settings along more de-contextualised, psychometric lines received criticisms with regard to validity. ^{Reference Kapp and Mossman21} Its authors developed the MacCAT-T, in part, to promote the incorporation of a tool into clinical and psychosocial assessment in a way that was both practical and flexible. ^{Reference Grisso and Applebaum22} The MacCAT-T, although potentially reliable, has short scales with limited ranges. The subscales have different ranges and we were forced to use a fairly crude definition of impairment (scoring below the halfway point on each scale). This limits statistical analysis and was complicated by the different scales we used for understanding in the two samples.

To our knowledge this is the first study to explicitly compare patients in two settings - psychiatric and medical. Other studies have described abilities in different diagnostic groups, but typically have not compared these with the criterion of whether the individual does or does not have DMC. ^{Reference Grisso and Appelbaum23-Reference Marson, Cody, Ingram and Harrell26} We therefore think this study adds helpful data to the literature. Next, we consider the four decision-making abilities in relation to our data before briefly discussing how our understanding of DMC might be advanced.

Understanding

The understanding ability is a pre-requisite for a patient's consent, or refusal, to be ‘informed’. To assess understanding it is necessary to ensure that information has been given to the patient in a manner clear enough to be understood. Grisso & Appelbaum advise beginning the capacity assessment with information disclosure and to test whether the patient has understood it by inviting them to paraphrase. ^{Reference Grisso and Applebaum22} Memory, language and attention impairments are likely to interfere with this understanding ability as are altered emotional states or thought disorder. There is a question about how far the assessor, or systems around the assessor, need to go to adjust for, augment or ‘scaffold’ ^{Reference Anderson, Forst, Hartmann, Jaeggi and Saar27} these disabilities to enable the patient to understand. The general principle is that assessment of DMC must ensure that lack of DMC is not founded on poor understanding when understanding has not been enabled or supported. The United Nations Convention on the Rights of People with Disabilities ^{28,Reference Bartlett29} emphasises a right to this kind of support and it also features in DMC law. ^{Reference Grisso and Applebaum22,30}

How is poor understanding related to DMC? Our data indicate that poor understanding is usually inconsistent with DMC; few patients with poor understanding were judged to have DMC. But poor understanding alone does not seem to be the determining factor. The few with poor understanding with DMC had good appreciation or reasoning and those with poor understanding without DMC almost always had poor appreciation or reasoning as well. So we suggest that if there is lack of DMC with poor understanding, this typically reflects not poor understanding alone, but poor scores in two or all three of the abilities in question.

What of good understanding? Is it enough for DMC? Our data indicate, in both medical and psychiatric settings, that it is not enough. A fair percentage show good understanding but lack DMC because of poor appreciation or reasoning. Our data therefore suggest that poor understanding alone is neither a strong necessary nor sufficient condition for a lack of DMC and that understanding needs to be considered in relation to appreciation and reasoning to have a bearing on DMC.

Appreciation

Grisso & Appelbaum ^{Reference Grisso and Applebaum22} suggest that a patient's non-acknowledgement of the diagnosis and recommended treatment should be counted as a failure of appreciation only when patients' choices are based on premises (beliefs) that: (a) are substantially irrational, unrealistic or a considerable distortion of reality, (b) are the consequence of impaired cognition or affect, and (c) are relevant to the treatment decision. Saks and colleagues ^{Reference Saks and Behnke31,Reference Saks, Dunn, Marshall, Nayak, Golshan and Jeste32} have elaborated on (a) with a concept of ‘patently false belief’ with the aim of finding a standard between, on the one hand, impossible belief and, on the other, making doctors the final authorities on truth. Grisso & Appelbaum characterise appreciation as ‘the ability to apply the information abstractly understood to his or her own situation’. ^{Reference Grisso and Applebaum22} Our data suggest that this ability is a more critical variable for DMC in the psychiatric hospital than in the medical hospital. The MacCAT-T however only captures this ability with two questions and a 0-4 scale (see the Appendix for the two questions and guidelines on rating inability).

Reasoning

Decision-making capacity law separates the decision-making process from the decision itself (otherwise the DMC test would become an ‘outcome test’). Grisso & Appelbaum ^{Reference Grisso and Applebaum22} address this by focusing on the ability to ‘manipulate information rationally’, that is, ‘to engage in logical processes when using the information that they have understood and appreciate in arriving at a decision’. The essence of the ability, according to Grisso & Appelbaum, is ‘engaging in a process of weighing treatment options’. Caution needs to be exercised in interpreting this as a procedural reasoning standard, as there is plenty of evidence to suggest healthy human decision makers do not conform reliably to formal reasoning. ^{Reference Sutherland33,Reference Owen, Cutting and David34} Caution also needs to be exercised in interpreting this as a substantive reasoning standard because the Mental Capacity Act 2005 states ‘a person is not to be treated as unable to make a decision merely because he makes an unwise decision’. Grisso & Appelbaum adapt a loose standard and draw attention to the connection that any reasoning inability must have to the disorder. An example of the questions the MacCAT-T uses to probe the reasoning ability is given in the Appendix together with the guideline on rating inability. Our data suggest that the reasoning ability is a more critical variable for DMC in the medical than in the psychiatric hospital.

Values and evaluation

Values and evaluation is arguably a feature of all of the above abilities ^{Reference Kim20} but in relation to appreciation this seems particularly salient. With regard to the patient with anorexia who values thinness over and above life, Grisso & Appelbaum argue that the evaluative inability is included in an inability to appreciate. ^{Reference Grisso and Appelbaum35} On MacCAT-T appreciation, the person is being assessed, in effect, on their ability to take reality as it is. How this is to be separated from the value judgements of society in general and from medical expertise in particular has been controversial as indeed has the question of whether it ought to be so separated. ^{Reference Freyenhagen and O'Shea36} The Mental Capacity Act specifies DMC abilities somewhat differently to the four-ability model. Instead of referring to appreciation and reasoning abilities it specifies an ability to ‘use or weigh’. On the question of evaluation we may ask what was in the minds of the legislators when these words were being examined by them. The following ^{Reference Hansard37} is an extract from an exchange between the government minister (Baroness Ashton) and the leader of the opposition (Earl Howe) in the House of the Lords during the parliamentary passage of the Mental Capacity Bill. Earl Howe sought to amend ‘use or weigh’ to ‘use, evaluate or weigh’.

Decision-making capacity assessment within the Mental Capacity Act thus seems to require an assessment of evaluative ability via ‘use or weigh’. ^{Reference Tan, Stewart and Hope38}

Advancing understanding of DMC

Although difficult concepts may only afford so much definitional precision, clinicians working in medical or psychiatric settings need to be aware of the relevant decision-making abilities and guard against idiosyncratic or erroneous interpretation. Our data suggest that clinicians, when assessing DMC, need to be sensitive not only to the decision at issue and its gravity or risk ^{Reference Kim, Caine, Currier, Leibovici and Ryan39} but also to the kind of disorder - i.e. whether it is medical or psychiatric.

Some may fear that this sort of sensitivity to diagnosis could allow prejudicial assumptions to compromise the objectivity of DMC assessment. We think two observations are in order in connection with this concern. First, it is important to emphasise that sensitivity to diagnosis in the assessment process is not at all equivalent to a ‘status’ account of capacity or competence. The best protection against prejudicial assessments is not blindness to diagnosis, but insistence on a high standard of reason-giving in the conduct of assessments. Second, we would echo Gadamer's warning about the so-called prejudice against prejudice. ^{Reference Gadamer40} The assessment of DMC is an interpretative endeavour, and interpretation always takes place in a context in which the interpreter arrives with orienting ‘pre-judgements’. Eliminating pre-judgement altogether would not lead to objective interpretation; it would make interpretation impossible. Objectivity is thus best served when pre-judgements are themselves responsive to evidence, subjected to critical scrutiny, and open to challenge and revision.

In advancing our understanding of DMC, it is worth pausing over one suggestive feature of Earl Howe's description of the way that depression can threaten DMC. He describes how the ‘world’ looks to the person with depression. Here the focus is not so much on the processing of information in the production of a decision, nor is the issue one of abnormal values. The question is rather whether DMC can be threatened by the distinctive way in which an individual encounters and makes sense of their world. Can the nature of that encounter itself disrupt DMC? If so, does it disrupt certain sorts of decisions but not others? Earl Howe's observation suggests that assessment of the ability to ‘use or weigh’ can sometimes require a distinctive skill-set on the part of the assessor: ^{Reference Martin and Hickerson41} for example, the skills of understanding and unpacking the ways in which people with depression encounter the world in which their decisions must be made. More work will be required to translate that understanding into publically attestable judgements of whether such individuals have the abilities to decide for themselves about treatment - literally, make that decision their own. ^{Reference Owen, Freyenhagen, Richardson and Hotopf42}

Implications

Decision-making capacity is a concept that now crosses psychiatric and medical practice. Using a validated interview - the MacCAT-T - we show some evidence that appreciation and reasoning differ in their association with DMC judgements in psychiatric and medical settings. Both the appreciation and reasoning abilities and the ‘use or weigh’ abilities are normative abilities in that they all derive their meaning from the right to self-determination, or individual autonomy. ^{Reference Owen, Freyenhagen, Richardson and Hotopf42,Reference Banner43}

Law can tend towards assumptions that medical in-patients choose treatment autonomously and that psychiatric in-patients are of ‘unsound mind’ in relation to treatment. ⁴⁴ Those assumptions are not well supported by evidence. ^{Reference Owen, Richardson, David, Szmukler, Hayward and Hotopf12,Reference Raymont, Bingley, Buchanan, David, Hayward and Wessely13,Reference Okai, Owen, McGuire, Singh, Churchill and Hotopf15,Reference Fassassi, Bianchi, Stiefel and Waeber45} Policy-making is moving beyond these assumptions and debating whether DMC law is evolving to recognise differences in kinds of mental disorder or disability that obviate the need for two streams of mental health law. ^{Reference Richardson46} This study adds data which, we hope, give some nuance to this policy debate and helps clinical and legal practitioners recognise that DMC assessment requires attention to the kind of mental disorder or disability present, as well as to the kind of treatment decision.

Appendix

Examples from the MacCAT-T of how appreciation and reasoning abilities are probed, with guidelines on inability

Ability	Example probe	Guidelines on inability
Appreciation	‘That is what we [your doctors] think is the problem in your case. If you have any reason to doubt that, I'd like you to tell me so. What do you think?’	Patient clearly does not agree that he or she has the disclosed disorder, with reasoning based on a delusional premise or some other belief that seriously distorts reality and does not have a reasonable basis in the patient's cultural or religious background.
		OR
		Patient believes that the symptoms are related to circumstances other than a medical/psychiatric disorder (e.g. psychotic symptoms seen as consequences of work-related stress).
		OR
		Patient clearly disagrees with symptoms or disorder, but with no comprehensible explanation offered.
	‘You might or might not decide that this is the treatment you want - we'll talk about that later. But do you think it's possible that this treatment might be of benefit to you?’	Patient acknowledges at least some potential for the treatment to produce some benefit, but for reasons that appear to be based on a delusional premise or a serious distortion of reality.
		OR
		Patient does not believe that the treatment has the potential to produce any benefit, and offers reasons that appear to be delusional or a serious distortion of reality.
Reasoning	‘You think that treatment X might be best. Tell me what it is that makes that seem better than the others.’	Patient can make no comparative statements.