The psychiatrists who serve on mental health review tribunals (MHRTs) are working in the fraught borderland between law and medicine. They have both to confront the potentially conflicting values of autonomy and beneficence and occasionally to resolve some fundamental incongruence between legal categories and clinical realities (Reference Eastman and PeayEastman & Peay, 1999). This much may be inevitable, given the context within which they are required to work, but arguably their position is made more difficult still by the variety of tasks they are given within the tribunal process. In essence, the medical member acts as an expert, as a witness and as a decision-maker. From both an ethical and a legal perspective this presents significant problems. In what follows we examine how these problems are treated in practice before the tribunal hearing, during the hearing and during the subsequent deliberations of the tribunal panel.

LEGAL CONTEXT

The primary role of MHRTs is to review the legality of a patient's detention in hospital and to direct the discharge of those to whom the statutory discharge criteria apply. Their current powers, duties and procedures are contained within the Mental Health Act 1983 and the Mental Health Review Tribunal Rules (Her Majesty's Stationery Office, 1983). To hear individual cases, tribunals sit as a panel of three: a psychiatrist, a lay person and a legal president.

As we have suggested elsewhere, MHRTs are designed to safeguard the patient's right to be free from unjustified detention in hospital (Reference Richardson and MachinRichardson & Machin, 1999). If they are properly to fulfil this role, they must be rigorous in their testing of medical claims and, to this end, the presence of the medical member can be seen as crucial: without an expert on the panel, the tribunal would be ill-fitted to its task. However, even in their role as experts, medical members are faced with a delicate balancing act: they must be sufficiently interventionist to enable tribunals to provide a rigorous challenge and yet sufficiently discrete to avoid dominating the panels' decision-making. The danger that the medical model may dominate what must essentially be a legal process has long been recognised (Reference PeayPeay, 1989).

In relation to the medical member's role as both witness and decision-maker, Rule 11 of the MHRT Rules presents particular problems. The rule requires the medical member “at any time before the hearing” to “examine the patient and take such other steps as he considers necessary to form an opinion of the patient's mental condition”. The medical member thus comes to the tribunal having conducted an examination of the applicant and presumably having formed a clinical opinion. He or she must then transform into a decision-making member of the tribunal: in the words of the Council on Tribunals, the medical member is “effectively a witness and a member of the tribunal deciding the validity of his own evidence” (Council on Tribunals, 1983, paragraph 322). From a legal point of view this presents significant problems in terms of impartiality or independence and fair process, and as a matter of professional ethics medical members are faced with a potential conflict between their duty as a doctor to the patient on the one hand and their additional duty to the tribunal on the other.

METHOD

The observations made in this article draw on data generated in the course of a project carried out in 1997-1998, with funding from the Economic and Social Research Council. The research was conducted in three stages: an analysis of 300 MHRT files from 1983 to 1997, data from which are not used in this article; the observation of 50 tribunal hearings and deliberations; and 37 interviews with tribunal members, patients' representatives and tribunal office personnel. The hearings that we attended during the second stage of the research were arranged through the tribunal offices in two regions and contained a mix of restricted and non-restricted cases: 18 restricted and 32 non-restricted cases. Information from the hearings was collected in relation to their three main components: preliminary meeting, oral hearing and deliberations. The data were organised using data collection instruments designed to allow us to record legal and procedural matters, the nature of the issues raised and the identity of the questioner and the respondent. Finally, 37 interviews lasting between 45 minutes and an hour were conducted over the telephone using an open-ended questionnaire. We make no claims for the statistical significance of these data and offer them simply to provide a picture of the tribunal decision-making that we observed.

BEFORE THE HEARING

It is customary for the three panel members to meet for about half an hour prior to the hearing. This enables them to identify any difficult issues likely to arise, to check the availability of reports and witnesses and to agree the order of proceedings. At this point the medical member will also give an account of his or her examination of the patient and it is here that the difficulties first begin to emerge. As a doctor, the medical member will have formed some clinical opinion as to the applicant's mental state; however, as a legal actor he or she should come to the tribunal hearing with a mind that is open to persuasion on the basis of the evidence presented and should not present his or her colleagues with a fixed clinical opinion before the hearing starts.

On appointment, all tribunal members are given an official guide to the work of the MHRT and their role within it (Department of Health, 1996). In its description of the functions of the medical member, this guide is careful to advocate limited disclosure at the preliminary meeting: “it is usually inappropriate for the medical member to give his opinion of the patient's mental condition at this stage”. However, the guide does comment that when it comes to identifying the “likely issues” that might arise during the hearing, the medical member will be in a particularly good position having already seen the applicant.

By encouraging medical members to limit their comments at the preliminary meeting to the likely issues rather than the patient's mental condition, the guide seeks to ensure that they perform appropriately as legal actors. Unsurprisingly perhaps, a number of the presidents implied that they approved of this advice and commented that they never asked the medical member for a clinical opinion at the preliminary meeting. Indeed, this was borne out by our observational data. However, even if they were not specifically asked to do so, the medical members at 50% of the hearings that we observed did give their opinion of the patient's mental state at the preliminary meeting; 50% also provided opinions of the patient's prognosis at that meeting. This suggests that the advice provided by the official guide may be unrealistic. It would appear to be hard in practice for psychiatrists to describe the “likely issues” relating to a patient in isolation from their professional opinion of that patient's clinical state. With considerable honesty, one medical member reflected the impossibility of giving facts free of opinion: “ it is important to avoid opinion at the preliminaries. I always avoid the word opinion”.

Clearly, there is confusion here between the professional expectations of law and psychiatry and both the formal structure and procedures of the tribunal and its practical reality. The tribunal rules require the medical member to examine the patient, yet the principles of fair procedure demand that the medical member come to the hearing with an open mind. It is a very fine, perhaps impossible, balance to strike.

In addition to the demands posed by legal values, medical members confront underlying ethical dilemmas as they transform from doctors to legal actors. A similar dilemma was emphasised by Stone (Reference Stone1994) in a slightly different context when he suggested that serious ethical difficulties can arise when a doctor employs interviewing skills to extract information from a patient that may be prejudicial if used subsequently in a formal legal setting. The medical member of an MHRT has to examine the patient in order to form an opinion of the patient's mental condition. In doing so the medical member acts as a doctor and uses his or her professional skills to acquire information from the patient that can ultimately influence the outcome of the tribunal decision. Yet at that point, the context being clinical, the patient has no representative present.

DURING THE HEARING

It is at the hearing that the medical members must deal with their dual role in a more public forum. During the hearing all three members of the tribunal put questions to the witnesses and consider the evidence. In offering advice to medical members at this stage, the official guide emphasises two issues that, while reflecting the formal legal position, serve also to illustrate the difficulties inherent in the medical member's dual role: the hearing is not a case conference and all evidence must be before the tribunal. We will take the second aspect first.

It is a fundamental principle of the common law rules of fair process that all evidence that is likely to influence the outcome of a decision be made available to all parties ( Mahon v. Air New Zealand, 1984). It is a principle that applies in all contexts where the rights of individuals are at stake. Thus, if the medical member of a tribunal takes a different view of the patient's condition than that taken by either the responsible medical officer (RMO) or any independent psychiatrist, and that view is likely to influence the outcome, then the medical member should reveal his or her opinion in the course of the hearing. This enables the patient or his or her representative to question that view. This principle was expressed by the courts soon after the introduction of the Mental Health Act 1983 ( R v. Mental Health Review Tribunal, ex p. Clatworthy, 1985), and is reflected in the advice given by the official guide. In the course of our research we recorded how this legal principle is dealt with in practice.

In interview, a number of presidents and patient representatives voiced concern about the role of the medical member as both witness and decision-maker. Of the fourteen presidents interviewed, nine said that there was a potential lack of fairness whereas four said that there was not. Of those who perceived a potential problem, most felt that the medical member was under an obligation to reveal his or her views at the hearing: “there may be a problem from a legal point of view. [The medical members] must reveal any knowledge to the hearing”, and “if the medical member has information from the patient they have to give it during the hearing”. Another president described one occasion “when the medical member refused to accept the diagnosis at the deliberations without mentioning it at all before”. In the event, according to the president, unfairness was avoided because the diagnosis was not central to the decision.

Of the thirteen representatives who we interviewed, nine felt that there was potential for lack of fairness in the medical member's role whereas three thought that there was not. Of those who feared unfairness, eight claimed that medical members never openly present their own views, thus “there is no way of testing what his view is”. In the face of this difficulty the representatives described how they had to imply the view of medical members from the direction of the questions they put to witnesses, particularly to the RMO.

These data suggest that the legal actors — the presidents and representatives — were generally alive to the problems of fairness and to some extent their concern was shared by the medical members themselves: “ I can't come in with concealed opinions because everything has to be on the table” and “if we have a view about the patient it is important that it is made clear during the hearing”. However, despite these expressed views, our observational data indicate that medical members rarely express a direct opinion during the hearing: in not one of the tribunals that we observed did the medical member directly express a clinical opinion during the hearing. Rather, those opinions had to be implied from the direction of the questioning in the way described by the representatives. The medical members may in practice be witnesses, as the Council on Tribunals suggests (Council on Tribunals, 1983, paragraph 322), but the patients do not have the opportunity to hear their evidence and nor do their representatives have the opportunity to examine them.

In the course of our research, however, we did observe many medical members make a conscious effort to encourage the patient to repeat the previous assessment interview. In this way medical members endeavoured to reproduce for the tribunal the evidence on which they had formed their professional opinions. At the hearing the medical member might typically bring the patient to difficult parts of his or her evidence by words such as “you told me when we chatted yesterday…”, and then encourage the patient to repeat what had occurred at the interview. For example, one medical member asked a patient about the relationship of his illness to a previous accident in order to reproduce a response made during the assessment that the medical member felt had demonstrated a lack of insight.

Although these replayed interviews can never fully replicate the original, they can go some way towards reproducing at the hearing those aspects of the evidence on on which the medical members have reached their clinical opinions. To that extent these exchanges in front of the tribunal serve to reduce the possibility of unfairness arising from the influence of undisclosed evidence. But, at the same time, they reintroduce the ethical dilemma referred to earlier: the medical member is using his or her clinical skills to obtain evidence for use in a legal forum. Although unlike the preliminary examination, on this occasion, at least, the patient is likely to be represented.

DURING THE DELIBERATIONS

Although our data suggest that medical members do not expressly reveal their clinical opinions during the hearings, it is evident that medical members do do so during the private deliberations of the panel after the hearing, although possibly not to the extent that might be expected, given their role as expert advisers to the panel. In 28% of the observed tribunals the medical member offered an opinion of the medical evidence during the deliberations; in 30% of the cases an opinion of the patient's mental state was offered; in relation to an opinion as to prognosis, the figure rose to 58%.

For the purposes of the present discussion, two issues emerge from this. In the first place, if the medical member takes a different view of the patient's condition to that taken by other medical witnesses but only reveals that opinion in the deliberations, there can be a breach of procedural fairness. In the course of our research we observed a number of cases where some degree of disagreement was revealed only in the deliberations. We offer here two of the more obvious examples. In one case the medical member expressed a strong opinion during the deliberations that the applicant's symptoms arose from the abuse of her medication rather than from a mental disorder. This opinion had not been expressly revealed during the hearing and ran counter to almost all of the evidence provided by witnesses. In fact, in this example the opinion of the medical member was supportive of the patient's discharge because, in his view, the patient was not suffering from a mental disorder under the Mental Health Act 1983. In the other case the medical member's opinion was adverse to discharge. In that case all the evidence given during the hearing supported discharge on the basis that the applicant was willing to remain as a voluntary patient in a staffed unit. Only the medical member felt firmly that the applicant was still a high risk, but this was not revealed during the hearing and came as a surprise to the other members when it was disclosed during the deliberations. In the event, the patient was not discharged.

The second issue to emerge from the finding that the medical member's opinion is revealed during the deliberations and not the hearing relates to the overall influence of the medical member. In the second case given above, for example, it would have taken a brave tribunal to discharge in the face of an adverse opinion from the medical member. Unfortunately, it is not possible from the bare figures alone to estimate accurately the influence of the medical members' opinion on tribunal outcomes; however, it is significant that in none of the observed tribunals was a decision reached in the face of opposition from the medical member. Further, from the interview data it was clear that the presidents perceived a danger that medical members might ‘take over’, and many lay members indicated that they had to guard against the medical member carrying too much influence. One lay member argued that it was dangerous when doctors let their opinions show during the deliberations, and she emphasised the tendency of doctors to favour medical interpretations of behaviour: “doctors can be very biased towards a medical answer, for example, regarding non-compliance with medication”.

However, if tribunals are to fulfil their role as guardians of the patient's right to be free from unjustified detention in hospital they must both apply the law impartially as legal actors and be rigorous in their testing of medical claims. Arguably, the presence of a medical member is central to a tribunal's ability to perform the second aspect of this role, but the doctor's influence must not be so great as to distort the former. It is a difficult balance to strike: the medical member must be rigorous without being dominating.

THE TRIBUNAL AS CASE CONFERENCE OR LEGAL ACTOR

We turn now to the nature of the task undertaken by the tribunal. Formally, the tribunal is required to determine a legal question: does the applicant fulfil the criteria for discharge? Is continued detention lawful or unlawful? It is a legal exercise governed by legal procedures and designed to protect the patient from unjustified detention. It is not, in the words of the official guide, “a seminar nor a case conference”. However, although it may be a legal exercise concerned to apply statutory criteria to an existing set of circumstances, those statutory criteria are highly dependent for their application on clinical judgment. A tribunal is under a duty to discharge a patient if it is satisfied “that he is not then suffering from mental illness, psychopathic disorder, severe mental impairment or mental impairment or from any of those forms of disorder of a nature or degree which makes it appropriate for him to be liable to be detained in hospital for medical treatment” (s72(b)(i)). Such a criterion can only be applied by reference to the opinion of doctors, nurses and social workers, and even then is open to wide interpretation (Reference Eastman and PeayEastman & Peay, 1999).

Thus, even though strictly the task is a legal one, it relies heavily on clinical judgment. This ambiguity is well illustrated by the official guide itself, which, having suggested that the medical member might think it appropriate to question the RMO about the patient's history, progress, treatment, prognosis and future care, reminds the medical member that the tribunal is not a case conference. The questions asked may be similar to those addressed by a case conference but formally, at least, the outcome sought is legal and not medical.

In the course of our research we observed 50 hearings and recorded each question that was put to witnesses both by tribunal members and by the patients' representatives. When analysing these data we were able to group the questions into 28 subject categories. Table 1 gives all the subject categories and the frequency with which each category of question was asked. The five most frequently asked questions all relate closely to the formal statutory task. The next two, which relate to after-care arrangements and the experience of leave or transfer, are also centrally relevant to the tribunal's need to determine whether treatment in hospital is ‘appropriate’ or ‘necessary’ for the patient's ‘health or safety or for the protection of other persons’.

Table 1 Frequency with which questions were asked

Subject of question	%
Diagnosis	90
Mental health symptoms	88
Cooperate with medication/treatment	88
Is section necessary	78
Non-statutory questions of risk/dangerousness	76
After-care arrangements	74
Leave, transfer	72
Behaviour on ward	64
Dose, type of medication	64
Treatment plan in hospital	64
Domestic, social skills	62
Future plans	60
Social life, family ties	58
Contact between patient and carers	52
Formal statutory questions	52
Prognosis	52
Incidents of inappropriate or aggressive behaviour	52
Side-effects	42
Overt questions about insight	40
Activities on ward	38
Substance misuse	36
Consequence if lapsed from medication	34
Index offence, incident	32
Previous discharge, transfer	30
Progress	28
Physical health symptoms	24
Facilities in community	16
Treatability	8

The remaining 21 categories of question range from those such as the statutory questions themselves, which are directly driven by the nature of the tribunal's formal task, to those such as activities on the ward, where any connection with the formal task is extremely tenuous. Between these two extremes, 16 categories of question address aspects of the patient's progress in hospital or ability to cope in the community, given the existing facilities, and three categories concern specific aspects of risk. Although many of these topics may closely reflect the issues typically addressed in a case conference, they also have relevance to the tribunal's formal task. The tribunal has to determine a legal question, but it is a legal question set in a health-care context and dependent for its interpretation on a clinical opinion. Thus, an examination of the statutory criteria can lead almost inexorably to a wider discussion of the patient's care and future plans. The picture revealed by our data merely reflects that reality.

In order to identify the medical member's role within this process of exchange between law and medicine, it is necessary to look more closely at the division of labour between the doctor and the president. Significantly, perhaps, the greatest difference between the input of the medical and the legal members occurs in relation to the statutory questions. It was the practice in a number of tribunals to organise the statutory criteria for discharge into four questions. In the course of our interviews with presidents and medical members it emerged that these questions were seen as a means of both explaining the complicated discharge criteria and identifying the legal issues. The questions were put to witnesses in 52% of all the hearings that we observed. In 48% of all cases they were put by the president, and in 6% by the medical member. They were never asked by the lay member.

Thus, in the vast majority of the cases where the statutory questions were used, the president took the initiative and the medical member was merely an observer. The position is very different in relation to type and dose of medication, which were addressed in 64% of cases. Such questions were put by medical members in 34% of all cases and by the presidents in 4%. Questions relating to the two most frequently covered topics - diagnosis and symptoms - were put by medical members in 60% and 68% of hearings, respectively, whereas the comparable figures for presidents were only 50% and 52%. When it comes to examining the nature of the patient's condition, which is an essential step in applying the legal criteria, the medical member takes the lead role.

In relation to the next three categories of question - voluntary compliance with medication, the necessity for a section and the nature of the risk - all three of which are also intimately tied to the statutory task, the figures are reversed, with the questions being put by the president more often than they are put by the medical member. But here, in contrast to the position in relation to the statutory questions, the medical member does deal with the questions in a significant proportion of cases: 46%, 40% and 32%, respectively.

To some extent these figures suggest that the division of responsibility between the doctors and the lawyers does follow disciplinary lines. But only in the case of the statutory questions (legal) and the type and dosage of medication (medical) is the boundary very starkly drawn. The doctors tend to concentrate on the direct medical questions but they certainly do not confine themselves to those alone.

DISCUSSION

Clinical Implications and Limitations