Mental Health Review Tribunals

Law and Human Behavior, VoL 5, No. 2/3, 198l

Mental Health Review Tribunals Just or Efficacious Safeguards?

Jill P e a y *

This paper is concerned with the operation of the Mental Health Review Tribunal system and summarizes the findings of an empirical investigation of tribunal decision making. The tribunals constitute an independent body which reviews the necessity for the continued compulsory detention of patients in psychiatric hospitals. The research focuses on the interpretation and application of mental health legislation by individual tribunal members and the manner in which these individual approaches are qualified in the group context. The study comprises three stages: a self-report questionnaire, a national statistical analysis of decisions, and a simulation study of decision making using a videotape of a hypothetical case. The research establishes considerable inconsistency in tribunal decision making at both an individual and group level and attributes some of this inconsistency to individually based factors isolated by the research. The findings further indicate that these factors remain influential within the group decision-making context.

INTRODUCTION This report summarizes research undertaken during the period from October 1976 to March 1980 into the operation of the Mental Health Review Tribunal system (Peay, 1980). These tribunals, established under the 1959 Mental HeaRh Act, review the necessity for the continued compulsory detention of patients in psychiatric hospitals. Their fundamental purpose has been succinctly described as "to act as a safeguard for the liberty of the individual and to insure against unjustified detention in hospital" (Aarvold Report 1973, Par. 35). Through an examination of the assumptions upon which the effective operation of the tribunal system is predicated, the research has attempted to ascertain the extent to which it is possible for these tribunals and other similar quasijudicial bodies to fulfill their prescribed roles and attain decisions that are both just and efficacious. *Centre for Criminological Research, Oxford. i61 0147-7307/81/0900-0161503.00/0 9 1981Plenum Publishing Corporation

162

PEAY

Fennell (1977) has traced the historical evolution of the tribunal system and analyzed some of the factors considered likely to influence tribunal decisions. As a result of his observations of 18 tribunal hearings during 1974-1975, Fennell argued that the legal concepts upon which tribunals might be assumed to be basing their decisions, those detailed in Section 123 of the 1959 Act, in fact form no more than "shorthand reference points," providing only organizing boundaries within which "a number of subordinate, non-legal, conceptual categories are indexed." Fennell suggested that tribunal decisions were being influenced more by "common sense" factors than by legal criteria, with the latter providing little indication as to the decision that would ensue in specific instances. The current research has provided an empirical investigation of this assertion, through a study of the factors that influence both the interpretation and application of mental health legislation by tribunal members and the manner in which these individual approaches are qualified by the process of decision making in a group context.

T H E F U N C T I O N , P O W E R S , AND O R G A N I Z A T I O N OF T R I B U N A L S When introducing the Mental Health Bill in Parliament, Walker Smith, the then Minister of Health, described the proposed tribunal system thus: Just as the right of application to a tribunal is one of the main safeguards against improper admission under compulsory powers, so it is a main safeguard against unduly protracted detention. (Hansard, 26 Jan. 1959).

However, it quickly became apparent that, as a safeguard against improper admission, tribunal review was likely to be ineffective since it could only take place after the implementation of compulsory procedures. Furthermore some compulsorily detained patients, for example those on short-term orders of up to 28 days duration (Sections 25, 29, 30, 135, 136 of the 1959 Mental Health Act), would not be eligible for tribunal review. For patients detained under Section 26 and Section 60 of the 1959 Act (treatment and hospital orders, respectively) the tribunal's powers as a safeguard against unduly protracted detention were to be limited to discharge and reclassification. For patients detained under Section 60 with a restriction order (Section 65) attached to their discharge, the tribunal's role was only to be advisory, with recommendations being made to the Home Secretary regarding the patient's discharge, transfer, or trial leave of absence. The tribunal's powers and their procedures, as governed by the Mental Health Review Tribunal Rules 1960, have been widely criticized (Gostin, 1975, 1977; Gostin & Rassaby 1980) and will not be reviewed further. However, the specific section of the 1959 Act that governs tribunal decisions (Section 123) will be given in full: Where application is made to a Mental Health Review Tribunal by or in respect of a patient who is liable to be detained under this Act, the tribunal may in any case direct that the patient be discharged, and shall so direct if they are s a t i s f i e d (a) that he is not then suffering from mental illness, psychopathic disorder, subnormality or severe subnormality; or

(b)

that it is not necessary in the interests of the patient's health or safety or for the

MENTAL HEALTH REVIEW TRIBUNALS

163

protection of other persons that the patient should continue to be liable to be detained; or

(c)

in the case of an application under subsection (3) of section 44 or subsection (3) of section 48 of this Act, that the patient, if released, would not be likely to act in a manner dangerous to other persons or to himself.

Subsection (c)relates to patients detained under Part IV of the Act with diagnoses of either subnormality or psychopathic disorder who are about to attain the age of 25 and to applications by a patient's nearest relative, when that relative has been denied the right to discharge the patient by the responsible medical officer. Applications under this subsection require that the tribunal applies to its decision the stricter criterion, outlined above, of "dangerousness." The tribunal system is organized on the basis of i5 regional panels comprising legal, medical, and lay members. For each region, one of the legal members is appointed as Chairman and assumes the responsibility for the organization and operation of the tribunal system in that region. In 1980, 482 individuals could be called upon to act as tribunal members. For each tribunal hearing, one representative of each of the three groups of members sit together, with the legal member acting as the President of the tribunal. The Mental Health Review Tribunal Rules (1960) provide considerable latitude for variation in the procedures adopted by different tribunals, although some features are common to all tribunals. Firstly, every tribunal receives a written report from the Responsible Authority, technically the hospital managers, although in practice normally prepared by the patient's responsible medical officer. This report details both the psychiatric case against discharge and gives an account of the facilities available for the care of the patient (usually, his likely future circumstances) if his order were to be discharged. Secondly, patients whose cases are under review are independently examined by the medical member of the tribunal prior to the hearing and are afforded some opportunity to put their case to the full tribunal.

TRIBUNAL DECISION MAKING Ideally, tribunal decisions should be attempting to reflect a balance between two competing demands. The first is that individuals should be treated equally before the law, with similar legal standards being applied to the assessment of cases, while the second is that the peculiarities of any particular case should be given due consideration. Traditionally, this has been achieved through the use of discretion, allowing decision makers sufficient flexibility to select the course of action that is most appropriate in the individual case considering both the pertinent facts and mitigating circumstances, within a framework of designated general principles that constrain decisions to accord with the policy of the legislature. Although tribunal disposal alternatives are limited, their decisions do comply with Reiss's (1974) definition of discretionary judgements as "the ability to make decisions and perform acts not subject to review by outside authorities." Since the tribunal has a duty to discharge patients unless prescribed criteria are met, it is necessary, in order to fulfill the minimum standards for just decisions, that these criteria are interpreted in a similar manner regardless of which members corn-

164

PEAY

prise the patient's tribunal or where in the country the patient is detained. Furthermore, it is to be hoped that every patient has an equal opportunity of benefiting from the discretionary decision to discharge permitted to tribunals. However, a number of factors seem likely to impede the attainment of consistency between tribunals in their application of Section 123. These relate firstly to the nature of the decision that members are required to make and secondly to the manner in which and by whom those decisions are made.

Factors Relating to the Decision At the time of the introduction of the tribunal system it was recognized that, as a safeguard, the tribunal could only operate effectively if presented with realistic decision alternatives. This required the provision of adequate community care facilities. In many parts of the country this community support has not been forthcoming, and neither does it seem likely to emerge in the foreseeable future. Therefore, it is to be expected that tribunal decisions will vary regionally because of the disparity in the availability of alternative facilities in the community. In Scotland, the Mental Welfare Commission, which has powers similar to tribunals regarding discharge, has recognized the adverse effects that inadequate community facilities can have on a patient's chances of leaving hospital. In a report of their work for the period 1972-1975, they stated For this reason, during the period under review, the Mental Welfare Commission have not discharged any patient on this ground. Too many patients remain in hospital because they have "no place to go" (1975, p. 5).

That these views are endorsed by some tribunal members is illustrated by the remark made during the course of the research by one medical member that "it's better for the patient to remain in hospital than to spend nights in third rate digs and days in the local shopping centre." It should similarly be noted that the nonavailability of places in local hospitals may influence the tribunal's willingness to recommend the discharge or transfer of patients detained in Special Hospitals (secure mental hospitals for patients with "dangerous, violent or criminal propensities," Section 97, 1959 Act). It is also suggested that, for a number of reasons, the legal criteria specified in Section 123 are likely to contribute to inconsistency in tribunal decisions. These criteria comply with Bean's (1975) description of "therapeutic legal rules" in that their emphasis on therapeutic goals requires considerable flexibility in their application. The flexibility is, in part, achieved through the use of terms that are either not defined (such as mental illness) or so loosely defined that they are capable of a number of interpretations (such as "in the interests of the patient's health or safety"). Flexibility is also achieved through the inclusion of criteria that only the medical profession can realistically be expected to assess, such as prognoses with reference to future treatment and the assessment of potential dangerousness. This results in an unnecessary reliance on medical integrity and a general emphasis on the medical approach. As Kittrie (1971) has suggested, when decisions are influenced by such therapeutic considerations there is likely to be a lessening of emphasis on legal procedures or on the control of medical discretion, since any action taken is justifiable on the basis that it is "in the patient's best interests." It should also be noted that the criteria in Section 123 do not specify objective measures of behavior that could be made more

MENTAL HEALTH REVIEW TRIBUNALS

165

accessible to nonmedical members, with the result that tribunals tend to rely upon such "evidence" as the patient's degree of insight into his condition. This is again an area where the medical member is presumed to be the expert. That such vague criteria do result in an application of the law at variance with its intended application has been illustrated by Bean's (1978) study of the operation of compulsory procedures by mental health professionals. Of 58 compulsory admissions included in the study, 31 were found to be made against the rules or spirit of the 1959 Act. Another factor likely to make the application of Section 123 by tribunals problematic is the essentially predictive nature of the decisions tribunals are required to make. Most other decision-making bodies are required only to assess the balance of the facts relating to past behavior. Tribunals, however, are required both to make assessments about a patient's potential for dangerous behavior, a notoriously difficult undertaking (Monahan, 1978), and to make a general assessment of the patient's likely behavior outside the structured hospital environment. Both of these assessments are necessarily speculative. The situation may be further confused by the tendency of the tribunal to address such questions as whether the patient, if discharged, will continue with his medication and, if not, whether this in turn would be likely to lead either to a relapse of his illness, if currently controlled by medication, or to an increase in his potential for committing dangerous acts. Furthermore, the quality of the decisions reached by tribunals may be affected by the necessity for members to base their decisions on evidence of the patient's conduct prior to admission or on evidence of his institutional behavior. It is rare for tribunals to possess information about the patient's behavior in the environment to which he is likely to be discharged. Indeed, taking decisions within the hospital environment may, in itself, affect the probability of the tribunal deciding that the patient is ill and in need of treatment (Rosenhan, 1973). The information on which tribunals have to base their decisions is wholly inadequate for predictive purposes. It might be suggested that the informational deficiencies, combined with the tendency to emphasize therapeutic rather than legal considerations, may unduly bias tribunals against the likelihood of reaching discharge decisions.

Factors Relating to the Decision Makers It has been suggested (Bell, 1970) that tribunals have a dual function to perform, in that they have to achieve a balance between the social interest in the liberty of the individual and the social interest in the treatment of mental disorder, and by implication, the protection for society this affords. Since members presumably reach decisions that comply with their understanding of the role they are required to fulfill, any variation that exists between members in their conception of that role is likely to contribute to variation in the decisions reached. It follows that if some members subscribe to a further aim of protecting society, as implied by Bell (1970), then the factors constraining their decisions will be subtly altered. Indeed, as Wood (1970) suggests, significant differences may exist anyway between the three groups of members in their conceptualization of their role even with reference only to the aim of "protecting the liberty of the individual." Since tribunal members are given no guidance about their role, and since there is not a body of case law relating to their decisions, nor even are

166

PEAY

the reasons for tribunal decisions published, it is possible that members may be unaware both of the function tribunals should fulfill and of the range of roles that other members may be attempting to fulfill. In view of the lack of feedback members receive about the outcome of their decisions and those of other tribunals, tribunal decisions may be influenced by inappropriate assumptions that may in turn contribute to inconsistency between them. The effective operation of the tribunal system is also predicated on a number of assumptions which, if erroneous, may similarly jeopardize the ability of tribunals to reach consistent decisions. The first of these assumptions is that, because of the professional and personal backgrounds of the individuals selected to become tribunal members, specific guidance or training for the task they are required to perform is superfluous since, as a group, they may have access to all the necessary relevant background information on which to base their decisions. The second assumption is that the combined decision of these individuals is informed by the relevant expertise that each brings with him. This assumption requires effective communication between the three members. The third assumption is that the tribunal's decision reflects equally the psychiatric, legal, and lay perspectives that constitute a balanced decision. This necessitates that the individual contributions are equally valuable and equally valued. It is apparent that the tribunal system makes considerable demands on the abilities of its members and, if its decisions are to reflect consistent legal criteria, places a particular burden on the role of the legal member. As such, variations between tribunal decisions may be partially attributable to variations in the quality of the members appointed. Other factors may contribute to a possible tendency for members to reach individually based decisions and to the selective processes that underlie all decision making. These include the relatively unstructured and ambiguous nature of tribunal proceedings, the inevitable presence of contradictory information from differentially valued sources, and the costs to the decision maker associated with mistaken decisions. A mistaken decision to discharge a patient may attract adverse publicity. A mistaken decision not to discharge will result only in the patient remaining in an environment which the decision maker may, in any case, regard as beneficial. The greater likelihood of tribunals making decisions of the latter type may be facilitated by the procedure whereby members are presented with the case against discharge, in writing, prior to the hearing. Evidence presented subsequently may be interpreted in the light of information, presumed accurate, contained in this report. Finally, as Cavenagh and Newton (1971) have illustrated, the individuals who comprise the tribunal membership are unlikely to be from the same peer group as the applicants. Consequently, the quality of justice they receive may be adversely affected (Schubert, 1965). Further sources of misconception may derive from some applicants' violent or criminal backgrounds. Any of these factors may systematically affect members' perceptions of realistic decision alternatives and thereby contribute to variations in the outcome.

Psycholegal Research and Decision Making Although it is accepted that review tribunals are in a particularly invidious position with regard to the attainment of consistent decisions, psycholegal research has

MENTAL HEALTH REVIEW TRIBUNALS

167

demonstrated that, even where specific guidelines exist, consistency in decision making is hard to achieve (Mitchell & Byrne, 1973). Amongst the legally irrelevant factors that have been identified as being likely to influence decisions in a significant and systematic manner, decision makers' attitudes have traditionally been given preeminence: It would be foolish to assert that when judges are engaged in solving problems all of their personal attitudes and values become dissipated in a bright glow of objectivity. (Fuller, 1966, p. 1619).

It has been suggested (Schubert, 1965) that, if attitudes do play a significant role in decision making, then individual decision makers may be capable of reaching decisions that are both internally consistent and rational, as defined by them, without necessarily being consistent with other decision makers in comparable cases. This attitudinal model has been developed to include the influence of individuals' perceptions of the objective facts (Hogath, 1971); information-seeking strategies (Wilkins & Chandler, 1965); the processes of selective perception (Carter, 1967); decision makers' conceptions of their aims and roles (Wheeler, 1968); and the acquisition of these aims and roles and their influence on the choice and interpretation of information (Hood, 1972). An important extension of this model would appear to be an analysis of the influence of the information decision makers possess which is relevant to the decision, namely, their beliefs about the facts and relevant background knowledge. These beliefs represent an individually based source of information additional to that with which members are formally presented. However, since tribunal decisions are effectively made in groups and not by individuals, it is possible that any individually based factors will have their influence moderated by the group decision-making process. Alternatively tribunal decisions may be systematically influenced by additional factors introduced by the very process of making decisions in groups, factors that are absent when individuals reach decisions alone (Moscovici & Zavalloni, 1969; Myers & Lamrn, 1976). Any attempt to understand inconsistency between tribunal decisions will have to take account both of these factors and of the plethora of legal and extralegal variables isolated by psy~ cholegal research as being likely to affect the tribunal's perception and assessment of a case. Research on Mental Health Review Tribunals

As Cavenagh and Newton (1970) have highlighted, the tribunal system has, like many administrative tribunals, remained relatively secluded from both the general public and social scientists. The exceptional inaccessibility of Mental Health Review Tribunals has meant that most of the research that has been carried out into their operation has been undertaken by individuals already associated with the system. Consequently, this research is largely of an observational nature. For example see Fleming (1963), Freer (1966), Cooke (1969), and Wood (1970), although the two latter contributions did adopt a more analytical approach. A review of the tribunal's legal provisions was undertaken by Stephens (1968). Cavenagh and Newton (1970, 1971) carried out a postal survey of members in the West Midlands. Their articles

168

PEAY

highlight both recruitment procedures and the variation in the qualifications and experience of the lay membership. The only completed work of any substantial research value is Greenland's (1970) statistical analysis of 1,250 applications which he supplemented by both observational work based on 180 tribunal hearings and follow-up of the 50 patients discharged. Greenland noted the occurrence of significant variations between tribunal regions in their discharge rates and highlighted the tendency for representation to increase a patient's chances of being discharged. A similar finding resulted from the joint MIND/National Council for Civil Liberties study of representation carried out in 1967. Finally, the preliminary findings of Hepworth (1980), resulting from his study of the formal assessment of dangerousness by tribunals held at Rampton Special Hospital, has indicated that members' concepts of fairness and the desire to limit retribution may influence discharge decisions when dealing with offenders convicted of sexual offenses or arson. Hepworth has also reported the relative lack of conflict that members experience when reaching decisions. Although research on tribunals has been limited in its scope and has failed, as yet, to provide an adequate understanding of the bases on which tribunal decisions are made, it has indicated that tribunals are unlikely to apply Section 123 mechanistically when reaching decisions.

THECURRENTSTUDY The research undertaken focused on the decision-making process in an attempt to establish the degree of consistency in the interpretation and application of Section 123 of the 1959 Act between individual tribunal members, between the three groups of panel members, and between different tribunals. The research also aimed, in the light of previous research, to explicate those factors that might be contributing to any disparity established, and subsequently to make recommendations for a clearer and more effective basis for those decisions. It should be stressed that such an undertaking was only possible as a result of the invaluable cooperation received throughout the course of the research from the Department of Health and Social Security, the Lord Chancellor's Department, the tribunal Chairmen, the tribunal officers, and particularly that of individual tribunal members. The willingness of the members to participate in the research was felt to be indicative of their general approach to their role--that of thoughtful, dedicated, and caring individuals. Any criticisms that have resulted from the research are not directed at the members as a whole, but at the tribunal system and its expectations of the membership. Following a preliminary pilot study, the research was undertaken in three major stages: a self-report questionnaire administered nationally to the tribunal membership, a retrospective statistical analysis of tribunal decisions during the period 1974-1978 inclusive, and an experimental study of individual and group decisionmaking processes using a videotape of a hypothetical application. The pilot study, undertaken in the West Midlands region and at Rampton Special Hospital, involved observations of 11 tribunal hearings and semistructured in-

MENTAL HEALTH REVIEW TRIBUNALS

169

terviews with nine representatives drawn from the three groups of members. Although highly subjective, this preliminary work provided a better understanding of the nature and variety of tribunal procedures and of the factors, as defined by individual members, that appeared to be influencing their decisions. The interviews particularly helped with the design of the standardized questionnaire, with many of the attitude statements included in the questionnaire being based on comments made by individual members. SELF-REPORT QUESTIONNAIRE STUDY The questionnaire aimed to establish the following: (i) The nature and range of members' attitudes, opinions, and knowledge about a variety of issues relevant to tribunal decisions, including their understanding of the tribunal's role and of the criteria underlying their decisions. As such, the primary aim of the questionnaire was to establish the nature of the information that members take with them to the tribunal setting. This information will remain relatively stable across different tribunal hearings. (ii) Whether any significant differences existed in the nature of the information possessed by representatives of the three groups of members or in their attitudinal approach. (iii) Whether any systematic relationships occurred between individual members' factual beliefs and their attitudes concerning the above issues. The questionnaire was in two parts. The first part comprised 71 attitude statements which members were required to rate on an agree/disagree basis. The attitude statements related to topics such as mental disorder, psychiatric treatment, the role of the tribunal system, and "dangerousness." The second part of the questionnaire required members to estimate mental health statistics, to weight evidence, and to interpret terms. These questions were largely of a forced-choice format. With the support of the Regional Chairmen, the questionnaire was sent by post to members in 13 of the tribunal regions. A response rate of 54% was achieved, with good distribution both by region and by member groups. The 32 legal, 102 medical, and 66 lay members who completed the questionnaire had a total of 9,781 tribunal experiences between them. This was felt to be an adequate sample from which to draw conclusions.

Principal Findings Rather than attempting to summarize individual and group responses to the attitudinal section of the questionnaire, some of the more striking findings that resulted from a factor analysis (J factor with unweighted least squares; see Nie et al., 1975) carried out on the responses to the 71 attitude statements will be presented. The factor which accounted for the largest percentage of variance in the data (16.4%) was interpreted, on the basis of the statements contributing to it, as representing members' overall attitudes toward discharge from psychiatric hospitals (factor 1). Significant group differences (p = 0.001) were established, with the medical group demonstrating attitudes relatively more disposed toward discharge and with the Iay

170

PEAY

group having attitudes least favorably disposed toward discharge. Other factors which emerged from the factor analysis indicated that, as a group, the medical members generally had more innovative and enlightened views with regard to stereotypes about mental disorder, the mentally abnormal offender, and treatment under secure conditions. The nonmedical members were significantly more likely to conceptualize the mentally disordered as dangerous, impulsive, showing little insight, and as having disabilities of a relatively permanent nature. Further analyses on an individual basis demonstrated that members who held these negative attitudes toward mental disorder were significantly more likely to be against discharge from psychiatric hospitals (as measured by their scores on factor 1). However, it was apparent that some negative beliefs were shared by all three groups. For example, patients made the subject of hospital orders were conceptualized as being most likely to have committed offenses of a violent or sexual nature. However, in reality the type of offenses committed by patients with Section 60 orders (with or without restrictions) do not differ in distribution from those committed by mentally normal offenders. Acquisitive offenses and criminal damage are the types of offense most commonly committed by both groups. It is conceivable that a significant proportion of tribunal applicants may be viewed unfavorably by some tribunal members on the basis of factually mistaken beliefs. Members generally expressed satisfaction with the operation of the tribunal system. For example, 76% agreed with the statement that "The M.H.R.T. provides an adequate safeguard for the liberty of detained patients." Factor analysis demonstrated that it was the legal group who felt most satisfied about the operation of the tribunal system, whilst it was the medical group who most favored changes in the system, particularly advocating a need for greater professionalism amongst the membership. It was felt that some of the members' negative attitudes toward the mentally disordered and their satisfaction with the effectiveness of the tribunal system could result from their limited experience with the mentally disordered. Responses to the questionnaire indicated that 51% of the legal group had either never visited a psychiatric ward or had only had a "conducted tour" of a hospital. It is desirable for members to have relevant experience of hospital conditions, as advocated by the Mental Health Review Tribunal Procedures Committee (Department of Health & Social Security, 1978), if their negative beliefs are mediated by the limited nature of their experiences. It is possible that the medical group's advocacy of greater professionalism indicates a recognition of the importance of their fellow members being more aware of hospital conditions. Response to section 2 of the questionnaire indicated that members generally felt confident about their use of the term "dangerous," although it was evident from the range of behaviors included in members' definitions of "dangerous behavior" (e.g., confusional states, embarrassing or incompetent behavior) that not all members were applying the term in a similar or strict manner. Not only did members express confidence in their own use of the term, but the nonmedical members also expressed a degree of confidence in the ability of the medical profession to predict the occurrence of dangerous behavior. This degree of confidence was not shared by the medical group. In many ways, this response typified the incongruity between the beliefs of the

MENTAL HEALTH REVIEW TRIBUNALS

171

nonmedical and the medical members, with the former expressing confidence in the medical approach to treatment and in the abilities of the profession to deal with such problems as psychopathic disorder, whilst the medical profession had less regard for their own abilities. It should also be noted that, when asked to rate the order of importance of the evidence influencing the tribunal's decision, all three groups of members emphasized the independent medical evidence in preference to the interview with the patient. This was a further illustration of the preeminence of the medical approach. It was also apparent that members generally drew little distinction between the types of dangerous behavior that would satisfy the statutory criteria included in subsections (b) and (c) of Section 123 [subsection (c) refers to a narrower range of behavior than subsection (b)]. Members' responses also indicated that considerations of "dangerousness" generally informed their decisions rather than being limited to specific cases. Finally, it was evident that, in cases where a patient had committed a dangerous act in the past, members became more conservative in their decisions; 42% agreed with the statement "Patients who have committed dangerous acts should be released from secure conditions only when it is absolutely certain that they no longer represent a danger." This is an extremely cautious approach. Although members' interpretations of the term "necessary in the interests of the patient's health or safety" generally stressed health and safety considerations, some members did seem to be interpreting this phrase as including welfare considerations. This may be illustrated by the paternalistic nature of some of the members' approaches, including one who commented just being there (in hospital) is often better than being harrassed by the complicated rules our society has evolved for the population to conform to both economically and socially.

Another expressed a similar view rather more succinctly: "Do I have the right to discharge a man to be unhappy?" Questions designed to elicit members' conceptions of their role indicated that they were primarily divided into two groups, with 56% selecting an option that stressed the necessity of reaching a balance between the safety of the public and the liberty of the individual and 35% emphasizing the protection which the review tribunal afforded the patient against unjustified detention. However, responses to another question indicated that even those members who chose the latter option had some sympathy for the former approach, since only 5% of the members disagreed with a statement that the tribunal had a dual role to protect both the liberty of the individual and society. Even given this flexibility in their conception of their role, it is unlikely that all the members would have endorsed those individuals who commented "liberty is not my major objective--the safety of the public comes first" and "the question ! ask myself is 'is this man fit to live next door to my children?' " Members did not appear to conceptualize their role as that of a judicial body, but rather as an informal reviewing panel intended to assess the most appropriate course of action, taking into consideration their conception of the patient's "best interests." In order to achieve this the tribunal rules were apparently frequently disregarded or circumvented. Indeed, 80% of the members agreed with the statement that "Tribunals

172

PEAY

should consider the spirit rather than the letter of the law when making decisions." Of those members disagreeing, many were vociferous in their objections: when questions of personal liberty and the safety of others are concerned, parliament should specify the standards . . . individual members' interpretations of the "spirit" vary widely.

One of the most disturbing findings to emerge from the questionnaire study resulted from a question concerning members' understanding of their specific powers under Section 123 of the Act. Only 29% were aware of their permissive power to discharge a Section 26 patient in any circumstances, as opposed to the circumstances under which they have a duty to discharge patients. It should also be noted that, of the legal members, less than half of them (48%) answered this question correctly. Taking into account the frequency with which members sat on tribunals, it was calculated that patients have a one in three chance of appearing before a tribunal where none of the members are aware of this permissive power. Thus, not all patients will have an equal chance of benefiting from discretionary practices. A combination of members' apparent ignorance of the legal criteria for their decisions and their willingness to apply the "spirit" rather than the letter of the law would seem likely to contribute to inconsistency between tribunal decisions. This inconsistency will be enhanced by the absence of any alternative formalized criteria for those decisions. Unfortunately, the response to the question concerning members' familiarity with Section 123 was not atypical of the degree of awareness demonstrated by members in areas where they might have been expected to be better informed. Only 31% of the membership were familiar with circumstances under which compulsory treatment could legally be given, with only 36% of the medical group answering this question correctly. A question designed to elicit members' understanding of Rule 13 in the tribunal rules (the circumstances under which confidential information can be disclosed to the patient) was answered correctly by 71% of legal members and 45% of medical and lay members. In contrast, a question concerning whether aftercare could be enforced by the tribunal was answered correctly by 100% of the legal members, 93% of the medical members, and 75% of the lay members. It might be suggested that, because aftercare is a topic frequently discussed when making tribunal decisions, members have acquired knowledge concerning the boundaries of what is legally permissible. In contrast, although confidential reports are frequently made about applicants, tribunals rarely discuss whether or not this information should be disclosed. Therefore, many members, as illustrated by their responses, may remain unaware of their powers on this topic. The disparity between the legal members' understanding and that of the nonlegal members is interesting since it suggests either that these issues are not discussed and do not generally inform tribunal decisions or that any discussion of them does not result in nonlegal members improving their understanding of the relevant legal issues or challenging preconceived ideas about them. Support for this latter argument, that of ineffective communication between the members, was highlighted by interviews in the pilot study. One of the legal members explained that he took great care at every hearing to ensure that his fellow tribunal members understood the relevant subsection of Section 123 to be applied. However, nonlegal members who had sat with him had apparently not benefited from his verbal explanations.

MENTAL HEALTH REVIEW TRIBUNALS

173

In order that comparisons could be made between the members' overall factual understanding, a crude measure was developed that rated individual members' performances on the factual sections of the questionnaire (known as a member's '~ al score"). A similar measure was derived from responses to those questions that had legally appropriate answers ("legal score"). For the factual score, no significant differences in performance were established between the three member groups, nor between members with special or local hospital experience, nor between members who had a great deal of tribunal experience and newly appointed members. Similarly, there was no evidence that the legal members were generally any more accurately informed about the relevant legal issues than either the medical or the lay group. One of the few significant correlations that was established was between a member's factual score and his conceptualization of the tribunal's role. Those stressing the safeguard option rather than the balance option were significantly (p = 0.005) better informed. There was also a significant correlation between a member's factual score and his score on factor 1, which measured individual attitudes to discharge. Those members more in favor of discharge were significantly (p = 0.009) better factually informed than members whose attitudes were generally against discharge. Finally, the questionnaire demonstrated that members were ill informed both of the outcome of tribunals they had attended and of the actions of other tribunals. When asked to estimate statistics concerning discharge, members tended both to underestimate the numbers of patients discharged by tribunals and to underestimate the likely success of those patients in the community. As such, it might be suggested that members have unjustifiably low expectations about the patients they encounter. This may help to explain some members' apparent satisfaction with the operation of the tribunal system. Underestimating patients' success in the community may also lead the tribunal to believe that their decisions are more risky than the objective evidence suggests. The importance of accurate feedback for informed decisions cannot be overemphasized. It is also interesting that those members who overestimated the numbers of patients discharged had attitudes significantly (p = 0.001) more in favor of discharge than members who underestimated discharge rates. Since the overestimating/positive attitude group were heavily weighted with medical members, it might be suggested that a member's actual discharging experiences constituted an intervening variable. The medical members would have had additional discharging experience outside the tribunal system, and their knowledge of previous successful discharges might have contributed both to their more positive attitudes toward discharge and to their tendency to overestimate tribunal discharge rates. Thus, there is evidence that the manner in which members' beliefs about the facts vary in relation to the objective nature of these facts may be affecting discharge decisions in predictable ways. Certainly, there is strong evidence that these systematized belief patterns a r e related to members' attitudinal preparedness to make discharge decisions. Thus, the questionnaire study demonstrated that some members took with them to the tribunal setting certain beliefs that might be considered not only inaccurate but also inappropriate. The additional demonstration that some members from each of the three groups were inadequately informed about issues considered fundamental to the tribunal's decision and that individual members from the professional groups did not appear to be in possession of information that they might reasonably be expected

174

PEAY

to have does cast considerable doubt upon the likely effective operation of the tribunal system. The questionnaire also demonstrated that significant attitudinal differences existed both between members as individuals and between members grouped by profession in their approach to discharge and that these attitudinal differences were significantly correlated with the nature and correctness of members' beliefs about factual issues related to their decisions. Finally, the questionnaire cast doubt upon the effective communication of information between the three members, both because of the existence of significant group differences in the information possessed about issues that might be expected to be discussed at tribunal hearings and because of the failure to demonstrate any improvement in members' factual knowledge as a result of attending tribunals. As such, the tribunal does not appear to be either an effective self-educating process or an effective basis on which to take informed group decisions. Since letters received from some of the members indicated that completing the questionnaire had reduced their selfconfidence in their abilities as tribunal members, it is conceivable that members are simply unaware of deficiencies in their knowledge.

R E T R O S P E C T I V E S T A T I S T I C A L A N A L Y S I S OF T R I B U N A L DECISIONS This stage of the research involved analyses of tribunal decisions made in 13 regions during the period 1974-1978 inclusive (a total of 4,218 decisions) and aimed, through a breakdown by participating members, to establish (i) whether variations in discharge rates occurred either on a regional basis or between members within a region; (ii) whether any variations in discharge rates related either to the theoretical measures of attitude to discharge or to any of the other individually based factors produced by the questionnaire; (iii) whether there was statistical evidence that any of the individual factors isolated could be influential in a group setting; (iv) through an analysis of any available follow-up data, whether empirically successful decision makers shared any skills or characteristics capable of identification.

Principal Findings For the period under review the analysis indicated that, nationally, patients had a relatively low chance of being discharged from their orders by a tribunal. The analysis also established the existence of statistically significant variations in regional discharge rates, thus confirming Greenland's (1970) findings (see Table 1). Similarly, significant variations occurred between the Special Hospitals in their tribunal discharge rates (Broadmoor 11%; Rampton 24%; Moss Side/Park Lane 13%). Thus, patients' chances of being discharged by a tribunal depended, to some degree, on where in the country they were detained. More surprisingly, the analysis established the existence of statistically signifi-

MENTAL HEALTH REVIEW TRIBUNALS

175

Table 1. Regional Discharge Rates for Nonrestricted Cases 1974-1978 Trent Mersey Oxford N.W. Thames S.E./S.W. Thames N.E. Thames/E. Anglia Wessex Northern S. Western Yorkshire N. Western Average discharge rate

24% 13% 11% 11% 27% 18% 20% 11% 17% 30% 13% 17%

cant variations between members in their individual discharge rates. This is illustrated by a comparison of individual discharge rates between two medical members sitting in different regions with comparable overall discharge rates. The members were likely by chance factors alone to have had an equal opportunity to experience discharge decisions, but one member had attended 40 hearings and never been involved in a discharge decision, whilst the other had attended 55 hearings and had been involved in 25 discharge decisions. Further analyses indicated that, in the majority of regions, most discharge decisions were apparently being taken by only a small proportion of the membership, with the majority discharging at below the average rate in those regions. In two regions this trend was reversed, with a small number of members apparently being more likely to influence the decision against discharge, and the majority discharging at slightly above the average rate. Thus, the initial analyses appeared to indicate that variations in individual members' discharge histories were unlikely to have been produced by chance factors alone and appeared to indicate the influence that some members could have upon group decisions, By a comparison of the discharge rates from Rampton and Moss Side/Park Lane Special Hospitals, where it was possible to control for some of the variation resulting from regional factors and from different discharge policies operated by the resident psychiatrists, it became possible to attribute some of the variation in the two tribunal panels to real differences in the criteria applied in their decisions. Although at this stage of the research it was only possible to suggest that members sitting at Rampton were being less cautious in their discharge policies than members sitting at Moss Side/Park Lane, further evidence emerged from the final part of the study, in which members at both hospitals were required to assess the same case. This indicated that members at Rampton would be prepared to take discharge decisions in cases that members of the Moss Side/Park Lane panel would consider too risky to discharge. Since the follow-up of patients discharged by tribunals from these two Specia| Hospitals indicated that their success rates in the community were comparable, it is conceivable that the tribunal at Moss Side/Park Lane were being unjustifiably cautious in their decisions.

176

PEAY

Table 2. Change Rates in Restricted Cases 1974-1978 for the Special Hospital Hospital Rampton Moss Side/Park Lane Broadmoor Total,

% Change

% Rejection

45% 25% 20% 30%

62% 59% 48% 56%

An analysis was also undertaken of the tribunals' advice to the Home Secretary in restricted cases for the Special Hospitals during the period 1974-1978 (see Table 2). The percentage change rates given indicate the frequency with which the tribunals advised conditional discharge, transfer, or removal of restrictions in Section 65 cases. The frequency with which that advice was rejected by the Home Secretary is also given. Comparing these figures with those given earlier of the discharge rates at these hospitals, it is apparent that in each of the hospitals the tribunal members have doubled the rate at which they discharge patients when giving advice to the Home Secretary. The relative rejection rate by the Home Secretary is felt to reflect this disparity. The overall rejection rate casts some doubt upon the value of tribunal review in restricted cases. The analyses of individual member change rates permitted a comparison with their individual discharge rates in nonrestricted cases. Individually, members appeared to be employing internally consistent policies across different types of cases, since those with an above average discharge rate also appeared to advise change to the Home Secretary significantly more frequently (p = 0.001). However, further analyses indicated that this stability in individual approach was only statistically significant for the legal and medical members. This ability to maintain individual consistency across cases indicates that the medical and legal groups were likely to be having a more influential effect on the direction of the group decision. In an attempt to establish the nature of the factors contributing to the variation between individual members in their discharge rates, intraregional correlational analyses were undertaken between individual discharge rates and three of the factors isolated by the questionnaire study: attitude to discharge (score on factor 1), factual score, and legal score. Significant positive relationships were established between each of these factors and a member's individual discharge rate (p = 0.05, 0.001, and 0.004, respectively). This demonstrated that factually informed members and those with a positive attitude to discharge had a significantly larger number of discharging experiences than either their less factually informed peers or members with a negative approach to discharge. More detailed analyses indicated that the relationship between individual factors and individual discharge rates was most prominent for the legal group. Indeed, it was only the legal members for whom a significant (p = 0.004) positive group relationship was established between attitudinal approach to discharge and individual discharge rates. Since the legal members preside over the tribunal proceedings, it would seem reasonable to suggest that their special position enables such attitudinal factors to play a more significant part in the decision than would be possible for either of the nonlegal groups.

MENTAL HEALTH REVIEW TRIBUNALS

177

It was also demonstrated that members with an extreme attitudinal approach, as reflected in their scores on factor 1, were more likely to have either very high or very low discharge rates during the period 2974-1978. This relationship was again most significant for the legal members (p = 0.002). Analyses by type of region (Special Hospital region/local hospital region) indicated that in the detailed analyses of the relationship between discharge rates and legal/factual scores, the legal group were again the members for whom the most significant relationships were demonstrated. It should be noted that no relationship was established between lay members' legal knowledge and their discharge rates in local hospital regions. Thus, the analyses have shown that members with extreme views, and some member groups, appear to be having a consistent and influential effect on the nature of the tribunal's decision. It was recognized that the individual factors isolated by the questionnaire study were unlikely to exert an independent influence, since they obviously exist in conjunction for any individual. Therefore, analyses were undertaken to investigate the effect of the combined influence of an individual's attitudinal approach and his factual understanding on discharge rates. Table 3 shows that highly significant relationships have again been established between the combined individually based factors and individual discharge rates. However, analyses by member groups established significant relationships for the legal (p = 0.05) and medical members (p = 0.02), but not for the lay group. These findings were felt to be indicative of the relative powerlessness of the lay members in the group decision-making process, in comparison with the medical and legal members who were seemingly in a better position to influence the group decision in a direction that accords with their personal view of the case. It is recognized that since the questionnaire study postdated the period for which tribunal decisions were analyzed it is possible that members' questionnaire responses could simply reflect their disparate experiences. Nevertheless, the significance and nature of these individual relationships appear to indicate that the reverse is more likely to be tenable, namely, that individual factors influenced decisions. Since the individual discharge rates derived from the study will have been moderated by the influence of the members with whom those decisions were made and yet were still in many cases significantly related to individually based factors, it is felt that they reflect the underlying strength of the relationship between individual factors and discharge rates. The ability of individual members to exert a significant, systematic effect on group decisions may reflect the active nature of the relationship between discharge

Table 3. Discharge Rates by Combined Member Characteristics Combined attitudes/facts

Discharge rates Above average

Below average

33 10

16 11

Positive Negative X2 = 14,83 p = 0.0005 (I tailed)

178

PEAY

rates and individual factors, rather than those factors resulting passively from disparate discharging experiences. Finally, the analysis of the tribunal files permitted some inferences to be drawn about factors associated with "successful discharge." Although it is accepted that an appraisal of discharge outcome is fraught with difficulties, crude measures of a patient's success could be derived from the files. A patient was considered to be a "success" if he did not reappear in court and was not readmitted to hospital within a two-year follow-up period. These measures indicated high levels of success for patients discharged by tribunals from Rampton and Moss Side/Park Lane hospitals (59% and 63%, respectively). Further analyses demonstrated that more experienced tribunal members at Moss Side/Park Lane and those with an above average discharge rate at Rampton were more likely to be associated with "successful" discharge outcomes (p = 0.01 and 0.05, respectively). Since the individual factors isolated by the _questionnaire were not found to be associated with "successful" discharge outcome, it is suggested that the ability to make such discriminatory decisions, rather than being inherent, may be acquired by individuals as a result of their tribunal experience. If this hypothesis could be validated further, it would indicate the desirability of members becoming more experienced in their tribunal work. For example, lay appointments should not only be for limited periods. I N D I V I D U A L A N D G R O U P D E C I S I O N - M A K I N G P R O C E S S E S IN A HYPOTHETICAL CASE This controlled study of decision-making processes utilized a film of a hypothetical application. By adopting a simulation approach, differences between tribunal decisions resulting from variations between regions or between cases should be eliminated. This aimed to establish (i) whether, given an identical case to assess, variation occurs either between individual members in their decision preferences or between group decisions; (ii) to what extent any individual variation established is (a) random or (b) associated with either the individually based factors isolated by the questionnaire or with a member's individual discharging experiences; (iii) the stage at which effective decision making takes place and the aspects of the information on which it is based; (iv) how any individual variation in decision preference is resolved in the group decision-making process, and whether each of the three individual decisions is reflected equally in the group decision. The hypothetical case members were required to assess concerned an application by a patient detained on a Section 26 order in a local hospital with a diagnosis of paranoid schizophrenia. Members who agreed to take part in the exercise were sent all the relevant documentation concerning the hypothetical case and, following the assessment of a real application, viewed a videotape of the hypothetical application with two other tribunal members. At various stages during the film members were required to complete a decision-making schedule that indicated both the direction of their individual decisions at that stage (discharge/no discharge) and the confidence they had in those decisions. At the end of the film, members were required to reach a

MENTAL

HEALTH

REVIEW TRIBUNALS

179

group decision. Subsequently, two changes were made in the case. The first altered the decision from a subsection 123 (b) application to a subsection 123 (c). The second provided further evidence concerning the patient's likely dangerousness, and, like the previous exercise, was designed to focus the tribunal's discussion on issues related to dangerousness. Again, members were required to reach individual and group decisions. The case was designed to ensure that there was no obvious decision, with the evidence capable of supporting either course of action. The case was assessed by 13 tribunals at Special Hospitals and 7 at local hospitals. The absence of a lay member at one of these hearings resulted in a total of 59 members completing the three exercises. Principal Findings Table 4 shows that, even given identical facts to assess, the exercises resulted in considerable variation in decision outcome, both on an individual basis and between different tribunals. For example, in exercise 1, 5 legal members favored discharge whilst 15 were against and in exercise 2, 6 tribunals reached discharge decisions whilst 13 were against discharge. It should also be stressed that this variation reflected confidence levels that ranged in exercises 1 and 2 from 100% discharge to 100% no discharge and in exercise 3 from 80% discharge to 100% no discharge. Thus, not only were individual members and those in groups reaching disparate decisions, but members were also expressing high levels of confidence in the appropriateness of those decisions. During the course of the final stage of the research, the film of the hypothetical case was viewed by a number of individuals who were not members of the tribunal system (n = 60). These individuals fell into two groups, those having some experience in mental health and those having no special experience (the general public group). This provided an opportunity for gross comparisons to be made between the standards applied by tribunal members and those of other individuals. The average discharge rate of 27% reached by the tribunal members was considerably lower than the rates of the mental health and general public groups (74% and 67% discharge respectively). Thus, there is evidence not only of variation in the criteria as applied by tribunal members, but also that the degree of caution reflected in their decisions would not necessarily be endorsed by individuals not associated with the system. Analysis of the variation in individual members' discharge preferences demonTable 4. Decision Outcome in Hypothetical Case by Tribunals and Member Groups Exercise a 1 Section 123(b)

Tribunals Legal members Medical members Lay members

2 Section 123(c)

3 Danger

Dis

No Dis

D.K.

Dis

No Dis

D.K.

Dis

3 5 5 6

17 15 15 13

0 0 0 0

6 7 7 6

13 12 12 12

1 1 1 1

2 2 3 2

aDis = discharge decision; No Dis = n o discharge; D.K. = no decision reached.

No Dis / D.K. 18 18 17 15

: : : :

0 0 0 2

180

PEAY

strated that this was not random, but was significantly related to those individual characteristics identified in the previous stages of the research. Members with high factual scores and a positive approach to discharge were more likely to discharge in the hypothetical case. When these factors were combined it was established that those members with low factual scores and a negative approach to discharge did not discharge in any of the three exercises. In addition, those with an above average discharge rate (as derived from the file analysis) were significantly more likely (p = 0.05) to discharge in the hypothetical case than those with a below average discharge rate. Thus, it is apparent that individual variables are responsible for some of the variation in decision outcome that occurs in the hypothetical case. The demonstration that ambiguity in the case is resolved in individually predictable ways lends further support to the proposition that decision making is an active process. The variations in individual members' discharge histories revealed by the file analysis are not likely to be the passive result of variations between the cases, but represent the imposition of individual preferences in the decision-making process. The analysis of members' confidence ratings on their decisions at various stages during the presentation of the evidence enabled some inferences to be drawn about those aspects of the evidence that had most influence on members' preferences. It was apparent from this analysis that the greatest influence on the eventual direction of members' decisions was effected by the Responsible Authority's report members received before they attended the mock hearing. Only 8 members out of a sample of 50 changed the direction of their decision during the period after having read the reports and before being required to reach a decision based on all the evidence. Thus, members' initial responses were excellent predictors of the direction of their eventual decision (p = 0.0005). With the exception of the evidence given by the patient's responsible medical officer which produced a clear swing in members' decisions toward "no discharge," all the evidence given by subsequent witnesses produced confidence swings in both directions with comparable strength. This suggested that evidence was capable of being selectively perceived to support either discharge or no discharge decisions. Therefore, it appeared that members had difficulty assessing the evidence objectively and tended to perceive any new information in accordance with their initially preferred view. Overall, there was little change in the direction of decisions during the first exercise. Similarly, the introduction of new evidence in exercises 2 and 3 resulted in only a small percentage of members (14% and 13%, respectively) changing the direction of their decisions. The demonstration that members actively select and interpret information in accordance with their individual predispositions lends further support to the proposition that the relationship between these individual variables and members' actual discharging behavior is causal in nature. Members' comments during the discussion phases of the exercise starkly illustrated the manner in which identical evidence could be interpreted to support diametrically opposed positions. For example, the evidence that the patient carried furniture into the garden whilst it was raining before attempting to set it alight was interpreted by some members to support a discharge decision (the patient had not set the furniture alight inside the house and was therefore not a dangerous arsonist) and by others to support the opposite position (the fact that it was raining made the patient socially irresponsible for taking furniture outside the house).

MENTAL HEALTH REVIEW TRIBUNALS

181

Similarly, members' comments revealed considerable variation between them in their assessment of the individuals giving evidence and of the value of that evidence. The contrast between their assessments of the responsible medical officer and the social worker are particularly interesting. Although the treatment proposed by the responsible medical officer was medically questionable (e.g., he indicated the possibility of using electroconvulsive therapy for the treatment of paranoid schizophrenia), the majority of members appear to have been influenced by the manner in which he gave his evidence, which was both authoritative and empathetic. In contrast, the social worker's oral evidence provided a strong case for discharge. However, his presentation and appearance was poor and members apparently felt his report to be inadequate, with the result that his evidence produced no significant unidirectional effect on members' decisions. Members' responses to the hypothetical case also showed considerable inconsistency in their application of both the legal and behavioral criteria included in Section 123. In the exercise members' assessments of the patient's likely dangerousness ranged from "ineffectual and placid" to "prime Park Lane [Special Hospital] material." Although some variation is to be expected in view of the lack of guidance about the kinds of behavior that justify the application of the label "dangerous," the degree of inconsistency in the interpretation of exactly the same objective evidence was unexpected. Furthermore, members were unable to decide whether the patient was primarily a danger to himself or a danger to others. There was a similar lack of agreement between the medical members regarding the patient's diagnosis and symptomatology. Assessments of the severity of his illness ranged from "an extreme case, in need of confinement" to "relatively s t a b l e . . , should be thinking of withdrawing medication." It was also evident that not only did doubt exist about the correct legal interpretation of the criteria included in Section 123, but there was also considerable doubt about the basic relevance and appropriateness of those criteria. Some members stated that it was more important to reach a decision that was in the patient's best interests than to adhere strictly to the legal criteria in Section 123. This has been illustrated by the relative lack of change in members' decisions, despite substantial changes in the legal criteria to be applied. Thus, members tended to emphasize the severity of the patient's illness in exercise 2 when justifying further detention although, legally, this decision should have been related to the criterion of dangerousness. Perhaps, however, members' unfamiliarity with the legislation [in the majority of cases the nature and purpose of Section 123 (c) had to be explained individually to members by the researcher] assists in explaining the apparent irrelevance of the legal criteria included therein. Finally, members' comments reflected the tendency in tribunals to focus only on the unsuccessfu! aspects of a patient's history. This may result from the fact that the report of the Responsible Authority tends also to stress these aspects. Since the hypothetical case was not legally represented; members were required to seek out information from the reports that was indicative of likely success. Only a minority did this. The exercise permitted certain inferences to be made about the nature and effects of group decision making. By comparing group decisions after discussion with members' individual decisions before discussion it was apparent that the group discus-

182

PEAY

sion had the effect of enhancing the average of the prediscussion decisions regardless of whether those decisions were for or against discharge. In the majority of the groups this process resulted in more cautious decisions than the members would have made individually. In a minority of cases the effect was reversed, with groups reaching "riskier" decisions. The preliminary observational work and Hepworth's (1980) study had both indicated the apparent ease with which the majority of tribunal decisions are made. The hypothetical case, however, produced a large number of split decisions (65% in exercise 1), or cases where all the members were not in agreement. It is possible that this results from the artificiality of the exercise and the inability of members to communicate their views during the course of the hearing. Achieving consensus may, in part, depend as much on the informal communication that occurs in real tribunals prior to the "decision-making stage" as on that occurring during the discussion subsequent to the formal presentation of the evidence. It should also be noted that although no minority view ever changed the direction of the group decision in the exercise, minority views did affect the confidence with which the group decisions were made. It was clear from the analysis that dissenting medical opinion had a much greater influence on the confidence with which the legal member recorded the group decision than did dissenting lay opinion (50% swing in group confidence compared with 2.5% swing). In real tribunal decision making, where members are in a position to influence each other's opinions during the course of the hearing, it is likely that the unequal influence of the medical and lay representatives highlighted in the exercise will be exaggerated.

C O N C L U S I O N S AND R E C O M M E N D A T I O N S The research has established the following: (i) the existence of significant differences between both individual members and member groups in their attitudes (particularly those relating to discharge), their factual knowledge, their conceptualization of the tribunal's role, and the legal criteria they consider applicable to their decisions; (ii) the existence of significant differences both between individual members within a tribunal region and between tribunals in different regions in their discharge rates; (iii) that the variation in discharge rate is partially attributable to the individual differences between the members outlined in (i) above; (iv) that these individual differences have a significant effect on the outcome of the tribunal's decision in a substantial proportion of cases. Individual factors are influential even within a group decision-making context; (v) that some individuals (e.g., those with extreme approaches) and some member groups (e.g., legal and medical members) are significantly more likely to influence the outcome of group decisions in a systematic manner; (vi) that the communicative processes that occur between members are not sufficiently explicit to moderate or challenge individually based preconceived views. Finally, the decision-making exercise has indicated that individual decisions tend

MENTAL HEALTH REVIEW TRIBUNALS

I83

to be formulated at an early stage during the hearing and that evidence presented subsequently is less critical with regard to the nature of the final decision. Thus, the research has questioned the validity of the consensus approach to decision making and has indicated that both the nature of the law, its present method of group application, and the nature of the task members are required to fulfill, facilitate the expression of individually based decision preferences. Since these individual preferences have been demonstrated to reflect not on objective assessment of the facts, but rather the often inappropriate opinions of those individuals, the efficacy and justice of the resulting decisions has been questioned. If a patient's chances of being discharged are significantly influenced by the composition of the tribunal reviewing the application, as the research has demonstrated, then tribunal decisions cannot evade the accusation that their nature is arbitrary. It should, however, be stressed that the demonstration that tribunal members tend to reach individually biased decision.s is not felt to be any reflection on the ability of individual members, but is rather attributable to the system within which members are required to reach decisions. The system is felt to be at fault because it provides neither unambiguous decision criteria, nor an appeal system for clarification of those decision criteria, nor even effective feedback about the outcome of the decisions taken. Indeed, the system is predicated on the assumption that guidance for members for the decisions they are required to make is unnecessary, since the group basis of the decisions compensates for any individual deficiencies in knowledge. This assumption has been demonstrated to be unsupportable. Furthermore, the research findings have illustrated how a combination of high expectations of the membership by the tribunal system, and a failure to ensure that members have access to information relevant to the difficult decisions they are required to make, results in an unacceptable degree of inconsistency between tribunal decisions. As Hogarth (1967) succinctly states: "Decision-making in ignorance of pertinent information is not the exercise of discretion, it is pure speculation." Perhaps the high level of rejection of the tribunal's advice by the Home Secretary reflects the inaccessibility to him of the basis on which such advice is given. The factors informing tribunal decisions are similarly inaccessible to applicants. The Review of the Mental Health Act 1959 (Department of Health and Social Security, Home Office, Welsh Office, Welsh Office, Lord Chancellor's Department, 1978) and the Committee on Mental Health Review Tribunal Procedures (Department of Health and Social Security, 1978) have made a number of recommendations for changes in the tribunal system, including the introduction of automatic reviews, increased powers for tribunals, greater feedback about decisions, more informative decisions, and a more structured format for the conduct of tribunal hearings. Although the research findings would appear to support the desirability of such changes, it should be emphasized that the research has indicated that any changes in either the legal criteria or procedural basis for tribunal decisions are unlikely to be either implemented or operated in the intended manner unless further account is taken of the decision-making processes of those individuals who apply the law. As such, a number of additional changes are recommended. Firstly it is suggested that tribunal members could benefit from either some form of training or greater guidance for their decisions or both. It is recognized that such

184

PEAY

proposals are likely to be controversial. For example, consider the negative reception of the Bridge Report's recommendations that judges should undergo study programs (Home Office, Lord Chancellor's Department, 1978). Nevertheless, their implementation is of vital importance if the tribunal system is to represent an "adequate safeguard." Other similar decision-making bodies, such as the Magistracy and the Parole Board, currently benefit from such schemes. Such guidance or training would have a number of advantages. By clarifying the criteria that should underpin tribunal decisions it could reduce the influence apparently exerted currently by extralegal criteria. Although it is accepted that a strict adherence to legal criteria is not always advantageous, if extralegal criteria are felt to be relevant to tribunal decisions then their nature should be explicated, so that members' decisions can at least be based on similar kinds of considerations. Again, guidance could clarify the behavioral bases of such criteria as "dangerousness." The tendency for some members to apply standards which reflect an attempt to identify good risks (individuals who are likely to be successful if discharged) rather than bad risks (individuals who are likely to be dangerous if discharged) not only makes it impossible for applicants to prepare a relevant case but also jeopardizes any attempt to improve the efficacy of tribunal decisions, since the bases for those decisions remain unidentifiable. Guidance or training would also permit each tribunal member to be in possession of equivalent, relevant information, thus lessening the current dependence on effective communication between the three tribunal members. The research has challenged the assumption that the nature of the selection process appoints individuals who have access to all the relevant information. Indeed, during the course of the research a number of members advocated either the introduction of "mock tribunals for training and maintenance of consistent standards" or greater access to statistical and other information. Secondly, if it is still considered desirable that tribunal decisions should be based on a consensus of medical and nonmedical opinion, then some strengthening of the lay element in those decisions needs to be achieved. Although training or guidance may help to alleviate the tendency within tribunal decisions to minimize the lay contribution to the deliberation process, if a more active role could be developed for the lay members then their particular approach might be given more weight. This might be achieved if the lay members were to adopt some of the responsibilities for preparing reports about the patient's likely whereabouts and future circumstances if discharged, currently undertaken primarily by hospital social workers. This proposal would have the additional advantage that the information presented to the tribunal would reflect more closely the kind of information that tribunals define as relevant to their decisions. Furthermore, the factual basis of those reports could more adequately be assessed. Finally, since the research has highlighted the tendency of members to base their decisions on selected aspects of the relevant information and to be influenced unduly by the Responsible Authority's report, it is suggested that some procedural changes in the tribunal's format need to be explored. It is evident that the tribunal would benefit from access, prior to the hearing, to a written report detailing the case in favor of discharge. With the advent of greater representation (e.g., MIND'S tribunal assistance scheme) this report might be prepared by the patient's legal representative.

MENTAL HEALTH REVIEW TRIBUNALS

185

The demonstration that the information-gathering stages of the tribunal are inextricably linked with the decision-making process also justified formal recognition. This might be achieved through a more structured format for the hearing that prescribes discussion at certain stages of the presentation of the evidence, rather than implicitly encouraging the informal exchange of opinion that currently occurs. The possibility that decisions are influenced more by the opinions of individual members than by the factual basis for those opinions has implications for the quality of tribunal decisions. Greater access to suitably qualified representatives and a more controlled discussion of the evidence could help to ensure that decisions are reached on the basis of an assessment of all the relevant information. Thus, the underlying theme of these recommendations has concerned the necessity for both legislation and the tribunal system to recognize formally strategies of human decision making and to attempt to compensate for the evident frailties of such decision makers. Although the research has demonstrated that tribunal members have the capacity to apply the law compassionately, the discretion which enables them to make such decisions also results in those decisions apparently mirroring neither the attributes of justice nor efficacy. If the tribunal system is to form our primary safeguard against the unjustified use of involuntary confinement in psychiatric hospitals, these additional attributes are essential. In view of the research findings, considerable improvements in the quality of decision making could be achieved through some relat!vely minor reforms in the nature of the information available to members.

REFERENCES Aarvold Report. Report on the review of procedures for the discharge and supervision of psychiatric patients subject to special restrictions. (Cmnd 5191.) London: Her Majesty's Stationery Office, 1973. Bean, P. Some issues concerning rules and rule enforcement. British Journal of Law and Society, 1975, 2, 225-235. Bean, P. Are the mental health guardians abusing their power? The Times, 1978, January 3rd. Bell, K. The Mental Health Review Tribunal:A question of balance. Case Conference, 1970, 16, 385-391. Carter, R. The pre-sentence report and the decision-making process. Journal of Research in Crime and Delinquency, 1967, 4, 203-211. Cavenagh, W., & Newton, D. The membership of two administrative tribunals. Public Administration, 1970, 48, 449--468. Cavenagh, W., & Newton, D. Administrative Tribunals: How people become members. Public Administration, 1971, 49, 197-218. Cooke, J.A. Mental Health Review Tribunals. Solicitors Law Journal, 1969, 7th November, 843. Department of Health and Social Security. A discussion paper by the Committee on Mental Health Review Tribunal Procedures. Unpublished, 1978. Department of Health and Social Security, Home Office, Welsh Office, Lord Chancellor's Department. Review of the Mental Health Act 1959. (Cmnd 7320.) London: Her Majesty's Stationery Office, 1978. Fennell, P. The Mental Health Review Tribunal: A question of imbalance. British Journal of Law and Society, 1977, 2, 186-219. Fleming, A.C. Appeals to Mental Health Review Tribunals. The Lancet, 1963, 7275, 263-264. Freer, C.J. Review Tribunals with special reference to Rampton Hospital. British Journal of Psychiatry, 1966, Suppl. 7, 12-13. Fuller, L. An afterword: Science and the judicial process. Harvard Law Review, 1966, 79, 1604-1628. Gostin, L.O. A Human Condition Vol I, The Mental Health Act from 1959 to 1975. Special Report MIND (National Association for Mental Health), 22 Harley Street, London, 1975.

186

PEAY

Gostin, L.O. A Human Condition Vol 2. The Law Relating to Mentally Abnormal Offenders. Special Report MIND (National Association for Mental Health), 22 Harley Street, London, 1977. Gostin, L.O., & Rassaby, E. Representing the Mentally Ill and Handicapped. A Guide to Mental Health Review Tribunals. Quartermaine House, Middlesex: National Association for Mental Health and the Legal Action Group, 1980. Greenland, C. Mental Illness and Civil Liberty. Birkenhead: Willmer Bros, 1970. Hansard's Parliamentary Debates. Mental Health Bill 1959. Presented 17th Dec. 1958; 2nd Reading 26th January 1959; 3rd Reading 5th and 6th May 1959. Hepworth, D. Assessment of Danger to Self and Others (Study of Mental Health Review Tribunals' Interpretations of Dangerousness). Unpublished Ph.D. thesis, University of Nottingham, 1980. Hogarth, J. Towards the improvement of sentencing in Canada. Canadian Journal of Correction, 1967, 9, 122-136. Hogarth, J. Sentencing as a Human Process. Toronto: University Press of Toronto, 1971. Home Office & Lord Chancellor's Department. Judicial Studies and Information. Report of the Working Party. (Bridge Report). London: Her Majesty's Stationery Office, 1978. Hood, R. Sentencing the Motoring Offender: A study of magistrates' views and practices. London: Heinemann, 1972. Kittrie, N. The Right to be Different: Deviance and Enforced Therapy. Baltimore: John Hopkins Press, 1971. Mental Health Act 1959 (7 & 8 Eliz.2,c.72), Mental Health (Amendment) Act (c.29) 1975. Mental Health Review Tribunal Rules. SI 1960/1139 as amended by SI 1974/241 and SI 1976/447, 1960. Mental Welfare Commission for Scotland. No Place To Go. Edinburgh: Her Majesty's Stationery Office, 1975. Mitchell, H.E., & Byrne, D. The defendant's dilemma: effects of jurors' attitudes and authoritarianism on judicial decisions. Journal of Personality and Social Psychology, 1973, 25, 123-129. Monahan, J. The prediction of violent criminal behavior; a methodological critique and analysis. In A. Blumstein, J. Cohen, & D. Nagin (Eds.) Deterrence and Incapacitation." Estimating the effects of criminal sanctions on crime rates. Washington D.C.: National Academy of Sciences, 1978. Moscovici, S., & Zavalloni, M. The group as a polarizer of attitudes. Journal of Personality and Social Psychology, 1969, 12, 125-135. Myers, D., & Lamm, H. The group polarization phenomenon. Psychological Bulletin, 1976, 83, 602-627. Nie, N., Hull, C., Jenkins, J., Steinbrenner, K., & Bent, D. Statistical Package for the Social Sciences 2nd ed. New York: McGraw-Hill, 1975. Peay, J. Mental health review tribunals--A study of individual approaches to decision-making. Unpublished Ph.D. thesis. Department of Psychology, University of Birmingham, 1980. Reiss, A.J., Jr. Discretionary justice in the United States. International Journal of Criminology and Penology, 1974, 2, 181-205. Rosenhan, D. On being sane in insane places. Science, 1973, 179, 250-258. Royal Commission on the Law Relating to Mental Illness and Mental Deficiency 1954-1957. (The Percy Report) (Cmnd 169). London: Her Majesty's Stationery Office, 1957. Schubert, G. The Judicial Mind: The attitudes and ideologies of Supreme Court Justices 1946-1963. Evanston, Illinois: Northwestern University Press, 1965. Stephens, D.J. Mental health review tribunals. Unpublished thesis submitted for degree of Magister in Legibus, University of Wales, 1968. Wheeler, S. Controlling Delinquents. New York: Wiley, 1968. Wilkins, L., & Chandler, A. Confidence and competence in decision-making. British Journal of Criminology, 1965, 5, 22-35. Wood, J.C. Mental Health Review Tribunals. Medicine, Science and the Law, 1970, 10, 86-92.