The new Mental Health Act for England and Wales is likely to extend the powers of mental health review tribunals (MHRTs) by giving tribunals the power to approve all compulsory treatment (Department of Health, 1999a , b ). The medical member may be dropped entirely from the tribunal's proceedings (Reference Richardson and MachinRichardson & Machin, 2000). In Ireland, a proposed new Mental Health Act will introduce MHRTs for the first time (Reference CalvertCalvert, 2000). The 1983 Mental Health Act contains no explicit statement of underlying principles, although some were introduced in the revised Code of Practice. The Expert Committee (Department of Health, 1999b ) suggested that the new Act should specify broad principles where these would help in statutory interpretation, particularly because a range of practitioners working in different settings will be required to understand and implement its provisions. The Green Paper initially suggested that the proper place for setting out principles should be a Code of Practice, but ended by inviting comments on the principles proposed by the Expert Committee, and on whether inclusion of principles would aid interpretation of the new Act.

In most ethical debates a number of principles may apply to a given case; the usual difficulty is in deciding which principles should take precedence. There are real differences between legal and medical ethical paradigms, and many clinically perverse tribunal decisions are the result of conflicts between legal and medical values. Medical ethics hold that the best interests of the patient are paramount. The physician is directed to ‘do no harm’. Under the legal principle of the rule of law, legal procedures and the liberty of the individual are paramount (Reference Jones, Berrios and FreemanJones, 1991). The medical counter to this is that the restoration of the capacity to exercise free choice and responsibility must come before liberty.

Because lawyers are likely to have an authoritative role in the process for compulsory admission to hospital as well as discharge, there is a risk that legalistic approaches to decision-making may seriously harm clinical decision-making, while legal styles of rhetoric and debate may seriously harm therapeutic relationships. By analogy with iatrogenic harm, this could be called juridogenic harm. Those reports of inquiries after homicide that involve examples of mishaps owing to legalistic decision-making (e.g. Reference Dixon, Herbert and MarshallDixon et al, 1999; Reference Ashtal, Kelly and DevauxAshtal et al, 1998), notably avoid levelling any explicit criticism of tribunal or court decisions, possibly because of respect for the judicial process. In the medical literature, evaluation of juridogenic harm (or therapeutic jurisprudence) is limited by the current lack of regard for case reports and the requirements to protect medical confidentiality.

We propose that the new Mental Health Acts should include Statements of Principles similar to those in the 1989 Children's Act (Reference White, Carr and LoweWhite et al, 1990), and the 1996 Family Law Act in order to safeguard therapeutic relationships and clinical practice. Explicit guiding principles would limit the adverse effect of tribunal hearings on therapeutic relationships between patients and clinicians. They would also aid judicial interpretation and promote consistent application of the legislation by clinicians. We have listed below four principles that might usefully be included in the new Act, and would welcome suggestions for others.

1. Nothing should be said or done in a MHRT that could harm an established or potential therapeutic relationship

Natural justice suggests that patients should be able to challenge the grounds for their detention through their legal advocates and that it is sometimes appropriate for this to involve rigorous cross-examination. However, patients' advocates frequently go beyond this, resorting to rhetorical or theatrical submissions, including expressions of mock indignation. Legal advocates are not obliged to disclose unfavourable reports and, having commissioned reports from independent experts or catchment area clinicians who know the patient well, may contrive to prevent the tribunal from coming to know the opinion of key clinicians. By contrast, a clinician who discloses information to a MHRT without the patient's consent is required to demonstrate a serious cause for concern (e.g. W v. Egdell [Reference Gunn1990]).

Legal advocates are under a broad ethical obligation to avoid actions that might conflict with their client's instructions (as opposed to acting in their client's best interests). In some hospitals particular legal firms can gain near monopolies of representation, often by apparently fierce and contentious, though not necessarily effective, advocacy. Gostin and Fennell (Reference Gostin and Fennell1992, p. 194) advise lawyers against using such strategies on the ground that they are relatively unsuccessful. A lawyer with a thin case against detention may none the less seek to satisfy clients and enhance his reputation among a hospital population by this approach. Clinicians trained in rational scientific discourse rather than in rhetorical skills are disadvantaged because Presidents of tribunals (lawyers) often permit this style of advocacy. Tribunal Presidents effectively treat clinicians as the prosecution side in an adversarial trial, while assuming that the patient will listen dispassionately to arguments against his or her release. Suggestible patients and patients with paranoia may interpret their lawyer's presentation of the case against detention as validating their belief in a conspiracy against them. Contentious styles of legal representation can extend to matters unrelated to the patient's detention under the Act. Solicitors commonly send letters demanding change of consultant psychiatrist, increased leave, reduction of medication etc., all more appropriately dealt with by independent lay advocacy services. Once lawyers establish this wider ‘extra-legal’ advocacy role (or market) for themselves, vulnerable patients may feel less able to negotiate directly with their psychiatrists.

The present MHRT system in England and Wales is expensive to the legal aid fund (Reference Blumenthal and WesselyBlumenthal & Wessely, 1994) and anything that increases the legal aid budget will be unpopular with politicians and civil servants. However, a solution is required. A possible remedy would be for hospital managers to provide lawyers to argue against discharge, with clinicians acting as witnesses only. Under any system all parties, including legal representatives, should be under an explicit obligation to act for the patient's best interests and to avoid harming the patient's health by saying or doing anything that puts current or future therapeutic relationships in jeopardy.

2. All decisions on evidence should be made on the balance of probabilities, not beyond reasonable doubt

The MHRT rules for England and Wales (Reference JonesJones, 1999, pp. 511-543) allow MHRTs to admit a much wider range of evidence than the criminal courts, where the standard of proof is ‘beyond reasonable doubt’. Despite this, Presidents of tribunals are sometimes susceptible to the argument that a history of violent acts by patients can be discounted if they have not been proven in a criminal court (Reference PeayPeay, 1989, pp. 217-220). This disregards evidence that the police and courts commonly present patients directly to hospital and then fail to pursue prosecutions, even after quite dangerous behaviour (Reference Humphries and JohnstoneHumphries & Johnstone, 1993). In practice good clinical decisions are made on the balance of probabilities, taking into account a wide range of sources of information and utilising observations over time rather than a snapshot of the patient's mental state. Provided such information is documented to a reasonable standard, it should be accepted and given substantial weight by the tribunal.

3. The evidence of clinicians should be given more weight when clinical responsibility results from the expressed opinion

There is a market in independent expert opinions, and some experts become known for having a particular view in relation to certain clinical issues. In America it has been argued that clinicians treating a particular patient should never give evidence as expert witnesses owing to conflicts between their legal obligations to the court and those arising from the therapeutic relationship with the patient (Reference ApplebaumApplebaum, 1990). This position has not been widely accepted outside the US (e.g. Reference GunnGunn, 1999). In relation to formal detention or release, we propose a principle of continuity of responsibility according to which the MHRT gives more weight to the evidence of the clinician who would take responsibility for the patient's ongoing care, rather than to independent opinions.

This problem regularly crops up in tribunals involving patients who are dangerous and are placed in hospitals outside their local services, such as special hospitals or private secure units. In practice tribunals seldom adjourn to obtain evidence from the clinicians likely to be responsible for the patient's care as a consequence of the tribunal's decision (e.g. R v. MHRT and others ex parte Hall [Reference Humphries and Johnstone1999]). Tribunals hearing applications from patients ought to hear evidence from catchment area services as a matter of course and from the clinicians who would take responsibility for the patient in the community. This would promote continuity of responsibility since the opinion of the clinicians would be linked to clinical responsibility for the patient. It would also minimise the risk of tribunal decisions that prejudice future work with the patient.

4. Legal processes should not be used to bring indirect pressure to bear on clinicians through legal threats to health authorities

Applications for judicial review are often intended to achieve some concession other than that specific to the point of law or procedure in the application. Lawyers profit substantially if granted leave to proceed with an application funded by legal aid, whereas the health authority must defend actions out of its own over-stretched funds. Patients' lawyers can therefore bring minatory pressure to bear on the health authority and on clinicians simply through the threat of judicial review. An example of this is the way that the judgement in the case of Fox (R v. Ealing District Health Authority [1993]) is commonly used. The only equitable solution in such cases would be to allow the legal representation of health authorities to be funded out of the legal aid fund. Before granting leave to applications for judicial review, the higher courts should give greater consideration to the background factual details of an application and the likely effect on therapeutic relationships.

A model solution to the problem of competitive legal advocacy may be found in the role of the guardian ad litem in relation to children or incapacitated patients. In matters relating to children, family courts appoint a guardian ad litem who is under a duty to safeguard the interests of the child in a manner prescribed by the rules of the court (Children's Act 1989, Section 41 (10)). It is expected that the guardian will normally instruct a solicitor to act for the child (Reference White, Carr and LoweWhite et al, 1990). This is a successful model that achieves a high standard of representation in the child's interests. It allows experts, including lawyers or psychiatrists, to be instructed by the court or the guardian ad litem to act in the best interests of the child. In appropriate cases the court may rule that a child can instruct for him or herself. Although this can be characterised as paternalistic, this system is employed in the family courts under the principle of parens patriae and can also justifiably be applied to incapacitated adults. Adopting this approach would limit costs while still allowing the patient access to the appropriate professional input when necessary.

Conclusion

We have concentrated here only on those principles relevant to the workings of existing MHRTs and their impact on therapeutic relationships and clinical decision-making. Eastman (Reference Eastman1994) has suggested a more far-reaching principle of reciprocity that balances the State's right to deprive a psychiatric patient of his or her rights against the State's duty to allocate the resources required for treatment to restore health or ameliorate suffering. The principle of non-discrimination, which aimed at treating patients with physical and mental illnesses equally by applying a capacity test to both, now looks unlikely to find its way into proposed mental health legislation in England and Wales. This may perhaps be owing to fears that a capacity test would simply be equated with insight, and lawyers tend to see this concept as no more than a measure of whether the patient agrees with the psychiatrist. However, clinical assessments of insight are more sophisticated than that (Reference Amador and DavidAmador & David, 1998), and objective tests based on the clinical history could readily be established (Reference White, Carr and LoweKennedy, 1999).

The new Act will add a formal appeal mechanism to a higher court, bringing MHRTs into line with the Human Rights Act. Clinicians and health authorities, faced with decisions by tribunals that they believe are dangerous, should use this and other available remedies, and may be required to do so by ethical considerations. Tribunals do make mistakes and clinicians who acquiesce in a tribunal's decision to override their clinical opinions may still be held responsible if the discharge goes wrong (e.g. Reference Ashtal, Kelly and DevauxAshtal et al 1998, pp. 124-131). It is right that clinicians and health authorities, faced with decisions that they find baffling, perverse or dangerous, should have recourse to the available remedies, particularly since MHRT decisions are not subject to any other form of audit or governance.

The legal system has a vital role as a balance against the powers given to psychiatrists in their operation of the Mental Health Act. This need not mean developing rigidly judicial tribunals that undermine therapeutic relationships. Nor should the new Act allow tribunals to abrogate the therapeutic roles and responsibilities of other professions. Extending the powers of tribunals may lead to the legal equivalent of iatrogenic harm. This could be avoided by the inclusion of principles founded on medical values and ethics in the new Act, helping to safeguard clinical practice and therapeutic relationships. If therapeutic principles are not explicitly stated in statute the new Acts may quickly be effectively rewritten in the higher courts in accordance with legal rather than medical principles.