Compulsory community psychiatric care remains the subject of intense international debate. Introduced into North American and Australian jurisdictions in the 1980s, community treatment orders (CTOs) were seen as a method of addressing problems created by deinstitutionalisation policies. ^{Reference Geller1} Originally supported by civil libertarians as being less restrictive than hospitalisation, ^{Reference Hiday2} their evolution into legal interventions and the lack of scientific evidence supporting their use has caused concern. ^{Reference Swartz and Swanson3,Reference O'Reilly4}

Until the introduction of the 2007 amendments of the Mental Health Act 1983, CTOs did not exist in England and Wales. Instead, three main forms of community treatment existed: extended Section 17 (s.17) leave, statutory guardianship and supervised discharge. The latter two options neither authorise compulsory community treatment nor provide sanctions for non-cooperation. Conversely, s.17 leave, which has been used as a de facto CTO since the 1980s, enables clinicians to recall individuals to hospital for compulsory in-patient treatment. This practice was deemed unlawful in 1986, ⁵ but subsequent legislation - the Mental Health (Patients in the Community) Act (1995) - and case law have altered this status. ^6-9

The 2007 amendments of the Mental Health Act introduced CTOs in the form of supervised community treatment (s.17A-G) aimed primarily at ‘revolving door’ patients. Theoretically, CTOs should replace the use of extended s.17 leave and provide patients with greater legal protection. This article examines why this may not necessarily be the case.

Extended s.17 leave

The Mental Health Act 1983 (s.17) sets out provisions for individuals compulsorily detained under sections 2, 3, 37 and 47 to be granted leave of absence. Its general purpose is to facilitate an (often institutionalised) individual's return to the community as part of a wider care package. Only the responsible clinician may grant leave, which may be indefinite (s.17(2)) or for a specified period. Leave may be extended (s.17(2)) or revoked (s.17(4)), but should ‘not be used as an alternative to discharging the patient’. ¹⁰ While on leave, individuals remain ‘liable to be detained’ and may be recalled in cases of failure to cooperate with treatment or of deterioration in the person's condition. Although it has a primarily facilitative function, extended s.17 leave is commonly used to enforce community treatment by ensuring patients remain ‘liable to be detained’. Patients may be discharged on leave, recalled before their section expires, have their section renewed and then be discharged as ‘liable to be detained’ back into the community. The renewal of sections while individuals are on leave, the increasingly minimal requirement for in-patient treatment and the criteria for s.3 renewal compared with those for detention have been challenged in the courts. ^5,7-9,11

In R v. Hallstrom [1985] the practice of renewing civil detentions while individuals were on extended leave was challenged. ⁵ Interpreting the words of s.20 narrowly, the judge ruled this use of s.17 unlawful (approved in R v. Canons Park MRHT [1995]), ¹¹ stating that individuals on extended leave cannot be in-patients, that patients detained under s.3 should be in-patients and that the Mental Health Act did not provide for compulsory community treatment.

In 1999, B v. Barking Havering and Brentwood Community Healthcare Trust again challenged this practice. ⁹ The patient B lived mainly in the community but was subject to random illicit drug testing. After her s.3 was renewed she challenged the decision, declaring that she received insufficient hospital medical treatment (defined as ‘nursing, and… care, habilitation and rehabilitation under medical supervision’; Mental Health Act 1983 s.145) to warrant continued detention. This argument was dismissed by the Court of Appeal, which held that ‘continues to be detained’ should be interpreted as ‘continues to be liable to be detained’; that regular illicit drug testing satisfied the renewal criteria; and that any element of hospital treatment that ensured a holistic care plan's success allowed lawful detention, even if the individual resided mainly in the community. The determination that an individual attending hospital once a week is detained is clearly questionable. Even in the broadest sense s.20(4) cannot realistically be interpreted to allow such a practice. Medical treatment ‘in a hospital’ is clearly that.

In R (DR) v. Mersey Care NHS Trust [2002] the court was asked to quantify the amount of hospital treatment needed to justify a section's renewal. ⁸ The patient DR resided in the community, returning to hospital for once-weekly occupational therapy, once-weekly clinical review and fortnightly antipsychotic administration. Her responsible clinician believed that continuation of her detention was essential to prevent non-adherence to treatment and subsequent deterioration. The court dismissed any distinction between ‘in a hospital’ and ‘at a hospital’, suggested that the test of s.20(4) was whether a ‘significant part’ of the treatment plan was ‘in hospital’, stated that twice-weekly hospital visits were ‘significant’ and declared detention renewal under these circumstances lawful. This case raised the issue of de facto CTOs, something that R v. Hallstrom had sought to prevent, the seemingly marked differences between the criteria for initial detention under s.3 and those for renewal, and the liberal definition of ‘hospital’.

In R (CS) v. Mental Health Review Tribunal and Another [2004], the patient was obliged to return to hospital only once a week for psychological treatment and once a month for a ward round. ⁷ While on extended leave, her s.3 was renewed by a mental health review tribunal, based on her poor insight and non-adherence to treatment which could lead to her deterioration. She appealed, stating that the tribunal should have discharged her or considered other options: guardianship (Mental Health Act 1983 s.7), which is not widely used, ^{Reference Shaw, Hatfield and Evans12} or supervised discharge (s.25A-J), which was used sparingly and has now been replaced by the CTO. ^{Reference Davies13} Rejecting her appeal, the judge acknowledged that re-integrating individuals into the community was difficult and that although the hospital/overall treatment plan relationship must be clearly visible, ¹⁴ it might be very narrow. In this case, civil detention with extended leave was required to allow recall to hospital should the patient's condition deteriorate.

The judgments of DR and CS liberalised what is legally permissible when a patient no longer requires in-patient treatment but the responsible clinician believes that ‘being liable to be detained’ provides back-up to protect the individual from deteriorating by allowing recall to hospital. Psychiatrists are familiar and comfortable with extended leave. Would existing CTOs have changed these cases' management?

Community treatment orders: overview

Mainly an Australasian and North American phenomenon, the CTO has existed for more than 30 years, ^{Reference Miller and Fiddleman15} and has recently been defined as:

Four main types exist:

1. • conditional leave following an involuntary admission - treatment, but not hospitalisation, compulsory;

2. • hospitalisation criteria met but the individual is treated involuntarily in the community;

3. • preventive order - involuntary admission criteria not met but may be if deterioration occurs;

4. • court-ordered treatment. ^{Reference Wales and Hiday16}

There are three main conceptual designs underpinning CTOs: whether the CTO and hospitalisation criteria are similar; whether the aim is to treat or prevent (‘preventive’ CTO) deterioration; and whether it is the ‘least restrictive’ (civil libertarian approach) option for an individual, ^{Reference Hiday2} or simply involuntary community care for ‘revolving door’ patients, used according to clinical and legal criteria. Preventive CTOs aim to prevent deterioration which might endanger the individual or the public, ^{Reference Geller1} have different criteria from those for hospitalisation, and provide a component of psychiatric management for which there are specific indications. ‘Least restrictive’ CTOs provide an alternative to hospital admission for anyone requiring compulsory treatment, have identical criteria to civil detentions and allow people with deteriorating mental health to be treated.

Their use remains controversial. Advocates of CTOs believe these orders are less restrictive, ^{Reference Pinfold and Bindman17} stabilise the lives of individuals with severe mental illness, ^{Reference O'Reilly18} and provide greater freedom for patients. Opponents fear that the original concept has been replaced by control, restraint and threat, affecting the therapeutic relationship and driving individuals away from mental health services; ^{Reference Pinfold and Bindman17} other concerns are that these orders are too severe and restrictive, ^{Reference Batty19} may discriminate against ethnic minority groups, and that vulnerable patients will be further deprived of civil liberties. ^{Reference Carvel20}

Ten years after initial consultation began, and after much controversial debate, CTOs were introduced into England and Wales by amending the 1983 Mental Health Act to include subsections 17A-G, which define supervised community treatment. Eligibility for supervised community treatment is relatively restricted: only patients detained under s.3, unrestricted Part 3 hospital orders or transfer directions are eligible. Such treatment aims to support ‘revolving door’ individuals and individuals requiring community support to prevent relapse and further prolonged hospitalisation. Community treatment orders, initiated by the responsible clinician and an approved mental health practitioner (AMHP), should be considered for any individual likely to be on s.17 leave for more than 7 days. Criteria include that: the individual's mental illness necessitates treatment; treatment is necessary for the individual or society but hospitalisation is unnecessary; and that the responsible clinician can recall the patient (as a last resort, if in-patient treatment is necessary and not recalling the individual would present a risk to the patient or the public; subsection 17A(5)). Recall is for 72 h maximum (unlike s.17 leave), after which the individual must be discharged into the community or the CTO revoked, with the individual remaining subject to their original section.

The CTO's conditions must state the date and time that the order begins, how the individual meets the criteria and the conditions to which the individual must adhere (including examination to consider CTO extension and examination by a second opinion appointed doctor). Patient collaboration is crucial. The CTO does not allow forced community treatment unless the individual lacks capacity, immediate treatment is required or treatment is necessary to prevent harm to the individual.

Community treatment orders run for an initial 6-month period, may be renewed for a further 6 months and then annually (examination to ensure criteria remain must accompany each renewal). Renewal criteria are similar to those that initiate a CTO: the responsible clinician and AMHP must agree, with an opinion being sought from another treating clinician. As soon as the criteria are not met, individuals should be discharged by their responsible clinician, relatives, hospital managers, a mental health review tribunal or the secretary of state (independent sector).

Part 4A of the 2007 revision of the Mental Health Act regulates community treatment and provides a limited scope for enforcing treatment compared with that available under Part IV of the original Mental Health Act for in-patients. It clarifies that professionals cannot override a competent individual's refusal of consent (s.64B,C for adults, s.64G for children), but can treat incapacitated patients (provided force is unnecessary; s.64D) if deemed essential and regardless of the existence of a valid advanced directive (s.58(3)(b)). When urgent treatment is justified by criteria set out in s.62 of the Mental Health Act 1983, force ‘as is a proportionate response to the patient's suffering harm and the seriousness of that harm’ may be used. Competent individuals can be compulsorily treated (regardless of their refusal; s.56(4)) following hospital recall, provided such treatment may be administered under Part 4, s.58 of the amended Act.

Evidence supporting the use of CTOs

The expectation is that CTOs should replace the equally controversial use of extended s.17 leave. The widespread practice of evidence-based medicine means that clinicians must be clear that a new treatment or practice is superior to existing practice. The efficacy of CTOs is unproven, however. Six literature reviews have examined CTO effectiveness. ^{21-Reference Churchill, Owen, Singh and Hotopf26} The Chief Psychiatrist's review lacked specific objectives, was mainly descriptive and failed to reach any useful conclusions. ²¹ Ridgely's review examined existing evidence for CTOs and other alternatives and concluded that evidence supporting the widespread use of CTOs was lacking. ^{Reference Ridgely, Borum and Petrila23} The National Association of State Mental Health Programme Directors' review examined national literature and policies on CTO usage and was unable to demonstrate any conclusive supporting evidence for their use. ²² Dawson's review, a comparative study of legislation governing CTO use, found similar negative aspects of CTOs as previous reviews (medication-dominated care, difficulties assessing alternative care, prolonged use of CTOs) but, unlike previous reviews, also found some benefits (therapeutic, improved adherence to community medication regimens, reduced hospitalisation rates). ^{Reference Dawson24} Others question whether his non-systematic approach might have produced these discrepancies.

The two largest reviews do not suggest any definite benefits of CTO usage. ^{Reference Kisely, Campbell and Preston25,Reference Churchill, Owen, Singh and Hotopf26} The Cochrane review considered clinical and cost-effectiveness of CTOs for severely mentally ill individuals. ^{Reference Kisely, Campbell and Preston25} It only analysed the highest form of evidence - randomised controlled trials. Two American trials, both examining court-ordered CTOs, were identified. ^{Reference Swartz, Swanson, Wagner, Burns, Hiday and Borum27,Reference Steadman, Gounis, Dennis, Hopper, Roche and Swartz28} Neither appears methodologically sound. In the first study, by Swartz et al, randomisation procedures were not described and the allocation procedure was not concealed, but the CTO provisions were well described and the source and type of patients were reported. Although outcomes were specified at the start of the study, outcome assessments were not conducted masked, there was no indication that the data analysers were masked and reports of losses to follow-up were unclear. ^{Reference Swartz, Swanson, Wagner, Burns, Hiday and Borum27}

In the second trial, by Steadman et al, randomisation was not described and the allocation procedures were not concealed from the investigators, but the CTO provisions were well described and the source and type of patients were reported. Outcome measures were not specified, outcome assessment was not conducted masked and it was not possible to determine whether those analysing the data were masked. ^{Reference Steadman, Gounis, Dennis, Hopper, Roche and Swartz28} Neither study demonstrated significant differences in service use, social functioning, quality of life or cost-effectiveness compared with standard care. Individuals were, however, less likely to be crime victims. Additionally, 85 individuals would need their freedom limiting to prevent one hospital readmission, something that the majority of society would deem unacceptable. The authors concluded that more robust randomised controlled trials were needed. ^{Reference Kisely, Campbell and Preston25}

The review by Churchill et al systematically reviewed all trials relating to CTOs. ^{Reference Churchill, Owen, Singh and Hotopf26} It identified a number of CTO arrangements (described above) and consistencies in the characteristics of patients subject to these orders (men; schizophrenic disorder; serious mood disorder; repeated admissions; non-adherence to treatment; complex aftercare needs; violent potential). Efficacy measurements, however, were inconsistent. All studies reported readmission rates and some reported length of hospitalisation, contact with services, medication adherence, social functioning and general mental state. No study provided good evidence that CTOs were effective in reducing readmission rates.

Although it is recognised that the hierarchy of research is not necessarily appropriate for the assessment of complex community-based interventions, ^{Reference Burns, Catty, Dash, Roberts, Lockwood and Marshall29-Reference Wright, Catty, Watt and Burns31} all the studies fell below the standards of evidence-based medicine (mostly cross-sectional, non-randomised comparative studies), therefore providing poor qualitative data.

Arguments for CTOs

Despite the lack of definitive evidence in favour of CTOs, they have been used internationally for a considerable amount of time. Although the Royal College of Psychiatrists has not rejected CTOs outright, it has pointed out that ‘studies from abroad do not show community treatment orders to be the panacea that the Government makes them out to be. Indeed the evidence is equivocal as to whether they bestow any benefit on a wide scale.’ ^{Reference Slade and Priebe32} With this statement emanating from a national body and the lack of robust evidence, it is difficult to find academic papers supporting their use. ^{Reference Lawton-Smith, Dawson and Burns33-Reference Burns and Dawson35}

A debate in the British Journal of Psychiatry presented arguments for and against the use of CTOs. ^{Reference Slade and Priebe32} Those opposing CTOs cited lack of evidence, the danger of increasing compulsion and prolonged medical supervision, unresolved ethical concerns (autonomy v. medical paternalism) and a distraction from finding voluntary solutions in the community. Those supporting CTOs suggested that a properly regulated CTO can be used in place of compulsory detention. They suggested that all methods of compulsory treatment can be flawed: too long, too widely interpreted, for the incorrect reasons and possibly in contravention of human rights. They also argue that properly organised involuntary care can be successful even in those with severe mental illness, and cite Australasian studies in support.

It can be argued that evidence-based medicine may not be appropriate in this complex area of psychiatry, particularly when patient outcomes are dependent on the ‘postcode lottery’ of care. Randomised controlled trials are difficult to apply to compulsory interventions, particularly as those suitable for CTOs differ from those receiving involuntary in-patient treatment. Lesser evidenced based studies, however, may provide a better method of researching this issue.

Patient surveys in New Zealand have demonstrated that patients prefer out-patient to in-patient treatment as the former allows more freedom and greater control over their lives. ^{Reference Owino36} Additionally, the community support provided them with a sense of security. A survey of psychiatrists believed CTOs to be a useful way of organising community treatment. ^{Reference McIvor37} These orders made continuing patient contact easier, improved medication adherence, made detection of relapse easier and allowed involvement of families. Reviews of Canadian ^{Reference Gibbs, Dawson, Ansley and Mullen38} and New York ^{Reference Steadman, Gounis, Dennis, Hopper, Roche and Swartz28} CTOs suggest benefits such as the maintenance of treatment while individuals lack insight, and the ability to secure better housing. The real question is whether these benefits are the result of legal intervention or of an increased ability to access a high level of services.

With so little ‘scientific’ evidence, it is difficult to see how psychiatrists would be persuaded to alter their clinical practice, although individual psychiatrists might wish to experiment with a different mechanism of treating individuals in order to prevent the use of increasingly scarce hospital beds. With international research guidelines stipulating that randomised controlled trials should be undertaken where there is doubt about treatment superiority, the amended Act provides an excellent opportunity for a high-quality trial. With two main options (extended s.17 leave and supervised community treatment) for ongoing community care and no definitive research evidence for either option, a trial commencing after the introduction of a new treatment option is ideally placed to provide this evidence. The Oxford Community Treatment Order Evaluation Trial (OCTET) aims to do just this. ^{Reference Romans, Dawson, Mullen and Gibbs39} The initial plan is to compare 300 patients being discharged from a civil detention by randomising them to an experimental group (CTO) or a control group (all other options) and comparing the outcomes at 12 months. The researchers hope to: provide rigorous and convincing evidence as to CTO effectiveness; demonstrate whether adding CTOs to high-quality community care reduces readmission rates and affects other patient outcomes; identify patient characteristics and care patterns associated with good outcomes; inform an economic analysis to model the national cost of introducing CTOs; and contribute to training for effective implementation. If the trial provides good conclusive evidence for CTO effectiveness, psychiatrists may be persuaded to change their practice.

Why should CTOs curtail the use of s.17 leave?

Traditionally, court-approved extended s.17 leave has been used to enforce community treatment. The introduction of supervised community treatment provides another way of administering community treatment, but critics believe that CTOs do not alter the creative use of s.17 leave despite the revised Mental Health Act stipulating that clinicians must ‘first consider’ the suitability of supervised community treatment for any patient likely to be granted s.17 leave for more than 7 consecutive days (s.17(2A)). ‘Consider’ is a vague, non-legal term and therefore the responsible clinician's legal responsibility is simply to mentally weigh up the pros and cons of each option. With no evidence to support either method, several differences may sway clinicians in either direction. Bowen, ^{Reference Dreezer, Bay and Hoff40} supported by the Mental Health Act Code of Practice, ^{Reference Burns, Rugkasa and Molodynski41} argues that supervised community treatment impinges less on human rights and provides better safeguards for individuals when compared with extended leave. In Bowen's opinion, several contrasting features exist:

1. (a) two individuals (responsible clinician, AMHP) must agree that specified supervised community treatment criteria are met, whereas the responsible clinician alone may grant extended s.17 leave;

2. (b) supervised community treatment specifically provides for expiry, renewal, discharge and the right to appeal, whereas s.17's safeguards apply to the compulsory treatment's application;

3. (c) supervised community treatment criteria for recall are specific;

4. (d) competent patients on supervised community treatment cannot be compulsorily treated (Part 4A), whereas Part IV allows compulsory treatment of competent individuals on s.17 leave;

5. (e) supervised community treatment is less likely to violate the European Convention on Human Rights:

   1. (i) s.17 powers are too vague to satisfy the requirement that any interference with rights under Article 8 must be ‘in accordance with the law’;

   2. (ii) s.17 recall may not meet the Winterwerp criteria (justifying detention based on recent evidence of lack of capacity). ^{Reference Dreezer, Bay and Hoff40}

Bowen's reasoning, however, may need further examination: points (c) and (e) are unlikely to raise legal consequences because this use of s.17 leave is court-endorsed; the criteria differences raised in points (a) to (c) are relatively insignificant and of no real practical significance; and the law provides safeguards for those on extended leave (renewal with two clinicians' agreement must mean that the individual remains ‘liable to be detained’). Point (d), however, raises a legitimate issue. The ability to force treatment on those on extended leave clearly violates their human rights. Conversely, treatment cannot be forced on competent individuals under supervised community treatment; instead, they must be recalled to hospital where treatment can be lawfully administered. Bowen's arguments in conjunction with inadequate evidence supporting the use of CTOs are unlikely to curtail psychiatrists' use of s.17 leave which is tried, tested and court-approved.

Conclusion

Despite their widespread international use, and their introduction by the amended Mental Health Act 1983 in 2007, CTOs remain a relatively new concept in community psychiatric care in England and Wales, partly as a result of varying funding for community resources. Extended s.17 leave has, however, been used for many years as a method of enforcing treatment in the community and granting responsible clinicians the right to recall non-cooperative or deteriorating individuals. Controversy over the amount of treatment required ‘in hospital’ persists. Although CTOs may have advantages there is still no robust scientific evidence to support their widespread use. Some clinicians, however, support the use of CTOs although there is no professional consensus. A study of psychiatrists in England and Wales showed that 46% favoured CTOs with 34% being opposed to them. ^{Reference Bowen42} A debate in the Institute of Psychiatry in 2000 concluded with a two-thirds majority against the introduction of compulsory community treatment. ^{Reference Pinfold and Bindman17} There is currently no answer to this debate. Where funding is available clinicians may choose to use CTOs. Where cynicism exists, clinicians may choose to wait for the results of the OCTET trial, as may those who are influenced by evidence-based medicine.