For
Every psychiatrist knows that the commonest cause of relapse of psychotic illness in patients living in the community is non-adherence with medical treatment. The inpatient bed pool has shrunk in Ireland and further bed reductions in line with Vision for Change policy and development of community supports, such as home treatment services, are in the pipeline. This transfer to a primarily community-based system of care is to be warmly welcomed. But it also requires that the laws governing inpatient treatment need to be extended into the community. Specifically, supports are necessary for that subgroup of patients with severe psychotic illness accompanied by cognitive dysfunction and impaired insight who are unable to adhere to even well resourced community care. We are talking here about patients whose behavioural stabilization historically was only achieved by coercive inpatient care. Such difficult to treat individuals can often do very well when treated under a community treatment order (CTO).
CTOs are legal statutes that require individuals, who suffer from severe mental illness, to attend clinical services and follow a plan of treatment, when living in the community. The use of CTOs is restricted to patients meeting specific criteria. For example, in some jurisdictions patients must meet inpatient committal criteria and also be deemed incapable of making treatment decisions.
CTOs were developed in the United States in response to the very situation we increasingly face in Ireland: the transfer of the locus of care and the treatment of people with severe mental illness from hospitals to the community. In the United States it quickly became apparent that many patients dropped out of follow-up care because they lacked any appreciation that they needed treatment. Some of these patients were able to cooperate with assertive outreach services even when lacking full capacity to provide informed consent for such services. However, a small group of patients refused to engage in any treatment or follow-up care without a legal requirement and possible sanction of return to hospital. CTOs in one form or other are now available in more than 75 international jurisdictions, including Australia, Canada, England and Wales, New Zealand, Scotland, and the United States.
The situation in Ireland at present is similar to that which pertained in the United Kingdom before the introduction of CTOs with the Mental Health Act 2007 – when patients could only receive care on a compulsory basis in the community as a condition of Section 17 leave (which could be for extensive periods) from an inpatient unit where they notionally remained detained under Section 3 (Mental Health Act 1983). Currently Ireland has provision for approved leave for involuntarily admitted patients under Section 26 of MHA 2001, with conditions specified by the patient’s consultant psychiatrist. The period of leave must not exceed the unexpired period of the relevant admission order. Conditions attached to these periods of approved leave usually include residing in particular accommodation, attending for review and adherence with psychotropic medication. Medication is not compulsorily administered in the community, and the main sanction for non-adherence is withdrawal of the approved leave and return to inpatient care.
In the absence of CTOs in Ireland, such approved leave under MHA 2001 has been used as a surrogate means to ensure adherence with community care for such patients and has led to marked improvements in the quality of life of selected patients as demonstrated by case report (Bainbridge et al. Reference Bainbridge, Byrne, Hallahan and McDonald2014). This option would be curtailed if the recommendation of the Expert Group on the Review of the Mental Health Act 2001 (Department of Health, 2014) to reduce periods of leave to a maximum of 14 days in order to prevent approved leave being used a ‘quasi community treatment order’ is implemented. The restriction of extended leave without a corresponding introduction of CTOs in Ireland would likely lead to prolonged involuntary admissions for a vulnerable cohort of individuals who cannot voluntarily adhere to community care.
It is neither ethical nor logical to withdraw the option of delivering care to such patients with psychotic illness who respond to inpatient care and treatment, but do not recover insight into the benefit of treatment so apparent to their families and clinical team. It makes no sense that coercive care stops at the door of the psychiatric inpatient unit for those patients who continue to lack capacity to consent to the treatment which could aid their further recovery and prevent repeated admissions. Indeed, the failure to extend coercive care to the community setting consigns such vulnerable individuals to a poorer long-term prognosis by replacing the necessary community care to support recovery by revolving door inpatient care and self-neglect in the community.
Distributive justice demands that all citizens have an equal right to treatment and should not be deprived thereof because they lack the capacity to recognize a need for treatment. We do not deprive demented or intellectually impaired individuals of the treatment they need when they lack capacity and equally we should not deprive people with mental illness. This parens patriae duty has long been accepted as a basic responsibility of a society to take care of vulnerable citizens.
The above ethical considerations underpin society’s rationale for inpatient committal. A second essential principle of mental health law is that treatment and care should be provided in the least restrictive location. If people can be treated in their community then that is where treatment should take place. Patients often require a sustained period of treatment before their symptoms resolve to a level where they can develop insight and engage consistently with supports. Due to pressures on the ever decreasing numbers of inpatient beds such sustained periods of treatment will be increasingly difficult to provide in hospital.
International research indicates that psychiatrists value the ability to use CTOs because they provide an alternative to simply letting vulnerable patients discontinue treatment and disengage from clinical services. Discontinuation of treatment is most concerning in those patients who have a history of deteriorating to the point where their illnesses causes them to pose a risk to themselves or to others. However, even in the absence of serious risk, repeated episodes of decompensation, or a single prolonged period of untreated psychosis derails the lives of our patients, preventing them achieving educational, vocational and social goals.
One of us has had 16 years of experience treating patients on CTOs in Canada. Many of these patients are able to break the pattern of revolving door admissions, regain insight, re-establish relationships with family and recover other important functions.
Of course, like any therapeutic approach CTOs do not work for every patient. Patients who respond poorly to medication do not to do well on CTOs and a few patients who have entrenched oppositional personality traits may actively resist the CTO. Often these poor candidates are recognizable a priori, but in some cases a trial of treatment under a CTO is required to determine if it will help.
One positive predictor of a good response to treatment under a CTO is previous robust response to antipsychotic medication and such an individual will almost always do well when the CTO requires that the antipsychotic is delivered by a long-acting injection. Further, patients who show respect or acknowledge the authority of the law typically do well even if they do not recover full insight. It is common to hear patients say ‘I don’t agree that I need this medication, but if I have to take it I will.’
CTOs are not a single entity. Provisions in different jurisdictions determine who can be placed on a CTO, the conditions that can be imposed, the consequences of non-adherence and the degree of bureaucratic burden placed on the clinicians implementing and supervising the CTO. Ireland should not introduce CTOs where the legislation is so bureaucratic that it is seldom used, or so cumbersome that it is ineffective. Fortunately, Ireland, as a late adopter of CTOs, has the advantage of being able to learn from the experience of other jurisdictions in drafting and implementing optimal legislation.
We suspect that Professors Kelly and Burns will ignore the research that shows that CTOs increase the likelihood that a patient will attend outpatient services, take prescribed medication and reduce the likelihood of criminal activity, victimization and hospitalization (Swartz et al. Reference Swartz, Wilder, Swanson, Van Dorn, Robbins, Steadman, Moser, Gilbert and Monahan2010; O’Reilly, Reference O’Reilly2011), and instead propose that only improvement in the primary outcome measure in a randomized-controlled trial (RCT) can be taken as evidence of the effectiveness of a CTO. This blinkered approach not only ignores the extensive research on the positive effects of CTO, it also ignores the fact that RCTs have inherent problems that limit their suitability for evaluating these complex legal interventions.
One major challenge for RCTs in the evaluation of CTOs is the need to obtain informed consent. Patients who are in most need, and therefore most likely to benefit from a CTO – those refusing treatment and follow-up – are especially likely to refuse to take part in research studies. Furthermore, many individuals most suitable for a CTO will be incapable of providing a valid consent.
A more fundamental problem is that RCTs, whilst ideally suited to evaluate simple discrete interventions such as drug treatment, are unsuitable for the evaluation of complex interventions such as CTOs. As noted CTO legislation is highly variable in determining the type of eligible patient and the powers the CTO gives to clinicians. Furthermore, CTOs require the participation and cooperation of many different stakeholders including patients, substitute decision makers, psychiatrists, non-medical clinicians, community agencies, police and review boards. Evaluation of complex interventions is based on the theory of how they work and researchers must determine if the critical aspects of the intervention are delivered as required. An example is a RCT of CTOs conducted in New York (Steadman et al. Reference Steadman, Gounis, Dennis, Hopper, Roche, Swartz and Clarke Robbins2001). To facilitate the study the New York legislature passed a bill to permit the use of CTOs in one hospital, which allowed researchers to randomize one group of patients to follow the law and the control group to be exempted from it. The New York Police Department refused to transport non-adherent patients back to hospital as provided for by the law, thus negating the effect of the law.
For these reasons evaluation scholars and institutions, including the Medical Research Council (Craig et al. Reference Craig, Dieppe, Macintyre, Michie, Nazareth and Petticrew2008), state that the evaluation of complex interventions cannot be accomplished by RCTs measuring a single outcome. Rather, evaluation scholars recommend posing a broader question for complex intervention, which in the case of CTOs would probably be ‘what type of legislation might be effective for what type of patient to achieve what type of outcome.’ A more comprehensive evaluation is needed to answer this type of nuanced question.
In qualitative research studies many patients say that they have benefited from treatment under a CTO. Patients and their families often indicate that it takes about a year of consistent treatment before a patient begins to develop the degree of stability necessary to relate the improvement to the consistent treatment (O’Reilly et al. Reference O’Reilly, Keegan, Corring, Shrikhande and Natarajan2006).
Uncontrolled before and after studies, so-called mirror image studies, avoid the need to obtain consent from participants and are thus able to include all patients on a CTO within a jurisdiction, thereby avoiding selection bias. Mirror image studies consistently find improved outcomes after placement on a CTO (O’Reilly, Reference O’Reilly2011). A weakness of this design is that it is vulnerable to the statistical phenomena of regression to the mean. The findings from controlled before and after studies, which offer some protection against regression, are more equivocal. Unfortunately, it is impossible to retrospectively match subjects and controls for insight or overt treatment refusal: critical reasons that patients are placed on a CTO.
Thus by their nature CTOs are a difficult type of intervention to evaluate. After reviewing all studies one senior group of researchers concluded that the evidence is ‘good enough’ (Swanson & Swartz, Reference Swanson and Swartz2014). We agree that the evidence of benefits is good enough to justify the introduction of legislation in a situation where we know patients do exceedingly poorly when left untreated. Following the introduction of CTOs, Ireland should contribute to research geared to determine how to best use CTOs in the treatment of specific groups of patients.
Most CTOs are initiated after involuntary inpatient care but we recommend including the option of initiating a CTO as an outpatient, to avoid the need for an unnecessary admission. The same checks and balances as underpin respect for human rights when coercive inpatient care is implemented should apply to the use of CTOs. These could include:
-
∙ assessment by two independent consultant psychiatrists before a CTO is applied for an initial period of up to 6 months;
-
∙ renewal for periods of up to 12 months by the treating consultant psychiatrist;
-
∙ specific requirements including that the patient lacks capacity to consent to treatment in the community and that treatment is likely to benefit the patient and prevent substantial harm or mental or physical deterioration;
-
∙ confirmation that a realistic plan to ensure adherence with treatment exists in the community, for example attendance for clinical reviews, administration of medication by long-acting intramuscular injection or supervised administration of medication by a healthcare worker or carer, residing in designated supervised accommodation;
-
∙ review by a tribunal under the auspices of the Mental Health Commission for the initial CTO and each subsequent renewal;
-
∙ the power of the CTO would be mediated through a mechanism to bring the patient to hospital for assessment for inpatient admission and would not extend to compulsorily administering medical treatment outside a hospital setting.
As with coercive inpatient care, CTOs should be used in a selective and tightly regulated manner. The focus should be on that small cohort of vulnerable individuals with psychotic illness who respond to psychiatric treatment, but do not regain capacity to consent to community care, are non-adherent with medical treatment in the community and consequently suffer repeated relapse. Such careful use in Ireland will ensure that CTOs provide less restrictive care by enabling sustained discharge from inpatient units and optimal recovery in the community.
Conflicts of Interest
Colm McDonald has no conflicts of interest to declare. Richard O’Reilly has no conflicts of interest to declare.
Against
Community Treatment Orders: Liberty, Rights and ‘Legislative Creep’
The history of psychiatry is a complex one, characterised by therapeutic enthusiasms and great disappointments, progress in human rights and setbacks in services, achingly gradual scientific advances and endless debates about psychiatry’s role in society (Kelly, 2016 Reference Kellya ). In the late 1800s, the Committee on Lunacy Administration (Ireland) examined the role of mental health law in some detail and concluded that ‘good lunacy laws should make it possible to obtain care and treatment in asylums with ease, but they should make unnecessary detention difficult’ (1891: 36).
Over a century later, Ireland’s Mental Health Act 2001 does not make explicit provision for community treatment orders (CTOs), but states that ‘the consultant psychiatrist responsible for the care and treatment of a [detained] patient may grant permission in writing to the patient to be absent from the approved centre concerned for such period as he or she may specify’ (within the duration of the order) (Section 26). This ‘permission may be made subject to such conditions as he or she considers appropriate and so specifies’, and is, on occasion, used as a de facto CTO (Bainbridge et al. Reference Bainbridge, Byrne, Hallahan and McDonald2014).
Against this background, should Irish legislation make explicit provision for CTOs? No, it should not.
In the first place, the key ethical and human rights issue raised by any proposed healthcare intervention is that it must be therapeutically effective. If a treatment or intervention is ineffective or of unproven efficacy, then providing it to patients is a violation of the basic precepts of medical practice and an abuse of trust. Other contributors to this debate in the Irish Journal of Psychological Medicine examine the clinical evidence relating to CTOs, but it is worth noting here that it is impossible to separate completely issues of efficacy from issues of human rights. If we provide an involuntary treatment that does not benefit the person, we are violating trust, dignity and rights.
Second, the history of psychiatry clearly demonstrates a need for caution with all interventions, especially involuntary ones. One of the reasons for this is the occurrence of ‘legislative creep’, a tendency for legislation to extend its remit and become more coercive over time. As a result, the standard of proof required to justify compulsory interventions must be very, very high. It is simply inadequate to hypothesise that CTOs diminish the need for compulsory inpatient treatment, and thus increase liberty. This certainly makes sense as a theoretical possibility, but the converse is also possible: that CTOs simply extend coercion and compulsion into new areas, with a resultant diminution in net liberty.
This is where ‘legislative creep’ comes in. The yet-to-be-commenced Mental Capacity Act (Northern Ireland) 2016 is a case in point, as it articulates vast powers of compulsion ranging from ‘deprivation of liberty’ to ‘requirements to attend for treatment with serious consequences’ to ‘community residence requirements’. The Act’s safeguards are reasonable but they do not change the fact that it permits extraordinary interference in virtually all areas of the life and liberty of individuals in the community, with exceptionally little by way of rights-based argument or evidence to support it (Kelly, 2016 Reference Kellyb ).
And there are very many compelling, unresolved human rights issues and reasons why the case for compulsion in the community falls far short of the standard of proof required for such draconian measures. For example, the precise criteria upon which CTOs can reasonably be based have never been satisfactorily established in any jurisdiction. Ireland’s Section 26 is impossibly vague, giving enormous discretionary powers with no accompanying guidance about criteria for leave, thresholds to be reached for revoking leave, or the nature or extent of conditions to be imposed. No jurisdiction has satisfactorily filled this gap, owing possibly to the profound contradictions inherent in a person being at liberty in certain ways (e.g. residing at home) but not others (e.g. being required to accept treatment).
Indeed, in Ireland, one of the requirements for involuntary status for many patients is ‘that failure to admit the person to an approved centre would be likely to lead to a serious deterioration in his or her condition or would prevent the administration of appropriate treatment that could be given only by such admission’ (Section 3(1)(b)(i)). It is extraordinarily difficult to see how ‘treatment that could be given only by such admission’ can be given at home.
Is it, then, proposed that clinical criteria for CTOs would be somehow less than those for involuntary admission? If so, such criteria would likely fall well short of the requirements of international human rights law, such as the European Convention on Human Rights, with which involuntary admission criteria in many countries, including Ireland, are now broadly aligned.
There are many other issues (Bartlett & Sandland, Reference Bartlett and Sandland2014). Introducing compulsion in communities would irrevocably change the tenor of community mental health services which should emphasise voluntary engagement and mutuality, rather than hierarchy or coercive practices. In addition, community services remain inadequate to deliver treatment of sufficient certainty and quality to justify compulsion; therapeutic relationships would be changed in negative and enduring ways; and ‘legislative creep’ would ensure that compulsion would soon become a convenient alternative to voluntary outpatient care rather than an alternative to involuntary inpatient care.
There is also a real risk of ‘institutionalisation in the community’, especially in the context of medication delivered under threat of readmission, mandatory requirements to attend for treatment or coercive domestic residential arrangements (as mandated by Northern Ireland’s 2016 Act). Many such patients would likely not be residing at home with their families anyway, but rather compelled to reside in other facilities operated by state services, such as homeless accommodation or mental health hostels. They would therefore likely be subject to other forms of regulation too, such as curfews and restrictions on guests, further darkening this ever-deepening Orwellian dystopia.
In 2015, the Expert Group on the Review of the Mental Health Act 2001 examined this matter, noting the use of Section 26 as ‘a kind of de facto community detention’ (2015: 51). The ‘majority of the Expert Group having looked at this again agreed that the evidence is not convincing that CTOs are effective’. The Group agreed ‘that a leave of absence can be an important part of a patient’s care plan’ but, ‘in keeping with the original purpose of [Section 26], such leave should be clearly subject to a specified time limit which the Group agreed should be at a maximum of 14 days. In addition, the Group would like it to be made clear that the provisions of this section should not be used as quasi–community treatment orders’.
Liberty is a fundamental human right and so the standard of evidence required to support any compulsory intervention must be very, very high. The case for CTOs is unproven at best. The history of psychiatry clearly indicates that, in this circumstance, we need to tend towards liberty.
The 1891 Committee was correct: good mental health ‘laws should make it possible to obtain care and treatment [with] ease, but they should make unnecessary detention difficult’. The argument that compulsory community treatment is ‘necessary’ remains unproven, to say the very least. We should not introduce CTOs in Ireland.
Acknowledgement
This paper contains Irish Public Sector Information licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) licence.
Conflicts of Interest
The present author (B.D.K.) was nominated to the Expert Group on the Review of the Mental Health Act 2001 by the College of Psychiatrists of Ireland and appointed to the group by Minister Kathleen Lynch, Minister for Primary Care, Social Care (Disabilities & Older People) and Mental Health, in 2012. This paper is written in his personal capacity, as a psychiatrist, and does not represent the views of the Expert Group on the Review of the Mental Health Act 2001, College of Psychiatrists of Ireland, or any other group. There is no other interest to declare.
Compulsion has always been a necessary, albeit controversial, feature of psychiatry. With the move out of the asylums over the last half-century it was inevitable that its use in the community would be considered. Applying compulsion to someone who is otherwise able to live independently outside hospital raises concerns about natural justice and Brendan Kelly elegantly lays out the ethical and legal arguments (ref).
Personally, I have long considered the anticipated benefits of community treatment orders (CTOs) to outweigh these very legitimate concerns. Psychoses are terrible disorders inflicting enormous suffering and disruption on patient and carer alike. They merit serious consideration and justify serious measures. Since my time on the RCPsych CTO working party in 1993 (Royal College of Psychiatrists, 1993) I have lobbied for testing them in the care of those psychosis patients with frequent relapses and poor compliance. We recognised the ‘lobster pot’ potential of CTOs and the difficulty of predicting the course of psychoses so proposed a 3-year limit for their duration. We took for granted that their introduction would be subject to careful monitoring and, hopefully, rigorous testing.
By the time of the 2007 CTO legislation in England and Wales our high hopes were tempered by a commissioned review of the evidence (Churchill et al. Reference Churchill, Owen, Singh and Hotopf2007). This identified 72 studies, of which only nine had an experimental design and only two were randomised controlled trials (RCTs). It concluded that there was no convincing evidence that CTOs had a significant impact on patient outcomes and that more rigorous research was urgently needed. By 2014 a further 18 studies had been added (Maughan et al. Reference Maughan, Molodynski, Rugkåsa and Burns2014) but only one of these was a RCT. OCTET (Oxford Community Treatment order Evaluation Trial) was a pragmatic RCT conducted across over 30 Trusts in England in clinical conditions likely to mirror those in the Irish Republic (Burns et al. Reference Burns, Rugkåsa, Molodynski, Dawson, Yeeles, Vazquez-Montes, Voysey, Sinclair and Priebe2013). Like the two RCTs before it, OCTET found no benefits from CTOs in the primary outcome of reducing rate of rehospitalisation over the 1 year follow-up, nor in any other of the hospitalisation outcomes (delay, duration, etc.).
The hierarchy of evidence
Conducting an RCT of CTOs is not at all easy and it is no surprise that evidence has been sought in other ways. The most common method is a before-and-after study in which hospitalisation is compared for the same duration before and after the imposition of the CTO for each patient. Most of these studies are conducted without controls, blithely ignoring the inevitable effects of regression to the mean in such fluctuating conditions. They should simply be discounted. A smaller number, however, do include carefully matched controls and some use propensity matching. In the absence of available RCTs such studies, which can establish correlation but not causation, can still provide strong evidence. Their quality varies markedly, however, and this is a murky literature with the same databases reported several times and in various ways. The most that can be said is that the results are conflicting. In some studies CTOs reduce admissions, in others increase them and in a few have no effect whatsoever. The most influential such study is from New York and shows a massive reduction in admission (Swanson et al. Reference Swartz, Wilder, Swanson, Van Dorn, Robbins, Steadman, Moser, Gilbert and Monahan2013). But read carefully! A CTO in this study was the required ticket to decent care provided for by millions of dollars of ring-fenced funding. No surprise they did better.
When cohort studies are conflicting or unclear then the accepted procedure in medicine is to move up the hierarchy of evidence. For CTOs this replaces uncertainty with consistent and clear findings. All three RCTs and all four published systematic reviews find no significant benefit in any of the hospitalisation outcomes (Rugkåsa & Burns, Reference Rugkåsa and Burns2017). No matter how you cut it CTOs simply do not deliver what they promise and what we hoped for. Would you trust a doctor who prescribed a drug for which all the rigorous evidence was negative but he still thought it was a good idea?
Criticisms of the RCTs
All three RCTs were pragmatic trials. They tested CTOs in real-life conditions, powered to allow for some inevitable mess. All have their limitations (exclusions and reduced follow-up in the US studies, cross overs in OCTET) but even after sensitivity analyses no significant differences emerge. All the patients in these trials were ‘proper’ CTO candidates (legally, of course, they have to be) with clinical characteristics and histories as severe as those in the cohort studies. They are not selected ‘easy’ patients.
Yes we want more trials and bigger trials, but these are not poor trials. Altogether 739 patients have been randomised; 155 of 373 CTO were readmitted in the follow-up year versus 153 of 366 controls (41.6% versus 41.8%). How close do the numbers have to be to accept that there really is nothing going on?
The three RCTs have also been criticised for all having admissions as the primary outcome and some have suggested more ‘relevant’ patient outcomes such as social functioning or satisfaction. RCTs have to have a single primary outcome that researchers and clinicians agree is clinically decisive. It is hard to imagine how longer-term social outcomes in psychosis could be improved without the illness first being stabilised. Searching for these benefits has also proved illusory, as have attempts to identify ‘responsive’ clinical subgroups.
The trials have also been criticised because of the limited follow-up of 1 year. Yet over a third of patients do relapse and are readmitted in that timeframe, so there is ample scope to demonstrate an impact if there were one. A particularly persistent myth is that there is some step change in effectiveness after 6 months. This stems from a non-randomised secondary analysis in the North Carolina study and you would be well advised to read this paper carefully (Swartz et al. Reference Swanson, Van Dorn, Swartz, Robbins, Steadman, Mcguire and Monahan1999). In particular, examine the graphs on which this, now very influential, misunderstanding is based. Notwithstanding the tiny samples of 30 patients it is patently obvious from their slopes that admission rates deviate from day one – this is a selection effect, not a treatment effect.
Consequences of introducing CTOs
Despite the strength and consistency of the evidence the appeal of CTOs remains strong. However, everyone apart from politicians will lose by their introduction. Patients will find their liberty restricted without any clinical benefit in return. Carers will have their hopes falsely raised. Clinicians have probably most to lose. CTOs bring with them bureaucratic and time-consuming obligations that distract us from other more effective and evidence-based activity. Perhaps most important it will perpetuate an image of our profession as one that follows fads and fashions and ignores the evidence when it is inconvenient.
Psychiatry, more than any branch of medicine, needs to keep its feet on the ground and stick to evidence. History is not on our side in this matter and we need to strengthen our credentials as trustworthy and reliable medical professionals. Introducing CTOs in the face of consistent evidence of their ineffectiveness discredits us. If it is to be done in Ireland then it should be in the context of a bigger, better RCT and a commitment to abide by the results. It is not enough simply to hope or believe. We owe more to our patients, and to ourselves.
Conflicts of Interest
None.
For
Prof. Burns bases his argument on the lack of difference in the primary outcome of hospitalisation in three RCTs examining CTOs and blithely discounts the other evidence in their favour. Due to the difficulties of randomising one group of subjects to be under a law and another exempt, all three studies involved compromises that limit applicability. In the recent English study CTOs were compared to S.17 leave – another form of involuntary community treatment – rather than to voluntary care. RCTs require subject consent and are therefore inappropriate as a primary evaluation of interventions for patients who lack capacity, refuse or default from treatment. Psychiatrists understand the utility of using long-acting injections of antipsychotic medication and observational studies show a clear advantage for this approach. In contrast a meta-analysis of RCTs showed no advantage over oral administration (Kishimoto et al. 2013): the same pattern seen in research on CTOs. Certainly more research should be performed, in particular to identify for which patients and community supports CTOs work best – which can only take place in the Irish context if they are introduced here.
Prof. Kelly agrees with us that the current use in Ireland of leave as a form of CTO for very difficult patients will likely be abandoned. Many of these patients will then remain in hospital. Others will inevitably be exposed to various types of informal coercion. CTOs provide mandatory oversight and statutory rights including the right to appeal – far better than unregulated informal coercion. Prof. Kelly raises the possibility of legislative creep but provides no evidence for it. We simply propose that, for patients who lack capacity, the requirements to follow a realistic treatment plan should not be restricted to the inpatient ward, and that established legal protections for inpatient involuntary treatment should continue to apply.
Against
Professors McDonald and O’Reilly provide an excellent overview of the history and use of CTOs. What they fail to provide is any convincing case for Irish Psychiatrists to use them. Why overlook the accumulated evidence that they simply do not work? Why abandon the benefits of 50 years of EBM practice? Why ignore the consistent finding of the three RCTs that CTOs (despite their obvious appeal) do not benefit patients?
Our interpretation of the evidence is not ‘blinkered’; it includes effectiveness studies, human rights and ethics. It is, however, disciplined. Discipline and care are necessary in interpreting medical evidence. Beware those who skate over the key findings to dig out some exciting secondary nugget that fits their case.
In fairness, Professors McDonald and O’Reilly propose some good principles for the implementation of community care in this difficult patient group and excellent human rights protections if CTOs are to be introduced. However, they still miss two fundamental points. First, an ineffective intervention, no matter how carefully implemented, is still an ineffective intervention and does nobody any good. Second, any intervention that does not work is a violation of rights, regardless of safeguards.
People with mental health problems are entitled to the same level of attention to evidence in their care as people with any other health problems. The current evidence from high-quality studies indicates CTOs simply do not work. Patients with severe mental illnesses should not have them imposed on them – they have more than enough disadvantages already to contend with.
