Community treatment orders are (somewhat) effective: Their future in the context of rights-based mental health law

A study by Harris et al. (2018) provides the best evidence to date that community treatment orders (CTOs), while in operation are, at least somewhat, effective in reducing psychiatric readmission. However, since the study was conducted between 2004 and 2009, it provides data from a period before any Australian mental health legislation hinged involuntary treatment to a patient’s decision-making capacity. Since 2013, every Australian jurisdiction, except the Northern Territory, has reformed its mental health legislation to either prohibit or strongly discourage forced psychiatric treatment in those who competently refuse it (Callaghan and Ryan, 2016). It is therefore important to examine the impact of the Harris et al. findings in the context of the contemporary legislative environment.

The study used health department data from New South Wales (NSW) to compare the outcomes of 5548 patients, who received a CTO, with controls selected via matching propensity scores derived from demographics and clinical care received in the 30 months prior to the first CTO. The investigators found that compared to controls, people on CTOs were less likely to be readmitted, had fewer hospitalisations, and the period prior to admission was longer. Despite this though, most CTOs had no effect on the number of days spent in hospital. The effects on readmission were statistically significant, but they were not strong. The number of people who needed to be subject to a CTO to result in one fewer hospital admission was 10, and 48 people needed a CTO to result in one additional person having no hospital readmissions. Though there is no way of knowing from the data how CTOs exerted their effect on readmission, it is reasonable to speculate that it may have been via increased contact with community health services. People on CTOs received significantly more input than controls, averaging 3.9 community contacts per month, compared to just 2.5.

The first thing to note is that during the period of the study, many people who were made subject to CTOs would have retained the ability to competently refuse outpatient psychiatric treatment as, at that time, NSW legislation made no reference to decision-making capacity as a determining factor in the use of CTOs. Recent legislative changes in Queensland, South Australia, Tasmania and Western Australia prohibit the use of coercive treatment if a person is able to competently refuse it (Callaghan and Ryan, 2016). In NSW, Victoria and the Australian Capital Territory, it is still possible to deliver coercive care to people competently refusing it, but legislative reforms should mean this is occurring far less frequently. Since 2015, NSW clinicians have been required to ‘make every effort that is reasonably practicable’ to obtain patients’ consent and consider their views and expressed wishes when developing management plans (Ryan and Callaghan, 2017). Changes like this should mean that individuals who competently refuse treatment should only very rarely be made subject to a CTO, although there is evidence that these reforms are yet to filter down into changes in the practice of many clinicians and mental health tribunal members (Maylea and Ryan, 2018).

The reforms should also have had an impact on the way the CTOs are imposed even when patients lack decision-making capacity. Though it is expressed differently in different jurisdictions, current Australian mental health legislation asks clinicians to carefully consider if the incompetent person that they plan to make subject to a CTO would have consented to a CTO, had he or she been competent. It is usually expected that this judgement will be made by reference to views of the person him or herself, as expressed at the time, or earlier via an advanced care directive. Clinicians are also expected to take into account the views of family and friends, to the extent that they can shed light on what the person would have wanted, if they had been competent.

Now, these discussions should be informed by the findings of the Harris et al. study. The patient, friends and family should be told that while people subject to CTOs do not on average spend less time in hospital, they do on average have fewer readmissions and on average remain out of hospital longer. They should be told that this effect is real, but very weak and most individuals will not derive these benefits – a message that can be quantified using the numbers needed to benefit set out above. Clinicians should also acknowledge that these modest benefits may not derive from the CTO’s coercive effect, but via an administrative mechanism that signals to community health services that patients on CTOs should have priority access to care. Finally, drawing on other literature, patients and their loved ones should be told many people object to being on a CTO and understandably feel coerced and controlled (Corring et al., 2017). Moreover, while some people on CTOs approve of them, this is often because they see the CTO as a better option than hospital (Corring et al., 2017), a rationale that is largely undermined by the Harris et al. findings.

Of course, it is possible that some patients and their families armed with all this information will still opt for a CTO. After all, many people take medications to avoid cardiovascular events, with numbers needed to benefit far larger than 48. However, few people will find statins as aversive as many find CTOs.

In conclusion, while the findings of the Harris et al. study strongly support the efficacy of CTOs, they are of limited direct applicability in an era where few people who competently refuse treatment should still be subject to CTOs. To the extent that they are relevant, it is likely they will lead to many people who are trying to anticipate what the patient would have wanted if competent, to conclude that he, or she, would have rejected CTO enrolment.

CTOs are diminished but, not yet, defeated. It is possible that CTOs are far more potent when their use is restricted to those who lack decision-making capacity. Perhaps the benefits of CTOs were so modest in the Harris et al. cohort because they were being used indiscriminately and a possible strong benefit when directed to incompetent patients was being ‘washed out’ by their use in many people capable of deciding for themselves.

There is still a place for ongoing CTO research in patients who are largely recovered, but who still, despite our best efforts at providing support and education, either do not understand the information relevant to the utility of treatment or cannot use and weigh that information as a result of the effects of their mental illness. The state has long recognised a responsibility to protect those who cannot, by reason of mental incapacity, protect themselves, so to the extent that CTOs might be efficacious this should surely be the population targeted. Perhaps, a replication of the Harris et al. study in coming years, enrolling only those who lacked decision-making capacity, may show a considerably stronger treatment effect, and perhaps then, finally, we will be able to confidently recommend CTOs to substitute decision-makers.

In the meantime, CTOs should only be used in those who lack decision-making capacity and then with all the caveats that the findings of this study entail.