NZLII Home | Databases | WorldLII | Search | Feedback

Otago Law Review

University of Otago
You are here:  NZLII >> Databases >> Otago Law Review >> 2018 >> [2018] OtaLawRw 6

Database Search | Name Search | Recent Articles | Noteup | LawCite | Download | Help

Christie, Alex --- "Is preventative use of Treatment Orders lawful in New Zealand?" [2018] OtaLawRw 6; (2018) 15 Otago LR 285

Last Updated: 12 May 2021

Is Preventive Use of Community Treatment Orders Lawful in New Zealand?

285


Is Preventive Use of Community Treatment Orders Lawful in New Zealand?

Alex Christie

I Introduction

This article considers whether the legal criteria governing compulsory treatment under New Zealand’s mental health legislation permit a person to be kept under a Community Treatment Order (CTO) for ‘preventive’ purposes. Can they be kept under compulsion even when their condition is currently stable and they present no immediate threat of harm to themselves or others, to prevent their future relapse? In particular, this article debates whether such a person can be considered to meet the legal test of posing a ‘serious danger’,1 even when stable, and considers the approach taken by New Zealand courts and tribunals to the application of this test.

A The CTO Regime

Compulsory psychiatric treatment in New Zealand is governed by the Mental Health (Compulsory Assessment and Treatment) Act 1992 (MHA). This Act provides the civil route of entry into compulsory psychiatric care. It permits people to be placed under two types of compulsory treatment orders: ‘inpatient orders’, where the patient is detained in the hospital for treatment,2 and community treatment orders’, where the patient must accept treatment administered in their home or a community facility, but cannot be detained there.3 CTOs were introduced in the wake of the deinstitutionalisation of psychiatric hospitals, following campaigns by civil rights activists and clinicians for the care of mentally ill patients in a less restrictive setting.4 The MHA gives priority to CTOs over Inpatient Orders, requiring the order to be a CTO unless the patient cannot adequately be treated as an outpatient.5
CTOs in New Zealand today have ‘less restrictive’ characteristics, as the patient is able to live in the community rather than in a hospital setting.6 However, they also have ‘preventive’ characteristics. This means

* Solicitor, Anderson Lloyd, Auckland. This article is adapted from the winning entry of the 2017 Joshua Williams Memorial Essay Prize.

  1. Paragraph (a) of the definition of “mental disorder” in s 2 of the Mental

Health (Compulsory Assessment and Treatment) Act 1992 [MHA].

  1. Section 30(1).
  2. Section 29(1).
  3. Tom Burns “Compulsion in Community Mental Health Care: Historical Developments and Current Provisions” in Andrew Molodynski, Jorun Rugkåsa and Tom Burns (eds) Coercion in Community Health Care: International Perspectives (Oxford University Press, Oxford, 2016) 13 at 14.
  4. Section 28(2).
  5. Burns, above n 4, at 16.


patients can be kept on CTOs in order to prevent them acting violently in the future, even if they do not present an imminent threat of harm to themselves or others, based on a longitudinal assessment of ‘risk’.7 This preventive use of CTOs can be both desirable and problematic. Maintaining a patient on a CTO can be advantageous as it may prevent their deterioration into a mental state in which they pose a ‘serious danger’ to self or others, thus addressing safety concerns. But the preventive use of CTOs can also be considered undesirable as it may disproportionately affect patients’ rights to keep them under compulsion when they are currently in a stable condition. In New Zealand, CTO use is high compared to international use, and is increasing annually.8
Whether a person can be kept under a compulsory order depends on whether they meet the MHA’s complex definition of ‘mental disorder’ (or a ‘mentally disordered’ person).9 This definition comprises two parts, or limbs. The first requires that a person has an ‘abnormal state of mind’, characterised by one of the listed disorders. The second limb requires that this abnormal state of mind ‘poses a serious danger’, or seriously diminishes the person’s capacity for self-care. A person can be placed on a compulsory treatment order if they are ‘mentally disordered’ in this sense and the order is deemed necessary.10 But if they cease to be mentally disordered at any time, they must immediately be taken off the order.11 Consequently, to remain on a CTO, the person must be mentally disordered at the present time. This presents the main interpretive problem to be considered in this article: ‘is’ a person ‘mentally disordered’ in the necessary sense, when their mental health is reasonably stable; or, in those circumstances, should they be released from control under the MHA, to protect their rights?

B Approaches to Interpretation

Two main approaches can be taken to the interpretation of the statutory definition of ‘mental disorder’: the purposive approach, and the rights- driven approach.12 The purposive approach, codified in s 5(1) of the Interpretation Act 1999, states that: “The meaning of an enactment must be ascertained from its text and in light of its purpose.” The rights-driven approach involves choosing an interpretation which is consistent with patients’ rights under the New Zealand Bill of Rights Act 1990 (NZBORA).

  1. Burns, above n 4, at 16.
  2. Anthony J O’Brien “Community Treatment Orders in New Zealand: Regional Variability and International Comparisons” (2014) 22 Australasian Psychiatry 352 at 353.
  3. Section 2.
  4. Section 27.
  5. Waitemata Health v Attorney-General [2001] NZCA 312; [2001] NZFLR 1122 (CA) at [95] per Elias CJ, Richardson P, Gault and Thomas JJ and at [118] per Tipping J.
  6. John Dawson “The Process and Criteria for Compulsory Psychiatric Treatment” in Peter Skegg and Ron Paterson (eds) Health Law in New Zealand (Thomson Reuters, Wellington, 2015) 425 at 435.


Section 6 of the NZBORA states that an interpretation consistent with the rights contained in the Act should be preferred if possible. However, the MHA has the dual purposes of allowing compulsion and protecting patients’ rights, so different purposes may be emphasised in different situations. Furthermore, different views of what promotes patients’ rights may be taken. For example, placing a patient on a CTO rather than an ‘inpatient order’ may be viewed as protecting their right to liberty by allowing them to live in the community. However, the same decision may be viewed as restricting their right to liberty by forcing them to accept treatment at all.
Consider, for example, the situation in Re PT.13 PT had a schizophrenic type mental illness, but did not believe he was mentally disordered. He was admitted to hospital following an application for a CTO, but was subsequently fully discharged from compulsion. He was later readmitted to hospital under the MHA following a severe relapse of his condition after he stopped taking medication. He was placed under a CTO and some months later this order came before the District Court for review. At the time of the review hearing, PT was well-nourished, living at home with family support, not considered to be actively dangerous to himself or others, and his mental health was reasonably stable under his medication regime. Taking one approach, PT should have been taken off the CTO as his condition was stable and it would have been disproportionate to his rights to keep him on it. Taking another approach, PT should have been kept on the CTO, as, though his condition was stable, if he was taken off the CTO he may have stopped taking his medication again and relapsed into a severe condition. But would he have been considered ‘presently’ mentally disordered in the necessary sense at the time of the hearing when his condition was reasonably stable? This example highlights the importance of interpretation, as PT’s liberty was directly dependent on the approach taken.
PT was placed on a CTO, despite his stable condition at the time.14 Whether this decision was legally correct requires analysis of the main interpretations given to the concept of ‘mental disorder’ by New Zealand courts and tribunals, and whether these interpretations are convincing. The most contentious part of the definition is the ‘serious danger’ test in the second limb. Therefore, the main legal question this article will address is whether the dominant interpretation of the concept of ‘poses a serious danger’ is correct, and whether the preventive use of CTOs is legal on that basis.


  1. Re PT [2001] NZFLR 79 (FC).
  2. Judge Walsh held that PT responded well to medication, but “his longitudinal history shows that when he is non-compliant with medication, he will suffer a relapse”: Re PT, above n 13, at [18]. Therefore a CTO was necessary to make sure PT was compliant with treatment: at [17]–[18].

II Analysis of the ‘Serious Danger’ Test

  1. The Dominant Approach to the Interpretation of ‘Serious Danger’ I will first provide an account of the current approach to interpretation of ‘serious danger’ taken by the courts and the Mental Health Review
Tribunal (MHRT). When reviewing the status of compulsory patients, the Tribunal considers the ‘serious danger’ test according to four main parameters: the ‘nature’ or ‘magnitude’ of the likely harm, and its ‘probability’, ‘imminence’, and ‘frequency’.15 Evidence that is both intrinsic to the patient, for example their mental state, and extrinsic, for example where they live and their family support, is relevant.16
A gradual widening of the notion of ‘serious danger’ has occurred since the introduction of this test in 1992. ‘Serious danger’ was initially limited to physical, imminent violence,17 but the MHRT soon abandoned these requirements.18 Currently, a patient does not have to score highly on all four parameters to pose a ‘serious danger’, so although imminence is often sufficient, its absence is not decisive. For example, if the predicted harm is very serious, or has occurred at a high frequency in the past, the patient may still pose a ‘serious danger’ even if their condition is currently stable.19 The MHRT has held that likelihood of relapse is critical in predicting future danger. It considers factors such as “the degree of causal connection between relapse and dangerousness”, the expected time lapse between discharge from the MHA and cessation of treatment, whether interventions by others can prevent a relapse in its early stages and the ability of clinicians to reinitiate assessment under the MHA.20

B Analysis of the ‘Serious Danger’ Test

The ‘mental disorder’ definition requires the patient to be currently ‘mentally disordered’. This suggests that they must pose a ‘serious danger’ right now, even when their condition may be relatively stable. Therefore, there needs to be a notion of danger that is enduring, or continuous in some latent sense, for this test to be satisfied. Is there such a concept of ‘dangerousness’, and can it be enduring even if the patient is not presently acting violently?

1 Danger as an enduring characteristic, or a situational occurrence?

If a person is to pose a ‘serious danger’ at the present time, even when

  1. Re Mental Health [Serious Danger] [2012] NZFLR 1 (MHRT) at [57].
  2. Re KMD MHRT Invercargill 04/139, 23 May 2005 at [41].
  3. In In the matter of JK, it was held that the “serious danger” test requires “a demonstrable risk of no less than serious physical violence”: In the matter of JK [1994] NZFLR 678 (FC) at 702.
  4. Re GM (Mental Health) [2001] NZFLR 665 (DC) at 670–671.
  5. Re Mental Health [Serious Danger], above n 15, at [59].
  6. Re KMD, above n 16, at [47]. In Re Mental Health [Serious Danger], above n 15, at [60]–[61], the Tribunal held that these considerations are in effect a restatement of the four parameters already referred to at [57].


their condition is relatively stable, this might seem to suggest they need to be inherently dangerous. It suggests that ‘dangerousness’ would be some enduring part of their character, which continues, or is present, right now, even though they may be calmly seated in a chair, having not assaulted anyone in months. However, the idea that ‘dangerousness’ is an enduring characteristic of this kind reflects outdated theories, and is not in line with contemporary psychiatric ideas of the concept of ‘danger’ in mental health settings.21 Instead, mental health professionals currently understand ‘dangerousness’ as a situational occurrence. This takes into account both intrinsic factors to the person and extrinsic, situational factors present at a certain time,22 and suggests biological and psychological factors are just parts of a multifactorial interaction which produces violence.23 The idea is that violence is produced when a person’s individual neurobiology and their mental state interact with situational variables that are present at a certain time, such as being intoxicated, or running into a loathed ex-partner.
The notion of ‘danger’ as a situational occurrence, rather than an enduring character trait, is therefore more in line with current psychological thinking. But we still face the question: what if the relevant situational factors are not present at the current time, and therefore no violence is occurring? It seems difficult in that case to describe the person as posing a ‘serious danger’ right now, as the definition seems to require. But it seems sensible to consider ‘danger’ as a situational occurrence, as that is in line with current professional thinking, and to use that approach in our interpretation of what posing a ‘serious danger’ means. So, can the concept of ‘danger’ in the MHA be understood as a situational occurrence, but one which also allows for current nonviolence?

2 The concept of ‘risk’

Understanding the notion of ‘danger’ within a ‘risk paradigm’ permits this interpretation to be taken of the ‘serious danger’ test. It may be legitimate to understand ‘danger’ in this way, as psychiatrists undertake risk assessment to determine whether a patient may act violently in future, and Parliament places psychiatrists at the centre of the assessment process under the MHA. For a person to pose a ‘serious danger’ right now, the concept of ‘danger’ would have to be equated with ‘risk’, and the concept of ‘risk’ would need to be considered enduring or continuing and therefore present even when a patient is stable and not acting violently. So, in this part I will analyse the concept of ‘risk’ to see whether it can be considered enduring, in this sense.
  1. Arie Bauer and others “Dangerousness and Risk Assessment: The State of the Art” (2003) 40 Isr J Psychiatry Relat Sci 182 at 183.
  2. Richard Whittington and Dirk Richter “From the Individual to the Interpersonal: Environment and Interaction in the Escalation of Violence in Mental Health Settings” in Dirk Richter and Richard Whittington (eds) Violence in Mental Health Settings: Causes, Consequences, Management (Springer, New York, 2006) 47 at 48.
  3. Whittington and Richter, above n 22, at 49.

(a) Definitions of ‘risk’

Defining ‘risk’ is more difficult than it seems, as there is no one recognised meaning for the concept. Although risk may be given a narrow definition, it is simultaneously “substantiated in myriad forms”.24 In the social sciences, ‘risk’ is commonly defined on two dimensions: one concerning probability, and the other concerning effect. It can be the probability of a negative event occurring during a certain time period, or a measure of the severity of the harm caused by the negative event.25 Although these conceptions of ‘risk’ can be differentiated, they have been amalgamated over time and the term ‘risk’ is often used to refer to both. Upon the emergence of ‘risk’ in the scientific realm, it was thought of as an objective concept, capable of quantitative assessment and prediction.26 However, the use of ‘risk’ in everyday parlance has moved away from probability calculations, and the term is now often loosely used to refer to threats, danger, or potential harm.27 The term ‘risk’ is also often used to refer to ‘uncertainty’, but these concepts are not the same.28 Whereas ‘risk’ is usually based on known probabilities, the term ‘uncertainty’ is used to denote situations where probabilities are unknown and not calculable.29 Despite this distinction, ‘risk’ is usually used to refer to situations where there is a likelihood of harm, regardless of whether the probability of this is calculable or not.30

(b) The case that ‘danger’ (or ‘risk’) is a continuing concept

In the context of risk assessment in the mental health setting, the Ministry of Health, in official guidelines, has described ‘risk’ as “the likelihood of an adverse event or outcome”.31 ‘Risk’ is viewed as capable of having a probability attached, rather than as an uncertainty where probabilities are unknown.32 Thus, despite the move away from probability calculations in other arenas, there is still a strong focus in the mental health setting on ‘risk’ as a probability of a certain person acting violently. Though it may fluctuate, it may still be present all the time. On this reasoning, it could be said that ‘risk’ in a mentally disordered person who presents certain factors is enduring and continuously present, and if it is serious enough, it will necessitate their compulsion.

  1. Glynis M Breakwell The Psychology of Risk (2nd ed, Cambridge University Press, Cambridge, 2014) at 3.
  2. At 3.
  3. Deborah Lupton Risk (2nd ed, Routledge, Abingdon, 2013) at 7. 27 At 10.
  4. Sven Ove Hansonn “Risk” (11 August 2011) Stanford Encyclopedia of

Philosophy <https://plato.stanford.edu/entries/risk/>.

  1. Sven Ove Hansonn “What is Philosophy of Risk” (1996) 62 Theoria 169

at 170.

  1. Hansonn, above n 28.
  2. Ministry of Health Guidelines for Clinical Risk Assessment and Management in Mental Health Services (July 1998) at 2.
  3. Hansonn, above n 29, at 170.


I submit that this conception of ‘risk’ as a probabilistic and enduring concept is a convincing one, suitable for use in this context, in light of the purposes of the MHA. One purpose of the MHA is to allow compulsion, and viewing ‘risk’ as an enduring concept fits this purpose as it will allow compulsion of all those who pose a serious enough ‘risk’ of causing harm in the future. It may also be fair to say that the MHA has some preventive purposes, as though it probably does not authorise indefinite compulsion, it is implausible that a patient should be released as soon as they stop acting violently. For example, if a patient was violent for three days, and then stopped acting violently for the two days following, should they immediately be released? This was surely not Parliament’s intention, so there must be some sense in which the ‘danger’ is meant to be viewed as continuing.

(c) The most convincing conception of ‘risk’ in this legal context Finding the best interpretation of ‘risk’ should turn on which is most

convincing in light of the purposes of the MHA, reflecting the context
in which the concept is used. Considering ‘risk’ as capable of having probability attached is most in line with psychiatric risk prediction. Thus there is a strong argument that this conception of ‘risk’ as an enduring probability is the correct interpretation in the mental health setting. Surely Parliament intended that psychiatrists would try to predict the probability of compulsory patients acting violently if not controlled under the MHA.
Furthermore, though the NZBORA requires a rights-consistent interpretation to be given, it only requires this where that interpretation is viable or ‘can’ be given.33 It may be argued that the only plausible interpretation of the ‘poses a serious danger’ test with regard to the purposes of the MHA is that of ‘risk’ as enduring, as this is in line with how risk prediction is conducted by psychiatrists. In addition, though this interpretation is less rights-friendly than ‘risk’ as a non-enduring, or occasional concept, the inclusion of the term ‘serious’ in the ‘serious danger’ test provides a separate safeguard against too much compulsion. It may therefore be legitimate to understand the notion of posing a ‘serious danger’ in this context in terms of current psychiatric conceptions of ‘risk’, provided the risk is of a necessary degree to be called ‘serious’.

III The Courts’ Approach to Interpretation in Preventive Cases

The next question concerns how the courts (and the MHRT) actually interpret and apply the ‘mental disorder’ definition, in cases where CTOs are used preventively. Do they interpret ‘serious danger’ within a ‘risk paradigm’, and is this consistent over time and between cases?



  1. New Zealand Bill of Rights Act 1990 [NZBORA], ss 4 and 6.

A Do the Courts Interpret ‘Serious Danger’ Within a ‘Risk Paradigm’?

In this analysis, I will focus mainly on the approach taken by the MHRT. As the Tribunal’s role is to determine whether patients should be discharged from compulsion or not,34 it is able to approve the preventive use of CTOs. Furthermore, it has considered the interpretation of the concept of ‘mental disorder’ in the most depth. There are many indications in preventive cases that the MHRT is interpreting ‘serious danger’ within a ‘paradigm of risk’, some of which are explicit, and some which are more implicit, in the reasoning. I will analyse these indications below.
  1. Direct references to ‘risk’ and risk assessment
In some cases, the MHRT makes direct reference to assessing ‘risk’.35 This can be seen even in older cases, where the stricter approach to ‘serious danger’ was used. For example, in In the matter of T, a case from 1994, though there is emphasis on the need for physical and imminent violence to meet the ‘serious danger’ test, the MHRT still makes reference to “predicting future conduct”, by undertaking what it calls an “assessment of risk”.36 The MHA Guidelines also describe the four parameters as useful in “conducting a risk assessment”.37 Often, the MHRT will not refer to risk assessment specifically, but will still refer to future ‘risk’ and the degree to which it exists, in its approach.38 This seems to suggest that the Tribunal’s determination of the ‘risk’ involved in each case is at the core of its decision whether a person ‘poses a serious danger’, even if the Tribunal makes no explicit reference to risk assessment itself.
  1. The use of probabilistic terms
The use of probabilistic terms by the MHRT has become increasingly common. Originally, patients were assessed using the parameters of the

  1. Nigel Dunlop “The Mental Health Review Tribunal” in John Dawson and Kris Gledhill (eds) New Zealand’s Mental Health Act in Practice (Victoria University Press, Wellington, 2013) 97 at 97.
  2. In Re RJ MHRT Dunedin SRT80/98, 30 October 1998 (date of hearing) at 8–9, the Tribunal referred to the need to “make a determination on the risks imposed”; and in In the matter of T [1994] NZFLR 946 (MHRT) at 950–951, the Tribunal referred to undertaking “the assessment of future risk”, and noted that “[p]redicting future conduct is a notoriously difficult task”, so “[w]hat is undertaken is an assessment of risk”.
  3. In the matter of T, above n 35 at 951.
  4. Ministry of Health Guidelines to the Mental Health (Compulsory Assessment and Treatment) Act 1992 (November 2012) at [1.1.5].

38 In Re CDW DC Nelson FAM-2003-009-004445, 29 June 2011 at [33], the

Court stated that “[p]ast history is always a good predictor of future risk”; in Re VAP MHRT Dunedin 09/122, 11 December 2009 (date of hearing) at [27], the Tribunal noted that “[t]he fact that the Applicant has not assaulted others since 1999 attests not to the absence of risk, but to the treatment which she has received”; and in Re GM (Mental Health) [2001] NZFLR 665 (DC) at 670–671, the Court states that serious danger will be established where there is a “real and substantial risk to the lives of other people”.



‘nature of likely harm’, ‘magnitude’, ‘imminence’ and ‘frequency’.39 This showed a kind of probabilistic analysis, as the Tribunal was collating evidence for each of these considerations, and then deciding on the likelihood (or probability) that a patient would act violently in the future. Recently, a slight alteration of these parameters demonstrates a further trend toward probabilistic thinking. In Re Mental Health [Serious Danger], the considerations of the ‘nature’ and ‘magnitude’ of the harm have been amalgamated and an extra consideration added: the ‘likelihood’ or ‘probability’ of the likely harm occurring.40 In that case, the Applicant had a schizophrenic type illness, was subject to an indefinite CTO, and lived independently in the community. Though he had no criminal record and no past history of physical violence, the Tribunal found that the nature of his delusions, threats made to others and their proximity to him, the absence of insight, and the likelihood of immediate relapse were he to go off the order necessitated the order being continued.41 It is unclear whether this shift was intentional or not, but, even if it was not, it suggests that probability, a central notion in risk assessment, is becoming a key part of the Tribunal’s approach. References to ‘prediction’ of future violence have also been made in some preventive cases. For example, in Re CDW, the District Court stated that “[p]ast history is always a good predictor of future risk”,42 showing the use of probabilistic thinking through prediction of the likelihood of the patient acting violently in future.
  1. Analysis of ‘risk factors’
The MHRT sometimes explicitly undertakes an analysis of ‘risk factors’ in order to determine whether a patient poses a ‘serious danger’. The most striking example of this is Re Mental Health [Serious Danger]. There, the MHRT directly considers indicators and contra-indicators as to whether the Applicant poses a ‘serious danger’, and concludes that the “indicator factors outweigh the contra-indication factors”.43 The factors considered include the Applicant’s mental state, previous history of violence, and situational factors such as the proximity of victims and the Applicant’s living situation, showing consideration of intrinsic and extrinsic factors, which is in line with current psychiatric thinking about ‘danger’ and ‘risk’.44
  1. Longitudinal approach to interpretation
In most preventive cases, the MHRT states that it must take a
“longitudinal” approach to determining whether a patient ‘poses a

  1. Re KMD, above n 16.
  2. Re Mental Health [Serious Danger], above n 15 at [57]. 41 At [63]–[64].
    1. Re CDW, above n 38, at [33].
  3. Re Mental Health [Serious Danger], above n 15 at [63]–[65]. 44 At [61]–[64].


serious danger’.45 This means the Tribunal considers not only the patient’s current state at the time of the hearing, but their history as well, including their mental state when unwell,46 history of relapses,47 past incidents of violence and drug use,48 and other factors relevant to their condition. This is effectively another form of risk prediction, involving analysis of ‘risk factors’. An example of this is Re CDW. In that case, the Court took a longitudinal approach, noting that “the absence of any overt symptoms of a mental disorder or violence at the time of [assessment]” would not stop a patient being kept on a CTO.49 It was further stated that the responsible clinician was “entitled” to consider past history, and that “[t] he clinical examination does not occur in a vacuum”.50 The Court found that the patient’s longitudinal history of assault and the threat he posed to others necessitated him being kept under a CTO.51
  1. Considering ‘risk’ as still present even when the patient’s condition is stable
In many preventive cases, it is clear the MHRT considers ‘risk’ as still present, even when a patient’s condition is currently stable. For example, in Re RJ, a patient with a bipolar disorder who was currently well was kept on a CTO on the reasoning that if her noncompliance and consequent relapses continued, there was “a real likelihood ... of the Applicant’s mental health being seriously undermined”.52 This suggests that though the patient’s condition was currently stable and there was no immediate danger of harm, the Tribunal considered there was still a serious ‘risk’, or ‘likelihood’ that she would deteriorate into a condition where she would be a danger to her own mental health. Similarly, in In

  1. In Re PT, above n 13, at [16]–[17], the Tribunal stated that it must consider “the longitudinal history of relapse”, and goes on to consider PT’s responses to medication over time; in In the matter of T, above n 35, at 955, the Tribunal noted that the “longitudinal history of the applicant” must be taken into account; and in L v Director of Mental Health Services [1999] NZFLR 949 (FC) at 955, the Court held that whether a patient poses a serious danger requires the Court to “take a longitudinal view of the available evidence”.
  2. In Re CDW, the Court held that the responsible clinician was “entitled to have regard to any history of mental health issues and the earlier findings which have been made about this by previous clinicians”: Re CDW, above n 38, at [29].
  3. Re RJ, the Tribunal considered that there was a clear longitudinal history of relapse, as though the Applicant had some consecutive years of stability in the past, in recent years he had been admitted to hospital more frequently and was spending shorter periods of time in the community: Re RJ, above n 35, at 9–10.
  4. In Applicant 11/019, the Tribunal considered that the Applicant’s history of cannabis was of concern, as it may “serve to destabilise his mental health”: Applicant 11/019 MHRT 11/019, 11 May 2011 at [56].
  5. Re CDW, above n 38, at [29]. 50 At [29].
  6. See discussion at [26]–[37].
  7. Re RJ, above n 35, at 11.


the matter of T, the MHRT found that there was a high “likelihood of any deterioration” if the patient was taken off the order, suggesting that the risk of deterioration into a state where he may cause ‘serious danger’ to self and others was still present, despite his presently improved condition.53 Furthermore, references to patients having a ‘propensity’ for violence in certain situations suggest that this ‘propensity’ is viewed as continuing even when the patient is stable, in line with the concept of ‘risk’ as an enduring probability.54

6 Compliance, insight and relapse

The final indication that the MHRT is interpreting ‘poses a serious danger’ within a ‘risk paradigm’ is its recurrent reliance on factors such as the likelihood that the patient will comply with treatment, their degree of insight into their condition, and the probability of their future relapse if they are discharged from compulsion and discontinue treatment. Kate Diesfeld describes these as “‘implicit’ factors governing compulsory status”, because they “seem to influence discharge decisions but are not expressly listed in the Act”.55 In the case of Applicant 11/019, the Tribunal had to decide whether to keep a patient with a schizo- affective disorder on a CTO, whose condition had been stable for almost 10 years.56 The Tribunal noted that it had “to consider such issues as the likelihood and consequences of a relapse were the Applicant released from compulsory status”, which necessitated consideration of “whether or not the Applicant has a sufficient understanding of his illness and the means to manage it such that future serious dangerousness may be avoided”.57 It was decided that the patient must be kept under compulsion because of his lack of insight into his illness, his future intentions to cease treatment and contact with mental health services, and the seriousness of his condition were it to relapse.58 The patient’s forensic history of manslaughter and heavy cannabis use were also considered indicators that he needed continuing compulsion.59 In Re Mental Health [Serious Danger], the Tribunal outlined relevant matters to consider regarding relapse, including the expected time lapse between
  1. In the matter of T, above n 35, at 957.
  2. In Re CDW, above n 38, at [32]–[33], the Court considered that the Applicant had a propensity for violence, and it was necessary to decide whether this was linked to his mental disorder; and in In the matter of T, above n 35, at 950, the MHRT stated that “[t]he history of an applicant is only relevant in so far as such proven reported acts of actual physical violence prove an indicator of the applicant’s propensity for future acts”.
  3. Kate Diesfeld “Implicit Factors for Discharge by the Mental Health Review Tribunal” in John Dawson and Kris Gledhill (eds) New Zealand’s Mental Health Act in Practice (Victoria University Press, Wellington, 2013) 114 at 114–115.

56 Applicant 11/019, above n 48, at [5]-[7].

57 At [54].

58 At [54].

59 At [56].



release and cessation of treatment and between noncompliance with treatment and relapse, the ability of clinicians to reinitiate the processes under the Act, and whether early intervention by clinicians or others may lessen the prospect of a relapse.60 Use of these implicit factors seems to be another way the Tribunal is assessing ‘risk’, as it is essentially trying to predict the likelihood of relapse and therefore the risk of the patient becoming unwell and acting violently again.

B Summing Up the Tribunal’s Current Approach

In summary, there are many explicit and implicit indicators that the Tribunal is interpreting ‘serious danger’ within a ‘risk paradigm’, and there is an underlying consistency in this approach. This is probably due to the fact that many of the leading cases I have analysed were decided under the guidance of long-term MHRT convener, Nigel Dunlop, who has had substantial input into the development of the ‘serious danger’ test. However, some variation over time and between cases is still evident. As discussed earlier, the test has been widened over time, and this has been described as reflecting “the apparent increasingly risk-averse nature of society”.61 This suggests that the ‘serious danger’ test has become more ‘risk’ focused, and this seems clear in my analysis, as recently there has been a more explicit focus on risk assessment and a greater emphasis on ‘danger’ as a continuing concept. The shift from the requirement that the ‘danger’ be immediate could also reflect the long-lasting effects of modern treatment such as injections of slow release anti-psychotic drugs. Patients receiving this treatment are unlikely to relapse immediately on ceasing medication, but may still in time relapse seriously with ‘dangerous’ consequences.62 There is less explicit consideration of ‘risk factors’ and risk assessment in the MHRT’s approach than psychiatrists would undertake when assessing mentally ill patients. However, most cases I have analysed were decided before 2012. The most recent assessment of the ‘serious danger’ test, undertaken in the 2012 case Re Mental Health [Serious Danger], shows a much more explicitly ‘risk-based’ approach. This is probably the best example of the approach the Tribunal is taking at present.
From this material, it is possible to predict how the Tribunal might reason about the presence of ‘serious danger’ when assessing whether a currently stable patient should be discharged from a CTO. It is likely to follow the approach described in Re Mental Health [Serious Danger]. It would likely place most importance on a patient’s longitudinal history of relapse and past incidents of dangerous behaviour, as well as their insight into their illness and predicted compliance with treatment. It would consider whether the patient would continue to stay well once off
  1. Re Mental Health [Serious Danger], above n 15, at [60].
  2. Kris Gledhill “Risk and Compulsion” in John Dawson and Kris Gledhill (eds) New Zealand’s Mental Health Act in Practice (Victoria University Press, Wellington, 2013) 62 at 68.

62 At 66.



an order, which is directly related to compliance, insight and therefore risk of relapse. The Tribunal would also consider external factors such as family support, and the likelihood of timely intervention if the patient were to start to relapse. These factors would be balanced against each other and a determination made of whether an order is still justified.
In summary, the Tribunal would make predictions of a patient’s future conduct, based on balancing factors intrinsic and extrinsic to the patient. This clearly demonstrates that the Tribunal is working within a ‘risk paradigm’.

IV Evaluating the Tribunal’s Approach

A Why Does the Tribunal Pay Little Attention to Rights Issues?

The widening of the ‘serious danger’ test since the introduction of the current MHA has undoubtedly resulted in greater encroachment on patients’ rights than the previous, stricter approach. It allows stable patients to be kept on CTOs even when they do not pose an immediate threat of danger, based on an ‘assessment of risk’. Nevertheless, the Tribunal seems to be paying less attention to patients’ rights than it did in earlier cases. Patients’ rights are not explicitly considered in many cases, and when they are, it is usually decided that state interests in preventing the ‘risk’ of a stable patient acting violently in the future outweigh that patient’s right to liberty. Why does the Tribunal pay so little attention to rights issues in its decisions? There are several reasons why this might be so.
Section 6 of the NZBORA requires that an interpretation consistent with the rights contained in it is to be preferred, where it can be given such a meaning. This means that the Tribunal must consider whether alternative interpretations exist that are more rights-consistent. It is possible that the Tribunal does not see this interpretive question as difficult, or controversial, so as to engage s 6 at all. It may view the ‘risk-based’ approach to interpretation, and the restriction of rights that comes with it, as necessary to protect state interests, therefore concluding that other more rights-consistent alternatives are not feasible. Another explanation is that the Tribunal may consider the MHA to clearly authorise the relevant rights to be limited. Section 4 NZBORA states that courts cannot hold another enactment impliedly repealed or invalid just because it is inconsistent with the NZBORA. So the Tribunal cannot hold the MHA invalid simply because it is not very rights-friendly. Thus the MHRT may be simply giving effect to the MHA, so as not to render it in any way ‘ineffective’.63 Alternatively, the Tribunal may not consider that there is a major limit imposed on rights in these circumstances. It may view the encroachment on rights as minimal compared to the alternative of the patient not being under compulsion and potentially causing harm to self or others. Finally, the Tribunal may consider the

63 NZBORA, s 4(a).



limits on rights ‘demonstrably justified’. Section 5 NZBORA states that the rights contained in the NZBORA are subject to justified limitations, so the MHRT may consider that the limits imposed on rights by keeping stable patients on CTOs are justified in the circumstances. The Tribunal may also consider it a ‘proportionate’ limit on rights, in light of the compelling purposes of the MHA to allow compulsion and provide treatment to patients, and in light of the other benefits a CTO may confer on the patient and those around them.
These are all plausible reasons why the Tribunal may not be giving much consideration to rights issues in its interpretation of the ‘serious danger’ test. But is it actually legitimate for the Tribunal to take such a view?

B Legitimacy of the Tribunal’s Rights-Restrictive Approach

Despite the restriction of rights that occurs when a stable patient is kept on a CTO, I submit that it is legitimate for the Tribunal to take such a view, because of the following reasons.
  1. Consistency with the text and purpose of the MHA
The Tribunal’s approach is consistent with the text and purposes of the MHA. Thus it is probably not especially controversial. Keeping stable patients on CTOs in order to prevent their deterioration into a state where they may harm self or others is consistent with the preventive purposes of the MHA, and with the purpose of providing treatment to keep patients in a stable condition. The Ministry of Health, in official guidelines, describes one of the purposes of the MHA as to “ensure that both vulnerable individuals and the public are protected from harm”, suggesting that ensuring the safety of all those involved is paramount.64 Thus a ‘risk-based’ approach to the ‘serious danger’ test in order to keep stable patients well is in line with these purposes.
  1. Mental health professionals are given significant powers under the MHA
Furthermore, I suggest it is fair for the Tribunal to interpret the notion of ‘serious danger’ in a manner consistent with the concepts and practices of mental health professionals, as Parliament has put them at the centre of patients’ assessment under the Act, and the legal criteria must be directly applied to a patient’s situation by these professionals in their functions. Responsible clinicians (usually psychiatrists) must conduct periodic reviews of the patient’s condition,65 and have a continuing duty to consider whether the patient is fit to be released (in other words, whether they still meet the ‘mental disorder’ definition).66 The MHA requires that one of the three people on the Tribunal is a psychiatrist,67 and psychiatrists are called upon to provide expert evidence of whether they

  1. Ministry of Health, above n 37, at 1.
  2. MHA, s 76.
  3. Section 35.
  4. Section 101(2).


consider a patient to be mentally disordered in Tribunal cases. It is clear that Parliament intended to give mental health professionals substantial powers under the MHA, including interpreting the ‘mental disorder’ test. Thus taking a psychiatric ‘risk-based’ approach to interpretation of the ‘serious danger’ test is consistent with this legal framework. Moreover, the MHA does clearly authorise limits to be placed on patient rights, via the powers conferred by CTOs. These powers allow mental health professionals and others to treat the patient, recall the patient if they cannot adequately be treated as an outpatient, and use force to treat the patient in emergency situations. It would be improper to render this regime ‘ineffective’ just because it restricts patients’ rights.

3 History of relapse justifies compulsion

Finally, the Tribunal could reasonably conclude that where the patient does have a history of repeated relapses following cessation of treatment, there are compelling justifications for placing such limits on rights. Dawson contends that as well as ideas about negative liberty, concerns about positive liberty are also relevant: that is, promoting a person’s capacity for self-governance, to set goals and meet these.68 Though keeping a patient under compulsion may encroach on a patient’s rights to autonomy and to refuse treatment (negative liberty), it can also aid in enhancing the patient’s ability to achieve their own ends (positive liberty) by keeping them in a stable condition over a long period of time.
Therefore, there are several compelling reasons why the approach taken by the Tribunal is legitimate. Overall, the approach is not ‘disproportionate’; it correctly balances patient rights and state interests.

C Residual Issues with Preventive CTOs

Though I have concluded that the Tribunal is interpreting ‘serious danger’ within a ‘risk paradigm’, and it is legitimate to interpret it in this way, there are still some dangers of this approach. It may be said that the Tribunal has moved too far away from the ‘harder’ approach to the ‘serious danger’ test used early in the Act’s life. In In the matter of JK, it was held that the ‘serious danger’ test would only be met if there was a “demonstrable risk of no less than serious physical violence”.69 This requirement that the ‘danger’ be imminent acted to protect patient rights, as only those who met this high threshold test could be kept under compulsion. It is arguable that if a harder approach was taken, as in In the matter of JK, patients may get too unwell before intervention can occur, exposing them and the public to the consequences of that. Furthermore, the harder test requiring imminence would not allow a stable patient who is intermittently unwell to meet the ‘mental disorder’ definition. This cannot be correct, as the definition expressly allows patients who have an intermittent abnormal state of mind to come under the definition.
  1. John Dawson “Concepts of Liberty in Mental Health Law” (2009) 12

Otago L R 23.

  1. In the matter of JK, above n 17, at 702.


Thus there are problems with the ‘harder’ approach which are difficult to reconcile. Nevertheless, some of the emphasis on patient rights in the ‘harder’ approach could be brought into the current approach taken by the Tribunal to make it more rights friendly.
The assessment of ‘risk’ is a difficult task, and carries a real danger of error. It involves consideration of dispositional, historical, and contextual factors,70 and should be based on “a thorough collection of information from all available sources and cover all aspects of the illness, background, behaviour and circumstances of the individual”.71 There are many guidelines as to ‘risk factors’ that should be assessed, but it is probably impossible to identify all the ‘risk factors’ present in a certain case. Short term predictions can be reliable, but long term predictions are much more difficult to make and may be unreliable. There is also danger of using actuarial assessments of the general population to assess specific patients, as the same ‘risk factors’ may not be generalisable.72 Ethical issues involved in risk assessment pose another problem. Restricting a patient’s liberty on the basis of prediction, rather than evidence of actual violence, raises a major ethical issue, and there is potential for discrimination when using actuarial factors such as gender or ethnicity, in the risk assessment of the mentally ill. Thus there are many issues faced when undertaking risk assessment which threaten the validity of such predictions and jeopardise patient rights. This emphasises the need for caution when basing decisions that affect people’s liberty on such assessments.
Moreover, the Tribunal has stated on multiple occasions that compulsory treatment under the MHA is not meant to be used as a ‘backstop’ measure in case of relapse,73 and the possibility of a patient coming under the Act again at a later stage is not sufficient reason to keep them under it now.74 Though a patient may have a lifelong mental illness, the Tribunal has stated that this does not necessarily mean that they will be kept under compulsion indefinitely, as the nexus between the ‘abnormal state of mind’ and ‘dangerousness’ may not always be






  1. Paul E Mullen “Assessing Risk of Interpersonal Violence in the Mentally Ill” (1997) 3 APT 166 at 166.
  2. Ministry of Health, above n 31, at 4.
  3. Mullen, above n 69, at 166.
  4. In Re MJB, it was stated that “orders should not be regarded as back up measures in case something goes wrong”: Re MJB MHRT Hamilton 06/090, 28 September 2006 at [25].
  5. Diesfeld, above n 55, at 121–122.


strong enough to meet the ‘mental disorder’ definition.75 It should be remembered that the test for compulsion is whether the patient is presently mentally disordered. Though the MHA probably allows some preventive use of CTOs, this should be confined to patients who pose a ‘serious’ risk of danger. Otherwise, there is a risk that CTOs will be used preventively more and more often, on patients who do not actually need to be compelled.
The element of ‘seriousness’ could be given greater emphasis, especially if the impact on patients’ rights is to be proportionate. There are several ways in which the dangers described in the last section may be minimised. Firstly, this could be achieved through a greater emphasis on the ‘serious’ component of ‘serious danger’. Parliament clearly intended that the danger of the patient causing violence in the future must be significant. This serves the purpose of protecting patient rights by only compelling those who meet the high threshold of the ‘mental disorder’ test. Secondly, greater effort should be made to undertake a proper balancing of ‘risk factors’ in all cases, as was done in Re Mental Health [Serious Danger]. I suggest that the Tribunal should make sure it is listing all the ‘risk factors’ for a patient that indicate ‘serious danger’, and all of those which do not, and balance these against each other to make sure patients are being fairly assessed. Thirdly, a more stringent risk assessment should be undertaken. This requires the assessment of information from an “adequate database”,and should be based on both factual information and informed opinion.76 The Ministry of Health stated that “reliance upon actuarial factors alone is unwise”, suggesting that a combination of actuarial and clinical methods should be used. 77 Finally, the Tribunal should conduct a more explicit balancing of the arguments for a person’s compulsion against the impact on their rights, to ensure that the impact is proportionate in the circumstances.

V Conclusion

The preventive use of CTOs on stable patients is a major concern for human rights advocates in New Zealand. This article has considered whether the MHA permits CTOs to be used preventively in order to keep stable patients well, and if so, whether this constitutes the correct balance between state interests and patients’ rights. I submitted that the concept of
  1. In Re TAK, the Tribunal stated that as the “mental disorder definition ... is a legal construct”, a patient with a lifelong illness may not always meet this test, as time and changing circumstances may mean they do not always pose a “serious danger”. It then went on to state that the Applicant was “moving toward the time when a Tribunal might be prepared to determine that he is no longer mentally disordered because the nexus between his abnormal state of mind and dangerousness is becoming progressively less clear as a result of a long period of stability”: Re TAK MHRT Wellington 03/047, 29 May 2003 at 6.
  2. Ministry of Health, above n 31, at 6.
  3. At 3.


‘risk’ as an enduring probability is convincing and a legitimate substitute for the notion of ‘danger’ in the ‘mental disorder’ definition, in light of the purposes of the MHA. In this fashion, the ‘serious danger’ test can be understood within a ‘paradigm of risk’, where a patient is assessed against a number of ‘risk factors’ in order to determine the likelihood of their acting violently in future. Analysis of the Tribunal’s decisions suggests that it is taking this ‘risk-based approach’, and, though this is not very rights-friendly, it may be legitimate to take this view. This is because the approach is consistent with the text and purpose of the MHA, mental health professionals are given significant powers under the MHA, the Act clearly authorises limits to be placed on patient rights, and the Tribunal could reasonably conclude that a patient’s history of relapses justifies compulsion. Despite the fact that it may be legitimate, there are still dangers in it, such as the lack of emphasis on rights compared to the original ‘hard’ approach to assessing ‘serious danger’, the dangers of error involved in risk assessment, and the use of the MHA as a ‘backstop’. There are ways to make the ‘serious danger’ test more rights-friendly, such as placing greater emphasis on the word ‘serious’, the Tribunal making sure it properly balances ‘risk factors’, undertaking more stringent risk assessment, and determining more explicitly whether a proportionate limit is being placed on rights.
This article also raises broader issues about the interpretation of mental health law. As the MHA involves the use of public powers, which directly affect patients’ rights, there is always a need to balance rights and state powers. Consequently, interpretation requires the balancing of rights-driven and purposive approaches. It seems that, in the context of preventive CTOs, the purposive approach often outweighs human rights considerations. The trend toward the use of risk assessment in determining whether a patient should be compelled also points to wider societal trends towards greater use of preventive laws and the language of ‘risk’. The use of ‘risk’ in conjunction with preventive laws has been described as a way for the government to manage fear of certain groups in society.78 Are we heading toward a time when precaution against the unknown is overly-favoured, and the consequent restrictions on human rights considered unavoidable? This is concerning in the context of mental health law, where those affected are vulnerable and reliant on others’ help. Although there are arguments that justify the Tribunal’s current approach, it should be remembered that there are also legitimate concerns associated with the trend towards prioritising prevention over individual rights when applying the coercive power of the state.
  1. Bernadette McSherry Managing Fear: The Law and Ethics of Preventive Detention and Risk Assessment (Taylor and Francis, New York, 2014) at ch 12.


NZLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.nzlii.org/nz/journals/OtaLawRw/2018/6.html