Risk management in the era of recovery and rights

Summary

Clinical decision-making in psychiatry is affected by many factors, including how best to reduce risks of harm while promoting autonomy and personal recovery. This article proposes guidance for clinical decision-making that is consistent with civil liability law. It emphasises collaboration, clarification of the available information and communication of decisions as a basis for recovery-oriented risk management.

LEARNING OBJECTIVES

After reading this article, you will be able to:

• • understand the influence of recovery and rights on clinical practice

• • identify relevant factors for clinical decision-making

• • use a three-step process for justifiable decision-making.

Mental health policy is increasingly influenced by the concept of ‘recovery’ (Skuse 2012), which refers to ‘a set of ideas and principles derived from the experiences of people with mental health problems’ (Boardman 2012). The underlying themes of the recovery approach encompass (Leamy 2011):

• • connectedness

• • hope

• • identity

• • meaning

• • empowerment.

In tandem with the recovery approach, an emphasis on human rights is shaping mental health law reform (McSherry 2010). The first guiding principle of the United Nations Convention on the Rights of Persons with Disabilities (CRPD) is, ‘[r]espect for inherent dignity, individual autonomy including the freedom to make one's own choices’ (Article 3). The CRPD requires states that have ratified it (which include the UK, Australia and Canada) to ensure that those with disabilities have the right to ‘enjoy legal capacity on an equal basis with others in all aspects of life’ (Article 12). Legal capacity encompasses making decisions about all aspects of life, including medical care and, if needed, having supports put in place to enable such decisions (McSherry 2012; Gooding 2013).

Human rights complement the focus on empowerment as a recovery theme: ‘people in recovery begin to demand the same rights (e.g., the right to decide where to live, whom to love, how to spend one's time) and take on the same responsibilities (e.g., paying taxes, voting, volunteering) as other citizens’ (Davidson 2005).

This article explores the implications of the focus on recovery and human rights for how mental health clinicians carry out their day-to-day work. It focuses on how clinicians might resolve dilemmas involving risk, while still meeting the requisite ‘standard of care’.

The new balance of power: the drive for patient autonomy and empowerment

The corollary of advocating for greater power to be afforded to people with mental illness is that they must also be afforded greater responsibility for their own lives. The recovery approach and human rights are potent drivers for patient choice and autonomy rather than subservience to the control of clinicians and services. Importantly, what is claimed is not simply empowerment to be able to make choices about simple and uncontentious matters, but also the autonomy to make ‘risky and potentially self-defeating choices’ (Parsons 2008). This has influenced legislation. For example, one of the ‘mental health principles’ outlined in the Mental Health Act 2014 of the Australian state of Victoria states that: ‘persons receiving mental health services should be allowed to make decisions about their assessment, treatment and recovery that involve a degree of risk’ (italics added).

There is potent therapeutic logic in this. The paternalistic default assumption of decision-making power by services whenever choices become complex or involve significant risk will impede patients’ personal growth (Parsons 2008).

Determining the circumstances under which patient autonomy may, or even must, be overridden raises complex challenges for clinicians. Recovery advocates (Deegan 1996) flag the need to ‘be careful to distinguish between a person making (from our perspective) a dumb or self-defeating choice, and the person who is truly at risk’ (italics added) in order to embrace the ‘dignity of risk’ and the ‘right to failure’. This begs the question not only of how clinicians ought to determine whether a patient is ‘truly at risk’ – the subject of a vast and controversial literature (Fazel 2012; McSherry 2020) – but also glosses over the complex array of values at play for clinicians in determining whether to intervene to override patients’ choices.

While recovery- and rights-oriented policy has come to influence practice, there has also been growing emphasis on the importance of effective risk management in mental healthcare (Department of Health 2009; Holmes 2013; McSherry 2014). These twin imperatives – patient empowerment allied with greater accountability of professionals – reflect broad contemporary societal trends.

The emphasis on risk, however, can render clinicians reluctant to cede power to patients to whom they owe a legal duty of care. Clinicians may be forgiven for thinking ‘If recovery is the person's responsibility, then how come I get the blame when things go wrong?’ (Amering 2009), meaning that caution may thus trump patient autonomy and choice. Practitioners may be risk averse because if patient empowerment results in negative outcomes, as inevitably from time to time it will, they need to be able to justify their actions (Meehan 2008). A fear of legal ramifications in the event of failure is a barrier to practitioners allowing patients ‘the dignity of risk’ (Parsons 2008). Failing to address the challenge leaves practitioners more concerned about what might be termed the ‘risk of dignity’.

In brief, ‘secondary’ risk management (Undrill 2007) – clinician anxiety regarding loss of professional reputation or being subject to civil litigation – may act as a (largely covert) driver for clinicians asserting excessive control. The logical basis for such anxiety is questionable in that it is predicated on an unhelpful conflation of ‘risk-averse practice’ and ‘best practice’. As will be explored below, the law itself endorses clinically justifiable risk-taking.

The challenge for contemporary mental health practitioners is to incorporate the recovery and rights approach, which is appropriately endorsed at a policy level, while discharging their ‘duty of care’ to patients by delivering care that meets acceptable standards. We will consider how recent medico-legal discourse may provide guidance in meeting the challenge to both respect patient autonomy and avoid neglectful care.

Recovery-focused risk management: achieving the requisite standard of care

Although laws of negligence differ between legal jurisdictions, in general, negligent conduct occurs when conduct falls below an acceptable ‘standard of care’. There must exist some form of proximate relationship between the claimant and the defendant, a duty of care established and the damage suffered must be reasonably foreseeable (Caparo Industries plc v Dickman [1990]).

The ‘standard of care’ applicable to the relationship between mental health practitioner and patient is determined with reference to ‘reasonable’ professional standards, by relevant practitioners, although the courts are able to step in if standards are considered too low (Bolitho v City and Hackney Health Authority [1998]; Dobler v Halverson [2007]; Brakoulias v Karunaharan [2012]). It is thus generally recognised that mental health practitioners have a duty to exercise reasonable care and skill in treating those with mental health problems. Although there may be slight differences in wording in different legal jurisdictions, what is considered to be reasonable care or skill is measured against professional practice and must be ‘logically defensible’ (Sappideen 2010).

In Australia, following a national inquiry into tort law reform (Commonwealth of Australia 2002), guiding principles were inserted into civil liability legislation regarding when a person is not negligent in omitting to take precautions against a risk of harm. For example, section 48 of Victoria's Wrongs Act 1958 now states:

1. ‘(1) A person is not negligent in failing to take precautions against a risk of harm unless –

   1. (a) The risk was foreseeable (that is, it is a risk of which the person knew or ought to have known); and

   2. (b) The risk was not insignificant; and

   3. (c) In the circumstance, a reasonable person in the person's position would have taken those precautions’ (italics added).

The section then adds:

1. ‘(2) In determining whether a reasonable person would have taken precautions against a risk of harm, the court is to consider the following (amongst other relevant things) –

   1. (a) The probability that the harm would occur if care were not taken

   2. (b) The likely seriousness of the harm

   3. (c) The burden of taking precautions to avoid the risk of harm

   4. (d) The social utility of the activity that creates the risk of harm’ (italics added).

Although this legislative section is concerned with negligence as it relates to members of the general public rather than those with special skills, these principles can nevertheless provide a useful framework for determining when care provided by mental health practitioners may or may not be considered to be negligent. Different jurisdictions in the English-speaking world may use slightly different terminology, but the utility of this framework is essentially universal: the core tasks when managing risk are similar across different fields of professional endeavour and across different jurisdictions.

Each of the relevant constructs will be discussed below and a model proposed as to how mental health clinicians might apply them in practice. To articulate the pragmatics of this model, four fictitious case examples will be considered (Box 1).

BOX 1 Four fictitious case examples

Ahmed is a 28-year-old man with a 10-year history of schizophrenia who, at a scheduled appointment in the clinic with his case manager, reports intermittent command hallucinations encouraging him to kill himself. He has known the case manager for 2 years but only in the past 6 months has he begun to be more open and honest about reporting symptoms. He has had three hospital admissions of several weeks’ duration each and is fiercely opposed to readmission. He is subject to an order under mental health legislation mandating treatment. The mental health service includes an intensive home treatment team that can provide visits to the home, including (in theory) supervision of medication up to twice daily.

Dilemma: whether or not to admit Ahmed to hospital on an involuntary basis.

Barbara is a 40-year-old woman with bipolar affective disorder who has been admitted on a voluntary basis to a private hospital after an attempted hanging during a depressive episode. She is in her second week of admission and showing some signs of recovery, including making plans for the future. However, she still has some transient periods of despair and suicidal thinking. She is keen to have day leave with her daughter, who is visiting from another city tomorrow.

Dilemma: whether or not to endorse day leave despite ongoing suicide risk.

Carlos is a 45-year-old man who was admitted to hospital 24 h ago with a relapse of schizophrenia with intermittent hallucinations and associated persecutory delusions. He lives 200 km away but is on holiday and went to the local emergency department with his friend. He is now keen to return home to his mother and she is willing to have him home with the support of input from his local mental health service. His friend is willing to drive him the 4 h home, provided that ‘the doctors’ feel that this is safe. Ten years ago when actively psychotic Carlos attempted suicide by lying on train tracks.

Dilemma: whether to continue to detain Carlos in hospital or allow him to go home to his mother.

Dave is a 25-year-old prison inmate on remand for drug trafficking. He was placed in a ‘management cell’ 2 days after incarceration, following an episode when he punched a wall and at review by the psychiatric nurse disclosed that he wanted to hang himself, since his long-term girlfriend had ended their relationship. He now appears despondent but is denying plans to follow through on these urges and is asking to return to a mainstream unit.

Dilemma: whether or not to recommend Dave's release into a mainstream unit, where access to means of self-harm cannot be entirely removed.

Foreseeability, likely seriousness and probability of risks

Notwithstanding the significant progress made in the field of risk assessment in mental health, it remains impossible to predict with confidence harmful outcomes at the level of the individual patient (Hart 2007). However, by drawing on the relevant evidence base and applying it to a given clinical situation, clinicians can foresee what kinds of situation involve a significant likelihood of adverse outcomes such as self-harm (Fagan 2009) and violence (Webster 2007). The concept of ‘foreseeability’ does not imply that the clinician can be expected to know with a high level of confidence that a particular outcome will occur in a specific case. Rather, it can be taken to mean that a clinician is expected to recognise the kind of scenario in which a harmful outcome of some kind might unfold, with a probability high enough to warrant consideration of precautionary measures to mitigate risk. The clinician should systematically consider ‘probability’ and ‘likely seriousness’ of future harms when making risk management decisions. This requires both a working knowledge of the relevant evidence-based risk/protective factors and an awareness of the case-specific data (such as the patient's clinical history and mental state).

In the four case examples outlined (Box 1), there are dynamic risk factors that raise the probability of harmful outcomes to an appreciable level, such that the clinician can be expected to have at least considered the possibility of an adverse outcome. The clinician should, at the very least, be aware of certain key details relating to the individual's history, current mental state and current (and anticipated) psychosocial context. Such key details would include the evidence-based risk factors for harm to self or harm to others in each case (Box 2).

BOX 2 Critical evidence-based information

Ahmed

• • History of violence

• • Recent use of substances and availability of illicit substances at home

• • Phenomenological details of command hallucinations, including a sense of how powerful he perceives them to be

• • Support or lack thereof from the family at home.

Barbara

• • History of suicide attempts

• • Family history of suicide

• • Reliability of daughter

• • Mental state just prior to episode of leave.

Carlos

• • Mental state over 24 h period since admission

• • History of adherence to prescribed medications and to other aspects of care

• • History of risk behaviours when psychotic

• • Reliability of mother

• • Likely delay to next psychiatric assessment.

Dave

• • Past psychiatric treatment

• • Past episodes of self-harm

• • Recent substance use

• • Level of social support

• • Current mental state

• • Recent behaviours in the cell

• • Level of support and monitoring available if released from management cell.

Although the use of particular risk assessment tools is not essential to achieving the expected standard of care, structured or semi-structured approaches using checklists and other tools may aid the rapid and efficient consideration of relevant risk and protective factors (de Vogel 2009). However, if such tools are utilised in a clumsy or bureaucratic way, rather than in a patient-centred fashion, they may paradoxically impede risk management by disrupting relational aspects of care (Holmes 2013; Carroll 2014).

The burden of taking precautions to avoid risk

For most situations, the detection of risk is generally the easy part of risk management. The clinical challenge is what to do about apprehended risk. The law recognises that clinicians often face ‘no-win situations’, wherein mitigating risk inevitably entails certain ‘burdens’. Commonly, such burdens include the creation of future problems due to overly cautious decisions being taken in the here and now (Box 3). Such future problems can include a diminution of therapeutic trust and slowed clinical recovery when patient autonomy is overridden. Consequently, longer-term risks can be aggravated as a result of measures taken to mitigate short-term risk.

BOX 3 Possible adverse consequence of taking the preventive option

Ahmed

Too readily admitting him to hospital against his will, although managing the short-term risk over the next week or two while he is actively hallucinating, could have the effect that next time he has breakthrough command hallucinations he will be less likely to disclose these to his case manager. The most dangerous symptom, of course, is one that is unknown to the treating clinician: a strong trusting long-term therapeutic alliance is the most effective risk management tool that there is in mental healthcare.

Barbara

Not only might overriding her desire for the family leave lead her to sign herself out of hospital, but even if she agrees to stay on a voluntary basis, a refusal to give her leave may have the effect of setting back her recovery from the depressive episode by inculcating feelings of frustration and despair. Again, short-term risk reduction may come at the cost of medium- to long-term risk exacerbation.

Carlos

A denial of his desire to return to his mother may both prolong his current relapse and also lead him to be less inclined to seek psychiatric care in the event of future episodes of psychosis.

Dave

Ongoing detention in a bare cell not only works against improvement in his mood state, but also discourages him (and co-prisoners) from disclosing suicidal ideation in the future.

Social utility of the activity that creates the risk of harm

Risk management dilemmas in mental health generally involve a choice between two alternatives:

• • adopting a preventive focus (Halvorson 2014) that privileges the reduction of short-term risk of harm; and

• • adopting a promotional focus (Halvorson 2014) that privileges longer-term recovery, even if the risk of harm in the short term is increased.

The Wrongs Act concept of ‘social utility’ may be particularly relevant when a promotional focus is adopted. In such instances, ideally the clinician considers the empirical evidence base regarding the ‘social utility’ of the promotional, as compared with the preventive, choice (Box 4). Logically, the social utility of a particular activity is judged through a social lens. Deinstitutionalisation and changes in how the general public perceives mental illness have influenced how the courts judge the risk management endeavours of mental health services and clinicians (see, for example, Hunter and New England Local Health District v McKenna [2014] discussed below). Thus, the transition of the core tenets of the recovery and rights approach from patient and mental health professional discourse into public policy documents, guidelines from professional bodies and even legislation, is of great importance (Wolfson 2009). This transition not only allows, but at times may require, clinicians to advocate for potentially risky choices that have significant social utility.

BOX 4 Social utility evidence to consider

Ahmed

• • What does the evidence say about the effectiveness of home treatment teams?

• • What is the evidence regarding suicide risk reduction and psychiatric in-patient admission?

Barbara

• • What is the evidence base about the role of family links in recovery from depression?

• • What is the evidence base about possible timing of suicide attempts in people in early stages of recovery from depression?

Carlos

• • Is the evidence regarding ‘expressed emotion’ and families relevant to his situation at home with his mother?

• • What does the evidence say about the likely time frame for recovery of clinical stability in somebody who presents as actively psychotic, as he was only 24 h previously?

Dave

• • What is the evidence base about the impact on well-being and morale of detention in bare isolation cells?

• • What is the evidence relating to periods of high suicide risk in recently incarcerated prisoners in withdrawal from substances and in receipt of bad news regarding intimate relationships?

‘Other relevant things’

This rather general term suggests that the courts will reserve the right to consider whatever issues it believes to be relevant in any given case. In practice, the challenges of mental health risk management commonly involve two particular issues that clinicians may consider and document when making decisions in the context of risk.

The first issue is ‘operational constraints’. These are pragmatic, real-world limiting factors that affect the feasibility of management of risk (Box 5). Risk management ideals predicated on professional knowledge must be applied in the real world of busy, imperfectly resourced service systems. Clinicians generally need to settle for ‘best feasible practice’ rather than ‘best practice’. It is good procedure to openly discuss, and even document, the reasons for the gap between the two in any specific risk management scenario.

BOX 5 Possible operational factors to consider

Ahmed

• • Does an intensive home-based treatment team exist for his area?

• • Is it currently able to cope with the need for twice-daily visits?

• • Is there an in-patient bed available at the local hospital or might admission involve a 300 km ambulance ride to a different service that does not know him?

Barbara

• • Is her treating psychiatrist, or a trusted psychiatric nurse, available to reassess her mental state in a detailed fashion prior to granting the leave?

• • Can leave with her daughter be postponed or is her daughter only visiting for this one day?

Carlos

• • Is the mental health service local to his mother's home able to visit and reassess as soon as he arrives or does he face a delay of several days before being seen?

Dave

• • If released from the management cell, are adequate observations in place?

• • Is his next placement free of means of suicide such as ligature points?

• • Can he be reviewed psychiatrically within the next 24 h or does high demand mean that the next review is several days away?

Legal constraints will also need to be considered. For example, involuntary hospital admission and treatment is obviously only lawful when certain criteria are met. Clinicians’ powers are very significantly constrained: a risk that can be foreseen, even of a serious nature, does not necessarily allow a clinician to admit a patient to prevent harm, unless other criteria are also fulfilled.

Similarly, powers to intervene in the context of risks in psychosocial domains, such as financial imprudence or a risk of inadvertent fire setting due to hoarding, are legally constrained. Legislation exists to allow for the formal appointment of an independent person to make choices on an individual's behalf, but only when certain criteria relating to impaired decision-making capacity are met.

The reasonable clinician

The Wrongs Act 1958 (Vic) criteria include the notion of the ‘reasonable person’. The ‘Bolam test’ of reasonableness (Bolam v Friern Hospital Management Committee [1957]) was predicated on endorsement by professional peers (presumably influenced in turn by both emerging evidence and societal expectations). In that landmark case, Justice McNair, in his direction to the jury, said: ‘[a doctor] is not guilty of negligence if he has acted in accordance with the practice accepted as proper by a responsible body of medical men [sic] skilled in that particular art’.

Subsequent case law (e.g. Bolitho v City and Hackney Health Authority [1998]) reflects more sceptical contemporary attitudes towards practitioners. The requirement to meet the requisite standard of care can no longer be assumed to have been met merely by adducing evidence that a group of peers endorse it as ‘proper’. Rather, as outlined in Bolitho by Lord Browne-Wilkinson:

‘The court has to be satisfied that the exponents of the body of opinion relied upon can demonstrate that such opinion has a logical basis. In particular, in cases involving, as they so often do, the weighing of risks against benefits, the judge before accepting a body of opinion as being responsible, reasonable or respectable, will need to be satisfied that, in forming their views, the experts have directed their minds to the question of comparative risks and benefits and have reached a defensible conclusion on the matter’ (italics added).

The reasonable clinician must therefore undertake a process of reasoning, weighing up risks and benefits to reach a logical and therefore defensible decision. Working through such dilemmas is at times a complex and challenging task. However, the ability and responsibility to deal with complex matters is at the heart of all professional work.

For serious risk management dilemmas, the formal documentation of clinical reasoning using a ‘risk–benefit table’ (Table 1) can help with processing the various considerations relevant to choosing between the two (or more) alternative courses of action. The result is a medico-legally robust record of how the dilemma was managed. The process of generating such a table, which can of course be conducted within multidisciplinary teams and (in some situations) with input from the patient, may also help to create possible ‘compromise’ choices.

TABLE 1 Risks–benefits table in the case of Ahmed, with some of the possible considerations

Promotional versus preventive decisions

As discussed above, risk management dilemmas invariably involve deciding between, on the one hand, an option with a ‘promotional’ focus (such as home-based care) and, on the other hand, one with a ‘preventive’ focus (such as compulsory hospital admission). The considerations involved in such decisional alternatives show certain commonalities, irrespective of the precise nature of the dilemma, as set out in Table 2.

TABLE 2 The promotion/prevention dichotomy

Preventive decisions generally involve a relatively coercive, often resource-intensive intervention to mitigate the likelihood of a short-term adverse outcome. The justification for such decisions hinges on an assessment that the imminence, probability and seriousness of the anticipated harm are sufficiently high as to mandate the taking of the precautions involved. Such precautions involve reducing acute (potentially short-lived) dynamic (changeable) risk factors and/or increasing protective factors.

The justification for such approaches may generally be the preservation of life (that of the patient and/or others) in the short term. The harms that may stem from the omission to take such precautions are concrete, manifest in the short term and likely to be subject to intense scrutiny.

The harms stemming from the commission of such preventive approaches are less readily apparent but need to be considered if a balanced approach to risk management is to be achieved. Such harms tend to be abstract and difficult to demonstrate in the specific case, but may include diminished self-efficacy and future reluctance to collaborate with clinicians and services. Resultant harms are more likely to be manifest in the medium to longer term, rather than immediately.

Conversely, promotional decisions generally involve supporting the patient's choice even in the face of the possibility of a short-term adverse outcome. Usually, the justification for such choices is that they will increase protective factors and/or reduce dynamic (changeable) risk factors – especially those that are expected to only slowly change for the better.

Risk management in practice: the three Cs guideline for risk dilemmas

The recovery and rights approach means that clinicians will tend to adopt a ‘default’ position of selecting choices with a promotional focus. However, clinicians must never of course neglect their professional responsibility to adopt preventive strategies where required. This section details a three-step process for clinicians faced with the need to negotiate the often narrow path between dereliction of duty and negation of patients’ rights: collaborate, clarify and communicate (Carroll 2018).

Collaborate

Fundamentally, recovery-focused risk management involves collaboration between clinicians and patients. Evidence increasingly supports recovery-based collaborative approaches to working with risk (Parsons 2008). This is unsurprising, since, over the long term, ‘danger can only be reduced within a trusting relationship within which each understands where the other is coming from, feels able to express their fears and concerns, and can share responsibility for safety’ (Perkins 2016).

Collaboration with family members, carers and/or other key nominated support people may add to the complexity of the task but is now mandatory in some jurisdictions.

There are inevitable constraints on collaboration. Some patients will decline or (at least temporarily) be unable to collaborate because of decision-making impairments. The clinical challenge is to see such situations as a starting point rather than necessarily a reason to default to coercive, paternalistic approaches. Tools such as advance directives, prepared by patients to address future preferences in the event of crises, can facilitate collaborative ways of working with risk (Weller 2015).

For complex risk scenarios in mental health, collaboration between disciplines will also generally be involved. A common error is to assume that the preferred values of these disciplines will be in perfect alignment. Effective risk management requires an acknowledgement that this may not be so. Optimal risk management is more likely when the contributions of all discipline groups, and all levels of seniority, are allowed a hearing (Prins 1999) in an atmosphere of mutual respect and consideration (Fulford 2012).

Clarify

Ideally, a clinician will clarify which information is relevant and use it in a way that supports the patient and minimises risk. The most important source of information is the subjective experience of the patient: their experience of symptoms, their experiences of mental health services and their experience of coping with life in a society that discriminates against those with mental health problems. On occasion, the patient will convey such information directly as a coherent narrative, readily translatable into clinical and risk management concepts (Starr 2002). More often, however, it is conveyed as a more inchoate, emotionally charged account. The task of the ‘reasonable’ clinician is to collaborate with the patient to obtain a rich, detailed subjective account and to utilise the patient's expertise and lived experience to the maximal degree possible.

Information derived from the patient must then be incorporated with other pertinent data, such as collateral information and the relevant empirical evidence, into a coherent and clinically robust formulation.

Communicate

Recovery-focused and legally defensible care requires the communication not only of clinical decisions themselves but also of their logical basis. Such communication (both verbal and written) will be directed to the patient and the treatment team, but may also involve carers or other support persons.

Ideally, communication will include careful documentation of risk management decisions – ‘thinking for the record’ (Gutheil 1980). This does not require overly verbose file entries, but does involve clinicians asking themselves ‘If the worst should happen, how would I justify this decision to my peers/managers/the coroner/carers/the general public?’ Such a ‘pre-mortem’ should not become a futile exercise in rumination and self-doubt, but rather provide a solid basis for proactive, patient-centred risk management.

Members of the public and the media may not appreciate the benefits of recovery-oriented promotional decisions as readily those of preventive approaches. It is therefore especially important that the basis for such decisions is proactively articulated and documented at the time of decision-making. This helps to offset the possibility (if the outcome is adverse) of the decision appearing to have been naive and reckless.

A risk–benefit table (as set out in Table 1) can assist with clarifying the evidence and values at stake, provide a tool for collaboration with a patient and/or a multidisciplinary team and help communicate the rationale for the decision that is ultimately made. Assuming that the considerations are consistent with the relevant evidence base and with the known facts of the situation, and that the subsequent management decisions logically flow from this, such a table provides powerful evidence that the requisite standard of care has been met.

Positive therapeutic risk-taking and its limits

One of the more challenging ideas found within the recovery literature is that of ‘therapeutic’ or ‘positive’ risk-taking. Policy guidelines in the UK (Department of Health 2009) have advocated that ‘positive risk management as part of a carefully constructed plan is a required competence for all mental health practitioners’ (p. 6) and defined the concept thus: ‘being aware that risk can never be completely eliminated [and that] management plans inevitably have to include decisions that carry some risk. This should be explicit in the decision-making process and should be discussed openly with the service user’ (p. 11).

A more radical interpretation of the concept of positive risk-taking is that such risk-taking is not simply seen as an inevitable (albeit non-ideal) concomitant of recovery. Rather, it is by being allowed to deal with risk that true personal recovery is achieved (Parsons 2008). This notion finds some support in the literature on coping and resilience: ‘resistance to hazards may derive from controlled exposure to risk (rather than its avoidance)’ (Rutter 2006). The underlying principle that risk-taking and recovery are intrinsically and inevitably linked, since ‘every opportunity for growth carries with it the potential for failure’ (Parsons 2008), certainly reflects the anxieties of thoughtful mental health practitioners.

The counterintuitive nature of the concept, however, must be acknowledged. Clinicians must take particular care to explain and communicate the rationale for therapeutic risk-taking. Also, approaches that focus exclusively on promoting ‘strengths’ (Rapp 2011) may neglect the role of prevention-focused ‘treatment goals’, which are an essential part of comprehensive mental healthcare (Slade 2009).

Supporting positive risk-taking may be viewed as the antithesis of overly defensive clinical practice. Rangarajan & McSherry (2009: p. 298) have pointed out that the quality of healthcare is likely to be diminished as a result of defensive decision-making that is based on the fear of litigation and liability rather than on clinical grounds. This fear is common among clinicians, despite the fact that case law does not support defensive practice. For example, the Australian High Court in Hunter and New England Local Health District v McKenna [2014] upheld an appeal against a (majority) decision of the New South Wales Court of Appeal concerning the alleged negligence of a New South Wales hospital and a psychiatrist in its employ in the discharge of a patient with mental illness who subsequently killed a man. The High Court found that the hospital and psychiatrist held no duty of care to the relatives of the man who had been killed. The Court found that the New South Wales legislative provisions which enable compulsory detention require ‘the minimum interference with the liberty of a mentally ill person’ (at para. 31). This finding reflected the dissenting judgment of Justice Garling in the Court of Appeal decision of McKenna v Hunter and New England Local Health District (2013), in which he pointed out (at para. 258) that the ‘burden on the Hospital in the event of such a duty [to detain] being imposed would be “intolerable”’.

Defensive practice also has resource implications. It is likely to result in prolonged periods of involuntary detention for an increased number of patients, thereby shifting available resources to keep them detained while a growing number of potential patients are left without access to an appropriate level of care (Freckelton 2003).

Facilitating and encouraging patients to negotiate challenges and manage risks for themselves (drawing on assistance from services where they deem fit) is now recognised as ‘best practice’. However, it is in the nature of severe mental illness that there will be times when the patient's abilities to make specific choices will be impaired by the effects of that illness. For example, it would clearly be a dereliction of duty for a clinician to fail to override choices made by a person who is psychotically depressed and influenced by suicidogenic command hallucinations. Although therapeutic or positive risk-taking emphasises the therapeutic value of handing over responsibility for risk management to the patient, this can never be an excuse for the abrogation of clinical responsibility.

Conclusion: balanced recovery-focused risk management

An effect of the shifting culture towards recovery-oriented values, as well as the CRPD's international human rights obligation to uphold the right to legal capacity of those with disabilities on an equal basis with others, is likely to be a far more questioning and rigorous approach to the removal from patients of responsibility for their own choices.

The explicit recognition in formal policy documents that risk can never be completely eliminated, together with case law limiting the legal concept of a duty of care, provides clinicians with important support for adopting interventions with a promotional focus that (inevitably) involve some degree of risk. Although the recovery approach and human rights are now providing the context for patient autonomy and empowerment, there is undoubtedly still a need for the education of the general public and media in terms of the social utility of activities that lead to positive, therapeutic risk-taking in an age where the prevention of risk is generally seen as paramount.

In this article, we have indicated how mental health clinicians working with risk, while adopting a recovery and rights approach, can usefully apply a framework derived from Australian tort law which emphasises: notions of foreseeability, likely seriousness and probability of risks; the burden of taking precautions to avoid the risk; the social utility of the activity that creates the risk of harm; ‘other relevant things’; and the notion of the ‘reasonable’ clinician.

Clarification of the available information, collaboration and communication of decisions are essential to risk management in practice. However, although risk may be mitigated, it is inevitable that adverse outcomes, even tragedies will occur. It is therefore especially important that promotional decisions that support the patient's choice are clearly documented at the time of making the decision.

Notwithstanding the default preference for patient-centred ‘promotional’ care, ‘preventive’ decisions are not inherently undesirable. The key lies in discerning when such approaches are indicated. There will inevitably be times when a preventive approach is appropriate, but civil liability law, mental health law and public policy frameworks now support an approach that encourages the ownership of power and responsibility by mental health patients rather than the default assumption of control by clinicians.