The PACT advance decision-making template: preparing for Mental Health Act reforms with co-production, focus groups and consultation

Background Advance decision making (ADM) in mental health is supported by stakeholders but faces significant barriers. These must be overcome, not least to support the UK government's commitment to introduce statutory mental health ADM in England and Wales. Aims To build understanding and address the gap between aspirations for ADM and actuality, with feasible co-produced ADM resources. Methods We used focus groups and consultation to explore experience and views of stakeholders on ADM processes and materials. Discussions included feedback on an ADM template which was adapted accordingly throughout the research process. Results Between September 2017 and December 2019, 94 individuals, representing stakeholders advised on design and process of ADM, alongside wider discussion at stakeholder events. Collaborative ADM was universally supported. Valued outcomes were diverse and combining aspirations with practicality required resolving dilemmas. A prototype template and guidance, the PACT (Preferences and Advance decisions for Crisis and Treatment) was co-produced, designed to help manage fluctuating mental capacity through collaborative decision making. The PACT enables direct engagement with medico-legal frameworks, with provision to facilitate person-centred assessments, treatment refusals and requests. Resources including supported engagement and cross-agency awareness and accessibility were seen as essential. Conclusion Our research confirms high stakeholder motivation to engage in ADM is hampered by multiple barriers. We identified enabling conditions for ADM and co-produced an ADM template and guidance which supports achievement of a range of valued outcomes. Further developments to support and evaluate the process of implementation are now needed to prepare for statutory change.


Introduction
Advance decision making (ADM) for mental health crises is widely approved, but under-used and under-resourced, particularly when compared to ADM in physical healthcare settings. Existing research on ADM in severe mental illness (SMI) has confirmed interest amongst key stakeholders (Hindley et al., 2019;Swanson, Swartz, Ferron, Elbogen, & van Dorn, 2006b), together with some effectiveness in reducing coercion and increasing therapeutic alliance (de Jong et al., 2016;Molyneaux et al., 2019;Thornicroft et al., 2013). However, difficulties in achieving uptake, engagement with legal provision and successful implementation are consistent (Hindley et al., 2019;Morriss, Mudigonda, Bartlett, Chopra, & Jones, 2017;Swanson et al., 2006b) and there is limited exploration of the diversity of potential outcomes beyond a reduction in compulsory admission. This situation is surprising given the relapsing/remitting pattern of many severe mental illnesses (SMI) lends itself to the development of highly informed ADM which harness service user lived expertise of repeated mental health crises.

Reform of the Mental Health Act 1983 to include statutory provision for Advance Decision Making
In England and Wales government has recently committed to introducing statutory provision for ADM using 'Advance Choice Documents' (ACDs) in upcoming Mental Health Act 1983 (MHA) reform. Unlike existing ADM provision within the Mental Capacity Act 2005 (MCA), ACDs will retain legal weight even for those detained under the MHA. This important, and service user supported (Rethink Mental Illness, 2017), policy aims to increase parity between the treatment of mental and physical illness .

Statutory provision for mental health Advance Decision Making supports human rights
Achieving legal parity for people with any form of disability is a human rights issue. Article 12 of the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD) states that 'persons with disabilities enjoy legal capacity on an equal basis with others in all aspects of life' (United Nations General Assembly, 2007). Therefore, introducing provision for formal mental health ADM that mirrors physical health ADM can be understood as redressing the balance and moving closer to the standards set in the UNCRPD.
However, this perspective has some complications, given that this proposed legal reform is seen as a way of increasing autonomy through the ethical concept of 'precedent autonomy'. This allows individuals to extend autonomy to future periods when their capacity for informed decision-making is impaired, through ADM when they are still well/ retain capacity. Here, ADM documents made by the service user when they had DMC-T would be activated when they lose DMC-T due to an episode of mental illness. This version of 'precedent autonomy' depends on accepting that an individual could be assessed as having or not having mental capacity due to disability, an idea itself viewed as a violation of the fundamental human rights for autonomy by the United Nations Committee on the Rights of Persons with Disabilities.
The Committee's General Comment 1 on Article 12 of the UNCRPD states that 'perceived or actual deficits in mental capacity must not be used as justification for denying legal capacity' and calls for the abolition of 'substitute decision-making' within mental health laws. A full discussion of the Committee's interpretation of the UNCRPD and whether using the notion of mental capacity advances or diminishes the aims of the articles of the UNCRPD is beyond the scope of this article. 2 However, a key point in the debate for mental health ADM is around whether ADM can be understood as support for the exercise of legal capacity and autonomy for purposes of Article 12, given that it operates across time rather than being the expression of that legal capacity at the specific point where the decision may be required.
This may be a particularly contentious issue in the context of 'selfbinding' advance decisions where a person makes an advance request for coercive treatment in acknowledgement that at the time when the treatment is required i.e. during an episode of illness they are likely to refuse it (Das et al., 2020).
Whatever the position on the notion of mental capacity, ADM is widely understood as a tool which can help service users realise the ideals of Article 12 but one that is challenging to implement, a difficulty recognised across multiple jurisdictions (Ward, 2017). Given the known challenges in implementing mental health ADM our research takes a pragmatic approach and aims to understand, anticipate and address the challenges within the current and reformed legal framework to enable service users to realise the benefit of these new legal provisions and support their right to exercise their legal capacity on an equal basis with others.

Study rationale and objectives
For a broad exploration of the complex subjective phenomena that may influence stakeholders' motivation and experience of engaging with ADM, we conducted a focus group study. This aimed to deepen understanding of desired outcomes, enabling conditions and barriers, while also developing co-produced ADM materials, which maximise engagement with current law (in England and Wales) and can be easily adapted to work as ACDs or within other medico-legal frameworks. Materials were further refined during a consultation process. We have already made these resources available for use in local QI and research projects, to aid future phases of outcome and process evaluation and implementation. We now aim to make them more widely available, in response to increasing stakeholder demand from both individuals and organisations.

Study design and participants
Based on the view that understanding ADM in a 'paradigm' case of fluctuating capacity (Owen et al., 2011) can create processes to support ADM in a broader and more complex range of contexts, we focussed on bipolar. The multidisciplinary research team included expertise in psychiatry, medical ethics, law, psychotherapy and lived experience of ADM in bipolar. Public and patient involvement (co-production) has been fundamental at all stages: within the core research team and through ongoing consultation with a Service User Advisory Group and third sector organisations.
The research team internally co-produced a prototype ADM template, using a conceptual model of collaborative ADM (Gergel & Owen, 2015) and informed by a survey of over 900 members of the leading national service user charity for people with bipolar (Bipolar UK) (Hindley et al., 2019) and a literature review. This prototype was presented to legal experts to ensure compatibility with existing legal frameworks.
Focus groups, a method considered useful for eliciting views from hard to reach groups, safe discussion of difficult topics, problem solving and generating new ideas (Barbour, 2008) were then held. The aim was both to explore participants' experience and opinions on ADM and to help develop the ADM materials. Participants were purposively selected from key stakeholder groups: service users (SU) with bipolar, family members/friends of service users with bipolar (F&F), consultant psychiatrists (P), AMHPs (S), and care coordinators (C). SU and F&F participants were recruited from Bipolar UK and from participants in 2 A full discussion of this issue can be found in this report: (Martin et al., 2016) L.A. Stephenson, et al. International Journal of Law and Psychiatry 71 (2020) 101563 previous studies who had signalled interest in future research participation. Clinicians were recruited by email contact with individuals from South London and Maudsley NHS Foundation Trust and professional training organisation lists. We held 7 focus groups with 10 service user, 3 family member and 19 clinician participants. Written informed consent was obtained from all subjects. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. All procedures involving human subjects/patients were approved by London -Surrey borders research ethics committee (ref: 17/LO/1071).
Following analysis of focus group data, a second prototype template, draft care pathway and clinician and service user guidance documents were co-produced and further refined through a consultation process involving 5 service user led organisations and 5 frontline community mental health teams at South London and Maudsley NHS Foundation Trust. Overall, 94 individuals took part in the focus groups and consultation process, as well as presentations at national clinical and service user conferences.
For further details on participants see Table 1a-g.

Focus groups
Focus groups, held between September 2017 and November 2018, were co-facilitated by LS and TG, with 2 also observed by GO. The facilitators' broad range of professional and lived experience and expertise allowed the experience and views of diverse participants to be clearly understood and deeply probed, within a secure and open environment. To provide as neutral an environment as possible, groups were held in university facilities. All groups started with a video     Stephenson, et al. International Journal of Law and Psychiatry 71 (2020) 101563 explaining ADM and a chance to request clarification of key concepts. Discussion of previous experience and opinions on ADM used a topic guide, developed by the research team and tailored to the nature of the participant group (Krueger & Casey, 2002). Participants were then shown the template and encouraged to provide candid feedback.

Consultation process
Structured field notes were taken during and immediately following consultation sessions conducted by LS and TG. Materials were also distributed to key stakeholders and any written or verbal feedback recorded.

Focus groups
Focus groups were audio recorded, transcribed and entered into coding software (NVivo 12). We used thematic analysis, (Braun & Clarke, 2006) to achieve flexibility to identify themes across a diverse dataset. The 'trustworthiness' of analysis was checked using criteria developed by Nowell et al. (Nowell, Norris, White, & Moules, 2017) and SRQR reporting guidelines were followed (O'Brien, Harris, Beckman, Reed, & Cook, 2014).
LS and TG read the raw data independently, discussed initial reflections then developed a preliminary coding framework. An inductive approach was used and both coding framework and themes were refined through an iterative process until saturation was reached. A thematic map was then used to create a logic model. This was refined by the research team until consensus was reached that themes and model accurately represented the data. Updated drafts of the template, accommodating feedback, were presented to successive groups, and revisions made until all team members were satisfied with the prototype template and supporting materials. Ongoing research was discussed and reflected upon at frequent team meetings throughout this process.

Consultation process
Feedback from the consultation process was discussed until team members agreed that all feedback had been captured and appropriately translated into development of the ADM model and materials.

Role of the funding source
The funder of this study had no role in the collection, analysis, or interpretation of data and no role in writing the report or submitting for publication. The corresponding author had full access to study data and final responsibility for decisions surrounding publication.

Results
Participants were united by enthusiasm for mental health ADM despite diverse motivations and aspirations. One predominant concern was how to reconcile aspirations with resource constraints and ensure clinical, legal and social practicability, in order to avoid unrealistic expectations about the accessibility and implementation of documents. From the initial legal consultation, the PACT was seen as a document to 'inform, not fetter, clinical judgement', and ways to avoid concerns Offers workshops and courses to SLAM service users and staff that are co-designed and co-run by trainers with lived experience and professional experience.
2 Promise Resource Network (USA) Extensive expertise and experience with providing peer support to service users who wish to create Psychiatric Advance Directives within existing US legal frameworks 2  about restricting exercise of clinical judgement were explored throughout.
Along with the presentation of the results below, a logic model ( Fig. 1) summarises participants' understanding of conditions for enabling successful ADM and desired outcomes. Illustrative quotes are included in the text and detailed for all themes and subthemes in Table 2. Key design dilemmas in creating a single feasible template, guidance and care pathway and how these were resolved during the consultation process are discussed below and summarised in Table 3. The product of this research is the PACT (Preferences and Advance decisions for Crisis and Treatment) -an ADM template with guidance materials (see supplementary materials).

Theme 1: template form and content
All participants identified format and language of materials as fundamental to the success or failure of ADM documents (subtheme). Key dilemmas involved naming the document, balancing the need for detail with manageable length and providing structured response prompts while avoiding becoming a 'tick box exercise'.
Participants were keen to include elements of informal crisis planning, the usual form of mental health ADM, while ensuring that information identifying relapse indicators and the recommended response is highly personalised.
Use of ADM documents to express treatment requests relevant to all stages of an episode, from early crisis through to discharge from hospital was less familiar but universally supported. While the most common category of preferences concerned particular medications, the form also included space for preferences around medical and nonmedical treatments and relevant care settings: community/home treatment teams, inpatient admission and discharge.
There was enthusiasm about personalising crisis communication preferences by changing the legally defined 'nearest relative' to the service user's 'nominated person', in line with expected MHA reform.

Theme 2: process and context
All participants saw making and using ADM documents as a reflective process rather than single event, with a built in review period (subtheme). Creating co-produced ADM documents was thought to require a series of thoughtful conversations with authentic engagement from all parties.
"has to be validity and ownership around it, not just the person writing it taking ownership, but collectively within the environment and the culture that you're working in….that these are actually valid, and people's views are actually heard, rather than it just being an exercise of making you feel better, that 'When you become unwell we might do this'" (AMHP).
Context was seen as critical to enabling quality of this process at three levels. First, the wider systemic context (subtheme), including legal frameworks, resources available (both to create documents and meet treatment preferences), NHS Trust level support and authentic clinical openness to co-production. Second, the nature of the interpersonal relationships (subtheme) between those making the   document, given potential difficulties of discussing coercion, concern about causing distress, conflict over treatment recommendations or harm definitions and preventing undue influence from e.g. family members. Third, the service user's personal context (subtheme): their past treatment experiences, acceptance of illness, motivation to avoid particular harms, and ensuring documents are created when sufficiently recovered from recent crisis. Addressing concerns around interpersonal power differentials within the drafting process proved a significant challenge, especially the question of how to allow for the potential added 'clout' that including clinician endorsement of service user treatment preferences might bring without devaluing non-endorsed preferences. A further complication was finding language which made clinicians comfortable with the responsibility involved in endorsing advance requests and commenting on refusals. The solution was to include distinct sections for service user requests, agreed requests and for each party to explain rationales. Signatures from all parties to affirm accuracy of discussion summaries was also requested.

Theme 3: accessibility
A major universal concern was ensuring clinician awareness of existing documents and accessibility, particularly for someone presenting out of area or lacking social support. More positively, many saw ADM documents as having potential to improve cross-agency communication and hoped that checking records for ADM documents could become standard procedure, particularly considering expected MHA reforms.

'I'm still convinced that not enough professionals learn from each other or communicate with each other and that this, depending on if you can get it onto some computer systems… would it cover enough computer systems?' (Service User).
A 'PACT access plan' section was introduced to help devise appropriate individualised storage plans, enabling accessibility. Confidentiality concerns were resolved by highlighting the confidential nature of the document on each page and including a declaration of permission for the document to be seen and used by professionals.

Theme 4: harnessing expertise
An idea highlighted repeatedly was the importance of 'harnessing expertise' from 'lived experience' to produce the document. Cyclical and often highly repetitive illness patterns create knowledge which can help inform future treatment decisions, a phenomenon unique to clinical contexts with marked and predictable episodes of illness and fluctuations in DMC. The reflective questions in the PACT aim to encourage service users to adopt a position of hindsight and draw out their expertise.

Theme 5: personalising medico-legal assessment
The idea of direct service user engagement with the medico-legal   Stephenson, et al. International Journal of Law and Psychiatry 71 (2020) 101563 framework was received positively and seen as a valuable distinguishing factor of the prototype PACT document.
'this form is very useful, and clearly all the case law that's coming along says that we need to take into account people's views when they have capacity, and what is known of them……it gives you valid reasons for setting aside some of the things they may be saying which appears to give capacity…certainly the judges are telling us through their case law that we should be taking these sorts of forms into account.' (AMHP).
Given that DMC-T assessment, especially in early crisis, was widely viewed as potentially complex and contested, a personalised capacity assessment component was introduced to help manage such difficulties.

'I think by having this document it will help that understanding…he doesn't have capacity and what the family is saying is actually what we've agreed with my dad when he is well. And that would help a lot.' (Family member).
Clinicians recognised the document's potential to increase confidence in detecting subtle signs of loss of DMC-T or managing discrepancies between expressed and enacted DMC-T.
Participants welcomed using the form to specify preferred conditions for MHA assessments. Determining thresholds for coercive treatment to manage risk was an emotive issue. Nevertheless, all participants saw a role for ADM in risk avoidance, despite likely disagreement about what constituted the kind of risk demanding intervention. While some participants assumed avoidance of admission as the main aim, there was also significant interest in using ADM to prevent harm through requesting treatment under the MHA, this type of arrangement is sometimes known as a 'self-binding directive'. In response the template was worded to encourage discussions around personalised definition of harms across multiple domains.
Some clinicians recounted positive experiences of how such 'selfbinding directives' make MHA assessments less traumatic for service users and more straightforward.
"'I will say and do anything to avoid admission, so just ignore me, and crack on and do your job'. He actually wrote that….So that works fantastically well." (AMHP).
In response to some concerns about uncertainty over the legal status of the ADM document guidance was included, drafted in collaboration with legal experts.

Theme 6: outcomes
A diverse range of potential outcomes, both positive and negative, were envisaged by all parties and understood to occur at multiple points during the process of making and using ADM documents. Participants' experience led them to believe the process of creating these documents could offer a space for service users to reflect on their experience of living with their illness leading to enhanced self-management. In addition, it was felt that a collaborative process of making ADM documents could increase understanding of the service user's experience and foster a stronger therapeutic alliance with family members and health professionals.
'the last time I was hospitalised I went in as a voluntary patient, and I was slowly getting to know more about myself…I reckon now is the time that I can be most honest with myself and work out what's going on with myself the most, and be more honest with those nearest to me so that they can pick up the signs' (Service User).
'it can help families to come together around the illness in a way that we hadn't before we did this kind of process….to realise that my view of what I had seen and understood of her illness was so different from her experience of it. And to come to some kind of shared understanding of it, and understanding the other person's memories of it' (Family member).
Conversely, (as discussed in section 3.2) there was a concern that reflecting on past traumatic experiences may be distressing and that a collaborative approach may potentiate conflict where there is disagreement about treatment options and undue influence.
If ADM documents are accessed in a crisis, participants believed this would be empowering for service users and clinicians in that they would help communicate high quality information and guide confident clinical decision making.
'if the advance directive could assist, as a kind of a quicker way to some of the things that actually are helpful and are known to be helpful, instead of each time that somebody gets admitted.' (Care coordinator).
If the contents of the document were applied to inform treatment choices participants saw the potential for service users to receive preferred and established treatment and avoid personally defined harms from the illness and unsuitable care and treatment.
Although the advantages of informing clinical decision making were well understood there was some concern, particularly amongst psychiatrists, that their decision making may be undermined, service users would receive sub optimal treatment and that as psychiatrists they could be liable. In response, clarity on this issue was sought during legal consultation and included in the guidance documents.
Participants did reflect on the impact ADM documents might have on compulsory admission to hospital. Some hoped they may encourage In the longer term service users hoped using ADM documents could offer 'peace of mind' that they could be meaningfully involved in shaping a reliable and helpful response to a crisis.
'I like this concept of legally binding advance decisions, and if I can make it… I would feel safer, and it's wonderful' (Service User).
Health professionals also hoped for this outcome but expressed concern that if service users were not able to make and use these documents within a trustworthy system the disappointment they may experience if their documents were not accessed or dismissed could be damaging.
'if you create something that gives people this hope, and then there isn't a provision…. is that even more damaging rather than helpful for them?' (Consultant Psychiatrist).
Service users and service user organisations were consulted about this point and agreed there was a potential for individuals to experience painful disappointment. However, they advised this possibility should not be a barrier to encouraging engagement in ADM but highlighted the importance of statutory reforms in promoting the status of ADM documents and the necessity of providing support and high quality information about their clinical and legal limits.

Summary of main findings
We identified a diverse range of outcomes and clarified understanding of enabling conditions and barriers for mental health ADM. Taking account of such factors, we have co-produced an ADM document, the PACT (Preferences and Advance decisions for Crisis and Treatment), a care pathway and guidance documents (see supplementary materials).

'PACTs' compared to other models of advance decision making
Mental health ADM has shifted greatly from early associations with the 'anti-psychiatry' movement and complete refusals of treatment (Szasz, 1982). Internationally, a large variety of tools and processes (Henderson, Swanson, Szmukler, Thornicroft, & Zinkler, 2008;Owen et al., 2019) now allow service users to make treatment requests and recommendations, alongside more specific refusals. The UK's most extensively researched model, the Joint Crisis Plan (JCP) supports facilitated, collaborative, informal ADM with a primary aim of reducing compulsory admission. Success in a small early trial of JCPs (Henderson et al., 2004) was not replicated in a larger trial, possibly due to reduced clinical engagement (Farrelly et al., 2016;Thornicroft et al., 2013). Similar studies on JCPs in the Netherlands (Ruchlewska et al., 2014) and Switzerland have also had some success (Lay et al., 2015)(). PACTs may build on the success of JCPs by offering supported formal ADM, within an environment where statutory reform may make clinician engagement more likely.
PACTs contain service user requests and refusals (rather than relying on proxy decision makers) and support collaborative decision making. They have some distinctive characteristics, especially their explicit and guided engagement with medico-legal frameworks, designed to cater for fluctuating capacity within SMI. The PACT is a model with explicit space and guidance for personalised input into medicolegal assessment, while also allowing 'authentication' of the document itself, through elements such as documented assessment of capacity when drafting and specification of review period. They are also de-signed to challenge the model of ADM as primarily a tool to avoid compulsory treatment (de Jong et al., 2016;Molyneaux et al., 2019), by identifying and faciliating diverse outcomes; building, for example, on law reform supporting 'self-binding' in jurisdictions such as the Netherlands and Washington State (Berghmans & van der Zanden, 2012;Varekamp, 2004). PACTs also take lessons from the US, where legal provision is more widely established (Van Dorn, Swanson, Swartz, Elbogen, & Ferron, 2008) yet studies consistently demonstrate disappointingly low levels of document completion, despite high levels of service user demand (Srebnik, Russo, Sage, Peto, & Zick, 2003;Swanson et al., 2006b;Swanson et al., 2003). Possible explanations from research in the US include difficulties within the documents and lack of adequate support (Srebnik & Brodoff, 2003;Van Dorn, Swanson, Swartz, Elbogen, & Ferron, 2008), with one RCT showing facilitation to be effective in increasing uptake (Swanson et al., 2006a). Therefore, while PACTs are clearly structured around legal provision, they are also designed as flexible documents, to be made collaboratively, with clinical and family support and with clear guidance available for all parties. By treading a flexible middle ground between minimal and highly intensive clinical support, PACTs may help to achieve the benefits of ADM whilst remaining feasible within resource constrained clinical services.

Strengths and limitations
A major strength of this study was consultation with a wide range of stakeholders who would be directly involved in making and using ADM documents. In addition, data was collected and analysed by researchers with diverse academic, professional and lived experience. Limitations include that, despite our efforts, not all service users recruited for focus groups had lived experience of MHA assessment and detention. Nevertheless, even these service users saw an ADM document as personally relevant. Our sampling with regard to fluctuating capacity was restricted to bipolar though this gave us more confidence that our user focus groups were experts by experience in fluctuating capacity. Given that most (but not all) participants came from London, there may also be uncertainties regarding wider generalisability.

Future directions
In terms of future directions, the PACT materials, underpinned by increased understanding of how to enable ADM with diverse outcomes, pave the way for further evaluation of process and outcomes of implementation within mental health settings. This is crucial, given the likely MHA reforms will apply across all clinical settings. A regional project -CPMS 43078 -PACTs for Bipolar -is exploring the experience of service users, their family members and treating clinicians in South London and Maudsley NHS Foundation Trust. Projects using the PACT in other regions of England or other jurisdictions (e.g. USA or Europe) may also be possible with suitable adjustments. Early results of this study suggests that the activity of making high quality, collaborative ADM documents could be difficult to absorb within the typical structure and resources of current mental health settings e.g. community mental health teams. One way to anticipate and make provision for ADM could be dedicated 'clinics' or 'workshops', which interface with local service user organisations and clinical teams. Such a clinic could also provide a hub within regional mental health, able to lead on education, training and increasing access. This may include using technology to develop online resources and remote consultation. Research linked to such a clinic could clarify the ways in which PACTs need to be adapted for those who have different experiences of mental illness, of losing DMC-T and of mental health services. For example, those who have a high level of trust in mental health services vs those who have had traumatic experiences. While PACTs are designed for those who have a fluctuating pattern of DMC-T, there may be service users who do not regain DMC-T between episodes of illness and still wish to express their wishes about treatment. Further work could involve co-producing resources to support collaborative ADM in this group to ensure that all who wish to engage in mental health ADM have options to participate.

Conclusions
Desired aims for ADM catering for fluctuating capacity within a mental health context may range from avoidance to facilitation of admission and include both requests and refusals. This needs clear and guided personalised engagement with medico-legal processes. To achieve this, The PACT ADM template and guidance materials have been co-produced and assessed for feasibility through extensive and innovative consultation and review with key stakeholders. We are making them available, in response to increasing demand from individuals and organisations, and also hope to stimulate further research, development and availability of mental health ADM resources and optimise future implementation.

Funding
This work was funded by the Wellcome Trust, United Kingdomgrant number 203376/Z/16/Z.

Declaration of Competing Interest
None. Alex Ruck Keene was the legal adviser to the Independent Review of the Mental Health Act.