key: cord-0795612-ruc7qopd authors: Brown, Christian; Keene, Alex Ruck; Hooper, Carwyn Rhys; O'Brien, Aileen title: Isolation of patients in psychiatric hospitals in the context of the COVID-19 pandemic: An ethical, legal, and practical challenge date: 2020-05-08 journal: Int J Law Psychiatry DOI: 10.1016/j.ijlp.2020.101572 sha: 61796e0c7e6460b62b42fee4da47c860aa9a9543 doc_id: 795612 cord_uid: ruc7qopd Abstract Psychiatric inpatients are particularly vulnerable to the transmission and effects of COVID-19. As such, healthcare providers should implement measures to prevent its spread within mental health units, including adequate testing, cohorting, and in some cases, the isolation of patients. Respiratory isolation imposes a significant limitation on an individual's right to liberty, and should be accompanied by appropriate legal safeguards. This paper explores the implications of respiratory isolation in English law, considering the applicability of the common law doctrine of necessity, the Mental Capacity Act 2005, the Mental Health Act 1983, and public health legislation. We then interrogate the practicality of currently available approaches by applying them to a series of hypothetical cases. There are currently no ‘neat’ or practicable solutions to the problem of lawfully isolating patients on mental health units, and we discuss the myriad issues with both mental health and public health law approaches to the problem. We conclude by making some suggestions to policymakers. principlethat of whether isolation of patients to prevent the transmission of disease ought to rely on mental health or public health law. Concluding that there are difficulties currently with both approaches, we then make some suggestions for policymakers based on our exploration of the issues. The COVID-19 pandemic is likely to affect mental health services in a multitude of ways. Increased levels of anxiety in the population may precipitate psychiatric crises (e.g. Holmes et al., 2020; Gunnell et al., 2020) . The effect of social distancing measures, and consequent isolation, means that social protective factors and safety nets will suddenly be absent (Reger et al., 2020) . Normal routines, occupations, and even the easy availability of food and other essentials, have been interrupted. Mental health services are changingday centres closing, routine appointments either being cancelled, put-off, or conducted remotely. All of these disruptions are likely to have destabilising effects on the mental health of those living with serious mental illness. The UK Government has passed new legislation 1 which, if it comes into force (the expressed hope being that it will not be necessary), would significantly lighten the usual legal safeguards around detention and ongoing treatment. 2 Already there have been changes to the ways that mental health review tribunals (Courts and Tribunals Judiciary, 2020) , and second opinion assessments take place (CQC, 2020) . Mental health trusts are under increased pressure to manage patients in the community (albeit with reduced availability of key professionals), likely altering thresholds for admission and discharge (NHSE & NHSI, 2020a) . social distancing, new le gislation in the UK 3 has made it a criminal offence to leave the place one is living without reasonable excuse. The need for isolation is particularly acute in mental health inpatient wards. Close living quarters provide ample opportunity for transmission between patients and staff (Campbell, 2020) . Whilst not a mental health unit, this risk was illustrated in the well documented case of a long-term care facility in Washington, US. Following identification of one case of COVID-19, a further 129 cases among patients, staff, and visitors were found, of whom 23 died (McMichael et al., 2020) . Those with serious mental illnesses often have poor physical health (De Hert et al., 2011) , and some common psychiatric medications increase the risk of developing serious respiratory infections (Kuo et al., 2012) . Those who lack the ability to understand or properly appreciate the risks of the disease, and the necessary behavioural modifications to stay well (e.g. physical distancing, frequent effective handwashing), are particularly vulnerable, and will be over-represented in psychiatric inpatient settings (Zhu et al., 2020) . Patients in an agitated state may have very close face to face contact, and spitting at other patients and staff is not uncommon. Mental health service providers have an ethical duty to prevent harm to their patients (and staff), and given the above considerations, it follows that there are compelling reasons to effectively prevent the introduction and spread of COVID-19 in these units. There are a number of different approaches to achieving this. At the most drastic, units can attempt to entirely shut themselves off from the world. Staff at one care home in Scotland have moved into temporary on-site accommodation to minimise the chance they would introduce COVID-19 from the community (O'Brien, 2020) . Prisons in England have stopped all visits, and all but essential reasons for leaving cells (Beard, 2020) . In a psychiatric hospital in Wuhan, China, new admissions were automatically isolated for fourteen days in order to observe for symptoms before being allowed into the main hospital (Zhu et al., 2020) . 'Cohorting' of patients, with certain facilities designated for use only by those with confirmed COVID-19, allows patients to live less restricted lives than those who are isolated individually. Where cohorting is not possiblepatients may need to be isolated in their rooms. The necessary duration of isolation is an important consideration. Current testing methods are understood to have a non-negligible false-negative rate (West et al., 2020) , and so may be falsely reassuring that patients don't have COVID-19 when in fact they do. We want to emphasise in light of this, that it may be justifiable for infection control policies to necessitate a set period of time for isolation on the basis of clinical judgement even in case of a negative test result. This would have the effect that review of the need for isolation could only meaningfully happen on a relatively infrequent basis, 4 e.g. initially after seven or fourteen days. Infection control policies should allow for the least possible restriction on individual liberty that adequately mitigates riskthough the science and ethics of this balancing exercise fall outside the scope of this paper. However, if these policies allowed scope to terminate seclusion early on the basis of test results and clinical presentation, this would underline the importance of regular review of those subject to isolation, as well as the availability of viral testing for mental health patients. In practice, then, on a psychiatric ward, should a patient have suspected or confirmed COVID-19, there are three main ways in which infection risk might be addressed:  The patient could be discharged home. So long as they do not live in a similarly shared environment (e.g. a care home), this option is likely to mitigate overall transmission risk, and will certainly reduce the risk of transmission within the mental health unit. This would limit the availability of usual ward provisions (e.g. occupational therapy) to them.  Cohorting. The patient could be moved to a ward or ward area which is designated only for those with suspected or confirmed COVID-19. This strategy is part of national infection control advice for care homes (DHSC et al., 2020) and hospitals (PHE, 2020) , and would apply similarly to mental health units. Ward activities would in theory be able to proceed fairly normally in a cohorted environmentthough staff (specially allocated to the unit) would be wearing appropriate protective equipment, and there would likely be additional restrictions on visitors to / leave from the unit.  Isolating the patient in their rooms. In practice this might take many different forms. If the patient is violent or aggressive, they may need a specially built room (see 4 As compared to, say, the usual requirements for review of seclusion as per the Mental Health Act 1983 Code of Practice. J o u r n a l P r e -p r o o f section 2.2.3)however most would be able to stay in their bedrooms, which may or may not have an en-suite bathroom. If there is no en-suite they would need to leave their room to use a facility designated for their use only. Food and drink would be brought to them by a member of staff. If the patient were unable to comply with isolation (e.g. constantly wandering out), it may be necessary to have a member of staff outside their room at all times to remind them to stay in the room and prevent them leaving, or it might be necessary to lock their door. In any case it would be important that isolated patients were able to contact staff quickly in order that their needs were metand in some cases this in itself might necessitate a dedicated staff member being nearby at all times. Many usual ward provisions would not be possible in isolation (e.g. sports, group sessions, communal mealtimes)though some would be, even if in an adapted form (e.g. medication, medical assessment, psychotherapy via video-call). A decision as to which strategy to employ will be led by clinical considerations, including the patient's care needs and risks. Guidance published by NHS England (NHSE & NHSI, 2020b) acknowledges that isolation of patients due to suspected or confirmed COVID-19 may be challenging, particularly if the patient refuses. It suggests that providers develop appropriate strategies, within legal constraints, on a case-by-case basis, and with reference to organisational ethics committees and medicolegal colleagues. It goes on to state that the 'key human right that is at risk when considering the management of people who will not self-isolate is the Right to Liberty which is a non-absolute right. This means that any restriction on this right has to be lawful, necessary and proportionate'. There are a multitude of frameworks in England and Wales which provide the necessary procedure prescribed by law to bring about a lawful deprivation of liberty for purposes of Article 5 of the European Convention on Human Rights ('ECHR'). The statutory frameworks pertinent to this discussion include the Mental Capacity Act ('MCA 2005') , the Mental Health Act ('MHA 1983') , and public health legislation. We will provide an overview of each in turn, before applying them to a series of hypothetical cases. First, we will consider the broad human rights implications of respiratory isolation, and the potential applicability of the common law doctrine of necessity. J o u r n a l P r e -p r o o f The decisions that are made in this context engage the positive and negative obligations upon the State imposed by the ECHRand are therefore of immediate relevance to the 47 states of the Council of Europe. Given the development of ECHR in relation to the Universal Declaration of Human Rights, particularly Article 3 thereof, we suggest that the following considerations are of even more widespread global concern. The ECHR was 'domesticated' as part of the law of the United Kingdom by the Human Rights Act 1998. The key positive obligation 5indeed, the justification for taking the draconian steps that may be necessaryis the obligation under Article 2 ECHR to take practicable steps to secure the right to life of others who may be affected by the transmission of COVID-19. The negative obligations are those arising under Article 2 (i.e. to ensure that the steps taken do not give rise to a risk to the life of the person themselves), Article 3 (that the steps taken do not give rise to torture, inhuman or degrading treatment), Article 5 (that, if the steps give to a rise to a deprivation of liberty, this can be justified by reference to the criteria contained within Article 5, and are accompanied by the requisite procedural protections), and Article 8 (that the steps do not represent either arbitrary or disproportionate interference with the right to private lifeincluding autonomyof the individual required to isolate). Isolation will also give rise to a positive obligation upon the State under Article 8 to put in place 'off-setting' measures to ensure, for instance, that the consequent additional hurdles that may be put in place for maintaining contact between the individual and their family are overcome insofar as possible. The steps that are outlined below will engage most directly with Articles 5 and 8, which means that an intense focus will always be required upon the question whether the steps required are necessary and proportionate. 6 One contextual aspect of some importance is that the proportionality of any restrictive intervention is sensitive to the wider context in which the restriction takes place. 7 International 8 and national advice, 9 and the aforementioned new legislation enforcing social distancing provides an unusual context, insofar as there have been mass restrictions on the movement of people in public throughout the country. In other 5 For an overview of the concept of positive obligations, see Akandji-Kombe (2007) . 6 A useful overview of the concept of proportionality in the psychiatric context can be seen in (Curtice, 2011). words, all individuals in the country are already subject to restrictions upon their movement. This does not mean that there is licence to ignore the proportionality of the interference with the inpatients who may be required to isolate within their roomsindeed, these patients include many who are already especially vulnerable and discriminated against by virtue of their mental illness. However, without wanting to overstate the point, we suggest that it does provide some relevant context. The common law doctrine of necessity could be invoked by providers to prevent patients from leaving their rooms, in order to prevent an immediate risk of serious harm to others. 10 Relying upon this doctrine (which, in law, operates as a defence to liability on the part of the relevant staff) is, however, problematic on a number of grounds. It is legitimate only in the very short term, and so could not be relied upon for the periods of time required for respiratory isolation. Furthermore, the isolation of patients in psychiatric hospitals due to COVID-19 may become widespread practice, and providers would be expected to apply existing frameworks or develop appropriate other policy in order to adequately safeguard ongoing restrictions on people's right to liberty. Especially when viewed through the prism of the ECHR, the courts would be likelyrightlyto take a very dim view upon reliance upon a residual common law defence to justify isolation, carrying with it as it does no clear procedural guarantees to ensure that it is not used arbitrarily. 11 The MCA 2005 has provisions to make decisions in the best interests of the individual who lacks capacity. Decisions can relate to a wide range of issues, including personal health and welfare, and financial matters. It also contains in Schedule A1 a frameworkthe Deprivation of Liberty Safeguards ('DoLS')for the authorisation of deprivation of liberty of those aged 18 and above in a hospital or care home. Where an individual is to be admitted to hospital for 10 See Munjaz v Mersey Care National Health Service Trust & Ors [2003] EWCA Civ 1036 at paragraph 46 per Lady Justice Hale: "[t]here is a general power to take such steps as are reasonably necessary and proportionate to protect others from the immediate risk of significant harm." 11 As the European Court of Human Rights did in HL v United Kingdom [2004] ECHR 471, finding that the common law defence of necessity could not operate as a basis to prevent confinement from being viewed as deprivation of liberty for purposes of Article 5 ECHR. J o u r n a l P r e -p r o o f purposes of assessing or treating mental disorder, and are to be deprived of their liberty, 12 thenif they lack capacity to consent to their admissioneither the MCA or the MHA 1983 can be used as the mechanism to render that deprivation of liberty lawful. For an overview of the complex interface which can arise in consequence, see Ruck Keene (2019) . The MCA 2005 defines what it means for someone to 'lack capacity' regarding a decision. The Act is unusual, in that it is governed by five statutory principlesthe most important to our discussion being the fourth and fifth: 4. "An act done, or decision made, under this Act for or on behalf of a person who lacks capacity must be done, or made, in his best interests. 5. Before the act is done, or the decision is made, regard must be had to whether the purpose for which it is needed can be as effectively achieved in a way that is less restrictive of the person's rights and freedom of action." 13 Decisions made on behalf of someone who lacks capacity must be made in their 'best interests'. In exceptional cases, the courts have been able to interpret best interests so as to be able to incorporate consideration of the interests of others if sufficiently linked back to the interests of the person. 14 However, outside the court room setting, and when it comes to selfisolation, the position is more difficult. Requiring a person to isolate where they do not understand what they are required to do and are not prepared to cooperate will undoubtedly amount to restricting their liberty. This is lawful under s.6 MCA 2005, but only where it is not only in their best interests, but also necessary and proportionate to the likelihood and seriousness of the harm that they would suffer otherwise. Here, the real reason for doing this would be for the interests of others, so s.6 MCA 2005 would not apply. In some cases, it might be possible to argue that leaving one's room whilst symptomatic might bring about a significant risk of assault by another patient who is concerned about infection. However, absent strong documented evidence that a particular patient or patients 12 The 'acid test' for determining whether a person is deprived of their liberty was articulated by Lady Justice Hale in P v Cheshire West & Chester Council & another; P & Q v Surrey County Council [2004] UKSC 19 at paragraph 54 as being "whether a person is under the complete supervision and control of those caring for her and is not free to leave the place where she lives". If they are under such control, and cannot (or do not) give their consent to that confinement and the state either knows or ought to know of the confinement, a deprivation of liberty. 13 J o u r n a l P r e -p r o o f are likely to take such a step, arguing best interests on these lines seems highly tenuous. Similarly it might be argued that it is in the best interests of an individual not to expose others to a potentially deadly pathogenif a fellow patient or resident died as a result of contact, it may cause serious moral injury to the individual. Once again, this is an interpretation of best interests which strays quite far from established lines of argumentand should not be relied upon to justify a deprivation of liberty. Respiratory isolation will almost inevitably be imposed for greater than 'non-negligible (2019)). The MHA 1983 is the main piece of legislation in England and Wales relating to the detention, treatment, and rights of mentally disordered persons. It is an expansive piece of law, but for the present purposes we will briefly outline 1) the Act's 'Code of Practice' ('CoP') and its principles, 2) the criteria for 'civil' compulsory admission, and 3) the CoP's existing provisions for isolating, or 'secluding', patients. The Act's widely interpretable definition of 'mental disorder' is given in s.1 as "any disorder or disability of the mind", making exceptions for learning disability (unless "associated with abnormally aggressive or seriously irresponsible conduct"), and dependence on alcohol or drugs. The 'nature' of a mental disorder relates to what is known about its character, usually 19 Such principles have been recommended by the independent Review of the Act: (Wessely et al., 2018) . 20 The CoP provides statutory guidance to professionals undertaking duties under the MHA 1983, and is prepared by the Secretary of State for Health in accordance with s.118 of the Act. Where the word 'should' is used, departures from the Code, if subsequently scrutinised in Court, will need to have been made with 'sufficiently convincing justification'. 21 MHA 1983 CoP 1.1. 22 MHA1983 s.2 2(2) J o u r n a l P r e -p r o o f over time and in relation to the environment or treatment. For example, the nature of someone's schizophrenia might be chronic, relapsing and remitting in course, responsive to medication, with relapses precipitated rapidly by significant life stressors. The 'degree' of a mental disorder relates to its present manifestations. For example, that someone who is currently manic has grandiose delusions, an elated mood, agitated behaviour, and no insight into their being unwell or needing treatment. 'Nature' and 'degree' are disjunctive conceptssuch that a patient may be detained on either alone, or both together. The patient's 'own health or safety' is broadly construed to cover a multitude of risks to the patient, including deterioration in their mental health, self-harm or suicide, the risk of accidents, or serious vulnerability (e.g. in the context of disinhibited behaviour). The protection of others is to be understood in lay terms, but might include physical violence or significant emotional harm. Section 3 is similarly constructed to s.2 but adds provisions such that it relates to medical treatment that cannot be delivered unless the patient is detained, and that appropriate treatment is available. Whilst the capacity status of a patient should be important to any decision regarding their treatment, notably, it is not a criterion for detention under any section of the MHA 1983. An application for detention under s.2 or s.3 is made by an Approved Mental Health Practitioner ('AMHP'), 23 on the basis of the recommendations of two doctors, one of whom should be previously acquainted with the patient, and one of whom must be approved "by the Secretary of State as having special experience in the diagnosis or treatment of mental disorder". 24 The assessing doctors and AMHP must have regard to the principles of the CoP, including that of 'least restrictive option and maximising independence'. In practice, this means for example, that where it is possible to "treat a patient safely and lawfully without detaining them under the Act, the patient should not be detained", 25 and that detention should be used for the "shortest time necessary in the least restrictive hospital setting available". 26 It also means that the threshold for detention may be relative to factors independent of the patient's 23 A health or social care practitioner (usually a social worker) specially trained in the MHA 1983 and recognised as such under the Act (s.114). 24 MHA1983 s.12 25 MHA 1983 CoP 1.2 26 MHA 1983 CoP 1.4 J o u r n a l P r e -p r o o f mental disorder, for example, the availability of care and treatment in the community, or the number of beds available locally (CQC, 2018). The MHA 1983 gives powers ancillary upon detention. For present purposes, two are relevant: 1. The implied power to seclude a patient within a hospital as a 'necessary ingredient flowing from a power of detention for treatment' 27 2. The power to seclude (at least in some circumstances) as an aspect of the power to provide medical treatment for mental disorder under s.63. 28 However, as the Court of Appeal emphasised in Munjaz, "[t]he fact that there exists a power to control or protect cannot mean that any and every use of that power is lawful. There must be limits. If there were not, it would still be lawful to confine patients in the shackles and other mechanical restraints which were commonly employed in the madhouses and asylums of the past". 29 We return below to the fact that the courts in Munjzaz's case did not conclude that the seclusion to which he was subject amounted to an additional deprivation of liberty over and above the deprivation of liberty to which he was subject by virtue of beinglawfullydetained under the MHA 1983. The courts were therefore not subjecting his circumstances, or those in a similar position, to the scrutiny that they would have been had they been considering questions of deprivation of liberty. In Munjaz's case, key to the seclusion to which he was subject being found lawful, both before the domestic courts and the European Court of Human Rights, was that there was a precise and accessible framework 30 which set out the circumstances in which the seclusion 27 [2003] EWCA Civ 1036 at paragraph 40, referring to R v Broadmoor Special Hospital Authority, ex parte S, H and D, unreported, 5 February 1998 and Pountney v Griffiths [1976] AC 314. 28 [2003] EWCA Civ 1036 at paragraph 44, Hale LJ noting that "seclusion aimed at addressing the risks to others presented by the behaviour of a patient in the manic phase of a bipolar affective disorder when the behaviour is itself the result of that disorder is treatment 'for' the disorder in the same way that force-feeding the anorexic patient was treatment for her disorder." 29 [2003] EWCA Civ 1036 at paragraph 47. 30 In fact, in the hospital policy, rather than in the Code of Practice, but the House of Lords found that there were sufficiently cogent reasons for the hospital's policy to have departed from the Code that this did not mean that the policy was unlawful. J o u r n a l P r e -p r o o f could take place, the monitoring of the seclusion, and the length of time for which it could run. Where, then, do we find a framework for seclusion? The obvious place to turn is the MHA 1983 CoP. Chapter 26 of the CoP, "Safe and therapeutic responses to disturbed behaviour", includes a provision for the isolation of patientsseclusion. This is defined as: "the supervised confinement and isolation of a patient, away from other patients, in an area from which the patient is prevented from leaving, where it is of immediate necessity for the purpose of the containment of severe behavioural disturbance which is likely to cause harm to others." 31 The Care Quality Commission ('CQC'), which monitors the MHA 1983, has yet to provide an official statement as to whether it considers isolation for purposes of securing against the risk of transmission of COVID-19 to fall within the scope of Chapter 26. As we discuss below, it falls outside the scope of the commonly understood purposes of seclusion. However, a running theme of this paper is the need for an accessible and precise framework so as to comply with the provisions of the ECHR, and, absent a court application in every case, 32 Chapter 26 provides the clearest such framework we can identify. Because seclusion represents very significant interference with the patient's rights, the Code mandates constant observation, intermittent nursing reviews, and regular medical and multidisciplinary team reviews. The purpose of reviews usually is three-fold: 1. Risk assess and ascertain whether seclusion remains necessary; 2. Assess the patient's mental health needs and adjust care plan accordingly; and 3. Assess the patient's physical health needs (including basic welfare, medication side effects, etc.) and adjust care plan accordingly. In addition to regular review, the secluded patients should be constantly observed in order to "safeguard the patient, monitor their condition and behaviour and to identify the earliest time at which seclusion can end". 33 Seclusion should be terminated as soon as it is no longer required to manage the patient's risk of harm to others. Seclusion is typically used in the context of severely physically aggressive and violent behaviour, and the provisions of Chapter 26 are clearly written with this kind of presentation 31 MHA 1983 CoP 26.103 32 As we discuss below, a court application may still be required in some cases. 33 MHA 1983 CoP 26.121. J o u r n a l P r e -p r o o f in mind. As such, it states seclusion should only take place in dedicated rooms, built to a robust specification, and able to withstand physical aggressionwithout exception, these are Spartan environments. Seclusion rooms should facilitate constant observation of the patient, and some now have systems which can remotely monitor certain vital signs. Remote monitoring of physical observations may be useful for patients isolating with COVID-19, however many other aspects of seclusion rooms will be unnecessary for non-violent patients, or needlessly restrictive of patients who will need to remain in them for many days at a time. Furthermore, seclusion rooms are generally few in number, and many units do not have one at all. The CoP's definition of seclusion describes it being used for 'severe behavioural disturbance', likely to cause harm to others. 'Severe behavioural disturbance' is not defined in the Code. Whilst in usual clinical practice this is usually taken to mean violent or extremely agitated behaviour, the phrase is open to wide interpretation. We have argued that failure to comply with respiratory isolation is likely to cause harm to others. There should be a coherent link between the behavioural disturbance and the harm to others. The CoP makes provisions for a less restrictive regimen, called 'segregation'. In this, due to a risk to others, a patient may be isolated from other patients on a ward. However, they should be allowed out of their room, have access at all times to a comfortable lounge area, and be able to use an outside space on a regular basis. By virtue of their design, the majority of inpatient psychiatric environments would not be able to support this degree of freedom whilst adequately protecting other patients from the risk of COVID-19 transmission. Segregation may apply in certain inpatient environments, but in relation to isolation, does not appear to be the answer. Schedule 21 to the Coronavirus Act 2020 ('CA 2020') gives a select group of people including Public Health Officers (PHOs) 34 extensive powers in relation to potentially infectious people. If a person has been assessed by a PHO and the officer has reasonable grounds to suspect that the person is potentially infectious, they can require the person to remain in a specified place in isolation from others for a specified period, of initially up to 14 34 Which includes a designated registered public health consultant. J o u r n a l P r e -p r o o f days. 35 Failure to comply with the requirement gives rise to a criminal offence. 36 On the face of it, therefore, the powers could be exercised so as to require individuals to remain in their hospital bedrooms in isolation, and would serve as authority to deprive them of liberty in isolation. 37 An individual's recourse would be to appeal to the Magistrates' Court. 38 At the time of writing, guidance as to the operation of Schedule 21 had not been published, although we understand that guidance had been provided to PHOs. It may be that either the internal or any future public guidance will address the position in respect of in-patient units. In any event, we do rather doubt the extent to which any PHO would wish to deploy these powers (as they stand) within the mental health setting. This is for two reasons: (1) there are already frameworks to authorise deprivation of liberty within such settings, more closely tied to the reason that the person is in hospital; and (2) the powers are of doubtful utility where the person lacks capacity to understand what it is that they are being required to do. Paragraph 14 requires the PHO, in deciding whether to impose an isolation requirement, to "have regard to a person's wellbeing and personal circumstances". "Personal circumstances" here couldand arguably shouldinclude whether they have capacity to understand what it is that they are being required to do, and the consequences if they do not. It would also be challenging to suggest that individuals lacking capacity with regards to compliance with selfisolation could or should be subject to criminal prosecution for breaching any requirement. Under the CA 2020, isolation could be enforced by the constant presence of a PHO or police officer on the ward, preventing the individual from leaving their room, or returning them to their room every time they leave. It seems highly unlikely that police resources could support the widespread enforcement of these powers, even if they were considered appropriate to be used for patients who lack capacity. There is no provision in the Act for delegation of this power to healthcare providers. Accordingly, the Act could not legitimately be used as a justification to 'lock the door' on a patient unwilling or incapable of complying with isolation. Providers would always be beholden to contact the PHO or police in case of noncompliance. Finally, for completeness, we mention the Health Protection (Coronavirus, Restrictions) restrictions on people leaving their main residence without a 'reasonable excuse', making it an offence to do so. This is enforced primarily by police, and can be punishable by a fine. Reasonable excuses include taking exercise, obtaining basic necessities, like food, and seeking medical help. It is unlikely that these restrictions apply to bedrooms in mental health inpatient units, and even if they did, that they would or could (practically) be enforced by the police. The police seem to be considering inpatient units as a "household", in which case the regulations would not apply (personal correspondence). The principle of these Regulations may apply to leave from hospital, but do not clearly apply to the issue of isolation within wards. As such, we suggest that the CA 2020 is the relevant public health legislation. Having considered the context and rationale for isolation of patients in mental health settings, as well as some of the legal background, we will illustrate the practical application of the law as it stands, using a number of hypothetical cases. These have been written to explore the issue of isolation in particularand we acknowledge therefore some artificiality in their construction. We emphasise at the outset that, where applicable, the principles of the MHA 1983 demand that all less restrictive options (e.g. cohorting, or community treatment) should have been considered before resorting to isolation. Caroline has been admitted voluntarily ('informally') due to concerns that she is experiencing a relapse in her bipolar affective disorder. She understands that she's unwell, that her erratic behaviour in the community was putting her at risk, and was distressing to her family. She understands the need for treatment, and appears to be able to consent validly to an informal psychiatric admission. She has begun coughing and has had a documented fever of 38.5C. She doesn't think that she has COVID-19, and believes that even if she did, the media has over-blown the risks, and that she shouldn't self-isolate in her bedroom. A viral swab has been sent but there is a 2 -3 day wait for results. There is no capacity to 'cohort' patients. As discussed in the introduction, in the absence of a cohorting option, there are compelling reasons for Caroline to be isolated from other patients on her wardshe presents a significant infection risk to a number of physically vulnerable people. We might start by considering whether she has capacity to make the decision not to selfisolate. There certainly is evidence of a disturbance in the functioning of her mind or brainshe is hypomanic, and her resultant grandiosity may have impacted her appraisal of the situation -or she may be engaging in a 'manic defence', denial in the face of significant health-anxiety. It is unclear whether she really understands the concerns of the team and her need for isolation (given her denial of the facts and risks), and her hypomania may also have affected her ability to weigh up her options appropriately. However, as discussed, even if on balance we felt that Caroline lacked capacity with regards to this decision, the MCA 2005 cannot be relied upon to justify a deprivation of liberty in this context. The relevant risk is to others, not to herself, and so it is a considerable stretch to argue that it is in her best interests to be locked in her bedroom for a week. Clearly, if it was felt that Caroline did have the relevant capacity, the MCA 2005 would not be of any assistance. In the very short-term, and to prevent her posing an immediate risk of serious harm to others, it may be necessary and proportionate to stop Caroline from leaving her roomrelying on common law to do so. 39 However, as discussed, this is far from satisfactory for the days-long period required for respiratory isolation. We may then reassess whether or not Caroline is truly consenting to being on the ward, if she is unable to comply with self-isolation. Whilst blanket restrictions and rules should usually be avoided, they may be justified on the basis of necessity and proportionality. For example, patients are usually not allowed to enter staff areas, other patients' bedrooms, or areas designated for the opposite sex. Patients are expected not to make lots of noise at night or constantly set-off alarms for no reasonto do otherwise would be a clear reason to reconsider whether they were consenting validly to being on the ward. In our view it is arguable that self-isolation of suspected or confirmed COVID-19 cases is a necessary and proportionate response to the risks of not doing soand an expectation on voluntary patients to comply with this as a condition for admission would therefore be justifiable. Given the above, and assuming that a satisfactory discussion was had with Caroline about the issue, and it was understood she had the relevant decision-making capacity, it would appear that she is no longer consenting validly (or indeed, agreeing) to the conditions of informal admission. If this is accepted, her risks to self / others should be reassessed and it considered whether or not she is detainable under the MHA 1983, or whether she could better be 39 See Munjaz at paragraph 46 per Lady Justice Hale: "[t]here is a general power to take such steps as are reasonably necessary and proportionate to protect others from the immediate risk of significant harm." J o u r n a l P r e -p r o o f managed in the community (taking into consideration the risks she now presents on the ward). If it was felt that Caroline had capacity with regards to the decision to stay on the ward or not, she could make this choice herself, though her capacity status does not preclude the use of the MHA 1983. If detained under the MHA 1983, questions and considerations about how best to review her deprivation of liberty and meet her physical and mental health needs remain, to which we return below. Once again, there is good reason for Owen to be isolated at present on the ward. He lacks capacity regarding the decision to self-isolate, but as discussed above, that is unlikely to affect our management. He is being locked in his room (which has an en-suite bathroom), though even if the door was unlocked, he'd be stopped from leaving by his 1:1 nurseeither way his movement has been restricted very severely. Owen's team may explore public health legislation, but for the same reasons as discussed in relation to Caroline, these would be of questionable application to someone who by virtue of their mental disorder cannot comply with isolation, and lacks relevant capacity regarding this decision. Owen's leaving his room presents a risk of harm to other patients, and is the result of disordered behaviour as a result of his mental illness. It is arguable as to whether his J o u r n a l P r e -p r o o f behaviour represents 'severe behavioural disturbance'. Compared to presentations which would usually warrant seclusion, he is not severely behaviourally disturbed. Compared to how he usually is at his voluntary job, his behaviour certainly is severely disturbed. It is arguable therefore that he meets criteria to be secluded. Owen currently has an en-suite bathroom, his room is well proportioned, and he has a selection of his own belongings, including his mobile phone, so he can call his friends. Furthermore, he may have COVID-19 and transporting him to a dedicated seclusion room on another unit will bring about a multitude of opportunities to pass this on. Given that he's not physically aggressive or overtly distressed, there seems to be cogent reason to depart from the CoP on this aspect of seclusion policy. Very few wards outside of the high secure estate would have adequate facilities to seclude patients requiring isolation in the long-term. As discussed previously, there are three broad reasons for regular seclusion reviews, the first being to consider whether seclusion can be terminated. Depending on local infection control policy, it may be clear at the outset that Owen needs to be isolated for seven days. If so, this aspect of the review becomes far less relevantindeed it would seem disingenuous to pretend otherwise. Despite this, it may, for example, be legitimate to 'terminate seclusion' if Owen's mental state improves such that he can self-isolate on a voluntary basisthough we will discuss the appropriateness of this in a later vignette. The other two reasons (assessment of psychiatric and physical health needs) remain relevant. Indeed, consideration of the patient's physical health most definitely increases in importance, given the possibility of a serious respiratory condition. However, in the context of COVID-19, the frequency and manner of reviews must also consider the infection risk to staff. As such, psychiatric aspects of the review may justifiably take place remotely using videoconferencing (as long as this was felt clinically sufficient). Physical health will need to be kept under close review (given that patients with COVID-19 can deteriorate rapidly), and concerns escalated to the medical team as needed. It should go without saying that adequate PPE ought to be used, both by staff, and by the patient if they are able to safely tolerate this. The CoP's minimum standards for regular review must not override clinical judgement if a patient needs closer medical supervision. It is quite possible that during the COVID-19 pandemic, a high proportion of patients will need respiratory isolation. If this were the case it would be particularly important to implement cohorting strategies, so as to minimise restrictions on liberty. However, there may be points at which there are simultaneously no provisions for cohorting, and large numbers of J o u r n a l P r e -p r o o f patients being isolated. If each were subject to normal seclusion review schedules, this would have a very significant resource implication for staffing. This may be unavoidable, and therefore first require preparation-for by providersthough may in extremis provide cogent reason for individual provider policies to temporarily deviate from the CoP. There is, however, an important legal issue arising, which has not yet been the subject of consideration by the courts in England & Wales. Owen is lawfully detained at the hospital, because he has been admitted under the MHA 1983. However, Owen is subject to additional restrictions at the hospital because he is in isolation. In Munjaz v United Kingdom, 40 the European Court of Human Rights ('ECtHR') recognised that individuals detained in hospital under the MHA 1983 enjoy residual liberty there, and that it is possible on the facts of any given case for additional restrictions imposed upon them to amount to an additional deprivation of liberty itself requiring authorisation by way of a procedure prescribed by law. The ECtHR found that the periods of seclusion that Colonel Munjaz was subject to (which were as long as 21 days at a time) did not cross the line. It placed weight upon four factors: 41 (1) that he was already in a high security hospital; (2) that the seclusion was not imposed as a punishment but was such as to contain severely disturbed behaviour which is likely to cause harm to others; 42 (3) the length of his seclusion each time was a matter of clinical judgment; and (4) (the factor upon which the court placed the greatest weight) the seclusion was implemented in as liberal a fashion as possible. The reasoning of the court is perhaps a little obscureafter all, whether or not a confinement should be seen as a deprivation of liberty arguably has nothing to do with whether it is said to be clinically necessary: that goes to whether the deprivation of liberty is justified. 43 The first of the factors will also not apply in most situations. However, the logic of the judgment does suggest that, especially if seclusion in isolation is carefully monitored, and the impact upon the patient minimised, the line may not always be crossed into the situation of an additional deprivation of liberty. This is to say that the manner in which providers safeguard Article 5 rights, e.g. using seclusion provisions of the MHA CoP, or otherwise, may be determinative of whether or not that person has been deprived of residual liberty. And, if they have been so deprived, then 40 [2012] ECHR 1704. 41 [2012] ECHR 1704 see paragraphs 69-72. 42 It also noted that Colonel Munjaz did not argue that the periods were unnecessary; his argument was, rather, that they should be recognised as deprivation of liberty. 43 That is very much the logic of the decision of the majority of the Supreme Court in Cheshire West [2014] UKSC 19. J o u r n a l P r e -p r o o f unless the public health mechanisms described above in relation to Caroline are invoked, it seems likely that an application to court would be required. Otherwise, there must be grounds to doubt as to whether there would be lawful authority for the additional deprivation of liberty to which Owen would be subject. 44 Although there are powers to seclude ancillary upon detention (discussed above), those powers have never been tested against the proposition that, for purposes of Article 5, the person is to be seen as deprived of their liberty, as opposed merely to having their liberty restricted. If they are so deprived, then whatever the position at common law, 45 there would have to be specific additional consideration, justification, and authorisation of the position to satisfy Article 5(1) ECHR and (arguably) specific additional appeal rights to a court so as to satisfy Article 5(4). Indeed, even if seclusion proceeded by reference to the procedure set down in Chapter 26 of the MHA 1983 CoP, it is arguable that would not be sufficient to satisfy Article 5 (for instance, Owen would have no stand-alone right of appeal to a court against his seclusion). The question therefore is whether seclusion itself (as applied to COVID-19 isolation for a fixed time duration), flexible application of seclusion, or indeed bespoke policies developed for COVID-19, would risk crossing this line. The decision in Colonel Munjaz's case is also of importance for reminding us that, whether or not Article 5 is engaged, Article 8 is undoubtedly engaged, and that: "the importance of the notion of personal autonomy to Article 8 and the need for a practical and effective interpretation of private life demand that, when a person's personal autonomy is already restricted, greater scrutiny be given to measures which remove the little personal autonomy that is left." 46 It would be infinitely preferable were the Government to issue an addendum to the CoP to address the position in relation to those such as Owen, because, without such addendum, Trusts (and private hospitals) are operating without any very clear framework. Even with such a framework, and as noted above, a question would still arise as to whether that would 44 That court would have to be the High Court exercising its inherent jurisdiction, whether or not Owen has capacity to consent to remaining in isolation. Even if Owen lacked capacity, the Court of Protection could not authorise the deprivation of liberty because of the operation of s. J o u r n a l P r e -p r o o f suffice to comply with Article 5 in the event that the line were crossed from restriction to deprivation of liberty, but it would make it much easier to contend that the line was not crossed. At a minimum, and pending any addendum, a very clear policy should be put in placeand, ideally, agreed with CQC. 47 , in which Mostyn J considered the situation of a person subject to Do LS at a care home who had his telephone calls and correspondence monitored and his room and person regularly searched. The Official Solicitor raised concerns as to whether the restrictions in this case were compliant with Article 8 on the basis that they were insufficiently prescriptive, carried insufficient safeguards, and lacked validation and oversight by a public body. To address these concerns the parties agreed a 52-page policy document that included specific policies governing searches of the man and his room, as well as monitoring of his telephone calls and correspondence. Additional layers of scrutiny were also agreed between the parties, including provision for the NHS Trust to periodically review each separate policy and receive monthly reports. The Official Solicitor submitted that the agreed policies and procedures put beyond doubt any question of compliance with Article 8 ECHR. The care home, which was said to have agreed the policies out of benign concern for the man, argued that the policies were not in fact necessary to legitimise the restrictions. This was not accepted by Mostyn J, who held (at paragraphs 20-21): "… not every case where there is some interference with Art 8 rights in the context of a deprivation of liberty authorised under the 2005 Act needs to have in place detailed policies with oversight by a public authority. Sometimes, particularly where the issue is one-off (such as authorising an operation), an order from the Court of Protection will suffice and will provide a sufficien t basis in law. But where there is going to be a long-term restrictive regime accompanied by invasive monitoring of the kind with which I am concerned, it seems to me that policies overseen by the applicable NHS Trust and the CQC akin to those which have been agreed here are likely to be necessary if serious doubts as to Article 8 compliance are to be avoided. Vaughn's relevant mental disorder (ASPD) is resulting in behaviour which is putting others at risk of harm. When asked to comply with a plan which would mitigate this risk, his behaviour becomes more aggressive, and could be described as 'severely disturbed'. In this context it seems that seclusion, in a dedicated seclusion room, may be the only way to manage his risk to others (both in terms of physical aggression, and infection risk). Important considerations remain as to how to meet his physical health needs and minimise infection risk. Given that he appears to have capacity with regards to his decision not to self-isolatepractical difficulties notwithstanding, it may be appropriate to seek to invoke the public health legislation set out above. This legislation is enforceable by the police, even if they may not necessarily be willing to seek to do so. It may, however, provide a deterrent to Vaughn in this situation. It is unlikely that it would be justifiable to remove a patient from a mental health ward to custody on public health groundsthough as discussed above, risk to other patients on the ward needs to be factored into the equation as to whether admission is the least risky way to manage the patient, and may give cause to discharge a patient. Were Vaughn to actively spit at another person he could be charged with assault. It is not impossible that Vaughn may decide, after being informed of the relevant public health legislation, that he is happy to self-isolate on the ward. We will return to the legitimacy of such 'voluntary' self-isolation in the final vignette. Ines is a 46 year old lady who is severely depressed and detained under S3 of the MHA 1983. She displays profound psychomotor retardation and very rarely leaves her room. She is provided meals in her room. She is mute, and appears distracted. It is not possible to engage her in conversation and it is not clear that she understands that given her new cough, she presents an infection risk to others and is being asked to self-isolate. Nonetheless, she 'passively complies', never trying to leave her room. Crucially, if she did try to do so, she would be stopped. The same issues will arise as in Owen's case, as to whether Ines' situation amounts to a deprivation of the residual liberty to which she is subject. Applying the logic of Colonel Munjaz's case, it is less likely that she should be seen as subject to an additional deprivation of liberty than Owen, because there are no actual steps being taken to seclude her by clinical staff. Equally, however, it will be just asif not moreimportant that steps be taken to J o u r n a l P r e -p r o o f ensure that her autonomy is maximised (and her physical and mental health do not suffer). As with Owen, if she is deprived of her liberty then it seems likely that an application to the High Court will be required for it to authorise that additional deprivation of liberty under its inherent jurisdiction. David is a 52 year old man who's been admitted voluntarily following a highly risky suicide attempt. He's recently lost his job, separated from his wife and family, and been made homeless. He has no support network. He is clinically depressed and consented validly to inpatient admission on this basis, recognising that until his mood improves and he has accommodation, he would present a very significant risk to himself if he weren't admitted. He develops a cough and fever and, being aware of the government guidance, is happy to self -isolate in his en-suite bedroom on the ward. He says he's happy to be prevented (by locking his door, if need-be) from leaving his room for the next 14 days. We have alluded throughout to the possibility that patients might voluntarily self-isolate, much in the same way that they may voluntarily be admitted to hospital. Consent to such measures is valid only if made by a person who is adequately informed (of the risks, benefits, and alternatives to the proposed intervention), has capacity with regards to the decision, and who has come to their decision freely, that is, in the absence of coercion. It would be wholly inappropriate to accept as valid the consent of an individual to admission, if this consent was given on the understanding that they would otherwise be detained (e.g. initially using s.5 holding powers). This amounts to de facto detention, in which patients are effectively detained but have no recourse to the safeguards of the MHA 1983. The CoP states that "The threat of detention must not be used to coerce a patient to consent to admission to hospital or to treatment (and is likely to invalidate any apparent consent)". 48 The CoP states that seclusion should be regarded as seclusion, even if the patient has 'agreed to or requested such confinement'. 49 By extension, even if the patient consents to isolation, and particularly if that isolation will be enforced for a set, non-negligible period of timethat 48 MHA 1983 CoP 14.17. 49 MHA 1983 J o u r n a l P r e -p r o o f restriction of their right to liberty should be accompanied by appropriate procedural safeguards. Further, a patient's initial consent to admission cannot be used to authorise someone's detention should they subsequently change their mind, or lack capacity to continue to consent to their admission. 50 If the next day, David said he had changed his mind and wanted to use the TV lounge on the ward, it would be unlawful to prevent him from doing so solely on the basis that he had consented to a week's isolation the day prior. There is no provision for Ulysses clauses such as these in English law. If we accept that isolation for COVID-19 symptoms is a reasonable 'blanket restriction' (as discussed in relation to Caroline's case), then potentially David could voluntary self-isolate in a room in the hospital, but only on the conditions that it is clearly understood that the choice he had is (1) stay in the room; or (2) leave the hospitaland that he was making this choice freely. For voluntary patients who are truly free from the threat of s.5 holding powers and subsequent detention, voluntary isolation may then be free from coercion. However, for detained patients, particularly those for whom immediate discharge is a very remote possibility (e.g. those subject to s.41 restriction orders), there can be no pretence that patients requiring respiratory isolation face a true choice. In the absence of cohorting arrangements, they will either 'voluntarily' agree to self-isolation or be secluded in their rooms. There are patients who are admitted voluntarily who would have been detained had they not consented to admission. It is widespread practice, and considered a less restrictive approach, to allow this, so long as the patient was not 'threatened' with detention and therefore coerced them into admission. This is ethically difficult territory, as without knowledge of their 'true' range of options (in order to ensure it is given freely), their consent is arguably not truly informed. It could be argued that if a detained patient were not to be informed that they would be made to isolate should they not comply with isolation (in an attempt to avoid coercion), that they could voluntarily self-isolate. However, it is clear that this faces similar ethical difficulties insofar as their choice is poorly informed, and the validity of their consent somewhat artificial. 50 At least, in respect of the loss of capacity, as the law currently stands. The Independent Review of the Mental Health Act 1983 led by Sir Simon Wessely proposed that there be consultation as to whether 'advance consent' to confinement should be accepted (Wessely et al., 2018) . If it is the case that detained patients cannot avoid being coerced regarding requests to selfisolate, then it follows that they cannot validly consent to self-isolation. Returning to the issue of voluntary patients, voluntarily self-isolatingour position on detained patients is instructive, given the fine line between the duties of providers to detained and voluntary patients, as highlighted in the case of Rabone. 51 This returns us to the issue of the voluntariness of consent if voluntary patients nonetheless fall within the scope of the MHA 1983 and therefore may be made subject to s.5 holding orders. In short, voluntary self-isolation on a psychiatric ward is likely to be a tenable legal position in few circumstances, and perhaps never when the patient is already detained. Having considered the law as it stands, it is clear that none of the usual paradigms for legally isolating and managing patients in inpatient wards are either practicable or neatly apply to the current situation. Common law and the MCA 2005 do not provide the answers. Public health legislation is of little practical use in its current state, is of questionable legitimacy with regards to those lacking capacity, and is deficient compared to the MHA 1983 in terms of associated guidance or procedures for caring for patients who are isolated. Seclusion as per the MHA 1983 appears to be available for use in most cases, though there are likely reasons for departing from the CoP. In this final section we seek to address a broader question of principlethat of how the law ought to beand specifically, whether it is problematic to have to rely on mental health legislation to underpin an approach to respiratory isolation. COVID-19 presents situations in which seclusion appears to be the best answer to circumstances markedly different to those which the seclusion provisions in the CoP were designed for. We have noted that (short of obtaining a court order in each case) these provisions are the only framework that is accessible to clinicians to have recourse to in respect of testing and justifying the necessity and proportionality of the interference with the rights of the patient.  It may be clear at the outset that a patient needs to be isolated from others on a ward for a set period of time, e.g. seven days. As such, one of the main reasons for constant observation and frequent review of patients, that of considering terminating seclusion, is largely redundant. For some patients it may be possible to eventually transfer them directly to the community from isolation (e.g. discharge or on leave), though it is unlikely to be useful to make this assessment more than once daily.  Harm to others is not mediated by physical violence, and it is unlikely that the patient requires the physical security of a dedicated seclusion room.  Infection risk to staff in a hospital setting would be a compelling reason to undertake some reviews using telepresence rather than in-person.  It is likely that procedural safeguards are needed for patients who comply with isolation but lack the capacity to make this decision.  It is likely that procedural safeguards are needed for patients who 'voluntarily' selfisolate, even those who are detained under the MHA 1983, given how easily it could be argued that their consent was invalid. There is little reason these patients should require different standards of medical or nursing review of their physical and mental health, and consent status, to those who are forcibly isolated.  There are so many simultaneous isolations that it is not possible to undertake reviews at the frequency specified by the CoP by medical staff, without compromising capacity to provide essential care elsewhere in the service. In our view, all of the above considerations could constitute sufficiently convincing justification to depart from the seclusion provisions within the CoP in relation to seclusion in many cases involving COVID-19. For compliant but noncapacitous patients, and those who voluntarily self-isolate, it may be that bespoke policies (which may nonetheless use seclusion as a starting point) would be preferable. Any such policy should be written in accordance with the principles of the MHA 1983, and therefore seek the least restrictive option (e.g. ensuring the patient has access to a phone or computer in their room, 52 so as to communicate with others), and to empower and engage the patient to the fullest degree. The aforementioned guidance from NHS England states that "MHA powers must not be used to enforce treatment or isolation for any reason unrelated to the management of a person's mental health". We presume that this would allow for the management of risks arising from the person's mental disorderotherwise seclusion would almost never be permissible. The management of risk arising from mental disorder (e.g. harm to self, or harm to others) is a critical aspect of psychiatric care. Most psychiatric interventions are given with the dual aim of treating underlying mental disorder, as well as reducing the associated harm. Whilst it may be absolutely necessary in some cases to use seclusion in order to facilitate treatment (e.g. the administration of medication), the only justification for doing so would be that the risks to others were such that the treatment could not be given in a less restrictive environment. Nonetheless, the two outcomes are conceptually separateit is possible to manage risk without treating mental disorder, and it is possible to provide psychiatric care without managing risk. Seclusionthe isolation of a patient away from other people, in a sturdily-built locked room clearly falls into the category of risk management, rather than treatment. Indeed, that is why the primary consideration for terminating seclusion, is whether the patient's risk of harm to others can be adequately managed in a less restrictive fashion. The fact that seclusion has historically been seen as falling within the scope of 'treatment' powers under the Act, is testament primarily to the broad understanding of these powers as has been interpreted by the Courts. This is legal language, not medical language. In many cases, and certainly those of Caroline (mania), Owen (thought disorder), and Vaughn (ASPD)mental disorder could be easily linked to a patient's inability to self-isolate, and therefore to a risk of harm to others. Of course, if there is no link whatsoever between a patient's mental disorder and their inability to isolate, it would be inappropriate to use the MHA 1983 to enforce isolation. Another argument against the use of the MHA 1983 is the contention that 'but for' suspected COVID-19, a physical illness, these patients would not be secludedand that therefore this is a misapplication of 'sanction' given to seclusion within the Act. However, this is a misframing of the rationale. Seclusion in this instance is not because of COVID-19, rather because there is a cogent link between disturbed behaviour (inability to self-isolate being a product of mental disorder) and a risk to others (mediated by a dangerous virus). Indeed, it is often the case that factors entirely independent of a patient's mental disorder are relevant to a J o u r n a l P r e -p r o o f decision regarding seclusion. A very large and strong individual is far more likely to require seclusion in order to manage the same degree of behavioural disturbance as someone who lacks that physical presence. Needless to say, having an imposing stature is neither a symptom of mental disorder nor something we aim to 'treat' in psychiatryyet it is relevant to risk of harm to others, and so also to a decision to seclude. Fundamentally, a risk of 'harm to others' necessarily relates to factors extrinsic to the individual's mental disorder. As such, we argue that seclusion may be permitted, even if it is accepted that, minus COVID-19, seclusion would not be necessary. If the issue driving the need for respiratory isolation is the safety of public health, then should it not be public health law, rather than mental health law, which sanctions it? The advantages of a practicable solution in public health law would be many:  Public health law could apply similarly in multiple contexts, e.g. care homes and general hospitalswhere (though outside the scope of this paper to explore) similar quandaries are likely to occur. A power to isolate persons separate to the MHA 1983 would on view be disability-neutral, 53 insofar as it would not only apply to those detained under the Act;  It could authorise the isolation of patients who cannot be discharged (e.g. those subject to s.37/41 restricted hospital orders) but for whom there is no cogent link between their mental disorder and their objection to isolation;  Current public health law provides a means of appeal (to the Magistrates' Court), a right of appeal absent from seclusion under the MHA 1983;  It might force providers to develop policies specifically for the isolation of patients with COVID-19, and therefore avoid problems associated with deviating from the MHA 1983 CoP (e.g. as outlined in section 4.1);  Authorising isolation on public health grounds would avoid further watering-down an already stretched (legal) definition of 'treatment' under the MHA 1983; and  Authorisation of respiratory isolation by public health legislation would simply be more intellectually honest than doing so using mental health law. 53 And hence not taking UK law further from compliance with the UN Convention on the Rights of Persons with Disabilities. For an overview of the obligations of the UNCRPD in the context of the MHA 1983, see the independent Review of the Act: (Wessely et al., 2018) . However, the way that public health legislation is currently framed means that its operation would be problematic, even if relevant powers to isolate could be delegated to health and care providers. In particular:  As previously discussed, it is unclear how the CA 2020 ought to be applied to those who lack capacity with regards to a decision to self-isolate, in the case of either compliant or objecting patients;  Public health law makes it a criminal offence not to comply with an order to selfisolate. For those lacking capacity or for whom their inability to do so is related to mental disorder, criminalising their behaviour does not seem like the appropriate paradigm. Indeed, unless enforcement powers were delegable to health and care providers, the consequence would be a significant police presence on psychiatric wardspotentially undermining therapeutic efforts; and  The principles of the MHA 1983 demand that isolation only be used if there are no less restrictive options. Public health legislation would need to replicate or align with these principles, otherwise its use would result in incongruence in the application of restrictive practice within mental health settings. Perhaps most importantly, the legal authorisation for a deprivation of liberty does not change patients' mental and physical health needs. Recourse to public health law would not obviate the need for a well-constructed schedule of observation, therapeutic engagement, and regular review of the patient's mental and physical health, and consent status. Given the similarity in the degree of restriction on liberty between respiratory isolation and seclusion, as well as the similarities in the needs of the patient group being considered, we suggest it would be reasonable to use provisions for seclusion as a starting point for this. Evidently, there are deficiencies in the law in its current state, with regards to the isolation of mental health patients who have suspected or confirmed COVID-19. At present it seems to us that the MHA 1983 is available, and provides the best framework for providing care to, and safeguarding, those patients whose mental disorder renders them incapable of complying with respiratory isolation. However, there remains a political question as to whether this ought to be the case, one which extends beyond the borders of England and Wales. J o u r n a l P r e -p r o o f In short, we recommend UK Government 1) issue clear and practical guidance as to how to regard isolation on mental health wards in the context of COVID-19, and 2) amend public health legislation such that it is both practical and equitable for use, if required, in health and care settings. This would bring much needed clarity to professionals working in England and Wales, but as suggested throughout, might help others elsewhere. Specifically, guidance, ideally in the form of an addendum to the MHA 1983 CoP, ought to address, in relation to patients who are unable to comply with, or consent to, respiratory If public health legislation ought to be used in some cases, or indeed, if our policymakers feel that it ought to be used in all cases of isolation in the context of COVID-19 infectionthe UK Government should amend the relevant legislation (the CA 2020) in order to make it more practicable. In particular, the powers to enforce isolation under the Act should be delegable to health and care providers. Guidance should be issued alongside any changes to the law, addressing:  How providers should regard the interface between the MHA 1983 CoP and any relevant public health legislation;  How the particular needs of mental health inpatients ought to be met, including recommended schedules for review (and whether or how these should be derived from existing policies on seclusion);  How the law should be applied to, and adequately safeguard the rights of, those lacking capacity; and J o u r n a l P r e -p r o o f  Any reporting and monitoring frameworks in relation to the use of isolation. Until formal guidance is issued, we would recommend individual Trusts and private providers develop clear policies on this issue and agree these with the CQC. Legally robust policies regarding restrictive interventions are only one aspect of what healthcare providers need to develop in order to safeguard the rights of patients being isolated due to COVID-19. As discussed, providers need to consider the interplay between patient liberty and the clinical rationale for their infection control and testing policies. They should create capacity for cohorting of patients, and ensure that both staff and patients are provided adequate protective equipment for the pandemic. Amidst this crisis, there is opportunity. As much of the population have reoriented their social lives around screens and phones, despite physical lock-down, some aspects of work, culture, and socialising have opened up to those with accessibility needs (Ryan, 2020) . Organisations across the world must now think creatively as to how to maximise people's liberties, and provide meaningful psychiatric care, even within the constraints of isolation and infection control. We have an opportunity now to find better ways of working with patients which we hope will out-live the COVID-19 pandemic. Positive obligations under the European Convention on Human Rights: A guide to the implementation of the European Convention on Human Rights Presumed Asymptomatic Carrier Transmission of COVID-19 39 Essex Chambers Rapid Response Guidance Note: COVID-19 and the Mental Health Act Psychiatrists fear surge of Covid-19 cases in UK mental health units. The Guardian The Mental Health Act: The rise in the use of the MHA to detain people in England COVID-19: Interim Methodology for Second Opinions Pilot Practice Direction: Health, Education and Social Care Chamber of the First-Tier Tribunal (Mental Health) The proportionality principle and what it means in practice Physical illness in patients with severe mental disorders. I infection-prevention-and-control/reducing-the-risk-of-transmission-of-covid-19-in-thehospital-setting#patient-placement---inpatient-settings (accessed 20 April 2020) and Social Care Coronavirus: stay at home, protect the NHS, save lives -web version NAPICU Guidance: Managing acute disturbance in the context of COVID-19 Lancet Psychiatry Multidisciplinary research priorities for the COVID-19 pandemic: a call for action for mental health science Second-Generation Antipsychotic Medications and Risk of Pneumonia in Schizophrenia Disability, coronavirus and international human rights Scientific and ethical basis for social-distancing interventions against COVID-19 COVID-19 in a Long-Term Care Facility Managing capacity and demand within inpatient and community mental health, learning disabilities and autism services for all ages Legal guidance for mental health, learning disability and autism, and specialised commissioning services supporting people of all ages during the coronavirus pandemic Coronavirus: Care home workers move into campervans infection-prevention-and-control/reducing-the-risk-of-transmission-of-covid-19-in-thehospital-setting Suicide Mortality and Coronavirus Disease 2019 -A Perfect Storm? JAMA Psychiatry 39 Essex Chambers Guidance Note: COVID-19 and Deprivation of Liberty in the Hospital Setting Public health restrictions and capacity Covid lockdown opening up world for people with disabilities. The Guardian Modernising the Mental Health Act. Increasing choice, reducing compulsion: Final report of the Independent Review of the Mental Health Act COVID-19 Testing: The Threat of False-Negative Results Coronavirus disease (COVID-19) advice for the public Nowcasting and forecasting the potential domestic and international spread of the 2019-nCoV outbreak originating in Wuhan, China: a modelling study The risk and prevention of novel coronavirus pneumonia infections among inpatients in psychiatric hospitals CB would like to thank Dr Mohsin Khan for his helpful comments on the manuscript during the preparation of this article , and Dr Frankie Sanders Hewett for her careful proofreading.