COVID-19: legal implications for critical care.

The COVID-19 pandemic has caused an unprecedented challenge for the provision of critical care. Anticipating an unsustainable burden on the health service, the UK Government introduced numerous legislative measures culminating in the Coronavirus Act, which interfere with existing legislation and rights. However, the existing standards and legal frameworks relevant to critical care clinicians are not extinguished, but anticipated to adapt to a new context. This new context influences the standard of care that can be reasonably provided and yields many human rights considerations, for example, in the use of restraints, or the restrictions placed on patients and visitors under the Infection Prevention and Control guidance. The changing landscape has also highlighted previously unrecognised legal dilemmas. The perceived difficulties in the provision of personal protective equipment for employees pose a legal risk for Trusts and a regulatory risk for clinicians. The spectre of rationing critical care poses a number of legal issues. Notably, the flux between clinical decisions based on best interests towards decisions explicitly based on resource considerations should be underpinned by an authoritative public policy decision to preserve legitimacy and lawfulness. Such a policy should be medically coherent, legally robust and ethically justified. The current crisis poses numerous challenges for clinicians aspiring to remain faithful to medicolegal and human rights principles developed over many decades, especially when such principles could easily be dismissed. However, it is exactly at such times that these principles are needed the most and clinicians play a disproportionate role in safeguarding them for the most vulnerable.

Introduction

On 31 December 2019, China reported the first cases of a pneumonia of unknown aetiology in Wuhan, Hubei province [1], subsequently identified to be due to a novel Coronavirus (2019‐nCov) [2]. Renamed as SARS‐Cov‐2 due to its potential to cause a severe acute respiratory syndrome [3], the World Health Organization (WHO) declared this virus and the disease COVID‐19 a public health emergency of international concern [4]. The virus spread globally, manifesting a pandemic [5], and outside China, outbreaks in Iran and Italy further illustrated the potential for disease and associated mortality [6, 7]. Although the UK diagnosed its index case at the end of January, the events that unfolded in Lombardy three weeks later vividly demonstrated the susceptibility of a fully functioning modern health service to this unprecedented challenge [8, 9].

Against this backdrop, in anticipation of an impending, sudden and unsustainable burden on the health service and wider civic society, the UK Government stepped up its response [10], culminating in the Coronavirus Act 2020 [11], and other legslative measures including those mandating social distancing (commonly referred to as ‘the lockdown’). These various legislative measures grant wide powers to the executive and enact deep changes in the ordinary functioning of systems across society, from health and social care to law and order and the food supply chain. In doing this, they introduce interferences with a wide range of human rights. The Act alone comes to 359 pages.

However, two critical points must be made. First, these measures do not purport to regulate every aspect of the crisis. Rather, they make targeted amendments to, and supplement, existing law. Second, they do not replace or remove certain fundamental standards relevant to intensive care, such as the Human Rights Act (HRA) 1998 or the Mental Capacity Act (MCA) 2005. Frontline clinicians should not therefore assume that they are acting in a legal vacuum where the new measures do not include specific provision for particular situations. Rather, the pre‐existing law, rights, standards, and guidance continue to apply. This existing framework will frequently adjust to take account of the present exceptional circumstances, but it is not simply extinguished – a point that was reiterated in the Parliamentary debates during the expedited passage of the Coronavirus Act [12].

This article aims to outline the broad framework within which we can consider the medicolegal and ethical aspects of some of the more readily identified issues experienced during the surge in demand for critical care. In particular, it analyses: legal aspects of patient care (standard of care; informed consent; restraint and isolation; family rights; personal protective equipment) and the legal and ethical aspects of rationing critical care. Abstract legal analysis is an imprecise art given how much turns on the specific circumstances, and so specific or final answers cannot be given.

Patient care

Negligence and the standard of care

COVID‐19 has imposed on clinicians a need to move, in certain cases, from delivering the best care they can to a lower standard of acceptable, safe care. The question thus arises as to the legal position encountered by medical professionals who knowingly, though reluctantly, find themselves providing a lower standard of care than that which they would have only recently delivered.

The English law of negligence typically defines the standard of care by reference to a ‘reasonable person’. Where a special skill is involved, the standard is that of the ‘ordinary skilled (person) exercising and professing to have that special skill’ [13]. This is assessed by reference to all of the circumstances in light of the facts known at the time. It follows that there is no absolute requirement to deliver the best possible care. The Bolam test holds that a doctor will not have been negligent where they have acted in accordance with a practice accepted as proper by a responsible body of medical opinion, even where another body of opinion takes a contrary view. The test accepts and accounts for disagreements as well as variations in the exercise of clinical judgement. Subsequent case‐law has nuanced this, noting that: any responsible body of opinion must be capable of withstanding logical analysis by the judge [14]; that professional guidance is important evidence in considering the standard [15, 16]; and that the standard is set by reference to the doctor's post, not the particular experience or ‘rank’ of the individual doctor [17, 18].

The law thus allows doctors discretion in the manner in which they provide care to patients, as well as allowing consideration of the breadth of factors which may influence their decision‐making. Over the decades, there has arguably been a ‘ratcheting up’ of the standard required as medical advances and dissemination of knowledge have occurred. In the pandemic situation, by contrast, it would allow for the degree of uncertainty and the particularly pressured circumstances under which doctors are operating to be taken into account. As one leading textbook suggests, ‘[o]ne can hardly expect the same meticulous attention in a hospital that is coping with a rail disaster or an epidemic as at normal times’ [17, 19]. In short, the test would seem to protect doctors who provide a reasonable standard of care, consistent with a logical body of medical opinion and taking into account the new circumstances specific to COVID‐19, notwithstanding the fact that this may be of a lower standard than that previously provided. This is consistent with the approach that the General Medical Council (GMC) have outlined in respect of regulatory law [20].

However, clinicians should keep three points in mind. First, this leeway is not unlimited. In particular, clinicians who have been redeployed to areas outside of their area of expertise will not be granted more leeway; they will be held to the same standard of care as others in the same post. Second, clinicians operating under such constraints need to evidence not only the treatment undertaken but also the rationale underpinning their decision‐making. A doctor asserting that they satisfied the requirements of the Bolam test may need to evidence the constraints placed on their ability to deliver care by the circumstances at the time. This may require a more extensive explanation of how decisions have been arrived at and a more thorough recording of the circumstances in which such decisions are made than was required before COVID‐19. Third, clinicians will need to ensure that their practice is acceptable to a reasonable body of medical opinion. Ensuring awareness of clinical guidance is likely to be important, and doctors retain their general duty to keep up to date with developments in medical knowledge [21]. Nevertheless, doctors are not required to read each and every publication, and practitioners may be excused ignorance of newly reported advances in practice [22]. Similarly, a clinician will not be negligent solely through a failure to practise in accordance with what is now accepted wisdom abroad [23]. These points may be of some reassurance to doctors who find themselves dealing with a rapidly evolving pandemic with great volumes of new research being published, much of it contradictory and of questionable quality.

Informed consent

Less flexibility can be expected of the law around informed consent. In the critical care setting, this is likely to be particularly important because a feature of COVID‐19 is that most patients arguably retain capacity until the point where they need sedation; contrast, in ordinary times, as per the evidence given in Ferreira [24].

Such patients must give informed consent to their treatment. Clinicians must inform them of the treatment's material risks (those that a reasonable patient would regard as significant, or that the doctor should be reasonably aware that the particular patient would regard as significant) and of reasonable alternative treatments [25, 26]. This standard is stricter than the ordinary Bolam/Bolitho one. Exceptions to this duty apply only where the doctor reasonably considers that disclosing a particular risk would be ‘seriously detrimental’ to the patient's health, or in ‘situations of necessity’, such as urgent treatment where the patient ‘is unconscious or otherwise unable to make a decision' [27, 28]. Moreover, advance care planning should be discussed where feasible [25, 29, 30]. Finally, though unlikely in the context of critical care, before carrying out further interventions it will be necessary to consider whether the patient can reasonably be woken up to provide consent [31].

A significant challenge for clinicians is the uncertainty around this new pathogen, with limited evidence on which to base decisions. That makes it difficult to judge which risks are ‘material’ and how treatment might evolve. Doctors are likely to best discharge their duty to patients by being honest about what is and is not known, within the limits of the available evidence. It will be for the patient and doctor to then decide whether to pursue a particular treatment or opt for a different approach, accepting the limitations of the evidence. One particular point that arises is the likely need to inform SARS‐Cov‐2‐negative patients of the risk of acquiring it through pursuing treatment, for example, surgery [32]. Doctors should ensure there is adequate written evidence as to the consent process.

Restraint and isolation

Patients receiving critical care are typically unable to leave. They also frequently lack capacity. That inability can arise directly from their life‐threatening condition, but it can also arise from treatment, sedation and/or restraints applied to them. Managing these issues and ensuring protection of the patient's health, life, liberty and autonomy give rise to a host of complex legal issues at the intersection of criminal, civil, regulatory and human rights law. A large body of guidance addresses these issues [33, 34]. We will focus on three important aspects of human rights law.

First, the sedation and restraint of a patient must remain tightly linked to the life‐saving treatment, be kept to the minimum necessary and not materially differ between those with and without an underlying mental disorder in order not to amount to a deprivation of liberty (Article 5 European Convention on Human Rights (ECHR)) [35, 36]. This is the principle established in Ferreira [37]. Restraints that are stricter than necessary, or that continue beyond the life‐threatening condition, are likely to amount to a deprivation of liberty, requiring authorisation by law (through the Deprivation of Liberty Safeguards process or court order). Prior consent to the restraints is highly unlikely to provide a sufficient legal basis [38]. Conversely, insufficient or inappropriate restraint may lead to harm to or even death of the patient or others, potentially breaching Articles 2 (right to life) or 3 (freedom from torture and inhuman and degrading treatment). Second, difficult issues as to the patient's autonomy and consent arise under Article 8 (right to private and family life), as addressed in the previous section. Third, particular risks of discrimination, notably on grounds of race and disability, arise in the application of restraint (Article 14 ECHR).

Against that background, two particular risks ensue in the COVID‐19 context.

First, resource limitations risk restraints going beyond what is strictly necessary. NHS England guidance envisages one critical care nurse (supervising a team of eight) per six patients in the most extreme possible ‘quadruple capacity’ situation [39]. That contrasts with ordinary guidance requiring one critical care nurse per patient [40]. In that context, the ordinary practice of monitoring sedation levels to ensure the patient does not awake may not suffice. In order to avoid this happening (and so protect the patient from foreseeable harm: Articles 2 and 3 ECHR), care teams may increase sedation levels for all patients; use physical restraint (with ‘bed buddies’ restraining any patient who wakes whereas the lead nurse attends); or use mechanical restraint (mittens). Moreover, the lower nurse‐ and doctor‐to‐patient ratios risk patients remaining sedated and restrained longer than strictly medically necessary.

In the first instance, the relevant law is sensitive to context. It is likely to tolerate stricter or longer restraint than that which would be applied when resources were optimal. So long as such restraint remains ‘the minimum required’ in that context, they are unlikely to cross the line into a deprivation of liberty. This leeway will not, however, be unlimited, and inadequate, inappropriate or clearly excessive restraint will raise Article 2, 3 or 5 issues (as appropriate). Careful thought must be given to modifying restraint policies to ensure a fair balance, based on evidence, meeting the above standards and including regular review. Notably, evidence from the US and Europe suggest that physical and mechanical restraints are ineffective at preventing forcible tracheal extubation, and increase the risk of inadequate supervision [41].

Before deploying such restraints clinicians should carefully consider a number of issues (Fig. 1).

Figure 1

Considerations for the lawful use of physical or mechanical restraints in critical care.

Second, critical care patients are subject to the same isolation and quarantine rules as others in hospital under the Infection Prevention and Control guidance [42, 43]. This guidance, which envisages possible and confirmed cases of COVID‐19 being grouped together where single/isolation rooms are unavailable, raises a number of human rights issues. The most notable are the risk that someone is infected as the result of this grouping, and the lawfulness of any deprivation of liberty to enforce the quarantine/isolation [44]; both of which are beyond the scope of this article. One critical point is, however, relevant: clinicians must be very clear as to what restrictions apply to a patient's movement, and what their legal basis is. As outlined above, the sedation and restraints that are incidental to life‐saving treatment fall outside Article 5. Once that treatment ends, any restrictions on a patient's movement must be justified on other grounds. That would apply, for instance, to continuing to isolate a patient post‐ventilation under the Infection Prevention and Control guidance [45]. Any control or restraint of the patient may amount to a deprivation of liberty and thus require a legal basis (notably Schedule 21 to the Coronavirus Act 2020).

Restrictions on family members' contact

Most hospitals have severely restricted family visiting rights to inpatients, including those in the IICU [46]. Public Health England's infection control guidance recommends permitting access to ‘essential visitors only, such as parents of paediatric patients’, but goes on to recommend local risk management with the need for proportionality [42, 47]. The way hospitals implement this raises a number of issues.

These hospitals' restrictions interfere with the family rights of both the patient and the relatives (Article 8(1) ECHR) [28, 48]. Moreover, visitors play a role in ensuring public scrutiny of standards of care; may act as an ‘advocate and defender’ of an incapacitated patient; and their face‐to‐face presence is likely to facilitate more rigorous consultation where best interest decisions are to be made than consultation by telephone or electronic mail. Their presence is likely to be particularly crucial in the case of children and vulnerable adults, above all when the visitor has the power to make decisions on their behalf (e.g. someone with parental responsibility or a lasting power of attorney).

To be justified, this interference must: pursue a legitimate aim; be prescribed by law; and be proportionate (Article 8(2) ECHR). Several issues arise. First, it should be clear which aim(s) the hospital's restriction pursues and how. The aim is likely to be the ‘protection of health’, but is this solely a question of infection control, or rather (or also) about maintaining patient care standards given reduced staff ratios? Second, is the restriction set out in adequately accessible and sufficiently precise rules [49]? This is likely to require visitor rules to be published online or to be otherwise accessible to visitors, and to make clear which categories of visitor are exempt from the general ban. Third, proportionality requires demonstration that the legitimate aim of health protection outweighs the interference with the right. Showing that less restrictive measures have been considered is typically important here. For instance, if infection control is the sole aim, hospitals are likely to need to show why providing personal protective equipment for the visitor would be impractical (for instance, due to diversion of resources) or fail to fulfil the aim (for instance because relatives are unlikely to use the protective equipment with sufficient rigour). Finally, still in respect of proportionality, ‘blanket’ bans are typically difficult to defend. That is particularly so here where, as outlined above, the degree of interference is much higher in some cases and the interests of vulnerable groups are in play, which in turn raises possible discrimination issues [50]. Policies granting exceptions for: children; persons with a learning disability; birth partners; and persons receiving end‐of‐life care are significantly more likely to be justified than ones containing no such exceptions [51]. Hospitals should consider alternative measures to mitigate the restrictions. Regular contact to update family members on the patient's situation (unless inconsistent with patient confidentiality) is required where reasonably practicable, and is in any event important to facilitating any best interest decisions and some version of an ‘advocate and defender’ role. Video calls with the patient raise interesting issues. This is best analysed through a General Data Protection Regulation lens [52]. It would constitute processing of sensitive personal data (relating to health) and so require legal bases under both Articles 6 and 9 of the Regulation. Regulatory guidance should also be complied with [53] (Fig. 2).

Figure 2

Video calls between ICU patients and their family: data protection issues. GDPR, General Data Protection Regulation; MCA, Mental Capacity Act [54, 55].

Personal protective equipment

Hospital Trusts, as employers, are under a range of duties to ensure the health and safety of staff. This includes providing safe places and systems of work. Here, we focus on the duty to provide adequate personal protective equipment (PPE). Under regulation 4(1) of the Personal Protective Equipment at Work Regulations 1992 [56], employers must ensure that ‘suitable PPE is provided to employees who may be exposed to a risk to health or safety at work except where and to the extent that such risk has been adequately controlled by other means which are equally or more effective.’

Difficulties inevitably arise during a pandemic, in the provision of resources for both patients and staff, especially in the case of a contagious risk. Whereas representative organisations re‐affirm the “ethical and legal responsibility to protect staff” and the need to “ensure appropriate and adequate protective equipment is available” [57], uncertainty has stemmed from a paucity of concrete guidance for clinicians to reconcile their regulatory and legal responsibilities with personal health and safety concerns. This uncertainty has been heightened by numerous iterations of guidance from Public Health England, seemingly suggestive of ‘an erosion in standards’ of the protective equipment required to work with patients with COVID‐19, in the absence of a change in scientific evidence [58].

Two issues arise. First, how do clinicians' duties to their patients apply where they consider that they have been provided with inadequate protective equipment? The professional duty to assist applies only as far as being able to provide treatment when it is safe for the treating clinician to do so [59]. The GMC has set out guidance as to how to proceed [20], recommending that the risks be balanced; consideration be given as to alternative courses of action; encouraging escalation of concerns and the need for ‘keeping a clear and contemporaneous record of decisions’.

Second, what this guidance does not provide is reassurance that if a bonafide decision was made not to carry out duties due to the inadequate provision of protective equipment, the clinician would be protected from regulatory and legal repercussions. The Employment Rights Act 1996 [60] offers some protection against dismissal on the sole or principal grounds of refusal to work in an area of work reasonably believed to be a serious and imminent risk to health. How this would be viewed by professional regulators, however, is unclear. This difficulty is exacerbated by the fact that should clinicians continue working with inadequate protective equipment, this could consequently present a risk to patients [61]. Ultimately, such a situation manifests a legal risk for Trusts, both for breach of the 1992 and other health and safety regulations and potentially in negligence or Article 2 ECHR claims by staff or patients. However, certain negligence claims may be indemnified by the government under section 11 of the Coronavirus Act.

Resource allocation (triage) and rationing

Legal aspects

The possibility of having to ration acute life‐saving care has perhaps been clinicians' greatest concern. The legal principles of rationing, particularly of acute life‐saving care, are not straightforward, drawing on a complex interaction of several legal regimens. This section will highlight four important areas.

Whilst there is no universally accepted definition to rationing, Scheunemann and White state that a unifying feature of most definitions is that of “denying a potentially beneficial treatment to a patient on the grounds of scarcity [62, 63]”. Using this definition, triaging or prioritising access to potentially life‐saving treatment, in the face of finite resources, is rationing.

First, rationing is an everyday part of healthcare provision, including in ICU [64]. The nationally agreed transplant benefit score [65] is one example of a rationing tool. It aims to be transparent and uses patient and clinical indices to maximise benefit; an urgent request for a life‐saving donor liver may not fit the system exactly, but still has its basis in clinical justification which acts symbiotically with resource considerations [66, 67].

In ICU, judgements as to whether a treatment is either clinically indicated in the best interests of the patient, or ‘futile’, take into account the quality of care that can reasonably be provided given the available resources. The courts have repeatedly upheld the lawfulness of resource considerations [68, 69]. In this sense, a degree of rationing has already occurred in this pandemic: reduced staff‐to‐patient ratios; use of sub‐optimal equipment; postponement of non‐urgent care; redeployment of non‐critical care staff [39]; and apportionment of organ support, all ration care. In the context of bestowing critical care during the COVID‐19 crisis, however, these measures remain within a ‘margin of acceptability’, the focus remaining on the clinical benefit or best interests of the individual patient within the new context [68, 69, 70]. The more difficult issue will come if, and when, resources are so scarce that critical care has to be refused to some patients against their best interests: when the line is crossed from ‘triage by outcome’ to ‘triage by resource’ (paragraph 2.1 [71]). This is a difficult and at times fuzzy line, but it should be drawn clearly. Dressing up a decision that is primarily based on resource limitations as a clinical or best interests decision risks illegitimacy in medicine and vulnerability to legal challenge [72].

Second, critical care networks should be used where possible to avoid the need to triage by resource. There are 18 critical care operational delivery networks across England, Wales and Northern Ireland charged with improving patient care through collaboration between provider organisations and clinical services. Use of these networks during surges is already strongly emphasised in NHS England's Surge Management Standard Operating Procedure [71], which envisages looking further to national resources where necessary ([71], paragraph 4.1]). Systemic dysfunction in these networks leading to patients being denied life‐saving treatment may give rise to a breach of Article 2 ECHR [73]. Systemic integrity may also be of relevance to the prospect of co‐ordination across networks at a supra‐regional level for the delivery of critical care [71].

Third, overt rationing by resource is likely to require an authoritative public policy and robust evidence supporting its criteria. A clear, agreed national policy has been repeatedly called for by commentators [74], concurrent with the argument that withholding clinically‐indicated ventilation may only be lawful in (i) conformity with a sufficiently robust and authoritative policy; or (ii) where multiple patients present at the same time. Furthermore, policies governing denial of life‐saving care must be clear and accessible to comply with Article 8 ECHR as per Lord Dyson in Tracey. (Liddell et al. unpublished observation) [75].

Judicial reviews of Clinical Commissioning Groups' rationing decisions on public law grounds have rarely succeeded (see Fig. 3). Policy decisions sympathetic to the ‘resources made available under current government policy’ are a legitimate consideration [77], though the courts' deference is not unlimited. Policies must not be overly rigid, in order to permit consideration of individual circumstances [78]. The need for robust evidence to support any criteria is heightened beyond the ordinary public law standards in certain cases by the human rights and equality implications of denial. Basing decisions solely on age or disability (including many underlying conditions) would amount to unlawful direct discrimination [79]. Including criteria that have a disproportionate impact on older persons or persons with a disability (for instance, using age, comorbidities or quality adjusted life years as indicators) amounts to prima facie indirect discrimination. This may be justified if the use is shown to be a proportionate means of achieving a legitimate aim (in this case, appropriately allocating scarce NHS resources), but adequate evidence of the appropriateness of these indicators will be necessary. As argued (by Liddell et al. unpublished observation), “[s]mall differences between patients may not be meaningful, and allocation decisions based on very limited evidence and poor science could be challenged”. Finally, in very exceptional circumstances, a denial of critical care to particular individuals leading to death may violate the state's Article 2 ECHR positive duties [80].

Figure 3

Judicial review and rationing decisions [76].

Fourth, perhaps the greatest difficulties arise from the prospect of withdrawing ventilatory support from a patient who is still deriving benefit from that treatment, in favour of another who may have greater survival prospects. This situation raises extremely difficult issues, with three specialist lawyers recently arguing that ‘removal of ventilation from a capacitous patient who does not consent, and withdrawal of ventilation from such a patient after he has lost capacity to communicate his decision, may be murder or manslaughter’ [74].

We make two observations. First, in certain cases, withdrawal will fit within the existing law. Where a patient with capacity consents to withdrawal, where treatment is ‘futile’ (as in the case of a patient in permanent vegetative state) [81], or where the treatment is no longer clinically‐indicated or in the best interests of the patient [81], withdrawal will be lawful under ordinary principles. By contrast, withdrawing ventilation from a patient whose best interests require continued treatment, in order to allow a different patient with better prognosis to use the ventilator, does not fall within that ordinary framework. Second, there is no clear authority on the position outside that framework. There is some limited support for the courts being flexible in an exceptional circumstance like this. As the specialist lawyers note, in Re A (Conjoined Twins) the court approved the separation of conjoined twins in circumstances which would lead to the death of the weaker twin [82]. Failure to pursue an elective separation was deemed likely to result in the death of both as they continued to grow. The court acknowledged that, where the need to act in the best interests of each child produced a conflict, it was appropriate to undertake a balancing exercise. Given the absolute non‐viability of the weaker twin, the exercise was in many ways simpler in respect of the conjoined twins than it would be for patients competing for ventilators. The principles of the case however may still be applicable: where one individual has much more to gain from a course of treatment than another; where performance or non‐performance has consequences for both patients; and where the duty towards one patient is irreconcilable with the duty to another. This interpretation would almost certainly require prior sanction by the courts. Second, the ordinary case‐law requiring that life‐prolonging treatment be given is underpinned by the duty to take ‘reasonable steps’ to protect life, rather than an absolute duty, and contains some suggestion that a different approach may apply where resources are inadequate [83] – although other case‐law appears to contradict this [84].

Our analysis illustrates the careful diligence clinicians must employ when making decisions relating to resource allocation vis‐à‐vis clinical judgement. A national emergency may recalibrate what is reasonable or even possible. Clinicians must consider resources optimally and efficiently, and avoid recognising resource considerations inappropriately in delivering life‐saving treatment. How thinly these resources are spread is a matter for policymakers, hospital Trusts and the wider NHS.

Ethics of rationing

Ethics and law do not neatly coincide. Yet, as the preceding section on legal aspects of rationing affirms, the difficult issue concerns a context in which scarce resources mean that critical care could be refused to some patients against their best interests. As previously narrated, what is needed is ‘clear guidance from an authoritative public policy and robust evidence supporting its criteria.’ It also needs ethical justification.

Unfortunately, rather than a single authoritative policy originating from an appropriate body, a number of guidelines have been published by organisations such as the British Medical Association [57], the Royal College of Physicians [85], the National Institute for Health and Care Excellence [86], and defended within academic publications [61]. This proliferation of guidelines is potentially confusing for clinicians as they are, in effect, left to choose between different sources. There should then be both clarity and a lack of ambiguity. Moreover, any authoritative statement of guidelines should be subject to broader public consultation and consistent across the health service.

Guidelines need of course to be legally unimpeachable and not, in their application, expose those who use them to legal liability. Yet, in this area where the law may not provide sufficient guidance, it is important at least that ethical advice is precise and transparent. This is not obviously the case with most of the guidelines published, for a number of reasons. First, the guidelines state a number of overarching or foundational moral principles such as: minimise suffering and loss of life; respect for the treatment of individuals; and equitability in provision. Yet often they are merely listed as relevant moral considerations without any statement of how they should be ranked. Principles can conflict. It may not be clear which principle should be applied in any particular situation. Most guidelines do not offer a means of determining which principle or consideration is decisive or of special importance. A plurality of different considerations without ranking or trade‐off rules is of no practical use.

Second, such principles do not, as generally stated, provide a clear and definitive answer to the question of who, here and now, should be offered treatment. What does it mean in such a circumstance to be equitable? Or how exactly should one seek to do the least harm and most good? This problem is compounded by the pressure of exigency.

Third, since guidelines disagree as to which criteria should be adopted, or in what order of importance, the requirement of authoritative ethical justification may not be met. The problem is that there can be reasonable disagreement as to which criterion should be decisive. It may be easy to set some aside: first come, first served; and the tossing of a coin or use of a lottery. ‘First come, first served’, whilst practical, is not a defensible moral principle in as much as the hour of first presentation should not trump considerations such as clinical need.

However, disagreement extends even to principles of prima facie plausibility. For instance, consider priority to the worst‐off patient. But this should not be without consideration of the kind of life the patient when recovered might enjoy. Prioritising the patient with the greatest chance of survival on its own fails to take account of the likely length and quality of life a surviving patient will have. A simple calculus of which treatment secures the greatest number of extra years of life is also insufficiently sensitive to the quality of life that can be led in these extra years. It is incredibly hard to make clear determinations of how to balance lives of both different lengths and different degrees of quality. In COVID‐19, clinicians are dealing with a disease of different phenotypes, and a weak clinical evidence base, particularly in regard to how the disease reconciles with patient characteristics. Indeed, much is being learned even as patients are being treated.

The use of age as a criterion for prioritisation does, as previously suggested, invite the charge of direct discrimination. Even the use of age as a marker of what are seen as relevant differences, such as ‘clinical frailty’ and the likelihood of survival, or of the prospect of fewer years of life after treatment, would be prima facie indirect discrimination. Yet, it is also hard to justify morally. Age as such is a morally irrelevant difference; as a signifier of something that is relevant it is crude and unreliable.

Yet, the use of age as such has intuitive attractiveness to many. One important reason is the ‘fair innings’ argument [87]. This holds that everyone should have an opportunity to lead a life of a certain duration. Resources should then be distributed (and care given selectively) to ensure that those who have yet to live that length of life are prioritised over those who have already managed to do so. It has an intuitive appeal: why shouldn't those who have not had an opportunity to lead a life of decent duration be preferred to those who have already done so?

Nevertheless, what exactly counts as a ‘fair innings’? And, even if we can agree, it is not clear why we should speak of fairness in this context [88]. Luck and circumstances play a major role in how long we live and it is not clear that we can speak of the length of a life as a good that can, and should be, distributed. The need for care, irrespective of age, might arise from bad luck. But it might also arise from choices for whose consequences an individual should rightly be held responsible.

A last moral consideration is this. The application of any rationing guidance will cause moral injury in those compelled by circumstances to utilise what they believe bears serious moral costs. Furthermore, rationing human resource, clinical expertise and the efforts of the multi‐disciplinary team adds to this burden. Recognition of these harms and support for those who suffer them is paramount. Support for clinicians in their decision‐making by ethics committees is desirable though it should be recognised that the provision of such services is variable.

Conclusion

It is perhaps with a twisted sense of irony that the European Convention on Human Rights this year celebrates 70 years since its inception. The COVID‐19 global pandemic will challenge rights and legal considerations in numerous and as yet unrecognised ways. These implications will not be unique to healthcare, but will be felt throughout society. Despite these challenges, such considerations are inherently malleable to context. Clinicians employing a commitment to the usual medicolegal and ethical principles, even whilst the ability to meet these commitments is being tested more so than ever, will make a disproportionately positive contribution in safeguarding such principles. In this sense, it is not ironic at all: now more than ever, we must look to the core values encoded in our rights and legal regimes in facing the new challenges raised by the pandemic. It remains to be seen exactly what final cost, not merely fiscal but also human, will be levied by this pandemic, especially on those most vulnerable in society. Undoubtedly, medical professionals, especially in ICU, will play a disproportionate role in setting this potentially intangible price.

Disclaimer

Whilst the authors have endeavoured to ensure the content herewith is correct, the information contained is intended for information only, are the views of the authors and is not intended to construe or replace formal legal advice.