Introduction

As many researchers have discussed, protocols for reverse triage are likely to be implemented if and when a pandemic overwhelms healthcare institutions and resources (Emmerich 2011; Truog et al. 2020; Wilkinson 2020; Sprung et al. 2020). The concept of reverse triage encompasses two distinct undertakings (Emmerich 2011, pp. 95–96). On the one hand there are reverse triage policies that free up resources by discharging patients either from the hospital or from Intensive Care Units (ICUs). The aim is to create additional surge capacity in anticipation of imminent and potentially overwhelming demand. The patients affected will be those who are, perhaps, not quite ready to be discharged but who are highly unlikely to be adversely impacted by such decisions. This is not the kind of reverse triage that concerns us here.

Here we are concerned with the kind of reverse triage that might be practiced during a disaster, such as when a pandemic has overwhelmed tertiary healthcare services in a particular location.¹ In this context resources—specifically care in an ICU and/or use of a ventilator—might be withdrawn from patients who are highly unlikely to recover without continued access to such resources. This would be done in order that another patient, one with greater potential for recovery, could receive treatment that would otherwise not be available to them.² Were the relevant conditions to arise, such policies would be implemented without regard for the patient’s specific ailment. Those suffering from, say, a brain injury would be considered alongside those suffering from the symptoms of SARS-CoV-2. On the face of it, this would suggest that someone living with a chronic illness or disability such that its proper management requires the ongoing support of a ventilator, could have that ventilator taken from them and allocated to another patient if they were to attend a hospital when such reverse triage policies were in effect.

That such circumstances might actually occur is not entirely theoretical. For example, although a member of the Task Force that produced it rejects the interpretation, a guideline issued by the New York State Department of Health arguably adopts such a policy (New York State Task Force on Life and the Law 2015). This has clearly been a source of concern for some (Ne’eman 2020a, b; Wong 2020).³ The fact that this issue does not seem to have been addressed by other triage protocols could be taken to imply that it is not a major concern. Equally it might be taken to imply that there is a lacuna in our thinking, and that we should take care to ensure that we fully consider the implications of such protocols, particularly when it comes to disadvantaged and vulnerable groups. Thus, the notion that a ventilator that belongs to—or has been in long-term use by⁴—a chronically ill or disabled individual might be removed from them in order to treat another patient is an issue that demands further reflection.

If we suppose that it is ethical to reallocate a ventilator from one patient to another who would otherwise go untreated, but unethical to reallocate a ventilator belonging to an individual who is chronically dependent on it, then there must be some morally relevant difference between such cases. One intuitive distinction we might make between such cases has to do with the role of ownership of a ventilator. This intuition is challenged, however, when considering the possibility of an otherwise healthy and financially privileged individual who holds a ventilator in reserve.⁵ It seems to us that most people would be reluctant to embrace the idea that an individual who happens to own a ventilator should have privileged access to it during a pandemic rather than it being used for the greatest good on the basis of clinical need. Thus, analysis of this issue should enable us to differentiate between chronically ill patients who are dependent on ventilators and those who simply happen to possess a ventilator.

The account presented in this essay begins with an exploration of a type of property rights that Sulmasy and Sugarman (1994) have previously drawn upon in their attempt to show that withdrawing and withholding treatment are not morally equivalent actions. Examining the claims they have made makes it clear that if chronically ill patients have a moral claim to the ventilators they rely on, then that claim cannot be dependent on a potentially defeasible property right alone. Our discussion then proceeds by examining the more promising terrain of phenomenology, something recently explored by Reynolds et al (2020). Their essay draws on the prominent phenomenological notion of embodiment and the idea that objects can be literally incorporated into an individual’s body and thereby into their corporeal sense of self. As a result, they argue that additional moral significance can be attached to the personal ventilators of chronic users, meaning that it is not permissible to reallocate them for the purposes of reverse triage. However, we argue that it is difficult to demonstrate that ventilators either can be or are in fact embodied in the manner supposed by phenomenological theory. Nevertheless, we suggest that alternative ground on which a comparable moral claim can be made by chronically ill patients to their own ventilators. This ground is the phenomenological notion of homelikeness, in which health can be interpreted in relation to the whole context of ease—or disease—related meaning and not exclusively as a matter of embodiment (Svenaeus 2011, 2019).  First, however, let us more clearly sketch the three scenarios that illustrate the matter at hand.

Three Scenarios

In setting forth an argument for the moral non-equivalence of withdrawing and withholding, Sulmasy and Sugarman appeal to Nozick's Principle of Original Acquisition of Holdings. They seek to claim that ownership or simple possession generates a moral claim to an object and, therefore, to its continued use. Given that this claim applies only once ownership or possession has been established, it is not relevant to decisions to commence or withhold treatment. However, it is a relevant, albeit defeasible, consideration in deciding whether to withdraw treatment. It is for this reason that Sulmasy and Sugarman claim withdrawing and withholding cannot be considered morally equivalent undertakings. Simple possession carries some kind of moral weight, thus withdrawing treatment always removes something of moral significance whilst withholding does not. Nevertheless, the claim is a relatively weak one. Even if it does militate against withdrawing treatment for arbitrary or clinically irrelevant reasons, such as those raised by Sulmasy and Sugarman’s thought experiment, the imperatives generated by clinical reality can overcome moral claims to continued treatment based on ownership or simple possession. Therefore, Nozick's principle does not rule out the withdrawal and reallocation of treatment when doing so is justified by sound clinical reasoning; it does not rule out the practice of reverse triage per se.
 
It seems a particularly fine balance will have to be struck if the notion of property rights, which is what the principle of original acquisition of holdings seeks to capture, is to offer some assistance when it comes to resolving the three scenarios outlined above. Given that the principle of original acquisition can be overridden when a relatively minor degree of greater benefit would be derived by reallocating treatment to another patient, the principle is not inconsistent with scenario A. Whilst the principle militates against arbitrary reallocation, it does not prevent it when there is a sound clinical justification for doing so. The only real concern is the relative prognosis of the patients and the degree to which they need to differ in order to justify the withdrawal and reallocation of a ventilator from one patient to another. However, on the face of it, this line of reasoning also applies to scenario B. Unless one thinks that duration of possession strengthens the moral claim generated, it would seem permissible to take a ventilator used by a chronically ill patient and reallocate it to someone else if the requisite differential in prognosis existed. Indeed, issues remain even if one supposes duration of possession to carry significance; a stronger moral claim would still be defeasible by a greater difference in prognosis. Furthermore, a chronically ill patient may have only recently acquired a ventilator, or they may have recently acquired a replacement machine. It would be odd to suppose that either of these facts could influence a decision to reallocate their machine to another patient who is more likely to recover.

Whilst the principle of original acquisition of holdings does not seem to differentiate between scenario A and B, both it and the notion of property rights more generally might have greater influence on scenario C. It may be that actual ownership, as opposed to simple possession, of a ventilator would generate a stronger (if not necessarily strong) moral claim to its continued use. Nevertheless, ownership is not absolutely inalienable. In the same way that land is acquired for the purposes of public infrastructure, a government could compel private owners of ventilators to give up possession of them and provide whatever compensation it deemed appropriate. However, presuming no such rules are in place, if someone without any pre-existing clinical reason for possessing a ventilator should arrive at a hospital with a unit of their own, should that mean they have priority over its use? Let us suppose that (1) immediate ventilation is the clinically indicated course of action for the owner of the machine but that (2) on clinical grounds, they are a less promising candidate for ventilation support than another patient already admitted to the hospital who could be treated with this new machine. Who should be ventilated: the patient who owns the ventilator or the patient with the better prognosis?

It seems to us that the property rights of ownership provide insufficient moral reason to neglect the just allocation of ventilators in accordance with clinical need. Any healthcare professionals encountering this set of circumstances would be justified in simply adding the new ventilator to the hospital’s own resources. After all, the ventilator is of little use without the necessary staff and expertise needed to care for patients attached to them. This point is underscored by the fact that the individual who owns the ventilator has arrived at the hospital. Even if ventilator-owning patients have some moral claim to their property, they have no moral claim when it comes to healthcare professionals, nor are they in a position to direct clinical practice or its underlying principles. This includes the principles that govern the social organization and delivery of care as well as those that direct the allocation of resources such as ventilators. Just because someone owns a ventilator does not mean healthcare professionals are required to help them use it.⁷ Furthermore, even if they elect to do so, as might be the case if the owner of the ventilator also happened to have the better prognosis, it continues to be ethically mandatory for healthcare professionals to assist in accordance with their own protocols and to treat the ventilator-owning patient in exactly the same way as patients who do not own such machines.

Given these comments we might conclude that, whilst Sulmasy and Sugarman may have helped to draw distinctions between withdrawing and withholding treatment, property rights are not that helpful when it comes to the actual (re)allocation of life-saving resources. Whilst the principle of original acquisition of holdings may favour the status quo, once this is disturbed by some clinically relevant factor withdrawing and reallocating treatment seems justified. Indeed, given that owners of ventilators cannot expect to control how their property is allocated not least because they cannot expect to redirect established protocols regarding resources and professional practice. Property rights do not seem to count for very much when it comes to the allocation and reallocation of treatment under conditions of reverse triage. This being the case, however, means that reallocating ventilators from long term users remains ethically possible.

Conclusion

The analysis offered in this article proceeded in light of what we take to be common moral responses to the three scenarios outlined. First, if we accept the implementation of reverse triage policies that withdraw treatment from those who will not survive without it, we acknowledge the propriety of reallocating ventilators to patients with a better chance of recovery. Second, we question the propriety of reallocating a ventilator belonging to a chronically ill patient who is reliant upon it in an ongoing manner. We feel they should have priority over its use even if its reallocation would likely mean that another patient would derive greater benefit. Third, we accept the propriety of allocating a privately owned ventilator in accordance with established protocols, meaning that its owner need not be given priority access if and when they need to be ventilated. We have argued that the moral significance of property rights, such as those elucidated by Sulmasy and Sugarman, do not resolve these three scenarios satisfactorily.

A phenomenological perspective provides resources that are more adequate to the task. Drawing upon them provides us with the means to recognise the ethical significance of embodiment as well as that which attaches to particular objects if and when they form part of our basic, health-related or embodied homelikeness. Seen in this light, the chronically ill patient’s ventilator should not be considered just another possession; it is not something that can be taken in order to pursue some other good, purpose or end in the way that is permissible for other ventilators, such as those owned by financially privileged individuals. Thus, a policy that reallocates a chronically ill person’s ventilator to another patient might be thought of as no more legitimate than one which reallocates an individual’s kidney, their prosthetic limb, or any other aspect of their embodiment to another person, even if it is more easily removed.