Fantastic, wonderful to be here in person and see actual human faces not scintillating on a monitor. So I'm going to be talking about allocating scarce ICU resources, how to ethically say no. And what I'd like to do is first lay out what I think are two obviously unethical strategies or approaches to doing this as a way to juxtapose with what I'll do next is give four recommendations. for the ethical allocation of scarce resources or saying no when there's not enough for everyone. These are my disclosures, none of which pose a conflict with what I'll be presenting. And here's the first example of what I think is clearly unethical treatment. And let's talk about it as a hypothetical first in a region of your state or province hospitals in economically disadvantaged areas are overwhelmed with critically ill patients while private hospitals nearby have available beds. The private hospitals are refusing to accept patients and transfer and the regional health authorities are also refusing to intervene to require transfer between the overwhelmed hospitals to the hospitals with available beds. And therefore physicians at the safety net hospitals are having to ration and refuse ICU care to patients who will die without it and would otherwise potentially survive. So what's the issue here? I would argue that the ethical violation on the part of both the private hospitals and the regional health authorities are that they fail to uphold their obligation to the public safety net which is to say to take steps to protect and prevent unnecessary loss of life which in this case given how ICU beds and per capita resources are allocated would disproportionately fall on disadvantaged populations. So that's example number one. And you don't have to actually look far to find that this hypothetical did in fact play out during the pandemic. These are several articles by Sherry Fink and Jim Dwyer in the New York Times and the LA Times detailing really clear examples when during the early waves of the pandemic hospitals safety net hospitals like Elmhurst Hospital in New York City were overwhelmed while nearby private hospitals had beds and the ability to transfer patients was seriously impaired such that there were in the ICUs at Elmhurst the nurse to patient ratios were ten to one. Dialysis was being split between patients and nearby at other hospitals standards of care were as per usual. All right. How about example number two and this we'll call disability discrimination. In responding to acute shortages of ICU beds and ventilators an individual physician decides that one of his patients with a moderate intellectual disability and severe physical disability from cerebral palsy won't be admitted to an ICU on the grounds that his quality of life is too poor and this patient parenthetically lives at home with his mother who addressed all his ADLs and felt that his quality of life was acceptable. So what's the ethical issue here? Well I think this one is also pretty straightforward. Triage according to one's disability status or based on the presence of a disability violates our social commitment to equal treatment of all persons and also stepping outside the ethical to the legal realm it violates the United States anti-disability Americans with Disabilities Act. Okay. And also unfortunately you don't have to look too far to see that this kind of discrimination was embedded in early triage protocols and this is a slide of Alabama's triage protocol that was developed and promulgated in 2010 and was on the books through March 2020 when the pandemic really hit. And this triage protocol includes guidance that says do not offer mechanical ventilatory support for patients with any one of the following and clearly the first two there are examples of discrimination based on intellectual disability, severe mental retardation, there are words, or moderate to severe dementia. Okay so with those two sort of things to get your appetite whetted for how not to do it let's talk about four things that should be in my view integrated into triage allocation strategies to fairly and ethically allocate scarce resources. Taking a step back thinking about what are the ethical goals of allocation in a public health emergency, Larry Gostin who's a health law scholar at Georgetown has really nicely articulated it when he wrote twin moral impulses animate public health. To advance human well-being by improving health and to do so particularly by focusing on the needs of the most disadvantaged. And what's embedded in here are two ethical goals, first promoting population health outcomes which might be thought of as a utility based goal and second promoting fairness which might be thought of as an equity based goal. So as you're thinking about triage these are I think two ethical goals at the very high level that should be front and center in your mind. Promoting the population health and doing so fairly. So the first way to accomplish this is to institute load balancing across hospitals. We along with Lisa Villarreal and John Hick we wrote an article in the New England Journal in December of last year in which we really laid out the case for this but the core recommendation was before any patient in need is refused ICU care, patients at overwhelmed hospitals should be transferred to hospitals with available beds elsewhere in the region. And this is the high level definition of load balancing. And there are a couple strategies to accomplish this. The first is in advance of acute scarcity hospitals should create what are called voluntary mutual aid agreements and the idea here is that hospitals, health systems in regions agree that they'll help each other out in the setting of one becoming overwhelmed and others having available beds and this is done as a voluntary agreement in advance not knowing who will be the health system that has the scarcity of beds. And I think in theory this is a great idea. I put a reference up here to the ASPR website that lays out a lot of the details and background about how this can be done and how it was done regionally during the pandemic. But they also when they did in-depth interviews with some of the stakeholders in this process several issues came up with this voluntary approach. First some in Michigan, some hospitals in the state just said no, we're not going to do that. We have our resources, typically these were hospitals in wealthier areas and not particularly interested in taking patients from disadvantaged inner city hospitals. So that's issue number one. Issue number two, in Utah there was a regional load balancing set up in place but there was great concern that it was tenuous in the sense that it really relied on the big hospitals in Utah, the big health systems voluntarily participating. There was some concern that if things got quite bad they would just withdraw from the agreement. And so these are the limitations of a voluntary approach to load balancing. The other approach, strategy two, is for public health officials in states to issue emergency orders to require hospitals to participate in load balancing efforts during public health emergencies and this would involve accepting patients in transfer who are not part of their normal covered population. From an ethical and legal standpoint the justification for this is quite clear. One of state's foremost responsibilities arguably is safeguarding residents' health and well-being and this is threatened if during a pandemic health systems fail to cooperate leading to preventable loss of life. And again Larry Gostin has articulated this approach to preserving the public health and he and colleagues developed a model state emergency health powers act that has been accepted in some form or taken up in some form by most states. I've put up here an example from the language in the California public health and medical emergency operations manual from 2019 which says there's a section on what they call commandeering of facilities and personnel during a proclaimed state of emergency the governor is authorized to plan for the use of private services, facilities and properties in addition to the normal approach which is procuring services and supplies by contract. So this is takings, right, a public governmental taking of private property for the expressed goal of preserving the public health in an emergency. And you might say well listen this is great in theory but not possible in practice and I would just point out that this has been done not only in numerous European countries but also here in the United States. The state of Arizona early in the pandemic created the Arizona surge line which was a cooperative system by law amongst all of the hospitals in Arizona that essentially coordinated the transfer of patients from overwhelmed hospitals many of which were actually on American Indian reservations where the Indian health services woefully underfunded and the pandemic had an especially hard impact transferring from those areas to urban centers with much higher levels of bed availability so into Phoenix and into Tucson. And this approach started with an executive order from a Republican governor, Doug Ducey, and was backed up with a lot of infrastructural support through the Arizona Department of Health and it resulted in more than 7900 transfers of patients most of whom were Native Americans. So again, an example that load balancing is actually possible in the United States in a Republican dominated state as well. All right, number two, moving away from strategies to prevent the need for triage and saying no to strategies for triaging in ways that are ethically acceptable and the idea here is to develop triage criteria that reflect, I'm sorry, that respect patients' rights and reflect community values. What do I mean by this? Well, from the standpoint of respecting citizens' rights, it's quite simple. Don't use triage criteria that are discriminatory based on, for example, disability status, gender, sexual orientation, race, ethnicity, religion, or assessments of broad social worth. All of these are clearly discriminatory and should not be used. Beyond that, there are a lot of questions, though, about permissible strategies and the idea here is that amongst these multiple ethically reasonable approaches, the way that we should determine how to triage is that doing so, it should be done in a way that respects community values and so I just put up here a list of questions around whether certain considerations should come into triage, all of which would conceivably be ethically permissible and non-discriminatory in terms of protected classes. So should patients expected to die sooner from an end-stage condition be given lower priority on the grounds of promoting more population health benefit? Should pregnant patients be given higher priority on the grounds that two lives may be saved rather than one? Again, another utilitarian consideration. Should there be efforts to mitigate inequities in population level outcomes by giving some priority to disadvantaged groups? And so this is an equity-based consideration. And then finally, should younger patients be given higher priority than older patients? Again, not on narrow utilitarian grounds, but on the grounds of equity, that dying young is a severe form of disadvantage. So all questions that could cut either way depending on the values of the populace and are the kind of things that really require careful engagement. And when one asks oneself, well, so how are we going to engage about this? How should we reason? There are different ways to think about this. And I want to draw a distinction between private reasoning and public reasoning. Public reason, I'm sorry, private reasoning would entail, for example, hospitals selecting triage criteria according to their own institutional determinations of beliefs and values, which may or may not be broadly representative of community values. An example of this would be a religiously affiliated hospital, for example, making triage criteria that are narrowly based on their particular religious directives. Public reasoning, by contrast, would involve hospitals having triage criteria grounded not in their own private values, but instead in the community's values and also made in ways that don't violate anyone's fundamental rights. So why public policies if these are private hospitals? This is, again, a fairly American issue. In countries where health care is nationalized, there's less of a public-private divide. But I just want to point out that although many hospitals in the United States are indeed private, they also almost all serve a very public function, which is providing a national health care safety net via EMTALA and other mechanisms. Also federal funding for Medicare and Medicaid go to these hospitals. So they're part of the public health care delivery system. And therefore, these hospitals have some obligation as part of the national safety net to be guided by public values and public deliberation rather than by private values. So what would this look like? Well, you know, not something we can get deep, deep into today, but at the high level, that would look like, number one, rather than individual hospitals developing their own triage criteria, that this would be done by legitimate governmental authorities. So public health officers, for example, developing statewide triage guidelines that are made with the input of citizens. So for example, key stakeholder panels or citizen advisory boards. And then those triage criteria would not apply to individual hospitals only, but would apply to all hospitals in the state. Easy to say, hard to do. We've had a number of intensivists have been involved in really exemplary public engagement processes. Lee Daugherty-Bitteson at Hopkins, in particular, has done a great job with this. And it is a huge amount of work definitely to be done in advance of the pandemic or now after the pandemic in expectation of the next. All right, number three, the third recommendation is to ensure procedural fairness in how triage criteria are applied. And as you're thinking about this, the counterpoint or the juxtaposition is there are good reasons that someone other than the individual treating physicians should make the triage decisions. And so when you think about what does it mean to have a procedurally fair approach to triage, there are a number of criteria that have been articulated over the years. Decision makers shouldn't have conflicts of interest or conflicts of commitment. Think about a doctor taking care of her own patients and also having to think about how to choose between her own patients, to whom she has commitments to both, and between her own patients and other patients. So that's clear conflicts of commitment there. Whomever is triaging should have situational awareness of the resources and patient load. So that would need, if you're at a hospital with five different ICUs or in a region with 10 different hospitals, you'd need to know what's the bed availability in all of these places. And that's something that an individual doc at the bedside is not likely to be able to know. Next, individuals who are doing triage need to be able to consistently and impartially apply triage criteria. And we've seen over the pandemic that these criteria are often quite complex to apply. Arguably without training, it's unlikely that individual docs at the bedside would be able to do that reliably. And then I've put a number of other important criteria for procedural fairness up here that I think all speak to transparency, legitimacy, accountability. These are all things that are much better done by a group or a small select number of people who have been trained in processes for how to proceed with triage. And so what could this look like? The way that this has been advocated in general in existing documents is that there should be what's called a triage team. So rather than the people engaged in the frontline care of patients, it should be a separate group who are not actively the attending of record for any patient, a physician, a nurse, support staff. And these individuals should have undergone extensive training in how to apply the triage framework. They should have had anti-discrimination and anti-bias training. They should be trained in how to record these triage decisions for retrospective review. And they should also be trained in how to manage appeals or objections to the triage decisions that are made. And then last, when we say no, which is to say when we can't provide intensive care to everyone, our obligations to non-abandonment persist. So we have ongoing responsibilities to provide whatever medical care that we can up to and including palliation if that's all that can be provided. But all patients, I think the ethical requirement is quite clear that all patients have the need and right to high-level palliation even if they can't receive intensive care. And so what could this look like? Well, of course, intensive symptom management for dying patients, psychosocial support of families, and then also something we saw in New York City when hospitals were overwhelmed, concerns about respectful handling of postmortem bodies. Okay. So I'm going to stop here and just summarize the four recommendations that I put forward are to institute load balancing across hospitals before any patient in need is denied ICU care, to develop triage criteria that respect patients' rights and reflect community values, to ensure procedural fairness in how triage criteria are applied, and fourth, to provide basic medical care and palliation to patients denied ICU care. Thanks very much. Thank you.

ethical allocation

scarce ICU resources

load balancing

triage criteria

public engagement