Bio(un)ethical
The podcast where we question existing norms in medicine, science, and public health.
Bio(un)ethical
#2 Govind Persad: How (not) to allocate resources during a pandemic
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this episode, we interview Dr. Govind Persad, an expert on resource allocation whose work influenced COVID-19 allocation policies, about how we should allocate scarce medical resources, what stood in the way of optimal allocation during the covid pandemic, and how we can improve resource allocation within the US healthcare system.
(00:00) Our Introduction
(04:05) Interview begins
(12:06) What is a “framework” for allocating health resources?
(14:20) What normative assumptions are baked into allocation frameworks?
(19:36) What principles are included in allocation frameworks?
(20:24) Overview of first two principles: Maximizing benefits and favoring the worst off
(26:11) Overview of second two principles: Equal treatment and rewarding social usefulness
(39:37) Which benefits matter when allocating health resources?
(43:33) Should we account for quality of life?
(49:22) Should we prioritize the youngest first?
(53:29) When does reciprocity matter?
(57:32) Putting principles into practice
(1:02:22) Legal considerations
(1:07:25) Improving resource allocation in the US healthcare system
(1:13:16) How living through the pandemic influenced Govind’s research
Mentioned:
- “Principles for Allocation of Scarce Medical Interventions” by Govind Persad, Alan Wertheimer, and Ezekiel Emanuel (2009)
- “Fair Allocation of Scarce Medical Resources in the Time of COVID-19” by Ezekiel Emanuel et al. (2020)
- “The Rebugnant Conclusion” by Jeff Sebo (Draft)
- “Equality and Priority” by Derek Parfit (1997)
- “Should the Numbers Count?” by John Taurek (1977)
- “In the Line For Scarce Covid Treatments, Immunocompromised Americans Should Go Before the Unvaccinated” by Govind Persad and Emily Largent (Opinion in The Washington Post, 2022)
- “How COVID-19 Hollowed Out a Generation of Young Black Men” by Akilah Johnson and Nina Martin (ProPublica, 2020)
Thoughts? Guest suggestions? Email us at biounethical@gmail.com
You can find more episodes of Bio(un)ethical at biounethical.com. Sign up for our email list at biounethical.com to receive episode alerts and submit questions for upcoming guests.
To support us, please subscribe, rate, and review our show wherever you get your podcasts, and recommend it to a friend. For updates, follow Leah and Sophie on Twitter (leah_pierson and sophiehgibert).
Bio(un)ethical is a bioethics podcast written by Leah Pierson and Sophie Gibert, with editing and production by Ambedo Media (previous production support by Audiolift.co). Our music is written by Nina Khoury and performed by Social Skills. We are supported by a grant from Amplify Creative Grants.
Hi, and welcome to Bio and Ethical, the podcast where we question existing norms in medicine, science, and public health. I'm Sophie Jabert, a PhD candidate in philosophy at MIT.
SPEAKER_02And I'm Leah Pearson, an MD PhD candidate at Harvard Medical School and the Harvard T.H. Chan School of Public Health. Today we're speaking with Dr. Govin Prasad, a lawyer and philosopher at the University of Denver Sturm College of Law. His work focuses on applying bioethical and distributive justice frameworks to address problems at the interface of health, law, and policy. Some of Govind's ongoing projects consider how to integrate health, justice, and equity into frameworks for the allocation of scarce medical resources, including resources like vaccines and medical treatments during the COVID pandemic. Govind earned a BA, a law degree, and a PhD in philosophy at Stanford University.
SPEAKER_03With Govind, we're going to talk about how bioethicists decide what principles should govern resource allocation, whether existing practices tend to align with those principles, and how we might bring the allocation of health resources into better alignment with them.
SPEAKER_02Issues related to the allocation of health resources received a lot of attention during the COVID pandemic. In March of 2020, I was a medical student on my surgery clerkship. And over a few days, as it became clear that medical grade masks would soon be in short supply, I went from casually ripping masks off my face between cases and throwing them in the trash to carefully untying them and tucking them in the pocket of my scrubs, knowing I might need them again.
SPEAKER_03As the pandemic continued, medical equipment became an increasingly short supply. Masks, of course, but also things like ventilators and ECMO machines for patients with severe COVID. Other resources were also implicated. For instance, many elective surgeries or procedures were postponed or canceled. Similarly, many routine primary care visits and preventative screenings were canceled. Although some canceled care had to do with the desire to keep healthy patients out of the risky hospital environment, much of it had to do with limitations on staff time and hospital capacity. Once treatments and vaccines emerged, these too were allocated. This led to contentious debates about which patients to prioritize for care and high-profile cases of patients who traveled thousands of miles to seek vaccines or hospital administrators with low-risk jobs who were able to access vaccines that were meant to be targeted to high-risk clinical employees but were often haphazardly allocated through imprecise policies.
SPEAKER_02In normal times, healthcare is always allocated. Whether this allocation occurs via making patients wait for hip replacements or MRIs, as happens in much of the world, or whether goods are instead allocated based on ability to pay, as too often happens in the U.S. healthcare system. Issues related to resource allocation remain perpetually fraught, as we often disagree with each other morally about the principles on which we should allocate health resources. For instance, whether we should focus on saving the most lives or prioritizing the most disadvantaged patients, whether we should prioritize patients who were in no way responsible for their medical conditions over those who plausibly were, and so on. When you couple these complex moral issues with equally complex practical issues related to how we craft and implement clear and effective policies, things can get messy quickly. As a philosopher and lawyer, Govind will cast light on both sets of issues.
SPEAKER_03With that said, let's get into the interview. As always, you can access the papers that we reference in the episode notes or at our website, biounethical.com. And you can submit feedback there or email us at bioenethical at gmail.com.
SPEAKER_02Thanks for coming on the podcast, Govind. Thanks. So our topic today is something you've written about extensively, issues pertaining to the allocation of health resources. In some of your work, you talk about scarce health resources. Can you say a bit about this notion of scarcity?
SPEAKER_00In terms of defining scarcity, I guess I would say you can think of scarcity as existing whenever one individual or patient's receipt of a medical resource has what economists would call opportunity costs or what like ordinary people might just say are negative effects to some extent on other people who are also looking for that same resource. So it's clear, right, for organs, if you give one person a kidney transplant, that means another person on the kidney transplant list can't get that kidney. They might be able to get a different kidney. For organs, the scarcity seems, at least in the near term, absolute, in the it's not easy to think of other resources we have around in society that we could immediately transform into organs. For other things, like for instance, you know, healthcare spending in different sectors, you might think of the scarcity as being a little more porous, where if there's not enough money to fund this many healthcare procedures, you might wonder about are there other sectors that we might move funding from, which raise questions about sort of trade-offs more generally. So I would say that perhaps scarcity is a special category of the more general species of trade-offs where it seems less possible to alleviate the trade-offs by sort of obtaining more resources in the near term from somewhere else.
SPEAKER_03I know that some people object to sort of this whole area of research because of the idea that it kind of takes for granted scarcity. So they think we should be demanding that more resources be devoted to things rather than thinking about how to allocate existing resources and treating the scarcity as fixed. Could you speak to that a bit?
SPEAKER_00So I've tended to see the project of trying to increase the available set of resources as being also an important project, but not in any kind of obvious tension with the project of trying to fairly allocate, you know, however many resources we've got. So take the example of organs. I think there are many ways in which different people have argued for improving the system of organ procurement. So you might use various incentives to try to promote organ donation. You might improve efficiencies in the supply chain to try to make sure organs don't arrive frozen and unusable or like arrive in a situation where they are no longer cold and suitable to transplant into somebody. Similarly, for actually a debate about this for the COVID-19 pandemic, people had said earlier, you know, maybe what we should do or should have done is have a lot more ventilators around. So it's not a problem with ventilator scarcity. I think it's important to have both discussions. Having discussions about increasing supply or procurement doesn't sort of eliminate the significance in the near term of allocating what we have, especially once you go beyond something like organs, but even for organs, there are often going to be trade-offs involved at that next level of trying to increase supply. So I think often in hospital settings, for instance, a scarcity that I didn't mention that was really the most common form of scarcity from talking to colleagues in medicine during the pandemic was scarcity of staffing time. And if you move more staffing time into, say, caring for COVID patients, um, that could be a good thing for that population of patients. But if that means delaying surgeries or even things like needed physicals and primary care checkups, you have trade-offs for other patients with other types of medical conditions. So I think often the apparently innocuous and I think often morally very sensible goal of increasing supply itself involves trade-offs because um increasing that supply itself draws on some semi-finite pool of resources. At the same time, there are certainly situations where sometimes there are pools of resources in society that are being allocated right now to like things that many people might agree are not sensible. So it's hard to say no to people in the face of trade-offs because you could say, well, we could just get money from having a more fair system of taxation. We're spending less money on certain military spending. And I think there's a lot of sense to that. But if you politically aren't actually going to be able to do those things in the near term, you still have to allocate. And then even if you got that money, you would still have a question about what you do with those funds next. And in the COVID pandemic, we're dealing with situations where no matter how much money you had, you couldn't convert that money immediately into certain of the scarce medical resources.
SPEAKER_02I mean, it's interesting that there's often this sort of adversarial relationship between people who are working on expanding the pot of resources and people who are working on allocating resources, because you might think that often this relationship could be symbiotic in that if you can demonstrate that you're using resources really well, people may be more inclined to fund you or give you more resources. And so it just seems sort of unfortunate that we've ended up in the situation where there's tension when it really feels like these groups could be working together.
SPEAKER_00Yeah, I mean, I think part of it may be that people come from different backgrounds and orientations toward these questions of allocation versus procurement. I do think it's a happy thought that it would motivate greater procurement to show that you are using resources well. I do think a sort of practical fear that people may have, I'm less qualified to speak about this talking from the allocation side, but that people may feel like if people focus a lot on allocation issues, that it may make the problems seem more solvable and less severe. And so it may reduce the impetus for pushing on procurement harder. I suppose I, as with other areas, I tend to think that we shouldn't sort of accelerate something bad in the near term to draw attention to the severity of a problem. But I could so something that I worked a little bit on during the pandemic was questions about should you, for instance, dose vaccines fractionally? So use a given quantum of vaccine material to give half or one-fifth doses, or spread out your vaccine supply by dosing at longer intervals if you have supply that's increasing. And I think some people might have pushed back against that and said, well, if we talk about fractional dosing, maybe that reduces the impetus for saying we have to ramp up vaccine production.
SPEAKER_03I see. So if I'm understanding you correctly, the question was whether we should take the standard vaccine doses that we had and either divvy them up into smaller doses so that each shot would contain less vaccine or space them out more by asking people to wait, say, months rather than weeks between doses. And if you did that, then you could potentially vaccinate more people more quickly, but in a suboptimal way. And in the meantime, you might make the public too satisfied with their access to vaccines, which would alleviate some necessary social pressure on the companies that manufacture vaccines. So let's transition now to your work in particular. Could you tell us a little bit about how you started working on issues related to the allocation of health resources and what initially motivated you to work on them?
SPEAKER_00The first paper I think came out in 2009 that I co-authored with Zeke Emanuel, who was at the time the chair of the NIH bioethics department and the late Alan Wertheimer, who is a colleague in that department as well. And this project was sort of developing some ideas that I had been interested in and that also built on a paper that Professor Emmanuel and Wertheimer had written in science in 2006 about allocating pandemic flu vaccines. So I think the impetus for us really was actually thinking about, in particular, pandemic flu, which was still sort of on the table, things like H1N1 or other pandemic scenarios at the time. But we're interested not just in guidance for a specific scarcity scenario, but whether we can learn from principles that had been raised in other contexts of scarcities, for instance, organs, or in the paper we talk about historical examples like scarcity of dialysis or pedicillin, even initially when a new technology was developed.
SPEAKER_02So it seems like that paper and some of the other work you've done has developed decision-making frameworks for allocating resources. So, for example, in some of your later work on COVID-19, there were frameworks developed about the allocation of vaccines. And often these frameworks would start with some high-level principles like maximize utility, and then would get a little lower level about how you actualize these principles in a more applied way, something like save the most lives. And then sometimes they get into the nitty-gritty of how to actually realize that principle. So, for example, prioritize healthcare workers or the people at highest risk of COVID. Before we get into the substance of the views endorsed in those papers, we'd like to step back and discuss the purpose of frameworks in general. So, can you say a little bit about how you go about developing those frameworks and how they're meant to be used?
SPEAKER_00So, yeah, I would say that the way that in writing I've developed these frameworks is really drawing on existing frameworks in the area and also thinking about how those frameworks connect with theories of fairness in distribution more broadly. So, from other areas of, say, social or political philosophy, where there are thinkers who have thought a lot about these questions of what principles we use to distribute other resources fairly. So you have people in broadly social contract traditions like John Rawls, Elizabeth Anderson, people like that. Then you have people in a sort of more consequentialist tradition, people like Derek Parfitt. And I think even if people who are creating, say, the pandemic flu policy for the WHO, which we looked at, aren't citing Rawls or Anderson or Parfit, some of the principles that those thinkers regarded as important also end up, we think, undergirding a lot of the ideas that are there. So I think probably influences come from both kind of policies developed by nation states or other sort of agencies like the WHO. And then also thinking about how those principles that we extract from those documents connect with broader ideas about fair distribution.
SPEAKER_03Got it. So it sounds like these allocation frameworks are roughly built on three literatures work by philosophers from the social contract tradition, by philosophers from the consequentialist tradition, and resource allocation policies that are developed by governments or multilateral organizations. Do you think the choice to develop a framework bakes in any normative assumptions? So you might think, for example, that it assumes some form of pluralism at the foundations of ethics. You mentioned that these frameworks build on ideas from both consequentialist and social contract traditions. And so you might think they therefore recognize multiple distinct moral values. And it seems like there's no single overarching theory or principle that's giving rise to the multiple values expressed in the framework, because if there were, then you might expect that overarching principle not just to explain why all the different values are being included, but also tell you how to weigh or balance them. And as far as I'm aware, proponents of frameworks don't usually tell you how to weigh or balance the different criteria that they think you should use to make decisions.
SPEAKER_00In terms of pluralism, I have tended in my work not to say, for instance, that one value should take decisive priority over the others. I have tended to say, no, they can be weighed and balanced, and that often how they're weighed and balanced is going to depend on facts about context. So I do think that there is some commitment to the idea that there are contextual factors or different sets of background normative values that could affect how one weighs the values. And so I do agree, actually, that in a sense, that is a normative choice. That I think my work with colleagues and my work individually has not generally tended, I think, to come from a place of saying this is a, for instance, social contractarian or utilitarian or what have you approached allocating medical resources. So why are you distributing things that way? There's a different approach, uh, which I am skeptical of, where people have often said, no, what you should do is just have doctors who are very wise decide who should get stuff. Or on the flip side, you could say, you know, this should just be sort of a purely political process decision. There shouldn't be these sort of independent constraining normative principles. And I so I think there is a commitment to the idea of there being these organizing values that also can be advanced sincerely as public justifications for why things are being done the way they are.
SPEAKER_02Yeah, that makes a lot of sense. I mean, another reason you might think that it makes sense to include a bunch of different principles in a framework doesn't have to do with pluralism so much as it might have to do with a view on moral uncertainty. And so one way you might interpret allocation frameworks is that they're recommending several distinct courses of action a government might take. Like hick prioritarianism in a case where the health benefits are pretty similar for two different populations, but vaccinating one population will greatly reduce health disparities. HIC utilitarianism in a case where both populations are similarly badly off, but one population stands to benefit much more from being prioritized and so on. So, in other words, on this view, frameworks aren't telling policymakers to blend utilitarianism and prioritarianism and the other ethical theories together. They're laying out options available to policymakers under moral uncertainty. And then the relevant ethical question would be how should they go about deciding which of these different ethical theories is the right one to act on in any given case, rather than how should we mush all these principles together all of the time?
SPEAKER_00Maybe I've always found moral uncertainty a challenging topic to know what to say about because there's some ethical theories that have included them moral claims that seem like they would have extremely strong weight compared to other claims. And depending on how you approach uncertainty, like there's a philosopher who I know who's writing a paper. The title of the paper is something like the rebugnant conclusion about, I believe, something having to do with the moral weight of claims of certain non-human animals like insects. And the thought is like there's a ton of insects, apparently, so I'm given to understand. And even if like I'm very uncertain about the moral weight of their claims, if those claims are very, very, very weighty claims because there's so many insects, it might turn out that even if I attach only like a 1% credence to their claims being ones that have a moral weight in, say deciding, you know, which theory to act on, the enormous magnitude of those claims might end up outweighing the very small credence I give to the theory. So I think one has to figure out if it's not something I've thought a lot about. Theories that can generate just extremely weighty claims that seem like they might outweigh a lot of other claims. What to do with those theories to keep them from just by being on the table, like eating all the other theories.
SPEAKER_02Yeah, no, my sense is that this is something that people who work on moral uncertainty are also worried about. Like, as you say, you might have to drop everything and devote a ton of resources to figuring out how to make bugs' lives go better because the stakes are so high because there's so many bugs, even though you're pretty sure that bugs do not matter morally.
SPEAKER_03Yeah. So why don't we turn away from bugs now to introduce the specific principles and processes that you think ought to govern allocation decisions? So in the paper that you mentioned earlier, the 2009 one called Principles for Allocation of Scarce Medical Interventions, you discussed four overarching principles of resource allocation. They were first maximizing total benefits, second, favoring the worst off, third, treating people equally, and fourth, promoting and rewarding social usefulness. And roughly the same four principles appear in your 2020 paper with Zeke Emanuel and others that's called Fair Allocation of Scarce Medical Resources in the Time of COVID-19. So I take it that you think the first two principles, maximizing total benefits and favoring the worst off, are the more important ones. Could you start by telling us about those and why you consider them to be so important?
SPEAKER_00Yeah. So, first of all, then in 09, we call, you know, maximizing benefits. In sort of more recent work, we've talked about trying to come up with a more pithy way of putting it that doesn't call it the only maximizing principle, but in some places I wanted to call it benefiting people and preventing harm, or in one draft, I think called it beneficence, but people felt like that didn't sort of capture the importance of the principle. I would say, like in practical contexts, we said in the 2020 paper that this principle is, by consensus, the most important one in pandemic contexts. And I would say that still often I would view it as the goal toward which allocation policies are aiming, as some of the other ones often function more as constraints. Although I think the next principle I'll talk about is the closest one to being one that I would say is genuinely also a goal. So as an example, like when we're trying to allocate organs, I think we often would say the point of allocating organs, the point is to keep people from dying of whatever it is they're going to die of if they didn't get these organs, so kidney failure or heart failure. Or if we're allocating COVID vaccines or antivirals, the point of these is to keep people from suffering certain kinds of harm. Like the way that we sort of make the case for a medical problem being a serious problem that people should take steps to address typically involves saying something about the importance of preventing harm. Part of why we thought that people should take all kinds of steps, for instance, at the start of the pandemic to mitigate COVID transmission, is that we thought it was really important to prevent serious harm. Or the case that's made to people at organ procurement for why they should donate organs is often supposed to be about, you know, you should donate your organs when you pass away because it could help to keep somebody from dying of kidney failure. So the next principle is a principle of prioritizing among those who might be protected against harm, people who are more disadvantaged in some way. So in philosophy, I think people often talk about this as a principle of priority to the least advantaged. The same thinkers I mentioned before, Rawls, for instance, or Parfit, give different weight to the importance of prioritizing those who are least advantaged. And there are different interpretations you can give to this question of who's least advantaged. So is it people who are likely to fear the worst if they don't get the scarce resource? Is it people who have been treated unjustly in some ways in the past? Is it people who just look like they're in a very bad situation right now? In my work, I've generally argued for a view something like people who look like they're going to be worst off if they don't get the resource. And in thinking about what it means to be worst off, looking at how badly their life will have gone if they end up getting very sick or dying for want of the scarce resource. I think this principle is also very important. Rawls himself did not think that in the case of allocating organs, this principle should have priority over maximizing benefits or over-benefiting people and preventing harm. But Rawls did say about organizing social structures that we should give what's called lexical priority. We should regard it as the most important thing to improve the situation of the worst-off group and how we organize social structures. And then once you do that, then you could figure out what to do in terms of improving benefit across society broadly. That makes the calculation easier. But I think unfortunately, and I think Rawls actually agreed, lexical priority is not a plausible way of resolving trade-offs and allocating a discrete body of resources, like, say, organs or vaccines. And so I think some weight would attach to both the idea of preventing more harm overall and in particular preventing harm to those who are less advantaged or have been subject to unfair disadvantage. But I don't in my work have a precise view about how that's to be done. My view is more like something that I think Parfitt said in Equality and Priority about giving priority to the worst off, which is that we should give some priority to people who are worse off. But in terms of deciding what degree of priority is sufficient, Parfitt said something like, in the end, we must simply use our judgment. And unfortunately, despite all the other very good work that Parfitt did, I don't know that he had a lot more to say about in what ways we should use our judgment to resolve that specific trade-off.
SPEAKER_02Yeah. And to just say a bit more about lexical priority, the idea is that if one principle has lexical priority over the other, then the less important principle doesn't even come into play until the more important principle is fully satisfied. So Rawls thought that when it comes to establishing the basic structure of society, benefiting the least advantaged is lexically prior to benefiting people more broadly. So an example might be that if you had to choose between giving one dollar to someone who is badly off or giving $1,000 to someone who is slightly less badly off, he'd say, give the $1 to the more badly off person. The importance of benefiting the second person doesn't even come into play until you benefited the least advantaged as much as you can. But as you note, this would have pretty weird implications in the case of most scarce medical resources. Like, do you give the organ to the slightly more disadvantaged person who will live one year with that organ, or to the slightly less disadvantaged person who could live 20 years? Seems like you should do the latter. Giving priority to the worst off should not always be prioritized over providing bigger benefits.
SPEAKER_03Right. Okay, so just to complicate things a little bit further, let's add the other two principles that you discuss into the mix. The other two principles are what you, at least in some of your work, called equal treatment and rewarding social usefulness. Can you say a bit about those and why they maybe don't seem as important to you as the first two?
SPEAKER_00Equal treatment is a principle I've become increasingly skeptical about. When people say that they care about fairness or justice and they think that that might conflict sometimes with preventing as much harm as we can, I've become increasingly persuaded that the way of doing that is to focus on mitigating unfair disadvantage or giving priority to those who are unfairly disadvantaged, and that just formally equal treatment, as opposed to what we in later work call equal concern, is not a principle that has particularly strong moral weight. So I'm just writing something that talks about this argument, which actually you find interestingly in some religious texts, but says things like to save one life is as if to save a thousand. And sounds very poetic. It sounds a lot nicer than most like work in bioethics. But taken literally, I don't think people, including probably the people who develop these very meaningful religious texts, literally meant that if you have to pick between saving a thousand people or saving one, you should regard them as equivalent. And that's what some views of equal treatment have. This is a famous paper in philosophy that almost everybody agrees argues for a conclusion that's wrong. They don't agree about why it's wrong, which is that this philosopher in the 1970s said, if you have, you know, a thousand people on one island and one person on the other island, you should just slip a coin as to where to go. And people have generally thought that's not required. They're not sure what is the best explanation of why it's not required, whether it has to do with fairness or just preventing more harm. But more broadly, it just seems like the requirements of fairness are more about treating people in light of relevant factors, as opposed to saying we should treat them identically, even if there are relevant differences in how much intervention you can protect them from harm or how disadvantaged they are. So I think that understanding it more as a requirement of equal concern that is treating people only in light of relevant differences makes more sense. There is one issue that I still wrestled with a little bit in the 2009 paper, which is that I had been thinking at that time about why people do give some weight to wanting to incorporate some kind of equal treatment requirement. So one argument for, for instance, doing a weighted lottery rather than just prioritizing people based on, for instance, how much we can protect them against harm and how disadvantaged they'd be without it, has been that there's some value in giving everybody some chance or spreading chances out more broadly among people. And I still have more work to do in figuring out how much weight that factor should get. I mean, if some situations maybe we think randomness is sort of fun. So, like in the NBA draft or something, if, you know, there was this year where the Warriors were bad one year and so they got the second pick in the lottery, and that seems fine to make basketball more exciting. But I've become less and less persuaded that there is some sort of justice value being served by just adding extra randomness to our algorithm. I think whatever the values we think are important, those should go into the framework. Importantly, actually, I feel like we didn't see people saying we should allocate the COVID vaccines to people using lotteries. People like to say this a lot about the ventilators and could talk about why people thought this, but in practice, we didn't actually see lotteries used. It turns out the public hate lotteries when they're used in those kinds of allocation processes.
SPEAKER_02Yeah. So the lottery idea would be rather than prioritizing all of the highest risk people for vaccination, we would say that everyone deserves some shot at getting a vaccine. So we do a lottery. But then no one actually thinks that you should literally allocate vaccines totally randomly, since that would result in a bunch of people like me getting vaccinated over my 90-year-old grandma. So instead, what people have suggested is something like a weighted lottery where my grandma gets, say, 15 lottery tickets since she's higher risk, and I get one lottery ticket since I'm lower risk, she's still more likely to win, but I get to have a chance. And I think what you're saying, which I would agree with, is that it's not clear why adding this chanciness to the allocation scheme is better than just saying, we're gonna prioritize my grandma since she's higher risk. Like it's unclear what the value add is of giving me a shot at winning the lottery if she needs the vaccine more than I do. And giving lower risk people a shot of getting prioritized will predictably lead to some older people getting deprioritized and correspondingly dying.
SPEAKER_03Can we return a minute to that paper that you mentioned, the one that advocates flipping a coin to decide whether to save one person or a thousand people? I assume you're talking about John Torricks. Should the numbers count?
SPEAKER_00Yes.
SPEAKER_03So in that paper, he does argue that the numbers shouldn't count, at least when you're allocating a privately owned resource, like it's your boat in your time, and you can decide whether to go save the one person stranded on the island or the 1,000 people stranded on the other island. Anyway, one reason that he thinks numbers don't count is that he thinks it's impossible to aggregate harms or benefits across people. And his argument there is based on the metaphysics of pain. So what kind of thing pain is fundamentally. His thought is that pain is like boxing skill, or physical beauty, I think he says. If you have one amazing professional boxer on this side of the room and you have 250 mediocre boxers over there, you can compare the boxing skill of the one professional to the boxing skill of each of the 250 amateurs. You can say the pro has more boxing skill than any one of the amateurs has. But it just doesn't make any sense, he thinks, to say, oh, but when taken together, the 250 amateurs have a lot more boxing skill than the pro has. Like that's just not how boxing skill works. That's how weight works. So it's true that the 250 have a bigger combined weight, but Torek thinks pain isn't like weight. It's more like boxing skill. So there's no such thing as the combined pain of multiple people. In philosophy, this idea about pain is, I take it, pretty well regarded. And I wanted to ask you: do most or all of the bioethicists working on research allocation accept that you can add together or aggregate benefits to different individuals?
SPEAKER_00There's certainly people who work in this area who don't think that, but I would say many people working in bioethics are not people who would have, for instance, had any reason to read that paper by Taurek or think about that degree of sort of abstract detail. So I'm not sure they would have a view about can you aggregate benefits or not. But I think a lot of people who think that there maybe isn't some like sum of pain across two people would still think something like the following, which is that if two people have a very weighty claim to get something, and then one person also has a weighty claim, we can kind of balance out the two. Say that you could take vaccines, for instance, to a place where they're likely to be more recipients and they can be used up faster, or a place where you're gonna be waiting a lot longer. You can weigh the claim of one of the people in the place where you can get the vaccines distributed more expeditiously against the claim of one person in the other place. And then those claims maybe are kind of on a par. And then if you look at the claims of the other people who are in the place who could get the vaccines as well, add on the second person in place number two, and that claim, as it were, kind of breaks the tie. An example in organ transplantation that I've found very compelling in this area is about transplants called multiple organ transplants, where sometimes you could transplant more than one organ into one person who would die fairly quickly if they didn't get both the organs, or you can sometimes split those and do them separately. Or it may be that the person needs two organs that are separate and those organs could have instead individually been given to people who were at risk of dying if they didn't get the one organ. And there, I think, even if one doesn't accept some sort of aggregation, there's a very strong case against giving two organs to one person, then instead it could have saved two people. And it doesn't depend on the two people having a super claim. It's just that, say, all three people each have the same claim, but you can fulfill one person's claim and then also another's, as opposed to only one person's claim. People have responded to this, I think, initially by saying something which seems completely wrong to me, which is to say that the person who can only survive if they get both the organs is somehow worse off or more unfairly disadvantaged. That just seems wrong to me in that someone is always going to fare less well than others where resources are scarce. But just the fact that one would require, say, a heart and a lung or something to survive, as opposed to you just need a heart, but if you don't get the heart, you'll die. You just get a lung, but if you don't get the lung, you'll die. Those people, I would say, seem like they're all in a very parallel position.
SPEAKER_03I see. That's helpful. So the idea is suppose you don't believe in aggregating benefits across people. So you don't think that the two people who need organs would, in aggregate, benefit more from getting them than the one person would benefit. And suppose you also don't believe in aggregating claims. So you don't think that the claims that the two people have to the organs somehow sum together into some super strong claim that's stronger than the one person's claim. Still, you can coherently think that we should give the organs to two people rather than giving them both to one person. And that's just because there are more claims on that side of the equation. So if we treat everybody's claim the same and give them equal weight, then the claim held by the one person cancels out with one of the claims held by the two people, and you're left in a situation where by allocating to the two, you can meet sort of one net claim. So you should do that. Anyway, you were saying a moment ago that some people think the person who needs two organs, say a heart and a lung, is somehow worse off or more unfairly disadvantaged than the two people who each need one organ, and that this would tell in favor of prioritizing the one. Do you think that what people are doing there is thinking about how unlikely it is to get any given organ? And so they're sort of multiplying the unlikelihood of getting a heart by the unlikelihood of getting something like a lung and thinking being less likely to get some resource you need itself makes you worse off.
SPEAKER_00It seems challenging to try to define being worst off in terms of unlikelihood of receiving the scarce resource, as opposed to how badly off you would be if you don't get it, irrespective of how likely you are to get it, as it were. But yeah, I think it's a good question about how exactly people would define that. I would tend to define their being worst off in terms of looking at how badly off they will have been overall if they don't get whatever it is that they would require in order to say continue being alive or in order to avoid having some sort of a health problem. We've talked about three principles so far, the preventing harm and benefiting people, principle one, mitigating disadvantage or prioritizing people who are unfairly disadvantaged, principle two. And then third, a principle which I had initially talked about equal treatment, but I think should be more understood as equal concern. The fourth value that I do think is relevant, but I think we've seen it talked about differently in different reports. One is a value that I put in the fourth category in the 2009 paper, but now that I actually think of really as more of a species of either benefiting people and limiting harm or potentially also mitigating disadvantage is what we call instrumental value in 2009, or what I would call indirect benefit, which is just sometimes you could prevent more overall harm or mitigate more disadvantage, not by directly helping the person who's the most disadvantaged or who we can best protect, but by helping people who, for instance, could go on to help others. So in the COVID pandemic, medical professionals were talked about in that way, or caregivers might be another example. So that's one other principle that I think is really more of a meta principle that says that sometimes we can indirectly promote the values that we regard as important, which I think, again, the two weightiest to me are preventing harm and prioritizing people who are more disadvantaged. The other value, which I think is genuinely a different value, is a backward-looking one of reciprocity, where we say, you know, to what extent should we take into consideration people's contributing to solving a health problem when we decide who gets resources? So an example in the COVID pandemic has been: should you have prioritized people who participated in vaccine clinical trials to get scarce vaccines? In organs, it's should you prioritize people who previously donated an organ or who have pledged to be organ donors in receipt? And I think reciprocity is a principle that does have genuine moral weight, and you do see a lot of reports referencing it. I tend to think that its weight should be subordinate to the first two principles that I mentioned. So it often ends up getting used as a principle that breaks ties. I think often tiebreakers proliferate in frameworks in ways that are not helpful. But I think that reciprocity has important weight, but is often less important than the forward-looking principles of benefiting people and preventing harm and mitigating disadvantage.
SPEAKER_02So now that you've given us a bit of an overview of these four principles, let's get a bit more into how they're actually realized. You've said, and it seems like people agree, that the maximizing total benefits or increasing benefits and reducing harms principle is very widely accepted. But it seems like the obvious interpretive question here is how you define benefits. In your work, you've identified two possible interpretations. One is the save the most lives principle, and the other is the prognosis or life years saved principle. Just to define these quickly, saving the most lives is pretty intuitive. You just try to keep people from dying regardless of how old they are. So an 80-year-old and an eight-year-old would count the same. A principle that prioritizes saving the most life years, though, would strongly prioritize saving the eight-year-old since, in expectation, they're going to live a lot longer than the 80-year-old. What are the pros and cons of each of these principles? And why do you think that both should be included in an allocation system?
SPEAKER_00The principles you mentioned are sort of raised, particularly in the context of interventions where a person is likely to die without receiving the resource. In other situations where, for instance, you're allocating pain medication or allocating, say, scarce resources for procedure like in vitro fertilization, you could have other benefits you're trying to realize or harms you're trying to prevent that aren't in sort of the register of preventing loss of life. So I think how you interpret what kinds of harms you are mitigating or preventing is often, you know, also sensitive to context. In our global vaccine allocation paper, we also talked about impacts in terms of, for instance, poverty or other things. But I'll focus on the ones that you were asking about, which are lives lost and life years. What we said in the 2020 paper still seems right to me as a first pass, but I think the empirical data have made me doubtful about a piece of it, is that in practice, it's often going to just gonna be easier to figure out are there gonna be some people that are likely to die right away if they don't get this, or they're likely to die pretty soon, even if they get this. Those seem easier and more tractable things to determine than to try to assess the sort of question of how long would somebody be likely to live if they get this resource, or how long will they be likely to live without the resource? Length, I think, can require more complexity in calculating and figuring out. So it is really more of an argument that is epistemic, it's about whether we can know how likely somebody is to be protected against harm if they get the resource. Some people have taken a further step that I've become increasingly unconvinced by, which is to say that there's something fairer about just looking at whether people would die at some point if they don't get this, as opposed to how much lost life you're averting. The problem with this is that the actual pandemic, as it turned out, people who thought of this as fairer thought that people who were more disadvantaged beforehand or more disadvantaged overall would sort of lose less life from the harms of the pandemic because they were facing so many other threats anyway. So I think their thought was that if we just focus on are they going to lose some life, this is gonna be fairer somehow than to look at how much. As it turned out, at least in the COVID pandemic, and I think in other scenarios, you actually saw that basically because people who were unfairly disadvantaged tended to get very sick earlier in life. This is particularly true in the US by racism and ethnicity, there were actually more future years of life lost per 1,000 African Americans or Latino Americans than per 1,000 white Americans. And so initially the pandemic, there were some folks who said we should ignore years of life lost because it will magnify the life-shortening injustices that people have faced because of poverty or ableism or racism. It turns out, I think, that actually ignoring that, if it had been put into practice, that actually would have magnified at least some dimensions of injustice. So I've tended to think that the right way is not just to focus on one dimension of preventing harm, but look broadly. And then if you're worried that when you look broadly at preventing harm, it's going to exacerbate disadvantage, then try to figure out. What framework would best balance the values of mitigating disadvantage and preventing harm, as opposed to the alternative of saying we're going to ignore some kinds of harm because we're worried that considering them would exacerbate disadvantage. If disadvantage is to be mitigated in a framework, that should be done through intentionally considering the way in which some harms are correlated with disadvantage and maybe attaching additional weight to preventing harms that are disadvantage-correlated rather than saying we should ignore some harms.
SPEAKER_03Okay, yeah, that seems reasonable. So in your work, you've explicitly disavowed using quality adjusted life years, qualities or disability adjusted life years, dahlies, which are metrics that account not just for the number of life years saved, but also for the quality of those life years. So for instance, if two cancer treatments extend life by a year, but one provides a year in good health and the other provides a year in poor health, considering just life years saved wouldn't account for the difference while considering dahlies or quallies would lead you to favor the treatment that improves both quality and length of life. Do you think that there is an appropriate role for the use of qualies and dallies in resource allocation?
SPEAKER_00I think it depends on what kind of health resource allocation and how exactly the quality metrics are being used. So we were very critical of the use of quallies, and I don't think that they were ever actually used in allocating scarce pandemic interventions, because I think the problem of calculating life expectancy is magnified for if you're trying to also do some sort of quality calculation. So I think it's just not very tractable to try to do some sort of broad quality of life calculation. There's also an ethical problem that exists that I think you're raising that goes beyond this question of just it's going to be very hard to calculate them, which is that there's a worry that if you're looking at sort of future quality of life that takes into account if somebody had, say, some chronic condition that pre-existed their need for, say, an organ, that then you would attach less importance to providing them the organ because they would still have whatever condition was leading to lower quality of life before, say if they had some type of condition that caused persistent, severe chronic pain before they became eligible for an organ, you might worry that allocating organs based on quality-adjusted life years would exacerbate that pre-existing disadvantage. I think that's a very legitimate concern. And there are metrics that recently have been developed by health economists that I think do a good job of bringing in quality differently. That I would say instead of looking at absolute quality of life afterward, look at what the procedure's effect would be incrementally on the person's quality of life as well as how much it would extend life. So, what that approach would say is, for instance, give my example of someone who had a severe chronic pain and was in line to receive an organ, look at how much the organ would extend their life. Then also look at not what their absolute quality of life would be afterward, but whether they would experience some increment or decrement in quality of life per year after receiving the organ or whatever the intervention is. And the idea there is that you're not disadvantaging them further on the basis of their pre-existing baseline quality of life. Instead, you're looking basically at the treatment's efficacy at extending life and what its implications are for quality of life. I think that there actually would be broad consensus that quality of life is important to consider in certain respects. So for instance, for COVID antivirals, or certainly for organ transplantation, we often say a reason against giving somebody an organ, even if it could work, or reason not to give certain antivirals to people is if they're likely to suffer severe side effects or organ rejection if they get this. And I think a lot of the concerns raised about quality of life metrics, disadvantaging people who were unfairly disadvantaged before, don't really seem to apply to the idea that we should regard it as important, not just will this organ help somebody get out of the hospital, but are they likely to end up rejecting the organ in a way that'll be very painful or bad for them? So I think that there can be a legitimate place for considering quality of life, but it depends both on whether the setting is one where you can make those types of calculations feasibly, which I think is more feasible for something like an organ waiting list than it is for something like trauma casualty situation, like an overwhelmed hospital. And then can you incorporate quality of life in a way that doesn't exacerbate pre-existing disadvantage? There are further puzzles about whose judgments about quality of life count. Do you look, for instance, at the judgments of people who have a given condition? Do you look at people who have had that condition recovered from it? Do you look at the general public? Do you look at revealed preferences or some other sort of assessment? But I do think that there's a legitimate place for it in making decisions about allocation. But I think the quality in particular has the challenge of just being what I call in some work in progress, sort of a first generation approach to doing this, that I think there are approaches that can do better at integrating some of the concerns for fairness that I think people have raised about.
SPEAKER_02Yeah, I mean, it's interesting because these two ideas almost seem to be intention. Like if we think that one of the problems with using something like life years saved is that it's complicated. Well, Dalli's is more complicated than that. And then obviously these more complex metrics are even more complex. And so it does seem that like the theoretical pull and the practical pull are kind of leading us to different metrics here, perhaps. But this also segues nicely into our discussion about prioritarianism or favoring the worst off, which is the second principle that you discussed. In some of your work, you've talked about this in terms of prioritizing youngest first or sickest first. So starting with the former, you might think that the young are disadvantaged and that they haven't had the opportunity to live for as long, or as some people have talked about this, they haven't gotten their fair innings. But you might also not think the young are disadvantaged for lots of other reasons, or at least that other groups intuitively face more disadvantage. For example, like why focus on the young rather than the poor or people who are marginalized in other ways.
SPEAKER_00Yeah. So I think in the context in particular of the COVID pandemic, but also more broadly, I think it also would apply for many other illnesses. What we're really going to be talking about is not just sort of young people writ large as a group, but people who get so sick earlier in life that they are in a situation where if they don't receive a scarce resource, they'll suffer very severe health consequences. So that subset of people is really who we're concerned with rather than younger people more broadly. So I guess what I would say is that it does look like, certainly in COVID and in general, like people who have this sort of very urgent medical need earlier in their life tend to be disadvantaged along some of these other dimensions as well. So in practice, it's turned out that, you know, if you look at, for instance, who are the people in America who were in need of ECMO or something like that because they got really sick from COVID when they were in their 30s or 40s, it was generally people who had experienced some form of social disadvantage or had some very serious preexisting medical conditions.
SPEAKER_02Yeah, that makes sense. So if I understand you right, you're saying that basically it's sort of this intersectional consideration of both sickest and youngest. And you said, like, you know, in the case of COVID, it just so happens that a lot of the people who were young and who were getting sick were socially disadvantaged in various ways. But if you're just using youngest as a proxy for something like social disadvantage, why not just say social disadvantage? Or is this something like a political choice where you're worried that people are not going to like that kind of principle? And so saying youngest first is a way of kind of getting around that.
SPEAKER_00No, I mean, I I think it is a very good question. One way of I think understanding what I was saying is that sickest first is a hard principle to apply in allocating many scarce resources because generally, for instance, when antivirals were scarce or when people need organs, it's not like people show up wanting organs who are not very sick and they're just like, you know, my organs are maybe not as the 100% best they could be. Maybe I should have a new one. It tends to be that everybody who's in line to get an organ is actually quite sick. So it's not clear that just being sickest at this time is going to do much helpful work in differentiating among people who are claimants now. And I think being at risk of dying early tends to be a pretty good proxy for having a life that has gone quite poorly overall. But I do agree also with the point that you raise about politics. Although in practice, one unfortunate fact is that I think people who are likely to die very early in their lives often are like not very politically powerful groups. But I do think an advantage of it is that it is a more tractable metric than if you were trying to say, you know, assess in medical care or in organ allocation or something like that, how poor or marginalized somebody has been over their whole life history. That would be very hard to assess. Doctors aren't well placed to do it. Making these judgments might require sort of intrusive assessments. Whereas if you look at, you know, is this person someone who's coming in very ill much earlier in their life than someone who typically would show up at this problem? That's at least one good metric for looking at that. I wouldn't say it should be the only factor, but it is one factor that I think um could help to capture multidimensional disadvantage readily.
SPEAKER_02Yeah, that makes sense. So you've talked about equal treatment uh a fair bit. So we're gonna skip that principle and just move on to the last principle, which is this promoting and rewarding social usefulness, or as you later call it, promoting and rewarding instrumental value. And before you distinguish between this prospective principle, which is like we should prioritize those likely to make relevant contributions in the future, and you sort of said this might have more to do with the increasing benefits and reducing harms principle. But then there's this backward-looking retrospective principle, which says we should prioritize those who have recently made relevant contributions. So, for example, participants in COVID vaccine trials. And this seems to be pretty controversial among bioethicists. In fact, I think a lot of people think that this is like pretty abhorrent to look at people's past medical decisions and say, actually, we should consider that in deciding whether they should be eligible for the allocation of future health resources. But my sense is that you're actually a little bit more sympathetic to this backward-looking principle. Can you say a few words about that?
SPEAKER_00I actually think for what it's worth among bioethicists, I don't think there was a lot of controversy about the idea that the clinical trial participants should get some degree of priority. I think that there was more controversy about the idea that people who, for instance, refused to be vaccinated when they medically could have been vaccinated, that that should affect their access to resources. There was a lot of controversy about that. Here's an example that I've defended in an op-ed in the Washington Post where I think it's at least, I think, in pretty strong tension with the things that some bioethicists have said, which is there was a period of time when there was a fairly effective monoclonal antibody infusion for people who had immunocompromise for COVID. The antibody was called AVUSHELD, was its trade name. Sadly, it doesn't really work anymore because the omicon strain evades this antibody. But when this was effective and scarce, it was prioritized for people who either were immunocompromised or who couldn't be vaccinated because they had allergies or other medical contraindications. Those people were prioritized over people who were unvaccinated but could have been vaccinated. They just didn't elect to be vaccinated. And it's not that there was some medical difference between the person who is allergic or the person who is immunocompromised and the person who chose not to be. The difference, I think, at least tacitly, I don't think it was said explicitly in the FDA authorization, is that the folks who are immunocompromised or allergic didn't have the choice to protect themselves with a resource that wasn't scarce at the time. Vaccines weren't scarce in the US anymore. By taking EVUSHALD, they are using a scarce resource that they could have through a pretty easy to assess and pretty easy to make choice, avoided being in need of. So in the post article, we had an analogy that I think is not perfect, but we said something like, you know, this is a much less weighty example than production against COVID for a person with severe immunocompromise. Say you have a picnic and you have some different sandwiches, and some sandwiches are on regular bread and some sandwiches are on gluten-free bread. It would be sort of a jerk move. Um jerk move is not a sophisticated enough term, but to be like, well, I just prefer the way the gluten-free sandwiches taste. So I'll take the gluten-free sandwich, even though I know that there are people with celiac disease that are going to come to this picnic. And in the same way, it seemed to Emily Largers, my co-author on this and I, to be something that should be relevant to whether you get the scarous resource, whether through a sort of readily accessible choice, you could have substantially mitigated your odds of needing the scarce resource. And obviously, people worry about all sorts of slippery slopes, which you could raise for that and also for the, you know, vaccine clinical trials. So you could say if you're going to give reciprocity for the people who help to mitigate the harms of the pandemic by being in clinical trials, should you give extra priority to people who were like really diligent about wearing their masks or people who had their birthday party outside instead of inside? And what I would say is the difference between that sort of thing and things like whether somebody was vaccinated or whether they were in a clinical trial is that those latter two things are just much more readily accessible and identifiable than some of the things that people sort of worry about on the slippery slope, like tracking where people have their birthday party or how much they wear their mask.
SPEAKER_03So we'd love to transition now to talking a little bit about how to put these principles into practice. We've covered the allocation frameworks that bioethicists develop to help people make allocation decisions. And I guess we're wondering whether the groups that actually make decisions about resource allocation are likely to use or pay attention to these frameworks.
SPEAKER_00So in practice, I think you certainly did see a lot of influence of bodies on which SSIS participated in the formal rules of allocating vaccines and antivirals, more so vaccines, I thought. So the National Academy of the Sciences developed a framework that had some influence on the way that states and the advisory committee for immunization practices made recommendations for vaccine allocations, although every US state had its own setup and every national jurisdiction also did. So certainly for that, and similarly for organ allocation, you have there's actually some overhaul of this process going on around right now, but you have organ allocation scores or frameworks for each specific organ that are informed in part by ethical decision making. A big issue that we haven't talked about and that I haven't really worked on is what people called contingency care during COVID, which was instead of there being situations where really you had one ventilator and two patients, which didn't really, that wasn't really the scarcity. It was to what extent do we delay a whole universe of other non-COVID procedures? To what extent do we have fewer nurses spending time with patients? Or to what extent do we rely on traveling nurses? I think that there has been less development of sort of multi-dimensional or across different dimensions of need or condition allocation frameworks. A lot of that decision making was not made with the benefit of having as much of a formalized arrangement. And a worry I have about that, this is not to say the biothesist would have gotten the answers right, but that making these decisions ad hoc runs into the problem that typically in a medical context, physicians are supposed to be looking out for their particular patients as opposed to the interests of patients more generally. And it puts them, I think, in a very challenging position to be asked to make what are called sort of bedside allocation decisions as opposed to having some framework on which to rely. So I would say that there has been uptake, but it's often been more for what seemed like discrete challenges than for sort of more overall questions of allocation and priority setting.
SPEAKER_02Yeah, that makes a lot of sense. And I think, yeah, we want to talk about those practical issues more because as you suggested, it's not like when you're a doctor, you have five patients who need ECMO show up on your floor at the exact same time and you get to like have this nice framework and you apply all the principles and make a reasoned decision. There's all this stuff floating around. Like you have empirical uncertainty and then considerations of potential unintended consequences. When you're talking on a more policy level, there's questions about political feasibility and so on. To what extent did these sort of practical considerations enter the conversation when you and other bioethicists are developing these frameworks?
SPEAKER_00Yeah, so what this was very valuable to me is talking to people from either other academic disciplines or people who are in clinical practice. So we did some work on developing with colleagues frameworks for allocating monoclonals during COVID. One thing that came, certainly came up, was the question of how understandable are the sort of inputs to this framework going to be and how possible is it going to be to sort of calculate or work these things out in real time. Um, something that I thought was really encouraging is I I worked with some economists in Boston who had developed what are called reserve systems or categorized priority systems. The idea of these is that you can say that some subset of your available pool of scarce resources are prioritized for a population that, on the basis of someone of these principles, we think is an appropriate priority population. So an example would be you might prioritize some subset of those for healthcare personnel, or you could prioritize some set, this is actually used in Massachusetts for antivirals, prioritize some set for people coming from zip codes or census tracts within Boston that had experienced or were experiencing a particularly high COVID burden or that were disadvantaged along some scale of multidimensional disadvantage. So you can build in these things using the categorized priority system without saying something like, we're going to attach a certain number of points to being from this zip code, which might mean that they always get priority or might be so insignificant as to mean that it doesn't ever end up mattering. In practice, something that we saw a lot in these policies is that everything that didn't seem like it was the most important consideration, like preventing more overall harm, got thrown into being tiebreakers. So I think the categorized practices gives a way of incorporating these values without calling them all tiebreakers.
SPEAKER_03One kind of practical consideration that I know you pay attention to is legal considerations. So are there any principles you advocate that you think are likely to face legal resistance when put into practice?
SPEAKER_00Yeah. So I think obviously this is going to depend on the law of the jurisdiction. In the US, that's the law of each state, as well as federal law. And certainly in other jurisdictions, the laws may be different. I think that in practice, people raised legal objections to all kinds of principles, but I tend to think that in terms of what is actually case law on the books, many of these objections would not have been viewed by courts as being decisive. So, for instance, there got to be a view that you shouldn't consider how comparatively likely two patients were to benefit beyond just will they get them out of the hospital? Or how much time will somebody require, say, on ECMO in allocation. And the worry, this was not really ever clearly stated what the legal problem was supposed to be, but some suggested that there was some tension between this and some provisions of the Americans with Disabilities Act. HHS considered issues related to this when they evaluated many years ago a plan in Oregon for setting sort of broader health system priorities. What they said is that you couldn't, for instance, say that people who had liver cirrhosis as a result of alcoholism were categorically deprioritized on that ground alone. But you could consider whether one type of cirrhosis as opposed to another was less likely to be responsive to treatment. Something else that they say is explicitly okay, not prohibited, is cost, you can consider that, and the length of hospital stays. So, my view, at least, this hasn't been tested in court, is that even though you have agencies taking sometimes positions, so you had this agency during the Trump administration, the head of it said, the civil rights laws of the U.S. bar ruthless utilitarianism. And the civil rights laws of the U.S. are not taking a position between different normative theories. They do say things like certain policies that might unfairly disadvantage people with disabilities are not allowed. But I think that sometimes agencies have been aggressive in wanting to exert power in these spaces beyond what I think the law actually enforces. So there was language, for instance, I think also from agencies saying that you couldn't use age in some of these calculations. And age is not a protected category under the Constitution, such that a law that differentiates by age receives higher scrutiny. And we see this all the time with, you know, recommendations for screening tests, for various other procedures in medicine. One area that I have thought about this a lot more recently is about the question of racial disparities. For antivirals, there got to be this interesting movement where some hospitals wanted to explicitly use individual patients' racial identity as one priority point factor in allocating resources. Because race, unlike age or disability, does receive this highest level of constitutional scrutiny, I do think that there would be in practice problems with policies like the ones that we saw in some hospitals that said if someone is a person of color, they receive one more point. There are really good reasons to want to mitigate disadvantages having to do with structural racism, but having a point system that lumps all people of color into one bucket as a person of color, I think it just doesn't track with the heterogeneity and disadvantages. Manage both within groups and across groups, and it would be unlikely to pass the Supreme Court's strict scrutiny test, where if you use an individual person's race to decide whether they get something, you have to show that it's narrowly tailored, meaning it's basically the only practicable way to achieve a very important objective. So I think in practice a better alternative would be something that I've been working on more recently, would be instead of using individual race, self-reported identity as the criterion, to look at metrics of multidimensional disadvantage, look at what census tract are they coming to the hospital from? Is it one that is disproportionately then burdened? And that I think both has the virtue of being legally on much sounder ground and also getting more effectively at the actual drivers of disadvantage. People often say, which I think is just not correct, that you know, race is the risk factor. It's not race that's the risk factor, it's being subject to effects of racism that drove these disparities. And so you can pick out the people who, whether or not they self-identify in a certain way, are the ones who have disproportionately been subject to those. At the same time, that kind of allocation can sometimes run up against the question of how do we most prevent harm. So for instance, for vaccines, there could be situations where the populations that have been most burdened is going to be harder to allocate as expeditiously in those places. So you can have genuine trade-offs and tensions in those principles.
SPEAKER_02That makes sense. So we've been talking a bit about the legal system in the US and practical issues around the implementation of these allocation schemes. But let's delve into that a bit more because notably, most people don't look at the US healthcare system as a paragon of good resource allocation. Our healthcare system uses vastly more resources than other healthcare systems in high-income countries and at the same time performs poorly compared to other countries on most axes, including access, equity, efficiency, and health outcomes, i.e., most of the things that you would care about. So why do you think there's such a disconnect between the allocation principles that US bioethicists endorse and what actually happens on the ground? And are we doomed to have bad resource allocation within the US unless we rebuild our healthcare system from the ground up? Or are there tweaks that could be made that might substantially improve things?
SPEAKER_00So I think I tend to be more incremental in terms of like ways in which the US system could be sort of improved along specific axes. One thing that I've been working on a lot recently has been thinking about pharmaceutical pricing. And in some ways, this seems to be a more tractable area than trying to apply some of these same ideas to like medical procedures, because I think maybe doing it in with pharmaceutical pricing seems like it may impinge less on norms of professional authority for physicians and aspects of the physician-patient relationship in the US. Pretty much every other country in making decisions about reimbursing from public insurance for drugs or even setting launch pricing for drugs looks at what benefit the drugs are going to deliver. And most other countries pay, as a consequence of this, much lower prices for drugs. I think that there would be a lot of potential gains in trying to build in some more metrics of value for pharmaceuticals at least to start with. It would have two good consequences. One is that it would align reimbursement more with benefit and it would also create more incentives, better incentives for pharmaceutical companies. There are various areas. I think vaccines are one prominent area where the consensus among people in health policy is that these areas are under incentivized because of our current reimbursement regimes. So I think that's one area where you could see progress. That's just sort of one thought.
SPEAKER_03Yeah, I mean, you might worry that the U.S. is actually heading in the wrong direction in this regard. For example, five Republican Congress people recently introduced a bill that aims to prohibit all federal healthcare programs and federally funded state healthcare programs, for example, Medicaid, from using prices that are based in quality adjusted life years to determine relevant thresholds for coverage, reimbursements, incentive programs. And our understanding is that this would also apply to drug pricing. Now, it's not at all clear that this bill will pass, but what does the introduction of that sort of bill suggest to you about the ongoing issues pertaining to resource allocation in in the US?
SPEAKER_00One important thing is that you do have to make clear to people what the broad social benefits would be of considering value. I think people often imagine that like only if you consider value will you have to set priorities or decide what's covered in a formulary. In practice, what happens right now is that instead of formularies being limited by value, people do other things like tiering by ability to pay or requiring things like a step therapy or fail first, where you have to suffer pain or do really badly on a cheaper drug before you can be at a drug that's a higher priced formulary tier. Another example beyond step therapy is that there's work on what people call rationing by ordeal. It's not that you formally don't have drugs on your formulary because they don't provide sufficient incremental value over the other drugs that are available for the condition. But instead, it's just that you have to go through innumerable insurance appeals, and most people don't have the resources to do that. I do think that there are legitimate concerns about the quality-adjusted life here as a metric. And if people want to be critical of it, which I think they're legitimate criticisms, I tend to think you have to compare it to the status quo of what you do without having quality-adjusted life hearers, which isn't that there's just unlimited access. It's easy to say, well, we should ban this way of doing things because it's not perfectly fair. You have to compare it to what the status quo is. But I also think there are innovations in health economics, for instance, instead of using the quality, could you use other value metrics that mitigate some of the worry about unfairly exacerbating pre-existing disadvantage? It's actually just very unfortunate because actually pharmaceutical companies that are confident about their ability to deliver value would stand to gain from value-based metrics. I think there could be gains for everyone, but you have this sort of coalition between companies who would rather keep doing what they're doing and not innovate. And people who I think for understandable reasons are worried that polys aren't perfectly fair, organizing together to be critical of using some of these value metrics. Some of these bills would have the effect not just of saying like you can't use a poly threshold in the US, but would try to sort of by the back door prohibit comparing it to the price in the UK or Canada, because those countries, unlike the US, ask pharmaceutical companies to show that they're delivering some value in order to have the taxpayers pay for the drug. But there's pretty broad consensus about using things like international reference pricing as one factor. I think both the Trump administration and Biden administration talked about looking at whether drugs are being priced here in ways that are higher than where they're being priced in Canada or the UK or Germany. So it's pretty clear that a lot of these bills, there is some influence from the subset of pharmaceutical companies who don't want to valuably innovate to try to avoid being held to account for value.
SPEAKER_02Yeah, I think one takeaway here is that when we don't allocate resources intentionally, we wind up allocating resources unintentionally. And I'm curious if that was sort of your impression of what happened during COVID too. So, right at the beginning of this interview, you said that one thing that inspired you to work on these issues early in your career was this theoretical concern about a flu pandemic and how we might respond to it if it happened. And then, of course, a decade later, there is a pandemic. And so, how did living through a pandemic change your perception of how willing and able we are to be intentional about resource allocation?
SPEAKER_00I think give me a much more insight into seeing the sort of challenges in translating between the formal framework and implementation, and also just very vividly the extent to which there was, I think, often resistance to acknowledging the existence of scarcity and addressing it through some sort of formal and publicly acknowledged mechanism. So many states had these, I think, ethically very reasonable frameworks for thinking about allocation. But state governors were very, very reluctant to authorize what are called crisis standards of care in their state. My sense is that they felt like to admit that there was scarcity in their state of something was some sort of admission of failure. The other thing that stood out was that I think that there was a lot of interest group-related policy making by all sorts of interest groups to the extent that they could exert policy influence. But the interest groups that in practice seemed to have influence were not always the interest groups that actually were the most disproportionately burdened compared to what would have happened in the absence of the COVID pandemic. The group that I think had the biggest divergence from in relative risk, from the risk of death otherwise, actually came in some like groups of workers middle-aged, basically. But you didn't see heterogeneous organizations of, I don't know, people who worked in public-facing jobs or something like that having a lot of influence over these policies. Some patient groups at times, I think, were able to have more policy influence, sometimes. And some businesses were able to have that influence. Very understandably, people who had severe lung conditions were really worried about allocation policies and got up in arms about some of their perceived as unfairly disadvantaged them. But you didn't see people who thought that they were healthy middle-aged men getting up in arms about allocation policies, even though some of the like disapproporate burdens, a great pro-public article about why COVID-19 was so disproportionately harmful to young African-American men in the US. And we didn't see as much some organization looking out for those groups' interests. So I think even if the programs were revised in ways that maybe departed from what policymakers thought before the pandemic, it wasn't always in a way that was, I would say, like responsive to the people who were actually most harmed, as opposed to the people who thought they were going to be most harmed.
SPEAKER_03All right. So we like to close the podcast by asking our guests, what is one rule or norm broadly related to what we've been talking about today that you would change if you could and why?
SPEAKER_00There's so many possibilities. I think it's just one thought at the level of sort of implementation, but I would like to see more work done on ways to both consider a broader range of harms and relevant ethical factors, but to combine those into a priority-setting framework that is both publicly understandable, but also like genuinely multidimensional. I think we often pay lip service to the idea that we should consider fairness or justice or values other than maximizing benefit. And in practice, often what that has ended up being is like in the service of ostensible fairness, picking some very narrow notion of benefit, like just get more people to survive out of the hospital. Where I think a better approach to fairness would be the following. This is a change I would make is think about, for instance, something like a categorized priority system. What are populations that you worry might be treated unfairly by a one-dimensional allocation, whether that's people with pre-existing disabilities, people from disadvantaged communities, people at risk of death early in life, health professionals, what have you. And then think about can you prioritize some subset of resources for those categories and then allocate within them maybe by your benefit-based category? I think there are other innovations we can look at from other areas like operations research. Like for COVID vaccines, you had these lines and you would only be eligible based on one criterion, and that was the only set of people who could get the vaccine. And so you ended up waiting. And could you have had, you know, 80% of your vaccines on that day were prioritized for people in that group, but 20% were general public allocation? Allocation procedure implementation is not a new thing to people that are not doctors and not philosophers. But I think there's a lot that we can learn in terms of implementation that could be really helpful in reconciling some of these different genuinely weighty ethical principles, as opposed to either trying to reduce them all to points or like make them all tiebreakers.
SPEAKER_02Yeah, that makes a lot of sense. So that brings us to the end. And thank you so much, Gowen, for coming on the podcast. Great.
SPEAKER_00Thank you, Leah and Sophie.
SPEAKER_03Biounethical is written and edited by me, Sophie Jabert, and Leah Pearson with production by audiolift.co. If you want to support the show, please subscribe, rate, and review it wherever you get your podcasts and recommend it to a friend. You can also follow us on Twitter at Leah underscore Pearson and Sophie H. Jabert to be notified about new episodes. And you can sign up on our website, bioenethical.com, to receive emails when new episodes are released. We promise we won't spam you, but we may reach out to let you know about upcoming guests and give you the opportunity to submit questions. Our music is written by Nina Corey and performed by the band Social Skills. We're supported by a grant from Amplify Creative Grants. Links to papers we reference and other helpful resources are available on our website, bioenethical.com. You can submit feedback there or email us at bioenethical at gmail.com. Thanks for listening and for your support.