The coronavirus disease 2019 (COVID-19) pandemic is forcing some health systems to choose who lives and who dies because of limited treatment capacity. Among patients with urgent life-threatening conditions, shortages require triaging or designating some to life-saving care and others only to supportive measures. Later terrible choices can involve whether to withdraw patients from life-sustaining support when others have better outlooks. Not surprisingly, raging discussions address the production and distribution of ventilators and other vital supplies.1,2
I argue below that triage decisions should not consider life-years saved beyond brief posthospital survival. The reasons are as follows: (1) historical and present inequities have reduced expected life-years in populations experiencing chronic disadvantage (racial/ethnic minorities are key examples); (2) justice requires avoiding policies that further increase inequities; and (3) greater priority for more predicted life-years saved will exacerbate those inequities. A related justice argument supports diverse representation among policymakers. The underlying moral framework is social justice that builds on respect for persons and people’s equal and substantial moral worth.
PUBLISHED GUIDANCE AND MAXIMIZING LIVES AND LIFE-YEARS SAVED
Recent documents and publications proffer and defend moral frameworks for guiding these terrible triage decisions. As a prime example, Emanuel et al3 in March 2020 published a framework for “Fair allocation of scarce medical resources in the time of COVID-19” in the New England Journal of Medicine. Their “ethical values and guiding principles” were as follows:
- “Maximize benefits” (“Highest priority” for saving most lives and life-years).
- “Treat people equally” (“random selection” for people with “similar prognosis,” ruling out “first-come, first-served”).
- Factor in “instrumental value” (“research participants and health care workers,” given equality of other key elements).
- “Priority to the worst off” (“sickest first” and “youngest first” “when it aligns with maximizing benefits”).
I recommend reviewing their detailed and thoughtful account.
Maximizing lives and life-years saved is also the lead emphasis in the March 2020 “Allocation of Scarce Critical Care Resources During a Public Health Emergency,” a model from the Department of Critical Care Medicine, School of Medicine, University of Pittsburgh.4,5 This document also aims to guide triage decisions about critical care that COVID-19 is generating. The proposal is another detailed account worth careful review. In this model, clinical triage personnel assess short-term and long-term prognoses, outcomes expressed through a scoring system. People with worse scores receive lower priority for critical care. This strategy aims to maximize lives and life-years saved.
ETHICS AND TRIAGE POLICY
In a time of key shortages, triage needs raise questions not only about allocation criteria, but first, who should decide guiding policies, how they should do so, and what ethically should inform policy decisions. Clarity about these foundational aspects then provides a lens for examining emphasis on lives saved and life-years maximized. In examining who should decide policies and how to do so, suppose we are a task force asked to design a provisional model for a state, health care system, or a health care facility. Where should we start? We can reasonably begin with a foundational moral framework open to reflexive elaboration and revisions.6 Task force composition, function, and ultimate triage policies should fit core ethical features such as principles and values.
Social Justice, Respect for Persons, Substantial and Equal Moral Worth7
This schema starts by asserting people’s equal and substantial moral worth, often typified as inherent dignity. One thing that the ethical principle of respect requires is empowering and recognizing everyone who has a stake in what is being addressed.8 Thus, respect for persons intrinsically entails equity in the sense of ensuring people’s equal opportunity for influence. (This respect principle supports respect for autonomy, not addressed here.) Hence, justice as fairness and respect for persons are inseparable. Social justice builds on this base to address equity and other justice issues concerning populations and groups. Although these are only summary points from a huge literature, I submit that they provide a defensible ethical foundation for triage decisions. Any acceptable triage model must be consistent with the justice-respect-worth framework.
Policy Formation and Triage Decision-makers
Social justice, resting on respect for persons and stemming from people’s substantial and equal moral worth, demands especially that historically marginalized and oppressed groups have a major voice in allocation policy. Of course, equitable representation means that all affected groups should have a voice about policies influencing their care. However, past inequities reinforce a major place for disenfranchised populations. Not only have they often been abused but they also have had limited or no power to advocate sufficiently for themselves. Thus, we could draw on a concept of reparative justice.9 Moreover, feminist and critical race theory analyses show how existing structures and discourse standards maintain male and majority White power and control.10 An important counter to such injustice is fair, empowered, and recognized voices at decision tables.8 Allocation policies generated by narrower groups are ethically suspect, even if the policies themselves are compatible with the ethical foundation. Thus, to represent interests of populations historically discriminated against, triage decision-makers must include diverse personnel.11
JUSTICE VERSUS MAXIMIZING SAVED LIVES AND LIFE-YEARS
The ethical framework of social justice, respect for persons, and equal and substantial moral worth demands that rationing policies must consider injustice in assessing possible criteria for allocation decisions about individuals. In that light, we should reconsider the above frameworks published by Emanuel et al3 and in the Pittsburgh (2020) document.4,5
Maximizing Saved Lives Generally Violates the Justice-Respect-Worth Framework
Assuming that prognostic assessments are accurate and reliable (not examined here), both frameworks will give lower priority on average to individuals for whom social/structural inequities are significant causes of worse health, increasing the injustice. That is, previous adverse and inequitable social conditions produce greater morbidity through underlying health conditions, unfavorable environments, reduced access, and much more.2 In short, stressing lives and life-years saved will often violate the justice-respect-worth framework. Counting prospective saved life-years further compounds inequities.11 Saving lives matters and saving more lives is important, all else equal. However, all else is not equal.
Saved Life-years Should Be Rejected as a Triage Criterion
Justice supports triage priority for those with better initial survival prognosis, but opposes considering subsequent life-years saved. Those with substantially lower probability of survival a few months after hospital discharge should receive lower priority for life-saving measures. Although living a few months more has meaning, such short-lived individuals generally will have very poor quality of life. Thus, these minimal gains seem an unjust use of scarce resources when others could benefit much more.
Populations experiencing social inequities, on average, will likely have worse outlooks for survival to hospital discharge or, say, for another 3 months. They already have higher frequencies of advanced chronic disease and will present relatively later for care of COVID-19 infection. Given higher concentrations in service jobs and congested housing, disenfranchised groups will disproportionally have greater infection rates.11 Hence, factoring short-term prognosis in triage priority will be an additional injustice for some. I see no immediate corrective for these inequities.
In contrast, justice requires ignoring life-years saved after predicted initial survival. Doing so avoids perpetuating inequities. But ignoring saved years seems unjust in another way. The elderly already had concrete opportunity for life. The young obviously not. Thus, a justice-based concept of fair equality of opportunity for living a life seems to require life-saving priority to the young over the old. The above schema of Emanuel and colleagues would give priority to the young when consistent with benefit maximization. The Pittsburgh document would count age as a “tiebreaker” in some cases when all else is equal, suggesting demarcations of “age 12–40, age 41–60; age 61–75; older than age 75.” (Why exclude below age 12 is not addressed.) However, sharp demarcations by age group are problematic at the margins, such as one day older or younger than 40. Thus, arbitrariness is a concern. But still, the opportunity basis of favoring the young apparently makes ethical sense. But consider further.
Suppose only my grandchild or I can get life support, assuming we could then survive for some years. I and many will insist that the child’s life should preferentially be saved. Also, I and probably many would do the same for a stranger’s child. Moreover, fair equality of opportunity to live a complete life here absolutely favors priority to the young, absent outweighing factors. Justice in such “pure young versus old cases” seems to demand that life-saving measures always go to significantly younger individuals, fitting our personal sense of fairness.
Instead, what if I, an older person, belong to a historically abused population (oppressed and lynched, for example), which continues experiencing inequities, and the child’s group were unjustly privileged, and previously murderous oppressors. Would I, or should I, then insist that the child get life-saving preference? I might so insist out of generosity, but fairness no longer seems to demand saving the child. Rather, past and continued injustices are strong reasons why potential life-years saved should not influence triage priorities for individuals who prospectively will likely survive more than a few months after hospitalization. (Bias against the elderly is another reason not to weight life-years saved beyond initial survival.12)
An objection is that the very young cannot be held responsible for their group’s actions and privilege. Thus, group membership is an unfair reason to deny their opportunity for a more complete life. Hence, the very young should receive triage priority over the much older. This objection’s reasoning has merit, but fully analyzing the issue is beyond the present scope. In terms of triage of the very young versus the much older, justice also grounds following the recommendations of those groups experiencing chronic injustices, through their representatives on the envisioned task force.
Triage Teams Should Be Diverse
Triage decisions for life-saving measures are fraught with possible inequity. In the Pittsburgh model, triage personnel are apparently not shuttered from patient demographics and potentially other aspects such as sexual orientation. A triage physician does the assessment. The guidance tries to ensure triage objectivity. However, implicit and unconscious negative bias can influence the best-intentioned assessments. Because such adverse bias is often directed toward groups already experiencing social inequities, average triage evaluations of them will be worse, enhancing the injustice. As a partial corrective, populations historically oppressed and disadvantaged must be represented on triage teams.11
Triage Accounts Incorporate Justice, but Differently
To my knowledge, published triage models and related analyses explicitly or implicitly incorporate social justice and background concepts of respect for persons and their equal and substantial worth. My stress on the justice-respect-worth framework does not imply otherwise. I presume that all value these core moral precepts. At issue is how we interpret them.
The foundational moral framework for triage of life-saving measures should be social justice coupled with respect for persons and their equal and substantial moral worth. This framework has important implications for who sets policies, how they do it, and the policies. Policy decision-makers must include empowered representatives of disenfranchised and oppressed groups, such as populations experiencing historical racial/ethnic bias, oppression, and other social inequities. Concrete allocation policies and strategies must build on this ethical framework. Triage policies can reasonably lend priority to people more likely to survive hospitalization and a brief time after. Priority for subsequent life-years saved perpetuates social injustice and generally should be avoided.
The author thanks Jeroan Allison, MD, MS, for suggesting the general topic, shared comments about the Pittsburgh model, and helpful feedback about an earlier draft that included 2 other reviewers’ input. One reviewer also contributed line-by-line review and comments. Many thanks are due to all.
1. Cha AE, McGinley L. Who gets a shot at life if hospitals run short of ventilators? The Washington Post
; 2020. Available at: www.washingtonpost.com/health/2020/04/07/ventilators-rationing-coronavirus-hospitals/
. Accessed April 15, 2020.
2. Schmidt H. The way we ration ventilators is biased. The New York Times
; 2020. Available at: https://www.nytimes.com/2020/04/15/opinion/covid-ventilator-rationing-blacks.html
. Accessed April 16, 2020.
3. Emanuel EJ, Persad G, Upshur R, et al. Fair allocation of scarce medical resources in the time of COVID-19. N Engl J Med. 2020. Doi: 10.1056/NEJMsb2005114.
4. Department of Critical Care Medicine, School of Medicine, University of Pittsburgh. Allocation of scarce critical care resources during a public health emergency; 2020.
5. White DB, Lo B. A framework for rationing ventilators and critical care beds during the COVID-19 pandemic. JAMA. 2020. Doi: 10.1001/jama.2020.5046.
6. Wendel ML, Garney WR, Castle BF, et al. Critical reflexivity of communities on their experience to improve population health. Am J Public Health. 2018;108:896–901.
7. Stone JR. Healthcare inequality, cross-cultural training, and bioethics: principles and applications. Camb Q Healthc Ethics. 2008;17:216–226.
8. Blacksher E. Redistribution and recognition—Pursuing social justice in public health. Camb Q Healthc Ethics. 2012;21:320–331.
9. Walker MU. Moral Repair: Reconstructing Moral Relations After Wrongdoing. New York, NY: Cambridge University Press; 2006.
10. Bohman JZalta EN. Critical theory. The Stanford Encyclopedia of Philosophy (Winter 2019 Edition). 2005. Available at: https://plato.stanford.edu/archives/win2019/entries/critical-theory/
11. Santiago J, Rollins R, Pressley A. Bold action needed to battle coronavirus in communities of color. The Boston Globe
; 2020. Available at: www.bostonglobe.com/2020/04/10/opinion/bold-action-needed-battle-coronavirus-communities-color/
. Accessed April 21, 2020.
12. Stone JR. Elderly & older racial/ethnic minority healthcare inequalities: care, solidarity, and action. Camb Q Healthc Ethics. 2012;21:342–352.