Fair Allocation of Scarce Resources in a Pandemic: Benefit, Reciprocity, Equality, and the Limits of Brudney’s “Principle”

Pandemics force health systems to make decisions that ordinary clinical ethics tries to avoid: who receives treatment when there is not enough to go around. Under those conditions, allocation criteria become more than policy; they become a public statement of what a system values, what it owes patients, and what trade-offs it’s willing to make under pressure. In this paper, I will present a COVID-19 pandemic scenario to keep in mind while offering three key considerations for resource allocation and explaining why these are important. Next, I will introduce Daniel Brudney’s “Principle” (a principle for the fair allocation of scarce resources) from his paper “Are Alcoholics Less Deserving of Liver Transplants?” (2007, 41). Then, I will analyze this principle in detail and argue that, although I believe it has value, it cannot be applied to the scenario I present. This essay aims to outline considerations for creating frameworks for future pandemics under limited-resource conditions. 

Consider the COVID-19 pandemic, when hospitals had limited resources for curative measures to treat patients infected with the virus. A vaccine has been made widely available to everyone in this hypothetical scenario, regardless of healthcare disparities, although not everyone is vaccinated. The patients arriving at these hospitals, therefore, include both vaccinated and unvaccinated individuals. How should these limited clinical resources be allocated among patients?

The first consideration is benefit maximization: increasing the number of patients who survive treatment with a reasonable life expectancy (meaning they are expected to live for a significant period after treatment, not just survive until discharge) and viewing improvements in length of life as a secondary goal (Emanuel et al. 2020, 2052). The length of life should be considered secondary because making it the primary focus shifts attention away from saving the most lives and instead emphasizes predicting and comparing long-term outcomes. Prioritizing length of life could result in selecting patients expected to gain more years, even if they require more time or support to recover from the illness. Meanwhile, another patient might survive with quicker or less intensive treatment, shifting the focus away from preventing the most deaths to predicting whose life will be longer later. During such times, the most urgent moral goal is to prevent the most avoidable deaths and give as many people as possible a fair chance at continued life. Framing benefit maximization this way also reflects the inherent value of human life by avoiding ranking people based on social worth or other subjective judgments. Only once patients have similar survival probabilities does it become appropriate to consider additional life-years as a secondary consideration. 

Shouldn’t the patient’s future quality of life or quality-adjusted life also be considered? Those specifics are only relevant when comparing patients with similar survival probabilities because, if one patient is much more likely to survive with treatment, that difference matters more than differences in predicted quality of life, aligning with my framing of maximizing benefits in this scenario. Due to limited time and information during a pandemic emergency, it is reasonable to advise against including patients’ future quality of life and quality-adjusted life years in benefit maximization (Emanuel et al. 2020, 2052). Doing so would require a time-consuming data collection effort that could otherwise be used to save other lives that are more urgent, and it could involve ethical and legal issues that I will not discuss further in this paper.

Implementing benefit maximization means prioritizing those who are sick but could recover with treatment over those unlikely to recover even with treatment—and over those who are likely to recover without intervention. Because young, severely ill patients often comprise many of those who are sick but could recover with treatment, this approach also has the effect of giving priority to those who are worst off, in the sense of being at risk of dying young and not having a complete life full of lived experiences (Emanuel et al., 2020, 2052; Reis-Dennis and Applewhite 2022, 505–506).

In light of this, a key scenario to consider is as follows: If patient A is using a scarce curative resource and patient B arrives with a prognosis indicating that reallocating the resource to patient B would maximize benefits, what should the care team do under this framework? Although this remains widely debated today, let’s assume that withdrawing a scarce resource to save others is not considered an act of killing and does not require the patient’s consent. I aim to show that reallocating the scarce curative resource from patient A to patient B would be permissible, but not obligatory. 

It is reasonable to argue that it is permissible because reallocating the scarce resource would maximize benefits, allowing the greatest potential to save the most lives with a reasonable life expectancy, thereby doing the most good overall in this scenario. What makes this choice not obligatory are the potential downsides that may result from such a decision. Removing scarce resources from one patient to give to another can be extremely psychologically traumatic for clinicians. This trauma can be severe and may leave lasting impacts on the clinician’s mental health. For clinicians sworn to do no harm, removing resources from someone in desperate need is a difficult and emotionally draining task. 

Beyond ethical considerations, there is also a practical limitation: in some systems, clinicians may face legal and institutional consequences for withdrawing or reallocating life support once it has been initiated. Even if reallocation practices could maximize benefits, it is unreasonable to consider it a mandatory practice in a scenario where the system itself may penalize clinicians for doing so. Therefore, reallocating scarce resources in cases like these should be considered permissible only under a clearly defined protocol that safeguards clinicians, rather than being an obligatory practice (Reis-Dennis and Applewhite 2022, 506).[1]

The second consideration is to prioritize healthcare workers and others who care for sick patients and maintain infrastructure, especially those at high risk of infection and for whom training makes them difficult to replace, over other patients (Emanuel et al. 2020, 2053; Reis-Dennis and Applewhite 2022, 511–512). This aligns with the practical idea of benefit maximization since protecting the workforce helps the system save more lives overall. These workers should be prioritized due to their critical role in responding to the COVID-19 pandemic. If these healthcare workers become incapacitated, all patients, including those with COVID-19, would face higher mortality rates and more lost years of life (Emanuel et al. 2020, 2053). Giving healthcare workers priority may also motivate them to return to work sooner after illness, as they might feel encouraged by the priority given to their treatment. 

However, what if a healthcare worker becomes terminally ill and cannot return to work? Should we still prioritize them over others, even if they are unlikely to rejoin the workforce? It is reasonable to argue that they should still be given priority in this case. This is because prioritizing them recognizes their willingness to undertake high-risk work to save others. It could also motivate other healthcare workers to continue their life-saving work, knowing the system would support them no matter how ill they become. Just imagine if they weren’t given priority in this situation. Healthcare workers might fear working and face increased stress, constantly worrying about what could happen if they became terminally ill. If that were the case, it’s reasonable to assume that fewer healthcare workers would be available to help us survive the pandemic, since many might stop working out of fear of falling terminally ill. Assuming fewer healthcare workers would have a harsher impact on the pandemic than prioritizing scarce resources to the terminally ill healthcare workforce, it becomes necessary to prioritize them to maximize overall benefits in this situation.

 The third consideration is to treat patients with similar prognoses equally, meaning that when there is a surplus of patients with comparable health conditions and a shortage of resources, a policy should be established to address this issue fairly. But how should this be done? When patients have similar prognoses, benefit maximization cannot clearly distinguish between them, so equality becomes the appropriate way to decide among tied cases. Equality can be achieved and implemented by randomly allocating scarce resources rather than using a first-come, first-served approach (Emanuel et al. 2020, 2053). Practically, this would not require delaying care to wait for patients to accumulate. Instead, when a scarce resource becomes available, and more than one patient with a similar prognosis needs it at that moment, the hospital can run a simple lottery among that small group. Because COVID-19 treatments address urgent needs, a first-come, first-served method would unfairly favor patients living closer to healthcare facilities and could even encourage crowding and violence, especially when social distancing is critical (Reis-Dennis and Applewhite 2022, 508). 

There is reason to believe that patients, especially those who may get sick but can recover on their own, might start crowding healthcare facilities out of fear that waiting to see if their symptoms worsen could result in no treatment at all, since they chose to wait instead of going to their doctor right away after noticing symptoms. Additionally, people who do happen to get sick later on, perhaps because of their strict adherence to recommended public measures, such as social distancing, may be excluded from treatment, which would worsen outcomes while also not improving fairness. A first-come, first-served system may also cause enough frustration to lead patients to act irrationally, either out of anger at the system or to pressure healthcare providers to see them sooner.

However, random selection would mitigate some of these problems. Given the limited time and information available during a pandemic emergency, a random selection system might be considered more suitable when deciding which patients with similar prognoses receive scarce resources. When patients have comparable prognoses, determining who would benefit more from limited curative resources often relies on subjective judgment, and different clinicians may arrive at different conclusions. A lottery system avoids this inconsistency. While stress and psychological hardship are inevitable in these situations, a first-come, first-served approach can add to this burden, as healthcare workers would be placed in a challenging position when deciding which patients should and should not receive scarce resources, while a lottery system leaves the decision to chance. Therefore, it is more practical for healthcare teams to randomly select which patients receive resources, reducing procedural stress and delays for caregivers, even though moral distress from denying care may still persist.

Now that I have outlined three considerations for this scenario, I would like to introduce Daniel Brudney’s “principle.” Daniel Brudney presents this principle based on his conversations with patients who require a liver transplant due to their alcoholism. When explaining my stance on his principle, I will frequently use the example of alcoholism to illustrate my argument.[2] He uses this principle to support the conclusion that most patients needing a liver transplant because of their alcoholism don’t satisfy the principle’s premises and therefore aren’t less deserving than others of receiving a scarce, lifesaving resource. 

Daniel Brudney’s principle states that “[a]n agent is responsible for knowingly, voluntarily, and repeatedly engaging in easily avoidable conduct that might significantly contribute to that agent’s needing a scarce, lifesaving resource” (Brudney 2007, 42). Was the agent sufficiently educated? Did they do it of their own free will? Were they able to repeatedly decide otherwise? Brudney claims that “[a]n agent responsible for such conduct may legitimately be given a weaker claim on scarce, lifesaving resources if their need for such resources is due to such conduct” (42). In other words, this principle provides a clear guideline for lowering someone’s priority on a transplant or treatment waitlist, not excluding them altogether. Specifically, if you knowingly, voluntarily, and repeatedly engaged in easily avoidable conduct that caused your injury or medical condition, then you are less deserving of a scarce, lifesaving resource compared to someone who is not responsible, perhaps excepting cases like substance abuse disorder, where autonomy is compromised.

Daniel Brudney’s principle has some validity. Knowingly damaging your liver from drinking excessively and then needing a transplant to stay alive can be seen as demonstrating a disregard for how your actions impact another person’s chances of survival (Brudney 2007, 45). Assuming we have a moral obligation not to harm others’ chances of survival, doing so is highly immoral and should weaken someone’s claim to a liver transplant compared to someone whose need arises from circumstances beyond their control. 

Moving a patient to the bottom of the transplant list significantly raises the risk of that patient’s death. One might wonder whether significantly increasing that risk is an appropriate response to the immoral behavior outlined above. Going back to the liver transplant analogy, all the patient did was drink alcohol, a common practice in many communities. Do they really deserve to have their chances of dying while waiting for a transplant dramatically increase?

While I agree that it would be disproportionate to increase this patient’s chances of dying solely because they knowingly, voluntarily, and repeatedly engaged in easily avoidable conduct that caused their illness, a factor being overlooked is that other people also need liver transplants. By moving this patient down the list, we increase the chances of someone else surviving, and decreasing this other patient’s chances of survival is solely a side effect of this good. 

Turning to how this principle could apply to my vaccine scenario, I want to begin by providing some background: a vaccine is a preventive measure that significantly increases a patient’s chances of survival and reduces the severity of symptoms if they ever contract the disease. Next, I’d like to address another question that critics might reasonably raise. If you believe that Brudney’s principle has merit, why don’t you consider it when discussing the scenario you present? In other words, if you think Brudney’s principle is valid, why isn’t it one of your considerations that, when allocating scarce resources, vaccinated patients have a stronger claim than those who voluntarily, knowingly, and repeatedly chose not to get vaccinated, despite the vaccine being readily available? 

While this consideration has merit, there is reason to doubt that healthcare providers can practically assess how knowledgeable a patient was about their ability to access a vaccine or, even if they were knowledgeable, how voluntary their decision to decline vaccination was. Here’s why. 

The challenge with enforcing Brudney’s principle is that patients can lie in three ways: about whether they have the vaccine, whether they knew its benefits, or, if they knew, how voluntary their decision not to get vaccinated was. Why would a patient lie about these things? If a patient knows that their claim to scarce resources depends on the answers to these questions, and, assuming they want to stay alive, some patients might feel pressured to misrepresent their information to their healthcare team. Patients who decide to lie risk losing their credibility, which can reasonably be seen as a risk they might be willing to take if they believed their life depended on it. Because of this, there is reason to believe that patients would frequently lie if it meant their lives could be saved. 

The issue now becomes how the healthcare team can reliably verify whether the patient is being truthful. The healthcare team must find a way to confirm that patients’ stories are accurate. Although there may not be a completely reliable method to verify a patient’s entire account, especially regarding their understanding or how voluntary their decision was, let’s assume for the argument’s sake that one exists; it would likely require significant time and effort from the healthcare team to make difficult judgments, which could otherwise be allocated to serve patients. Since it is reasonable to assume that these judgments could not be made quickly or consistently across time-sensitive patients in urgent need, it is reasonable to conclude that Brudney’s principle becomes impractical in this COVID-19 scenario, which explains why it wasn’t included in any of the recommendations I gave. 

In this paper, I have presented a COVID-19 pandemic scenario to keep in mind while offering three key considerations for resource allocation and explaining why these were important. Then, I introduced Daniel Brudney’s “Principle” from his paper “Are Alcoholics Less Deserving of Liver Transplants?” (Brudney 2007, 41). After analyzing this principle in detail, I argued that, although I believe it has value, it cannot be applied to the scenario I present. While I list three considerations, the latter two often serve as ways of applying the first: prioritizing healthcare workers maintains the system’s capacity to save lives, and lotteries among similarly situated patients implement benefit maximization fairly when medical differences are minimal. This essay aimed to outline considerations for creating frameworks for future pandemics under limited-resource conditions.

End Notes

[1] For example, some jurisdictions were reluctant to endorse ventilator reallocation (a curative resource) during the modern COVID-19 pandemic. New York State guidance explicitly rejected the redistribution of ventilators from long-term care facilities, demonstrating how legal and institutional constraints can limit what clinicians can do, even when reallocation could maximize benefits.

[2] However, it is widely debated today whether substance use disorders can be considered “voluntary” under Brudney’s principle, so take the alcoholism example as an analogy to help with understanding its application to vaccines.

Bibliography

Brudney, Daniel. “Are Alcoholics Less Deserving of Liver Transplants?” Hastings Center Report 37, no. 1 (2007): 41-47. https://dx.doi.org/10.1353/hcr.2007.000 

Emanuel, Ezekiel J., Govind Persad, Ross Upshur, Beatriz Thome, Michael Parker, Ari Glickman, et al. “Fair Allocation of Scarce Medical Resources in the Time of Covid-19.” New England Journal of Medicine 382, no. 21 (2020): 2049–2055. https://doi.org/10.1056/NEJMsb2005114 

Reis-Dennis, Samuel, and Megan K. Applewhite. “Rationing Ventilators.” In Difficult Decisions in Surgical Ethics, edited by Victor A. Lonchyna, Peter Kelley, and Peter Angelos, 503–518. Cham, Switzerland: Springer, 2022. https://doi.org/10.1007/978-3-030-84625-1_36