by B. Corbett Walsh, MD, MBE; Anna Nolan, MD, MSc
As intensivists practicing in New York City at the epicenter of the COVID-19 pandemic, there was a palpable concern that there may not be enough ventilators for every patient that required it. If rationing were to occur, it should utilize a principled morally sound algorithm to aid physicians’ resource allocation decisions. While there has been much written about this timely topic, we would like to focus on a recent manuscript: Drs. White and Lo describe allocating scarce resources via a predicted mortality score using the Sequential Organ Failure Assessment (SOFA), with a subsequent modification for comorbidities to prioritize saving the most life-years. After patients are prioritized, ties are adjudicated by age, usefulness during the pandemic, and finally by lottery. The proposed framework should be praised for its helpfulness, simplicity, inclusivity, and anticipation of future reallocation difficulties.
However, this allocation tool does raise concerns. Most problematic is that a SOFA score remains unvalidated in predicting mortality in COVID-19. Additionally, the mental status exam or blood oxygenation components can be confounded by the deep sedation, paralysis, or prone positioning that are often necessary in refractory hypoxemia. The proposed framework may be preliminarily favorable to those without life-limiting comorbidities, but, it becomes deeply concerning when chronic poorly controlled comorbidities are included that might seem punitive to those of low socio-economic status or poor healthcare access.Lastly, it is unclear how pregnant, chronically ventilator dependent, or other special populations would be addressed.
In addition to the increasing attention that comorbidities has on COVID-19 related mortality, similar focus has been devoted to the identification of biomarkers, such as d-dimer or CRP, that might suggest elevated risk of morbidity and mortality. We applaud the authors for excluding disease specific biomarkers and would caution their inclusion for three reasons. First, there may be insufficient evidence to support their incorporation into a resource allocation tool. Second, while alternative mortality prediction tools exist for other disease processes, we would challenge that COVID-19 related biomarkers would meaningfully contribute to predicting morbidity and mortality that is not already captured in the SOFA score. Lastly, it may be difficult to interpret disease-specific biomarkers in an inclusive allocation tool when their elevation in other diseases does not suggest such risk of morbidity or mortality.
Providing a preference to healthcare providers combating the pandemic rests on the utilitarian principle that the sick need caretakers to facilitate their healing and reciprocity. Such reciprocity exists between providers, having a strong ethical duty to provide care for the ill even at elevated risk to their own health and their community. Discharging this duty requires the community to safeguard providers’ health and well-being while at work (food, water, rest, mental health, measures to limit occupational hazards including personal protective equipment, prophylactic and proportional access to treatment recognizing the elevated risk providers take), and minimize providers’ legal exposure during a crisis. This preference should be flexible. We agree with authors that healthcare workers should receive heightened priority. However after this principle of reciprocity has been discharged and in a supreme emergency where all of society is singularly focused on combating disease, resources should be prioritized proportionally to the likelihood that the individual will be able to return and meaningfully contribute to curbing the pandemic. This can be difficult to operationalize as healing is facilitated by more than doctors and nurses but also pharmacists preparing medications, environmental custodians and security who ensure a clean, safe work environment, in addition to many others. Additionally in circumstances where resources are strained but not maximally scarce, priority for healthcare workers should not be unilaterally overriding.
Finally, the authors emphasize the challenges of tertiary triage, particularly during a novel pandemic where our understanding of disease trajectory is evolving rapidly in real time. Once a pandemic triggers a declaration of crisis standards of care, all scarce resources should enter and be allocated by the same allocation system. Scarce resources will likely differ slightly by facility, and will be dynamic as a pandemic unfolds. These can include space (ICU beds), staff (nurses, physicians, respiratory therapists), or specific items that are needed to support end-organ dysfunction (high flow oxygen delivery devices, ventilators, and dialysis machines). In addition, hospital supplies such as instrument specific tubing and disposable patient specific items needed for ventilators, dialysis, and oxygen delivery devices (oxygen, facemasks and nasal cannulas, or filters). Tertiary triage should ensure that allotment does not revert to a first-come first-serve policy that might disadvantage low socio-economic individuals, those who have unreliable transportation, or those who distrust the healthcare system. Despite these challenges the authors should be praised for their work. The importance of their paradigm and resource allocation as a whole is further highlighted as conditions continue to evolve and healthcare institutions finalize disaster preparedness policies.