Rebuttal From Dr White et al

We appreciate the opportunity to respond to our colleagues’ arguments against incorporating equity considerations when allocating scarce critical care resources during a pandemic. We respond to four objections they raised.

First, the authors claim there is no justification to incorporate equity considerations into ICU allocation frameworks because there is no evidence that disadvantaged groups have worse ICU outcomes once hospitalized and when resources are adequate. However, these data from times of adequate ICU resources are not informative because they do not illuminate what would happen in the setting of scarcity that requires allocation decisions. The use of purely prognosis-based allocation will result in disadvantaged populations, especially Black patients, being disproportionately and inequitably denied ICU care. As we detailed in our “Pro” position, this is the case because disadvantaged populations generally present with higher severity of illness than advantaged populations—because of the impact of structural inequities on health—and they will consequently receive worse triage priority scores. Moreover, their argument ignores the fact that during the COVID-19 pandemic, disadvantaged groups have required ICU care at 2 to 3 times the rate per capita than other groups, because of their excess burden of the negative social determinants of heath. As we illustrated in Table 1 in our Pro article, under these conditions of disproportionate burden, using prognosis-based triage would actually widen the disparities in death rates across racial groups. This is the case even if, counterfactually, we assumed that baseline triage scores were similar across groups. Therefore, unless society believes that it is acceptable for ICU triage to widen disparities, equity considerations must be built into triage frameworks. Hick et al¹ have elsewhere declared that crisis care protocols “should not exacerbate underlying disparities”.¹

Second, Hick and Hanfling are concerned that incorporating an equity correction into allocation frameworks would violate the bedside physicians’ duty to equally treat all patients in front of them. However, this objection implies that triage should be governed by the ethics of the doctor-patient relationship. In our view, it should not. Allocation of scarce resources during a declared public health emergency is a public health matter, concerned with equity and efficiency at the population level.² It is not a private matter to be negotiated between a physician and their individual patients. Consequently, concerns about physicians’ fiduciary obligations to individual patients are not a compelling reason to disallow equity considerations in triage under crisis conditions. Public health ethics is concerned with both using scarce resources to save lives while also striving for equitable distribution of those benefits. In this view, incorporating an equity correction into the allocation framework to prevent a worsening of disparities is well justified.

Third, Hick and Hanfling claim that using the ADI or other population-based indices to aid in equitable resource allocation would violate the US Office for Civil Rights (OCR) requirement that patients receive individualized evaluations. However, OCR’s recent actions suggest that they support the use of population-based measures. For example, OCR reviewed and did not object to the Commonwealth of Pennsylvania’s ventilator allocation framework, which includes the use of a population-based mortality prediction model to predict hospital survival.³ The Commonwealth of Pennsylvania also recommends use of the ADI as part of its multi-principle framework to allocate scarce COVID therapeutics, without objection from the OCR.⁴ Race is not included as a consideration in the ADI.

Finally, Hick and Hanfling object to the use of equity corrections on the grounds that there is not “community consensus to correct historical inequities at the bedside.” However, the purpose of the equity considerations we propose is not to correct the massive historical inequities arising from our country’s racist social policies. Redress for these past and present injustices will require much larger policy interventions across numerous policy domains.⁵ The purpose of the equity corrections we propose for ICU triage is to mitigate the extent to which present-day health disparities would cause disadvantaged groups to be disproportionately denied ICU care under prognosis-based triage. ICU triage based only on prognosis would widen the disparities in COVID death rates across racial groups, an outcome Hick and Hanfling agree is unacceptable. Therefore, equity corrections are necessary to achieve a triage strategy that both saves lives and fairly distributes these health benefits across the population.