Editor’s note: Find the latest COVID-19 news and guidance in Medscape’s Coronavirus Resource Center.
Approximately twice as many Black patients as patients of other ethnicities scored in the lowest priority group for allocation of critical care resources, as determined from data from nearly 500 patients treated at six hospitals during a COVID-19 surge.
Regional surges in critically ill patients during the COVID-19 pandemic prompted concerns about the allocation of critical care resources, such as ventilators and available intensive care unit (ICU) beds, write Elisabeth D. Riviello, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, and colleagues. Crisis standard of care (CSOC) plans developed in response to the pandemic focused on the greatest number of lives saved and life-years saved, but whether CSOC triage scores exacerbated racial inequities has not been thoroughly explored, they say.
In a study published online March 15 in JAMA Network Open, the researchers retrospectively analyzed scoring data from a regional surge in COVID-19 cases between April 13, 2020, and May 22, 2020, at six hospitals in a single healthcare network in greater Boston, Massachusetts.
The study population included 498 adults admitted to the ICU, 225 of whom had COVID-19. Patients were scored by acute severity of illness using the Sequential Organ Failure Assessment score and by chronic severity of illness using scores for comorbidity and life expectancy. The median age of the patients was 67 years; 38.4% were women, and 15.9% were Black patients.
Overall, Black patients were significantly more likely than any other patients to be in the lowest priority group for resource allocation (15.2% vs 8.1%; P = .046). Several patient outcomes differed significantly according to race. Notably, Black patients had a longer median duration of mechanical ventilation (15 days vs 8 days), longer ICU stay (8 days vs 5 days), and longer hospital stay (13 days vs 10 days) compared to White patients; however, no significant difference appeared between Black and White patients in the likelihood of death or discharge home, the researchers write.
The researchers also conducted an exploratory simulation model using the score for ventilator allocation (244 patients), with only highest priority group receiving ventilators. In this model, they found that there were 43.9% excess deaths among Black patients, compared to 28.6% among all other patients (P = .05). When they analyzed the highest and intermediate groups, these percentages were 4.9% and 3.0%, respectively.
In this model, providing ventilators to all patients in the highest priority group yielded 31.2% excess deaths; by comparison, using a random allocation system yielded 35.9% excess deaths, the researchers say.
“If higher severity of illness scores among Black patients reflect the impact of systemic racism on health, the CSOC system could perpetuate the impact of racism on health by deprioritizing patients with more severe illness,” the researchers write.
The study findings were limited by several factors, including the use of a single health system that is not the primary hospital system for the region, which could cause selection bias, the researchers note. Other limitations include the relatively small sample size, the data collection in the context of a regional patient surge with a lack of quality control, a high rate of missing data, and the inability to assess long-term mortality, they say.
Unpredictability of Resource Constraints Remains a Challenge
“Even before the COVID-19 pandemic, crisis standards of care planning was an important part of disaster preparation; however, a detailed and practicable approach becomes even more critical throughout the country in the midst of regional crises in the COVID-19 pandemic,” said lead author Riviello in an interview. “To be very clear, we never needed to implement crisis standards of care plans in our hospital network. But we knew that we could see resource crises again,” she said. “We felt it was important to use the time between crises to evaluate how we would have done as a health system and what the impact would have been on health inequities,” she added. “We wanted to know whether any crisis standards of care strategies could mitigate the impacts of racism on health outcomes during a resource crisis,” she added.
Dr Elisabeth Riviello
“We were surprised by the findings in the sense that researchers are always surprised; we don’t actually know the answer we will find when we ask a question,” Riviello said. When conducting the study, “we did wonder whether crisis standards of care systems based on mortality prediction scores might deprioritize people of color,” she said. “Given baseline inequities in health outcomes due to racism, we knew people of color might arrive to the hospital acutely sicker and with more comorbidities than others. This fact alone could make them less likely to receive resources in a system that prioritizes people who are most likely to benefit from resources based on predictive scores,” she added.
“Our exploratory simulation of a random lottery showed that more lives might be lost with a random lottery as compared with a score-based system,” Riviello noted. “It demonstrated that a lottery is not a clearly better alternative to scoring systems because more people might die overall, and even within each racial group,” she said. “That finding was striking to us. It is important to understand the potential impacts of different systems that have been proposed over time,” she emphasized.
“Each [COVID-19] surge brought different potential resource constraints. In the first surge, we were concerned that ventilators and staffed ICU beds could become inadequate to meet need. In subsequent surges, it was hospital beds and staffing that were potentially at risk to be inadequate,” Riviello said. “We always needed an ethical approach that included equity, but each surge had particular frames and challenges. We did approach each of the surges with equity as a key concern, and our working on this data helped to keep equity at the forefront of planning efforts during each surge,” she noted. “The deep challenges of using scoring algorithms for prioritization in times of resource scarcity became very clear,” she said.
Depending on resource constraints, different policies might be needed, which could be one barrier to improving current CSOC protocols, said Riviello.
“While the ethical principles should remain the same, planning for inadequate ventilators is actually different from planning for inadequate staffed hospital beds,” she explained. “We would like to plan with perfect data on how patient outcomes will be impacted, but even with studies like ours, we can only be guided by what prior data shows; given differences in each crisis, we cannot know with certainty what system will produce the most equitable outcomes,” she said.
As for additional research, “We need to know the health outcomes of different proposed systems of prioritization for scarce resources,” said Riviello. “This work can be based on retrospective data collected for planning purposes, as in our study, as well as modeling to predict outcomes in future crises,” she said. “As we note in our paper, citing the work of Ibram X. Kendi, ‘Antiracist policies are not defined by the ethics or intentions underlying them, but are defined as those that produce or sustain racial equity,’ ” Riviello emphasized.
Inconsistent Studies Call for Better Assessment
“These findings add to a growing yet inconsistent evidence base pertaining to CSOC equity,” writes Hayley B. Gershengorn, MD, of the University of Miami, Florida, in an accompanying commentary. Previous studies of COSC policies are conflicting as well; some show no clear differences in priority for patients of different races or ethnicities, whereas others show the opposite. Cohort definitions, timing of priority assessment, and CSOC policies vary widely, she said.
Gershengorn emphasized three “truths to which our best approach [to CSOC] must adhere.” First, CSOC policies should not be abandoned; they are needed in order to manage resources, and they should be revisited once the COVID-19 pandemic wanes she said. In fact, “many efforts to establish equitable CSOC policies were taken up during the epidemics of H1N1 influenza in 2009 and Ebola in 2014 and subsequently abandoned,” she noted. Second, “we must do all we can to enhance the likelihood that CSOC policies will not exacerbate disparities,” Gershengorn said. Third, CSOC policies must be “feasible, useful, and acceptable to all stakeholders,” she emphasized.
The COVID-19 crisis must not prompt a further crisis of inequitable resource allocation, Gershengorn concluded; “our patients, their families, and we ourselves deserve better.”
The study was supported by the Agency for Healthcare Research and Quality. The researchers have disclosed no relevant financial relationships.Gershengorn has received grants from the National Institutes of Health/National Heart, Lung, and Blood Institute and various personal fees from Gilead Sciences, Inc, the Annals of the American Thoracic Society, and the SE Critical Care Summit, all unrelated to the current study.
Heidi Splete is a freelance medical journalist with 20 years of experience.