During the worst of the COVID-19 pandemic, overwhelmed doctors and nurses in some places have had to make agonizing decisions about which patients should receive scarce health care resources. Some hospitals have contemplated using a standard measurement to guide these decisions. But two new Yale-led studies suggest that the protocol known as SOFA could promote racial disparities in treatment outcomes.
Developed two decades ago to assess illness severity among patients with sepsis, SOFA — Sequential Organ Failure Assessment — was designed to help predict which patients have the best chances of survival. A patient with a high SOFA score, according to the protocol, has a lower chance for survival.
In one of the new studies, published Sept. 16, researchers found that Black patients with COVID-19 admitted to Yale New Haven Health System hospitals typically have higher SOFA scores than white patients do.
The other study showed that while Black patients admitted in the Yale New Haven Health System tended to have higher SOFA scores than those from other racial or ethnic groups, their mortality rate was about the same.
“If we adopt this protocol [SOFA], we could end up disproportionally steering resources away from Black patients and towards other groups,” said Benjamin Tolchin, director of Yale New Haven Health’s Center for Clinical Ethics, and corresponding author of the second paper, published on Sept. 17.
The Yale New Haven Health System is not using SOFA to guide patient care or inform allocation of resources. But other U.S. hospitals have used the scores as a tool for rationing care during the recent surge in COVID-19 cases.
“SOFA scores are already being used to triage COVID-19 patients in Idaho as they experience a surge in COVID-19 hospitalizations,” said Shireen Roy, corresponding author of the Sept. 16 paper and a recent graduate of Yale College (B.S. ’19) and Yale School of Public Health (M.P.H. CDE ’20). “Our study shows that it is extremely important for health care systems to create triage protocols that prioritize equity, both for this pandemic and going forward.”
Both papers were published in the journal PLOS One.
Since the earliest days of the pandemic, some health facilities have been forced to triage patients based on limited supplies. During the first wave of infections in the spring of 2020, some hospitals in northern Italy and New York City lacked the staff and equipment — such as ventilators and hemodialysis machines — needed to assist all emergency cases.
“What happened was that clinicians were making resource allocations decisions on a case-by-case basis, with some preference given to younger patients who were less likely to die from severe infection,” Tolchin said.
While some hospitals began contemplating the use of crisis standard protocols — including those that rely on criteria such as SOFA scores — these protocols had not been used until recently.
Developed in 1996, the SOFA system calculates the mortality of patients based on the function of six organ systems.
“When we looked at this earlier this year, we were thinking of devising crisis standard of care recommendations that might be adopted in future pandemics,” Tolchin said. “We didn’t think we would be needing them so quickly.”
Roy became involved in the research during her 2020 summer internship with Yale’s Equity Research and Innovation Center or ERIC. She continued to work on the project as she completed her MPH degree in her second year at YSPH.
“My training at YSPH was invaluable to me as I worked on this study, for everything from understanding the statistical methods to analyzing the data to writing the paper,” said Roy, who is now a medical student at the Washington University School of Medicine. “I’ve been fortunate to learn valuable skills at ERIC during my MPH studies and I’ve found multiple incredible mentors that have helped me grow, such as Dr. Saria Hassan and Assistant Professor Carol Oladele (a core faculty member at ERIC, who served as one of the papers’ lead authors).”