Doctors often make life-changing decisions about patient care based on algorithms that interpret the results of the test or the risk of weight, such as to perform a particular procedure. There are some formulaic factors in a person’s race, which means that patients’ skin color can affect their access to care.
A new study of patients in the Boston area is one of the earliest documented causes of harm. It investigated the impact on the care of a widely used but controversial formula to assess renal function, which by design provides healthy scores to black people.
The study analyzed health records for 57,000 people with chronic kidney disease from the Mass General Brigham Health System including the Harvard Teaching Hospital Massachusetts General and Brigham and Women. One-third of black patients out of more than 700 people would have been classified as more serious in kidney disease if their kidney function had been estimated using the same formula as white patients.
This can affect decisions such as sending a person to a kidney specialist or referring them for a kidney transplant. In 64 cases, the patients’ restructured scores would have made them eligible for the kidney transplant waiting list. Doctors did not question race-based recommendations, suggesting no one was referred or evaluated for the transplant.
“It was really shocking,” says Mallika Mendu, an assistant professor at Harvard Medical School and an expert on kidney and kidney in women, as her work on the study led her to stop using race-based calculations with her own patients Assured to “We know that there are already other disparities in the care and management of the condition. It is not helping.”
The study is the most recent of several indications that math tools increase health disparities. Last year, software used by many health systems was used to systematically privilege white patients over black patients to prioritize access to specialized care for chronic conditions. It was clearly not taking account of race, but replicated patterns of access to health care due to factors such as poverty.
In contrast, the renal algorithm is one of several diagnostic decision algorithms that explicitly take care of race. More than a dozen such devices have recently been listed in areas including cancer and lung care. In August, a group of black retired NFL players sued the league, claiming that it uses an algorithm that believes whites have higher cognitive function to decide compensation for brain injuries .
The issue is attracting more attention from federal lawmakers. Representative Richard Neal (D-Massachusetts), chairman of the House Ways and Means Committee, states that the study of the kidney underscores the need for the rethinking of race in all medical algorithms. “Many clinical algorithms can cause delayed or misdiagnosis for patients with black and latex, leading to lower quality care and poor health outcomes,” he says.
Neil calls for medical committees and Medicare & Medicaid Services to examine the impact on patients with clinical algorithms that use race. Last month, Senator Elizabeth Warren (D-Massachusetts) and others asked the Department of Health and Human Services to investigate race-based medical algorithms.
The new study examined a standard calculation called CKD-EPI that is used in blood tests for the level of the waste product creatinine, in the estimation of kidney function called an estimated glomerular filtration rate, or EGFR. Lower scores indicate impaired kidney function; The score is used to classify the severity of a person’s illness and to see what care they receive. Equation factors in a person’s age and gender. Black patients receive an additional 15.9 percent.
The design is under fire from academics and medical residents who fear it discriminates in kidney care. Researchers who created the formula in 2009 added “race correction” to their data to smooth statistical differences between younger patients and others. But that project and subsequent studies have not revealed why the relationship between creatinine and kidney function appears to be different in Black patients, or the role of factors that affect creatinine levels, such as diet, a subsidiary of the University of Pennsylvania Prof. Namamka Enanya also worked on the new Boston study. A person’s race is a social category, not a physical one, she says, and it makes no sense to use it to explain blood tests.