Since 2007, ob-gyns have used a calculator to predict how likely it is for patients who’ve had cesarean births to subsequently have healthy, complication-free vaginal births (called “VBAC”). These predictions are based on a number of different variables. For a long time, one of them was race: The calculator was more likely to dissuade doctors from recommending vaginal birth to Black and Hispanic patients, compared to white ones.
Vaginal births are generally considered safer than C-sections for mothers. While we don’t know exactly how, or if, the calculator influenced maternal health outcomes, Black women are still three times more likely than white women to die from pregnancy-related causes. So the idea of a widely used clinical tool effectively pushing nonwhite patients into a higher-risk labor scenario didn’t sit well with many experts. But they secured a win this summer, when various medical institutions removed race from the VBAC calculator.
The underlying reason for the change is that, unlike hereditary disease or personal health history, race is not a medically meaningful detail. But race made it into the calculator because becoming a mom has statistically been more dangerous for nonwhite women, specifically Black women.
“This disparity is not because of some inherent genetic or biological factor; it’s a reflection of the systemic racism that affects many obstetrical outcomes,” says gynecologist Dr. Veronica Ades, an associate professor at NYU’s Grossman School of Medicine. “By including race, we are further entrenching and augmenting that systemic racism.”
The same argument applies to a number of other clinical “calculators.” Sometimes called diagnostic algorithms or risk predictor tools, these calculators help doctors and other providers make treatment decisions based on up-to-date guidelines — and it’s not uncommon for race to be a variable. This practice, known as “race correction” or “race norming,” is the subject of a decades-long debate receiving renewed interest because of the VBAC calculator revision. While a growing, vocal group of critics see the practice as outdated and misleading, others say race is an imperfect but useful way to account for broader health trends rooted in genetics and ancestry.
“Dropping race correction will have effects on medical practice: Any time you move a threshold, some people will be helped, others will be harmed,” says David Shumway Jones, a professor of epidemiology at the Harvard T.H. Chan School of Public Health. “But there’s not much reason to believe that the current race corrected thresholds are the right ones.”
The tip of the iceberg
By itself, race doesn’t tell you much about a person’s biology. This fact is essential to the argument against using race in clinical calculators. In recent years, scientists have learned a lot about the human genome and how it works. As it turns out, race is only a tiny part of the equation. Two people of European descent, for instance, might have more in common genetically with someone of Asian descent than with each other.
Most health conditions stem from a complicated mix of genetics and ancestry, says Giorgio Sirugo, a geneticist at the University of Pennsylvania. “Race, which is broadly used in medicine and by medical sciences, is really a bad way to categorize ancestry,” he says. “Ancestry is such a melting pot.”
In medicine, it’s common to use race as a proxy to show how ancestry or sociodemographic factors like poverty affect someone’s risk of developing a medical problem. In the past, researchers have often used race corrections when they can’t quite pinpoint what’s causing a health trend or they need a simple way to bucket a group of people together.
Race has appeared in at least 13 clinical calculators over the years. Here are four that either currently include race or used to.
The tool: The estimated glomerular filtration rate, or eGFR Calculator, helps clinicians assess kidney filtration in people with chronic kidney disease. This calculator, which divides patients into the categories “nonblack” and “Black,” tends to indicate that Black people with chronic kidney disease have better functioning kidneys than their white counterparts. This means they’re less likely to receive important treatments early on.
Rationale for using race: The calculator’s race correction stems from a 1999 study that drew on data from the 70s and 90s. Researchers concluded that Black populations naturally have more muscle mass, and thus higher kidney function, than white people.
The debate: A growing number of health systems have eliminated race from this tool. They say the original race correction was based on insufficient evidence from flawed, outdated studies, and argue that sociodemographic differences, not race, likely shaped the data.
University of Washington Medicine, one of the first health systems to update the calculator, says including race can lead to disparities such as longer times on kidney transplant waiting lists for Black patients. This past March, the presidents of the American Society of Nephrology and the National Kidney Foundation urged clinicians to replace race with a “substitute that is accurate, representative, unbiased, and provides a standardized approach to diagnosing kidney diseases.”
The tool: Doctors use the Get with the Guidelines–Heart Failure Risk Score, developed by the American Heart Association, to determine mortality risk in patients admitted to the hospital with heart failure. The tool categorizes patients as “Black” or “nonblack” and associates being Black with lower mortality rates.
Rationale for using race: The calculator was created using AHA patient data from 2005 to 2007. It’s unclear why race is a variable, as current data directly contradicts the race-mortality rate association in the algorithm.
The debate: Recent studies say being Black puts people at higher risk of death from heart failure. Many say this algorithm promotes disparate care. Here’s one illustrative example: A 2019 study found that Black heart-failure patients were less likely than white ones to be admitted to a Boston ER.
In a November 2020 letter, the American Association of Black Cardiologists urged Congress to exclude race from heart health algorithms in general: “Algorithms learn from historical patterns to make predictions and decisions, but if they learn from biased data, they will produce biased outputs.”
The tool: In the 1990s, psychologists developed Heaton norms to determine patients’ baseline levels of cognitive functioning. Clinicians calculate this number when patients have cognitive impairment and they need to figure out if the cause is brain injury or disease. Heaton norms correct for race by automatically assuming Black patients have lower cognitive ability than white ones.
The rationale: In this equation, race is used as a placeholder for sociodemographic factors, such as education inequities and poverty, that disproportionately affect nonwhite populations and can interfere with cognitive development.
The debate: Defenders of Heaton norms say they prevent healthy Black people from being diagnosed with brain disease. But other healthcare professionals say they’re “a crude proxy for lifelong social experience” that tend to underestimate cognitive capacity in Black people. Not to mention, the norms are based on one small sample of Black people in San Diego.
Currently, we’re seeing a fierce debate over Heaton Norms unfold in real time between NFL and some former players. In 2013, the league settled a class action lawsuit with players who said they weren’t protected from the danger of head injuries on the field. But, to qualify for settlement money, players had to be diagnosed with a certain level of dementia. The league used Heaton norms, which made it harder for Black players to qualify for payouts. In 2020, a few players filed suits alleging discrimination.
The tool: FRAX helps clinicians identify the likelihood of a patient developing an osteoporosis-related bone fracture. It includes race corrections for Asian, Black and Hispanic women by classifying them as lower risk.
The rationale: In general, women have a greater risk of osteoporosis than men. But, in 1992, the WHO identified an osteoporosis epidemic and found fewer cases reported by women of color. So it added a race correction to prevent unnecessary treatment in women.
The debate: Several studies have shown that Black women are less likely to be treated for osteoporosis than white women, and are more likely to die or become disabled following osteoporosis fractures. Proponents of removing race from the tool say it will help clinicians catch osteoporosis in women of color earlier, so they can receive better treatment.
To boycott or not?
A large movement wants to boycott the inclusion of race in these algorithms altogether. “In many cases race can simply be dropped,” says Jones, adding that researchers can create universal algorithms that apply to everyone — with no mention of race.
But many warn we can’t uniformly remove race without carefully re-crafting these equations to account for other risk factors. Jones says we’ll likely want to put more of a focus on social determinants, such as class.
This is what happened with the VBAC calculator. Race was dropped from the equation and history of chronic hypertension — an objective variable — was incorporated into the adjusted calculator.
“I think it will allow doctors and midwives to give more accurate information, and allow patients to make decisions based on stronger evidence,” Ades says. “In reality, this calculator is just one factor in how a patient makes a decision of whether to VBAC. But if we are going to use a calculator, it should be one that doesn’t worsen existing inequities. Patients also consider how strongly they feel about having a vaginal delivery and how many more children they want, among other things.”
Experts caution that while some changes have already happened in theory, it may take time for them to show up in patient care. These formulas have been programmed into many hospitals’ computers for years, and they’re fully embedded into many clinicians’ practices.
Nevertheless, proponents for change are excited to see progress, however incremental, toward rooting out racial bias in healthcare. “The occurrence of race in the tools perpetuates the old assumption that race is real, biological, medically actionable,” says Jones. “I think that’s pernicious, and that alone is a good argument for me to drop race from them.”