New research calls into question the high-profile conclusion of the first major study to show that the race of physicians influences health outcomes.
In August of 2020, a study published in the Proceedings of the National Academies of Science made headlines with its finding that Black infants were half as likely to survive to their first birthday when they were cared for by white doctors instead of Black ones.
But a new analysis in the same journal upends that result, showing that the survival difference seen in the original study was almost entirely due to infants with very low birth weight; the race of the physician did not have a statistically significant effect on the mortality rates of the babies they cared for.
While other studies have shown that Black patients are more likely to follow medication guidelines and get cancer screenings and vaccinations when asked to do so by Black doctors, and a 2023 study showed that people of all races in counties with at least one Black physician lived longer, the 2020 paper attracted wide attention.
Published just after George Floyd’s death, the study gained traction with many groups who said it showed the urgent need to diversify the nation’s physician workforce. Just 5.2% of U.S. physicians are Black. The study was cited by Supreme Court Justice Ketanji Brown Jackson in her 2023 dissent in the highly watched Students for Fair Admissions v. Harvard case that severely limited the use of affirmative action in college admissions.
Robert VerBruggen, a co-author of the new research and a fellow at the Manhattan Institute, a right-leaning think tank, said he’d been initially skeptical about the study when he read a CNN article about it. But when he saw how comprehensive the paper was and how many variables it included, he was “at a loss how to explain the findings.”
The 2020 study, which includes the 65 most common contributors to infant mortality, as well as factors such as insurance status and how many very sick babies a hospital treats, has been described as “painstaking” and “meticulous” even by critics.
Verbruggen said he decided to look more closely at the study after it resurfaced in the Supreme Court decision. The new analysis included one variable the original study had not: very low birth weight (babies weighing less than 3.3 pounds.) Such babies are rare — just 1.2% of white babies and 3.3% of Black babies are born so small — but are extremely vulnerable and least likely to survive.
“That’s where the death rates really start skyrocketing,” VerBruggen said. “This is an extremely strong predictor.”
The new study found that very low birth weight babies are both more likely to be Black and cared for by white doctors, explaining much of the original finding that physician race played a role in infant survivorship, he said in an interview with STAT. It was easy to see why the factor had not been included originally, VerBruggen said, as birth weight is not simple to pull from the data. The original authors had used some measures of low birth weight and prematurity, but did not focus on the very lowest weights.
The sickest babies, the new study pointed out, are most likely to be treated in neonatal intensive care units (NICUs), which are largely staffed by white physicians. Just 3.8% of the nation’s neonatologists are Black. The study also noted that it is not fair to attribute the outcomes of babies in NICUs to a single physician since those babies are cared for by large (and presumably multiracial) teams that would override any implicit or explicit bias that might exist in a single clinician.
The new study found that physician race still appeared to play a role in infant survivorship, but that it was not statistically significant. “You can’t say the effect is zero, but it’s much smaller than what they had,” VerBruggen said. He also praised the original study’s lead author for being open to sharing the data and walking him through the original analysis.
That researcher, Ben Greenwood, said he was happy to share his data. “We encouraged them to study this because it’s so important,” said Greenwood, a professor of business at the Costello College of Business at George Mason University. “I don’t think it’s helpful to be defensive about your work.”
Greenwood told STAT that he thought the new analysis still showed that physician race played some role in infant mortality, even if it was smaller than the effect his research found, and that the conflicting findings highlight the need for more research. Further studies using different statistical models or including different variables could come to yet other conclusions on the complex role a physician’s race might play, he said.
“If the real goal is to reduce infant mortality, then what we need to do is make sure people continue to work this out,” he said.
Greenwood said he believed his study has been scrutinized more heavily than other topics in health research because issues of race are so politically divisive. He added that he has not been happy with how the research has been used by either the right or the left, including people who interpreted the finding to conclude that some white doctors are racist.
“The first thing we said in every interview is you shouldn’t select your doctor based on race. There are great doctors and bad doctors of any race,” he said. “We also said you shouldn’t set policy on this.”
VerBruggen is very open that politics — and the role the 2020 study could play in setting policy — was a driving factor in his doing a new analysis. The Manhattan Institute is a “free-market oriented, right leaning” think tank, he said. VerBruggen is not an academic. He is a journalist but said he has taught himself to use statistical-analysis software. For the new research, he teamed up with George J. Borjas, a Harvard economist.
Borjas, a Cuban-American who focuses on immigration, has been involved in several research controversies, including his supervision of a 2009 Harvard dissertation that claimed Latino immigrants in the U.S. are and will remain less intelligent than “native whites.” His work showing that Cuban immigration to Florida reduced wages in the state was cited regularly by the Trump White House and politicians seeking to curtail immigration, though it has been largely disproven by other economists.
Politics aside, Verbruggen said this new study showed that future research on infant mortality needed to include very low birth weight as a factor and should also examine hospital effects — some hospitals have more technology that can save the lives of premature infants. “The racial gap [in infant mortality] is worthy of study,” he said.
Black infants are twice as likely to die in their first year, compared to white infants. Understanding which factors could curb those numbers, and which will not, are critical topics for future research, said Ted Joyce, an economist at Baruch College, in a commentary accompanying the new paper.
Black infants are also three times more likely to be born at very low birth weight than white infants, Joyce noted, numbers that other research shows are clearly tied to socioeconomic status and have not markedly changed in the past three decades.
STAT’s coverage of health inequities is supported by a grant from the Commonwealth Fund. Our financial supporters are not involved in any decisions about our journalism.