A report by the nonprofit Do No Harm has revealed that a study from 2020 alleging white doctors exhibit “spontaneous racial bias” leading to worse outcomes for black newborns continues to serve as a pillar of diversity, equity, and inclusion (DEI) efforts, despite major methodological flaws.
The study, published in the Proceedings of the National Academy of Sciences (PNAS), concluded that the disparity in mortality rates between black newborns and white newborns dropped by 58 percent when the black babies were treated by black physicians.
However, the paper drew criticism after it was revealed by the Manhattan Institute that the co-authors had made a methodological error, which revealed that controlling for very low birth weight newborns nullified the effect of racial concordance on infant mortality. According to Do No Harm, the authors buried that data in the appendix. Additionally, one of the study’s co-authors left her university in April amid allegations of plagiarism and incompetence from her staff.
Despite the study’s flaws, Do No Harm says it remains influential, as PNAS has not retracted the study. It has been cited to support affirmative action in medical school admissions and within physician DEI training. The study even appeared in an amicus brief filed by dozens of medical associations led by the Association of American Medical Colleges in the Students for Fair Admissions v. Harvard case. Associate Justice Ketanji Brown Jackson cited the study in her dissenting opinion.
Ian Kingsbury, director of Research at Do No Harm, has argued that the study is an example of how scientific literature has used improper science to promote a certain policy goal.
“This was holding up much of the DEI enterprise. It’s the foundational study in racial concordance mythology,” Kingsbury said, according to the Daily Caller.
Do No Harm’s report also noted that several prestigious journals have cited the study to support the claim that DEI policies improve clinical outcomes, including the American Academy of Pediatrics’ journal Pediatrics, Academic Medicine, the Journal of Neurosurgery and the Journal of Graduate Medical Education.
“You can see the ideological bent of this," Kingsbury added. "For some reason it’s completely acceptable for journals – in order to make the claim that racial concordance is beneficial – to cite one or two bad studies, ignoring the weight of the evidence."
