Addressing racism in health services research

At the 2020 Academy Health Research Meeting, presenters dug deep into the subject on everyone’s minds: How do we tackle structural racism within health services research? They discussed how researchers can include and explain racial disparities within their research, and how leadership can reduce structural racism within institutions themselves.

Measuring racism, not just race

It has become much more common for race to be included as a variable in health services research. However, it is not enough to merely include race as a variable; we must be thoughtful about explaining what it is we are actually measuring.

Researchers that find racial disparities in their results need to make it clear to readers that differences are not due to biological factors but environmental and sociological factors created by racism. Without this clarification, it is easy for readers and the media to perpetuate the idea that genetics or behavioral factors are to blame, not racism. For example, an Ohio lawmaker speculated that racial differences in rates of Covid-19 were because “‘the colored population’ did not wash their hands as well as other groups.”

“Leaving things unsaid in your research reporting leaves it ripe for racist inference,” said Dr. Ruha Benjamin, sociologist and an Associate Professor in the Department of African American Studies at Princeton University, and a keynote speaker at the Academy Health conference.

“Leaving things unsaid in your research reporting leaves it ripe for racist inference.”

Dr. Ruha Benjamin, Associate Professor in the Department of African American Studies at Princeton University

Linda Goler Blount, President and CEO of the Black Women’s Health Imperative, agreed. “Describing differences in data by race is fine, so long as we are not leading the reader to think, ‘It’s because they’re black,'” said Blount. “Race is not a risk factor; racism is a risk factor!”

Rather than just recording racial differences in the data, researchers should strive to capture the results of racist structures that lead to these differences. “Think about proximate causes. If you think there will be racial disparities, think about the causal factors that are behind that instead of just measuring race,” said Steven Brown, Research Associate at the Urban Institute.

“Race is not a risk factor; racism is a risk factor!”

Linda Goler Blount, President and CEO of the Black Women’s Health Imperative

The same considerations should be given when thinking about risk adjustment. There is a current debate in health policy about whether provider reimbursements should adjust for measures of social risk, and whether these should include race. On the one hand, clinicians that care for patients dealing with the health effects of racism should not be financially punished; but on the other hand, there is a fear that adjusting for race would let clinicians off the hook for treating patients of color poorly.

Benjamin suggested looking behind the binary of either including or not including race in the risk adjustment. “What would it look like for the inputs to reflect how providers treat patients? How do patients feel seen and heard by providers?” she said. “Data needs to capture the relationship, not just the racial identities.”

Including race in research tools without understanding why disparities occur can lead to biased algorithms that reproduce inequities. But removing race altogether (making an algorithm “race neutral”) can still be discriminatory. One example is an algorithm that was widely used to allocate health care to patients based on their supposed medical need, but used cost as a proxy for need, which made Black patients seem less sick than white patients. This is what Benjamin calls “The New Jim Code” (a variation on Michelle Alexander‘s book, The New Jim Crow), referring to technology that appears more fair by bypassing human decision-making, but in fact bakes inequality further into the system.

Look inward for change

Presenters stressed the importance of examining structural racism in one’s own workplace, as a way to fight racism. There is a tendency within medical and research institutions to deal with topics of race without looking at evidence, despite being “data-driven” on every other issue.

“When it comes to addressing racism in research, why do we go by ‘feelings’?” said Dr. Sherilynn Black, Associate Vice Provost for Faculty Advancement at Duke University. “We need to treat this topic with the rigor it deserves.”

Measuring racism within institutions is critical toward reducing it, but too often there is no attempt to even measure things like who gets hired, who gets promoted, which projects get funded, etc. “We need to set targets and have the benchmarks come from the top,” said Brown. “Accountability structures are fundamental.”

“We recognize there are differences in funding rates, applicant pools, differences in the way review occurs for people of difference backgrounds.”

Dr. Nakela Cook, Executive Director at the Patient-Centered Outcomes Research Institute (PCORI)

Presenters also identified ways in which institutions can tackle racism inside their walls, beyond just “Diversity and Inclusion” trainings, which don’t change the fundamental structure of institutions. Looking at board composition, project funding, the application and review process, the tenure and promotion process, committee composition, event planning, and mentoring and sponsorship, are all potential avenues for change.

Alan Weil, editor-in-chief of Health Affairs, noted that there is a bias in medical research toward quantitative results, but that leaves out important research. He plans to broaden the types of methods they cover in the journal, to break down this barrier.

Similarly, Dr. Nakela Cook, Executive Director at the Patient-Centered Outcomes Research Institute (PCORI), acknowledged that inequities are prevalent in the way we fund public health research. “We recognize there are differences in funding rates, applicant pools, differences in the way review occurs for people of difference backgrounds,” said Cook. Cook also identified opportunities to “experiment in the application evaluation process” and “elevate communities as partners in research.”

And of course, it is incredibly important to recognize that the solution is not individual, but structural. Our academic institutions were not designed to help everyone succeed; they were originally designed to exclude, said Black. “We should not try to fix people to fit a racist structure; we should fix the racist structure,” she said.