LITTLE ROCK -- Clare Brown, assistant professor in the University of Arkansas for Medical Sciences Fay W. Boozman College of Public Health, has received a K01 grant from the National Institute on Minority Health and Health Disparities.
Brown will use the award, which is worth $496,000 over four years, to conduct the study, "Algorithmic fairness in predictive models to eliminate disparities in adverse infant outcomes: A case for race."
"Getting this award is a major honor for a researcher because you must convince a panel of peer researchers that you're worthy of receiving the award," she said in a UAMS news release.
"For the next four years, this grant will allow me to focus my research specifically on disparities in adverse infant and maternal outcomes in relation to race. Because this is a training grant, a portion of my time will also include training and learning more about the cultures of Black, Hispanic and Marshallese women and ways to reduce adverse infant outcomes."
Brown's study will have two primary aims, the release said.
The initial aim will focus on creating predictive algorithms for low birthweight births using the Arkansas All-Payers Claim Database, which has insurance claims and birth certificates for infants in Arkansas.
"I'll create algorithms to predict low birthweight and will run tests to see if using racial and ethnic information helps with algorithmic fairness," she said.
Brown's second aim will address when -- and how -- insurance companies collect race and ethnicity data. That effort will include interviewing women from three racial/ethnic subgroups in Arkansas.
"We'll have focus groups of Black, Hispanic and Marshallese women," Brown said. "The info from the focus groups will help us determine when race and ethnicity data should be collected and what the supportive data should be used for."
Brown said that most insurance companies do not collect racial and ethnic information, which makes it difficult for the industry to recommend methods that can reduce racial and ethnic health disparities in infant and maternal outcomes.
"Insurance companies may be hesitant to collect race data because of fears of being blamed for redlining or charging higher premiums to populations of minority race and ethnicity," she said. "Meanwhile, minority populations might be fearful of providing racial and ethnic information to insurance companies because many populations may be fearful of racial targeting."
"In Arkansas, adverse infant and maternal outcomes are abnormally high for Black infants and their mothers. However, this race-related health outcome is not a new development. In fact, Brown's interest in addressing the issue dates back to 2014 when she was a student in the college pursuing a Master's in Public Health.
"For that reason, not only is Brown excited about the K01 grant -- but she's also enthusiastic about the chance to focus on creating the most predictive and fair algorithms for predicting low birthweight birth. Meanwhile, she'll also create guidelines for insurance companies about the best and most respectful ways to collect race and ethnicity information," the release said.
"By identifying who's at risk, we can help reduce those outcomes," she said. "The United States and Arkansas specifically have large health disparities in infant and maternal outcomes -- and this is the only way I know to help. I'm not a clinician. However, by doing this research I can help provide data to clinicians, health policymakers and to program officials that can help reduce adverse outcomes and disparities. This is a truly great opportunity."