Experts in data science and health shared expectations on Nov. 1 for the newly established Advanced Research Projects Agency for Health (or ARPA-H), including that it should incorporate social determinants of health into its research and think about a reusable data structure.
The Department of Health and Human Services (HHS) announced the official launch of ARPA-H, which sits within the National Institutes of Health (NIH) as an independent entity, in May of this year. President Biden made ARPA-H official through the fiscal year 2022 omnibus appropriations bill, which included $1 billion to create the new agency.
President Biden appointed Dr. Renee Wegrzyn as the agency’s inaugural director in September, but experts this week agreed the agency has much to consider as it begins its research.
“We have the unique opportunity in this country to fund research that will be the high-risk funding,” Dr. Bill Kassler, chief medical officer at Palantir Technologies, said during an ACT-IAC event on Nov. 1.
“My worry is that ARPA-H will solely focus on the biomedical side and miss an opportunity for the biopsychosocial side to incorporate social determinants of health, to incorporate the socio-economic factors, the racism, the stigma,” he added. “If ARPA[-H] only focuses on the biomedical, it will be addressing 20 percent of the contribution to healthcare. So, ARPA-H must take a broader view.”
Additionally, Dr. Kassler said the Federal government must also consider the role data science will play in relationship to the new agency, and stressed that “ARPA-H needs to think about reusable science.”
He went on to explain that oftentimes, researchers will put a lot of work into a study only for their published paper to sit on a shelf. Dr. Kassler said these papers are public assets that need to be “reused and repurposed, sometimes for equity research.”
“ARPA-H needs to think about a data infrastructure that allows that public asset of research to be more widely used and disseminated,” he said. “It needs to think about research beyond biomedical research. And then it needs to think about how that data can be used for operational excellence within the agency, how the agency can deliver these funds fast, fail fast, [and] repurpose to other promising types of studies.”
Dr. Belinda Seto, Ph.D., deputy director of the Office of Data Science Strategy at NIH, agreed with Dr. Kassler and said that approach to data science is known as the “FAIR” principle.
Within the FAIR acronym, Dr. Seto said, “F is findable, A is assessable, I is interoperable, and R is reusable. And that’s really a primary principle for data science.”
“For me, the important point for health equity is that these are approaches, policy decisions made based on data science – we shouldn’t just be throwing darts in the dark,” she added.
Dr. Seto also explained that the idea for ARPA-H is based on the Defense Advanced Research Projects Agency (DARPA), which is a research and development agency within the Department of Defense.