Purdue team excelling in health-systems improvement competition

Nicole Adams, Nan Kong, and Baijian Yang
Team members Nicole Adams, clinical associate professor of nursing; Nan Kong, professor of biomedical engineering; and Baijian Yang, professor of computer technology.

WEST LAFAYETTE, Ind. – Purdue University researchers are top contenders in a national competition to provide new tools that help health care providers systematically address social factors that negatively affect health.

Just 15 of the original 134 competing teams are still in contention following two rounds of the three-round Health Resources & Services Administration competition. HRSA is a federal agency whose mission is to provide equitable health care to the nation’s highest-need communities.

The HRSA competition challenges teams to create, develop and demonstrate solutions that “will support health centers to enhance access to primary care and to improve care coordination with other local providers and social service organizations to address social determinants of health.”

Social determinants of health, as defined by the Centers for Disease Control and Prevention, are nonmedical factors that greatly influence health outcomes, such as where and how people are born, grow, work, live and age. These include socioeconomic status, education, race, housing and other nonmedical factors that research indicates contribute around 80% to a person’s health, while medical care accounts for just 20%.

Purdue team members in the competition have combined their unique disciplinary strengths to create and hone a software solution with the working title, “Bridges,” which is short for “Bridges to Community Resources.” They are Nicole Adams, clinical associate professor of nursing; Nan Kong, professor of biomedical engineering; and Baijian Yang, professor of computer technology.

“We’ve created a software system that helps case managers in primary care clinics assign or refer patients to services,” Adams says. “It also will track where they’ve been referred and if they get the service they need. And we are going to use machine learning to continually improve the recommendations of services that we measure. Using all of that data, the system will continually learn which services should be assigned first to get the biggest improvements and health outcomes. The software gives us the potential to create the first national-level, unique data set on social determinants of health.”

Adams adds that data gathered by the Bridges software promises to provide critical information that’s unattainable today. “If I want to look at how social determinants of health influence the outcomes of people with diabetes, I can’t find data to do that analysis,” Adams says. Bridges will first gather this kind of data throughout Indiana before expanding to the national level.

Kong says the software’s machine-learning component must be developed over time because it requires a critical mass of data.

“We are in a good position to develop an offline trained general model to start with,” Kong says. “This process will also be facilitated by careful mechanistic modeling of patient referrals and transitions. The offline trained model will be periodically recalibrated with state-of-the-art transfer-learning techniques.”

Kong adds that this work will contribute to a social well-being data center at Purdue. “This data center can put Purdue on the map of community health and well-being management analytics,” Kong says. “It will provide a central point of access for community health and well-being data from many community health centers in Indiana and around the nation and provide valuable data sources for many communities aiming to reduce health disparities.”

Phase three of the HRSA competition ends July 28 with a top prize of $150,000. Adams says that regardless of the outcome, the Bridges project will continue in partnership with the Regenstrief Center for Healthcare Engineering.

“We will be able to help federally qualified health centers better refer their patients to social services and give them a way to track where people are getting or not getting services,” Adams says. “This will give health centers tools to make better decisions for patients while improving patients’ health by addressing their social needs.”

Writer/Media contact: Amy Raley, araley@purdue.edu

Sources:
Nicole Adams, adams417@purdue.edu
Nan Kong, nkong@purdue.edu