August 10, 2020
Over five million Americans have tested positive for COVID-19, and one thing is certain: individuals of certain race and ethnic groups are being disproportionately affected by the virus. These disparities are likely driven by underlying social risk factors, such as healthcare access or housing conditions, which can impact a person’s health.
IUPUI’s Joshua Vest, professor of health policy and management in the IU Richard M. Fairbanks School of Public Health and director of the Center for Health Policy, is working with colleagues at Regenstrief Institute to develop and evaluate technologies that help organizations better understand which patients have social risk factors.
"We are not looking at COVID-19 the disease, but rather how do we better deal with the social factors that increase an individual's risk for getting COVID-19," Vest said. "Understanding how widespread social, economic, and behavioral risk were among those affected by COVID-19 will support communities', organizations', and policy makers' prevention, response and recovery efforts."
Information regarding patients' social factors and risks are typically gathered via surveys, but that process can be a burden for healthcare providers during emergency situations like the pandemic. Instead, Vest and his team are looking to natural language processing, a branch of artificial intelligence which focuses on computational processing of written text. Their system will read clinical notes in electronic health records provided by doctors about the patient's social history and classify that information based on specific social factors, including inadequate financial resources, employment status and food insecurity. That information will then be validated through surveys conducted at Eskenazi Hospital in Indianapolis to test for accuracy of the process. The process will be applied to a larger set of patients who were hospitalized or had visits related to COVID-19 to get a statewide view of issues.
Vest and his team are utilizing data available through the Indiana Health Information Exchange – an electronic network that allows hospitals, public health agencies, health care organizations and more to quickly share patient information in an effort to deliver efficient and high-quality healthcare to Hoosiers. This near real-time data on health and needs is a vital source of information for researchers and other individuals trying to inform policy and decision-making during the pandemic.
The researchers hope this project will help identify the types of needs people will have even after the pandemic, and the information will be vital to assisting in the development of prevention and mitigation strategies.
"A better understanding of social factor data would support ongoing projects like seasonal flu and other projects on health disparities," Vest said. "Ultimately, we want our work to be used even beyond Indiana."
Collaborators on the project include Wei Wu, Suranga Kasthurirathne, Eneida Mendonca, and Katie Allen.
Their project, "Measuring patients' social risk factors to understand infectious disease and health disparities", is one of 17 IUPUI projects to receive funding as part of the OVCR COVID-19 Rapid Response Grant Program.
