September 26, 2022 | Erin Bluvas, email@example.com
The National Institute of Allergy and Infectious Diseases has awarded more than $400K to health services policy and management assistant professor Chen Liang as part of the Long COVID Computational Challenge. Liang, along with partners at the South Carolina SmartState Center for Healthcare Quality and UofSC Big Data Health Science Center, will use the two-year grant to develop an innovative graph model of electronic health records (EHR) and a hybrid machine learning algorithm for identifying and characterizing individuals with long COVID.
“Preliminary studies in China, Europe and the United States estimate that 47 to 87 percent of patients experience long COVID symptoms, such as fatigue and chest pain, 3-6 months after the initial acute phase of infection,” Liang says. “As the U.S. approaches 100 million known COVID cases, we can expect that millions of Americans could potentially experience long COVID.”
As the U.S. approaches 100 million known COVID cases, we can expect that millions of Americans could potentially experience long COVID.
-Chen Liang, assistant professor of health services policy and management
The current and future health problems associated with long COVID pose a significant challenge for clinicians and health systems in caring for this population. This profound public health impact is further complicated by the fact that there are no clear standards for identifying individuals with this condition.
This new challenge from the National Institutes of Health was established to rapidly improve our understanding of long COVID. The initiative relies on the expertise of scientists like Liang, who can leverage big data sets and advanced methodologies to uncover patterns and trends that researchers were not able to capture during the frenzied early stages of the pandemic.
UofSC’s role in this nationwide project will be to address one of its most urgent questions: How do we identify individuals with long COVID?
“Timely identification of long COVID is critical – both for individuals who are part of an existing COVID-19 cohort and for those who have been newly infected,” Liang says. “Doing so will help us better understand transmission, biomarkers and other risk factors, and the different types of long COVID as well as how to better predict outcomes.”
The team will use advanced biomedical informatics methodologies to analyze data pulled from EHR, community-based health services data, medical claims and other sources. They will borrow this big data set from their other projects, which established a statewide South Carolina COVID-19 cohort.