New Analytics Tools Predict COVID-19 Patient Mortality Posted on : Sep 10 - 2020

One of the most urgent needs in the care of patients with COVID-19 is a better understanding of which patients will require more intensive treatment and attention. Now, researchers from Oklahoma State University’s Center for Health Systems Innovation (CHSI) are applying big data analytics to build predictive models of COVID-19 patient risk that could help physicians better manage patient care during the pandemic.

Zhuqi Miao (the health data science program manager at CHSI) and Meghan Sealey (a doctoral student studying statistics at Oklahoma State) worked with anonymized data from nearly 19,000 COVID-19 patients from healthcare IT firm Cerner’s COVID datasets. Using this data, they developed two tools for modeling mortality risk: one based on patient data at time of admission, and one based on patient data from the first data of hospitalization.

“The models identified a similar set of medical conditions suggested by the Centers for Disease Control and Prevention as the essential risk factors for death, such as history of diabetes, respiratory disorders and hypertension, and onset of respiratory or kidney failures,” Miao said, “but we also found some unique ones.”

Using these tools, the researchers say they are able to accurately predict mortality for almost 70% of patients (with the first model) and nearly 75% of patients (with the second model). The team sees a slew of benefits in store for the healthcare industry if it were to deploy such tools in the field. View More