Skip to main content

News and Media

Open Main MenuClose Main Menu
Researchers at the OSU Centers for Health Systems Innovation have created a model of COVID-19 mortality using health data analytics.

Parsing COVID-19 data

Friday, October 23, 2020

Spears researchers use data analytics to model risk factors

An analysis of digital health records from thousands of COVID-19 patients admitted to hospitals around the country has led to the development of technology to predict the risk of death for those with COVID. Researchers at the Oklahoma State University Center for Health Systems Innovation (CHSI) in Tulsa and Stillwater are using digital health data to build predictive models of disease risk that could improve treatment and outcomes.

Dr. Zhuqi Miao
Dr. Zhuqi Miao

“There is an urgent need to determine which COVID patients are at highest risk for bad clinical outcomes, as early as possible, so that plans and actions can be made to save more lives,” said Dr. Zhuqi Miao, the health data science program manager with CHSI.

Miao and Meghan Sealey, an OSU statistics doctoral student, analyzed anonymous data of 18,742 hospitalized COVID-19 patients from the Cerner COVID-19 data cohort, a collection of HIPPA-compliant hospital and clinic records donated to CHSI by Cerner for research.

The CHSI researchers created two models of potential mortality risk, one based on patient data at the time of admission using demographic and historical medical conditions, and the second at the end of the first day of hospitalization using demographics, procedures, medications and known conditions.

“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 a history of diabetes, respiratory disorders and hypertension, and onset of respiratory or kidney failures, but we also found some unique ones,” Miao said.

Dr. William Paiva
Dr. William Paiva

The models show the potential of using predictive data analytic technology, such as risk models, to identify patients who are at highest risk of death if not treated appropriately and promptly. This allows health care organizations to focus their resources on those patients who need it most.

“These kinds of analytic tools are the wave of the future to diagnose, stage and monitor disease progression and save lives. It can also help alleviate the financial burdens for both patients and healthcare systems alike,” said Dr. William D. Paiva, CHSI executive director.

CHSI operates at the intersection of the OSU Spears School of Business and the OSU Center for Health Sciences and focuses on business, information technology and clinical innovations to improve the delivery of health care.

Story By: Jeff Joiner | Discover@Spears Magazine

 MEDIA CONTACT: Terry Tush | Director of Marketing and Communications | 405.744.2703 |

Back To Top
SVG directory not found.