Lerner Research Institute News
Read about the latest advances from Lerner Research Institute scientists, including new findings, grant awards, innovations and collaborations.
Cleveland Clinic researchers have developed an algorithm to predict which COVID-19 patients are at highest risk for becoming seriously ill or dying from the disease. This prediction model will help physicians and healthcare systems efficiently allocate resources, including COVID-19 vaccines.
“As the number of new COVID-19 cases climbs higher every day, it is critical to develop an informed strategy to help physicians identify who should be treated most aggressively, including with forthcoming vaccines, and who should be most protected from settings where social distancing is challenging, including classrooms,” said Michael Kattan, PhD, chair of the Department of Quantitative Health Sciences and lead author on the study.
Published in Critical Care Explorations, the algorithm—the latest in a suite of COVID-19-related risk prediction models from the Cleveland Clinic team—helps to identify those individuals who are at greatest risk for severe symptoms and disease complications and should be prioritized for swift and aggressive treatment.
The newest prediction model was developed using data from more than 7,600 patients from Cleveland Clinic’s COVID-19 registry who tested positive for the disease from early March to mid-July. By comparing patient outcomes over time, the team’s mathematical algorithm identifies those at high risk for COVID-19-related admission to the intensive care unit (ICU) or death in the two weeks immediately following a positive test.
While the model can be used to identify those at high risk, the data more broadly show that 91 percent of patients who test positive for COVID-19 will not develop severe disease that requires ICU care or results in death.
These study findings will help healthcare systems responsibly allocate a host of resources, ranging from personal protective equipment to ventilators and care providers’ time. Perhaps of greatest interest, however, will be how the algorithm can be used to help inform to whom and how vaccines should be administered, particularly as pharmaceutical companies begin reporting promising early results from COVID-19 vaccine clinical trials and shots begin going in arms.
“Our research suggests that considering who is at risk for serious, potentially lethal disease and not just who is most likely to transmit the virus may be an effective approach to allocate resources and do the greatest good with what we anticipate to be a limited number of vaccine doses,” said Dr. Kattan. “To our knowledge, ours is the first peer-reviewed research to suggest such an approach.”
It is important to note that although the model was based on and validated using Cleveland Clinic COVID-19 registry data, which includes patients from both northeast Ohio and Florida hospitals, a critical next step will be to validate the model with other data sources.
This algorithm is the third prediction model developed by the Cleveland Clinic team, including Dr. Kattan and his collaborator Lara Jehi, MD, Cleveland Clinic’s Chief Research Information Officer. The first model, which was first published in a June edition of CHEST, predicts an individual’s likelihood of testing positive for COVID-19. Notably, that model is now available to health systems around the world through Epic.
This study was funded internally and with support from the National Institutes of Health, through the Clinical and Translational Science Collaborative of Cleveland. Dr. Kattan holds the Dr. Keyhan and Dr. Jafar Mobasseri Endowed Chair for Innovations in Cancer Research.