![Using an algorithm compiled from thousands of Cleveland Clinic patients, researchers developed a tool that helps predict whether a patient may test positive for the coronavirus. A nurse holds a swabs and test tube kit to test people for COVID-19, the disease that is caused by the new coronavirus, at a drive-thru station set up in the parking lot of the Beaumont Hospital in Royal Oak, Mich., Monday, March 16, 2020. [Paul Sancya / AP]](https://npr.brightspotcdn.com/dims4/default/b1101d2/2147483647/strip/true/crop/3000x2004+0+0/resize/840x562!/format/webp/quality/90/?url=http%3A%2F%2Fnpr-brightspot.s3.amazonaws.com%2Flegacy%2Fuploads%2F2020%2F6%2F16%2FAP_20076696305899_0.jpg)
Cleveland Clinic researchers have developed a new tool that can help predict a patient’s likelihood of testing positive for the coronavirus and becoming seriously ill from it. They said the prediction model, or nomogram, is the first of its kind.
The prediction is based on multiple factors, such as a patient’s age and race, as well as their medications and vaccine history.
“Without a tool that combines all of those different risk factors for any given patient, it is difficult for a physician to know whether a patient in front of them is at high or low risk,” said Cleveland Clinic’s Chief Research Information Officer Dr. Lara Jehi, who co-authored the study.
Jehi said the tool can help providers make more intentional decisions about patient care and better allocate resources.
“We’re going to use it to help us identify those who are at [a] higher likelihood of having it so that we can follow them up more closely, instead of sending them home and telling them, 'well, come back to us if your symptoms get worse,’” she said.
Jehi also said the model can help experts make more informed policy decisions while COVID-19 testing is still limited.
“As we talk about reopening and getting people back to work, short of universal testing, you won’t really know who has the disease and who doesn’t,” she said. “This tool could help identify the people who are at high risk of having it, so it could guide some of those decisions.”
The model is based on data collected from nearly 12,000 Cleveland Clinic patients who were tested for the coronavirus in March and April. Researchers are also using this data to follow up on some new observations made during the study. For example, patients taking medications such as melatonin and over-the-counter sleep aids were less likely to test positive than patients not taking it, and researchers want to explore if this means anything.
The tool is now in use at the health system, and the online calculator is available for anyone to use. Jehi said at this point, however, the nomogram is most useful for medical providers for decisions related to patient care.
“If they have a tool like this that can help strategize with how these resources should be allocated in a scientific way … then that would be a much more informed way of delivering care, than us making our best guess, which is what we have been doing so far,” she said.
The study was recently published in the peer-reviewed journal Chest.
