From news anchors to medical experts put on camera every evening, it seems everyone has an interest in identifying the profile of who will and won’t become extremely ill (or worse) if infected with COVID-19, the disease caused by coronavirus.

At present, no one knows for sure. But more information is becoming available daily and researchers are working hard to find answers. An international research team has designed a computer program that can predict, with up to 80% accuracy, which COVID-19 patients will develop serious respiratory disease.

Developed by U.S. and Chinese researchers, the artificial intelligence (AI) program has been tested at two hospitals in China with 53 patients who were diagnosed in January with COVID-19. The new tool is still considered experimental and is now in its testing phase.

The goal of this program is to help doctors make the best use of limited resources by identifying early on which patients will likely need hospitalization versus those who can be sent home for self-care.

Dr. Megan Coffee is a clinical assistant professor of infectious disease and immunology at NYU Grossman School of Medicine in New York City. She is a co-author of the study. "Of those who have symptoms, 80% — maybe up to 85% — will have mild disease; around 15% to 17% will have severe disease and need to be hospitalized; and a further 3% to 5% will need intensive care, usually due to Acute Respiratory Distress Syndrome [ARDS]," said Dr. Coffee.

ARDS is a potentially deadly complication where fluid leaks into the lungs, making breathing increasingly difficult. At least two-thirds of COVID-19 patients who go on to need treatment in an intensive care unit develop ARDS, said Dr. Coffee, which is the "underlying process leading to death in many of the cases."

The goal of this study was to develop an artificial intelligence version of a "master clinician." In other words, a very experienced doctor dealing with a well-known disease. Working with researchers at two research hospitals in Wenzhou, China, the NYU research team developed a computer model based on the kind of "predictive analytics" used to forecast stock market activity and voting patterns.

Researchers fed the program important patient information, including results of lung scans and blood tests, muscle ache and fever patterns, immune responses, age and gender.

Researchers were surprised to find that the factors most clinicians would likely focus on, such as lung status, age and gender, were not all that helpful in predicting outcomes. But what was?

The most accurate predictors were small elevations in a liver enzyme called alanine aminotransferase (ALT), deep muscle aches and higher levels of hemoglobin (a protein that facilitates blood transport of oxygen throughout the body). "That's the value of this approach," Dr. Coffee said, "to look for what we, as clinicians, might not notice."

While the program is still in testing and needs to be validated on much larger populations, she said it would be easy to roll out if future testing results show similar accuracy. Study findings can be found online in the March 30 issue of the journal Computers, Materials & Continua.