New App Predicts Severity of COVID-19

In order to aid the effort against the malevolent disease which has gripped the world, NYU (New York University) department of dentistry has created an app which uses artificial intelligence to review biomarkers and risk factors, from a single blood test that produces a score from 1-100, where the latter suggests a heightened severity from COVID-19.

It is an improvement on current diagnostic tests, which only attempt to detect viral RNA, which indicates whether or not a person is diseased, but it does not tell the person or doctors how severe this particular carrier’s disease will become. Therefore, those who are asymptomatic are often lumped in with those who do present symptoms, therein misjudging their subjective risk.

A quick overview of the app’s layout and functions. (Credit: New York University)

This app helps determine those who are at higher risk than others, allowing for reduced hospital admissions, as those with less of a need could be safely monitored at home.

To develop this app, researchers from NYU identified four key biomarkers measured from blood tests that appeared most prevalently in those who died compared to those who recovered; these were C-reactive protein, myoglobin, procalcitonin, and cardiac troponin I. These biomarkers can signal complications that would be relevant to COVID-19, including acute inflammation, lower respiratory tract infection, and poor cardiovascular health.

A model was constructed thereupon, using the aforementioned biomarkers as well as sex and age, two established risk factors, and trained using a machine learning algorithm, to predict the patterns of COVID-19 and predict its severity. When a patient’s biomarkers and risk factors are entered into the model, it produces a numerical COVID-19 severity score ranging from 0 (mild or moderate) to 100 (critical).

The model was validated using data from 12 hospitalised patients from Shenzhen, China, which confirmed that the model’s severity scores were significantly higher for the patients that died compared to the scores for the patients that were successfully discharged. These findings are published in Lab on a Chip, a journal of the Royal Society of Chemistry.         

Thumbnail credits: Getty Images/iStockphoto