Machine learning tool can predict severe illness, death from COVID-19 in patients

Researchers developed a machine learning calculator that provides predictions of whether patients hospitalized with COVID-19 will progress to severe illness or death, according to a March 2 study published in Annals of Internal Medicine.

Most COVID-19 prognostic tools only use data given during admission and don't include changes that occurred after admission, the report said. The study's researchers developed a machine learning tool to predict severe illness or death in patients with COVID-19 during their first 14 days in the hospital.

The study followed 3,163 patients who were hospitalized between March 5 and Dec. 4, 2020, for being at risk for severe COVID-19. The study took place in five hospitals within the John Hopkins Medicine system in Maryland and Washington, D.C.

The interactive tool developed rapid and accurate probabilities of whether a patient will progress to severe illness or death based on clinical data available. The tool's purpose is to assist front-line clinicians with prognostic tools to better allocate a hospital's resources.

 

© Copyright ASC COMMUNICATIONS 2021. Interested in LINKING to or REPRINTING this content? View our policies by clicking here.