COVID-19 Risk Score Might Predict Critical Illness in Hospitalized Patients

Earlier prediction of critical illness in patients admitted with COVID-19 could improve patient outcomes by leading to prompt therapy, proactive intensive care unit (ICU) transfers, or better allocation of limited resources. Researchers in China retrospectively reviewed data on 1600 patients (mean age, 50) who were admitted with polymerase chain reaction–confirmed COVID-19 to 575 hospitals. The investigators constructed a calculation tool (COVID-GRAM) to predict which patients would develop critical illness, defined as ICU admission, mechanical ventilation, or death.

Eight percent of patients (131) in the data set developed life-threatening illness; mortality was 3.2%. From 72 potential predictors, 10 variables* were found to be independent …