Predicting COVID-19 Prognosis in Hospitalized Patients Based on Early Status

Predicting which patients are at greatest risk of severe disease from COVID-19 has the potential to improve patient outcomes and improve resource allocation. We developed machine learning models for predicting COVID-19 prognosis from a retrospective chart review of 969 hospitalized COVID-19 patients at Robert Wood Johnson University Hospital (RWJUH) during the first pandemic wave in the United States. We generated two models, PLABAC and PRABLE, that use age and five common laboratory tests to predict mortality (PLABAC: AUC ROC=0.796; PRABLE: AUC ROC=0.793), which outperformed CURB-65, a commonly used clinical prediction rule for pneumonia severity (AUC ROC =0.722). We validated PLABAC using data from the National COVID Cohort Collaborative Data Enclave (NC3), which performed well on patients from both before and after the advent of COVID-19 vaccines. This study demonstrates that our models can accurately predict COVID-19 outcomes from information obtained early and can serve as a useful clinical decision-making tool.