Development and Validation of the Acute PNeumonia Early Assessment Score for Safely Discharging Low-Risk SARS-CoV-2-Infected Patients from the Emergency Department

Abstract

A continuous demand for assistance and an overcrowded emergency department (ED) require early and safe discharge of low-risk severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-infected patients. We developed (n = 128) and validated (n = 330) the acute PNeumonia early assessment (aPNea) score in a tertiary hospital and preliminarily tested the score on an external secondary hospital (n = 97). The score’s performance was compared to that of the National Early Warning Score 2 (NEWS2). The composite outcome of either death or oral intubation within 30 days from admission occurred in 101 and 28 patients in the two hospitals, respectively. The area under the receiver operating characteristic (AUROC) curve of the aPNea model was 0.86 (95% confidence interval (CI), 0.78–0.93) and 0.79 (95% CI, 0.73–0.89) for the development and validation cohorts, respectively. The aPNea score discriminated low-risk patients better than NEWS2 at a 10% outcome probability, corresponding to five cut-off points and one cut-off point, respectively. aPNea’s cut-off reduced the number of unnecessary hospitalizations without missing outcomes by 27% (95% CI, 9–41) in the validation cohort. NEWS2 was not significant. In the external cohort, aPNea’s cut-off had 93% sensitivity (95% CI, 83–102) and a 94% negative predictive value (95% CI, 87–102). In conclusion, the aPNea score appears to be appropriate for discharging low-risk SARS-CoV-2-infected patients from the ED.