LAUSR

Study Objective: To externally validate a risk-stratification tool—the Quick COVID-19 Severity Index (qCSI)—developed by Haimovich et al. to predict 24-hour respiratory decompensation in admitted patients with COVID-19. Methods: This was a retrospective observational cohort study of COVID-19 patients admitted from the emergency department between Feb 29, 2020 to Feb 1, 2021. The health care system is composed of a mix of 2 community and 4 academics EDs in a major metropolitan area. Patient demographics, vital signs, laboratory results were extracted from our institutional COVID-19 Data Warehouse. Following the convention of qCSI variables, respiratory rate (breaths/min), pulse oximetry (%), and oxygen flow rate (L/min) were used to calculate points between 0 to 12, with higher points associated with highly likelihood of respiratory decompensation within 24 hours. Results: 35,696 COVID-19 patients were admitted via the emergency department during the study period. The mean qCSI was 1.73 (SD 1.82) for non-ICU admissions (n=34,647). The mean qCSI was 2.83 (SD 2.53) for ICU admission (n=1,049). As of the time of submission, ED treat and release patients, as well as decompensation results are pending. Conclusions: In this validation study of qCSI using a large system cohort of COVID-19 patients, qCSI appears to correlate strongly with clinical triage for admission decision to regular floor vs. ICU level care. Further analysis is needed to identify 24-hour respiratory decompensation after regular floor admission.