Triaging acute chest syndrome clinical decision-making using bedside SaO2/FiO2 ratio

Abstract

Acute chest syndrome (ACS) severity is inconsistently defined, and its clinical course is difficult to predict. This retrospective observational study evaluated the utility of the ratio of pulse oximetry oxygen saturation (SaO2) to the fraction of inspired oxygen (SaO2/FiO2) in adult patients with ACS and its association with the clinical outcome of intensive care unit (ICU) transfer. Across all ACS hospitalizations at a tertiary medical center from 2017 to 2021, we characterized the SaO2/FiO2 ratio at 3 time points: emergency department (ED) presentation, ACS diagnosis, and antibiotic initiation. Of the 227 hospitalizations identified, 54% were female, the mean age was 29 years, 70% had hemoglobin SS, and 9% had obesity. Although ICU transfer was not strongly associated with the SaO2/FiO2 ratio at ED presentation (area under the curve [AUC], 0.59), it was strongly associated with the ratio at ACS diagnosis (AUC, 0.73) and antibiotic initiation (AUC, 0.74). Given the highest sensitivity at ACS diagnosis, a diagnostic SaO2/FiO2 cutoff of 310 was proposed for triaging likely ICU transfer (sensitivity, 63%; specificity, 82%; adjusted odds ratio, 8.94; 95% confidence interval [CI], 2.12-37.6; adjusted hazard ratio, 4.86; 95% CI, 1.91-12.4), with models adjusted for obesity, lung disease, and blood counts. This cutoff corresponds to an SaO2 acquired from pulse oximeter saturation <90% on 2 L/min nasal cannula support. We propose using the SaO2/FiO2 ratio cutoff of 310 prospectively as a simple bedside triage tool for adult patients with sickle cell disease hospitalized with ACS to be transferred to a higher level of care.