Choice of respiratory therapy for COVID-19 patients with acute hypoxemic respiratory failure: a retrospective case series study.

Background: In the treatment of acute hypoxemic respiratory failure (AHRF) due to coronavirus 2019 (COVID-19), physicians choose respiratory management ranging from low-flow oxygen therapy to more invasive methods, depending on the severity of the patient's symptoms. Recently, the ratio of oxygen saturation (ROX) index has been proposed as a clinical indicator to support the decision for either high-flow nasal cannulation (HFNC) or mechanical ventilation (MV). However, the reported cut-off value of the ROX index ranges widely from 2.7 to 5.9. The objective of this study was to identify indices to achieve empirical physician decisions for MV initiation, providing insights to shorten the delay from HFNC to MV. We retrospectively analyzed the ROX index 6 hours after initiating HFNC and lung infiltration volume (LIV) calculated from chest computed tomography (CT) images in COVID-19 patients with AHRF. Methods: We retrospectively analyzed the data for 59 COVID-19 patients with AHRF in our facility to determine the cut-off value of the ROX index for respiratory therapeutic decisions and the significance of radiological evaluation of pneumonia severity. The physicians chose either HFNC or MV, and the outcomes were retrospectively analyzed using the ROX index for initiating HFNC. LIV was calculated using chest CT images at admission. Results: Among the 59 patients who required high-flow oxygen therapy with HFNC at admission, 24 were later transitioned to MV; the remaining 35 patients recovered. Four of the 24 patients in the MV group died, and the ROX index values of these patients were 9.8, 7.3, 5.4, and 3.0, respectively. These index values indicated that the ROX index of half of the patients who died was higher than the reported cut-off values of the ROX index, which range from 2.7-5.99. The cut-off value of the ROX index 6 hours after the start of HFNC, which was used to classify the management of HFNC or MV as a physician's clinical decision, was approximately 6.1. The LIV cut-off value on chest CT between HFNC and MV was 35.5%. Using both the ROX index and LIV, the cut-off classifying HFNC or MV was obtained using the equation, LIV = 4.26 × (ROX index) + 7.89. The area under the receiver operating characteristic curve, as an evaluation metric of the classification, improved to 0.94 with a sensitivity of 0.79 and specificity of 0.91 using both the ROX index and LIV. Conclusion: Physicians' empirical decisions associated with the choice of respiratory therapy for HFNC oxygen therapy or MV can be supported by the combination of the ROX index and the LIV index calculated from chest CT images.