High-Flow Nasal Cannula and Prone Positioning in Awake Patients with COVID-19 Related Respiratory Failure: An Observational Study.
Abstract Background: We report an observational study on the use of High-flow nasal cannula (HFNC) and awake prone position in patients with Covid-19 related severe ARF.Methods: chart analysis of consecutive patients with confirmed Covid-19 and severe ARF (PaO2/FiO2 ratio <150) who received HFNC. Patients were systematically encouraged to lie in the prone position if tolerated. We calculated initial ROX index (the ratio of SpO2/FIO2 to respiratory rate) while in supine position and at the end of the first HFNC session whether in prone or supine position, and their difference (delta ROX). The success/failure of HFNC (need for intubation) was recorded. Predictors of HFNC failure were identified using ROC curve and logistic regression.Results: HFNC was administered to 213 out of 360 patients with COVID-19 related severe ARF (71% male, median age:59 years (IQR:50-68), median PaO2/FIO2: 104(73-143). At the start of HFNC, median ROX index was 4(3.4-5). Among included patients, 178 (83.5%) could tolerate prone position and had a median of 4.4(2-6) prone sessions during a median of 10(4-16) hours/day each, for a median of 4(2-7) days. Overall, HFNC failure occurred in 61 patients (28.1%) with similar proportions in patients who had HFNC in prone position and in patients who did not tolerate prone position (29% and 26%, respectively; relative risk:1.14. 95%CI:0.62-2.1). In the prediction of HFNC outcome, AUC was highest for delta ROX (AUC=0.83); AUC for baseline ROX (0.71), PaO2/FiO2 (0.73), and SpO2 (0.67), were significantly lower. The delta ROX cut-off ≤1.8 had the best Youden index indicating the best combination of sensitivity (0.89) and specificity (0.61) with a PLR (2.33) and a NLR (0.17) to predict HFNC failure. Logistic regression disclosed the following predictors of HFNC failure: delta ROX: RR=0.44, 95%CI=0.32-0.62; p-0.0001); baseline ROX index: RR=0.58, 95% CI:0.39-0.85, p=0.005); SOFA score (RR=1.6 for each point; 95%CI: 1.1-2.2, p=0.007); and PaO2/FiO2 at admission: RR=0.96, 95%CI=0.94-0.99). Prone position was not related to HFNC success.CONCLUSION: Awake HFNC in prone position is feasible in most patients with severe hypoxemic COVID-19. Indicators of ARF severity and the early response to HFNC, rather than prone position are independently associated with HFNC outcome.