Non-invasive ventilation for acute hypoxemic respiratory failure, including COVID-19

Abstract

Optimal initial non-invasive management of acute hypoxemic respiratory failure (AHRF), of both coronavirus disease 2019 (COVID-19) and non-COVID-19 etiologies, has been the subject of significant discussion. Avoidance of endotracheal intubation reduces related complications, but maintenance of spontaneous breathing with intense respiratory effort may increase risks of patients' self-inflicted lung injury, leading to delayed intubation and worse clinical outcomes. High-flow nasal oxygen is currently recommended as the optimal strategy for AHRF management for its simplicity and beneficial physiological effects. Non-invasive ventilation (NIV), delivered as either pressure support or continuous positive airway pressure via interfaces like face masks and helmets, can improve oxygenation and may be associated with reduced endotracheal intubation rates. However, treatment failure is common and associated with poor outcomes. Expertise and knowledge of the specific features of each interface are necessary to fully exploit their potential benefits and minimize risks. Strict clinical and physiological monitoring is necessary during any treatment to avoid delays in endotracheal intubation and protective ventilation. In this narrative review, we analyze the physiological benefits and risks of spontaneous breathing in AHRF, and the characteristics of tools for delivering NIV. The goal herein is to provide a contemporary, evidence-based overview of this highly relevant topic.

Keywords: Hypoxemic respiratory failure; Non-invasive ventilation; Self-inflicted lung injury.

Figure 1

Pendelluft during spontaneous breathing. Representative patient tracings during spontaneous breathing, showing the pendelluft phenomenon between the ventral and dorsal lung regions. A: Esophageal pressure tracings showing the start of muscular inspiration corresponding to the point of esophageal pressure deflection and ΔPes representing inspiratory effort. B: Global ΔZ, expressed in arbitrary units, showing overall lung insufflation during inspiration. C–F: Ventral ROI ΔZ from C and D, demonstrating an initial emptying of these lung areas (blue rectangle), with air moving toward the dorsal ROI (E and F), which are instead characterized by an increased electrical impedance tomography signal. The opposite phenomenon occurs during expiration. The ΔZ of each ROI, corresponding to their relative tidal volume and occurring at different timepoints across respiratory cycles, are highlighted by the red dotted lines. DD: Dorso-dorsal; MD: Mid-dorsal; MV: Mid-ventral; ROI: Regions of interest; VV: Ventro-ventral.