Influence of leaks and mask design on CO2 clearance during non-invasive ventilation: a combined bench and physiological study.

BACKGROUND: Non-invasive ventilation (NIV) interface can generate carbon dioxide (CO2) rebreathing, potentially contributing to NIV failure. The impact of leaks and mask design remains unclear. This study evaluates how these factors influence CO2 rebreathing in oronasal face masks. DESIGN, SETTINGS, PARTICIPANTS: A bench study was followed by a physiological study. The bench study used a mannequin head connected to a breathing simulator exhaling CO2. Six oronasal face masks designed for single-limb circuits with varying exhalation port positions were tested with three levels of fit. A CO2 rebreathing index was calculated, along with intentional and unintentional leaks. The best-performing mask was then compared with two clinically used masks in healthy volunteers. Ventilatory ratio (VR, reflecting dead space) was derived from transcutaneous CO2 (PtcCO2) and ventilation was assessed by electrical impedance tomography. Analyses used Friedman test and linear mixed-effects model. RESULTS: The bench study showed (1) mask design as a whole, rather than intentional leak flow rate, was the main determinant of CO2 rebreathing: exhalation ports situated on the mask body had lower CO2 rebreathing; (2) unintentional leaks decreased CO2 rebreathing only with poor mask design and (3) a new under-nose mask with nasal nozzles had the lowest CO2 rebreathing index. The latter was compared with two other masks in ten volunteers, median aged 32 (IQR 29-34): PtcCO2 decreased from 37.3 (33.7-39.5) mm Hg at baseline to 32.5 (28.9-35.4) with the new mask (p=0.023). VR with the new mask (0.96 (0.93-1.12)) was lower than with the classical masks (1.40 (1.20-1.50) and 1.38 (1.19-1.45); p<0.001 for both). CONCLUSIONS: Mask design significantly influences CO2 rebreathing during NIV, with leaks' effects dependent on exhalation port position. A novel under-nose hybrid mask with embedded nozzles reduces CO2 rebreathing and dead space when compared with classical masks. TRIAL REGISTRATION NUMBER: NCT03902639.