Comparison of ventilation techniques for compensation of mask leakage using a ventilator and a regular full-face mask: A bench study.

BACKGROUND: The use of noninvasive ventilation (NIV) during and after extubation is common. We designed this study to determine the optimal strategy to compensate for mask leaks and achieve effective ventilation during NIV by comparing commonly used operating room ventilator systems and a regular facemask.

METHODS: We tested four operating room ventilator systems (Dägger Zeus, Dägger Apollo, Dägger Fabius Tiro, and General Electric Healthcare Carestation 650) on a lung model with normal compliance and airway resistance and evaluated pressure control ventilation (PCV), volume control ventilation (VCV), and AutoFlow mode (VAF). We set the O2 flow at 10 L/min and the maximal flow at 13, 16, or 26 L/min. We simulated five leak levels, from no leak to over 40 L/min (I to V levels), using customized T-pieces placed between the lung model and the breathing circuit. We recorded the expired tidal volume (Vte) from the lung model and peak inspiratory pressure via two flow/pressure sensors that were placed distally and proximally to the T-pieces.

RESULTS: 1. Comparison of four ventilators: with any given ventilation mode, an increase in leak level caused a decrease in Vte. With PCV, only Zeus produced Vte larger than 150 ml at leak level V. 2. Effect of ventilation mode on Vte: across all four ventilators, PCV resulted in a higher Vte than VCV and VAF (P < 0.01). PCV mode with all ventilators at leak level II provided Vte values that were equal to or greater than those obtained with no leak. 3. Effect of O2 flow on Vte Using PCV mode: only Carestation 650 Vte at leak level II during PCV were significantly greater with 16 L/min O2 flow compared with 10 L/min O2 flow (P < 0.01). 4. Actual leak: increasing the O2 flow from 10 L/min to the maximum O2 flow dramatically increased the real leak with all 4 ventilators at any fixed leak level (P < 0.01). 5. Preset PIP vs. actual PIP with PCV: at low preset PIP and leak levels such as leak II and III, the discrepancy between preset PIP and actual PIP was small. The disparity between the preset and actual PIP grew when the target PIP and the leak level were raised.

CONCLUSION: For NIV using a mask, the ventilator is preferred whose Pressure generator is Turbine, the PCV mode is preferred in the ventilation mode and the oxygen flow is set to 10 L/min or maximum oxygen flow.