LAUSR

Background The use of high flow oxygen therapy (HFOT) has significantly escalated during the COVID-19 pandemic. HFOT can be delivered through both dedicated devices and ICU ventilators. HFOT can be administered to a patient via a nasal cannula (NC). In intubated patients, a tracheal cannula (TC) is used instead. In this study, we aim to compare the work of breathing (WOB) using a TC or NC and to explore whether differences exist among HFOT devices. Methods Seven HFOT devices (three dedicated and four ICU ventilators) were connected to a manikin head (Laerdal Medical) through a NC (Optiflow 3S, large size, Fisher and Paykel Healthcare) or a TC (OPT 970 Optiflow+, Fisher and Paykel Healthcare). Each device was also attached to a manikin head that was connected to a lung simulator (ASL5000, Ingmar Medical), set at 40 ml/cmH2O compliance, 10 cmH2O/L/s resistance, and sinusoidal inspiratory effort (muscular pressure 10 cmH2O, rate 30 breaths/min). HFOT was delivered at 40 L/min and at 21% inspired oxygen fraction. The total WOB per breath and its resistive and elastic components were automatically analyzed breath by breath over the last 20 breaths by using Campbell's diagram. Results The WOB and its resistive and elastic components were significantly lower with the TC than with the NC for every device, and systematically lower with the reference device than with others. These differences were, however, very small and may be not clinically relevant. Conclusion The WOB is lower with the TC than with the NC and with the reference device, compared with the most recent devices.