High flow nasal cannula and low level continuous positive airway pressure have different physiological effects during de novo acute hypoxemic respiratory failure

Background Large tidal volumes during de novo acute hypoxemic respiratory failure (AHRF) may promote patient self-inflicted lung injury. Tidal volume assessment under high flow nasal cannula (HFNC) is not routinely feasible at the bedside. Our objective was to determine whether tidal volume during low-level continuous positive airway pressure (CPAP) could predict tidal volume during HFNC and to compare the physiological effects of HFNC and low-level CPAP. Methods Prospective, single-center study including 29 de novo AHRF patients treated with HFNC (50 to 60 L.min(- 1)). Patients were monitored using electrical impedance tomography during HFNC then CPAP at 4 cmH(2)O. Tidal volume during HFNC was calculated based on tidal impedance variation. The ability of tidal volume under low-level CPAP to predict tidal volume under HFNC was explored using Bland-Altman analysis. CPAP and HFNC were compared in terms of tidal volume, minute ventilation, respiratory comfort, dyspnea, oxygenation, ventilation distribution, end-expiratory lung volume, thoraco-abdominal asynchrony and recruitment. Results Under HFNC, patients had a tidal volume of 6.6 (5.9-8.7) mL.kg(- 1) PBW. 20 (69%) patients exhibited a tidal volume between 4 and 8 mL.kg(- 1) PBW, while in 5 (17%) patients it exceeded 9 mL.kg(- 1) PBW. Tidal volume under CPAP was higher (9.4 (8.3-11) mL.kg(- 1) PBW, p < 0.001). Tidal volumes under CPAP and under HFNC were modestly correlated (Spearman r = 0.50, p = 0.005). Bland-Altman analysis showed a bias of 2.4 mL.kg(- 1), with limits of agreement ranging from - 1.1 mL.kg(- 1)to 5.9 mL.kg(- 1). Nevertheless, a larger (> 11.5 mL.kg(- 1) PBW ) tidal volume under low-level CPAP predicted a larger (> 9 mL.kg(- 1) PBW ) tidal volume under HFNC with 80% sensitivity and 96% specificity. Low-level CPAP was associated with increased minute ventilation, end-expiratory lung volume, and oxygenation as compared to HFNC. It decreased signs of respiratory distress in the most severe patients but was associated with lower comfort compared to HFNC. Conclusion Among ICU patients with de novo AHRF, tidal volume under HFNC was mostly protective. Tidal volume during CPAP at 4 cmH(2)O did not predict tidal volume during HFNC. Such low-level CPAP was associated with increased tidal volume, minute ventilation, end-expiratory volume, and oxygenation.