Physiological comparison of high-flow oxygen via endotracheal tube and T-piece strategies during spontaneous breathing trials: a randomized crossover study.

BACKGROUND: The optimal spontaneous breathing trial (SBT) strategy remains uncertain because commonly used approaches involve distinct physiological trade-offs. Pressure support ventilation (PSV) may reduce inspiratory effort, whereas T-piece trials more closely approximate unsupported breathing after extubation. Neither approach has shown clear superiority for extubation outcomes. High-flow oxygen (HFO) delivered via endotracheal tube has been proposed as an alternative SBT strategy, but the physiological effects remain insufficiently defined. METHODS: In a randomized crossover physiological study, patients receiving mechanical ventilation for at least 24 h and deemed ready for weaning underwent five SBT conditions: T-piece and four HFO via endotracheal tube settings combining two interfaces with different expiratory port diameters (9.8 and 6.9 mm) and two flow rates (40 and 60 L/min). Airway and esophageal pressures, respiratory variables, and electrical impedance tomography parameters were measured during baseline mechanical ventilation and under each SBT condition. Esophageal pressure was also measured after extubation for comparison with post-extubation inspiratory effort. A complementary bench study assessed 24 simulated conditions across four interface diameters and six flow rates. RESULTS: Twenty patients were analyzed. Compared with T-piece, HFO-6.9 at 60 L/min produced the largest increases in mean expiratory airway pressure (median paired difference, 7.0 cm H2O [95% CI, 6.7-7.4]) and PEEP (6.0 cm H2O [5.8-6.1]). HFO-6.9 at 60 L/min also attenuated global end-expiratory lung volume loss (321 mL [267-523]), reduced dynamic transpulmonary driving pressure (3.3 cm H2O [1.8-4.1]), improved PaO2/FiO2 (28 mmHg [14-52]), and reduced respiratory rate (2 breaths/min [1-4]) compared with T-piece. No statistically significant difference in ΔPes was observed across the five SBT conditions and post-extubation measurements (p = 0.24). Bench simulations supported the flow- and diameter-dependent pressure effects. CONCLUSIONS: During SBT, HFO via endotracheal tube generated flow- and expiratory port diameter-dependent increases in airway pressure and lung volume, with lower dynamic transpulmonary driving pressure and improved oxygenation; no statistically significant difference in ΔPes was observed across HFO, T-piece, and post-extubation measurements. These findings support tracheal HFO as a physiologically distinct SBT strategy that may inform individualized SBT selection. Its relationship to clinical outcomes requires further evaluation. TRIAL REGISTRATION: ClinicalTrials.gov (NCT06816706). Registered on 6 February 2025.