The objective of this clinical trial is to assess whether early extubation of patients with hypoxemia during the spontaneous breathing trial (SBT) shortens the duration of ventilatory support. The trial will also evaluate the safety of this approach. The key research questions include:

Does early extubation of hypoxemic patients reduce the total duration of ventilatory support (both invasive and non-invasive) by 36 hours? Does early extubation of hypoxemic patients increase the number of ventilator-free days by day 28? Can the safety of early extubation in hypoxemic patients be ensured by confirming no significant differences in rates of reintubation, tracheostomy, or mortality? The trial will compare ventilatory outcomes between two groups: hypoxemic patients who undergo early extubation (hypoxemic extubation group) and those who remain on invasive ventilation until hypoxemia resolves (conventional extubation group).

International guidelines recommend extubating patients after correction of hypoxemia, meaning if the SpO2 measured during the spontaneous breathing trial is above 92%. However, there is strong rationale for modifying this practice to extubate patients earlier, particularly those presenting with hypoxemia after major surgery, by using alternating non-invasive ventilation (NIV) and high-flow oxygen therapy:

Several studies have found no link between patient oxygenation and extubation success, where outcomes for hypoxemic and non-hypoxemic patients are similar. Isolated hypoxemia thus does not appear to be a predictor of reintubation.

Hypoxemia is very common following major surgery, primarily due to shunts caused by atelectasis. Treatment for these atelectasis includes airway pressurization, bronchial secretion drainage, mobilization, and reducing factors that lead to diaphragmatic dysfunction.

In patients on invasive mechanical ventilation, secretion drainage is impaired, and mobilization to a seated position is more challenging. It has also been shown that diaphragmatic dysfunction occurs with prolonged ventilation. Hypoxemia can therefore be sustained by invasive ventilation, increasing the risk of therapeutic escalation.

Current guidelines do not account for the widespread use of non-invasive assistance techniques (such as high-flow oxygen therapy and non-invasive ventilation) that are now routinely employed in intensive care. These techniques allow for adequate oxygenation with high patient comfort and good tolerance.

Prolonging invasive ventilation in hypoxemic patients, as recommended by guidelines, could lead to associated complications. In contrast, early extubation of patients with hypoxemia may reduce the duration of both invasive and non-invasive ventilation, as well as complications related to prolonged invasive ventilation, without increasing the risk of reintubation.

Compared to continuing mechanical ventilation until hypoxemia is corrected, extubating purely hypoxemic patients and transitioning them to non-invasive ventilation techniques could represent a significant improvement in patient care.

Inclusion Criteria:

• Francophone patient affiliated to a health insurance plan;
• Patient having granted free, informed and written consent to participate in the study;
• Patient with hypoxemia defined as SpO2 ≤ 90% under 6 L/min or FiO2 40% during spontaneous breathing trial.

Exclusion Criteria:

• Presence of hypercapnia at the end of SBT (PaCO2 above 50 mmHg);
• Presence of severe hypoxemia during SBT defined by SpO2 below 86% under 9 L/min or FiO2 = 50%;
• Presence of poor clinical tolerance of SBT defined by polypnoea above 30/min, agitation, sweating, hypertension (PAS above 180 mmHg), tachycardia (HR above 140 bpm) or arrhythmia;
• Presence of an ineffective cough or major bronchial congestion;
• Patient already included in a type 1 interventional research protocol (RIPH1), modifying the procedure for ventilatory weaning and/or ventilatory support after extubation;
• Anatomical factors precluding the use of NIV or high-flow oxygen therapy, notably facial or cervico-facial malformations;
• Tracheostomized patient;
• History of obstructive ventilatory disorders with indication for NIV post-extubation, chronic obstructive pulmonary disease (COPD) GOLD score III/IV;
• History of obstructive sleep apnea syndrome with equipment;
• cardiogenic pulmonary edema;
• Patient on extracorporeal membrane oxygenation (ECMO) at the time of inclusion;
• Patient under guardianship or curatorship;
• Minor patients;
• Patient deprived of liberty or under court protection;
• Pregnant or breast-feeding women;
• Patient in therapeutic limitation with decision not to re-intubate.