Critically ill patients are predisposed to oxyhaemoglobin desaturation during intubation. For the intubation of hypoxemic patients, preoxygenation using non invasive ventilation (NIV) is more effective at reducing arterial oxyhaemoglobin desaturation than standard method.

Objectives: To find out whether NIV, as a preoxygenation method, is more effective at reducing the degree of organ dysfunction/failure than standard preoxygenation during the week following endotracheal intubation.

During the inclusion period (at least 10 min and maximum 30 min), the patients ware a high FiO2 mask, driven by 10-15L/min oxygen and are randomly assigned to control or NIV group. Preoxygenation is then performed for a 3 minute period prior to a standardized rapid sequence intubation. For the control group, preoxygenation use a non-re-breather bag-valve mask driven by 15L/min oxygen. Patients allow to breath spontaneously with occasional assists (usual preoxygenation method). For the NIV group, pressure support mode is delivered by an ICU ventilator through a face mask adjusted to obtain an expired tidal volume of 7 to 10 mL/kg. The fraction of inspired oxygen (FiO2) was 100% and we used a PEEP level of 5 cmH2O.

Inclusion Criteria:

• Adults patients with acute respiratory failure requiring intubation

Exclusion Criteria:

• Encephalopathy or coma, cardiac resuscitation