High Tidal volume is one of the main mechanisms that lead to lung injuries under mechanical ventilation (ventilator induced lung injury: VILI). It could also induce lung damage during spontaneous or assisted ventilation (patient-self inflicted lung injury: P-SILI). Different non invasive oxygenation devices are available to deliver oxygen during acute hypoxemic respiratory failure: high concentration mask, high flow nasal canula and non-invasive ventilation (with bucco-nasal mask or helmet). The investigators hypothesized that the device may influence the tidal volume. Therefore, the objective of this study is to measure and compare the tidal volume during the use of each device. Tidal volume will be measured using Electrical impedence tomography.

Background : De novo acute respiratory failure with moderate to severe hypoxemia is associated with high intubation and mortality rates. Under non-invasive ventilation (NIV) with a bucco-nasal mask, a high Tidal volume is associated with NIV failure and poor outcome. It is suspected that non-control of the Tidal volume may worsen the prognosis of the patient. Indeed, it is logical to consider that the physiopathological mechanisms that lead to ventilator induced lung injury (VILI) under mechanical ventilation, especially over-distension, can in the same way aggravate pulmonary lesions under NIV and even during spontaneous breathing. The value of the Tidal volume is available in non-invasive ventilation with the bucco-nasal mask, but it is unknown with the use of other non-invasive oxygenation techniques : oxygen therapy with the high concentration mask, non-invasive ventilation with the helmet, and high flow nasal canula.

Objectives : The main objective is to compare the Tidal volume received under different oxygenation techniques (oxygen therapy with high concentration mask, NIV with bucco-nasal mask and with helmet, high flow nasal canula), in patients with de novo acute hypoxemic respiratory failure and moderate to severe hypoxemia.

Methods : All patients admitted in intensive care for de novo acute hypoxemic respiratory failure, non hypercapnic, with moderate to severe hypoxemia (PiO2/FiO2 < 200 mmHg) will be included. Patients with one of the following items will be excluded : chronic respiratory disease, cardiogenic pulmonary edema, hemodynamic instability, coma, contraindication to NIV, indication for immediate intubation.

After inclusion, the Tidal volume will be measured for each technique with the use of electrical impedance tomography, in a random order. Randomisation will be carried out by using sealed envelopes containing a predetermined order. Each technique will be used for 15 minutes. For NIV with the bucco-nasal mask, the inspiratory pressure will be progressively adjusted to maintain the Tidal volume between 6 and 8 ml/kg of ideal body weight without dropping below 7 cm H2O and the positive expiratory pressure will be gradually increased to 5 cm H2O. For NIV with the helmet, the inspiratory pressure will be progressively increase to 12 cm H2O and the positive expiratory pressure will be gradually increased to 5 cm H2O. For high flow nasal canula,the gas flow will be set at 50 L/minute. FiO2 will be adjusted to keep SpO2 above 92%. To convert the thoracic impedance variations into absolute values of Tidal volumes, the airway flow signal will be collected with a pneumotacograph during the NIV with bucco-nasal mask session. The following data will also be collected : respiratory rate, comfort, regional Tidal volume in dependent and non-dependent areas and global inhomogeneity index, Pa02, PaCO2 and pH (If an arterial catheter is present). Signed consent will be required prior to any inclusion.

Expected benefits : This study will provide a better understanding of P-Sili and help define the place each non-invasive oxygenation techniques in the management of de novo acute respiratory failure.

Inclusion Criteria:

• Respiratory rate > 25 per minute
• PaO2/FiO2 < 200 mmHg
• PaCO2 < 45 mmHg
• Signed consent by patient or relative

Exclusion Criteria:

• Chronic respiratory disease
• COPD exacerbation
• Acute asthma
• Cardiogenic pulmonary edema
• Hemodynamic instability (systolic blood pressure < 90 mmHg or mean blood pressure < 65 mmHg or use of catecholamines)
• Coma (Glasgow coma scale < 12)
• Contraindication to non-invasive ventilation
• Indication to immediate intubation according to the physician

Patients with de novo acute hypoxemic respiratory failure