With the birth of Mechanical Ventilation in the 1950s came the ventilation induced lung injuries (VILI). Numerous works have since then shown the benefit of "protective ventilation", notably by controlling the delivered tidal volume and pressures. However, as the respiratory condition improves and the weaning is started by shifting to Pressure Support Ventilation (PSV), these parameters stop being tightly controlled. This study aims to determine whether there is a relationship between the driving pressure measured in PSV and the weaning time.

Scientific justification :

As mechanical ventilation developed since the 1950s, researchers started to recognize characteristic lung disease associated with it, Nash et al giving an anatomopathological description of "Respiratory Lung" on post-mortem examination of lungs after mechanical ventilation in 1967 [2]. It progressively led to the concept of VILI and of the protective ventilation to minimize it, enhancing lower tidal volume and plateau pressure [3], controlled Driving Pressure < 15cmH2O[4], neuromuscular blockade[5] and prone positioning[6]. However, these parameters can only be controlled for sedated patients in Controlled Ventilation. As the respiratory conditions improve, the onset of spontaneous breathing uses PSV [7] but because pressure support is added to the inspiratory effort of the patient, tidal volume and driving pressure stop being tightly controlled. It is therefore possible for the driving pressure to be higher than 15 cmH2O in case of a major inspiratory effort. One ought to wonder whether a high driving pressure is associated with a prolonged weaning phase following a moderate to severe ARDS.

Strategy description:

Patients that enter the weaning phase following a moderate to severe ARDS equipped with a nasogastric allowing measures of EAdi will be included. Driving Pressure will be measured following the method used by Bellani et al [1]. A weaning test will be conducted daily.

Follow up description:

• Daily measures of End Inspiratory Pressure with respiratory synchronisation optimised by use of EAdi
• Daily spontaneous breathing trial using low levels of pressure support
• Pplat, Respiratory System Compliance, Driving Pressure, PEEP, Tidal Volume will be monitored daily as well as clinical and other routine ventilatory data. Data concerning initial severity of ARDS, and duration of ARDS, controlled mechanical ventilation, sedation and neuromuscular blockade and date of first spontaneous breathing trial will be collected. A weaning test will be conducted daily.

Inclusion Criteria:

• Patient over 18 years old
• Patients hospitalized in the ICU and who have suffered moderate to severe ARDS (Berlin criteria) equiped with nasogastric feeding tube allowing measures of Eadi

Exclusion Criteria:

• COPD or measured or suspected auto-PEEP higher than 3 cmH2O at beginning of Pressure Support
• Admission in the ICU more than 15 days before the initiation of weaning
• Broncho-pleural leaks
• ECMO
• Pregnant woman.

Patients hospitalized in the thoracic resuscitation department of the Haut-Lévêque Hospital, Bordeaux University Hospital, who will be consecutive patients undergoing respiratory weaning following a moderate-to-severe ARDS according to the Berlin criteria, and not presenting the exclusion criteria mentioned above will be proposed the protocol.