Acute respiratory distress syndrome (ARDS) is a severe pulmonary insult responsible for major, life-threatening hypoxemia. The alteration of hematosis is secondary to alveolar edema, following damage to the alveolocapillary barrier in response to a systemic inflammatory process. The presence of fluid effusion within the alveolar sacs and the modification of type II pneumocyte activity due to the presence of numerous pro-inflammatory mediators will lead to a quantitative and qualitative alteration of the surfactant. At the same time, leukocyte infiltration will lead to an alteration of the support tissue and to the accumulation of cellular debris. All these elements will lead to a heterogeneous loss of aeration of the lung. In addition, the alveolar units are compressed by the entire lung parenchyma due to the effect of gravity on the edematous tissue.

The treatment of ARDS is based on the antagonistic need to maintain hematosis and reduce parenchymal insult secondary to mechanical ventilation. Optimization of mechanical ventilation consists in reducing the volume of gas administered at each respiratory cycle and in limiting thoracic parietal stress by the use of curares. More recently, the interest of the ventral decubitus position has been demonstrated. During such a maneuver, the posterior pulmonary parenchymatous zones, usually subjected to gravity in the supine position, will be able to re-expand under the effect of the prone position and of the positive pressure induced by the ventilator. The increase in parietal elastance, due to the compression of the thorax between the posterior part of the trunk and the bed, also contributes to an improvement in the distribution of inhaled gases within the pulmonary parenchyma by limiting the loss of energy, transmitted directly to the wall. The ventral decubitus position allows to redistribute the ventilation in territories which were not aired before but which participate to the respiratory exchanges because they are still perfused and thus to improve the pulmonary compliance measured.

Although described as an atypical form, SARS-CoV-2 infection can lead to ARDS with severe forms of viral pneumonia and thus require prone positioning.

While this results in improved oxygenation and compliance, prone positioning is accompanied by a risk of complications such as pressure sores, described as the most frequent. In addition, the massive influx of patients and more generally the lack of personnel during pandemic peaks has made the application of prone position sometimes complex because it requires human resources. As a result, the benefit/risk ratio of the maneuver is difficult to determine because not all patients respond in the same way to prone positioning. It appeared essential to be able to predict the expected benefit of the prone position before performing the procubitus maneuver.

The application of thoracic and abdominal pressures, as part of the respiratory management of patients, is a technique commonly used by physiotherapists. Investigators have demonstrated a similar change in measured lung parenchymal compliance during manual compression of a patient's chest and during prone positioning. In the context of the epidemic, investigators used this test systematically to determine which patients were most likely to benefit from prone positioning and for whom the available resources should be concentrated at any given time.

Inclusion Criteria:

• Patient whose age ≥ 18 years
• Patient hospitalized in the ICU for acute respiratory distress syndrome at Covid-19
• Ventilated intubated patient for whom prone sessions were offered
• Patient who received respiratory physical therapy sessions on the same day as the prone position
• French-speaking patient

Exclusion Criteria:

• Non-intubated patient performing vigorous prone sessions
• Patient on invasive mechanical ventilation for another etiology than Covid-19
• Patient under guardianship or curatorship
• Patient deprived of liberty
• Patient under court protection
• Patient objecting to the use of his/her data for this research

Patients were admitted to the ICU for acute respiratory distress syndrome at COVID-19 between 04/10/2021 and 04/01/2022.