ARDS is an acutely induced respiratory failure characterized by the appearance of bilateral alveolar opacities on imaging and hypoxemia Etiologies are divided into two classes: pulmonary, including all infectious pathologies, aspiration pneumonia, and drowning, and extra-pulmonary, induced by sepsis or acute pancreatitis. The mortality rate of ARDS remains high in unselected patient populations Among strategies that have proven beneficial in terms of patient outcome, prone positioning (PP) is associated with the greatest impact in terms of reduction in mortality.

PP is currently recommended in the European guidelines for ARDS associated with a PaO2/FiO2 ratio < 150 mmHg in patients in whom ventilatory settings have been optimized beforehand, The failure of early PP studies to demonstrate a survival benefit in ARDS was attributed to insufficient session duration. The PROSEVA study was the first to demonstrate that a PP duration of 17 h is associated with a reduction in mortality

During the COVID-19 pandemic, several centers have reported the implementation of longer PP sessions. Two strategies have emerged from these studies. In one case, the patient was left in the prone position until the criteria for stopping PP were met. Thus, the PP/supine position alternation was completely suppressed. In another published strategy, PP sessions were maintained for a period covering two nights. Furthermore, in a multicenter retrospective study, PP sessions were maintained until clinical improvement was associated with reduced mortality. In this study of 263 patients, the median duration of PP in the extended duration group was 40 h, and 75% of the sessions lasted 48 h or less. Using a propensity score, the authors showed that patients treated with an extended PP duration had a lower 3-month mortality rate than patients in the standard duration group . This protocol was also associated with a 29% cumulative incidence of pressure sores, similar to the 25% cumulative incidence reported in the PROSEVA study Other data published on pressure sores and PP of duration > 24 hours are also reassuring. Finally, a recent review recently reported that an extended PP session of > 24 h had also been used before the COVID-19 pandemic.

PP sessions had a median duration of 47-78 hours and were applied mainly to ARDS secondary to community-acquired pneumonia. All pre-COVID studies were retrospective, monocentric, without a control group.

Description of procedures performed according to standard of care:

Patients will be ventilated with a tidal volume of between 6 and 8 ml/kg ideal body weight. FiO2 will be set to target oxygen saturation between 92 and 96%. Patients will be curarized. The positive end-expiratory pressure (PEEP) will be set at least at 8 cm H2O, and a PEEP trial will be performed to find the best compromise between being slightly below maximum lung compliance, maximizing oxygenation, minimizing hemodynamic effects, and maintaining a plateau pressure strictly below 30 cm H2O.

Patients will be turned in the prone position, with a cushion placed transversely opposite the shoulders and thorax, to relieve the neck and abdomen of any overpressure. The pressure points will be protected. The protocol complies with the recommendations of the Intensive Care Society and the Faculty of Intensive Care Medicine. The response to PP will be assessed by taking an arterial blood gas before PP, after 16 hours of PP, just before return to the supine position and 4 hours after the return to the supine position. In the event of a P/F ratio below 80 mm Hg, inhaled nitric oxide (iNO) will be added. In the event of non-response to iNO in terms of PaO2, the use of extracorporeal membrane oxygenation (ECMO) will be discussed. Patients will also benefit from etiological treatment specific to the underlying pathology. Finally, pressure sores are prevented through daily care and the use of anti-sore mattresses. Regular repositioning and daily skin assessment will be carried out by a nurse.

Description of procedures added by the research protocol Two procedures will be added to this usual patient management procedure. Patients will be maintained in PP until all the following criteria have been met for 4 consecutive hours: P/F ratio > 150 mm Hg, FiO2 < 60%, PEEP ≤ 10 cm H2O. We also impose a minimum session duration of 16 hours and a maximum session duration of 48 hours. Following this first session, if PP sessions are still indicated, for example in case of mechanically ventilated acquired pneumonia (MVAP), these sessions will be carried out according to the usual protocol, with sessions maintained for 16 to 24 hours, alternating with 8-hour periods in dorsal decubitus. The criteria for stopping this PP / supine alternation will be the usual ones: P/F < 150 mmHg with FiO2 < 60% and PEEP ≤ 10 cm H2O.

The second procedure added by the research is ventilation monitoring with the Dräger PulmoVista 500 electrical impedance tomography (EIT). This monitoring will only be carried out during the first prone session. Two sets of sixteen electrodes will be placed around the patient's thorax. These electrodes will capture 50 images per second, reflecting variations in pulmonary electrical impedance as a function of ventilation. The data will be recorded during 4 successive sessions: one before turning the patient on the prone position, one at H+16 after PP started, one before being turned back to the supine position and one 4 hours after being turned back in the supine position. The raw signals will be extracted from the EIT and analyzed using dedicated external software.

Summary of the main benefits and expected risks associated with the protocol

The main benefits expected for the patient are:

• The pursuit of a more homogeneous distribution of lung ventilation volumes associated with fewer overdistended pulmonary units, atelectatic pulmonary units and pulmonary units opening and closing cyclically.
• Delayed derecruitment induced by return in the supine position after a PP session.
• In the event of acute right heart failure, prolonged PP session is likely to further allow the unloading of the right ventricle.
• possibly reduced mortality

The main risk associated with this procedure is the potential increase in the incidence of pressure sores. Our published data are rather reassuring on this subject and show that increasing the duration of all PP sessions to a median duration of 40 hours is associated with a 2.5% cumulative incidence of grade ≥ III pressure sores, a cumulative incidence of pressure ulcers grade ≥ II of 26%, which was no greater than the cumulative incidence of 25% reported in the PROSEVA ancillary study. Finally, the main risk factor for pressure sores is not so much the duration of individual sessions as the cumulative duration.

Other risks include:

• a deterioration in the homogeneity of ventilatory volume distribution, with a re-increase in the percentage of atelectatic pulmonary units, over-distended pulmonary units and pulmonary units opening and closing cyclically.
• in the event of cardiac arrest, cardiac massage is delayed by the need to return the patient in the supine position before cardiopulmonary resuscitation can be started

Course of research Inclusion visit At the inclusion visit, the patient will be intubated and mechanically ventilated. The indication for ventilation in the prone position will have been established but will not have been turned in the prone position yet. Demographic variables and patient history (COPD, pulmonary emphysema, asthma, etc.) will be collected. A clinical examination will also be carried out on the patient to ensure that there are no pressure sores. Monitoring by EIT will be initiated once the ventilator parameters have been stable for 20 minutes. All the ventilator's anonymized data will be retrieved on an encrypted external hard disk. Afterwards, the patient will be placed in the prone position.

Research follow-up visits

At least four follow-up visits will be organized:

1. One between H+12 and H+24 after the start of PP, trying to be as close as possible to H+16.
2. One just before return in the supine position
3. One between one and four hours after return in the supine position
4. One on the day of discharge or death. In the event of a subsequent indication to be returned to the supine position, a visit will be made after each return to the supine position.

During the first three visits, the same information and complementary examinations will be carried out: monitoring by electrical impedance tomography with initial ventilation parameters and during a long inspiratory time. All anonymized ventilator data will be retrieved on an encrypted external hard disk.

At the 3rd follow-up visit, i.e. between H+1 and H+4 after the patient has been put back on his back, we will note the exact date and time when the patient was put back in the supine position, and whether the patient was curarized, put on NO, put on ECMO or put on norepinephrine at any time during the PP session. If one of these interventions was carried out, we'll note the start and end times. Finally, we'll check whether the patient has any pressure sores.

After this first PP session, if subsequent sessions are indicated, the patient will be examined after each return to the supine position, and any pressure sores will be recorded.

End-of-research assessment During the end-of-research visit, complete information on the patient's care in the intensive care unit will be collected.

If the patient has not died, we will record whether he or she has brachial plexus or pressure sore damage. The doctor in charge of the patient in intensive care will be responsible for the various visits.

Method for collecting efficacy assessment parameters:

The distribution of ventilatory volumes during PP sessions will be performed using an electrical impedance tomograph (EIT). The EIT used will be the Dräger PulmoVista500, consisting of two 16-electrode belts. The EIT measures variations in electrical impedance in different areas of the lung by sending electrical pulses at a speed of 50 Hz. This variation in electrical impedance is proportional to the variation in ventilation of the zones in question. So, the greater the delta in ventilation of a lung zone between the end of inspiration and the end of expiration, the greater the variation in its impedance.

The PulmoVista is set to a recording speed of 50 Hz. The filter for suppressing low-frequency signals and the high and low limits of the bandwidth filter will be set individually for each patient to personalize image quality.

To limit the noise associated with patient mobilization and changes in ventilatory parameters, the electrode belt should be installed 20 minutes before each series of measurements. Similarly, before each series of measurements, ventilatory parameters should have been stable for 20 minutes.

The variation in electrical impedance will be recorded by the EIT for 5 minutes.

The inspiratory time is then increased by decreasing the respiratory flow to 10 L/min, the respiratory frequency to 15/min and reducing the ratio of inspiratory time to expiratory time to 1:1. Immediately afterwards, the EIT parameters are recorded for 5 minutes and then the ventilator parameters are reset to their initial values.

Methodology

This is a prospective, bicentric, cohort study. The two recruiting centers will be:

• The intensive care station of Louis-Mourier university hospital in Colombes, part of the greater Paris Hospital, chief of station: Pr Jean-Damien Ricard.
• The intensive care station of Saint-Antoine university hospital in Paris, part of the greater Paris Hospital, chief of station: Pr Eric Maury.

The inclusion period will last a maximum of 12 months. Depending on the length of ICU stay, the total duration of the study will be maximum 14 months.

Inclusion Criteria:

• patient with with ARDS with a P/F ratio < 150 mm Hg regardless of etiology
• invasive ventilated patient
• indication for DV placement determined by the treating practitioner
• BMI≥ 18 kg/m²
• Affiliated to social security

Exclusion Criteria:

• Patients who have already undergone two prone position sessions
• Pregnant women
• Refusal to participate in the research
• Patients under guardianship and trusteeship
• Patients under State Medical Assistance (AME)
• Uncontrolled intracranial hypertension
• Unstable spinal fracture
• Hemodynamic instability defined by MAP < 65 mmHg
• Presence of a pacemaker
• Presence of an implantable defibrillator
• Unresolved pneumothorax or bronchopleural fistula