Rationale: Patients with severe heart failure increasingly benefit from extracorporeal circulatory assistance (Extracorporeal Membrane Oxygenation - ECMO) that completely replaces the function of the heart and is in addition to mechanical ventilation (veno-arterial ECMO). In clinical practice, some of these patients have been found to be dyspneic, although circulatory assistance ensures satisfactory hematosis. Indeed, nearly half of the patients undergoing mechanical ventilation in intensive care have a generally intense dyspnea. In addition to the immediate discomfort caused by dyspnea, which can be assimilated to pain, this symptom is associated with a prolonged duration of mechanical ventilation and contributes to the genesis of neuropsychological sequelae of resuscitation such as post-traumatic stress syndrome.

Objectives :

1. To study the effectiveness and impact of an increase in ECMO sweep gas flow through the membrane lung or positive tele-expiratory pressure (PEP) on dyspnea in patients under ECMO.
2. To understand the mechanism involved in this relief of dyspnea. Methods: in 30 patients of Intensive Care Unit, dyspnea will be evaluated by psychophysiological scales and scores as well as by neurophysiological tools (electromyography of inspiratory muscles) during sweep gas flow through the membrane lung incrementation and PEP increase.

Ethical framework: Changes in ventilation parameters and ECMO will be performed in the usual manner, without additional or unusual diagnostic, treatment or monitoring procedures.

Expected results and outlook: the investigators hypothesize that half of the patients on veno-arterial ECMO have dyspnea related to pulmonary edema. Therefore, an increase in PEP would alleviate this dyspnea. The investigators hypothesize that by using a relief method related to increased sweep gas flow through the membrane lung, the investigators can reduce the discomfort without inducing any associated complication.

This study will improve the knowledge of the mechanisms determining dyspnea in patients under ECMO and the principle of therapeutic strategy to improve it.

Half of the patients undergoing mechanical ventilation in intensive care have moderate to severe dyspnea. This dyspnea has a negative impact on the prognosis and quality of life in the medium term.

As ECMOs, techniques for extracorporeal assistance to supplement cardiac and pulmonary functions, are developing rapidly, it seems essential to study the dyspnea associated with the use of these techniques. In particular, understanding the main mechanisms involved in the genesis of dyspnea in patients under ECMO would allow the development of therapeutic solutions aimed at minimizing it.

The investigators hypothesize that, in patients under veno-arterial ECMO with dyspnea, the dyspnea decreases in response to an increase in ECMO sweep gas flow through the membrane lung via modulation of central and peripheral chemoreceptors sensitive to variations in PaO2, PaCO2 and pH. Since dyspnea experienced by patients on veno-arterial ECMO is related to cardiogenic pulmonary edema, increased discharge from the left cavities by increasing positive expiratory pressure should also alleviate dyspnea.

The main objective of the present study is to evaluate the impact of sweep gas flow through the membrane lung and PEP incrementation on dyspnea in patients with veno-arterial ECMO with dyspnea.

Relief of dyspnea will be carried out by the clinician in charge of the patient. He will have complete control of his behaviour. He will carry out this test according to the practices in force in the department.

One of the investigators will be present and will collect the data listed above. This investigator will not intervene in the conduct of the test.

Each patient will participate in a dyspnea relief test according to the protocol of the department and according to the management of the practitioner in charge of the patient.

Initially a recording of 5 minutes will be made in the basal state, before any therapeutic action.

If a sweep gas flow through the membrane lung increment has been decided upon to relieve dyspnea, a new recording will be made after each scan increment and the patient will be asked at each step.

In ventilated or non-ventilated patients in whom the decision to implement non-invasive ventilation has been made, an increase in PEEP in 2 cmH2O steps without exceeding a plateau pressure of 25 cmH2O and a VT of 10ml/kg of the patient's theoretical weight will be achieved.

Inclusion Criteria:

• Circulatory support by veno-arterial ECMO for cardiogenic shock;
• Presence of severe respiratory distress;
• Presence of dyspnea ≥ 40 mm on a visual analogue scale of dyspnea (VAS-dyspnea)
• Decision by the clinician in charge of the patient to either modify the ECMO parameters or to administer non-invasive ventilation in order to decrease the intensity of the dyspnea;
• presence of an arterial catheter; :
• RASS score between 0 and +1 ;
• Absence of delirium on the CAM-ICU score;
• being socially insured.

Exclusion Criteria:

• patient unable to express opposition;
• protected minor or adult patient;
• pregnant woman;
• double veno-arterial ECMO; central ECMO; left ventricular assist or long-term bi-ventricular assist

Patients under veno-arterial ECMO with dyspnea