Difficulties for the recruitment of patients

Rapid volume expansion using repeated intravenous fluid boluses is a very common intervention performed in the intensive care unit (ICU) in the early days of resuscitation of patients with shock. Once passed the initial phase of resuscitation, the fluid boluses administered fail to effectively increase the patients' cardiac output in about 50% of cases. Pulse pressure changes or stroke volume changes induced by a Passive Leg Raising (PLR) test have acceptable/good ability to predict fluid responsiveness (in terms of cardiac output change) and may be systematically used in patients with persistent shock with the aim of limiting the total amount of fluid administered to patients by avoiding undue fluid boluses. One may suppose that such a volume expansion management policy could impact morbidity and mortality of shocked patients. Among the predictive indices available in clinical practice, the PLR test has the advantages of being usable regardless of the patients' respiratory status and cardiac rhythm. Changes in left ventricular stroke volume during the PLR test perform better that changes in pulse pressure to predict fluid responsiveness. However, in counterpart, pulse pressure changes during PLR can be assessed without the need of other hemodynamic exploration such central venous pressure measurement or cardiac output monitoring. The investigators hypothesized that strategies using either stroke volume changes or pulse pressure changes induced by the PLR test to decide wether a fluid bolus clinically deemed indicated should or should not be administered, may limit the amount of fluid received by the patients during the first 5 days of shock, improve their oxygenation index, and shorten the time passed under mechanical ventilation, as compared to a "liberal" strategy (usual care) that does not use predictive indices of fluid responsiveness.

A pilot, single-center, randomized, open-label, 3 parallel groups clinical trial with 1:1:1 assignment of interventions, comparing outcomes between patients with persistent shock assigned to either 1) the "Volume expansion guided by PLR-induced changes in Stroke Volume" strategy, or 2) the "Volume expansion guided by PLR-induced changes in Pulse Pressure" strategy, or 3) to usual care (i.e., without the use of any predictive index of fluid responsiveness). Patients in shock (either of septic, cardiac or other origin) will be included once passed the first hours of resuscitation. The time window for inclusion will be from 6 to 24 hours after the beginning of resuscitation, a delay necessary to ensure that initial hypovolemia has been fully compensated.

The randomization will be stratified according to the presumed origin of shock (cardiac, septic, or other) and according to the PaO2/FiO2 ratio (<200 or ≥ 200 mmHg).

The randomly assigned intervention will be used during the first 5 days of shock (120 hours).

Pre-eligibility criteria:

• Patients presenting on Intensive Care Unit (ICU) admission or later during their ICU stay with arterial hypotension (Mean arterial pressure [MAP]<65mmHg or systolic arterial pressure [SAP]< 90 mmHg at least two times over a 15-minute interval), regardless of the method used to measure blood pressure (invasive or non invasive),
• Or patients already receiving continuous iv vasopressor therapy on ICU admission, regardless of the blood pressure level, will be considered as potential candidates for inclusion in the study, depending on his/her clinical evolution after 6 to 24 hours of care according to current guidelines for shock management.
• In case arterial hypotension is present and/or the patient is already receiving continuous iv vasopressor at ICU admission, the date and time of admission will be considered as the time of shock onset (H0).
• In case arterial hypotension occurs after ICU admission, the date and time of the first episode of hypotension will be considered as the time of shock onset (H0).

Inclusion Criteria:

• Patients will be eligible for inclusion if 6 hours after H0 (and before H24) all the following conditions are present:
• Informed consent obtained (or emergency inclusion possible when legal representatives and patient's family are not present, as allowed by the Ethic Committee and by the French Law)
• Persistence of arterial hypotension (as defined above) or continuous iv infusion of norepinephrine
• Patient under invasive mechanical ventilation
• At least one of the following conditions is present
• Alteration of consciousness
• Mottling skin
• Cyanosis of the extremities despite SaO2>90%
• Oliguria defined as urine output below 0.5ml/kg of actual body weight over at least 1 hour
• Arterial lactate > 2 mmol/L
• ScvO2 < 70%
• Central venous (internal jugular or subclavian) or femoral venous catheter in place
• Arterial catheter in place

Exclusion Criteria:

1. Shock began more than 24 hours ago
2. Age < 18 yrs.
3. Pregnancy
4. Uncontrolled haemorrhage
5. Intra-aortic balloon counterpulsation in place
6. Patient under veno-arterial extracorporeal membrane oxygenation (V-A ECMO)
7. Cardiogenic pulmonary oedema during the past 24 hours
8. Patient equipped with a cardiac output monitoring device
9. Brain death
10. Moribund patient
11. Traction of a lower limb
12. Amputated lower limb above the ankle level
13. Documented intra-abdominal hypertension

13. Cardiac arrest motivating ICU admission or cardiac arrest since admission 14. Patient without social security number 15. Consent refusal