Acute respiratory failure represents a frequent cause of admission to intensive care units (ICUs). In the absence of tailored interventions, it poses an imminent threat to patients' lives. Most patients admitted in ICU undergo fluid expansion to enhance oxygen delivery and preserve cellular function.

This practice is grounded in the concept of " preload responsiveness ". However, the accrual of positive fluid balance resulting from fluid administration is now acknowledged as an autonomous risk factor for mortality.

Consequently, preload unresponsiveness assumes a pathological character, potentially indicative of fluid overload or right ventricular dysfunction, both deleterious conditions linked to unfavorable outcomes.

Maintaining patients in a preload-responsive state may be interesting to limit fluid expansion and the need of invasive mechanical ventilation.

The objective of this prospective observational study is to evaluate the prognostic significance of preload responsiveness in patients admitted to the ICU with hypoxemic, non-hypercapnic respiratory failure.

• Main objective: To evaluate the association between fluid responsiveness, assessed by the inferior vena cava collapsibility index (cIVC) with trans-thoracic echocardiography within the initial 48 hours post-ICU admission, and mortality or the need for invasive mechanical ventilation by day 28 in patients admitted to the ICU for hypoxemic, non-hypercapnic acute respiratory failure.
• Secondary objectives: To evaluate the association between fluid responsiveness and mortality at day 28 and day 90, the need of invasive mechanical ventilation, and the number of days free from organ support (vasopressors, mechanical ventilation and renal replacement therapy) by day 28.

Upon receipt of both oral and written information, patients will provide non-objection to participate in the study. This prospective single-center study has obtained approval from the Regional Ethics Committee of Ile de France III approval (No. 2022-A02813-40).

Video loops of the inferior vena cava (IVC) will be recorded during a transthoracic echocardiography performed for diagnostic purpose. This method is routinely employed in our ICU to assess preload responsiveness in spontaneously breathing patients admitted for acute respiratory failure. For study purposes, IVC diameters will be measured remotely on anonymized recordings by operators blinded to patients' outcomes. The cIVC will be calculated at 4 cm from the IVC-right atrium junction, using the following equation: (Maximum expiratory diameter - inspiratory diameter)/ Maximum expiratory diameter.

According to prior research conducted by our group (Caplan et al., Ann Intensive Care 2020), a cIVC ≥44% will be utilized to diagnose preload responsiveness.

Inclusion Criteria:

Age > 18 years Registered in the French National Health Insurance system

Presenting all the following criteria:

• Admission in ICU for less than 48 hours
• Type 1 acute respiratory failure: hypoxemia (PaO2< 60 mmHg on room air) without hypercapnia (PaCO2 < 45 mmHg)
• Requiring oxygen support of ≥ 5L./min
• Necessitating transthoracic echocardiography for diagnostic purposes

Exclusion Criteria:

Pregnancy Adult with disability or without social protection BMI > 35 kg/m² Withhold decision regarding invasive mechanical ventilation Moribund state diagnosed as a SOFA score > 20

This is a monocentric observational study performed in the University Hospital of Lille, among adult patients admitted in ICU for hypoxemic and non hypercapnic acute respiratory failure for less than 48 hours