Fluid expansion in critically ill patients following high risk surgery may induce fluid overload and worse outcome. Several tools have been developped to predict fluid responsiveness in such situation in order to avoid inappropriate fluid administration but with several limitations. Inferior vena cava (IVC) distensibility is one of those tools which has the advantage to be non-invasive, dynamic and safe, is usually measured by subcostal (SC)approach. In post surgical setting this acess is limited du to practical reasons (scar, dressing...), therefore a transhepatic (TH) approach is used but has not been validated as a fluid responsiveness prediction tool. The correlation between SC approach with the TH approach vary according to studies. Therefore the performances, the threshold identified for SC approach can not be translated to the TH approach. Further, fluid congestion status measured before IVC analyses, may be a useful confounder and safety endpoint for fluid responsiveness interpretation.

The primary objective of this study is therefore to study the performance of the IVC measured using TH approach (IVCth) in predicting of fluid responsiveness defined as an increased of 10% and over of stroke volume.

Secondary objectives intend to analyse the correlation between TH and SC approaches, to compared their performances for fluid responsiveness prediction, and to analyse the weight of venous congestion on fluid responsiveness prediction.

Inappropriate volume expansion in the critically ill patient may leads to an increase in fluid balance, fluid overload and worse outcome. This increase in fluid balance is associated with a poor prognosis in the ICU patients, especially after cardiac surgery and high risk surgery .

Generally, the primary objective of an adequate volume expansion is a significant increase in stroke volume called fluid responsiveness (> 10% of increase). This increase in cardiac output promotes tissue perfusion, thus avoiding the occurrence of organ dysfunction.

The identification of fluid responsiveness in these patients is thus a cardinal element of haemodynamic management in intensive care. Among the non-invasive tools to assess this fluid responsiveness, the assessment of the ventilatory or forced inspiration distensibility of the inferior vena cava (IVC) has been proposed with convincing results in several clinical studies. However with various diagnostic performances and threshold.

The measurement of the diameter of the IVC is classically performed via the subcostal (SC) or subxiphoid approach. However, this approach is not always easily accessible in intensive care patients, particularly after cardiovascular surgery or laparotomy, due to practical constraints (algic scars, dressings, prone position). The transhepatic approach (TH) is a technically feasible approach in such cases. However, it has little validation in the literature, particularly in its ability to predict response to volume expansion. knowing that the correlation between the SC approach with the TH approach varies according to studies, the performance and the threshold identified for SC approach can not be translated into that of the TH approach. Further, fluid congestion status measured before IVC analyses, may be a useful confounder and safety endpoint for fluid responsiveness interpretation.

Inclusion Criteria:

• High risk cardiovascular surgery patients admitted in ICU
• With transhepatic IVC and stroke volume measurement feasible.
• Informed consent

Exclusion Criteria:

• Righ ventricle failure
• Significant tricuspid regurgitation
• Tamponade
• Severe hypoxemia
• Age <18 ans
• No health care coverage

Patients admitted to our intensive care unit following cardiac surgery or aortic surgery, under mechanical ventilation or not.