Central venous pressure as an adjunct to flow-guided volume optimisation after induction of general anaesthesia.

BACKGROUND Although the central venous pressure (CVP) is often used as a guide to volume status in major surgery and intensive care, fluid therapy should be guided by the response of the stroke volume (SV) to a fluid bolus. The present study evaluates whether the central venous pressure (CVP) can serve as an adjunct to decisions of whether or not fluid should be infused. METHODS Stroke volume (SV) and stroke volume variation (SVV) was monitored with FloTrac/Vigileo and the CVP were measured in 80 patients just before general anaesthesia was induced (baseline) and then, before each of three successive bolus infusions of 3 mL kg⁻¹ of 6% hydroxyethyl starch 130/0.4. A patient showed fluid responsiveness and was denoted a "responder" if SV increased by ≥10% from the bolus infusion. RESULTS The CVP was higher in non-responders (mean 7.2 mm Hg) than in responders (mean 5.8 mm Hg, P < 0.0001). In non-responders but not in responders, the absence or presence of a rise in CVP improved the prediction of whether the patient would show fluid responsiveness during the next fluid bolus. For example, if no rise in CVP occurred the chance was 48% of subsequent fluid responsiveness, while this chance was only 9% for those who had an increase in CVP (P < 0.004). There was only a fair concordance between fluid responsiveness as indicated by SV and SVV (Cohen´s kappa 0.28). CONCLUSIONS A low CVP suggests that the patient is lower on the Frank-Starling curve than indicated by SV as measured by FloTrac/Vigileo.