Comparison of positive end-expiratory pressure-induced increase in central venous pressure and passive leg raising to predict fluid responsiveness in patients with atrial fibrillation.

Abstract

Positive end-expiratory pressure (PEEP)-induced increase in central venous pressure (CVP) has been suggested to be a robust indicator of fluid responsiveness, with heart rhythm having minimal influence. We compared the ability of PEEP-induced changes in CVP with passive leg raising (PLR)-induced changes in stroke volume index (SVI) in patients with atrial fibrillation after valvular heart surgery. In 43 patients with atrial fibrillation after cardiac surgery, PEEP was increased from 0 to 10 cm H2O for 5 min and changes in CVP were assessed. After returning the PEEP to 0 cm H2O, PLR was performed for 5 min and changes in SVI were recorded. Finally, 300 ml of colloid was infused and haemodynamic variables were assessed 5 min after completion of a fluid challenge. Fluid responsiveness was defined as an increase in SVI ≥10% measured by a pulmonary artery catheter. Fifteen (35%) patients were fluid responders. There was no correlation between PEEP-induced increases in CVP and changes in SVI after a fluid challenge (β coefficient -0.052, P=0.740), whereas changes in SVI during PLR showed a significant correlation (β coefficient 0.713, P<0.001). The area under the receiver operating characteristic curve of the PEEP-induced increase in CVP and changes in SVI during PLR for fluid responsiveness was 0.556 [95% confidence interval (CI) 0.358-0.753, P=0.549) and 0.771 (95% CI 0.619-0.924, P=0.004), respectively. A PEEP-induced increase in CVP did not predict fluid responsiveness in patients with atrial fibrillation after cardiac surgery, but increases in SVI during PLR seem to be a valid predictor of fluid responsiveness in this subset of patients.