Diagnostic accuracy of the peripheral venous pressure variation induced by an alveolar recruitment maneuver to predict fluid responsiveness during high-risk abdominal surgery

Abstract Background In patients undergoing high-risk surgery, it is recommended to titrate fluid administration using stroke volume or a dynamic variable of fluid responsiveness (FR). However, this strategy usually requires the use of a hemodynamic monitor and/or an arterial catheter. Recently, it has been shown that variations of central venous pressure (ΔCVP) during an alveolar recruitment maneuver (ARM) can predict FR and that there is a correlation between CVP and peripheral venous pressure (PVP). This prospective study tested the hypothesis that variations of PVP (ΔPVP) induced by an ARM could predict FR. Methods We studied 60 consecutive patients scheduled for high-risk abdominal surgery, excluding those with preoperative cardiac arrhythmias or right ventricular dysfunction. All patients had a peripheral venous catheter, a central venous catheter and a radial arterial catheter linked to a pulse contour monitoring device. PVP was always measured via an 18-gauge catheter inserted at the antecubital fossa. Then an ARM consisting of a standardized gas insufflation to reach a plateau of 30 cmH2O for 30 s was performed before skin incision. Invasive mean arterial pressure (MAP), pulse pressure, heart rate, CVP, PVP, pulse pressure variation (PPV), and stroke volume index (SVI) were recorded before ARM (T1), at the end of ARM (T2), before volume expansion (T3), and one minute after volume expansion (T4). Receiver-operating curves (ROC) analysis with the corresponding grey zone approach were performed to assess the ability of ∆PVP (index test) to predict FR, defined as an ≥ 10% increase in SVI following the administration of a 4 ml/kg balanced crystalloid solution over 5 min. Results ∆PVP during ARM predicted FR with an area under the ROC curve of 0.76 (95%CI, 0.63 to 0.86). The optimal threshold determined by the Youden Index was a ∆PVP value of 5 mmHg (95%CI, 4 to 6) with a sensitivity of 66% (95%CI, 47 to 81) and a specificity of 82% (95%CI, 63 to 94). The AUC’s for predicting FR were not different between ΔPVP, ΔCVP, and PPV. Conclusion During high-risk abdominal surgery, ∆PVP induced by an ARM can moderately predict FR. Nevertheless, other hemodynamic variables did not perform better.