| Background:Pleth variability index (PVI) is a novel non-invasive algorithm capable of, automated and continuous calculation of the respiratory variations in the pulse oximeter waveform amplitude. Some studies showed that it can predict fluid responsiveness and the hemodynamic effect in mechanically ventilated patients under general anaesthesia. Most studies were conducted under general anesthesia and confined during the period after induction and before the beginning of the operation to eliminate the influence of incision. According to its mechanism, accuracy of PVI is influenced by respiratory cycle, volume statue and vascular tone. Peripheral vascular tone can be alerted by the stress during the operation and the cadiovascular effect of epidural block, which can affect the accuracy of PVI. To our knowledge, no researches have been done to assess the accuracy of PVI to predict intraperative fluid responsiveness in patients under combined general and epidural anesthesia.Objective:1、To assess whether PVI can predict fluid responsiveness during the operation in patients under combined general and epidural anesthesia.2、To compare the ability of PVI to predict fluid responsiveness with stroke volume variation(SVV), mean arterial pressure (MAP) and heart rate (HR)Methods:Eighty-one patients undergoing elective hepatectomy were randomized into3groups:general anesthesia group (group G:epidural drug was normal saline), low concentration bupivacaine group (group L:epidural drug was0.25%bupivacaine) and high concentration bupivacaine group (group H:epidural drug was0.5%bupivacaine). All these three groups followed the same inductive protocol. Operation started immediately after10ml loading dose of epidural local anesthetic in each group. Volume challenge with250ml Hydroxyethyl starch was done20min after bolus administrations of epidural local anesthetic. PVI, SVV, MAP, HR and cardic index (CI) were measured before epidural local anesthetic administration (T1), before (T2) and after volume expansion (T3). The patients with an increase of CI (△CI) equal to or larger than15%after volume challenge were defined as responders. Receivers operating characteristic (ROC) curves were generated for PVI, SVV, MAP and HR to verify the discriminating threshold of each and areas under the ROC curves (AUC) were calculated and compared. Heamodynamic variables were analyzed with one-way ANOVA among groups. Paired t-test was performed to detect the inter-group difference. Pearson linear correlation and analysis of Kappa consistency was used to test correlation between PVI and SVV in each group. P-value of<0.05was considered as statistically significant.Results:1、In group G, PVI>12%before volume expansion discriminated between responders and non-responders with87.50%sensitivity and83.33%specificity. Its AUC was0.8819. SVV>10%before volume expansion discriminated between responders and non-responders with75.00%sensitivity and100%specificity. Its AUC was0.8681. In group L, the cutoff of PVI to discriminate between responders and non-responders was8%with75.00%sensitivity and55.56%specificity. AUC was0.6285. As for SVV, the cutoff was8%with75.00%sensitivity and55.56%specificity. AUC was0.7326.In group H, the AUC of PVI was less than0.5(0.3922). However, SVV>8%before volume expansion discriminated between responders and non-responders with76.47%sensitivity and55.56%specificity. Its AUC was0.7549. It meant that PVI was not a reliable variable to predict fluid responsiveness under this situation.2、Analysis of Kappa consistency between SVV and PVI suggested that the agreement between these two variables was moderate in group G(κ=0.412), while low in group L and group H (κL=0.199, κH=0.154)Conclusions:Combined general and epidural anesthesia greatly affect the accuracy of PVI to predict fluid responsiveness during operation. We must consider the impact of the change of vascular tone on the accuracy of PVI in clinical application. |
PDF Full Text Request