Stroke volume variation as a predictor of fluid responsiveness in patients undergoing brain surgery

Changes in arterial blood pressure induced by mechanical ventilation allow assessment of cardiac preload. In this study, stroke volume variation (SVV), which is the percentage change between the maximal and minimal stroke volumes (SV) divided by the average of the minimum and maximum over a floating period of 30 s, continuously displayed by the PiCCO continuous cardiac output monitor, was evaluated as a predictor of fluid responsiveness. Fifteen patients undergoing brain surgery were included. During surgery, graded volume loading was performed with each volume loading step (VLS) consisting of 100 mL of 6% hydroxyethylstarch given for 2 min. Successive responsive VLSs were performed (increase in SV > 5% after a VLS) until a change in SV of <5% was reached (nonresponsive). A total of 140 VLSs were performed. Responsive and nonresponsive VLSs differed in their pre-VLS values of systolic blood pressure, SV, and SVV, but not in the values of heart rate and central venous pressure. By using receiver operating characteristic analysis, the area under the curve for SVV (0.870, 95% confidence interval [CI]: 0.809 to 0.903) was statistically more than those for central venous pressure (0.493, 95% CI: 0.397 to 0.590, P = 7 × 10^-10), heart rate (0.593, 95% CI: 0.443 to 0.635, P = 5.7 × 10^-10), and systolic blood pressure (0.729, 95% CI: 0.645 to 0.813, P = 4.3 × 10^-3). An SVV value of 9.5% or more, will predict an increase in the SV of at least 5% in response to a 100-mL volume load, with a sensitivity of 79% and a specificity of 93%.