Prediction of fluid responsiveness for patients in shock using a ventilator disconnection test combined with the pulse contour-derived cardiac index

Abstract

BACKGROUND:

Finding a simple and reliable method to predict and assess fluid responsiveness has long been of clinical interest.

OBJECTIVE:

To investigate the predictive value of a ventilator disconnection (DV) test combined with the pulse contour-derived cardiac output (PiCCO) index on fluid responsiveness for patients in shock.

METHODS:

Thirty-two patients were chosen for the study. Patients who were in shock, received mechanical ventilation, and met the inclusion criteria were selected. Patients were divided into a fluid-responsive group (14 patients) and fluid-unresponsive group (18 patients) based on whether the increase in cardiac index ( $\mathrm{\Delta }$ CI) was $>$ 10% or not, respectively, following the fluid challenge test. Changes in heart rate, pulse oximeter-measured oxygen saturation, mean arterial pressure (MAP), and CI before and after passive leg raising (PLR), DV, and fluid challenge tests were observed. We used Pearson’s correlation coefficient to analyze an increase in the MAP ( $\mathrm{\Delta }$ MAP) and $\mathrm{\Delta }$ CI before and after the PLR, DV, and fluid challenge tests; the sensitivity and specificity of the $\mathrm{\Delta }$ MAP and $\mathrm{\Delta }$ CI in the PLR and DV tests for predicting fluid response were also analyzed by plotting the receiver operating characteristic (ROC) curves.

RESULTS:

CI results in the PLR and DV tests, as well as the fluid challenge test, were significantly higher in the fluid-responsive group compared with before the test ( $P<$ 0.05). The $\mathrm{\Delta }$ CI before and after the PLR, DV, and fluid challenge tests were positively correlated among patients in the fluid-responsive group. The area under the ROC curve for the post-PLR test CI and the post-DV CI for predicting fluid responsiveness was 0.869 (95% confidence interval (CI) [0.735–1.000, $P=$ 0.000]) and 0.937 (95% CI [0.829–1.000, $P=$ 0.000]), respectively, in patients in the fluid-responsive group. The sensitivity and specificity of the post-DV CI for predicting fluid responsiveness in all patients was 100.0% and 88.9%, respectively, using a 5% increase as the cut-off value.

CONCLUSION:

Application of DV, combined with PiCCO, has a high predictive value for fluid responsiveness among patients in shock.