Background: In this study we quantify the ability of dynamic cardiovascular parameters measured by the PulseCO™ algorithm of the LiDCO™plus monitor to predict the response to a fluid challenge in post-operative patients.
Methods: Surgical patients, admitted to the Intensive Care Unit from the operating theatre were monitored with the LiDCO™plus system. A number of static and dynamic cardiovascular measurements were recorded before and after a fluid challenge. Receiver Operator Characteristic (ROC) curve analysis was used to identify the baseline values, with optimum sensitivity and specificity, to predict responsiveness to a fluid challenge.
Results: Thirty-one patients were enrolled, and received protocol-based fluid challenges. Twelve (38%) responded by demonstrating an increase in stroke volume of >15%. Heart rate (HR) and central venous pressure (CVP) were not statistically different between responders and non-responders. Mean arterial pressure (mAP), systolic pressure variation (SPV), pulse pressure variation (PPV) and stroke volume variation (SVV) were statistically different between responders and non-responders. Parameters with a ROC area under the curve (AUC) significantly >0.5 included SPV 0.70 (0.52-0.88) P=0.046, PPV 0.87 (0.76-0.99) P<0.0002 and SVV 0.84 (0.71-0.96) P=0.0005. The best cut-off values (sensitivity and specificity) to predict fluid were SPV >9 mmHg (73%, 76%), PPV >13% (83%, 74%) and SVV >12.5% (75%, 83%). ROC analysis did not show the AUC to be significantly >0.5 for HR, mAP and CVP CONCLUSION: Dynamic indices measured by PulseCO™ (LiDCO) have a high sensitivity and specificity in predicting fluid responsiveness in sedated and mechanically ventilated patients. A cut-off value for PPV of 13% is the most sensitive and specific indicator of fluid responsiveness.