Dynamic preload indicators fail to predict fluid responsiveness in open-chest conditions

Objective: Dynamic preload indicators like pulse pressure variation (PPV) and stroke volume variation (SVV) are increasingly being used for optimizing cardiac preload since they have been demonstrated to predict fluid responsiveness in a variety of perioperative settings. However, in open-chest conditions, the value of these indices has not been systematically examined yet. We, therefore, evaluated the ability of PPV and SVV to predict fluid responsiveness under open- and closed-chest conditions. Design. Prospective, controlled, clinical study. Setting: University hospital. Patients. Twenty-two patients scheduled for elective coronary artery bypass graft surgery. Interventions: Defined volume loads (VL) (10 mL kg(-1) hydroxyethyl starch 6%) intra- and postoperatively. Measurements and Main Results. Stroke volume index was measured 1) before and after a VL intraolperatively in open-chest conditions, and 2) under closed-chest conditions within 1 hour after arrival in the intensive care unit. Central venous pressure and global end diastolic volume were assessed as static preload indicators. In addition, PPV and SVV (both obtained with PiCCO system) were recorded. Fluid-responders were defined by an increase in stroke volume index >= 12% subsequent to the VL. Receiver operating characteristic analysis showed that all preload indicators failed to predict fluid responsiveness in open-chest conditions. Under closed-chest conditions, the areas under the receiver operating characteristic curve for PPV and SVV were 0.884 (p = 0.004) and 0.911 (p = 0.003), respectively, whereas the static and volumetric preload parameters failed to predict fluid responsiveness. A PPV of >= 10% identified fluid-responders with a sensitivity of 64% and a specificity of 100%, while a SVV of >8% identified fluid-responders with a sensitivity of 100% and a specificity of 78%. Conclusions. Our results suggest that the dynamic preload indicators PPV and SVV are able to predict fluid responsiveness under closed-chest conditions, whereas all static and dynamic preload indicators fail to predict fluid responsiveness under open-chest conditions. (Crit Care Med 2009; 37:510-515)