Objective: Fluid therapy after resuscitation from cardiac arrest is challenging since both hypovolemia and fluid overload may cause circulatory failure. Therefore, prediction of fluid responsiveness is a major issue in optimizing hemodynamic therapy. The aim of the present study was to evaluate the performance of stroke volume variation, pulse pressure variation, variation of Doppler-derived velocity time integral, and global end-diastolic volume index to predict fluid responsiveness in the postcardiac arrest period.
Design: Prospective animal study.
Setting: University-affiliated research laboratory.
Subjects: Twenty anesthetized and ventilated Goettinger minipigs.
Intervention: Animals were equipped with a central venous catheter, a thermistor-tipped arterial catheter, and a transesophageal echo probe. Electrically induced cardiac arrest of 8 mins was followed by cardiopulmonary resuscitation. Hemodynamic measurements were performed before and after a two-step fluid bolus at baseline and both 1 and 4 hrs after return of spontaneous circulation. Fluid responsiveness was defined by an increase in stroke volume of at least 15%. Performance of variables was analyzed using receiver operator characteristics analysis.
Measurement and main results: Variables reliably indicated fluid responsiveness at baseline. Fifteen animals were successfully resuscitated. Left ventricular ejection fraction was significantly reduced 1 hr after return of spontaneous circulation (52.6% ± 6.4%; p < .01) compared with baseline (69.9% ± 5.3%). One hour after return of spontaneous circulation, fluid responsiveness could not be predicted by any variable. In contrast, pulse pressure variation, variation of the velocity time integral, and global end-diastolic volume index, but not stroke volume variation, were able to predict fluid responsiveness 4 hrs after return of spontaneous circulation, since area under the curve was 0.85 (p < .01), 0.94 (p < .01), 0.77 (p = .02), and 0.68 (p = .12), respectively.
Conclusions: Prediction of fluid responsiveness failed 1 hr after successful cardiopulmonary resuscitation from cardiac arrest. Four hours after return of spontaneous circulation, however, the variables pulse pressure variation, variation of the velocity time integral, and global end-diastolic volume index, but not stroke volume variation, enabled prediction of fluid responsiveness and may, therefore, be considered for subsequent hemodynamic optimization after successful cardiopulmonary resuscitation.