Noninvasive stroke volume variation using electrical velocimetry for predicting fluid responsiveness in dogs undergoing cardiac surgery

Objective To evaluate the ability of a noninvasive cardiac output monitoring system with electrical velocimetry (EV) for predicting fluid responsiveness in dogs undergoing cardiac surgery. Study design Prospective experimental trial. Animals A total of 30 adult Beagle dogs. Methods Stroke volume (SV), stroke volume variation (SVV) and cardiac index were measured using the EV device in sevoflurane-anaesthetized, mechanically ventilated dogs undergoing thoracotomies for experimental creation of right ventricular failure. The dogs were considered fluid responsive if stroke volume (SVI; indexed to body weight), measured using pulmonary artery thermodilution, increased by 10% or more after volume loading (10 mL kg^–1). Relationships of SVV, central venous pressure (CVP) and pulmonary artery occlusion pressure (PAOP) with SVI were analysed to estimate fluid responsiveness. Results Better prediction of fluid responsiveness, with a significant area under the receiver operating characteristic curve, was observed for SVV (0.85 ± 0.07; p = 0.0016) in comparison with CVP (0.65 ± 0.11; p = 0.17) or PAOP (0.60 ± 0.12; p = 0.35), with a cut-off value of 13.5% (84% specificity and 73% sensitivity). Conclusions and clinical relevance SVV derived from EV is useful for identification of dogs that are likely to respond to fluids, providing valuable information on volume status under cardiothoracic anaesthesia.