LAUSR

OBJECTIVE To compare electrical velocimetry (EV) noninvasive measures of cardiac output (CO) and stroke volume variation (SVV) in dogs undergoing cardiovascular surgery with those obtained with the conventional thermodilution technique using a pulmonary artery catheter. STUDY DESIGN Prospective experimental trial. ANIMALS Seven adult Beagle dogs with a median weight of 13.6 kg. METHODS Simultaneous, coupled cardiac index (CI; CO indexed to body surface area) measurements by EV (CIEV) and the reference pulmonary artery catheter thermodilution method (CIPAC) were obtained in seven sevoflurane-anaesthetized, mechanically ventilated dogs undergoing experimental open-chest cardiovascular surgery for isolated right ventricular failure. Relationships between SVV or central venous pressure (CVP) and stroke volume (SV) were analysed to estimate fluid responsiveness. Haemodynamic data were recorded intraoperatively and before and after fluid challenge. RESULTS Bland-Altman analysis of 332 matched sets of CI data revealed an overall bias and precision of - 0.22 ± 0.52 L minute-1 m-2 for CIEV and CIPAC (percentage error: 30.4%). Trend analysis showed a concordance of 88% for CIEV. SVV showed a significant positive correlation (r2 = 0.442, p < 0.0001) with SV changes to a volume loading of 200 mL, but CVP did not (r2 = 0.0002, p = 0.94). Better prediction of SV responsiveness (rise of SV index of ≥ 10%) was observed for SVV (0.74 ± 0.09; p = 0.014) with a significant area under the receiver operating characteristic curve in comparison with CVP (0.53 ± 0.98; p = 0.78), with a cut-off value of 14.5% (60% specificity and 83% sensitivity). CONCLUSIONS AND CLINICAL RELEVANCE In dogs undergoing cardiovascular surgery, EV provided accurate CO measurements compared with CIPAC, although its trending ability was poor. Further, SVV by EV, but not CVP, reliably predicted fluid responsiveness during mechanical ventilation in dogs.