Background: Monitoring of cardiovascular function is essential during major pediatric and pediatric cardiac surgery. Invasive monitoring of cardiac output (CO) and oxygen delivery is expensive and sometimes associated with adverse events. Therefore, we investigated the accuracy of a new noninvasive CO monitoring device using electrical velocimetry (EV) in comparison with the more invasive transpulmonary thermodilution (TPTD) method.
Methods: In five fasted, anesthetized and mechanically ventilated piglets, CO was measured simultaneously using EV and TPTD under normal conditions, volume loading, inotropic support and exsanguination.
Results: In five piglets, 169 measurements could be performed. The correlations between EV-CO and TPTD-CO were significant for absolute values (P < 0.0001, r = 0.82) and relative changes from baseline (P < 0.0001, r = 0.93). The receiver operating characteristic (ROC) curve analysis of the relative changes of the EV-CO values in relation to the first EV-CO measurement showed a sensitivity of 91% and specificity of 94% (AUC 0.974, 95% CI 0.96-0.99). Changes in TPTD-CO greater than 15% lead to a change of EV-CO in the same direction in 93%. Bland-Altman analysis showed a mean difference between the two methods of -0.63 l x min(-1) with an sd of 0.64 l x min(-1). The lower and upper limits of agreement were -1.88 and 0.62 l x min(-1), percentage limit of agreement was +/-82.8%.
Conclusions: The results show that EV is a safe, simple, noninvasive and cost-effective method for continuous trend monitoring of CO in piglets. The agreement of the EV-CO with TPTD-CO is not good enough to replace the standard method in our animal model. A correction factor for body habitus in piglets may be beneficial.