Cardiac Output Measurement by Bioimpedance and Noninvasive Pulse Contour Analysis Compared With the Continuous Pulmonary Artery Thermodilution Technique

Objective: The aim of the present study was to compare 2 noninvasive cardiac output measurement methods with the continuous cardiac output thermodilution (CCO-TD) method. Design: A single-center prospective design. Setting: A university hospital. Participants: Fifty-three consecutive patients scheduled for elective, non-emergent cardiac surgery. Interventions: With each participant the cardiac output was measured using 3 methods: CCO-TD, the Endotracheal Cardiac Output Monitor (ECOM), and the Nexfin monitor. Measurements and Main Results: Measurements were performed simultaneously at 7 time points: After induction, before cardiopulmonary bypass, after cardiopulmonary bypass, after protamine, after arrival in the intensive care unit, and before extubation on postoperative day 1. Statistical analysis was performed using Pearson's correlation, Bland-Altman, percent error, and polar plots. Compared to CCO-TD, ECOM showed significant correlation of R0.619 with a bias of -0.13 L/min (95% confidence interval -2.19-1.93 L/min), a percent error of 40%, and trending ability of 87% and 97% within 0.5 L/min and 1.0 L/min, respectively. The Nexfin monitor showed significant correlation of R0.535 with a bias of -0.35 L/min (95% confidence interval -3.36-2.66 L/min), a percent error of 58% and trending ability of 84% and 97% were within 0.5 L/min and 1.0 L/min limits of agreement. Conclusions: Neither the ECOM nor the Nexfin had the ability to replace the thermodilution-based continuous cardiac output monitor. The ECOM did not have acceptable accuracy or trending ability and only could be utilized for intubated patients. The Nexfin lacked reliability and trending ability. Also, the Nexfin did not provide consistent results.