Impedance cardiography in the measurement of cardiac output: Studies in rabbits

A thoracic electric bioimpedance device with improved signal processing was used to noninvasively measure cardiac output in eight New Zealand White rabbits (average wt = 4.7 kg). Prospective correlation was performed between aortic thermodilution and impedance cardiography in a closed chest model. Aortic thermodilution was compared to the electromagnetic flowmeter in an open chest model. In four rabbits, the change in the impedance signal (dZ/dt) was quantified after repeated mechanical occlusion of the aorta and pulmonary artery. The mean cardiac output as measured by the impedance device was 0.56 ± 0.01 liter/min (range 0.29-1.16 liter/min) compared to 0.53 ± 0.01 liter/min (range 0.25-0.83 liter/min) by aortic thermodilution. For the 116 data pairs, regression analysis revealed a statistically significant agreement (r = 0.82, P < 0.001) between the two techniques. The mean difference between the techniques (bias) was 0.03 liter/min and 81% of the impedance values were within 0.1 liter/min of the individual thermodilution measurements. A statistically significant decline in the mean magnitude of the dZ/dt signal tracing (1.16 ± 0.10 V-pre, 0.31 ± 0.4 V-post, P < 0.005, n = 21) was observed upon aortic arch occlusion. Conversely, pulmonary artery occlusion did not have a statistical effect on the impedance signal (1.07 ± 0.09-pre, 0.95 ± 0.08-post, P > 0.05, n = 20). In conclusion, a significant correlation was observed between impedance cardiography and aortic thermodilution in measurement of cardiac output in sedated, anesthetized rabbits. This simple technique which involves application of skin electrodes may prove useful in measurement of cardiac output in surgical experimental small animal models. Our data support the hypothesis that the impedance signal, dZ/dt, which is an essential component of stroke volume estimation in impedance cardiography, results in great proportion from aortic outflow during the cardiac cycle.