Validity of cardiac output measurement by computer-averaged impedance cardiography, and comparison with simultaneous thermodilution determinations.

The accuracy and reproducibility of noninvasive cardiac output determinations by computer-averaged impedance cardiography were compared with those of simultaneously performed thermodilution cardiac output. In all, 43 patients (14 men and 29 women = 201 pairs) were studied by simultaneously performed impedance and thermal determinations. Individual impedance values correlated with paired thermodilution determinations (r = 0.75; p less than 0.0001). Each patient's average thermodilution values correlated with the average impedance values (r = 0.86; p less than 0.0001). Mean thermodilution output was 4.6 +/- 1.37 liters/min. Mean impedance output was 4.5 +/- 1.27 liter/min Reproducibility was comparable for impedance (0.0059 +/- 0.639) and thermodilution cardiac output (0.023 +/- 0.556). There was high agreement between methods by plot of the difference against mean of the 2 methods. Impedance cardiac output values agree and correlate highly with quality-controlled thermodilution outputs across a wide range of clinical conditions and hemodynamic values.