Objective: To assess the degree of error of the BoMed NCCOM3 model revision seven (R7) impedance cardiograph in determining stroke volume and estimated cardiac output.
Design: Three-group, within-subject, repeated measures design.
Sample: Group 1: patients (n = 17) with heart disease undergoing an elective coronary angiogram; group 2: patients (n = 28) after elective heart surgery; and group 3: healthy volunteers (n = 28).
Measurement: Cardiac output was determined by the BoMed NCCOM3-R7 impedance cardiograph, Fick principle, and thermodilution method. The NCCOM3-R7 was compared with the direct Fick and thermodilution methods in groups 1 and 2, respectively, to estimate validity coefficients. In group 3, repeated measures were obtained with the NCCOM3-R7 to calculate reliability coefficients.
Results: The NCCOM3-R7 underestimated Fick measurements by 1.050 +/- 1.529 L/min at rest and 1.505 +/- 2.214 L/min during exercise. Correlation coefficients of 0.684 at rest (p = 0.001) and 0.219 during exercise (p = 0.248) were obtained. The NCCOM3-R7 underestimated thermodilution values by 0.425 +/- 1.325 L/min in subjects initially after heart surgery and 0.358 +/- 1.235 L/min 2 to 4 hours later. Correlation coefficients of 0.547 (p = 0.002) and 0.505 (p = 0.004) were obtained for the two time periods, respectively. A reliability coefficient of 0.837 was calculated with healthy subjects.
Conclusion: The NCCOM3-R7 has a clinically unacceptable level of error for evaluating cardiac performance in patients with heart disease.