Determination of cardiac output using ensemble-averaged impedance cardiograms

Although impedance cardiography provides safe and reliable noninvasive estimates of stroke volume in humans, its usefulness is limited by the necessity for subjects to be apneic and motionless. In an effort to circumvent this restriction we studied the validity of ensemble-averaging of impedance data in exercising normal subjects and in intensive-care patients. The correlation coefficient (r value) between 128 ensemble-averaged and standard hand-digitized determinations of stroke volume index from the same records taken during rest and exercise in six normal male subjects was +0.97 (P less than 0.001). The r value for ensemble-averaged stroke volume indices during free breathing and breath hold in the same subjects was +0.92 (P less than 0.001), suggesting that breath hold did not significantly affect the stroke volume estimation. In 14 freely breathing hospital intensive-care patients the r value between simultaneous thermodilution cardiac output readings and ensemble-averaged impedance determinations was +0.87 (P less than 0.01). The results indicate that ensemble-averaging of transthoracic impedance data provides waveforms from which reliable estimates of cardiac output can be made during normal respiration in healthy human subjects at rest and exercise and in critically ill patients.