Noninvasive cardiac output: simultaneous comparison of two different methods with thermodilution.

The authors attempted to simultaneously measure cardiac output by thermodilution (COtd), thoracic bioimpedance (CObi), and suprasternal Doppler ultrasound (COdopp) in 68 patients. Subgroups separately compared included patients whose lungs were mechanically ventilated, patients undergoing cardiac surgery, aortic surgery, patients with dysrhythmias, and patients with sepsis. The authors also studied the value of the ventricular ejection time (VET) in evaluating the agreement of CObi and COdopp with COtd. Simultaneous CObi and COtd were available in a total of 56 patients (416 data sets) with an overall correlation coefficient r = 0.61, regression slope (m) of 0.52, intercept (y) of 2.46, and mean (CObi-COtd) difference (bias) of -0.67 +/- 1.72 (SD) l/min. Simultaneous COdopp and COtd were available in 59 patients (446 data sets) with an overall r = 0.51, m of 0.53, y of 2.05, and bias of -0.79 +/- 1.95 l/min. CObi agreed most closely with COtd in patients whose lungs were mechanically ventilated, who had not undergone cardiac or aortic surgery, and with VET difference less than 40 ms (16 patients, 99 data sets; r = 0.74; m = 0.97; y = 0.15; bias = -0.02 +/- 1.53 l/min). COdopp agreed most closely with COtd in patients whose lungs were mechanically ventilated, who had not undergone cardiac or aortic surgery, and in sinus rhythm with VET difference less than 40 ms (10 patients, 45 data sets; r = 0.82; m = 0.98; y = -0.07; bias = -0.82 +/-1.03 l/min). VET by radial artery can help evaluate the reliability of CObi and COdopp.