Lack of agreement between esophageal doppler cardiac output measurements and continuous pulse contour analysis during off-pump cardiac surgery

Objective: Transesophageal echo-Doppler cardiac output as well as arterial pulse contour analyses cardiac output are increasingly used for cardiac output monitoring. No data are available whether both techniques may be used interchangeably in patients undergoing cardiac surgery. Design: Prospective, observational study Setting: Operating rooms of a university affiliated hospital. Patients: 30 patients undergoing elective coronary artery bypass grafting surgery. Interventions: None Measurements: 900 paired cardiac output measurements were obtained by pulse contour analysis following transpulmonary thermodilution equilibration by the PiCCO system (PiCCO, Pulsion, Munich, Germany) and by the HemoSonic esophageal doppler monitor (HemoSonic 100; Arrow International, Reading, PA). Measurements were performed within the first hour after induction of anesthesia. Results: Bland-Altman analysis of the complete data set showed a mean difference (bias) of 0.12 l/min (95% CI -0.06 to -0.18) with limits of agreement + 1.8 l/min to -1.6 l/min (upper 95% CI 1.78 to 1.98; lower 95% CI -1.74 to -1.54), the percentage error was + 37% to -44.5%. Transesophageal echo-Doppler cardiac output closely correlated (r = 0.75, p < 0.0001) with pulse-contour analyses cardiac output. Conclusions: Several studies have shown the accuracy of calibrated pulse contour cardiac output measurements in patients undergoing cardiac surgery. Thus, the present data question the reliability of transesophageal echo-Doppler derived cardiac output measurements in this setting and may have implications for using transesophageal echo-Doppler during goal-directed hemodynamic optimization.