Esophageal doppler is discribed as a non-invasive alternative to cardiac output (CO) estimation by thermodilution, the current bedside "gold standard". This study was designed to evaluate the accuracy of CO estimations performed by esophageal doppler (EDCO), compared to those obtained using a continuous CO pulmonary flotation catheter (TDCO).

In 16 patients undergoing off-pump coronary artery bypass surgery, CO was measured simultaneously by the esophageal doppler and the thermodilution method, after induction (A), after sternotomy (B), after coronary revascularization (C), and after sternal closure (D). Agreement between the TDCO and EDCO estimations was assessed by analyzing their mean differences and the distribution of these differences. Relative CO changes (percentages of the previous value) was analyzed by the same method.

Both absolute CO values and relative CO changes by esophageal doppler showed a considerable scatter compared to those obtained using the thermodilution method. The bias (EDCO-TDCO) between the two mehtods was -0.8 ± 2.7 L/min for A, -0.9 ± 2.5 L/min for B, -0.9 ± 3.6 L/min for C, and -0.6 ± 2.7 (mean ± 2 SD) L/min for D. On analyzing changes in CO, no significant method bias was found but 2 SD of the bias were ± 74% for A to B, ± 100% for B to C, and ± 83% for C to D.

These results suggest that CO estimations by esophageal doppler cannot replace estimations by the thermodilution method in patients undergoing off-pump coronary artery bypass graft surgery.