Objectives: To study the reproducibility of and agreement between perioperative transesophageal echocardiographic (TEE) and radionuclide (RN) assessment of ventricular volumes and ejection fraction (EF).
Design: A prospective, blinded comparison of two methods of measurement.
Setting: A surgical intensive care unit in a university hospital.
Participants: Patients after coronary artery bypass surgery.
Interventions: Left ventricular volumes and ejection fraction were simultaneously measured by radionuclide ventriculography and transesophageal echocardiography. Two sets of measurements were made in rapid succession with a third set after an interval.
Results: Reproducibilities of EF and ventricular volumes by TEE and RN were similar. At each set of measurements, the bias for radionuclide EF and TEE Simpson's rule EF, 0.03 +/- 0.05, 0 +/- 0.06, -0.01 +/- 0.07, respectively, for radionuclide EF and TEE area length EF 0.01 +/- 0.05, -0.01 +/- 0.05, -0.03 +/- 0.08, respectively, were significantly less than for radionuclide EF and TEE FAC 0.07 +/- 0.05, 0.05 +/- 0.05, 0.03 +/- 0.09. Poor agreement was observed between RNTD-EDV, and both of the TEE EDV measurements.
Conclusion: EF measured by TEE area length and Simpson's rule method are as reproducible as TEE FAC and are more accurate estimates of RN EF. Poor agreement between methods of measuring end-diastolic volume was observed.