Measurements of cardiac output obtained with transesophageal echocardiography and pulmonary artery thermodilution are not interchangeable.

Abstract

Echocardiography is increasingly becoming an integrated tool for circulatory evaluation in the intensive care unit and the operating room. Therefore, it is imperative to know the reproducibility of measurements obtained by echocardiography. In this study, a comparison of cardiac output (CO) measurements obtained with transesophageal echocardiography (TEE) and pulmonary artery catheter (PAC) thermodilution (TD) was carried out to test the precision, accuracy and trending ability of CO measurements obtained with TEE. Twenty-five patients completed the study. Each patient was placed in the following successive positions: supine, head-down tilt, head-up tilt, supine, supine with phenylephrine administration, pace heart rate 80 beats per minute (bpm), pace heart rate 110 bpm. TEE CO and PAC CO were measured simultaneously. The agreement was analysed by Bland-Altman plots, and to assess trending ability, a polar plot was constructed. Both methods showed an acceptable precision 8% (PAC TD) and 16% (TEE). In comparison with PAC TD, the TEE was associated with a bias of -0.22 l/minute [95% confidence interval: -0.54; 0.10], wide limits of agreement (-1.73 l/minute; 1.29 l/minute), a percentage error of 38.6% and a trending ability with a radial degree of 53.6°, corresponding to a poor trending ability. In comparison, CO measurements obtained with TEE and PAC TD had wide limits of agreement, a larger percentage error than would be expected from the precision of the two methods, and a poor trending ability. Thus, TEE is not interchangeable with PAC TD for measuring CO.