Transesophageal Echocardiography Cardiac Output in FlowTrak/ Vigileo Study

Abstract

To the Editor: In the article by Concha et al.1 comparing cardiac output (CO) measurements of the FlowTrak/Vigileo monitor with those obtained by transesophageal echocardiography (TEE), the authors state that they used the transesophageal CO method previously validated by Perrino et al.2 However, an important difference between the validated technique and that used in their study has implications with regard to the interpretation of their findings. Stroke volume by the Doppler method can be understood as the product of blood transit (distance traveled, which is derived from velocity measurements) and the cross-sectional area of the flow stream during systole. Concha et al. chose continuous wave (CW) Doppler measures of blood flow velocity in conjunction with a 2-dimensional measurement of left ventricular outflow tract (LVOT) diameter to calculate the cross-sectional area. This is not the method described by Perrino et al.2 in the cited article, nor is it as Perrino and Maslow describe in the text, “A Practical Approach to Transesophageal Echocardiography.”3 The method validated in the 1998 article does use CW Doppler, but matched to the triangular-shaped aortic valve area, not LVOT area, to best estimate the average cross-sectional area of the flow stream. The concern with using LVOT area is the overestimation of the cross-sectional area of the flow stream across the aortic valve that was assessed by the CW Doppler. The overestimation of cross-sectional area leads to a proportional overestimate of CO. Darmon et al.4 showed that a circular measurement of the aortic valve area overestimates the average cross-sectional area of the flow stream by 36% (3.28 vs 2.41 cm2). Because LVOT area equals or exceeds the area of the aortic valve annulus, Concha et al. substantially overestimate CO by TEE when using LVOT area in conjunction with CW Doppler velocities from the aortic valve. For this reason, a pulsed wave Doppler measurement with the sample volume located within the LVOT is preferred for CO calculations based on LVOT area. The overestimation of CO measurements by the TEE method used by Concha et al. affects their findings because they observed a consistent positive bias; TEE measurements exceeded those derived by FlowTrak/Vigileo (mean TEE CO 6.21 L/min; mean FlowTrak CO 4.84 L/min; mean difference 28%) (Figs. 1 and 2, Table 3). We believe the mean bias between TEE and FlowTrak/Vigileo CO measurements to be substantially less than reported (and perhaps a negative value) because of the TEE CO method used in this study. Interpretation of this study should recognize that the chosen TEE method affects the reported mean bias between techniques but does not alter the SD of the differences between techniques.5 In conclusion, the method used to measure TEE CO described in the study by Concha et al. is not recommended as a reference technique because it will overestimate CO. Interpretation of their results and the performance of the FlowTrak/Vigileo should take into consideration overestimations of TEE CO. William J. Perez, MD, MA Department of Anesthesiology The Ohio State University Medical Center Columbus, Ohio [email protected] Albert C. Perrino, Jr., MD Yale University School of Medicine VA Connecticut Healthcare System New Haven, Connecticut