Two- or 3-Dimensional Echocardiography–Derived Cardiac Output Cannot Replace the Pulmonary Artery Catheter in Cardiac Surgery

Abstract

Three-dimensional (3D) transesophageal echocardiography (TEE) has been shown to be more accurate than 2D TEE for the evaluation of the left ventricular outflow tract area. The aim of the present study was to compare the agreement of 3D echocardiography-derived cardiac output (CO) with thermodilution-derived CO (TDCO) before and after cardiopulmonary bypass (CPB). This was a prospective observational study of patients who underwent cardiac surgery between 2016 and 2018. Weill Cornell Medicine, a single large academic medical center. The study comprised 78 patients undergoing elective cardiac surgery. CPB, TEE, pulmonary artery catheter, and elective cardiac surgery. Two-dimensional CO, 3D CO-diameter, and 3D CO-area values pre-CPB were strongly correlated with one another both pre-CPB and post-CPB. The 3D CO-diameter and the 3D CO-area were mildly correlated, with TDCO measurements pre-CPB (r = 0.46 and 0.39, respectively) and post-CBP (r = 0.43 and 0.47, respectively). Pre-CPB 3D CO-diameter had the most agreement with TDCO in terms of bias (-0.13 L/min); however, the limits of agreement (LOA) were wide (-2.2- to- 2.45 L/min). Post-CPB, 3D CO-diameter had the most agreement with TDCO in terms of bias (0.41) but with wide LOA (-3.29 to 2.47). All pre-CPB echocardiography-derived CO (2D CO, 3D CO-diameter, 3D CO-area) had more agreement with TDCO than did post-CPB measurements. Three-dimensional CO measurements were only modestly correlated with pulmonary artery catheter-derived CO pre-bypass and post-bypass. Despite low bias, the wide LOA from 2D CO, 3D CO-diameter, and 3D-area compared with TDCO suggested that the 2 methods are not interchangeable.