O-27 Agreement between 4-D echocardiography computed cardiac output and standard PAC technique is poor

Abstract

Introduction. Pulmonary artery catheter (PAC) and 2D echocar-diography are considered standard clinical methods for stroke volume (SV) and cardiac output (CO) computation. The strength of echocardiography is to evaluate structures and function, whereas assessment of CO is less reliable. However, 4D ultrasound systems have now been optimized, giving hope for faster and more consistent measurements. Method. One hundred and one measurements from 30 patients were evaluated. Measurements were done using a GE Vivid E9 system fitted with a 4D probe. Subsequently, single heart beat analyses were performed off line using the auto LVQ semiautomatic function in Echopac software allowing for manual alignment of the end-diastolic/end-systolic endocardial border to calculate SV and CO. The echo measurements were done blindly by two independent sonographers and compared to continuous CO obtained with a thermistor-tipped, flow-directed PAC. Results. The Bland-Altman plot revealed a mean bias of 1.6 L (95% safety limits -0.3 - 3.4). The inter-observer bias was 0.29 L/min (7.75%). Conclusion. As the inter-observer bias is within reasonable limits, the difference of 1.6 L/min (30.6%) between PAC and echo-cardiographic measurements suggests that the methods are not interchangeable.