Comparison of esCCO and transthoracic echocardiography for non-invasive measurement of cardiac output intensive care.

BACKGROUND The esCCO monitor (ECG- estimated Continuous Cardiac Output, Nihon Kohden(®)) is a new non-invasive tool for estimating cardiac output (CO). It derives CO from the pulse wave transit time (PWTT) estimated by the ECG and the plethysmographic wave. An initial calibration is needed to refine the relation linking pulse pressure (measured by arterial pressure cuff) to PWTT. To assess the accuracy and reliability of the esCCO system, we performed an analysis of agreement of CO values obtained by transthoracic echocardiography (TTE). METHODS Thirty-eight intensive care unit patients were prospectively included. CO was determined simultaneously using esCCO (CO(esCCO)) and TTE (CO(TTE)) as our reference method. RESULTS A total of 103 paired readings from 38 patients were collected. The correlation coefficient between CO(esCCO) and CO(TTE) was 0.61 (P<0.001). The Bland and Altman analysis corrected for repeated measures showed a bias of -1.6 litre min(-1) and limits of agreement from -4.7 to +1.5 litre min(-1), with a percentage error (2 sd/mean CO) of 49%. The correlation for CO changes was significant (R=0.63, P<0.001), but the concordance rate was poor (73%). Polar plot analysis showed an angular bias of -9° with radial limits of agreement from -54° to +36°. The bias appeared to correlate with systemic vascular resistance (R=-0.45, P<0.001). CONCLUSIONS In critically ill patients, the performance of the esCCO monitor was not clinically acceptable, and this monitor cannot be recommended in this setting. Moreover, the esCCO failed to trend CO data reliably.