Lack of agreement and trending ability of the endotracheal cardiac output monitor compared with thermodilution

    loading  Checking for direct PDF access through Ovid



Minimally invasive monitoring systems of central haemodynamics are gaining increasing popularity. The present study investigated the precision of the endotracheal cardiac output monitor (ECOM) system and its agreement with pulmonary artery catheter thermodilution (PAC TD) for measuring cardiac output (CO) during steady state and with induced haemodynamic changes in patients scheduled for elective cardiac surgery.


Twenty-five patients were enrolled. After induction of anaesthesia, endotracheal intubation using a dedicated ECOM tube, and insertion of the pulmonary artery catheter (PAC), the patient was placed in the following successive positions: (a) supine, (b) head-down tilt, (c) head-up tilt, (d) supine, (e) supine with phenylephrine administration. CO was measured simultaneously using the ECOM and the PAC.


Both methods showed an equally good precision < 10%. Compared to PAC TD, the ECOM system was associated with a bias in supine position of −0.45 l/min (95% confidence interval: −0.86; −0.05), limits of agreement −2.40 l/min to 1.49 l/min and a percentage error of 41.0%. There was no agreement in trending ability between the two methods, with a concordance rate of 30%, shown in a four-quadrant plot.


In a direct comparison with PAC TD, the ECOM system did not show an acceptable agreement, with wide limits of agreement, a much larger percentage error than should be expected from the precision of the two methods and a very poor trending ability. Thus, the ECOM does not replace measurements done by thermodilution using a pulmonary artery catheter in cardiac surgery patients.

Related Topics

    loading  Loading Related Articles