Lack of reliability of cardiac output measurements derived from arterial pressure waveform analysis in on-pump cardiac surgery patients

Abstract

The arterial pressure waveform-based device for cardiac output (CO) measurement FloTrac/Vigileo™ offers the possibility of minimal-invasive CO monitoring without the need for invasive calibration. The agreement of the results of this device with bolus thermodilution remains controversialas is the choice of the best arterial cannulation site. After approval of the local ethics committee 14 Patients scheduled for elective cardiac surgery with cardiopulmonary bypass were enrolled. Routine arterial access inone radial and one femoral artery were used for CO measurement with the FloTrac/Vigileo device. Pulmonary artery catheter was inserted for determination of CO with thermodilution bolus technique. All CO was indexed to body surface area resulting in CI-FEM, CI-RAD and CI-PAC. Measurements were performed at least four times before and three times after CPB. A total of 262 data pairs were analysed. Including all data before and after CPB bias and precision for CI-PAC vs. CI-FEM were 0,09 L min-1m-2 and ±1,01 Lmin -1m-2, with a percentage error of 40%. For CI-PAC vs. CI-RADbias and precision were 0,34 Lmin-1m-2 and ±1,08 Lmin-1m-2 with a percentage error of 45%. CI-FEMvs. CI-RAD revealed a bias of -0,25 Lmin-1m-2 and a precision of ±0,48 Lmin-1m-2 resulting in a percentage error of 20%. Subgroup analysis of data pairs obtained before CPB showed a percentage error of 32% and 34% for CI-PAC vs. CI-FEM and CI-PAC vs. CI-RAD respectively. After CPB percentage error was 46% for CI-PAC vs. CI-FEM and 55%for CI-PAC vs. CI-RAD. We conclude that the agreement between CO measured with the FloTrac/Vigileo™ device and the thermodilution bolus method is not sufficient for clinical diagnosis and therapy guidance in the setting of cardiac anesthesia with CPB. © 2009 Springer-Verlag.