The method of determining continuous cardiac output (CO) with beat-to-beat pulse-contour analysis calibrated by transthoracic thermodilution is gaining much wider clinical acceptance. However, some questions have been raised regarding the reliability of this method during periods of profound hemodynamic instability. We validated the original calculation of pulse-contour analysis and a new, improved algorithm against thermodilution-derived measurements of CO in patients with changes of CO >20%.Design
Cardiac surgical intensive care unit of a university hospital.Patients
Twenty-four patients after cardiac surgery who experienced changes of CO >±20% during their postoperative course.Interventions
CO was measured by transthoracic thermodilution and pulse-contour analysis (PiCCO, PULSION Medical Systems, Munich, Germany) at serial intervals every 60 mins during study periods of 8–44 hrs. During this time, no recalibration of the pulse-contour computer was performed.Measurements and Main Results
A total of 517 simultaneous measurements of thermodilution CO and pulse-contour CO measured by the two different algorithms were compared by regression, structural regression, and Bland-Altman analyses. Mean change of CO was 40% ± 27% (range, 20% to 139%), range of systemic vascular resistance was 450–2360 dyne·sec/cm5. Correlation of the original pulse-contour algorithm to thermodilution CO was r = .76, with p = .027; bias was 0.08 L/min, with 1.8 L/min single sd. Correlation of the improved pulse-contour algorithm to thermodilution CO was r = .88, with p = .0001; bias was 0.2 L/min, with 1.2 L/min single sd. Mean CO by the new pulse-contour algorithm did not differ significantly from CO by thermodilution during the study period. The difference between the methods was not influenced by variations of heart rate or arterial pressure.Conclusions
CO measurement by arterial pulse-contour analysis based on a new, improved algorithm is reliable, even in patients with profound changes of CO and during periods of hemodynamic instability.