Disagreement between pulse contour analysis and transpulmonary thermodilution for cardiac output monitoring after routine therapeutic interventions in ICU patients with acute circulatory failure

Background and objective The present prospective study was aimed at assessing the reliability of the pulse contour method for measuring cardiac output (CO) after different routinely used therapeutic interventions that can influence vascular compliance and systemic vascular resistances in ICU patients (fluid challenges, changes in norepinephrine or dobutamine infusion rates and changes in ventilatory settings). Methods In ICU patients requiring CO monitoring, transpulmonary thermodilution CO (CO(TD)) and pulse contour CO (CO(PC)) were measured with a PiCCO device after therapeutic manoeuvre-free periods (<= and > 1 h) and after therapeutic interventions without recalibration. Results Three hundred fifty-two sets of CO measurement pairs in 63 ICU patients were performed. The biases (and percentage errors) between CO(PC) and CO(TD) for the overall paired measurement, therapeutic manoeuvre-free periods and therapeutic manoeuvres were 0.20 +/- 1.09 (33%), -0.01 +/- 0.93 (29%) and 0.37 +/- 1.18 (34%), respectively. The percentage errors were 36 and 39% for fluid challenges and changes in norepinephrine infusion rate, respectively. Conclusion In ICU patients requiring therapeutic interventions, CO(PC) is frequently in disagreement with CO(TD). Eur J Anaesthesiol 2011; 28: 664-669 Published online 12 May 2011