Comparison of pulse contour analysis by Pulsioflex and Vigileo to measure and track changes of cardiac output in critically ill patients

Abstract

Background We compared the new Pulsioflex and the Vigileo devices to measure cardiac index (CI) in critically ill patients. Both devices measure CI by pulse-contour analysis. The Pulsioflex device also allows an auto-calibration (not based on thermodilution). Methods Patients were included if we administered fluids (20 patients), reduced (20 patients), or increased (20 patients) the dose of norepinephrine. Before and after interventions, we measured CI provided by the Vigileo (CIVig) and Pulsioflex (CIPfx) devices before and after its auto-calibration. CI measured by transpulmonary thermodilution (CIthermo) was used as the reference. Results Considering absolute values of CI (n=120), the percentage error was 59% for CIVig vs CIthermo and 40% for CIthermo vs CIPfx. Auto-calibrating CIPfx after interventions did not improve the percentage error between CIPfx and CIthermo (39%). Considering the fluid-induced changes in CI, the coefficient of correlation with changes in CIthermo was 0.50 for CIVig, and 0.73 for CIPfx (P=0.27). It was not significantly improved if CIPfx was auto-calibrated (r=0.64). Considering the norepinephrine-induced changes in CI, the coefficient of correlation with changes in CIthermo was 0.41 for CIVig. It tended to be better for CIPfx (r=0.71, P=0.07). It was not significantly improved by auto-calibration (r=0.53). Conclusions The Pulsioflex did not reliably estimate the absolute values of CI. For tracking fluid-induced changes in CI, the Pulsioflex was reliable, and also the Vigileo. For tracking norepinephrine-induced changes in CI, it was also reliable and tended to be better than the Vigileo. Auto-calibration allowed by the system did not improve its reliability.