Semi-invasive measurement of cardiac output based on pulse contour: a review and analysis

The aim of this review was to provide a meta-analysis of all five of the most popular systems for arterial pulse contour analysis compared with pulmonary artery thermodilution, the established reference method for measuring cardiac output (CO). The five investigated systems are FloTrac/Vigileo(®), PiCCO(®), LiDCO/PulseCO(®), PRAM/MostCare(®), and Modelflow.

In a comprehensive literature search through MEDLINE(®), Web of Knowledge (v.5.11), and Google Scholar, we identified prospective studies and reviews that compared the pulse contour approach with the reference method (n = 316). Data extracted from the 93 selected studies included range and mean cardiac output, bias, percentage error, software versions, and study population. We performed a pooled weighted analysis of their precision in determining CO in various patient groups and clinical settings.

Results of the majority of studies indicate that the five investigated systems show acceptable accuracy during hemodynamically stable conditions. Forty-three studies provided adequate data for a pooled weighted analysis and resulted in a mean (SD) total pooled bias of -0.28 (1.25) L·min(-1), percentage error of 40%, and a correlation coefficient of r = 0.71. In hemodynamically unstable patients (n = 8), we found a higher percentage error (45%) and bias of -0.54 (1.64) L·min(-1).

During hemodynamic instability, CO measurement based on continuous arterial pulse contour analysis shows only limited agreement with intermittent bolus thermodilution. The calibrated systems seem to deliver more accurate measurements than the auto-calibrated or the non-calibrated systems. For reliable use of these semi-invasive systems, especially for critical therapeutic decisions during hemodynamic disorders, both a strategy for hemodynamic optimization and further technological improvements are necessary.