Background: In this study we compared the accuracy of three continuous cardiac output (CCO) measurement methods, with intermittent transcardiopulmonary thermodilution (TPTD-CO) as the gold standard. The three studied CCO measurement methods were: uncalibrated peripheral pulse contour measurement (FCCO), calibrated central pulse contour measurement (PCCO), and CCO obtained by indirect Fick principle (NCCO).
Methods: We performed an observational study in 23 critically ill patients. Statistical analysis was done using Pearson's correlation and Bland-Altman analysis. A review of the relevant medical literature was performed.
Results: Only PCCO showed good correlation (R = 0.9) and agreement with a bias of 0.0 ± 0.8 L min⁻¹ and percentage error of 24.5% when compared to TPTD-CO. In patients with normal systemic vascular resistance index (SVRI > 1,700 dyne sec cm⁻⁵ m⁻²), NCCO (R = 0.8 and bias 0.4 ± 1.3 L min⁻¹) and FCCO (R = 0.8 and bias 0.1 ± 1 L min⁻¹) also produced reliable results.
Conclusions: These results indicate that in our patient population, CCO can be most reliably monitored by calibrated central pulse contour measurements. All other methods appeared less accurate, especially in situations of low SVRI.