Assessment of the accuracy of continuous cardiac output and pulse contour cardiac output in tracking cardiac index changes induced by volume load

Objective: To assess the ability to track changes in cardiac index (ΔCI) induced by volume loading using continuous pulsed heat thermodilution (CCO), and pulse contour (PCCO) cardiac output (CO) with transpulmonary thermodilution (TD_tp) CO as reference. Design: Prospective observational clinical trial. Setting: Intensive care unit. Patients: Twelve ventilated and sedated post-operative cardiac surgery patients. Measurements and results: Each patient had a 7.5F CCO pulmonary artery catheter (Edwards Lifesciences) and a 5F, 20 cm PCCO femoral artery catheter (Pulsion Medical Systems). Forty-five data sets were taken before and after 25 volume loadings of 5 mL/kg of 4% albumin. Volume loading resulted in an increase in CI (2.84 L/(min m²) versus 3.12 L/(min m²), p < .05) although only nine volume loadings changed CI (ΔCI) ≥ 14%. The change in CI using PCCO (ΔPCCI) was correlated with ΔCI (TD_tp) (R² = .50, p < .0001), whilst ΔCI using CCO (ΔCCI) was not (R² = .14). The bias and limits of agreement (LOA) between ΔTD_tpCI and ΔPCCI was 6.2% (95% CI, ±5.8%) and 28.4% (95% CI, ±38.2%) respectively. ΔTD_tpCI and ΔCCI has a bias of 2.6% (95% CI, ±8.3%) and LOA of 39.6% (95% CI, ±63%). Both ΔPCCI and ΔCCI reliably tracked ΔCI ≥ 14%. Conclusion: In this small group of patients the continuous cardiac output methods tracked changes in CI, although, in individual cases they did not change in the same direction as the thermodilution method. Critical care nurses need to critically appraise the accuracy and clinical relevance of continuous CO data within the clinical context.