Estimation of changes in cardiac output from the arterial blood pressure waveform in the upper limb

We have developed a new pulse contour cardiac output (PulseCO) algorithm based on frequency analysis studies of the arterial system. PulseCO was compared with thermodilution cardiac output (TDCO) in 10 patients undergoing cardiac surgery. Results from one patient were unsuitable for analysis. In the remaining nine patients, 142 TDCO determinations were compared with PulseCO after logarithmic transformation and after being normalized by the initial cardiac output in each patient. Each determination was usually the average of three measurements. Least squares regression gave y = 0.77 x (r2 = 0.81) and the limits of agreement were from -26% to +21%. The accuracy of PulseCO in determining short-term changes in cardiac output was assessed by comparing the ratios of consecutive PulseCO determinations with the ratios of the corresponding, consecutive TDCO determinations. Least squares regression gave y = 0.71 x (r2 = 0.70) and the limits of agreement were from -21% to +25%. After phenylephrine had been given to five patients, PulseCO showed an increase in systemic vascular resistance consistent with the known pharmacological actions of the drug. The PulseCO algorithm was incorporated into a computer program that acquires arterial pressure data from an analogue-to-digital converter and displays beat-to-beat trend values.