In critically ill patients, cardiac output (CO) is used as a parameter for assessing hemodynamic status and efficacy of treatment. Continuous CO (CCO) could facilitate this assessment during general anesthesia. A new method of arterial pulse wave analysis has been introduced, which estimates beat to beat CO from arterial pressure via Modelflow. It remains uncertain how well this method performs in high output states. We analyzed the relationship between CCO and Modelflow computed from radial and femoral pressures (MFCO_RA, MFCO_FA) during liver transplantation (LT).

Measurements were performed in 100 liver transplant patients. Groups A had 36 patients, and group C had 64 patients with both groups composed of Child-Turcotte A, B and C patients Eighty patients had CCO < 10 L/min (group D), and 20 patients had CCO > 10 L/min (group E) during anhepatic phase.

CCO ranged from 5.0 to 15.4 L/min (MFCO_RA 3.2 to 10.7 L/min, MFCO_FA 4.3 to 11.8 L/min). Bland-Altman analyses showed the limit of agreement of MFCO_RA (-1.5 to 5.2, bias = 1.9 L/min) and of MFCO_FA (-2.6 to 4.4, bias = 0.9 L/min). CO measured by the two methods was significantly different in groups, except for MFCO_FA in group C. In group D, bias was 1.5 L/min (SD 1.3 L/min) for MFCO_RA and 0.9 L/min for MFCO_FA (SD 1.4 L/min). In group E, biases of 3.5 L/min and 2.4 L/min were obtained for MFCO_RA and MFCO_FA, respectively.

These results suggest that the group-average value of MFCO is not an accurate parameter for estimating CO during LT, with the exception of MFCO_FA in groups C and D.