Evaluation of an uncalibrated arterial pulse contour cardiac output monitoring system in cirrhotic patients undergoing liver surgery.

BACKGROUND The pulmonary artery catheter is invasive and may cause serious complications. A safe method of cardiac output (CO) measurement is needed. We have assessed the accuracy and reliability of a recently marketed self-calibrating arterial pulse contour CO monitoring system (FloTrac/Vigileo) in end-stage liver failure patients undergoing liver transplant. The pattern of alterations known as cirrhotic cardiomyopathy, and the transplant procedure itself, provided an evaluation under varying clinical conditions. METHODS The cardiac index was measured simultaneously by thermodilution (CI(TD): mean of four readings) using a pulmonary artery catheter and pulse contour analysis (CI(V): mean value computed by the FloTrac/Vigileo over the same time period). Readings were made at 10 time-points during liver transplant surgery (T1-T5) and on the intensive care unit (T6-T10). CI(V) was computed using the latest Vigileo software version 01.10. RESULTS A total of 290 paired readings from 29 patients were collected. Mean (SD) CI(TD) was 5.2 (1.3) and CI(V) was 3.9 (0.9) litre min(-1) m(-2), with a corrected for repeated measures bias between readings of 1.3 (0.2) litre min(-1) m(-2) and 95% limits of agreement of -1.5 (0.2) to 4.1 (0.3) litre min(-1) m(-2). The percentage error (2SD(Bias)/meanCI(TD)) was 54%, which exceeded a 30% limit of acceptance. Low peripheral resistance and increasing bias were related (r=0.69; P<0.001). The Vigileo system failed to reliably trend CI data, with a concordance compared with thermodilution below an acceptable level (at best 68% of sequential readings). CONCLUSIONS In cirrhotic patients with hyperdynamic circulation, the Vigileo system showed a degree of error and unreliability higher than that considered acceptable for clinical purposes.