Injectate port selection affects accuracy and reproducibility of cardiac output measurements with multiport thermodilution pulmonary artery catheters

OBJECTIVE: To compare the accuracy and reproducibility of thermodilution cardiac output measurements obtained from the injectate and infusion ports of a multilumen pulmonary artery catheter. The thermodilution results were compared with an independent measure of flow obtained from an electromagnetic flow meter. METHODS: In an experimental study conducted at an animal research laboratory of a health sciences university, two virgin western breed ewes were surgically instrumented with an inferior vena cava occluder, which reduced venous return and thus lowered cardiac output, and an ascending aortic electromagnetic flow probe, which provided an independent reference measure of cardiac output. On the day of study, a multilumen pulmonary artery catheter was inserted. Cardiac output was manipulated over a range of 2.9 to 12.1 L/min with i.v. isoproterenol or inferior vena cava occlusion. Approximately 30 simultaneous thermodilution and electromagnetic flow meter measurements of cardiac output were obtained from both the infusion and injectate ports in each of the two subjects. RESULTS: Correlation coefficients were lower and standard error of the estimates was higher for the infusion port (r = .83; SEE = 1.19 L/min) vs the injectate port (r = .94; SEE = .74 L/min), indicating reduced reproducibility with infusion port thermodilution cardiac output determinations. Accuracy was also adversely influenced in the infusion port results. The linear regression for the infusion port data intercepted the Y axis at +2.64 L/min, indicating significant overestimation of cardiac output at flows of less than 5 L/min and underestimation of flow when cardiac output exceeded 8 L/min. CONCLUSIONS: The use of the infusion port for the measurement of thermodilution cardiac output measurements may result in nonreproducible and inaccurate results.