407 Variation in Cardiac Index by Thermodilution Methods in Heart Transplant Recipients.

Abstract

BackgroundWhile several methods are used to measure thermodilution (TD) cardiac index (CI), most catheterization laboratories use proprietary systems that have not been recently validated. Because cardiac rejection may present with decreased CI, heart transplant recipients routinely undergo hemodynamic assessment, making accurate CI assessment essential in these patients.MethodsAs a quality improvement initiative, we measured CI in 39 stable heart transplant recipients without tricuspid regurgitation by one or more of three different TD methods: with the injectate temperature probe in the water bath through which the injectate was drawn (“open” systems) at room temperature (“RT,” n = 26) or iced (“Iced,” n = 13) or with direct injectate temperature measurement (“Inline,” n = 14). As a control, Fick CI (with assumed O2 consumption) was measured in all patients. Standard nonparametric tests were used for comparison.ResultsFick CI was overestimated by each TD method by median (interquartile range) values of 36% (21-61%) for Open-Iced (p = .0012), 26% (14-39%) for Open-RT (p < .0001), and 11% (0-18%) for Inline-Iced (p = .002). Correlation with Fick-CI was better for Inline-Iced CI (r = .79) and poorer for Open-RT CI (r = .56) and Open-Iced CI (r = .61). Among the TD methods, there were significant differences between median Inline-Iced CI values and those for either the Open-Iced (p = .017) or Open-RT (p = .0005) methods.ConclusionsAs compared to Fick, each TD method significantly overestimated CI in stable patients, often to an extent that may be clinically relevant. This was especially true if the injectate temperature was measured with an “open” system.