Validation of continuous thermodilution cardiac output in critically ill patients with analysis of systematic errors

Purpose: Bolus thermodilution cardiac output (BCO) measurements are affecte d by variations in injectate volume, rate, and temperature. These variation s are eliminated when CO is measured by a continuous automated thermal tech nique (CCO). Further, CCO eliminates the need for fluid boluses, reduces co ntamination risk, requires no operator, and provides a continuous CO trend. We prospectively evaluated CCO versus BCO in a population of critically il l adults with low, normal, and high CO states. We sought to discern any sys tematic effects of temperature fluctuations or signal-to-noise-ratios (SMR) on disparities between BCO and CCO measurements and also sought to assess the relative cost effectiveness of the CCO system. Materials and Methods: Pulmonary artery catheterizations were performed in a convenience sample of 20 patients over 6 months. BCO data were obtained u sing a standardized protocol. Three bolus injections of 5% dextrose were gi ven when each CO was within 10% of the median before averaging; otherwise f ive boluses were given, with the high and low values eliminated before aver aging. Injectates were administered randomly through the respiratory cycle and at 1-minute intervals. CCO measurements were recorded from a Vigilance monitor pre and post BCO measurements, yielding an average CCO value, Also recorded were pre- and post-core temperatures and SNR during the first CCO measurement. Cost data included estimates of operator time for BCO determin ations as well as costs of Intellicath (Baxter-Edwards, Irvine, CA) pulmona ry artery catheters, Vigilance (Baxter-Edwards, Irvine, CA) monitors, conve ntional catheters,and injectates. Results: Of the 20 patients, 15 were mechanically ventilated. A total of 30 6 paired CO values were obtained for analysis. CCO ranged from 2.5 to 14.4 L/min and BCO from 2.4 to 13.3 L/min. Absolute differences between CCO and BCO measurements increased with increasing CO, but percentage differences d id not. Of the paired values, 77% were within 1 L/min of one another. Tempe rature instability and SNR independently had weak correlations with CCO/ BC O disparities. The Vigilance system had a slightly higher net cost than con ventional BCO, although no economical value was assigned to the clinical us efulness of continuous, as opposed to intermittent, CO monitoring. Conclusions: Continuous CO is a reliable and cost-effective alternative to bolus thermodilution CO for critically ill patients in low, normal, and hig h CO states. This is a US government work. There are no restrictions on ifs use.