Objective: Bolus thermodilution (BTD) is the most commonly used bedside method of a cardiac output measurement in critically ill patients. A noninvasive alternative method for determining cardiac output and selected derived variables is that of impedance cardiography (ICG), however, conficting results in the literature regarding degree of agreement between BTD and ICG have been published in the literature so far. The aim of the study was to assess the degree of correlation and agreement between cardiac output by impedance cardiography and thermodilution method and to evaluate the relationship between selected variables derived from ICG and those derived from standard hemodynamic measurement. Design: Prospective study. Setting: General intensive care unit at tertiary referral hospital. Material and Methods: A prospective 2 months study in patients requiring ventilatory and hemodynamic support was conducted. All patients requiring pulmonary artery catheter on clinical grounds during a study period were included and simultaneous cardiac output measurements by BTD and ICG (BioZ ICG Monitor, CardioDynamics) were performed at least every four hours. Patients demographic data, cardiac index (CI), central venous pressure (CVP), wedge pressure (PCWP), total chest fluid content (TFC) derived from ICG and paO2/FiO2 index were recorded or calculated. Data are expressed as mean ± SD and were analyzed by SPSS 9.0 and MedCalc 7. The following statistical analyses were performed: Pearson correlation, regression analysis and Bland Altman analysis for bias and precision. Results: A total of 100 paired data points from 12 patients (9 male, 3 female) during a study period were collected.The range of CI (L/min/m2) measurements was 0.5–5.3 by the ICG (mean 2.33 ± 0.8) and 2–6.5 by BTD (mean 3.67 ± 1.08), the correlation coefficient between BTD and ICG was r = 0.268, the bias was -1.3 (95% CI -3.6 to 0.9 L/min/m2) and SD of the differences was 1.16 L/min/m2. The correlation coefficients between TFC and CVP, PCWP and paO2/FiO2 were -0.13; resp. -0.37; resp. -0.53. Conclusion: The ICG showed poor correlation and agreement with the BTD method in mechanically ventilated patients requiring hemodynamic support. Factors that can explain our findings may be the differences between populations usedfor calibration of ICG and the study population, the influence of changing peripheral perfusion due to vasopressor therapy and tissue edema on the bioimpedance signal. There was significant correlation between TFC and paO2/FiO2. The obtained results cannot support the routine use of ICG to replace BTD in this population of patients.

        Key words: intensive care – cardiac output – thermodilution – impedance cardiography – thoracic bioelectrial impedance