Comparison of cardiac output measurements by thermodilution and thoracic electrical bioimpedance in critically III versus non-critically III patients

Thoracic electrical bioimpedance (TEB) has been proposed as an alternative to thermodilution (TD) for the measurement of cardiac output in settings such as the Emergency Department where invasive monitoring is not available. Validation studies comparing TEB with TD suggest a wide range of variability in the agreement between the two methods. This prospective study tests the hypothesis that this variability may be related to the severity of patient illness. Fifteen non-critically III patients undergoing cardiac catheterization and 13 critically ill patients who underwent Swan-Ganz catheterization in the medical intensive care unit (MICU) were enrolled. Fifty-one pairs of data from the catheterization laboratory and 49 pairs of data from the MICU were obtained. The patients were graded retrospectively according to the APACHE II scoring system. The mean difference (bias) between TEB and TD results was calculated for each patient using the method suggested by Bland and Altman. A pooled t-test was performed to determine whether there was any significant difference between the APACHE II scores or cardiac output measurements obtained by TEB and TD in the two groups. APACHE II scores were 4.7 ± 1.2 for the catherization laboratory and 14.2 ± 5.0 for the intensive care unit patients (P< .001). The catheterization laboratory (cath lab) group bias was 0.23 ± 2.19, whereas the MICU bias was .002 ± 2.33. There was no significant difference in the bias between the two groups despite significant differences in the APACHE II scores. Standard deviations of the bias were less than 15% different from each other. There is no statistically significant difference in the bias in the two groups, indicating that the systematic error is low. However, the high standard deviation suggests imprecision. The variability of bias results for the comparison of TEB and TD that is reported in the literature may not be related to the severity of the patientʹs illness. The results of our data indicate that the TD/TEB bias is similar across a wide range of patient conditions and severity. This would suggest that TEB would perform equally across the spectrum of patients seen in an Emergency Department.