Reproducibility and comparison of cardiac output measurement by transthoracic bioimpedance and thermodilution methods in critically ill patients.

The short-term reproducibility in cardiac output (CO) and stroke volume (SV) measurements by transthoracic electrical bioimpedance (TEB) and thermodilution (TH) and their agreement were studied in 31 consecutive patients in the Intensive Care Unit (ICU). For comparison of changes in CO and SV, six patients were studied separately. TEB data were not obtainable in four patients due to interference with impedance signals or heart rate detection. Both methods were reproducible. For TEB, the mean difference of duplicate measurements was 0.03 (SD 0.3) l/min for CO and 0.3 (SD 3.2) ml for SV. For TH, it was -0.04 (SD 0.5) l/min for CO and 0.6 (SD 6.6) ml for SV. The coefficient of reproducibility for CO was thus 0.6 l/min for TEB and 1.0 l/min for TH and that for SV was 6.4 ml for TEB and 13.2 ml for TH. There was no agreement in absolute CO or SV between the methods; TEB giving lower absolute readings. The mean difference (TH-TEB) was 1.4 (SD 1.4) l/min for CO and 14 (SD 13.4) ml for SV. There was, however, good correlation between the two methods. For CO, the regression equation was TEB=0.623+0.739 (TH); (p less than 0.0005, r=0.87). For SV, the regression equation was TEB=-0.23+0.823 (TH); (p less than 0.0005, r=0.86). When only changes were considered, the mean difference (TH-TEB) was -0.5 (SD 0.9) l/min for CO and -4 (SD 8) ml for SV. There was significant correlation in the changes detected by each method.(ABSTRACT TRUNCATED AT 250 WORDS)