Use of noninvasive bioelectric impedance to predict cardiac output in open heart recovery

Abstract

Cardiac outputs (CO) measured by bioelectric impedance ( Z) and thermodilution (TD) were compared in ten stable, non-ventilated male coronary artery bypass patients (mean age 59 ± 12 years) in an open heart recovery unit. The measurements were obtained blindly in three sequential body positions (supine, 45°, final supine) using either a calculated value for resistivity ( p) (based upon hematocrit with blood sampled at the time of the study) to estimate CO( Z), or assumed values of p = 135.5 Ω cm and p = 150 Ω cm. The results indicate high correlations between the two measurement methods (range: r = 0.97 to 0.99) in the initial supine position for all resistivity conditions followed by a progressive decline when body position was changed to 45° and supine (range: r = 0.74 to 0.90). The highest overall correlations and closest absolute mean cardiac output values were obtained when p was calculated from actual hematocrit values obtained at the time of the study. Applying a two-way ANOVA to assess the simultaneous effects of method (TD vx. Z) and position change (supine, 45°, supine), no significant main effects or interactions were found when cardiac output values were estimated using the calculated measurement of p. However, significant main effects of method were found when p was assumed to be either 135.5 Ω cm ( p ⩾ 0.005) or 150.0 Ω cm ( p ⩾ 0.0001), with impedance showing a tendency to overestimate cardiac output. In conclusion, our findings suggest that impedance is a valid method to estimate cardiac output in this subpopulation of patients in open heart recovery provided that p is calculated at the time the study is performed.