ABSTRACT
When calculating cardiac index (C.I.) by the Fick method, oxygen consumption (VO2) is often unknown, so assumed values are typically used. This practice introduces a known source of inaccuracy into the calculation. Using a measured VO2 (mVO2) from the CARESCAPE E-sCAiOVX module provides an alternative that may improve accuracy of C.I. calculations. Our aim is to validate this measurement in a general pediatric catheterization population and compare its accuracy with assumed VO2 (aVO2). mVO2 was recorded for all patients undergoing cardiac catheterization with general anesthesia and controlled ventilation during the study period. mVO2 was compared to the reference VO2 (refVO2) determined by the reverse Fick method using cardiac MRI (cMRI) or thermodilution (TD) as a reference standard for measurement of C.I. when available. 193 VO2 measurements were obtained, including 71 with a corresponding cMRI or TD measure of cardiac index for validation. mVO2 demonstrated satisfactory concordance and correlation with the TD- or cMRI-derived refVO2 (ρc = 0.73, r2 = 0.63) with a mean bias of - 3.2% (SD ± 17.3%). Assumed VO2 demonstrated much weaker concordance and correlation with refVO2 (ρc = 0.28, r2 = 0.31) with a mean bias of + 27.5% (SD ± 30.0%). Subgroup analysis of patients < 36 months of age demonstrated that error in mVO2 was not significantly different from that observed in older patients. Many previously reported prediction models for assuming VO2 performed poorly in this younger age range. Measured oxygen consumption using the E-sCAiOVX module is significantly more accurate than assumed VO2 when compared to TD- or cMRI-derived VO2 in a pediatric catheterization lab.
