A practical method of measuring oxygen consumption in children with complex mixing circulations by the use of thermodilution cardiac output studies

ObjectiveWe describe a method to measure oxygen consumption (VO2) and pulmonary vascular resistive index (PVRI) based on thermodilution cardiac output in patients with complex mixing circulations. We apply this method to patients with a bidirectional cavopulmonary anastomosis (BDCPA). We compare our measured VO2 with a predicted VO2 based on a formula using height and weight.MethodsWe reviewed data based on cardiac catheterization and thermodilution cardiac output in a series of 25 catheterizations in patients with BDCPA. We used this data to measure VO2 and PVRI, and looked for a correlation between the predicted and measured VO2. We also assessed whether any other hemodynamic parameter correlated with measured VO2.ResultsThere was no significant correlation between the predicted and the measured VO2 (correlation coefficient = −0.258, P = .21). We did find a significant correlation in the difference between the measured and predicted VO2 against the measured VO2 such that at a lower measured VO2 the predictive formula tended to overestimate VO2 and at a higher VO2 the formula underestimated VO2 (correlation coefficient = 0.963, P < .0001). Body surface area did not correlate with measured VO2 (correlation coefficient = 0.28, P < .16). Mixed venous oxygen content showed a weak negative correlation with VO2 (correlation coefficient = −0.54, P = .005).ConclusionsAn assessment of PVRI that is based on a predicted VO2 is unreliable in this patient population. No hemodynamic parameter correlated well with VO2. The use of a measured VO2 is necessary in determining PVRI in these patients.