Real-time gas-exchange measurement of oxygen consumption in neonates and infants after cardiac surgery

Abstract

Objectives: The purposes of this study were: a) to measure oxygen consumption (VO2) in ventilated neonates and infants after cardiac surgery, utilizing a real-time gas exchange system; and b) to assess this new method by comparing the measured VO2 with calculated VO2 (using the Fick equation and simultaneously determined thermodilution cardiac output, measured hemoglobin, and measured mixed venous and arterial saturations).

Design: Prospective, comparison study. Comparison of measured VO2 and calculated VO2 using correlation coefficient, linear regression analysis, and bias and precision.

Setting: Cardiac intensive care unit in a children's hospital.

Patients: A total of 60 direct comparisons were made between measured and calculated VO2 in 15 patients (ages ranging from 4 days to 14.1 months with median age of 2.4 months) who were receiving mechanical ventilation after undergoing corrective cardiac surgery.

Interventions: a) Direct measurement of VO2 using gas exchange method involving a pneumotachograph and a gas sampling system; b) determination of cardiac output by the thermodilution technique; c) measurement of arterial and mixed venous oxygen content by blood sampling.

Measurements and results: The absolute measured VO2 ranged from 19 to 154 mL/min with a mean of 52 +/- 32 mL/min (when indexed, the range was 81 to 367 mL/min/m2 with mean 185 +/- 69 mL/min/m2, or range 4.7 to 18.8 mL/min/kg with mean 10.4 +/- 3.3 mL/min/kg). While 34 (57%) of 60 measured VO2 values were within 10% of their respective calculated VO2 values, 58 (97%) of 60 were within 25%; the mean percent difference between measured and calculated VO2 values was 10.6 +/- 7.7%. In comparing the measured VO2 and calculated VO2, the correlation coefficient was good (r2 = .87; p < .01) and the linear regression equation was: measured VO2 = 1.1 x calculated VO2 -9.0 mL/min/m2. The mean difference, or bias, was 0 mL/min/m2 with precision of 26 and 52 mL/min/m2 (at 1 and 2 SD). As an alternative means of examining the measured VO2 data, we also directly compared the thermodilution cardiac output with cardiac output derived by using the measured VO2 and the Fick equation. The range of Fick-derived cardiac output was between 1.69 to 8.11 L/min/m2 (mean 3.72 +/- 1.56) and the range of thermodilution cardiac output was between 1.75 to 7.42 L/min/m2 (mean 3.71 +/- 1.36). The correlation coefficient between thermodilution cardiac output and Fick-derived cardiac output was good with r2 = .88 (p < .01) and the linear regression equation was: thermodilution cardiac output = 0.81 x Fick-derived cardiac output + 0.71 L/min/m2. The bias was -0.01 L/min/m2 with a precision of 0.54 L/min/m2 at 1 SD (or 1.08 L/min/m2 for 2 SD).

Conclusions: Measured VO2 using a gas-exchange system compared favorably with calculated VO2 values using the Fick equation and simultaneously obtained thermodilution cardiac output and arterial and venous oxygen concentrations. By employing this breath-by-breath gas-exchange system, real-time VO2 measurement in ventilated neonates and infants is now feasible.