Assessment of three methods to reduce the influence of mathematical coupling on oxygen consumption and delivery relationships

Abstract

Study objectives: To determine if oxygen consumption/oxygen delivery (VO2/DO2) relationships derived using calorimetry (which are not influenced by shared measurement error) agreed with those obtained using the pulmonary artery (PA) catheter alone. To evaluate three strategies to reduce the influence of shared measurement error to determine if agreement between the two methods could be improved.

Methods: Twenty-seven patients were studied following coronary artery bypass surgery. Calorimetric VO2, six thermodilution cardiac outputs (COs), and arterial and mixed venous oxygen content measurements were made at baseline and were repeated 30 min following dobutamine administrations of 3 microg/kg/min and 5 microg/kg/min.

Results: Dobutamine produced a dose-dependent increase in DO2, from 378+/-65 mL/min/m2 to 446+/-78 mL/min/m2 (p<0.01) and in both PA catheter and calorimetric-derived VO2, from 104+/-18 mL/min/Mi2 to 114+/-22 mL/min/m2 (p<0.05) and from 117+/-15 mL/min/m2 to 126+/-19 mL/min/m2 (p<0.01), respectively. Agreement was poor (bias=12%, SD=21%) between the calorimetric and PA catheter methods of determining VO2/DO2 slope. When three CO measurements were used to calculate VO2, and three separate CO measurements were used to calculate DO2, the level of agreement between the two methods improved (bias=2%, SD=15%). Increasing the number of COs resulted in a similar improvement in the level of agreement between the two methods. Weighting the slope to the observed change in DO2 was the best method to improve the level of agreement (bias=2%, SD=6% for three COs).

Conclusions: To reduce the influence of shared measurement error, the best strategy to improve the measurement of VO2/DO2 slope is to maximize the change in DO2 (optimally over 100 mL/min/m2).