Accuracy of an indirect calorimeter for mechanically ventilated infants and children: The influence of low rates of gas exchange and varying FIO2

Objective: To test the accuracy and validity of the Deltatrac II MBM-200 me tabolic monitor for use in mechanically ventilated infants and children in the pediatric intensive care unit. Design: Laboratory validation of an indirect calorimeter with a ventilated lung model. The influence of low tidal volumes and low levels of oxygen con sumption ((V)over dot O-2) and carbon dioxide production ((V)over dot CO2) in combination with different levels of inspired oxygen concentrations (FIO 2) was investigated. Setting: University research laboratory. Subjects: Low tidal volumes were provided with two intermittent flow types of ventilators, a Servo 300 and a Servo 900C. Interventions: A butane flame with a (V)over dot O-2, approximating 20 mL/m in and 40 mL/min was ventilated. To investigate the effect of different lev els of FIO2 on the accuracy of (V)over dot O-2, (V)over dot CO2, and respir atory quotient (RQ), measurements were performed at FIO2 target values of 0 .25, 0.40, and 0.60. Measurements and Main Results: No significant differences were found betwee n the ventilators regarding (V)over dot O-2, (V)over dot CO2, and RQ measur ements. The mean deviation of (V)over dot O-2 increased significantly with increasing FIO2 to -7.98% with a (V)over dot O-2 of 21.0 mL/min and to -8.4 6% with a (V)over dot O-2 of 38.9 mL/min (FIO2, 0.558) with a variability ( 2 SD) of +/- 4.86% and +/- 6.82%, respectively. The mean deviation and vari ability of (V)over dot CO2 in all tests remained within 8%. The mean deviat ion of RQ increased significantly with increasing FIO2 to 5.5% with a (V)ov er dot O-2 of 21.0 mL/min and to 5.69% with (V)over dot O-2 of 38.9 mL/min (FIO2, 0.558) with a variability (2 SD) of +/- 5.62% and +/- 5.76%, respect ively. The minute to minute delivered FIO2 fluctuated significantly when in creasing the level of FIO2. Conclusions: The Deltatrac II MBM-200 metabolic monitor appears accurate fo r low levels of (V)over dot O-2 and (V)over dot CO2 during mechanical venti lation with FIO2 levels up to 0.390, With increasing FIO2 to 0.558, the inc rease in deviation of (V)over dot O-2 for single measurements can be of cli nical relevance for mechanically ventilated infants and children. The incre ased fluctuation of delivered FIO2 on higher levels of FIO2 is likely the c ause of the inaccuracy.