Technical and methodologic considerations for performance of indirect calorimetry in ventilated and nonventilated preterm infants

Objective: To evaluate and refine indirect calorimetry measurement techniques so that accurate metabolic measurements can be performed in mechanically ventilated and convalescing preterm infants who require supplemental oxygen, Design: Laboratory validation of an indirect calorimeter; clinical and laboratory assessments of technical problems in performing metabolic measurements; and clinical indirect calorimetry studies in mechanically ventilated and nonventilated preterm infants, Setting: Neonatal intensive care unit (ICU) in a tertiary care university hospital. Patients: Level II and level III mechanically ventilated (n = 10) and nonventilated (n = 14) neonatal ICU patients who required FIO2 levels ranging from 0.21 to 0.42, Interventions: None, Measurements and Main Results: System calibration was assessed by combustion of 100% ethanol; the mean respiratory quotient was 0.667 +/- 0.001 (SEM). In addition, oxygen consumption (VO2) and CO2 production (VCO2) were simulated by CO2/nitrogen infusions within the range expected for 0.5- to 7-kg infants. Mean relative errors were 0.6 +/- 0.3% and 1,8 +/- 0.3% for expected VO2 and VCO2 values, respectively. In 27 mechanically ventilated patients with no audible endotracheal tube leak, measured endotracheal tube leak ranged from 0.0% to 7.5%. Fluctuations in FIO2 during mechanical ventilation were monitored in 30-min studies, using wall-source (n = 27) or tank-source (n = 11) supplemental oxygen. Mean FIO2 variation was 0.00075 +/- 0.00013 vs. 0.00011 +/- 0.00001 using wall-source and tank-source oxygen, respectively, Some of the difficulties of obtaining accurate measurements in supplemental hood oxygen studies were overcome by using tank-source vs, wall-source oxygen and a unique hood design. Conclusions: Accurate indirect calorimetry studies can be performed in both ventilated and nonventilated infants weighing as little as 500 g, providing that sufficient attention is paid to technical and methodologic measurement details.