Reliability of, and Agreement Between, two Breath-by-Breath Indirect Calorimeters at Varying Levels of Inspiratory Oxygen

• Published in: Nutrition in clinical practice Vol. 34; no. 5; p. 767
• Main Authors: Poulsen, Mathias Krogh, Thomsen, Lars Pilegaard, Kjaergaard, Søren, Rees, Stephen Edward, Karbing, Dan Stieper
• Format: Journal Article
• Language: English
• Published: United States 01.10.2019
• Subjects: Adult; Calorimetry, Indirect - instrumentation; Calorimetry, Indirect - standards; Energy Metabolism; Healthy Volunteers; Humans; Inhalation; Male; Oxygen Consumption; Reproducibility of Results; Respiration, Artificial - methods; Respiration, Artificial - standards; reproducibility of results; resting energy expenditure; mechanical ventilation; bias; indirect calorimetry
• ISSN: 1941-2452
• DOI: 10.1002/ncp.10250

Indirect calorimetry (IC) is considered the accurate way of measuring energy expenditure (EE). IC devices often apply the Haldane transformation, introducing errors at inspiratory oxygen fraction (FiO ) >60%. The aim was to assess measurement reliability and agreement between an unevaluated IC (device 2) (Beacon Caresystem, Mermaid Care A/S, Noerresundby, Denmark) not using Haldane transformation and an IC that does (device 1) (Ecovx, GE, Helsinki, Finland) at varying FiO . Twenty healthy male subjects participated, with 16 completing the study (33 ± 9 years, 83.3 ± 16 kg, 1.83 ± 0.08 m). Subjects were mechanically ventilated in pressure support (3cmH O; positive end-expiratory pressure: 3cmH O) at FiO of 21%, 50%, 85%, and 21% for 15 minutes at each FiO . Mean EE, oxygen consumption (VO ), and CO production (VCO ) were compared within and between devices across FiO levels. Device 2 showed within-device EE significant differences at 21% vs 50% FiO and device 1 for VCO at 50% vs. 85% FiO . For all variables, both devices showed reliable measurements at 21% and 50% FiO , but at 85%, FiO bias and limits of agreement increased. Between devices, there were significant differences for EE at both 21% and 85% FiO for VO and for VCO at 85% FiO . Both systems measured EE, VO , and VCO at 21%-85% FiO reliably but with bias at 85% FiO . The devices were in agreement at 21% and 50% FiO , but further studies need to confirm accuracy at high FiO .