(Re)assessment of the COSMED Quark CPET and VO2Master Pro Systems for Measuring Pulmonary GasExchange and Ventilation.

Abstract

Listen

Translate article

We assessed the validity, reliability, and transferability of gas exchange and ventilatory variables from two commonly used metabolic measurement systems (COSMED Quark and VO2Master Pro). Two identical devices from each system were independently connected to a metabolic simulator (VacuMed), and 2 minof steady-state data was recorded at simulated oxygen uptake (V̇O2) of 1, 2, 3, and 4 L∙min-1 achieved through minute ventilation (V̇E) of 30, 60, 105, and 150 L∙min-1. Each metabolic analyzer recorded data three times for each "intensity" in a randomized order, and assessments were completed on two separate days. Douglas bag-based measurements were also made once at each simulated "intensity". Measured steady-state data (average of final 1.5 min) for both V̇O2 (STPD) and V̇E (ATPS) were compared with simulated values to assess validity, repeated values between and within days assessed reliability, and between-device comparisons provided transferability. Including both COSMED devices at all intensities, the mean percent error for V̇O2 was 3.5% (range: -2.5%-8.1%) and, for V̇E, was 2.0% (-0.5%-7.6%). For the VO2Master, these values averaged 0.6% (-9.3%-4.8%) and 1.1% (-6.3%-4.0%) for V̇O2 and V̇E, respectively. Mean percent error for Douglas Bag was 1.5%, -3.7%, -3.1%, and -2.0% for 1, 2, 3, and 4 L∙min-1, respectively. Between-day differences (reliability) for V̇O2 of both COSMED devices ranged from -4.1% to 2.2% (mean 0.1%) and, for both VO2Masters, between -1.6% and 11.1% (mean 1.2%). Between-device differences (transferability) ranged from -3.5% to 0.5% (mean 1.3%) for COSMED and from -11.0% to 3.6% (mean 0.0%) for VO2Master. Mean values and ranges for V̇E were similar. When used appropriately in laboratory settings, the COSMED Quark and VO2Master Pro systems provide gas exchange and ventilatory data within an acceptable range for metabolic testing equipment that are both reliable and transferable between optimally performing devices.