Comparison of 2 portable respiratory gas analysers

The aim of this study was to compare respiratory measures taken simultaneously using the Cosmed K4b(2) and Cortex Metamax II portable metabolic systems. 10 trained male cyclists performed a graded exercise cycle test to exhaustion (40w.3 min (- 1)) under standardized conditions. The measured respiratory variables were significantly correlated between both devices: r=0.97 and 0.98 (n=10, p<0.01) for oxygen uptake ( V˙O(2)) and ventilation (VE), respectively. Further Bland and Altman plots revealed a good level of agreement for measures of V˙O(2) expressed in mL.min (- 1) [- 670 to 486] (mean bias of - 91.7, i. e., - 3.1%) or in mL.min (- 1).kg (- 1) [- 7.3 to 9.0] (mean bias of 0.85, i. e., 2.3%) and V(E) [- 23.1 to 18.2] (L.min (- 1), mean bias of - 2.4, i. e. - 4.1%). However, poor agreement was found for measures of carbon dioxide (VCO(2), mL.min (- 1)) [- 280 to 1 394] (mean bias of 671 i.e., 20.3%). VCO(2) at maximal exercise intensity was also significantly (p<0.01) greater in the Cortex compared to the Cosmed system. The higher measured CO(2) concentrations led to significantly (p<0.01) higher calculated respiratory exchange ratio (RER) values with the Cortex device. In conclusion, there was satisfactory agreement between the Cosmed K4b(2) and Cortex Metamax II systems for most respiratory measures; however there was a poor level of agreement between VCO(2) and calculated RER measurements between the 2 systems.