The level of accuracy and agreement in measures of FEO2, FECO2 and VE between the cosmed K4b2 portable, respiratory gas analysis system and a metabolic cart

This study aimed to assess the accuracy of the Cosmed K4b(2) (Cosmed, Italy) portable metabolic system that measures FEO2, FECO2 and (V)over dot(E) on a breath by breath basis. For gas concentration comparisons, expired air from 20 subjects performing treadmill running was collected in a 600 litre chain compensated Collins Tissot tank and analysed for FEO2 and FECO2 using a laboratory metabolic cart and the Cosmed K4b(2) metabolic system. For ventilation comparisons, serial steady state (V)over dot(E) (STPD) values were measured on 10 subjects using the Cosmed K4b(2) ventilation turbine and a Morgan ventilation monitor during a continuous treadmill running protocol at ascending speeds of 8, 11 and 14 km.h(-1). The Cosmed K4b(2) FEO2 and FECO2 measures were significantly lower (P < 0.001) than the metabolic cart values. Pearson correlation coefficients (r) and the standard error of measurement (SEM) demonstrated a high association between the Cosmed and the metabolic cart measures (FEO2 r = 0.971, SEM 0.071: FECO2 r = 0.925, SEM 0.087). Cosmed (V)over dot(E) (l.min(-1)) measures were significantly greater than Morgan values at running speeds of 8 km.h(-1) (P < 0.001) and 11 km.h(-1) (P < 0.001) but not significantly different at 14 km.h(-1) (P > 0.05). When (V)over dot(E) measures at the three running speeds were combined. the mean difference between instrument measures ranged between 3.5 - 4.0 l.min(-1) but the values were highly correlated (r = 0.982. P < 0.01: SEM 3.03). Linear regression analysis revealed the following regression equations to predict metabolic cart values from Cosmed measures: FEO2 = 0.852 + 0.963 Cosmed (R-2 = 0.940, P < 0.001). FECO2 = 0.627 + 0.878 Cosmed (R-2 = 0.856, P < 0.001), (V)over dot(E) = -2.50 + 0.984 Cosmed (R-2 = 0.965, P < 0.001). The results indicated that the Cosmed K4b(2) Unit assessed here produced measures of FEO2, FECO2 and (V)over dot(E) that had strong correlation to values obtained from a metabolic cart. However. linear regression analysis may further improve the accuracy of Cosmed K4b(2) measures when compared to metabolic cart values.