Gas exchange measurement during pediatric mechanical ventilation--agreement between gas sampling at the airway and the ventilator exhaust.

BACKGROUND & AIMS: A variety of indirect calorimetry (IC) devices are used for gas exchange measurement and calculation of resting energy expenditure (REE) in the pediatric intensive care unit. The aim of this investigation was to compare oxygen consumption (VO2), carbon dioxide elimination (VCO2), REE and respiratory quotient (RQ) in mechanically ventilated children, obtained by 2 devices using distinct gas sampling methods.

METHODS

Mechanically ventilated children were targeted for IC and gas exchange measurements were recorded for a 30 min period, simultaneously using the E-COVX(®) (gas sampling at the airway) and the Vmax(®) (gas sampling at the humidifier and ventilator exhaust). Steady state gas exchange measurements by the 2 devices were tested for agreement using Spearman correlation and Bland-Altman analysis.

RESULTS

Steady state data from both devices were available in 19 tests and were included in the analysis. The correlations coefficients for measurements by the 2 devices were r = 0.903(P < 0.001), 0.955(P < 0.001), 0.944(P < 0.001) and 0.484(P < 0.05) for VO2, VCO2, REE and RQ, respectively. The mean percentage bias (limits of agreement) for VO2, VCO2, REE and RQ values between the two methods (Vmax-E-COVX) was 0.2 (-41.8-42.3), -0.8 (-21.8-20.1), -2.2 (-33.9-29.6) and 1.9 (-21-24.9) respectively.

CONCLUSIONS

Despite strong correlations and small mean biases for VO2, VCO2 and REE obtained by the Vmax(®) and E-COVX(®), the limits of agreement were beyond the clinically acceptable range. These devices should not be used interchangeably for gas exchange measurements in mechanically ventilated children.