Modeling of end-tidal and arterial PCO2 gradient: comparison with experimental data.

Abstract

The aim of this study was to test whether a tidally ventilated homogeneous lung model can correctly describe arterial and end-tidal gas partial pressures and thus the difference in end-tidal and arterial gas partial pressures at rest and during exercise. The implemented mathematical modeling described variations during the breathing cycle in CO2 and O2 fractions, alveolar volume, and pulmonary capillary gas exchange. Experimental data were obtained from measurements performed by 17 healthy subjects at rest and during 40, 50, 65, and 75% exercise .VO(2max) on a cycle ergometer. VO2, VCO2, and PET,CO2 were continuously measured using the MedGraphics CPX/D gas exchange system. Arterial gases were measured in brachial artery blood samples drawn simultaneously with gas exchange. Cardiac output was measured using the CO2 rebreathing method corrected by the blood sample data. The model was driven using experimental data for ventilation, VO2, VCO2, and cardiac output. The mean difference and the upper and lower limits of agreement between measured and simulated data were -0.004, +0.84, and -0.84 Torr for Pa,CO2; -0.06, +0.64, and -0.76 Torr for Pa,O2; -1.96, +2.84, and -6.76 Torr for PET,CO2; and +7.20, +25.80, and -11.40 Torr for PET,O2. Actual PET,CO2-Pa,CO2 difference increased significantly with workload (P < 0.0001) from 0.3 +/- 3 Torr at rest to 4.7 +/- 2.5 Torr at 75% .VO(2max). Model-simulated PET,CO2-Pa,CO2 difference also increased significantly with exercise (P < 0.0001) from 0.7 +/- 1.7 Torr at rest to 9.1 +/- 3.4 Torr at 75% .VO(2max). The lung model described actual arterial CO2 partial pressures better than variations in end-tidal CO2 partial pressures and thus better than the gradient in end-tidal arterial CO2 partial pressures.