Evaluation of cardiac output from a tidally ventilated homogeneous lung model

We used the direct Fick measurements to validate a method for estimating cardiac output by iteratively fitting V̇_CO2 at the mouth to lung model values. This model was run using a series of 50, 30 and 10 breaths to test sensitivity to number of breaths used for fitting. The lung was treated as a catenary two-compartment lung model consisting of a dead space compartment connected with a single alveolar space compartment, perfused with constant pulmonary blood flow. The implemented mathematical modeling described variations in O₂ and CO₂ compartmental fractions and alveolar volume. This model also included pulmonary capillary gas exchange. Experimental data were collected from measurements performed on six healthy subjects at rest and during 20, 40, 60 and 85-90% of peak V̇_O2. The correlation between the two methods was highest and the average agreement between the methods was best using 50 breaths (R = 0.95; P<0.0001; Q̇_model = 1.1 Q̇_Fik - 2.3). The mean difference and lower to upper limits of agreement between measured and estimated data were 0.7 l/min (-2.7 to 4.1 l/min) for cardiac output; -0.9 ml/100 ml (-1.3 to -0.5 ml/100 ml) for arterial O₂ content; -0.8 ml/100 ml (-3.8 to 2.2 ml/100 ml) for mixed venous O₂ content and -0.1 ml/100 ml (-2.9 to 2.7 ml/100 ml) for arteriovenous difference O₂ content. The cardiac output estimated by the lung model was in good agreement with the direct Fick measurements in young healthy subjects.