Restructuring Riley's Historic 3-Compartment Lung Model for Evaluation of Pulmonary Gas Exchange

In 1951, Riley's classic 3-compartment model of gas exchange estimated pulmonary shunt, alveolar dead space, and an effective compartment representing the functioning lung. But trial-and-error steps and conversion charts made its application impractical. We implemented estimates of alveolar and effective PCO₂ to simplify computations, making it useful when more advanced technologies are unavailable. Using stepwise computations, we studied 10 healthy individuals and 43 outpatients with mild to severe chronic obstructive pulmonary disease and, in another study, 32 healthy subjects during 12 h of hypobaric hypoxia at 426 mm Hg (ALT). The "effective" PaCO₂ due to pulmonary shunt and Haldane effect when breathing increased O₂ was calculated via the CO₂ dissociation curve. The model was applied while breathing air and 25% O₂ to simulate sea level in outpatients at 1620 m. Pulmonary shunt rose significantly with increasing hypoxemia (P<0.001), whereas alveolar dead space remained high. Breathing 25% O₂ reduced the shunt (P<0.001) by elevating systemic PO₂. The effective compartment in healthy subjects was 0.87, but only 0.41 in patients with severe hypoxemia, increasing to 0.45 on 25% O₂ (P=0.031). In ALT, a scoring system demonstrated that 16 subjects experienced acute mountain sickness (AMS) after 1 h with a significant increase in pulmonary shunt compared with 16 subjects without AMS. The model shows that hypoxemia in patients is associated with perfusion redistribution from high to low V/Q regions, consistent with reports using more sophisticated techniques. Subjects susceptible to AMS also increased shunt, suggesting autonomic instability.