Inhomogeneity of ventilation leads to unpredictable errors in measured D(L)CO.

We evaluated the effects of inhomogeneity of ventilation on single-breath (SB), rebreathing (RB) and open circuit (OC) D(L)CO using a mathematical model consisting of two alveolar compartments and a common dead space. Inhomogeneity in ventilation was studied by altering inspired volume, initial alveolar volume and compartment size independently. When distribution of inspired volume between alveolar compartments was inhomogeneous (9:1), D(L)CO was underestimated by 35% for SB, 25% for RB, and 16% for OC, and there was an underestimation in V(A) of 9%, 15% and 9%, respectively. With inhomogeneity in initial alveolar volume there was an overestimation in D(L)CO of 13%, 7% and 11% for SB, RB and OC techniques and an underestimation of V(A) of 7%, 12% and 9%. Finally inhomogeneity of compartment size led to an underestimation of D(L)CO of 18%, 35% and 36% with no change in measured V(A). These results suggest D(L)CO measurements are sensitive to inhomogeneity of ventilation, and importantly, all techniques were at times, significantly in error.