Intrapulmonary shunt in excised dog lobes: comparison of oxygen and inert gas methods

We evaluated the contribution of measurement errors in PO2 to the discrepancy between the intrapulmonary shunt values simultaneously obtained by the conventional O2 method (O2-shunt) and the multiple inert gas elimination technique (IG-shunt). Excised left lower lobes of dogs were perfused at a constant flow and ventilated at constant volume. Variable degrees of intrapulmonary shunt were created by O2 absorption while the lobes were collapsed. IG-shunt ranged from 0 to 66%. Although the O2-shunt and IG-shunt values correlated well (r = 0.97), the slope was only 0.74 and the regression line crossed the line of identity at 12% IG-shunt. The gas-to-blood ratio of the oxygen electrode was constant over the range 80-687 Torr and therefore did not seem to be an important factor in the shunt discrepancy. Errors due to nonlinearity of the O2 electrode also had only a minor effect. However, PO2 history of the electrode just before injection of blood samples into the cuvette had a profound and complex effect on the measured PO2. We conclude that PO2 electrode hysteresis can be a major factor in the wide variability of PO2 measurements and could account for most of the differences between our O2-shunt and IG-shunt data.