Separate lung blood flow in anesthetized dogs: a comparative study between electromagnetometry and SF6 and CO2 elimination.

Anesthetized, prone dogs were intubated with a double-lumen endobronchial tube, and the lungs were ventilated independently. Three methods of recording differential blood flow were compared during unilateral lung hypoxia: electromagnetic flow measurement, flow probes being fitted onto each main pulmonary artery after thoracotomy (QPr); SF6 elimination from each lung, the inert gas being continuously infused into a central vein (QSF6); and CO2 elimination (QCO2). During control conditions (100% O2 to both lungs), the test lung QPr was 54% of cardiac output, and corresponding QSF6 and QCO2 were 56% and 52%, respectively. Hypoxic challenge with 8% O2 to the test lung reduced QPR, QSF6, and QCO2 by 25%, 27%, and 7%, respectively. Ventilation of the test lung with pure nitrogen reduced its blood flow further, QPr, QSF6, and QCO2 being reduced by 39%, 42%, and 23%, respectively, from initial control. A strong correlation between test lung QPr and QSF6 was seen with a slope of 0.90 (r:0.89, P less than 0.001). Only 60% of the reduction in test lung blood flow was detected by CO2 elimination, as compared to electromagnetic flow measurement or SF6 elimination. The poor results obtained with CO2 elimination can be explained by its dependence on the ventilation-perfusion ratio and the effect of oxygen tension on the CO2 binding capacity of blood (Haldane effect). The findings emphasize the necessity of using an inert, poorly soluble gas for the measurement of separate lung blood flow.