Comparison between alveolo arterial gradient predicted from mixing studies and the observed gradient

¹ Department of Medicine, Columbia University, College of Physicians and Surgeons, and Cardio-Pulmonary Laboratory (Columbia University Division), Bellevue Hospital, New York City

The component of the A-a gradient due to uneven V/Q ratios would average 9 mm in 18 studies on six normal subjects if the observed uneven ventilation coexisted with even perfusion of the lung. This component constitutes the largest part of the A-a oxygen gradient of the resting man breathing air. In cases where the A-a gradient, predicted on the basis of helium mixing studies and the assumption of even perfusion, exceeds the observed total gradient, there must be either an error in the measurements or else the less ventilated alveoli must be underperfused with blood. In a group of eight subjects with cardiopulmonary dysfunction, the A-a gradient predicted from the helium mixing studies averaged 23 mm, with a range of 5–65 mm. The measured total gradient averaged 25 mm, a nonsignificant difference. Comparison with total measured A-a oxygen gradient is difficult because, this gradient is underestimated to a variable degree when computed on the assumption that arterial and alveolar CO₂ tensions are the same. It is clear, however, that the predicted component due to uneven Va/Qc ratios accounts for the largest part of the total observed A-a gradient.