Interpretation of alveolar-arterial oxygen tension difference

In this study, an assessment is given of errors resulting from the use of the alveolar-arterial oxygen difference determined on both room air and 100% oxygen in estimating changes in total shunt fraction of 41 patients hospitalized with long bone fractures. In 113 studies, changes in alveolar-arterial oxygen difference in 29 patients were in the opposite direction to changes in shunt fraction. Based upon these studies, changes of less than 45 millimeters in the arterial oxygen tension determined with a patient breathing 100% oxygen are not reliable indicators of direction of change in shunt fraction. In 126 studies, the shunt fraction determined from arterial and mixed venous oxygen contents in 71 patients was greater when determined on 100% oxygen than when determined on room air, a possible indication of the induction of alveolar or small airway collapse. The errors in estimation of shunt fraction due to assuming a value for arteriovenous oxygen content difference become larger as total shunt fraction increases; in particular, use of the alveolar-arterial oxygen difference as a guide to serial changes in pulmonary dysfunction can be particularly misleading when the alveolar-arterial oxygen tension difference is so large that the arterial hemoglobin is less than fully saturated on 100% oxygen. Use of mixed venous samples was found necessary, in these instances, to avoid large errors in estimation of total shunt fraction.