Theoretical basis of alveolar sampling.

The conditions under which the partial pressure of a solvent in the alveolar gas is likely to provide a valid index of its partial pressure in the mixed venous blood, and thus of whole-body exposure, is explored on a theoretical basis. Under steady-state conditions, providing the solvent's blood/gas partition coefficient exceeds 10, its mixed venous and alveolar pressures will agree within 10% and become virtually identical during the prolonged expiration necessary to obtain an alveolar sample. With less soluble solvents, however, this is not necessarily so, and when the partition coefficient is less than three alveolar sampling is unlikely to provide a valid index of mixed venous levels. Abnormalities of pulmonary function, particularly ventilation/perfusion imbalance, are unlikely to alter these conclusions significantly.