On the estimation of pulmonary capillary pressure from arterial occlusion

We have attempted to evaluate some approaches for estimating pulmonary capillary pressure from the transient pressure data obtained from the tip of the Swan-Ganz catheter after inflation of the balloon. To this end experiments were carried out in anesthetized dogs in which pulmonary vasoconstriction was induced by infusion of histamine, serotonin, or norepinephrine or by hypoxia. In a group of closed chest dogs, the time course for the fall in the catheter tip pressure toward the wedge pressure was different for the different vasoconstrictors. For example, during histamine infusion the fall in pressure was slow, whereas during serotonin infusion it was rapid. To evaluate the information in the decay curves, a group of dogs was studied with open thorax to permit the simultaneous occlusion of both the artery and vein of one lung lobe (double occlusion) as well as occlusion of the lobar artery alone (arterial occlusion). The pattern of response to the various vasoconstrictors observed with lobar arterial occlusion was similar in these animals to the closed chest animals. The equilibrium pressure after double occlusion (the double occlusion pressure), expressed as a fraction of the mean arterial-venous pressure difference before occlusion, was correlated with the mean decay time of the arterial pressure curve after arterial occlusion. We also found significant correlations between the double occlusion pressure and the estimates of capillary pressure based on model interpretations of the pressure decay curve after arterial occlusion. However, there was wide scatter in the individual comparisons. Overall, the results suggest that the pressure decay curve at the tip of the Swan-Ganz catheter after inflation of the balloon in a lobar or sublobar artery is strongly influenced by the same features of the pulmonary vascular bed that determine the double occlusion pressure. However, the calculation of a specific value of the capillary pressure from the arterial occlusion data with presently proposed methods needs to be interpreted carefully since there may be the potential for considerable variability in an individual estimate.