A comparison of pulmonary arterial occlusion algorithms for estimation of pulmonary capillary pressure.

Using the arterial occlusion method, we compared five literature-based estimates of pulmonary capillary pressure (Ppc) with the corresponding double occlusion pressures (Pdo) in anesthetized dogs whose chests had been closed after sternotomy for instrumentation. Arterial occlusions were performed with a balloon-tipped pulmonary artery catheter that housed pressure transducers immediately proximal and distal to the balloon. Separation of the proximal and distal pressure waveforms during balloon inflation allowed us to precisely define the moment of occlusion. We fit a monoexponential curve to pressure data beginning 200 ms after the onset of occlusion and a biexponential curve to data beginning at the instant of occlusion, with data obtained over a range of vascular states (control, serotonin infusion, histamine infusion). In addition, we investigated the use of sampling of the raw data to estimate capillary pressure. Three of the five literature-based estimates of Ppc yielded values similar to Pdo. The optimal (least average difference from Pdo) interpolation/extrapolation time points of the curve fits varied, depending on the type of curve fitting and the state of the pulmonary vasculature. We also determined that a close approximation of Pdo may be derived from the raw data, as an alternative to exponential curve fitting.