ObjectivesAssessment of left ventricular preload and left ventricular compliance changes in septic shock using pulmonary artery occlusion pressure (PAOP) presumes that this pressure accurately reflects left heart filling pressure. We tested the hypothesis that Starling resistor forces render PAOP inaccurate as an index of left heart filling pressure, resulting in misleading assessments of left ventricular compliance changes.DesignProspective, randomized, controlled study.SettingLarge-animal research laboratory at a university.SubjectsFourteen anesthetized domestic pigs weighing 20 to 25 kg.InterventionsPulmonary artery flotation catheters and systemic arterial catheters were placed via right cervical vessels. The left atrium was directly catheterized for left atrial pressure measurements. Left ventricular end-diastolic diameter was measured using sonomicrometry. Other measured or calculated variables were mean arterial pressure, mean pulmonary arterial pressure, PAOP, pulmonary capillary pressure, and pulmonary arterial and venous resistances. Pigs received endotoxin (0.5 mg/kg iv over 30 mins), or an equivalent volume of saline. At t = 60 mins, pigs were resuscitated with lactated Ringer's solution (40 mL/kg over 30 mins). Measurements were taken before and after endotoxin administration, and immediately and 30 mins after lactated Ringer's solution administration. Data were analyzed by two-way analysis of variance (p <or=to .05).Measurements and Main ResultsPAOP, mean pulmonary arterial pressure, and pulmonary capillary pressure increased after endotoxin infusion, while left atrial pressure and left ventricular end-dlastolic diameter decreased. Left atrial pressure and left ventricular end-diastolic diameter returned to baseline immediately after lactated Ringer's solution administration, while PAOP remained increased. Pulmonary arterial resistance and pulmonary venous resistance increased after endotoxin administration, with pulmonary venous resistance showing the greater percent increase. Pulmonary venous resistance decreased transiently immediately after lactated Ringer's solution administration. These changes were not observed in the control group. Accordingly, comparisons of PAOP vs. left ventricular end-diastolic diameter, and left atrial pressure vs. left ventricular end-diastolic diameter yielded divergent results.ConclusionsThe dissociation between PAOP and left atrial pressure, while left ventricular end-diastolic diameter (preload volume) decreased, and changes in pulmonary venous resistance, are strong evidence for Starling resistor forces (venocompression) rather than active venoconstriction. These data indicate that PAOP overestimates left atrial pressure during endotoxin shock, making it an inaccurate index of left ventricular preload. This overestimation can cause misleading assessments of left ventricular compliance.(Crit Care Med 1996; 24:1618-1625)