ObjectiveTo investigate whether rapid or slowly infused intravenous fat emulsions affect the ratio of prostaglandin I2/thromboxane A2in arterial blood, pulmonary hemodynamics, and gas exchange.DesignProspective, controlled, randomized, crossover study.SettingOperative intensive care unit of a university hospital.PatientsEighteen critically ill patients. Ten patients were stratified with severe sepsis, and eight patients had acute respiratory distress syndrome (ARDS).InterventionsPatients were assigned randomly to receive intravenous fat emulsions (0.4 × resting energy expenditure) over 6 hrs (rapid fat infusion) or 24 hrs (slow fat infusion) along with a routine parenteral nutrition regimen, by using a crossover study design.Measurements and Main ResultsSystemic and pulmonary hemodynamics as well as gas exchange measurements were recorded via respective indwelling catheters. Arterial thromboxane B2and 6-keto-prostaglandin-F1&agr;plasma concentrations were obtained by radioimmunoassay, and 6-keto-prostaglandin-F1&agr;/thromboxane B2ratios (P/T ratios) were calculated. Data were collected immediately before and 6, 12, 18, and 24 hrs after onset of fat infusion. In the ARDS group, P/T ratio increased by rapid fat infusion. Concomitantly, pulmonary shunt fraction, alveolar-arterial oxygen tension difference [P(a-a)o2]/Pao2, and cardiac index increased as well, whereas pulmonary vascular resistance and Pao2/Fio2declined. After slow fat infusion, a decreased P/T ratio was revealed. This was accompanied by decreased pulmonary shunt fraction, lowered P(a-a)o2/Pao2, and increased Pao2/Fio2. Correlations between plasma concentrations of 6-keto-prostaglandin-F1&agr;or thromboxane B2and measures of respiratory performance could be shown during rapid and slow fat infusion, respectively. In the sepsis group, the P/T ratio remained unchanged at either infusion rate, but pulmonary shunt fraction and P(a-a)o2/Pao2decreased after rapid fat infusion, whereas Pao2/Fio2increased.ConclusionPulmonary hemodynamics and gas exchange are related to changes of arterial prostanoid levels in ARDS patients, depending on the rate of fat infusion. In ARDS but not in sepsis patients clear of pulmonary organ failure, a changing balance of prostaglandin I2and thromboxane A2may modulate gas exchange, presumably via interference with hypoxic pulmonary vasoconstriction.