ObjectiveWe examined the effects of arginine vasopressin, when substituted for norepinephrine as a vasopressor in septic shock, on global and hepatosplanchnic hemodynamic and oxygen transport variables.DesignExperimental study.SettingIntensive care unit.SubjectsTwelve septic shock patients.InterventionsNorepinephrine was replaced by vasopressin in a dose sufficient to keep mean arterial blood pressure constant. Blood flow, oxygen delivery, and oxygen consumption of the hepatosplanchnic region (calculated by a hepatic venous catheter technique using the Fick principle during continuous infusion of indocyanine green), global hemodynamics (by thermodilution), and gastric regional Pco2gap (by air tonometry) were calculated during infusion of norepinephrine (mean, 0.56 &mgr;g·kg−1·min−1; range, 0.18–1.1 &mgr;g·kg−1·min−1) and again 2 hrs after replacement by vasopressin (mean, 0.47 IU/min; range, 0.06–1.8 IU/min).Measurements and Main ResultsCardiac index decreased significantly from 3.8 ± 1.3 to 3.0 ± 1.1 L·min−1·m−2, heart rate decreased from 96 ± 14 to 80 ± 16 min−1(p< .01), and global oxygen uptake decreased from 248 ± 67 to 218 ± 75 mL/min (p< .05). Absolute splanchnic blood flow tended to increase, although not significantly, whereas fractional splanchnic blood flow increased from 10.8 ± 7.6 to 25.9 ± 16.6% of cardiac output (p< .05). Gastric regional Pco2gap increased from 17.5 ± 26.6 to 36.5 ± 26.6 mm Hg (p< .01).ConclusionVasopressin, in doses sufficient to replace the vasopressor norepinephrine, had mixed effects in septic shock patients. Hepatosplanchnic blood flow was preserved during substantial reduction in cardiac output. An increased gastric Pco2gap suggests that the gut blood flow could have been redistributed to the disadvantage of the mucosa. Based on these limited data, it does not appear beneficial to directly replace norepinephrine with vasopressin in septic shock.