Fifty-four cirrhotic patients with refractory ascites were treated with one-session large-volume paracentesis and randomly assigned to two groups. The first group was infused with human albumin, and the second group was infused with hemaccel at doses with comparable oncotic power. The two groups were compared for incidence of complications, recurrence of massive ascites after hospital dismissal and survival rate.

The incidence of complications traditionally related to paracentesis, the probability of requiring readmission to the hospital for ascites (p = 0.48) and the probability of survival after entry into the study (p = 0.85) were the same for the two groups. A multivariate analysis of 16 parameters, including treatment modality, identified absolute unresponsiveness to diuretics as the only independent predictor of mortality. These results indicate that hemaccel infusion may safely replace albumin infusion after total paracentesis for cirrhotic patients with refractory ascites. (HEPATOLOGY 1991;13:707–713.)