Annals of Internal Medicine

Ascites occurs in most patients with severe cirrhosis. Although the presence of ascites is often obvious, smaller amounts of ascitic fluid can be difficult to diag-nose. Ultrasonography has been helpful in diagnosing the presence and determining the cause of ascites (1). Hepatic outflow obstruction is central to ascites forma-tion. Of continued debate is whether ascites formation is a primary event resulting in relative hypovolemia and secondary renal salt retention or is secondary to pri-mary renal sodium avidity. There is evidence support-ing both the "underfill " (2, 3) and "overflow " (4) the-ories, and various therapies are based on correcting the abnormalities postulated by these theories. Before the 1950s, repeated large-volume paracentesis was done with impunity on patients with ascites. Clin