Serum/ascites albumin gradient: its value as a rational approach to the differential diagnosis of ascites.

BACKGROUND: The utility of differentiating ascites into 'transudate' and 'exudate' has recently been challenged. The aim of the present study was to compare the diagnostic accuracy of the serum/ascites albumin gradient, proposed as a new biochemical criterion for the differential diagnosis of ascites, with the markers traditionally used for the classification of peritoneal fluid into transudate and exudate.

METHODS: Paired ascitic fluid and serum samples from 51 patients were examined with an established method for the diagnosis of the cause of ascitic fluid collection. Included in the study were 32 patients with ascites related to portal hypertension (cirrhosis, n = 28; 'cardiac' ascites, n = 2; Budd-Chiari, n = 2) and 19 patients with ascites not related to portal hypertension (peritoneal carcinomatosis, n = 17; tuberculous peritonitis, n = I; secondary bacterial peritonitis, n = 1). Specimens were collected during an episode of spontaneous bacterial peritonitis in 7 of 28 patients with cirrhosis. The serum/ascites albumin gradient was compared with ascitic fluid total protein, ascites/serum total protein ratio, ascites lactic dehydrogenase concentration, and ascites/serum lactic dehydrogenase ratio.

RESULTS: The diagnostic accuracy was 98% for the serum/ascites albumin gradient compared with only 52%-80% for the four other markers tested. In patients with infected ascites, diagnostic accuracy was 89% for the albumin gradient and < or = 50% for the four other markers.

CONCLUSIONS: The classification of ascites into transudate and exudate appears to be based on markers with low diagnostic accuracy. Differential diagnosis of ascites should be based on the serum/ascites albumin gradient, which is a reliable marker distinguishing ascites related to portal hypertension from all other causes of ascitic fluid collection, regardless of the presence of bacterial infection.