How this instrument works
The serum-ascites albumin gradient is exactly what it sounds like: draw blood and sample the ascitic fluid on the same day, measure albumin in both, and subtract the fluid's value from the serum's value. Albumin sits in the bloodstream at a concentration the liver actively maintains, and how much of it leaks into the abdominal cavity depends heavily on whether high pressure in the portal venous system is forcing fluid — but not the larger albumin molecule — out of the vessels and into the peritoneal space.
For decades the standard approach classified ascites as an exudate or a transudate, based on the ascitic fluid's total protein content, borrowing a framework built for pleural fluid. Runyon and colleagues tested that approach directly against the albumin gradient in a prospective study of 901 paired serum and ascites samples published in Annals of Internal Medicine in 1992, and found the gradient correctly separated portal-hypertensive causes from everything else about 97% of the time — meaningfully outperforming the older protein-based split, which stumbled on cases like ascites complicated by infection.
A gradient of 1.1 g/dL or higher points toward portal hypertension as the underlying driver — cirrhosis and heart failure are the two most common examples, since both raise pressure in the vessels feeding the peritoneal cavity, pushing fluid out while the liver's albumin mostly stays in circulation. A gradient below 1.1 g/dL instead suggests a cause unrelated to portal pressure, such as peritoneal carcinomatosis, tuberculous peritonitis, or pancreatitis, where the fluid's own local protein content — including its own albumin — tends to run higher.
- Enter Serum albumin in g/dL, from blood drawn the same day as the paracentesis if possible.
- Enter Ascites fluid albumin in g/dL, measured from the same paracentesis sample.
- Read the resulting gradient in g/dL.
- Compare it against the 1.1 g/dL cutoff: at or above suggests portal hypertension; below suggests a non-portal-hypertensive cause.
Worked example — serum 4.0 g/dL, ascites 1.5 g/dL
A patient with known heavy alcohol use develops abdominal swelling. Serum albumin comes back at 4.0 g/dL, and the ascitic fluid drawn by paracentesis measures 1.5 g/dL. The gradient is 4.0 − 1.5 = 2.5 g/dL — well above the 1.1 g/dL cutoff, consistent with portal hypertension from cirrhosis as the driver, exactly the picture this patient's history would suggest.
Contrast that with a second patient whose serum albumin is 3.0 g/dL and whose ascites albumin is 2.2 g/dL: 3.0 − 2.2 = 0.8 g/dL, below the 1.1 threshold. Even though this second patient's raw ascites albumin number sits closer to the first patient's, the gradient points away from portal hypertension and toward a cause such as peritoneal infection or malignancy — a distinction the older total-protein exudate-transudate method would have been considerably more likely to get wrong.
Questions
What replaced the older exudate-versus-transudate method, and why?
The serum-ascites albumin gradient replaced it. The exudate-transudate framework classified ascitic fluid by its total protein content, a system borrowed from pleural fluid analysis. Runyon and colleagues compared both approaches head-to-head against 901 paired samples in a 1992 study and found the albumin gradient correctly identified portal hypertension as the cause about 97% of the time, clearly ahead of the older protein-based cutoff, which misclassified a meaningful share of cases — particularly ascites complicated by spontaneous bacterial peritonitis.
What does a gradient of 1.1 g/dL or higher mean?
It points toward portal hypertension as the reason fluid is accumulating — most commonly cirrhosis, but also conditions like heart failure or portal vein thrombosis that raise pressure in the same vascular bed. High pressure pushes fluid out of the vessels while albumin, a larger molecule, tends to stay behind in the blood, widening the gap between serum and ascites concentrations.
What does a gradient below 1.1 g/dL suggest?
A cause not driven by portal pressure. Peritoneal carcinomatosis, tuberculous peritonitis, and pancreatitis are classic examples — in each, the ascitic fluid itself tends to carry more locally produced protein, including albumin, narrowing the gap between the two measurements rather than widening it.
Do the two albumin samples need to be drawn at the exact same time?
Same day is the standard, and closer together is better. Albumin concentrations shift gradually rather than by the hour, so a same-day serum draw paired with the paracentesis sample gives a reliable gradient; a serum sample from days earlier introduces error that isn't necessary to accept.
Can a patient have more than one cause of ascites at once, and can the gradient still work?
Yes, and mixed cases are a recognized limitation. A cirrhotic patient who develops a second process, such as peritoneal carcinomatosis on top of portal hypertension, can produce a gradient that sits in an ambiguous zone or trends toward one cause while imaging and cytology point toward another. The gradient is a strong first-pass tool, not a replacement for the rest of the clinical workup when the picture doesn't add up cleanly.
References
- Runyon BA, Montano AA, Akriviadis EA, Antillon MR, Irving MA, McHutchison JG. The serum-ascites albumin gradient is superior to the exudate-transudate concept in the differential diagnosis of ascites. Ann Intern Med. 1992;117(3):215-220 (PubMed)
- MedlinePlus — Peritoneal Fluid Analysis
Read this first: This instrument computes a screening figure from population formulas — it is not a diagnosis, and it cannot see the whole picture a clinician can. Use it to inform a conversation, not to replace one.