Core concept2 exam ›
- SAAG = serum albumin - ascitic albumin, both drawn the same day
- Measures portal pressure, not the ascites protein content
- High oncotic gradient is needed to balance a high hydrostatic portal pressure -> the gap widens
- SAAG >=11 g/L = portal hypertension (~97% accurate)
- SAAG <11 g/L = non-portal-hypertensive - peritoneal or lymphatic disease
- Not affected by diuresis, therapeutic paracentesis, or infection - the gradient holds
- Total ascitic protein is the second axis, not a competing test - it separates sinusoidal from post-sinusoidal causes
Key detail
Two-axis classification
| Ascitic protein <25 g/L | Ascitic protein >=25 g/L | |
|---|---|---|
| SAAG >=11 | Cirrhosis (sinusoidal), late Budd-Chiari, fulminant hepatic failure | Cardiac ascites, constrictive pericarditis, early Budd-Chiari, IVC obstruction, sinusoidal obstruction syndrome |
| SAAG <11 | Nephrotic syndrome, protein-losing enteropathy | Peritoneal carcinomatosis, TB peritonitis, pancreatic ascites, serositis, chylous |
- Cirrhosis leaks a low-protein fluid - damaged sinusoids are already leaky, so albumin is lost upstream
- Cardiac ascites leaks a high-protein fluid - sinusoids intact, hydrostatic backpressure only
- The classic exam pairing: high SAAG + high protein = the heart, not the liver
- Mixed ascites (cirrhosis + peritoneal TB or malignancy) occurs in ~5% - SAAG stays high
Adjuncts on the same tap
- PMN >=250/mm3 = SBP, treat before culture returns
- Ascitic amylase >>serum -> pancreatic ascites
- Triglycerides >200 mg/dL -> chylous
- Cytology ~97% sensitive for peritoneal carcinomatosis, but near-useless for hepatocellular carcinoma or massive liver metastases - those cause ascites by portal hypertension, not by seeding
- ADA / mycobacterial culture for TB peritonitis (low yield in cirrhotics)
Clinical relevance1 exam ›
- A high SAAG mandates a portal-pressure answer - do not chase cytology first
- dec sodium intake + spironolactone/frusemide works only for high-SAAG ascites; diuretics are ineffective and harmful in malignant ascites
- Low SAAG + high protein + lymphocytes = TB or carcinomatosis -> laparoscopy with peritoneal biopsy
- Nephrotic ascites is the low-SAAG low-protein exception - loss, not leak
Pleural equivalent (same logic, different cutoffs)
- Light's criteria overcall exudate in diuresed heart failure - use a gradient to rescue
- Serum - pleural protein >31 g/L -> transudate
- Serum - pleural albumin >12 g/L -> transudate (most sensitive; useful in hepatic hydrothorax)
- Hepatic hydrothorax - transudate crossing diaphragmatic defects, usually right-sided, may occur without clinically detectable ascites
Correlations
- Cirrhosis and portal hypertension; SBP; hepatorenal syndrome; TIPS
- Budd-Chiari and sinusoidal obstruction syndrome
- Constrictive pericarditis vs restrictive cardiomyopathy
- Peritoneal carcinomatosis; ovarian malignancy (Meigs)
- Pleural effusion - Light's criteria and transudate/exudate
- Nephrotic syndrome; protein-losing enteropathy
Study aid only. These notes are written with the help of AI. Not for guiding clinical decisions.