In the United States, among patients with ascites, approximately 85% of have cirrhosis and 15% have a nonhepatic cause. Ascites is the most common complication of cirrhosis, with about 50% of compensated cirrhotics developing ascites over a 10-year period. Further, ascites is the leading cause of hospital admission for patients with cirrhosis. Development of fluid retention is a hallmark of hepatic decompensation and is associated with significant morbidity and mortality in cirrhotic patients. Successful treatment of ascites depends on identifying the correct cause, since some nonhepatic causes of ascites, such as peritoneal carcinomatosis, do not respond to diuretic therapy. The following discussion will focus on the causes of ascites (Figure 1) and the initial evaluation of a patient with new-onset ascites. The treatment of patients who have ascites will be discussed in the case Management of Ascites.
A careful history and physical examination is essential in the overall diagnosis and management of ascites. The history should elicit common risk factors for liver disease, such as alcohol use and viral hepatitis infection. With the rise in obesity in the United States, non-alcoholic fatty liver disease-related cirrhosis has become more common; thus risk factors for non-alcoholic fatty liver disease should be looked for, namely obesity, diabetes, hypertension, and hyperlipidemia. In addition, the evaluation should determine whether the patient has a history of cancer, renal disease, heart failure, or tuberculosis. If the physical examination shows full and bulging flanks (Figure 2), then evaluation of "shifting dullness" should be performed (Figure 3). First, identify the tympany-dullness interface when the patient is lying supine, and then have the patient lie on their side and assess whether the interface shifts. The presence of "shifting dullness" has 83% sensitivity and 56% specificity for the diagnosis of ascites. In most situations, however, a definitive diagnosis requires an abdominal ultrasound, especially in obese patients. In general, when a patient complains of sudden increase in abdominal girth, over weeks, then ascites is more likely than when increase in girth over months, which is most likely due to fat-related weight gain.
All adult patients with new onset ascites of uncertain cause should undergo paracentesis. Performing a diagnostic paracentesis is the most rapid and cost-effective method to diagnose the cause of new onset ascites. With appropriate fluid analysis one can usually differentiate between ascites due to cirrhosis/portal hypertension and other causes. Furthermore, since patients presenting with new onset ascites often have concomitant infection, obtaining ascitic fluid can also diagnose infected ascitic fluid (spontaneous bacterial peritonitis). An appropriately trained clinician can quickly and safely perform a diagnostic paracentesis, even in an outpatient clinic setting. It is essential, however, that clinicians performing a paracentesis receive (or have received) supervision on how to appropriately perform this procedure. Special caution should be given for patients who are pregnant, or those with bowel obstruction, bowel adhesion, organomegaly, or a distended urinary bladder. The following provides an overview of the key steps in performing a diagnostic abdominal paracentesis (Figure 4Performing a Diagnostic ParacentesisObtaining Informed ConsentPositioning Patient at 30-45° AngleAppropriate Sites for Needle InsertionIdentify Location for Needle InsertionPreparation: Cleaning SitePreparation: Placing Sterile Drapes Over FieldPreparation: Anesthetize Skin and Deeper TissuesNeedle Insertion: Angular Insertion TechniqueNeedle Insertion: Z Tract Technique).
Complications of Diagnostic Paracentesis
Complications are rare (approximately 1%) if a relatively smaller gauge needle is used (typically 21-gauge) and the needle-insertion site for the paracentesis is appropriately identified. Bleeding is the most common complication and infrequently hemoperitoneum can occur. In one study of 4729 paracentesis procedures, only 8 (0.2%) had a bleeding complication. Further, in a separate study of 1,100 paracentesis procedures performed without pre-procedure transfusion of blood products, no bleeding complications were noted despite a platelet counts less than 50,000 cells/mm3 in 54% of the procedures and an increased prothrombin time (international normalization ratios [INR] greater than 1.5) in 75% of the procedures. Because bleeding complications occur rarely, the routine use of fresh frozen plasma or platelets prior to performing paracentesis is not recommended. Other complications, which are rare, include bowel perforation.
Analysis of Ascitic Fluid
Approach to Ordering Laboratory Studies: For patients with new onset ascites, the AASLD has categorized ascitic fluid diagnostic tests in four categories: routine, optional, usual, and unhelpful (Figure 5). If uncomplicated ascites due to cirrhosis is suspected then only a minimal number of studies are required: cell count and differential, albumin, and total protein concentration of the ascitic fluid. If any of these tests are unexpectedly abnormal, then additional tests can be ordered.
Serum-Ascites Albumin Gradient (SAAG): Measuring the ascitic albumin level in addition to a serum albumin level (obtained on the same day) allows one to calculate the serum-ascites albumin gradient (SAAG), which is simply the difference between the serum and ascitic fluid albumin values (SAAG = albumin concentration of serum - albumin concentration of ascitic fluid). A SAAG of 1.1 g/dL or greater is consistent with portal hypertension as the cause of ascites and values less than 1.1 less indicate other disease processes (Figure 6)[5,7,12]. The SAAG correctly differentiates causes of ascites due to portal hypertension from causes not due to portal hypertension with approximately 97% accuracy and clearly is superior to the old transudate-exudate concept.
Culture: If infection of the ascitic fluid is suspected (signs and symptoms that include fever, leukocytosis, abdominal pain, encephalopathy, or hypotension), then the ascitic fluid should be placed directly into bacterial cultures bottles and sent for culture analysis. Furthermore, a presumptive diagnosis of spontaneous bacterial peritonitis (infected ascitic fluid) can be made if the absolute neutrophil count of the ascitic fluid is 250 cells/mm3 or greater. Gram's staining of ascitic fluid generally has low yield, except in cases of intestinal perforation. In suspected cases of peritoneal tuberculosis, ascitic fluid can be sent for mycobacteria smear and culture, but this test should be reserved for highly suspicious cases of peritoneal tuberculosis since the diagnostic sensitivity for acid-fast smear is extremely low (less than 1%) and only 50% for culture.
Cytologic Analysis: Rarely, cytologic analysis of the ascitic fluid is necessary to evaluate for malignancy. Cytologic analysis is only useful when peritoneal carcinomatosis is suspected, where the accuracy of a positive test is nearly 97%. In patients with ascites, the cancer antigen CA 125 should not be measured in ascites fluid; the CA 125 level is elevated in essentially all patients with ascites because ascites places pressure on mesothelial cells, which leads to release of CA 125. Therefore, checking CA 125 in women is especially problematic in this setting, since an elevated CA 125 will lead to unnecessary gynecologic evaluation to evaluate for ovarian cancer.