EASL Clinical Practice Guidelines

Uncomplicated ascites

Evaluation of patients with ascites

Approximately 75% of patients presenting with ascites in Western Europe or the USA have cirrhosis as the underlying cause. For the remaining patients, ascites is caused by malignancy, heart failure, tuberculosis, pancreatic disease, or other miscellaneous causes.

Diagnosis of ascites

The initial evaluation of a patient with ascites should include history, physical examination, abdominal ultrasound, and laboratory assessment of liver function, renal function, serum and urine electrolytes, as well as an analysis of the ascitic fluid.

The International Ascites Club proposed to link the choice of treatment of uncomplicated ascites to a classification of ascites on the basis of a quantitative criterion (Table 2). The authors of the current guidelines agree with this proposal.

Table 2
Grading of ascites and suggested treatment.

Grade of ascites Definition Treatment
Grade 1 ascites Mild ascites only detectable by ultrasound No treatment
Grade 2 ascites Moderate ascites evident by moderate symmetrical distension of abdomen Restriction of sodium intake and diuretics
Grade 3 ascites Large or gross ascites with marked abdominal distension Large-volume paracentesis followed by restriction of sodium intake and diuretics (unless patients have refractory ascites)

A diagnostic paracentesis with an appropriate ascitic fluid analysis is essential in all patients investigated for ascites prior to any therapy to exclude causes of ascites other than cirrhosis and rule out spontaneous bacterial peritonitis (SBP) in cirrhosis. When the diagnosis of cirrhosis is not clinically evident, ascites due to portal hypertension can be readily differentiated from ascites due to other causes by the serum–ascites albumin gradient (SAAG). If the SAAG is greater than or equal to 1.1 g/dl (or 11 g/L), ascites is ascribed to portal hypertension with an approximate 97% accuracy [[8], [9]]. Total ascitic fluid protein concentration should be measured to assess the risk of SBP since patients with protein concentration lower than 15 g/L have an increased risk of SBP [10].

A neutrophil count should be obtained to rule out the existence of SBP [10]. Ascitic fluid inoculation (10 ml) in blood culture bottles should be performed at the bedside in all patients. Other tests, such as amylase, cytology, PCR and culture for mycobacteria should be done only when the diagnosis is unclear or if there is a clinical suspicion of pancreatic disease, malignancy, or tuberculosis [[8], [9], [10], [11]].

Recommendations A diagnostic paracentesis should be performed in all patients with new onset grade 2 or 3 ascites, and in all patients hospitalized for worsening of ascites or any complication of cirrhosis (Level A1).

Neutrophil count and culture of ascitic fluid (by inoculation into blood culture bottles at the bedside) should be performed to exclude bacterial peritonitis (Level A1).

It is important to measure ascitic total protein concentration, since patients with an ascitic protein concentration of less than 15 g/L have an increased risk of developing spontaneous bacterial peritonitis (Level A1) and may benefit from antibiotic prophylaxis (Level A1).

Measurement of the serum–ascites albumin gradient may be useful when the diagnosis of cirrhosis is not clinically evident or in patients with cirrhosis in whom a cause of ascites different than cirrhosis is suspected (Level A2).

Prognosis of patients with ascites

The development of ascites in cirrhosis indicates a poor prognosis. The mortality is approximately 40% at 1 year and 50% at 2 years [7]. The most reliable factors in the prediction of poor prognosis include: hyponatremia, low arterial pressure, increased serum creatinine, and low urine sodium [[7], [12]]. These parameters are not included in the Child-Turcotte-Pugh score (CTP score) and among them, only serum creatinine is included in the Model for end-stage liver disease (MELD score). Furthermore, since serum creatinine has limitations as an estimate of glomerular filtration rate in cirrhosis [13], these scores probably underestimate the mortality risk in patients with ascites [14]. Since allocation for liver transplantation is based on the MELD score in several countries, patients with ascites may not receive an adequate priority in the transplant lists. Therefore, there is a need for improved methods to assess prognosis in patients with ascites.

Recommendations Since the development of grade 2 or 3 ascites in patients with cirrhosis is associated with reduced survival, liver transplantation should be considered as a potential treatment option (Level B1).

Management of uncomplicated ascites

Patients with cirrhosis and ascites are at high risk for other complications of liver disease, including refractory ascites, SBP, hyponatremia, or hepatorenal syndrome (HRS). The absence of these ascites-related complications qualifies ascites as uncomplicated [11].

Grade 1 or mild ascites

No data exist on the natural history of grade 1 ascites, and it is not known how frequently patients with grade 1 or mild ascites will develop grade 2 or 3 ascites.

Grade 2 or moderate ascites

Patients with moderate ascites can be treated as outpatients and do not require hospitalization unless they have other complications of cirrhosis. Renal sodium excretion is not severely impaired in most of these patients, but sodium excretion is low relative to sodium intake. Treatment is aimed at counteracting renal sodium retention and achieving a negative sodium balance. This is done by reducing the sodium intake and enhancing the renal sodium excretion by administration of diuretics. Whilst the assumption of the upright posture activates sodium-retaining systems and slightly impairs renal perfusion [15], forced bed rest is not recommended because there are no clinical trials assessing whether it improves the clinical efficacy of the medical treatment of ascites.

Sodium restriction

A negative sodium balance can be obtained by reducing dietary salt intake in approximately 10–20% of cirrhotic patients with ascites, particularly in those presenting with their first episode of ascites [[16], [17]]. There are no controlled clinical trials comparing restricted versus unrestricted sodium intake and the results of clinical trials in which different regimens of restricted sodium intake were compared are controversial [[17], [18]]. Nevertheless, it is the current opinion that dietary salt intake should be moderately restricted (approximately 80–120 mmol of sodium per day). A more severe reduction in dietary sodium content is considered unnecessary and even potentially detrimental since it may impair nutritional status. There are no data to support the prophylactic use of salt restriction in patients who have never had ascites. Fluid intake should be restricted only in patients with dilutional hyponatremia.

Recommendations Moderate restriction of salt intake is an important component of the management of ascites (intake of sodium of 80–120 mmol/day, which corresponds to 4.6–6.9 g of salt/day) (Level B1). This is generally equivalent to a no added salt diet with avoidance of pre-prepared meals.

There is insufficient evidence to recommend bed rest as part of the treatment of ascites. There are no data to support the use of fluid restriction in patients with ascites with normal serum sodium concentration (Level B1).


Evidence demonstrates that renal sodium retention in patients with cirrhosis and ascites is mainly due to increased proximal as well as distal tubular sodium reabsorption rather than to a decrease of filtered sodium load [[19], [20]]. The mediators of the enhanced proximal tubular reabsorption of sodium have not been elucidated completely, while the increased reabsorption of sodium along the distal tubule is mostly related to hyperaldosteronism [21]. Aldosterone antagonists are more effective than loop diuretics in the management of ascites and are the diuretics of choice [22]. Aldosterone stimulates renal sodium reabsorption by increasing both the permeability of the luminal membrane of principal cells to sodium and the activity of the Na/K ATPase pump in the basolateral membrane. Since the effect of aldosterone is slow, as it involves interaction with a cytosolic receptor and then a nuclear receptor, the dosage of antialdosteronic drugs should be increased every 7 days. Amiloride, a diuretic acting in the collecting duct, is less effective than aldosterone antagonists and should be used only in those patients who develop severe side effects with aldosterone antagonists [23].

A long-standing debate in the management of ascites is whether aldosterone antagonists should be given alone or in combination with a loop diuretic (i.e., furosemide). Two studies have assessed which is the best approach to therapy, either aldosterone antagonists in a stepwise increase every 7 days (100–400 mg/day in 100 mg/day steps) with furosemide (40–160 mg/day, in 40 mg/day steps) added only in patients not responding to high doses of aldosterone antagonists or combined therapy of aldosterone antagonists and furosemide from the beginning of treatment (100 and 40 mg/day increased in a stepwise manner every 7 days in case of no response up to 400 and 160 mg/day) [[24], [25]]. These studies showed discrepant findings which were likely due to differences in the populations of patients studied, specifically with respect to the percentage of patients with the first episode of ascites included in the two studies [26]. From these studies it can be concluded that a diuretic regime based on the combination of aldosterone antagonists and furosemide is the most adequate for patients with recurrent ascites but not for patients with a first episode of ascites. These latter patients should be treated initially only with an aldosterone antagonist (i.e., spironolactone 100 mg/day) from the start of therapy and increased in a stepwise manner every 7 days up to 400 mg/day in the unlikely case of no response. In all patients, diuretic dosage should be adjusted to achieve a rate of weight loss of no greater than 0.5 kg/day in patients without peripheral edema and 1 kg/day in those with peripheral edema to prevent diuretic-induced renal failure and/or hyponatremia [27]. Following mobilization of ascites, diuretics should be reduced to maintain patients with minimal or no ascites to avoid diuretic-induced complications. Alcohol abstinence is crucial for the control of ascites in patients with alcohol-related cirrhosis.

Complications of diuretic therapy

The use of diuretics may be associated with several complications such as renal failure, hepatic encephalopathy, electrolyte disorders, gynaecomastia, and muscle cramps [[20], [21], [22], [23], [24], [25], [26], [27], [28], [29]]. Diuretic-induced renal failure is most frequently due to intravascular volume depletion that usually occurs as a result of an excessive diuretic therapy [27]. Diuretic therapy has been classically considered a precipitating factor of hepatic encephalopathy, yet the mechanism is unknown. Hypokalemia may occur if patients are treated with loop diuretics alone. Hyperkalemia may develop as a result of treatment with aldosterone antagonists or other potassium-sparing diuretics, particularly in patients with renal impairment. Hyponatremia is another frequent complication of diuretic therapy. The level of hyponatremia at which diuretics should be stopped is contentious. However, most experts agree that diuretics should be stopped temporarily in patients whose serum sodium decreases to less than 120–125 mmol/L. Gynaecomastia is common with the use of aldosterone antagonists, but it does not usually require discontinuation of treatment. Finally, diuretics may cause muscle cramps [[28], [29]]. If cramps are severe, diuretic dose should be decreased or stopped and albumin infusion may relieve symptoms [29].

A significant proportion of patients develop diuretic-induced complications during the first weeks of treatment [24]. Thus, frequent measurements of serum creatinine, sodium, and potassium concentration should be performed during this period. Routine measurement of urine sodium is not necessary, except for non-responders in whom urine sodium provides an assessment of the natriuretic response to diuretics.

Recommendations Patients with the first episode of grade 2 (moderate) ascites should receive an aldosterone antagonist such as spironolactone alone, starting at 100 mg/day and increasing stepwise every 7 days (in 100 mg steps) to a maximum of 400 mg/day if there is no response (Level A1). In patients who do not respond to aldosterone antagonists, as defined by a reduction of body weight of less than 2 kg/week, or in patients who develop hyperkalemia, furosemide should be added at an increasing stepwise dose from 40 mg/day to a maximum of 160 mg/day (in 40 mg steps) (Level A1). Patients should undergo frequent clinical and biochemical monitoring particularly during the first month of treatment (Level A1).

Patients with recurrent ascites should be treated with a combination of an aldosterone antagonist plus furosemide, the dose of which should be increased sequentially according to response, as explained above (Level A1).

The maximum recommended weight loss during diuretic therapy should be 0.5 kg/day in patients without edema and 1 kg/day in patients with edema (Level A1).

The goal of long-term treatment is to maintain patients free of ascites with the minimum dose of diuretics. Thus, once the ascites has largely resolved, the dose of diuretics should be reduced and discontinued later, whenever possible (Level B1).

Caution should be used when starting treatment with diuretics in patients with renal impairment, hyponatremia, or disturbances in serum potassium concentration and patients should be submitted to frequent clinical and biochemical monitoring. There is no good evidence as to what is the level of severity of renal impairment and hyponatremia in which diuretics should not be started. Serum potassium levels should be corrected before commencing diuretic therapy. Diuretics are generally contraindicated in patients with overt hepatic encephalopathy (Level B1).

All diuretics should be discontinued if there is severe hyponatremia (serum sodium concentration <120 mmol/L), progressive renal failure, worsening hepatic encephalopathy, or incapacitating muscle cramps (Level B1).

Furosemide should be stopped if there is severe hypokalemia (<3 mmol/L). Aldosterone antagonists should be stopped if patients develop severe hyperkalemia (serum potassium >6 mmol/L) (Level B1).

Grade 3 or large ascites

Large-volume paracentesis (LVP) is the treatment of choice for the management of patients with grade 3 ascites. The main findings of studies comparing LVP with diuretics in patients with grade 3 ascites are summarized as follows [[30], [31], [32], [33], [34], [35], [36]]: (1) LVP combined with infusion of albumin is more effective than diuretics and significantly shortens the duration of hospital stay. (2) LVP plus albumin is safer than diuretics, the frequency of hyponatremia, renal impairment, and hepatic encephalopathy being lower in patients treated with LVP than in those with diuretics, in the majority of studies. (3) There were no differences between the two approaches with respect to hospital re-admission or survival. (4) LVP is a safe procedure and the risk of local complications, such as hemorrhage or bowel perforation is extremely low [37].

The removal of large volumes of ascitic fluid is associated with circulatory dysfunction characterized by a reduction of effective blood volume, a condition known as post-paracentesis circulatory dysfunction (PPCD) [[31], [36], [38]]. Several lines of evidence indicate that this circulatory dysfunction and/or the mechanisms activated to maintain circulatory homeostasis have detrimental effects in cirrhotic patients. First, circulatory dysfunction is associated with rapid re-accumulation of ascites [35]. Secondly, approximately 20% of these patients develop HRS and/or water retention leading to dilutional hyponatremia [31]. Thirdly, portal pressure increases in patients developing circulatory dysfunction after LVP, probably owing to an increased intrahepatic resistance due to the action of vasoconstrictor systems on the hepatic vascular bed [39]. Finally, the development of circulatory dysfunction is associated with shortened survival [36].

The most effective method to prevent circulatory dysfunction after LVP is the administration of albumin. Albumin is more effective than other plasma expanders (dextran-70, polygeline) for the prevention of PPCD [36]. When less than 5 L of ascites are removed, dextran-70 (8 g/L of ascites removed) or polygeline (150 ml/L of ascites removed) show efficacy similar to that of albumin. However, albumin is more effective than these other plasma expanders when more than 5 L of ascites are removed [36]. Despite this greater efficacy, randomized trials have not shown differences in survival of patients treated with albumin compared with those treated with other plasma expanders [[36], [40], [41]]. Larger trials would be required to demonstrate a benefit of albumin on survival. Although there are no studies on how fast and when albumin should be given to patients treated with LVP, it seems advisable to administer it slowly to avoid a possible cardiac overload due to the existence of a latent cirrhotic cardiomyopathy and at the end of LVP when the volume of ascites removed is known and the increasing cardiac output begins to return to baseline [42].

As far as alternative plasma volume expanders are concerned, it should be noted that polygeline is no longer used in many countries because of the potential risk of transmission of prions. Despite some evidence of the fact that the use of saline is not associated with an increased risk to develop PPCD after small volume paracentesis [40], there are no randomized controlled studies comparing saline versus albumin in patients who require LVP of less than 5 L. Few data exist on the use of starch as a plasma expander in patients with cirrhosis and grade 3 ascites treated with LVP, while there are some concerns about the possibility for starch to induce renal failure [43] and hepatic accumulation of starch [44].

Furthermore, a recent health economic analysis suggested that it is more cost-effective to use albumin after LVP compared with alternative but cheaper plasma volume expanders since the administration of albumin post-paracentesis is associated with a lower number of liver-related complications within the first 30 days [41].

Although LVP is the treatment of choice for large ascites in patients with cirrhosis, it is important to realise that LVP does not address the underlying cause of the condition, namely renal sodium and water retention. Therefore, patients treated with LVP require diuretic treatment after the removal of ascitic fluid to prevent the re-accumulation of ascites [45].

LVP should be performed under strict sterile conditions using disposable sterile materials. It is generally agreed that there are no contraindications to LVP other than loculated ascites, although studies have excluded several subsets of patients. Hemorrhagic complications after LVP are infrequent. In one study, which also included patients with INR >1.5 and platelet count <50,000/μl, only two patients experienced minor cutaneous bleedings out of 142 paracenteses [46]. The frequency of bleeding complications in patients with coagulopathy after LVP are also reported to be low in other studies and do not support a relation between risk of bleeding and the degree of coagulopathy [37]. Thus, there are no data to support the use of fresh frozen plasma or pooled platelets before LVP, yet in many centers these products are given if there is severe coagulopathy (prothrombin activity less than 40%) and/or thrombocytopenia (less than 40,000/μl). Nevertheless, caution should be exercised in patients with severe coagulopathy and LVP should be avoided in the presence of disseminated intravascular coagulation.

Recommendations Large-volume paracentesis (LVP) is the first-line therapy in patients with large ascites (grade 3 ascites) (Level A1). LVP should be completed in a single session (Level A1).

LVP should be performed together with the administration of albumin (8 g/L of ascitic fluid removed) to prevent circulatory dysfunction after LVP (Level A1).

In patients undergoing LVP of greater than 5 L of ascites, the use of plasma expanders other than albumin is not recommended because they are less effective in the prevention of post-paracentesis circulatory dysfunction (Level A1). In patients undergoing LVP of less than 5 L of ascites, the risk of developing post-paracentesis circulatory dysfunction is low. However, it is generally agreed that these patients should still be treated with albumin because of concerns about use of alternative plasma expanders (Level B1).

After LVP, patients should receive the minimum dose of diuretics necessary to prevent the re-accumulation of ascites (Level A1).

Drugs contraindicated in patients with ascites

The administration of non-steroidal anti-inflammatory drugs (NSAIDs), such as indomethacin, ibuprofen, aspirin, and sulindac to patients with cirrhosis and ascites is associated with a high risk of development of acute renal failure, hyponatremia, and diuretic resistance [47]. The impairment in glomerular filtration rate is due to a reduced renal perfusion secondary to inhibition of renal prostaglandin synthesis. Thus, NSAIDs should not be used in patients with cirrhosis and ascites. This represents an important therapeutic limitation for these patients when analgesis are needed. Preliminary data show that short-term administration of selective inhibitors of cyclooxygenase-2 does not impair renal function and the response to diuretics. However, further studies are needed to confirm the safety of these drugs [48].

Angiotensin-converting enzyme inhibitors, even in low doses, should be avoided in patients with cirrhosis and ascites since they can induce arterial hypotension [49] and renal failure [50]. Likewise, α1-adrenergic blockers, such as prazosin, should be used with great caution because despite a reduction in portal pressure, they can further impair renal sodium and water retention and cause an increase in ascites and/or edema [51]. Among cardiovascular drugs, dipyridamole should be used with caution since it can induce renal impairment [52]. Aminoglycosides alone or in combination with ampicillin, cephalothin, or mezlocillin should be avoided in the treatment of bacterial infections, because they are associated with a high incidence of nephrotoxicity [[53], [54]].

Nephrotoxicity induced by the administration of contrast media is a frequent cause of renal failure in the general population of hospitalized patients. However, it has been shown that cirrhosis with ascites and substantially normal renal function does not appear to be a risk factor for the development of contrast media-induced renal failure [55]. Nevertheless, the possibility that contrast media administration can cause a further impairment of renal function in patients with pre-existing renal failure cannot be excluded.

Recommendations Non-steroidal anti-inflammatory drugs (NSAIDs) are contraindicated in patients with ascites because of the high risk of developing further sodium retention, hyponatremia, and renal failure (Level A1).

Drugs that decrease arterial pressure or renal blood flow such as ACE-inhibitors, angiotensin II antagonists, or α1-adrenergic receptor blockers should generally not be used in patients with ascites because of increased risk of renal impairment (Level A1).

The use of aminoglycosides is associated with an increased risk of renal failure. Thus, their use should be reserved for patients with bacterial infections that cannot be treated with other antibiotics (Level A1).

In patients with ascites without renal failure, the use of contrast media does not appear to be associated with an increased risk of renal impairment (Level B1). In patients with renal failure there are insufficient data. Nevertheless, contrast media should be used with caution and the use of general preventive measures of renal impairment is recommended (Level C1).