Ascites and Spontaneous Bacterial Peritonitis

Meet the patient

A 58-year-old man with alcohol-related cirrhosis is on day 3 of an admission for grade 2 ascites when the night SHO flags him as "off". Temperature 37.9, heart rate 104, blood pressure 96 over 60, new confusion, and a creatinine that has climbed from 90 to 140 since yesterday. He has ascites, so he has a tap waiting to be done.[10]

The single thought that must cross your mind on that ward round is spontaneous bacterial peritonitis until proven otherwise — and the only thing that proves or disproves it is a diagnostic paracentesis within hours. Up to a third of patients with SBP have no abdominal symptoms at all, so the diagnosis must be considered in every clinical decompensation of a cirrhotic patient.[9][10]

The peripheral-arterial-vasodilation cascade — why cirrhosis makes fluid

The body reads underfilling that is not there. That one sentence is the mechanism of cirrhotic ascites, and every drug in the ladder attacks a limb of it. The peripheral arterial vasodilation cascade runs portal pressure into splanchnic vasodilation, vasodilation into a perceived fall in effective arterial blood volume, and that false signal into RAAS, sympathetic and ADH-driven sodium and water retention — the fluid that becomes ascites.[20]

Step 1 — portal hypertension. Architectural distortion raises sinusoidal hydrostatic pressure and drives fluid across the sinusoid into the space of Disse and the peritoneal cavity. Portal pressure is the engine; relieving it (TIPS) or replacing the liver (transplant) are the only definitive moves.[16]

Step 2 — splanchnic vasodilation, the pivotal step. High portal pressure triggers sustained splanchnic vasodilation, so the effective arterial blood volume the baroreceptors sense falls — even though the patient is volume-overloaded. The patient is drowning and reading thirst; that is the whole insight.[20]

Step 3 — neurohormonal activation and avid sodium retention. The perceived underfilling activates endogenous vasoactive systems: aldosterone reclaims distal sodium, sympathetic drive reclaims proximal sodium and fires renin, and non-osmotic vasopressin retains free water — the basis of dilutional hyponatraemia. In its extreme form this reduction of effective circulating volume, driven by splanchnic arterial vasodilation together with a falling cardiac output, is what precipitates type-1 hepatorenal syndrome.[20]

Step 4 — hypoalbuminaemia, the permissive factor. The failing liver makes less albumin, oncotic pressure falls, and the retained fluid leaks into the peritoneal space rather than staying intravascular. The cycle is self-perpetuating — more vasodilation, more RAAS, more sodium, more ascites — until the cause is treated or the kidney fails. This is exactly why spironolactone (aldosterone blockade) is the cornerstone diuretic, and why only TIPS (relieving portal pressure) or transplant (replacing the liver) are definitive.[4]

SBP is the infective complication of this same gut. Bacterial translocation carries enteric organisms — gram-negative rods from the colon above all — across the oedematous, hyperpermeable cirrhotic gut wall to mesenteric nodes and into ascitic fluid that is too protein-poor to opsonise them: low ascitic protein, lack of serum complement and gastrointestinal bleeding all predispose to SBP.[9][10]

The classic trap: a polymicrobial ascitic culture — multiple organisms including anaerobes — is not SBP. It is secondary peritonitis from a perforated viscus, and it needs surgical review, broad anaerobic cover and a CT, not a 5-day cephalosporin course. The course of the ascitic PMN under therapy and persistently positive cultures help discriminate the two.[9]

SAAG — the one test that does the work

Serum albumin minus ascitic albumin — that is the SAAG, and the 1.1 g per dL threshold splits portal from peritoneal. It classifies by mechanism rather than by disease, which is why one number does the work of a long differential. In a prospective series of 901 paired samples the albumin gradient correctly differentiated causes of ascites due to portal hypertension from those that were not, 96.7 percent of the time — far better than the old exudate-transudate total-protein concept it replaced (55.6 percent). At 1.1 g per dL or more the problem is portal hypertension; under 1.1 g per dL it is a non-portal cause such as malignancy, tuberculous peritonitis or nephrotic syndrome.[5][6]

[5] [6]

Then add total protein as the adjunct. The old exudate-transudate concept should be discarded as a classifier, but ascitic fluid total protein remains a useful adjunct in the differential: the total protein of most cardiac ascites samples is high (traditionally expected to be a transudate) while that of most spontaneously infected cirrhotic samples is low (traditionally expected to be an exudate). Everyone forgets the total protein and misses the heart.[5]

Etymology for viva gold: it is a gradient (subtract), not a ratio (divide) — the SAAG replaced the old exudate-transudate concept because the gradient tracks portal hypertension far more faithfully than fluid total protein ever did. Subtract, do not divide.[5]

[5] [6]

PMN 250 — treat now

Ascitic neutrophils 250 cells per mm3 or more is SBP, and it is treated within hours — not after the culture, not after the meeting. The cell count is the one test you hand-carry to the laboratory; it triggers or withholds the antibiotic bundle. Culture takes days, detects the organism in fewer than half of cases, and treatment should be initiated without delay — a negative culture with neutrocytic ascites (culture-negative neutrocytic ascites) is still SBP.[7][10]

[7] [9] [8]

Two practical points examiners test. First, inoculate a large volume (10 mL in toto) of ascitic fluid into blood-culture bottles at the bedside — positivity rises from 42 percent with the old technique to 91 percent. Second, treatment starts on the cell count, not the culture: bacteriological examination detects causative agents in fewer than half of cases and the result takes days, so empirical therapy after the tap is the standard. The threshold of 250 PMN per mm3 — irrespective of the culture result — is the line every candidate must reproduce.[8][7][10]

The tap itself. Diagnostic paracentesis goes in the left lower quadrant, two finger-breadths cephalad and two medial to the anterior superior iliac spine — it avoids the inferior epigastric vessels, and the left side avoids the caecum. Coagulopathy is not a contraindication to a fine-needle diagnostic tap, and prophylactic plasma or platelets are not routinely given. Use ultrasound guidance when fluid is loculated or the patient is obese.[2]

What is the diagnostic threshold for SBP?

Ascitic fluid PMN count of 250 cells/mm3 or more, with no evident intra-abdominal surgical source — universally accepted as the best surrogate marker for the diagnosis, irrespective of the ascitic culture result.

[7]

What is culture-negative neutrocytic ascites (CNNA)?

Raised PMN (250 cells/mm3 or more) with a negative culture in a patient not on antibiotics. It is a variant of SBP — same organisms, same treatment — and reflects the fastidiousness of the organisms or the limits of culture, not a different disease. Bacteriological examination of ascitic fluid detects causative agents in fewer than half of cases.

[9] [10]

What distinguishes SBP from secondary (surgical) peritonitis?

Secondary peritonitis is polymicrobial (multiple organisms including anaerobes), with chemical markers of severity: glucose under 50 mg per dL, LDH and lactate above serum levels — implying a perforated viscus or abscess that needs surgical review and broad anaerobic cover. SBP is monomicrobial; the course of the ascitic PMN under therapy and the time cultures stay positive discriminate the two.

[9]

Classification — by severity and by response to diuretics

Ascites is graded three ways: by SAAG (above), by severity, and by response to diuretics — and only the third predicts whether you escalate to TIPS or transplant.[4]

Refractory ascites is the turning point. Defined as ascites that cannot be mobilised by, or recurs early after, paracentesis despite sodium restriction and diuretics, it carries 50 percent mortality within 6 months of its development and is an absolute indication to consider TIPS or transplant — liver transplantation being the only effective therapy for refractory ascites associated with cirrhosis.[4]

Two scores grade the cirrhosis itself and drive transplant listing. Child-Pugh (bilirubin, albumin, INR, encephalopathy, ascites; A 5 to 6, B 7 to 9, C 10 to 15) and MELD-Na, which folds serum sodium into MELD because dilutional hyponatraemia and ascites both mark advanced disease — and which now governs allocation.[1]

[5]

The differential table — one discriminator each

The SAAG plus a handful of confirmatory tests closes the differential. High SAAG with high protein points to the heart or the hepatic veins; low SAAG with positive cytology is carcinomatosis, with high ADA is tuberculosis, with high amylase is pancreatic, with high triglycerides is chylous. Cytology, amylase, triglycerides, ADA and culture are the tests that finish the job the SAAG starts.[5][6]

Hepatic hydrothorax deserves its own line: ascitic fluid tracking through small diaphragmatic defects into a (usually right-sided) pleural effusion, in a cirrhotic with minimal or no detectable ascites. It is not cardiac or pulmonary — it requires the same treatment as ascites (salt restriction, diuretics, paracentesis), and refractory cases are the hard problem.[25]

How common, who, and the numbers that decide disposition

Ascites is the commonest presentation of decompensated cirrhosis, and its appearance is the decompensation landmark. In a cohort of 293 patients with compensated cirrhosis the ten-year probability of decompensation was 58 percent, and the development of ascites heralds a poor prognosis with a 50 percent 2-year survival rate. SBP complicates cirrhosis with ascites in up to 25 percent, recurs in 68 percent of untreated survivors at one year, carries hospital mortality of 10 to 20 percent, and 1-year mortality after an episode approaches 70 percent.[24][4][9][13][10]

[24] [4] [13]

Who gets it: the cirrhotic with clinically significant portal hypertension. In India, tuberculosis sits higher in the differential of low-SAAG ascites than in the West, and the underlying aetiology mix (hepatitis B, alcohol, MASLD) drives the aetiological workup the examiner expects you to name.[6]

SBP risk tracks how poorly the fluid kills bacteria. Gastrointestinal bleeding, lack of serum complement and a low ascitic protein all predispose to SBP. The combination the Fernández trial used for primary norfloxacin prophylaxis is ascitic protein under 15 g per L (1.5 g per dL) plus advanced liver failure (Child-Pugh at least 9 with bilirubin at least 3 mg per dL) or impaired renal function (creatinine at least 1.2 mg per dL, BUN at least 25 mg per dL, or sodium 130 mEq per L or less) — in that group norfloxacin cut the 1-year SBP rate from 61 to 7 percent, delayed hepatorenal syndrome, and improved survival.[9][14]

The combination diuretic principle — the pharmacological cornerstone

A loop diuretic (furosemide) plus a distal aldosterone antagonist (spironolactone), titrated together, is the pharmacological cornerstone of cirrhotic ascites. The combination is not arbitrary: in the randomised comparison of combined versus sequential therapy, sequential distal-diuretic monotherapy caused significantly more hyperkalaemia (18 versus 4 percent) and more adverse effects overall (38 versus 20 percent), while more patients on the combined regimen resolved their ascites without changing step (76 versus 56 percent). Titrate the pair together rather than pushing one drug to its maximum.[4][21]

[4] [21]

Dietary sodium stays at 2 g per day; fluid restriction is reserved for dilutional hyponatraemia. Monitor weight, sodium, potassium and creatinine while titrating, and interrupt diuretics when encephalopathy, worsening hyponatraemia or a rising creatinine appears. The classic trap — over-diuresis: the fastest way to precipitate prerenal AKI, hepatic encephalopathy and hyponatraemia is to chase a dry abdomen too quickly.[4][21]

Large-volume paracentesis over 5 L — give albumin, then escalate

[4]

Large-volume paracentesis relieves tense ascites immediately, and when 5 litres or more is removed, albumin 6 to 8 g per litre of fluid removed should be given intravenously to minimise haemodynamic and renal dysfunction. The randomised comparison of plasma expanders showed why: post-paracentesis circulatory dysfunction occurred in 18.5 percent given albumin versus 34.4 percent with dextran 70 and 37.8 percent with polygeline — and it is not spontaneously reversible, shortening time to readmission and survival. Albumin is the best plasma expander for preventing it.[4][15]

When ascites becomes refractory, the ladder is serial LVP with albumin, TIPS in selected patients, and transplant as definitive therapy. TIPS drops portal pressure directly — the only non-transplant treatment that addresses mechanism rather than consequence. It is most effective when liver function is relatively preserved and may hasten death in advanced liver failure.[4][16]

The Salerno meta-analysis of individual patient data (305 patients, 4 randomised trials) confirmed that TIPS controls refractory ascites better than serial paracentesis — tense ascites recurred in 42 percent with TIPS versus 89 percent with paracentesis — with significantly better transplant-free survival, at the cost of more hepatic encephalopathy episodes per patient (1.13 versus 0.63). TIPS is not a substitute for transplant evaluation.[16]

[4] [5] [21] [13]

SBP bundle — the time-critical ladder

PMN 250 or more: cefotaxime 2 g IV every 8 hours for 5 days, plus albumin 1.5 g per kg on day 1 and 1 g per kg on day 3, then norfloxacin 400 mg once daily for secondary prophylaxis and a transplant referral. That is the whole bundle in one line, and every dose is examinable.[11][3][13]

[11] [12] [3] [13] [9]

Why albumin in SBP? The infection worsens the hyperdynamic circulation, renal function frequently becomes impaired, and that impairment is probably related to a reduction in effective arterial blood volume — with a high mortality rate. The Sort trial (NEJM 1999) proved that albumin 1.5 g per kg at diagnosis and 1 g per kg on day 3 on top of cefotaxime cuts renal impairment (33 to 10 percent, P = 0.002), in-hospital mortality (29 to 10 percent, P = 0.01) and 3-month mortality (41 to 22 percent, P = 0.03). This is the albumin dose candidates forget.[3][4]

HRS-AKI — recognise it early, then terlipressin plus albumin

A rising creatinine in decompensated cirrhosis with ascites is hepatorenal syndrome-AKI until you prove otherwise. HRS-AKI is the specific form of acute kidney injury in patients with advanced cirrhosis and ascites; it carries an especially high mortality risk, early recognition is crucial, and the first moves are to exclude and treat other causes of AKI — hypovolaemia, obstruction, parenchymal disease, nephrotoxins — while stopping diuretics and nephrotoxic drugs.[19][20]

The kidney is structurally normal but vasoconstricted. HRS-AKI is a functional, potentially reversible renal failure: type 1 develops as a consequence of a severe reduction of effective circulating volume due to both an extreme splanchnic arterial vasodilatation and a reduction of cardiac output; type 2 is stable or slowly progressive, its main clinical consequence being refractory ascites rather than acute renal failure, with a less negative impact on prognosis. The 2024 ADQI-ICA consensus (Nadim) anchors the modern definition.[20][19]

[20]

Treatment is a splanchnic vasoconstrictor plus albumin. Terlipressin is given either as an intravenous infusion starting at 2 mg per day, or as intravenous boluses of 0.5 mg every 4 hours, titrated up to a maximum of 12 mg per day — the infusion is better tolerated and effective at lower doses — always with albumin. Where terlipressin is not available, midodrine plus octreotide plus albumin is the alternative, though terlipressin plus albumin is significantly more effective at improving renal function. The CONFIRM trial (Wong, NEJM 2021) confirmed terlipressin plus albumin achieves verified HRS reversal in 32 versus 17 percent with placebo (P = 0.006) — at the cost of death from respiratory disorders within 90 days in 11 versus 2 percent, so avoid it in fluid overload. Liver transplantation remains the only curative option.[22][23][18][19]

Hepatic hydrothorax and the chest-tube trap

A right-sided pleural effusion in a cirrhotic is hepatic hydrothorax — fluid tracking through diaphragmatic defects — and it requires the same treatment as ascites: salt restriction, diuretics and paracentesis. Refractory hydrothorax is the hard problem, and its management is not well established. Everyone forgets: manage it as ascites, not as a primary pleural disease.[25]

The classic traps — the preventable-harm list

Four traps recur in exams and in morbidity meetings. Name them and you will not fall into them.[4]

• Over-diuresis — chasing a dry abdomen precipitates prerenal AKI, encephalopathy and hyponatraemia; titrate the combined regimen and monitor electrolytes and renal function.[4]
• Hepatic hydrothorax managed as a primary chest problem — it requires the same treatment as ascites; escalate to TIPS for refractory cases.[25]
• Polymicrobial ascitic culture equals secondary peritonitis — look for a perforated viscus, add anaerobic cover, call the surgeon; this is not SBP.[9]
• NSAIDs, ACE inhibitors, ARBs and aminoglycosides in decompensation — they blunt renal prostaglandins, drop renal perfusion, and tip the patient into AKI; stop them on admission.[1]

The trials that built the bundle

Five trials anchor every recommendation on this page. Name the trial, then the number it changed.[3]

[3] [14] [16] [18] [17]

The ATTIRE trial (UK multicentre, NEJM 2021) is the cautionary tale on long-course albumin. Targeted 20 percent infusions to keep serum albumin at 30 g per L or more in hospitalised decompensated cirrhosis did not reduce the composite of infection, renal dysfunction or death versus standard care (29.7 versus 30.2 percent) — so albumin is a drug with specific evidence-based indications, not a routine tonic.[17]

[7] [3] [4] [4] [23] [7] [11] [3] [4] [23] [7] [12] [6]

The mnemonic and the high-yield one-liners

[7] [3] [11] [12] [13]

High-yield one-liners, in the order the viva asks them:[4]

• Ascites is the most common presentation of decompensated cirrhosis; 50 percent 2-year survival once it appears.[4][24]
• SAAG 1.1 or more equals portal hypertension; under 1.1 equals non-portal — malignancy, TB, nephrotic. The gradient discarded the exudate-transudate concept.[5][6]
• SBP equals PMN 250 or more equals cefotaxime 2 g IV every 8 hours for 5 days plus albumin 1.5 g per kg day 1 and 1 g per kg day 3.[7][11][3]
• When 5 L or more is removed at paracentesis, give albumin 6 to 8 g per litre of fluid removed.[4]
• Diuretics: combine furosemide with spironolactone and titrate together — combined beats sequential, with less hyperkalaemia.[21][4]
• Refractory ascites means 50 percent mortality within 6 months — serial LVP, TIPS in selected, transplant referral.[4]
• HRS-AKI means terlipressin plus albumin (CONFIRM: 32 versus 17 percent verified reversal), or midodrine plus octreotide where terlipressin is unavailable.[18][23]
• Hepatic hydrothorax takes the same treatment as ascites; a polymicrobial culture means secondary peritonitis.[25][9]

The mantra

The mantra, in one line: SAAG tells you why the fluid is there; PMN tells you whether to treat now.[5][7]

Ward-round test — three stems, thirty seconds each

Stem 1 — the man from the top of the topic (answer)

The 58-year-old cirrhotic from Meet the patient: day 3, T 37.9, HR 104, BP 96 over 60, new confusion, creatinine 90 to 140. The ascitic PMN comes back at 320. What is the bundle in the next hour? Model: This is SBP. Within the hour: cefotaxime 2 g IV every 8 hours for 5 days, albumin 1.5 g per kg on day 1 and 1 g per kg on day 3 (Sort), stop any NSAID or diuretic driving the creatinine, follow the PMN course on therapy, and plan norfloxacin 400 mg once daily secondary prophylaxis with a transplant referral — 1-year mortality after an episode approaches 70 percent.[11][3][13][4]

Stem 2 — the polymicrobial tap (answer)

A cirrhotic with tense ascites has a diagnostic tap that grows Escherichia coli, Bacteroides and Enterococcus, with PMN 600, glucose 35 mg per dL and a markedly raised LDH. SBP, or not? What changes? Model: Not SBP — secondary peritonitis. Polymicrobial growth plus glucose under 50 mg per dL plus LDH and lactate above serum levels means a perforated viscus. Add anaerobic cover (metronidazole), broaden for enterococci, order a CT abdomen, and call the surgeon. Do not discharge on a 5-day cephalosporin course.[9]

Stem 3 — the refractory abdomen and the climbing creatinine (answer)

A 49-year-old with diuretic-resistant ascites needs 8 litres drained every two weeks. Today the creatinine is 1.6, up from 1.1. What do you do, and what must you not give? Model: Draining 8 L needs albumin 6 to 8 g per litre of fluid removed to prevent post-paracentesis circulatory dysfunction. The rising creatinine raises HRS-AKI: exclude other causes of AKI, stop diuretics and nephrotoxins, and start terlipressin plus albumin (CONFIRM: verified reversal 32 versus 17 percent) — but do not give terlipressin if there is fluid overload or respiratory compromise (deaths from respiratory disorders 11 versus 2 percent at 90 days). This patient also needs transplant referral: refractory ascites carries 50 percent mortality within 6 months.[4][18][19]

References

1. [1]European Association for the Study of the Liver EASL Clinical Practice Guidelines for the management of patients with decompensated cirrhosis J Hepatol, 2018.PMID 29653741
2. [2]Runyon BA Introduction to the revised American Association for the Study of Liver Diseases Practice Guideline management of adult patients with ascites due to cirrhosis 2012 Hepatology, 2013.PMID 23463403
3. [3]Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis N Engl J Med, 1999.PMID 10432325
4. [4]Garcia N, Sanyal AJ Ascites Curr Treat Options Gastroenterol, 2001.PMID 11696279
5. [5]Runyon BA, et al. The serum-ascites albumin gradient is superior to the exudate-transudate concept in the differential diagnosis of ascites Ann Intern Med, 1992.PMID 1616215
6. [6]Khandwalla HE, et al. The utility of evaluating low serum albumin gradient ascites in patients with cirrhosis Am J Gastroenterol, 2009.PMID 19491852
7. [7]Koulaouzidis A, et al. Diagnosis of spontaneous bacterial peritonitis: an update on leucocyte esterase reagent strips World J Gastroenterol, 2011.PMID 21448413
8. [8]Runyon BA, et al. Inoculation of blood culture bottles with ascitic fluid. Improved detection of spontaneous bacterial peritonitis Arch Intern Med, 1987.PMID 3541825
9. [9]Press AG, et al. Spontaneous bacterial peritonitis Z Gastroenterol, 1992.PMID 1413938
10. [10]Šenkeříková R, et al. Spontaneous bacterial peritonitis Vnitr Lek, 2015.PMID 25813257
11. [11]Runyon BA, et al. Short-course versus long-course antibiotic treatment of spontaneous bacterial peritonitis. A randomized controlled study of 100 patients Gastroenterology, 1991.PMID 2019378
12. [12]Rimola A, et al. Two different dosages of cefotaxime in the treatment of spontaneous bacterial peritonitis in cirrhosis: results of a prospective, randomized, multicenter study Hepatology, 1995.PMID 7875666
13. [13]Ginés P, et al. Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis: results of a double-blind, placebo-controlled trial Hepatology, 1990.PMID 2210673
14. [14]Fernández J, Navasa M, Planas R, et al. Primary prophylaxis of spontaneous bacterial peritonitis delays hepatorenal syndrome and improves survival in cirrhosis Gastroenterology, 2007.PMID 17854593
15. [15]Ginés A, et al. Randomized trial comparing albumin, dextran 70, and polygeline in cirrhotic patients with ascites treated by paracentesis Gastroenterology, 1996.PMID 8831595
16. [16]Salerno F, Cammà C, Enea M, et al. Transjugular intrahepatic portosystemic shunt for refractory ascites: a meta-analysis of individual patient data Gastroenterology, 2007.PMID 17678653
17. [17]China L, Freemantle N, Forrest E, et al. A Randomized Trial of Albumin Infusions in Hospitalized Patients with Cirrhosis N Engl J Med, 2021.PMID 33657293
18. [18]Wong F, Pappas SC, Curry MP, et al. Terlipressin plus Albumin for the Treatment of Type 1 Hepatorenal Syndrome N Engl J Med, 2021.PMID 33657294
19. [19]Nadim MK, Kellum JA, Forni L, et al. Acute kidney injury in patients with cirrhosis: Acute Disease Quality Initiative (ADQI) and International Club of Ascites (ICA) joint multidisciplinary consensus meeting J Hepatol, 2024.PMID 38527522
20. [20]Angeli P, et al. Hepatorenal syndrome Contrib Nephrol, 2011.PMID 21921608
21. [21]Angeli P, et al. Combined versus sequential diuretic treatment of ascites in non-azotaemic patients with cirrhosis: results of an open randomised clinical trial Gut, 2010.PMID 19570764
22. [22]Cavallin M, et al. Terlipressin given by continuous intravenous infusion versus intravenous boluses in the treatment of hepatorenal syndrome: a randomized controlled study Hepatology, 2016.PMID 26659927
23. [23]Cavallin M, et al. Terlipressin plus albumin versus midodrine and octreotide plus albumin in the treatment of hepatorenal syndrome: a randomized trial Hepatology, 2015.PMID 25644760
24. [24]Ginés P, et al. Compensated cirrhosis: natural history and prognostic factors Hepatology, 1987.PMID 3804191
25. [25]Xiol X, et al. Hepatic hydrothorax Curr Opin Pulm Med, 1998.PMID 10813241