SBP Prophylaxis in Cirrhosis: Who Needs Primary vs Secondary Prevention — and When Should It Be Stopped?

Long-term antibiotics can prevent spontaneous bacterial peritonitis in selected patients with cirrhosis, but deciding whom to treat — and when prophylaxis can safely be withdrawn — requires balancing infection risk against antimicrobial resistance and drug toxicity.

Clinical bottom line

Do not give primary SBP prophylaxis simply because ascites is present. The strongest conventional indication is low ascitic fluid protein (<1.5 g/dL) plus advanced liver failure, renal dysfunction or significant hyponatraemia.

After a documented episode of SBP, secondary prophylaxis remains recommended by major guidelines because recurrence risk is high; the landmark trial reduced recurrence from 68% to 20% with norfloxacin.

Common long-term regimens include norfloxacin 400 mg once daily, ciprofloxacin 500 mg once daily, or trimethoprim–sulfamethoxazole (TMP–SMX) one double-strength tablet daily, although availability, resistance patterns, renal function and adverse effects should guide selection.

Cirrhosis with acute gastrointestinal bleeding is a separate indication for short-course antibiotic prophylaxis, generally ceftriaxone 1 g IV every 24 hours for up to 7 days; it should not automatically become indefinite SBP prophylaxis once the bleeding episode has resolved.

Stopping primary prophylaxis is reasonable when ascites has genuinely disappeared following sustained clinical improvement. EASL explicitly recommends this, and NICE 2023 advises prophylaxis only until ascites resolves.

When secondary prophylaxis can be stopped is much less certain. Traditional AASLD practice is to continue until liver transplantation or death, whereas EASL acknowledges that evidence is insufficient to know whether treatment can be withdrawn following substantial improvement in liver disease.

Which patients actually need SBP prophylaxis?

There are three clinically distinct situations, and they should not be conflated:

The distinction matters because the evidence, expected benefit and appropriate treatment duration are different in each group.

Who should receive primary prophylaxis?

Primary prophylaxis should be reserved for patients at sufficiently high risk of developing their first episode of SBP rather than prescribed routinely to everyone with cirrhotic ascites.

The conventional AASLD/EASL high-risk profile is:

Ascitic fluid protein <1.5 g/dL

PLUS either:

Advanced liver failure

• Child–Pugh score ≥9, and
• serum bilirubin ≥3 mg/dL (≈51 μmol/L)

OR

Renal/circulatory dysfunction

At least one of:

• serum creatinine ≥1.2 mg/dL (≈106 μmol/L)
• blood urea nitrogen ≥25 mg/dL
• serum sodium ≤130 mmol/L

This risk-enrichment strategy derives largely from the Fernández trial, in which 68 high-risk patients were randomised to norfloxacin or placebo. One-year SBP probability was 7% versus 61%, respectively, although the small sample and analytical limitations make the magnitude of benefit uncertain.

Is low ascitic protein alone enough?

Probably not for many patients.

This is an important area of disagreement.

EASL and AASLD emphasise low ascitic protein together with markers of advanced hepatic or renal dysfunction. The 2021 BSG/BASL guideline was somewhat broader, suggesting prophylaxis could be offered when ascitic protein is <1.5 g/dL, while explicitly acknowledging uncertainty and the potential harms of long-term antibiotics.

NICE took a still more individualised position in its 2023 update: antibiotics should not routinely be given to everybody with cirrhosis and ascites. They may be considered in patients at high risk, for example because of severe liver disease, or where an infection would substantially disrupt planned transplantation or TIPS.

For practical purposes, an otherwise stable patient with low-protein ascites but preserved renal function and relatively preserved hepatic function should therefore not automatically receive lifelong antibiotics solely because the ascitic protein is below 1.5 g/dL.

Who needs secondary prophylaxis?

A patient who has recovered from a documented episode of SBP should generally be offered secondary prophylaxis.

Recurrence risk is historically estimated at approximately 70% within one year without effective preventive therapy. The pivotal placebo-controlled study enrolled 80 patients following SBP and found recurrent SBP in approximately 20% with norfloxacin versus 68% with placebo.

This remains the basis for recommendations from AASLD, EASL and BSG/BASL despite the age and relatively small size of the trial.

A previous episode of SBP should also trigger consideration of liver transplantation evaluation, because SBP is a marker of advanced decompensation and poor prognosis. EASL specifically recommends considering transplantation in patients who recover from SBP.

What should be prescribed?

Choice should be guided by local resistance patterns, previous cultures, drug availability, renal function, allergy history and toxicity rather than treating all agents as interchangeable.

AASLD lists norfloxacin, ciprofloxacin and TMP–SMX as options for long-term primary or secondary prophylaxis.

Daily dosing is generally preferable to intermittent fluoroquinolone regimens because intermittent exposure has been associated with greater selection of quinolone-resistant organisms.

What about rifaximin?

Rifaximin is attractive because of its predominantly intestinal action and potentially different resistance profile, and some trials suggest efficacy.

However, EASL concluded that available evidence was insufficient to recommend rifaximin routinely instead of norfloxacin for secondary prophylaxis. More recent trials have continued to generate interest, and the 2026 Korean Association for the Study of the Liver (KASL) guideline allows rifaximin or norfloxacin to be considered in high-risk primary prophylaxis and following recovery from SBP.

Thus, rifaximin should not yet be interpreted as an internationally established replacement for conventional prophylaxis.

What about patients presenting with gastrointestinal bleeding?

Patients with cirrhosis and acute upper gastrointestinal bleeding need short-term antibiotic prophylaxis regardless of whether they meet the usual long-term primary-prophylaxis criteria.

A commonly recommended regimen is:

ceftriaxone 1 g IV every 24 hours for up to 7 days.

This reduces bacterial infections associated with variceal and other upper GI bleeding and is particularly appropriate in advanced cirrhosis or settings with substantial quinolone resistance.

Importantly:

GI-bleed prophylaxis is not the same as indefinite primary SBP prophylaxis.

Once the acute bleeding period has ended, reassess the patient. Long-term prophylaxis should continue only if a separate indication exists — previous SBP or sufficiently high-risk ascites.

When can primary prophylaxis be stopped?

The clearest reason to stop primary prophylaxis is durable resolution of ascites together with clinical improvement.

EASL explicitly states that norfloxacin prophylaxis should be stopped in patients with long-lasting improvement in their clinical condition and disappearance of ascites.

NICE similarly recommends continuing prophylaxis until the ascites has resolved.

This is biologically and clinically sensible: SBP requires susceptible ascitic fluid. If a patient recompensates after treatment of the underlying liver disease and no longer has ascites, the original target condition is no longer present.

Therefore, primary prophylaxis should not automatically become a permanent entry on a medication chart.

Reassess long-term primary prophylaxis when:

• ascites has disappeared;
• hepatic function has substantially improved;
• previous renal dysfunction or severe hyponatraemia has resolved;
• liver transplantation has occurred;
• toxicity develops;
• resistant infection or recurrent C. difficile changes the risk–benefit balance;
• the original indication is no longer clear.

The exact duration of “sustained” improvement is not defined by high-quality evidence, so this remains a clinical judgement rather than a validated time threshold.

Can secondary prophylaxis ever be stopped?

This is the most difficult question — and current evidence does not provide a definitive answer.

Traditional AASLD recommendations continue secondary prophylaxis until liver transplantation or death.

EASL, however, explicitly acknowledges uncertainty over whether prophylaxis must continue uninterrupted until transplantation or death in every patient who has previously had SBP, or whether it could be discontinued when liver disease markedly improves.

This distinction matters increasingly as some patients achieve meaningful recompensation following:

• sustained alcohol abstinence;
• cure or suppression of viral hepatitis;
• substantial improvement following treatment of the underlying liver disease;
• selected portal-hypertension interventions.

A practical interpretation

If clinically significant ascites persists, secondary prophylaxis should generally be continued unless toxicity, resistant infection or another compelling contraindication develops.

If the patient has durably recompensated and ascites has completely disappeared, the continued absolute benefit of lifelong prophylaxis becomes uncertain. In this situation, withdrawal can reasonably be discussed with hepatology/infectious-disease input, recognising that this is not supported by definitive randomised stopping data.

This should therefore be described as an individualised clinical decision — not as an established guideline stopping rule.

Why has long-term antibiotic prophylaxis become controversial?

Because the evidence supporting a practice that may continue for years is considerably weaker than its widespread use might suggest.

A 2020 Cochrane network meta-analysis included 29 randomised trials involving 3,896 participants and concluded that evidence for the benefits and comparative effectiveness of antibiotic prophylaxis was very low certainty. Effects on mortality in particular remained uncertain.

Meanwhile, the microbial environment of contemporary cirrhosis care is very different from that of the landmark norfloxacin trials.

Long-term prophylaxis can select for:

• fluoroquinolone-resistant Gram-negative organisms;
• Gram-positive infections;
• multidrug-resistant organisms;
• Clostridioides difficile;
• organisms against which traditional selective intestinal decontamination is less effective.

Fluoroquinolones also carry important non-infectious toxicities, while TMP–SMX can cause hyperkalaemia, renal dysfunction, cytopenias and hypersensitivity.

A large 2025 observational analysis of two US national cohorts found that secondary prophylaxis was associated with a higher rate of non-SBP infections. In the Veterans Affairs cohort, the adjusted odds ratio for any non-SBP infection was 1.26 (95% CI 1.10–1.44); a similar association occurred in the independent TriNetX cohort. Because this was observational, confounding by indication cannot be excluded, and the study should not by itself overturn guideline recommendations. It does, however, strengthen the argument for periodically reassessing rather than automatically continuing antibiotics indefinitely.

How should prophylaxis be reviewed at the bedside?

Patients receiving long-term prophylaxis should not simply have the prescription automatically renewed.

At each meaningful clinical review, ask:

Does the indication still exist?

Then review:

• current ascites status;
• previous SBP history;
• hepatic function and trajectory;
• renal function;
• serum sodium;
• previous microbiology and antimicrobial resistance;
• recent breakthrough infections;
• previous C. difficile;
• drug adverse effects;
• transplantation status or candidacy;
• other antibiotics and potential interactions.

A patient developing SBP while adherent to prophylaxis should not simply be restarted indefinitely on the same agent without review. Culture history, adherence, local antimicrobial susceptibility and the possibility of prophylaxis-selected resistant organisms should be assessed.

Practical decision algorithm

Practice pitfalls

Areas of uncertainty

How much benefit does modern primary prophylaxis actually provide?

The striking reductions in SBP reported in selected older trials have not been consistently reproduced, and meta-analytic certainty is low. The baseline microbiology and standard care of cirrhosis have also changed substantially.

Should secondary prophylaxis still be universal?

Major guidelines continue to recommend it, but the pivotal placebo-controlled trial is more than three decades old. Recent observational evidence raises concerns about resistance and other infections but cannot establish that withholding prophylaxis is safer.

When can secondary prophylaxis be stopped after recompensation?

No adequately powered randomised trial has established a stopping rule. EASL explicitly identifies this as uncertain.

Will rifaximin replace fluoroquinolones?

Possibly in selected populations, but comparative evidence remains insufficient for a universal recommendation. Notably, KASL's 2026 guideline is more permissive regarding rifaximin than older European and North American recommendations.

What to do on the next ward round or clinic

Identify why every patient receiving chronic SBP prophylaxis is taking it: primary prevention, previous SBP, or merely continuation after a GI bleed.

For primary prophylaxis, confirm that the patient actually remains high risk rather than relying on low ascitic protein alone.

In previous SBP with persistent ascites, continue secondary prophylaxis unless harms or microbiological considerations change the balance.

If ascites has disappeared after sustained clinical improvement, actively reconsider primary prophylaxis rather than automatically renewing it.

Review previous cultures, resistant organisms, renal function, potassium, adverse effects and C. difficile history whenever prolonged antibiotic therapy is continued.

Evidence review and evidence map

Clinical scope

This review addresses adult patients with cirrhosis, principally those with portal-hypertensive ascites, in inpatient and specialist outpatient settings.

The clinical decisions addressed are:

• who should receive primary SBP prophylaxis;
• who requires secondary prophylaxis following SBP;
• which regimens are supported by major guidelines;
• how GI-bleed prophylaxis differs from chronic prophylaxis;
• when primary prophylaxis should be stopped;
• whether secondary prophylaxis can be withdrawn;
• how antimicrobial resistance and newer evidence affect decision-making.

Treatment of active SBP, albumin therapy during active SBP, management of non-cirrhotic ascites and paediatric cirrhosis are outside the principal scope.

Evidence search cutoff: 15 September 2026.

Guideline comparison

Major supporting evidence

Claim-level evidence map

Evidence synthesis

The most consistent international recommendations concern secondary prophylaxis after SBP and short-course antibiotic prophylaxis during acute GI bleeding.

Primary prophylaxis is more controversial. The most defensible approach is to restrict it to patients with low-protein ascites plus evidence of advanced hepatic or renal dysfunction rather than treating low ascitic protein as an automatic indication.

The greatest uncertainty concerns duration. Primary prophylaxis can reasonably be withdrawn when ascites disappears following sustained improvement. Evidence supporting withdrawal of secondary prophylaxis after recompensation is inadequate.

Recent evidence has not yet overturned guideline recommendations, but it has strengthened concerns surrounding lifelong antibiotic exposure, antimicrobial resistance and non-SBP infections. This makes regular reassessment of the indication increasingly important.

References

Biggins SW, Angeli P, Garcia-Tsao G, et al. Diagnosis, Evaluation, and Management of Ascites, Spontaneous Bacterial Peritonitis and Hepatorenal Syndrome. Hepatology. 2021. AASLD Practice Guidance.

European Association for the Study of the Liver. EASL Clinical Practice Guidelines for the management of patients with decompensated cirrhosis. J Hepatol. 2018. doi:10.1016/j.jhep.2018.03.024.

Aithal GP, Palaniyappan N, China L, et al. Guidelines on the management of ascites in cirrhosis. Gut. 2021;70:9–29. doi:10.1136/gutjnl-2020-321790.

National Institute for Health and Care Excellence. Cirrhosis in over 16s: assessment and management. Updated recommendations on prevention of spontaneous bacterial peritonitis. 2023.

Korean Association for the Study of the Liver. KASL clinical practice guidelines for liver cirrhosis: ascites and related complications. 2026.

Ginès P, Rimola A, Planas R, et al. Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis: results of a double-blind, placebo-controlled trial. Hepatology. 1990;12:716–724. doi:10.1002/hep.1840120416.

Fernández J, Navasa M, Planas R, et al. Primary prophylaxis of spontaneous bacterial peritonitis delays hepatorenal syndrome and improves survival in cirrhosis. Gastroenterology. 2007;133:818–824. doi:10.1053/j.gastro.2007.06.065.

Komolafe O, Roberts D, Freeman SC, et al. Antibiotic prophylaxis to prevent spontaneous bacterial peritonitis in people with liver cirrhosis: a network meta-analysis. Cochrane Database Syst Rev. 2020;1:CD013125. doi:10.1002/14651858.CD013125.pub2.

Praharaj DL, et al. Rifaximin vs norfloxacin for spontaneous bacterial peritonitis prophylaxis: a randomized controlled trial. 2022.

Silvey S, Patel N, O'Leary JG, et al. Secondary spontaneous bacterial peritonitis prophylaxis is associated with a higher rate of infections other than spontaneous bacterial peritonitis in two US-based national cirrhosis cohorts. Clin Transl Gastroenterol. 2025;16:e00837. doi:10.14309/ctg.0000000000000837.

Evidence-integrity note

Several aspects of SBP prophylaxis remain less certain than their widespread use might imply.

The evidence underpinning secondary prophylaxis comes predominantly from an old, small randomised trial conducted before the present era of multidrug-resistant infection. Primary-prophylaxis trials are also relatively small and heterogeneous, and the 2020 Cochrane analysis rated the overall evidence as very uncertain.

There is no high-quality prospective evidence defining exactly when secondary prophylaxis should be withdrawn after hepatic recompensation. The recommendation to stop primary prophylaxis following disappearance of ascites is explicitly supported by EASL and NICE, whereas extending this approach to secondary prophylaxis remains a clinical extrapolation rather than an established recommendation.

The 2025 observational studies questioning secondary prophylaxis are important but cannot establish causality and therefore should not be interpreted as sufficient evidence to abandon current guideline-supported secondary prophylaxis.

Antibiotic selection should also reflect local antimicrobial susceptibility patterns. This is particularly relevant in regions where quinolone resistance is common and where norfloxacin is unavailable.