Skip to content

Rifaximin and the SIBO question

TL;DR — Rifaximin is the only antibiotic with replicated efficacy in IBS, and its rationale is a construct most of the field does not accept. In TARGET 1 and 2 (two identical phase 3 trials, NCT00731679 and NCT00724126), 550 mg three times daily for 2 weeks gave adequate relief of global IBS symptoms in 40.7% versus 31.7% of placebo patients over the following 4 weeks (p<0.001 combined) and relief of bloating in 40.2% vs 30.3% (p<0.001) — a 9-percentage-point absolute difference (Pimentel 2011, PMID 21208106). TARGET 3 (NCT01543178) showed retreatment works but less well: of 1,074 open-label responders, 64.4% relapsed, and among 636 randomised, repeat rifaximin gave 38.1% response vs 31.5% placebo (p=0.03), with abdominal pain response 50.6% vs 42.2% (p=0.018) but no significant difference in stool consistency (51.8% vs 50.0%, p=0.42) (Lembo 2016, PMID 27528177). It is also the safest IBS drug measured: its number needed to harm for adverse-event discontinuation was negative and non-significant, i.e. no worse than placebo (Busam 2026, PMID 40471839). The SIBO construct that justified the trials is where the trouble lies. Breath-test-defined SIBO is present in 35.5% of IBS patients and 29.7% of controls (Shah 2020, PMID 31913194); lactulose breath testing yields a 3.6-fold higher prevalence in patients and a 7.6-fold higher prevalence in controls than glucose breath testing; jejunal aspirate culture puts prevalence at 4% (95% CI 2–9) (Ford 2009, PMID 19602448). The construct's diagnostic thresholds are set by consensus, not by outcome data, and the North American Consensus that set them attracted a published conflict-of-interest challenge (Maltz 2017, PMID 29215615). Rifaximin works; whether it works because of SIBO is unestablished — its measurable effect on the microbiota is "modest, largely transient" (Fodor 2019, PMID 29708822).

The efficacy record

Trial Design Primary endpoint Result
TARGET 1 & 2 (Pimentel 2011, PMID 21208106) Two identical phase 3 double-blind placebo-controlled trials; IBS without constipation; rifaximin 550 mg tds or placebo for 2 weeks, then 10 weeks' follow-up Adequate relief of global IBS symptoms for ≥2 of the first 4 weeks after treatment TARGET 1 40.8% vs 31.2% (p=0.01); TARGET 2 40.6% vs 32.2% (p=0.03); combined 40.7% vs 31.7% (p<0.001). Bloating relief 39.5% vs 28.7% (p=0.005), 41.0% vs 31.9% (p=0.02), combined 40.2% vs 30.3% (p<0.001). Daily ratings of global symptoms, bloating, abdominal pain and stool consistency also favoured rifaximin. Adverse-event incidence similar between groups
TARGET 3 (Lembo 2016, PMID 27528177) 270 centres, USA and Europe, Feb 2012–Jun 2014; 2-week open-label rifaximin, then randomisation of relapsing responders to repeat rifaximin or placebo Responder = ≥30% decrease in abdominal pain and ≥50% decrease in loose-stool frequency for ≥2 weeks in a 4-week post-treatment period 1,074 patients (44.1% of those entering open-label treatment) responded; of those, 382 (35.6%) did not relapse and 692 (64.4%) did; 636 randomised. Repeat treatment: 38.1% vs 31.5% (p=0.03). Abdominal pain 50.6% vs 42.2% (p=0.018); stool consistency 51.8% vs 50.0% (p=0.42, not significant). Improvements also in prevention of recurrence, durable response and urgency. Adverse-event rates low and similar
Pooled antibiotic analysis (Ford 2018, PMID 30294792) 5 similarly designed trials in non-constipated IBS RR of symptoms persisting 0.84 (0.79–0.90)
Post-hoc pain analysis (Lembo 2020, PMID 32352714) TARGET programme Abdominal pain response Detailed abdominal-pain response characterisation in IBS-D

Three features of this record are worth naming precisely. First, the absolute treatment–placebo difference is 9 percentage points in the registration trials and 6.6 points on retreatment — smaller than for linaclotide in IBS-C (19.8 points; Chey 2012, PMID 22986437) and comparable to eluxadoline. Second, the benefit is durable only in the sense that retreatment works; 64.4% of initial responders relapsed within 18 weeks (Lembo 2016, PMID 27528177). Third, retreatment loses the stool-consistency effect entirely while keeping the pain effect — which is difficult to reconcile with a simple bacterial-load model.

Safety and resistance

Rifaximin is minimally absorbed and its safety record is the best of any IBS drug quantified: NNH for adverse-event discontinuation was negative and statistically non-significant, and the 2026 safety meta-analysis names it "the safest pharmacotherapy studied" among IBS drugs (Busam 2026, PMID 40471839). Repeat courses did not produce clinically significant changes in stool microbial antibiotic sensitivity in the TARGET 3 programme (Pimentel 2017, PMID 28589238). Post-marketing pancreatitis reports for rifaximin ran at 0.5% of adverse-event submissions, versus 16.4% for eluxadoline (Gawron 2018, PMID 28804032).

The SIBO construct

Prevalence depends entirely on which test is used

Test SIBO prevalence in IBS In controls Source
Any breath test 35.5% (33.6–37.4) 29.7% (27.6–31.8) Shah 2020, PMID 31913194
Lactulose hydrogen breath test 54% (32–76) — Ford 2009, PMID 19602448
Glucose hydrogen breath test 31% (14–50) — Ford 2009, PMID 19602448
Jejunal aspirate and culture 4% (2–9) — Ford 2009, PMID 19602448
Culture, lower CFU/mL cut-off 13.9% (11.5–16.4) 5.0% (3.9–6.2) Shah 2020, PMID 31913194
Culture, higher CFU/mL cut-off 33.5% (30.1–36.9) 8.2% (6.8–9.6) Shah 2020, PMID 31913194

The odds ratio for SIBO in IBS versus controls is 3.7 (2.3–6.0), rising to 4.9 (2.8–8.6) against healthy controls only (Shah 2020, PMID 31913194); an earlier pool gave 3.45 (0.9–12.7) or 4.7 (1.7–12.95) depending on the positivity criterion, with significant heterogeneity for both (Ford 2009, PMID 19602448). Lactulose breath testing yields 3.6-fold higher prevalence in patients and 7.6-fold higher in controls than glucose breath testing (Shah 2020, PMID 31913194) — a test whose false-positive rate scales with the substrate is not measuring a stable biological entity. Shah and colleagues' own verdict: the evidence quality is low, "mainly due to substantial clinical heterogeneity due to lack of uniform selection criteria for cases and controls and limited sensitivity and specificity of the available diagnostic tests."

Subtype associations do exist and are internally consistent: SIBO odds are higher in IBS-D than IBS-C (OR 1.86, 1.83–2.8), while methane-positive breath tests are commoner in IBS-C (OR 2.3, 1.2–4.2) (Shah 2020, PMID 31913194) — the latter now framed as intestinal methanogen overgrowth and associated with delayed transit on wireless motility capsule (Talamantes 2024, PMID 39068378). Proton pump inhibitor use was not associated with SIBO within IBS (OR 0.8, 0.5–1.5, p=0.55) in that meta-analysis (Shah 2020, PMID 31913194), although a separate meta-analysis reports PPIs moderately increase SIBO risk generally (Su 2018, PMID 28770351).

The thresholds are consensus, and contested

The North American Consensus set the operative cut-offs: hydrogen rise ≥20 ppm by 90 minutes on glucose or lactulose breath testing, methane ≥10 ppm at any point; substrate doses of 10 g lactulose, 75 g glucose, 25 g fructose and 25 g lactose (Rezaie 2017, PMID 28323273). Those statements attracted immediate published objection on two grounds: the diagnostic role of lactulose in SIBO (Usai-Satta 2018, PMID 29535443, with reply Rezaie 2018, PMID 29535445) and conflicts of interest among the consensus authors (Maltz 2017, PMID 29215615). The consensus paper's own disclosure statement records that the meeting was supported in part by a breath-testing laboratory and that the convening institution held licensing agreements with breath-testing and pharmaceutical companies (Rezaie 2017, PMID 28323273).

Thresholds are also not internationally agreed. In 90 consecutive Malaysian IBS patients undergoing glucose hydrogen breath testing, SIBO was diagnosed in 44.4% by the Asia-Pacific criterion (hydrogen rise ≥12 ppm) and 37.8% by the North American criterion (≥20 ppm). Hydrogen positivity associated with IBS-D and methane positivity with IBS-C. Only the Asia-Pacific criterion was significantly associated with higher symptom severity scores (median 200 vs 165, p=0.009); the North American criterion was not (200 vs 170, p=0.087) (Loh 2026, PMID 42235989). ACG's own SIBO guideline codifies definitions and testing but acknowledges that several key concepts rest on expert consensus rather than GRADE-able evidence (Pimentel 2020, PMID 32023228). The dysbiosis framing has since broadened beyond bacterial "overgrowth" to small-intestinal microbial dysbiosis generally (Damianos 2026, PMID 41547360).

The antibiotic story before rifaximin, and the IBS-C branch

The line of work that produced TARGET began with a different antibiotic and a different gas.

Neomycin (2003). One hundred and eleven IBS subjects were randomised to neomycin or placebo, with blinded lactulose breath testing before and after. 84% of IBS subjects had an abnormal lactulose breath test versus 20% of healthy controls (p<0.01) — a figure that already contains the specificity problem this page turns on. Neomycin improved a composite score by 35.0% versus 11.4% (p<0.05) and self-reported bowel normalisation by 35.3% versus 13.9% (p<0.001), with a graded response: 75% improvement where neomycin normalised the breath test (one-way ANOVA, p<0.0001). Crucially, gas type tracked subtype — methane excretion was 100% associated with constipation-predominant IBS, and methane excretors had mean constipation severity 4.1 versus 2.3 in everyone else (p<0.001) (Pimentel 2003, PMID 12591062).

Neomycin plus rifaximin in methane-positive IBS-C (2014). A double-blind randomised placebo-controlled trial at three tertiary centres, 2010–2013, in 31 Rome II IBS-C subjects with breath methane >3 ppm. Constipation severity was lower with neomycin plus rifaximin (28.6 ± 30.8) than neomycin alone (61.2 ± 24.1), p=0.0042, with greater improvement in constipation (p=0.007), straining (p=0.017) and bloating (p=0.020) — but not abdominal pain. Within the combination arm, subjects whose methane fell below 3 ppm had lower constipation severity (30.5 ± 21.8) than those with persistent methane (67.2 ± 32.1), p=0.020 (Pimentel 2014, PMID 24788320).

Two features carry over into the modern argument. First, the gas-normalisation gradient — response tracks the change in breath gas, in both the 2003 and 2014 trials — is the strongest evidence anywhere that breath testing measures something causally connected to symptoms, and it sits uncomfortably beside the finding that 29.7% of healthy controls are breath-test positive (Shah 2020, PMID 31913194). Second, the 2014 trial is small (n=31) and targets a phenotype — methane-positive IBS-C — that no guideline recognises as a treatment category. Targeted antibiotic and dietary approaches across IBS subtypes have been reviewed more recently (Iftequar 2026, PMID 42310284).

Cost, access and value

Rifaximin's problem in practice is price rather than evidence. In a decision-analytic cost-benefit model of IBS-D treatments comparing guideline-recommended drugs, recently approved drugs, supplements, the low-FODMAP diet and CBT (Shah 2022, PMID 33010413):

  • From an insurer perspective, on-label prescription drugs (rifaximin, eluxadoline, alosetron) were significantly more expensive than off-label treatments, low FODMAP or CBT; insurer preferences were driven by average wholesale price and were not changed by health gains in sensitivity analysis up to a willingness-to-pay of $150,000 per QALY gained.
  • From a patient perspective, prescription drugs and neuromodulators appeared preferable, because effective therapy reduces lost wages, and because low-FODMAP food and attending CBT appointments carry their own out-of-pocket costs.

A separate analysis argued that value-based pricing for rifaximin would increase patient access (Shah 2019, PMID 30831219), and a treatment-free-interval framework has been proposed to capture rifaximin's episodic-dosing pattern in real-world effectiveness and economic terms (Lacy 2024, PMID 38619720). The insurer-versus-patient divergence is the honest summary: the payer and the patient rank IBS-D treatments differently, and neither ranking is driven primarily by efficacy.

Does rifaximin work through the microbiota?

The mechanistic test is the microbiome substudy of TARGET 3. Sixteen-S sequencing of stool from 103 randomly selected patients found that two weeks of open-label rifaximin significantly lowered the relative abundance of seven taxa (including Peptostreptococcaceae, Verrucomicrobiaceae and Enterobacteriaceae) at a 10% false-discovery-rate threshold — but there was "little evidence of significantly different changes in taxa relative abundance at the end of the study (up to 46 weeks) versus baseline." The authors' conclusion: rifaximin has "a modest, largely transient effect across a broad range of stool microbes," and whether those taxa are causally linked to IBS-D is future work (Fodor 2019, PMID 29708822).

Two observations follow. Stool is not small bowel, so a null stool result does not exclude a small-bowel mechanism. But the drug's clinical effect also outlasts its microbial effect and survives retreatment while losing its stool-consistency component (Lembo 2016, PMID 27528177) — a pattern more consistent with an anti-inflammatory, barrier or bile-acid-related action than with eradication of an overgrowth. Faecal bacterial composition has been reported to predict rifaximin response in IBS-D (Li 2020, PMID 32470562), which if replicated would be the first microbiome-based treatment-selection test in this condition.

Where guidelines land

AGA gives rifaximin a conditional recommendation at moderate certainty for IBS-D (Lembo 2022, PMID 35738725). BSG's position and the position of other societies are catalogued on guidelines. The pragmatic reading — treat non-constipated IBS with a safe antibiotic that helps about 1 in 11 patients beyond placebo, without believing the SIBO story — is coherent but leaves clinicians without a rule for whom to treat, because breath testing does not identify responders. BSG-aligned guidance is explicit that there is "no role for routine hydrogen breath tests for lactose malabsorption or small intestinal bacterial overgrowth" in suspected IBS (Black 2020, PMID 32133113), and Aziz and colleagues posed the question as an open verdict — "guilty or not guilty?" (Aziz 2017, PMID 28257307). Quigley's framing is that much of what is called SIBO "is not" SIBO (Quigley 2014, PMID 24406476).

Open questions

  • If rifaximin works, does SIBO exist? The drug's efficacy is the strongest argument for the construct, and its microbial effect is transient (Fodor 2019, PMID 29708822).
  • Why does retreatment preserve the abdominal-pain benefit but lose the stool-consistency benefit (Lembo 2016, PMID 27528177)?
  • Can responders be identified before treatment? Faecal composition has been reported to predict response once (Li 2020, PMID 32470562); a targeted PubMed search on 2026-09-02 retrieved no randomised trial showing that breath-test selection improves rifaximin response.
  • What threshold defines a positive breath test? The North American (≥20 ppm) and Asia-Pacific (≥12 ppm) criteria disagree and only the latter tracked symptom severity in one prospective series (Loh 2026, PMID 42235989).
  • Why is breath-test SIBO present in 29.7% of controls (Shah 2020, PMID 31913194)? Either the test is non-specific or asymptomatic overgrowth is normal — both readings undermine its diagnostic use.
  • How should the conflict-of-interest critique of the diagnostic consensus be weighed (Maltz 2017, PMID 29215615; Rezaie 2017, PMID 28323273)? This is a live methodological dispute, not a settled one.
  • Is intestinal methanogen overgrowth a distinct entity in IBS-C with its own treatment (Talamantes 2024, PMID 39068378; Pimentel 2014, PMID 24788320)? The neomycin-plus-rifaximin trial that supports it enrolled 31 patients.
  • Why does response track normalisation of breath gas (Pimentel 2003, PMID 12591062; Pimentel 2014, PMID 24788320) if the breath test is non-specific (Shah 2020, PMID 31913194)? These two facts are hard to hold simultaneously and nobody has reconciled them.
  • Should rifaximin be priced for value? Insurer rankings of IBS-D treatments are driven by wholesale price and unmoved by health gain up to $150,000/QALY (Shah 2022, PMID 33010413).

References

  1. Pimentel M, et al. Rifaximin therapy for patients with irritable bowel syndrome without constipation. N Engl J Med. 2011;364(1):22-32. PMID 21208106
  2. Lembo A, et al. Repeat Treatment With Rifaximin Is Safe and Effective in Patients With Diarrhea-Predominant Irritable Bowel Syndrome. Gastroenterology. 2016;151(6):1113-1121. PMID 27528177
  3. Lembo A, et al. Abdominal Pain Response to Rifaximin in Patients With Irritable Bowel Syndrome With Diarrhea. Clin Transl Gastroenterol. 2020;11(3):e00144. PMID 32352714
  4. Pimentel M, et al. Repeat Rifaximin for Irritable Bowel Syndrome: No Clinically Significant Changes in Stool Microbial Antibiotic Sensitivity. Dig Dis Sci. 2017;62(9):2455-2463. PMID 28589238
  5. Fodor AA, et al. Rifaximin is associated with modest, transient decreases in multiple taxa in the gut microbiota of patients with diarrhoea-predominant irritable bowel syndrome. Gut Microbes. 2019;10(1):22-33. PMID 29708822
  6. Ford AC, Harris LA, Lacy BE, Quigley EMM, Moayyedi P. Systematic review with meta-analysis: the efficacy of prebiotics, probiotics, synbiotics and antibiotics in irritable bowel syndrome. Aliment Pharmacol Ther. 2018;48(10):1044-1060. PMID 30294792
  7. Shah A, et al. Small Intestinal Bacterial Overgrowth in Irritable Bowel Syndrome: A Systematic Review and Meta-Analysis of Case-Control Studies. Am J Gastroenterol. 2020;115(2):190-201. PMID 31913194
  8. Ford AC, Spiegel BM, Talley NJ, Moayyedi P. Small intestinal bacterial overgrowth in irritable bowel syndrome: systematic review and meta-analysis. Clin Gastroenterol Hepatol. 2009;7(12):1279-86. PMID 19602448
  9. Pimentel M, Saad RJ, Long MD, Rao SSC. ACG Clinical Guideline: Small Intestinal Bacterial Overgrowth. Am J Gastroenterol. 2020;115(2):165-178. PMID 32023228
  10. Rezaie A, et al. Hydrogen and Methane-Based Breath Testing in Gastrointestinal Disorders: The North American Consensus. Am J Gastroenterol. 2017;112(5):775-784. PMID 28323273
  11. Maltz C. Conflicts of Interest in the North American Consensus on Breath Testing. Am J Gastroenterol. 2017;112(12):1892. PMID 29215615
  12. Usai-Satta P, Oppia F, Cabras F. The North American Consensus on Breath Testing: The Controversial Diagnostic Role of Lactulose in SIBO. Am J Gastroenterol. 2018;113(3):440. PMID 29535443
  13. Rezaie A, Pimentel M. Reply to Satta et al. Am J Gastroenterol. 2018;113(3):440-441. PMID 29535445
  14. Loh AJC, et al. Reassessing Small Intestinal Bacterial Overgrowth in Irritable Bowel Syndrome: A Comparative Study of Diagnostic Guidelines and Their Clinical Relevance. J Gastroenterol Hepatol. 2026;41(7):2082-2088. PMID 42235989
  15. Talamantes S, et al. Intestinal Methanogen Overgrowth (IMO) Is Associated with Delayed Small Bowel and Colonic Transit Time on the Wireless Motility Capsule. Dig Dis Sci. 2024;69(9):3361-3368. PMID 39068378
  16. Su T, et al. Meta-analysis: proton pump inhibitors moderately increase the risk of small intestinal bacterial overgrowth. J Gastroenterol. 2018;53(1):27-36. PMID 28770351
  17. Damianos JA, et al. Mechanisms and pathophysiology leading to development of small intestinal microbial dysbiosis. Lancet Gastroenterol Hepatol. 2026;11(3):256-266. PMID 41547360
  18. Aziz I, Törnblom H, Simrén M. Small intestinal bacterial overgrowth as a cause for irritable bowel syndrome: guilty or not guilty? Curr Opin Gastroenterol. 2017;33(3):196-202. PMID 28257307
  19. Quigley EM. Small intestinal bacterial overgrowth: what it is and what it is not. Curr Opin Gastroenterol. 2014;30(2):141-6. PMID 24406476
  20. Black CJ, Ford AC. Rational investigations in irritable bowel syndrome. Frontline Gastroenterol. 2020;11(2):140-147. PMID 32133113
  21. Lembo A, Sultan S, Chang L, Heidelbaugh JJ, Smalley W, Verne GN. AGA Clinical Practice Guideline on the Pharmacological Management of Irritable Bowel Syndrome With Diarrhea. Gastroenterology. 2022;163(1):137-151. PMID 35738725
  22. Busam JA, et al. The Safety of Pharmacotherapy for Irritable Bowel Syndrome: A Systematic Review and Meta-Analysis. Am J Gastroenterol. 2026;121(3):745-753. PMID 40471839
  23. Gawron AJ, Bielefeldt K. Risk of Pancreatitis Following Treatment of Irritable Bowel Syndrome With Eluxadoline. Clin Gastroenterol Hepatol. 2018;16(3):378-384.e2. PMID 28804032
  24. Li Y, et al. Fecal bacteria can predict the efficacy of rifaximin in patients with diarrhea-predominant irritable bowel syndrome. Pharmacol Res. 2020;159:104936. PMID 32470562
  25. Chey WD, et al. Linaclotide for irritable bowel syndrome with constipation: a 26-week, randomized, double-blind, placebo-controlled trial. Am J Gastroenterol. 2012;107(11):1702-12. PMID 22986437
  26. Pimentel M, Chow EJ, Lin HC. Normalization of lactulose breath testing correlates with symptom improvement in irritable bowel syndrome: a double-blind, randomized, placebo-controlled study. Am J Gastroenterol. 2003;98(2):412-9. PMID 12591062
  27. Pimentel M, et al. Antibiotic treatment of constipation-predominant irritable bowel syndrome. Dig Dis Sci. 2014;59(6):1278-85. PMID 24788320
  28. Iftequar Y, et al. Targeted antibiotic and dietary approaches in managing small intestinal bacterial overgrowth across irritable bowel syndrome subtypes. Intern Emerg Med. 2026 Jun 17 (online ahead of print). PMID 42310284
  29. Shah ED, Salwen-Deremer JK, Gibson PR, Muir JG, Eswaran S, Chey WD. Comparing Costs and Outcomes of Treatments for Irritable Bowel Syndrome With Diarrhea: Cost-Benefit Analysis. Clin Gastroenterol Hepatol. 2022;20(1):136-144.e31. PMID 33010413
  30. Shah ED, et al. Value-based Pricing for Rifaximin Increases Access of Patients With Irritable Bowel Syndrome With Diarrhea to Therapy. Clin Gastroenterol Hepatol. 2019;17(13):2687-2695.e11. PMID 30831219
  31. Lacy BE, et al. Treatment-Free Interval: A Novel Approach to Assessing Real-World Treatment Effectiveness and Economic Impact Among Patients with Irritable Bowel Syndrome with Diarrhea. Adv Ther. 2024;41(6):2253-2266. PMID 38619720
  32. Takakura W, Pimentel M. Small Intestinal Bacterial Overgrowth and Irritable Bowel Syndrome - An Update. Front Psychiatry. 2020;11:664. PMID 32754068