In this guide
What is Saccharomyces boulardii?
Saccharomyces boulardii is a live yeast used as a probiotic. Many researched products identify the strain as CNCM I-745. “Probiotic yeast,” “S. boulardii” and CNCM I-745 are not automatically interchangeable descriptions: a useful evidence match still requires the strain, viable finished product, amount, schedule, population and outcome. [1]
Its yeast identity has two practical consequences. Antibacterial drugs do not directly target it the way they can target susceptible bacterial probiotics. Antifungal medicines can inhibit yeast, so concurrent use may defeat the purpose of the product. More seriously, a live yeast can enter the bloodstream in vulnerable patients. That risk is rare in ordinary reports but is concentrated enough around central venous access and critical illness to define a clear avoid-or-specialist-review boundary. [6], [7]
Organism identity
A live yeast changes the interaction check
The exact medicine, patient risk and product still need review.
Typical Saccharomyces boulardii Protocols
| Study and population | Exact product | Amount each time | Frequency | Daily total and duration | Primary interpretation |
|---|---|---|---|---|---|
| Ehrhardt 2016 · 477 hospitalized adults receiving antibiotics | Perenterol forte 250 mg capsule | 250 mg | Twice daily | 500 mg/day from within 24 hours of the antibiotic through 7 days after it ended; 6-week follow-up | Stopped early for futility; no reduction in antibiotic-associated diarrhea |
| McFarland 1995 · 193 hospitalized adults receiving beta-lactam antibiotics | S. boulardii preparation | Not reported in the abstract | Not reported in the abstract | 1 g/day, started within 72 hours of the antibiotic and continued until 3 days after it ended | AAD occurred in 7/97 with S. boulardii and 14/96 with placebo |
| 2025 adjunct trial · 120 adults with mild or moderate C. difficile infection | CNCM I-745, two 250 mg capsules per administration, plus standard vancomycin | 500 mg | Every 12 hours | 1 g/day for 10 days; vancomycin 125 mg every 6 hours in both groups | Clinical cure was similar; global cure and recurrence favored adjunct S. boulardii |
The 2016 inpatient trial is the most important counterweight to broad prevention claims. It found antibiotic-associated diarrhea in 21 participants assigned S. boulardii and 19 assigned placebo, with a hazard ratio of 1.02 and P=0.94. C. difficile-associated diarrhea occurred in two participants in each group. The study stopped after an interim futility analysis, so it did not reach the planned sample size, but the observed result does not suggest a missed large benefit. [2]
- Amount versus viability
- A 250 mg capsule is not a universal number of living cells. Product moisture, excipients, manufacturing and viable count differ; preserve both mg and CFU when a label provides them.
- Antibiotic timing
- The trials started S. boulardii early during antibiotic treatment and continued for a defined period afterward. They did not test replacing, delaying or shortening the prescribed antibiotic.
- C. difficile context
- The 2025 schedule was an adjunct to vancomycin in selected mild or moderate infection. It was not S. boulardii monotherapy and excluded many high-risk patients. [5]
What is S. boulardii commonly used for?
Antibiotic-associated diarrhea
This is the best-studied use, but the answer depends on the setting. The older beta-lactam trial found fewer antibiotic-associated-diarrhea cases with 1 g/day. A 2015 meta-analysis combining 21 randomized trials reported a reduction from 18.7% to 8.5%, a pooled risk ratio of 0.47 and a number needed to treat of about 10. Yet the later, larger hospital trial found no benefit at 500 mg/day. [2], [3], [4]
Those findings can coexist. A pooled average can be favorable while one well-run trial in a specific inpatient population is negative. Differences in antibiotic exposures, diarrhea definitions, baseline risk, dose, product and study conduct prevent a single guaranteed effect estimate for every course.
C. difficile infection
S. boulardii is not a substitute for diagnostic testing or guideline-directed antimicrobial treatment. In a 2025 randomized adjunct study, adding CNCM I-745 to vancomycin did not improve initial clinical cure, which was about 98% in both groups. Global cure was 96.6% versus 85.3%, and recurrence was 1.7% versus 13.1%, favoring the adjunct group. These are promising secondary and follow-up results from one 120-person trial, not a basis for unsupervised treatment. [5]
IBS and general “microbiome restoration”
Small IBS studies have reported quality-of-life or biomarker signals without consistent superiority across bowel symptoms. Changes in stool microbiota during antibiotic exposure can support a mechanism, but they do not prove durable colonization or patient-important benefit. The guide therefore does not assign an IBS protocol from mechanistic or mixed-outcome studies.
Antibiotic-associated diarrhea
Two valid findings, one mixed record
Population, product, dose and outcome definition decide how closely either result matches a new use.
How do you read the form and label?
Look for Saccharomyces boulardii, a strain code such as CNCM I-745, the amount per capsule or sachet, a viable count and when that count is guaranteed, the serving size, expiry and storage directions. “Probiotic blend” without organism-level amounts cannot be matched to these trials. A milligram label and a CFU label describe different properties and should both be recorded rather than converted with a generic ratio.
Products can be foods, dietary supplements or licensed natural health products depending on jurisdiction and claims. A familiar brand name does not establish that the current formulation is identical to an older research product.
How should S. boulardii be stored and handled?
Follow the finished-product label and keep capsules or sachets dry and sealed. Heat, moisture and time can reduce viability. In hospitals or homes with a person who has a central line or profound immune compromise, handling is also a safety issue: case reports and an ICU outbreak describe contamination of hands, surfaces or catheter sites when live-yeast capsules or sachets were opened nearby. [7]
What does the human evidence show?
| Question | Best direct evidence here | Answer | Main limit |
|---|---|---|---|
| Does 250 mg twice daily prevent inpatient AAD? | Ehrhardt, 477 adults | No detectable benefit; stopped for futility | Early stop reduced planned precision |
| Does the whole AAD trial record favor S. boulardii? | 2015 meta-analysis, 21 trials | Yes, pooled RR 0.47 | Heterogeneous populations, products and methods |
| Does adjunct CNCM I-745 improve initial CDI cure? | 2025 trial, 120 adults | Initial clinical cure was similar | Single selected mild/moderate population |
| Did the adjunct trial reduce recurrence? | Same 2025 trial | Favorable secondary/follow-up signal | Needs independent replication |
The evidence is stronger for saying “some tested regimens reduced AAD in some populations” than for promising prevention to any person taking antibiotics. It is also stronger for adjunct research in C. difficile infection than for S. boulardii alone, which the cited trial did not test.
What are the risks and interactions?
Gas, bloating, constipation, altered stools, nausea, itching or rash can occur. Stop and seek assessment for severe or persistent diarrhea, blood in stool, high fever, dehydration, marked abdominal pain, repeated vomiting, confusion or rapid deterioration. These signs can indicate the infection, medicine complication or another condition rather than a “die-off” response.
Fungemia is the distinctive serious risk. A review of 60 published fungemia cases found central venous catheters in 93% and broad-spectrum antimicrobial exposure in 88%; overall mortality was 28%. Those proportions describe selected reported cases, not the incidence among users, and cannot estimate ordinary outpatient risk. They do identify critical illness, central access, immune compromise and impaired intestinal barriers as settings where a live yeast should be avoided unless the treating specialist explicitly directs it. [6]
Cross-contamination matters. An ICU report described fungemia in three patients sharing a room with people receiving S. boulardii, even though the affected patients were not themselves prescribed the probiotic. The authors implicated environmental or hand transmission during capsule opening. Do not open or prepare a live-yeast product around central-line care. [7]
Antifungal medicines can inhibit S. boulardii. Do not stop or separate a prescribed antifungal to preserve a probiotic regimen. Antibacterial drugs create the use context for AAD trials, but S. boulardii does not prevent all antibiotic complications and does not replace medical review.
What is the legal and regulatory status?
In the United States, S. boulardii products sold as dietary supplements are not FDA-approved treatments for antibiotic-associated diarrhea or C. difficile infection. In Canada, a marketed natural health product needs a product licence and approved label claims; that licence does not make every S. boulardii product or off-label treatment claim equivalent. Hospital treatment decisions follow clinical policy and infection-control rules, not retail supplement status.
What happens when you stop?
S. boulardii is not known to produce a drug-like withdrawal syndrome. Stool recovery of the yeast typically falls after use ends. A return of diarrhea can reflect the underlying infection, antibiotic effect or another cause and deserves clinical review when severe, persistent or accompanied by warning signs. Restarting repeatedly can delay diagnosis.
What should you monitor?
Record the antibiotic name and dates, the exact S. boulardii product and strain, each administration, stool frequency and consistency, hydration, fever and any rescue treatment. Use the study-defined window: symptoms during antibiotics and for several weeks afterward mattered in the inpatient trial. A central line, hospitalization, immune suppression or new antifungal prescription should trigger immediate clinician review rather than routine tracking.
What does community experience add?
Public reports often focus on loose stools, bloating, constipation, itching, cost and whether symptoms returned after stopping. These reports can identify tolerability questions but cannot estimate efficacy or fungemia risk because diagnosis, antibiotic exposure, strain and product handling are rarely verified. No systematic current community sample was assembled for this refresh, so no consensus percentage is reported.
How do nearby probiotic options compare?
| Option | Organism | Main evidence boundary |
|---|---|---|
| S. boulardii CNCM I-745 | Yeast | Mixed AAD evidence; distinctive fungemia and handling risk |
| Lactobacillus strains | Bacteria | Results stay attached to exact strain and indication |
| Bifidobacterium strains | Bacteria | Negative and positive results vary by strain and outcome |
| Synbiotic | Live organism plus substrate | Effect belongs to the complete tested combination |
Common questions
Can S. boulardii be taken with an antibiotic?
That was the tested setting in the AAD trials. It does not guarantee benefit, and the exact product, schedule and patient risk still matter. Do not alter the antibiotic prescription. [2], [3]
Does it need to be separated from antibiotics?
Antibacterial drugs do not directly target yeast in the same way they target susceptible bacteria. Follow the product or pharmacist instructions because timing may still matter for adherence and other medicines. Antifungals are a different interaction.
Is 250 mg the same as a specific CFU count?
No universal conversion exists. Milligrams describe product mass; CFU estimates viable cells. Use the values on the exact product rather than importing a ratio from another brand.
Can it treat C. difficile infection?
The cited 2025 trial studied it only as an adjunct to vancomycin in selected mild or moderate cases. Initial clinical cure was similar. Diagnosis and treatment require clinical care. [5]
Who should avoid it?
People with a central venous catheter, critical illness, major immune compromise or severely impaired gut barrier should not self-start a live yeast. The treating team should decide whether any use is justified. [6], [7]
Glossary
Plain explanations of the trial and safety terms used here.
- Antibiotic-associated diarrhea (AAD)
- Diarrhea that begins during or after antibiotic exposure, with several possible causes.
- Fungemia
- Yeast or another fungus detected in the bloodstream, a potentially life-threatening invasive infection.
- CNCM I-745
- A deposited S. boulardii strain identifier used by several studied products.
- Futility stop
- Ending a trial early because interim results show a low chance of demonstrating its planned benefit.
- Global cure
- In the cited C. difficile trial, initial clinical cure without recurrence during follow-up.
How this guide was researched
The protocol table uses full trial reports or PubMed records to preserve product mass, frequency, duration, population and outcome. The large negative inpatient RCT was retained beside the favorable older trial and meta-analysis. The C. difficile trial was treated as adjunct research, and the fungemia literature was used to define risk factors without converting case proportions into incidence.
Research record · checked October 9, 2026
Protocol verification: the Ehrhardt hospital trial, McFarland beta-lactam trial and 2025 CNCM I-745 adjunct trial were checked for amount per administration, frequency, daily total, duration and co-treatment. Missing per-dose details in the older abstract remain marked not reported.
Negative evidence: the 477-person futility-stopped trial and its equal C. difficile-associated-diarrhea count remain explicit.
Safety calibration: proportions from a 60-case fungemia review describe selected cases, not population incidence. The ICU report supports a handling boundary but cannot estimate frequency.
Deliberate exclusion: mechanistic microbiome changes, testimonials and mixed probiotic products were not converted into product-specific efficacy claims.
Explore the sources
Primary trials, evidence synthesis and safety reports supporting this guide.
Showing 7 sources
- 01
Probiotics: Health Professional Fact Sheet ↗
Official evidence summary
Used for probiotic identity, strain specificity and high-risk-host context.
- 02
Saccharomyces boulardii to Prevent Antibiotic-Associated Diarrhea: A Randomized, Double-Masked, Placebo-Controlled Trial. ↗
Randomized placebo-controlled trial
Full report reviewed; stopped for futility with no AAD benefit.
- 03
Prevention of beta-lactam-associated diarrhea by Saccharomyces boulardii compared with placebo. ↗
Randomized placebo-controlled trial
Abstract reviewed; favorable older 1 g/day result, per-administration details not reported there.
- 04
Systematic review with meta-analysis: Saccharomyces boulardii in the prevention of antibiotic-associated diarrhoea. ↗
Systematic review and meta-analysis
Used for pooled AAD effect and heterogeneity context.
- 05
Efficacy and safety of Saccharomyces boulardii as adjunct therapy with Vancomycin in treating Clostridioides difficile infection: A randomized controlled trial. ↗
Randomized adjunct trial
Full report reviewed; initial cure similar, global cure and recurrence favorable.
- 06
Saccharomyces cerevisiae fungemia: an emerging infectious disease. ↗
Case series and literature review
Used to identify central-line and critical-illness patterns; not an incidence estimate.
- 07
Outbreak of Saccharomyces cerevisiae subtype boulardii fungemia in patients neighboring those treated with a probiotic preparation of the organism. ↗
ICU outbreak report
Used for cross-contamination and handling risk around central lines.
No matching sources
Try a different term or show the full list.
Editorial note
This guide preserves the large negative inpatient AAD trial beside favorable pooled and older results. It separates initial clinical cure from recurrence in the adjunct C. difficile trial and presents fungemia case patterns as a high-risk boundary rather than a population rate. It was authored in English for translation from the same source record.