Antibiotic-Associated Diarrhea
Probiotics' Best-Evidence Use Case
Evidence grade 4 of 5, Strong. Read how the Atlas grades evidence.
63 human trials, about 300 studies, Tens of thousands across Cochrane reviews. The written rationale for this grade is still being prepared.
- FDA
- Probiotics used adjunctively; not a drug claim
- WADA
- Not applicable
Reference only. Not medical advice.
What can go wrong
For most healthy outpatients, specific well-studied strains taken with antibiotics have a good short-term safety record, and some meta-analyses even report fewer overall adverse events than placebo. The serious caution is population-specific: in severely immunocompromised, critically ill, or central-venous-catheter patients, probiotics carry a real if rare risk of bloodstream infection (probiotic sepsis or fungemia). One ICU analysis found probiotic-associated catheter bloodstream infections in about 0.37 percent of recipients and none in non-recipients, with increased mortality, and guideline experts specifically advise avoiding S. boulardii in these groups. Benefit is also conditional: the wrong strain, a low-risk patient, or a late start may yield no benefit at all. Anyone who is pregnant, immunocompromised, seriously ill, or caring for a hospitalized or vulnerable patient should decide about probiotics with a qualified healthcare provider rather than self-prescribing.
Antibiotic-associated diarrhea, including some Clostridioides difficile infection, is the closest probiotics come to a flagship evidence-based use, and it is also where the field's internal disagreement is sharpest. Large meta-analyses point one way: Hempel 2012 in JAMA (63 RCTs) found a relative risk of 0.58 for AAD, and the Goldenberg 2017 Cochrane review (31 RCTs) found a relative risk of 0.40 for C. difficile diarrhea, with Saccharomyces boulardii CNCM I-745 and Lactobacillus rhamnosus GG the best-studied strains. Yet the large pragmatic PLACIDE trial of 2,941 older inpatients found no benefit at all, and the guideline bodies split: the AGA conditionally suggests specific strains, the ACG recommends against them, and IDSA/SHEA call the evidence insufficient outside clinical trials. The most consistent reading is that any benefit is strain-specific, timing-dependent (started early with the antibiotic), and concentrated in higher-risk populations, not a universal effect. This is the category's best evidence, presented honestly, rather than unconditional proof.
Preventing diarrhea caused by antibiotics is the strongest, best-evidenced use of probiotics, with specific strains clearly ahead.
What people take it for
What it is taken for. Approval, where it exists, covers only some of these.
What people get wrong
Often saidAny probiotic taken with antibiotics prevents diarrhea
ActuallyOnly specific strains have real supporting evidence, notably S. boulardii CNCM I-745 and Lactobacillus rhamnosus GG. The AGA 2020 guideline names particular strains and combinations, not probiotics in general, and pooled analyses that mix strains are exactly what the ACG criticized. Generic or unstudied products may do nothing.
Often saidThe science is settled that probiotics prevent C. difficile
ActuallyIt is genuinely disputed. The Goldenberg 2017 Cochrane review (relative risk 0.40) and Hempel 2012 JAMA support prevention, but the pragmatic PLACIDE trial of 2,941 older inpatients found no benefit, and major bodies disagree: AGA conditionally suggests specific strains, ACG recommends against, and IDSA/SHEA call the evidence insufficient outside trials.
Often saidIf probiotics help, they help everyone equally
ActuallyThe Cochrane review found the C. difficile benefit was concentrated in patients with a baseline risk above 5 percent (number needed to treat about 12) and not statistically significant at 5 percent or below. Timing also matters: a 2017 meta-regression found benefit mainly when probiotics were started within 2 days of the first antibiotic dose.
Often saidProbiotics with antibiotics are safe for absolutely everyone
ActuallyFor most healthy outpatients the safety record is good, but in ICU patients with central venous catheters one analysis found probiotic-associated bloodstream infections in 86 of 23,015 recipients (0.37 percent) versus none in non-recipients, with higher mortality. Guideline experts advise avoiding S. boulardii in immunocompromised, critically ill, or central-line patients. Anyone unsure should consult a qualified healthcare provider.
Who profits, and who pushes back
Cochrane Collaboration
Multiple reviews support specific probiotics for preventing antibiotic-associated and C. difficile diarrhea, the category's strongest evidence.
S. boulardii / LGG manufacturers
Sells this. Has a commercial interest in a product sold as this compound, as of July 2026.
Sell the best-evidenced strains; here the marketing and the evidence actually align reasonably well.
Gastroenterology guidelines
Cautiously support specific strains for AAD/C. diff prevention while noting population and timing caveats.
American College of Gastroenterology (2021 guideline)
The ACG recommends against probiotics for the prevention of C. difficile infection in patients being treated with antibiotics (conditional recommendation, moderate quality of evidence), and against probiotics for the prevention of recurrence.
Stephen Allen and PLACIDE investigators (Lancet, 2013)
The authors concluded that this preparation of lactobacilli and bifidobacteria did not appear to prevent antibiotic-associated diarrhoea or C. difficile diarrhoea in older inpatients, the largest such trial to that point.
IDSA/SHEA C. difficile guideline panel
There are insufficient data at this time to recommend administration of probiotics for primary prevention of C. difficile infection outside of clinical trials.
Where the research comes from
| Place | Share of published research |
|---|---|
| Europe | 32% |
| USA | 32% |
| Asia | 20% |
| Other | 16% |
Research published from: Global.
How we got here
Antibiotic-associated diarrhea was recognized as antibiotics spread through mid-twentieth-century medicine worldwide, and the idea of countering it with beneficial microbes drew on a much older lineage. Saccharomyces boulardii itself was isolated in the 1920s by French microbiologist Henri Boulard from lychee and mangosteen skins in Indochina (present-day Vietnam), where he observed local people brewing the fruit peel to calm diarrhea, an East Asian folk practice that predated the Western product. Through the 2000s and 2010s, as C. difficile emerged as a costly hospital-acquired infection, probiotics were tested more heavily here than almost anywhere else in the field. McFarland's meta-analyses, successive Cochrane reviews, Hempel's 2012 JAMA synthesis, and Szajewska's strain-level reviews converged on a preventive signal for specific strains. The 2013 PLACIDE trial then complicated that story, and the indication became both the showcase for strain-specificity and the clearest illustration of how pragmatic trials and pooled meta-analyses can disagree.
- 1920sHenri Boulard isolates S. boulardii from fruit peel used against diarrhea in Indochina (Vietnam)
- 2000sRising hospital C. difficile drives heavy probiotic testing for antibiotic-associated diarrhea
- 2012Hempel meta-analysis in JAMA: 63 RCTs, relative risk 0.58 for AAD, NNT 13
- 2013PLACIDE RCT (Lancet), 2,941 older inpatients, finds no benefit for AAD or C. difficile diarrhea
- 2017Goldenberg Cochrane review: relative risk 0.40 for C. difficile diarrhea, benefit concentrated in high-risk patients; Shen meta-regression shows early start matters
- 2020AGA conditionally suggests specific strains for C. difficile prevention (low certainty)
- 2021ACG recommends AGAINST probiotics for primary and secondary C. difficile prevention
At a glance
- Best-evidence use
- Preventing antibiotic-associated diarrhea
- Top strains
- S. boulardii CNCM I-745, LGG
- C. diff
- Specific strains/mixtures reduce risk
- Caveat
- Timing and population matter
What this entry does not yet cover
What follows is unfinished work on our side, not a finding about Antibiotic-Associated Diarrhea. A short entry is not a safe entry.
- Cost and market. We have not documented what it sells for or who sells it.
- Supply chain. We have not mapped where it is made.
- Who uses it. We have not documented the user groups.
- What could change. No forward-looking summary yet.
- How it compares. We have not set it against the alternatives.
- Children and teenagers. We have not checked what is known about younger people.
Sources
Hempel 2012 JAMA meta-analysis: 63 RCTs, 11,811 participants, relative risk 0.58 (0.50-0.68) for AAD, NNT 13, evidence insufficient to say which strains or populations
Goldenberg 2017 Cochrane: 31 RCTs, 8,672 participants, C. difficile diarrhea relative risk 0.40 (0.30-0.52), moderate certainty, benefit concentrated in baseline risk over 5 percent (NNT 12), no significant benefit at 5 percent or under
Cochrane Database of Systematic Reviews (Goldenberg et al.), via PubMed
PLACIDE pragmatic RCT: 2,941 older inpatients, no benefit for AAD (10.8 vs 10.4 percent) or C. difficile diarrhea (0.8 vs 1.2 percent)
PLACIDE detailed numbers, strains, dose, mean age 77, and low-event-rate power limitation
AGA 2020 conditionally suggests specific strains for C. difficile prevention (low certainty); recommends against probiotics for most GI conditions
AGA 2020 full clinical practice guideline on the role of probiotics in GI disorders
ACG 2021 recommends against probiotics for primary (conditional, moderate quality) and secondary (strong, very low quality) prevention of C. difficile
Why AGA and ACG differ: AGA accounts for strain specificity across 39 RCTs, ACG pools strains and leans on one underpowered trial
IDSA/SHEA 2017 (upheld 2021): insufficient data to recommend probiotics for primary prevention of C. difficile outside clinical trials
Shen 2017 meta-regression: probiotics started within 2 days of first antibiotic dose cut C. difficile risk (RR 0.32) versus later start (RR 0.70)
S. boulardii CNCM I-745: 21 RCTs, ~4,780 people, ~53 percent AAD reduction; multifactorial mechanism; well tolerated
Probiotic-associated central-line bloodstream infections in 86 of 23,015 ICU recipients (0.37 percent) vs none in non-recipients, higher mortality; avoid S. boulardii in immunocompromised, critically ill, or central-line patients
Critical Care Medicine (ICU probiotic BSI study), via PubMed
Regulatory oversight and safety of probiotic use, including fungemia risk context
Real-world OTC framing of S. boulardii as evidence-based option to reduce AAD risk