The authors concluded that probiotics were generally beneficial in treatment and prevention of gastrointestinal diseases. The conclusions reflect the evidence presented, but potential for publication bias and error and bias in the conduct of review processes should be considered.

To assess the efficacy of probiotics for treatment and prevention of gastrointestinal diseases.

PubMed, MEDLINE, Google Scholar, EMBASE, Biological Abstracts and Science Direct were searched for studies published between 1970 and January 2011. Search terms were reported. No language limitation was applied. Reference lists of identified articles were handsearched for additional studies.

Randomised controlled trials (RCTs) that used a specified probiotic in the treatment or prevention of Pouchitis, Infectious diarrhoea, Irritable Bowel Syndrome, Helicobacter pylori, Clostridium difficile Disease, Antibiotic Associated Diarrhoea, Traveller’s Diarrhoea or Necrotizing Enterocolitis were eligible for inclusion. The primary outcome measure was prevention in overall symptoms or treatment of the gastrointestinal diseases.

Most studies enrolled adults; few enrolled children or infants and one enrolled elderly patients. Probiotics included VSL#3, Lactobacillus rhamnosus GG, Saccharomyces boulardii, Bifidobacterium infantis, L. acidophilus, L. casei, Clostridium butyricum, Enterococcus faecium, L. plantarum, B. lactis and L. acidophilus combined with B. infantis. The duration of treatment in most studies (where reported) was less than four weeks. Most studies included patients with Antibiotic Associated Diarrhoea, H. pylori or Irritable Bowel Syndrome; remaining studies included patients with C. difficile Disease, Infectious diarrhoea, Necrotizing Enterocolitis, Pouchitis or Traveller’s Diarrhoea.

One author assessed study eligibility for inclusion.

Studies were assessed with the Jadad scale. Key criteria assessed the adequacy of randomisation, concealment of treatment allocation, blinding of investigators and outcome assessors and completeness of follow-up.

One author assessed study quality. Inconsistencies were resolved by discussion with other reviewers.

Data on the type of disease, probiotic species, the dose, treatment length, age group, number of patients who received the probiotic or the control, and the number of patients that improved following probiotic/control were extracted and used to calculate risk ratios (RRs) and 95% confidence intervals (CIs).

Authors did not state how many reviewers extracted data.

Pooled risk ratios and 95% confidence intervals across diseases, probiotic species, single versus multiple species, patient ages, dosages and length of treatment were calculated using random-effects meta-analysis where there was evidence for significant heterogeneity; otherwise fixed-effect meta-analysis was used.

Heterogeneity was assessed using Ι². Publication bias was assessed using a funnel plot, Begg and Egger's tests and fail-safe N.

Eighty four studies were included (10,351 participants; range 18 to 756). All trials included in the meta-analysis had a Jadad score of at least 3 except four trials which had a Jadad score of 2.

Probiotics were effective in the treatment and prevention of all gastrointestinal diseases and probiotic species (RR 0.58, 95% CI 0.51 to 0.65; Ι²=61.24%).

Effect by type of diseases: Significant effects of probiotics were observed in the following diseases: Pouchitis, Antibiotic Associated Diarrhoea (27 studies), Infectious diarrhoea (three studies), Irritable Bowel Syndrome (16 studies), Helicobacter pylori (13 studies) and Clostridium difficile Disease (six studies). No significant effects were observed for probiotics for Traveller’s Diarrhoea (six studies) and Necrotizing Enterocolitis (nine studies).

Effect by probiotic species: Across all diseases, significant effects of probiotics were observed in the following species: VSL #3, E. Faecium, C. butyricum, L. acidophilus combined with B. infantis, B. lactis, LGG, L. Casei and S. boulardii. No significant effects were observed for L. acidophilus, L. plantarum and B. infantis.

Effects of age: Across all diseases and probiotic species, significant effects of probiotics was observed for all age groups (infants, children and adults).

Effects of dose: Across all diseases and probiotics species, significant efficacy was observed for three doses: 1- 5x10¹⁰ colony-forming units (CFU)/day, 1–5.5^6, 10⁶, 10⁷, 10⁸ CFU/day and 1–9x10⁹ CFU/day. No significant effect was observed for one dose (1-9x10¹¹, 10¹² CFU/day).

Effect of treatment length that probiotic was administered: significant efficacy was observed across all groups (one to two weeks, three to four weeks, five to eight weeks and nine to 240 weeks).

There was no significant difference in efficacy between single and multiple species probiotics. There was evidence of publication bias.

Probiotics were generally beneficial in treatment and prevention of gastrointestinal diseases. Efficacy was not observed for Traveller’s Diarrhoea or Necrotizing Enterocolitis or for the probiotic species L. acidophilus, L. plantarum and B. infantis.

The review question was broadly stated. Major databases were searched but limited efforts were made to search grey literature sources. Funnel plot suggested evidence of publication bias (some relevant papers could have been missed). Review processes were conducted by one reviewer which raised the possibility of error and bias. Study quality was assessed using the Jadad scale; most studies were reported to have been high quality. Random-effects meta-analysis was used to incorporate moderate heterogeneity across studies.

The conclusions reflect the evidence presented and are likely to be reliable, but the potential for publication bias and error and bias in the conduct of review processes should be considered.

Practice: The authors stated that the type of disease and probiotic species (strain) should be considered when choosing to use probiotics in the treatment or prevention of gastrointestinal disease.

Research: The authors stated that additional studies across gastrointestinal diseases were needed to assess which disease conditions respond to L. plantarum. Future studies should also compare different strains of L. acidophilus, and report adverse effects of probiotics. Also, future research should perform fecal samples before and after treatment to distinguish microbial changes.

National Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant.