Nine RCTs (n=765) were included in the review.
Duration: 7 studies provided a measure of variance. A meta-analysis of these studies showed a reduction in the duration of diarrhoea of 0.7 days (95% confidence interval, CI: 0.3, 1.2) for those treated with Lactobacillus compared with placebo. A positive significant linear association between the logarithm of the Lactobacillus dose and reduction in diarrhoea duration was reported (P<0.01).
Frequency: 3 studies provided a measure of variance. A meta-analysis of these studies showed a reduction in the frequency of diarrhoea on day 2 of 1.6 fewer stools per day (95% CI: 0.7, 2.6) for those treated with Lactobacillus compared with placebo.
Volume: no data were found for changes in the volume of diarrhoea.
Subgroup analyses: the authors planned 7 a priori subgroup analyses. The results of two were reported, while a third was not undertaken due to the heterogeneity of the included studies.
Five studies carried out in developed countries showed a reduction in the duration of diarrhoea of 0.8 days (95% CI: 0.1, 1.5) for those treated with Lactobacillus compared with placebo. Six studies using live Lactobacillus preparations showed a reduction in the duration of diarrhoea of 0.8 days (95% CI: 0.3, 1.3 days) for those treated with Lactobacillus compared to placebo. Due to heterogeneity between the studies, the effects of different strains of Lactobacillus could not be determined.
Adverse reactions: overall, these were similar in both the control and treatment groups. Two studies reported a decrease in vomiting in children given Lactobacillus. One study reported myoclonic jerks in two patients, one in each arm of the trial.
Publication bias: the results of the tests were not statistically significant.