Probiotics for treating acute infectious diarrhoea
April Padua-Zamora,Shelui Collinson,L. Dans,Stephen J Allen,A. Deans,Chao Li,G. Gregorio
DOI: https://doi.org/10.1002/14651858.cd003048.pub4
IF: 8.4
2020-12-08
Cochrane Database of Systematic Reviews
Abstract:Abstract Background Probiotics may be effective in reducing the duration of acute infectious diarrhoea. Objectives To assess the effects of probiotics in proven or presumed acute infectious diarrhoea. Search methods We searched the trials register of the Cochrane Infectious Diseases Group, MEDLINE, and Embase from inception to 17 December 2019, as well as the Cochrane Controlled Trials Register (Issue 12, 2019), in the Cochrane Library, and reference lists from studies and reviews. We included additional studies identified during external review. Selection criteria Randomized controlled trials comparing a specified probiotic agent with a placebo or no probiotic in people with acute diarrhoea that is proven or presumed to be caused by an infectious agent. Data collection and analysis Two review authors independently applied inclusion criteria, assessed risk of bias, and extracted data. Primary outcomes were measures of diarrhoea duration (diarrhoea lasting ≥ 48 hours; duration of diarrhoea). Secondary outcomes were number of people hospitalized in community studies, duration of hospitalization in inpatient studies, diarrhoea lasting ≥ 14 days, and adverse events. Main results We included 82 studies with a total of 12,127 participants. These studies included 11,526 children (age < 18 years) and 412 adults (three studies recruited 189 adults and children but did not specify numbers in each age group). No cluster‐randomized trials were included. Studies varied in the definitions used for "acute diarrhoea" and "end of the diarrhoeal illness" and in the probiotic(s) tested. A total of 53 trials were undertaken in countries where both child and adult mortality was low or very low, and 26 where either child or adult mortality was high. Risk of bias was high or unclear in many studies, and there was marked statistical heterogeneity when findings for the primary outcomes were pooled in meta‐analysis. Effect size was similar in the sensitivity analysis and marked heterogeneity persisted. Publication bias was demonstrated from funnel plots for the main outcomes. In our main analysis of the primary outcomes in studies at low risk for all indices of risk of bias, no difference was detected between probiotic and control groups for the risk of diarrhoea lasting ≥ 48 hours (risk ratio (RR) 1.00, 95% confidence interval (CI) 0.91 to 1.09; 2 trials, 1770 participants; moderate‐certainty evidence); or for duration of diarrhoea (mean difference (MD) 8.64 hours shorter, 95% CI 29.4 hours shorter to 12.1 hours longer; 6 trials, 3058 participants; very low‐certainty evidence). Effect size was similar and marked heterogeneity persisted in pre‐specified subgroup analyses of the primary outcomes that included all studies. These included analyses limited to the probiotics Lactobacillus rhamnosus GG and Saccharomyces boulardii. In six trials (433 participants) of Lactobacillus reuteri, there was consistency amongst findings (I² = 0%), but risk of bias was present in all included studies. Heterogeneity also was not explained by types of participants (age, nutritional/socioeconomic status captured by mortality stratum, region of the world where studies were undertaken), diarrhoea in children caused by rotavirus, exposure to antibiotics, and the few studies of children who were also treated with zinc. In addition, there were no clear differences in effect size for the primary outcomes in post hoc analyses according to decade of publication of studies and whether or not trials had been registered. For other outcomes, the duration of hospitalization in inpatient studies on average was shorter in probiotic groups than in control groups but there was marked heterogeneity between studies (I² = 96%; MD ‐18.03 hours, 95% CI ‐27.28 to ‐8.78, random‐effects model: 24 trials, 4056 participants). No differences were detected between probiotic and control groups in the number of people with diarrhoea lasting ≥ 14 days (RR 0.49, 95% CI 0.16 to 1.53; 9 studies, 2928 participants) or in risk of hospitalization in community studies (RR 1.26, 95% CI 0.84 to 1.89; 6 studies, 2283 participants). No serious adverse events were attributed to probiotics. Authors' conclusions Probiotics probably make little or no difference to the number of people who have diarrhoea lasting 48 hours or longer, and we are uncertain whether probiotics reduce the duration of diarrhoea. This analysis is based on large trials with low risk of bias.
Medicine