A meta-analysis of treatment of infantile diarrhea with bifid triple viable bacterial tablet / 中华儿科杂志

ABSTRACT

<p><b>OBJECTIVE</b>To assess the efficacy and safety of bifid triple viable bacterial tablet in treatment of infantile diarrhea.</p><p><b>METHODS</b>According to the requirements of Cochrane systematic review, a thorough literature search was performed among Chinese Digital Hospital Library (www.chkd.cnki.net) and Chinese Biomedical Literature Disk Database (CBMdisk). A meta analysis was performed on a total of 1326 patients involved in 12 papers which met the inclusion criteria.</p><p><b>RESULTS</b>Publication bias analysis showed that the funnel plot was symmetrical. Test for heterogeneity showed that the groups treated with bifid triple viable bacterial tablet and antibiotics or anti-viral agents or placebo control had clinical homogeneity and statistical homogeneity (P = 0.66, 0.67, 0.85, respectively, I(2) = 0%), which allowed to use fixed effect model analysis. The bifid triple viable bacterial tablet and Smecta did not have statistical homogeneity (P = 0.02, I(2) = 70.9%), therefore random effect model analysis was applied. Incorporation analysis showed that in comparison of the bifid triple viable bacterial tablet versus antibiotics or anti-viral agents or placebo control, the odds ratios were 5.34, 4.74 and 6.43, respectively, and 95% CIs were [2.81, 10.16], [2.47, 9.09], and [2.61, 15.83], on test for overall effect, Z = 5.11, 4.67 and 4.04, P < 0.00001. In the forest plot the 95% CI horizontal line of incorporation odds ratio droped to the right side of the vertical line indicating the border of effectiveness. However, no statistically significant difference was found between bifid triple viable bacterial tablet and Smecta.</p><p><b>CONCLUSION</b>The clinical evidences available so far indicated that treatment of infantile diarrhea with bifid triple viable bacterial tablet is safe and effective although rigorously designed large sample size randomized double blind clinical trials are required to further demonstrate and support the conclusion.</p>