Interesting science. But, I don't feel that the second is a legitimate use of meta-analyses. Due to publication bias, the conclusions it draws are only useful

Interesting science. But, I don't feel that the second is a legitimate use of meta-analyses. Due to publication bias, the conclusions it draws are only useful if the efficacy of probiotics is a forgone conclusion. Moreover: I have a huge question regarding their primary measure, specifically regarding their use of number from this study: https://www.ncbi.nlm.nih.gov/pubmed/14608267 https://www.ncbi.nlm.nih.gov/pubmed/14608267 Which show up as the 8th line of their risk-ratio calculation, but whose results have no relation to the values on that graph. Moreover, this isn't peer reviewed, as far as I could see.