Background:Clinical decision-making requires synthesis of evidence from literature reviews on a specific theme. Meta-analysis and Network meta-analysis are two important tools that provide an efficient and comprehensive way of solving conflicts on one health problem statistically. Researchers from different countries are using English, which is generally considered as the universal language of science when publishing studies. A systematic review of empirical studies found no differences between summary treatment effects in English-language restricted meta-analyses and non-English inclusive meta-analyses. To date, the effect of excluding non-English studies of network meta-analysis haven’t been studied yet.

Objectives:We aim to assess the impact of excluding non-English studies on the network geometry and conclusions in Cochrane network meta-analysis.

Methods:We searched Cochrane Library for network meta-analysis. We extracted the original data from the attached forest plots and also the publication language of included studies. We then re-conducted network meta-analysis with same approach to the original Cochrane network meta-analysis after excluding non-English studies. We compared the network geometry and results between network meta-analysis including all studies and those only including English studies. We will report the change of network plots and results, and also calculate the relative ratio of network estimates between including all studies and including only English studies. All statistical analysis will be done using Stata 15.1 software.

Results:This study is ongoing and results will be presented at the Cochrane Colloquium as available.

Conclusions:This study is ongoing and results will be presented at the Cochrane Colloquium as available.

Patient or healthcare consumer involvement:No patient or healthcare consusmer is invovled in this study.