Meta-analysis is the statistical method for combining both qualitative and quantitative outcomes from multiple studies to draw statistically stronger conclusions. Researchers may want to use meta-analysis because of the greater statistical power it creates by means of having a larger number of subjects and results than any individual study.
- To develop estimate of the effect size based on multiple studies
- To establish statistical significance with studies that have inconsistent results
- To provide analysis of safety, efficacy, and benefits
- To compare subgroups of studies
X-meta is an open-sourced, well-documented and interactive toolbox for meta-analysis. There are three main components to this toolbox: An R package called XMETA, video tutorials and documentation for the package and an online analysis platform. XMETA package offers several functions for performing meta-analysis and visualizing outcomes, allowing users to conduct robust multivariate meta-analysis (mmeta), publication bias test (PB), outcome reporting bias test (ORB) and novel visualization tool (galaxy). Through the tutorials, reference documents and sample code, users can have a comprehensive exploration of the features found in XMETA and how they may apply to analytical work.
The web-based xmeta online platform supports efficient and accurate meta-analysis studies without requiring any programming skills. The xmeta platform mainly focuses on the meta-analysis by offering comprehensive methods for study-level summary data, with model diagnoses, data visualizations, sensitivity analyses, and an automated generated data analysis report. It features univariate, multivariate and network meta-analyses with different types of outcomes including continuous, binary and time-to-event outcomes. In addition, it contains advanced model diagnosis tools such as evaluation of publication bias and outcome reporting bias, in various types of meta-analysis models. Xmeta also allows the users to create all kinds of meta-analysis results plots, including forest plot, funnel plot, bubble plot, and galaxy plot [Chuan et al, 2020]. In addition, the platform can generate an automated dynamic data-analysis report for the users. This report contains visualization of the data, meta- analysis of average effect size, between-study heterogeneity, sensitivity analyses, evaluation of risk of publication bias, and etc.
Ref: Hong C, Duan R, Zeng L, Hubbard RA, Lumley T, Riley R, Chu H, Kimmel SE, Chen Y. Galaxy Plot: A New Visualization Tool of Bivariate Meta-Analysis Studies. American Journal of Epidemiology. 2020 Jan 13.
- Questions about the study you would like to conduct:
- How many outcomes do you potentially want to study?
- How many treatment options are you comparing?
- Do you want to conduct ARM-BASED meta-analysis or CONTRAST-BASED meta-analysis? [Ref: https://training.cochrane.org/resource/comparison-arm-based-and-contrast-based-approaches-network-meta-analysis]
- Prepare your data in pre-specified formats (please see our sample data/templates page [link to be added])
- Login to your own user profile
- Create a new project or continue working on an exciting project
- Upload dataset
- Ready to use the platform to conduct meta-analysis and create results/figures/report
The development of the xmeta online platform was divided into five main parts. The first part is the interface design by R shiny, which requires the R programming language. The second part is the backend computation. All of the meta-analyses are conducted on the backend with the well-built R packages such as meta, xmeta, mmeta. The next part is the database management, for which we use Elasticsearch and Kibana. Kibana also offers exploratory data analysis visualization to be embedded to the xmeta platform. Finally, we built all of the above parts as an interconnect technology stack architecture hosted by the Amazon Web Service (AWS).
Xmeta online analysis platform is a free web-based tool, which does not require the users to download and install locally.
The manuscript will be ready to be cited soon. [add link here once we have the manuscript upload to arXiv]