Use R! An Introduction to Meta-Analysis with R

Article type
Authors
Sommer H1, Schwarzer G2, Rücker G3
1Cochrane Germany
2Institute for Medical Biometry and Statistics, Freiburg
3Medical Center - University of Freiburg
Abstract
Objectives: Drawing on our forthcoming book 'Meta-Analysis with R', part of the Use R!-series by Springer, this course will:
- provide a brief introduction to R (getting started: R as a calculator; getting help; importing data from RevMan; R packages for meta-analysis).
- present basic analyses (meta-analysis with continuous/binary outcomes; forest plot; fixed-effect and random-effects model).
- describe methods to evaluate bias (funnel plot; tests for funnel plot asymmetry; trim-and-fill method).
Description: Review Manager (RevMan) is the main software used to prepare and maintain Cochrane Reviews but its statistical methods are limited. Accordingly, review authors have to conduct more advanced techniques using external software. The rapid development of the open-source software R makes it an ideal environment for keeping up with the latest methodological developments in meta-analysis.
For each of the topics listed above, a short presentation will be followed by hands-on application using publicly available R packages. At the end of the workshop, participants should be able to import data from RevMan 5, perform necessary calculations, and produce appropriate plots and summary tables using R.
This workshop will be based on our book 'Meta-Analysis with R' which is likely to be available by summer 2015. Participants should have their own laptop with a current installation of RStudio (http://www.rstudio.org/) and have already installed R packages meta and metafor.