This is one of four related projects designed to document the current standards and methods used in the meta-analysis of diagnostic tests, validate newly proposed methods, develop new statistical methods to perform meta-analyses of diagnostic tests, and then to incorporate these insights into computer software that will be available to all EPCs and others conducting reviews of diagnostic tests. The related projects can be accessed on the right side of this page.
Background: Meta-analysis, the quantitative synthesis of information from independent sources (studies), forms a cornerstone of evidence-based medicine (EBM). Indeed, by quantitatively summarizing the available evidence pertaining to precise clinical questions, meta-analyses can inform clinical practice at all levels of healthcare. Given the vital role of meta-analyses, it is therefore imperative that they are performed correctly, i.e., that appropriate statistical techniques are used to guide meta-analyses so that the conclusions are sound.
Despite this need, however, most meta-analysts are familiar with only basic meta-analytic methods, limiting the accuracy and power of analyses. This is primarily a consequence of software; meta-analysts tend not be programmers, and so use whatever is made available to them via graphics-driven (point-and-click) software. Unfortunately, the methods available in such software are usually dated; new methods are implemented in statistical programming languages such as R.
Approach: OpenMetaAnalyst is a free, cross-platform, open-source program for performing meta-analysis of efficacy and diagnostic test accuracy studies that aims to combine the ease-of-use of graphical, spread-sheet driven software with more advanced analytic options. More specifically, the software will include methods for performing Bayesian meta-analysis, multivariate meta-analysis and network meta-analysis in addition to all standard fixed and random effects methods. OpenMetaAnalyst features a modern graphical user interface (GUI), but interfaces with R “under the hood”. Leveraging the R environment allows us to exploit the wealth of existing meta-analytic methods already developed and maintained by the meta-analysis community as well as the superb graphical capabilities of the R environment that includes resources for user customization of graphics. Additionally, OpenMetaAnalyst provides an API for statisticians-methodologists to "plug-in" their new methods into the system; these methods will then be made available to end users via the graphical user interface (GUI) automatically.
The project Web site is at http://www.cebm.brown.edu/openmeta. Please contact Tom Trikalinos (firstname.lastname@example.org) with any questions. Different aspects of this program are currently supported by funding from AHRQ, grant number R01HS018574, and Task Order 2 to the Tufts EPC (contract number HHSA 290 2007 10055 I).