This workshop is given on behalf of the Statistical Methods Group.

Workshop type: Training.

Objective: To train reviewers and editors in some basic concepts of meta-regression.

Summary:
Meta-regression is an extension to meta-analysis, and a generalization of subgroup analyses, that can be used to investigate heterogeneity of effects across studies.  It examines the relationship between one or more study-level characteristics and the sizes of effect observed in the studies. Characteristics of studies might be, for example, aspects of the interventions, the settings or the designs (such as geographical location, dose of a drug, or length of treatment).  The sizes of effects are the usual measures available for a meta-analysis, such as odds ratios or differences in means.

The workshop will provide an introduction to the method of meta-regression. We will consider when it might be use ful for application in a Cochrane review and will illustrate it using several examples. We will discuss, in particular, the limitations of meta-regression, including problems associated with looking at characteristics of participants, issues of power, the risk of false-positive results and availability of software. There will be ample opportunity to discuss the issues raised.

Intended audience: (i) Review authors; (ii) editorial staff; (iii) methodologists.

Technical level expected of participants: Intermediate/Advanced.