Article type
Year
Abstract
Background:
Researchers often wish to identify which individuals benefit most (or least) from particular interventions and this idea provides the basis for 'stratified medicine'. However, single studies are typically underpowered for exploring whether participant characteristics, such as age or disease severity, determine an individual's response to treatment. A meta-analysis can provide greater power to investigate relationships, or 'interactions', between such characteristics and the effect of an intervention. Whilst individual participant data (IPD) provides the most power and analytical flexibility, aggregate data (AD) can also often be used. However, approaches to the analysis, presentation and interpretation of interactions vary widely.
Objectives:
In this workshop we aim to demystify interactions in meta-analysis and show how they can be explored using both AD and IPD. In particular, participants will:
1) understand the purpose of interaction analysis in trials and meta-analysis, and its strengths and limitations;
2) acquaint themselves with some important subtleties, such as the distinction between 'within-trial' and 'across-trial' effects;
3) extract and calculate a simple interaction effect using AD;
4) explore and critique the various approaches seen in the literature, and discuss best practice.
Much of the material and examples are taken from Fisher et al, BMJ 2017;356:j573.
Description:
The 90-minute workshop will consist of short slide presentations, group discussion and practical activity. We will begin by considering subgroups and interactions within a single randomised trial. Using real examples, participants will discuss the presentation and interpretation of results; in particular, what can and cannot be concluded from the given data. Participants will then learn how interactions from multiple trials may be pooled using a standard meta-analysis approach, which is the basis of 'within-trials' IPD interaction testing. We will demonstrate how this can often be done with AD using examples from the literature. Working in small groups, participants will then perform a simple AD interaction analysis, learn how to do this in Stata, and discuss the results and interpretation. Finally, we will explore and critique alternative approaches seen in the literature, and discuss the situations in which they are appropriate.
Researchers often wish to identify which individuals benefit most (or least) from particular interventions and this idea provides the basis for 'stratified medicine'. However, single studies are typically underpowered for exploring whether participant characteristics, such as age or disease severity, determine an individual's response to treatment. A meta-analysis can provide greater power to investigate relationships, or 'interactions', between such characteristics and the effect of an intervention. Whilst individual participant data (IPD) provides the most power and analytical flexibility, aggregate data (AD) can also often be used. However, approaches to the analysis, presentation and interpretation of interactions vary widely.
Objectives:
In this workshop we aim to demystify interactions in meta-analysis and show how they can be explored using both AD and IPD. In particular, participants will:
1) understand the purpose of interaction analysis in trials and meta-analysis, and its strengths and limitations;
2) acquaint themselves with some important subtleties, such as the distinction between 'within-trial' and 'across-trial' effects;
3) extract and calculate a simple interaction effect using AD;
4) explore and critique the various approaches seen in the literature, and discuss best practice.
Much of the material and examples are taken from Fisher et al, BMJ 2017;356:j573.
Description:
The 90-minute workshop will consist of short slide presentations, group discussion and practical activity. We will begin by considering subgroups and interactions within a single randomised trial. Using real examples, participants will discuss the presentation and interpretation of results; in particular, what can and cannot be concluded from the given data. Participants will then learn how interactions from multiple trials may be pooled using a standard meta-analysis approach, which is the basis of 'within-trials' IPD interaction testing. We will demonstrate how this can often be done with AD using examples from the literature. Working in small groups, participants will then perform a simple AD interaction analysis, learn how to do this in Stata, and discuss the results and interpretation. Finally, we will explore and critique alternative approaches seen in the literature, and discuss the situations in which they are appropriate.