Published online Aug 28, 2020. doi: 10.13105/wjma.v8.i4.345
Peer-review started: May 20, 2020
First decision: July 4, 2020
Revised: July 4, 2020
Accepted: August 15, 2020
Article in press: August 15, 2020
Published online: August 28, 2020
Meta-analysis is an important statistical tool, and it is often used to solve clinical problems. However inevitably when conducting a meta-analysis, the included studies often have heterogeneity. This paper suggests the inclusion of relevant background data or contextual variables into the model. The contextual variables are those variables not explicitly measured in the studies included in a meta-analysis; thus, these must be very well-described and justified as parameters for analyses.
Core Tip: This letter call for the use of contextual variables, that are typically not in use for covariate analyses. Contextual variables are introduced and defined as variables not immediately/directly measured by the original studies in the meta-analysis but rather can be estimated knowing the background of each study. For example in a meta analysis of clinical trails one might want to adjust for studies from high income vs low income countries or studies that were funded vs independent.