Meta-regression models describe associations between the summary effects and studylevel
data; that is, it describes only between-study and not between-patient variation. We
would use multilevel models, which boost the power of the analysis by sharing strengths
across subgroups for variables where it makes sense to do so, or subgroup analysis (with
random effects meta-analysis) to explore heterogeneity if there are a sufficient number of
studies. When the sizes of the included studies are moderate or large, each subgroup
should have at least 6 to 10 studies for a continuous study-level variable and a minimum
of 4 studies for a categorical study-level variable. These numbers serve as a rule of thumb
for the lower bound for number of studies that investigators would consider for a metaregression,
but power will vary according to the size and variability of the effect.