Keywords: meta-analysis, orphan disease, non-inferiority
Meta-analysis is useful in a regulatory setting for the estimation of a non-inferiority margin, the estimation of efficacy in subgroups, and as supporting evidence in in product development. Meta-analysis can provide information in an orphan disease setting, when the orphan designation is defined by a genetic mutation and the goal is an assessment of the efficacy of a treatment within the subgroup possessing the mutation of interest. Meta-analysis can also be used when designing non-inferiority studies in order to provide information for the specification of a non-inferiority margin. Methodology for combining studies is readily available; however, these techniques assume that a significant number of studies are available for the meta-analysis. Less attention has been given to the common problem where few studies are available and meta-analysis is needed to provide an overall estimate of treatment effect.
This talk focuses on the problems encountered when a meta-analysis includes a small number of studies. Statistical approaches for this situation are presented and examples of the use of meta-analysis in regulatory settings are provided.