A review of frequentist methods for combining results of series of trials

Background: A randomized controlled trial is considered the gold standard in clinical research, also in rare diseases. However, in small populations, single large-scale well-powered trials are often not possible. For valid decision-making, both on efficacy and on safety, evidence generated from series of trials could be exploited. Methods: We conducted a review over a five-year period between January 2009 and December 2013 to identify relevant methodology on combining results of series of trials. Results: Sixty-two papers were identified and summarized in this review. Its focus is on frequentist methodology. Most papers deal with meta-analyses on aggregated data. Only few papers discuss multivariate outcomes. We categorized the relevant methods according to the type of (meta-) analysis. Conclusions: Only a few papers dealt directly with series of trials in small populations. The results of the review lead to some directions for further investigation on evidence-based decision-making from a (small) number of trials in small populations.