USA – Meta-Analyses of Randomized Trials: FDA Offers Draft Guidance

The conducting of meta-analyses to evaluate safety questions presents a unique opportunity to detect and quantify the risk of a safety event and may help provide a more precise estimate of the risk of uncommon serious adverse events by combining information from multiple trials, the US Food and Drug Administration (FDA) said Tuesday as it released new draft guidance.

The 26-page draft defines meta-analysis as the “combining of evidence from relevant studies using appropriate statistical methods to allow inference to be made to the population of interest.”

The draft says meta-analyses are most useful in detecting and quantifying an increased risk over the background rate of a safety event. And with aggregating data, a sponsor may be able to provide a higher statistical power and more robust estimate than may be possible with an individual study.

But investigators must make important decisions on studies to include in a meta-analysis, whether they should be included based on prospective and objective criteria, how to deal with incomplete data, analyze the results and account for publication bias.

FDA Commissioner Scott Gottlieb said: “Our goal is to encourage more pooled studies to evaluate important safety events. Meta-analyses can be conducted in the pre-and post-market setting, and the recommendations in this draft guidance, when finalized, will apply to both. Sponsors apply the principles described in this draft guidance when planning and conducting a meta-analysis of randomized controlled trials to better assess the safety of a drug product.”…