Overview analysis issues using combined androgen deprivation overview analysis as an example.

ABSTRACT

The validity of an overview analysis depends on addressing specific concerns, including identification of the trials, collection of the data, trial design heterogeneity, data incompatibilities, treatment arm comparability, trial maturity heterogeneity, data quality, funding, and administrative problems. This article reviews various issues in metaanalysis and overview analysis methodologies and then discusses these issues in the context of the combined androgen deprivation (CAD) overview analysis effort. Following small clinical series reports of promising efficacy for CAD, over 20 randomized clinical trials designed to confirm this result were initiated. Subsequently, an overview analysis was planned, resulting in an international effort to collect data from these confirmatory trials. The overview analysis was motivated in part by a realization that the size of the CAD benefit was smaller than originally thought and that the confirmatory trials were underpowered relative to revised estimates of benefit.