Keywords: Historical Data, Bayesian methods
Room: Madison A
Historical data are increasingly used to expedite drug development process or enhance strength of evidence. They can be used to partially substitute or supplement a control arm in early phase or confirmatory trials, establish a reference threshold value for comparison in single- arm trials, create a synthetic control using patient level data, or extrapolate to another population (eg, from adult to pediatric). Bayesian approach is considered as the default one but frequentist approach can also be applied. In this session, any type of experiences of historical data applications in drug development are to be shared among participants. Statistical methodology development is also welcome. The chair of the session will prepare some materials to facilitate the discussion.