Tool-set to support Bayesian evidence synthesis. This includes meta-analysis, (robust) prior derivation from historical data, operating characteristics and analysis (1 and 2 sample cases). Please refer to Weber et al. (2021) for details on applying this package while Neuenschwander et al. (2010) and Schmidli et al. (2014) explain details on the methodology.
To get a quick overview over the package functionality, please refer
to the introductory vignettes. We recommend
starting with the Getting started with RBesT
(binary) vignette as a first read.
For a brief 15 minute introduction to using RBesT, have a look at
the ESMARConf 2023
tutoral.
The rbest-assistant AI skill helps assistants such as Claude, Posit
Assistant, GitHub Copilot or ChatGPT Codex guide RBesT workflows, from
MAP priors to operating characteristics. See the AI assistant
skill
article for installation. Then try:
Using the RBesT assistant skill: give me a short introduction to MAP priors. How can I use historical control data for a new trial? No analysis yet, just an overview. Then briefly tell me what else you can help me with in RBesT.
The agent's output is not validated by RBesT and can differ between
runs; check the code and results, and have any analysis or design
reviewed by a qualified statistician. The assistant is an experimental
feature and may change; watch for updates, and we welcome your feedback
via GitHub issues.