A Fast, Flexible Simulation Framework for Bayesian Adaptive Designs-The R Package BATSS

Area of Science:

• Biostatistics
• Clinical Trial Design
• Computational Statistics

Background:

• Bayesian adaptive designs offer flexibility in clinical trials but have lacked accessible software tools for statisticians.
• Efficient simulation is crucial for evaluating the operating characteristics of complex adaptive trial designs.

Purpose of the Study:

• To introduce the Bayesian Adaptive Trials Simulator Software (BATSS) package for the R statistical environment.
• To provide a flexible and efficient tool for simulating Bayesian adaptive designs in clinical trials.

Main Methods:

• Development of the BATSS software package in R.
• Simulation of Bayesian adaptive designs with various primary outcome distributions (normal, binary, Poisson, negative binomial).
• Incorporation of common adaptations: stopping for efficacy/futility and Bayesian response-adaptive randomization.

Main Results:

• BATSS facilitates the definition and evaluation of operating characteristics for diverse Bayesian adaptive designs.
• The package supports common adaptive strategies, including early stopping and response-adaptive randomization.
• Features include (Integrated Nested) Laplace approximations, parallel processing, and covariate adjustment.

Conclusions:

• The BATSS package significantly enhances the practical application of Bayesian adaptive designs in clinical trials.
• This software empowers statisticians to more readily design and analyze complex adaptive trials.
• BATSS promotes the use of advanced adaptive methodologies through accessible simulation capabilities.