BayesCTDesign: An R Package for Bayesian Trial Design Using Historical Control Data.

This article introduces the R (R Core Team 2019) package BayesCTDesign for two-arm randomized Bayesian trial design using historical control data when available, and simple two-arm randomized Bayesian trial design when historical control data is not available. The package BayesCTDesign, which is available on CRAN, has two simulation functions, historic_sim() and simple_sim() for studying trial characteristics under user defined scenarios, and two methods print() and plot() for displaying summaries of the simulated trial characteristics. The package BayesCTDesign works with two-arm trials with equal sample sizes per arm. The package BayesCTDesign allows a user to study Gaussian, Poisson, Bernoulli, Weibull, Lognormal, and Piecewise Exponential (pwe) outcomes. Power for two-sided hypothesis tests at a user defined alpha is estimated via simulation using a test within each simulation replication that involves comparing a 95% credible interval for the outcome specific treatment effect measure to the null case value. If the 95% credible interval excludes the null case value, then the null hypothesis is rejected, else the null hypothesis is accepted. In the article, the idea of including historical control data in a Bayesian analysis is reviewed, the estimation process of BayesCTDesign is explained, and the user interface is described. Finally, the BayesCTDesign is illustrated via several examples.

Bayesian clinical trial design using Markov models with applications to autoimmune disease.

Immune Thrombocytopenia is an autoimmune disease associated with bleeding that is treated by increasing the platelet count to a level where the chance of uncontrollable bleeding is low. Failure occurs when platelet counts are not raised sufficiently (initial failure), or when high platelet counts are not maintained after initial success (relapse). In this paper, we propose a Bayesian clinical trial design that uses a Markov multistate model along with a power prior for the parameters which incorporates historical control data to estimate transition rates among two randomized groups as defined by the model. A detailed simulation is carried out to examine the operating characteristics of a trial to test whether a new treatment reduces the relapse rate by 40% relative to standard care when data from 60 historical controls treated with standard care is available. We also use simulated data to demonstrate effects of discordance between historical and randomized controls on the estimated hazard ratios. Finally, we use a simulated trial to demonstrate briefly what type of results the model can give and how those results can be used to address hypotheses regarding treatment effects. Using simulated data, we show that the model yields good operating characteristics when the historical and randomized controls are from the same population, and demonstrate how discordance between the control groups affects the operating characteristics.