Author: Joost van Rosmalen
Date: November 2016

This file contains the files of the simulation study in the paper 'Including historical data in the analysis of clinical trials: Is it worth the effort?' in Statistical Methods in Medical Research. The main file for running the simulations is AnalysisSimulation.r. This file calls the function in SimulationOneDataSet.r using the parallel package. The function SimulationOneDataSet.r generates a single data set and then applies all methods to it.  The output of AnalysisSimulation.r consists of the tables in the manuscript, which are printed on the screen using both a standard format and a LaTeX format.

To be able to run the code, the working directory has to be specified on line 38 of AnalysisSimulation.r. Also, the number of cores (for parallel computations) in line 77 should be set by the user.  JAGS should be installed.  With the current value of 75 cores the code can only be run on a server. The parameters of the simulation study should be specified from line 25 onwards in both AnalysisSimulation.r and SimulationOneDataSet.r. On a desktop computer, the program may require several weeks to run.

Please contact the author (j.vanrosmalen@erasmusmc.nl) for questions on how to use this code.