How to Simulate Realistic Survival Data? A Simulation Study to Compare Realistic Simulation Models

Abstract

In statistics, it is important to have realistic data sets available for a particular context to allow an appropriate and objective method comparison. For many use cases, benchmark data sets for method comparison are already available online. However, in most medical applications and especially for clinical trials in oncology, there is a lack of adequate benchmark data sets, as patient data can be sensitive and therefore cannot be published. A potential solution for this are simulation studies. However, it is sometimes not clear, which simulation models are suitable for generating realistic data. A challenge is that potentially unrealistic assumptions have to be made about the distributions. Our approach is to use reconstructed benchmark data sets %can be used as a basis for the simulations, which has the following advantages: the actual properties are known and more realistic data can be simulated. There are several possibilities to simulate realistic data from benchmark data sets. We investigate simulation models based upon kernel density estimation, fitted distributions, case resampling and conditional bootstrapping. In order to make recommendations on which models are best suited for a specific survival setting, we conducted a comparative simulation study. Since it is not possible to provide recommendations for all possible survival settings in a single paper, we focus on providing realistic simulation models for two-armed phase III lung cancer studies. To this end we reconstructed benchmark data sets from recent studies. We used the runtime and different accuracy measures (effect sizes and p-values) as criteria for comparison.