Abstract:
|
An important task in drug development is to identify patients, which respond better or worse to an experimental treatment. While a natural question it is also a challenging task (EMA, 2019) as typically clinical trials are not specifically designed for the purpose (Senn, 2016). In recent years the literature on subgroup identification methods exploded (see e.g. Lipkovich et al 2017), with new methods and approaches appearing at a very high rate. With this project we would like to provide a benchmark framework to test and compare new methodologies on realistic simulated clinical trial data. Specific emphasis is placed on making the simulated test cases realistic and defining the performance metrics that are important for decision making in drug development (Loh, et al 2019). Initial results comparing some recent algorithms will be presented.
|