ECRIN hosted on the 6th of December the e-workshop “Statistical environment for in-silico trials”. The workshop was organised as part of the H2020 SIMCor project and was attended by 50 participants, from the major EU-funded in-silico projects currently running (SIMCor, In Silico World, SimInSitu, SimCardioTest). The objective of the workshop was to specify the requirements for the implementation of an R-statistical analysis environment for planning and managing in-silico trials and to explore implementation strategies within the project. Experts from in-silico research, as well as experienced biostatisticians, data managers and data scientists, discussed potential issues to overcome and desired functionality.
An in-silico trial refers to the use of individualised computer simulations in a cohort of patients during the development or regulatory evaluation of a medicinal product, medical device, or medical intervention. Different tools have been developed to support statistical management of in-silico trials (e.g., QSP, Simulo, Hect, InSilico trial platform, UISS). The requirements for virtual cohort generation and in-silico trials defined by the SIMCor project can only be partially fulfilled by the existing tools. In order to cover the full cycle of development in the SIMCor project, a need was identified, to implement a more flexible and openly available statistical environment for the analysis of in-silico trials. Thus, it was decided to explore the use of free and open-source software for statistical planning and analysis, such as R and RStudio. The R-statistical environment provided for the SIMCor project should cover the following functionalities: specification of trial design and of trial data structure, importing virtual cohort data to the R-environment, analysis of in-silico trial, assessment of results, export of results to the computing environment and systematic variation of in-silico trials.