The R programming language is becoming increasingly important for pharmaceutical companies in the planning and analysis of clinical trials. Due to the wide range of modern statistical methods and the large and very active community that is constantly developing and publishing new methods, R is very attractive both technologically and professionally. However, with regard to the qualification and validation of applications required in the GxP environment, the hurdles are much higher compared to standard software. While, for example, standardized IQ/OQ (Installation Qualification / Operational Qualification) procedures for the respective operating environment are included in the scope of functions, with R a large number of dependencies on the operating system, other R packages and their versions must be taken into account for the successful execution of the unit tests supplied. It is also important to ensure that the high regulatory security requirements can be met in the day-to-day work of users.
The R environment was designed with three levels: While users are free to select and use R packages on the development environment, the test and production environments have been classified and validated as GxP-relevant. The definition of clearly defined and documented deployment processes ensures regulated operation: In order to install new packages or versions in the production environment, these are first evaluated in the test environment according to standardised guidelines and then transferred to the production environment as part of a change procedure. Changes to the R installation itself and the available packages are not possible in the test and production environment for standard users, but are reserved for a special user, the super user. Used packages are stored in a local repository using the miniCRAN package. In combination with the Checkpoint package, all packages are assigned to a date (date snapshot). This means that historical program executions can also be reproduced, as they are based on a defined date snapshot.
HMS Analytical Software GmbH
Grüne Meile 29
69115 Heidelberg
Germany