Tailor-made solutions for statistics and data management
At the Department of Clinical Reseach (DKF), the statistical tool R is being optimised for use in clinical research.
R is a programming environment for statistical calculations and graphics. The tool runs on different operating systems and is licensed under an open source license allowing userd to extend and modify the software. Developed in 1992 for statistical computing tasks, R has since become a standard application with a large fan base. This community continuously provides newly developed statistical solutions for R online.
These so-called "packages" extend the scope of R's functions in order to analyse and present data more efficiently and more quickly using the latest methods. Among the active users of R are Thomas Fabbro, Milica Markovic and Patrick R. Wright. They have programmed two innovative R packages: one that enables an efficient and professional implementation of statistical methods for sample calculation and one that provides a simplified interface between the study database and the evaluation of results.
Sample size calculation with sse
In the planning of a clinical trial, the calculation of the sample size is a central step with far-reaching consequences. Nevertheless, very little time is usually available to evaluate different methods and to study the sensitivity of the estimate to the assumptions made. This is where the R package "sse" for "Sample Size Estimation" comes in. It allows to evaluate the sample size calculations for different scenarios. All methods available in R can be used. In addition, the framework for complex study designs allows to efficiently estimate the sample size by means of so-called "resampling" methods. Thomas Fabbro received a grant from the Swiss Clinical Trial Organisation (SCTO) for the development of sse.
![Graphik sse R-Paket](/fileadmin/_processed_/e/d/csm_Graphik_sseRPaket_199a7a9f9c.png?1614730943)
Instead of just a naked number, the R packet "sse" additionally provides a figure that illustrates how sensitive the calculation is with regard to the assumptions made.
Efficient data management with secuTrialR
"secuTrialR" is a joint project of Patrick Wright, Data Scientist at the DKF, Milica Markovic, former Data Scientist at the DKF and Software Engineer at Paranor Inc. today, as well as Alan Haynes, Senior Statistician at the Clinical Trial Unit Bern. As the name suggests, the package was developed, as the name suggests, for data export from "secuTrial®", a data management system for clinical studies. The system is suitable for complex data sets and is recommended by the DKF as a standard solution for data management. Using this new package, the research data recorded in the "secuTrial®" database are read directly into the "R" program and then directly evaluated. Through the standardisation of redundant processes, the technical effort is reduced and more time is left for content-related issues.
Through the standardisation of redundant processes, the technical effort is reduced and more time is left for content-related questions.
Milica Markovic, Software engineer at Paranor Inc., and Patrick R. Wright, Data Scientist at the DKF
![Graphik SecuTrialR](/fileadmin/_processed_/8/b/csm_Graphik_SecuTrialR_24aa8dbeb2.png?1614730944)
The secuTrialR package facilitates the visualisation of study recruitment data. The shown example can be illustrated with two lines of programming code and is therefore easily accessible for both experts and R novices.
Thomas Fabbro, PhD, Head of Research Infrastructure at the DKF
Useful links
Department of Clinical Research, October 2019, updated January 2020