Working at the intersection of Biostatistics and Software Engineering in a Pharma company


Increasingly biostatisticians in pharma companies would like to use R on a daily basis, e.g. the growing number of participants in R/Pharma conferences is one metric showing this trend. As R programs replace proprietary software in this regulated industry, the requirements for quality and reliability increase, e.g. validation of packages might be required. More importantly, programs used in multiple projects or over a longer time period with changing biostatisticians need to be maintainable, intuitive and reliable. Getting this done takes experience and time. In particular, it is challenging to do this as a side project next to other daily biostatistics work and without the right tools in place. Therefore forming a dedicated team, working together closely with the applied biostatisticians, methods experts and IT professionals, can help. We share our Statistical Engineering team experience in Roche, covering agile ways of working, resourcing, product prioritization, and open source collaborations. We believe that taking the Research Software Engineering aspects seriously and adopting an Open Source mindset is key to sustainably and reliably use R in Pharma.

Presented at 2022 Conference