Keywords: Advanced Visual Analytics, Visualizations, Graphics, R-Shiny, Dynamic graphics
With the virtual tsunami of data being generated these days, not just in the pharmaceutical / healthcare industry, but in many other business sectors, it is a great time to be a statistician! However, methods that may have worked well to explain data in the past are today, not fit for purpose. For example, for the past few decades, Novartis has used SAS to pre-specify and produce hundreds of static tables, listings and figures to report out and submit our clinical trial data. In almost all cases, teams must produce additional static output to explore subgroups of interest or help explain unexpected results. With the advent of R – Shiny, we are now shifting our focus to developing dynamic displays of our data, known as Shiny Applications or Apps, that engage our statisticians, clinicians and clinical team members to interactively interrogate the data collaboratively and ultimately make better inferences and decisions from our trials.
Novartis has invested in the potential of R – Shiny with the intent to scale its use across the entire drug development portfolio. This talk will describe our experiences with R – Shiny and showcase some of our success stories and describe the challenges we have faced along the way. Finally, it should be stressed that creating effective Shiny Apps requires thought, rigorous planning, as well as adherence to strong graphical principles. By doing so, one maximizes the impact of their work, whether it is static or dynamic in nature.