The Journey to a Successful Shiny App Submission đź›°
Teal
- “Challenge study results delivery: from paper to interactive” Webinar
- “Gaining efficiency in the study results delivery internally & externally”
- “Challenging the way analyses are delivered”
FDA Uses Shiny
Shiny is Transforming Clinical Workflows
- Accelerating internal decisions
- Efficient review of clinical data
- Intuitive interfaces to complex statistical methods and pipelines
How can Shiny bring similar benefits to a clinical submission package?
R Submissions Working Group
- Cross-industry working group with members from life sciences organizations and regulatory agencies
- Evaluate clinical submission packages leveraging R for analysis programs and results
- Identify potential gaps and solutions using targeted pilots that mimic a clinical submission process
- Open-source: All materials available on GitHub!
Foundation - Pilot 1 Submission
- November 2021 - Successful R-based Test Package Submitted to FDA
- Adhered to Electronic Common Technical Document (eCTD) specifications
- Analysis Data Reviewer Guide
- R Package converted to text file using
{pkglite} - ADaM datasets
- TLF output
Technical Overview
- Application components created with Shiny Modules
- Assembled as an R package using
{golem} - Package dependencies managed by
{renv}ensuring reviewers can replicate R environment used to build application - Version control and collaborative development with Git and GitHub
- ADRG specifying precise instructions to execute application on a local machine with R available - created with Quarto!
With Great Power…
- Dynamically updates p-values based on filtered observation
- Potential for confusion and misinterpretation
- Compromise: Allow filtering within visualization module only
Acknowledgements
- Gregory Chen
- Hye Soo Chen
- Robert Devine
- Ning Leng
- Tadeusz Lewandowski
- Edgar Manukyan
- Eli Miller
- Joseph Rickert
- Paul Schuette
- Heng Wang
- Yilong Zhang
Thank You!
