Attendees: Bharath, Eric, Ineborg, Heather, Jasmine, Rinki, Sharon, Srinivas, Lauren, Paul, Mike
Apologies: Zac, Karnika
- Welcome Elena Rantou from the FDA
- Collaboration Framework for interactive analysis POC project - an update based on discussions from last meeting
- Align with the Areas that the FDA currently focuses on - Sponsor representative to become a member of the PhUSE meetings
- Planning: example use case for interactive analysis - OSI listing - how to get connected with the FDA?
- Share with FDA how an interactive approach could make the data listing review easier
- Test out a framework for interactive approach
- FDA has been looking into interactive tools we need to make sure FDA are on the same page as us and maybe start with small projects
- Could we submit Spofire listings instead of Jreview as this is easy to link together
- Give a sense of what interactive tool can help in this use case
- Eric- Ways to share this information is by providing access to data and exploring Jreview definitions and then sending those to the FDA so they can import them into their Jreview environment
- Next step - Sharon to reach out to Paul Mike and Elaine to show how use cases
Discussion topic: what does traceability and reproducibility look like in interactive analysis?
- Eric- Jreview when anyone defines any reports they encourage to save the definition to be able to re execute against the same data as you will get the same results. Traceability- how did you get to that point? what did you do leading up to that?
- Jreview and SAS we do not get the code to be able to reproduce
- -Eric - The reproducibility is not looking at code but looking more from a testing perspective than a QA perspective.
- Sharon - Run the definition and get the same results there should be a way to show how the results are created Jreview
- Paul - Standard references now include the date accessed one of the problems with this is when using SAS you have to specify the version and any modifications which have been made. With R you can issue a command it will list all of the packages and the versions, so that you can archive that piece of information
- Sharon - The key is to talk about documentation and the results so that when the auditor comes you can show the results and how you came to gather those
- Paul- Reviewers were not able to replicate sponsors codes with sponsors programmes so an archivable with a list of commands would make things more reproducible
- Heather- Look at the bookmark functionality
- We need to capture in process- Systems, Data, and individual user is taking the in interactive visualisation to get to a point where they draw to a conclusion
- Invite more FDA members to join the next meeting
- Talk about the OSI Listing and how the interactive tool can help and invite more FDA members to the next meeting (Sharon to invite Roche colleague who will be presenting on OSI)
- Sharon to send Paul meeting objective so Paul can forward on meeting invites to FDA members
- Sharon and Lauren to work together to arrange the next meeting