Hi, my name is Dave Radtke, I co-chair the current working committee within CDISC focused on development of guidelines and standards for population PK and PK/PD datasets with a focus on the creation of ADaM standards for datasets. I will post presentations and discussion topics here from time to time for review and comments. Thank you in advance for your thoughts and input.
Dave
Much thanks…The work that the CDISC group does in this regard will be be extremely beneficial, and this forum can act as a great resource for bouncing ideas off of the broader community.
Notes from first meeting, 06-Apr-16
(sorry, I do not have permission to upload files yet, here is some summary slide content):
Opportunities for Exploration
- Rules
- How to implement ADDL/II
- Definitions
- Common Processes
- Sequence/order of data combination
- Exclusions?
- Early analysis
- Rules, guidelines, blinding
- How does that differ between companies
Additional Thoughts
-
Standard datasets are not intended to force a standard analysis, however, to restrictive of a format may be perceived as forcing an analysis preference
-
A possible solution would be to drive to a high level guidance, then move toward specific tool standards
-
Do standard tools need to adapt to a data standard as opposed to the data standard fitting all tools?
-
What does the agency(agencies) really need? How do they really review our data/analysis?
-
How do things differ between the FDA, EMA, PMDA, HCA, Other countries`
-
What are various companies currently submitting?
-
Standards
- Data standards (data types, rules around nomenclature, implementation) - Rules (data handling, exclusion, data correlation) - Workflow (may be difficult)
Moving forward
-
Team decided to meet bi-weekly
-
May try and divide efforts between ISOP groups and CDISC
-
The goal should be both a high level standard and specific ADaM standard files maybe more focused on a specific program’s requirements.
-
Need to look at how population PK program needs fit within existing ADaM standards
- i.e., what does an ADaM dataset look like compared to a standard NONMEM dataset