PhUSE Working Group RGvD - Meeting Minutes 2015-06-09
From PHUSE Wiki
- Sandy (Astra Zeneca)
- Kiran (Roche)
- Steve (Cytel)
- Ying (Genentech)
- Wei (Gilead)
- Janet (Merck)
- Ellen (Merck)
- Discuss SDRG template section 3
- Sandy met with Mat Bryant who heads up Best Practices. We are the only BP group that has started meeting so far. Mat will head up a group for lab challenges and the Trial Design group isn’t likely to happen this year as nobody has volunteered to lead.
- We are working on getting a Wiki site set up to post our internal notes and share documents.
- There is a PhUSE webinar on Thursday to give team updates.
- Section 3.1/Overview:
- Primary and secondary endpoints as separate sections? They are needed in ADRG. SDRG also has an area for this – is the duplication needed? Maybe, for medical reviewers using the SDTM data.
- Do these guides work for integrated data (ISS/ISE)?
- Include data flow diagrams for non-standard data flow (i.e. ADaM not created from SDTM, extension studies).
- Section 3.2/aCRFs:
- unique forms – annotating versus submitting (submit all and annotate just the unique)
- common to submit only the unique forms
- bookmarking domains and visits
- guidelines aren’t clear on headers to use
- refer to ‘topics’, so ‘domains’ may be discrepant
- order bookmarks alphabetically in the by-domain section
- use “domains” or “topics” for bookmark header?
- Suggest wording change in SDRG completion instructions
- Do we annotate by domain or by CRF topic/module?
- Suggest to use the form name and change header to “CRF Topics” (not “Domains”)
- Section 3.3/Subject Domains:
- Note the primary/secondary efficacy here – flag using different codes….P=primary, S=secondary? Would need footnote or key to identify the meaning of these codes.
3.3 SDTM Subject Domains
|Dataset - Dataset Label||Efficacy||Safety||Other||SUPP--||Related Using RELREC||Observation Class|
|AE – Adverse Events||X||Events|
|DM – Demographics||X||Special Purpose|
|DS – Disposition||P||X||Events|
|EX – Exposure||S||X||Interventions|
P=primary, S = secondary
- Observation class needed? Yes, to help identify structure, particularly for custom domains.
- Typo in instructions at top of page 9…2.2 should be 3.2. (Note to SDRG team)
- SUPPQUAL tables not needed?
- Check with team on reason for including them in the template.
- CBER wanted to see all SUPPQUAL in one place
- CBER data interpretation guidelines lists qualifiers with reason they’re included.
- If including, suggest to add a column to explain why they are mapped to SDTM.
- Reasons may be repeated, such as “data not available in standard variables”, “required for analysis”, etc.
|QNAM||Description QLABEL||Explanation for Inclusion|
|LBCVRESC||Character result in conventional units||Conventional units needed for additional analysis|
|LBCVRESU||Conventional unit||Conventional units needed for additional analysis|
|LBCVNRLO||Reference range lower limit in conventional units||Conventional units needed for additional analysis|
|LBCVNRHI||Reference ranges upper limit in conventional units||Conventional units needed for additional analysis|
- Sections 3.3.x:
- Derivations should be in the computational algorithms in define.xml only
- Unless there are special characters, images?
- Hyperlink from define.xml in that case to specific section of RG if possible
- Explanation for origin values – how they are determined, e.g. Assigned versus Derived.
- Would this be domain-specific or general? If general, then note in 3.1, not here.
- Large table or lot of detail for a dataset – put in an appendix and include a hyperlink (rather than within a section 3.3.x)
- Next meeting: 2015-06-23