SEND Implementation User Group Minutes 2016-07-25

From PHUSE Wiki
Jump to: navigation, search


When: 2016-07-25, 10:00am-11:00am EST
Place: Telecon

For other minutes, see the SEND Implementation User Group Minutes page.

Logistics

Attendance

Participant Attended
Troy Smyrnios X
Audrey Walker
Bill Houser
Brian Argo
Carrie Neeley
Cheryl Sloan
Christy Kubin
Dan Potenta
Debra Oetzman X
Donna Danduone
Frederic Mura
Jamie Megna
Jennifer Feldmann
Jeff Foy
Kathy Powers
Kev Martin
Linda Hunt
Lou Ann Kramer
Louis Norton
Lynda Sands
Mike Wasko X
Pam Hills-Perry
Peggy Zorn
Sarah Obbers
SJ
Sue DeHaven
Wenxian Wang


Meetings

Next meeting: 2015-08-22 10-11am EST

Meetings: 4 week schedule

Minutes

General

Forum

Links:

Review Forum

  • Question on differing STRESU in same TEST
    • (ACTION ITEM - Troy) Add wiki item for multiple STRESU in TEST issue
  • Question on timelines and GLP
    • Stock answers, mention of FAQ for first part, and second part will be answered when FAQ is updated
  • Question on PC/PP wrt analyte
    • Gitte's answer answers it
    • Lingering issue with ambiguity of the subject to begin with - potential picture to help make it clearer?
    • (ACTION ITEM - Debra) - Draft up a picture that could make it clearer
  • Question on AUC codes - OK
  • Question on PC BLFL and ORRES/STRESC - OK
  • Question on TFRESCAT, etc. with metastasis in SUPPTF - answered
  • Question on VISITDY, etc. for 3.1 LB variables - OK
  • Report from Bill on question to edata for text LBORNRLO/HI:

What data type(s) will you accept for LBORNRLO and LBORNRHI in SEND dataset submissions? FDA validation rule FDAN032 states “Variable Data Types in the dataset should match the variable data types described in SEND.” I see that the SEND IG 3.0 specifies that they should be numeric, but this is in conflict with the base SDTM model which says they should be text. The CDISC SEND team recognizes this as an error in SEND IG 3.0 and is planning to make these variables text in SEND IG 3.1 to comply with the SDTM base model.

The answer was: You can set up those variables as text and then explain it in the study data review guide.

    • (ACTION ITEM - Troy) Create FAQ Item for textual LBORNRLO and LBORNRHI example
  • Question on FC (collected daily, reported weekly) - Answered
  • Question on DRVFL - OK
  • Question on timing of 3.1 - OK
  • Question on extensible terms warning - OK

Review News Page

SEND Implementation News

  • Carryover*
    • Concept of shared industry calendar
      • Laura has a google calendar and is curating
      • Might be issue on ownership (plenty can update, but still one owner)
      • Ready for mass use
      • Would need steps for people who aren't familiar with
      • (ACTION ITEM - Laura) Set public settings?

Wiki

  • Tumor.xpt item
    • Current wiki text says not required
    • Probably is still desired if not required by FDA
    • (ACTION ITEM - Troy) Get core team meeting item added for tumor.xpt expectancy
  • Handling of SEND in Study Documentation
    • Current text is now wrong, after some paradigm shifts/discussions/etc.
      • SEND datasets not done under the time span of the GLP study
      • but if datasets are used for interpretation, then they are part of the GLP study, and thus need to abide by all that entails
      • sponsor responsible for ensuring that the datasets accurately represent the study
      • Text from Debra:
        Appropriate verifications need to be done to ensure that data in the SEND datasets are an accurate representation of study data. A positive statement needs to be included in this section. In the event of a directed audit of the data set integrity, the actual verification documentation would most likely be helpful to the auditor.
        An example of a statement is:
        “Data in the SEND datasets are an accurate representation of the data for Study No. 12345. Any differences between the data sets and the report are described in section 6.2. Verification procedures and documentation supporting this are available upon request.”
    • If..
      • a) the datasets are used as basis of analysis and/or study decisions, then GLP, then GLP treatment.
      • b) If not, then do not necessarily need hardcore GLP/validation treatment. This is also the case when the files are generated as a post-reporting process (e.g., legacy studies, years later, etc.)

Resulting draft pieces:

  • Should you consider it GLP?
    • Are the datasets used to make study decisions / used to generate the tabulations on a GLP study?
      • Then should be considered GLP
      • If you have validated the software creating datasets in general, then you're good
    • If not (e.g., generation of SEND files: after finalization of study / outside of reporting process / not used for analysis / generating ahead of time for later package), then it does not fall under GLP regulations
  • What can you expect the FDA to expect?
    • In every submission, methods should be in place to check validity and accuracy of the SEND datasets (validating your software/process is one way of ensuring this, although validation is not necessarily required), and the results from that verification need to be available upon Agency request.
    • Additionally, a statement should be placed in the SDRG to cover this, for example:
“Data in the SEND datasets are an accurate representation of the data for Study No. 12345. Any differences between the data sets and the report are described in section 6.2. Verification procedures and documentation supporting this are available upon request.”
  • (ACTION ITEM - Troy) Draft GLP question thing


  • Handling of SEND in Study documentation:
    • Maybe add thing about whether it's GLP straightaway
      • Add the two draft pieces above
    • Protocol
      • General gist is still good (good to list as a deliverable, from a statement of work perspective)
      • ADD: If GLP does not apply (e.g., (a) not a GLP study or (b) the SEND datasets will not be used for study decisions or to generate the tabulations), then include a statement such as "SEND datasets will not be subject to Quality Assurance review."
    • Post-finalization piece - nuke
    • Study director piece
      • caveatize with if GLP verbiage
      • if GLP, then SD is signing for them
      • if not GLP, then not signing
      • The Study Director signature on a study report indicates accountability for the validity of the content of the datasets, indicating responsibility for the datasets at the same level as any other part of the report.
      • Remove 2nd paragraph onward
  • In general, make the sections start with the non-GLP
  • Known Issues
    • 3.0 items given a heading to make it clear they are 3.0 (because they are still valid for a while after 3.1 is in place)
    • Future evaluation of where errata go...


Action Items

Responsible Task Timeframe
 ? Update pages discussing QA/Validation/SD (see above) Next meeting
Debra Draft up a picture that could make the PC/PP analyte situation clearer Next meeting
Troy Add wiki item for STRESU/TEST issue Next meeting
Troy Create FAQ Item for textual LBORNRLO and LBORNRHI Next meeting

Last revision by Troy.smyrnios, 2016-07-26