Meeting Minutes-DSTU-Testing Team 4-8-2010
Bernadette Billet, Liquent; Bob Birmingham, Johnson & Johnson; Joel Finkle, ISI; Jill Geist, Abbott; Matthew Leceva, Teva; Harv Martens, Extedo; Joe Montgomery, FDA; Lenore Palma, Datafarm; Ragunathan Pathmanathan, Abbott; Rich Ware, AstraZeneca
Action Item Review
- Virtify and Octagon have identified representatives to participate on the Technical Sub-team.
- Bernadette Billet and Lenore Palma met to begin drafting a DSTU/US NDA-specific implementation guide.
- Bob Birmingham will contact Jason Rock to schedule the Data Model Walkthrough (post-meeting note: this has been scheduled for April 28)
Controlled Vocabulary Formatting
Joel Finkle has begun to review the Excel-formatted controlled vocabulary to determine potential translation to XML. Additionally, SPL has a format for controlled vocabulary that may address the need. Joel will continue to investigate.
The team discussed the difference and value of GUIDs vs. OIDs, OIDs being controlled with a pre-determined pattern that requires an organization to register for an OID identifier. For testing purposes, GUIDs should be sufficient as long as we maintain a record of the GUIDs assigned to the lists, however, we will explore what is necessary to register for an OID and the organizational labels necessary.
Validation and Acceptance Criteria
As discussed previously, the messages created will require a level of evaluation for technical acceptance criteria. Using vendor-provided tools for this may prove challenging as they are open to interpretation, and if a tool is used, it must be one that is made available free-of-charge to all participants. In lieu of that, Bernadette will create a 'shell' RPS XML that includes instructional/mapping information to identify what information is expected to be included in the messages.
Jill Geist raised the question of how sponsors who are partnering with vendors can provide feedback while maintaining compliance with non-disclosure agreements. The team agreed that there needs to be an outline of appropriate feedback that focuses on requirements of the message itself and does not provide information about the message creation experience (tool usage) itself. We should identify any discrepancies of interpretation, such as required vs. optional information, but not a preference to how that information is captured. An independent (vendor-free) review board may be necessary to review the feedback and ensure blinded feedback to the full team.