WG4 18JAN2012

From PHUSE Wiki
Jump to: navigation, search

When: hh:mm , 18JAN2012
Attendance: Helena; Liz; John
Agenda: Co-leads planning discussion


Who's there?


General discussion


Here are few notes from today’s call, which included Helena, Liz and me.

Helena expressed concern that if she provided her own list of issues that it would become the “official” list. Liz and I countered that the rest of us were only guessing at what CDER’s implementation issues are, so her list is essential. In fact, it would be great if there were some way for her to solicit issues from reviewers. She said that would be hard to do formally, but she did some of that informally last year and distilled it into a presentation she gave at the DIA annual meeting. She will send us her “Top 10 List” of issues from that presentation.

We discussed some of the issues I raised in the document I e-mailed to the group last week. As we were talking, a few more topics came up that may qualify as issues for discussion:

1. In the category of practical and specific issues with SDTM domains, Helena mentioned that Scott B. had sent an e-mail raising the issue that he had use cases in which the SDTM standard could not express the relationships between records in PC and PP. Helena has had the same problem with the microbiology domains, and we know several companies that have had similar problems with the tumor domains. My personal observation about this is that the domain structures are relational and require unambiguous identifiers and record relationships, but neither tumors nor microbes come with barcodes or behave according to SDTM rules.

2. Helena brought up again the problem of identifying deaths in SDTM data. Pretty sure that will be on her Top 10 list.

3. We discussed the general lack of understanding by people at sponsor organizations of how reviews are done at CDER, and what toolsets are available to reviewers and/or are actually used. There is a similar lack of understanding within the agency about the processes and lead times involved in preparing data and analyses for submission. Both lead to false expectations that can easily result in failed policies or unintended consequences.

4. Helena and others within CDER recognized the potential for the Trial Design Model (TDM) datasets, especially TS, to provide characterizations of studies that CDER could use to identify studies of interest within a data pool or data repository, to address some of the agency’s key mandates. However, additional business rules and controlled terminologies, and more reliable use of the domains by sponsors, would be necessary to realize the potential. Helena has been involved in the efforts to evolve those rules and terminologies.

We had some discussion of how we might harness the energies of industry participants, what categories of participation we might have, what the roles of different types of participants might be, and what the WG’s work products might be.

1. Categories of participation. I threw out for consideration a notion of

a. a small-ish core group of participants who sign up to contribute significant effort, that holds regular conversations and produces work products and/or engages others to help produce the work products.

b. a larger group of more casual participants who contribute ideas, comment on work products, and generally take an active interest.

c. A still larger group that follows the group’s activities and wants to stay abreast of developments and may provide comments, but is mostly silent.

We sort of recalled Steve saying that a substantial number of people had expressed interest in participating in WG4, but could not recall the number.

2. Work products. We talked around this topic for awhile, without vectoring in on a specific concept of work products. Then Helena suggested the idea of creating Use Cases as one type of work product, and that idea seemed to have legs. Another idea was to enlist participants to create and/or test procedures and work instructions (and utilities?) to address CDER use cases. An example we discussed was a work instruction for creating a computable datetime column in JMP from ISO 8601 datetime values in a SAS xpt file. Food for thought.

We’re scheduled for an hour at the same time next Wednesday, and I think everyone should be able to attend that one.


Follow up

Last revision by SaschaAhrweiler,02/2/2012