test code valuelists for no data in Define v2.0

j jim

on June 19, 2024

As I understand, empty datasets should not be documented in the define.xml (v2.0). A current study has a set of conditional CRFs (30+ pages) that are event driven (if the event does not occur then there is no data). The event driven CRFs are mapped to standard domains (FA, CE, PR, etc) which contain data from other sections of the CRF (so, there are no 'empty' datasets). At end of study, there were no data collected for the event driven CRFs for any subject.

There is a significant amount of FATESTCDs/FAOBJs annotated for the event driven CRFs. Should these FATESTCDs/FAOBJs be included in the define.xml valuelists when there was no data collected? thanks

Forums: Define.xml

j Jozef

on June 20, 2024

test code valuelists

This is more a question for e.g. the LinkedIn "SDTM experts" forum (https://www.linkedin.com/groups/2758697/) or the LinkedIn Define-xml forum (https://www.linkedin.com/groups/4975366/)
It has nothing to do with Pinnacle21 ...
But ok ...
Essentially define.xml is about what is planned to be done. That is why we also always recommend to generate a (prototype) define.xml even when do data is available yet.
IMO it is bad practice to generate the define.xml at the end of the study, based on the collected data.

Whether you add ValueLists is your choice: there is no obligation at all except for the SUPPxx datasets. As long as you document everything very well, taking care that everything is clear for the reviewer, all is fine.
Remember that you can always use @def:CommentOID, pointing to a def:CommentDef on several elements, to add comments.

j jim

on June 20, 2024