According to page 4 of http://cdisc.org/system/files/members/standard/adam/ADaM%20OCCDS%20v1Provisional.pdf
you should use this for the Class
ADaM 1.0 IG – ADAE
ADaM 1.1 IG – OCCURRENCE DATA STRUCTURE
Of course, NCI does not have ADAE in its GNRLOBSC codelist. According to ALT (adam leadership team) and ADaM Validation subteam, you should use ADAE. And since you can’t possibly be using ADaM 1.1 since it has not been released, i would go with ADAE. Interested in your take on this.
EVENTS
|
C103372
|
FINDINGS
|
C103373
|
INTERVENTIONS
|
C103374
|
RELATIONSHIP
|
C103376
|
SPECIAL PURPOSE
|
C103377
|
TRIAL DESIGN
|
C103379
|
SUBJECT LEVEL ANALYSIS DATASET
|
C103378
|
BASIC DATA STRUCTURE
|
C103371
|
OCCURRENCE DATA STRUCTURE
|
C123454
|
The Define-XML specification rules that def:Class needs to follow the NCI GNRLOBSC codelist. ADAE was indeed never part of the GNRLOBSC CodeList.
So, I still think we have to follow whatever is in the GNRLOBSC CodeList.
Note that the Analysis Metadata Results example as published by CDISC also uses OCCURRENCE DATA STRUCTURE.
This is all about Define-XML v2 of course, as in Define-XML v1 this was not strictly defined. The older document (ADaM IG 1.0) were published before the release of Dfine-XML v2.
Be aware that more Define-XML metadata will be driven by NCI Codelists in Define-XML 2.1 (eg. Origin), so can not be hardcoded in the application.
I hear you. We will account for the lastest terminology coming soon so that define.xml issues are corrected. However if you plan on using a Define.xml for ADaM, the ALT and OCCDS documentation is very clear that you should use ADAE with ADaM 2.1/IG 1.0.
I don’t see why you have to use ADAE, but i’m also not very passionate about issues dealing with labels or categories that do not really impact regulatory review. i.e. ADAE vs OCCDS or …Standard unit vs Std. Unit. Sometimes we don’t give humans sometimes enough credit to deciper on their own 
It looks like the ItemGroupDef/def:Class values are hard-coded, since “OCCURRENCE DATA STRUCTURE” for ADaM is deemed invalid (ERROR). However, the values are governed by CDISC/NCI CodeList C103329 (GNRLOBSC, “General Observation Class”). “OCCURRENCE DATA STRUCTURE” has been part of the CodeList since 2015-09.
Thanks,
Lex Jansen