This attribute for (RACE) codelist was modified for validation needs.
Validation messages for CDISC non-extensible CT are Errors, while those are Warnings for extensible CT. There are several exceptions.
Almost all studies have subjects with Race which does not fit existing CDISC (RACE) non-extensible CT. For example, ‘MULTIPLE’ , ‘UNKNOWN’ or ‘OTHER’. Validation messages for (RACE) CT are expected to be reported as Warnings rather than Errors. Another similar case is (ETHNIC) CT.
The issue here is that the CDISC-CT team decided that the RACE codelist may only contain values that are “real” races (whenever that exists - “race” is not scienctific). So they decided that “MULTIPLE” and “OTHER” do not go into the list as e.g. “MULTIPLE” is not a race. The SDTM-IG however states that also “MULTIPLE” is allowed.
Other organizations like HL7 have the better approach that a codelist is the list of ALL allowed values for a term, with less emphasis on whether the semantics is 100% right.
The CDISC-CT approach is at least questionable, as it makes the codelist not very well usable in software: either one must than hardcode in the software that “MULTIPLE” is allowed although it does not appear in the non-extensible codelist, OR one has to change the codelist type from non-extensible to extensible.
Pinnacle21 choose for the second solution.
Also remark that “warning” in Pinnacle21 means that something is less usual, not that it is wrong.
Personally, I consider this as a serious design error of the CDISC-CT team.