This is a pretty old and famous one (see https://www.pinnacle21.com/forum/sd0072-relrec-when-more-one-fa-domain) that comes over and over again when I review SDTM submissions for customers. In that forum contribution, on the statement "SDTM IG 3.1.3 says: 'If a dataset-level RELREC is defined (e.g., between the CE and FACE datasets), then RDOMAIN may contain up to four characters to effectively describe the relationship between the CE parent records with the FACE child records'", the reaction of Pinnacle21 at that time (2016) was: "I believe that this statement in SDTM IG is incorrect." (Wow!). The statement is also present in SDTM-IG 3.2 and in SDTM-IG 3.3.
Personally (Pinnacle21 may disagree), I consider your case as a false positive and that is also what I state in my review reports to my customers when this happens (very frequently).
Also: NEVER, NEVER start manipulating your data in order to satisfy any validation software when you are sure that you have it right. The correct behavior is to document this in the "reviewers guide".
Pinnacle 21 Community Validator raises a rule SD0072 error "Related Domain Abbreviation (RDOMAIN) must have a valid value of Domains included in the study data".
I am following the Submitting Study Datasets for Vaccines to the Office of Vaccines Research and Review use of FACE as a domain. I am also following the SDTM Vaccine TAUG v1.1, which explicitly provides an example where RDOMAIN='FACE'. A dataset face.xpt exists with DOMAIN='FACE'.
Is there anything I can do to avoid the SD0072 error and still comply with the OVRR and Vaccine TAUG guidance?
Thanks in advance for help on this.