Hi Harini,
As a disclaimer, I do not have enough experience with LOINC coding. It is still a new concept in our industry with very low level of utilization across studies.
In general, you can provide Lab-specific version of LOINC dictionary using VLM in define.xml file based on LBNAM or a combination of LBTESTCD/LBSPEC/LBMETHOD/... values.
My assumption is that for practical purposes, LOINC coding is not very sensitive to a version of dictionary. So, the latest version utilized in your study should be also applicable for other cases from different labs. At least, it's expected to have the same LOINC code for the same test/assessment regardless of what lab it came from.
Some example would be beneficial to better understand your case.
Kind Regards,
Sergiy
As a disclaimer, I have a LOT of experience with LOINC. I disagree completely that it is a "new" concept in our industry (xxLOINC is in SDTM for already about 20 years), and it definitely SHOULD be used much more in submissions. Especially with the raise of RWD, LOINC is becoming extremely important. Expect that FDA will mandate it for more domains (e.g. MB) in future.
I would not recommend adding the version of LOINC used in any datasets themselves, but I would recommend to declare the version in the define.xml in the "ExternalCodeList" element. If you don't know how to do that (I am a define.xml specialist and CDISC Define-XML trainer too) just drop me a mail.
LOINC coding is indeed not very sensitive to a version. LOINC is extremely stable, and unlike other coding systems, it is very seldom that codes are deprecated or removed. BTW, the latest version (2.69) has a lot of new codes for Corona-virus testing, which is of course very important. As in the US reporting of Corona-virus testing is required to use the LOINC code as identifier, my personal expectation is that FDA will in future requiring LOINC codes for Covid-19 studies.
With best regards,
Jozef Aerts
XML4Pharma
Unlike Meddra and WHO, Loinc codes are collected from lab and dictionary version would not be single one. Is it required to be kept in SDTM to be decoded?