Cannot convert to LBSTRESU

Cannot convert to LBSTRESU

I presume this question is well beyond the scope of Pinnacle21 ...

"U/L" is an ARBITRARY unit, and its value (and thus also conversion factors for them) depends on the analyte that is measured. See e.g. https://myhealth.alberta.ca/Health/Pages/conditions.aspx?hwid=stu3135&lang=en-ca. What was your analyte? Is the molecular weight known? Is there a LOINC code for the test?

Here is a RESTful service you might have luck with if you have sufficient information to use the service: https://ucum.nlm.nih.gov/ucum-service.html.

Best regards,
Jozef Aerts
XML4Pharma

Cannot convert to LBSTRESU

I don't think MinJi was asking for guidance on conversion factors, but how to handle cases like this broadly in SDTM (hence the posting to an SDTM forum).

If the raw data reports a unit that cannot be converted to the standard unit for the parameter, how should this be handled in SDTM?

I believe MinJi's solution is the correct one (leave LBORRES/LBORRESU populated, but leave LBSTRESC/LBSTRESN/LBSTNRLO/LBSTNRHI null), but I found this page while searching for confirmation of the same.

Obviously, if the data are perfectly clean, this shouldn't happen and probably in most (if not all) cases, the unit given is incorrect or the test is reported under the incorrect TESTCD. Nonetheless, this is a frequent occurrence.

Can someone from P21 provide official guidance on how to handle cases like this?

Thanks!

Cannot convert to LBSTRESU

"Can someone from P21 provide OFFICIAL guidance on how to handle cases like this?" ...
Do you really presume P21 is the single source of ultimate wisdom and representing CDISC and/or FDA???
If you want a reaction from CDISC or the CDISC community, there is a much better forum on LinkedIn (https://www.linkedin.com/groups/2758697/)

There is no such "official" guidance!

I was a bit surprised in the post von Minji that she states that the unit as collected is "U/L", but the SI unit is "ug/L". In the scope of the FDA, the SI unit would be "molar", like "umol/L".
Further on, "U/L" is, as said, an arbitrary unit, and in 99.9% of the cases, there would be no SI or "conventional" unit for exactly the same test. So I wonder whether he/she uses the right "corresponding" test with "ug/L" ... @Minji: about exactly which test are you speaking. What is the LOINC code for it?

Anyway, if there is no unit to which the collected data can be standardized, whether as there is no such or the conversion factor is unknown, the common practice is to copy -ORRESU to -STRESU and copy -ORRES to both -STRESC and -STRESN. Leaving -STRESC, -STRESN and -STRESU empty will surely lead to validation messages, which then need to be explained.

If CDISC and FDA would be using UCUM notation for units, we would have much less problems anyway, as unit conversions can then be fully automated (see https://ucum.nlm.nih.gov/ucum-service.html).

Cannot convert to LBSTRESU

Jozef, thank you for coming back to my posting.

This was post in 2021. Nowadays, I communicate through linkedin you mentioned.

For the above issue, the lab is local lab and LOINC code was not shared.

And, thank you for sharing how to find conversion factor using UCUM URL.