P21 Check (DD0116) 'FATESTCD/FATEST mismatch in Codelist 'FATESTCD'

From the message it looks…

From the message it looks like FATESTCD codelist in your Define-XML does not match values in the dataset. Have you checked it? But the full rule description is a bit confusing and the reason may be different. It can also depend on the version of the CT and SDTM IG you are using. For example in 3.2 FATESTCD has a specific codelist, while in 3.3 it does not. Although that should trigger a different kind of error.

Sergiy

on January 29, 2020

P21

Hi Colm, A value for…

Hi Colm, 

A value for FATEST should be 'Occurrence Indicator' which represents NCI code = 'C127786'.

FATESTCD='OCCUR' is a standard term utilized in 7 'test code' SDTM CT codelists. A term 'OCCUR' has NCI code 'C127786'. A decoded synonyms of 'OCCUR' term in 'test name' CT codelists have the same NCI code and standard term 'Occurrence Indicator' instead of your 'Occurrence'.

Kind Regards, 

Sergiy

P21 Check (DD0116) 'FATESTCD/FATEST mismatch in Codelist 'FATEST

The 7 codelists Sergiy is taking about are about "Asthma Findings", "Ebola Virus Finding", "Malaria Findings", "Nutrition Findings", "Schizophrenia Findings", "Tuberculosis Findings", and "Vaccines Findings". There is NO "general" FATESTCD nor FATEST codelist, at least not in the latest versions of CDISC-CT. So, extrapolating from these 7 codelists to a "general" FATESTCD/FATEST is pretty tricky and surely not in the spirit of the SDTM-IG 3.2, where "Occurrence" is given as an example anyway (but that is not normative). I would say this extrapolation is an overinterpretation of the SDTM-IG.

My own long experience with FATESTCD is that, unless it is about one of the above diseases, one should create a custom codelist, so WITHOUT NCI codes, and put that in the define.xml. Do so both for FATESTCD and FATEST. To keep the 1:1 relation between FATESTCD and FATEST, one should use the "CodeListItem/Decode" mechanism for FATESTCD. This should then be picked up by any validation software as "the sponsor's truth" (the define.xml IS "the sponsor's truth"). As there is NO codelist for FATESTCD nor for FATEST in the latest controlled terminology, this is the only correct way, unless the submission is about one of the above diseases and the define.xml explicitely listens (and references) one of these 7 mentioned codelists.

Hi Dmitry, Sergiy, Jozef …

Hi Dmitry, Sergiy, Jozef

Many thanks for your replies.This study is not about those disease noted above so I don't think their specific codelists will be suitable. However the FATESTCD has a codelist associated with it but is not linked currently to any CT code in our SDTM's. However I do see that both FATESTCD and FATEST have a codlist by those same names that are not of decode type. I am thinking the FATEST codelist is not needed and then convert the FATESTCD codelist into a decode type might fix this ?

Thanks

Colm

CodeList for FATEST

For the latter, I am afraid not ...
The SDTM standard still requires to provide a separate CodeList for --TEST, even when a codelist for --TESTCD with decodes is provided. At least, that is how interprete it.

This is due to the 2D thinking of the SDTM team and the 1D thinking of the CT teams ;-)