How to handle planned versus actual study data in value level metadata?

n Nienyin

on July 1, 2024

Hi,

I have a question for value level metadata. If a --TESTCD was planned to be collected in a study, but there was no data after DBL, do we need to include this --TESTCD into the VLM? Under my understanding, we should present all possible value in codelist, but it seems to have no rule for VLM? Do you have any suggestion for how to handle this condition in VLM? If we need to present this kind of no data in VLM, how to decide its length, type?

Thank you so much.

Best Regards,

Nien

Forums: Define.xml

j Jozef

on July 2, 2024

How to handle planned versus actual study data in VLM

This has nothing to do with P21 ... There are better forums for this, such as LinkedIn Define-XML (https://www.linkedin.com/groups/4975366/) and LinkedIn SDTM experts (https://www.linkedin.com/groups/2758697/).

To answer your question: Defíne-XML is ALWAYS about "planned".
Therefore, people should really learn that the define.xml is the "submission plan", and can be set up as soon as the study design is final, even possibly before the study starts.
Unfortunately, many people use "black box" software to generate the define.xml from the SDTM data when these are final, in worst case a few days before the submission.
This is however "worst practice".

So, look at your study design, look at your CRFs, and generate the define.xml from that information, not from the SDTM data after DBL, but much much earlier.
There is very good and user-friendly (GUI-based, using "wizards") software on the market at very affordable prices to do so.

In your case, for "length" and "type", this should be obvious from your CRF or EDC system settings.
Also please read the define.xml specification which clearly states "Length is the maximum expected variable length. It is not based on the length of the data collected".

With best regards,
Jozef Aerts
CDISC Define-XML instructor