TAETORD in multi-zone studies

l Laura

on May 2, 2017

I all,

I'm working on dermatological studies where the same subject can be treated with treatment A on the right side and treatment B on the left side. Both, right and left treatment are given at the same time. Treatment side is randomized.

For me, we have 4 elements:

- Screening

- Treatment A

- Treatment B

- Follow-up

I'm facing an issue with the TAETORD variable:

We have 2 options:

- as the treament are applied at the same time, TAETORD is not unique (TAETORD is respectively 1 / 2 / 2 /3) but this create an issue in P21

- we define a sequence order with TAETORD unique (respectively 10 / 21 / 22 / 30) but this could not match with order in SE as it happen that a subject treat his left side and forgot the right side so we are still not compliant with P21.

Has someone a suggestion ?

Thank you for your help.

Forums: SDTM

j Jozef

on May 3, 2017

TAETORD

I have seen this issue in the past, also for a dermatological study. We did not find a final answer, also as the study was discontinued.
The issue again shows one of the weaknesses of the current SDTM model. The latter assumes that a subject can never obtain 2 different investigational treatments in the same period of time in parallel. As I argued in my presentation at the CDISC Interchange in London, the SDTM model is on a dead track. For example, we see more and more studies that have no visits, but the SDTM model requires us to submit VISITNUM. How can we do that?

One of the ideas that we had for the issue you describe here was to make TAETORD of data type "character" in both TA and SE and then assign 2A and 2B for the elements. Whether this would be acceptable to the FDA we don't know, as the study did not make it to the submission state.

As this is not a Pinnacle21 question, I would recommend to also submit your question to the "CDISC - SDTM Experts" forum on LinkedIn. There are people there with much more SDTM experience than I have.