SDTM

Dear all,

I have a question about SD0052: Inconsistent value for VISITNUM within VISIT.

There were at least two subjects had unscheduled visit, and I presented my SV dataset as following:

USUBJID  VISITNUM  VISIT  VISITDY  SVSTDTC  SVENDTC

STUDYXX101  1.0  SCREENING  -21  2013-07-12  2013-07-12

STUDYXX101  2.0  VISIT2  1  2013-07-25 2013-07-25

STUDYXX101  3.0  VISIT3  7  2013-07-31  2013-07-31

HI All
We could like to get some clarifications from you about the EPOCH variable for SDTM. There are many warning messages about "SD1077 FDA Expected variable not found" when we use the OpenCDSIC validator v1.4.1 for SDTM 3.1.3. However, EPOCH is not specified on the SDTM Implementation Guide v3.1.3(IG3.1.2+3.1.3 Annotation) for these domains on AE CM EG LB MH PE SU VS. Does the EPOCH is the permissive variable and do we have to include EPOCH on the SDTM transfer for the domains for AE CM EG LB MH PE SU VS?

Thank you!
Bessie

Dear all,

I understand the new request from the FDA to reduce unnecesary space usage in variables by limiting variable lengths to the length of its longest value. I'm just wondering how in practice this can be attained for --TESTCD's.

According to SDTMIG 3.1.3: "--TESTCD and IDVAR will never be more than 8, so length can always be set to 8." I can imagine in certain circumstances, like in an LB domain, most LBTESTCD values are shorter than 8 characters.

For example: the maximum length of all values in LBTESTCD is 4. Which rule does then apply?

Hi,

I added SESTDY and SEENDY to the SE domain.  They were added after SEENDTC, so that the order looks like:

..SESTDTC, SEENDTC, SESTDY, SEENDY, TAETORD, EPOCH, SEUPDES 

This order is similar to how SVSTDY and SVENDY are placed after SVENDTC.

Please let me know if there is a problem with the rule derivation, or the correct location for where the two variables should be placed in the dataset.

Thanks,

Cathy

Hi,

I'm getting a warning 'Missing value for LBREASND, when LBSTAT is NOT DONE' when using v1.4.1 config 3.1.3.

However, the SDTMIG 3.1.3 section 4.1.5.1.2 gives examples which state: 

 --STAT should be 'NOT DONE'

 --REASND, if collected, might be .......

Note, it says 'if collected' for REASND because it is not necessary to have a reason.

Thanks,

Cathy 

Hi,

In v.1.4.1 using config-sdtm-3.1.3, I'm getting error SD0056 for a custom events domain, saying xxDECOD is a required variable not found.  However, xxDECOD is not a required variable. For example, in MH, DV, CE, it is Permissible.

Thanks,

Cathy

Hi,

I have the case where I have a LBTESTCD = HCG and LBTEST = Pregnancy Test. HCG is part of CT, but Pregnancy Test cannot be found in LBTEST codelist. There is also no other similar value. Therefore, I get this error message. Did someone have the same issue. If so, please let me know. Are such cases considered for future`update of SDTMIG?

Furthermore, what is the difference between the validation checks CT0082 and CT0082R. They seem to be the same...

Best regards,

Manolya

Hi,

I have populated DSDTC and DSSTDTC and only included DSSTDY. Do I really need DSDY in the dataset also? If so, why is it not in the SDTMIG?

Best regards

For TR, it needs to take into account the TRLNKID before deciding its a duplicate.

Dear Sergiy,

Thanks a lot for the new release.