j John


Hi, I'm creating PP domain for the PK parameters calculated per study period, and I get SD1117 check from P21 report if I put the PERIOD information into PPGRPID.

I've tried to add it to PPCAT to avoid this error but it seems not a good solution because PPCAT should be populated the Analyte from PCTEST according to SDTMIG.

Is there any suggestion? Combing the PERIOD information with Analyte in PPCAT, like "Analyte: Period 1", is acceptable?

Forums: SDTM

on March 8, 2022

Hi John,

In general, both PPGRPID and PPCAT variables are not designed to capture Timing information. 

Kind Regards,

j John
on March 9, 2022

Hi Sergiy,


So the solution is to add the timing information to SUPPPP and provide an explanation in the reviewer guide?

However, in SDTMIG v3.3, the Day information for the parameters calculated for Day 1 and 14 was added to the PPGRPID.


Kind regards,


on March 9, 2022

I would use PPTPT variable. --GRPID is designed for grouping records. Grouping of all records for Period 1 is OK! But utilization of --GRPID variable as primary timing information may be not a good practice.

Another alternative could be VISIT variable.


j John
on March 9, 2022

Many thanks for your suggestion! My case is more like grouping the records by period.


The PK sampling was drawn on Day 1 and Day 2 in each period with TV.VISIT=PERIOD x DAY 1 or PERIOD x DAY 2, but the PK parameters were calculated per Period.

Should we define new TV.VISIT for this case? PPTPT may be a solution, but it would be a little overlapping with the TV.VISIT.

j Jozef
on March 10, 2022

This is an interesting discussion!
First of all, NEVER, NEVER start changing your data (or where you put them) just to avoid validation errors (like SD1117) when you are sure that there is no violation of the SDTM standard. Much better is to document false positives in the SDRG.
Rule SD1117 does not say anything about PPGRPID or any -GRPID, so the messages you get must be considered as false positives, except of course when there is another underlying cause.
Furthermore, -GRPID can be anything, so there cannot be any rules on it. If you look in the by CDISC published validation rules (for 3.4 for example), there is not a single rule about -GRPID either.
Another discussion is of course about best practices, but validation rules should not enforce one best practice or another. That can be done e.g. on the company level, by adding additional company-internal rules.


j John
on March 11, 2022

Thanks, Jozef, and sorry I didn't make it clear enough.

I'm seeking a solution to fix the problem about duplicate records caused by calculating PK parameters for multiple dosing periods, becuase the Key variables, PPCAT and PPSCAT, have been defined as the "Analyte" and "Method type" of the paramters in the SDTM IG.

According to the example from SDTM IG, I tried the variable PPGRPID for the period information but the error is still fired; the reason is PPGRPID is not a key variable to identify the unique record, abivously.

I will leave this error message to cSDRG and explan it.

j Jozef
on March 12, 2022

Hi John,
Essentially, it is YOU who defines what the key variables for record uniqueness are, using the "KeySequence" attributes in the define.xml.
But I don't know whether SD1117 uses this. I think it only uses the surrogate key (i.e. derived in a post-processing step) --SEQ variable.

Want a demo?

Let’s Talk.

We're eager to share and ready to listen.

Cookie Policy

Pinnacle 21 uses cookies to make our site easier for you to use. By continuing to use this website, you agree to our use of cookies. For more info visit our Privacy Policy.