1777 Sentry Parkway West
Building 17, Suite 405
Blue Bell, PA 19422
+1 (888) 507-2270
SD2239 wrongly references CG0240
SergiyYes, SD2239 and CG0240 are different rules
Hi Thierry,
I agree with you.
SD2239 (Inconsistent value for --TPT) rule ensures consistency of --TPT value for the same DOMAIN, USUBJID, VISITNUM and --DTC
CDISC CG0240 (--TPT and --TPTNUM have a one-to-one relationship) rule is different and covered by SD1126 and SD1127 rules.
SD2239 looks like data quality rule.
Thank you for reporting this issue!
Kind Regards,
Sergiy
SD2239
Hi Thierry,
I looked up in my evaluation of the FDA rules version June 2019 I made at that time, and found my (old) notice that this rule is nonsense, just like some other FDA rules (e.g. SD0029 "Standard Units (--STRESU) should not be NULL, when Character Result/Finding in Std Units (--STRESC) is provided").
At that time, I have send my evaluation to the FDA, but never got any answer. I will send an additional follow-up mail to R.F. at the FDA and ask to immediately retreat this rule. As you state, it makes no sense to compare a planned time point with an actual datetime.
SD2239 wrongly references CG0240
Thank you!
If it is a data quality check, then maybe the level should be "warning" or even "notice"?
SergiyHi Thierry,A tricky thing…
Hi Thierry,
A tricky thing is that all properties are quite subjective. Users may have even different interpretations of the same words. Plus, everything is changing overtime. For example, FDA has stopped providing their Severity for validation rules.
At Pinnacle we always have 2 different attributes for validation rule:
- Message Type (Error, Warning, Notice)
- Review Impact/Risk (High, Medium, Low)
In terms of Message Type, SD2239 is an Error. If there are two or more different Planned Time Points for the same actual date/time, then definitely it’s an issue due to the mapping and/or programming mistake. I think that Review Impact of this issue is rather Low, as you suggested by ‘Warning/Notice’.
I do not agree with you that Data Quality issues are less important than Standard Compliance. For regulatory review process, it’s opposite. Standard Compliance facilitates an analysis of study data. Data Quality issues affect the analysis results.
That’s why I understand PMDA which considers SD2239 issue with Severity=Error in sense of ‘High Review Impact’.
Kind Regards,
Sergiy
SD2239
Whether two time points (planned or actual) having the same -DTC is a quality issue or not depends on the study and its design. I have seen studies where two samples needed to be (or where) taken at the same time (point), e.g. one on the left arm and one on the right arm (tissue samples). The only thing that is different between them is the laterality (-LAT). Then having two records (even with the same result, e.g. "NEGATIVE") is surely NOT a quality issue.
Now, if the mapper would believe the P21 error message, and removes one of the two records, then we DO have a real quality issue.
In my opinion, such quality issues should be tested on by rules developed and implemented by the sponsor for that specific study, not by software from an external vendor that doesn't understand the pecularities of that study.
For me (personal opinion), rule SD2239 is and remains a nonsense rule.
We got SD2239 firing (different --TPT for the same --DTC) for SDTM-IG 3.2, FDA flavor. It references CG0240, but the latter is about one-to-one relationship between --TPTNUM and --TPT, and does not mention --DTC. Furthermore, since --TPT is a planned timepoint, so it makes no sense to compare it with an actual datetime. (You could have two samples drawn on the same date/time, but which were planned to be taken on different timepoints, for instance.) Apparently this is a PMDA rule, but it is present as FDA rule as well?