Pinnacle21 check ADaM vs. SDTM

validation for ADaM IG 1.1 includes traceability rules

Hi Manolya, 

Yes, P21 validation for ADaM IG 1.1 includes traceability rules similar to ADaM IG 1.0.

The execution of these rule requires adding AE, DM and EX datasets when validating ADaM data. This is very important  especially when preparing your submissions to PMDA.  

Unfortunately, there are still many users which do not run validation of ADaM data correctly. Therefore, we are introducing new set of rules. For example, AD1024 : Traceability rules not executed due to missing DM dataset.

Kind Regards, 

Sergiy

P.S. For your reference this is a list of data traceability rules:

• AD0053 (Error) - The combination of STUDYID and USUBJID value does not exist in the SDTM DM domain
• AD0061 (Error) - SDTM.EX is present but neither ADSL TRTSDT nor TRTSDTM are present
• AD0061A (Error) - SDTM.EX is present but neither ADSL TRTEDT nor TRTEDTM are present
• AD0204 (Error) - For the same USUBJID, the ADSL.AGE != DM.AGE
• AD0205 (Error) - For the same USUBJID, the ADSL.AGEU != DM.AGEU
• AD0206 (Error) - For the same USUBJID, the ADSL.SEX != DM.SEX
• AD0207 (Error) - For the same USUBJID, the ADSL.RACE != DM.RACE
• AD0208 (Error) - For the same USUBJID, the ADSL.SUBJID != DM.SUBJID
• AD0209 (Error) - For the same USUBJID, the ADSL.SITEID != DM.SITEID
• AD0210 (Error) - For the same USUBJID, the ADSL.ARM != DM.ARM
• AD0253 (Error) - Record key from SDTM AE is not traceable to ADaM ADAE (not enough ADAE recs)
• AD0258 (Error) - Record key from ADaM ADAE is not traceable to SDTM.AE (extra ADAE recs)
• AD0367 (Error) - For the same USUBJID, the ADSL.ACTARM != DM.ACTARM

Traceability

hi Sergiy

thank you very much for your input! Greatly appreciated.

I have just downloaded the new P21 community and will have a look into the new rules. 

Best regards

Manolya

Question about AD0253 - Record key from SDTM AE is not traceable

Hi, I have 2 questions about the rule AD0253.

Let's say I get this error because I have some adverse events for screen failures that are in AE but not in ADAE. Is it ok even if those subject are in ADSL (since we need them for a few outputs)? Or do all ADSL subjects (even SF ones) need to have their AEs added to ADAE?

And what about randomized subjects who have not been treated. Do we need to have their AEs in ADAE, or is it ok to not keep them?

Thank you very much in advance,

Jessica

Sergiy

on June 1, 2020

P21

We just create rules to assist the FDA/PMDA reviewers

Hi Jessica,

Your questions are not for some third-party consultants on our forum. Such decisions are expected to be pre-specified in study protocol and statistical analysis plan. What is needed for your analysis of study data? Do you consider Adverse Events applicable for Screen Failures? If not, then why did you collect them and keep in tabulation data? In general, randomized subjects are considered as Randomized and/or Intent-To-Treat populations in study data analysis. Please refer to your study SAP if safety analysis includes these populations.

We just create rules to assist the FDA/PMDA reviewers in knowing when AE records didn't make it into ADAE, or when there are ADAE records that for some reason weren't included in AE, and then the sponsor/CRO needs to at least explain why the situation exists so it is clear to a reviewer.

Kind Regards,
Sergiy

Thank you.  I guess it is…

Thank you.  I guess it is the "not in ADSL" that confused me, as when I read it I understood that if your subject was in ADSL, then it's AEs have to be included in ADAE.  Sorry for the confusion.

However, subjects not analyzed (e.g. screen failures) who have AEs recorded in SDTM AE but not in ADSL do not need to be included in ADAE.  Currently this check does not take into account the latter.  Please check your ADSL file to see if these issues are for screen failures.

Sergiy

on June 11, 2020

P21

You are proposing a new custom rule

Hi Jessica, 

All rules represent a particular business case and answer to a specific question / test. 

Pinnacle 21 implemented official rules from CDISC team. AD0253 checks if any records in AD domain are missing in ADAE dataset. Your example is a different business rule to ensure that subjects in analysis data did not lose their records from AE domain. Your rule is a subset of AD0253 rule and represents your study-specific implementation. A value of AD0253 rule for FDA reviewers is that you will need to report its validation findings and explain your case in Reviewer's Guide

You are proposing a new custom rule. I agree that it could be very handy for many users. However, our Community Validator provides executable implementation of official rules from CDISC and FDA. You can submit a proposal of your new rule to CDISC ADaM team.

Thank you,
Sergiy