CT0037 - errors with MedDRA SOC Terms in sentence case format

RE: CT0037 - errors with MedDRA SOC Terms in sentence case

Hi Dave,

CT0037 checks against the CDISC-published Controlled Terminology list, which lists these values as being all uppercase. There is a question of whether checks against Controlled Terminology should be case sensitive which we are trying to get an answer on, but until that time we've taken the position that they should be case sensitive (which results in the error that you're getting).

It seems from your description that the FDA advice might conflict with the Controlled Terminology case format provided by CDISC. Is this variable specifically mentioned in their guidance, or have you coded --BODSYS against MeDRA, and need to ensure its casing based on that association?

We're reviewing this issue on our end to try and understand what the authoritive response to this issue is, so if we come up with anything more definitive, we'll let you know. Until then, take everything I've said with a grain of salt. In a worst case scenario, CT0037 is only a Warning, so it's permissible for you to ignore it if you've reviewed the data and consider it to be valid.

Regards,

Tim

Curious to know ...

Hi David,

I have seen similar agency request for Preferred Term (AEDECOD), but not SOC (AEBODSYS).  I am curious to know - that is, if you don't mind sharing - the origin of such request so the community is more informed, e.g., CDER vs. CBER, review division, etc.

On a general thought to share ...

While their request makes logical sense (i.e., MedDRA being an externally controlled dictionary), it conflicts with the essence of what's documented on the NCI spreadsheet, i.e., for column E [CDISC Submission Value], "Currently (as per SDTMIG 3.1.2) this is the specific value expected for submissions."  I read that verbiage to literally mean matching the text along with the choice of case shown.  Fortunately, in SDTMIG v3.1.2 Appendix C1 pg 273, codelist C66783 is cited to be optional, i.e., "Could be used for --BODSYS variables but not required to be used."

Nonetheless, it is a good business practice to retain all original dictionary encoding information in SDTM datasets without text manipulation (changing case, padding extra zeros, or worse truncation), e.g., MedDRA (PT, LLT, HLT, etc), WHO-DD (Preferred Name, ATC levels), LOINC.

A more important question to me is whether it is necessary to repeat the content of AEBODSYS as a codelist in the define.xml, epsecially when an external dictionary like MedDRA is used to re-encode verbatim terms.  My opinion is not, just as we would not do so for AEDECOD.

Given 1) the agency's sensitivity to the correctness of MedDRA data; and 2) an SDTM validation check exists to verify PT in AEDECOD, I would go to further suggest a new check to ensure the [SOC, PT] pair is valid in a given version of MedDRA (assuming primary SOC).

Hello Tim, We received this

Hello Tim,

We received this in a letter from the FDA to the sponsor. It was from

Bob A. Rappaport, MD

Director Division of Anesthesia, Analgesia and Rheumatology Products

Office of Drug Evaluation II

Center for Drug Evaluation and Research

It was a standard letter with an attachment that was several pages it contained

The Division requests the following for the submitted datasets: (there were a number of points here)

8. The spelling and capitalization of MedDRA terms must match the way the terms are presented

in the MedDRA dictionary. For example, do not provide MedDRA terms in all upper case

letters.

So we provide the MedDRA PT and SOC in the MedDRA standard dictionary listing -- a more sentence case type of data.

Can I change the configuration file and substitue the sentence case text? This could fix the error and provide some additional checking.

Hello Anthony This was

Hello Anthony

This was from

Bob A. Rappaport, MD

Director

Division of Anesthesia, Analgesia and

Rheumatology Products

Office of Drug Evaluation II

Center for Drug Evaluation and Research

It appeared to be a standard response letter in preparation for filing our NDA. The MedDRA PT and SOC are both in a sentence case in the standard MedDRA dictionary.

MedDRA Case

I have also had a similar request from the FDA regarding the case for MedDRA coding.  The following is an excerpt from a document provided by the Office of Oncology titled "OODP's End-of-Phase 2 General Advice for Planned Marketing Applications":

10) Use of Adverse Event Terminology (i.e., MedDRA, etc)

When using MedDRA for adverse events and past medical history terms, sponsors should exactly follow the spelling and case of the MedDRA terms.

RE: Hello Tim, We received this

Hi,

Thanks for the additional information. I've personally been a bit busy the last week, but I just wanted to let you know that we are still looking into what the best solution here is. It makes sense that the casing should match that of the MedDRA data, but arguably in that case so should the Controlled Terminology file.

I'd recommend temporarily changing the casing in that file so that you can ensure that your data complies with the FDA guidance. It's important that we find a permanent solution though, since we can't rely on people making these special case changes all the time. I'll see what I can do on our end to speed that process up.

Regards,

Tim

Hello Tim, I agree that

Hello Tim,

I agree that providing a permanent solution is better for our potential worldwide user base. I will modify the terminology case for the SOC while we wait for this to be done.

I very much appreciate your help on this.

Thanks

Dave Borbas

CT0037 - errors with MedDRA SOC Terms in sentence case format

Hi,

I have one suggestion. We could split the check in two parts.

Check if the term match regardless of capitalization.

False==>   Term do not match MEdra Dictionnary

True==>   if the term match exactly the medra term.

                  FALSE==>Term macth but case do not match

Jacques 

Hey Jacques this would work

Hey Jacques

this would work and be more informative

-- Dave

To MedDRA PT case sensitive match or not?

Agreed, as long as we are talking about a more intelligent message, i.e., not having 2 checks.  I hope to see:

When AEDECOD = "Influenza" then no message, but;

When AEDECOD = "FLU" then message reads "Not an entry in MedDAR Preferred Term (Version x.y)"
When AEDECOD = "INFLUENZA" then message reads "Case mismatch to MedDRA Preferred Term (Version x.y)"

Both are output from SD0008, which should remain an Error finding.

RE: CT0037 - errors with MedDRA SOC Terms in sentence case forma

Hi Jacques,

This makes sense. There are some implementation issues that we need to work around to make this possible, but I've put this on my tenative to-do list for the next release and can hopefully get something working by then.

Regards,
Tim

FDA CDER's stance

I hope this further convince you there shall be no wishy-washy. When it is not an exact match, it is a mismatch regardless what dictionary it is.

Excerpt from CDER Common Data Standards Issues Document, pg 5-6, May 2011:

Common Dictionaries
It is expected that common dictionaries are used across trials and throughout the submission for each the following: adverse events, concomitant medications, procedures, indications, study drug names, and medical history. Implementation of such dictionaries should be careful to exactly follow the terminology conventions (e.g., spelling and case) specified by the dictionary or according to a single consistent sponsor specification if no pre-existing terminology exists. CDER has frequently received data in which terminology conventions were not followed, for example, misspelling of MedDRA or WHO Drug terms or lack of conformance to upper/lower case or the use of hyphens. This makes it difficult to use or develop tools to analyze this data in an automated fashion. Again, the use of a standard dictionary that is sponsor defined or extended should be documented using the define.xml.

Terminology Change requested

Hello Anthony and Tim,

Given the recent FDA document I made a Terminology request to change the case of the SDTM SOC codelist. Currently it is in full upper case. I asked for a change to MedDRA sentence case to match the FDA CDER document. Once this change happens it should be reflected in the terminology file that will be used in the validator program. This may be the best approach to this problem.

Thanks -- Dave