SD0061 - Domain referenced in define.xml but dataset is missing ?

Pinnacle 21 Product Suggestions Validation Rule Suggestions
1

Hi!

We need to inverstigate the issue.

However I guess that it because the tumor.xpt dataset is not a standard CDISC data type (e.g. SDTM or ADaM) and the validator doesn’t recognize such data.

We are planning to add a tumor.xpt validation soon, which wil fix the reported issue.

Thank you,

Sergiy

2

Hi Sergiy,

Looking forward to see the Validator for tumor.xpt :slight_smile: What is the aprox. date of release?

If a temporary version is already on SVN, I would like to give it a try.

Best regards,

Amelia

3

The v1.4 release will be very soon.

However I am not sure that tumor.xpt validation will be included.

4

Hi Amelia,

A tumor.xpt file is not a Tabulation Findings domain. It’s a legacy Analysis data. You should not keep it in an eCTD “send” folder.

e.g.,

m4

_ datasets

_ _ examplestudyname

_ _ _ analysis

_ _ _ _ legacy

_ _ _ _ _ tumor.xpt

_ _ _ tabulations

_ _ _ _ send

_ _ _ _ _dm.xpt, etc.

Best Regards,

Sergiy Sirichenko

5

Hi Sergyi,

Yes, you are right - the tumor.xpt file must be placed separately since it’s not a CDISC data but a legacy data.

If I’m not mistaking, it should be accompanied by define.xml and define.pdf files, placed on the same level as the tumor.xpt file. Do you think that you could provide a sample of define.xml / define.pdf for tumor.xpt, please?

Best regards and a happy New Year to all of you, guys!

Amelia

6

Hi Amelia,

Happy New Year to you too!

Yes, you need to provide a define file for all your data and programs.

Unfortunately I do not have good examples to share. You can find one on FDA website:

See http://www.fda.gov/Drugs/DevelopmentApprovalProcess/FormsSubmissionRequirements/ElectronicSubmissions/ucm149719.htm

Find, download and unzip “Example.zip [683 Kb]” file.

Go to “…example\N123456\pharmtox\datasets\101” folder.

However keep in mind that those example is more about eCTD structure, rather than define file content.

  1. At this point I would recommend you to use only define.pdf file in favor of define.xml for tumor.xpt dataset.

  2. Use Comments fields to provide Source and Computational Algorithm for each variable. E.g., SPECIES variable Comments:

"Source: DM.SPECIES;

Computational Algorithm: if DM.SPECIES = ‘MOUSE’ then SPECIES = ‘M’; else if DM.SPECIES = ‘RAT’ then SPECIES = ‘R’"

or “… if DM.SPECIES = ‘MOUSE’ then TUMOR.SPECIES = ‘M’ …”

  1. Add any other relevant details in Comments to provide clear and full description.

  2. Consider to include a program which was used to generate your tumor.xpt dataset.

BTW have you ever considered to participate in FDA SEND pilot project? See details on http://www.fda.gov/OHRMS/DOCKETS/98fr/E7-19468.pdf

Kind Regards,

Sergiy Sirichenko

7

Thank you for the feedback, Sergiy!

8

Hello all,

When processing a dataset that contains tumor.xpt and define.xml entries for it, the OpenCDISC Validator reports:

SD0061 - “Domain referenced in define.xml but dataset is missing”: “Domains referenced in data definition document (define.xml) should be included in the submission”.

The dataset contains the tumor.xpt file and define.xml references it correctly (I think):

... tumor.xpt

Can anyone please share some light on why is this warning reported?

Thank you,

Amelia