Senior Customer Success Manager Jenn Fukuyama and Senior CDISC Consultant Seiko Yamazaki guided a deep dive into best practices for submitting data to multiple regulatory agencies using P21 Enterprise.

Here's the recap:

This Office Hours session focused on how to use P21 Enterprise features can be used to submit the same study data to different regulatory agencies efficiently - and without duplicating work.

Important: Users should always refer to each agency’s specific documents such as the technical conformance guides or data standards catalogs for more detailed information for submissions.

The recommended process for preparing a data package for multiple agencies is to use one data package folder.

First, create Data Packages, for example, one for SDTM and one for ADaM, within one Study in P21E. You will need the Data Package Editor role or above to confirm this and make any corrections needed.

Next, select the P21 Engine on the Data Package configuration page to validate the datasets.

This new set of permissions was added to P21 Enterprise with version 5.2.0. Previously, to configure a Data Package, you needed to have permission to add a Data Package, which many organizations restrict to certain supervisory positions.

However, many of our users needed to be able to update Data Package configurations for just this purpose – to update the Engine selections.

Users can now request access to edit Data Packages by navigating to their profiles. There, they can review their current access and make the request for permission to do more things, like Configure Data Packages. Once submitted, the request is sent immediately to your organization’s administrators.

Learn more about how to request access you don’t currently have in Enterprise.

We recommend using the P21 engine for ongoing, in-progress studies since it has the latest version of the Validation Rules, and reflects the highest severity or impact across agencies. Another huge benefit of the P21 engine is that it contains all of the available validation rules from all agencies. The users can be sure they're seeing all issues and won't get blindsided by a new issue popping up when the engine is switched from one agency to another.

With modifications and refinements of rules (reducing false-positives and fewer bugs) the new engine has the latest algorithms and validation rules. This should allow you to focus your efforts on actual issues only.

After validating your data, work on the Issues list and try to fix as many as you can - P21E's issue management tools make organizing, delegating, and prioritizing your list a breeze.

For Issues that cannot be fixed, set the Status to “Closed,” then prepare explanations for them. You can prepare custom, reusable Fix Tips and Explanations to save time on both issue management and populating your Reviewer’s Guides. For more information, check out our recent Office Hours session, Using Issue Metadata to Save Time and Improve Consistency, in the Help Center.

One way to know if your data is nearing submission readiness is to check your Data Fitness Score; we recommend a score of 90 or greater by submission time.

You can prepare up to ten Fix Tips and ten Explanations, so if you need to prepare agency-specific Fix Tips or Explanations, you could use those five columns for each Rule ID in the template.

This example shows a general explanation in the Explanation 1 column for Rule ID: SD0008C and an agency-specific explanation in the Explanation 2 column. In this case, the user wants to create an Explanation in Chinese for submitting to NMPA.

When you re-validate, P21E shows issue metadata on the issues tab across multiple validations. If you change the engine and validate but some issues did not fire in the latest validation, it does not mean that issue metadata info is gone. Once you switch back to the original engine, if those issues arise again, the related issue metadata will show up.

For example, suppose you ran a validation and SD0008C fired. Since you added Explanations for that rule ID, you can simply copy and paste the suggested Explanation that works best for your submission. Every time the Data Package is validated, P21 Enterprise retains all the issue information from Issue Management, so the Explanation that was selected for SD0008C in that Data Package will always be there aslong as the issue is present on the validation report.

Suppose now it is time to prepare a data package for FDA submission.

After you update the Data Package configuration selection to the FDA engine and run the validations.

For SD0008C, you can overwrite the text in the Explanation field by clicking on the "Copy/Paste" link next to the suggested Explanation. Even if the issue disappears from the validation report after validating with a different engine, the Explanation will still be present if that issue should ever reappear on the report for that Data Package.

When it is time for submission, use the same Data Package and change its configuration settings to use an agency-specific engine, such as the latest FDA engine, and validate your datasets again.

If you plan to submit to TWO agencies, use the same Data Package again and change the configuration settings again to use a different agency-specific engine, such as the latest PMDA engine, and validate your datasets again.

The biggest advantage of using the same Data Package folder in P21E and simply changing the engine settings during final validation is that all Issue Metadata, including Comments and Explanations, is remembered die each item on the Issues list every time a validation is run.

Another major advantage is that the Progress & History of the Data Package's fitness/quality is tracked over time. You can see all validation results from “Validation history” page. You can also see “Progress,” “Data fitness score”, and “Projected score” in the graph as well as in the tables.

You can also compare two validation results. Select validation results by two different engines and compare to see the differences. For example, if you switch to a different agency-specific engine, this is where you can compare the validation results of two different agency-specific engines.

There are three categories in the comparison page:

After adding Explanations for your Closed Issues, export your Reviewer's Guide in one of two ways:

After switching to a different agency specific engine, P21E remembers issue metadata, so if you have the same issues from before switching the engine, the same explanations will be applied to the same issues even after the engine is changed.

To create define.xml, there are six options available in P21 Enterprise. The version of define.xml can be set on the Data Package configuration page. For example, if this Data Package is for FDA submission, you can set the version to define.xml 2.1.

We recommend “From Scanned Data” because P21E can scan your uploaded datasets for validation to autofill up to 80% of define.xml.

Learn more about options for creating define.xml.

Once in the Define designer, check any current issues for the define in the “Issues Tab.” Choosing “View Issues” will direct users to the source of those issues to begin researching - and hopefully fixing - what's wrong directly in P21E.

When it is time for submission, export the define.xml and stylesheet from P21E and include them with other xpt files and documents in a zip file, then validate it with an appropriate engine.

If you plan to submit to two agencies, once the submission package is finalized for one agency and it is time to work on another agency submission, just update your Data Package's configuration with the appropriate agency-specific engine.

Since requirements can differ from one agency to the next, you may need to update the define version. To do this, toggle between Define v2.0 or v2.1 in the same configuration screen as you select the engine. For example, if your define version was set to version 2.1 for FDA submission, the Data Package configuration can be set to version 2.0 for PMDA submission.

When you up-version your define.xml from v2.0 to v2.1, it's important to make adjustments to account for the version differences. For example, you'd need to add values for “subclass,” “standard,” and “has no data” columns in the Dataset tab, and “source” and “has no data” in the Variable tab, “source” needs to be added in VLM as well as “terminology” in the Codelist tab when you change define.xml version from v2.0 to v2.1.

Note: If you change a document filename for a different agency, check the Documents section on the Properties tab to ensure the file name is accurate. For example, for FDA submission, a user creates adrg.pdf, but if a user changed the filename to “analysis-data-reviewers-guide.pdf” for PMDA submission, you need to ensure that the href is also updated.

If you make TWO different Data Packages in P21E for the same SDTM datasets (for example, one each for the FDA and the PMDA), the disadvantage is that NONE of Issue Metadata will be remembered across those two Data Packages, the Progress & History will be wrongly separated, and your teams will have to do DOUBLE the work in managing the Issues.

If you create different Data Packages based on different agencies, you would also be unable to compare validation results in Validation History either.

Creating two Data Packages could also impact your organization's license count.

If you are planning for simultaneous submissions to two separate agencies, it is appropriate to create separate and distinct Data Packages for each regulatory agency. Check with your Customer Success Manager about how licensing works in this case.

If you’re interested in learning more, or if you have a topic you'd like to see in an upcoming Office Hours session, please contact your Customer Success Manager.