What to Expect When Submitting Your First SEND Dataset to the FDA
With the December 17, 2016* requirement for the FDA Standard for the Exchange of Nonclinical Data (SEND*) fast-approaching, our SEND action team prepared a dataset for test submission to the FDA. This helped us to better understand the FDA’s SEND submission requirements, build experience and confirm our readiness to help clients submit their SEND datasets.
During this process, we uncovered a couple of significant learnings:
- Allow for adequate time to prepare and submit to the FDA
- The process to deliver our first test submission took more than two months from kickoff to FDA notification
- It’s important to start early to understand the preparation time needed for submissions
- For test submissions only, the dataset must be submitted on a physical CD and sent to the FDA via postal mail
- This came as a surprise to us, since SEND is a streamlined electronic format of the data
- (Note: In a real submission, the SEND datasets will appear in a specific location labeled “tabulations” in the submission folder structure as described in section 7 of the FDA CDER/CBER Study Data Technical Conformance Guide).
The following is a detailed log of our real-life SEND test submission. It’s a sneak peek at what you can expect when you go through this process. We’ve also offered some ideas on navigating challenges and best practices to consider along the journey.
The Tale of a Real-Life SEND Test Submission
Week 1: Project Set-Up
1: Kick off the SEND project. The best way to start is by researching the SEND submission requirements. We created this SEND Checklist you may find helpful:
- Obtain submission identification number (one for each study)
- Identify study(ies) to be submitted
- Gather documents (protocol, amendments, report, key dates)
- Obtain Study Data Reviewers Guide (SDRG) template
- Generate SEND dataset following our established procedures
- Run the SEND dataset through the validator tool to identify errors, warnings, etc.
- Prepare the SDRG and Define-XML files
- Electronically package the SEND dataset in submission file structure
- Obtain shipping address, prepare accompanying letter, shipping paperwork
- Package and ship to FDA
2: Request submission identification numbers.
- If you know you’ll have multiple submissions—for example, we planned to submit six separate datasets—request an identification number for each submission at the onset.
3: Select your study.
- We chose a 104-week carcinogenicity internal validation study because we had already prepared the SEND dataset for other purposes and thought that the volume of data would be a good stress test for the pilot submission.
- When you conduct your SEND test submission, you should consider the common study types that your organization will encounter and select one of them to pilot.
4: Gather the study documents that must accompany the submission. Include the study protocol, amendments and final report.
- It’s a good idea to review your documents to become familiar with the study design and to ensure you are knowledgeable about what is being submitted in the event you get questions from the FDA.
Week 2: Dataset Generation
1: Prepare the SEND dataset. Identify data terms for mapping to controlled terminology (CT).
- One recommendation would be to pre-map your terminology to the SEND CT so that you can utilize your system administrator to help with additional, study-specific CT mapping that may be needed.
- This also allows you time to start working on the Define-XML file.
2: Draft the SDRG. You can do this easily using the latest available template from the PhUSE SDRG team’s wiki.
Week 3: Dataset Review
1: Initiate Quality Checking (QC) activities. All manual portions of the dataset, Define-XML file and SDRG need to be quality checked.
- If possible, we recommend assigning a team of reviewers to spread out the workload.
2: Start prepping for the shipment.Get the shipping address ready, research how to package the data, start prepping the accompanying letter, etc.
- If you are not shipping from a US-based location, take note that a physical shipment of a CD is required to be sent to the FDA—and proper customs paperwork for CDs will be necessary as well.
Week 4: Dataset Submission
1: Validate the dataset. Following QC completion, use the validator tool on the dataset so that you can explain any warnings in the SDRG.
- During preparation of our test case, the team found an error in the dataset, which we expected would prevent its acceptance into the FDA’s software system. While we recommend you pre-address any errors found in the validation process, in our case, this was actually helpful in enabling us to proactively communicate potential concerns.
2: Complete the submission packet. This includes presentation of the dataset in the submission file structure.
3: Submit the SEND dataset. Send in your pilot submission via mail. For us, this was–approximately 30 days after project initiation.
- If your dataset is complex, contains data from multiple sources or validation errors or warnings to resolve, you may want to plan for it to take a bit longer.
- Remember, you MAIL your electronic data to the FDA, so plan to include at least one day for overnight express and tracking of your SEND data.
Week 5+: FDA Results
1: Receive initial findings.
- 23 days after mailing, we received an email from the FDA eData mailbox with the findings of the pilot submission. As expected, the dataset was rejected and the FDA’s validation report indicated the error we previously noted, as well as a few other errors that we would need to address.
- If you receive a similar response, it is good practice to pull your internal team together to review the FDA’s findings and determine the next best steps.
2: Address FDA feedback, if necessary.
- It was determined in our internal meeting that it would be best to contact the FDA about the validation report findings.
- We noted several inconsistencies and wanted to make sure we were interpreting the feedback properly and were able to make the appropriate corrections before we resubmitted.
- The FDA eData mailbox is designed to facilitate questions and requests, such as our concerns with the validation report.
3: Final Confirmation.
- We received a reply from the FDA eData mailbox that indicated that the SEND dataset and our validation reports now matched.
- This test process captured some possible errors in our system and confirmed technical conformance alignment with the FDA validation.
- Going through a test run with a test SEND dataset submission months ahead of December 17, 2016will allow you to uncover issues in your process that you will need to correct, as well as ways to save time in the future with your actual SEND submissions.
- Get started as soon as possible with a test SEND dataset submission to the FDA. You’ll not only be able to establish your own internal standard operating procedures (SOPs) well in advance of the December SEND deadline, but you’ll also become familiar with the FDA submission process.
- While your first submission may go without incident, it is likely that you’ll encounter some form of initial set-up problems. A test-run or pilot submission will help you learn how to handle follow-up and resolution with the FDA.
- It’s also a good idea to identify and correct any potential errors—ahead of the deadline and with a test submission—so that none of your future studies are held up due to a SEND data rejection or “Refuse to File”.
For more information:
Visit our SEND webpage to access regulatory guidelines, SEND education and other valuable information about SEND.
Ready to take a SEND test-drive? Your move to SEND doesn’t have to be stressful. We can help.
*The Standard for the Exchange of Nonclinical Data (SEND) is the new FDA format for submitting your nonclinical study data. All carcinogenicity and toxicology submissions to the FDA for studies that begin after December 17, 2016 must comply with this format.