View Categories

Correspondence Tracking

The Correspondence Tracking report identifies and organizes all agency correspondence within a dossier — cover letters, meeting requests, meeting minutes, FDA questions (RFIs), and commitments. It uses a two-phase AI extraction pipeline to find and classify these documents automatically.

Correspondence Tracking report with letter and meeting tabs

Correspondence Tracking report with letter and meeting tabs

Accessing Correspondence Tracking #

  1. Click Reports in the sidebar.
  2. Click the Correspondence Tracking tile.
  3. Select a dossier from the application selector.

Two-Phase Extraction #

Correspondence Tracking uses two extraction steps that you trigger manually:

Phase 1: Scan M1 #

Click Scan M1 in the toolbar. This scans Module 1 (Administrative) documents to identify cover letters, form FDA 356h, and other administrative correspondence. This is a quick scan of the structural metadata.

Phase 2: Crawl Content #

After Scan M1 completes, click Crawl Content. This performs a deeper AI analysis of the document text to identify meetings, questions, commitments, and classify each correspondence item. This step takes longer because it reads the full text of each document.

Tip: Always run Scan M1 first, then Crawl Content. The content crawl builds on what the M1 scan identifies. Running Crawl Content without scanning M1 first will miss structural references.

Correspondence Categories #

After extraction, results are organized into four tabs:

Tab What It Contains Examples
Letters Cover letters, agency responses, information requests FDA Complete Response Letter, Cover Letter for sequence 0005
Meetings Meeting requests, meeting minutes, pre-submission conferences Pre-IND Meeting Minutes, Type A Meeting Request
Questions / RFIs Requests for information, clinical questions, chemistry questions FDA Information Request dated 2024-03-15
Commitments Post-marketing commitments, agreed-upon actions, follow-up items Commitment to submit 12-month stability data

Working with Entries #

Click any entry in a tab to expand it and see the full details — document text, source file, classification confidence, and linked submission sequence.

Manual Entries #

You can manually add entries that the AI did not detect. For example, to add a meeting that was scheduled outside the dossier documents:

  1. Go to the Meetings tab.
  2. Click Add Meeting.
  3. Fill in the meeting type, date, participants, and notes.
  4. Click Save.

Correcting AI Classification #

If the AI misclassified a document (e.g., categorized a meeting request as a cover letter), click the entry and change its category using the type dropdown. Your correction is saved immediately.

Exporting #

Click Export to download all correspondence entries as a spreadsheet. This is useful for populating IND Annual Reports or preparing for FDA meetings where you need a correspondence history.

Frequently Asked Questions #

Why does “Scan M1” find nothing?
The dossier may not have Module 1 documents, or the eCTD structure may not have recognizable cover letter or form sections. Try Crawl Content to do a text-based search instead.

Can I link a response to its original question?
Yes. Click an entry and use the Link to field to associate it with another entry (e.g., link an FDA response to the original information request).

How do I refresh after importing a new submission?
Click Refresh in the toolbar, then re-run Scan M1 and Crawl Content to include the new sequence’s documents.

Leave a Reply

Your email address will not be published. Required fields are marked *