Skip to content

FDA_CRL_API.md

Latest commit

 

History

History
372 lines (288 loc) · 9.92 KB

FDA_CRL_API.md

File metadata and controls

372 lines (288 loc) · 9.92 KB

FDA Complete Response Letters (CRL) API Documentation

Table of Contents

Overview

The FDA Complete Response Letters API provides public access to correspondence regarding New Drug Applications (NDAs) and Biologics License Applications (BLAs). This transparency initiative, launched on July 10, 2025, represents a major shift in FDA operations toward radical transparency.

Key Statistics:

  • Total Letters: 392 (as of November 6, 2025)
  • Approved Applications: 295
  • Unapproved Applications: 97
  • Time Coverage: 2020-2024

What are Complete Response Letters?

Complete Response Letters (CRLs) are decision letters issued by the FDA when it completes its review of a drug or biological product application but cannot grant approval in its current form.

Common reasons for CRLs:

  • Safety and efficacy concerns
  • Manufacturing deficiencies
  • Bioequivalence issues
  • Prescribing information/labeling deficiencies
  • Incomplete safety data

The letters detail:

  • Specific deficiencies that must be addressed
  • FDA recommendations for resolving issues
  • Required resubmission timeline (typically within one year)

Why This Matters

FDA's Transparency Initiative

Previously, pharmaceutical sponsors often minimized or omitted FDA concerns when publicly announcing application rejections. A 2015 FDA analysis found that sponsors avoided mentioning 85% of the FDA's concerns about safety and efficacy in their public disclosures.

Benefits of Public CRLs:

  1. Industry Learning: Companies can learn from others' mistakes rather than repeating them
  2. Predictability: Better understanding of FDA expectations
  3. Faster Development: Clear guidance accelerates getting treatments to patients
  4. Public Trust: Transparency in the drug approval process

Important Caveats:

  • Letters are redacted to protect trade secrets and confidential commercial information (CCI) per FOIA regulations
  • Data has not been validated for clinical or production use
  • Should not be used to make medical care decisions

API Access

Base Endpoint

https://api.fda.gov/transparency/crl.json

Authentication

No API key required (public access)

Rate Limits

Not specified in documentation (standard openFDA limits likely apply)

Alternative Access Methods

Bulk Downloads:

  • Approved CRLs: https://download.open.fda.gov/approved_CRLs.zip
  • Unapproved CRLs: https://download.open.fda.gov/unapproved_CRLs.zip

Interactive Search:

  • Web interface: https://open.fda.gov/crltable/

Data Structure

Response Format

All API responses contain two main sections:

{
  "meta": {
    "disclaimer": "...",
    "terms": "https://open.fda.gov/terms/",
    "license": "https://open.fda.gov/license/",
    "last_updated": "2025-11-06",
    "results": {
      "skip": 0,
      "limit": 1,
      "total": 392
    }
  },
  "results": [...]
}

Available Fields

Field Name Type Description Example
letter_date String Date the letter was issued "04/25/2025"
letter_year String Year of the letter "2025"
letter_type String Type of letter "COMPLETE RESPONSE"
file_name String PDF filename "CRL_NDA215818_20250425.pdf"
approval_status String Current status "Approved" or "Unapproved"
application_number Array FDA application number(s) ["NDA 215818"]
company_name String Applicant company name "Fresenius Kabi USA, LLC"
company_address String Company mailing address "Three Corporate Drive\nLake Zurich, IL 60047"
company_rep String Company representative "Jennifer Gross, MS"
approver_name String Approving organization "U.S. Food and Drug Administration"
approver_center Array FDA center(s) involved ["Center for Drug Evaluation and Research"]
approver_title String Approver location "Silver Spring, MD 20993"
text String Full text content of the letter (lengthy text)

Query Syntax

Basic Query Structure

All queries use URL parameters joined by &:

https://api.fda.gov/transparency/crl.json?<parameter>=<value>&<parameter>=<value>

Search Operations

Single Field Search:

search=field:term

Boolean AND (both conditions must match):

search=field1:term1+AND+field2:term2

Boolean OR (either condition matches):

search=field1:term1+field2:term2

Phrase Matching (exact phrase):

search=field:"exact phrase"

Exact Field Matching:

search=field.exact:"whole phrase"

Pagination

Limit results:

limit=25

Skip results (offset):

skip=50

Counting & Aggregation

Count unique values:

count=field_name

Count exact phrases:

count=field_name.exact

Sorting

Sort results:

sort=field_name:desc    # descending
sort=field_name:asc     # ascending

Examples

1. Get First Record

GET https://api.fda.gov/transparency/crl.json?limit=1

2. Search by Approval Status

GET https://api.fda.gov/transparency/crl.json?search=approval_status:unapproved&limit=10

3. Search by Company Name

GET https://api.fda.gov/transparency/crl.json?search=company_name:"Pfizer"&limit=5

4. Search by Year

GET https://api.fda.gov/transparency/crl.json?search=letter_year:2024&limit=20

5. Complex Search (AND operator)

GET https://api.fda.gov/transparency/crl.json?search=approval_status:unapproved+AND+letter_year:2024&limit=10

6. Count by Approval Status

GET https://api.fda.gov/transparency/crl.json?count=approval_status

Response:

{
  "results": [
    {"term": "approved", "count": 295},
    {"term": "unapproved", "count": 97}
  ]
}

7. Count by FDA Center

GET https://api.fda.gov/transparency/crl.json?count=approver_center.exact

8. Search with Pagination

GET https://api.fda.gov/transparency/crl.json?search=letter_year:2023&limit=25&skip=0

9. Search Text Content

GET https://api.fda.gov/transparency/crl.json?search=text:"manufacturing deficiency"&limit=5

10. Multiple Criteria with OR

GET https://api.fda.gov/transparency/crl.json?search=letter_year:2023+letter_year:2024&limit=10

Data Coverage & Limitations

Coverage

  • Time Period: Applications from 2020-2024
  • Application Types: NDAs (New Drug Applications) and BLAs (Biologics License Applications)
  • Total Documents: 392 letters (as of November 6, 2025)
  • Update Frequency: Infrequent updates; FDA continues to publish archived CRLs

Limitations

Redactions:

  • Trade secrets removed per FOIA regulations
  • Confidential commercial information (CCI) protected
  • Some sections may show "(b)(4)" redaction markers

Data Validation:

  • Not validated for clinical or production use
  • Should not be relied upon for medical care decisions
  • Informational purposes only

API Constraints:

  • No real-time updates
  • Historical data only (2020-2024)
  • Some fields may contain redacted or incomplete information

Important Disclaimers

From FDA documentation:

"Do not rely on openFDA to make decisions regarding medical care. While we make every effort to ensure that data is accurate, you should assume all results are unvalidated."

Example Response

{
  "meta": {
    "disclaimer": "Do not rely on openFDA to make decisions regarding medical care...",
    "terms": "https://open.fda.gov/terms/",
    "license": "https://open.fda.gov/license/",
    "last_updated": "2025-11-06",
    "results": {
      "skip": 0,
      "limit": 1,
      "total": 392
    }
  },
  "results": [
    {
      "letter_date": "04/25/2025",
      "approver_title": "Silver Spring, MD 20993",
      "file_name": "CRL_NDA215818_20250425.pdf",
      "letter_year": "2025",
      "approval_status": "Unapproved",
      "approver_name": "U.S. Food and Drug Administration",
      "approver_center": [
        "Center for Drug Evaluation and Research"
      ],
      "company_rep": "Jennifer Gross, MS",
      "company_address": "Three Corporate Drive\nLake Zurich, IL 60047",
      "company_name": "Fresenius Kabi USA, LLC",
      "text": "[Full letter content - typically several paragraphs describing deficiencies and required actions]",
      "application_number": [
        "NDA 215818"
      ],
      "letter_type": "COMPLETE RESPONSE"
    }
  ]
}

Technical Implementation Notes

API Technology

  • Built on Elasticsearch
  • Returns JSON format by default
  • RESTful API design

Best Practices

  1. Use pagination for large result sets (limit + skip)
  2. Use count endpoint for statistics before fetching full results
  3. Cache responses when appropriate (data updates infrequently)
  4. Handle redactions in text field - expect "(b)(4)" markers
  5. Validate data - don't assume completeness or accuracy for critical decisions

Common Use Cases

  • Researching FDA rejection patterns
  • Understanding common deficiencies by therapeutic area
  • Tracking specific companies' application histories
  • Analyzing trends in FDA requirements over time
  • Educational/academic research on drug approval process

Additional Resources

Last Updated: November 9, 2025