📋 FDA Warning Letters + 483 Inspection Observations Tracker

The joined regulatory feed pharma compliance teams actually need. Every Warning Letter and Form 483 the FDA has issued in the last 5 years, mapped to the same facility via FDA Establishment Identifier (FEI), with a per-facility severity score derived from prior actions. Competing Apify actors ship raw HTML scrapes of a single Warning Letters page; this actor ships the joined CMO/CDMO supplier-risk feed that Verisk SteriQual and 3E sell at five figures a year.

The differentiator is the join — match Warning Letters on Pfizer's New Jersey plant to the four Form 483s the same plant has received in the last 18 months, and you get a "recurring" severity tag that drives supplier-risk decisions. Raw HTML scrapers can't do this because they only see one side of FDA's enforcement universe.

🧠 What you get

Field	Notes
type	warning_letter or form_483
date	Letter issue date or Form 483 issuance date (ISO-8601)
manufacturer / facility	Company Name as published by the FDA
fei_number	FDA Establishment Identifier — canonical facility join key
state, country	From the 483 FOIA index (Warning Letters feed has implicit US default)
product_type	drug / device / food / biologic / tobacco / veterinary / any — inferred from FDA Issuing Office (Warning Letters) or Establishment Type (Form 483s)
issuing_office	Full FDA office string for Warning Letters (e.g. "Center for Drug Evaluation and Research (CDER)")
subject / citation_summary	Subject line of the Warning Letter, or Form 483 — <establishment-type> inspection
citations[]	List of 21 CFR cites parsed from the Warning Letter HTML (e.g. 21 CFR 211.113, 21 CFR 820.30(a)) — Warning Letters only, when the letter URL is resolvable
fdc_act_citations[]	FD&C Act section cites (e.g. Section 501(a)(2)(B))
response_deadline_15_working_days	Boolean — whether the letter requires the standard 15-working-day response
fda_division	The specific FDA division issuing the letter, when parseable from the HTML
response_letter_received / closeout_letter_issued	Booleans from the Warning Letters feed
full_text_url	Direct link to the FDA's published letter page (or FOIA index for 483s)
establishment_type	For Form 483s — e.g. "Outsourcing Facility", "Producer of Sterile Drug Products"
facility_severity_score	{score, label, prior_warning_letters, prior_form_483s} — clean (0 priors), single-prior (1), recurring (2-3), systemic (4+)

🎯 Use cases

• Pharma QA/QC director — Daily monitoring of CDER Warning Letters against current finished-pharmaceutical CMOs; the severity score surfaces any contractor with a recurring 21 CFR 211.x posture before audit week.
• CMO/CDMO supplier risk team — Cross-reference every active supplier's FEI against the joined feed; flag any facility with severity_label="recurring" or "systemic" for re-qualification or fallback-sourcing.
• FDA defense counsel — Pull all Warning Letters + 483s for a client facility, sorted by date, in one CSV — replaces the multi-tab spreadsheet that junior associates spend a day building per matter.
• Investment due diligence — Pre-IPO/M&A: scan the target's manufacturing footprint against the feed; systemic severity is a Section 8 closing condition issue.
• Plaintiff law / mass tort intake — Tie individual product complaints (FAERS / MAUDE) to the issuing facility's regulatory posture for proximate-cause arguments.

📊 Sample input

{
  "manufacturer_name": "",
  "product_type": "drug",
  "days_back": 180,
  "severity_filter": "both",
  "max_records": 50
}

📊 Sample output

{
  "type": "warning_letter",
  "date": "2026-04-07",
  "posted_date": "2026-04-28",
  "manufacturer": "Lexia LLC",
  "facility": "Lexia LLC",
  "state": null,
  "country": null,
  "fei_number": null,
  "product_type": "drug",
  "issuing_office": "Center for Drug Evaluation and Research (CDER)",
  "subject": "CGMP/Finished Pharmaceuticals/Adulterated",
  "citation_summary": "CGMP/Finished Pharmaceuticals/Adulterated",
  "citations": ["21 CFR 211.22", "21 CFR 211.100", "21 CFR 211.113"],
  "fdc_act_citations": ["section 501(a)(2)(B) of the FD&C Act"],
  "response_deadline_15_working_days": true,
  "fda_division": "Center for Drug Evaluation and Research (CDER)",
  "response_letter_received": false,
  "closeout_letter_issued": false,
  "full_text_url": "https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/lexia-llc-...",
  "facility_severity_score": {
    "score": 2,
    "label": "recurring",
    "prior_warning_letters": 0,
    "prior_form_483s": 2
  },
  "source": "FDA Warning Letters"
}

And a Form 483 record on the same facility:

{
  "type": "form_483",
  "date": "2026-04-17",
  "publish_date": "2026-05-27",
  "manufacturer": "BPI Labs, LLC",
  "facility": "BPI Labs, LLC",
  "state": "Florida",
  "country": "USA",
  "fei_number": "3015156709",
  "product_type": "drug",
  "establishment_type": "Outsourcing Facility",
  "record_type": "483",
  "observation_count": null,
  "severity": "single-prior",
  "citation_summary": "Form 483 — Outsourcing Facility inspection",
  "full_text_url": "https://www.fda.gov/about-fda/office-regulatory-affairs/ora-foia-electronic-reading-room",
  "facility_severity_score": {
    "score": 1,
    "label": "single-prior",
    "prior_warning_letters": 1,
    "prior_form_483s": 0
  },
  "source": "FDA OII FOIA Electronic Reading Room (Form 483 index)"
}

🚀 How to use

Run the actor from the Apify console, or programmatically:

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("nexgendata/fda-warning-letters-483-inspections").call(
    run_input={
        "manufacturer_name": "Intas",
        "product_type": "drug",
        "days_back": 365,
        "severity_filter": "both",
        "max_records": 100,
    }
)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["date"], item["type"], item["manufacturer"],
          item["facility_severity_score"]["label"])

cURL:

curl -X POST "https://api.apify.com/v2/acts/nexgendata~fda-warning-letters-483-inspections/runs?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"product_type":"drug","days_back":180,"max_records":50}'

⚖️ Compliance & data sources

All sources are published by the U.S. Food and Drug Administration under federal public-disclosure mandate (Federal Food, Drug, and Cosmetic Act § 705, 21 U.S.C. § 375) and are explicitly machine-readable via the XLSX export endpoints:

• Warning Letters: fda.gov/inspections-compliance-enforcement-and-criminal-investigations/compliance-actions-and-activities/warning-letters/datatables-data — covers ~1,000 letters, ~5-year rolling window.
• Form 483 Inspection Observation Index: fda.gov/about-fda/office-regulatory-affairs/ora-foia-electronic-reading-room/download — OII FOIA Electronic Reading Room export, ~2,000 records, covering 483s + EIRs + closeout reports.

All HTTP requests are sent with the polite-bot User-Agent NexGenData FDA Warning Letters Actor hello@thenextgennexus.com. No anti-bot measures encountered. No FDA rate limits documented for these XLSX endpoints. The actor downloads each XLSX once per run (no per-record requests against FDA).

Future enhancements (paid add-ons under consideration): OCR of the underlying Form 483 PDFs to extract per-observation numbers (currently the FOIA index publishes the count of inspections but not the per-observation cite count from each 483 PDF), HTML enrichment of older Warning Letters whose URLs aren't on the recent-letters HTML index, and a Schedule-A export of the facility severity index alone (for supplier-onboarding KYC use).

🔗 Related Actors — NexGenData FDA / Healthcare Cluster

Actor	What it does
⚠️ FDA Recall Monitor	Drug + device safety alerts (recalls, withdrawals) — pairs with this actor for full FDA post-market surveillance
Federal Register Rules Scraper	Daily federal-register rule publications across all agencies — catches FDA proposed-rules that drive future Warning Letter citations
NIH Reporter Grants Scraper	Federal biomedical research funding — counterpart to FDA enforcement for the full healthcare-funding-and-enforcement picture
⚕️ Doctor Directory Lead Finder	NPI-registry-backed physician lookups — useful for the pharma-payment + adverse-event analyst stack

🏷️ About NexGenData

NexGenData publishes the deepest set of public-record data actors on Apify — SEC EDGAR forms (NPORT-P, 8-K, Form 4, Form D, 13F, 13D/G, ADV), CISA KEV + EPSS, DOL Form 5500, FDA Warning Letters + 483s + Recalls, FDA Adverse Events (FAERS+MAUDE), FINRA BrokerCheck, EUIPO trademarks, WIPO Patentscope, and more — all priced as pay-per-event so you only pay for the data you actually pull. Visit apify.com/nexgendata for the full catalog.