OpenFDA Adverse Events & Recalls Scraper

Extract FDA safety data at scale — drug adverse events (FAERS), device adverse events (MAUDE), drug / food / device recalls, device clearances and approvals, drug labels, and the NDC/UDI directories — one record per row, ready for a pharmacovigilance, risk, or litigation pipeline.

Built for pharma and medical-device safety teams, insurers, medical-malpractice and product-liability attorneys, and life-science analysts.

---

Why use this actor

• No account, no API key required.
• Enormous volume — tens of millions of records available (drug adverse events 20M+, device adverse events 24M+, plus recalls, clearances, and the device UDI directory at 4.9M+).
• One record per row, with the FDA's own fields kept as-is.
• Targeted or broad — pull everything, or narrow with a search query (by drug, device, date, manufacturer, or reaction).

How it works

1. You pick one or more FDA datasets from the list (drug events, device events, recalls, clearances, labels…).
2. (Optional) Add a search query to focus on a specific drug, device, manufacturer, or date range.
3. The actor collects every matching record and streams it into your dataset, ready to download as JSON, CSV, or Excel.

You don't manage any scrapers, accounts, or infrastructure — just pick datasets and press Run.

Use cases

Drug & device safety monitoring (pharmacovigilance) Pull adverse events for a specific product and track report volume and seriousness over time to spot emerging safety signals.

Recall intelligence Pull drug/food/device recalls filtered by date to monitor competitor and supplier recalls as they're posted.

Litigation & claims research Pull adverse events and recalls tied to a product line to build an evidence base for product-liability or malpractice work.

Input

Field	Type	Description
endpoints	multi-select	FDA datasets: Drug events (FAERS), Device events (MAUDE), Drug/Food/Device recalls, Device clearances, Drug labels, NDC/UDI directories.
search	string	Optional FDA query, e.g. patient.drug.medicinalproduct:"aspirin" or receivedate:[20260101 TO 20261231].
apiKey	string	Optional free FDA key — raises the daily limit for big jobs.
pageSize	integer	1–1000 (default 1000).
maxItemsPerEndpoint	integer	0 = up to the per-query maximum.

endpoints is a dropdown of fixed datasets — no codes to remember.

Output — one record per FDA record

{
  "_input": "drug-recall",
  "_source": "S1-openfda",
  "_scrapedAt": "2026-06-03T10:00:00Z",
  "recordType": "FDA_RECORD",
  "fda_endpoint": "drug-recall",
  "recall_number": "D-1234-2026",
  "status": "Ongoing",
  "recalling_firm": "...",
  "reason_for_recall": "...",
  "product_description": "...",
  "classification": "Class II",
  "report_date": "20260214"
}

(Fields vary per dataset — the FDA's field names are passed through as-is.)

Tip for very large pulls

A single search returns up to ~26,000 rows. To pull more for a broad dataset, split by date and run the actor per slice, e.g. receivedate:[20260101 TO 20260331], then the next quarter. The run log tells you when a query is larger than one pull.

Known limits

• Public FDA data — no account needed. Add a free FDA key for higher daily throughput on big jobs.
• Verified live: drug adverse events 20.3M, device adverse events 24.7M, device UDI directory 4.9M.