US FDA Safety Recalls & Enforcement Actions — 45,000 Records (Food, Drug, Device)

About this data

Wide-coverage dataset of 45,000 US FDA enforcement actions spanning food, drug, and medical device recalls. Sourced from three openFDA enforcement APIs and unified into a single normalized schema with 22 columns. **Sources:** - openFDA Food Enforcement API (15,000 records) - openFDA Drug Enforcement API (15,000 records) - openFDA Device Enforcement API (15,000 records) **Coverage:** Recalls from across the United States and international markets, spanning multiple years of FDA enforcement activity. **Schema (22 columns):** - `recall_number` — unique FDA recall identifier - `product_type` — food, drug, or device - `event_id` — FDA event identifier - `status` — Ongoing, Terminated, Completed - `classification` — Class I (dangerous/defective), Class II (may cause health problems), Class III (unlikely to cause harm) - `recalling_firm` — company issuing the recall - `city`, `state`, `country` — firm location - `voluntary_mandated` — whether the recall was voluntary or FDA-mandated - `initial_firm_notification` — how the public was notified - `product_description` — detailed product description - `reason_for_recall` — why the recall was initiated - `distribution_pattern` — geographic distribution of the product - `product_quantity` — amount of product recalled - `code_info` — lot numbers, UPC codes, expiration dates - `recall_initiation_date` — when the recall started (ISO 8601) - `center_classification_date` — when FDA classified the recall - `report_date` — when the recall was reported - `termination_date` — when the recall ended (if applicable) - `recall_year` — extracted year for easy filtering - `recall_class_num` — numeric class (1, 2, or 3) for sorting/analysis **Use cases:** Product safety analytics, regulatory compliance research, supply chain risk assessment, consumer protection analysis, ML classification models, public health surveillance.

Schema

Sample Data

Preview a sample of the data before downloading.