FDA data sources

Why source transparency matters

A derivative index is only as trustworthy as its sources and its honesty about them. Because this site reorganises public FDA data rather than originating it, the most important thing it can do is be explicit about exactly where each field comes from and what that source does and does not guarantee. The four sources below are the entirety of the site's data lineage. We do not blend in undisclosed sources, and we do not present any value that cannot be traced back to one of them.

1. openFDA Drug Enforcement API (recalls)

• Official URL: https://open.fda.gov/apis/drug/enforcement/
• What it contains: The FDA's drug recall enforcement reports — recall number, event ID, classification (Class I/II/III), status, product description, the recalling firm's stated reason, distribution pattern, geographic fields, and the report, initiation and termination dates.
• Limits: openFDA notes that its data is provided for research and is not a substitute for the official, legally controlled record; field completeness varies between reports; and there can be a lag between an action and its appearance in the API. The API is rate-limited.
• Expected update frequency: openFDA refreshes enforcement data periodically; an exact cadence is not guaranteed by the agency, so this site treats the data as best-effort and timestamps every ingestion.

This source powers the drug recalls hub and the per-class recall pages.

2. openFDA Drug Shortages API (shortages)

• Official URL: https://www.accessdata.fda.gov/scripts/drugshortages/
• What it contains: The FDA's drug-shortage records — affected product (generic and brand name where available), reporting company, shortage status, presentation, availability text, and initial-posting and last-updated dates.
• Limits: Field names and availability of the shortages dataset have varied across versions; not every entry populates every field; and a shortage's live status can change faster than any periodic snapshot. The data describes supply, not clinical impact.
• Expected update frequency: The FDA updates shortage information as situations change; a fixed cadence is not guaranteed, so currency is best confirmed against the official source for time-sensitive decisions.

This source powers the drug shortages hub and the current-shortages page.

3. Drugs@FDA (approvals)

• Official URL: https://www.accessdata.fda.gov/scripts/cder/daf/
• What it contains: FDA-approved drug products — application numbers, sponsors (application holders), products, product forms, and marketing status. It is the canonical reference for what has been approved for marketing.
• Limits: A dependable, browser-free, machine-readable ingestion of the full dataset is not guaranteed within this MVP's scope, which is why the approvals layer may be deferred. The dataset does not, by itself, describe shortages or recalls.
• Expected update frequency: The FDA updates Drugs@FDA regularly, but a guaranteed machine cadence for this project is not guaranteed; the approvals layer is populated only when a reliable source path is available.

This source underpins the drug approvals page.

4. Orange Book data files (approvals reference)

• Official URL: https://www.fda.gov/drugs/drug-approvals-and-databases/orange-book-data-files
• What it contains: Approved drug products with therapeutic-equivalence evaluations, plus patent and exclusivity data, published by the FDA as downloadable data files.
• Limits: The therapeutic-equivalence codes are regulatory classifications and are not, on this site, used to make any clinical or substitution recommendation. Bulk-file ingestion is subject to the same dependability constraint as Drugs@FDA.
• Expected update frequency: The FDA publishes Orange Book data files on a periodic basis; the exact machine cadence for this project is not guaranteed.

This source is a reference for the approvals layer and is cross-referenced from the approvals page.

A note on update cadence and "not guaranteed"

You will notice that the expected update frequency for every source above is described as not guaranteed. This is deliberate and honest. The FDA refreshes these datasets on its own schedule, and openFDA documents that its data is provided for research rather than as a real-time, legally controlled feed. Rather than imply a precise cadence the agency does not promise, this site timestamps every ingestion and exposes that timestamp on the freshness page and in /freshness.json. The pipeline itself runs daily, but how much new data each run finds depends entirely on what the FDA has published since the previous run.

The practical implication for you is simple: for anything time-sensitive, treat this site as a fast index and confirm the live value at the official source. The freshness page tells you when the data was last checked, and every record links to the regulator's current copy. This combination — a quick, navigable index plus an always-available path to the authoritative record — is the whole design intent of the project.

Verification and attribution

Every record page on the site carries a source-and-provenance block linking to the specific openFDA query or FDA page that returns the original record, together with the source dataset name, the ingestion timestamp and the record's content hash. This is intentional: the site's role is to help you find and date the official record quickly, then hand you straight to the regulator's authoritative copy for verification. If a value here ever disagrees with the FDA's live record, the FDA's record governs. For definitions of the terms used across these sources, see the regulatory intelligence glossary; for the full pipeline, see the methodology.