Skip to content

Medical-device intelligence

Data sources, coverage, and methodology

How MedTech Watch collects FDA records, publishes complete snapshots, counts activity, and distinguishes source dates from detection dates.

What the public catalog covers

We collect complete snapshots of openFDA’s device classification, qualifying 510(k) clearance, and device recall datasets. Public history extends through the available records in those sources. It is not a claim to every device authorization or every record held by FDA.

De Novo identifiers and decisions identified as withdrawn, deleted, or not substantially equivalent are excluded from the current 510(k) clearance catalog. PMA and De Novo decision catalogs are outside current coverage.

Complete imports, then publication

Imports follow pagination, preserve checkpoints, and verify record counts and the upstream version before publication. The public catalog uses a complete release. Failed or interrupted imports leave the last successful release available and are shown as delayed updates.

Records no longer present in a later source release are retained as labeled archives outside current totals. Source checks and content revisions do not automatically rewrite historical publication dates.

Company names and device categories

Public company profiles group FDA names with equivalent normalized spelling. Punctuation and case can be normalized; subsidiaries, acquisitions, and parent-company relationships are not inferred. Customer-entered watchlist aliases remain private.

Device-category hubs use explicit curated lists of FDA product codes, checked against classification records. The scope of each grouping is displayed. These research groupings are not classifications of a particular new device.

Recall classification

The device-recall dataset is enriched with a recall class only when its recall number matches an enforcement record. Without a confirmed match, classification remains unknown. Device class and recall class are not interchangeable.

Dates, freshness, and counts

Clearance pages label the decision date. Recall listings use the FDA classification date where available, then the record creation date or initiation date, with the applicable label. Monthly reports use clearance decision dates and recall initiation dates.

Counts describe source records. Several recall records may belong to one recall event. Monthly totals exclude records without the required grouping date and disclose catalog-wide undated counts. Counts are not market share, product launches, incidence, causation, or safety comparisons.

openFDA’s clearance data updates monthly; recall and MAUDE data updates weekly. We display both the upstream dataset date and the last successful check. Daily checks do not imply daily FDA publication.

Paid watchlist coverage

Initial paid matching covers twelve months of clearances and ninety days of recalls and MAUDE reports. Historical matching establishes a baseline. Subsequent discoveries and updates retain account-specific matching dates.

MAUDE collection follows active customer criteria, uses overlapping change windows, and reconciles recent history. It is not a complete public MAUDE mirror. Reports may be incomplete, inaccurate, delayed, unverified, or biased; report counts cannot establish incidence, causation, or comparative device safety.

Primary sources

Device classifications

FDA dataset:

Last successful check: Nov 21, 2026

Source classifications may change.

510(k) clearances

FDA dataset:

Last successful check: Nov 21, 2026

FDA publishes this dataset monthly.

Recall records

FDA dataset:

Last successful check: Nov 21, 2026

FDA publishes this dataset weekly.

Keep this research moving.

Follow ten competitors and ten product codes. Get matching changes in your dashboard and a weekly brief, for $49/month.

Start a 14-day free trial ↗