Our Data

Public data, fused and graded.

Six public datasets linked per provider, 2.5M+ U.S. physicians scored in depth — and a provenance model that tells you exactly where every field came from and how much to trust it.

01  /  Sources

Six datasets, linked via NPI.

NPI Registry (NPPES)
The provider backbone: identity, taxonomy/specialty, practice addresses, credentials.
Every NPI-registered provider in the United States
CMS Part B utilization
Procedure volumes and service utilization — who actually performs the procedures you care about.
Medicare Physician & Other Practitioners public-use files
CMS Part D prescribing
Prescribing patterns by drug and drug class.
Medicare Part D prescriber public-use files
MIPS quality data
Quality performance measures and participation status.
CMS Quality Payment Program public data
Open Payments (Sunshine Act)
Industry relationships: speaker fees, consulting, research funding, royalties, ownership — rolled up by manufacturer and payment type, with engagement flags.
Payment-level records, five-year rolling window
State medical boards
Licenses, board certifications, multi-state licensure, and disciplinary actions.
State-by-state board records

Agentic research adds cited enrichment on demand: PubMed (publication history and research activity) · ClinicalTrials.gov (clinical trial involvement and investigator roles) · Google Places (practice business details, reviews, and hours).

02  /  Provenance

Every field knows where it came from.

Each enriched field — a direct email, a publication count, a hospital affiliation — carries its source, its verification status, and when it was last checked. Contact points get reach labels (direct, office, switchboard, mobile) and confidence grades, and when sources disagree the record shows the conflict instead of hiding it. Research findings that make a claim about a provider must cite a source — the agents are not allowed to assert a clean disciplinary record without an official source to point to.

Verified

Confirmed against a cited source, with a checked-at date.

Inferred

Derived from signals — always labeled as such, never dressed up as verified.

Unknown

We say so. An honest gap beats a confident guess your rep dials.

03  /  Stewardship

No PHI. Real opt-out.

Everything above is public professional data — never patient data. Providers who don't want to be in the platform can opt out, and suppression applies everywhere: search, campaigns, dossiers, and exports. Read how we handle provider data and our security overview.

04  /  FAQ

Data questions.

Is any of this patient data?

No. Every dataset is public professional data about providers — identity, specialties, procedure volumes, prescribing patterns, industry payments, licensure. There is no patient data and no PHI anywhere in the platform.

Which datasets does Oystercatcher use?

Six public datasets linked per provider via NPI: the NPI Registry (NPPES), CMS Part B utilization, CMS Part D prescribing, MIPS quality data, Open Payments (Sunshine Act), and state medical boards — plus cited on-demand enrichment from sources like PubMed and ClinicalTrials.gov.

How do I know whether a field is accurate?

Every enriched field carries its source, a verification status — verified, inferred, or unknown — and the date it was last checked. Contact points add reach labels and confidence grades, and when sources disagree the record shows the conflict instead of hiding it. Read the methodology.

Can providers opt out of Oystercatcher?

Yes. Providers who do not want to be in the platform can opt out, and suppression applies across search, campaigns, dossiers, and exports. Provider data & opt-out.

Closing  ·  spend your time selling

Let the agents
do the research.
Your team does
what it does best.

Want proof first? Read an example dossier — sample data, real mechanics.