Cross-source identifiers
Every current identifier attachment, with the link method and confidence carried verbatim — deterministic links are 1.000; probabilistic links surface their real matcher confidence, never rounded up.
| Scheme | Value | Source table | Link method | Confidence | Since | Provenance |
|---|---|---|---|---|---|---|
| cms_op_profile_id | 11550205 | cms_open_payments | exact_npi_in_open_payments | 1.000 | 2026-06-13 | evidencerecorded: 2026-06-13 methodology: entity-graph/v1 evidence: {"npi":"1003391046","rule":"D-OP-NPI","recipient_profile_id":"11550205"} |
| npi | 1003391046 | nppes_providers | deterministic_pk | 1.000 |
Graph edges
One-hop relationships — ownership, chain membership, operation — each backed by a federal source row.
| Relationship | Entity | Confidence | Source table | Provenance |
|---|---|---|---|---|
| ← pays | Novartis Pharmaceuticals Corporation (NJ) | 1.000 | cms_open_payments | evidence |
Current record
Every current claim asserted about this entity, grouped by provenance subject. Each cell carries its upstream federal source.
provider · 1003391046 JSON
Provider screening signals
NPI-keyed source families read directly for 1003391046: federal and state integrity lists, CMS Open Payments, Order and Referring eligibility, Medicare opt-out, and enrollment revalidation. Each cell carries its own source and 14-tuple provenance. A screening aid — re-confirm any match against the primary source before acting; absence means nothing on file in the data Fonteum currently holds, never a guarantee none exists.
Sanctions & integrity
None on fileNothing on file across the 14 integrity lists Fonteum currently holds — not a guarantee none exists elsewhere.
- Screened: OIG LEIE, SAM.gov Exclusions, NY Medicaid (OMIG), OH Medicaid (ODM), GA Medicaid (DCH OIG), PA Medicaid (Medicheck), NC Medicaid (DHHS), MD Medicaid, WA Medicaid, IA Medicaid, MS Medicaid, MT Medicaid, OIG CIA, CMS Penalties
14-tuple provenance (14)
Point-in-time
Reconstruct this record as of any date, or diff it between two dates — both axes of the bitemporal store (bitemporal-read/v1): valid time (when a fact was true per the source) and recorded time (what the store knew when).
Pick a date to reconstruct, or two dates to diff. Equivalent JSON: /api/v1/history/{subject_type}/{subject_id}?at=… and …/diff?from=…&to=…