Physician Knowledge Graph.

Identity. Professional graph. Institutional graph. Entity-reconciliation chain.

The Physician Knowledge Graph entry contains four layers. Identity: name, identifiers including NPI, image, jobTitle. Professional graph: specialty alignment to MedicalSpecialty, sub-specialties, board-certification status, active state licensure. Institutional graph: hospitalAffiliation, memberOf for professional society membership, alumniOf for medical-school education. Entity-reconciliation chain: sameAs pointing to NPPES, ABMS verification, state medical board licensure profile, LinkedIn professional profile, peer-reviewed publication index ¹ .

Each property does specific work. usNPI ties the identity to the federal registry. medicalSpecialty aligns to the controlled vocabulary. hospitalAffiliation chains the practice to an institutional entity. sameAs distributes the trust load across multiple independent surfaces. The byline alone is just a claim; the chain is the proof.

Federal entity key. Persists across employment.

The NPI is the load-bearing federal entity key. The 10-digit intelligence-free number issued by CMS through the NPPES registry persists with the physician across employment changes, location changes, and specialty additions ² . Schema.org's usNPI property on the Physician type allows direct machine-readable linkage to the NPPES entry.

Google's Knowledge Graph reconciliation uses the NPI to cross-reference the physician identity against state medical board licensure databases ⁵ , third-party physician directories (Healthgrades, Zocdoc, Vitals), and the AMA Physician Masterfile ³ . Without the NPI in the schema, Google has to infer identity from name and address, which is fragile and error-prone for common names (the 'Dr. John Smith' problem).

Independent surfaces. Distributed trust load.

sameAs on the Physician type takes an array of canonical URLs identifying the same entity across the web. The chain typically includes: NPPES profile, ABMS board-certification verification ⁴ , state medical board licensure profile (per state of active licensure), Healthgrades profile, Zocdoc profile, Doximity profile, LinkedIn professional profile, and the physician's PubMed or Google Scholar author profile when applicable.

Each URL is independently verifiable. Google's Knowledge Graph reads the chain as 'these surfaces describe the same person.' The trust load distributes across the chain. A strong ABMS surface plus a strong NPPES surface plus a strong state-board surface stabilizes the entity even when one surface ages or moves. The chain is also forward-compatible: when the physician adds a sub-specialty, claims a new directory profile, or publishes a peer-reviewed article with a Scholar profile, the new URL extends the chain rather than disrupting it.

hospitalAffiliation. memberOf. Institutional EEAT carried into the byline.

hospitalAffiliation on the Physician type takes a Hospital entity (typed as a MedicalBusiness subtype) the physician practices at or has admitting privileges at. The property establishes institutional EEAT: a physician affiliated with a teaching hospital or a recognized medical center carries the institution's EEAT in the chain. The property takes a typed entity (not just a text string), so the chain extends through the affiliated institution's own Knowledge Graph entry.

memberOf takes professional society memberships: American College of Cardiology, American Academy of Dermatology, American Society of Plastic Surgeons, Royal College of Surgeons, ESC. Society membership is a credential-adjacent signal the Knowledge Graph reads alongside board certification. The property takes the Organization entity of the society, again extending the chain through the society's own Knowledge Graph entry.

alumniOf takes the medical-school entity. The medical-school chain is the historical depth signal. A physician's residency-training program (separate from medical school) gets surfaced through the JSON-LD hasCredential property when the program is documented in a verifiable directory.

The Physician Knowledge Graph entry is the load-bearing structural input to the medical practice SEO architecture at Praxis. The byline on every editorial article and every commercial author surface carries the chain.