MDClarity · Data foundation

Data stewardship

Four explainers on one problem: a price is only correct as of a date. Who the provider was affiliated with, which location the service happened at, which TIN billed it, and which contract governed it are all facts that changed — and a claim priced against today's picture of the world is priced against the wrong world.

They read in order — the problem, why the obvious public source can't solve it, the data model that can, and the human workflow that keeps that model true — but each stands alone.

The explainers

4 pieces
  1. 1 The problem Point-in-Time Claim Pricing Explorer

    What a claim actually observes versus what must come from effective-dated master data — and how the same claim prices differently depending on which point in time you resolve its context against.

    interactive scenario walkthrough /point-in-time-pricing/
  2. 2 Why the registry isn't the answer NPPES vs Master Data

    NPPES is real identity and reference data — the risk is treating it as historical contractual truth. Six cases where that substitution prices a claim wrong: moves, concurrent sites, credentialing lag, service-location drift, taxonomy, and TIN/NPI pairing.

    interactive 6 failure cases /nppes-pricing/
  3. 3 The model that can answer it Temporal Healthcare Affiliation Graph

    Durable entities are nodes; pricing-relevant affiliations are effective-dated edges. Move the date of service and watch which relationships are valid, then open an edge for its provenance and its pricing significance.

    interactive 4 scenarios · DOS slider /affiliation-graph/
  4. 4 Keeping it true Location stewardship — sketch prototype

    Two systems joined at two points: the app fetches candidates from gold, a steward accepts / rejects / matches each source identity, and the ledger is promoted back to bronze. Live view of what every decision changes.

    prototype steward workflow /location-stewardship/