The Blueprint: Inside the Four Components of the Healthcare Clearing Commission
Spoke 3

The Blueprint: Inside the Four Components of the Healthcare Clearing Commission

Here’s how the HCC works in practice: a Universal Clearing Layer, a Shared‑Risk Vault, Transparent Arbitration, and civic oversight. It’s plumbing for fairness—tech that makes trust verifiable.

The Architecture at a Glance

Flow: Service Node → Clear on UCL → Apply corridor → Settle member/share → Route outliers to Arbitration → Publish trace → Rebalance risk via SRV.

Each care event becomes a Node in an enumerated graph—drug, lab, imaging, visit. The UCL verifies parties, price elements, and corridors; SRV absorbs catastrophe; arbitration resolves disputes; oversight keeps models and margins honest.

1) Universal Clearing Layer (UCL)

The UCL is health finance’s equivalent of a settlement rail. It doesn’t deliver care; it verifies and settles truth before money moves.

Enumerated Health Graph & Node Model

A Node is more than a billing code. For a drug like insulin, the Node includes:

  • Clinical identifiers: RxNorm concept, strength, route, device, pack size; crosswalks to NDC (pharmacy) and HCPCS J‑codes (medical drug).
  • Cost stack: acquisition (e.g., WAC/NADAC/contracted), distribution markup, dispense fee/margin, admin fee, site‑of‑care factor.
  • Context: channel (retail/mail/specialty), site (outpatient/home infusion), geography, provider type, service date.
  • Governance: corridor parameters (reference cost ± percentage), audit trail, versioned rules.

Standards Integration (not replacement)

  • FHIR carries clinical context (MedicationRequest, CoverageEligibilityResponse). The UCL references FHIR resources to bind clinical intent to settlement.
  • X12 (837/835/270/271) and NCPDP (telecom, SCRIPT) remain payload channels. The UCL provides a translation layer that maps their fields to the Node schema, adding the cost‑stack and corridor metadata they lack.
  • Where standards are insufficient, extensions add fields (e.g., corridorId, referenceCostSource, siteOfCareWeight) without breaking legacy rails.

What clears per transaction

  • Identity & integrity of parties (NPI/Plan IDs) and Node metadata.
  • Cost stack elements and the corridor‑checked allowed price.
  • Member responsibility calculated off net (not list) with instant crediting if rebates or adjustments arrive later.
  • Signed trace (hash) posted to the public ledger for audit.

2) Shared‑Risk Vault (SRV)

The SRV is a pooled reserve that reinsures big shocks while routine care settles directly on the UCL. Two contribution logics matter:

Model How contributions are set Pros Tradeoffs
Tax‑Bracket Style % of income across bands (e.g., 0.8%, 1.2%, 1.8%…) Capacity‑based fairness; stable revenue; predictable for households Requires verified income data; policy choices on bands
Community‑Rated Flat rate per member in a region/plan Simplicity; easy enrollment Cross‑subsidy may feel blunt; can undercharge high‑risk groups

Actuarial logic (sketch): expected catastrophic cost per member (E[C]) is estimated from pooled variance; SRV sets layers (e.g., triggers at $50k annual spend) and collects contributions so that the probability of shortfall remains below a policy threshold (e.g., 0.5%). Surpluses generate dividends of wellness (lower future contributions or community benefits).

Routine care clears locally; catastrophe clears collectively.

3) Transparent Arbitration

Disputes are inevitable. What changes is where and how they resolve.

  • Open venue: cases are filed to a public docket; decisions are published and precedent‑linked.
  • Explainable algorithms: anomaly detection and corridor logic are open‑sourced; parties can simulate outcomes with the same code that will adjudicate them.
  • Time‑boxed SLA: minor disputes (coding/units) resolve in hours; complex cases in days; all with auditable timers.
  • Restitution by rule: if a member overpaid relative to net, refunds/credits are automatic—no “pay and chase.”

Justice, but operational: the goal isn’t lawsuits—it’s rapid truth, published once, reused forever.

4) Auditors & Committees (Civic Oversight)

Algorithms propose; humans ratify. Oversight balances speed with legitimacy.

  • AI Ethics Office: audits corridor models for bias (age, zip, comorbidity), drift, and disparate impact; publishes reports.
  • Rotating Committees: clinicians, patients, economists, ethicists, employers—each seat has a shadow seat (civic observer) who can trace and contest decisions.
  • Public Dashboard: corridor updates, anomaly rates, SRV solvency metrics, and arbitration stats visible to all.
  • Capture resistance: term limits, conflict disclosures, and open-source deliberation notes.

Practitioner Appendix: Technical Chops

A) UCL ↔ Standards: FHIR, X12, NCPDP

  • FHIR binding: tie MedicationRequest/Claim/ExplanationOfBenefit to Node IDs; use FHIR extensions for corridorId, referenceCost, costStack.
  • X12 mapping: 837 (claim) → Node + units; 835 (remit) → settlement record; 270/271 (eligibility) → corridor eligibility and member caps.
  • NCPDP: Telecom fields for ingredient cost/dispensing fee map directly to acquisition and dispense elements in the cost stack; UCL adds provenance + corridor validation.
  • Why not replace? Compatibility. We add an overlay schema (Node + cost stack + corridor metadata) that can be serialized alongside legacy transactions.

B) Insulin Node: more than NDC

  • Identifiers: RxCUI, NDC set, HCPCS J‑code (if buy‑and‑bill), GS1 barcode.
  • Clinical facets: analog vs human insulin, device (pen/vial/pump cartridge), titration guidance reference, interchangeability flag.
  • Cost stack facet: acquisitionBasis (WAC, NADAC, 340B, contract), distributionMarkup, dispenseFee, adminFee (medical), siteOfCareWeight.
  • Corridor policy: reference cost sourced from pooled data; allowed float ±X%; exceptions: clinical hardship or shortage rules.

C) SRV Contribution Models (sketch math)

Tax‑Bracket style: contribution per member = f(income) = Σ bandRate × bandIncome, capped at policy %. Community‑rated: = flatRate(plan, region). In both, catastrophic layers are priced via simulation of tail risk; solvency target sets reserves (e.g., 99.5% VaR at 1‑year horizon).

D) Transparent Arbitration Algorithms

  • Outlier flagging: Z‑scores or robust MAD on node price residuals; human review threshold configurable.
  • Similarity search: compare to peer nodes (same route/strength/site) to propose fair price within corridor.
  • Explainability: publish feature importances, residuals, and counterfactuals per case so parties can understand and challenge.

Next — Spoke 4: Cents & Sensibility — Economic Reality & the Human Layer

Missed the setup? Read Spoke 2: Collective Solvency, or return to the HCC Hub.

Join the Founding Circle