Payer Policy and Provider Screening Architecture

Coverage in dentistry isn't one rulebook — it's several, stacked. The ADA defines the procedures; each payer layers its own coverage policy on top; and a practice or group may have its own overlays besides. Accord Dental models this explicitly, as a three-layer hierarchy with no coverage values hardcoded into logic, and pairs it with provider-integrity screening built into the pipeline. This document describes both.

The three-layer rule hierarchy

Accord resolves every coverage question by reading down a deliberate hierarchy, from the most general standard to the most specific policy:

ADA baseline. The CDT code set and the ADA's clinical guidelines define what a procedure is and the professional standard around it. This is the shared foundation every determination starts from.
Payer rules. On top of the baseline sits the payer's coverage policy: coverage percentages per CDT code, bundling rules, the downgrade matrix, frequency limits, coordination-of-benefits rules, and fee schedules. This is where "the procedure is valid" becomes "here's whether and how we pay for it."
Tenant overlay. Finally, a practice or DSO's own policies and expectations layer on top of the payer rules — the tenant-specific overrides and configurations that reflect how a particular organization operates.

A determination is the result of reading down this stack: start from the correct code, apply the payer's policy, then the tenant overlay. When a result surprises someone, the explanation is always locatable in a specific layer.

Zero hardcoded coverage values

A defining architectural choice: coverage values are data, not code. Coverage percentages, bundling relationships, downgrade mappings, and frequency limits live in the reference layer as rules the engine reads — they are not embedded in the application's logic. This has two important consequences.

First, every signal is traceable. Because a determination resolves to specific rule rows, Accord can point to the exact ADA guideline, payer rule, or tenant overlay behind any flag. Nothing is a black-box constant.

Second, updating policy is a data change, not a code change. When a payer revises a rule, the update happens in the rule data — no engineering release required to reflect it. That keeps the engine both auditable and maintainable as payer policies evolve, which they constantly do.

Provider screening in the pipeline

Alongside coverage, Accord runs provider-integrity checks as part of processing rather than as a separate manual step:

NPPES — NPI verification. The rendering provider's National Provider Identifier is verified against the national registry, so the provider on a submission is real and correctly identified. (Accord's own reference data is validated this way — for example, verifying a provider's NPI against NPPES rather than trusting a supplied number.)
OIG — exclusion screening. Providers are screened against the OIG exclusion list, so a practice is alerted if a provider appears on a federal exclusion — a check healthcare organizations are expected to perform.

Running these at the point data enters the system means provider-integrity problems surface early, as part of the record, rather than being discovered later.

Why "compliant by design" means architecture, not a claim

The phrase describes how the system is built, not a certification it holds. Accord's architecture is designed so that determinations rest on sourced rules rather than hidden constants, so that policy is transparent and updatable, and so that provider screening is automatic. That design supports a practice's own compliance and revenue-integrity work — it doesn't replace it. The organization using Accord remains responsible for its own compliance program; what Accord provides is a rules engine whose structure makes coverage decisions transparent, traceable, and maintainable.

The through-line

From ingestion to determination, the same principle holds: a coverage decision should be explainable down to the rule that produced it. The three-layer hierarchy makes the rules explicit and ordered; keeping values as data keeps them traceable and current; and provider screening adds integrity checks the domain expects. Together they're what let Accord stand behind the promise that every pre-determination is explained, cited, and defensible.