The administrator is not the authority

A dental manual can be operationally correct and still be the wrong authority.

The Michigan evidence makes the distinction useful. DentaQuest publishes an office reference manual for Molina’s Michigan dental lane. It tells a provider how to work inside that lane: which programs are in scope, how benefit tables are arranged, where authorization duties appear, how electronic attachments move, what emergency documentation has to prove, and how noncovered treatment and hold-harmless language affect billing. The guidance is operational, and a practice cannot ignore it.

The same source does not become the state Medicaid authority just because it is the document a biller opens most often. Michigan MDHHS owns the Medicaid policy surface. Its Medicaid Provider Manual, bulletins, provider enrollment rules, CHAMPS claims-processing infrastructure, eligibility rules, prior authorization paths, claims and encounter instructions, and state-policy precedence have to remain visible as their own source layer.

The transferable rule is simple: preserve who administers the work separately from who has authority over the rule.

The administrator runs the lane

The DentaQuest/Molina manual is valuable because it is close to the work.

It names Michigan Adult Medicaid, Healthy Michigan Plan, and Molina Dual Health Link. It shows how the dental benefit appears in the administrator’s office-reference table. It carries the rows a front desk team and billing team have to translate into a daily workflow.

For tomosynthesis, the manual gives the operator a useful shape. D0372 and D0387 sit with D0210 and D0330 in a 60-month per-patient frequency group. D0373 and D0388 sit with D0270 through D0274 in a 12-month per-patient group. D0374 and D0389 appear as their own periapical image rows.

The source also carries non-code duties. It points providers toward prior authorization for out-of-network treatment. It names emergency authorization documentation. It routes electronic attachments through NEA or similar electronic submission handling. It discusses mounted radiographs. It carries hold-harmless and noncovered-treatment language that affects whether the provider can shift cost to the patient.

These details are the difference between a code answer and a billable workflow.

A weak system stores the manual as if it were just another payer policy PDF. A useful system records the administrator role: DentaQuest as vendor administrator, Molina Michigan as the plan sponsor or payer lane, Adult Medicaid, Healthy Michigan, or Dual Health Link as the program lane, and the operating duty attached to the row. The record should also keep the frequency counter named by the manual, because the grouped code family and measurement window tell a trainee how the work moves. It does not yet answer who can overrule the work.

The state owns the authority surface

The MDHHS Medicaid Provider Manual has a different job.

It is the official state Medicaid policy surface behind Michigan dental coverage and administration. It is not a DHP office-reference guide and should not be collapsed into one. It establishes the state program posture: MDHHS policy, Medicaid manuals, bulletins, provider enrollment, eligibility, prior authorization, claims and encounter handling, and CHAMPS as the Medicaid claims-processing system.

A vendor manual can summarize a state rule. It can implement state requirements in a plan lane. It can tell the provider which portal, attachment route, or authorization path to use. It can expose the operational version of a frequency rule.

The authority record still needs to point upstream.

If MDHHS changes a policy bulletin, the vendor grid is not the safest source of truth. If CHAMPS is the verification route for code and rate lookup, the vendor manual should not be the only lookup node. If state policy has precedence, the system should show that precedence rather than leaving it as something the experienced biller remembers.

Good operators already make this distinction in their heads. They know the manual they use every day. They also know when the answer has to be checked against the state, the plan lane, the member, the claim route, or the proof package.

Hadto’s job is to make that judgment inspectable.

Flattening creates hidden dependence

Most automation fails here because it wants a single answer.

It wants to turn D0372 into covered or not covered. It wants to turn D0373 into annual or not annual. It wants one payer field, one policy field, one authorization flag, one attachment rule.

The Michigan evidence refuses that shape.

D0372 is not only a code. In the DentaQuest/Molina lane it participates in a 60-month group with D0210 and D0330. That group is administered through a vendor manual for named Michigan programs. The provider still has to know whether the member is in Adult Medicaid, Healthy Michigan, or Dual Health Link. The provider still has to verify eligibility and billing route. The policy authority still traces back to MDHHS.

Flattening those facts makes the business dependent on private memory at the worst point in the workflow. A senior biller knows when a vendor table is enough and when the state source has to be checked. A trainee sees only a row. An AI sees only a retrieved snippet. The owner sees a report that looks complete because it has a source URL and a confidence score.

The missing field is authority.

Once authority disappears, every downstream step looks more certain than it is. A routing instruction can be mistaken for coverage. A benefit table can be mistaken for state policy. A prior-authorization workflow can be mistaken for permission to treat. A hold-harmless row can be separated from the condition that made it safe to use.

These mistakes do not always show up as denied claims on day one. They show up as training debt, appeals debt, patient-balance risk, and a business that still needs the same expert to bless the final decision.

Store the job each source is allowed to do

Owner-ready automation needs a source-role model before it needs a prettier summary. For this Michigan lane, the record should keep at least five things apart.

Authority source means the Michigan MDHHS Medicaid policy surface, including manuals, bulletins, CHAMPS-related verification, and state-policy precedence. Administrator or vendor means DentaQuest administering a Molina Michigan dental lane. Plan and program lane means Adult Medicaid, Healthy Michigan Plan, or Molina Dual Health Link, because the same administrator can operate different benefit contexts.

Verification route means where the operator must check eligibility, authorization, code/rate facts, claim status, or attachments before acting. Billing and proof duty means the documentation, radiograph, emergency, out-of-network, noncovered-treatment, and hold-harmless facts that decide whether the workflow is defensible. Those fields should travel with the code row. The operator should not have to infer them from a PDF title.

The same structure helps outside dental. A payroll provider may administer tax filings without owning tax authority. A benefits broker may administer enrollment without owning plan terms. A field-service platform may route warranty work without owning the warranty obligation. A lender portal may collect documents without owning underwriting policy.

Administration is part of the workflow. Authority is part of the rule.

When a system preserves both, the business can train people on the real decision path. When it merges them, the business trains people on shortcuts and hopes the expert is nearby.

Authority is what makes ownership teachable

Hadto’s thesis is not that every small business needs a larger database. The thesis is that domain experts become owners when the business can preserve their judgment in infrastructure another person can inspect.

The source split is exactly where that either happens or fails.

The expert does not merely know that a manual exists. The expert knows what the manual is allowed to decide. They know which parts are vendor administration, which parts are plan-lane implementation, which parts require state verification, and which parts change the billing proof package.

An owner-ready system should make that order visible:

1. Identify the member and program lane.
2. Identify the administrator and the manual that controls the operating route.
3. Identify the authority source that controls state policy.
4. Verify the route, authorization, attachments, and proof duties before billing.
5. Preserve the source trace so the next operator can review the decision.

Private memory becomes business memory when the order is recorded.

The DentaQuest/Molina manual is not less useful because MDHHS is the authority. It is more useful when the system records its proper job. It administers the lane. It gives the provider workflow detail. It teaches the practice how the benefit is handled.

MDHHS remains the authority surface. The state policy layer decides what can override, where official verification lives, and why the vendor manual should be modeled as an administrator source instead of the top of the rule tree.

The administrator is not the authority. A business that records that sentence as structure can let more people make correct decisions without waiting for the one person who remembers why it is true.

---

Source evidence used in this note: public source DentaQuest Molina Michigan Office Reference Manual, reviewed through Hadto’s 2026-05-12 ontology research cycle; public source Michigan MDHHS Medicaid Provider Manual, reviewed through Hadto’s 2026-05-12 ontology research cycle. Internal research notes, not public evidence: 2026-05-12 ontology research-cycle notes on file, sections 06:41Z and 09:18Z, including preserved manifest references research/manifests/dental/2026-05-12-dentaquest-molina-michigan-office-reference-manual.yaml and research/manifests/dental/2026-05-12-michigan-mdhhs-medicaid-provider-manual-dental-authority.yaml.