Admission decision
Should we accept this resident?
An accept-or-decline decision should be assembled from nine governed inputs — clinical fit, staffing capacity, incremental labor cost, payer economics, comparable historical outcomes, current census, building capability, regulatory constraints, and expected contribution margin — each traceable to a named authoritative source and time-stamped at the moment of decision. Regulatory and capability constraints act as hard gates. The remaining inputs produce a recommendation with its exceptions named; a human accepts or declines, and the record retains who decided, on what, and when.
- Inputs
- Nine governed inputs, each with a named source
- Hard gates
- Regulatory and building capability
- Decision authority
- Human. Clinical authority stays clinical.
- Maturity
- DESIGN INTENT — not deployed, not exercised on operator data
Evidence posture. The resident economic and clinical journey record is a DESIGN INTENT architecture: the canonical object, its authority rules and its lineage requirements are decided, and no part of it is deployed, integrated with a third-party system, or exercised on operator data. Every figure on these pages is a worked example on synthetic inputs, not a benchmark, forecast, or observed result. Last reviewed August 30, 2026.
The governed inputs
| Input | Authoritative source | Governance rule |
|---|---|---|
| Clinical fit | Clinical system of record (SeniorCRE clinical record, or an incumbent read one direction) | Clinical eligibility is a clinical determination. The record carries it; it does not compute it around the clinician. |
| Staffing capacity | Scheduling and worked-hours source | Capacity is stated by shift and discipline, not as a building-level average. |
| Labor cost | Payroll source | Expected incremental labor is separated from fixed coverage already scheduled. |
| Payer economics | Rate structure and authorization record | Estimated reimbursement is an estimate with a stated method, never a promised amount. |
| Historical outcomes | Closed journey records for comparable stays | Comparables are disclosed with their count. A thin comparable set lowers confidence rather than being smoothed. |
| Current census | Census and bed-board source | Census is read at decision time and time-stamped into the decision record. |
| Building capability | Service-capability register for the specific building | Capability is per building and per license, not per portfolio. |
| Regulatory risk | Compliance and licensure record | Regulatory constraints act as hard gates, not as weighted factors. |
| Predicted margin | Reconciled reimbursement and labor definitions | Predicted contribution margin is decision support. A human accepts or declines. |
The shape of a governed disposition
Worked example — synthetic inputs, illustrative only
- Estimated reimbursement
- $18,700
- Expected length of stay
- 23 days
- Expected incremental labor
- $6,800
- Therapy
- $2,100
- Historical denial rate in comparable set
- 8%
- Readmission rate in comparable set
- 14%
- Facility staffing capacity
- Constrained
- Expected contribution margin
- $5,900
- Clinical-fit confidence
- 92%
- Operator acceptance policy
- Satisfied
ACCEPT — with staffing exception flagged
A disposition is a recommendation with its exception named. The operator, not the model, accepts the resident — and the record retains who decided, on what inputs, at what time.
Exception flags
An exception that is not named is an exception that becomes a variance nobody can explain a quarter later. Each flag below stays attached to the decision record.
Staffing constrained
The shift and discipline required by the referral is already at or beyond planned coverage. Acceptance implies incremental agency or overtime cost that must be named in the decision record.
Capability gap
The specific building does not hold the service capability or license the clinical characteristics require. This is a hard gate, not a weighted factor.
Authorization risk
Authorization is absent, partial, or historically contested for this payer and clinical profile. Estimated reimbursement is flagged as unstable.
Thin comparable set
Too few closed journey records resemble this referral to support the historical inputs. Confidence is lowered and disclosed rather than smoothed.
Policy departure
The disposition conflicts with the operator’s own written acceptance policy. The departure, its author, and its rationale are retained.
Margin below threshold
Expected contribution margin under the operator’s definition falls below the threshold the operator set. Accepting anyway is a governed exception, not a silent one.
Closing the loop
The decision is only half of the record. After discharge, actual reimbursement, actual labor, actual length of stay, denials and readmission are reconciled against the pre-admission expectation in the same definitions — per field, with lineage. Governance first. Intelligence second. Execution last.
Frequently asked questions
How should an operator decide whether to accept a referral?
On a governed set of inputs, each with a named authoritative source: clinical fit, staffing capacity by shift and discipline, incremental labor cost, payer economics, comparable historical outcomes, current census, building capability, regulatory constraints, and expected contribution margin. Regulatory and capability constraints act as hard gates; the remaining inputs produce a recommendation that a human accepts or declines.
Does SeniorCRE make the admission decision automatically?
No. The record assembles the inputs, names the exceptions, and retains the decision. A human accepts or declines. Clinical eligibility remains a clinical determination made by the clinical authority.
What is an exception flag?
An exception flag is a named, retained departure from operator policy, capacity, capability, or confidence thresholds. Its purpose is to keep the departure visible in the record instead of absorbed into an aggregate.
How is the decision evaluated afterwards?
After discharge, the same record reconciles actual reimbursement, actual labor, actual length of stay, denial history and readmission against the pre-admission expectation. The variance is retained per field, which is what makes the next decision better informed.
Are the figures shown here real?
No. They are a worked example on synthetic inputs, published to show the shape of a governed disposition. They are not operator data, a benchmark, a forecast, or an observed outcome, and this capability is DESIGN INTENT.
Continue
- Canonical record: The operating record · One Operator-Controlled Operating Record
- Authority: Operator Authority Chain™ · Data integration
Coherence before Intelligence. When systems disagree, the operator governs.
Bring one contested admission decision
The useful version of this conversation is a referral you accepted or declined last quarter, and whether your systems can still reconstruct why.