We model the operation, not the diagnosis.
AnantState is used on the running of a healthcare organization: patient flow, capacity, scheduling, claims and billing. It does not diagnose, recommend treatment or make clinical decisions, and keeping that line clear is what makes it safe to deploy.
In plain termsOperations only. Never clinical care.
Where capacity pressure travels
An illustration of one ward’s pressure and the places it comes from.
The entities are operational: a patient pathway, a unit or ward, an operating room, a roster, a claim, a referral, an appointment slot. Aggregate occupancy hides the unit whose state is degrading fastest, which is why the picture is a set of connections rather than a single number.
Where capacity pressure travels · illustrative
- Emergency inflowlinked to Imaging, Ward 4
- Imaginglinked to Emergency inflow
- Operating roomslinked to Ward 4
- Ward 4linked to Emergency inflow, Operating rooms, Night roster, Discharge
- Dischargelinked to Ward 4
- Night rosterlinked to Ward 4
What moves in a unit
| State | Baseline drivers | Unexpected drivers |
|---|---|---|
| Flow and throughput | Case mix, schedule, day of week, seasonality | Cancellations, emergency inflow, staff absence |
| Capacity pressure | Booked load, discharge pattern | Length-of-stay outliers, isolation requirements |
| Roster adequacy | Rostered hours, skill mix | Sickness, agency availability |
| Claim posture | Payer rules, coding pattern | Denial trends, payer policy change |
| Referral backlog | Arrival rate, triage capacity | Referral source shift, seasonal surge |
Healthcare operations are dominated by schedule, so the baseline carries most of the state. That is a good fit for a baseline plus a small correction: useful forward state with very little training history.
Where it goes wrong today
- Throughput problems become visible when the operating-room schedule slips, not when the schedule started to become impossible.
- Revenue-cycle teams work denials one at a time rather than preventing the pattern that produces them.
- Roster gaps are often found on the shift itself, when cover can no longer be arranged cheaply.
A worked example
Two thousand claims at risk of denial, and a fix applied before the payer decides.
Round, illustrative figures for claims at risk: they show the arithmetic, leave out the subscription, and are for you to change in the calculator.
| Input | Value | What it stands for |
|---|---|---|
| Claims at risk in scope | 2,000 | Everything the team is responsible for |
| Value at stake per one | $900 | Revenue lost if one claim is denied and not recovered |
| Cost of the action per one | $60 | Correcting and resubmitting a claim before it is denied |
| Share acted on | 30% | The claims with the highest denial risk: 600 of 2,000 |
| Modeled, USD | Pessimistic | Expected | Optimistic |
|---|---|---|---|
| Share of value at stake avoided | 10% | 25% | 40% |
| Cost to act, in total | $36,000 | $36,000 | $36,000 |
| Avoided loss | $54,000 | $135,000 | $216,000 |
| Net position | $18,000 | $99,000 | $180,000 |
Break-even is 6.7%: each action costs $60 against $900 at stake, so it has to avoid that share of the loss to pay for itself.
What exists today
Healthcare operations exists as a catalog template (assess to treat, detect to respond and budget to report are all in the catalog) and as one of the six domains compiled into the engine. It is not a live simulated world.
Find out which unit is about to break
Bring a month of scheduling and throughput events. We will show where capacity pressure was building before the list slipped.