Hoyani Review Room

Patients across Indian Country travel significant distances for care that does not always require a hospital or clinic visit. Tribal and Indian health programs carry persistent workforce shortages, transportation barriers, geographic isolation, continuity challenges and growing demands on existing clinical teams. Vacancies take years to close, and patients need care in the interval.

There is a second opportunity alongside the first. Many Tribal citizens are eligible for services they never end up using. A ride to an appointment, a prescription that reaches the house, an aide for an elder living alone, a veterans benefit nobody has walked them through. Eligibility is established. What is missing is the coordination between programs that would turn it into a service at the door.

Hoyani is built around one question.

The answer is yes, and only if clinical governance, payment authority, workforce, information exchange, patient safety, Tribal data governance, program integrity, vendor independence and working capital are designed together rather than added later. That combined model is set out here.

Why the question is answerable now

In the CY2025 hospital outpatient final rule, published 27 November 2024 and effective 1 January 2025, CMS created a mandatory exception to the Medicaid clinic services four walls requirement for Indian Health Service and Tribally operated clinics, codified at 42 CFR 440.90(c). Clinic services from those providers no longer have to be furnished inside the building. Separately, services received through an Indian Health Service or Tribal facility can draw a 100 percent federal medical assistance percentage under a care coordination agreement, described in State Health Official Letter 16-002.

Those facts make a home-delivered encounter coherent in a way it was not before. They do not make it automatic. A mandatory federal exception still has to appear in a state’s own Medicaid plan before a state agency pays against it, and the 100 percent federal share is a federal match on qualifying state expenditure rather than a pool of startup money or a promise of a preferred rate. Section 09 sets out how a pilot site is chosen because of the first point. Section 10 sets out the second.

Hoyani is a care and benefit coordination system, plus a credentialed field workforce, supporting a participating healthcare organization. The scope of what Hoyani itself claims to do is deliberately narrow, because the broader the claim the harder it is to defend and the less likely it is to be true.

Hoyani runs as a mobile application and as a full desktop application in a browser. Both carry the same data and the same rules. The difference is the job in front of the person using it. There is also a fourth route that is not software at all, because a system that only works for people with a working smartphone and a signal does not work in Indian Country.

How a patient enrolls

Enrollment is open to any enrolled member of a federally recognized Tribe, and to anyone else eligible for services from the participating program under its own eligibility rules. A patient, or an authorized representative, creates an account in the app or on the desktop, or by telephone with program staff, and adds the documents that establish identity and eligibility once, so that no program has to ask for them again. Each document is captured by camera or upload, encrypted on the device and in transit, stored in a document vault encrypted at rest with keys held separately from the data, and viewed only by program staff with a verified role and a logged reason. Nothing is stored on the device after upload. Program staff, not software, verify the documents and confirm the patient’s relationship to the program. A patient can see every person who has opened a document and when.

Location decides the program, because a mailing address cannot

A large share of homes in Indian Country have no street address and receive mail at a post office box, sometimes fifty miles from the house. Hoyani therefore assigns a patient to a participating Tribal 638 or Indian Health Service program by the location of the home itself. The patient or a family member drops a pin on a map, or the first clinician confirms the coordinates at the door, and the pin can be described in the local way, the chapter, the road, the landmark, the turn. That location places the patient inside a program’s service area, sets drive times and safety planning for the clinician, and is what emergency responders receive if a clinician ever has to call them. A patient who lives inside more than one program’s area is asked which program they use, and the answer is confirmed by that program.

How a clinician enrolls

A clinician creates an account, uploads licensure, certifications, government identification and the documents the credentialing organization requires, and completes the background investigation under the Indian Child Protection and Family Violence Prevention Act. Verification is performed by the independent credentialing organization or the program’s own credentialing office, never by the staffing company that recruited the clinician, and the result is a privilege for specific work at a specific program. Clinicians are assigned to participating programs by the location of their home base and the service areas they can safely reach, and they see assignments only for programs that have privileged them. Their documents sit in the same vault under the same controls, and a lapsed license blocks assignment automatically.

Hoyani is not an emergency line

Every screen that accepts a request states it. Hoyani is not 911, does not dispatch emergency services and is never the route for an emergency. A request that describes an emergency is met with the instruction to call 911 or the local emergency number, the program’s on-call number, and the location the patient has already confirmed so it can be read aloud to a dispatcher. A clinician who finds an emergency in a home calls emergency services directly under the section 14 protocol.

Accessibility and language

The patient application is built and tested to WCAG 2.2 Level AA and Section 508, with large-type and screen-reader modes, and it is tested with elders on the devices they actually own rather than on a developer’s phone. Written content is held to a plain-language standard. Spoken-language support is confirmed before an appointment rather than after, with interpreter arrangements for Diné bizaad and the other languages the participating community names, and a patient may choose another arrangement. Telephone access through program staff is a first-class route, and pilot results are reported for assisted and telephone users alongside app users. Accessibility testing is an item in section 23 with its own owner and date.

A Tribal citizen’s benefits do not arrive from one place. Care comes from the Indian Health Service or a Tribal 638 program. A veteran carries a separate set of rights through the Department of Veterans Affairs. A service member’s family may hold TRICARE. Medicaid covers rides and, for some, home and community based support. Tribal programs run their own aging, transport and community health services. Each has its own authority, its own payer and its own front door, and a citizen is expected to find all of them alone.

Hoyani’s job is to make one request open the right door. Every line below names the program that administers the benefit. Hoyani creates no benefit, becomes no payer, and every claim is the participating program’s, submitted under its own enrollment. The status column states exactly what a first pilot enables and what is mapped authority with a named responsible program.

Who Hoyani is built for, and who a first pilot serves

Hoyani is built for every enrolled member of a federally recognized Tribe, across the whole of a life. Veterans, elders, parents, teens, children, families and people in recovery. The platform is built to serve all of them, and a first pilot begins with adults established with the program while the others are staged and prepared for the partner’s approval, because each carries a precondition that the partner’s clinical and legal leadership signs off on its own terms. The preconditions are named here so that each group is switched on deliberately and none is added quietly.

Software does not take a blood pressure. A home visit program is a physical operation, and the equipment is where clinical scope, infection control duty, the connectivity problem and asset accountability all become real. Three kits, each owned deliberately.

Ownership follows accountability. Kits are program property, or leased to the program under the services agreement, so that a program ending its participation keeps the capability it has come to rely on rather than losing it with the vendor. That is the same principle as the transition rights in section 17, applied to equipment.

Every serious reviewer eventually asks one compound question. For this patient and this visit, who is the provider of record, who authorized the care, who may legally perform it, does it generate a billable encounter, who pays, who carries clinical risk, who is responsible after the clinician leaves, where does the record go, and what agreement gives Hoyani authority to do any of it?

Answering that correctly and repeatably, across the specific rules of Indian Country, is the product. Here is the answer.

One patient, four requests

A composite, not a real person. A 74 year old Diné Army veteran who uses a wheelchair lives forty miles from the Tribal health program where he is an established patient. He is enrolled in his state’s Medicaid program under the American Indian fee for service option and is enrolled in VA health care. He was discharged from a regional hospital three days ago. His daughter, an authorized representative on his account, makes four requests in one evening. Here is what the model does with each, and where it stops. Every step below can be reproduced in the section 08 engine.

Care delivery, coordination technology and payment do not sit in one box. A structure in which a single entity supplies the clinicians, verifies their credentials, coordinates the visit and collects the payment is a closed loop with no independent check, and it puts a vendor in conflict with the Tribal health systems it is meant to extend. Hoyani separates the three, and section 16 sets out how that separation is enforced rather than asserted.

Hoyani is connective tissue. It is not a replacement health system. That distinction is the center of the project, and every other design decision here follows from it.

Why each of those rails benefits

Every program on the third layer gains from a request that arrives correctly routed. The Tribal 638 or Indian Health Service program captures an encounter it could not previously reach, with the eligibility checked, the authorization documented and the claim prepared under its own enrollment. The state Medicaid agency receives a claim that carries the right pathway and the 100 percent federal share documentation, with nothing routed to a fund that should not pay it. Veterans Affairs sees a Veteran’s benefit used under the right agreement with the verification done first. A Tribal program sees its own transport, aging and community health services reached by the citizens they exist for. Purchased and Referred Care is protected rather than drawn down. The purpose of the platform is to relieve staffing shortage stress, expand what a medical center can reach without adding beds or buildings, reduce paperwork for staff and patients, increase the benefits citizens actually use, and lower the cost of each of those to the programs that carry it.

Ownership of the platform

Ellsworth develops, owns and manages the Hoyani platform, including the coverage rules logic, the routing and dispatch design, the credential and visit verification workflows, the interface implementations and the operational tooling. Ellsworth is responsible for maintaining it, securing it, and holding the agreements that govern its use. That ownership is stated plainly because a participating health program needs to know who is accountable for the software, who maintains it, and who answers when something needs support.

The boundary matters as much as the ownership. Tribal health data is never Hoyani intellectual property. Clinical content, patient records, encounter documentation and any information about Tribal citizens belong to the health program and the Tribe, governed under section 17. Nothing a Tribe contributes through participation becomes a Hoyani asset, and platform ownership creates no claim over the data flowing through it.

This is a working implementation of the gate sequence Hoyani runs before any request is acted on. Change the inputs and read the trace. Coverage is checked as all that apply, because a Tribal citizen may hold Medicaid, Medicare, VA enrollment, TRICARE and a commercial plan in any combination, or none of them, and the engine resolves the order of payment across every coverage held. Every gate has a named legal or clinical basis, and an unsatisfied gate produces a routing decision with a reason attached rather than a dead end.

Eligibility, authorization, network participation, medical necessity, coding, documentation, coordination of benefits and adjudication are separate functions, and an eligibility response is never a payment guarantee. The engine does not represent a service as covered. It represents that every precondition has been evaluated, and it names the ones that are not satisfied. Selecting a ride, a refill or benefit help shows the benefit route rather than a clinical encounter, because those are different authorities with different payers.

A federal exception permitting a clinic service outside the building is necessary and not sufficient. A state Medicaid agency pays against its own approved plan, and the conforming authority has to be on file in that state before a home-delivered clinic encounter is payable there.

CMS approved the first conforming state plan amendments on 13 June 2025 for Minnesota, New Mexico, Oregon, South Dakota, Washington and Wyoming, covering 89 Tribes and Alaska Native villages. Arizona followed. AHCCCS submitted SPA 25-0001, an exception to the clinic services four walls requirement for IHS and Tribal facilities and behavioral health clinics, on 31 March 2025, and CMS approved it on 24 September 2025. Oklahoma’s SPA 25-0010 was approved on 3 October 2025. The published record, rather than sentiment, decides where a responsible first pilot runs, and on that record both Arizona and New Mexico qualify.

What this rule means in practice

Ellsworth’s existing clinical staffing footprint covers Arizona and New Mexico, and both states carry the authority. Site one is therefore chosen by the willing partner rather than by the state, with the state’s own confirmation of encounter qualification obtained in writing before the first visit in either. Selecting a site by the state’s Medicaid authority rather than by proximity is the clearest available signal that this project is built to survive an audit rather than to launch quickly.

Where a state’s position is unconfirmed, as it is in Utah on the published record, the posture is to ask the state agency directly and in writing, share the answer with the Tribal partner, and wait. A reviewer who wants to test that can set the pilot state control in section 08 to Utah and read what the engine does with it.

A state plan amendment answers one question. It does not answer which practitioners and which services the state treats as a qualifying encounter when delivered in a home, how the state’s electronic visit verification requirements apply to a service, or how the state’s Tribal consultation policy expects a new arrangement to be raised. Those three questions are put to the state agency together, in one written request, so that the answer arrives as one document the Tribal partner can rely on.

The financing test is a transaction-level explanation. The patient, the service, the rendering professional, the billing entity, the location, the coverage, the required authorization, the rate, and the party responsible if payment does not resolve. Listing several programs is not a payment model.

How the payers actually relate to each other

These pathways are not a ranking by size. They are different mechanisms, and treating them as one list is the mistake that costs credibility with any payer official in the room. State Medicaid is the fee for service spine and the only clinical pathway in a first pilot. The Indian Health Service reaches an outside party through a contract or through Purchased and Referred Care with an authorization, so it is a funding and authorization relationship rather than a claims payer for this model. Veterans health reimburses a participating facility for direct care it furnishes itself to eligible American Indian and Alaska Native Veterans, acts as payer of last resort, treats a Medicaid payment as final, and places home based services in a category requiring benefit verification before a claim. Commercial coverage requires a network agreement that does not exist.

How Purchased and Referred Care is treated

PRC is an authorization gate, not a selectable payer. It requires eligibility, residence in the delivery area, medical priority, notification within the required window and exhaustion of alternate resources, and it is generally payer of last resort. Without an authorization number on file, the engine stops and routes the request to the program’s PRC office.

There is a second point here that matters more to a Tribal health director than to a payer. PRC funds at many service units are committed before the year ends, and a platform generating new demand against a fixed fund is a liability rather than a benefit. PRC sits last in the ordering rather than first, the program sees its PRC exposure before a request is scheduled rather than after, and PRC avoidance is a reported pilot measure.

The all inclusive rate is an encounter rate paid to a facility for a qualifying visit. Each state plan limits how many encounters a day are payable, from one per date of service with exceptions in South Dakota and Utah to as many as five in New Mexico, and each generally requires a qualifying practitioner. A registered nurse arriving at a home, however clinically valuable that visit is, does not automatically produce a billable encounter. A model quietly assuming it does will not survive a first conversation with a state Medicaid analyst or a program’s revenue cycle director.

So the model states it the other way around. Some visits generate an encounter. Some do not, and are funded deliberately. Both are shown here, every one of them is routed and tracked through Hoyani, and the engine classifies every visit before it is dispatched.

There are two sets of books in this model and a reviewer should see both. The health program’s ledger receives the encounter payment. From it the program pays the clinician’s time and travel at cost, passed through and itemized, plus a modest coordination fee per completed visit. Hoyani’s ledger receives the pass-through and the fee, pays the clinician and the travel out of the pass-through, and keeps the fee less its own direct coordination cost. Hoyani makes a little on every visit and the program keeps most of every encounter, which is the intended shape of the model. The figures below are a sensitivity example rather than a forecast, a wage benchmark or a reimbursement estimate.

Assumptions, stated so they can be argued with. A staffed clinician day costs $480 in clinician pay and $90 in travel, $570 passed through to the program at cost. A billable encounter collects $826 at the CY2026 rate, discounted to $700 after denials and coordination of benefits. A supporting visit collects nothing. Hoyani’s coordination fee is $75 per completed visit, billable or supporting, about nine percent of the published rate, and Hoyani’s own direct coordination cost is $60 per staffed day. Central overhead, software development, insurance, equipment and taxes are excluded from both ledgers and must be added to any real model.

Three things follow from reading both ledgers. The program keeps more than Hoyani at every row, and the gap widens with every additional billable encounter, because the fee is flat and the encounter is not. The program’s contribution depends on the billable share, which is set by the state’s encounter rules and the service mix, so the pilot reports it rather than assuming it. And Hoyani’s contribution depends on visit density, not on the rate, which is why the day below is tracked the way it is.

Mileage, documentation time, cancellations and route density determine viability more than the rate does. Clinical duration and safe travel time set real capacity, and the participating program’s clinical leadership sets both. So every minute and every mile of a staffed day is tracked, paid and reported, and the pass-through the program sees is built from that record rather than from an estimate.

A nurse practitioner’s day on a rural route, as the clinician app records it and as the program’s pass-through statement shows it. Times and miles are illustrative. The tracking is not.

The same record feeds the pilot measures. Drive minutes per completed visit and miles per completed visit show route density. Cancellations show up with their reasons, so a program can see that dialysis days should be blocked before scheduling rather than after. Documentation minutes show what the program’s record system costs a clinician in the field. And because the clinician is paid for every worked minute on a fixed cycle regardless of the program’s payment timing, the working capital above is a real number with an explicit funding source and a loss allocation policy, both settled in the services agreement before the first visit.

What a pilot proves with data rather than projection

Who bills, who receives reimbursement, what Hoyani charges, what the clinician costs, what administrative expense is created, what travel or facility utilization is avoided, what happens when a claim is denied, what share of completed encounters produce the expected financial result, what share of visits are supporting visits rather than billable encounters, and what value remains from a visit that is not directly reimbursed. Those figures come from operations and are reported to the participating program and Tribe before any expansion is proposed.

A promise a patient can rely on has to name the mechanism, the funding, the exclusions and the complaint route. Anything short of that is marketing language a program’s compliance officer has to defend later. Both patient-facing commitments below are written the way they appear in the services agreement.

Eligibility verification and claim adjudication are separate events. A claim can be denied after a completed visit for medical necessity, coding, benefit limits, coordination of benefits, a retroactive eligibility change, prior authorization or a payer edit. Section 10 states who receives that denial, who prepares the correction, and who bears the loss if it remains unpaid.

The participating organization’s authorized system is the clinical system of record. Hoyani transmits or receives only the information necessary to support authorized workflows, and never becomes a second chart. The architecture uses nationally recognized standards including HL7 FHIR, with the implementation profile depending on the participating organization and its exchange environment.

Working with the record systems that actually exist

Indian Country runs a mix of record environments. Federal facilities are moving from the decades-old Resource and Patient Management System to an Oracle Health enterprise record deployed by a large systems integrator, and the Lawton Service Unit in Oklahoma went live with that record’s registration and scheduling capabilities on 1 August 2026, with clinical capabilities and wider rollout to follow on the program’s own schedule. Many Tribal 638 programs run their own commercial records, including Oracle Health, formerly Cerner, and Epic, some still run the legacy system, and urban programs run a third set. Statewide health information exchanges sit over all of them. Hoyani is being designed for compatibility with the existing and updated record and exchange environments partners operate, including Oracle Health, Epic, the legacy federal system and the statewide exchanges, through the standards those environments already expose, HL7 FHIR R4, C-CDA and HL7 version 2 interfaces, rather than through anything proprietary. Compatibility is a design commitment and a standards choice. It is not a statement that any vendor has agreed to anything, and no such agreement is claimed. A first implementation plans for the systems a partner actually operates today, independent of any modernization timetable.

A first pilot asks the federal technology program for nothing, and that decides who the first partner can be

This is a deliberate concession rather than a limitation. A national deployment in active rollout should not acquire an external dependency during its deployment years, and no agency chief information officer should be asked to approve one. A first pilot therefore runs with a Tribal 638 health organization operating its own record system as the participating program. It requires no federal network connection, no agency interface, no place in a modernization roadmap and no security authorization sponsored by a federal program office. It produces a structured visit summary the participating program imports through its own normal process.

The consequence is stated rather than hidden. A federally operated Indian Health Service service unit runs on federal systems, and any workflow that reads or writes its record, or gives a field clinician access to it, sits inside the agency’s security program and its authorization process. A service unit is therefore a staged configuration, not a first-pilot option, and it becomes available only through the agency’s own route, with the sponsor and path recorded in section 23. A standards-based path to integration is preserved for later, and only where a partner asks for it.

What is custom, what is licensed, and who supports it

A reviewer should be able to see which parts of the system are built, which are bought, and who answers when something stops working. The coverage and authority rules engine, the decision trace, the routing and assignment logic and the pilot reporting are custom, because they are the part that does not exist elsewhere. Identity and authentication, consent management, telehealth video, mapping and routing, secure messaging, mobile device management and the interface engine are licensed from established vendors and named in the vendor inventory, with configuration owned by Hoyani and disclosed to the program. Support runs on published hours agreed with the program, a clinician-facing on-call route during operating hours, a defined severity scale with response commitments, a monthly patch cadence with an emergency path, and a change log the program receives before each release. The dependency inventory, the versioned rule set and a synthetic-data demonstration of a failure and its recovery are part of the technical diligence package.

What can be read, what can be written, what a person handles

The distinction is documented per partner environment rather than described in general. Demographics and problem, medication and allergy lists are read where the environment exposes them. Encounter documentation is written back through the environment’s standard interface surface. Orders are never written by Hoyani. Anything the environment does not expose is handled by a named person with a tracked task, and that list is published to the partner rather than hidden inside the integration.

Rules, not guesses

Routing and coverage determination are deterministic and versioned. Every encounter records which rule set version evaluated it, so a determination can be reconstructed a year later. Rules change when a rate is republished, a state plan changes or a program’s protocol changes, and each change is reviewed and approved by a named owner before release. No probabilistic model decides coverage, authority or clinical routing, and none will be introduced to make the system sound more advanced.

Exchange participation

Association with an Indian Health Service or Tribal organization does not by itself create information exchange privileges. Exchange requires the appropriate participating organization, agreements, a permitted purpose, authentication, testing, security controls and technical onboarding. The participating health program is the exchange participant, and Hoyani operates as its authorized technology and business associate layer, seeking no independent network role. A record arriving successfully does not establish authority to use it for every business purpose, so permitted purpose is evaluated per transaction and logged.

A closed referral, end to end

Request received, service performed, findings returned, the receiving team acknowledges responsibility, unresolved items tracked to closure. Duplicate and stale records are identified at match time, and a correction propagates to every downstream recipient that received the original, with the propagation recorded.

Security

Least privilege access, role-based authorization, multifactor authentication, encryption in transit and at rest, comprehensive audit logging, secure development practices, security monitoring, incident response, vendor risk management, business associate agreements where applicable, formal risk assessment under the HHS risk analysis guidance, data minimization and a defined authorization boundary. Substance use disorder treatment records are segregated and handled under 42 CFR Part 2 with their own consent and redisclosure controls, which is one reason recovery services are staged and prepared for the partner’s approval rather than switched on in the first week. A diligence package includes the data inventory, the legal-role assessment per workflow, tested backup and recovery, the remediation record, the incident process and the vendor inventory, with independent testing appropriate to the deployed scope. Information collected outside the clinical encounter, such as location and support logs, is reviewed against the minimum needed for each purpose. Hoyani holds no federal security authorization and describes itself as federally authorized only after obtaining one. Because a first pilot connects to no federal system, none is required to run it, and the sponsor and path for a later stage appear in section 23.

Software does not decide when home-based care is clinically appropriate. A participating program operates under a clinical governance structure established before any patient care begins. That structure names a medical director or clinical authority, defines ordering and referral authority, sets written criteria for which patients are appropriate and which are excluded, specifies scope protocols and standing orders where legally appropriate, sets emergency escalation procedures accounting for real drive times, and defines quality review, incident reporting, professional liability requirements, documentation requirements, the legal medical record, and the route by which documentation returns to the program.

Care exceeding the approved scope is escalated rather than completed. Home care becomes unsafe the moment convenience overrides clinical judgment.

The first service, defined precisely

One clinically approved service set and a defined population. Adult follow up, wound care and catheter change visits for established patients are the proposed starting point, subject to partner selection and clinical approval. That protocol set does not extend automatically to newborns, pregnancy, acute illness or any materially different care need, each of which is staged with its own protocol. Credential checks are translated into privileges for the specific work, with evidence of current licensure, relevant competence, malpractice coverage, exclusion checks where applicable, supervision and continuing review. Every clinician entering a home also clears the character and background investigation the Indian Child Protection and Family Violence Prevention Act requires for positions with regular contact with children, whichever population a first pilot serves, because a household is not a controlled clinic room. A learner’s presence is never treated as replacement clinical capacity, and who signs the record is stated in advance.

How this relates to the nursing framework that already exists

Indian Health Service Public Health Nursing already performs home visits, and Tribal Community Health Representative programs already conduct home visits and connect people to care. Hoyani is designed as added capacity inside those frameworks rather than a parallel service. The partner names the work currently delayed or uncompleted, and the proposed intervention is compared against the simpler alternatives of adding a coordinator, clinician time, or transportation support. If a simpler option achieves the result, that is the honest recommendation.

Staged behind the first pilot

Obstetric and prenatal assessment, behavioral health assessment and crisis response, complex wound management outside the wound protocol, any unscheduled or urgent request, any patient with a documented fall risk requiring two-person assist, and any first encounter with a patient not yet established with the program, which routes to enrollment first. Each is prepared and switched on with a protocol and a named supervising clinician once the partner’s clinical leadership approves it. Final scope belongs to that leadership.

Hours, after hours, and the acceptance test

Operating hours, response commitments and responsibility outside those hours are defined before enrollment, and rural emergency contingencies are designed with the local clinical and emergency-response partners rather than by the vendor alone. A disclaimer does not resolve an unsafe queue or an incomplete handoff. The operational acceptance test follows the patient beyond the visit. Who reviewed the findings, what action occurred, how the patient learned the next step, and how the episode closed. A completed appointment is an operational event. A completed care plan is a clinical result. Both are measured.

Liability and worker classification

Federal employees and Tribal 638 program employees carry Federal Tort Claims Act coverage under the Indian Self-Determination Act. A clinician dispatched by a private platform does not, unless the arrangement is deliberately structured so that coverage attaches. That is a legal question preceding software spending, and it is tracked in section 23.

Worker classification turns on control and economic reality rather than on a label. Even where classification is correct, payer enrollment, group affiliation, credentialing, privileging, professional liability and clinical supervision still have to be resolved. Clinicians are privileged under the participating program and, where that program prefers it, placed under its own employment or existing staffing arrangement. Legal classification and licensing depend on the actual services and arrangement, so a service-specific determination is obtained rather than assumed from the word platform.

Any model involving American Indian and Alaska Native Medicaid members, non-facility care, mobile clinicians, digital outreach or new billing arrangements deserves unusually strong program integrity controls. Arizona has recent and painful reason to demand them following member exploitation and provider fraud, and a model that does not acknowledge that openly is not ready to be evaluated anywhere.

The standard is not that an honest operator can use the platform properly. The standard is that the system is difficult to abuse by a dishonest one.

Controls

No direct-to-member solicitation of any kind, and a clear separation from any enrollment steering. Facility-directed eligibility and authorization workflows. Independent clinician credential verification. Verified provider identity. Verified visit location, date and time, feeding the state verification system where a service falls within its scope. Documented ordering or referral authority for every encounter. Consent records retained and producible. Clear separation between clinical delivery, credential verification and payment controls. Immutable audit logs. Duplicate-billing controls and claims reconciliation. Outlier monitoring. Human review of exceptions. Complaint handling with a named owner. Defined fraud, waste and abuse escalation. And the full decision trace retained and available on every encounter.

Visit verification aligns with the state rather than competing with it

Federal law requires electronic visit verification for Medicaid personal care and home health services, and each state runs its own system. Arizona moved from a single named vendor to the AHCCCS Aggregator on 1 October 2025, a vendor neutral model in which alternate systems integrate through a standard interface, and its requirement attaches to attendant care, homemaker, home health aide and personal care codes rather than to a practitioner’s clinic encounter. That distinction matters. A nurse practitioner’s home visit in a first pilot is a clinic encounter, so the state requirement does not reach it, and Hoyani verifies location, time and patient attestation on that visit by policy rather than by mandate. Where a staged service such as homemaker support does fall inside the state’s scope, the delivering agency’s verification feeds the state system through the state’s own route, and Hoyani never substitutes a private record for it. Confirming that route is one of the six site selection tests in section 09.

Credentialing independence

Ellsworth has deep experience recruiting healthcare professionals for Indian Country. Pilot clinicians are nonetheless verified independently, either through an appropriately qualified third-party credentialing organization or through the participating program’s existing credentialing structure.

The principle is simple. The party supplying clinicians is not the party attesting that those clinicians are qualified.

This states a problem before anyone has to ask about it. Ellsworth is a federal medical staffing company holding active clinical staffing contracts with Indian Health Service and Tribal facilities. Ellsworth also built and owns Hoyani, which decides how a visit is routed and to whom. A contracting officer looking at that is obligated to consider organizational conflict of interest under FAR subpart 9.5, in both forms. Impaired objectivity, where the party recommending who staffs a visit benefits from the answer. Unequal access to information, where the coordination layer sees facility demand data that competitors do not.

The concern is legitimate. It is not answered by good intentions. It is answered by structure.

What a government buyer would actually receive

Software development, clinical services, workflow validation, implementation support and evaluation are separate deliverables with separate acceptance criteria, so an agency can tell what it received for each. A cost basis is presented as a work breakdown covering labor assumptions, subcontractor costs, travel, security work, integration, evaluation and contingency, with an explicit control preventing the same expense or service from being charged both to a contract and to insurance reimbursement.

On the acquisition mechanism

FAR subpart 15.6 distinguishes an unsolicited proposal from advertising and from an ordinary commercial offer, and requires an innovative, independently originated approach with sufficient technical and cost information. Preliminary agency contact is permitted, and a favorable evaluation alone does not authorize a noncompetitive award. Supplemental clinical staffing is already acquired competitively, and telehealth services have been solicited as set-asides in more than one Area, so a coordination pilot presented as an unsolicited proposal invites a correct and unhelpful answer. The mechanism comparison belongs with the contracting office, and the alternatives include a competitive acquisition, a small-business pathway, a partner-led purchase, or an eligible funding opportunity. A private business does not assume a grant available to Tribes is available to it directly, and Buy Indian eligibility does not resolve clinical readiness, payment eligibility or proposal novelty.

The ask, stated accurately

The request is a determination of the appropriate evaluation and acquisition route, followed by a bounded proposal. Not a completed unsolicited proposal, and not sponsorship. A meeting is a meeting. Introductions and interest are recorded as exactly that in a factual chronology, and are never described as agency endorsement.

Key person concentration

Ellsworth is owned by one person who is also the primary business development lead for this project. That is a real diligence finding rather than a detail, and the remedy is named succession for the clinical, technical and contracting functions before any expansion.

The Tribal leadership question is a governance question before it is a branding question. What authority, capability and benefit remain with the Nation if it participates? A Native owned business can be a valuable partner, and each Nation still retains its own decisions about participation, care delivery, data and community representation.

The CARE Principles for Indigenous Data Governance emphasize collective benefit, authority to control, responsibility and ethics. They are a useful framework rather than a substitute for the relevant Nation’s own law, policies, decisions and agreements. A contract translates the partner’s chosen principles into enforceable operational terms.

What the Nation approves, and what it can stop

Service locations, outreach, language, clinical scope, community hiring and publication of results are approved by the partner-designated authority. The same authority can change or stop the service, and the community can raise concerns through a route that does not pass through anyone selling the service. Local approval precedes representing a community as a launch partner.

Cultural legitimacy in practice

Community advisors are compensated for defined work. Language, imagery and public claims about cultural meaning receive locally appropriate review. One Nation’s identity or practices are never treated as representative of every Arizona community, or of Indian Country. Interpreter availability is confirmed before an appointment rather than after, and a patient can choose another arrangement. Practical community benefit is tracked and reported to the Nation. Local jobs and retention, patient-reported respect, travel burden, continuity with the usual care team, grievances resolved, and useful information returned to the partner. The Nation and Hoyani agree in advance on who interprets and publishes those findings, and the partner can use the results for its own planning even if Hoyani never expands.

Why not build this yourselves, under your own 638 authority

This is the fairest question a Tribal health director can ask, and the honest answer starts with a concession. A Nation wanting to build and run this itself should, and several have the capacity. What Hoyani offers is a shorter path for the first one, because the policy work is already encoded and the operating model has already been tested against the objections a state agency, a contracting officer, a medical director and a payer will raise. Building that from a standing start costs a program time it would rather spend on patients.

A model ending there is still extraction with better vocabulary. So the agreement carries a transition path, named in advance rather than negotiated from weakness later.

Platform ownership and data governance are separate questions with separate answers. Ellsworth owns the Hoyani software and the rules logic. The Nation governs its citizens’ health information under this section and holds the transition rights above. Neither claim reaches the other, and the agreement says so in terms.

Hoyani does not compete with permanent Indian Health Service or Tribal hiring. Flexible capacity acts as a bridge while permanent positions remain vacant, or where geographic access makes traditional staffing insufficient, and the workforce strategy includes pathways from flexible service into permanent Indian Health Service, Tribal and rural health employment. A model converting clinicians into a program’s permanent staff supports a federal hiring campaign rather than arguing against one.

Where the clinicians come from

Ellsworth maintains a roster of registered nurses, nurse practitioners and other licensed clinicians who have served or applied to serve at Indian Health Service and Tribal facilities, and who want to practice in Indian Country without a facility posting or a long term full time commitment. Ellsworth reports more than 65 clinician placements at Chinle and Shiprock in the past year and a steady supply of clinical staffing applications for Arizona, New Mexico, Utah and South Dakota. That is company reported experience and it is stated as such. A roster is a reserve, not a guarantee of capacity for recurring assignments, so the pilot verifies the workforce before enrollment and reports fill rate, continuity and cancellations as measures. Clinicians choose the engagement form the participating program and applicable law support, including placement under the program’s own employment or existing staffing arrangement, and are paid on a fixed cycle regardless of the program’s payment timing. Cultural orientation specific to the participating community is completed before a first visit, delivered by people the community chooses and compensates.

Tribal employment law

Where a Tribal employment rights ordinance or a Tribal preference in employment law applies, including the Navajo Preference in Employment Act on the Navajo Nation, the program complies with it, registers with the Tribal employment rights office where required, and reports Native and local hiring to the Nation.

Native employment

Being Native owned is not the commitment. The commitment is Native benefit through actual employment, training, leadership, contracting and economic participation. A pilot establishes measurable Native and local workforce objectives in cooperation with the participating Nation, reported publicly to it. Where Tribal employment laws or preferences apply, the program complies with them. Where they do not legally apply, Hoyani voluntarily adopts Native recruitment, development and retention goals of its own. The Title VII Indian preference exception applies under specific statutory conditions and does not create a universal hiring preference for any Native owned company in any setting, and nothing here claims otherwise.

Effect on the partner’s existing capacity

The purpose of the platform is to relieve the pressure a staffing shortage puts on a program’s own nurses and providers, to let a medical center reach patients it has no room or staff to reach inside its walls, to take paperwork off the people who deliver care, to increase the benefits citizens actually use, and to lower the cost of each visit to the program that carries it. Each of those is measured and reported during a pilot rather than asserted. Clinician hours added to the program’s capacity, encounters completed that the program could not previously reach, referrals and rides fulfilled, minutes of documentation returned to the record without rekeying, PRC referrals avoided, travel time and cancellations, and continuity with the usual care team. The program’s own nurses see fewer uncovered home visits, not more, because a Hoyani clinician takes the visit the program could not staff. A result that shows the program better off is reported with the same care as one that shows a line to fix.

Community Health Representatives

Hoyani supplements rather than privatizes the Tribal Community Health Representative workforce. A participating Nation determines the role its CHRs play. Hoyani may support scheduling, communication, escalation, documentation or coordination around that workforce, and CHRs remain employees or agents of the Tribal program unless the Nation deliberately chooses another arrangement. Hoyani does not contract them, pay them or place them under its dispatch.

Hoyani does not yet have outcomes of its own, and this page does not claim any. What exists is a strong published record for each of the things Hoyani coordinates, home based primary care, post discharge follow up, transportation to treatment, community health work, and workforce relief. That record is set out here, service by service, with the figures and the sources, so that a reviewer can see what the model builds on and what a pilot will add to it.

Home based primary care

The largest home visit program in the country is run by the Department of Veterans Affairs. In an analysis of nearly 9,500 new Home Based Primary Care enrollees published in the Journal of the American Geriatrics Society in October 2014, Edes and colleagues found hospitalizations more than 25 percent lower than in the period without the program, Medicare costs 10.8 percent below projection, and combined VA and Medicare costs 11.7 percent below projection, with veterans reporting better access and quality. Medicare’s Independence at Home demonstration, which paid practices to deliver primary care in the homes of chronically ill beneficiaries, reported $16.3 million in savings in its third performance year across 11,382 beneficiaries, $1,431 per beneficiary, with 14 of 15 practices improving on at least one quality measure. Those are the programs whose model a Hoyani home visit most resembles, a scheduled clinician visit for an established patient with a chronic condition.

Post discharge follow up

Follow up after a hospital stay is the first service in a Hoyani pilot because it is where the evidence is deepest. Transitional care programs that reach a patient soon after discharge are consistently associated with fewer 30 day readmissions in the peer reviewed literature, and a 2024 systematic review and meta-analysis in Preventing Chronic Disease found that early outpatient follow up reduced 30 day all cause readmissions for heart failure, COPD, myocardial infarction and stroke. The program’s post discharge assessment in section 14 is built on that pattern, a clinician in the home within days of discharge, medication reconciliation, and findings owned by a named nurse on the care team.

Transportation to treatment

A ride is a health outcome when the alternative is a missed dialysis session. In the 2018 return on investment study commissioned by the Medical Transportation Access Coalition, dialysis patients with non-emergency medical transportation attended an average of twelve treatments a month against 4.1 without it, and after the cost of the rides the net saving was $3,423 per member per month. For diabetic wound care the net saving was $792 per member per month. Those figures are why the transportation line is enabled in a first pilot and why every ride is tracked to completion.

Community health work

The Community Health Representative program is the oldest community health worker program in the country and it is Tribally directed. A matched cohort study of Navajo adults living with diabetes, published in 2020, found that community based outreach by CHRs working alongside clinic teams was associated with increased use of health services. Hoyani hands work to that program rather than duplicating it, and tracks the hand-off, for exactly that reason.

The workforce gap the model is built for

The Government Accountability Office reported in August 2018 an average vacancy rate of 25 percent for physicians, nurses and other providers across the Indian Health Service areas it examined, ranging from 13 to 31 percent by area. On 29 January 2026 the agency launched the largest hiring effort in its history, with its Chief of Staff citing a near 30 percent vacancy rate across the service. A flexible clinician reserve that can be dispatched to a home visit a program cannot staff is a direct response to that number, and section 18 states how it supports permanent hiring rather than competing with it.

What the pilot adds

None of the studies above evaluated Hoyani, and the site does not present them as though they did. They establish that the services Hoyani coordinates work when delivered well. What the pilot establishes is that this coordination layer delivers them well in this setting. For an initial feasibility pilot the reporting is counts, denominators, completion, documentation return, follow up ownership, financial reconciliation, cancellations, adverse events and patient feedback, for app users and assisted users alike. Clinical benefit is claimed after an appropriate outcome over a realistic horizon, and cost effectiveness after the full incremental cost is counted against a credible comparison. Company operating experience, partner experience, pilot results and external research are four different things and are labeled separately in every document.

One willing operating partner, one defined population, one service protocol and one primary payment arrangement. Complexity is added only when it answers a specific evaluation question. The pilot produces evidence useful to the partner even if no broader rollout follows.

An illustrative starting scope is 50 to 100 consenting adult participants over roughly 90 days of operations. Those are planning assumptions rather than a statistical sample-size recommendation or an approved protocol. Final scale depends on clinical risk, workflow volume, staffing, budget and partner approval, and startup and claims follow-up extend beyond the operating period.

Entry conditions

Before enrollment, the pilot documents clinical accountability, permissions and agreements, patient protections, trained staff, the approved service area, data handling, payment arrangements and a realistic stop procedure. The workflow is tested with synthetic cases before any patient information is used. If the evaluation constitutes research requiring additional review, the appropriate determination is obtained first.

Staged decisions and the stop procedure

Preparation approves the workflow and governance, verifies the workforce, establishes the baseline and completes contingency rehearsals, releasing only the resources the next justified stage needs. Operations review safety and unresolved care issues continuously, and review operations and billing at a cadence agreed with the partner, maintaining a log of every exception and the person responsible for resolution. Protocol changes run through controlled approval rather than informal adjustment by individual staff. The decision point evaluates feasibility after the operating period with enough claims runout to judge collections, against the partner’s actual baseline or a suitable contemporaneous comparison, and expansion happens only if the same approach is likely to work in the next location and that community independently agrees.

New enrollment pauses for a material unresolved privacy incident, an unsafe handoff process, loss of essential clinical coverage, or recurring unauthorized billing. The precise triggers and the restart authority are set before launch. Low margin alone may justify redesign. Unresolved patient safety requires a different and immediate response.

If the pilot ends

No patient is left depending on Hoyani for anything the program cannot continue. Every participant remains an established patient of the program throughout, every finding lives in the program’s own record, and every open item has a named owner on the program’s care team. On a stop, open episodes are closed through the program’s normal pathway within a defined number of days, patients and representatives are told in plain language what changes and what does not, equipment stays with the program under section 17, and the data handling in the governance agreement runs. The Indian Health Service absorbs no dependency and inherits no obligation. That is the answer to the question of what happens to the patients who came to rely on it, and it is designed in before the first one enrolls.

Hoyani starts small on purpose and is built to grow. The purpose of the platform is to scale, because the more of Indian Country it reaches with the right routing, the more benefit flows to Indian health and Tribal health programs and the citizens they serve. Each stage below opens only with the approval of the community and the program it serves, and only after the stage before it has reported its results. The order is the same everywhere. Respectful, small, approved, then wider.

A favorable meeting with one participant does not settle the others. These are separate decisions made by separate people, and the request to each is stated in the form that person can act on.

Funding

A first demonstration is deliberately small enough to be supported several ways, including a participating program’s own resources, a philanthropic or foundation partner, a state rural health demonstration, a federal demonstration, or private investment. The structure of the model does not depend on which is chosen. Development spending and clinical reimbursement are accounted for separately, and no expense is charged to two funding sources.

These are the questions a Tribal leader, a medical director, an agency official, a contracting officer, a payer and an architect ask first. Each has an answer on this page, and the table says where. The decisions that belong to the reviewer rather than to Hoyani are set out in section 20, with what is prepared for each.

A reviewer should be able to distinguish a concept, a working demonstration, a restricted pilot and a production service without reconciling different parts of a page. Every capability named on this site appears below with its status, the artifact that supports it, the person accountable for it and a target date. A reviewer who finds an unstated gap reasonably assumes there are others, so the unresolved items are listed here rather than left out. A missed date is reported as missed, with the reason, rather than moved quietly.
Revenue cycle integration | Partner/program RCM pathway not yet selected | Required before first billable pilot encounter | Owner: Hoyani + participating program revenue cycle lead