Research

Check multistate telehealth licensure exposure

Builds the state-by-state exposure map and a sourced research plan with blank finding fields, so you get a worksheet you can actually verify instead of a confident chart that is eighteen months stale.

About 30 minadvancedHealthcare, Regulatory

Your prompt5,523 characters

Still to fill in: States where patients are located, Clinicians and their licenses, Service model

RoleYou are a regulatory lawyer who builds multistate licensure surveys for telehealth companies and has watched more than one vendor's 50-state chart turn out eighteen months stale. You do not answer licensure questions from memory, because the answer changes by board rule and by year. You produce the research plan, the sources, and the blank finding fields, and treat any rule you cannot cite as a question.What I needBuild the licensure exposure map and research plan for the service model below across States where patients are located. Do not tell me what the law is. Tell me what to check, where, in what order, and what a wrong answer costs.InputsStates where patients are located: States where patients are located Clinicians and their licenses: Clinicians and their licenses Service model: Service model Entity and ownership: Entity and ownership structure Sources I hold: Sources I already holdHow to work this1. Build the exposure map first. One row per state: volume, how patient location is established, license type, service type, modality, prescribing. Rank by volume times consequence and name the five states to work first. 2. For each state produce a worksheet, not an answer: the question, the primary source, the exact search terms, and a blank finding field with "checked by / date" underneath. Blank fields are the correct output. 3. Run the same question set in every state, every answer left blank: full licensure required to treat a patient located there; any telehealth registration or consultation pathway; exceptions that might reach this model; which compact covers the license type and whether the state is operational in it, verified at the compact's own roster (enacted, effective, and operational are three different facts); modality limits including audio-only; what establishes the patient-provider relationship; telehealth consent requirements. 4. Give prescribing its own worksheet: the federal controlled-substance telemedicine framework and its current status, the state's in-person examination requirement, prescription monitoring duties. This area has moved repeatedly. The date on a source matters as much as its text. 5. Flag corporate practice of medicine, fee splitting, and professional entity requirements separately where Entity and ownership structure makes them live. Keep payer coverage out. Reimbursement is a different question. 6. Close with interim controls that cut exposure while research runs: geofenced scheduling, location attestation at intake, established-patient limits, routing to a licensed clinician, and say what each does not solve.Ask me firstBefore you produce anything, ask me these questions, then stop and wait. The states, the license roster, and the service model are above. These are the things the paperwork cannot tell you: 1. Where is the patient physically sitting during the encounter: the state list I gave you, or billing addresses? 2. If a state comes back closed, what will the company actually do: stop booking there, sponsor a license, hand the panel to someone already licensed, or keep going? That decides whether I am building a research plan or a wind-down plan. 3. Does any encounter involve prescribing, are controlled substances involved, and was the patient ever seen in person? 4. Is this a launch or volume already flowing? If flowing, how much per state and since when. That is planning versus remediation. 5. Has any board, patient, payer, or plaintiff already raised this (a complaint, an inquiry, a demand) and has anyone at the company put an answer in writing? Do not begin until I answer. If I tell you to proceed anyway, state each assumption at the top and mark it [ASSUMPTION - verify].Output formatAn exposure map ranked by priority; one worksheet per state with source, search terms, and blank finding fields; a separate prescribing worksheet; a compact verification checklist; interim controls; and a list of every question you refused to answer from memory. End with one line naming the two of my answers that most changed the exposure ranking and the interim controls, and how the map would have been ordered without them. If an answer changed nothing, say so. It means I should not have been asked.Never do this- If this plan would fit any telehealth company in any state, it is too generic. Anchor it to these license types, this modality, this volume. - No hedging filler. Cut "arguably," "it should be noted," and "requirements vary by state" offered as a finding. Do not tell me to consult regulatory counsel. I am regulatory counsel. - Never state a licensure rule, registration pathway, exception, compact membership, fee, or deadline from memory. Every legal statement carries a source from my inputs or is marked [UNVERIFIED - confirm at the board]. A filled-in table you cannot source is the failure here; a blank one is not. - Where you do not know, leave the field blank and say you do not know. Do not smooth the gap over with fluent prose, and never infer one state's rule from another's. - Do not pad. A three-state map is three rows. Length is not value.Before you answer- Did I leave finding fields blank rather than filling them from memory? - Does every legal statement carry a source or an [UNVERIFIED] mark? - Did I point to the compact roster instead of recalling membership? - Would this plan fit a different company? It should not.

Adds driver's-seat tunes: options instead of answers, questions before work, every citation flagged. Your values come with it.

Pressure-test it

Makes the AI switch hats and attack its own answer.

A patient complaint opened the file, and the investigator at the state medical board has the platform's scheduling logs. Read this as the investigator would. Walk the encounters state by state and name where the board would say the clinician practiced without a license, which exception the company will claim and why it fails on these facts, what the platform's own records prove against it, and which single state carries the worst first-mover risk. Then tell me what to stop doing this week.

Go deeper

Pushes the work further once the basics are right.

Engineering cannot enforce a memo; it can enforce a rule. Convert the completed worksheets into the operating artifacts: a go / go-with-conditions / no-go matrix by state and license type, the scheduling rules engineering has to enforce at booking, the intake script that captures and records patient location, the escalation path when a patient joins from an unapproved state mid-session, and the quarterly re-verification calendar with named owners.

Before you run it

What to gather first

Encounter volume by state where the patient was physically located, not by billing address
Every clinician license type in play, and which are compact-issued versus state-issued
Whether prescribing occurs, whether controlled substances are involved, and whether any in-person visit ever happened
The entity's ownership structure and whether any non-licensee holds equity
Any survey, compact roster, or board correspondence you already hold, with its date

Watch for

Treat every licensure statement in the output as unverified until you read the practice act, the board's current rules, and the board's own telehealth guidance. This is the single highest-hallucination area in health regulation.
Compact membership changes several times a year, and enacted, effective, and operational are different states of the world. Verify at the compact's own roster, never from a survey or from memory.
Federal controlled-substance telemedicine authority has been extended and revised repeatedly. Confirm what is in force on the date you advise, and note that state in-person examination requirements apply on top of it regardless.
Board FAQs and staff emails are helpful and are not law. Cite the statute or rule they rest on before relying on them.
Do not paste patient-identifying encounter data. Volumes, states, and modalities are enough to build the map.

What comes back

An exposure map ranked by volume times consequence; one research worksheet per state carrying the question, the named primary source, the exact search terms, and a blank finding field with checked-by and date; a separate prescribing worksheet; a compact verification checklist naming the roster to open; interim controls with what each does and does not solve; and an explicit list of every question the model refused to answer from memory.

See an example of what you’ll get

*(After you answer the five questions: 18 states, patient-location basis confirmed from scheduling logs; volume already flowing since Jan 2025; stimulant prescribing in TX and FL, with about 40% of psychiatry patients never seen in person; if a state comes back closed the company stops booking there rather than sponsoring licenses; and the Montana board emailed in February, which nobody has answered.)* EXPOSURE MAP: research order | Rank | State | Encounters/yr | Clinician license basis | Prescribing | Why this rank | |---|---|---|---|---|---| | 1 | Montana | ~40 | No clinician holds an MT license | Yes: 6 stimulant patients | Lowest volume, highest consequence. Unlicensed practice plus controlled substances plus a board that already emailed you | | 2 | New York | ~1,100 | 2 psychiatrists licensed; 9 LCSWs are not | Yes | Large volume against a partially unlicensed panel | | 3 | Texas | ~4,200 | Licensed | Yes | Volume plus the corporate practice question raised by the MSO structure | | 4 | Florida | ~2,900 | Licensed; 3 via IMLC | Yes | Verify the IMLC licenses are active and state-issued, not assumed | | 5 | 11 states under 25 encounters | ~180 total | Mixed | Mostly no | Aggregate, then triage after the top four | WORKSHEET: MONTANA (findings intentionally blank) | # | Question | Where to look | Search terms | Finding | Checked by / date | |---|---|---|---|---|---| | MT-1 | Does treating a patient physically located in MT require an MT license? | MT medical practice act; board administrative rules | "telemedicine" "located in this state" "practice of medicine" | | | | MT-2 | Is there a telemedicine registration or out-of-state consultation pathway? | Board licensure pages; the act's exemption section | "telemedicine license" "consultation exception" "out-of-state physician" | | | | MT-3 | Is MT currently an operational IMLC member for the physicians' license types? | The IMLC roster itself, not a survey | State list; effective date; any implementation note | | | | MT-4 | Is a compact license available for LCSWs? Which compact, and is MT in it and operational? | Social work compact roster | Member states; effective date | | | | MT-5 | What establishes the patient-provider relationship, and is a prior in-person or synchronous visit required? | Board rules; any telehealth standard-of-care rule | "physician-patient relationship" "telehealth" "in-person" | | | | MT-6 | Does audio-only qualify as a permitted modality? | Same rule set | "audio-only" "telephone" "synchronous" | | | PRESCRIBING LAYER: separate worksheet. Federal controlled-substance telemedicine authority has been extended and amended repeatedly; confirm what is in force on your advice date and record the effective date next to the finding. Separately confirm each state's own in-person examination requirement and PMP query duty, which apply regardless of the federal position. [UNVERIFIED - every element of this layer.] COMPACT CHECKLIST. Open the IMLC roster and the relevant behavioral health compact rosters directly. For each state, record three separate facts: enacted, effective, and issuing licenses today. Your Q3 2024 vendor survey is not a source for any of the three. INTERIM CONTROLS - Geofence booking to the states on the approved list: solves new exposure; does nothing about the 40 Montana encounters already in the record. - Capture and log patient location at the start of every session: solves proof; creates a record that will be produced if a board asks. - Suspend new controlled-substance starts for patients never examined in person, in every state, until the prescribing worksheet closes. WHAT I REFUSED TO ANSWER FROM MEMORY. Whether Montana requires full licensure. Whether any consultation exception reaches recurring therapy. Whether Montana is an operational IMLC state. Whether audio-only is permitted anywhere on your list. Every prescribing question. Whether your MSO structure survives Texas corporate practice doctrine. Six of these are answerable in under an hour at the primary sources named above. WHAT YOUR ANSWERS CHANGED. Montana is ranked first at 40 encounters, ahead of Texas at 4,200, because of two things you told me: the board there has already written and no one has written back, and the stimulant prescribing runs to patients nobody has examined in person. On volume alone the order is Texas, Florida, New York, and Montana sits near the bottom. That is the map you would have received if I had built it off your state list and stopped there. That volume has been flowing since January 2025 is why the interim controls read as remediation rather than launch gates: the geofence entry says outright what it does not fix, and the location-logging entry warns you it builds a record a board can ask for. Your answer that a closed state means you stop booking rather than sponsor licenses changed nothing about this map. It decides what happens when a worksheet closes, not what order the worksheets get worked.

Why this prompt is built the way it is

## Framework 1. **Patient location governs.** Build the map on where the patient physically sits at the time of service, not on billing address or clinician location. 2. **Exposure map before research.** Volume times consequence tells you which five states to work first. 3. **Worksheets, not answers.** Question, primary source, search terms, blank finding field, checked-by and date. A blank field is a correct output. 4. **Compacts get verified at the roster.** Enacted, effective, and operational are three different facts. 5. **Prescribing is its own layer.** Federal controlled-substance telemedicine rules have moved repeatedly; the date on the source matters as much as the text. 6. **Keep the layers apart.** Licensure, corporate practice of medicine, and payer coverage are three questions. Merging them produces a model built on the wrong memo. 7. **Interim controls buy time.** Geofencing, location attestation, and established-patient limits reduce exposure while the research runs. Say what each does not solve.