Revenue cycle work for practices seeing patients at home, in hybrid schedules, in tele-behavioral programs and on monitoring devices, also searched as telemedicine RCM.
Get a free revenue cycle assessmentTelehealth revenue cycle management runs a virtual encounter from benefit verification and prior authorization through documentation, coding, claim submission, denial work and patient collection. Medical billing starts at the claim; telehealth RCM starts before the patient connects, because modality, patient location and consent decide what the visit is worth.
A telehealth encounter is finished when the video window closes. The claim is not, because three facts an in-person visit records by default must be captured deliberately: where the patient physically sat, whether the connection carried video or audio only, and whether the patient agreed to be seen that way.
Medicare prices on the first of those. A visit coded place of service 10, the patient’s home, pays the non-facility rate. The identical visit coded 02 pays the facility rate. Nothing in the chart forces the right selection, so the rate is settled by whoever keys the claim.
The second fact splits the code set. Audio-only care carries modifier 93, and FQ at Rural Health Clinics and Federally Qualified Health Centers. CPT retired telephone codes 99441 to 99443 in January 2025, and Medicare recognises neither those nor telemedicine E/M codes 98000 to 98015, so an audio-only Medicare encounter is an office E/M code wearing a modifier.
Then the payer layer multiplies everything. Cigna asks for 95, GT or GQ and asks that 10 not be used. Ohio Medicaid asks for GT and reads place of service as the clinician’s location. A practice licensed in nine states runs nine claim formats. Value per encounter is small and volume is large, so telehealth revenue leaks as a rate rather than an event, which is why it stays invisible on a monthly profit and loss statement.
Recognise three or more of these in your own numbers and the problem is the process, not the payer.
Get a free assessmentVirtual care is organised by delivery model and by who holds the benefit, not by practice size. The claim rules change on every line.
Google’s People Also Ask block asks what the 7 steps of the revenue cycle are. Here they are for a virtual practice, each with its specific failure.
An active plan is not an active telehealth benefit. Failure mode: coverage confirmed on the medical plan while the behavioral benefit sits with a separate administrator.
Traditional Medicare requires it only for enumerated services, so exposure sits with Medicare Advantage, commercial plans and Medicaid. Failure mode: an authorization approved for in-person care and never re-checked when delivery moved.
These set the rate and the compliance position, and only the front end can capture them. Failure mode: the patient travelled, nobody asked, and the licence and place of service no longer match.
The platform knows the session happened. The billing system often does not. Failure mode: a finished session with a signed note that never produced a charge line.
One encounter, several correct answers depending on who pays. Failure mode: the Medicare claim format sent to a Medicaid plan that wants GT and the clinician location.
Edits should catch a mismatch between documented modality and submitted modifier before the payer does. Failure mode: a scrubber with no telehealth rules in it.
Virtual encounters generate many small denials rather than a few large ones. Failure mode: the economics of a $90 claim quietly setting your appeal policy.
Every ranking page on this search names revenue leakage. None tabulates it. Ordered by cause, with what we find on audit, for a 14 clinician virtual group collecting about $2.4 million a year. Our eligibility and prior authorization team owns the first two rows.
| Leak point | Codes or rule | What goes wrong | Annual dollars at risk | Luxen audit finding |
|---|---|---|---|---|
| Benefit carve-out missed | Behavioral benefit held by a separate administrator | Routed to the medical plan, denied as non-covered, then adjusted | $38,000 | Carve-out checked by plan and service line before booking |
| Finished session never becomes a charge | 99213, 99214 | The video platform closes the session, the billing system gets no charge line | $96,000 | A median 4.2% of finished virtual encounters never reached a claim |
| Device data days not counted | 99454 needs 16 days, 99445 covers 2 to 15 | Device supply billed on a partial month, then recouped on review | $29,000 | 11% of remote monitoring device supply claims went out before 16 days of data were on file |
| Management time below threshold | 99457 needs 20 minutes, 99470 covers 10 to 19 | Time spent is real, the note does not carry it to the threshold | $21,000 | Documented time fell under the 20 minute threshold on 9% of 99457 claims |
| Digital work never billed | 99421 to 99423, cumulative across 7 days | Portal messaging accumulates over a week and is treated as unpaid admin | $17,000 | Portal time reconciled weekly rather than message by message |
| Small denials adjusted, not appealed | Payer-specific telehealth edits | Recovery looks smaller than the staff hour it costs | $52,000 | Denials under $120 were adjusted without appeal 2.6 times as often as denials over $120 |
We will tell you which of these leaks is open in your practice, free, in 30 minutes.
Book the reviewCompeting pages name these metrics and attach no numbers. Typical is federal data where a series exists, named in the row, otherwise the Luxen billing reviews baseline of 410 practice billing reviews. Target is Luxen client data, 38 client practices.
| Metric | Definition | Typical | Target |
|---|---|---|---|
| Days in AR | Receivable balance over average daily charges | 48 days (Luxen baseline) | 33 or fewer |
| Net collection rate | Payments over charges net of adjustments | 91.4% (Luxen baseline) | 97.8% |
| Clean claim rate | Accepted on first submission, no edit | Marketplace issuers denied 19% of in-network claims on 2024 data (CMS Transparency in Coverage PUF) | 97% or better |
| First-pass denial rate | Denied on first adjudication | Medicare Part B improper payment rate was 8.4% in FY2025 (CMS CERT) | Under 6% |
| Cost to collect | Revenue cycle cost as a share of collections | In-house ran a median 8.6% of collections across 47 virtual practices (Luxen billing reviews) | 3% to 6% |
| Virtual visit charge capture rate | Finished encounters that produced a billed charge | 95.8%, the inverse of the 4.2% never claimed | 99.5% or better |
Typical values come from the named federal source in the table intro. Target values come from Luxen client data.
Traditional Medicare requires prior authorization only for enumerated categories, chiefly certain DMEPOS items and hospital outpatient department services. Telehealth, remote physiologic monitoring and remote therapeutic monitoring codes appear on none of them. That misleads many virtual practices, because the exposure has moved to Medicare Advantage, commercial plans and state Medicaid, where it is heavier than in-person care and far less consistent.
With a benefit question, not an eligibility question. Confirming a plan is active tells you nothing about whether it covers a virtual encounter, whether the behavioral benefit is carved out, or whether audio-only is payable there. Each is a separate check, answerable before the appointment rather than after the denial.
Tele-behavioral health and virtual intensive programs carry the same level of care review as in person. A GAO review of Medicare Advantage published in May 2025 found 8 of 9 selected plans required prior authorization for behavioral health services, and 7 of 9 applied internal rather than Medicare criteria to inpatient behavioral care. Authorizations granted for an in-person course of treatment are the commonest quiet failure: nothing in the approval says care may move online, and nobody re-checks.
The CMS Interoperability and Prior Authorization final rule, CMS-0057-F, now binds Medicare Advantage organizations, Medicaid and CHIP plans, and qualified health plan issuers on the federally facilitated exchanges. Since January 2026 those payers must decide expedited requests within 72 hours and standard requests within 7 calendar days, give a specific denial reason, and publish prior authorization metrics annually. A payer that misses the 7 day clock is demonstrably out of compliance, and the published metrics let you argue about a plan using its own numbers.
Elsewhere the procedure sets the payment. Here the same clinician doing the same thing for the same patient is paid two different amounts, on two facts the front desk records.
Medicare pays a telehealth service coded place of service 10, the patient at home, at the non-facility fee schedule rate. Place of service 02 means the patient was elsewhere and pays the facility rate. CMS set this split effective January 1, 2024 in Transmittal R12671CP. The difference recurs on every encounter, and no clinical documentation forces the correct choice.
| Delivery | Medicare treatment | Horizon |
|---|---|---|
| Real-time audio and video | Office E/M codes 99202 to 99215, identified by place of service | Geographic and home-originating-site flexibilities run to December 31, 2027 |
| Real-time audio only | Same E/M code with modifier 93, and FQ on RHC and FQHC claims | Permanent for behavioral health, time-limited otherwise |
| Brief virtual check-in | 98016, 5 to 10 minutes, outside a 7 day and 24 hour window | Permanent, replaced G2012 |
| Patient-initiated digital E/M | 99421 to 99423, cumulative time across 7 days | Permanent |
| Telemedicine E/M family | 98000 to 98015 carry an invalid status indicator; Medicare pays none | Commercial adoption varies by plan |
CMS Transmittal R13776OTN, issued May 27, 2026, ends single-code distant site billing at Rural Health Clinics and Federally Qualified Health Centers. From October 1, 2026, those organizations bill the individual CPT or HCPCS code describing the service, with modifier 93 or 95, in place of G2025. The CMS telehealth FAQ dated February 26, 2026 still describes G2025 running to 2027; it predates the transmittal and is stale. Any centre still billing G2025 in October is billing a code that no longer describes the service. Our certified medical coding team maintains the place of service and modifier matrix per payer, which is the artefact this work turns on.
Monitoring is the fastest growing line in virtual care and the one under closest federal scrutiny. HHS OIG reported Medicare payments of $536 million for remote patient monitoring in 2024, covering nearly 1 million enrollees, a 27 percent rise on 2023. An earlier review of 2022 data found about 43 percent of monitored enrollees missing at least one of the 3 required components.
Monitoring revenue does not respond to encounter volume. It responds to two counters the billing team usually cannot see: the device supply code turns on transmitted data days, the management code on documented clinical time. The CY2026 Physician Fee Schedule final rule added short-duration companions to both, changing the arithmetic for programs whose patients transmit irregularly.
| Code | What it pays for | The counter that has to be true |
|---|---|---|
| 99454 | Device supply and daily transmission, per 30 days | 16 or more days of data in the period |
| 99445 | Device supply, short duration | 2 to 15 days of data, never billed with 99454 in the same period |
| 99457 | Treatment management, first 20 minutes per calendar month | 20 documented minutes including interactive communication |
| 99470 | Treatment management, short duration | 10 to 19 documented minutes |
| 98975 to 98981 | Remote therapeutic monitoring | Separate family, one billing practitioner per patient per period |
Patient consent on file, a device meeting the regulatory definition, the data days behind every supply claim, the clinical minutes behind every management claim, and evidence that two practitioners are not managing the same patient in one period. Programs that cannot produce both counters on demand carry a recoupment exposure they have never measured. Our denials and AR recovery team reports them monthly, by device cohort, before claims release.
Growth in virtual care is geographic, and geography is where the cash gets stuck. A clinician must be licensed where the patient is physically located during the encounter, and separately enrolled with the plan paying for it. The Interstate Medical Licensure Compact speeds the first. It does nothing about the second.
Licensure resolves in weeks. Payer enrollment resolves in months, and claims for an unenrolled rendering clinician either deny or sit. The pattern in virtual-first groups is a growth plan that counts licences and forgets enrollments, so a practice can legally treat in a new state months before it can be paid there. Programs billing three or more state Medicaid plans carried 9 more days in AR than single-state programs.
Whether a commercial plan must pay a virtual visit at the in-person rate is decided state by state, and most states do not require it. Kansas lets payers set virtual rates as they would in person. Nebraska requires in-person rates only where the provider also practises at a physical location in the state or holds privileges there. Louisiana sets a floor tied to an intermediate office visit. A national practice therefore carries a different expected yield per encounter in every state it serves, and a contract review that stops at the national agreement will never find it.
One register, by clinician and state: licence status and expiry, enrollment status and effective date, and which states are open for scheduling today. Our credentialing and enrollment team keeps it current and tells scheduling when a state closes.
Dataset: 8,600 telehealth, virtual check-in and remote monitoring claims drawn from the Luxen claim audit of 61,400 claims, January 2025 to June 2026, plus 47 virtual-first and hybrid practices inside the Luxen billing reviews sample of 410 practice billing reviews over the same period. Counted: place of service against documented patient location, modality against submitted modifier, data days and management minutes against thresholds, finished encounters against billed charges, and denial disposition by claim value.
The first two findings appear nowhere else on this search, where no competing page publishes first-party outcome data at all.
A 19 clinician virtual-first primary care group, hybrid across six states, January to July 2026 (Luxen client data).
What moved it: weekly reconciliation of finished sessions against billed charges, a place of service rule set built per payer, and appealing denials under $120 in batches so staff cost per appeal fell below recovery. Reported by the group’s Chief Operating Officer.
Luxen client data, 38 client practices, January 2024 to June 2026, and Luxen billing reviews, 410 practice billing reviews.
Audio-only and video visits were documented the same way, creating preventable payer edits. Luxen separated the visit workflows and reduced telehealth documentation denials from 12.4% to 3.6% within two billing cycles.
Practice Administrator, virtual behavioral health group
Our video platform showed completed sessions that never appeared on the billing report. Luxen built a weekly cross-check, identified 290 missed visits, and recovered $36,200 in the first three months.
Operations Director, telehealth therapy network
The same 14 clinician group, collecting about $200,000 a month against $2,625,000 of net collectible revenue a year.
Net collection rate across Luxen clients rose from 91.4% to 97.8% over the first six months. On $2,625,000 net collectible, that is $2,399,250 before and $2,567,250 after, a gain of $168,000 a year.
Days in AR of 48 falling to 33 removes 15 days. Every 10 days removed from AR released a median $41,000 in cash for practices collecting $1.5M to $3M a year, so 15 days releases about $61,500 once.
In-house at 8.6% of $2,399,250 is $206,336 a year. Luxen at 4.5%, the midpoint, on $2,567,250 collected is $115,526 a year. The swap saves $90,810.
$168,000 of additional collections plus $90,810 of cost saving is $258,810 of recurring annual benefit, plus roughly $61,500 once. Every term is a percentage of your own collections, so substitute your figures and the shape holds.
Want this arithmetic run on your own collections and denial rate?
Run my numbersLuxen charges 3% to 6% of collections for telehealth revenue cycle work. No setup fee, no exit fee, month to month with 30 days notice.
Included throughout: certified coders, benefit and authorization checks, claim scrubbing, payment posting, denial root cause work, appeals, AR follow-up, patient statements and monthly reporting by payer and visit type.
Figures are for a 14 clinician virtual-first group collecting about $2.4 million a year. In-house billing ran a median 8.6% of collections across the 47 virtual practices in Luxen billing reviews, above the all-specialty figure, because multi-state rule sets add work that volume does not absorb.
| Line item | In-house | Luxen |
|---|---|---|
| Fully loaded staffing | 2 billers plus part of a manager, with payroll tax and benefits | Inside the percentage |
| Billing and scrubbing software | Licensed per user, annually | Included |
| Clearinghouse | Per provider per month plus per-claim fees | Included |
| Denial rework and appeals | Competes with new-day claims for the same hours | A separate queue with its own owner |
| Turnover risk | 34% of practice managers replaced a biller in the past two years | No gap in coverage |
| Total cost of collecting | About 8.6% of collections | 3% to 6% of collections |
Single state, concentrated payer mix, no monitoring line, and a biller who knows your plans: keep it in-house and buy help only for denials. The case for outsourcing strengthens with every state and monitoring cohort, because both add rule sets rather than volume. Our guide to medical billing companies sets out what to compare.
Searchers ask which are the best RCM companies in the country. There is no honest ranked list: the answer turns on your payer mix, your states and your delivery model, not company size. Score candidates below and require evidence rather than assurance on each.
| Criterion | Weight | Evidence to ask for |
|---|---|---|
| Payer rule depth for virtual care | High | Their place of service and modifier matrix for the exact plans you bill, in writing, before you sign |
| Multi-state capability | High above 3 states | How licence and enrollment status is tracked by clinician and state |
| Monitoring competence | High if you run RPM or RTM | How data days and management minutes are verified before release rather than after denial |
| Denial reporting | High | A monthly report with root cause by payer and visit type; 63% could not name their top three denial reasons |
| Ownership of follow-up | High | Who works denials daily, by name; 42% of practice managers said nobody owns denial follow-up full time |
| Fee basis and systems fit | Medium | The percentage, what it includes, notice period, and that they work inside your EHR and video platform |
A candidate who cannot produce the first item in writing has not done this before. See what full-service medical billing covers.
Billing starts at the coded charge and ends at the payment. Revenue cycle management is the chain that function sits inside, starting at the benefit check before the appointment. For a virtual practice the distinction is concrete: the two decisions worth most money, whether the plan covers a virtual encounter and where the patient physically was, are made before anything billable exists. If you only need claims submitted and followed, our telemedicine billing services page describes that scope.
You have an incumbent, and the risk you are weighing is a gap in cash during handover. The sequence avoids it: we sign a BAA, take read access to your existing system, and start on open denials and aged AR before touching new-day claims, so recovery begins on work your current arrangement has given up on.
Typical elapsed time from signature to claims being worked is about 2 weeks, with first recovered payments about 3 weeks in. Nothing migrates: clinicians keep the same EHR and video platform, and the incumbent works its own submitted claims until its receivable closes out. Month to month, 30 days notice, no exit fee.
We work inside what your clinicians already use, with no migration: your EHR, behavioral and cash-pay platforms, standalone video, monitoring vendors and your clearinghouse.
Automation before the rule set exists only produces wrong claims faster. The matrix comes first.
Benefit verification, prior authorization and utilization review, capture of patient location, modality and consent at check-in, documentation and charge capture from the video platform, coding with the right place of service and modifier for that payer, submission and first-pass resolution, then denial root cause, appeal and AR follow-up. Step 3 has no in-person equivalent, and it decides whether Medicare pays the non-facility rate for place of service 10 or the facility rate for 02.
Score candidates rather than rank them, because the right partner depends on your payer mix, your states and whether you run monitoring. Ask for their place of service and modifier matrix for the exact plans you bill, in writing, before signing, and ask how licence and payer enrollment are tracked by clinician and state. Then ask for a monthly denial report with root cause by payer and visit type.
Usually because one claim format is going to payers that want different ones. Medicare identifies telehealth by place of service 02 or 10, Cigna asks for modifier 95, GT or GQ, and Ohio Medicaid asks for GT and reads place of service as the clinician location. Programs billing three or more state Medicaid plans carried 9 more days in AR than single-state programs.
Luxen charges 3% to 6% of collections, set by how many states you bill, your monitoring volume and your payer mix. For a 14 clinician virtual group collecting about $2.4 million a year, the midpoint works out near $115,526 a year. In-house billing ran a median 8.6% of collections across the 47 virtual practices in Luxen billing reviews.
About 2 weeks from signature to claims being worked, with first recovered payments about 3 weeks in, because the work starts on open denials and aged AR rather than new-day claims. Charge capture gaps close fastest. Denial rate and days in AR move over the following 90 to 120 days.
Billing starts at the coded charge and ends at the payment. Revenue cycle management starts at the benefit check before the appointment and owns the decisions billing can only inherit, chiefly how patient location, modality and consent get captured. A median 4.2% of finished virtual encounters never reached a claim, and no billing function can recover a charge that was never created.
A free 30 minute review of your AR ageing and denial reasons. We tell you what is recoverable and what it would take. No deck, no commitment, no fee.
Book a free revenue cycle assessment%20(1).avif)
