We agree with CMS's goals, all of them. But the rule regulates two programs and disconnects a million seniors. This plan regulates the whole remote care family (RPM, RTM, CCM, PCM, BHI) with more accountability and roughly double the savings, and not a single senior loses care.
The HHS Office of Inspector General documented it in September 2024: cold-calling operators enrolling patients they never treated, 43% of enrollees not receiving all three required service components, and Medicare unable to see who ordered the monitoring or which companies furnished it. Meanwhile spend compounded: $15M (2019) → $536M (2024) for RPM alone, with CCM, PCM, and BHI riding the same curve. Nobody serious disputes any of this. We don't either.
Cold-call operators signing up patients with no clinical relationship, the OIG found practices with no prior relationship to more than 80% of the patients they billed.
Medicare cannot see who ordered monitoring, what data is monitored, or which companies furnish it. Waste hides in the dark.
A physician spends a decade earning the right to bill Medicare. Anyone can start a remote care company tomorrow, no registration, no standards, no bond.
RPM +31%/yr; CCM bigger still at just ~4% penetration. Left alone, the family compounds toward $7B+ a year by 2031.
From January 1, 2027, Medicare pays only when monitoring is performed by the practice's own employed staff, supervised clinical partnerships are cut off, with no exemption for patients already enrolled.
A new face-to-face initiating visit is required, and the rule is silent on the million-plus patients already in care, no exemption, no transition.
Code re-valuations engineered in two steps: the statute caps year-one cuts, so a monitored month falls only ≈$104 → $91 in 2027 — then CMS's own phase-in tables take it to ≈$40–58 in 2028, and a floated consolidation into flat G-codes would land at ≈$35.
RPM software and device providers estimate that 75–80% of the market is delivered through third-party monitoring partners, because most practices are incapable of running these programs themselves. On January 1, that is who loses access. Vendors can attest to the share; we'll help CMS assemble that record.
CMS moved remote monitoring from direct to general supervision precisely because practices could not staff 24/7 monitoring in-house. The partnership model isn't a loophole; it's the delivery architecture CMS's own rulemaking created. One nurse covers ~300 patients (≈$94K median salary, BLS), more patients than most practices have, and round-the-clock coverage no single hire can provide.
Pure-play RPM companies that raised tens of millions of dollars have gone out of business, the devices, logistics, software, and 24/7 clinical staffing are that hard to operationalize. And the largest health systems (Providence, LifePoint, Community Health Systems, Sentara, UnityPoint, Mount Sinai) publicly contract specialized partners; the systems that self-build run subscale programs their patients outgrow. Their care teams are centralized anyway, not employed by each billing practitioner, so the W-2 test fails even them.
The deep cut is deferred, not avoided: section 1848(c)(7) of the Act caps year-one reductions at 19%, so 2027 shows a survivable −12% for a typical month — then CMS's own "Fully Implemented" tables complete the phase-in, taking device supply from $41.38 to $9.85 and the month to ≈$40–58 in 2028. At those rates — let alone the ≈$35 flat G-code floated in the rule — no delivery model, employed, partnered, or health-system, can furnish 24/7 clinical monitoring. The payment provisions end the benefit by arithmetic; the staffing ban just gets there first. The rule's own tables cannot say within 45% what a monitored month pays in 2028 — the valuation record is as incomplete as the access record. And a W-2 test never distinguishes a fraudulent operator from a legitimate one, it disqualifies both alike.
In its September 2024 report, Additional Oversight of Remote Patient Monitoring in Medicare Is Needed, the OIG gave CMS five recommendations. Not one of them says “ban the delivery model.” Every one of them maps to this framework:
1 · “Implement additional safeguards to ensure that remote patient monitoring is used and billed appropriately”
→ The Remote Care Organization standard: registration, required leadership, filed protocols, surety bonds, revocation
2 · “Require that RPM be ordered and that information about the ordering provider be included on claims”
→ The initiating-visit gate, every patient starts with a physician order, on the claim, in the record
3 · “Develop methods to identify what health data are being monitored”
→ Certified software with audit-grade visibility: every reading, alert, and intervention timestamped and auditable
4 · “Conduct provider education about billing of remote patient monitoring”
→ Filed clinical protocols and marketing-conduct standards, with education built into registration
5 · “Identify and monitor companies that bill for remote patient monitoring”
→ The heart of it: register and credential the companies. The OIG asked CMS to identify and monitor them, the rule skips oversight and goes straight to elimination.
Source: HHS OIG, OEI-02-23-00260 (September 2024)
Medicare has regulated its way out of exactly this problem three times: DMEPOS suppliers, IDTFs, and clinical labs (CLIA). Each time, the fraudulent exited and the legitimate industry kept serving every beneficiary. The same playbook, applied to all of remote care: RPM, RTM, CCM, PCM, BHI. The result: more accountability than the rule, roughly double the savings, and nobody disconnected.
A physician-ordered start for every new patient, cold-call fraud becomes structurally impossible, while grandfathering the million-plus already enrolled.
CMS credentialing for remote care organizations, like DMEPOS, IDTFs, and CLIA labs. Bad actors exit; visibility becomes total. No registration, no payment.
Outcomes reported with every claim, reviewed annually, waste can't hide, and CMS tightens patient eligibility with evidence, like it did for CCM.
What Medicare actually pays, by code and locality
What providers contract to partners, the real split between the provider's share and the partner's share
Software · security · development · FDA-cleared devices · cellular transmission · clinical labor, per patient, per month
What's left for each party, and payback periods per enrolled patient
The principle: valuations should be set on these numbers, and they must leave a workable margin for both the ordering provider and the specialist organization. A program that isn't economically viable for either party delivers care to no one. There is room for a measured re-valuation; there is no room for $35/month.
Specific enough to draft from. Adopt these six lines and CMS gets the accountability, roughly double the savings, and a program nobody has to lose:
Finalize the initiating-visit requirement prospectively. Exempt patients enrolled before the effective date; authorize telehealth or a consent-portal-of-record for initiation when clinically indicated.
Withdraw the employment restriction. In its place, require enrollment of Remote Care Organizations under a new standard: CMS registration and screening, a physician medical director and designated quality manager, filed clinical protocols, certified software with audit-grade visibility, a surety bond, and revocation authority.
Condition payment on annual outcomes reporting with claims, adopting the ACCESS model's claims-as-attestation architecture for all registered organizations. Use the annual outcomes record to refine patient-selection criteria over time, covering the patients the data proves benefit, exactly as CMS tightened CCM eligibility.
Hold current valuations and open a formal cost-data collection. The industry will populate it, reimbursement, contracted fees, provider/partner splits, delivery costs, payback periods, and CMS re-values on the record in CY2028.
Decline the flat G-code consolidation. Retain acuity-based increments pending the transition plan, so the sickest patients stay funded.
Extend the standard to CCM, PCM, and BHI, and announce the transition workplan: per-minute thresholds retire as AI proves outcome parity under supervision; legacy codes sunset into ACCESS on a 3–5 year schedule.
CMS wants care programs that get cheaper as they scale, not more expensive. This framework builds that in: inside registered organizations, voice-AI clinical support runs beside human teams, and outcomes-with-claims reporting compares the two cohorts in real time. Supervision winds down only when the evidence says it can, and cost per patient falls as autonomy is earned.
Share of AI-supported monitoring under direct human supervision, each step down is an outcome gate, not a date
AI-supported cohorts must meet or beat human-cohort benchmarks in the same outcomes-with-claims reporting that governs every registered organization. Where AI doesn't prove it, supervision stays. Autonomy is granted by CMS's own evidence, and revocable.
Cost per monitored patient falls as autonomy is earned, that is what turns the glide path downward and makes ACCESS economics work at national scale, with the evidence trail built in.
Time-threshold billing pays for human minutes rather than results, structurally preventing the least-cost path to the same outcome. Retire the thresholds; keep the accountability.
The credentialing standard makes the comparison auditable: same platforms, timestamped interventions, parallel cohort reporting. Nothing is taken on faith.
The rule targets RPM and RTM. But CCM is the family's biggest spend at just ~4% penetration, the next RPM-sized curve, already compounding, and untouched by the rule. Regulate the whole remote care family at once and the trajectory doesn't just flatten, it declines: the first genuinely deflationary care framework in Medicare.
$ billions per year, RPM + RTM + CCM + PCM + BHI combined · RPM actual (OIG); other programs industry-estimated, blended growth ≈ +22%/yr · framework = measured re-valuation + initiating-visit growth gate in 2027, then decline via AI-earned efficiency into ACCESS
Illustrative projection. 2024 family base ≈ $1.8B: RPM $536M (OIG actual, +31%/yr documented); CCM ≈ $1.0B, RTM ≈ $150M, PCM + BHI ≈ $150M (industry estimates from enrollment × published rates, we will substantiate with claims data in the working session). Blended growth ≈ +22%/yr. Framework: 2027 holds near $2.5B (measured re-valuation + growth gate across the family), declining toward ≈$1.65B by 2031 via AI-earned efficiency and ACCESS migration. All assumptions open to CMS actuaries, we will run this jointly.
Estimated 2026 Medicare spend by program, $ billions · the rule regulates only the two smaller curves
Leaving CCM, PCM, and BHI outside the framework doesn't just shrink the win, it hands the same integrity problems a bigger, faster-growing place to live. One standard, one gate, one glide path, the whole family.
Independent analysis by Optum (UnitedHealth Group), May 2026, propensity-score-matched, difference-in-difference, 12 months of claims, monitoring reimbursement excluded from the savings math. 2,800+ monitored Medicare patients vs. 13,000+ matched controls.
| Cohort | Medical cost savings, per member per year | Sample |
|---|---|---|
| Hypertension with complications | $2,467 | n = 1,001 vs. 4,864 matched controls |
| Diabetes | $3,012 | n = 703 vs. 3,438 |
| Chronic kidney disease | $2,483 | n = 1,112 vs. 5,418 |
Results are directional actuarial findings pending larger samples. Full methodology open to CMS actuaries, that review is part of our ask.
average systolic blood-pressure reduction across monitored hypertension patients, up to −29.6 mmHg in Stage 2 hypertension
National monitoring program data, 2026glucose reduction in high-risk diabetes patients (>175 mg/dL cohort) by 12–18 months of monitoring
National monitoring program data, 2026reduction in critical weight alerts among monitored heart-failure patients over six months, decompensation caught before the ER
National monitoring program data, 2026reviewed by licensed nurses in six months in a single national program, 246,785 critical alerts resolved, 41,000+ per month
Jan–Jun 2026Published literature shows reduced heart-failure hospitalizations and lower total cost of care for remotely monitored populations.
University of Michigan Health reported $12M in avoided hospitalizations across 1,700+ remotely monitored patients.
Sustained blood-pressure reductions of this magnitude translate directly into fewer strokes and heart attacks across a monitored population.
At the ~$94K median RN salary (BLS 2025), more monitored patients than most practices have, and 24/7 alert coverage no single hire can provide.
Providence, LifePoint, Community Health Systems, Sentara, UnityPoint, and Mount Sinai all publicly contract specialized monitoring partners, and their care teams are centralized, not employed by each billing practitioner.
RPM software and device providers estimate roughly 80% of the market runs through third-party partners, because most practices cannot operate these programs themselves. Vendors can attest, we'll help assemble that record.
Nsight Health, on behalf of the industry, will bring the Optum actuarial study with its methodology open to CMS actuaries · industry-wide outcomes and utilization data · complete unit economics from our programs and our competitors' · our clinical protocols as a drafting template · and the joint spend model above.
We are asking to help write standards we will then be held to.
Nsight Health · nsightcare.com · Campaign & public evidence: protectrpm.org
Sources: Federal Register 2026-14327; HHS OIG OEI-02-23-00260 (Sept 2024) & 2025 data snapshot; Optum Actuarial Study (May 2026); 42 CFR 410.33 (IDTF); DMEPOS supplier standards (BBA 1997/MIPPA); CLIA (1988); CMS ACCESS Model (2026); BLS OEWS (2025); Mathematica (CCM penetration); public health-system RPM partnership announcements.