Blog

The CY 2027 Physician Fee Schedule Would End Remote Monitoring.

The Evidence Doesn’t Support Doing So.

An analysis for remote monitoring organizations, physician practices, and the therapy providers who depend on RPM and RTM. Comments on the proposed rule are due September 14, 2026.

The CY 2027 Medicare Physician Fee Schedule proposed rule, published in the Federal Register on July 16, 2026, contains the most consequential remote monitoring proposals since the code set was created in 2019. Anyone who furnishes Remote Patient Monitoring (RPM) or Remote Therapeutic Monitoring (RTM), and any practice that relies on monitoring to manage chronic disease, should read Section (48) closely. Taken together, the proposals would raise the cost of starting monitoring, cut the payment for delivering it, and outlaw the staffing model that makes it feasible for small and rural practices.

Five changes stand out:

1. A mandate that clinical staff furnishing RPM/RTM be direct employees of the billing practice — a de facto ban on third-party monitoring support.
2. A separately billed, face-to-face initiating visit before monitoring can begin.
3. An established-patient requirement extended to RTM for the first time.
4. Deep payment cuts achieved by crosswalking RPM codes to self-measured blood-pressure codes.
5. A solicitation for comment on collapsing seventeen remote-monitoring codes into four all-or-nothing bundled G-codes.

These are justified by citation of two HHS Office of Inspector General (OIG) reports — reports that did not review medical records, quantified no improper payments, and flagged outlier billing at fewer than 1.2 percent of routine RPM billers on every measure. The central 43 percent statistic did not match the claim that improper care was delivered and drew a retraction demand from the Alliance for Connected Care. The proposals are simply not supported by the data they reference. 

Proposed pricing changes often lack sufficient evidence

On valuation, the Proposed Rule states that “due to lack of information regarding the typical device used to perform these procedures, these services are overvalued… we have received very little invoice or pricing information from interested parties for the specific devices used in RTM and RPM services.”

Without device pricing data, the agency cannot know that devices are overvalued. Proposed changes should be based on data, and we request the opportunity for multiple RPM providers and data suppliers to supply that information. 

The direct-employment mandate contradicts CMS’s own regulation

The proposal would allow RPM/RTM payment “only… when furnished by clinical staff employed by the practice,” expressly excluding contracted staff. This contradicts the text of the “incident to” regulation CMS itself cites. Under 42 C.F.R. § 410.26(a)(1), auxiliary personnel are defined as “any individual who is acting under the supervision of a physician (or other practitioner), regardless of whether the individual is an employee, leased employee, or independent contractor of the physician.”

The regulation defines “independent contractor” and “leased employment” precisely so that supervised third-party staff can furnish incident-to services. CMS reaffirmed in the CY 2021 PFS that auxiliary personnel furnishing RPM may be employees of a third-party entity under contract with the billing practitioner, and it has permitted the identical model in Chronic Care Management for a decade. The proposed rule offers no evidence that employment status affects care quality, oversight, or fraud incidence. The regulation’s existing safeguard is supervision: the billing practitioner orders the service, directs and controls it, reviews the data, and certifies the claim. A W-2 requirement changes who signs the nurse’s paycheck, not who supervises the nurse.

The data CMS relies on actually exonerates the vendor model. OIG’s 2024 report found roughly one in ten RPM enrollees — on the order of 100,000 patients — received services involving one of 41 identified RPM companies. If third-party staffing were the fraud vector the rule implies, the outlier measures would show it. Instead, every measure flagged fewer than 1.2 percent of routine billers, and nearly 99 percent of practices behaved normally on every dimension during the very years the vendor model scaled. The “cold calling” consumer alert CMS invokes describes criminal actors impersonating providers — conduct already illegal, and conduct an employment mandate does nothing to stop, because fraudsters do not follow staffing regulations.

The access arithmetic is stark. Median physician-practice staffing is about 2.19 clinical FTEs per provider, support-staff turnover runs 12 to 26 percent annually, recruiting a single clinical hire takes four or more months, and the country faces a projected physician shortage of up to 86,000 by 2036 and an RN shortfall exceeding 60,000 by 2030. A practice with two clinical FTEs per physician cannot staff continuous device monitoring, triage, and monthly interactive outreach for a 70-patient RPM panel — the median panel in OIG’s own data — without either a vendor or a new hire it would be unlikely to find. The mandate functions as a wealth test: large systems internalize monitoring; small and rural practices exit. Roughly 100,000 current patients would need their monitoring re-homed or discontinued on January 1, 2027. This is a forced disruption of ongoing chronic-disease care, concentrated in the practices least able to absorb it.

The initiating visit: a redundant requirement

RPM already requires an established patient relationship — a face-to-face or telehealth encounter within the look-back window — plus consent at the time services are furnished. The proposal adds a second, separately billed gate: a dedicated visit at which RPM must have been discussed, with the explicit caveat that an otherwise-qualifying recent visit “cannot count” if RPM was not raised.

Three consequences follow. First, cost. With roughly one million Medicare enrollees starting or receiving RPM annually, requiring even one additional E/M visit (roughly $75 to $115 at 2026 rates) for a substantial fraction of initiations adds tens of millions of dollars in new Medicare spending, plus 20 percent beneficiary coinsurance — an odd companion to a rule whose valuation section is premised on cost concern.

Second, delay. The clinical windows where RPM earns its keep are short: post-discharge heart-failure monitoring must start within days, and published programs report 30-day readmission reductions of 50 percent and admission reductions up to 65 percent. A scheduling queue for a formality visit sits directly in that window. Third, audit exposure. The “was RPM discussed” test converts a chart-documentation nuance into a claims-denial and False Claims Act hook, inviting exactly the pattern-based enforcement the OIG snapshots foreshadow.

The established-patient requirement already guarantees the practitioner has seen the patient. Requiring that the visit include a scripted topic adds no clinical safeguard that an order requirement would not add more cheaply. CMS could simply require a documented practitioner order — which the OIG itself recommended — at zero patient cost and zero delay.

Established-patient for RTM: importing a rule the code set was built to avoid

CMS created RTM in CY 2022 without an established-patient requirement, deliberately. RTM is ordered predominantly by physical and occupational therapists and other qualified health professionals receiving referrals, and a referred patient beginning a therapy episode is, almost by definition, new to the ordering therapist. Musculoskeletal RTM episodes track short post-operative and rehabilitation windows measured in weeks. Layering an established-relationship requirement plus the proposed initiating visit onto a six-to-twelve-week therapy episode consumes the front of the episode with administrative gating.

The rule cites no RTM-specific integrity finding. The OIG reports it references studied RPM, not RTM. Extending an RPM-derived restriction to a different benefit, on the basis of reports about the first benefit, is not evidence-based. 

The valuation cuts: crosswalking device economics to codes with no device

The proposed crosswalks re-price RPM around CPT 99473 and 99474 — the self-measured blood-pressure codes, which involve a patient-owned cuff, no supplied device, no cellular connectivity, no logistics, and no continuous monitoring platform. The arithmetic is plain:

Code	2026 national rate (approx.)	Proposed treatment	Directional effect
99453 setup/education	$22	Practice-expense crosswalk to 99473 (in-office cuff training)	Cut; ignores shipping, provisioning, remote training
99454 device supply (16–30 days)	$47	Crosswalk to 99474 (total rate ≈ $26)	Cut plausibly ~50%+ on a PE-only code
99445 device supply (2–15 days)	≈ 99454 rate	Same 99474 crosswalk	Same cut, one year after the code was created
99457 management, 20 min	$52	Eliminate all PE inputs; keep 0.61 work RVU (≈ $20)	Cut plausibly ~40–60%
GRPM2 bundle (if adopted)	—	99457 work + 99474 PE, replacing 99454+99457+99458	A full 40-min month drops from ~$140–185 to ~$50 or less

A cellular-enabled blood-pressure cuff or weight scale, its connectivity, fulfillment, replacement logistics, and the monitoring platform behind it must all be funded from the device-supply payment. 99454 is a practice-expense-only code with no physician work component to cushion a cut. Re-pricing it against a code whose “device” is the patient’s own cuff guarantees under-valuation by construction. CMS’s stated basis is that it lacks pricing data. The remedy for missing data is a data call, which the Proposed Rule makes. If reimbursement falls below the fully loaded cost of supplying and monitoring the device, the supply of monitoring does not become cheaper. It becomes zero, practice by practice, starting with the smallest.

The treatment-management proposal contains the rule’s starkest internal contradiction. CMS proposes to strip practice-expense inputs from 99457, 99458, and 99470 because “we do not believe the typical clinical workflow for these services would involve clinical staff time.” Yet the CPT descriptor for these codes begins: “Remote physiologic monitoring treatment management services, clinical staff/physician/other qualified health care professional time…” Clinical staff are not incidental to the code; they are the first noun in it. And one subsection earlier, the same rule proposes an employment mandate governing that same clinical staff. Both positions cannot be true. If clinical staff do not typically furnish RPM management, the employment mandate regulates a workflow CMS believes does not exist. If they do, the valuation predicate is false. Commenters should force CMS to choose.

The bundled G-codes: rebuilding a cliff CMS just tore down

The proposed GRPM2 would pay a single monthly rate only when the practice delivers, in the same calendar month: device supply, at least 2 days of data transmission, at least 20 minutes of treatment management, and at least one real-time interactive communication. Miss any element — most commonly the 20 interactive minutes, which depend on the patient answering the phone — and payment for the entire month, including the device the practice already bought, shipped, and monitored, is zero.

This is an engagement problem escalated from partial to total. Under current codes, a practice that cannot reach a patient still recovers device costs through 99454. Under GRPM2 it recovers nothing. Real-world adherence data show how often that will happen: in a published Medicaid diabetes cohort, roughly half of monitored patients fell below adherence thresholds despite active outreach. The response of a margin-constrained practice will likely be to disenroll low-engagement patients — disproportionately the rural, low-income, dual-eligible, and minority beneficiaries whom OIG’s own 2024 report shows RPM currently reaches at roughly double the rate of white beneficiaries. A payment design that makes revenue contingent on patient behavior selects against the hardest-to-serve patients. It also creates the opposite integrity risk: a 20-minute, all-or-nothing threshold manufactures pressure to inflate interactive time, the exact falsity pattern enforcement should want to avoid incentivizing.

The timing compounds the error. CMS finalized codes 99445 and 99470 in the CY 2026 final rule — published November 2025, effective January 2026 — precisely to pay for legitimate below-threshold service months. The bundling concept would repeal that architecture barely one year later, before a single full year of utilization data on the new codes exists, and would eliminate 99458 add-ons entirely, removing any payment gradient for high-acuity patients needing 40 to 60 minutes of monthly management. 

The stated rationale recycles the OIG 43-percent statistic, a figure computed on 2022 claims under the old code structure, rebutted by CMS’s own CY 2024 clarification that the components are separately billable by design, and subject to a standing retraction request. The rule even concedes the point mid-argument: “Although we have not required that providers bill for all three components, this data from the OIG Report raises questions.” A statistic that measures compliance with the code architecture cannot simultaneously justify demolishing it.

The measurable evidence the Proposed Rule does not engage

Scale. RPM is $536 million of a $1.1 trillion program — 0.05 percent — reaching under 1.5 percent of enrollees while more than 60 percent have hypertension. The policy problem visible in the data is under-use, not overuse.

Integrity. OIG’s 2025 snapshot flags 0.4 to 1.2 percent of routine billers on every outlier measure; roughly 99 percent of practices are normal on every dimension. There was no medical record review, no improper-payment estimate, and an explicit disclaimer of any fraud finding.

Outcomes. Tang et al. (Annals of Internal Medicine, 2023, cited by OIG itself) and subsequent cohorts report uncontrolled hypertension falling from 66.3 to 40.2 percent after 90-plus days of RPM, systolic reductions of 7.3 to 16.7 mmHg, and heart-failure programs reporting 50 percent readmission and up to 65 percent admission reductions with positive net savings.

Equity. Black and Hispanic beneficiaries and dual-eligibles receive RPM at roughly double the rate of white beneficiaries. Every proposal in Section (48) either raises fixed costs or conditions payment on engagement — both of which fall hardest on those populations.

Congressional direction. The RPM Access Act advanced through committee markup on July 14, 2026, two days before this rule published. CMS is proposing to constrict a benefit Congress is actively moving to expand.

Where CMS makes strong points

Not every concern in the rule is misplaced, and the industry’s comments will be stronger for acknowledging it. Device-pricing transparency is a fair ask. Monitoring organizations and practices should answer the invoice-data solicitation with real, auditable cost data covering hardware, connectivity, software, logistics, and replacement, because silence invites exactly the blind crosswalk proposed here. A practitioner-order requirement and an ordering-provider identifier on claims are cheap, targeted integrity measures worth supporting, as is enforcement against the cold-calling schemes described in OIG’s consumer alert. Even the impulse to simplify seventeen codes is understandable — but the right simplification is consolidation that preserves component payment, such as merging duplicative set-up codes, not an all-or-nothing bundle. 

What you can do before September 14

Comments on the CY 2027 proposed rule are due September 14, 2026. This is the moment for remote monitoring organizations, physician practices, therapy providers, and the patients they serve to be heard. The most persuasive comments will pair the rule’s own admissions — “very little invoice or pricing information,” uncertainty over “how often third-party billing currently occurs” — with concrete, auditable data: real device cost stacks, actual patient-month payment scenarios under GRPM2, staffing and rural-shortage figures, and the outcomes and equity evidence the rule leaves unaddressed. Filing through established coalitions such as the Alliance for Connected Care and ATA Action, and aligning with physician, cardiology, and rural-health organizations, amplifies individual voices into a record CMS cannot ignore.

The stakes are not abstract. Under these proposals, the predictable result is not less fraud. It is less monitoring, concentrated among exactly the patients who benefit most.

---

This analysis reflects a review of the CY 2027 Medicare PFS proposed rule and the public evidence base on remote monitoring, prepared with input from the ChartSpan policy team. It is provided for informational purposes and does not constitute legal advice.

Primary sources: CY 2027 Medicare PFS Proposed Rule, Section (48) Remote Monitoring (Federal Register, July 16, 2026); 42 C.F.R. § 410.26(a)(1); CY 2021, CY 2024, and CY 2026 PFS Final Rules; HHS-OIG reports OEI-02-23-00260 and OEI-02-23-00261; Tang et al., Annals of Internal Medicine (2023); and published RPM/RTM outcomes, staffing, and equity literature cited throughout.