Remote patient monitoring has grown into a meaningful Medicare program. It has also become difficult to oversee.
In its 2025 report on Medicare remote patient monitoring, the HHS Office of Inspector General found that nearly 1 million Medicare enrollees received RPM in 2024. Medicare payments reached $536 million that year.
Growth creates an opportunity to help more people manage chronic conditions between visits. It also makes weak billing, poor clinical integration, and patient recruitment that begins with a cold call rather than a care relationship more consequential.
That is the context for CMS’s CY 2027 Physician Fee Schedule proposed rule, published on July 16, 2026. The rule is proposed, not final. The relevant changes would take effect on January 1, 2027 only if CMS finalizes them.
The proposal would make Medicare payment for RPM and RTM dependent on clinical staff being direct employees of the billing practitioner or practice. CMS would also require a separately reportable initiating visit, extend the established-patient requirement to RTM, revalue several services, and seek comments on replacing the existing code structure with four bundled G-codes.
The Federal Register text lays out the proposed changes in detail.
This is a proposed condition for Medicare payment. It is not a general ban on remote-monitoring software, devices, data infrastructure, or technology vendors that support a provider’s workflows. A practice could still use outside technology.
The policy question is whether the clinical staff performing the billable monitoring should belong to the provider responsible for the patient.
Many healthcare organizations and technology groups oppose the direct-employment requirement. They argue that it could disrupt legitimate programs, especially in rural areas and smaller practices that do not have enough internal staff to review incoming data and respond to patients.
Those concerns deserve a serious answer.
The answer should still begin with a principle that has been missing from too much remote-monitoring growth: the provider responsible for the patient should remain responsible for the monitoring.
CMS is responding to a real program-integrity problem
The case for reform does not depend on assuming that every remote-monitoring vendor is exploiting Medicare. The 2024 OIG review did not make that claim.
It found oversight weaknesses that make improper billing and poor-quality service harder to identify.
About 43 percent of enrollees in the OIG review did not receive all three components required for RPM. OIG also found that Medicare claims did not contain key information about who ordered monitoring. Its recommendations included identifying the ordering practitioner on claims, developing ways to determine what health data are being monitored, educating providers about billing requirements, and identifying companies that bill for RPM.
CMS has also cited concerns about third parties cold-calling beneficiaries, enrolling people who may not have an established relationship with the practice, and billing for services without enough evidence of meaningful clinical integration or practitioner oversight. These are exactly the kinds of problems that a payment program should be designed to prevent.
The $536 million figure does not prove that all spending outside a provider’s direct employment structure is wasteful. The OIG’s 2025 analysis itself notes that its billing measures do not establish that a particular practice committed fraud or abuse. It does show why CMS cannot treat rapid expansion as a reason to relax the connection between a claim, a patient, an ordering practitioner, and an actual clinical service.
CMS is right to ask a basic question: who is monitoring this patient, and where is the accountable care team?
The opposition is about access as well as business models
The response from the healthcare industry has been broad. ATA Action, the American Hospital Association, Fairview Health Services, and a 229-organization stakeholder coalition have urged CMS to reconsider the proposal or replace it with other safeguards.
Some hospitals and practices use contracted clinical staff because they cannot recruit, train, and retain enough people to review data every day. A contractor can work under a practitioner’s protocols, follow an escalation pathway, document activity, and refer concerns back to the treating team.
If CMS requires every clinical worker involved in monitoring to become a direct employee, these organizations warn that programs could shrink or close.
Fairview Health Services makes an important point in its comment submitted to the CMS-2026-2377 docket, although its full submission is separate from the Federal Register notice. Employment status alone does not prove clinical integration.
A directly employed worker can still operate inside a weak process, while a contracted worker can operate under a strong one. Fairview recommends auditable requirements such as a practitioner order and plan of care, documentation in the billing practitioner’s medical record, written protocols, escalation pathways, and identification of the ordering practitioner on the claim.
The Paragon Health Institute supports CMS’s integrity rationale while warning that a blanket prohibition on contracted clinical staff could reduce access for smaller and rural practices.
These organizations are making forecasts and policy arguments, not reporting settled outcomes. CMS should test those forecasts carefully. A policy that closes legitimate rural programs without improving clinical accountability would be a poor result.
There is still a difference between acknowledging an access problem and defending a model in which an outside company owns the practical work of monitoring a Medicare patient. The care team should know the patient, order the service, define the plan, review the information, and make the clinical decision.
That responsibility should remain visible in the medical record and auditable by Medicare.
The clinical work should stay with the care team
Remote monitoring is often described as a technology service. In practice, the valuable part is the clinical relationship around the data.
A blood-pressure reading has limited value on its own. Its usefulness depends on why it was collected, what the practitioner is watching for, which threshold matters for that patient, who reviews the result, and what happens when the result requires attention. The same applies to weight, glucose, oxygen saturation, symptoms, medication response, and other information collected between visits.
The Agency for Healthcare Research and Quality’s Patient Safety Network guidance on remote patient monitoring emphasizes the operational conditions that make these programs safer: patient-specific parameters, defined staff responsibilities, training, response procedures, and escalation pathways. A systematic review of RPM research likewise describes a heterogeneous evidence base in which technology alone does not explain outcomes.
Follow-up, patient education, and the way a program is implemented matter.
That is why we support CMS’s direction. A provider-led model gives the public a clearer answer when a monitoring program fails. The clinical team cannot point to a vendor as the party that was really responsible for reviewing the patient.
The vendor cannot present automated collection as a substitute for clinical care.
CMS should still refine the rule before finalizing it. A narrow, clearly auditable pathway may be necessary for rural and safety-net providers that need outside staffing support. But any exception should preserve the same core requirements: the practitioner orders the service, establishes the plan, documents the work, defines escalation, and retains responsibility for the patient.
Contractors can support a provider’s operation. They should not become a parallel care team that sits between the patient and the practitioner.
AI and connected tools can reduce the work without taking over the care
Keeping monitoring with the provider does not mean asking clinicians to perform every repetitive task manually. It means giving care teams practical tools for the work around monitoring, so they have more time for the decisions that require clinical knowledge and a relationship with the patient.
We can already see how this model works across different parts of healthcare. The tools below handle documentation, digital-care operations, hospital follow-up, revenue-cycle work, and interoperability. The provider remains responsible for reviewing the patient and deciding what happens next.
1. Actuvi already operates the way CMS wants clinics to operate
Actuvi already operates within the provider-accountable model CMS is trying to encourage. It uses AI and automation to make remote care easier for patients and providers, while automating the workflows around assessments, communication, data collection, monitoring plans, thresholds, alerts, dashboards, reporting, and integrations.
Actuvi builds and launches digital-care programs for clinics, connects the necessary data sources, and gives providers the tools to deliver remote care and monitor and review patients through their own organization.
That is why Actuvi belongs at the center of this discussion. The clinic keeps the patient relationship, the monitoring responsibility, the review process, and the clinical decisions. Actuvi supplies the digital infrastructure that makes those responsibilities easier to manage.
It supports remote physiologic monitoring and remote therapeutic monitoring without creating a separate outside layer that takes ownership of the care. We see Actuvi’s model as evidence that CMS’s direction is workable: provider accountability and useful technology can operate together.
2. Ambient documentation can give clinicians time back
Abridge turns clinical conversations into draft documentation that clinicians can review and edit. Its platform also connects documentation with revenue-cycle workflows and clinical context. This matters in remote monitoring because the clinicians reviewing patient data also have to document encounters, update records, and close the loop with patients.
A scribe can reduce documentation work, but it does not decide what a reading means or what treatment a patient should receive.
A 2025 JAMA Network Open study of 263 physicians and advanced practice practitioners across six health systems found that reported burnout fell from 51.9 percent to 38.8 percent after 30 days of ambient AI scribe use. The study had no control group and relied on self-reported measures, so the result is encouraging rather than definitive.
Documentation support can still create more capacity for clinical work.
3. Operations automation can keep follow-up from falling through the cracks
Qventus focuses on hospital and health-system operations. Its AI Operational Assistants can help with discharge planning, phone calls, follow-up items, and EHR updates. That work sits outside monitoring itself, but it determines whether a provider can act on what monitoring reveals.
If a patient needs an appointment, a referral, a follow-up call, or an update to the record, operations automation can help the care team complete that work instead of leaving it in a queue.
The clinical team should decide which follow-up is appropriate. Qventus can support the operational steps that follow that decision without taking responsibility for the patient.
4. Revenue-cycle automation can remove work that does not belong in the monitoring queue
AKASA applies generative AI to healthcare revenue-cycle work, including authorization, documentation, coding, and claims processes. Those tasks are separate from clinical monitoring, but they consume the same organizational capacity providers need to run care programs.
Automating parts of the revenue cycle can help staff spend less time searching for records, checking status, and resolving routine administrative work.
That separation is valuable. A revenue-cycle platform can help a practice document and bill correctly without becoming the party that decides whether a patient’s condition has changed. Clinical work stays with the care team while administrative work becomes easier to manage.
5. Interoperability can connect the systems that care teams already use
Redox connects EHRs, digital-health applications, devices, and other healthcare systems. Its platform supports the exchange of information such as patient history, medications, scheduling, laboratory results, vitals, and orders.
For remote monitoring, that connection can determine whether patient data reaches the provider’s existing workflow or sits in another isolated dashboard.
Interoperability is not a substitute for clinical review. It is the infrastructure that helps the right information reach the right care team in a usable form. When data can move between systems, providers have a better chance of keeping the monitoring relationship inside their own clinical operation.
The FDA’s January 2026 clinical decision-support guidance points in the same direction. Software can support decision-making, but the clinical professional must be able to review the basis for the output and exercise independent judgment.
That is the right standard for remote-monitoring AI as well.
The better answer is accountable automation
CMS should finalize a rule that makes provider accountability impossible to hide. The agency should require a documented relationship between the patient, the ordering practitioner, the monitoring plan, the data being collected, the staff reviewing it, and the action taken when a threshold is crossed.
The agency should also give legitimate practices enough time to adapt and should study the impact on rural and safety-net providers. If CMS keeps a direct-employment requirement, it should explain how organizations with limited staffing can comply without withdrawing a useful service.
If it creates an exception for contracted support, that exception should be tied to auditable provider oversight rather than a loosely defined vendor arrangement.
The policy debate should move beyond a false choice between outsourced monitoring and no technology. There is a third path. Providers can keep the clinical work in-house while using accountable platforms and AI to remove the administrative friction around it.
That approach spends technology where it helps most: patient engagement, documentation, data organization, scheduling, reminders, reporting, and workflow coordination. It keeps diagnosis, treatment, escalation, and clinical judgment with the care team.
Remote monitoring needs a clearer chain of responsibility
CMS’s proposal is a signal that remote monitoring has reached a point where growth alone is no longer enough. The program must show that the money attached to a claim supports a real clinical service for a real patient under the oversight of a real care team.
The concerns raised by hospitals, physician groups, rural providers, and technology organizations should shape the final rule. Access matters. Staffing constraints matter.
A direct-employment test may need refinement.
The central direction should remain. Clinical monitoring belongs with the provider accountable for the patient. Technology vendors can help that team reach patients, collect information, manage workflows, and respond to the volume of data that modern care produces.
AI can make the model more practical. It should not make responsibility less clear.
The future of remote care will depend less on who owns the software than on whether every patient has a visible chain of responsibility from data collection to clinical action. CMS is right to demand that chain. The healthcare industry should build the tools that help providers maintain it.
