The remote patient monitoring (RPM) sector is entering a pivotal phase in 2025, marked by a decisive shift from pilot programs to scaled, reimbursable clinical workflows. While the pandemic-era surge in at-home monitoring has stabilized, the industry is now grappling with a new set of pressures: Medicare payment recalibration, the integration of generative AI into alert management, and the expansion of RPM into chronic disease prevention rather than just post-acute follow-up. This week’s developments underscore a maturing market that is no longer about hardware novelty but about demonstrable outcomes and interoperability.

CMS Proposed Rule Signals a Tighter but More Flexible Reimbursement Landscape

The most consequential news this quarter comes from the Centers for Medicare & Medicaid Services (CMS), which released its proposed physician fee schedule for 2026, including significant changes to RPM billing codes (99453, 99454, 99457, 99458). The proposal, currently in a 60-day comment period, suggests consolidating the “initial setup and education” code (99453) with the “device supply” code (99454) into a single bundled payment. Industry analysts estimate this could reduce per-patient revenue by approximately 12-15% for vendors using high-cost cellular-enabled devices, but it may also lower administrative burdens that have historically deterred small practices from launching RPM programs.

More notably, CMS is floating a new add-on code for “asynchronous remote evaluation of patient-generated health data with AI-assisted clinical decision support.” This would be the first time Medicare explicitly recognizes algorithmic triage as a billable service, provided a licensed clinician reviews the AI’s output within 24 hours. Dr. Elena Vasquez, a health policy fellow at the Brookings Institution, commented: “This is a double-edged sword. It legitimizes AI in the reimbursement pathway, but it also puts the onus on vendors to prove their algorithms are not introducing bias. The ‘black box’ problem just became a billing problem.”

The Virtual Ward Boom: From COVID Stopgap to Standard of Care

Across the Atlantic, England’s National Health Service (NHS) announced a five-year contract extension for its virtual ward program, which currently manages over 24,000 patients daily at home. The new framework emphasizes “hospital-at-home” for frailty, heart failure, and COPD, but with a critical upgrade: integration with local electronic health records (EHRs) via the FHIR standard. Previously, many virtual wards operated on proprietary dashboards that required manual data entry by community nurses. The new mandate requires bidirectional data flow—meaning a patient’s home blood pressure readings automatically populate their hospital record, and medication changes push back to the home monitoring device.

This interoperability push is a direct response to a scathing review from the Care Quality Commission, which found that 38% of virtual ward discharges had documentation gaps. “The technology was never the problem,” said NHS Digital’s interim chief clinical information officer, Dr. Alistair Finch. “It was the handoff. Now we are treating RPM not as a separate silo but as a vital sign that travels with the patient. The next step is enabling primary care physicians to initiate RPM without a hospital referral—that’s the true test of sustainability.”

AI Alert Fatigue: The Hidden Cost of Too Much Data

While reimbursement and interoperability dominate policy discussions, a quieter but equally critical trend is emerging in clinical operations: alert fatigue caused by high-frequency RPM devices. A study presented at the American Telemedicine Association’s annual meeting this month analyzed 1.2 million RPM alerts from 14,000 patients across three large health systems. It found that 61% of alerts were false positives or clinically insignificant, yet each required a nurse to acknowledge and document them. The average nurse in a large RPM program now spends 3.2 hours per shift on alert review—up from 1.8 hours in 2023.

Vendors are responding with “smart suppression” algorithms that learn patient baselines. For example, a blood pressure reading of 150/90 might trigger an alert for a patient with a baseline of 120/80, but not for a patient whose baseline is 145/85. However, clinicians remain wary. “We cannot allow the algorithm to decide what is urgent without a human in the loop,” argued Dr. Marcus Chen, medical director of virtual care at Providence Health. “The FDA has cleared some of these algorithms for ‘decision support,’ but that does not mean they are ready for autonomous alert suppression. We are seeing a push toward ‘AI-first, human-second’ that scares me.”

The Consumerization of RPM: Wearables vs. Medical-Grade Devices

One of the most contentious debates in the industry is whether consumer wearables (Apple Watch, Smart Scales, Oura Ring) should be eligible for RPM reimbursement. Currently, CMS requires that devices be “medical grade” and meet FDA clearance for the specific measured parameter. But a coalition of 14 digital health companies, including several major tech firms, filed a petition with CMS this week arguing that modern wearables with photoplethysmography (PPG) and electrocardiogram (ECG) capabilities provide clinically equivalent data for conditions like atrial fibrillation and sleep apnea.

The petition cites a recent peer-reviewed study inJAMA Cardiologyshowing that Apple Watch ECG readings had a 94% sensitivity for detecting AFib compared to standard 12-lead ECGs. However, the American College of Cardiology (ACC) has formally opposed the petition, citing concerns about data quality in low-resource patients and the lack of standardized validation protocols. “A consumer device is designed for lifestyle, not for life-threatening events,” said ACC president Dr. Patricia O’Malley. “We are not Luddites. We just think the bar for medical-grade should remain high, especially when we are using the data to adjust anticoagulation dosing.”

The Global South: A Leapfrog Opportunity with New Risks

Outside the U.S. and Europe, RPM is expanding rapidly in low- and middle-income countries, but with a different business model. In India, the government’s Ayushman Bharat Digital Mission has partnered with four private RPM vendors to provide blood pressure and glucose monitors to 1 million hypertensive patients in rural areas. The twist: devices are not owned by patients but by community health workers (ASHAs) who conduct weekly home visits. This “hub-and-spoke” model reduces device loss and improves data quality, but it also raises questions about data ownership. Who owns the patient’s continuous glucose data—the patient, the community worker, or the government?

Dr. Amara Okafor, a global health informatics researcher at the University of Lagos, notes that RPM in emerging markets is often driven by donor funding, which can evaporate. “We are seeing a race to deploy devices without a clear sustainability plan. In Nigeria, we have 40,000 RPM devices in the field, but only 12% are actively transmitting data after six months. The batteries die, the SIM cards expire, and the community health worker moves on. The real innovation needed is not a better sensor; it is a better service contract.”

What’s Next: The 2025-2026 Outlook

Looking ahead, three trends are likely to dominate the RPM news cycle. First, the final CMS rule on AI-assisted remote evaluation, expected in November, will set the precedent for how algorithms are audited and paid. Second, the FDA is expected to release draft guidance on “software as a medical device” for RPM that includes continuous learning algorithms—a controversial move that would allow devices to update their own thresholds without re-submission. Third, the hospital-at-home waiver, which allows acute-level care at home with RPM, is set to expire at the end of 2025 unless Congress acts. A bipartisan bill, the “Hospital at Home Act,” has been introduced, but its chances remain uncertain given budget constraints.

For providers, the message is clear: RPM is no longer a novelty. It is a core component of chronic care management, but it demands rigorous clinical governance. As Dr. Vasquez put it, “The technology has matured. Now the question is whether our payment models, our regulatory frameworks, and our clinical workflows can mature just as fast. If they do, RPM will be the backbone of a truly proactive health system. If they don’t, we will just be collecting more data that no one uses.” The next 18 months will determine which path the industry takes.