Is RPM a Good Investment? An Evaluation Guide

A remote patient monitoring program is worth the investment when three things are true: you have established patients whose readings would change what a clinician does, you have staff time to review those readings and talk to patients every month, and your billing can show that work happened. When any of the three is missing, the program tends to become device logistics that neither improves care nor survives an audit.

This guide is a framework for making that call before you sign a vendor contract, and for judging the program once it is running. It does not include revenue projections. Payment varies by locality and changes every year, and the 2027 rules are still being written, so the only reliable model is one built with your own patients, staff costs, and current fee schedule rates.

Start with patients whose readings change decisions

The evidence points to the follow-up, not the device. A systematic review and individual patient data meta-analysis of blood pressure self-monitoring in PLoS Medicine found that self-monitoring alone was not associated with lower blood pressure or better control, while monitoring combined with co-interventions such as medication titration, education, or lifestyle counselling led to clinically significant reductions that lasted at least 12 months. A program earns its cost when someone acts on the numbers.

That gives you a practical test for each candidate group. Ask whether a clinician would adjust treatment, order something, or call the patient based on the readings. Uncontrolled hypertension under active titration, heart failure with weight and symptom monitoring, and diabetes on changing insulin doses usually pass. Stable patients whose readings would be filed and never acted on usually do not. Medicare also sets eligibility limits: the CMS remote monitoring booklet (MLN901705) requires an established patient relationship for RPM, an acute or chronic condition, and monitoring that is reasonable and necessary.

Then size the group honestly. Count established patients who meet the clinical test, subtract those who are unlikely to use a device at home most days, and treat that number as your ceiling rather than your target.

The work each enrolled patient creates

RPM has three parts, and each needs an owner:

• Setup and education: getting the device to the patient, pairing it, and teaching them to take a reliable reading. Plan for repeat visits, replacement devices, and caregivers who need the same training.
• Device supply and data: tracking which patients transmitted and on how many days, and chasing those who went quiet. The data must upload automatically from a device that meets the FDA medical device definition.
• Treatment management: reviewing readings, deciding what is out of range for this patient, talking with the patient or caregiver live at least once a month, escalating to the prescriber, and documenting minutes and outcomes.

Treatment management is where staffing gets underestimated. Someone has to own the review queue every working day, know when to escalate, and reach the patient by phone or video. Write down who reviews readings, how quickly, who the escalation goes to when the ordering clinician is out, and what happens to a reading that arrives on a Friday evening. If those answers are vague, the program will be too. Our post on reducing alert fatigue covers how to set thresholds the team can sustain.

Costs to put in your model

List each cost line with your own numbers before you compare it to expected payment:

• Devices and connectivity: purchase or lease price, cellular or hub fees, shipping, replacements, and devices that never come back
• Software: platform subscription, per-patient fees, and EHR integration work
• Staff time: setup, daily review, monthly patient contact, escalations, documentation, and chasing low transmitters, at loaded hourly cost
• Clinician time: reviewing escalations and changing treatment plans
• Billing: claim build, eligibility checks, denials, and appeals
• Compliance: consent workflow, documentation audits, and policy updates each January
• Patient cost sharing: patients may owe Medicare cost sharing each month, and those who decline or drop out because of it affect enrollment and retention

Then model revenue per patient month using your locality's current rates and realistic assumptions about how many enrolled patients will meet the data-day thresholds and receive a monthly interactive communication. Run it at several adherence levels. If the program only works when nearly every patient transmits every month, it does not work. Our 2026 Medicare RPM billing guide explains the codes, day counts, and time thresholds you will need for the model.

Device choice affects several of these lines at once. Connectivity, ease of use for older patients, and replacement terms all matter; see how to choose remote monitoring devices for the tradeoffs.

Why RPM programs fail

The clearest public data on failure comes from the HHS Office of Inspector General. Its 2024 report on Medicare remote patient monitoring found that about 43 percent of enrollees who received RPM did not receive all three components (setup and education, device supply, and treatment management), and that Medicare lacked key oversight information such as who ordered the monitoring. A 2025 follow-up reported Medicare RPM payments above $500 million in 2024 and described billing patterns that warrant scrutiny, including billing for a high share of patients with no prior history with the practice and billing for multiple monitoring devices for one patient in a month.

Inside practices, the same patterns show up in more ordinary forms:

• Enrollment driven by device shipments rather than by a clinician's decision that monitoring will change care
• Readings that land in a portal no one checks daily, with no named owner for escalations
• Patients who stop transmitting after the first month, with no outreach plan
• Treatment management billed from time logs that do not show a live conversation or any resulting decision
• A monitoring vendor whose staff and notes sit outside the practice record, so the billing clinician cannot show oversight
• No plan for the annual fee schedule change, so the model quietly stops working

The 2027 proposal is a planning risk

In July 2026, CMS published the CY2027 Physician Fee Schedule proposed rule. If finalized as proposed, it would require a separately reportable face-to-face initiating visit at the start of RPM, pay for RPM only when clinical staff time comes from direct employees of the practitioner or practice rather than third-party contractors, revalue the device codes, and possibly replace the current codes with four new HCPCS G-codes. CMS tied the proposals to the two OIG reports above.

None of that is final. But a program that depends on contracted monitoring staff, or that enrolls patients without a visit where the clinician discusses monitoring, should be modeled both ways. We look at the staffing question in more detail in our analysis of the proposed rule.

A pre-launch checklist

Before committing budget, you should be able to answer each of these in writing:

• Which conditions and which established patients qualify, and roughly how many are there
• What decision each reading type is meant to support, and the threshold that triggers action
• Who reviews readings each day, who escalates, and who covers absences
• How staff will reach every enrolled patient live at least once a month, and how that is documented
• How consent is obtained and recorded, including the cost-sharing conversation
• How data days and treatment management minutes are counted, and by whom
• How the RPM record connects to the EHR so the clinician sees readings next to the rest of the chart
• What the program costs at low, medium, and high adherence, using this year's rates
• What changes if the 2027 proposals are finalized
• What would make you pause or stop the program

Start small. A pilot with one condition, one clinician group, and one care team member who owns the queue will teach you more than a broad launch.

How to judge it after six months

Judge the program on whether the work happened and whether it changed care. Revenue follows from those, and on its own it does not tell you the program is sound. Track:

• Share of enrolled patients meeting each data-day threshold each period
• Share receiving a documented live interactive communication each month
• Time from an out-of-range reading to review, and from review to clinician action
• Treatment changes made because of monitoring data
• Clinical measures for the target condition, such as blood pressure control in a hypertension cohort
• Disenrollment and its reasons
• Denial rate and the reason codes behind it
• Staff hours per enrolled patient per month, compared with your model

If the clinical and process numbers are strong and the cost per patient is close to plan, the program is doing what it should. If enrollment is high but interactive contact and treatment changes are low, fix the workflow before adding patients. The complete RPM workflow post lists what the record should capture at each step.

Hoss Care runs remote monitoring inside the same platform as video visits, follow-ups, documentation, and billing, so the readings, the staff time, and the clinician's decisions live in one record.