Remote Monitoring Reimbursement in 2026, and What to Do When Devices Disagree
Two questions come up in almost every conversation we have with clinics about wearable data. The first is whether any of it is billable. The second arrives about ninety seconds later: what a physician is supposed to do when a patient turns up wearing three devices that disagree with each other.
They are worth answering together, because the honest answer to the first shapes how you should think about the second.
The short answer on billing
Most consumer wearables do not qualify for remote physiologic monitoring reimbursement. CMS requires a medical-grade, FDA-cleared device that automatically collects and transmits physiological data, and devices requiring manual data entry by the patient do not meet the standard.
That rules out a substantial portion of what patients actually wear. It does not rule out everything, and the boundary is worth understanding rather than guessing at, since some connected devices in the blood pressure, weight and glucose categories do meet the requirement while the general-purpose activity tracker on the same patient’s wrist does not.
What changed for 2026
The 2026 Medicare Physician Fee Schedule was the most consequential rulemaking cycle for remote monitoring since the original codes were created.
| Code | Covers | Note for 2026 |
|---|---|---|
| 99453 | Initial device set-up and patient education | One-time, per episode of care |
| 99454 | Device supply with daily recording or transmission, 30 days | Requires 16 days of data in a 30-day period |
| 99445 | Device supply for shorter monitoring periods | New for 2026; covers 2 to 15 days of transmission in a 30-day period |
| 99457 | First 20 minutes of treatment management with interactive communication | Interactive communication is a documented requirement, not a formality |
| 99458 | Each additional 20 minutes of treatment management | Billed alongside 99457 |
The addition of 99445 is the most practically significant change, because the 16-day threshold under 99454 had made shorter monitoring episodes uneconomic even where they were clinically appropriate.
Two requirements are worth restating because they generate most of the audit exposure. The interactive communication under 99457 means a documented, live touchpoint with the patient. Passive review of transmitted data does not satisfy it. Remote physiologic monitoring can also be billed concurrently with chronic care management, which many practices still treat as mutually exclusive.
Clinically useful is not the same as billable
The reimbursement rules describe what Medicare will pay for. They do not describe what is worth knowing.
A patient’s Oura data showing sleep duration falling from seven hours to five over eight weeks is not billable under RPM. It is also one of the more useful pieces of information a clinician could have going into that appointment, and it costs nothing to collect. Treating the billing question as the value question is a category error that leads practices to ignore data they already have.
The reasonable posture is to separate the two explicitly. Build the RPM programme around devices and patients that meet the requirements, document it properly, and bill it. Use everything else as clinical context, without pretending it is a billable service.
The three-device problem
A patient arrives wearing an Apple Watch, an Oura ring and a continuous glucose monitor. The watch says they slept six hours and forty minutes. The ring says seven hours and twenty. Their sleep quality scores point in different directions.
The instinct is to work out which one is right. That framing produces no useful answer, since neither device is measuring sleep directly and both are inferring it from movement and cardiovascular signals through proprietary algorithms with different assumptions.
A more workable approach assigns each metric a source and stays with it.
| Metric | Sensible primary source | Reasoning |
|---|---|---|
| Nocturnal heart rate variability and resting heart rate | Ring, where one is worn | Finger-based measurement shows the strongest validated agreement with electrocardiogram overnight |
| Sleep duration and continuity | Whichever device the patient wears most consistently | Consistency of instrument matters more than the modest accuracy differences between devices |
| Daytime activity and exercise | Watch or sports device | Built for it; GPS and workout detection are considerably better |
| Cardiac rhythm events | Device with a cleared ECG feature | The feature exists for this purpose and others do not |
| Glucose | Continuous glucose monitor | No comparison |
Choose the source, record the choice, and read the trend from it. The other devices become corroboration rather than contradiction.
Where the work should not fall on the physician
Reconciling this by hand, chart by chart, is not a reasonable use of clinical time, and it is the sort of task that gets abandoned quietly after a few weeks.
It is a large part of what we built Longevitix to absorb. The platform reads from Apple Health, WHOOP, Garmin, Oura and Fitbit, with continuous glucose monitoring in development, and presents a physician with a trend rather than with five application interfaces. Where devices disagree, the underlying sources remain visible, so the clinician can see what is driving a divergence rather than being handed a smoothed number with no provenance.
We take the same position on that as on everything else the platform infers. A conclusion the clinician cannot trace back to its source is a conclusion they cannot properly disagree with, and a system that makes disagreement difficult is a system that will eventually be wrong in a way nobody catches.
Details above reflect the 2026 Physician Fee Schedule as published. CPT codes and CMS requirements change annually and this article is reviewed after each final rule. Practices should confirm current requirements with their billing team before relying on any summary, including this one.