Find out everything about the vision, mission, and the people who shape CompuGroup Medical worldwide. Investors will also find helpful information, documents, and other publications.
Remote patient monitoring has been a Medicare-reimbursable service since 2020, but two billing thresholds have quietly left a significant amount of real clinical work uncompensated since the beginning.
Prior to 2026:
Work that fell short of either number generated no billing, even when it was documented and clinically meaningful.
Two new CPT® codes introduced for 2026 lowered both of those thresholds. For practices already running an RPM program, the impact is immediate—it doesn’t require a single new patient enrollment.
The updates came out of an Office of Inspector General review of RPM utilization that ran through most of 2025—triggered not by concerns about fraud or overuse, but by six years of outcomes evidence that consistently supported the program.
A 2025 study in Health Affairs tracked 754 primary care practices over that period and found a 20% increase in Medicare revenue at RPM-adopting practices compared to matched controls. Of that, 12.4% came from direct RPM billing and roughly 25% from additional office visits and downstream testing the monitoring generated. The study also found no reduction in access or care quality for patients not enrolled in RPM—capacity expanded rather than shifted.
CMS took that body of evidence as reason to issue the 2026 Physician Fee Schedule final rule, broadening the billing framework specifically to cover clinical work that was already happening at practices but going uncompensated.
Medicare heard from practices that they were doing a lot of real work for free. If a patient hit 17 minutes of engagement instead of 20, that time was unreimbursable. If they took their readings 10 times in a month instead of 16, the device supply code didn't trigger. The new codes fix that.
CPT 99445 — Device supply, 2–15 days of patient readings
The previously existing device supply code, 99454, requires patients to transmit data on 16 or more unique days per month.
The new CPT 99445 covers patients who transmit on 2 to 15 days. The two codes are mutually exclusive—a patient bills under one or the other in a given month, not both.
The national average reimbursement for 99445 is $48, the same as 99454. That parity is notable and, in all likelihood, temporary.
I wouldn't count on equal reimbursement holding. My expectation is that CMS reduces the 2-to-15-day rate once they have a full year of claims data behind it. But right now it's the same, and that's a window worth taking seriously.
CPT 99470 — Treatment management, 10–19 minutes of engagement
The previously existing treatment management code, 99457, requires a minimum of 20 minutes of combined clinical staff time per month—reviewing readings, working the patient’s chart, and direct patient contact.
The new CPT 99470 covers engagement that runs 10 to 19 minutes, reimbursed at $26. Sessions that previously fell short of the 20-minute mark went uncompensated entirely. Under 99470, a 12- or 15-minute session now generates revenue.
No existing remote patient monitoring or chronic care management codes were removed or reduced in 2026.
The 2026 changes add 99445 and 99470 on top of the prior structure. Every code that was billing in 2025 continues to bill at the same rates.
| CPT | Description | 2026 Rate | Notes |
|---|---|---|---|
| 99453 | Initial setup and patient education | $19 | One-time per patient |
| 99445 | Device supply, 2–15 days of readings | $48 | New for 2026 |
| 99454 | Device supply, 16+ days of readings | $48 | Unchanged |
| 99470 | Treatment management, 10–19 min/month | $26 | New for 2026 |
| 99457 | Treatment management, first 20 min/month | $54 | Unchanged |
| 99458 | Treatment management, additional 20 min | $41 | Unchanged |
| G0506 | CCM enrollment/comprehensive assessment | $62 | One-time per patient |
| 99490 | CCM, first 20 min/month | $43 | Requires 2+ chronic conditions |
| 99439 | CCM, additional 20 min/month | $39 | Add-on to 99490 |
National averages. Geographic adjustment applies. RPM and CCM may be co-billed for the same patient when documentation supports separate services.
Consider a practice with 200 patients enrolled in remote patient monitoring. Under the pre-2026 structure, say 140 patients consistently hit the 16-day reading threshold and 20-minute engagement threshold. That meant the remaining 60 were generating clinical activity—monitoring, outreach, chart time—that produced no billing.
Under the 2026 code set, those patients bill under 99445 for device supply and 99470 for engagement. At $74 combined per patient per month, would generate approximately $4,400 in monthly revenue from work the practice was already doing.
At scale, the difference is more significant. A program generating $22,000 a month in 2025 can realistically generate $44,000 in 2026 on the same enrolled panel, based on ARIA Population Health program data.
The 2026 changes don't require re-enrollment, new devices, or additional staff—but they do require reviewing your current program with fresh eyes, because existing billing workflows may not automatically account for the new thresholds.
Start with patients who were falling below the 16-day reading mark. These weren't necessarily disengaged—some took readings regularly, just not at the frequency the old code required. Under 99445, patients transmitting on as few as two days per month now generate device supply revenue. Check whether those patients have been properly classified in your billing workflow or quietly written off as non-billable.
Then look at engagement time documentation. If your care management notes don't capture sessions running 10 to 19 minutes as distinct billable events, that's a gap worth closing now. The 99470 code doesn’t capture itself—your workflow needs to route those sessions correctly rather than discarding them or stretching documented time to hit 20 minutes.
Finally, review device non-transmittal patterns across your panel. Patients who were categorized as non-compliant under the 16-day rule may simply have been sitting below a threshold that no longer defines the boundary. Some belong in a re-engagement workflow; others are billable under 99445 right now.
We don't hand patients a device and move on. We use every channel—phone, text, app—and we re-engage if we haven’t seen readings in five to seven days. That consistency is what gets you to 90% billable rates. The codes are only as valuable as the engagement behind them.
One reason practices have held off on RPM is the compliance math: enroll 100 patients, and some percentage won't hit the reading or engagement thresholds, generating clinical overhead with no corresponding Medicare reimbursement.
The 2026 codes don't eliminate that dynamic, but they dramatically reduce it by creating a billing tier for the middle range of patient participation—exactly the tier where most hesitation has been concentrated.
The clinical outcomes data is solid. A quasi-experimental cohort study of Medicare patients found RPM associated with a 75% reduction in stage 2 hypertension over 12 months. A separate prospective cohort study found 71.5% of RPM-enrolled patients achieved controlled blood pressure at 12 months, compared to 58.1% under standard care. Cardiac readmission rates drop roughly 65% after 90 or more days of monitoring for appropriate patients. Both figures come from programs that have been running under Medicare authorization since 2020, with results validated in peer-reviewed research.
For practices also managing patients with multiple chronic conditions, remote patient monitoring pairs naturally with chronic care management—RPM handles the device-driven, condition-specific data layer while CCM provides the broader care coordination wraparound. The two programs can be co-billed for the same patient, and combining them tends to improve engagement in both.
The 2026 Medicare RPM reimbursement changes don't alter the clinical value of the program. They make the billing structure more honest about how patients actually participate, and for practices that have been on the fence, that's a meaningful shift in the calculation.
