RPM program · Obesity medicine

Obesity and incretin therapy

Half of patients stop within a year. Almost nobody is watching the months where they decide to.

This is the highest-leverage monitoring population in the country right now and the one with the worst instrumentation. Real-world persistence on incretin therapy is poor, discontinuation clusters in the first four months, and the clinical system finds out at the refill gap. The signals that predict a stop are tolerability and plateau, both of which are measurable daily and neither of which shows up in a chart. A program built around those two signals changes the conversation from 'you stopped' to 'let us adjust before you do.'
Obesity and incretin therapy · observed transmission profile Representative, not best case
Day 1Day 15Day 30
Transmitted days 0 / 30
Device supply code —
Start this program Model the economics
Signal stack

What gets measured, from where, and why it earns its place

Every source is tagged at ingest as billable or context-only. Context-only data is clinically valuable and never enters a device-supply line.

Weight
Connected scale, daily
Daily weighing with a smoothed trend, not weekly point weights. The trend is what distinguishes a plateau from a fluctuation, and plateaus are where people quit.
Blood pressure and heart rate
Cellular cuff
Resting rate rises modestly on incretins, and falling pressure during rapid loss means dose adjustments to antihypertensives that nobody remembers to make.
Body composition proxy
Bioimpedance scale, optional
Directional only. Consumer bioimpedance is not a lean-mass measurement, and we label it that way.
Tolerability and side effects
Structured daily check-in
Nausea, vomiting, and constipation scores. This is the actual reason people stop, and it precedes the stop by two to three weeks.
Injection taken
Patient-reported, weekly
Adherence is a therapeutic signal, not a physiologic one. Recorded, and deliberately kept out of the RPM device-supply line.
Escalation design

Thresholds, and who has to do something about them

Red marks a same-day or immediate pathway. Amber marks a review within the stated window. Each one is assigned to a named clinician with a response clock, not to a queue.

Tolerability score in the top quartile for 5+ days
ResponseContact before the next dose escalation
Weight trend flat for 21 days at a sub-target dose
ResponseTitration review
Systolic < 105 during active loss
ResponseAntihypertensive deprescribing review
Two consecutive missed injections
ResponseSame-week outreach

What we will not oversell

Adherence monitoring is therapeutic monitoring and lives in the RTM family. Physiologic monitoring lives in RPM. You cannot bill both families for the same patient in the same calendar month, so the program has to pick a track per month. Gathermed forces that choice at enrollment and records it.

Program operation

How the month actually runs

Enroll at initiation, not at the first plateau. The window you care about is weeks 4 through 20.
Scale plus cuff plus daily tolerability check-in.
Tolerability spike routes to a titration decision before escalation, which is the highest-value intervention in the whole program.
Plateau detection at 21 days flat triggers a review rather than an automatic dose increase.
Monthly management time with an interactive call. Bill 99470 for the light months, 99457 for the real ones.
Code path · CY2026

What this program can bill

Rates are national non-facility averages, rounded, for orientation. They vary by locality and payer. Verify against your fee schedule before you model revenue on them.

CodeWhat it coversNational avgCadence
99453Setup and patient education
One-time device setup and patient education per episode of care.		$22once per episode
99454Device supply, 16-30 days
Device supply with daily recordings or programmed alerts, 16 or more days of data in 30.		$52per 30 days
99445new 2026Device supply, 2-15 days
New for 2026. Device supply where 2 to 15 days of data were transmitted in 30. Paid at the same rate as 99454. Mutually exclusive with 99454.		$52per 30 days
99470new 2026Treatment management, first 10 min
New for 2026. 10 to 19 minutes of management time. Mutually exclusive with 99457.		$26per calendar month
99457Treatment management, first 20 min
Clinical staff or QHP time, 20 minutes, with at least one interactive communication.		$52per calendar month

Diagnosis anchors

ICD-10-CMDescription
E66.01Morbid obesity due to excess calories
E66.9Obesity, unspecified
Z68.41BMI 40.0-44.9, adult

Illustrative anchors only. Code to the documented clinical picture, not to a list on a vendor website.

The rules that break programs

One billing clinician per 30-day episode. RPM and RTM are not additive in the same calendar month. Duration codes are mutually exclusive. Management time requires at least one real-time interactive communication. Medical necessity precedes enrollment.

Full rule set and calculator

Sponsor view

If you are a manufacturer looking at this condition

The single clearest sponsored-program case in the portfolio. Persistence is the commercial bottleneck for every incretin franchise, and the causal chain from tolerability to discontinuation is currently invisible. A consented program produces dose-escalation trajectories, tolerability-to-discontinuation timing, plateau response, and structured stop reasons. That is a durable real-world evidence asset and a payer-negotiation asset. Read the sponsorship structure note carefully: this is exactly the arrangement that draws scrutiny if the economics flow the wrong direction.

Read this before you fund anything

Sponsor-funded monitoring that produces billable revenue for prescribers is remuneration to referral sources under the federal Anti-Kickback Statute. The structures that survive scrutiny keep the sponsor's data interest and the provider's billing interest completely separate. The three structures we will build, and the one we will not.

Questions

Obesity and incretin therapy monitoring, answered directly

Should a GLP-1 program be RPM or RTM?
If the billable measurement is weight and blood pressure from cleared devices, it is RPM. If it is drug adherence and therapy response, it is RTM. Not both in the same month for the same patient.
Is daily weighing harmful for patients with disordered eating history?
It can be. Screen at enrollment and use a weekly cadence with hidden daily values for anyone with a history. The platform supports suppressing the number while keeping the trend clinician-visible.
Can a manufacturer pay for the monitoring for our patients?
Not without careful structuring. Free or subsidized services that generate billable revenue for prescribers implicate the federal Anti-Kickback Statute. Get counsel before the first dollar moves.
Adjacent programs

Conditions that travel with this one

Comorbidity is the norm. One episode, one billing clinician, one device-supply line per 30 days, even when the patient has four diagnoses.

Type 2 diabetesContinuous glucose data made the signal free. The scarce thing is the person who acts on it. RPM Endocrinology, primary care 99453 99454 99445 HypertensionThe highest-volume remote monitoring program in Medicare, and the easiest one to run badly. RPM Primary care, cardiology, nephrology 99453 99454 99445 Obstructive sleep apneaAdherence is the entire clinical outcome, and it is decided in the first ninety days. RTM Sleep medicine, pulmonology, ENT 98975 98976 98984
Start this program All 18 conditions