The vendor’s projection probably looked simple. Multiply the panel size by the monthly reimbursement per patient, subtract a platform fee, and RPM turns into a reliable new revenue line. The spreadsheet had one row per month and every month was green.
Then the program ran for six months, and the deposits never matched the projection. Some patients didn’t take enough readings. Some months, nobody had the required conversation with them. A handful of claims came back denied, and the coinsurance statements went out to patients who didn’t expect a bill. Meanwhile, nobody had counted the hours your MA spent chasing cuffs that stopped transmitting.
Remote patient monitoring can be a profitable service line for an independent practice. Its ROI depends far less on how many patients you enroll than on how much of each billable dollar actually reaches your bank account. This guide shows how to calculate that number honestly. It covers where RPM revenue comes from, the five places it leaks before the deposit, a worked 100-patient example, and the two kinds of break-even you need to know before you commit staff time or sign a contract.
The RPM Projection That Promised $100 per Patient
A four-physician internal medicine group in Tennessee came to us eight months into its RPM program. The vendor’s projection had assumed every enrolled patient would bill the device code and the first management code every month, about $100 per patient. With 150 patients enrolled, the spreadsheet promised roughly $15,000 a month.
Their actual deposits were averaging a little over half that. Nobody had made a mistake you could point to. The program just leaked in several small places at once:
- About one patient in five fell short of the data-day threshold in a typical month.
- More patients missed the interactive-time threshold, because calls got squeezed out on busy clinic days.
- The patient coinsurance portion went partly uncollected.
- The MA who ran the program had quietly become a half-time RPM coordinator, and none of her hours appeared in the projection.
Once we rebuilt the model from billable to banked, the program was still profitable, by less than the partners had been promised. More useful, the model showed exactly which leak to fix first.
How to Calculate Remote Patient Monitoring ROI
Remote patient monitoring ROI is the net revenue the program brings in, minus everything it costs to run, divided by those costs, measured over a defined period. To calculate it for your practice:
- Estimate billed revenue. For each code, multiply the number of patients who will realistically qualify each month by the code’s payment.
- Subtract claim leakage. Remove denials, rework, and claims that can’t be supported.
- Separate the payer share from the patient share. Apply your realistic collection rate to the patient’s cost-share portion.
- Subtract every cost. Include platform and device fees, connectivity, all staff minutes (not only the billable ones), device loss, and your one-time startup costs.
- Divide net by cost. Calculate it monthly at steady state, and separately over the first 6-12 months, including ramp-up.
The rest of this guide fills in each step with the numbers that most projections leave out.
Where RPM Revenue Comes From
RPM reimbursement comes from a small set of CPT codes. One pays once per episode of care. The rest pay monthly, and only when a patient meets data-day or time thresholds in that month, which is why the qualification rate matters so much.
The Monthly Codes That Drive Revenue
The rates below are approximate Medicare national averages from our complete 2026 RPM billing guide. Your actual payments vary by locality and payer.
| Code | What it pays for | Monthly requirement | Approx. payment |
|---|---|---|---|
| 99454 | Device supply and data | 16+ days of readings | ~$50 |
| 99445 | Short-duration device supply | 2-15 days of readings | Lower than 99454 |
| 99457 | Treatment management, first 20 minutes | 20+ minutes including interactive communication | ~$50 |
| 99470 | Light-touch management | 10-19 minutes | Lower than 99457 |
| 99458 | Each additional 20 minutes | Beyond the first 20 | ~$40 |
A patient who meets the standard thresholds bills roughly $100 a month for the device and first management codes. That is the ceiling most projections treat as the average.
The One-Time Setup Code
The setup and education code (99453) pays roughly $20 once per episode of care. It helps in the first month, but it is too small to carry the economics, and a projection that leans on it is a projection built on enrollment churn.
How the 2026 Short-Duration Codes Change the Math
Before 2026, a patient with 14 days of readings billed nothing for device supply, and 18 minutes of management time billed nothing either. The short-duration codes now recover part of that revenue. They soften the qualification leak described below, but they don’t close it, because they pay less than the standard codes. Model them as partial recovery, not as a reason to stop chasing full qualification.
The Billable-to-Banked Waterfall: Five Leaks Between Projection and Deposit
Every RPM program loses money between the theoretical maximum and the bank deposit. The question is how much, and where. We trace it through five leaks in the order they happen. Each one shrinks the base the next one applies to, which is why small percentages add up quickly.
Leak 1: Qualification
Patients who don’t meet the data-day or time threshold in a given month bill less, or nothing, for that code. This is usually the largest leak and the one most within your control. It shows up first as a column on the platform report that someone checks on day 28, when it is too late to help. The patients who fall short are rarely refusing. Their device went silent, or their call kept getting pushed to tomorrow.
Leak 2: Claims
Some of what you bill won’t be paid. The causes are denials for missing consent or orders, time logs that don’t show who did the work, code combinations that conflict, and payer-specific rules. Rework costs staff time on top of the lost revenue.
Leak 3: Cost-Share
RPM services fall under Medicare Part B, so patients typically owe coinsurance unless secondary coverage picks it up. A patient who wasn’t told about it at enrollment may not pay it, and may leave the program when the first statement arrives. Model the Medicare share and the patient share separately, with a realistic collection rate on the patient portion.
Leak 4: Labor
Billable minutes are only part of the time RPM consumes. Shipping devices, troubleshooting connections, chasing missed readings, and preparing month-end reports all take staff time that never appears on a claim. If your model counts only the 20 billable minutes, it is understating your largest controllable cost.
Leak 5: Attrition and Devices
Patients leave: they are hospitalized, move into a facility, lose interest, or pass away. Each departure leaves behind setup effort that won’t be repaid and often a device that never comes back. New patients also bill less in their first month, because they join partway through it.
A Worked Example: 100 Patients, Month by Month
The model below is a practice of 100 enrolled patients. Every number is an illustrative assumption you should replace with your own, but the structure is the point. Run your numbers through the same lines and you will see your real ROI.
Steady-State Monthly P&L
Assumptions:
- Qualification: 80% of patients meet the device threshold, 70% meet the first management threshold, and 15% bill one additional 20-minute unit.
- Claims: Denials and rework take 4% of billed revenue.
- Cost-share: Medicare pays 80% of the allowed amount, and the practice collects 60% of the patient portion.
- Costs: Platform, device, and connectivity fees run $15 per patient per month. Staff average 30 minutes per patient per month at a $32 loaded hourly cost. Device loss and replacement average $2 per patient per month.
| Line | Amount |
|---|---|
| Billed: 80 × 99454, 70 × 99457, 15 × 99458 | $8,100 |
| Denials and rework (4%) | ($324) |
| Allowed and supported | $7,776 |
| Medicare share (80%) | $6,221 |
| Patient cost-share collected (60% of $1,555) | $933 |
| Banked revenue | $7,154 |
| Platform, devices, connectivity | ($1,500) |
| Staff time (50 hours) | ($1,600) |
| Device loss and replacement | ($200) |
| Net monthly contribution | $3,854 |
That works out to about $38.50 net per enrolled patient per month. The projection promised about $100 per patient. Roughly $28 of the gap disappears before the deposit. The other $33 is what it costs to run each patient.
Ramp-Up and Startup Payback
A program doesn’t start at 100 patients. Assume $6,000 in one-time startup costs: protocol and workflow build hours, staff training, platform configuration, and an initial device buffer. Enrollment runs 30, 40, and 30 patients over the first three months. New patients bill the setup code but qualify for the monthly codes far less often in their first partial month, and they need more staff time for onboarding.
| Month | Patients (new) | Banked | Costs | Net | Cumulative after startup |
|---|---|---|---|---|---|
| 1 | 30 (30) | $1,192 | $1,230 | ($38) | ($6,038) |
| 2 | 70 (40) | $3,736 | $2,630 | $1,106 | ($4,932) |
| 3 | 100 (30) | $6,200 | $3,540 | $2,660 | ($2,272) |
| 4 | 100 (0) | $7,154 | $3,300 | $3,854 | $1,582 |
In this model, startup costs are paid back during month four. Your first month will likely be roughly break-even at best, and that is normal. Judging the program on month one or two is like judging a new hire on their first week.
Your Break-Even per Patient
The second kind of break-even is the one most guides skip: what each enrolled patient has to generate every month just to cover what they cost.
In this model, each enrolled patient costs about $33 a month ($15 in technology, $16 in staff time, $2 in device loss), whether they qualify or not. Each billed dollar banks about 88 cents after denials and cost-share collection. So each enrolled patient needs to generate about $37 in billed charges a month to break even. That is a little more than a third of what a fully qualifying patient bills.
The flip side matters just as much. A patient who never transmits still costs you the full $33 every month. Panel size multiplies whatever your per-patient margin is, and that includes a negative one.
What Changes When Qualification Moves
| Scenario | Device / first management / extra unit qualification | Net monthly contribution | Net per patient |
|---|---|---|---|
| Struggling | 60% / 50% / 10% | $1,911 | ~$19 |
| Typical | 80% / 70% / 15% | $3,854 | ~$39 |
| Well run | 90% / 80% / 20% | $4,914 | ~$49 |
Moving from struggling to well run more than doubles the margin without enrolling a single new patient. In our experience, that is where most of an independent practice’s RPM upside lives.
Five ROI Mistakes That Distort the Picture
Some of these make RPM look far better than it is, and others make it look worse. Either way, the practice ends up making the wrong decision about staffing, pricing, or whether to keep the program at all.
Using Gross Reimbursement as Revenue
Multiplying panel size by the full monthly payment is the single most common error in RPM projections. Revenue starts with the share of patients who actually qualify each month, not with the share who are enrolled.
Treating Staff Time as Free
“Our MA will handle it” doesn’t make labor free. It moves the cost into a part of the budget nobody looks at. Count every minute spent on RPM, including the minutes that never appear on a claim.
Forgetting the Patient’s Cost-Share
If the model treats the full allowed amount as cash, it overstates revenue by whatever share of coinsurance you don’t collect. It also misses the patients who leave the program after their first statement.
Assuming Every Patient Qualifies Every Month
Even well-run programs have patients who travel, get hospitalized, or simply have a bad month. A model with no qualification assumptions is a best case presented as a forecast.
Deciding the Program’s Future in Month Two
Ramp-up months always look weak, because new patients join partway through a month and need extra onboarding time. Judge ROI on steady-state months and on cumulative payback, not on the first billing cycle.
How to Improve RPM Program Profitability Without Enrolling More Patients
Every lever that improves RPM profitability sits somewhere in the waterfall. Working the leaks almost always pays back faster than adding enrollment, because each new patient brings their own leaks with them.
Raise the Qualification Rate
Work a daily list of patients with no transmission in 72 hours. Check every patient’s data-day count around day 20, while there is still time to recover the month. Schedule the monthly interactive call early in the month, not in its last week. Our five-phase RPM workflow for clinics shows where those checks fit in the daily routine.
Lower Your Cost per Minute
The protocol-driven half of RPM work can be done well by trained, supervised support staff, at a lower cost per hour than nurse time. That covers onboarding calls, silent-device outreach, time logging, and month-end preparation.
That is exactly what our remote patient monitoring support provides. A dedicated, HIPAA-compliant Care VMA handles that work inside your platform and EHR, while your nurses keep the clinical escalations. It moves the labor line in the waterfall and, through better follow-up, the qualification line too. If you are weighing that model, our guide to measuring the ROI of a virtual medical assistant applies the same discipline to the staffing decision.
Have the Coinsurance Conversation at Enrollment
Explain the patient’s possible cost-share in the enrollment conversation and record it with the consent. Patients who know about the coinsurance up front are less likely to drop out after the first statement and more likely to pay it.
Count Value-Based Upside Only Where You Hold the Risk
Fewer admissions and emergency visits are real benefits of RPM. They show up in your revenue only if you share in the savings, through a Medicare Advantage risk contract, an ACO, or a similar arrangement. If you hold that risk, model the upside separately. If you don’t, keep it out of the ROI and treat it as a clinical benefit.
Track Banked Dollars per Enrolled Patient
RPM’s real economics come down to one number: banked revenue per enrolled patient per month, compared against what that patient costs you to run. Track it monthly, compare it to your break-even line, and work the leaks in order before you push enrollment. A program that nets $40 per patient on a stable panel is worth more than one that grosses $100 per patient on paper and leaks most of it before the deposit.
If you want to see your own program through this lens, book a free consultation with the Care VMA team. We will run your numbers through the billable-to-banked waterfall with you and show you which leak is costing you the most.
Frequently Asked Questions
These are the ROI questions practice owners ask most often, usually right before a partner meeting.
How do you calculate remote patient monitoring ROI?
Start with billed revenue based on how many patients realistically qualify for each code each month. Then subtract denials and uncollected patient cost-share to get banked revenue. Subtract every cost, including platform fees, devices, all staff minutes, and device loss, and divide the net result by total cost for the period.
How many patients do you need to break even on RPM?
Panel size matters less than per-patient margin. In our illustrative model, each enrolled patient needs to generate about $37 in billed charges a month to cover their own costs, and a 100-patient program paid back $6,000 in startup costs in about four months. Your figures will differ based on your payer mix, vendor fees, and staffing costs.
How long does an RPM program take to pay back its startup costs?
With a disciplined launch and steady enrollment, many practices recover startup costs within the first four to six months. Ramp-up months are weaker, because new patients join partway through a month and need extra onboarding time. Slow enrollment or low qualification rates stretch the payback period.
Does staff time make RPM unprofitable?
Not necessarily, but it is often the largest controllable cost and the one most projections leave out. Counting every RPM minute, not only billable ones, gives an accurate picture. Shifting protocol-driven work to lower-cost trained support staff is one of the most reliable ways to improve the margin.
Do patients pay anything for remote patient monitoring?
Under Medicare Part B, patients typically owe coinsurance on RPM services unless secondary coverage pays it. Explaining that cost at enrollment and documenting it with consent reduces both uncollected balances and dropouts after the first statement.
