RTM Devices and Wearables: A Technical Guide for PT Clinics

What counts as an RTM device

An RTM device is any hardware or app that captures non-physiological data about a patient's therapy, meaning information about how they move, exercise, and function rather than their vital signs. Medicare uses that exact distinction to separate remote therapeutic monitoring (CPT codes 98975 through 98981) from remote patient monitoring. RPM devices measure physiological data like blood pressure, glucose, and heart rate. RTM devices track the musculoskeletal and behavioral signals a physical therapy plan depends on, including exercise adherence, pain scores, and range of motion.

That line matters because the code you bill depends on the type of data your device produces. A connected blood pressure cuff feeds RPM. A phone that counts squat repetitions or a pain log inside a patient app feeds RTM. The two categories rarely overlap in a therapy setting, so choosing hardware built for physiological vitals adds cost without supporting the codes a clinic actually bills.

Three device categories capture RTM data in practice. Smartphones track motion and gait using their built-in sensors and camera. Dedicated wearables such as clip-on accelerometers and IMUs record continuous movement. App-based self-report tools collect structured check-ins on pain and adherence. Each captures data differently, carries a different setup burden, and fits different CPT codes, and the sections that follow break down all three.

Smartphone-based motion and gait tracking

Smartphone-based motion tracking turns the phone a patient already owns into an RTM sensor, using three components that ship in nearly every modern handset. The accelerometer measures linear movement and impact, the gyroscope measures rotation and orientation, and the front camera drives pose estimation that maps joints in the video frame to a skeletal model. Together these let a patient app count repetitions, estimate range of motion, and time a walking task without any extra hardware.

Each sensor answers a different question about a home exercise session. The accelerometer and gyroscope, working through the app, detect that a patient started a movement, how many times they repeated it, and how fast they moved. Camera-based pose estimation adds movement quality, since it tracks joint angles frame by frame and flags whether a squat reaches depth or an arm raise clears shoulder height. Physitrack's Motion Capture works this way, processing video on the device and storing structured numeric data rather than recording the session.

Phone-based tracking measures some things reliably and others only as an approximation. Rep counting, cadence, and gait speed over a fixed distance hold up well, because they depend on clear, repeated motion the sensors detect cleanly. Range-of-motion readings are better treated as a proxy than as a goniometer replacement, since camera angle, lighting, and clothing all shift the joint estimate. A clip-on inertial sensor placed directly on the limb will beat a phone for precise, continuous joint measurement, so a clinic tracking small ROM changes after surgery may want more than a phone can give.

The practical advantage is that the patient already owns the device, which removes the two steps that stall wearable programs. You skip shipping hardware, and the patient skips pairing and charging a separate sensor before the first session. Onboarding becomes downloading an app and following an on-screen setup, which raises the odds a patient logs the sessions RTM billing depends on.

Wearable sensors: accelerometers and IMUs

Dedicated wearables capture movement that a phone in a pocket cannot see. An inertial measurement unit, or IMU, combines an accelerometer, a gyroscope, and often a magnetometer into a small sensor a patient clips to a limb or straps across a joint. Because the sensor sits on the body part you want to measure, it records joint angles, repetition quality, and limb velocity with more fidelity than a phone estimating movement from across the room. A clip-on IMU on the thigh tracks knee flexion during a squat far more reliably than camera pose estimation catching a partially blocked view.

Continuous monitoring is the second advantage. A patient wears an IMU through daily activity, and it logs steps, cadence, and time spent moving without asking the patient to open an app and start a session. For a clinic tracking whether a post-op patient actually loads a repaired knee across a full day, that passive stream answers questions a scheduled phone check-in never reaches.

The fidelity comes at an operational cost the clinic absorbs. You have to buy the sensors, ship or hand them out, and get them back. Patients forget to charge them, misplace them, or return them dirty, so your staff spends time on cleaning, re-pairing, and chasing down units that walked out the door. Every lost or broken sensor is a replacement cost and a gap in the data record. For a small clinic running RTM across dozens of patients, that inventory work adds up quickly.

On billing, IMU-based programs most often support the CPT codes tied to musculoskeletal system status, since the data they produce describes movement and function directly. The higher-resolution motion record can make functional progress easier to document and defend. Whether that resolution justifies the hardware overhead depends on your patient population and clinical questions, which the evaluation section below works through in detail.

App-based patient self-report tools

Structured self-report tools are the simplest RTM device category to deploy, because the "device" is software the patient already runs on a phone they own. Inside a patient app, a clinician sets up scheduled check-ins that ask the patient to log pain on a numeric scale, confirm which exercises they completed, and note symptoms or function since the last session. Each logged response becomes a discrete data point, and those data points are what CMS treats as non-physiological therapeutic data under CPT codes 98975 through 98981.

The billing math is why this category matters. RTM device supply codes require at least 16 days of collected data in a 30-day period, and a daily self-report prompt maps directly onto that threshold. A patient who answers a short check-in most days generates enough qualifying data without wearing or charging anything. That low friction is also what makes self-report the easiest to scale across a full caseload, since staff do not distribute, pair, or track hardware.

Self-report data works best alongside functional outcome tracking rather than as a lone signal. Patient-reported outcome measures (PROMs) such as validated pain and disability questionnaires capture how a condition changes over a course of care, while daily adherence and pain logs capture the day-to-day pattern that supports the billing threshold. Together they give a clinician both the trend and the detail behind it.

Self-report cannot verify movement quality the way sensors can, so it complements motion data instead of replacing it. In PhysiApp, patients log exercise completion, pain, and PROMs responses in one place, which gives clinicians a review-ready record of both adherence and outcomes without adding any hardware to the clinic.

Device categories compared

The three device categories differ most in what they capture and what they cost you to run. The table below sets them side by side so you can match a category to your clinical need and staffing reality before you evaluate any specific platform.

Read the table by your priority. If continuous, high-fidelity movement data drives a clinical decision, wearables earn their overhead. If broad deployment and a clean path to the 16-day billing threshold matter more, smartphone tracking and self-report carry most clinics with far less friction.

How device choice affects billing and compliance

CMS pays the RTM device supply code, 98977, only when a device records and transmits at least 16 days of data within a 30-day period. That threshold makes patient behavior the deciding factor in whether a month is billable, and device choice shapes that behavior more than any other setup decision.

Proprietary hardware raises the risk of missing the 16-day mark. When a platform ships a clip-on IMU that a patient has to wear, charge, and remember to sync, every one of those steps is a chance for the data stream to break. A dead battery over a long weekend, a device left at a relative's house, or a sensor that ends up in a drawer all produce the same result. The month falls short of 16 days, and you cannot bill it. The clinic absorbs the cost of the hardware while collecting no reimbursement for the effort.

Device choice controls two variables directly, and both feed billing. The first is data accuracy, which determines whether the numbers you capture support the clinical picture and hold up to payer scrutiny. The second is patient adherence, which determines whether you capture enough days to bill at all. A high-fidelity sensor that patients abandon produces beautiful data for eight days and nothing for the rest of the month.

Weigh those two variables against your patient population before you commit. A smartphone-native approach lowers the adherence barrier because patients already carry and charge the device daily, which protects the 16-day count. A dedicated sensor buys accuracy, but only if your patients will actually keep it running.

Evaluating device compatibility before choosing a platform

Before you compare RTM platforms on features, check whether their device requirements will actually work in your clinic. Device compatibility decides your setup burden, your ongoing costs, and how consistently patients generate the data you bill against. Work through the four questions below for any platform on your shortlist.

Does the platform require proprietary hardware, or does it work with the patient's own smartphone?

Ask this first, because the answer shapes everything downstream. A platform tied to a proprietary wearable means you buy, distribute, and eventually replace the device for every enrolled patient. A smartphone-native platform lets the patient use hardware they already own and know how to charge.

What is the onboarding burden on staff and patients?

Count the steps between enrolling a patient and collecting their first day of data. Proprietary sensors add pairing, firmware updates, and a charging routine you have to teach. App-based tools usually need a download and a login. The more steps you add, the more patients drop off before they reach the 16-day billing threshold, so onboarding friction is a compliance issue, not just a convenience one.

How much data accuracy does your clinical use case actually need?

Match device fidelity to the decision the data supports. Tracking whether a patient completed their prescribed exercises needs adherence logs and rep counting, which a phone or self-report tool handles well. Measuring subtle joint-angle changes across a return-to-sport program may justify a dedicated IMU. Buying lab-grade sensors for routine adherence monitoring adds cost and cleaning work without improving the clinical picture.

What is the total cost of hardware ownership?

Add up procurement, charging infrastructure, cleaning between patients, and replacement for lost or broken units. A wearable-dependent program carries these costs indefinitely, and they scale with every patient you enroll. A smartphone-first program removes them entirely.

Physitrack fits this framework as a smartphone-first RTM platform with no proprietary hardware. Patients capture adherence, pain scores, and functional outcomes through the PhysiApp patient app on their own phone, so you avoid device procurement, charging routines, and the replacement overhead that comes with a wearable fleet. That model suits clinics that want to enroll patients quickly and keep the setup burden on staff low. Clinics with a genuine need for continuous high-fidelity sensor data should weigh that requirement against the ownership costs above.

Conclusão

Your device choice should follow the clinical need in front of you and the patients you actually treat, not the hardware a vendor wants to sell you. A knee rehab cohort that needs objective range-of-motion data has different requirements than a chronic low back caseload managed through structured check-ins and pain tracking. Start from what you need to measure, then find the lightest device category that captures it reliably. Hardware you have to procure, charge, clean, and replace only earns its place when phone-based tracking or self-report cannot answer the clinical question.

Physitrack's smartphone-first model fits clinics that want to run RTM without standing up a device inventory, since patients use hardware they already own and staff skip the pairing and distribution step. For a full platform view of remote therapeutic monitoring, including billing workflows and integrations, see Physitrack's RTM platform comparison content. This guide stays scoped to the device and sensor question on purpose.

Perguntas frequentes

Do patients need to buy a wearable to bill RTM? No. CMS's RTM codes 98975 through 98981 require non-physiological data, which structured self-report and smartphone-based tracking can capture. Physitrack's smartphone-first approach lets you collect billable adherence and pain data without distributing proprietary hardware, which removes procurement and replacement cost for the clinic.

Can RTM be done with just a smartphone? Yes, for most musculoskeletal caseloads a smartphone handles adherence logging, pain scales, and app-based check-ins. Physitrack runs RTM through the patient app on a phone the patient already owns, so onboarding takes minutes rather than a hardware handoff. The practical benefit is faster patient uptake and fewer abandoned devices threatening your data threshold.

What data does CMS actually require for RTM? CMS requires non-physiological data such as therapy adherence, musculoskeletal system status, and patient-reported pain or function, distinct from the vitals that remote patient monitoring captures. Physitrack organizes this into structured records tied to prescribed programs, so the captured data maps to the code you intend to bill. That structure gives you a defensible record for the 16-days-of-data requirement.

How accurate does device data need to be for reimbursement? Reimbursement depends on capturing qualifying data across the required days, not on lab-grade sensor precision. Physitrack pairs self-report with in-app tracking so the data is consistent enough to support billing while staying practical for patients to submit. Accuracy still matters for clinical judgment, so match sensor fidelity to the decision you need to make.