Why Outcome Measures Are the Clinical Backbone of RTM

TL;DR

• Patient-reported outcome measures (PROMs) capture how a patient rates their own pain and function, which is different from adherence data like login streaks and completed-exercise counts.
• Adherence data shows a patient did the work. PROMs show whether the work changed their condition.
• CMS built RTM's CPT codes, 98975 and 98977, around collecting patient-reported outcome data, not passive device readings, tying reimbursement to clinical outcomes.
• A 2025 retrospective case-control study in Archives of Rehabilitation Research and Clinical Translation found in-person PT combined with RTM produced better patient-reported outcomes and engagement than in-person PT alone.
• Consistent PROM collection makes RTM billing defensible under audit and gives referring physicians and payers evidence they trust, which adherence logs cannot supply.

What PROMs are, and why RTM isn't just adherence data

Patient-reported outcome measures, or PROMs, are standardized questionnaires that capture how a patient experiences their own pain, function, and disability in their own words. A patient rates their back pain on a numeric scale, reports how far they can walk before symptoms stop them, or answers whether they can lift a grocery bag without difficulty. Each answer is scored against a validated instrument, so a clinician can track whether a patient is measurably better in week six than they were at intake.

That measurement is the whole point, and it separates PROMs from the data that dominates how remote therapeutic monitoring usually gets marketed. Most RTM sales pitches lead with step counts, login streaks, exercise completion percentages, and readings pulled from a connected device. All of that is adherence data. It tells you a patient opened the app, tapped through their exercises, and logged a session.

Adherence data answers a narrow question. It confirms the patient did something. It says nothing about whether the something worked. A patient can complete every prescribed exercise for four weeks with perfect app engagement and still report the same shoulder pain and the same functional limits they walked in with. The completion log looks excellent, and the clinical result is a plateau.

PROMs answer the question that actually matters to the patient, the referring physician, and the payer. A validated pain scale that drops from 7 to 3, or an Oswestry Disability Index score that improves past its minimal clinically important difference, shows treatment producing a real change in the patient's life. That is evidence of clinical value, not just evidence of activity.

The distinction has consequences for how you build an RTM program. A program built on adherence metrics can generate impressive dashboards while proving nothing about outcomes. A program built on consistent PROM collection generates a record of whether patients recovered. When you evaluate RTM around clinical value rather than billing mechanics, PROMs are the layer that carries the argument, and adherence data becomes context rather than the headline.

Why CMS built 98975 and 98977 around patient-reported data

CMS built RTM's core CPT codes around patient-reported outcome data because it wanted to reimburse clinical signal, not passive telemetry. When CMS created the remote therapeutic monitoring codes in the 2022 Physician Fee Schedule, it deliberately scoped them to musculoskeletal and respiratory therapy response, and it named non-physiologic data, including patient-reported outcomes and therapy adherence, as the monitored data. That choice separates RTM from remote physiologic monitoring, which tracks device readings like blood pressure or glucose. RTM asks a different question. It asks whether the patient is getting better.

Two codes carry the setup and monitoring work, and each has requirements you have to meet to bill defensibly. Code 98975 covers the initial setup and patient education on using the monitoring equipment or software, billed once per episode of care, and CMS only allows it once the patient has engaged with monitoring for at least 16 days. Code 98977 covers the device or software supply for musculoskeletal monitoring across the 16-to-30-day tier of a 30-day period. Sixteen days of reported data is not a formality. It is the evidentiary floor CMS set to confirm the patient engaged with the monitoring long enough to produce a meaningful clinical picture.

The 16-day threshold explains why adherence logs alone leave your documentation thin. A login streak or an exercise-completion count tells a payer the patient opened an app. It does not tell the payer what changed. A validated pain score or functional index collected across those days shows the trajectory of recovery, and that trajectory is the thing CMS treats as the reimbursable unit of value. When an auditor reviews an RTM claim, structured outcome data answers the question the code was written to capture. Completion metrics restate the question without answering it.

Read alongside CMS's broader push toward outcome measurement and value-based payment, the design of 98975 and 98977 reads as a clear signal. CMS is steering reimbursement toward programs that document clinical response, and it built RTM's billing structure to reward exactly that. Patient-reported outcomes are not an optional enhancement to an RTM program. They are the data CMS priced the codes to collect, and a program that skips them collects engagement metrics while billing for something the codes never intended to pay for.

What the evidence shows: the 2025 case-control study

A 2025 retrospective case-control study in Archives of Rehabilitation Research and Clinical Translation found that patients receiving in-person physical therapy alongside RTM reported better outcomes and higher engagement than patients receiving in-person physical therapy alone. The RTM group improved on patient-reported measures and stayed more involved in their care across the treatment episode. Both groups received the same core clinical care, so the added monitoring layer is what separated their trajectories.

The comparison design makes that finding worth taking seriously. A case-control study matches patients who received one treatment against similar patients who did not, which lets researchers isolate the effect of RTM instead of crediting it with improvements that better patients would have shown anyway. When the two groups start from comparable baselines and the only meaningful difference is the monitoring layer, a gap in outcomes points to the intervention rather than to selection.

This is independent clinical literature, not a vendor case study, and that distinction carries weight when you present RTM to a skeptical referring physician or a payer. A software company reporting its own success invites the obvious question about who chose the patients and how the numbers were framed. A peer-reviewed retrospective study built on a control group answers that question before it is asked.

The engagement finding matters as much as the outcome finding, because the two reinforce each other. Patients who stay engaged report their status more consistently, which gives clinicians earlier signals when recovery stalls and more chances to adjust the plan before a setback becomes a plateau. The monitoring loop and the outcome gain are connected, and the study shows both moving together rather than one standing in for the other.

One study does not settle a clinical question on its own, and the retrospective design has real limits. Retrospective work cannot randomize patients the way a prospective trial can, so residual differences between the groups may still influence the result. Read honestly, the study is a strong signal that pairing RTM with hands-on care produces measurable gains, and it gives clinics a defensible reason to build outcome collection into their programs rather than treating monitoring as an administrative add-on.

Choosing the right outcome measure for MSK RTM programs

No single scale fits every musculoskeletal patient, and the choice of measure shapes what your RTM data can actually prove. A validated PROM works because it captures the construct that matters for a specific condition and care phase. Pick the wrong one, and you collect numbers that move without telling you whether the patient improved in ways that matter to them.

Start with a general pain measure for nearly every intake. The Numeric Pain Rating Scale (NPRS) asks the patient to rate pain from 0 to 10, and its simplicity is the reason it travels across body regions and diagnoses. Use it as a baseline anchor at intake and repeat it at every check-in, since a two-point change on the NPRS is widely treated as clinically meaningful and gives you a consistent thread through the whole episode.

Pain alone rarely tells the full story, so layer a disability index on top for conditions where function drives the treatment plan. The Oswestry Disability Index (ODI) quantifies how low back pain limits daily activities like sitting, lifting, and walking, and it separates the patient who reports high pain but functions well from the one whose life has narrowed around the injury. Collect the ODI at intake, again mid-course to confirm the plan is working, and at discharge to document the change you are billing against.

For extremity injuries, a region-specific functional tool captures recovery that general scales miss. The DASH measures upper-limb disability across shoulder, elbow, wrist, and hand, while the Lower Extremity Functional Scale (LEFS) tracks hip, knee, ankle, and foot function through tasks like squatting and climbing stairs. These tools respond to the specific movements you are rehabilitating, so they detect progress a broad pain score would flatten out.

Match the measure to the care phase, not just the diagnosis. At intake, you need a baseline detailed enough to prove change later, which usually means pairing a pain scale with the relevant functional index. Mid-course, the same measures confirm whether the current plan earns another few weeks or needs revision. At discharge, the delta between baseline and final scores becomes the outcome you report to referring physicians and defend to payers.

Running this well means administering the right instrument on schedule and scoring it consistently, which is hard to sustain on paper across a caseload. Platforms like Physitrack build validated PROMs libraries and outcome analytics directly into the remote therapeutic monitoring workflow, so the correct measure fires at the right interval and the scores populate a trend you can act on. That same PROMs library and analytics have been used in published clinical studies, which is a reasonable bar for the tools you rely on to prove clinical value.

Turning PROM data into defensible billing and referral credibility

Defensible documentation under 98975 and 98977 rests on a paired record that links what a patient did to a validated measure of whether it worked. An adherence-only log shows session completion, exercise minutes, and login streaks. An auditor reviewing that log sees engagement and nothing else. A record built on repeated PROM collection shows a baseline Oswestry Disability Index score at intake, a mid-course score, and a discharge score, each timestamped and tied to a clinical decision you documented in response. That paired record answers the question a payer actually asks, which is whether the monitored care changed the patient's condition.

The distinction matters at the code level. 98975 covers initial setup and patient education, and a clean record shows you established the device or platform, oriented the patient, and captured a baseline measure. 98977 covers ongoing data collection over a 30-day period, and it holds up when the collected data includes patient-reported outcome scores you reviewed, not just a tally of completed sessions. Regular PROM entries give each billing period a defensible clinical anchor instead of a raw activity count that any reviewer can dismiss as automated.

That same record becomes leverage outside the audit context. When you send a discharge summary to a referring physician, a documented drop in an LEFS or DASH score tells that physician the referral produced a functional gain. Claims data and completion logs cannot make that case, because they never measured the patient's function. A surgeon deciding where to send post-operative patients trusts a clinic that reports outcomes, and outcome data earns repeat referrals in a way engagement metrics never will.

The same evidence strengthens your position with payers. As commercial plans and Medicare tie more reimbursement to demonstrated quality, a clinic that can show aggregate PROM improvement across a patient panel walks into contract and appeal conversations with proof rather than assertion. You can point to the share of patients who reached a minimal clinically important difference, segment results by condition, and defend a denied claim with the specific score changes that justified continued monitoring.

Consistent PROM collection turns RTM documentation from a compliance record into an asset. It satisfies the auditor reviewing 98975 and 98977, and it gives you the outcome evidence that referral sources and payers weigh when they decide who to trust. Adherence data alone does neither.

Where CMS and payers are heading on outcome-linked reimbursement

CMS has spent the last decade moving reimbursement away from paying for activity and toward paying for results. The Merit-based Incentive Payment System already ties a share of clinician payment to quality measures, and patient-reported outcome measures sit at the center of how CMS defines quality in rehabilitation. RTM codes fit that direction because they were written to capture outcome data, not just contact hours. A program that collects validated PROMs today already produces the kind of evidence future payer models will ask for.

Adherence-only monitoring runs the opposite way. Login streaks and completed-exercise counts prove a patient engaged with a plan, but they say nothing about whether pain fell or function returned. When a payer moves to reward measured improvement, an engagement log has nothing to offer. A clinic that built its RTM program around device metrics alone will find itself rebuilding documentation from scratch, while a clinic collecting NPRS, ODI, or LEFS scores at intake, mid-course, and discharge already holds the record payers want.

Commercial payers tend to follow CMS on outcome-linked design, so the shift will not stay confined to Medicare. Bundled payment arrangements and value-based contracts increasingly ask physical therapy providers to show functional change over an episode of care, and PROM data answers that question directly. Platforms with built-in validated PROMs libraries and analytics, including those used in published clinical studies, let clinicians capture that data as part of routine RTM rather than as a separate reporting burden.

Outcome measures are the clinical backbone of RTM, not a compliance box to check. Adherence data tells you a patient showed up. PROM data tells you whether the care worked, and that is the evidence every payer is moving to reward.