How PTs and OTs Win in Value-Based Care

Your Playbook to Navigate, Partner, and Succeed in New CMS Innovation Center Models

This is blog 3 of 3 in our series on the future of PT and OT payment. Blog 1 defined the problem. Blog 2 handed you the advocacy levers to change the current environment. This one shows you where the whole system is heading and how to move with it.

The U.S. healthcare system is changing. And while change can be hard, it also creates opportunity. For PTs and OTs, that opportunity could be huge. But it is also fleeting.

I know it seems like we have been talking about value-based care for years. I went into great detail on the history of PT and OT payments in the first piece of this series, but the gist is that it’s been a long and winding road to the current state of our reimbursements.

While I’m a firm believer in the power of advocacy (and provided several levers the rehab professions can collectively “pull” in my second piece in the series), I also know the system is moving in this direction with or without us, and it’s crucial that we move with it.

Case in point: The Innovation Center models, created under the Affordable Care Act, are currently testing new ways to deliver and pay for care. These models are changing what actually gets paid for, creating a massive open door for therapy. But, it won’t stay open forever.

You’ll leave this blog with three things:

1. A way to read any model, because models change. They get added, amended, and retired, and memorizing lists makes no sense (though we’ve built a table for you to keep as a resource).

2. A shortlist of models that matter most for PT and OT today.

3. One habit that will keep you current in the years ahead.

This piece is about more than policy and payment, even though both shape everything we're going to talk about. This is about your paycheck. It’s about the work

Why You Should Care About These Models

In fee-for-service, therapy is a de facto cost line for payers and those responsible for financial resourcing in healthcare. We’re a unit to be counted (or denied if the value isn’t obvious), and our services are rarely seen as something that prevents bigger costs down the road.

So, we get squeezed. Hard. The outpatient side of fee-for-service is essentially a treadmill of units to bill. The inpatient side is a race to optimize each therapist’s productivity, since we’re part of the conditions of participation for those sites of care. We are the ones with the most to gain when the math flips, which is exactly what these new models do.

When a payment model changes, it can change whether a particular type of care is seen as an investment in the model’s success (e.g., by supporting better patient outcomes at a lower cost) versus an expense to be minimized. In value-based care models, therapy becomes a savings engine instead of a cost center. The catch is that the people shouldering the risk in these models (e.g., policymakers, payers, and other providers), haven’t all figured that out yet. But, I believe we can show them. And we already have several ways to do that.

To that end, here are the currencies we currently hold. These are things we do every day that have no real reimbursement mechanism in fee-for-service, but align perfectly with the incentives inherent to value-based models.

The first is access. We extend the overwhelmed primary care sector. We can be the front door to healthcare for function, falls, movement, and musculoskeletal issues in a whole-person environment. Right now, a lot of those referrals go nowhere. (Go back to the “missing leg” section in blog two for more on this.)

The second is avoiding high-cost care. We interrupt a chain of events that would otherwise keep going. Falls prevention reduces fractures, ED visits, expensive surgeries, hospitalizations, costly post-acute stays, and risky transitions of care. Early therapy helps patients avoid imaging, opioids, injections, and surgery. We play a real role in preventing readmissions. You already know all of this, but the takeaway is that in value-based care, these levers carry weight, and you will need to point them out in that specific context.

The third is function and independence. Aging in place is an imperative now. Keeping people home (and out of both hospitals and long-term care) matters enormously, because that kind of long-term assistance is expensive and often creates real financial hardship for families and the healthcare delivery system. The medical and behavioral health providers who commonly sit on advanced primary care teams don’t routinely produce tangible value in this area. That’s okay! That’s not their role. We do; it is our role. We have the expertise.

For us, this doesn’t just translate to better pay. It also means the work itself gets more satisfying, because value-based care models pay for things fee-for-service has no code for:

• More time with patients to build trust

• Clinical judgment

• Care coordination

• Managing and nurturing a longitudinal relationship

• Behavior change

• Patient education

In fee-for-service, the job is counting units and hitting a productivity percentage. It’s part of an old culture that traces straight back to the structure we covered in blog one, built when third-party payers first showed up. Structures can become extinct over time.

The work that’s burning out PTs and OTs, the work that pushes people to leave clinical care, is the work fee-for-service rewards. Value-based care is a way out of that. When therapists work in transformed systems, the work matches what they were trained to do, and that leads to greater satisfaction as a clinician. You become more tied to the whole care team, and the value of every clinician on it goes up.

Now, here’s how that raises the salary ceiling for therapists over time.

Salaries follow revenue models. If you work in outpatient therapy, you’re generally paid against an expectation that you’ll maintain a certain volume of billable units, maybe with a bonus if you exceed it. In fee-for-service, your ceiling is the conversion factor and the value of the RVUs assigned to therapy services. Your revenue is directly tied to what you can deliver in X amount of time. (We explained where that’s headed in the first two blogs, and while there’s real work we can and should do on the fee-for-service side, it’s a limiting factor.)

In value-based care, when you can show that your work with patients creates savings, you have negotiating leverage with the people who own the risk for the spend. That leverage doesn’t exist in a unit-based world. You become worth more to the system because you save it more (or because you help it avoid higher-cost care that often wasn’t a great experience for the patient anyway).

To capture that value, though, you first have to understand the models. You don’t have to memorize them, but you do have to learn how to read them.

How to Read Any Model So You Never Fall Behind

This takes a mindset shift. Start by understanding the categories of models, and learn to contrast each category against fee-for-service.

The Innovation Center has launched dozens of models over the last couple of years. You probably can’t memorize all the acronyms, and the good news is that you don’t need to. The skill is reading new models, changes to models, and evaluations of models, and that starts with a simple question: what category does this fit into? Then: who holds the financial risk for the model’s success? And specific to us: where do the therapy dollars, or the shared savings that could flow to therapy, hide in the model?

I’ll commit to you that four categories cover most of what you’ll ever evaluate. Not quite everything, but most of it, and that’s fine.

Here’s why these categories exist at all. Every model is trying to answer the same question: how do we stop paying for volume and start paying for value? Another way to ask it is: what would the structure of reimbursement have to look like, and what would the incentives have to be, for outcomes to improve and for clinicians to be rewarded for changing how they deliver care? The four categories are really just different units of accountability, and sometimes a different answer to who is accountable.

Accountable Care Organizations put accountability on a population over time (specifically, on the total cost and quality of care for that population on an annual basis).

Acute episodes of care (what people often call bundles) put accountability on entity (like a hospital or group of physicians, to manage the cost and quality of an acute episode of care. Event-triggered episodes (like an inpatient or outpatient surgery) often include a recovery window ranging from 30 to 90 days depending on the model.

Condition-specific models put accountability on a clinician. This clinician often is not a primary care provider, but the provider the patient thinks of as their main doctor because of a specific disease (e.g., cancer, end-stage renal disease, or dementia).

Novel models test a new mechanism itself. Examples include tech-enabled care, using artificial intelligence to standardize assessments, and using technology to run prior authorization.

Each one answers the same two questions in different ways. What should we hold a provider responsible for? And what are we trying to change by changing the incentives?

Models compound over time

This is important to hold onto. Models of care are often next generations of models tested in the past. The Innovation Center tests a model, learns from it through the various participants, and then extends it, retires it, pulls components into other models or into fee-for-service, or builds a second generation. They roll the lessons forward. And what we’re starting to see is that the next generation gets bigger in terms of who participates (and sometimes it becomes mandatory).

Look at the lineages.

On the acute episode side, we had BPCI, the Bundled Payments for Care Improvement model. Then Comprehensive Care for Joint Replacement (CJR). Then BPCI-A, while CJR became mandatory in parts of the country. Out of all of that we now have the Transforming Episode Accountability Model (TEAM), plus the CJR-X model, which as of this writing is in proposed-rule status and expected to be finalized. So acute episodes are going mandatory, and they’re folding in outpatient episodes too, leaning hard into site-neutral policy.

On the ACO side, The Medicare Shared Savings Program (MSSP) is the permanent ACO program in statute. But inside the Innovation Center, we’ve kept testing more advanced versions. It started with the Pioneer ACO model, then Next Generation ACO. The ACO REACH model is running right now and ends at the end of this year. Then comes the Long-term Enhanced ACO Design (LEAD) model, starting January 1, 2027. Each generation tests more flexibilities, adds the learnings, and adjusts the incentives to get more participation and better results.

Remember, this only started in 2012. It hasn’t been going on long. We know from areas like pharmaceuticals how long testing and development takes, and here we’re trying to change a payment and delivery structure that’s been in place for about a century.

Two things make all of this matter more than it looks. Models that prove savings at equal or better quality can be expanded nationwide by the Secretary of Health and Human Services (HHS), with no act of Congress. That comes from the 1115A authority. And mandatory models are CMS telling you it has strong conviction that there’s a better way to deliver care in a specific area. When a model goes mandatory, it’s hard to miss that permanence is on the horizon. TEAM and CJR-X are good examples.

How to stay current

So how do you keep up with models that are announced, changed, evaluated, or updated? It’s straightforward.

Sign up for the CMS Innovation Center email updates and subscribe to both general and model-specific updates. There are a lot of options, so subscribe broadly and unsubscribe later from what you don’t want. Subscribe to the general CMS.gov updates too. And follow the APTA and AOTA regulatory and policy briefings, which translate new model releases, fee schedule changes, and Medicare Shared Savings Program updates that increasingly carry more innovation inside them as this whole trajectory moves forward.

Think of it this way: By the time a model becomes a continuing education course, the application window may already be closed. You want this in your inbox the day CMS posts it.

Now let’s walk the four categories and go a little deeper on one model in each that PTs and OTs should be watching.

Category One: ACOs, or Accountability for a Population

An ACO is a group of providers accountable for the total cost and quality of care for a defined population, what CMS calls attributed beneficiaries. Those are patients with a recent history of using that type of provider, or who voluntarily chose them as their provider.

CMS sets the ACO a spending benchmark. The inputs vary by model, but they can include risk-adjusted historical spend on the same or a parallel population, regional spend, and national spend. If the ACO beats that benchmark for the year while hitting its quality minimums, it shares the savings with CMS. In two-sided risk models, the ACO also owes money if its patients have poor quality outcomes or overspend the benchmark. So the team has to have a way to manage the highest-cost areas, like hospitalizations, post-acute care, readmissions, and unnecessary utilization.

Here’s the mental model. The ACO wins when patients stay healthy and out of the hospital and out of institutions. That is exactly what therapy aims to produce and what we’re capable of helping produce.

The Medicare Shared Savings Program is the permanent backbone of accountable care. It started in 2012, it’s the only permanent ACO program, and it covers a huge swath of Medicare ACO lives, a number that keeps growing. About half of traditional Medicare beneficiaries are now in an ACO.

You should know about MSSP because you’re very likely already treating patients attributed to one without knowing it. Depending on how long the ACO has participated and how seriously it mines its data, it can track whether your care helps it beat its benchmark, and it can compare your impact against your competitors. So the takeaway is to find out which ACOs are participating in your region. There’s a published list of MSSP and ACO REACH participants (and soon LEAD), and from there you can start to understand the impact you may or may not be having on the patients that ACO is responsible for.

ACO REACH tested capitated risk. It’s mostly provider-led organizations, with some health systems, and and it has a “Track” that handles high-needs and complex populations a little differently. It ends at the end of this year. But the lessons from REACH are the foundation for what replaces it. That’s the compounding concept in action.

What replaces it is the LEAD model, the Long-term Enhanced ACO Design model, and this is the ACO I want you to pay attention to. It’s the longest model CMS has ever tested, a full 10 years. That length matters, because the benchmark gets set and holds for 10 years, instead of dropping every cycle as providers succeed at lowering total cost of care. It’s a stable runway to actually transform care and see it pay off. I wouldn’t be surprised if LEAD eventually leads to a permanent structure in statute. It includes global risk, where you’re on the hook for up to 100% of savings and losses, and professional risk, up to 50%.

I want therapists to look at this model and think, this one is for me to participate in directly.

LEAD is focused on practices new to ACOs. It incentivizes care for dually eligible patients, homebound and home-limited patients, rural patients, and underserved patients. There are even add-on payments for some new participants to help them build the infrastructure to run an ACO.

LEAD also carries a structure we’re seeing for the first time, the CMS Administered Risk Arrangements, or CARA. CARA supports episode-based contracting with other providers, and it’s different from the acute episodes we’ll cover next, because a CARA episode can run a year, for example, for managing end-stage heart failure, or musculoskeletal and low back pain patients.

A Real Example: CARA’s Falls Prevention Episode

If all of this still feels abstract, here’s a concrete example:

Starting in 2028, CMS plans to launch the CMS-Administered Risk Arrangements (CARA) initiative as part of the LEAD model. One of CARA’s first episodes is a falls-prevention program called RISE (Resilience and Independence in a Safe Environment).

Read that again.

One of the first episode-based arrangements CMS chose to support isn’t a surgery. It isn’t a hospital stay. It isn’t a drug. It’s preventing falls so older adults can remain safely at home.

The episode includes interventions such as falls-risk assessments, strength and balance training, home safety interventions, communication plans, and functional support delivered by interdisciplinary teams that include occupational and physical therapists and registered nurses. The model is explicitly designed around reducing acute events and helping beneficiaries maintain independence in their homes.

That’s important for two reasons.

First, it validates something therapists have known for decades: preserving function and preventing decline creates enormous value for patients and the healthcare system.

Second, it creates a business opportunity.

The therapists who learn these models will identify the ACOs operating in their markets, learn who is building preferred-provider networks, and start conversations now. They’ll be asking questions like:

• How are you approaching CARA?

• Who owns your falls-prevention strategy?

• Do you have therapy partners capable of managing these patients?

• How are you planning to identify and intervene with high-risk beneficiaries?

Some therapy organizations will become part of these arrangements.

Unfortunately, others will discover that the contracts, relationships, and networks were formed without them.

That’s the bigger lesson behind CARA. The future of therapy participation in value-based care won’t be determined only by clinical skill. It will be determined by who understands where the payment models are going and gets in the room while they’re still being built.

Walk Through the Door

This is where we have an open door. ACOs have a lot on their plate, and they’re going to decide which specialty and post-acute providers to bring into these arrangements. CMS built CARA to make that simpler for the ACO. You want to be fluent in how it works, and you want to be one of the providers contracted in. Some therapists will figure this out, and the ones who don’t will miss the upside and the chance to practice in this kind of transformed care.

There’s also the chance to become a formal Preferred Provider under contract with the ACO, which can let you participate in capitation plus upside for the outcomes you influence. There are a lot of ways to play this. The point is to understand how the model works, because this is the clearest example in this piece of getting in the room while things are happening, so you can be part of what’s being built.

Category Two: Episodes, or Accountability for an Event

An episode of care is structured like this. CMS sets a target price for a clinical event plus a recovery window. The event can happen in the hospital or in the hospital outpatient department. The episode includes the cost of the DRG in the hospital, or the APC payment in the outpatient department, plus the spend CMS anticipates over the recovery window, usually 30 or 90 days. Whoever holds the risk benefits or pays the price depending on whether total spend comes in under or over target.

To be clear, this isn’t one lump payment handed to a participant to manage. Everyone still bills for themselves, and CMS reconciles the result of the episode afterward, holding the participant responsible for the spend and the quality. Most of what can be influenced is the site of care a patient goes to after the event (inpatient, home health, or outpatient) and how long they stay if it’s skilled nursing.

Post-acute care is our home turf. We know a lot about managing what happens after an event. If you’ve worked in almost any setting, this is familiar: something happened to a patient, then they recovered, and we were a big part of that recovery.

BPCI-A is the most recent episode model, and it’s concluded. It ended December 31, 2025, the second of two 90-day bundle models. So CMS now has about a decade of experience and knows two things for certain. Episodes can save money. And discharge disposition is what drives the margin. What I saw on the ground lines up with exactly that.

Now we have TEAM, and it’s live right now. This first year you can opt into no downside risk, so a hospital that wanted to delay big changes may have done that, but they know changes are coming if they don’t want to pay money back. There are five surgical episode categories: lower extremity joint replacement, surgical hip and femur fracture, spinal fusion, coronary artery bypass graft, and major bowel procedure. There’s a 30-day post-discharge accountability window, and the hospital is accountable for the downstream spend in that window.

And it’s mandatory. Roughly a quarter of hospital markets across the country were selected, with no opt-out. CMS isn’t asking, it’s telling.

In TEAM, we may be the most influential clinician on whether a bundle makes margin for the hospital or becomes a cost. Early mobilization. Pre-episode education where it’s possible. Discharging to home instead of an inpatient setting. Adherence to home programs and protocols. Readmission prevention. Caregiver training. These are things therapists often take the lead on.

If your hospital is a TEAM hospital, your discharge decisions are financial decisions. When you say a patient needs acute rehab, you are determining the financial impact of that model, even if nobody’s told you. I’m telling you. So if you don’t know, find out. If you work in a hospital, find out if it’s in TEAM. If you work in a post-acute setting, find out if the hospitals in your area are mandated into TEAM. If you work in outpatient therapy, same thing. That’s your first step.

Then, starting October 1, 2027, we have the third generation of the joint replacement model, the expanded version, CJR-X. It isn’t finalized yet. It was proposed in the fiscal year 2027 Inpatient Prospective Payment System rule. If it’s finalized as proposed, it’s mandatory for about 2,500 hospitals, and it has no end date, which makes it permanent in practice. It includes hip, knee, and ankle replacements like its predecessors, with both inpatient and outpatient triggers, and it keeps the 90-day episode. TEAM hospitals are excluded from CJR-X, but only until TEAM ends in 2030, if it ends then. More compounding.

CJR-X matters because the indefinite duration tells you CMS is treating mandatory joint replacement bundles as the new normal. Our leverage points span the entire episode: prehab, day-of-surgery mobilization, discharge-to-home rates, SNF avoidance, SNF length of stay, home health coordination, outpatient therapy at home, outpatient therapy. We touch all of it.

And here’s the part I want to be clear about, because it’s easy to get this backwards. In a bundle, the easy thing to do is spend the most. Send the patient to the most intensive, most expensive setting and check the box. That’s not value, that’s just spend. Value is figuring out what a patient actually needs and getting them prepared to go there at the right cost, so they get the most out of their insurance dollars. That is exactly the question we are trained to answer.

What does this patient need from a medical lens? Who in their life needs to be trained to help them once they leave? What other support do they need at home? Can this person make real gains with therapy three to five days a week that builds in the mobility work they need to go home instead of to a facility?

And when someone says a patient is “unsafe” to go home, it’s worth asking honestly, were they unsafe before this surgery or this hospitalization? Will anything make them safe to be alone? That’s the skill of the therapy prognoses. We are the only ones who can do that. We’re also the clinicians who should actually know what patients’ real experience in a skilled nursing facility versus an inpatient rehab unit versus their own home, because we’ve worked in those settings and watched it firsthand. If we’re not in the room, that call gets made by someone who hasn’t. When a PT or OT learns this model, we help make sure patients get what they truly need, and we help the hospital succeed at the same time. Those two things are not in conflict.

Category Three: Condition-Specific Models, or Accountability for a Diagnosis

These put accountability on a patient with a specific diagnosis, someone who, at least for a stretch of time, isn’t seeing a primary care provider as their main doctor but is seeing another physician type. Some conditions are so costly and so driven by a specific specialist that CMS builds the whole model around them.

Look at the pattern, because this is the skill. Nearly every condition-specific model contains a functional health gap, and that gap is therapy-shaped. Learn to spot it.

The first one to know is the GUIDE model, Guiding an Improved Dementia Experience, running July 2024 to June 2032. This is the first time CMMI explicitly included therapists as part of the care team in the examples it gave for how the model is structured. It pays a monthly per-beneficiary care management payment to dementia care programs, plus additional money for respite and caregiver support.

GUIDE is an easy place for OTs and PTs to show value: helping patients preserve function, setting up the home safely for someone with dementia, training caregivers (a key part of the model), fall prevention, and cognitive-functional integration. So GUIDE tells us something useful. CMS will name us in a model when the model is built around function and staying home. Try to map that in your head to wherever else that gap exists.

Two others, briefly. The Kidney Care Choices model (KCC) is winding down, built around nephrologist-led kidney contracting entities. It’s likely we’ll see a next-generation model after it ends, which is more compounding at play. Patients with end-stage renal disease frequent have a lot of deconditioning and fatigue, and there’s real fall risk in dialysis populations.

Then the Enhancing Oncology Model, July 2023 to June 2030, runs in six-month chemotherapy windows through oncology practices. Cancer rehab is a real thing, and cancer patients who lose function, safety, and mobility can have complications, readmissions, and ER visits. Those cost money, which is the responsibility of the oncology practice to manage.

So don’t wait to be invited into a condition model. Find the gap and bring the evidence to whoever holds the risk that you can help them close it.

Category Four: Novel Models, or Testing the Mechanism

The fourth category isn’t organized around a population, an event, or a disease. It’s testing a mechanism for managing outcomes and spend, and that mechanism is artificial intelligence. These models aren’t really about who you’re accountable for. They’re testing how accountability and payment can work when technology drives it.

There are two live models right now, and they point AI in opposite directions.

The Wasteful and Inappropriate Service Reduction (WISeR) model, , runs in six states from January 1, 2026 through December 31, 2031. The participants are technology companies, using AI and machine learning plus human clinician review to run prior authorization and prepayment review on about 17 select outpatient services that CMS considers overused and high in variability. A few things to know if you’re learning this model. Payment and coverage rules don’t change. Any denial has to be made by a clinician. Decisions come within 72 hours, or 48 if expedited. And a participant with a 90%-plus affirmation rate can earn an exemption from the process. There is controversy around WISeR. For those of you who don’t like prior authorization, I’d gently push you to stay open to the value of a second look at whether an intervention should be done at all.

Then there’s the Advancing Chronic Care with Effective, Scalable Solutions (ACCESS) model, a 10-year national voluntary model. The first cohort begins in July 2026, with rolling cohorts through 2033, and there’s a small control group built in so CMS can test the model the traditional way. Instead of total cost of care, ACCESS participants are paid condition-specific, outcomes-based payments tied to measurable clinical improvement for each patient. Generally, these participants use technology to keep patients engaged and to manage disease between in-person visits.

This is the first model built on the premise that technology can verify improvement, patient by patient, and pay on it directly. Outcomes that become measurable lead to payment. That’s the structure to watch, and it should excite PTs and OTs. The model’s tracks center on chronic disease, and there’s a musculoskeletal track where functional improvement is part of the outcome measures. The profession that shows up with validated functional measures and digital fluency will own that track, because ACCESS rewards exactly what we’re trained to do: produce a measurable functional improvement and get paid for the result.

There are a few things about ACCESS most therapists haven’t thought through, and they matter.

First, if you’re already billing for therapy services, you can’t also bill ACCESS for that same care. So this isn’t simply a new code to add to your clinic. Few therapy clinics are likely to apply to become ACCESS Orgs because of that overlap dilemma. Second, ACCESS can become the front door for care. Picture it positioned as the digital front door for therapy, where for some patients, the patient starts there and gets referred out to therapy only if the ACCESS provider decides they need it or the patient wants a different treatment path. Referring providers, physicians included, get paid for steering patients into ACCESS and co-managing alongside them. They can count on the mandated updates an ACCESS Org is required to provide digitally. There’s almost no friction in that handoff. So a patient who would have walked into your clinic may now go to ACCESS first, and you may or may not be who they get referred to afterward.

You can treat that as competition that quietly reroutes your patients, or you can partner and make it seamless for patients to get the care they actually want from you. That means building the tools patients increasingly expect into your own offering: remote therapeutic monitoring, asynchronous communication, telehealth visits, and a more catered, convenient experience. We are in the age of the empowered healthcare consumer, and I don’t think most of the field realizes how much that single shift is going to drive change.

Models at a glance

Note: CJR-X is a proposal as of this writing. The final rule is expected around August, and details may change.

AI Is Accelerating All of This

WISeR and ACCESS aren’t two odd models off in a corner. They’re the leading edge of something reshaping the whole landscape. Because AI doesn’t only change how we practice. It changes how the market sees care, including ours, and that second effect is what’s accelerating value-based care.

Start with AI inside the practice, because that transformation is coming no matter what. Think about the parts of your day that are the same every time. The note that says the same things. The standard exercise handout. The protocol you run for a given diagnosis. The scheduling, the authorization paperwork, the documentation. Those repetitive, predictable tasks are exactly what software is getting good at. In fee-for-service, a lot of what generates revenue is the volume of those repeatable visits and tasks. The work that’s hardest for software to touch is the judgment: reading a complex patient, adjusting a plan, building trust, coordinating with a family and a physician, getting someone to actually change their behavior. That harder work is what value-based care pays for. So the routine parts of our day are going to be automated either way. The real choice is whether we move toward the work only we can do, or get squeezed out of the repetitive work that barely paid us to begin with.

The part a lot of therapists haven’t considered is that AI makes healthcare legible. Price transparency data that’s been sitting dead in spreadsheets becomes a comparison tool anyone can query. Outcomes data becomes searchable, plain-language, and personal, which lets us look patient by patient and get to the root cause of outcomes and spend variation. Risk stratification, value measurement, and provider comparison get done by ACOs and payers at a speed fee-for-service billing was never built for.

Then there’s patient awareness. Picture a 67-year-old weighing a knee replacement against a structured therapy program that might let them avoid it. They can ask an AI assistant exactly that, and get an answer grounded in evidence, cost, recovery data, and function. So patients are about to be far more aware of their options, including skilled therapy as an alternative to imaging, injections, and surgery, and far more equipped and confident to act on it. That’s demand-side pressure for what we do, and it’s a good thing.

All of it accelerates value-based care, because value-based care’s bottlenecks include patient engagement and  measurement. Proving that an intervention actually produced an outcome used to take piles of data, teams of analysts, and years of claims runout. AI and more data availability through more sources, including that which digital health providers create, shrinks that to months, and eventually to real time. Once value is cheap and easy to see, the models that pay for value win by default, because the technology can finally identify what’s necessary, what’s valuable, and who can deliver it.

So here’s the plainest way I can put it. The “volume side” is the old world where you get paid per service, per unit, per visit. WISeR is AI pointed at that world. Its whole job is to question whether a service should happen or should have happened, to catch overuse, and to slow down or deny what is considered wasteful. If you live on the volume side, AI shows up as the thing scrutinizing your claims.

The “value side” is the world where you get paid for getting a patient a result. ACCESS is AI pointed at that world. Its job is to measure whether the patient actually got better, and to pay based on that. If you live on the value side, AI shows up as the thing that proves your work mattered and gets you paid for it. Same technology, two completely different jobs depending on which side you’re standing on. On one side it polices you. On the other side it pays you. That’s the choice, and ACCESS is what the value side looks like in model form. If we don’t move toward it, AI will simply be used to ration the volume side instead.

The Risk of Sitting This Out

This whole blog has been an invitation to learn, explore, and eventually participate. The other side of that invitation is a warning about what happens if you don’t.

Here are the ways we get excluded from these models.

Primary care teams get built without us. Behavioral health earned its embedded seat in total cost of care models. Pharmacy is starting to. Nurses and social workers are already on many advanced primary care teams. Those teams are being built without a functional health leg, and that becomes the template everyone copies once they see it work. LEAD networks will start forming, CARA arrangements begin in 2028, and the preferred-provider networks can form on an annual basis once organizations get their arms around them. The value-based-care-savvy practices are going to get in, start those conversations now, figure out what they need to win the contracts, and begin participating.

Bundles define what good therapy looks like without us in the room. The participants in TEAM and CJR-X are choosing their post-acute partners. If you work in a post-acute setting, and almost any setting is post-acute for some share of patients, you want to be part of shaping what that looks like, so it’s built around what patients actually need rather than around whatever’s cheapest or easiest.

And remember the patient-driven payment model. We covered it in blog one. SNF payment changed in 2019, and therapy staffing dropped within months. Part of what drove the creation of the Patient Driven Payment Model (PDPM) was that nearly every patient was being assessed as needing the highest-paying therapy levels, and CMS took that incentive away. We didn’t have the evidence ready to show that the volume of our services was needed to drive good outcomes. And therapists were trained in that setting that “more therapy is better,” in terms of minutes per week and total SNF days.

Payment models will change whether or not we’re ready, and not being ready means you can’t give input when the window is open, so you end up the victim of whatever gets shifted. Do not assume for a second that any stakeholder knows the value you bring, even the ones who should. It’s a harsh reality. But knowing it is the first step.

And AI is the accelerant under all of it. It’s shortening every timeline, which means the window to get in closes faster than it used to.

So let’s be hopeful, because this is a genuinely open door with real potential for PTs and OTs. We just have to look outside the therapy world and into the full continuum of care, and pay attention to how policy shifts through CMS and CMMI.

So What Do We Do?

Here’s what different people in therapy can do.

Every clinician: Learn the vocabulary. Total cost of care, attribution, benchmark, episode, two-sided risk. Find out now what models you or your employer already touches because of your geography, and be the first to know who’s participating in LEAD and in next year’s MSSP. Volunteer for the work that drives model performance: falls, readmissions, discharge planning, functional outcome collection. If you’re in a setting that’s participating, get involved. And go back to the alerts section and sign up for email notifications from CMS and CMMI.

Managers and directors: Build the internal data story, even with just your own data. Discharge-to-home rates. How you contribute to readmissions. Functional outcomes you can quantify. Map your market: who’s participating, in which models, with what historical success, and where the openings are to start an ACO or get in early as one forms with providers who aren’t participating yet but might want to partner. Then start the preferred-provider conversation and the episode-arrangement conversation early. Even if the participants aren’t quite ready, plant the seeds.

Owners and the profession: Pursue contracting, and make sure the profession sees these opportunities while the networks are forming, especially around post-acute networks, TEAM, the CARA episodes, and preferred-provider relationships with ACOs. Engage CMMI’s RFIs and stakeholder calls. They genuinely inform the direction and structure of models, and an RFI is often a signal that CMS is developing something that will affect you whether you know it or not. And learn what outcome measurement really looks like, not just the outcomes in your own setting, but across a whole episode beyond where you touch the patient. That’s the foundation under every ask when we want to participate.

A few things worth asking at work this month:

• Are our patients attributed to an ACO?

• Is my hospital in TEAM?

• Who owns our post-acute network strategy?

• Are we tracking functional outcomes in a way a payer would recognize?

• Have we looked into the LEAD model?

Conclusion

Therapy has been undervalued for decades because of how the payment system was built, and that’s largely what created the structure we’re suffering under today. But healthcare is being rebuilt in real time, and the way it gets paid for is changing with it. As these models compound, as the learnings compound, and as permanent change eventually arrives, those models should be paying for exactly the kind of work we’re capable of doing.

Blog one named the problem. Blog two handed you the levers to pull. Blog three showed you where everything’s heading, and a few things to do today. The dates on these models should create some urgency. What’s your role going to be with patients in ACCESS? How can you contribute to ACO success in LEAD? How can you help participants succeed in TEAM and CJR-X?

Therapists who practice in transformed systems describe work that matches their training, paid for their time, their judgment, their relationships, and their results. That future is available. It’s earned through participation. The status quo isn’t going to fix your pay, your satisfaction, the turnover, or the career ceilings.

So ask the questions, and get in the room.