Which Medicare Care Management Programs Should Your Practice Actually Run?

Most of the practices I talk to aren’t deciding whether to run a Medicare care management program. They’re deciding whether to add a second one, and that’s a different question with a much worse answer available online.

The first program is straightforward. One workflow, one consent conversation, one clock to watch. Your care coordinator learns it in a few weeks and the billing team stops asking questions by month three. The second program is where it gets difficult, because now the same patient can sit inside two sets of documentation rules at once, and the people who have to keep them apart are the people you already have.

What usually stalls that decision isn’t capability. It’s that nobody inside the practice agrees on which program should come next, and the material available to settle it is mostly written for somebody else. Search for this and you’ll get health plans describing benefits to their members, a state Medicaid manual, and at least one page about how to become a care manager.

This is the version I’d give a practice that has one program running and is looking at the next one. It carries no code table and no fee schedule, because those belong on each program’s own page and I’ve linked them where they matter.

Everything here is fee-for-service Medicare. If a large share of your panel is in Medicare Advantage, the plan sets its own care management rules and none of this transfers to those patients.

If you want the short answer to the title: run the Annual Wellness Visit properly first, because it tells you who qualifies for everything else. Add Chronic Care Management or Advanced Primary Care Management second, depending on whether your team can sustain minute tracking. Add Transitional Care Management third if you have a reliable discharge feed. Treat the remaining four as situational. The rest of this piece is why.

Start with the source, because it explains why the reading is confusing. CMS maintains a page listing the care management services it pays for under the Physician Fee Schedule. It’s short, and it names five: Advance Care Planning, Advanced Primary Care Management, Behavioral Health Integration, Chronic Care Management, and Transitional Care Management.

Remote Physiologic Monitoring, Remote Therapeutic Monitoring, the Annual Wellness Visit, and Principal Care Management are all real, all billable, and none of them is on that page. They sit elsewhere in the fee schedule, under monitoring and preventive services rather than care management.

The overlap is partial in both directions, which is worth being precise about. Four of CMS’s five are on every practice’s evaluation list. The fifth, Advance Care Planning, is a conversation billed per encounter rather than a program you enroll a panel into, so it sits outside the eight below and has its own page. Four programs that aren’t on CMS’s list are on the evaluation list. That’s how five becomes eight.

This matters more than a filing quirk. When a vendor says its platform covers care management, it may mean CMS’s five or the eight practices actually evaluate, and those are different products with different staffing needs. When your billing lead searches for guidance and lands on something that omits monitoring entirely, that’s why. The set has never been defined the same way twice. The definition worth planning around is every program your patients are eligible for that your staff could realistically deliver.

The Eight Programs, And Who Each One Is Actually For

Here’s the whole set, one line each, so you can rule most of them out quickly. Each row links to the page carrying that program’s codes, thresholds and documentation rules, because those change annually and belong somewhere they get maintained..

A few of these deserve a note.

1. Chronic Care Management is the one most practices start with and the one most written about. It has its own program build guide, and everything about how to stand one up lives there rather than here.
2. Advanced Primary Care Management is the newest, and it’s the one people most often describe incorrectly. It’s common to read that only primary care providers can bill it. CMS says the opposite in the rule text: “We are not limiting APCM services to practitioners in specific specialties” (89 FR 97867), and it names obstetrics and cardiology as examples. The gate is whether the practitioner is responsible for the patient’s primary care and acts as the focal point, not what their specialty says.
3. Transitional Care Management is the one with a clock attached, and the one where claims most often die on a contact window rather than on the care itself. The 30-day episode and the contact requirements are on its own page.
4. Remote Therapeutic Monitoring is the one most often confused with its physiologic sibling, and the distinction is the data type rather than the device.
5. The Annual Wellness Visit is the one that pays for itself twice. It’s the visit where you find out which of the other seven a patient qualifies for, which is why practices that treat it as a standalone preventive line item tend to under-enroll everything else. Its mechanics are in the Annual Wellness Visit guide.

For the two that sit closest together in practice, behavioral health integration and the collaborative care variant alongside it, the distinction is staffing rather than intensity, and the codes page walks through it.

This is also where risk stratification enters, and where I’d send you elsewhere. The AAFP’s care management page covers how to rank a panel by risk before you enroll anyone, and it does it well. That work sits upstream of this decision. Sort the patients first, then pick the program that matches what the top of that list actually needs.

Running Advanced Primary Care Management Alongside Something Else Depends On Which Clinician Bills

This is the question every practice asks second, and the published answers keep attaching it to the wrong subject.

What you’ll read, on vendor blogs and in the AI summaries built on top of them, is that Advanced Primary Care Management cannot be billed alongside Chronic Care Management, Principal Care Management, or Transitional Care Management for the same patient in the same month. Stated that way, at the level of the patient or the practice, it makes a multi-clinician group look far more constrained than it is.

CMS did consider that version. It proposed restricting these codes at the practice level and then didn’t finalize it. The restriction as written binds the individual practitioner. CMS put the point beyond argument in the rule text itself: “we are not finalizing the concurrent billing restrictions, except with respect to the one practitioner who is furnishing APCM services” (89 FR 97896). The services it is talking about sit in the rule’s Table 26, and they include Chronic Care Management, Principal Care Management and Transitional Care Management. Its worked example on the same page is a practice with two clinicians: “an oncologist could primarily manage care, including providing TCM services, for a patient who is recently discharged after an admission related to chemotherapy side effects while another practitioner in the same practice could appropriately continue to furnish APCM services for the same patient during the same month.”

For a solo practitioner those two readings are identical. For a group of eight they aren’t close, and the gap is revenue you’re entitled to and aren’t claiming.

Two caveats worth holding. Only one practitioner may bill Advanced Primary Care Management for a given patient in a given calendar month (89 FR 97868). And the pairs above are the Advanced Primary Care Management rules specifically; every program pair has its own, and some of the widely repeated ones are wrong in the other direction. The full pairwise picture is in the program stacking rules guide, and I’d rather link it than reproduce half of it here. If the pair you care about is remote monitoring alongside chronic care management, the revenue math for that specific combination is modeled separately.

What Running One Costs You In Staff Hours

The revenue side of these programs is published everywhere. The staffing side isn’t, and it’s the number that actually decides whether a second program works.

Seven of the eight split into three shapes, each doing something different to a coordinator’s week. The Annual Wellness Visit sits outside them as a single scheduled visit per patient per year.

1. Time-tracked programs require documented minutes per patient per month. The care is usually not the hard part. The documentation is, because a month where the work happened and the minutes were not logged correctly is a month you cannot bill, and your coordinator finds out at the end of it.
2. Bundled programs pay a flat monthly amount per enrolled patient with no minute threshold, which removes the timer from the coordinator’s day and replaces it with a different discipline: keeping enrollment and tier status accurate as patients’ conditions change.
3. Event-triggered programs only run when something happens, so they cost nothing on a quiet week and everything on a busy one. Transitional Care Management is the clearest case. The work is bounded and predictable per episode, but it starts when a discharge lands rather than when your staffing allows, and the clock doesn’t pause.

The question I get next is always the same one: how many patients can one coordinator carry? I don’t have a defensible number to give you, and neither does anyone else publishing on this, because it moves with program shape, panel acuity, how much of the documentation your system does automatically, and whether outreach is happening in-house. What I’d do instead is measure it on the program you already run. Take one coordinator, one month, and the enrolled patients they actually closed documentation on. That ratio is worth more than any published benchmark, because it’s measured on your panel and your systems.

Whichever way that ratio comes out, it tells you what your team can absorb before you commit to a second program.

Work out your fee-for-service share before you size any of this, for the reason noted at the top: a practice that’s two-thirds Medicare Advantage is planning against a third of the panel it thinks it has.

The Three You Probably Should Not Add Next

Almost everything published on this topic is additive. Here’s the other half, and these are judgment calls rather than rules, so treat them as a starting position rather than a verdict.

1. Remote Therapeutic Monitoring, if you’re not already running therapy services. It’s a genuinely good program in the right practice, and the right practice usually has physical therapy, occupational therapy, or speech pathology in it already. Adding it purely because it’s adjacent to remote monitoring means buying devices and building an alert workflow for a patient population you don’t really have.
2. Principal Care Management, if you already run Chronic Care Management well. These two overlap heavily in workflow and differ mainly in which patients qualify. The realistic gain is a handful of single-condition patients your existing program cannot cover. That’s worth doing eventually and it’s rarely worth doing second.
3. Collaborative care, if you don’t have a psychiatric consultant lined up. The model requires a behavioral health care manager working cases and a consulting psychiatric provider reviewing them on a set cadence. Practices that commit before securing the consultant end up delivering the lighter behavioral health integration service and billing for something they aren’t structured to provide, which is the worst of both.

The general test behind that ordering: add the program your current workflow already half-does. Published rate and vendor module availability both make poor tiebreakers. If the honest answer to “who does this work today” is nobody, you’re making a hiring decision and calling it a billing decision.

What Changed In 2026

Two things landed this year that most guidance has not caught up with, and both matter to primary care specifically.

Behavioral health add-ons for Advanced Primary Care Management. CMS finalized three optional codes, G0568, G0569, and G0570, crosswalked from the existing collaborative care and behavioral health integration codes (90 FR 49469-49471). They carry no time threshold, and they’re billable only when an Advanced Primary Care Management base code is reported by the same practitioner in the same month. For a primary care practice already on the bundle, this is the cheapest behavioral health entry point currently available. Removing the minute threshold doesn’t remove the documentation requirement; CMS declined a request to make these attestation-only (90 FR 49470).

The ACCESS Model. CMS launched ACCESS, Advancing Chronic Care with Effective, Scalable Solutions, on 5 July 2026. It’s a ten-year Innovation Center model testing outcome-aligned payments in Original Medicare for technology-supported chronic care, aimed at hypertension, diabetes, chronic musculoskeletal pain, and depression. Rolling start dates follow; as of early August 2026 the next two were 17 August and 1 October 2026, so check the model page for the current cohort before planning around one. If you’re weighing a monitoring-heavy program build, understand how this model would interact with it before you commit, and read the Request for Applications rather than a summary of it. The interaction between model participation and existing fee-for-service care management billing is the part to get advice on, and it’s specific enough to your situation that a blog post is the wrong place to settle it.

That November note isn’t a throwaway. Rates get checked every year because they’re easy to check. Descriptors don’t, and a code that quietly starts meaning something else is far more damaging than one that moves a few percent, because the workflow around it keeps running as though nothing happened.

The clearest way I have seen this decision laid out was not written for Medicare at all. AHRQ’s guidance on Medicaid care management programs sorts the options into buy, build, and assemble, and walks through the staffing, data, and monitoring capacity each one demands. The program rules are Medicaid’s and don’t transfer, but the selection logic does, and it’s more honest than most of what’s written on the Medicare side.

In practice the three options look like this.

1. Rent a platform. Fast, generic, and the vendor owns the logic. What to look for if this is the route you take is covered in the care management software guide. This is the right answer for a single program where you want it running next quarter and you’re content to work inside somebody else’s configuration. The trade is that the workflow is tuned to a median customer rather than your panel, and your team logs into a second system beside the EHR.
2. Configure a pre-built use case. Most of the work in these programs is the same across practices, which is why we deliver it through ConnectHealth as pre-built use cases that get configured rather than written from scratch. Post-discharge follow-up is the clearest example: the discharge feed, the medication reconciliation with write-back into the record, and the readmission monitoring are all built. What gets configured is the part that’s genuinely yours, which is the two-business-day deadline timer, your complexity rules, and the audit trail your payers ask for.
3. Build it. Some of this doesn’t come pre-built, and the concurrency logic is the plain example. Enforcing which programs a given patient can be in this month, given who billed what, is a rules engine. Wiring systems together does not deliver it. If that enforcement is what you want, it’s a custom build, and worth asking any vendor to demonstrate live before you sign.

Which one is right depends on how many programs you’re running and how much of the logic sits between them rather than inside any one of them. One program, rent it. Several programs with patients moving between them, the logic between the programs is the part worth owning.