Medicare Will Pay You $116 A Month Per Patient. Why Is Nobody Collecting It?

In a previous life I watched a Stripe dashboard cross a million dollars in a single day.

I remember where I was sitting. I remember thinking, this is it, we've done it, whatever it was.

The product was bad. Not catastrophically bad, just quietly bad in the way that shows up eighteen months later as churn and refunds and a support inbox nobody wants to open. Scale hides problems beautifully. It also taught me the thing I keep coming back to, which is that a number on a screen tells you almost nothing about whether the underlying machine works.

Care management in 2026 is the exact inverse of that, and it's driving me slightly mad.

The machine works. The money is real and CMS is actively trying to hand it over. The dashboard is empty.

What is Medicare actually paying for care management in 2026?

More than most practices realise. Under Advanced Primary Care Management, CMS pays a bundled monthly fee tiered by patient complexity: roughly $16 per patient per month at G0556, about $54 at G0557 for patients with two or more chronic conditions, and around $116 at G0558 for those same patients when they're Qualified Medicare Beneficiaries. Traditional chronic care management still pays $66.13 nationally for 99490, the first twenty minutes of clinical staff time.

Run that on a panel. Four hundred eligible patients at the moderate tier is more than $250,000 a year for work your team is largely doing already and currently giving away.

So why has almost nobody collected it?

Because for a decade the price of admission was counting minutes. CCM required at least twenty documented minutes of clinical staff time per patient per calendar month, which meant somebody had to log every phone call, every refill coordination, every chase of a specialist's office, and then defend it. Practices tried. Most quietly stopped.

APCM removed that specific obstacle. No time threshold, no monthly minute log, no month lost because a coordinator came in at nineteen minutes. In exchange the practice has to genuinely be capable of the bundled service elements, including 24/7 access, a care plan, transition coordination and population level risk stratification.

Which moves the bottleneck. It doesn't remove it.

Where do these programmes actually die?

Enrolment and the cost sharing conversation. Not coding. Never coding.

This is the part I find genuinely interesting, because it's a human problem wearing a billing costume. CCM and APCM are Part B services, so patients owe coinsurance, roughly thirteen dollars a month on the base code. Practices that skip that conversation get a signature, then get an angry call in week six when the statement lands, then get a disenrolment. Practices that name the dollar figure before asking for consent enrol fewer patients on the first pass and keep far more of them.

Two sentences in a phone script. That's the whole difference between a programme and a spreadsheet.

And the other failure is simpler still. Nobody makes the calls. You need consent captured in a retrievable field, an initiating visit confirmed, one practitioner per patient per month verified so the claim doesn't bounce, and then actual monthly contact. That's a lot of phone calls to patients who are older, often unwell, and frequently not by the phone.

Is this a staffing problem or a contact problem?

It's a contact problem that everyone keeps trying to solve by hiring.

I built a mental health app once, for people living with bipolar disorder, and it was beautiful and got 15% weekly engagement. Then we stopped waiting for patients to come to us and called them instead. Eighty-five percent, every week, and it hasn't moved since across more than a million interactions in five countries and three languages, with zero critical adverse events.

The care management version of that insight is unglamorous. Most of a monthly care management touch is structured: how are the new meds sitting, did you get to the cardiology appointment, any swelling, any falls, do you need a refill. Those are answerable questions with escalation rules attached. Run them outbound, consistently, log them, escalate the ones that matter to a named human, and you've got a programme that actually delivers the service rather than one that bills for it and hopes.

Our use case breakdowns by specialty go through where the escalation thresholds sit, and the integration docs cover how this hangs off an existing EHR without a migration.

What does the arithmetic look like?

Take the moderate tier at roughly $54 per patient per month. Two hundred enrolled patients is about $130,000 a year. The question is what the delivery costs you, because that's the entire margin.

If it's a full time coordinator at loaded cost, you're spending most of it and the programme is a wash that you keep for the quality metrics. If the routine contact runs automatically and the coordinator only handles escalations and the genuinely complex conversations, the maths changes shape completely. That's where the 31 to 1 return we see across deployments comes from, and it isn't clever. It's just that the expensive person stops doing the cheap work.

One thing worth watching: CMS has proposed that from January 2027, remote physiologic and remote therapeutic monitoring may only be billed when furnished by clinical staff directly employed by the billing practice. It's proposed, not final, and as written it doesn't touch CCM or APCM. But if you're building on outsourced staffing for RPM, read the final rule carefully before you commit.

Key Takeaways

There is a well funded, fully legitimate, CMS designed revenue line sitting in nearly every primary care panel in the country, and the reason it goes uncollected has almost nothing to do with billing knowledge. APCM removed the time tracking that killed CCM adoption for a decade, which was the right fix for the wrong bottleneck, because the actual constraint was always whether anybody reliably contacts the patient every month and documents it. The programmes that fail do so on undocumented consent, on a surprise coinsurance bill in week six, and on months that quietly go unbilled because nobody owned the calendar. The programmes that work treat the monthly touch as an operational certainty rather than an aspiration, and they name the patient's out of pocket cost out loud before asking for a signature.

The money isn't hiding. It's just on the other side of a phone call that nobody has time to make. If you want to work out what your panel is actually worth and what it would take to run it, grab twenty minutes with me.

FAQ

Can specialty practices bill APCM?

No. CMS restricts APCM to primary care specialties, which means family medicine, general internal medicine, geriatrics and pediatrics. Specialty practices can still bill CCM and principal care management, and a specialist can bill PCM for a separate condition even while a primary care practice bills APCM for that same patient.

Can I bill APCM and remote patient monitoring in the same month?

Yes. RPM and RTM are both billable alongside APCM, as are behavioral health integration add ons, principal illness navigation and community health integration. What you cannot stack with APCM is CCM, PCM, transitional care management or virtual check ins by the same practitioner in the same month.

Do patients have to pay for this?

Yes, standard Part B cost sharing applies, roughly twenty percent after the deductible. Medigap and Medicaid cover it in full or in part for a large share of patients, but not everyone, and the single highest leverage thing you can do is say the actual number during the enrolment call rather than letting the statement say it first.