The Ambulatory Specialty Model Is Not MIPS With a New Name. It Is a Tournament, and the House Keeps 15%.
CY 2027 PFS Proposed Rule (CMS-1848-P) and what the Shared Savings Program proposals mean for ambulatory practices
Comments due September 14, 2026
There is a specialist somewhere reading this who is already in a mandatory two-sided risk model and does not know it.
That is not a rhetorical device. CMS published the preliminary participant list for the Ambulatory Specialty Model in February 2026. Participation is by TIN/NPI combination, selected by CMS. There is no application, no opt-in, and no opt-out. CMS's own participant readiness materials tell selected clinicians to monitor their email for model communications. In our conversations with specialty groups this year, awareness of the model has lagged well behind selection for it.
The model starts January 1, 2027. The first payment adjustments land in 2029, at up to ±9% of all Medicare Part B professional services not just the episodes being measured.
In the CY 2027 Physician Fee Schedule proposed rule, released July 14 and published in the Federal Register on July 16, 2026, CMS proposes to refine ASM. It does not propose to delay it, soften its mandatory character, or reduce the risk.
What ASM actually is
ASM is a CMS Innovation Center model, finalized in the CY 2026 PFS final rule and codified at 42 CFR Part 512, Subpart G. It runs five performance years, 2027 through 2031, with payment adjustments running 2029 through 2033.
It covers two conditions: heart failure and low back pain. Critically, neither is defined clinically. Each is defined by the corresponding MIPS episode-based cost measure episode specification, imported wholesale under § 512.710(e). Your episode volume is therefore driven by evaluation and management and therapy coding patterns, not by your procedure mix. This means a practice's attributed volume can move meaningfully without any change in what it actually does clinically.
Four eligibility criteria apply cumulatively under § 512.710(b). A clinician must bill under the PFS, hold a selected specialty designation, have at least 20 attributed episodes for the condition in the calendar year two years prior to the performance year.
The specialty list is narrow and literal:
Heart failure: cardiology, and only cardiology.
Low back pain: anesthesiology, interventional pain management, neurosurgery, orthopedic surgery, pain management, and physical medicine & rehabilitation.
Specialty is determined by the code appearing on the plurality (not the majority) of a clinician's Part B claims (§ 512.710(d)), as assigned by the MAC from the physician's PECOS designation. That is a lower bar than it sounds; a clinician can be pulled in on a code accounting for well under half their claims.
It also has a consequence worth pausing on: because "cardiology" is the only listed heart failure specialty, cardiologists carrying their own Medicare specialty codes. These cover electrophysiology, interventional cardiology, advanced heart failure and transplant, and adult congenital. CMS effectively concedes the point in the CY 2027 rule by proposing a permanent redesignation exception for exactly those subspecialties. A stale PECOS specialty designation is now a payment-risk variable.
Geographic selection is by core-based statistical area, stratified by episode spending and volume, with CMS selecting from each stratum. CMS describes the result as roughly one quarter of CBSAs nationally.
Non-physician practitioners are excluded entirely.
Key takeaways
Selection is automatic, by TIN/NPI, there is no application and no way to opt out.
All four eligibility criteria must be true simultaneously; PY 2027 eligibility is already locked to CY 2025 volume.
Specialty is assigned by plurality of claims, not majority which is a smaller share of your billing than you'd expect can trigger it.
A stale PECOS subspecialty code can pull a cardiologist into a designation that shouldn't apply to them.
The scoring is where the MIPS analogy breaks
ASM uses the four MIPS performance categories. It does not use them the way MIPS does.
Read the bottom two rows again. In MIPS, Improvement Activities and Promoting Interoperability are places to earn points. In ASM they are places to lose them. Perfect performance on both is worth exactly zero; it simply avoids a deduction of up to 30 points.
ASM also does not carry MIPS's reweighting machinery. Section 512.740 contains no small-practice PI reweighting and no MIPS-style hardship exception application. The relief valves are narrower: § 512.780, extreme and uncontrollable circumstances, and a PI measure suppression policy CMS proposes to add for CY 2027. Short of those, a PI failure is a straight deduction.
There are two adjustments on the upside. One is a complex patient adjustment worth up to 10 points. The other is a small practice adjustment worth 10 points for practices of 2–15 clinicians or 15 points for solo practitioners.
There is one cliff. Under § 512.745(a)(2)(ii), failing to submit quality data produces a final score of zero. Not a reduced score. Zero.
The measure sets are fixed, and they straddle two reporting rails
Unlike MVPs, ASM participants do not select measures. The sets are prescribed.
Heart failure (§ 512.725(b)): Q492 risk-standardized acute unplanned cardiovascular-related admission rates (administrative claims, CMS-calculated); Q008 beta-blocker therapy for LVSD; Q005 ACE inhibitor / ARB / ARNI therapy for LVSD; Q236 controlling high blood pressure; and Q377 functional status assessments for heart failure, eCQM only.
Low back pain (§ 512.725(c)): Q238 use of high-risk medications in older adults; Q134 screening for depression and follow-up plan; Q128 BMI screening and follow-up plan; and Q220 functional status change for low back impairments, MIPS CQM only.
That eCQM-only / CQM-only split is the operational detail most likely to produce an unpleasant surprise. A cardiology group without eCQM capability cannot score Q377 at all. A low back pain practice reporting purely through eCQMs cannot score Q220. Data completeness is 75% and the case minimum is 20 per measure, the same thresholds MIPS uses.
Cost is calculated by CMS from administrative claims. Like always, nothing to submit and nothing to fix after the fact.
The two Improvement Activities are ASM-specific and both are required. One covers connecting beneficiaries to primary care and health-related social needs screening. The other requires at least one executed Collaborative Care Arrangement with a primary care practice you share beneficiaries with, covering at least three of five specified elements. That second one is the long-pole item, because it requires a willing external partner and a signed agreement not an internal workflow change.
The financial structure is a tournament, and the house takes 15%
The payment adjustment scales up across the model: ±9% for performance years 2027 and 2028, rising to ±12% by the final performance year, per CMS.
The mechanism is what matters. The incentive pool equals participants' Part B payments multiplied by the risk level and then by an 85% redistribution percentage under § 512.750(c)(1)(iii). CMS retains the other 15% as guaranteed Medicare savings.
That is not a rounding detail. Scoring is relative to peers, and the pool pays out less than it collects. The AMA's read is that this guarantees a majority of participating physicians receive a reduction; CMS has not published the distribution that would settle the question either way. What is not in doubt is the structure: absolute clinical performance is not what sets your adjustment. Relative rank is.
Both the AMA and the American College of Cardiology have pushed back. The AMA has asked CMS to raise the redistribution percentage to 100% and to cut the maximum risk level in the model's first two years from 9% to 2%. The ACC has asked for that same ±2% cap on a phased-in basis, at least one dedicated educational and testing year before financial accountability begins, and a built-in goal score tied to guideline-directed clinical care rather than pure peer ranking. The American College of Surgeons has opposed the model outright.
Then there is the finding that surprises people most. On CMS's own Comprehensive List of APMs, ASM is marked not an Advanced APM and not a MIPS APM. Participation earns no QP status, no APM incentive payment, and no qualifying-APM conversion factor. Several law firm and consultant write-ups have described ASM in Advanced APM terms. The CMS list governs, and it says otherwise: this is downside risk without the MACRA upside.
What ASM does do is replace MIPS. Section 512.775(a) waives the MIPS statute for participants meeting eligibility criteria. Here, participants are exempt from MIPS and cannot receive a MIPS payment adjustment for those years. That exemption applies to the ASM years only. Clinicians who appear on the preliminary 2027 participant list must still report MIPS for CY 2026, and CMS has said so explicitly.
Key takeaways
The redistribution pool pays out 85% of what it collects: CMS keeps 15% regardless of clinical outcomes.
Scoring is relative to peers, not an absolute clinical bar: and CMS hasn't published the resulting distribution.a
ASM is officially not an Advanced APM or MIPS APM: downside risk without QP status or APM incentives.
MIPS is waived only for ASM years: CY 2026 MIPS reporting is still required for everyone on the preliminary list.
What the CY 2027 rule proposes to change
The proposals fall into three buckets.
Genuine burden relief. One Collaborative Care Arrangement can now cover multiple ASM participants billing under the same TIN, rather than requiring a separate agreement per physician. The shared-beneficiary requirement is met if at least one beneficiary is shared, and the "established patient" requirement comes out. "Payment" broadens to "remuneration," permitting in-kind support. On the Promoting Interoperability side, the Security Risk Analysis attestation and the ONC direct review attestation come out, and a measure suppression policy is added for external circumstances. Heart failure specialists who redesignate to a board-certified procedural cardiac subspecialty gain a permanent exception. An earned payment multiplier now follows the clinician to a new TIN.
Changes that raise the stakes. CMS proposes to add a claims-based lumbar spine MRI overuse measure. This is the measure it declined to finalize for CY 2026 because it was still in development. It proposes to remove Q220 and add Q182, Functional Outcome Assessment, swapping an outcome measure for a process measure. And it proposes to score claims-based measures at the individual TIN/NPI level even when a group reports collectively, which the ACC confirms applies to Q492 for heart failure beginning in 2027.
New adjustments that are contested. CMS proposes a +5-point rural adjustment to the final score, and a separate +5-point bonus inside the quality category for voluntarily submitting patient-reported outcome and risk-adjustment data. The AMA has objected to both on the same ground: in a zero-sum tournament, every point added for one group is a point of exposure added for everyone else, and the voluntary-reporting bonus advantages practices with the resources to submit optional data.
Electronic prior authorization follows the MIPS track: the medical services measure is optional in 2027 and required in 2028, and a new prescription-drug measure arrives as a requirement in 2028.
CMS also proposes new authority to terminate participants for program integrity, patient safety, legal, or compliance concerns, and clarifies that losing eligibility in a later year does not erase adjustments already earned.
What we'd do between now and September 14
Confirm your status and re-confirm it. Check the ASM participants dataset on data.cms.gov against every TIN/NPI in your group. CMS intends to release the final PY 2027 dataset in summer 2026, re-determined against more recent claims. Absence from the preliminary list is not a clearance: CMS has not said the final list can only shrink, and a clinician who crosses the 20-episode threshold in updated data could appear on it. Check again when the final dataset posts.
Audit your PECOS specialty designations. Especially in cardiology, where subspecialty codes fall outside the heart failure designation entirely.
Resolve the eCQM/CQM split, but stage it. Cardiology: confirm you can produce Q377 as an eCQM. It has no MIPS CQM fallback and nothing in the CY 2027 proposal changes that. Low back pain: Q220 is MIPS CQM-only today, but CMS proposes to swap it for Q182 for CY 2027 — confirm you can produce both collection types, and hold any Q220-specific build until the final rule lands in November.
Start the Collaborative Care Arrangement conversation now. It requires an external primary care partner and an executed agreement covering at least three of five elements. The CY 2027 proposal to let one CCA cover a whole TIN makes this dramatically easier — but only if it is finalized, and only if you have a partner.
Watch your email for ASM Participant Portal registration instructions. CMS has said registration is coming; separately, CMS plans to use the QPP portal for actual data reporting.6. Do not skip MIPS for CY 2026. The waiver starts with the ASM performance years, not before.7. Comment. File code CMS-1848-P at regulations.gov, docket CMS-2026-2377. The lumbar MRI measure, the Q220→Q182 swap, individual-level claims scoring, and the 85% redistribution percentage are all live.
Mark one more date: March 31, 2028, the PY 2027 data submission deadline. Miss it and § 512.745(a)(2)(ii) produces a final score of zero.
Why ReportingMD is watching this one closely
We have been doing population health and quality reporting work with CMS since 2007, and we have a seat at the table for most of the program architecture ambulatory practices live inside today.
We were part of the group that helped wire-frame MIPS. We built one of the first bidirectional interfaces for MIPS submission, because we did not think a one-way file drop was a real answer for practices that needed to act on their data during the year rather than read about it afterward. And we began working with CMS on eCQMs while they were still in development, starting in 2011.
ASM is the reason that bidirectional argument matters more now than it did in 2017. In MIPS, a specialist who discovers in March that a measure fell short has lost a year of scoring. In ASM, that same discovery arrives against a fixed measure set with no substitutions, a 20-case minimum, no automatic reweighting, and a submission failure that scores zero. There is no version of this model that works well for a practice reading its quality data retrospectively.
Under § 512.720(a)(1), ASM quality data may be submitted by a third-party intermediary. We are an eCQM-certified vendor serving ambulatory practices, including the cardiology, orthopedic, neurosurgery, PM&R, pain management, and anesthesiology specialties ASM selects. If you are on the participant list or even think you might be we would welcome the conversation.
Think you might be on the ASM participation list?
Talk to our team about checking your TIN/NPI status and building a submission-ready data pipeline before PY 2027 starts.
This post summarizes the Ambulatory Specialty Model as finalized in the CY 2026 PFS final rule and codified at 42 CFR Part 512, Subpart G, and proposals in CMS-1848-P, published at 91 FR 43842 on July 16, 2026. Proposals are not final policy and may change in the final rule, typically issued in early November. Participant lists, measure specifications, and portal timelines are subject to CMS updates — verify current status at cms.gov before acting. Nothing here is legal or compliance advice.
Comments