The Only Plan Built For Facility Residents Is 1.5% Of The Program CMS Is Reviewing

A Review That Starts With A Question About Sameness

The Centers for Medicare and Medicaid Services is working through whether the three Special Needs Plan types should stay three types. The agency is in an information-gathering posture — requests for information, conversations with experts — rather than a rulemaking one, and the framing of the question is what matters to anyone who owns a licensed building.

Rose Mollitor, managing director at ATI Advisory, put the agency's question plainly: CMS "is wrestling with the question of whether SNPs are delivering value, and to whom, and are the existing SNPs, specifically the dual plans versus the chronic needs plans versus the I-SNPs, are they delivering care in a differentiated enough way to be different products" ([Skilled Nursing News, September 4, 2026](https://skillednursingnews.com/2026/09/as-special-needs-plans-surge-cms-weighs-consolidating-plan-types-against-specialized-care-in-nursing-homes/)).

Read that as an underwriting input, not a policy abstraction. A review that opens by asking whether products are differentiated enough to remain separate products has a predictable asymmetry: the burden of proving distinctiveness falls hardest on the smallest one.

The Size Problem, In Numbers

As of the August 2026 CMS Special Needs Plan Comprehensive Report, 8,437,333 people are enrolled in a Special Needs Plan across 1,803 plan records. The split:

• •D-SNP (dual eligible): 6,500,322 enrollees, 347 contracts, 1,082 plans
• •C-SNP (chronic condition): 1,806,608 enrollees, 146 contracts, 562 plans
• •I-SNP (institutional): 130,403 enrollees, 72 contracts, 156 plans

(All figures: CMS Special Needs Plan Comprehensive Report, August 2026, as compiled at [medicareinsights.org](https://medicareinsights.org/market/snp/).)

I-SNP is about 1.5 percent of total SNP enrollment — a figure we computed from the CMS counts above rather than one the agency publishes as a headline. D-SNP is roughly fifty times larger.

It is also the only one of the three built specifically around people who live in a licensed setting. Ben Lerer, CEO of Longevity Health, drew that line in the same report: "While C-SNPs and D-SNPs may serve portions of a similar population, I-SNPs are designed specifically for individuals living in institutional settings," and the future of the program "should preserve the distinctions that make each model successful, including the specialized, facility-based infrastructure that allows I-SNPs to bring proactive, coordinated care directly to some of Medicare's most vulnerable members" (Skilled Nursing News, September 4, 2026).

Note what he is defending. Not an enrollment number — an infrastructure. The nurse practitioners who round in the building, the after-hours coverage that keeps a resident out of an ambulance, the staffing arrangement that exists because a plan is paying for it.

Why The Small One Is The Exposed One

There is a second number in the reporting that explains where the pressure is coming from, and it is not I-SNP's.

C-SNP enrollment is up 11.5 percent since January 2026, per ATI Advisory data cited in the same article. Mollitor's read: "Any time you have exponential growth in a product that wasn't as common a few years ago, it invites a lot of attention." CMS is asking who is enrolling in C-SNPs and whether they could be offered for more commonly diagnosed conditions like cardiovascular disease and diabetes.

So the growth that triggered the review is in the chronic-condition plans. The consolidation logic, if CMS pursues it, would be applied across all three. That is the structural risk: a review provoked by one product's expansion can end up resolving a different product's separate existence, simply because the smaller product is harder to justify as its own line item.

Mollitor also flags the overlap that makes I-SNP absorbable on paper: "Members who may be appropriate for an I-SNP oftentimes are also dual eligible." If a resident qualifies for both, an analyst can ask why two products are needed. The answer is the care model rather than the eligibility category — but care models are harder to defend in a spreadsheet than enrollment counts are.

The Concentration Nobody Prices

Divide the enrollment by the plan counts and the three products stop looking like versions of one thing. Using the CMS August 2026 figures above, average enrollment per plan works out to roughly 6,000 for D-SNP, 3,200 for C-SNP, and 840 for I-SNP (our computation).

That is not a scale failure. It is what a facility-based model looks like. Mollitor described why it cannot be fixed by wanting it bigger: I-SNP is "a very high touch clinical model, and the ability to grow in that space is geographically concentrated in a way that D-SNPs and C-SNPs are not. You need to be physically present where those individuals are."

Physically present. In buildings. That sentence is the entire investment thesis and the entire policy vulnerability in one line — the model's value and its inefficiency have the same cause.

She expects growth to continue regardless, and notes it has historically come from provider-owned plans, with national carriers now likely to scale their own offerings off local partnerships.

What This Changes For An Owner

Three things, in order of how soon they bite.

An I-SNP relationship is an asset attribute, not a footnote. If a building has a plan partner putting clinical staff on site, that arrangement affects acuity mix, hospital transfer rates and length of stay. It belongs in the diligence file next to the rent roll. If the plan structure is under federal review, so is part of that operating model — and a buyer who has not read the plan agreement has not finished reading the deal.

Provider-owned means owner-exposed. Where the I-SNP is owned by the operator rather than a national carrier, consolidation is not a vendor issue — it is a change to a business the seller may partly own. Ask who owns the plan entity, and whether it conveys.

The timing is knowable, and it is not now. Mollitor points to the 2028 Medicare Advantage and Part D proposed rule, expected this fall, as where SNP policy changes would surface if there are any. She also notes that unlike the SNF Prospective Payment System rule, the MA and Part D rule is not statutorily required and functions "more along the lines of a recommendation." So: a fall proposed rule is the next real signal, and even then it is a proposal.

What We Are Not Saying

We are not forecasting that CMS eliminates I-SNPs. Nothing in the reporting says that. The agency is gathering information, stakeholders are actively arguing the value of each model, and Mollitor's own expectation is continued incremental I-SNP growth.

What we are saying is narrower and more useful: a federal review is openly asking whether three plan types are different enough to stay three, the type built for facility residents holds 1.5 percent of the enrollment, and the specific thing at stake if the answer comes back "no" is the facility-based clinical infrastructure — the part that touches census, acuity and the operating model of the building itself.

For anyone underwriting a skilled nursing or higher-acuity assisted living asset this fall, that is not a headline to react to. It is a question to add to the list: is there an I-SNP in this building, who owns it, and what happens to the on-site clinical staffing if the plan type is folded into another one?

Owners who can answer that will price it. Owners who cannot will find out from the buyer.