D-SNP Enrollment Triples as Medicare Advantage Carriers Chase Higher Margins

KFF's February 2026 enrollment data shows more than 8 million people in a Medicare Advantage Special Needs Plan, up nearly 900,000 in a year. That increase is 83% of all MA net growth, in a year when total MA enrollment grew 3%, the weakest pace in close to two decades. The dual-eligible book is now carrying the program's growth, and part of the number is a reclassification rather than a sale.

Key Takeaways

• 83% of MA net growth came from Special Needs Plans, against total MA enrollment growth of 3%, down from a 9% long-run average between 2007 and 2024.
• 7.5% D-SNP margins against 3.6% for all MA in MedPAC's 2022 data, with C-SNPs at 7.4%, which is the differential the strategy is built on.
• Roughly $30 PMPM of value added came out between 2025 and 2026 after VBID terminated, with about $26 PMPM of that in the nonuniform benefit component.
• The look-alike threshold fell from 70% to 60%, forcing dual-heavy standard plans to redesignate as D-SNPs and moving enrollment into the category by rule.
• 54% of SNP enrollment sits with UnitedHealth Group and Humana, with UnitedHealth alone at 40% overall and 51% of C-SNPs.

What the February Data Says

D-SNPs became permanent under the Bipartisan Budget Act in 2018 with 2.2 million enrolled. That reached 5.8 million by 2024 and crossed 6.0 million in February 2025. Total SNP enrollment has risen 215% since permanent authorization, from 2.6 million to over 8 million, against 49% growth in general enrollment MA plans.

Roughly 82% of SNP enrollees are in D-SNPs. The remainder splits between Chronic Condition SNPs at about 1.15 million and Institutional SNPs at roughly 115,000, with I-SNPs essentially flat since 2024. C-SNPs are the fastest-moving piece, up from 674,500 in 2024 to 1.15 million in 2025, a 71% increase concentrated in diabetes, chronic heart failure, and cardiovascular disease.

The margin data explains the direction of capital. MedPAC put D-SNP margins at 7.5% in 2022 against 3.6% for all MA plans, with C-SNPs at 7.4%. Dual-eligible members carry higher risk scores because their clinical profiles include more diagnosed conditions, and Medicaid cost-sharing assistance moves part of the member's out-of-pocket burden onto the state, which improves the plan's loss ratio on the Medicare side.

The V28 risk model matters here in a way that cuts in D-SNPs' favor. MedPAC estimates coding intensity still inflates MA risk scores by about 4% overall. Where a population's clinical complexity is genuine rather than coded, the coding intensity adjustment bites less, so the 2027 rate environment compresses standard plans harder than it compresses duals.

What VBID's Termination Did to the Bid

CMS ended the Value-Based Insurance Design model after 2025, citing costs to the Trust Funds. Milliman's 2026 D-SNP valuation report sizes the result: total value added fell approximately $30 PMPM from 2025 to 2026, with about $26 PMPM of the decline in the nonuniform benefit component.

The replacement is not equivalent. Around 90% of D-SNP members were in plans offering VBID benefits in 2025, and SSBCI prevalence rose to approximately 90% in 2026, so plan-level coverage looks unchanged. But VBID targeted socioeconomic criteria while SSBCI requires a specific qualifying chronic condition. The same share of plans offers a benefit that fewer members can actually claim.

That is a pricing problem rather than a benefit-design problem. The eligible-member denominator changed shape, so estimating supplemental benefit cost now requires a clinical classification step that VBID did not need. An actuary who carries the old take-up rate into the June bid overstates cost; one who assumes SSBCI eligibility tracks the D-SNP population understates it.

The mix moved accordingly. Combo benefits rose from 36% to 42% of Part C benefit value while VBID, uniform flexibility, and SSBCI together fell from 29% to 21%. Plans shifted from defined standard designs toward Enhanced Alternative structures with $0 cost sharing on selected drug tiers, and 278 Enhanced Alternative D-SNP plans, 33% of the total, filed negative basic premiums. D-SNP plan count still grew about 10%, so carriers competed on footprint while per-member benefit value fell.

Standard MA moved the other way in the same season. Non-SNP MA and MA-PD plan counts fell 10%, from 3,719 to 3,373, and roughly 2.9 million enrollees were displaced by county exits at a forced disenrollment rate of 10% against a historical 1%. UnitedHealth lost 535,000 total MA enrollees while adding 266,800 in SNPs; CVS Health lost 28,600 overall and grew SNPs by 39,600. Molina is exiting traditional MA prescription drug plans entirely in 2027 to concentrate on a $5 billion dual-eligible business it says the roughly $1 billion MAPD line does not fit.

Part of the Growth Is a Rule, and the Rest Sits With Two Carriers

The D-SNP look-alike threshold, which limits how dual-heavy a non-SNP plan may be, dropped to 70% in 2025 and to 60% in 2026. Plans over the line must formally redesignate as D-SNPs, which pulls state contracting, benefit coordination, and CMS reporting obligations with them.

The effect on the headline is mechanical. Enrollment previously counted as standard MA is reclassified into the D-SNP category without a member changing plans, which inflates measured D-SNP growth at the same time as it raises the compliance floor. Any trend fitted to reported D-SNP enrollment across the 2025 and 2026 boundary is fitting a definitional change alongside real growth.

The concentration underneath is the harder constraint. UnitedHealth Group and Humana hold 54% of SNP enrollment, with UnitedHealth at 40% overall, 38% of D-SNPs, and 51% of C-SNPs. Elevance follows at 10%. Nonprofit insurers account for 14% of SNP enrollees, so most of the dual-eligible population sits with publicly traded carriers making capital allocation decisions on a quarterly cycle, as UnitedHealth's own medical cost dynamics have made visible.

Those two facts compound. The members being concentrated are the ones least able to absorb a plan transition: care coordination needs, cognitive impairment, limited English proficiency, and a Medicaid enrollment that has to realign with any new plan. Both dominant carriers have already demonstrated in standard MA that they will withdraw from counties when the economics turn, and the Star Ratings settlement directing $18.6 billion to MA insurers rewards rated quality rather than the population a plan serves.

Integration requirements pull the same way. CMS tightened enrollment, benefit, and coordination rules for FIDE and HIDE SNPs in 2025, and of 35 states with coordination-only D-SNPs in 2024, just over half added state-specific provisions. Each contract carries fixed compliance cost, which is absorbed more easily at 40% share than at 1%.