Medicare Advantage V28 risk adjustment, tighter Star Ratings, and a flat rate are forcing plan exits in markets and a pivot to D-SNPs. This post is the follow-up to my previous post (https://rhislop3.com/medicare-advantage-snf-denials/)…by Reginald Hislop, III
CMS spent three years phasing in a stricter, less generous way of paying for how sick Medicare Advantage enrollees are, tightened the quality-bonus system at the same time, and slowed the baseline rate increase to below medical cost trend. The result is not subtle, and it is not a surprise. Insurers are retreating from the geography and the products that no longer pencil out, while pushing harder into the ones that still do. The county exits are the geography. The D-SNP push is the products. Both are the same decision, made against the same payment model.
The mechanism is the risk-adjustment model that assigns the money in the first place. Medicare Advantage pays plans by risk score, and the entire structure of Medicare Advantage risk scores is being redefined. A plan whose enrollees register as sicker generates more revenue, and the model that assigns those scores has just been replaced. CMS has been phasing in a new version, V28, since 2024, and it became fully weighted — 100 percent of risk-adjustment payments — starting payment year 2026. The prior model, V24, is gone from the payment formula. Understanding CMS-HCC Model V28 Changes And Their Impact In 2026
V28 is structurally less generous by design. It cuts the number of valid diagnosis codes that generate a risk score from 9,797 to 7,770. It expands the number of condition categories from 86 to 115. And it is built to reward documented clinical severity rather than the sheer volume of diagnoses. The era one industry source calls the “add-only chart review” is over, and the plans that built their revenue on code volume rather than acuity are the ones absorbing the hit first.
The scale of the hit is now measurable. CMS projects a 3.12 percent average reduction in Medicare Advantage risk scores nationally, which translates to roughly $11 billion in savings to the Medicare Trust Fund — revenue that was flowing to plans as risk-adjustment payments and is now staying with the program. One industry estimate puts the cumulative V28 phase-in impact at more than 7 percent out of MA base rates across 2024 through 2026.
The effect is not evenly distributed. Plans serving sicker populations are hit hardest, with some seeing as much as 10 percent revenue loss and an average impact around 3.5 percent even after sophisticated coding efforts. The plans with the densest risk-score books — the ones that looked most profitable under V24 — are now the ones with the most to lose under V28.
The Medicare Advantage Star Ratings squeeze is compounding, not offsetting
The risk-model change is being layered on top of the Medicare Advantage Star Ratings changes that operate in the same direction. CMS has adjusted Star measure weightings, introduced new measures, and set more stringent cut points, making the high ratings and their bonus payments harder to earn. The average bonus payment per member fell nearly 10 percent, about $40 per member per year, from 2023 to 2025, as fewer beneficiaries landed in four-star-plus plans.
Looking ahead, CMS is shifting Star Ratings toward a 65 percent clinical-weighting model for 2027. That makes Star performance even more sensitive to exactly the clinical documentation pressure V28 already created. The two changes do not offset each other. They reinforce each other, and the plan that was already losing risk-adjustment revenue is now also watching its quality-bonus revenue thin out at the same time.
The Medicare Advantage rate environment tightened on its own
Even without the model and ratings changes, the forward-looking rate picture is weak, and the result is a sequence of Medicare Advantage payment cuts in real terms. CMS finalized a 5.06 percent rate increase for 2026 but proposed just a 0.09 percent increase for 2027. The Medicare Advantage rate increase 2027 is a number that matters because it sits below the trend: even with expected risk-score growth factored in, total payment growth for 2027 projects around 2.5 percent — below prevailing medical cost and utilization trends. For most plans, expenses are likely to outgrow revenue on the base rate alone. CMS makes structural changes to Star Ratings system for Medicare Advantage and Part D plans | ReedSmith
CMS has also moved forward with extrapolated RADV audits, applying statistically extrapolated recoveries to past risk-adjustment claims, with an estimated $425 million in annual recoveries once audits of payment year 2018 and beyond begin. That adds retroactive financial risk on top of the forward-looking rate pressure. A plan is now managing a present in which rates are flat and a past in which coding decisions can come back as recoveries.
What the carriers are actually doing
The carrier responses are the proof that connects the payment change to the exits. UnitedHealth Group now projects 1.3 to 1.4 million member exits from Medicare Advantage in 2026, revised sharply up from an initial internal projection of 600,000. The company called its Q1 2026 results “unusual and unacceptable” and said it is prioritizing profitability over membership growth, including cutting PPO benefits.
CVS Health and Aetna framed their 90-plan, 100-county pullback explicitly as a margin-recovery strategy centered on benefit cuts and portfolio optimization rather than growth. Humana reduced its county footprint to 85 percent from 89 percent in 2025 while working to improve its Star Ratings — directly trading membership breadth for quality-bonus eligibility.
None of these are the actions of plans that anticipate a revenue recovery. They are the actions of plans that have re-read the payment formula and concluded that the counties and the members at the margin no longer pay for themselves. The exit counts are the arithmetic of V28 working its way through a plan’s P&L, county by county.
The pivot that tells you the rest of the story
The telling detail is where the plans are now going. Even as standard Medicare Advantage retrenches, carriers are pivoting product mix toward D-SNPs, the dual-eligible special needs plans, and the Medicare Advantage D-SNP growth is one of the clearest signals in this cycle. One industry source describes it as carriers chasing higher margins in the dual-eligible segment. Special Needs Plans (SNP) | Medicare
That pivot is not a separate story from the pullback. It is the same strategic reallocation. The plans are not abandoning Medicare Advantage. They are moving their capital away from the standard book, where V28 and Star changes are compressing margin, and toward the dual-eligible segment, where the payment structure still compensates acuity. The D-SNP concentration is going to be one of the defining features of the next MA cycle, and it is happening because the same payment pressure that is emptying the rural counties is filling the dual-eligible lanes.
What this means for operators
The operator who only reads the county map will misunderstand the next two years. The V28 phase-in is already fully weighted, the Star methodology is tightening into 2027, and the RADV audits are pulling recoveries out of past years. The plans are behaving like organizations that expect flat revenue and rising costs, because that is what the rate card now says.
For the post-acute and skilled nursing side, the consequence runs through the payer mix. The plans are not going to negotiate more generously in the counties where they remain — they are going to manage the network harder, because the margin pressure that pushed them out of the rural counties is still present in the urban and suburban ones. The dual-eligible pivot means the growth in MA membership is increasingly concentrated among the population with the highest post-acute utilization, which reshapes what a facility’s MA census actually looks like.
The V28 model is not a reimbursement tweak. It is the end of a coding era, and it is the reason the maps changed. CMS tightened the payment for acuity, tightened the reward for quality, and slowed the base rate below the growth in medical costs, all inside the same three-year window. The plans read that sequence correctly: retreat from what no longer pays, lean into what still does. The rural counties lost their plans, and the dual-eligible segment gained the capital, because the formulas changed in the same direction at the same time.
For the operator, the instruction is the same whether you run a plan or a facility. Do not price the old model. The payer mix you carried under the add-only chart review era is not the payer mix you will carry under V28, tighter Stars, and a flat base rate. Re-underwrite it now, against the payment change, before the shift in geography and product mix reaches your county and your census in a single plan year.
