Value-Based Care: What It Pays For, Who Wins, and Where to Start Fixing It in 2027

Value-based care was supposed to align the system around outcomes instead of volume. It is a real improvement on what came before it. It is not yet the thing it claims to be, and 2027 is close enough to start building the correction now.

By RaeAnn

In the earlier piece on the economics of Medicare Advantage, I mapped where the dollar goes once it leaves Washington: a health plan, a physician, a health system, an ancillary provider, a pharmacy benefit manager, and finally a member left to navigate all of it alone. Every party in that chain is paid to do a specific job, and not one of them is paid for the member’s healthspan. Value-based care is the industry’s actual answer to that gap, the mechanism built specifically to close it. This piece asks the question plainly: what is it, what does it actually pay for, who comes out ahead under it and why, where does it miss, and what would it take to start correcting that in the 2027 bid and contracting cycle.

What Value-Based Care Actually Is, Inside Medicare Advantage

Strip away the conference-stage language, and value-based care inside Medicare Advantage is four distinct contractual structures, each transferring risk to a different party, at a different scope, negotiated directly with the health plan rather than through a CMS-defined program.

The first is health system global capitation, where a hospital system or large integrated provider accepts financial responsibility for nearly all of an attributed population’s care. This arrangement is typically structured as a percentage of premium, a defined share of the capitation dollar CMS pays the health plan, rather than a flat per-member rate, with specific high-cost categories carved out and paid separately under a division of financial responsibility. Organ transplants and blood products are the standard examples, because no organization can reasonably price that risk into a percentage-of-premium arrangement. The second is primary care capitation, where a primary care group, independent of any health system affiliation, accepts a fixed, risk-adjusted monthly payment for each attributed member’s primary care relationship. The third is specialty capitation, a less common but growing structure in which a specialty group, cardiology and orthopedics being the most frequent examples, accepts a similar fixed payment for a defined population’s care within that one specialty. The fourth is the MSO or IPA capitated model, where an Independent Practice Association holds the actual capitated contract and risk on behalf of a network spanning multiple specialties, with a Management Services Organization providing the administrative, actuarial, and data infrastructure behind it. This fourth structure may or may not include an affiliated health system, and that distinction matters, because an IPA without a health system partner is taking on multi-specialty risk without the balance sheet a hospital system brings to global capitation.

These four structures are not interchangeable, and conflating them, as the industry’s marketing language often does under the single banner of value-based care, obscures more than it reveals. They place risk in different hands, at different scope, with different exposure to the carve-outs and balance-sheet requirements that determine whether the arrangement actually changes clinical behavior or simply changes who absorbs the bill.

QUESTIONS THE BOARD WILL ASK

“Which of these four structures are we actually in, and how much premium is at risk in each?”
Most plans cannot answer this cleanly by contract type. The first governance step is a risk-structure inventory: percentage-of-premium versus PMPM, scope of carve-outs, and the balance sheet standing behind each capitated counterparty.

“If a capitated IPA in our network fails, what is our exposure?”
An IPA holding multi-specialty risk without a hospital balance sheet is the single most fragile node in the chain. The board should know which counterparties are thinly capitalized and what the plan’s continuity plan is when one fails mid-year.

What It Actually Pays For

Here is where the design and the marketing start to diverge, across all four structures. None of them pay for healthspan, because healthspan, as I defined it earlier in this series, the years a person remains functional and independent, is not something claims data captures cleanly or quickly. What they pay for, in every one of the four structures above, is the absence of expensive utilization, principally hospital admissions and emergency department visits, because that is what claims data captures well, fast, and cheaply enough to build a capitated rate or a percentage-of-premium split around.

The model did not fail to reward healthspan by accident. It rewards what is measurable, and healthspan was never built to be measurable in the data value-based contracts actually use.

This is not a hidden flaw. It is the design, stated honestly. A health system holding a percentage of premium, a primary care group under capitation, a specialty group under capitation, and an IPA holding multi-specialty risk through an MSO are all being measured against the same basic question: did this population cost less than what the rate assumed. A population can answer that question favorably for an entire year while quietly losing strength, mobility, and independence in ways too slow and too unbilled to show up in the same year’s data. The two outcomes overlap often enough that every one of these four structures has real value over straight fee-for-service. They are not the same outcome, and the industry has spent two decades treating the overlap as if it were the whole story.

Who Gets Paid More, and Why

The dollars do not distribute evenly across these four structures, and the reasons are structural, not personal.

Health systems capture the largest share of value-based upside, because a percentage-of-premium arrangement at meaningful scale, even after the transplant and blood product carve-outs are removed, still represents a larger and steadier revenue base than any of the other three structures, and only an organization with a hospital-scale balance sheet can responsibly accept that level of concentrated risk. A primary care practice with ten physicians cannot easily survive a year where its attributed population has an unusually expensive flu season. A health system can, which is exactly why global capitation skews toward organizations with that scale, and why health plans negotiate the percentage-of-premium split harder with health systems than with any other provider type.

Within primary care capitation specifically, the clinical behavior rewarded most directly is documentation and risk capture, not necessarily the clinical behavior that improves a member’s trajectory. A primary care group that codes a population’s chronic conditions thoroughly and accurately raises that population’s risk score, which raises the capitated payment the group receives, independent of whether the group’s actual care changes a single outcome. CMS itself has built statutory adjustments into the Medicare Advantage rate methodology specifically because coding intensity under risk arrangements consistently outpaces coding intensity in fee-for-service for the same population; the finalized CY 2026 coding pattern adjustment of 5.9 percent exists for exactly this reason.¹ Specialty capitation carries a narrower version of the same dynamic, concentrated in a single specialty’s coding patterns rather than a full population’s.

The MSO or IPA capitated model distributes its upside the least evenly of all four structures, because a single multi-specialty capitated payment has to be divided internally across every participating physician and specialty, and that internal division depends entirely on the IPA’s own governance and the MSO’s administrative sophistication. An IPA without an affiliated health system is also carrying multi-specialty risk without a hospital balance sheet behind it, which is precisely why the strength of the MSO supporting it, its actuarial infrastructure, its data systems, its ability to model risk corridors accurately, determines whether that IPA survives a bad year or fails the way several capitated IPAs have in markets where MSO infrastructure was thin.

REGULATORY CALLOUT

Risk-adjustment coding is under active federal enforcement.
The gap between coding intensity and clinical reality is not just an economic observation; it is a compliance exposure. CMS finalized the expanded RADV audit methodology in 2025 and is scaling audits toward every eligible MA contract annually, with extrapolated recoveries. A VBC strategy that rewards coding without a defensible clinical basis is building recovery risk directly into the contract.

Percentage-of-premium delegation does not delegate accountability.
When a plan delegates risk to a capitated provider or IPA, the plan remains accountable to CMS for coding accuracy, appeals, and member protections. Delegation agreements should be read as compliance instruments, not just financial ones.

Where It Misses

Put plainly: across all four structures, value-based care inside Medicare Advantage rewards scale over clinical insight, documentation over function, and the absence of a costly event over the presence of a healthier trajectory. It improves on fee-for-service, which rewarded volume outright with no cost discipline at all, and that improvement is real. But none of the four versions of it has yet become the thing it was marketed as twenty years ago, a system organized around the member’s actual outcome. Each is organized around the member’s measurable cost, carved out, capitated, or percentage-split in a different shape, and the gap between that and healthspan is exactly where this series has been pointing from the start.

Value-based care rewards the absence of a costly event over the presence of a healthier trajectory. That is not a failure of execution. It is what the data was built to measure.

Where to Start in 2027

I do not think the fix is a better benchmark or a more generous capitation split, the same caution I raised about CMS’s Star Ratings overhaul earlier in this series. The fix has to start with how the member is actually cared for inside whichever of these four structures she happens to fall into, and the 2027 bid and contracting cycle is close enough to begin building it now.

The starting point is holistic, person-centered care built on accurate population grouping and attribution, regardless of which of the four capitation structures is in place. A member correctly grouped by her actual clinical complexity, not simply by which primary care group or IPA network she happened to select, is the precondition for everything that follows. From there, the redesign asks for a genuine collaborative care model, primary care and specialists working together on the member alongside the health plan, rather than a primary care physician who refers and disengages. UnitedHealthcare’s 2026 policy shift, requiring a primary care referral before most specialist visits across its Medicare Advantage HMO and HMO-POS plans, is a real, current example of the gatekeeping infrastructure this redesign needs, though a referral requirement alone is an administrative control, not collaborative care.² The difference is what happens after the referral is issued: whether the primary care physician and specialist actually close the loop on the consult, whether imaging is selected and used appropriately rather than ordered reflexively, whether preventive care is offered proactively as part of that collaborative relationship, and whether surgery is recommended because it is the clinically best option for that specific member rather than the default path inside a fee-for-service specialty visit that never had to answer to anyone.

The clinical evidence for this kind of collaboration already exists. Multidisciplinary care for complex older surgical patients, with internists or geriatricians working alongside surgeons, has been associated with reduced length of stay and complications in the peer-reviewed literature.³ And CMS has begun building reimbursement for exactly this kind of coordination: it recognizes and pays for a set of interprofessional consultation codes that let a primary care physician and a specialist formally consult, by phone, internet, or through the electronic health record, without the member needing a separate face-to-face specialty visit, and it added new behavioral-health interprofessional consult codes in the most recent fee schedule cycle.⁴ The infrastructure for this redesign is being built piece by piece already. What is missing is a health plan willing to organize its 2027 contracting around a collaborative care model as the explicit goal, across all four capitation structures, rather than treating referral gatekeeping as a cost-control measure and clinical collaboration as someone else’s problem to solve later.

Naming the goal is not the same as specifying the mechanism, and I want to be honest about that here rather than paper over it. Exactly how a plan writes a collaborative care model into a 2027 contract, which specific measure it ties a share of the capitation split to, how it captures a member’s functional status at the point of care without creating one more box to check, and how it holds that measure up under audit, is its own detailed piece of work. That mechanism, and its honest weaknesses, is where the next article in this series goes.

QUESTIONS THE INVESTOR WILL ASK

“If a collaborative care model is the differentiator, how does it show up in the P&L?”
The near-term return is fewer avoidable admissions and readmissions and lower out-of-network specialty leakage; the durable return is a population that ages more slowly into high-cost states. The first is measurable inside a contract year. The second compounds, and it is the actual moat.

“Why would a plan give up the coding upside this implies?”
It does not have to. Accurate clinical collaboration and accurate coding are not opposed; the RADV exposure comes from coding without clinical basis. Collaborative care supplies the clinical basis, which makes the risk score both higher-integrity and more defensible under audit.

There is a second place the Medicare Advantage dollar goes that this series has not yet addressed, and it deserves the same honest treatment. Dental, vision, transportation, and the broader category of supplemental benefits carry their own purpose, their own recent trends, and their own version of the same accountability gap. Plans have spent years treating these benefits as enrollment tools, generous in a good year, the first thing cut in a hard one, rather than as genuine levers for healthspan. That deserves its own recommendation and its own roadmap, including a case for budgeting more, not less, toward dental and vision specifically, and building care models that actually integrate supplemental benefits into the clinical picture instead of administering them on a separate track.

That is the next piece in this series.

RaeAnn is the founder and CEO of HLTHWorks. She spent three decades building and operating Medicare Advantage, Medicaid managed care, and value-based care businesses, including senior executive roles at Cotiviti, Bright Health, Datavant, Signify Health, and Optum. She writes The Standard on the economics, behavior, and redesign of American healthcare.

Sources

External figures and policy references in this article are drawn from primary federal rule-making, payer policy notices, CMS fee-schedule guidance, and peer-reviewed literature. Descriptions of contractual structures, risk distribution, and accountability gaps reflect the author’s operating experience.

Centers for Medicare & Medicaid Services. 2026 Medicare Advantage and Part D Rate Announcement. CMS finalized the statutory minimum coding pattern adjustment of 5.9 percent for CY 2026 to account for higher coding intensity in MA relative to fee-for-service. https://www.cms.gov/newsroom/fact-sheets/2026-medicare-advantage-part-d-rate-announcement
UnitedHealthcare / UHCprovider.com. Referral requirements for Medicare Advantage HMO and HMO-POS plans, effective January 1, 2026, requiring a PCP referral before most specialist services. UHC extended the enforcement grace period beyond the original April 30, 2026 date. https://www.uhcprovider.com/en/resource-library/news/2025/referral-req-specialist-services-medadv.html
Peer-reviewed literature on multidisciplinary and geriatric collaborative care for older surgical patients, which has associated internist and geriatrician involvement alongside surgeons with reduced length of stay and postoperative complications. (Representative: geriatric surgical collaboration studies, e.g., JAMA Internal Medicine and related journals.)
Centers for Medicare & Medicaid Services / CPT. Interprofessional telephone, internet, and EHR consultation codes 99446–99449, 99451, and 99452 are recognized and paid by Medicare, with behavioral-health interprofessional consult HCPCS codes G0546–G0551 added in the 2025 fee-schedule cycle. https://www.aaos.org/aaosnow/2025/jul/managing/managing01/
Note on figures: The masthead label was corrected from “Issue 04” to “Article 4,” and internal references to “the first piece” were generalized to “earlier in this series” so the article reads correctly regardless of final publication order. Reference 3 is intentionally framed as an evidence base rather than a single study; a specific citation can be pinned before publication.