CMS Bans Unlinked Chart Reviews: $7B Hit to MA Plans Starting CY 2027

From analyzing risk adjustment submissions across a dozen MA contracts over multiple payment years, the pattern was visible long before CMS acted: organizations with heavy unlinked chart review record reliance consistently showed diagnosis-to-encounter ratios that diverged from provider-sponsored plans by 15 to 20 percentage points. The question was never whether CMS would address this divergence, but when, and how much financial disruption the correction would create.

On April 2, 2026, CMS published the Contract Year 2027 Medicare Advantage and Part D final rule, confirming the exclusion first proposed in the January 26, 2026 Advance Notice. Diagnoses submitted on unlinked chart review records will no longer factor into CMS-HCC risk score calculations for payment year 2027. CMS estimates the exclusion will reduce MA payments by $7.12 billion, representing a 1.53% aggregate payment reduction. For an industry serving 35.7 million beneficiaries and processing over $450 billion in annual capitated payments, the policy represents the most significant single-year risk adjustment methodology change since the V28 model phase-in began in CY 2024.

What Unlinked Chart Review Records Are and Why They Existed

To understand the magnitude of this change, plan actuaries need to distinguish between the two types of chart review records that MA organizations submit through the CMS Encounter Data System.

Linked chart review records tie back to a previously submitted encounter data record (EDR). When a plan conducts a retrospective medical record review and identifies a diagnosis that was present during a documented face-to-face visit but was not captured on the original claim, the linked CRR corrects that gap. The diagnosis is associated with a specific beneficiary encounter on a specific date of service. CMS considers these legitimate documentation completeness tools, and they remain fully eligible for risk adjustment in CY 2027.

Unlinked chart review records do not reference any previously submitted EDR. They introduce diagnoses into the risk adjustment data without tying them to a particular billable service or face-to-face encounter. In practice, this meant that a plan could review a beneficiary's medical records, identify chronic conditions documented somewhere in the chart, and submit those diagnoses for risk adjustment regardless of whether the beneficiary received a service related to that condition during the payment year.

The mechanism was straightforward: third-party coding vendors would review medical records in bulk, often working from scanned documents rather than live clinical encounters. They would identify hierarchical condition categories that appeared in the documentation but had not been captured on encounter submissions. These diagnoses would then be submitted as unlinked CRRs, increasing the beneficiary's risk score and, by extension, the plan's capitated payment from CMS. The scale was staggering: in 2023 alone, MA organizations submitted 88.8 million unlinked chart review records for 2024 payment. CMS analysis found that 85% of those records could not be matched to any encounter data record, even when matching on beneficiary, billing provider, and dates of service within a three-day window.

From a revenue optimization perspective, the economics were compelling. A single additional HCC capture could increase a beneficiary's annual risk-adjusted payment by $3,000 to $12,000 depending on the condition hierarchy. Multiplied across hundreds of thousands of beneficiaries, unlinked CRR programs generated substantial incremental revenue with relatively low operational cost. The coding vendors typically worked on a percentage-of-revenue-capture basis, aligning their incentives squarely with volume of added diagnoses rather than clinical accuracy.

The Scale of Reliance: OIG Findings and Industry Adoption

The HHS Office of Inspector General published its foundational report on chart review practices in December 2019, examining payment year 2017 data (2016 dates of service). The findings quantified what plan actuaries already suspected from internal benchmarking.

Key statistics from the OIG report:

Finding	Value	Implication
Diagnoses reported only on chart reviews (no service records)	$6.7 billion in risk-adjusted payments (2017)	Over 1% of total MA program spending attributable to chart-review-only diagnoses
Payments from unlinked CRRs specifically	$2.7 billion estimated (2017)	Diagnoses with no documented service to the beneficiary in the payment year
Chart review direction	Over 99% added diagnoses	Functionally a one-way revenue ratchet; deletions almost never occurred
MAOs with payments driven by unlinked CRRs	Approximately one-third of all MAOs	Concentrated but widespread practice

By 2022, CMS data showed that 58% of MA contracts submitted unlinked chart review records. The practice had expanded significantly since the OIG's 2017 payment year analysis. Whether measured by contract count or beneficiary coverage, unlinked CRRs had become embedded in the revenue infrastructure of a majority of MA organizations.

CMS disputed some of the OIG's methodology in its response to the 2019 report, estimating the actual overpayment value at approximately $675 million rather than $2.7 billion after applying proper blending adjustments between encounter data and RAPS submissions. This methodological disagreement, however, did not alter the directional conclusion: unlinked CRRs were generating risk-adjusted payments for diagnoses without corresponding clinical encounters.

MedPAC's March 2026 Report to Congress provided broader context. MA payments in 2026 exceed what traditional Medicare would spend for the same beneficiaries by $76 billion (14% above FFS equivalent spending). Coding intensity in MA runs 10.3% higher than fee-for-service. While unlinked chart reviews represent only one component of the coding intensity gap, they are the component most clearly disconnected from clinical activity, making them the natural first target for regulatory correction.

Quantifying the 1.53% Payment Reduction

CMS calculated the $7.12 billion impact by modeling the removal of all unlinked CRR diagnoses from the CY 2027 risk score computation. The 1.53% figure represents the aggregate payment reduction across all MA contracts, but the distribution is highly non-uniform.

Plans fall into distinct exposure tiers:

High exposure (estimated 2.5% to 4.0% revenue reduction): National plans with centralized coding operations that built systematic unlinked CRR programs at scale. These organizations typically contracted with large retrospective coding vendors (Optum360, Cotiviti, Inovalon, Reveleer) and operated chart review programs covering 80% or more of their membership. Their diagnosis capture rates from unlinked CRRs frequently exceeded 0.15 additional HCCs per member per year.

Moderate exposure (estimated 1.0% to 2.5% revenue reduction): Regional plans that adopted unlinked CRR programs selectively, often targeting high-acuity members or specific condition categories (diabetes with complications, chronic kidney disease, heart failure hierarchies). These plans typically had smaller vendor contracts and lower penetration rates across their books.

Low exposure (estimated 0% to 1.0% revenue reduction): Provider-sponsored plans and staff-model HMOs where documentation improvement programs were embedded in clinical workflows rather than operating through retrospective chart review. These organizations captured diagnoses primarily through face-to-face encounters and linked CRRs tied to documented visits. Kaiser Permanente's model, where physicians and coders work within the same integrated system, exemplifies this archetype. Provider-sponsored plans that relied on their network physicians for prospective documentation accuracy rather than retrospective vendor review will see minimal payment disruption.

The practical consequence is that the 1.53% average obscures significant variation. A national plan with heavy unlinked CRR dependence faces a materially different bid recalibration challenge than a provider-sponsored plan that barely utilized the mechanism. For the most exposed organizations, the revenue reduction may exceed the entire margin they built into their CY 2026 bids.

The Narrow Exception: Beneficiaries Switching MA Organizations

In response to industry comments during the February 2026 comment period, CMS introduced a limited exception to the unlinked CRR exclusion. When a beneficiary transitions from one MA organization to another, the receiving plan may submit unlinked chart review records for that beneficiary's pre-existing conditions.

The rationale is straightforward: a new plan inheriting a beneficiary has no encounter history with that individual. Without access to unlinked CRRs, the receiving plan would need to wait for the beneficiary to present for clinical services before capturing chronic conditions that were actively managed by the prior plan. For conditions like chronic kidney disease stage IV or major depressive disorder (recurrent), a delay in diagnosis capture could create a multi-month gap where the plan bears the cost of managing these conditions without receiving the corresponding risk-adjusted payment.

The financial offset from this exception is more substantial than it may appear. CMS quantified the difference: the average payment impact with the switcher exception is negative 1.24%, compared to negative 1.78% without it. That 0.54 percentage point spread translates to approximately $2.5 billion in preserved payments, reducing what would have been a larger gross exclusion impact down to the finalized $7.12 billion net estimate. The exception does not apply when a beneficiary transitions from original Medicare (fee-for-service) to MA; it is narrowly defined to cover transitions between MA organizations only.

For plan actuaries modeling the exception's value, the key variable is member churn rate. Plans in markets with high MA plan switching (often driven by Star Ratings changes, benefit reductions, or service area exits by competitors) will capture more value from the exception than plans in stable enrollment markets.

Revenue Projection Recalibration for CY 2027 Bids

The June 1, 2026 bid submission deadline leaves plan actuaries approximately eight weeks from the April 6 final rule publication to quantify and incorporate the unlinked CRR exclusion into their revenue projections. This is not a simple top-line adjustment; it requires granular analysis at the contract and plan benefit package (PBP) level.

The recalibration framework involves four steps:

Step 1: Identify historical unlinked CRR volume. Pull all chart review record submissions for payment years 2023 through 2025 and classify each as linked or unlinked based on whether the CRR references a previously submitted EDR. Calculate the proportion of total HCC captures attributable to unlinked CRRs by contract, by PBP, and by condition category.

Step 2: Quantify the risk score impact. For each beneficiary with unlinked CRR diagnoses, recalculate the CMS-HCC risk score excluding those diagnoses. The difference between the all-source risk score and the no-unlinked-CRR risk score, aggregated across the membership, gives the contract-level revenue at risk. Apply the CY 2027 benchmark rates to convert risk score reduction into dollar impact.

Step 3: Model the switching exception recovery. Using historical member churn data, estimate how many incoming switchers would generate unlinked CRR submissions under the exception. Apply the same per-member risk score differential methodology to quantify the partial offset. For most plans, this recovery will be 5% to 15% of the gross unlinked CRR revenue loss.

Step 4: Adjust the bid revenue line. The net revenue reduction (gross unlinked CRR loss minus switching exception recovery) must flow through to the bid's A/B revenue projection. This affects the plan's ability to fund supplemental benefits, maintain premium stability, and generate target margins. Plans with thin margins may need to reduce supplemental benefit offerings (dental, vision, hearing, fitness, OTC allowances) to maintain actuarial soundness in the bid.

Plans that began constructing their CY 2027 bids against the January Advance Notice's 0.09% baseline faced the additional complexity of CMS reversing to the 2.48% final rate. The rate increase partially offsets the CRR exclusion for most plans, but the net effect varies by plan archetype. A plan with 3% revenue exposure to unlinked CRRs receiving a 2.48% rate increase still faces a net 0.5% revenue reduction relative to CY 2026 baseline, before accounting for medical trend.

Strategic Responses: From Retrospective Review to Prospective Documentation

The unlinked CRR exclusion does not eliminate chart review as a risk adjustment tool. It eliminates one specific mechanism (diagnoses unconnected to documented encounters) while preserving the broader ecosystem of documentation improvement strategies. Plans have several strategic paths forward.

Invest in provider documentation quality at the point of care. The highest-ROI response is shifting coding improvement investment upstream. Rather than paying vendors to find missed diagnoses in records after the fact, plans can deploy clinical documentation improvement (CDI) specialists into provider practices, embed coding logic into electronic health record (EHR) workflows, and create provider incentive programs tied to documentation completeness at the encounter level. Every diagnosis captured prospectively during a face-to-face visit is fully eligible for risk adjustment under the new rules.

Convert unlinked programs to encounter-based models. Some chart review vendors are repositioning their services. Rather than conducting retrospective reviews disconnected from encounters, they now facilitate "chart review encounters" where a clinician reviews the medical record and conducts a brief assessment, creating a billable service that can anchor the diagnosis to a dated encounter. CMS has signaled skepticism toward in-home health risk assessments (HRAs) that serve primarily as diagnosis capture vehicles (a separate $7.5 billion annual payment category per Georgetown analysis), so plans pursuing this strategy should ensure the encounters involve genuine clinical value beyond coding.

Redesign value-based provider contracts to incorporate documentation metrics. Plans with delegated risk arrangements can add documentation quality measures to their provider scorecards. When a primary care physician captures all active HCCs prospectively during annual wellness visits and chronic care management encounters, the plan achieves the same risk score accuracy that retrospective chart review provided, without the regulatory vulnerability. DLA Piper's February 2026 analysis specifically highlighted this approach as the most sustainable path: converting the CRR exclusion from a pure payment loss into a catalyst for deeper plan-provider integration.

Accept the revenue reduction in specific markets and reassess viability. For plans operating in counties where margins were already thin, the combined effect of the CRR exclusion, rising medical trend, and Star Ratings changes may push certain plan benefit packages below breakeven. Plans facing this math will need to decide whether to exit specific service areas, reduce supplemental benefits to the minimum viable level, or cross-subsidize from higher-margin markets. The pattern of service area exits observed in 2025 and 2026 (UnitedHealthcare withdrawing from 109 counties, Humana from 194 counties, Aetna from 100 counties) may accelerate if plans cannot replace unlinked CRR revenue through alternative coding strategies.

The Legislative Context: No UPCODE Act and Long-Term Trajectory

The CMS administrative action on unlinked CRRs exists within a broader legislative environment targeting MA coding practices. The No UPCODE Act (No Unreasonable Payments, Coding, or Diagnoses for the Elderly Act), introduced in March 2025 by Senators Cassidy and Merkley, proposes more aggressive reforms including using two years of diagnostic data for risk adjustment rather than one year and limiting the ability to use old or unrelated conditions in risk score calculations.

The Congressional Budget Office scored the No UPCODE Act's provisions at $124 billion in savings over ten years. That figure dwarfs the $7 billion single-year impact of the CRR exclusion, indicating that CMS's administrative action is a down payment on a larger correction rather than the final word.

For plan actuaries projecting beyond CY 2027, the trajectory points toward continued tightening of diagnosis sourcing rules. The CMS decision to defer the risk adjustment model recalibration (using 2023 diagnoses and 2024 expenditures) to a future payment year means that additional risk score reductions remain pending. When that recalibration eventually takes effect, plans that have already diversified away from unlinked CRRs toward encounter-based documentation will face less incremental disruption than those that found workarounds to preserve retrospective coding revenue.

MedPAC's position adds pressure. The commission's consistent finding that MA coding intensity exceeds FFS by over 10%, generating $76 billion in annual excess payments, provides the analytical foundation for further administrative or legislative action. CMS's own minimum coding intensity adjustment, which reduces aggregate risk scores to partially offset favorable coding patterns, has been criticized by MedPAC as insufficient. Whether CMS increases that adjustment in CY 2028 or beyond depends partly on whether the CRR exclusion measurably narrows the coding intensity gap.

Implications for Risk Adjustment Data Validation (RADV) Audits

The unlinked CRR exclusion intersects with CMS's ongoing Risk Adjustment Data Validation audit program. RADV audits verify whether diagnoses submitted for risk adjustment are supported by medical record documentation. Under the pre-2027 framework, unlinked CRR diagnoses were subject to RADV validation alongside encounter-based diagnoses.

Starting in CY 2027, the interaction changes. Unlinked CRR diagnoses will not contribute to risk scores, so they cannot generate RADV overpayment findings. However, the OIG's active work plan includes "Audits of Medicare Part C Unlinked Chart Review Diagnosis Codes" (project SRS-A-25-018), targeting payment years prior to CY 2027 where these diagnoses did contribute to risk scores. Plans with material historical unlinked CRR submissions face retrospective audit exposure even as the prospective payment mechanism closes.

The actuarial reserve implication is straightforward: plans should maintain contingent liability estimates for RADV audit exposure related to historical unlinked CRR submissions. Given OIG's $2.7 billion aggregate estimate for a single payment year, and CMS's narrower $675 million methodology-adjusted figure, individual plan exposure depends on the volume and supportability of their historical submissions. Plans that operated large-scale unlinked CRR programs with third-party vendors should assess whether their documentation would survive a medical record review under RADV's one-best-code methodology.

Impact on the MA Competitive Landscape

The non-uniform distribution of unlinked CRR reliance creates competitive dynamics that will reshape market share over time. Plans that built their margin assumptions on retrospective coding revenue face a structural disadvantage relative to plans whose pricing already reflected encounter-based documentation accuracy.

Provider-sponsored plans and staff-model organizations, which historically relied less on unlinked CRRs due to their integrated documentation workflows, gain a relative advantage. Their cost structures already incorporated the provider investment needed for prospective coding accuracy. They do not face the transition cost of building new documentation improvement infrastructure from scratch.

National plans with large retrospective coding operations face both the revenue loss and the investment requirement. Converting from a vendor-driven retrospective model to a provider-embedded prospective model requires capital deployment (CDI specialist hiring, EHR integration, provider incentive funding) during the same period that revenue is declining. The timing mismatch creates a multi-quarter earnings headwind that may affect stock prices, credit ratings, and the ability to bid competitively on supplemental benefits.

The MA enrollment market in 2026 already showed strain: three of the five largest national plans reduced service areas, and total MA enrollment growth decelerated to 3% year-over-year (35.7 million in January 2026, per CMS monthly enrollment data). If the CRR exclusion triggers additional service area exits or benefit reductions by national plans, provider-sponsored plans in those markets may absorb displaced membership, accelerating a structural shift toward locally integrated care models.

Why This Matters for the Actuarial Profession

The unlinked CRR exclusion forces a reexamination of fundamental assumptions embedded in MA plan actuarial models. Revenue projections that treated unlinked chart review captures as a stable, recurring revenue component must now treat them as zero for CY 2027 forward. This is not a trend assumption or a sensitivity parameter; it is a binary policy change that eliminates an entire revenue category.

For pricing actuaries, the bid recalibration is immediate and deadline-driven. The June 1, 2026 submission date creates compressed timelines for the granular analysis described above. Plans that had already incorporated conservative assumptions about CRR sustainability into their prior bids (patterns we have seen increasingly among well-managed regional plans since the January 2026 Advance Notice) face less rework than those that assumed the proposal would be withdrawn or significantly modified during the comment period.

For valuation actuaries, the change affects MA block valuations in M&A contexts, embedded value calculations, and goodwill impairment testing. An MA block's revenue sustainability depends partly on the durability of its risk adjustment practices. Blocks heavily dependent on unlinked CRRs are worth less in CY 2027 forward than historical earnings would suggest.

For consulting actuaries advising MA plans, the CRR exclusion creates a rich engagement cycle: CRR revenue exposure quantification, bid recalibration support, documentation improvement program design, RADV contingent liability estimation, and strategic market assessment work all flow from a single policy change. The plans that invested in encounter-based documentation quality before CMS mandated it will navigate this transition with the least disruption. For everyone else, the eight weeks between the April 6 final rule and the June 1 bid deadline define the window in which the actuarial work must be completed.