2027 Medicare Advantage Risk Adjustment: Why Encounter-Linked Documentation Matters More Than Ever

Introduction: Medicare Advantage risk adjustment is moving closer to the encounter

Beginning with calendar year 2027 risk score calculation, CMS is generally excluding diagnoses submitted through unlinked chart review records from risk score calculation, with an exception for certain beneficiaries who switch from one Medicare Advantage organization to another.

The timing matters now. Medicare Advantage risk adjustment is prospective, and final risk-adjusted payments for 2027 will use diagnoses from 2026 dates of service.

For health plans, ACOs, MSOs, physician groups, health systems, CDI teams, and risk adjustment leaders, this is more than a technical encounter-data change. It raises a larger operational question: can a supported condition be identified, evaluated, documented, and connected to the appropriate encounter while the clinical context is still available?

That question points to a broader shift in risk adjustment. Finding a potential HCC is no longer the difficult part. The harder problem is turning relevant longitudinal evidence into complete, encounter-linked, defensible documentation without creating more work for clinicians.

1. What is an unlinked chart review record?

Medicare Advantage organizations submit encounter data to CMS describing items and services provided to beneficiaries. CMS also permits chart review records, or CRRs, to add or delete diagnosis information used in risk adjustment.

A linked chart review record connects diagnosis information to a previously submitted encounter data record. With an unlinked CRR, the submitted diagnosis information is not associated with a specific previously submitted encounter data record.

For CY 2027 risk score calculation, CMS is excluding diagnoses from unlinked CRRs, with an exception for beneficiaries who switch from one Medicare Advantage organization to another.

Medicare Advantage organizations may continue to submit unlinked CRRs. The important change is that, outside the finalized exception, diagnoses from those records will not be used in calculating CY 2027 risk scores.

2. Why the 2027 change matters beyond encounter-data submission

The immediate policy change concerns the source of diagnoses used for risk score calculation. The operational implication is broader.

Many risk adjustment programs still separate opportunity identification from the clinical encounter. Analytics identify potential conditions, coding or CDI teams review them, providers may be contacted later, and compliance teams evaluate support further downstream.

Every handoff creates another point where context can be lost. A patient may have substantial longitudinal evidence of chronic disease, but evidence existing somewhere in the record is not the same as having an appropriately documented diagnosis connected to the relevant encounter.

The organizations best positioned for this environment will increasingly need to connect evidence discovery with the encounter itself rather than depend on retrospective discovery alone.

3. Prospective risk adjustment has to be more than a suspect list

Prospective risk adjustment is often described as finding HCC opportunities before a visit. That is useful, but simply moving a longer suspect list into the EHR does not solve the underlying workflow problem.

Before an opportunity reaches a clinician, a strong workflow should establish whether there is sufficient clinical evidence to warrant review, whether the opportunity is appropriate for the patient and encounter, and what still requires provider judgment or documentation.

That distinction matters because a diagnosis appearing in a longitudinal record does not automatically make it appropriate for the current encounter.

The objective should not be to maximize how many conditions are shown to clinicians. It should be to bring forward the most relevant clinical decisions while filtering unnecessary provider noise.

• Is there sufficient clinical evidence to warrant review?
• Is the opportunity appropriate for this patient and encounter?
• What, if anything, still requires provider judgment or documentation?

4. Evidence should come before the HCC opportunity

Risk adjustment programs have historically measured success by how many opportunities they can find. But an opportunity has little value if the underlying diagnosis cannot ultimately be supported.

Longitudinal data can contain old diagnoses, incomplete documentation, conflicting information, and levels of specificity that are not appropriate for a new encounter. Those distinctions need to be considered before information reaches the provider or moves downstream to coding.

This is why evidence governance matters. The workflow should be capable of separating clinically relevant opportunities from diagnoses that require more support or should not move forward.

Done well, this protects both sides of risk adjustment. Organizations can surface meaningful opportunities while reducing unnecessary queries, inappropriate capture, and downstream rework.

• Conditions that are historical, resolved, or otherwise not current
• Opportunities that lack sufficient or consistent clinical support
• Diagnoses or specificity that are not appropriate for the current encounter

5. Better risk adjustment should mean fewer provider interruptions

Providers do not need another list of HCCs to reconcile. They need relevant clinical context at the point where an actual decision can be made.

Much of the preparation can happen before the clinician is involved. Longitudinal information can be organized, relevant evidence can be brought forward, and documentation needs can be clarified before the provider sees anything.

The provider should remain responsible for clinical judgment. Technology should reduce the administrative work surrounding that judgment rather than try to replace it.

That changes the experience from asking a provider to investigate a list of potential conditions to presenting the relevant clinical context and the specific question that still requires their decision.

6. Finding the condition and making it coding-ready are different problems

Even when a provider determines that a condition is current, the workflow may not be finished.

A clinically appropriate diagnosis can still have incomplete documentation, insufficient context, or other gaps that create uncertainty downstream. That is why prospective clinical documentation improvement and coding readiness need to operate as part of the same workflow.

The goal is not more documentation for its own sake. It is enough clear, clinically appropriate documentation for the resulting diagnosis to move downstream with confidence.

Resolving those gaps while the encounter is still active is fundamentally different from discovering them weeks or months later during coding review, reconciliation, or audit preparation.

7. A provider click is not proof

One of the most important distinctions in risk adjustment is the difference between workflow activity and documentation proof.

A provider can acknowledge an alert. A CDI specialist can close a query. A coder can review an opportunity. A workflow can mark an item complete.

None of those events necessarily proves that the resulting medical record actually contains documentation supporting the diagnosis.

A stronger control is to verify the resulting record after the action occurs rather than simply record that somebody interacted with the workflow. The relevant question is whether the final documentation supports the diagnosis and remains complete when the record is signed.

That turns documentation improvement from an activity-tracking exercise into a closed-loop control.

8. Why signed-record verification changes the control model

The signed medical record is ultimately more important than the workflow that preceded it.

Documentation can change between an initial provider action and final signature. Information can be added, removed, revised, or fail to appear where teams expected it to appear.

That makes final-record verification an important part of documentation integrity. Instead of assuming an intervention worked, organizations can determine whether the resulting record actually supports downstream coding and review.

For CDI, coding, compliance, and risk adjustment teams, that creates a common definition of completion: not that somebody acted, but that the required documentation exists in the record.

9. What this means for RADV and audit readiness

Medicare Advantage Risk Adjustment Data Validation puts the focus on whether diagnoses used for payment are supported by medical-record documentation.

That means audit readiness should not begin when an audit request arrives. Organizations benefit from preserving the connection between the clinical evidence, the encounter, the resulting documentation, and the final record as part of the operating workflow.

CMS has also continued expanding contract-specific RADV audit activity. Its Payment Year 2024 audit methodology is designed to calculate extrapolated overpayment amounts, although the legal and collection framework around extrapolation remains unsettled. CMS has stated that the prior rule governing extrapolation was vacated by a federal district court and is on appeal, and that it has not yet decided whether PY 2024 audits will collect extrapolated amounts or only overpayments associated with sampled enrollees.

Regardless of how that legal question develops, the operational lesson is the same: organizations need confidence in the diagnosis-level documentation they submit and the ability to produce the supporting record without reconstructing the reasoning months or years later.

10. The stronger model is prospective documentation control

Retrospective analytics will continue to matter. Organizations still need to identify missed opportunities, perform quality assurance, examine coding patterns, and understand where documentation is breaking down.

But retrospective discovery cannot replace the moment when a provider can appropriately evaluate the patient's current clinical status.

The stronger operating model connects longitudinal evidence with prospective clinical action and then verifies the outcome.

This is the shift from finding more HCCs to controlling the documentation workflow from evidence through the final record.

• Longitudinal evidence
• Supported clinical opportunity
• Appropriate encounter
• Provider decision
• Complete documentation
• Coding readiness
• Signed-record verification
• Audit-ready evidence

11. Questions risk adjustment leaders should be asking for 2027

The change to unlinked CRRs provides a timely reason to look beyond submission mechanics and evaluate how the entire documentation workflow operates.

The important question is not whether an organization has prospective analytics. It is whether evidence, provider workflow, documentation, coding, and compliance operate as one connected process.

• Can we identify clinically supported opportunities before or during an appropriate patient encounter?
• Can we keep low-confidence or inappropriate opportunities from creating provider noise?
• Can providers evaluate relevant clinical evidence without leaving their normal EHR workflow?
• Can we identify documentation gaps while the encounter is still actionable?
• Can CDI and coding teams distinguish an identified opportunity from one that is actually documentation-ready?
• Can we verify the resulting record instead of assuming an intervention worked?
• Can compliance teams trace supporting evidence back to the encounter and final documentation?
• Can leadership see where opportunities are being resolved, held, or lost before year-end reconciliation?

12. What the 2027 policy does not mean

The change does not mean retrospective risk adjustment disappears. Retrospective analysis remains useful for identifying missed conditions, detecting workflow problems, conducting QA, and improving future performance.

It also does not mean all chart review records are excluded. The finalized policy addresses diagnoses from unlinked CRRs, and CMS retained an exception involving beneficiaries who switch from one Medicare Advantage organization to another.

And it does not eliminate the need for coding or CDI expertise. Instead, it increases the value of connecting that expertise to the clinical encounter earlier, while documentation can still be completed appropriately.

The larger shift is from relying on retrospective discovery as the primary control to building a workflow that can act on supported evidence prospectively and verify the result.

13. How SynchroLink AI approaches the problem

SynchroLink AI is designed as a clinical evidence and workflow control layer connecting the patient record with provider documentation, CDI, coding readiness, and downstream audit evidence.

Before the encounter, SynchroLink reviews available longitudinal clinical information to help teams identify relevant documentation opportunities while filtering unnecessary provider noise.

During the encounter, Provider Companion brings relevant evidence and unresolved clinical questions into the EHR so providers can make the decisions that require clinical judgment without moving to a separate workflow.

After the encounter, SynchroLink helps teams determine whether documentation is ready to move downstream and verifies the resulting record rather than treating workflow activity as proof of completion.

The objective is not simply to find more HCCs. It is to help organizations get supported documentation right while the patient and clinical context are still available, then maintain visibility from the encounter through coding and audit readiness.

The shift: from retrospective discovery to encounter-linked documentation control

CMS's treatment of unlinked chart review diagnoses in 2027 is one more signal that Medicare Advantage organizations should examine where risk adjustment work happens and how diagnoses connect back to actual encounters.

Finding a potential diagnosis after an encounter can still create useful intelligence. But identifying relevant evidence before or during the encounter, enabling an appropriate provider decision, completing the documentation, determining coding readiness, and verifying the final record creates a much stronger control environment.

Risk adjustment performance should not depend on finding more opportunities after the fact. It should depend on making supported opportunities actionable at the right moment and knowing what actually survived into the record.

That is the opportunity for the next generation of prospective documentation workflows: less provider noise, stronger coordination across CDI and coding, and clearer evidence from the clinical encounter through audit readiness.

CTA

If your risk adjustment workflow still separates opportunity identification, provider documentation, CDI, coding readiness, and audit proof, it is worth pressure-testing where those handoffs create gaps.

See how SynchroLink AI brings those workflows together around the clinical encounter without adding another destination for providers.