Rethinking Risk Adjustment: Treading Carefully into the Next Chapter of Risk Adjustment (Part II)

As CMS finalizes the 2026 payment policy updates for Medicare Advantage and Part D programs, the shift toward using MA encounter data for risk adjustment is gathering momentum. While Part I of our series elaborated on the theoretical benefits of a more data-driven approach, the finalized changes also briefly outlined the pitfalls in Part I.

In this part two of the series, we will do a deep dive and examine the shortfalls & limitations of relying solely on MA encounter data and explore the broader implications this shift may have on the future of risk adjustment.

MA Encounter Data-Based Model: Worth a Thoughtful Revisit

1. Data Quality and Inconsistencies

A. Historical Data Gaps and Incomplete Reporting

A significant concern with the transition to an MA encounter data-based model is the persistent issue of data completeness. Transitioning to this model is complex as it involves gathering, standardizing, and ensuring data quality. Despite efforts to improve reporting processes, many MA plans have not achieved the robust data capture necessary for accurate risk scoring.
This can lead to:​

• Underestimation of Patient Risk: Incomplete encounter data may result in lower risk scores, leading to under-compensation for plans serving higher-risk enrollees.​
• Inconsistent Plan Comparisons: Variability in data reporting standards across plans can undermine fair competition and obscure accurate performance metrics.​
  

B. Coding Variability and Impact on Risk Scores

Differences in clinical coding practices between MA plans challenge the consistency and reliability of risk adjustment calculations. Shifting the calibration basis to encounter data may cause issues. This variability can lead to:​

• Misaligned Incentives: Disparities in coding accuracy may inadvertently encourage over- or under-reporting diagnoses, affecting payment fairness.​ Plans might change their coding behaviors if the data accurately reflects resource use.
• Quality vs. Quantity Concerns: An emphasis on comprehensive data capture might shift focus away from quality improvement initiatives toward meeting reporting requirements.​ Patient selections may unintendedly be altered or coding practices might not always align with optimal clinical care.

2. Transparency and Oversight Concerns

A. Methodological Ambiguity in Data Mapping

Stakeholders have expressed the need for greater transparency in how CMS maps clinical information to condition categories within the risk adjustment model.
Without clear guidelines, plans face difficulties in validating risk score calculations, leading to:​

• Enhanced Disclosure Needs: Clear, publicly accessible methodologies are essential for plans to understand and trust the risk adjustment process.​
• Robust Oversight Mechanisms: Strong oversight is critical to ensure that discrepancies in coding practices do not translate into significant financial inequities.​
  

B. Limited Stakeholder Engagement

The brief 30-day window for public comment on the proposed changes has been deemed insufficient by many industry stakeholders. This limited timeframe restricts the ability to conduct thorough analyses and provide meaningful feedback, resulting in:​

• Hasty Reactions: Stakeholders may not have adequate time to assess the full impact of proposed changes, leading to potential oversights.​
• Increased Uncertainty: Limited engagement can result in a lack of consensus and clarity on the path forward, affecting long-term planning.
• Insufficient Input: Time constraints limit meaningful engagement, particularly from smaller or resource-constrained plans.
• Planning Disruptions: Lack of stakeholder consensus fosters uncertainty and hampers long-term operational alignment.
  

3. Unintended Consequences and Financial Implications

A. Impact on Payment Accuracy and Plan Revenue

Transitioning to an MA encounter data model may introduce volatility in plan payments due to potential inaccuracies in data reporting.
Concerns include:​

• Payment Volatility: Minor inaccuracies in encounter data could lead to significant risk-score fluctuations, complicating MA providers’ financial planning.​
• Short-term Financial Risks: Investments in data infrastructure and potential disruptions during the transition could adversely affect revenue streams.​
  

B. Challenges in Advancing Value-Based Care

While the shift aims to support value-based care models, reliance on encounter data may inadvertently reinforce fee-for-service (FFS) assumptions, hindering progress toward outcome-focused care. 

• Increased Administrative Burden: Aligning MA encounter data with traditional fee-for-service (FFS) benchmarks may increase administrative complexity, thereby delaying the benefits of value-based care initiatives.

4. Impact on Specialized Plans

A. Special Needs Plans (SNPs) and Dual Eligible SNPs (D-SNPs)

SNPs and D-SNPs serve beneficiaries with complex medical and social needs. Since an encounter-based model is designed to capture detailed clinical data, this change could reveal nuances in patient risk that the traditional model might have smoothed over.

• Complexity of Patient Profiles: Risk scores could be lower if the new data in practice does not fully capture the depth and chronicity of needs in their populations.

• Documentation Challenges: If the data submitted isn’t standardized or comprehensive, there is a greater risk of under-coding or misrepresentation.
  This may adversely affect reimbursement levels, making it even more critical for SNPs to ensure accurate documentation.

5. Key Updates from the CMS 2026 Final Payment Policy

The CMS 2026 Final Rule introduces several significant updates that impact MA and Part D plans:​

• Payment Increases: Payments to MA plans are expected to increase by an average of 5.06% from 2025 to 2026, a notable rise from the initially proposed 2.23%. This adjustment reflects updated data on fee-for-service expenditures.
• Completion of Risk Adjustment Model Phase-In: CMS is completing the three-year phase-in of the updated risk adjustment model, which began in CY 2024. This model aims to improve payment accuracy by better accounting for enrollee health status and expenditures.
• Part D Redesign Implementation: In alignment with the Inflation Reduction Act, CMS is implementing changes to the Part D program, including capping annual out-of-pocket prescription drug costs at $2,100 for beneficiaries in 2026.

6. Charting a Path Forward

To navigate these changes effectively, MA plans, and policymakers should consider the following strategies:​

• Invest in Data Integrity: Prioritize data accuracy and standardization enhancements for reliable risk adjustment calculations.​ Also, collaborate closely with EHR vendors to ensure systems capture all relevant clinical data consistently and accurately.
• Advocate for Transparent Methodologies: Engage with CMS to promote clear and open disclosure of data mapping and risk adjustment methodologies.​
• Training and Compliance Programs: Develop regular training sessions for providers and coders. By having regular internal audits, one can ensure that documentation aligns with the evolving CMS requirements.
• Develop Robust Financial Models that ensure the creation of adaptable, data-driven financial planning tools that help Medicare Advantage (MA) plans anticipate and respond to:
  
•  Volatile payment outcomes tied to encounter data accuracy
•  Fluctuations in risk scores due to data quality and coding changes
•  Policy shifts, like updates to the risk adjustment model and Part D redesign
•  Investment needs, such as data infrastructure and compliance upgrades

Gradually Phase In: Consider piloting the new encounter-based approach on a subset of data, allowing the organization to compare results with existing models.

This approach can help identify issues and facilitate smoother scaling. 

End Note

While CMS’s vision of a modernized, encounter-data-driven risk adjustment model is bold, the 2026 Final policy updates confirm what many feared: we’re not ready yet.

Data gaps, coding variation, and socioeconomic blind spots still persist.

Add to that new pressures from Health Equity Index (HEI) measures and Star Rating reforms, and it becomes clear that risk adjustment—if not carefully calibrated—could become more opaque, not less.

Plans that invest now in Advanced AI solutions vis-à-vis robust IT infrastructure, policy advocacy, and continuous learning will be best positioned to adapt and lead in this new era of Medicare Advantage.

As we await the Advance Notice for CY 2027, the industry must confront this critical outlook: Can MA encounter data stand on its own as the cornerstone of risk adjustment, or will hybrid models and transitional safeguards need to stay in play longer?

Source

CMS

CMS