Inferred Risk: Reforming Medicare Risk Scores To Create A Fairer System
Abe Sutton Gabriel Drapos
April 24, 2024
Chat GPT 4
TL;DR: Reforming Medicare Risk Scores to Create a Fairer System
Authors: Abe Sutton, Gabriel Drapos
Published: April 24, 2024
Source: Health Affairs Forefront
Overview
The article critiques the CMS-HCC Risk Adjustment Model used to determine the health care costs of Medicare beneficiaries via the Risk Adjustment Factor (RAF).
While originally intended to avoid sicker patient discrimination, the system now faces issues like over-coding in Medicare Advantage (MA), inflated premiums, and burdensome documentation.
The authors propose a data-driven, inference-based approach for calculating RAF scores to improve accuracy, reduce administrative burdens, and ensure fairness across providers. The system would analyze claims, prescriptions, and demographics to infer diagnoses and assign risk scores, moving away from yearly manual documentation.
Key Advantages of an Inferred RAF System:
Improved Accuracy & Fairness
- Removes manual inefficiencies and levels the playing field for smaller entities.
- Accounts for social determinants of health.
Reduced Gameability
- Ties RAF to actual utilization, minimizing the incentive to manipulate scores.
Lower Administrative Burden
- Frees up resources for patient care by reducing documentation and audits.
Challenges:
Incentivizing Utilization
- Could unintentionally promote unnecessary care. [which counts as a disease signal]
- Solutions: Use lagged data or maintain the current system for Original Medicare while piloting inferred RAF for MA.
Historical Undertreatment
- Underserved patients may appear healthier in inferred models.
- Solutions: Incorporate broader demographic data to adjust scores.
Implementation Hurdles
- Requires testing (via CMS Innovation Center models like ACO REACH) and phased adoption.
Proposed Pilot:
The CMS Innovation Center could test this model by adapting ACO REACH, which already uses baseline risk scores, to include inferred RAF. After testing and refinement, a phased implementation could allow a smooth transition.
Conclusion:
A reform of the RAF system is necessary to achieve greater equity, efficiency, and transparency. The proposed approach leverages data science to create a more precise and less burdensome model, reducing taxpayer costs while improving care quality.
Author Disclosures: Both authors work for organizations that may be impacted by changes to RAF systems.
