CMS-HCC Risk Adjustment Model – Proposed changes in 2017
The Centers of Medicare and Medicaid Services – Hierarchical Condition Category or more commonly refereed to as the acronym CMS-HCC, in response to concerns about the accuracy of CMS-HCC risk adjustment model for predicting costs of dual eligible beneficiaries, CMS has undertaken an evaluation to assess how well the model performs for these beneficiaries. CMS will share the analysis with stakeholders and, if appropriate, propose modifications to the model to improve predictive accuracy in a future year?s process.
The memo shall serve the following purposes:
- Findings from the analysis to date, describes model development work CMS has been conducting in response to these findings and provides an early update on their plans for a revised model to be proposed for Payment Year 2017.
- The CMS-HCC risk adjustment model is used to calculate risk scores to adjust payments made for aged and disabled beneficiaries enrolled in Medicare Advantage (MA) plans and certain demonstrations. The CMS-HCC model has been calibrated using two full risk segments with separate coefficients to reflect the unique cost patterns of beneficiaries in the community and beneficiaries residing in long term care institutional facilities.
- The CMS-HCC risk adjustment model is prospective: it uses health status in a base year to predict costs in the following year.
- To measure model performance, CMS calculates predictive ratios for key subgroups of beneficiaries.
- A predictive ratio?the ratio of a group?s predicted cost to its actual cost? measures the accuracy of the model in predicting the average cost of a group. A predictive ratio close to 1.0 indicates that the model is accurately predicting that group?s average cost. A ratio greater than 1.0 indicates over-prediction, while a ratio less than 1.0 indicates under-prediction.
- The model predicts accurately overall, as well as for diseases and characteristics included in the model.
- Model Development: In order to improve the predictive ratios for full benefit and partial benefit dual eligible beneficiaries in the community, CMS is considering changes to the structure of the model. Specifically, rather than using a single segment for all community beneficiaries that includes factors for Medicaid status, we are developing a model that includes separate community segments for the following six populations:
- Full benefit dual aged;
- Full benefit dual disabled;
- Partial benefit dual aged;
- Partial benefit dual disabled;
- Non-dual aged; and
- Non-dual disabled.Full benefit dual eligibles are those who are eligible for full Medicaid benefits under title XIX of the Social Security Act.