Private Payers News

Medicare Advantage Risk Adjustment Model Accuracy May Vary By Race

The Medicare Advantage risk adjustment model consistently underestimated American Indian/Alaska Native beneficiaries’ costs.

Medicare Advantage, risk assessment, CMS, chronic disease

Source: Getty Images

By Kelsey Waddill

- The CMS-Hierarchical Condition Category (HCC) model used for Medicare Advantage risk adjustment may inaccurately estimate beneficiaries’ spending depending on the race or ethnicity of the beneficiary population, an Avalere study found.

The researchers tested the CMS-HCC model by Medicare fee-for-service Parts A and B files for 2019. The study considered all HCPCS codes in 2018 that could be risk-adjusted and claim spending. They partnered with Inovalon Inc. to access the CMS data and received funding from Arnold Ventures.

With this data, the researchers compared the actual FFS beneficiary spending to the projections the CMS-HCC community model made. They assigned a ratio to the results. A ratio of 1.0 meant that the costs were accurately predicted, any higher ratio indicated that the projections were overestimated, and lower ratios showed how much the projections underestimated.

“Previous research shows that, because the model predicts healthcare costs based on FFS Medicare spending, it can underpredict costs for beneficiaries who have had low spending as a result of systemic access barriers. In other words, the model coefficients may build inequities into the MA plan payment structure and perpetuate health disparities,” the researchers noted at the beginning of the study.

The model produced varying results based on race and ethnicity, the researchers found.

For non-Hispanic White beneficiaries and Black beneficiaries, the model’s predictions closely aligned with the beneficiaries’ actual costs. Non-Hispanic White beneficiaries had a ratio of 0.99, and Black beneficiaries had a ratio of 1.01.

However, for minority populations excluding Black beneficiaries, the results were less accurate. Asian/Pacific Islander and Hispanic beneficiaries’ healthcare spending was overestimated. The model produced a ratio of 1.25 for Asian/Pacific Islanders and 1.09 for Hispanic beneficiaries.

The CMS-HCC model consistently missed the mark on healthcare spending among American Indian/Alaska Native beneficiaries. Overall, the model underestimated American Indian/Alaska Native spending, giving a ratio of only 0.93. For partial dual eligible beneficiaries in this population, the model achieved a ratio of 0.98, and for full dual eligibles, the ratio was 0.95.

Although the model accurately represented costs for Black beneficiaries, it did not accurately portray costs for Black dual eligible beneficiaries, whether fully or partially dual eligible. For Hispanic dual eligible beneficiaries, the model continued to overpredict costs but at a lower rate than for non-dual eligible beneficiaries.

For chronic disease spending, the model underpredicted spending for Black beneficiaries with eight out of ten of the most common HCC diseases, and it overpredicted spending for Hispanic beneficiaries.

However, if beneficiaries have five HCCs or more, some of the prevailing trends flip. The model overpredicted spending for non-Hispanic White beneficiaries and underpredicted spending for Black and Hispanic beneficiaries with five conditions or more. The researchers attributed these patterns to the 2016 changes to the 21 Century Cures Act which affected cost predictions for multiple HCCs.

Avalere noted that CMS changed the CMS-HCC model in the 2024 Medicare Advantage and Part D Advanced Notice, which could affect future trends.

“An optimal risk-adjustment model accurately predicts costs overall and for subpopulations, including racial/ethnic minorities. Avalere’s research indicates that the CMS-HCC risk-adjustment model may incorrectly predict costs for certain subpopulations, which might perpetuate disparities by overpaying for some low-cost populations and underpaying for some high-cost groups of beneficiaries,” the researchers concluded.