The bars in the graph represent the percentage difference in rates between dual eligible and non-dual eligible members. A bar below 0% indicates dual eligible members perform worse than non-dual eligible members, while a bar above 0% indicates dual eligible members perform better.
Source: 2013 MORE2 Registry®
The more we study the risks associated with dual eligible members, the more interesting things get. For example, our recent study finds that a dual eligible member is at higher risk for hospital readmission compared to a non-dual eligible member with the same characteristics—and this added risk is not accounted for in the current Star Ratings.
Our study results, published in “An Investigation of Medicare Advantage Dual Eligible Member-Level Performance on CMS Five-Star Quality Measures,” demonstrate that dual eligible members have significantly worse outcomes on a majority of Medicare Advantage (“MA”) Star Rating measures evaluated. Researchers found that socio-demographic characteristics—such as living in a high poverty area—and community resource factors—such as a shortage of mental health professionals—were a main contributor to the disparity between dual eligible and non-dual eligible members, explaining 30% or more of the observed differences in outcomes.
Given the results, the report authors call for adjusting measures in the Star Rating system that account for clinical, socio-demographic and community resource factors to ensure a fair evaluation of plans and provide a more accurate quality comparison.
In 2013, Inovalon investigated the association of dual eligible populations and the performance of MA health plans. We found that dual eligible members had significantly lower scores compared to non-dual eligible members on nine of the 10 Star Measures evaluated.
The recent follow-up study endeavored to identify the factors associated with the lower Star Ratings among dual eligible members. Research included a claims database of more than 2.2 million MA beneficiaries, detailed sources of socio-demographic and community resource data, and monthly data on dual eligible status.
The study was conducted by Inovalon’s Division of Statistical Research, with additional funding provided by Cigna HealthSpring, WellCare, Healthfirst, Gateway Health, BCBS Minnesota and Blue Plus, and Health Care Service Corp.
This was the first study to leverage a database large enough to compare the outcomes of dual eligible members and non-dual eligible members enrolled in the same plan benefit package. That allowed the authors to answer the question: “To what degree are MA beneficiary outcomes related to the dual eligible status of individual plan members versus the quality of care provided by the plan?”
Here are the key findings, in brief:
- Dual eligible beneficiaries have significantly worse outcomes than non-dual eligible members enrolled in the same plan benefit package for five of the eight Star Measures analyzed.
- Worse outcomes of dual eligible members are not statistically related to the proportion of dual eligible members enrolled in a plan.
- Worse outcomes of dual eligible members appear to be related to a higher prevalence of risk factors among dual eligible members.
- Clinical and socio-demographic risk factors affect outcomes in all MA members, but the impact is magnified for MA health plans with higher numbers of dual eligible members.
- Differences in socio-demographic characteristics were consistently a main contributor to the differences in outcomes between dual eligible and non-dual eligible members
Importantly, the research shows that if MA Star Measures were adjusted to control for characteristics statistically associated with higher risk for the outcomes evaluated, the observed performance gaps between dual eligible and non-dual eligible members would be reduced by 70% or more.
Call to action
This investigation suggests that, under the current Star Rating system, MA plans serving disadvantaged members may be providing a higher quality of care than they currently appear to provide.
The results of the study support the need to adjust the Star Rating measures to account for clinical, socio-demographic and community resource factors. This would provide a more accurate comparison of quality across all MA plans by accounting for factors statistically associated with higher risk of worse outcomes. Consider:
- The appropriateness of risk adjusting for clinical risk factors is well accepted in quality measurement, but the majority of clinical Star Measures have no such adjustments.
- The appropriateness of adjusting quality measures for socio-demographic factors is under debate, but the results of this study demonstrate that these factors are important contributors to disparities in outcomes and affect all MA members with those characteristics.
- The appropriateness of adjusting for community resource availability is less controversial, but no Star Measures include adjustments for such factors.
With risk adjustment, plans doing a relatively worse job at achieving good outcomes among members with high-risk characteristics will still have lower performance scores relative to plans doing a relatively better job at achieving good outcomes among members with similar risk profiles. The adjustments would allow fair comparisons of quality of care across health plans serving different populations.