As mentioned in previous blogs, after turning a cold shoulder to Medicare Advantage Special Needs Plans (SNPs) just a decade ago, the Centers for Medicare and Medicaid Services (CMS) now seems to be positively immersed in all things “dual eligible.”
As we know, SNPs and their cousins, the Medicare-Medicaid Plans (MMPs), are often entirely or predominantly serving dual eligible beneficiaries. CMS now sees these vehicles as the best way to rein in costs and improve quality in the worlds of Medicare and Medicaid acute and long-term care. As expensive as acute care services are and are projected to be, the aging of America makes the provision of long-term care services (largely funded by Medicaid) an even more daunting task in the future. The combination of these future liabilities threaten to destabilize the financial underpinnings of both systems of care.
A few weeks ago, CMS released the results of a study that provides empirical evidence that SNPs and those whose products are dominated by dual eligible beneficiaries face major challenges in achieving quality ratings under the current Star methodology. This clearly puts these plans at a competitive disadvantage in the Medicare Advantage market. We predict that, in the not-too-distant future, CMS will either provide some methodological adjustments in the current Star program to account for these documented issues or fold SNPs into a broader MMP-SNP Star program alternative.
Now, CMS is tackling yet another complaint surrounding dual eligibles in the Medicare Advantage (MA) program. By and large, the CMS Hierarchical Condition Category (HCC) risk adjustment model does a good job of predicting overall Medicare beneficiary costs. At the same time, it clearly needs some improvement when it comes to risk recognition for dual eligible beneficiaries. CMS noted the potential deficiency in its 2016 Rate Announcement and has now provided greater detail in terms of possible refinements to better predict costs for duals.
CMS actuaries studied the 2014 HCC model and calculated predictive ratios to ascertain whether scores for full and partial dual eligible beneficiaries are reasonably predictive. (Full dual eligibles have all of their premium, cost-sharing and certain supplemental benefits paid by their state Medicaid agency; partial dual eligibles usually benefit only from premium or cost-sharing support from the Medicaid program. CMS included the Qualified Medicare Beneficiary (QMB) Plus and Specified Low-Income Medicare Beneficiary (SLMB) Plus sub-categories as full duals.) As CMS noted in its recent announcement, a predictive ratio close to 1.0 indicates that the model is accurately predicting that dual group’s average cost. A ratio greater than 1.0 indicates over-prediction, while a ratio less than 1.0 indicates under-prediction. FFS population data were used given the continuing paucity of managed care data. As part of the study, CMS also looked at how predictive HCC is based on disability and institutional status. Disability and Institutional status is factored into the HCC risk adjustment system in addition to dual status.
Over 80% of those who are institutional (in a facility 90 days or more) are also dual eligible beneficiaries. CMS found that the predictive cost for institutional dual eligible beneficiaries is very close to 1. Therefore, there is no issue here and CMS is planning no changes to the institutional segment of the HCC system.
But while HCC does a good job of predicting costs for those with a community status who are non-dual eligible beneficiaries (1.015), it actually over-predicts for partial benefit dual beneficiaries (1.092) and under-predicts for full benefit dual beneficiaries (0.914). Even when the dual groupings are taken as a whole, there is a substantial under-reimbursement against actual cost (.957).
Given the clear need to adjust for full duals (and realign for partials), CMS is undertaking a re-modeling in the community segment of the HCC system. Currently the community segment is calculated as one segment, with Medicaid status factored in. Since that is not adequately predictive, CMS is proposing to include 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
- Non-dual disabled
In this way, the model can be targeted to actual costs for full, partial, and non-dual beneficiaries to better predict costs. Disability substantially impacted costs within the three groups as well, necessitating six groupings. In the proposed final model – moving from one community segment to six – the predictive cost is 1.0 for all groupings.
The Special Needs Plans and dual eligible markets will only grow in time as America continues to age. Notwithstanding some of the current bias regarding quality and risk adjustment, insurers are making investments here. But they need to continue if CMS’ goals with dual eligible beneficiaries are to be achieved. The changes CMS is looking at on risk adjustment and quality should create an even greater value proposition for plans to continue to control costs and improve quality for these complex, co-morbid populations with socio-economic and health literacy barriers to health care.