Laura Skopec
In the Medicare Advantage (MA) program, private insurance companies hold one or more contracts with the Centers for Medicare & Medicaid Services (CMS) to provide Medicare benefits. As described in a previous blog, Medicare Advantage Organizations (MAOs) can include multiple, sometimes hundreds, of MA plans in one contract. Under the Quality Bonus Program (QBP), MA contracts are assigned a star rating of 1 to 5 based on about 40 quality measures. Payment bonuses are then applied to MA contracts based on overall star ratings, including boosts to payment benchmarks and higher rebate percentages that allow MAOs to offer lower cost-sharing or more supplemental benefits.
Contract-level quality ratings can create confusion for analysts and policymakers because MA plans are paid based on 1) estimates of care provided by the individual plan (i.e, the plan “bid”), and 2) on quality scores that are averaged across multiple MA plans in the contract. Contract-level quality ratings can also be misleading for Medicare beneficiaries who may not realize that the 5-star rating they see on the Medicare Plan Finder reflects an average across multiple MA plans, not necessarily the quality of the plan they are considering enrolling in.
Pros: Why Does CMS Measure MA Quality at the Contract Level?
The short answer is that MAO contracts tend to have substantial enrollment, providing a large, stable sample size for calculating quality measures.
- Ensuring an adequate sample size is particularly important for quality measures that only apply to specific populations of plan members. For example, measures of blood sugar control are only reported for members 75 and younger with diabetes; and measures of osteoporosis management are only reported for women aged 67-85 who have had a bone fracture.
- The large sample sizes in MA contracts also mean scores on quality measures are relatively stable year-to-year, avoiding large fluctuations that are unrelated to plan activities. This predictability makes it easier for MAOs to develop long-term quality improvement strategies and helps stabilize payment.
- Because bonus payments are used by MAOs to provide supplemental benefits to members, the contract-level quality payments allow MAOs to offer consistent benefit packages across a service area.
Cons: Why Is Contract-Level Quality Measurement a Problem?
The Medicare Payment Advisory Commission (MedPAC), researchers, and MA organizations themselves have identified many issues with contract-level measurement of quality star ratings and bonuses. These were summarized in a previous blog and include:
- Contract-level star ratings may not be meaningful for beneficiaries or reflect their likely experience in their local plan.
- Contracts can include plans with a mix of different eligibility rules, such as regular MA plans, special needs plans for people with chronic conditions or who are eligible for Medicaid, and employer-sponsored retiree plans, all combined under a single star rating score.
- Calculating star ratings and quality bonuses at the contract level invites MAOs to “game the system.” MAOs have a strong financial incentive to achieve the highest possible star ratings for their contracts, either by improving quality broadly or by selectively combining high- and low-quality plans to maximize ratings. There is evidence that MAOs selectively combine plans into contracts to achieve higher star ratings, undermining the purpose of the program and obscuring true plan performance from beneficiaries.
Should CMS calculate star ratings and quality bonuses at the plan level instead?
CMS has the authority to measure star ratings and apply bonuses at either the contract or the plan level. Measuring quality at the plan level would have several advantages over the contract-level approach. First, plan-level star ratings would be more transparent for beneficiaries and allow them to directly compare the performance of plans they are eligible to enroll in. The plan-level approach would also be harder for MAOs to game by selectively combining plans and contracts, though MAOs would still have the ability to consolidate plans.
Despite its positive attributes, however, there are several drawbacks to plan-level quality measurement. First, some plans may not have sufficient enrollment to calculate some quality measures, particularly those measures focused on specific populations. This may be particularly true for smaller local plans or for plans in rural areas, making it more difficult for beneficiaries to compare among plans based on star ratings. Lower sample sizes would also mean that star ratings and quality bonuses might change more year-to-year, resulting in large swings in plan payment and supplemental benefits. This issue could potentially be addressed by setting a minimum enrollment threshold for plan-level quality measurement wherein plans with lower enrollment would continue to have their quality measured at the contract level. However, it is unclear whether such an approach would create unfair advantages and disadvantages across MAOs based on contract and plan size.
Moving to plan-level star ratings may also require significant changes to the way CMS transforms quality measures into the 5-point star ratings system. Right now, CMS sets national “cut points” for each measure to decide how many stars to award to a contract based on its quality measure performance. Because MA plans generally have a narrower geographic reach than contracts, it may not be fair to set these cut points nationally. Many of the star ratings measures are focused on clinical quality, and measuring quality at the plan level using national cut points may put plans in areas with lower-performing health systems at a financial disadvantage relative to plans in areas with higher-performing health systems. This problem could be addressed by setting cutpoints at the local level, however.
Are there other alternatives to plan or contract level quality measurement that should be considered?
MedPAC, researchers, and local and regional MAOs have suggested several alternatives to plan- or contract-level quality measurement that could improve transparency for beneficiaries and better incentivize MAOs to improve the quality of their products.
First, CMS could calculate star ratings for MAOs or parent organizations (e.g., United Healthcare) at the local level, such as by core-based statistical areas or metropolitan statistical areas. This approach, recommended by MedPAC, would provide beneficiaries with comparisons among MAOs in their market, rather than among contracts that may span multiple states or include plans they may not be eligible for. Setting cut points at a regional level would also better control for regional differences in beneficiary needs and practice patterns, encouraging competition for higher quality within markets that takes into account market-specific conditions. MedPAC has similarly recommended adjusting MA plan payment and quality measurement approaches to use multi-county service areas, like Hospital Service Areas or Metropolitan Statistical Areas, to ensure MAOs can not pick and choose the counties they operate in to gain higher quality scores or avoid beneficiaries with higher health care needs.
Other recommendations focus on streamlining the star rating system to include fewer measures. One association of local and regional MAOs suggested removing process measures for which many contracts already receive high scores from the star ratings system. Similarly, a roundtable of small and mid-size MAOs suggested focusing more narrowly on measures that are important for beneficiary shopping, such as member satisfaction and disenrollment rates. Alternatively, MedPAC has recommended replacing the QBP with a system of bonuses and penalties based on performance on a few population health measures. Finally, researchers have suggested limiting the QBP to small bonuses for exceptional performance on well-documented measures like prevention, combined with new incentives like quality improvement grants.
Proposals to measure quality at the local level could be combined with streamlining star ratings measures to help address any issues that arise with sample size. Further, setting regional cutpoints would enhance local competition by encouraging MAOs to focus on quality at a level closer to where enrollees live, rather than on strategically combining contracts across regions and states.
Additional proposals are available in the Medicare Policy Initiative’s Compendium of Policy Proposals for Medicare Advantage and Part D.