Imagine eating at a restaurant and having to choose from more than 100 different salads on the menu. Slightly different ingredients, different sizes, different prices. Maybe some have the same ingredients but with different names. Perhaps you have allergies.
Now imagine you’re reading it in a language you barely speak, and your decision could have a significant impact on your health.
Which one would you choose?
There are 82,155 Medicare Advantage plans available when you include each unique plan and county combination across the nation. Read that again: 82,155. It’s a large number for policymakers to grapple with and CMS to oversee; it’s even more for Medicare beneficiaries to try to understand. At the county level, a Medicare beneficiary can expect an average of 26 different plans to choose from, with as many as 110 in Los Angeles, California.1
Choosing the right health plan can be overwhelming — navigating and comparing concepts like deductibles, benefit coverage (down to the hour in some instances), premiums, cost-sharing, network considerations. And if you’re a dually eligible beneficiary with both Medicaid and Medicare coverage, this decision is even more difficult. Do the plan’s Medicare providers participate in Medicaid? Are they willing to serve dual-eligible individuals? Do the out-of-pocket costs apply to you? Does the plan coordinate with your Medicaid benefits? The considerations go on and on.
Additionally, unrestricted Medicare plan choice may undermine policy efforts to coordinate or integrate Medicare and Medicaid. Often, even when an integrated, seamless plan option is available, dual-eligible beneficiaries must sift through a complex health plan marketplace to find it.
To better understand and quantify how “choice” impacts dual eligible beneficiaries beneficiaries, ATI Advisory researched how many Medicare plan options are available to this population in every county nationwide. 2 We assessed the number and types of plans available (excluding Medicare Savings Account plans, Part B only plans, and Employer-Group Waiver plans, as well as plans with fewer than 11 enrollees), and the accessibility of plan comparison information on Medicare Plan Finder.
In summary: there are a lot of plan options. And, dual-eligible individuals typically have access to more choices than Medicare-only beneficiaries because of plans targeted toward them, such as a dual-eligible special needs plan (D‑SNP), Medicare-Medicaid Plans (MMP), and the Program of All-Inclusive Care for the Elderly (PACE). Dual-eligible beneficiaries also are likely to qualify for institutional special needs plans (I‑SNP), which are Medicare Advantage plans limited to beneficiaries with an institutional level of care need (i.e., they meet a Medicaid program’s functional criteria for long-term services and supports). D‑SNPs, MMPs, and I‑SNPs all offer targeted models of care that can be incredibly meaningful to dual-eligible individuals, but in the case of D‑SNPs and I‑SNPs, the parent organization may or may not have any line of sight into the beneficiary’s Medicaid benefits. And ultimately, this adds to the total plan choices to contemplate (more detail provided in Figure 13).
Granted, there are also counties with limited or no choice. While we identified eligible plans in 3,102 of the possible 3,220 counties across the 50 United States plus Puerto Rico, no plans are currently available in Alaska (Figure 2). Of the 3,102 counties, fifteen percent (or 453) currently have no D‑SNP or MMP product, although PACE is available in 18 of these. Another 307 counties have only a single D‑SNP or MMP product.
Enrollment Channels
There are several channels that attempt to educate Medicare beneficiaries about enrollment choices, including insurance agents, Medicare’s online tool Plan Finder, and State Health Insurance Assistance Programs (SHIPS). However, these channels don’t always suggest options for dual-eligible individuals that are designed specifically for their needs (recall: there’s an average 26 plans per county). For example, Medicare Plan Finder aims to provide tailored decision support, and as a result, it only shows D‑SNP options if a beneficiary identifies that they are eligible for supports such as Medicaid or the Medicare Savings Program. If a beneficiary responds “I’m not sure” to this question, no D‑SNP options are shown. Understandably, the Centers for Medicare and Medicaid Services (CMS) doesn’t want to advertise plan options to beneficiaries who aren’t eligible for them. However, the consequence is that dual-eligible beneficiaries may not see D‑SNP as an available plan option, because they don’t understand the differences between Medicare Savings Programs, Medicaid, Social Security, and Medicare generally, or which programs they qualify for. Even when selecting “Medicaid” as a program you qualify for, the default plan order doesn’t prioritize D‑SNP. I tried this in my own county where there are nine D‑SNPs, and the first D‑SNP option was tenth on the list of 35 options.
It’s complicated within other channels too. For example, insurance agents typically aren’t commissioned to enroll a beneficiary in an integrated MMP product but may be financially incentivized to promote a standard Medicare Advantage plan. And SHIP volunteers tend to be well-versed in Medicare but may not be trained on the nuances of enrolling in a product that aligns with Medicaid, or the unique needs of dual eligible beneficiaries. All of this means a dual eligible beneficiary may be left to navigate an average of 26 choices themselves.
This raises the question, is too much choice harmful? The abundance of options highlights the downside of too much choice, and the negative impact it can have on the individuals who need to choose. As of June 2021, 3.9 million dual-eligible individuals (34%) were enrolled in a D‑SNP or MMP product. Approximately 1.6 million (14%) were in standard MA or other SNP types, and 5.9 million (52%) in Traditional fee-for-service Medicare (Figure 4). 4 To what extent is choice influencing this, and/or overwhelming dual-eligible individuals to the extent they remain in fee-for-service?
Policy Options
Among policymakers, Medicare choice is often referred to as a “third rail.” Preserving access to the Medicare benefits to which people are entitled is important, which makes well-intentioned efforts to simplify consumers’ options difficult.
But when too many options come at a cost, perhaps it’s time to think about guardrails. Today, beneficiaries may select fee-for-service or any Medicare Advantage plan available in their county (with limited exception). As a result, beneficiaries end up weighing 20, 30, or even 100 plan options side-by-side, each with slightly different benefits and participating providers.
Experience tells us that without meaningful guardrails around plan options, choice grows exponentially. Until 2019, Medicare Advantage plans were subject to a “meaningful difference” requirement that prevented a company from offering substantially similar plan options in a county. The goal of eliminating this requirement was to improve competition, innovation, and available benefit options. And the result was a 10-fold increase in the growth of plans year-over-year (from 1.2% growth between 2017 and 2018 to 11.2% growth the following year). 5 Perhaps a first step to simplifying plan selection is reducing “duplicate” plans.
At a minimum, policymakers should ensure enrollment channels are equipped to educate dual eligible beneficiaries about the most appropriate choice. Products that target the medical, functional, and social needs of dual eligibles should be the primary options a dual eligible beneficiary sees or hears about. SHIP counselors should receive training on the nuances of dual eligible beneficiaries and the important role of different Medicare plan types in addressing this nuance. Plans should be permitted (or required?) to incentivize agents to prioritize integrated and targeted products to dual eligible beneficiaries. For example, if a duals-specific option is available in a county, agent commissions for standard (non-SNP) Medicare Advantage plans could be discounted by ten percent for dual eligible beneficiaries.
I understand and appreciate the value of choice – but at what point does it become too much? I’m not sure I would be able to make it through the 35 plan options in my own county, and I work in health policy for a living. CMS and states have taken serious strides to promote products that best meet the complex care needs of dual eligibles; the fear of limiting “choice” shouldn’t undermine those efforts.
This insight was developed with support from Arnold Ventures.
-
1
MedPAC’s July 2021 data book identified an average of 32 plan offerings per beneficiary using an alternative calculation method. http://www.medpac.gov/docs/default-source/data-book/july2021_medpac_databook
↩︎ -
2
Given the target population of dual eligible beneficiaries, we included Medicare-Medicaid Plans and the Program of All Inclusive Care for the Elderly, even though these plan types are not subject to the same Medicare Advantage regulations as traditional Medicare Advantage and Special Needs Plans.
↩︎ -
3
C-SNP is a chronic condition special needs plan limited to Medicare beneficiaries with qualifying chronic conditions; D-SNP is a dual eligible special needs plan limited to dual eligibles (and may be further limited based on state program design, for example limited to individuals with Medicaid long-term services and supports eligibility); I-SNP is an institutional special needs plan limited to Medicare beneficiaries living in an eligible facility type, or residing in the community with an institutional level of care need; MMP is a Medicare-Medicaid Plan, a temporary demonstration program limited to certain dual eligible beneficiaries in select states; and PACE is the Program of All Inclusive Care for the Elderly, a day-center program that currently serves individuals aged 55 and older residing in the community with an institutional level of care need.
↩︎ -
4
ATI Advisory analysis of CMS Medicare Advantage enrollment files (July 2021) and the 2020 Master Beneficiary Summary File. Because these data are a point in time (June 2021), they may not be directly comparable to other published estimates
↩︎ - 5 ↩︎
