Overview

The Medicare drug benefit (Medicare “Part D”), provides federal subsidies to pay premiums and cost sharing for low-income beneficiaries—almost 10 million in 2009. Yet there are several policy issues concerning these low-income beneficiaries under Part D. First, over 2 million individuals who may qualify for the subsidies have not enrolled. Second, in some states, low-income beneficiaries have little choice of plans (while non low-income beneficiaries have dozens of choices), unless they pay out-of-pocket for premium amounts above what the subsidy covers. And third, millions of those who have enrolled in the benefit face the prospect each year of switching drug plans or paying more to keep their current drug plan. What led to this state of affairs? Are there lessons to be learned from Medicare Part D as Congress debates how to provide health insurance subsidies on behalf of low-income individuals?

Drug Benefit Basics

Over 45.24 million Medicare beneficiaries are eligible for Medicare Part D, including low-income and non low-income beneficiaries. Of those, nearly 41 million—over 90 percent—have prescription drug coverage through sources including “stand-alone” PDPs or MA-PDs, a former employer receiving a Medicare retiree drug subsidy,⁴ or other sources such as the Veterans Administration, a current employer, a Medicare supplemental (or Medigap) policy, or the Indian Health Service. About 4.46 million beneficiaries, including roughly 2 million low-income individuals, have no known source of drug coverage. (See Figure 1, right.)

Figure 1

Sources of Drug Coverage for Medicare Beneficiaries, 2009. * Other coverage includes: TRICARE, FEHBP, VA, active workers with Medicare secondary payer, retirees in plans not receiving the retiree drug subsidy, Medigap, Indian Health Service, and state (more...)

The vast majority of Medicare beneficiaries are enrolled in PDPs and MAPDs—17.48 million and 9.17 million, respectively, in 2009. Those beneficiaries can choose among dozens of plan options available in each of 34 regions across the country. While the number of options varies from region to region, most beneficiaries can choose from of about 50 PDPs and dozens of MA-PDs. Beneficiaries enrolling in a drug plan pay monthly premiums, in addition to cost sharing and any deductible.⁵ The standard benefit package in 2009 consists of a $295 yearly deductible, 25 percent coinsurance for drug spending between $295 and $2,700, and 100 percent coinsurance after the beneficiary reaches $2,700 until $4,350 is spent out of pocket (a gap in coverage popularly known as the “donut hole”). After a beneficiary spends $4,350 out of pocket, catastrophic coverage begins which reduces cost sharing to a minimal amount. Most beneficiaries are enrolled in plans that have an alternative benefit package that has the same benefit value (that is, it is actuarially equivalent) to the standard benefit.

Medicare subsidizes the cost of the drug benefit for all enrollees, paying 74.5 percent of program costs. Federal expenditures for Medicare Part D are expected to total almost $63 billion in 2009.⁶ Medicare makes additional payments to drug plans to subsidize the premiums, cost sharing, and deductibles of beneficiaries with limited income and assets. This is known as the low-income subsidy (LIS). In 2009, drug plans are expected to be paid an average of $1,105 per enrollee for drug benefits and another $1,950 per low-income enrollee to subsidize the additional benefits available to them.⁷ About one-third of total Medicare outlays for Part D are for these additional LIS payments. Of the 26.65 million beneficiaries enrolled in PDPs and MA-PDs in 2009, over one-third (9.67 million) are enrolled in the LIS program.⁸

Low-Income Beneficiaries in Part D

An estimated 12.5 million beneficiaries are eligible for extra benefits under Part D including payment of drug plan premiums, co-pays and deductibles. However, only 9.67 million of those potentially eligible for Part D and the extra benefits are enrolled in 2009, despite federal and local efforts to increase awareness of the benefit and enroll them.

Eligibility and Benefits

The LIS program significantly reduces out-of-pocket costs, including no or lower beneficiary premiums and limited cost-sharing (Figure 2, next page). Unlike non low-income beneficiaries, beneficiaries qualifying for the LIS continue to have their prescription drugs paid for through the donut hole.

Beneficiaries become eligible for the LIS when their incomes are at or below 150 percent of the federal poverty level ($16,245 for an individual and $21,855 for a couple in 2009) and they have limited assets ($11,010 for an individual and $22,010 for a couple in 2009).⁹ The amount of assistance varies, depending on the income and assets of the beneficiary. In other words, lower income beneficiaries receive the most assistance, including payment of the monthly drug plan premium if they enroll in an eligible plan.

Figure 2

The Low-Income Subsidy: Extra Assistance in 2009 for Beneficiaries with Limited Means.

As Figure 2 shows, beneficiaries with incomes below 135 percent of the federal poverty level and few assets generally pay no monthly drug premium, most pay no deductible, and all have lower cost sharing amounts than non low-income beneficiaries. Beneficiaries with incomes at or above 135 percent of FPL but below 150 percent of FPL pay premiums on a sliding scale according to income, and have lower deductibles and coinsurance than beneficiaries not receiving the LIS.

Enrollment and Auto-enrollment

Some low-income beneficiaries automatically qualify for LIS, whereas others must apply for it. Groups of beneficiaries automatically qualifying for LIS include: beneficiaries eligible for both Medicare and Medicaid (“dual eligibles”), beneficiaries who receive supplemental security income (SSI) but are not dually eligible, and beneficiaries enrolled in Medicare Savings Programs (federal-state programs that help low-income beneficiaries pay Medicare premiums, deductibles, and copayments). Beneficiaries who do not fall into one of these three groups but believe they meet the income and asset requirements may apply for LIS benefits, either through a state office or the Social Security Administration (SSA).

LIS beneficiaries receive assistance with enrollment in a drug plan. Those who automatically qualify for the LIS but do not choose a prescription drug plan on their own are automatically enrolled, or “auto-enrolled,” in one by Medicare when they first qualify for LIS. Beneficiaries who apply for and receive LIS, but then fail to select a plan, are also enrolled by Medicare (often referred to as “facilitated enrollment”). When a beneficiary does not choose a plan on his own, he is randomly assigned to a drug plan with a basic benefit package that has a premium at or below the low-income benchmark premium. Low-income beneficiaries, including those randomly assigned to a plan, are free to change plans during the year, whereas beneficiaries not receiving LIS may only change plans during an open enrollment period in late fall.

Auto-enrollment helps ensure that as many known low-income beneficiaries as possible are enrolled in a drug plan. However, there is no guarantee that the plan to which a beneficiary is randomly assigned has a formulary that includes the drugs the beneficiary is taking. For this reason, some have suggested that plan assignment take into account what drugs a beneficiary is taking and other factors unique to an individual beneficiary.¹⁰ Conversely, so-called “beneficiary-centered assignment” could increase the number of beneficiaries changing plans each year as beneficiaries are moved into plans that better meet their needs. And, plan costs could increase as more beneficiaries are auto-enrolled into plans because their formulary includes the needed drugs. These unintended consequences would need to be thought through before a policy change is made.

As Table 1 indicates, many beneficiaries—an estimated 2.34 million in 2009—may be eligible for low-income assistance but are not enrolled. There are many reasons why such a large share of eligible individuals would not enroll in Part D or fail to enroll in the LIS; factors that may play a role include lack of awareness, perceived lack of need (particularly for beneficiaries with no drug spending), language barriers, mental illness, low literacy, other coverage not known to Medicare, and personal preference. Beneficiaries may also experience a change in circumstances that qualifies them for LIS, such as loss of a spouse, or change in income or assets, but fail to connect the change to potential eligibility.

Table 1

Medicare Beneficiaries Eligible for Low-Income Subsidy (by Source of Coverage), 2006–2009.

CMS (the agency that oversees Medicare), SSA, and numerous advocacy organizations have undertaken efforts to reach out to and enroll beneficiaries potentially eligible for LIS. SSA, for example, sends letters to beneficiaries who the agency believes may be eligible for LIS based on the amount of their Social Security benefit. CMS has worked with beneficiary groups to identify and implement best practices for locating and enrolling potentially eligible beneficiaries, particularly those with limited English proficiency and low literacy. CMS has been trying multiple approaches to targeting potentially eligible individuals, including identifying and focusing outreach efforts on ZIP codes that have high numbers of beneficiaries who are potentially eligible for the LIS, but not enrolled.¹¹

CMS has also conducted focus groups, telephone interviews and other qualitative research in an effort to better understand reasons why low-income beneficiaries have not enrolled in Part D. For example, CMS research indicates that urban beneficiaries prefer to learn about LIS directly from Medicare or the SSA, rather than local beneficiary groups that seem to be a trusted information source for other beneficiaries. Rural beneficiaries who have not enrolled seem to be wary of providing personal information to anyone, have a strong sense of independence that might hinder enrollment in a public program, and are less likely than urban beneficiaries to contact Medicare or SSA for information about federal programs.¹² CMS is using this type of feedback to target the method, means, and type of communication it has with potentially eligible beneficiaries. Despite targeted interventions, however, many beneficiaries, including low-income beneficiaries, remain unenrolled.

How Part D Works for Low-Income Residents of U.S. Territories and Indian Reservations

While many of the features of Medicare Part D are the same for non low-income beneficiaries who reside in the territories or receive their health care from Indian Health Service facilities, there are a few important differences.

Residents of U.S. Territories — Over 650,000 Medicare beneficiaries reside in the U.S. territories (including Puerto Rico, the U.S. Virgin Islands, Guam, American Samoa and Northern Mariana Islands).^* Much of the design of Medicare Part D is the same in the territories as it is on the U.S. mainland. Plan design (including cost sharing and deductibles), bidding, and plan contracting are all similar. However, low-income beneficiaries residing in the territories are not eligible for the LIS. Instead, the territories receive a grant of federal Medicaid funds to permit each territory’s Medicaid program to “wrap around” Medicare drug coverage for beneficiaries du-ally eligible for both Medicare and Medicaid. This wrap-around benefit is only available to full dual eligible beneficiaries. Non–dual eligible beneficiaries under 150 percent of the federal poverty line—who would be eligible for assistance with premiums and cost sharing if they resided in the U.S. mainland—are not eligible for the extra benefits which are funded with Medicaid dollars, and the non-duals are not Medicaid recipients.

Indian Health Service — While Medicare Part D works the same for American Indians and Alaska natives as for any beneficiary, some do not enroll in Part D or the LIS. Most American Indians and Alaska Natives receive health care services free of charge at Indian Health Service (IHS) facilities, including prescription drugs. From the beneficiary perspective, there is little need to enroll in Part D because prescription drugs are already free. However, the IHS encourages beneficiaries to enroll in Part D, particularly if they may qualify for the LIS, so that Medicare is the primary payer for these Medicare beneficiaries.

*

State Health Facts, “Distribution of Medicare Beneficiaries by Age, 2004,” Kaiser Family Foundation”; available at www​.statehealthfacts​.org/comparetable.jsp?ind​=292&cat=6.

The failure of low-income persons to enroll in available subsidy programs is not unique to Medicare Part D. While it is difficult to estimate the number of potentially eligible individuals for any program, some analysts believe that millions of Americans are eligible for, but not enrolled in, other federally subsidized health insurance programs including Medicaid, the Children’s Health Insurance Program, and Medicare Savings Programs.¹³

Variation in Drug Plan Availability

Low-income beneficiaries make up a sizeable share of enrollment in PDPs and MA-PDs. About 45 percent of the beneficiaries enrolled in PDPs and about 17 percent of beneficiaries enrolled in MA-PDs receive the LIS.¹⁴ However, not all individual drug plans have a large share of low-income beneficiaries. Because Medicare only pays the full premium of basic benefit package drug plans that have premiums at or below the average in a geographic area, the great majority of LIS beneficiaries are in below-average cost plans.¹⁵ The average beneficiary premium in an area, weighted by enrollment of LIS beneficiaries, is called the low-income benchmark premium.¹⁶ If a low-income beneficiary selects a plan with a premium above the low-income benchmark premium, the beneficiary pays the difference.¹⁷ Beneficiaries are not auto-enrolled into plans with premiums above the benchmark. Drug plan premiums change each year; therefore, a plan with a premium below the benchmark this year may have a premium above the benchmark next year. If this happens, low-income beneficiaries then must decide between staying in the same plan—and paying the difference between the plan’s premium and benchmark premium—or enrolling in another plan with a premium below the benchmark. In 2009, about 2 million LIS beneficiaries are enrolled in non-benchmark plans and are paying Part D premiums.¹⁸ Once a beneficiary chooses a plan on his or her own, CMS will not reassign the beneficiary to another plan, even if the beneficiary’s plan has a premium above the benchmark in a future year.

The policy rationale for Medicare paying the full premiums only for basic benefit plans with premiums below the low-income benchmark is to ensure that Medicare is not paying for the most costly plans, and to encourage plans to bid as low as possible in order to attract enrollment. If Medicare paid the full premium of plans with higher-than-average premiums, program costs would be higher, and plans would have less incentive to bid competitively.

The number of plans with premiums under the benchmark (and therefore available to most low-income beneficiaries at zero premium) and the variation in the number available from region to region are policy concerns. The number of drug plans with premiums below the benchmark generally increased between 2006 and 2007; since then, however, the number of such plans generally declined—with steep declines in some regions. Indeed, all regions except Wisconsin experienced a decline in the number of plans with premiums under the benchmark in 2009.¹⁹ Figure 3 highlights notable patterns in the availability of plans with premiums below the benchmark. The 2006–2009 experience of Region 1 (New Hampshire and Maine) is typical: an increase in plans for 2007 followed by rather steep declines in 2008 and 2009. (See also Appendix 1 for 2006–2009 regional low-income benchmark premiums and the number of plans with premiums below the benchmark).

Figure 3

Plans with Premiums At or Below the Benchmark, 2006–2009. Sources: Centers for Medicare & Medicaid Services, 2009 Landscape Source and 2008 Landscape Source, www.cms.hhs.gov/PrescriptionDrugCovGenIn. 2007 and 2006 Landscape Source files (more...)

From a policy perspective, it is not clear what the “right” number of plans with premiums below the benchmark should be. While some might argue that many low-income beneficiaries would have difficulty comparing coverage among many plans and making a choice, others believe that choice should be as broad as possible. Some analysts and advocates believe that providing plan choice allows individuals the opportunity to select a plan whose formulary and other features best meet their needs while encouraging plan innovation. Others note that prior to implementation of Part D in 2006, most low-income beneficiaries received their drug coverage through Medicaid—with no choice of plans at all. In addition, states with the lowest numbers of below-the-benchmark plans also have a high percentage of low-income beneficiaries in Medicare managed care plans, so having more choices of PDPs is less important. In any event, there is some degree of discomfort among policymakers with the significant variation in the number of available zero-premium plans from one region to another, as well as the discrepancy between the amount of choice available to these beneficiaries compared to other beneficiaries.

Drug Plan Availability Over Time

A second major policy issue related to the bidding and benchmarks is that, because benchmark is calculated annually, each year the mix of plans with premiums at or below the benchmark has changed. According to CMS, 2.2 million beneficiaries—about 23 percent of low-income beneficiaries enrolled in PDPs and MA-PDs—were either switched or were notified that they needed to switch to avoid paying a premium (or an increased premium) for 2009. For 2008, 2.6 million low-income beneficiaries were affected by these changes. Further, recent analysis indicates that, of the 409 plans with premiums below the benchmark the first year of the drug benefit (2006), only 96 of them—23 percent—had premiums below the benchmark in 2007, 2008 and 2009.²⁰ Some beneficiaries have been enrolled in two, three, or more plans since the start of the program in 2006.

The number of beneficiaries switching plans is troubling from a policy perspective for several reasons. First, it is confusing for beneficiaries. This population tends to have higher drug utilization, more chronic health conditions (including mental health conditions), and cognitive impairments than the average American, all of which can make frequent changes in drug plans more difficult. Understanding which drugs are on plan formularies, which pharmacies are preferred, how to manage an appeal or grievance, and how to contact the plan is particularly challenging for many in this population.

Second, beneficiaries may be subject to utilization management techniques (prior authorization or step therapy, for example) in one plan, and be subject to different techniques in another plan the following year. While utilization management techniques are widely used and accepted tools, annual changes in the techniques and the drug substitutions may not be desirable, especially for this population. CMS has requirements for short-term supplies of a drug during a transition period, but staying on a preferred drug long-term sometimes requires effort by both the physician and the beneficiary. It is not yet clear whether changing plans leads to widespread quality-of-care issues, such as medication reactions because of therapeutic substitution or a confused beneficiary not refilling a prescription on time.

Third, the percentage of individuals switching plans is much higher among low-income beneficiaries than in the non low-income population: only about 6 percent of non low-income beneficiaries change Medicare drug plans in a given year. The experience of non low-income beneficiaries under Part D is consistent with that of of enrollees in the Federal Employees Health Benefits Program—about 6 percent of federal employees nationwide switch health insurance plans in any given year.²¹ Non low-income individuals most often elect to stay in their current plan, even if it costs more. It is worth noting that many Medicare beneficiaries are faced with difficult choices and trade-offs in spending each year, particularly when it comes to health care spending, and deciding whether to switch prescription drug plans because of cost is only one such decision. However, some may feel that the prospect of switching plans is a “price to pay” for the federal subsidy.

Fourth, there may be some modest hidden costs involved in switching plans. Plans incur costs educating new beneficiaries through printed material and fielding beneficiary phone calls about new formularies and plan requirements. Pharmacy costs include counseling beneficiaries regarding formulary changes at the point of sale and updating computer systems with new plan information. Physician offices may find that annual drug plan changes result in more of their time being spent on pharmacy issues as they handle plan requests for medication changes. For some patients, drug switches require an office visit, perhaps adding to Part B costs.

A Closer Look behind Annual Changes in Plan Options

It is instructive to understand why plan options change for low-income beneficiaries from year to year. The low-income benchmark premium amount is key to this. It determines how many plans, and which particular plans, are offered to low-income beneficiaries premium-free. The multi-year example in Appendix 2 illustrates how the low-income benchmark is calculated, and how the number and mix of plans with premiums below the benchmark change annually.²²

Several issues contribute to the fluctuation of the premium and the policy concerns described above that accompany it. These mostly technical issues are worth considering as policymakers weigh options to reduce the number of low-income beneficiaries experiencing a disruption of prescription drug coverage.

• Payment adequacy and risk adjustment. Payments to drug plans are adjusted to reflect certain characteristics of all beneficiaries including age, gender, and health status. Payments to plans on behalf of low-income beneficiaries are further adjusted to reflect the fact that low-income beneficiaries use more prescription drugs than other beneficiaries. Some researchers and plans believe that the adjustments to plan payments may not be adequate to account for these higher costs.²³ If payments to a plan for low-income beneficiaries do not accurately reflect the plan’s actual costs in one year, then the plan may not bid as competitively as possible in a future year to avoid the risk of covering low-income beneficiaries. In the Part D bidding process, plans must bid based on expected expenses and utilization, and plans must not bid based solely on a strategy of including or excluding certain groups. However, there is some judgment in putting the many pieces of bid information together to ensure that expenses are covered. If plans feel that the payment is inadequate to account for the costs of low-income beneficiaries, then their bid may reflect that belief. If the plan’s bid places its premium above the benchmark, then beneficiaries would need to decide whether to stay in the plan or enroll in a new plan below the benchmark. If many below-benchmark plans conduct themselves in this manner, the result can be instability in plan choices for beneficiaries and perhaps higher premiums over time.
• Calculation of the low-income benchmark premium. Although highly technical, the manner in which the low-income benchmark premium is calculated is very important because even seemingly modest changes may result in more plans or fewer plans having premiums below the benchmark. For the first several years of the drug benefit, the low-income benchmark premium was calculated in a manner that some analysts believe kept the benchmark premium artificially high. CMS changed the calculation for 2009 and future years. The calculation now involves multiplying the plan’s premium by the percentage of LIS beneficiaries enrolled in the plan.²⁴ The policy rationale for this change was to better reflect the premiums of plans that actually enroll LIS beneficiaries, and to reduce the effect of MA-PD premiums in the calculation (see below). The overall result is intended to be a reduction in the number of beneficiaries switching plans. It is not yet clear what the result over time will be. However, some analysts are concerned that the benchmark will actually be depressed over time.
• Inclusion of MA premiums in the calculation of the benchmark. The drug portion of the MA premium is included in the benchmark calculation, consistent with the statute. The inclusion of MA premiums depresses the benchmark because the MA-PD premium amounts used in the calculation are lower on average than the premiums of free-standing drug plans. The MA-PD premiums are considered artificially low for purposes of the calculation because MA-PDs are permitted to apply money saved providing benefits under Parts A and B (known as rebates) to the premium amounts. If MA plans were excluded from the benchmark, or if the premium amount did not include rebate dollars, the low-income benchmark premium likely would increase, and more plans would have premiums below the benchmark.

Conclusion

Reliance on competitive bidding to determine premiums and overall program costs is a hallmark of Medicare Part D. But the nature of the bidding process has introduced instability in the number of plans and the particular plans offered to low-income beneficiaries. Consquently, low-income beneficiaries might not remain in a prescription drug plan over the long term. This fluctuation may lead to beneficiary confusion around plan choices and enrollment. It is not clear whether or not frequent plan changes result in quality-of-care issues, but that would be of great concern. And the instability may prove costly both for pharmacies dealing with beneficiaries at the point of sale and for physicians and their staffs who field phone calls from pharmacies regarding medication changes at the request of drug plans.

Congress is considering health reform legislation that would involve providing subsidies for the purchase of health insurance on an income-related basis. Medicare’s Part D experience to date may be instructive as proposals are refined. Providing the highest subsidy level only to plans with average or below premiums has been a successful tool for keeping federal costs as low as possible. However, there has been disruption in plan enrollment and choice for low-income beneficiaries. Policymakers may want to consider the experience of low-income beneficiaries under Part D as they debate health reform measures.

Endnotes

1

Beneficiaries enrolled in one type of Medicare managed care plan, called a private fee-for-service plan, may also enroll in a PDP because the private fee-for-service plan is not required to offer a prescription drug benefit.

2

For a discussion of how the bid is converted into a monthly beneficiary premium, please see Medicare Payment Advisory Commission (MedPAC), “Part D Payment System,” paymentbasics, revised October 2008; available at www​.medpac.gov/documents​/MedPAC_Payment_Basics_08_PartD.pdf.

3

Some beneficiaries switch to a lower-premium drug plan offered by the same company or sponsor which may have the same formulary.

4

Employers offering a qualified drug benefit to Medicare-eligible retirees are eligible to receive a subsidy from Medicare known as the retiree drug subsidy to defray the cost of providing a drug benefit. In 2009, Medicare will pay roughly $600 per beneficiary enrolled in such plans. See 2009 Annual Report of the Boards of Trustees of the Federal Hospital Insurance and Federal Supplementary Medicare Insurance Trust Funds, May 12, 2009, p.163; available at www​.cms.hhs.gov/ReportsTrustFunds​/downloads/tr2009.pdf.

5

For more background information on the Medicare drug benefit please see “The Medicare Drug Benefit (Part D),” National Health Policy Forum, The Basics, January 15, 2009, available at www​.nhpf.org/library/details.cfm/2708, and background and presentation material from the fundamentals briefing, “Understanding Medicare and Medicaid: Fundamentals and Issues for the New Congress,” January 15, 2009, available at www​.nhpf.org/briefbook09Tab5.cfm.

6

2009 Trustees Report, p. 120. Amount includes federal expenditures for the retiree drug subsidy.

7

2009 Trustees Report, p. 163.

8

Nearly one-half million additional low-income beneficiaries have drug coverage from sources other than Medicare, such as the Veterans Administration (VA), Indian Health Service (IHS), and state pharmaceutical assistance programs (SPAPs).

9

For all states, except Alaska, Hawaii, and DC, where separate guidelines apply. Centers for Medicare & Medicaid Services (CMS), “Guidance to States on the Low-Income Subsidy,” February 2009; available at www​.cms.hhs.gov/LowIncSubMedicarePresCov​/Downloads/StateLISGuidance021009.pdf.

10

See “The Role of Beneficiary-Centered Assignment for Medicare Part D,” a study conducted by Georgetown University and NORC at the University of Chicago for MedPAC, June 2007, available at www​.medpac.gov/documents​/June07_Bene_centered​_assignment_contractor.pdf. Also see “Beneficiary-Centered Assignment and Medicare Part D,” presentation to MedPAC by Jack Hoadley et al., September 4, 2008, available at www​.medpac.gov/transcripts​/Hoadley%20MedPAC​%20presentation%2009%2004%2008.pdf.

11

See, for example, CMS, “LIS Outreach Toolkit Numeric Maps,” April 2009; available at www​.cms.hhs.gov/Partnerships​/99_LIS_Outreach​_Toolkit_Numeric_Maps.asp#TopOfPage.

12

CMS, “Formative Research on the Low Income Not Enrolled Population“; available at www​.cms.hhs.gov/Partnerships​/Downloads/Low​_Income_Initial_Qualitative_Research​.pdf.

13

See, for example, “Secretary Sebelius Announces Availability of $40 Million in Grants to Help Insure More Children,” U.S. Department of Health and Human Services, press release, July 6, 2009; available at www​.hhs.gov/news/press​/2009pres/07/20090706a.html.

14

MedPAC, Report to the Congress: Medicare Payment Policy, March 2009, p. 283; available at www​.medpac.gov/documents​/Mar09_EntireReport.pdf.

15

There are 34 regions nationwide. Many regions encompass only one state, and a few are multi-state.

16

The premium of a plan with very high LIS (low-income subsidy) enrollment counts more in calculating the average than a plan with lower enrollment.

17

For 2007 and 2008, CMS permitted plans with premiums slightly above the benchmark (a so-called de minimus amount above the benchmark)—$2 in 2007 and $1 in 2008—to offer the plan at zero premium. The policy was discontinued beginning in 2009.

18

“Medicare Prescription Drug Plans in 2009 and Key Changes Since 2006: Key Findings,” Elizabeth Hargrave et al. for Kaiser Family Foundation, June 2009; available at www​.kff.org/medicare/upload/7917.pdf.

19

MedPAC, Report to the Congress, March 2009, p. 291.

20

Laura Summer et al., “Low-Income Subsidy Plan Availability,” Kaiser Family Foundation, November 2008; available at www​.kff.org/medicare/upload/7836.pdf.

21

MedPAC, Report to the Congress, March 2009, p. 277.

22

This example is for illustrative purposes only; the actual calculation is more complex.

23

See, for example, work conducted for MedPAC by Dr. John Hsu, transcript from MedPAC Public Meeting, October 2, 2008, pp. 7–21; available at www​.medpac.gov/transcripts​/1002-1003MedPAC.pdf.

24

The regulation (CMS-4133-CN) making this change can be found at www​.cms.hhs.gov/PrescriptionDrugCovContra​/Downloads/CMS4133C.pdf.

APPENDIX 1. Low-Income Premium Subsidy Amounts and Number of Plans At or Below Benchmark, 2006–2009

There is a fair amount of variation in the low-income benchmark premiums nationwide. For 2009, regional low-income benchmark premiums ranged from a low of $16.22 in Arizona to a high of $38.15 in Wisconsin. There is also much variation in the number of drug plans with premiums at or below the low-income benchmark premium in a region. For example, in 2009, Wisconsin beneficiaries may choose among 16 plans below the benchmark, whereas Nevada beneficiaries only have 1 plan with premiums below the benchmark. Most regions have between 8 and 12 zero-premium plans.

APPENDIX 2. How the Low-Income Benchmark Premium Affects Options for Beneficiaries: A Multi-Year Example

This example illustrates the low-income benchmark calculation for the third, fourth and fifth year of the drug benefit in a fictional region. There are 14 PDPs and 3 MA-PDs with a range of premiums, and the low-income benchmark is $30.23. Low-income beneficiaries qualifying for a zero-premium plan in this region have a choice of four PDPs and three MA-PDs. A fifth PDP, whose premium is $31.00, will be permitted to offer their plan to LIS beneficiaries for $30.23, since the premium is a de minimus amount—less than $1.00—above the benchmark. In future years, the de minimus rule will not apply, similar to what actually occurred under Part D.

Year 3Illustrative Low-Income Benchmark Premium Calculation

* PDP 8 qualifies as a de minimus plan in year 3 (see above).

View in own window

In year 4 of this multi-year example, depicted below, several pieces of the puzzle change, quite similar to what actually occurred under Part D. First, plan premiums increase. Second, the calculation of the benchmark premium changes, as it did under Part D for 2009. Rather than using enrollment of all beneficiaries, the new calculation only includes the percentage of low-income beneficiaries enrolled in each of the plans. In addition, the de minimus policy is no longer in effect. PDP 4, which had been the lowest premium plan with the largest market share, increases its premium to $38.00. Perhaps PDP 4’s cost experience warrants the increase, or perhaps the plan does not bid as competitively as possible. The result is that PDP 4 now has a premium above the low-income benchmark of $28.07. There are now two PDPs and three MA-PDs with premiums below the benchmark. PDP 4—the most popular plan—and PDP 7 and PDP 8 are now no longer available to low-income beneficiaries at zero premium. Low-income beneficiaries qualifying for zero premium may stay in the plan and pay the additional premium, select and enroll in another plan, or be automatically enrolled in another plan by Medicare.

Year 4Illustrative Low-Income Benchmark Premium Calculation

View in own window

In year 5, three MA-PDs now have premiums, perhaps because Medicare managed care payments decreased and plans decide to increase or add a premium for prescription drug coverage. Premiums have increased across the board, as one would expect. In addition, PDP 4 has lost considerable market share, and drops its premium. The overall result is a shift in the benchmark premium to $25.88, with one PDP and three MA-PDs having premiums below the benchmark, somewhat similar to the real-world experience of Nevada in 2009. Notice that PDP 4 is back under the benchmark, so it is available to low-income beneficiaries for zero premium. Some of the beneficiaries who were enrolled in PDP 4 in year 3, disenrolled for year 4, are once again in PDP 4 for year 5.

Year 5Illustrative Low-Income Benchmark Premium Calculation

View in own window

In this simplified multi-year example, the low-income benchmark premium has declined, and with it the number of options available to most low-income beneficiaries for zero premium. While reality may differ from this example for many beneficiaries, for others it mirrors their experience under Part D. Plan premium amounts, plan enrollment, including MA-PDs in the calculation, plan willingness to attract low-income beneficiaries, and plan payment (including risk adjustment) all play a role in determining which plans have premiums below the benchmark.

The National Health Policy Forum is a nonpartisan research and public policy organization at The George Washington University. All of its publications since 1998 are available online at www.nhpf.org.