Medicare Part D

Medicare Part D Represents a bold experiment to publicly fund, but privately sell and administer, prescription drug insurance. Part D is only a relatively small part of the overall Medicare program; Medicare spent $432 billion overall in 2007, of which Part D accounted for only 11.6% ($50 billion).¹ However, the Medicare program is so large that Part D alone should account for 1% of the entire economy in less than 2 decades.¹ Furthermore, there is little room for more spending, given the trustees’ recent forecast of $12 trillion in unfunded liabilities over the next 75 years—a debt just slightly smaller than the entire US economy.¹ So much appears to be riding on Part D’s success or failure, including possible models for the provision of health insurance to the currently uninsured. Two articles^2,3 in this issue of JAMA raise important concerns about the program, but must be rightly viewed in this larger context.

Before the introduction of Part D, officials were justifiably concerned that the program would not appeal to all types of beneficiaries or entice a sufficient number of health plans to offer a benefit. Failing to attract a broad distribution of beneficiaries could induce adverse selection, which could have jeopardized the long-term viability of the program. Similarly, failing to attract a sufficient number of plan sponsors could diminish competition, which was rightly viewed as a cornerstone of a successful program. At the same time, consumer advocacy groups were concerned that low-income beneficiaries, particularly those with both Medicaid and Medicare, would be worse off under Part D.

More than 2 years into the program, Part D has broad appeal to beneficiaries and plan sponsors. Policies to protect plans from excessive losses in the first few years (through reinsurance and risk corridors) as well as efforts to educate beneficiaries about plan choices led to a large number of sponsors and a wide array of options that beneficiaries can choose from. In particular, low users of prescription drugs could select a drug plan with little or no annual premium in most parts of the country, mitigating the financial risk of enrollment. As a result, approximately 40 million Medicare beneficiaries now have prescription drug coverage at least as generous as the Standard Part D benefit.⁴ Depending on where they live, beneficiaries have a choice of at least 45 standalone prescription drug plans that offer coverage for prescription drugs only or they can choose to obtain drug coverage from Medicare Advantage plans that offer the drug benefit in addition to other Medicare benefits.

Although most enrollees rate their Part D plans favorably, many express concerns over increasing co-payments or premiums and having to switch medications.⁵ Many participants recall the complex process of choosing a plan when they first enrolled in Part D for 2006 and would rather “make do” with their current plan rather than revisit that decision.⁵ Such inertia may reduce competition over time if beneficiaries are reluctant to switch into lower-cost or more appropriate plans. In fact, as Hsu and colleagues² report in this issue of JAMA, enrollees in a large Medicare Advantage plan have limited knowledge of the plan’s cost-sharing requirements. For example, only 40% of beneficiaries surveyed in early 2007 were aware that their Part D drug plan in 2006 included a coverage gap. Although this lack of knowledge is a concern, similar knowledge gaps have been widely observed in non–Part D plans and should diminish over time. Nonetheless, this observation underscores one of the tradeoffs associated with offering beneficiaries a wide array of plan options and benefit designs.

After the Medicare Modernization Act was signed, but before Part D was implemented, there was a very public discussion that the 10-year cost of the program would be substantially greater than original Congressional Budget Office estimates. But soaring costs have not materialized. According to the 2007 Medicare Trustees report, the average 2007 plan bid was about 10% lower than in 2006.⁶ These savings likely reflect a variety of factors, including vigorous plan competition, increased generic drug use, and a general slowing of spending relative to earlier in the decade. Initial studies using 2006 claims data indicated that Part D was associated with a 16% decrease in out-of-pocket spending for prescription drugs and a 6.5% increase in the number of prescriptions filled in 2006, and these changes were concentrated among low-income beneficiaries.^7–9

Still, cost appears to be an issue for some. In their article in this issue of JAMA, Madden and colleagues³ use responses from the Medicare Current Beneficiary Survey from 2004 to 2006 to assess the association of Part D with cost-related medication nonadherence (CRN). The authors find only small reductions in CRN overall in 2006, and no change among those beneficiaries in poor health or with multiple morbidities. Some of the discrepancy may be a “first-year” effect, particularly because many beneficiaries were only enrolled in a Part D plan for part of the year.

One key policy question is whether the doughnut hole, a well-known gap in the standard benefit under Part D, is driving some of this nonadherence. In 2008, beneficiaries have drug coverage up to $2510 in drug costs, then coverage suddenly stops, at which time beneficiaries pay all of their drug costs up to a certain level ($5726 or $4050 in “true out-of-pocket” costs). Insurance coverage then resumes at a higher level. As a result, beneficiaries with moderate to high drug expenses will pay a substantial fraction of their costs out-of-pocket.

Little data on the effect of the doughnut hole exist, but in other private insurance plans pharmacy benefit caps have been associated with higher rates of medication discontinuation—nearly 25% for individuals taking cardiac drugs (15% for those taking antihyperlipidemic agents)—and only a minority of those discontinuing reinitiated therapy when coverage resumed.¹⁰ This nonadherence is associated with more emergency department use and more nonelective hospitalizations.^11,12 However, a recent report from IMS Health using retail pharmacy claims suggests that the “doughnut hole” affected only 6% of Part D participants in 2006.⁹ Low-income and dual-eligible beneficiaries, 61% of Part D enrollees, were unaffected by the coverage gap (because there is no gap in coverage for them), and 33% of enrollees did not reach $2250 in drug spending over the year and thus never reached the coverage gap. Of those beneficiaries who did enter the gap, a substantial fraction entered in the last days of the year. Among those who entered the gap earlier in the year, many continued through the gap allowing them to be insured at the low, catastrophic rate.

More generally, poor adherence underscores a fundamental problem with Part D. Stand-alone prescription drug plans have little incentive to improve medication adherence, because this means more prescription drug use. Any long-term savings in the form of reduced hospitalizations or less emergency department use accrue to the other parts of Medicare rather than the plan. Medicare Advantage plans, on the other hand, do face these incentives, and it will be important to monitor medication adherence in both types of plans and to evaluate how benefit design develops across these 2 types of products.

The primary objective of the Medicare Modernization Act was to provide seniors with affordable coverage for their prescription medications. This goal has been largely achieved, as more than 90% of Medicare beneficiaries now have prescription drug coverage at least as generous as the Standard Part D benefit.⁴ The fact that average premiums are substantially lower than initial projections provides some evidence that the market is working. The articles by Hsu et al² and by Madden et al³ document beneficiary confusion and problems with adherence—issues that are not likely to be endemic but still must be monitored. It may be that some features of the Part D benefit, especially the dreaded doughnut hole, are ripe for redesign.