0 Speaker List


Evaluating the Senate Finance Committee proposal to restructure Medicare Part D

Scott Gottlieb, M.D.

Senior citizens are facing increasing and often significant burdens from rising out-of-pocket costs on drugs that they need to improve their health while many observers worry about the long run stability of the Part D program. As part of the Prescription Drug Pricing Reduction Act (PDPRA), the Senate Finance Committee is considering two significant reforms to the Medicare Part D program aimed at reducing overall program outlays and beneficiary out-of-pocket spending. Most notably, the proposal would substantially redesign the structure of the Part D benefit to encourage more efficient management of drug spending and reduce incentives that encourage rapid growth of list prices. In addition, it would establish mandatory rebates in Medicare Part D if list prices on drugs grow faster than the rate of inflation (as measured by CPI-U) to reduce enrollee out-of-pocket spending.

This proposal is a well-intentioned effort to improve the incentive structure of the Part D program and protect seniors from rising out-of-pocket burdens. However, these proposals introduce a number of important tradeoffs, that in some cases may challenge the stated policy goal. In this piece, we outline the interaction between Part D benefit design, list price growth, and out-of-pocket spending. In addition, we discuss some of the potential effects of the policies under consideration, and offer alternatives proposals that Congress could consider. These policy alternatives could help achieve the desired policy goals of reducing financial burdens on seniors and bringing more fiscal discipline to Part D drug spending, while creating potentially more desirable tradeoffs.

Incentives Under the Current Part D Benefit

The PDPRA includes an ambitious plan to redesign Part D benefits in order to improve incentives that affect government and beneficiary out-of-pocket spending. Under current law, insurers must offer standard Part D coverage which consists of four phases (whose descriptions we take directly from the Chairman’s Mark of the PDPRA):

  • deductible ($415 in 2019).
  • Initial coverage in which the enrollee is responsible for 25% of the cost of drugs 
(with the plan covering the remaining 75%).
  • The coverage gap (“doughnut hole”) in which the enrollee is responsible for coinsurance of 25% of the cost of brand-name drugs and 37% of the cost of generic drugs, with plans covering the remaining 63% of generic drug costs and 5% of brand-name drug costs and manufacturers providing discounts for the remaining 70% of brand-name drugs.
  • Catastrophic coverage (reinsurance) in which the enrollee is responsible for 5% of drug costs, plans are responsible for 15% of costs, and Medicare subsidizes 80% of costs (the reinsurance subsidy).

Note: This figure taken directly from the Chairman’s mark of the PDPRA.

Importantly, Part D plans reimburse pharmacies as a function of the net price for drugs which reflect a variety of price concessions like rebates or discounts. Yet enrollee cost sharing is often based on the list price of drugs. This has created varied incentives for insurers. On one hand, there is a relatively strong incentive to negotiate for low net prices, since these affect insurers’ costs, and in turn, premiums (and profits). Part D plans typically market themselves to consumers based on the plan’s ability to secure and maintain low premium costs for beneficiaries. At the same time, plans have less incentive to negotiate for low list prices since list prices have muted effect on premiums and insurer costs.

The only instance where the list prices can be directly relevant to consumers is for the calculation of cost sharing. As a result, rising list prices have a much less noticeable, but potentially pernicious effects on the subset of patients who will take drugs with high list prices and reach certain coverage limits.

Moreover, understanding one’s exposure to these out of pocket costs is made challenging by the fact that consumers do not necessarily know which medicines they will need. Nor do they generally know the list prices of these drugs, or how much these prices may increase each year.

The fact that net prices are relevant to insurers’ costs, and list prices largely impact patients’ cost sharing, is likely one of the unfortunate reasons that list and net prices have diverged substantially in recent years. For example, the Centers for Medicare and Medicaid Services estimated that net costs to Part D plans averaged 17.5% lower than wholesale acquisition cost prices in 2014 — a number that has likely grown in recent years. In other words, there is often a very large difference between the list price facing patients and net price facing insurance plans.

As the Medicare Payment Advisory Commission (MedPAC) has noted, the Part D benefit structure includes a number of incentives that likely further contribute to this. Specifically, “because plan sponsors are not liable for much benefit spending in the coverage gap, Part D’s structure may provide a financial advantage to sponsors when they select certain drugs with high prices and large post-sale rebates over lower cost alternatives.” [1] Drugs with high list prices and large rebates move enrollees through the coverage gap more rapidly, since the drug’s price concessions count towards enrollee out-of-pocket spending. Once enrollees reach the catastrophic phase, the federal government covers 80% of costs. Further, enrollees have no cap on the amount they can spend out of pocket. Right now, as the list price of a drug rises, so does beneficiary out-of-pocket costs, with no limit on their total liability.

Insurers are meant to be well-informed agents that act on behalf of enrollees to design policies that mitigate unexpected financial risk (in exchange for premiums). Within this framing, it is odd that insurance plans would be disincentivized from preventing these kinds of volatile and potentially large financial shocks to patients. The fact that such a design can persist in a market suggests that the feature of plans that expose patients to high and unpredictable cost sharing is, perhaps unsurprisingly, not transparent to consumers.

Lowering Program and Beneficiary Spending Through Part D Redesign

In an effort to improve the design of the Part D benefit, the PDPRA includes two sets of changes.

First, it includes the following explicit changes to out-of-pocket spending:

  • Change enrollee cost sharing in the initial coverage limit and the coverage gap.
  • Cap enrollee cost sharing above the catastrophic out-of-pocket threshold.
  • Change the amount of annual out-of-pocket spending needed to trigger catastrophic coverage (from $8,139 to $3,100).

In addition, the financing structure of the benefit would be modified as follows:

  • Lower federal reinsurance during the catastrophic coverage period to 20%.
  • Sunset the existing manufacturer discount program in the coverage gap.
  • Institute a new manufacturer discount program in the catastrophic coverage phase of the benefit (equal to a 20% discount off of net prices for branded drugs, presumably structured in the form of a mandatory rebate to be paid by the manufacturer to the federal government).

Note: This figure taken directly from the Chairman’s mark of the PDPRA.

Anticipating Effects of Benefit Redesign

These changes mirror many of those suggested by MedPAC and represent a major change to the benefit design of Part D. These reforms will directly alter beneficiary spending and, because of the size of the Part D program, meaningfully change the incentives facing drug manufacturers.

This reform would reduce many of the problematic incentives in the current Part D program. Because insurers bear a substantial portion of costs throughout the entire benefit schedule, they have clearer incentives to control total spending. Moreover, they are no longer incentivized to steer patients towards drugs with high list prices and large rebates in order to push them into the catastrophic coverage phase. There will also be more incentive for insurers to try and closely manage beneficiary spending on high-priced drugs, through tools to control utilization, in order to keep beneficiaries out of the catastrophic tier where plans will now be liable for 60% of the total drug spending. This will lower the number of beneficiaries reaching the catastrophic tier, thereby reducing federal spending.

The effect of these changes on premiums is not necessarily immediately obvious. Some early reports suggest that premiums may remain flat, or even decline. However, there is the possibility that other scenarios could unfold as the market participants absorb these changes. Under this proposal, Part D plans would clearly have greater liability in the catastrophic phase of coverage, while Medicare’s subsidy would now be delivered more through capitated payments to insurers, and less through open-ended reinsurance payments. Whether this translates to higher premiums will depend on a variety of factors, like whether the total federal subsidy for plan participants remains constant and certain behavioral responses from insurers. For example, there is some uncertainty around the regulatory ability and market effectiveness of plans to adopt various tools, such as formulary design and step therapy, to more closely manage the catastrophic risk. In addition, plans may build in a larger risk premium — especially in short run — as they take on new liabilities. On balance, we think it is possible that premiums increase, at least in the near to medium term, as market actors absorb these changes.

In addition, the new benefit design could also, over time, change the investment decisions made by drug manufacturers in terms of the particular areas of drug development that they pursue.

When the specialty tier was first created in the Part D benefit, this structure was — in part — an effort to preserve market-based pricing for first-in-class drugs targeted to unmet medical needs. The benefit was designed at a time when there was substantial criticism that drug makers didn’t pursue enough genuine innovation, and too many new medicines were “me too” iterations on existing drugs. The creation of the specialty tier was the result of an effort to maintain the cost sharing required by law, in a fashion that was perceived as being fairer to Medicare beneficiaries. There was also a view that the overall structure of Part D, and the creation of the specialty tier in particular, would help make the market more highly competitive for drugs where there were “good enough” generics; while shifting financial incentive toward reimbursement for first-in-class drugs targeted to significant, unmet needs.

These decisions in the initial design of the Part D benefit have yielded certain intended benefits. Over the ensuing 15 years there has been a significant shift in investment capital toward drugs that make it onto the specialty tier, where reimbursement is more protected, with oral oncology drugs making up an outsized proportion of new investment, and in turn, innovation. This has resulted in many successful new drugs that deliver meaningful benefits to patients. On the other hand, there has been a shift away from investment in drugs that target primary care conditions and, in many cases, medicines that would compete with generic drugs that may offer comparable or slightly inferior benefits but at substantially less cost. Moreover, as outlined at the outset, there have also been other unintended consequences; including significant out-of-pocket costs for beneficiaries and rising program spending.

Based on this background, in addition to capping patient out-of-pocket costs and creating more incentive for Part D plans to more actively manage spending on specialty drugs, one thing to watch for is how the structure proposed in PDPRA may also begin to shift investment decisions anew.

On the margin, the new benefit design may steer investment away from drugs that would fall onto the specialty tier, and toward medicines that are generally priced below an annual cost that would trigger the substantial mandatory rebating under the new framework. This would typically be drugs that are priced at or below around $25,000 in annual list cost, based on a very rough assessment of how the new benefit could take shape. This might include new oral drugs that target complicated aspects of more prevalent diseases like pulmonary, cardiovascular, metabolic, or rheumatological conditions. In these therapeutic areas, there are highly effective generic drugs but still plenty of unmet medical need and the opportunity for improvement over the current standard of care.

Potential Effects of Limiting List Price Inflation

In addition to redesigning the Part D benefit, the PDPRA would explicitly limit the growth of list prices to a measure of inflation (specifically, CPI-U). Limiting the growth of list prices to inflation is one way of potentially controlling the evolution of beneficiary out-of-pocket spending. In the short run, this policy is likely to mechanically reduce the growth in consumer out-of-pocket drug spending by restricting the rate at which already-launched therapies can increase their list prices. The net prices are also based on already negotiated discounts off these list prices. If this reduces the number of beneficiaries entering the catastrophic phase of coverage — where the federal government currently pays 80% of costs — then federal spending will likely fall in the near term relative to a baseline scenario. Given that catastrophic spending has been growing at nearly 17% annually, changes to trends could result in non-trivial, short-term budgetary savings.

Over the long run, it is less obvious how this would affect enrollee and program spending. There are some considerations that will need to be closely evaluated and monitored for. Manufacturers would be incentivized to increase the initial launch list price of new drugs, knowing they won’t be able to take list price increases later on. The mandatory rebate under PDPRA for drug costs above the catastrophic cap could provide further pressure on drug makers to try and launch new medicines at higher prices, knowing that they will have to rebate a potentially substantial portion of the net price back to the government.

If such increases at launch were sufficiently large, they could offset some, or much, of the policy’s intended effect on out-of-pocket spending. Thus, it is not clear where the balance in this proposal will ultimately rest between its goal of substantially changing the market incentives versus the proposal’s ability to provide a meaningful, but perhaps one-time, reduction in beneficiary out-of-pocket spending on already-marketed medicines. Moreover, setting an inflation cap on list price increases could have the unintended effect of making CPI-U both a floor and a ceiling on price increases by sanctioning CPI-U as a politically permissible limit on annual, portfolio-wide price increases.

Alternative Options for Addressing Beneficiary Cost Sharing

There may be other mechanisms to achieve similar goals as the Finance Committee proposal, with more direct benefits to the patients, and fewer potential distortions introduced by the current plan.

First, if Congress wants to better align the growth of list and net prices, they could effectively force plans to do so. Health plans and PBMs demand — and in most cases secure — caps on the growth in the net price they pay for drugs as part of contracts they enter into with manufacturers. Under these agreements, if a drug maker increases their list price, then they must also increase the rebate they pay to the health plans in order to provide an effective cap on the costs to the plan. In most cases these agreements peg net price increases at rates that are flat, and generally below inflation.

Congress could leverage the outcome of these negotiations by mandating that inflation caps that are already negotiated between the drug makers, health plans, and PBMs must also be applied to the list price for the purposes of calculating beneficiary out of pocket costs. In the short term, this would protect beneficiaries from large and persistent increases in the cost sharing and remove the requirement that policymakers determine an appropriate inflation rate. However, it shares the same basic challenge facing Senate Finance’s current proposal: Drug makers would still be incentivized to launch new drugs at higher prices, knowing they could not raise them much later if list price growth were tethered to the flatter net pricing. In the long run, this may undermine the goal of this policy.

A second, and perhaps more natural approach, would be to require that patient out-of-pocket spending on cost sharing be a function of net prices. Such an approach would be akin to how health insurance works in other markets. Patient cost sharing for hospital care, for example, is typically not a function of hospital list (or “chargemaster”) prices. Linking drug costs for plans and beneficiaries would better align incentives between the two.

There is at least one operational challenge to such an option, however. Net prices reflect a host of price concessions from manufacturers, some of which occur with considerable delay (like rebates, for example). At any moment in time it is not clear what the true net price will ultimately be for that transaction once all price concessions are calculated. Thus, this type of proposal would almost surely need to base cost sharing on some lagged net price so that it reflected prices after all concessions had taken place. This would be similar to the current two-quarter lag in calculating Average Sales Prices in Part B. One potential benefit of basing the cost sharing on lagging net prices is that such an approach would limit the potential ability to use the calculation of cost sharing as a way to discover the current net price paid on a drug (and thus the concessions that drug makers are providing to health plans). If the net prices were easily discoverable, it could reduce the competitiveness of these negotiations.

Finally, Congress could consider altering the type of cost sharing offered by Part D plans, and fix cost sharing at a clearly defined set of co-pays based on the tier that a drug is placed in. This has the advantage of making cost sharing more predictable and consistent across drugs in a similar class. Plans can still have the opportunity to incentivize the use of cheaper drugs via larger or smaller copayments.

This option could help alleviate a core market friction — the interaction between list prices and consumer out-of-pocket spending is not salient to individuals purchasing coverage. Co-pays expressed in dollar terms are a far more transparent way of communicating information about expected out-of-pocket costs and may allow plans to better compete over out-of-pocket spending instead of largely competing on premiums.


The core principle of the well-intentioned efforts being pursued by the Senate Finance Proposal is to put patient cost sharing on a more predictable trajectory and reduce total spending by taxpayers. The proposal attempts to do this by limiting the list price increases — on which cost sharing is generally currently based — to CPI-U and restructuring the Part D benefit to shift more of the cost of catastrophic drug spending away from beneficiaries and taxpayers and onto health plans and drug makers.

But such an approach will have to be evaluated closely to make sure it does not introduce unintended consequences that undermine stated policy goals. For example, will efforts to limit out-of-pocket prices facing beneficiaries fundamentally change long-run pricing dynamics, or simply represent a one-time savings that is offset by novel pricing strategies? Reforming the underlying shortcomings in this market requires that we ameliorate these market distortions and avoid introducing new ones.

Moreover, capping beneficiaries’ total liability at $3,100 may still set a patient’s total out-of-pocket spending at a level that is unaffordable to some seniors. This is especially true when one considers that patients who reach this coverage limit typically suffer multiple morbidities. They may also have substantial out of pocket costs related to physician office visits and inpatient hospital stays.

The proposed legislation is likely to generate substantial savings through mandatory rebates being paid by drug makers and CPI-U caps on future list price increase. The immediate impact of these changes could come in the form of changes to Part D premiums, an eventual re-allocation of investment capital toward different public health priorities, and potentially more restricted patient access to some higher-price specialty drugs categories like oral oncology drugs. Set against the savings generated by PDPRA, and considering the potential unintended consequences that could flow from any change of this magnitude, we should make sure that efforts to lower drug costs and offer protection from catastrophic risk ultimately provide seniors meaningful assistance.

Get A Quote For: