|
A correction has been published: N Engl J Med 2009;360(9):944.
|
| |||||||||||||||||||||||||||||||||||||||
|
Patients who face life-threatening illnesses are also facing out-of-pocket costs for cancer therapies that can threaten their family's financial security.7 Their physicians, trained to manage the former set of challenges, are poorly equipped to help with the latter.8,9 Health economists are concerned, too, both by the rising expenditures for cancer care and because the prices of cancer drugs appear to be rising faster than the health benefits associated with them, at least in some cases. Both Schrag6 and Meropol and Schulman10 note such a pattern of diminishing returns in the treatment of metastatic colorectal cancer. With each advance in treatment, the magnitude of the increase in the cost of treatment exceeded the magnitude of improvement in efficacy, thus making each treatment advance less cost-effective than the one that preceded it.
Given these concerns, it is worth assessing why cancer-drug prices and spending have risen so quickly. I believe the growth can be attributed primarily to a unique legislative and regulatory framework that shields cancer drugs (as well as a few other specialty drugs) from the strategies that health care payers such as Medicare typically use to hold down the price and utilization of drugs and other health care goods.
In this article, I illustrate this point, focusing on the Medicare program. I first review the major strategies that Medicare uses to rein in spending on health care goods such as drugs and delineate those strategies into three general categories: those that reduce utilization, those that reduce prices, and those that have both effects (Table 1). For each of the strategies, Medicare requires a particular type of administrative flexibility, which I outline. I then review the laws and regulations that specifically limit these flexibilities when it comes to cancer drugs.
|
Control of Utilization
The primary strategy Medicare uses to hold down utilization of a drug (or another health care good or service) is to limit coverage of payment for it. The program does so by actively determining in which settings the drug is or is not "reasonable and necessary" through either a single national or one or more local coverage decisions. When these coverage decisions result in restricted guidelines for the use of the drug, the result is decreased utilization. For instance, in 2007, Medicare narrowed the coverage of erythropoiesis-stimulating agents (ESAs) for cancer treatment. Medicare limited not only the types of patients who could receive ESAs but also the clinical scenarios in which they could be used.11 Amgen reported to their investors in August 2007 that changes in coverage for ESAs by the Centers for Medicare and Medicaid Services (CMS) would reduce annual sales of the company's ESA, Aranesp, from approximately $1 billion to $200 million among Medicare patients.
Control of Prices
Medicare uses other strategies when trying to hold down the prices of drugs. All rely on the program's ability to conclude that several drugs are clinically interchangeable. Once concluded, Medicare can either encourage price competition among the manufacturers of interchangeable drugs or engage third-party intermediaries to negotiate directly with manufacturers to obtain lower prices.
Blended reimbursement is an example of the application of a reimbursement formula to groups of interchangeable drugs in order to obtain lower prices. Under this approach, Medicare reimburses for the use of a particular drug on the basis of a weighted average of prices for that drug and other similar drugs, with the weighting linked to the sales volume of each drug.
Under this approach, lower-priced drugs garner market share because their price is below the blended-reimbursement rate. As the sales of the lower-priced drugs increase and sales of the higher-priced drugs decrease, the blended-reimbursement rate declines. In July 2007, Medicare implemented blended-price reimbursement for two nebulized beta-agonist drugs that were not identical but that Medicare, through one of its contractors, deemed were clinically equivalent and thus interchangeable. Generic albuterol and branded levalbuterol (Xopenex, Sepracor) were the two drugs.
At the time the policy was implemented, the reimbursement rate for levalbuterol was higher than the rate for albuterol by a factor of 19 ($0.20 per unit of albuterol and $3.84 per unit of levalbuterol), whereas the blended rate was $0.525 per unit (1 mg of albuterol or 0.5 mg of the twice-as-potent levalbuterol).12 Because the reimbursement rate was higher than the purchase price of albuterol and lower than that of levalbuterol, the market shifted to albuterol. From July through September 2007 (the first quarter under blended reimbursement for the two drugs), the shift in sales lowered the blended rate by 16% (to $0.442 per unit), thus reducing Medicare's costs by the same amount. Reimbursement for the two compounds was later "unblended," so it is hard to know how much further prices would have fallen ultimately.
Similar to the strategy of blended reimbursement is that of least-costly-alternative (LCA) reimbursement. With this strategy, Medicare reimburses at the price of the least costly drug among all those that are interchangeable, no matter which drug is actually used. Manufacturers respond to LCA reimbursement by lowering prices so that their prices are not higher than the amount of reimbursement a provider can receive after administering the drug. Many of the local contractors to the Medicare program have instituted LCA reimbursement for the clinically interchangeable prostate-cancer drugs leuprolide acetate (Lupron, Abbott) and goserelin acetate (Zoladex, AstraZeneca). The Office of the Inspector General estimated that in 2002, full use of LCA reimbursement for the two drugs would have saved Medicare $40 million, because the reimbursement rate for the least costly drug (goserelin acetate) was 27% lower than the alternative drug (leuprolide acetate).13
Today, it is unclear whether Medicare has the authority to establish LCA reimbursement at the national level, and its authority to do so at the local level has been thrown into doubt by a recent court decision.14 In that decision, the court ruled that Medicare was required to reimburse for the combination brand-name inhaled drug Duoneb (Dey) on the basis of the price the company charged for the drug, rather than on the basis of the cost of the drug's two components, albuterol and ipratropium, which are both available as generics.
Medicare can also engage a third-party intermediary to obtain competitive bids when clinically interchangeable drugs are available. Under competitive bidding, Medicare puts out an offer to various intermediaries to bid to supply health care goods for clinical care. These intermediaries negotiate with various competing manufacturers, using their purchasing leverage and the intrinsic competition among the manufacturers to obtain low prices. They then present to Medicare bids that reflect the cost savings they have achieved. Medicare recently completed a pilot program of competitive bidding for durable medical equipment and other supplies (such as home hospital beds, wheelchairs, and home oxygen-delivery machines). Through the bidding process, the Government Accountability Office estimated that Medicare saw a 26% reduction in prices.15 A recent law (the Medicare Improvements for Patients and Providers Act of 2008) stopped Medicare from implementing competitive bidding for these goods at the national level.
Control of Both Utilization and Prices
In general, Medicare's strategies for simultaneously holding down utilization and prices involve third-party purchasers. The prospective payment system that is used to pay for inpatient care in most hospitals is a prominent example. Under prospective payment, hospitals receive a single payment that encompasses the costs of an entire hospitalization episode for a patient. With few exceptions, the drugs, devices, hardware, and disposables that are used to care for the patient during hospitalization are paid for by the hospital with the funds included in that single payment. As such, hospitals have an incentive to obtain needed goods at lower prices and actively manage their utilization.
The prospective payment system is credited with slowing hospital expenditures for both Medicare and private payers.16 Russell and Manning estimated that the savings in the first and second decade after the system was instituted were on the order of $12 billion and $18 billion, respectively.17 A number of organizations are developing methods by which prospective payment could be instituted over an extended period of outpatient care, but there are few examples of the approach being used currently.18 Some other terms that are used to describe this approach are "episode-based" payment, "case-rate" payment, and "bundled" payment.
Medicare's Part D (prescription-drug) program employs a third party to obtain drugs at low prices and manage their utilization. Under Part D, drug plans are given significant negotiating leverage when trying to obtain low prices from drug manufacturers, because their formularies need contain only a few drugs from each class of interchangeable drugs, rather than all drugs in that class.19 The plans may also attach different copayment amounts to different drugs within a category, thus holding down utilization of costly drugs and steering market share toward lower-priced drugs when they are available.19,20 Along with varying copayments (in which beneficiaries pay a set amount for each prescription), a recent trend is to introduce coinsurance (in which beneficiaries pay a percentage of the total drug cost, rather than a set amount) for very expensive drugs.7 The RAND Health Insurance Experiment and other studies have shown that increased cost sharing reduces utilization.21,22
There is general agreement that the negotiating leverage and the use of tiered copayments and coinsurance are to be credited for the fact that the Part D program is less expensive than was originally projected. For instance, plan bids for 2009 came in 37% lower than had been projected at the time the program was created.23 Still, policy analysts disagree about the relative strengths and weaknesses of the program as it is currently constituted.
Laws and Regulations Inhibiting Medicare's Flexibility
Drug Coverage
Several provisions of the law address Medicare's coverage of cancer drugs, and Medicare has traditionally interpreted these as mandates that it provide coverage, thus undoing the use of coverage restrictions to limit utilization. For cancer drugs that are covered under Part B, which are generally drugs that are administered in a physician's office, the law requires Medicare to cover any drug used in an "anticancer chemotherapeutic regimen," as long as the use is "for a medically accepted indication" (Table 2). The law defines "medically accepted indication" broadly as uses approved by the Food and Drug Administration (FDA), uses listed in one of several drug compendia, and uses supported in the peer-reviewed medical literature.
|
Drug Interchangeability
A complex set of laws and regulations prevents Medicare from deeming that related cancer drugs are interchangeable, thus undoing strategies that depend on this designation (Table 2). This occurs because the laws have two related effects. First, they classify nearly all newer cancer drugs as "sole-source" drugs. Second, they require that Medicare give each sole-source drug its own unique payment rate, which effectively requires Medicare to give each new drug its own unique billing code. These twin outcomes are the result of the legal definition of sole-source drugs versus multiple-source drugs and of the legal requirements with respect to payment and coding for the two respective drug categories (Table 2).
The law states that drugs qualify as multiple-source only if there are multiple compounds that are pharmaceutically, therapeutically, and biologically equivalent, as classified in the FDA's Orange Book of approved drugs. This standard is strict enough to ensure that no new cancer compound will be sufficiently similar to an existing compound to be classified as a multiple-source drug. On the issue of payment and coding, the law simply states that sole-source drugs must be given their own payment rate. As a result, blended reimbursement for cancer drugs is nonexistent in Medicare, and as noted above, LCA reimbursement may be on shaky legal footing.
A related law undid a previous attempt at competitive bidding for Part B drugs. The law required that all bids contain all sole-source drugs and thus stripped the bidders of any real negotiating leverage that could have come from an ability to designate related clinical compounds as interchangeable.
Medicare has not tried to implement prospective payment for cancer care that stretches over the course of an episode of illness. If it were to do so, it might consider providing to a treating oncologist a lump-sum payment that would cover all the costs of doctor visits, chemotherapy treatments, and the chemotherapy itself over a period of care (e.g., a lump-sum payment for a course of adjuvant chemotherapy). My interpretation of the laws governing the reimbursement of cancer drugs is that they would not allow Medicare to implement such a payment change, because the program is required to reimburse separately for the costs of each drug when it is used.
Private Payers and Cancer Drugs
Although the focus of this article is the Medicare program, it is worth noting that some state laws limit the ability of private payers to curb the utilization of cancer drugs. These laws mandate coverage of cancer drugs in many cases (Table 3).
|
College economics courses clarify why drug manufacturers seek both to charge high prices for their products and to encourage utilization: for-profit corporations have a fiduciary responsibility to their shareholders to maximize their profitability. Medicare has several strategies that counterbalance this profit-seeking behavior. But for cancer drugs and some other high-priced specialty drugs, these strategies have been disabled by laws and regulations that specifically constrain Medicare from making use of flexibilities that it has in other settings.
Policymakers are now in a quandary. To be sure, they are worried both by the increased total costs of cancer care and by the fact that high prices are leading to higher out-of-pocket expenses for patients. Yet, it is difficult for them to discount the progress that has been made overall in the treatment of cancer and to wholly disassociate that from the high prices that innovative cancer treatments are able to capture. Progress that has been made in the treatment of patients with chronic myelogenous leukemia and renal-cell carcinoma, for example, has come as a result of the discovery of new therapies that garner high prices.25,26 Similarly, the doubling of the median survival time among patients with metastatic colon cancer can be attributed in large part to new treatments, which are substantially more expensive than the older treatments they displaced.6 Some health economists argue for a causal link between high prices and the pace of innovation, which would suggest that efforts to reduce spending on cancer drugs would retard the pace of innovation.27,28,29
Thus, policymakers must decide whether they want to see lower prices and reduced spending on cancer drugs by Medicare, and if they do, they must determine how to achieve these ends without stifling beneficial innovation. One possible approach is to judiciously amend or reverse the laws that limit Medicare's flexibility with respect to cancer drugs, while moving rapidly toward the creation of a center for comparative effectiveness that could guide Medicare's actions.30 Such a center could help Medicare ensure that cancer drugs are covered when their use is reasonable and necessary but are not covered when it is not. This center could also create a robust framework for classifying cancer drugs and other specialty drugs as interchangeable when they are, basing the judgment on clinical rather than pharmacologic criteria. This would achieve two ends. Manufacturers would seek to prove through clinical research that their products are not interchangeable with other compounds on the market but, rather, are superior to their competitors. Meanwhile, Medicare and other payers could apply formulas for reimbursement that would encourage lower prices when products are in fact clinically interchangeable.
Giving Medicare the authority (or the mandate) to pursue small demonstration projects of prospective payment in cancer care could lead to an eventual alignment of incentives for both payers and providers. In part, this would help to undo some of the problematic incentives that physicians face because of the "buy and bill" system of reimbursement. Under this system, oncologists profit when they can buy a drug from a manufacturer for less than the reimbursement amount they obtain by billing for the drug after it is administered.31,32 Such demonstration projects would have to include mechanisms and measurements ensuring that patients are not undertreated, an evaluation that could be achieved by monitoring whether oncologists follow practice guidelines.33
Medicare could use a number of other strategies for containing costs and utilization. For instance, UnitedHealthcare instituted prior authorization for physicians ordering trastuzumab (Herceptin, Genentech). Their chief oncology officer reported that the program would probably reduce utilization by about 15 to 20%.34 "Reference pricing" is a version of LCA reimbursement in which the payer covers only the cost of the lowest-priced alternative, and patients pay the difference if they want a higher-priced item.35,36 "Payment for results" is a somewhat new idea that involves reimbursement for a drug by the payer only when it "works." One prominent example: under a preliminary agreement, Britain's National Health Service (NHS) will reimburse for doses of bortezomib (Velcade, Johnson & Johnson) when it has a therapeutic effect but not reimburse for the drug when it does not.37 The NHS has also experimented with limiting coverage only to drugs that are "cost-effective." The National Institute for Clinical Effectiveness in the United Kingdom is the entity charged with making such determinations.38
So, if policymakers seek to slow the upward rise in cancer-drug prices and spending, they could enable Medicare to use its existing approaches or provide legislation that would empower the program to experiment with new ones. Medicare's previous successes in several related areas suggest that the potential reductions in spending on cancer drugs could be meaningful. Determining whether innovation is hampered or helped by such actions should be an important part of the review process.
Dr. Bach reports receiving consulting fees from Genentech, Wyeth, and Abraxis and lecture fees from Genentech, Biogen Idec, and Johnson & Johnson. He reports serving as a senior advisor to the CMS from February 2005 through November 2006. No other potential conflict of interest relevant to this article was reported.
I thank Jacqueline Snow at Memorial Sloan-Kettering Cancer Center for her research assistance and Joan Sokolovsky at the Medicare Payment Advisory Committee for her helpful comments.
Source Information
From the Health Outcomes Research Group, Memorial Sloan-Kettering Cancer Center, New York.
This article (10.1056/NEJMhpr0807774) was published at NEJM.org on January 27, 2009.
References
| |||||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
