Payment for Oncolytics in the United States: A History of Buy and Bill and Proposals for Reform

Abstract

The authors review and discuss the opportunities and challenges raised by policy options for Average Sales Price reform or replacement.

Introduction

Treating patients with cancer with infused or injected anticancer prescription drugs (ie, oncolytics) is a central component of an outpatient oncology practice. Practices must purchase these drugs and then bill insurers for their use to treat specific patients, a system colloquially called buy and bill. Third-party reimbursement and patient cost sharing for oncolytics are the largest sources of gross revenue for oncology practices (> 50% of gross practice revenues are related to billing for oncolytics).¹ As the prices of cancer drugs have grown over time, these purchases have had a significant impact on the financial health of practices and created significant practice risk, jeopardizing the ability of many practices to operate and provide patient care.² In this article, we explore the history of the Medicare reimbursement policy for drugs provided in outpatient oncology practices, with a focus on the intended and unintended consequences of the current average sales price (ASP) –based model. Options for ASP reimbursement reform or replacement are presented.

Physicians who treat Medicare fee-for-service beneficiaries are and have been reimbursed for most oncolytics under the Part B benefit. Currently, Medicare pays for Part B–covered drugs using an ASP cost-plus-reimbursement system, implemented in 2006 after the passage of the Medicare Modernization Act (MMA) in 2003. ASPs for all Part B–covered drugs are calculated and reported by manufacturers to the Centers for Medicare and Medicaid Services (CMS).³ The MMA set CMS reimbursement for these drugs at the ASP plus 6%, which preserved the buy-and-bill system while reducing the potentially significant profits that could be reaped from the difference between the acquisition costs of Part B–covered drugs and reimbursement rates.⁴ Medicare beneficiaries treated with these drugs are subject to a 20% coinsurance requirement, which may be covered under secondary insurance plans (so-called Medigap). The Part B deductible also applies. There are no out-of-pocket caps for beneficiaries, and coverage of oncolytics is essentially guaranteed for US Food and Drug Administration–approved indications and many off-label uses as well.⁵

The ASP plus 6% reimbursement has recently come under attack. In the 2012 budget sequester, Congress mandated that CMS cut Part B drug reimbursement by 2%. As a result, practices are now reimbursed for oncolytics at ASP plus 4.3%. Intended to be temporary, in February 2014, Congress extended these sequester cuts through 2024. In addition, the 2014 presidential budget proposed to cut Medicare Part B reimbursement from ASP plus 4.3% to ASP plus 3%.

Current Part B reimbursement policy faces this type of political scrutiny because of federal spending pressures. Part B spending represented 18% ($498 billion) of the 2013 federal budget, and oncolytics accounted for 25% of total Part B spending.⁶ Both the President and Congress look to the Medicare program as a major source of funds to pay for other legislative changes or deficit reduction efforts. In 2014, the Congressional Budget Office (CBO) estimated President Obama's 2014 proposed cut would result in savings of $20 billion over 10 years.⁷

How to reduce Medicare Part B spending is limited to two options: extracting savings from Medicare beneficiaries through reduced benefits, reduced access, increased premiums, or cost sharing or from providers through lower physician fee schedules or reimbursements for covered drugs. Congress typically focuses on cuts that are the easiest to implement and are estimated to produce a positive score by the CBO. Physician fee schedules are already below feasible levels, and additional reductions are neither practically nor politically tenable. Curtailing Medicare beneficiary benefits has similar challenges. This leaves reductions in ASP-based reimbursement as a convenient target of reform—scored to save billions of dollars over the 10-year budget window for each percentage-point reduction.

Providers may be able to advocate successfully against additional Part B reimbursement cuts in the short term by citing major disruptions to how outpatient oncology care is delivered. However, this strategy is untenable over the long term, prompting real debate both by policymakers and within the oncology community regarding the merits of preserving buy and bill.

History of ASP Plus 6%

Facing a drug reimbursement system that was believed to be costing CMS and Medicare beneficiaries billions of dollars in overpayments,^8–14 in 2003, Congress enacted the MMA. The MMA altered incentives on both the buy and bill sides of Part B–covered drugs. On the bill side, the MMA required Medicare Part B payment for drugs administered to patients by outpatient oncology practices (excluding outpatient hospital departments) to be reimbursed on the basis of 106% of the ASP. This change established Medicare reimbursement for these drugs on a more readily verifiable and market-based price measure—their acquisition costs plus a 6% margin to cover overhead costs for drugs administered in physician offices¹⁵ or plus an annually updated margin (also currently 6%) for separately payable drugs administered in hospital outpatient settings.¹⁶ (Separately payable drugs are those not packaged within ambulatory payment classification group, because their average cost per day of treatment exceeds $80; Medicare Hospital Outpatient Prospective Payment System has typically reimbursed these drugs at ASP plus 4% to 6% margin.)

On the buy side, the MMA created the Competitive Acquisition Program (CAP), implemented in July 2006 and designed to produce significant cost savings by reducing the acquisition costs of Part B–covered drugs for physician practices and passing those costs on to Medicare. Under the CAP, participating physicians acquired oncolytics administered in their office through a vendor. The vendor would directly bill Medicare for the use of the drug. The program was intended to use competitive bidding by multiple vendors to lower drug costs for Medicare, similar to Medicare Part D, the outpatient pharmaceutical benefit created in the MMA, while substantially reducing the role of practices as middlemen in the system.

CMS suspended the CAP at the end of 2008. With fewer than 4,500 physicians (most of whom were not oncologists) and one lone participating vendor, the CBO projected savings would not occur. When the vendor withdrew from the program, the CAP was suspended (Appendix, online only, provides short history of CAP).

Opportunities and Challenges In Using ASP to Reimburse Oncology Practices

Although ASP-based MMA reimbursement has led to some of the most dramatic reductions in Medicare Part B spending in its history, there are considerable flaws in the methodology that have become increasingly apparent. First, the prices of many drugs covered under the Part B benefit have risen substantially since the implementation of the MMA.⁴ CMS posts a new ASP every quarter based on information submitted by drug manufacturers 6 months earlier. As a consequence, physician reimbursement remains stagnant for two quarters after drug acquisition costs rise.¹⁹ Thus, the risk of a drug price sinking underwater is borne by outpatient practices. Furthermore, ASP-based reimbursement does not seem to have curbed the initial pricing of a branded cancer drugs, most of which now meet or exceed $10,000 per month of treatment. It has been suggested that the launch prices of these drugs are high in part because manufacturers understand the risks practices face if pricing of a drug rises after launch.

Second, the ASP plus 6% calculation includes prompt pay discounts intended to benefit distributors. Their inclusion in the calculation effectively lowers the ASP (by some estimates to ASP plus 4%), even though providers do not benefit from the discounts in their purchases. With sequestration cuts in place, the inclusion of prompt pay renders provider reimbursement for these drugs as low as ASP plus 2.3%.

Third, reliance on the ASP as the basis of outpatient oncology practice drug reimbursement also has had heterogeneous and unequal effects across providers. On the bill side, practices can have dramatically different payer mixes because of socioeconomic characteristics of their locales. Commercial payers were slow to adopt ASP pricing and generally continue to reimburse more generously than Medicare, making practices that have more commercially insured patients less vulnerable to the ASP and its changes over time.

On the bill side, the ASP incentivizes outpatient practices to seek the lowest acquisition cost available for a given drug. Some practices have purchasing advantages unavailable to others. For example, institutions that are eligible for 340B drug discounts access drugs at relatively low costs, insulating them from many of the financial pressures independent practices face.²¹ Large group and institutional practices with lower drug acquisition costs have a competitive advantage over smaller practices. Large practices typically have better access to capital and favorable commercial contracts compared with smaller practices, allowing them to more easily benefit from any spread between insurer reimbursements and the acquisition costs of drugs.

The risks, inequities, and incentives in the ASP system have become exacerbated in recent years, because the acquisition costs for Part B–covered oncolytics have risen exponentially. Many small to medium-size practices recognize that they cannot risk providing new, expensive therapeutics in the office. An entire clinic can be jeopardized by a failure to be wholly or partially reimbursed in a reasonable timeframe for a given dose or cycle of an expensive drug. This is particularly problematic in rural and other underserved areas, where disadvantaged practices must seek size, capital, and management to survive. Given the lack of alternatives, such practices often must join with competitors, practice management organizations, or hospital or institutional networks. These decisions influence the number of available oncology practices in a community and also ultimately affect the prices that payers will pay for those services.^22–24

Finally, ASP-based reimbursement may have some perverse effects on the supply of generic oncolytics. Once multiple manufacturers produce a given generic, competitive market forces drive their acquisition costs to levels that leave manufacturers thin margins. If a generic manufacturer faces increases in the prices of raw material acquisition or costs of manufacturing and/or distribution, it may wish to increase the acquisition costs of its drugs. However, the lag in ASP reimbursement forces generic manufacturers to either assume all or some of the financial consequences of increased production costs or pass the price increases on, creating an underwater scenario for practices. Facing these alternatives, manufacturers may opt to cease production of these drugs or outsource them to contract manufacturers. This dynamic may have contributed to ongoing generic cancer drug shortages.²⁵

Alternatives to ASP

In part because of these problems, efforts to modify ASP-based reimbursement began even before the MMA was enacted. For example, advocacy to remove prompt pay discounts and shorten the two-quarter lag in ASP calculation has been a persistent goal of the American Society of Clinical Oncology (ASCO) and other professional oncology organizations. CMS has repeatedly rebuffed these suggestions, arguing, for example, that administrative complexity prohibits reducing the two-quarter lag in reimbursement.²⁶

Now there is growing interest in alternatives to the ASP-based reimbursement system. In effect, policy options entail either changing the reimbursement benchmark while preserving the general structure of buy and bill or replacing the system altogether. Here we discuss viable policy options:

Preserve Buy and Bill but Alter Reimbursement Metric

Invoice pricing and average acquisition cost.

Conceptually, the simplest model to replace ASP-based reimbursement of oncolytics while preserving buy and bill would be to reimburse practices for their drug acquisition costs. There are several possible benchmarks that like the ASP, purport to approach acquisition costs (Appendix Table A1, online only, lists alternative benchmarks and advantages and disadvantages).

The benefits of invoice-based reimbursement are manifold. First, any incentive to prescribe more expensive drugs or to overuse supportive care pharmaceuticals would be eliminated. Second, smaller practices with limited buying power would face a level playing field with larger or hospital-affiliated practices.

A major criticism of this reform is that it would provide little incentive for practices to seek out the lowest price for a drug, because they would be reimbursed fully regardless of the acquisition cost of the drug. An alternative would be to use average invoice prices across all oncologists (potentially stratified by oncology size and practice setting to deal with 340B and other cost distortions in market). This would preserve the incentive to seek out the lowest price for a drug and, if stratified, provide some relief for small, rural practices that provide critically important access to care but lack the market clout to obtain discount pricing. A pharmaceutical management fee would be required in such a system to offset unreimbursed practice costs, including support for procuring, handling, and maintaining inventory of the drugs.

More fundamentally, the average acquisition price approach shares many of the same problems as the ASP. The lag as CMS acquires acquisition cost data from individual practices would remain and probably be longer, because the process would be more cumbersome. Practices would also have to assume the cost of switching to the new system.

Least costly alternative and dynamic pricing.

The least costly alternative is an alternative to the ASP in which practices would be reimbursed based on the level of evidence that supports the use of a given drug to treat a given patient. It is a concept recently considered by the Medicare Payment Advisory Commission, an independent, Congressionally chartered organization that advises Congress on the Medicare program. Evidence for a given drug-indication pair would be judged by CMS to fit three general levels: one, more effective than current standard of care; two, equally effective as the standard of care; or three, relative effectiveness unknown based on current information. A drug deemed to be more effective would be paid on the current ASP basis. New drugs deemed equally effective would be paid the ASP reimbursement of the standard-of-care drug. If no information were available, the drug manufacturer would be required within a certain timeframe to produce data on the effectiveness of the drug compared with alternatives. The new drug would be paid on the current ASP plus percentage basis during this timeframe, and at its conclusion, the drug would continue to be paid at its ASP plus percentage rate if found to be more effective than the prior standard or at the lower alternative rate if greater efficacy was not demonstrated. Comparative-effectiveness models also could vary beneficiary cost-sharing requirements based on the effectiveness of given therapy compared with its alternative.

This model has significant limitations, because it is predicated on the assumption that comparators exist. For many drugs in oncology, this is not true, and the comparator is supportive care. Even if a comparator exists, defining “more effective” must be addressed. Is it overall survival, progression-free survival, response rate, or lower toxicity? This model is further complicated by molecularly defined subgroups of patients with a common disease. A drug may be more effective for one molecular subgroup of patients but equally effective for others. This raises the question of whether reimbursement should be patient, rather than drug, specific. Political considerations have also limited the use of comparative effectiveness for payment purposes; for example, the Affordable Care Act restricts CMS from using it for reimbursement decisions.²⁷

Bundling and episode-of-care payment.

Another attractive option for many payers is bundled payment—a set payment for all cancer care over an episode of care that includes estimated cancer drug costs. Under this policy, oncologists would determine how to manage care costs to stay within the budget of the bundle. This policy would discourage oncologists from using expensive drugs that are not part of the usual care for a particular cancer diagnosis and stage. Outlier payments would be available for atypical cases. Bundles could be recalculated each year or could grow by a defined inflation index akin to the diagnosis-related group system. This set of incentives would likely put significant downward pressure on spending.

Like the least costly alternative policy, bundling suffers from many logistic hurdles, particularly as cancer becomes subdivided into thousands of different diagnoses that have vastly different drug costs. Separate bundles could be set for each subtype, or the system could rely on outlier payments for expensive therapeutics. Medicare would still face the question of what drug prices to include in the bundled payment calculation. Regardless of how the bundle were calculated, oncologists could face the choice of denying patients newer, more expensive drugs (which exceeded bundled payment) or accept the financial consequences for prescribing nonreimbursed drugs.

Replace Buy and Bill and Remove Physicians, Practices, and Hospitals From Oncology Drug Purchasing Business

There are multiple reimbursement policies that would eliminate buy and bill for practices. The key design question is whether the new purchaser should be the government itself or commercial intermediaries.

The health ministries of the United Kingdom and many European Union member countries either directly negotiate drug acquisition costs on behalf of physicians and hospitals or set payment limits and do not require patients to pay copayments or coinsurance at the time of their treatment (Appendix Table A2 [online only] reviews key features of drug access policies of selected countries).^28–30 In the United States, the adoption of a similar policy would likely cede drug price setting authority to the Secretary of the Department of Health and Human Services.

These policies have many advantages, the largest being that the consolidation of negotiations for all drug purchases may result in national cancer drug treatment spending reductions per capita. However, part of the ability of these countries to extract cost concessions from branded pharmaceutical manufacturers (generic drugs are cheaper in United States than in other countries^29,30) lies in the complementary adoption of formulary restrictions, which may include significant delays in drug access for patients. For example, some drug industry critics cite the British National Health Service as a model for restraining drug prices in the United States. The National Institute for Clinical Effectiveness evaluates the cost effectiveness of new drugs and restricts access of National Health Service beneficiaries to cancer drugs that are not sufficiently cost effective. These restrictions are quite unpopular, and in 2011, Prime Minister David Cameron created the £200 million Cancer Drugs Fund to pay for otherwise noncovered cancer drugs.³¹ Furthermore, the long-term implications of the adoption of these policies in the United States could be quite negative, because consolidated government purchasing policies for many goods can lead to stagnant investment in innovation.³² For these and other reasons, such a policy approach has limited political viability in the United States and was explicitly rejected when Medicare Part D and the CAP were created through the MMA.

A more politically palatable set of policies would cede central negotiating power to commercial entities for the purchasing of these drugs in place of individual provider practices and central government control: one, transfer Medicare coverage of these drugs from Part B to the prescription drug benefit (Part D); or two, resurrect the CAP.

Move Part B drugs to Part D.

The MMA created a new benefit for fee-for-service Medicare beneficiaries (Part D) that includes coverage for most oral oncolytics, starting in 2006. Before then, the Part B outpatient benefit covered only oral oncolytics that were reformulations of older infused or injected drugs. According to a recent report, oral oncolytics currently rank first in spending among drugs covered under third-party payer pharmacy benefits.³³

Acquisition and coverage policies are quite different for drugs currently covered under the Part D benefit. Commercial insurers that participate in Part D determine the coverage level within Medicare requirements and either negotiate prices with drug manufacturers or delegate this function to pharmacy benefit managers (PBMs). These negotiations typically include discounts off the average wholesale prices of drugs based on purchase volume. Beneficiary copayments are collected at the point of sale in retail pharmacies and differ based on plan design. Part D plans typically include formularies, but they are required by law to cover at least one oncolytic for every class of medication. After meeting their pharmacy plan deductible, beneficiaries pay 25% of the drug cost up to the catastrophic limit (this will be true by 2020 when so-called doughnut hole is eliminated) and are then responsible for 5% of the cost.

There are potential advantages to moving Medicare coverage of oncolytics from Part B to Part D. The policy does have intuitive political appeal and was considered during the Congressional negotiations that ultimately culminated in the MMA. The main economic rationale for transferring the coverage of drugs from Part B to Part D is that it would open up the competitive market for the pricing of parental oncolytics by allowing large health plans and PBMs to negotiate deep discounts off list prices with manufacturers. According to a 2011 Zitter Group report, approximately 25% of national sales of Part B–covered oncolytics are already reimbursed under Part D in a practice termed white bagging.³⁴ White bagging occurs when a physician prescribes a drug-based treatment for a patient, and the payer directly contracts with a specialty pharmacy to purchase and deliver the product to the clinic or practice for use by that specific patient. The clinic or practice does not purchase the drug, nor does it receive reimbursement for the cost of the drug. The clinic or practice does receive reimbursement from the payer for administering the drug to the given patient. Intriguingly, the commercial plans surveyed in this report expect white-bagging sales to increase to > 50% in the next several years. This expectation is based on the savings and control inherent in Part D as opposed to Part B coverage. This also suggests that regardless of Congressional action, Medicare and other payers may continue to pay for an increasing share of infused or injected oncolytics under the pharmacy benefit.

From a provider's perspective, the main advantages of this approach are simplicity and parity. An oncologist does not as a rule choose a chemotherapeutic agent for a patient based on its formulation but rather based on the evidence for that particular patient. The current reimbursement system can impose different burdens on patients and physicians based on whether drugs are intravenous or oral. This disparity will only grow larger as more oral therapies enter the market.

A major challenge with shifting Part B drugs to Part D is the potential for patients with cancer to be priced out of the marketplace because of high copays and deductibles. In general, patients pay more under Part D than Part B because of the differences in cost sharing (25% v 20%) and deductibles (maximum, $310 v $147). This is more complicated for oral chemotherapy drugs, where many patients will reach the catastrophic limit quickly ($4,550 in 2014) and will then be subject to a lower cost-sharing amount. Acumen projected an average annual increase of $391 for beneficiaries receiving anticancer drugs if they were moved from Part B to Part D.³⁵ However, those patients whose costs ended up in the catastrophic coverage range saved $166 per year. Acumen estimated that premiums for Part D plans would increase < 1% because of the small percentage of patients actually receiving anticancer drugs. In addition, this increase might be offset by a similar decrease in the Part B premium. Any move from Part B to D would need to be done in a cost-neutral manner for Medicare beneficiaries.

A related issue is the use of tiered specialty formularies. Although Part D providers are required by law to cover at least one cancer drug in each class, there are few restrictions on copayment or coinsurance amounts for those drugs. Restrictions on the level of beneficiary responsibility for each course of drug would need to be included in any legislation.

The final hurdle to moving from Part B to Part D relates to the net cost to the Medicare program. Given the monopoly position of drug manufacturers for many chemotherapeutics, it is unclear how much negotiating clout Part D plans would have. Indeed, Acumen predicted that anticancer drug costs would be 3% higher under Part D than Part B. Assuming the CBO agreed with this assessment, Congress would have to either find offsets for this amount or be willing to spend more for Medicare. This may prevent the current Congress from acting on such a proposal.

Resurrect modified CAP.

For several reasons, a restructured CAP may be received by providers and potential vendors quite differently than it was a decade ago. Capital costs and risks associated with high-priced drugs have acted to promote just-in-time inventory management, with drugs being drop shipped on a nightly basis to practices. Practices have also adopted tight vetting processes to determine patient insurance coverage and collection of any copayments and deductibles before ordering drugs and scheduling patient treatment.

To resurrect the program, CMS would need to develop rules that would ensure that multiple specialty drug vendors are attracted to participate. Multiple vendors would be needed to provide choices for practices and foster competition that would promote excellent service and put downward pressure on drug costs. A new CAP should replicate the current drug supply chain like current group purchasing organizations, with the exception that the billing and copays would be transferred from physicians' offices to the CAP (and CMS). This would require special consideration of what to do with bad debt (where beneficiary coinsurance is not paid).

Under a resurrected CAP, an administrative payment should still be made to practices by payers for inventory management and billing. Under the original CAP, it was assumed that no additional payment would be needed. However, the administrative burdens on practices are still substantial, and physicians will not likely participate without it. In exchange, practices should accept drug delivery to ensure quality control for patients and payers. Unlike the original CAP regulations, CAP vendors should be responsible for tracking unused drugs (Data Supplement provides 2005 ASCO memo to CMS regarding changes necessary to make CAP successful).

Value-based purchasing algorithm to set prices.

There has been considerable discussion regarding the use of care pathways that incentivize the use of specific chemotherapeutic agents and supportive care drugs based on a patient's particular condition. The concept of explicitly rating these pathways based on efficacy, cost, and toxicity is evolving. It is beyond the scope of this article to address the details of value-based reimbursement, but any drug reimbursement model such as CAP or shifting from Part B to Part D could incorporate value based pricing. This could be done by altering cost-sharing requirements for patients or by providing financial incentives for providers to prescribe higher-value drug treatment regimens.

Discussion

As the cost of oncolytics and other drugs covered under Medicare Part B inexorably rise, the accompanying risks and inequities in the buy-and-bill system for providers, patients, and payers become apparent. As enticing as it may seem, a system in which the government sets prices for these drugs is not politically viable in the United States. However, failure to reform or replace this system will see further attrition to the margins that accompany the ASP and the infrastructure that has been built around it. Providing cancer drugs through Medicare Part D and PBMs or launching an updated CAP are two viable reform options. We exhort provider organizations, the pharmaceutical industry, drug distributors, policymakers, and payers to work together to replace buy and bill with a reimbursement model that will stabilize and preserve access to quality care for patients with cancer. For access to quality community-based cancer care to be preserved, it is critical that a portion of any savings accrued by way of the ASP be used explicitly to reimburse practices for the significant (and largely uncompensated) care costs of managing a complex disease population.

Appendix

Competitive Acquisition Program

In 2003, the Medicare Modernization Act (MMA) established the Competitive Acquisition Program (CAP). This program implemented a competitive bidding process allowing the Centers for Medicare and Medicaid (CMS) to choose vendors that would supply selected Medicare Part B drugs and biologics directly to physician practices. A practice elected to participate on an annual basis by choosing a CMS-approved vendor. Providers who participated in the CAP obtained designated Part B drugs through the vendor in a manner similar to obtaining them from a group purchasing organization, with the exception that the CAP vendor billed CMS for the drug rather than the physician office. The goals of the program were two-fold: to remove perceived conflicts of interest of physicians in the buy-and-bill model, and to save money through a consolidation of large wholesale purchasers (ie, CAP vendors).

Fifteen vendors expressed interest in participating when CMS set proposed rules to govern the program in 2004; however, ultimately only one vendor (BioScrip) signed a contract with CMS. The CAP began on July 1, 2005.

The drugs designated by CMS to be included in the CAP represented 169 of the > 500 agents covered by Part B. The selected drugs represented 85% of the allowed charges for physician Part B drugs. Specifically excluded were oral agents, intravenous immune globulins, controlled substances, clotting factors, leuproide, tissue, and agents with < $1,000,000 in allowed charges per year. Practices would procure and bill drugs not included in the program through the existing average sales price (ASP) model.

In 2008, BioScrip decided not to renew its contract with CMS, citing contractual terms that presented unacceptable short- and long-term profit risk. At the end of 2008, after 42 months, CMS suspended the CAP but reserved the option of future reinstatement. A CMS evaluation of the CAP noted that it had not saved CMS money. In fact, the CAP drug costs exceeded the 106% of the ASP model used to pay physician practices for buy and bill by 3.2% in the aggregate for 2006 and 2007 (http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Reports/downloads/CAPPartB_Final_2010.pdf). Some of the increase in cost was the result of a technical difference in the cost of drugs to the CAP vendor as opposed to the ASP model. Under the contractual terms with CMS, the CAP vendor's purchase price of drugs was adjusted by an inflation factor (ie, producer price index) that was not included in the ASP model. The result was that drug acquisition costs for the vendor were somewhat higher.

The CMS review assessed beneficiary and provider views of the program. A small beneficiary sample (most of whom were not patients with cancer) reflected a positive experience with the CAP vendor, with little change in billing or drug availability, except for having to return on another day to receive an intravenous agent, because it had to be shipped by the CAP vendor to the physician office. However, they did not think this was an impediment care (http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Reports/downloads/CAPPartB_Final_2010.pdf).

Physician satisfaction with the CAP was derived from a 2008 survey of 1,201 participation physicians who were compared with 1,200 nonparticipating physicians. Overall satisfaction of participating oncology physicians with the CAP was 68.8% (very satisfied, 14.3%; somewhat satisfied, 54.5%), although oncologists were less satisfied with the program than other specialists (http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Reports/downloads/CAPPartB_Final_2010.pdf). There was a high attrition rate, with 45% of participating practices in 2006 opting out for 2007 and 53% of participating practices in 2007 opting out for 2008. Smaller practices were more likely to continue in with the CAP after their first year, noting that it was less costly for them to acquire Medicare Part B drugs (http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Reports/downloads/CAPPartB_Final_2010.pdf). The most common reason for opting out was that the program did not fit the needs of the practice. Although practices generally did enjoy the lack of financial exposure with expensive drugs, the requirement to purchase all designated drugs through the vendor—including inexpensive generics—was perceived by many practices to increase the administrative burden of caring for patients. At the peak of the CAP, approximately 3,500 physicians in roughly 1,100 practices participated; however, oncologists made up < 10% of participating physicians (http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Reports/downloads/CAPPartB_Final_2010.pdf).

Table A1.

Drug Pricing Benchmarks

Benchmark	Purpose	Methodology	Flaws or Caveats	Suitability to Estimate Provider Acquisition Costs
AWP	Manufacturer's suggested retail price; historically used by many payers as basis for payment	Set by pharmaceutical company	Various publishers have or will stop publishing AWP	Bears little relationship to any pricing actually appearing in market, particularly in generics
ASP	Created as Medicare Part B reimbursement replacement for AWP	Reported by pharmaceutical company; actual price manufacturer is paid to include rebates, discounts, allowances, or other price concessions; based on all classes of trade but excludes government, 340B, Part D, and state pharmaceutical assistance sales; published with 6-month lag	Does not cover drugs not covered by Medicare Part B	Prompt pay discounts to distributors are not excluded in calculation; 6-month lag can create over- or underestimation of drug price when pricing fluctuates; volume or other preferential discounts can disadvantage other providers
WAC	Report of actual price distributors pay to manufacturers	Reported by manufacturer; does not include 340B or government sales or any rebates, discounts, allowances, or other price concessions offered by manufacturer	Policymakers have been reluctant to endorse WAC for fear it would be subject to manipulation through large rebating and discounting by manufacturers	For branded drugs, it is fairly close to hospital or physician costs, although prices to large purchasers can be well below WAC if distributors cede material share of service fees to these customers; for generic drugs, WAC may substantially overestimate acquisition costs, resulting in significant markup
AMP	Price reported to Medicaid by manufacturers for purposes of collecting statutory rebates	Redefined by ACA; based on sales to retail pharmacies, it excludes sales to hospitals, HMOs, PBMs, mail-order pharmacies, and physicians; prompt pay discounts are excluded, but pharmacy discounts are included; multiple source drugs are reported as single weighted average	Many Part B drugs cannot be calculated by this methodology, whereupon calculation reverts to methodology similar to ASP	Although it does exclude prompt pay discounts, it specifically excludes providers in its methodology and is unlikely to be any more accurate than ASP
WAMP	Defines drug price, independent of manufacturer input, through market surveillance	Has been variable; last collection was conducted by OIG; previous collections focused only on Part B drugs; it has not been collected in several years	Methodology, if done robustly, could be credible with CMS	There is no practical way to use WAMP without up-to-date and broader source of price surveillance
RSP/NADAC	Developed in 2013 to provide CMS and states alternative to AWP for pharmacy reimbursement	Contracted survey based on invoice prices to pharmacies; rebate information to be captured broadly and separately, but currently no mechanism exists to adjust more detailed survey	Still work in process; significant logistic hurdles exist	Although methodology is intriguing, it would have to be applied to providers to estimate their costs
AAC	NADAC-like survey of providers proposed in ASCO Payment Reform Workgroup discussions	Survey of invoice prices to physician and hospital providers of oncology care; 340B institutions and government contracts such as VA medical centers would be excluded; tiering, based on size and type of institution, could be done to improve accuracy	Purely theoretic at this juncture; if cost of survey was born by practices, it could prove insurmountable; tiering may help smaller practices, but it would still be subject to vagaries of margins inherent in buy-and-bill chemotherapy	Would have to take account of rebates, making calculation complicated; anticipate that like ASP, it would have significant time lag

NOTE. Data adapted (www.cms.gov/Medicare/Medicare-Fee-for-Service-Part-B-Drugs/McrPartBDrugAvgSalesPrice/index.html?redirect=/mcrpartbdrugavgsalesprice; Levinson DR: Washington, DC, Department of Health and Human Services, 2005; Wright S: Washington, DC, Department of Health and Human Services, 2012; http://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Benefits/Prescription-Drugs/FUL-NADAC-Downloads/NADACMethodology.pdf).

Abbreviations: AAC, average acquisition cost; ACA, Affordable Care Act; AMP, average manufacturer price; ASCO, American Society of Clinical Oncology; ASP, average sales price; AWP, average wholesale price; CMS, Centers for Medicare and Medicaid Services; HMO, health maintenance organization; NADAC, national average drug acquisition cost; OIG, Office of the Inspector General; PBM, pharmacy benefit manager; RSP, retail survey price; WAC, wholesale acquisition cost; WAMP, widely available market price.

Table A2.

Selected Countries Drug Purchasing and Access Policies

Country	Price Negotiation Responsibility	National Formulary?	Pharmaceutical Sales per Capita in 2005 ($)	No. of Drugs Available in 2005	No. of New Molecules (launch date post 1996) in 2005	No. of Biologic Molecules Available for Clinical Use in 2005	No. of New Biologic Molecules Considered Antineoplastics (launch date post 1996) in 2005
United States	Complex, depends on patient insurance and provider typei	No	921	1,872	301	117	18
Canada	Patented Medicines Prices Review Board sets ceiling price	Yes; provincial	481	1,341	173	74	9
United Kingdom	National Health Service limits the profits that a drug manufacturer can earn on drug sales	Yes; set by National Health Service	400	1,678	236		12
France	Government committee sets the prices at which drugs must be purchased	Yes	658	2,344	227	102	15
Germany	Government agency reference prices drugs based on others in a therapeutic class	Yes	518	2,581	282	104	13
Italy	Government committee sets the prices at which drugs must be purchased	Yes	432	1,811	227	98	12
Spain	Government committee sets the prices at which drugs must be purchased	Yes	447	1,840	231	101	12

NOTE. Data adapted (www.cms.gov/Medicare/Medicare-Fee-for-Service-Part-B-Drugs/McrPartBDrugAvgSalesPrice/index.html?redirect=/mcrpartbdrugavgsalesprice; Levinson DR: Washington, DC, Department of Health and Human Services, 2005; Wright S: Washington, DC, Department of Health and Human Services, 2012; http://www.medicaid.gov/Medicaid-CHIP-Program-Information/By-Topics/Benefits/Prescription-Drugs/FUL-NADAC-Downloads/NADACMethodology.pdf).

Authors' Disclosures of Potential Conflicts of Interest

Disclosures provided by the authors are available with this article at jop.ascopubs.org.

Author Contributions

Conception and design: Blase Polite, Jeffery C. Ward, John V. Cox, Roscoe F. Morton, John Hennessy, Rena M. Conti

Administrative support: John V. Cox

Collection and assembly of data: Jeffery C. Ward, John V. Cox, Roscoe F. Morton, John Hennessy, Rena M. Conti

Data analysis and interpretation: Jeffery C. Ward, John V. Cox, Roscoe F. Morton, John Hennessy, Ray Page, Rena M. Conti

Manuscript writing: All authors

Final approval of manuscript: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jop.ascopubs.org/site/misc/ifc.xhtml.

Blase N. Polite

Honoraria: Sirtex Medical

Speakers' Bureau: Bayer Healthcare Pharmaceuticals/Onyx Pharmaceuticals

Research Funding: Merck

Other Relationship: Gerson Lehrman Group

Jeffery C. Ward

Honoraria: Genentech, Prometheus, Bayer Healthcare Pharmaceuticals, Celgene

John V. Cox

Employment: Texas Oncology

Leadership: Texas Oncology

Stock or Other Ownership: Amgen, MedFusion Labs Dallas, Merck, Pfizer

Research Funding: US Oncology Research

Other Relationship: Mary Crowley Research Center, Dallas, TX, American Society of Clinical Oncology, Methodist Health System, Dallas, TX

Roscoe F. Morton

No relationship to disclose

John Hennessy

Employment: HCA Healthcare

Stock or Other Ownership: HCA Healthcare

Ray D. Page

Stock or Other Ownership: Oncology Metrics

Consulting or Advisory Role: International Oncology Network, via oncology

Speakers' Bureau: Biodesix, Celgene

Research Funding: Gilead Sciences, Celgene, Bristol-Myers Squibb, Genentech/Roche, Pfizer, Novartis

Travel, Accommodations, Expenses: raintree

Rena M. Conti

No relationship to disclose