This expert analysis and rich historical perspective on AMP offers valuable insights into the new AMP Proposed Rule and how it affects specialty pharmacy.
This expert analysis and rich historical perspective on AMP offers valuable insights into the new AMP Proposed Rule and how it affects specialty pharmacy.
The deadline to submit comments on the proposed rule recently published by the Centers for Medicare & Medicaid Services (CMS) in the Federal Register has come and gone. This proposed rule revises requirements pertaining to Medicaid reimbursement for covered outpatient drugs to implement provisions of the Affordable Care Act of 2010, as well as other requirements related to covered outpatient drugs including “key aspects of Medicaid coverage, payment, and the drug rebate program.” One of those “key aspects” is the calculation methodology used by manufacturers to determine Average Manufacturer Price (AMP). This marks the third time CMS has proposed regulations for the Medicaid Drug Rebate Program since it was established by the 1990 Omnibus Budget Reconciliation Act (OBRA 90). What once seemed a relatively simple and esoteric quarterly calculation to determine an element of the Medicaid Unit Rebate Amount (URA) has transpired into a hotly contested series of complex monthly and quarterly calculations for manufacturers. Now, there are TWO distinct calculations (determined by the drug’s route of administration) performed monthly and quarterly that involve ratios, calculation of a weighted average of lagged pricing concessions for a rolling 12-month period, and implementation of elaborate data filtering processes that consider customer types and transaction types, among other things.
The enhanced awareness of the AMP calculation seems driven by the Final Rule published by CMS in July 2007 that implemented provisions of the Deficit Reduction Act of 2005 pertaining to prescription drugs under the Medicaid Program. The Final Rule, which became widely known as “The AMP Rule,” not only established the methodology for manufacturers to calculate AMP, but also established AMP as the basis for the Federal Upper Limit (FUL). This provision, along with the fact that AMP would be publicly available, suggested to many that AMP might also be used for pharmacy reimbursement on a broader scale. Now, it seemed, almost everyone involved in any aspect of the delivery of pharmaceutical care—manufacturers, payers, pharmacy providers, and prescribers—became an “expert” on AMP. Since the publication of the 2007 AMP Rule, much has been written, discussed, and debated on what AMP is and what it is not. It is not the intent of this article to further debate the issues or revisit the discussions that occurred over the past few years. After all, the Affordable Care Act of 2010 superseded most, if not all, of the 2007 AMP Rule, leading to the withdrawal of the corresponding AMP regulations by CMS in late 2010. Instead, this article hopes to provide a historical perspective on how issues surrounding the AMP calculation evolved into the situation that confronts manufacturers and the practice of pharmacy today.
While in pharmacy school, a question my father, a civil servant and federal employee, often asked was, “Just how do pharmacists decide what to charge for prescriptions?” I sidestepped his question by saying that topic wasn’t something that was specifically addressed by the pharmacy school curriculum. When I began working in the community pharmacy setting, first as a student, then as a pharmacist, I felt a greater obligation to answer the question he was still asking the best I could. I explained there were 2 pricing references used in the pharmaceutical supply chain: WAC (wholesale acquisition cost), which was understood to be the price manufacturers sold their products to wholesalers, and AWP (average wholesale price), which was understood to be a price wholesalers used to establish the selling price for products sold to their retail pharmacy and hospital customers. I also explained to my dad that pricing of pharmaceuticals is similar to the pricing of new cars: the AWP is analogous to the window sticker price, and WAC is analogous to the dealer’s invoice price. So, much like the price actually paid by the customer for a new car is usually somewhere between the dealer’s invoice price and the window sticker price, the price paid for pharmaceuticals by pharmacies is usually somewhere between WAC and AWP. The credit worthiness of the wholesaler’s customer, the pharmacy’s purchase volume from a particular wholesaler, and the number of wholesalers serving a particular market are also factors that determined the price actually paid by the pharmacy. I also explained that the pharmacy may determine the price for a prescription based on what it considered to be its cost of the product (usually AWP), plus a dispensing fee or professional fee to compensate the pharmacist for his professional services, and related overhead.
As the number of third-party payers increased, this pricing model (drug cost plus dispensing fee) was used to determine the amount reimbursed to the pharmacy for the prescription, and the price determination for a prescription shifted away from the pharmacy to the third-party payers. While the reliance on the AWP as a reference point in determining the price (or reimbursement) continued, the fact that the actual cost of the drug was typically not AWP, but some discount off AWP, was increasingly recognized. This was further acknowledged by the increasing references to AWP as meaning “Ain’t What’s Paid” in a context meant to be humorous. Nevertheless, reimbursement to pharmacies from payers (ie, the “price” for the prescription) was determined as some discount off AWP on an increasing basis.
PRESCRIPTION DRUG PRICING
After leaving community pharmacy to pursue a career in the pharmaceutical industry, my father’s questions about prescription drug pricing continued, but the questions now included why prescription prices seemed to increase as they did, and how manufacturers determined the prices they charged to their customers. During the mid-1980s, other civil servants and federal employees— more specifically, members of Congress and Senators—began asking similar questions. Their questions included why the price for a prescription for the same product and quantity varied so much from one pharmacy to the next, why manufacturers were willing to give discounts on their products to certain group purchasing organizations and HMOs, and why the price paid for prescriptions by state Medicaid agencies always seemed to be among the highest of the prices surveyed. They also wanted to know why manufacturers were often willing to offer such steep discounts to the US Department of Veterans Affairs (VA) on the Federal Supply Schedule and through VA depots, while Medicaid was paying among the highest prices. After all, they reasoned, the federal government is the largest volume purchaser of prescription drugs, so it should get the lowest price—or at least a price as low as the manufacturers’ best price offered to other customers. The frequency and magnitude of price increases levied by manufacturers also did not go unnoticed by Congress. As if to provide the answers to these questions, Congress incorporated the Medicaid Drug Rebate Program (MDRP) into the budget (OBRA 90) to become effective January 1991.
Manufacturers had some input into how the MDRP was structured, including the pricing basis on which the mandatory price discounts were to be determined. It was suggested that any discounts provided to Medicaid through a rebate should not be based on prices for products that were already discounted to another customer. In other words, discounting of the same product twice was to be avoided. Furthermore, it was argued that the discount provided by Medicaid rebates should only reflect prices that manufacturers charged to their customers, ie, wholesalers and other direct purchasers. Thus, AWP and any price reflecting a discount off AWP were ruled out because those prices (ie, the prices paid by retail pharmacies) were determined by the wholesalers. WAC was also ruled out, because that was a price typically discounted further to wholesalers in return for a prompt payment. Prices for products sold direct to hospitals, HMOs, and the federal government were also excluded from the Medicaid rebate pricing basis, in part on the thinking that those prices were typically already discounted, and because prescriptions for Medicaid patients were not usually dispensed by these customers. The “direct” qualifier was thought to be included because some manufacturers did not have a chargeback system in place at the time. Therefore, any discounts to customers were reflected directly on the invoice from the manufacturer.
So, with AWP and WAC both excluded from consideration, a new pricing metric descriptor was necessary to determine the Medicaid rebate, and that price became AMP. The paraphrased definition of AMP in the original version of the statute is “the average price paid to the manufacturer for the drug in the States by wholesalers for drugs distributed to the retail pharmacy class of trade (excluding direct sales to hospitals and HMOs). Federal Supply Schedule prices are not included in the calculation of AMP.” This relatively simple and seemingly straightforward definition of AMP would soon be complicated by the fact that the term “retail pharmacy class of trade” was not defined in the statute or in the manufacturers’ Rebate Agreement. Since the definition of AMP explicitly excluded sales to certain customer types, an assumption many manu facturers adopted (the ability to make “reasonable assumptions” is one of the provisions in the Rebate Agreement) was to include all sales to wholesalers or other customer types to which the manufacturer sold that were not specifically excluded. The customary prompt pay allowance typically provided to wholesalers and other direct customers was also factored into the calculation, along with any discounts to retail pharmacies (either through net invoice prices or chargebacks paid through wholesalers).
The MDRP required manufacturers to provide states a rebate for products dispensed to Medicaid beneficiaries in order to have those products reimbursed by the states. The program was structured to ensure that the Medicaid program received the same discount as the manufacturer’s “best price,” or at least a minimum rebate percentage discount. The initial rebate in for innovator products was the greater of AMP minus Best Price (BP) or AMP X 12.5%. This rebate percentage later increased to 15%, then 15.7%, then 15.1%, and currently is 23.1%. An additional rebate is required by when the percent increase in AMP for the current quarter is greater than the corresponding percent increase in CPI-U compared with a baseline period. This ensures price increases that exceed the percent change in the CPI-U are not incurred by the Medicaid program. Initially, and because the VA reportedly testified before Congress on its success in negotiating discounts from pharmaceutical manufacturers, the MDRP subjected prices offered to the VA to BP consideration. This was done to ensure the Medicaid program would also benefit from those discounts. However, an unintended consequence of this policy was that some manufacturers canceled or renegotiated some of those prices offered to VA, presumably to avoid the Medicaid rebate liability that would result. Likewise, it was reported that many safety-net providers (ie, community health centers or other providers of pharmaceutical care to low-income citizens) saw their discounted prices disappear, presumably to avoid the Medicaid rebate liability associated with those prices. Thus, when the metaphorical line began forming to testify before Congress on who was hurt by the MDRP, ie, how the Medicaid Drug Rebate Program resulted in financial hardship to those involved in the delivery of pharmaceutical care, one would expect to see the VA and the safety-net providers at the front of the line. Such unintentional consequences no doubt contributed to the passage of the Veterans Healthcare Act of 1992. This legislation established the 340B Drug Pricing Program, and required manufacturers to sell to the VA, Department of Defense, the Public Health Service, and later, the Coast Guard (commonly referred to as “the Big 4”) at a price no higher than a ceiling price calculated as 24% of Annual Non- Federal Average Manufacturer Price (NFAMP). The Medicaid Rebate statute was also changed to exempt the VA and any other federal purchaser, along with the covered entities described in the 340B Drug Pricing Program, from BP consideration.
Although responsibility for regulatory oversight for the Medicaid Drug Rebate Program was assigned to CMS, proposed regulations were not published for comment until 1995. The regulations proposed in 1995 were not finalized, and comments submitted were never addressed by CMS, as they were reportedly lost during the process of CMS moving to its new (and current) location. No attempt was made to request resubmission of comments or to update the 1995 proposed rule. Instead, CMS continued to provide regulatory guidance to manufacturers and the states through written notifications commonly referred to as “Releases.” Manufacturers were instructed to document and consistently apply any “reasonable assumptions” about the AMP and BP calculations in the absence of specific regulatory guidance, or in situations where such guidance was not clear.
As the customer base to which manufacturers offered discounts continued to increase, so did the questions on how to treat those discounts for the calculation of AMP. Most questions seemed divided between what CMS considered to be “the retail pharmacy class of trade” and the treatment sales and discounts to pharmacy benefit managers (PBMs) in the AMP calculations. Absent specific regulatory guidance, many manufacturers relied on “reasonable assumptions” to determine the treatment of PBMs in their AMP calculations. Arguments were made for considering treatment of PBMs like HMOs (and hence, excluded from the AMP calculations) or, as retail, since the payment transactions conducted through PBMs occurred in the retail pharmacy setting (sales and discounts included in the AMP calculations). This is just one example of how the implementation of reasonable assumptions by manufacturers resulted in a calculation of AMP that was not consistent across all manufacturers.
In 1997, CMS published a chart in its Release 29 to manufacturers listing the customer types for inclusion or exclusion from the AMP calculation. Among the customer types described for inclusion (and therefore, considered to the “retail pharmacy class of trade”) were mail order pharmacies, nursing home primary/ contract pharmacy sales, and the outpatient pharmacy of a hospital for which the hospital bills Medicaid for reimbursement. CMS also provided clarification that to the extent they could be identified (ie, through chargeback transactions), both direct and indirect sales to customer types excluded from the calculation in the original definition of AMP (eg, hospitals, HMOs, FSS purchasers) could be excluded from the AMP calculation. CMS also advised manufacturers that all pricing adjustments affecting the price of any sales must be taken into account if the sales were included in the calculation of AMP and BP. Guidance from CMS on treatment of sales and discounts to PBMs was also provided in Release 29 with CMS stating, “We generally consider drug prices to PBMs as having no effect on the AMP calculations unless the PBM is acting as a wholesaler as defined in the rebate agreement.” CMS later clarified that a PBM acts as a wholesaler when it purchases drugs. Thus, if a PBM also owned a mail order component, sales and discounts to the mail order portion of the PBM are treated as “retail” and included in the AMP calculation. However, shortly thereafter in Release 30, CMS qualified its direction in Release 29 with the following statement: “This information was published to clarify, not change or further confuse the issue with respect to the inclusion or exclusion of sales to and through PBMs. CMS intended no change to the requirements of the drug rebate program by our latest release. We are currently reexamining the issue and hope to clarify our position in the near future.”
In 2002, based on concerns that manufacturers were not reporting BP correctly— and thus depriving the Medicaid program of untold sums of money— Representative Henry Waxman (D, California) and Senator Chuck Grassley (R, Iowa) requested that the Department of Health and Human Services Office of Inspector General (OIG) conduct an audit of several manufacturers. As the story goes, the OIG responded that they could not perform an audit because no regulations covering the Medicaid Drug Rebate Program existed, and therefore, the OIG could only “survey” the methodologies used by manufacturers. Apparently unsatisfied with this response, the General Accounting Office was enlisted to perform the audit. While the original focus of the audit was the BP reported by 10 manufacturers (one of which included my employer at the time), the GAO later expanded the scope to include AMP. When the GAO reviewed its preliminary findings with us, the investigators had several questions regarding the treatment of various transaction types in the AMP calculation compared with what they observed from other manufacturers. We explained to the GAO that these differences were likely due to the “reasonable assumptions” adopted by manufacturers due to statutory guidance that either did not exist or was found to be unclear by manufacturers. The investigators noted that our explanation was consistent with what was reported by other manufacturers. It seems that the GAO had also solicited feedback on AMP and BP from the OIG. In the report that was issued to the congressional requestors in February 2005, the GAO noted that “OIG officials told us that, despite the program releases by CMS, they remain unable to evaluate AMP because of the lack of clear guidance, particularly to the retail pharmacy class of trade…” The GAO report also stated in that “There was considerable variation in the methods that manufacturers used to determine best price and AMP.” The GAO concluded that the “current rebate program oversight does not ensure that manufacturer reported prices or price determination methods are consistent with program criteria specified in the rebate statute, rebate agreement, and CMS program memoranda.” In response to the GAO’s report, Senator Grassley issued a statement that said, “…it (the Medicaid Drug Rebate Program) is a program virtually without oversight. The GAO found that the OIG has issued only four audit reports on Drug Company reported prices since the inception of the program. According to the GAO, even when the OIG has managed to identify problems related to the drug companies’ reported pricing and methodologies, CMS has not done much of anything to resolve them.” Senator Grassley concluded his statement saying that “the GAO report confirms that neither CMS nor the OIG know the extent to which Medicaid overpays for prescription drugs because the program lacks effective management and oversight. A worse state of affairs is not likely. The Medicaid Drug Rebate Program is quite simply a mess—a Medicaid mess.” The Deficit Reduction Act of 2005 (DRA 2005) was signed into on February 8, 2006, and contained the following requirement: “Not later than July 1, 2007, the Secretary of Health and Human Services shall promulgate a regulation that clarifies the requirement for, and manner in which, average manufacturer prices are determined under section 1927 of the Social Security Act…” While I can’t be certain, I continue to believe that the information gleaned by the GAO from the manufacturers they investigated directly contributed to the statutory language that required the publication of the AMP Rule in July 2007.
Roughly coinciding with the GAO report were OIG reports on other areas related to the Medicaid program and pharmacy reimbursement, notably the perceived ineffectiveness of the Federal Upper Limit methodology used by CMS. These reports likely contributed to the FUL methodology that specified AMP (or more accurately, 175% of AMP) serve as the basis for the FUL determination also included in the DRA 2005. While not entirely clear what factors resulted in the AMP to become the pricing basis for the FUL, it is probably safe to assume they included the continued criticism levied at AWP in that it was not an average of anything, was not a price paid by wholesalers, and increasingly acknowledged as a price that “Aint What’s Paid” by any purchaser. So, that acknowledgment begs the question of just what IS the price that’s paid? One needs only to review the original definition of AMP paraphrased earlier in this article to conclude that AMP was probably thought to be a price that reflects what is paid more than any other defined pricing element, even though the original (OBRA 90) intent of AMP was in the computation of the Medicaid URA.
A PERFECT STORM
With the publication of the AMP Rule, there seemed to exist a “perfect storm” of issues that established a previously little known pricing factor originating with OBRA 90 (which by the way, is the same legislation that apparently requires the cashier who rings your prescriptions up to ask if you have any questions for the pharmacist). When retail community pharmacists and their advocates studied the 2007 AMP Rule to learn more about just what this AMP thing was, and more importantly, how it was going to partly determine the FUL, and possibly be a new standard for reimbursement in the future, it’s easy to imagine their dismay when they read all the customer types that CMS considered “the retail pharmacy class of trade.” Mail order pharmacies, home health care, outpatient hospital pharmacies, surgical centers, ambulatory care centers, and physicians all were considered to be a part of the retail pharmacy class of trade, and therefore, their sales and discounts are included in the AMP calculation. Furthermore, the calculation used a “presumed inclusion” methodology, meaning unless sales and discounts to an entity could be identified as specifically excluded from the AMP calculation, they were to be included. Suddenly, much like AWP, AMP had a second meaning: “Ain’t My Price.” This, of course, implied that the price that was about to become the basis of reimbursement for many multiple source products to retail community pharmacies seemed destined to be determined by a calculation that used sales and discounts to customers that arguably were NOT retail community pharmacies, and such calculations reflected discounts that were thought to be generally not available to retail community pharmacies. The concerns about the AMP calculation defined in the AMP Final Rule ultimately led to a lawsuit against CMS by the National Association of Chain Drug Stores and the National Community Pharmacists Association. This lawsuit asserted (among other things) that “The AMP rule includes a significant number of transactions that do not belong in the calculation of AMP because they are not prices paid to manufacturers by wholesalers for drugs distributed to the retail pharmacy class of trade” and “FULs calculated using these ‘flawed AMPs’ result in reimbursement for generic drugs that will be cut by a staggering 78.7%, and well below the prices retail pharmacies pay for drugs.”
We know that CMS was prohibited from implementing the FUL methodology described in the 2007 AMP Rule. In addition, the approach used in the Affordable Care Act of 2010 to address concerns about the definition and calculation of AMP was to change the original statute, not merely issue new or revise existing regulations. As a result, the AMP methodology in the 2007 AMP Final rule was withdrawn by CMS because those regulations were superseded by the statute. Thus, the wait for a new proposed rule from CMS began. Nevertheless, this once relatively simple calculation has become increasingly complex. For example, the statute resulted in the implementation of a new and second AMP calculation methodology for those drugs that are infused, instilled, injected, implanted, or inhaled, and the term “5i drugs” is now part of the established government pricing jargon.
When the most recent proposed rule was published for comment by CMS, there appeared to be alignment with industry and pharmacy on the new statutory definition of “retail community pharmacy.” What remained a point of contention was the calculation methodology. That is, should AMP be calculated using a “presumed inclusion “methodology as before or should the calculation include sales and discounts ONLY for those entities that could be definitively identified as retail community pharmacies by manufacturers? Not unexpectedly, there are compelling arguments and strong opinions for both. In fact, in the proposed rule, CMS favored the methodology that included only sales and discounts for entities identifiable as retail community pharmacies. After reviewing some of the comments submitted in response to the proposed rule, it appears that the manufacturers and retail pharmacy groups are aligned on this issue as well, in that a “presumed inclusion” methodology appears to be the more feasible approach that in that it will result in an calculated AMP that is both verifiable and does not appear to adversely impact reimbursement based on a FUL determined using those AMPs. Meanwhile, the wait for the new final rule continues. As for me, I continue to keep my fingers crossed that my father, who is still going strong at 88 years old, will not ask me to explain what AMP is.
Jerry Wolf, RPh, is senior director, government pricing compliance consulting at Compliance Implementation Services. Mr. Wolf has extensive experience with pharmaceutical government pricing program requirements and assessment of risk and compliance associated with the Medicaid Drug Rebate Program and 340B Drug Pricing Program. He has also advised pharmaceutical manufacturers on pricing and contracting strategies, using supplemental rebates and targeting the Medicaid fee-forservice patient population. As a registered pharmacist, Mr. Wolf’s background includes more than 25 years in the pharmaceutical and managed health care industries.