Marianne Messer, Elsevier's Gold Standard

Fig. 1. PAC vs. AWP for Tracking Actual Acquisition Cost

		In the mid-1990s, Gold Standard, a drug information publisher, launched Clinical Pharmacology to fill a need in pharmacies for reliable and affordable drug reference information, easy to use at the point of care. Over the next few years, we worked closely with many retail pharmacies, from the largest chains to small independents, PBMs, and health insurance companies. More and more, our customers were expressing the need for non-clinical drug information, especially price information, to support their business operations. This is not to say that price information was not available from other sources—it was&mdash;but many of our customers were not satisfied with the price information they had access to. Their concerns included timeliness and accuracy, but the word they used most to describe what they wanted was transparency.

		This seemed reasonable enough. People who are in the business of supplying a product need to know what their costs will be in order to determine a profitable price. We quickly determined that we could provide our customers with more timely and accurate prices from our partner manufacturers, but transparency was a taller order. As we all know, the drug supply chain is complex, and price obscurity across the continuum has been an accepted, if not entirely desirable, state of affairs for decades. I always say the first question to ask when considering a change is: How does it work now and why?

		Thanks to an astonishingly productive period in the research and development of pharmaceuticals during the second half of the 20th century, drugs became an essential component of medical treatment and a major health insurance benefit. During the same period, computer technology advanced even more rapidly than pharmaceuticals and drug compendia became available, allowing for greater automation and electronic processing of prescriptions and claims adjudication. Third-party payers and prescription benefits managers emerged to assist health insurance companies, adding another layer to the supply chain. The sheer volume of transactions combined with the number of players involved with getting drugs to consumers required agreement on what drug price to use for each drug in negotiating contracts and adjudicating claims.

		Almost by default, the industry settled on the average wholesale price (AWP), developed by George Pennebaker to adjudicate claims and establish proper drug reimbursement for the California Medicaid Drug Program (called Medi-Cal) in the 1970s. The price was originally based on survey data from pharmaceutical wholesalers about what they charged pharmacies for their products. However, in the absence of reported data, First Databank (FDB) and other compendia calculated an AWP price using 1.20 or 1.25 multiples of the WAC (wholesale acquisition cost) price.

		It has long been commonly agreed that AWP is an imperfect standard, varying from true acquisition cost by a wide range, and jokingly referred to as “ain’t what’s paid” since the 1980s. Nevertheless, AWP has been the accepted starting point to determine reimbursements and contract purchase price. For example, a contract might state that payment will be the usual and customary, but no more than AWP minus 15%, or the estimated acquisition cost, but no more than AWP minus 10%.

		Criticism of AWP was not confined to its failure to represent a true average. More strenuous complaints have been directed at how easily AWP can be manipulated. This assertion ultimately led to a Medicare Part B fraud investigation and a lawsuit against its original publishers.

		Medicare Part B primarily covers physician-administered drugs. The investigation focused on whether AWP was being used to bill Medicare for more than the actual cost of drugs. AWP was being used to calculate claims, and the charge was made that the price was being artificially manipulated to ensure higher reimbursements. In 2005, the Prescription Access Litigation (PAL) project claimed there was a conspiracy to increase AWP. As a result, in 2011, FDB ceased publishing AWP in its drug price data.

		Because FDB is commonly used for price data, the looming, and now actual, absence of AWP from its price file sparked debate about a new drug price benchmark to replace it. As imperfect as AWP is, it is still the standard for many contracts and claims.

		Industry leaders and advocates have proposed various alternatives to AWP and, while no consensus has yet emerged, there is general agreement on the criteria that a new standard should meet. It should be:

			Transparent, easily understood, unambiguous, and reflect the true acquisition cost of a drug.

			Accessible, administratively simple and efficient, so that it can be easily adopted by the industry.

			Comprehensive, and offer drug pricing information for all branded and generic products.

			Trustworthy, timely, and updated frequently enough to reflect quickly changing acquisition costs, particularly for generic products.

			Immune to manipulation, auditable, not anti-competitive, stable, and not result in more litigation.

			Administratively simple and practical.

		A review of the drug price types available today shows that none meet the desired criteria for a new standard:

			Actual Acquisition Cost (AAC) is the final price that a pharmacy pays after all discounts have been subtracted. Were it accessible, this would be the ideal standard.

			Average Manufacturer Price (AMP) is the price that manufacturers report to the Medicaid drug rebate program. It is only reported monthly and quarterly, and even the monthly information is three months old.

			Average Sales Price (ASP) is the calculated price for Medicare Part B drugs. The use of ASP is problematic because its accuracy is questioned. In addition, it’s not available for all drugs, and not specific to NDC.

			Estimated Acquisition Cost (EAC) is either the estimated cost of a product or a pharmacy’s usual and customary charge.

			Federal Upper Limit (FUL) is a calculation by CMS (Centers for Medicare and Medicaid Services) for the amount that will be paid in aggregate on multisource drugs. No FULs have been published since September 2009.

			Maximum Allowable Cost (MAC) is defined by each payer/state for its multisourced drugs only.

			Manufacturer List Price (MLP) is the price listed by the drug company.

			Wholesale Acquisition Cost (WAC) is the only price type defined in regulations. It is the list price from a manufacturer to a wholesaler or a direct purchaser without discounts.

		In recent months, two new drug price options, each taking quite different approaches, have emerged as potential new standards for drug acquisition cost:

			National Average Drug Acquisition Cost (NADAC) is based on a voluntary CMS survey of 2,500 pharmacies requesting invoice data to gather their acquisition costs and compute a public average acquisition cost (AAC). The level of participation and transparency from pharmacies remains to be seen. **

			Predictive Acquisition Cost (PAC) combines a predictive analytics model with multiple factors associated with the cost of a drug, including industry MAC benchmarks, published price lists, existing price benchmarks, supply/demand measures and survey-based acquisition costs. The statistical model is trained to synthesize various known attributes into an overall estimation of acquisition cost.

		PAC, developed by Glass Box Analytics, a Springfield, VA, consulting company, is currently available, and in the interest of full disclosure, Elsevier’s Gold Standard is its exclusive publisher at this time. PAC leverages proven concepts from other industries and applies the power of predictive analytics to drug pricing. By using various factors associated with the cost of a drug, it deploys a multidimensional predictive analytics model to track the acquisition cost of drugs with sufficient accuracy to support pricing activity. The statistical model is trained to synthesize various known attributes into an overall estimation of acquisition cost.

		As a publisher, our role isn’t to tell our customers what price type they should use in their business transactions. Our job is to make available to them whatever price or prices they might wish to use and to be transparent in how those prices were acquired. In the case of AWP, we do this by publishing three versions of the price—reported, WAC x 1.20 and WAC x 1.25. This makes it clear where the price type comes from, but unfortunately, it does not help our customers know how close AWP is to the true drug acquisition cost.

		Glass Box Analytics has made a good case in showing that PAC more closely tracks true drug acquisition cost as historically compared with AWP, as shown in the generics example, using generic drugs, below. Of course, proving that PAC is a better price type than AWP does not necessarily mean that it is the best option for a new standard. PAC estimates or predicts drug acquisition cost using sophisticated analytics, in much the same way that FICA scores estimate or predict an individual’s credit worthiness. While we believe that the model is very strong and produces accurate results, it is different than actual reported payment data.

Articles

Pricing systems in US drug supply chain have many limitations; an analytics-based method might avoid some of them