Determining the Cost of Pharmaceuticals for a Cost-Effectiveness Analysis
There is no published schedule of reimbursements for individual drugs. The analyst can estimate cost by adjusting the costs in the Federal Supply Schedule (FSS), the VA cost schedule, or the Average Wholesale Price (AWP) given by proprietary sources of pharmaceutical prices.
To find a cost of brand name prescription medications that represents the usual cost in the U.S. health care system, we recommend using 121% of the drug costs reported in the FSS, 152% of the VA cost, or 64% of AWP. To find the cost of a generic label prescription drug, we recommend using 27% of AWP.
Medicare publishes a reimbursement schedule for medications that are ordinarily administered in a hospital clinic or physician's office, which are reimbursable under Part B. This schedule includes payment for both the drug and its administration. Not all drugs administered in clinic have a reimbursement listed in the Medicare schedule, however.
Pharmaceutical manufacturers give rebates to health care sponsors such as insurance plans or government programs, such as Medicaid. These rebates refund to the sponsor some of the cost that they have paid. The amount of rebates is not disclosed. A study of rebates to Medicaid by the Congressional Budget Office found that rebate represented a significant reduction in the cost of medications. Because the amount of rebates are large but unknown, it is difficult to know the true cost that health sponsors pay for the medications received by their members.
Available sources of information include the Federal Supply Schedule (FSS), the Big 4 federal supply schedule, and proprietary sources of drug prices.
The Federal Supply Schedule (FSS) is a publicly available source of information on the relative cost of prescription medications. It is available via the VA National Acquisition Center; it is no longer available at the web site of the VA pharmacy benefits program. Pharmaceuticals may be searched by name or the VA drug class at the following link: https://www.va.gov/nac/Pharma/List. The Second Panel on Cost-Effectiveness in Health and Medicine recommend using the FSS for acquisition (transaction) costs for pharmacueticals (Basu (2017)).
The FSS price is the cost to most Federal agencies. The price to the “Big 4 Agencies” is not always disclosed. These agencies (VA, Department of Defense, Public Health Service, and the U.S. Coast Guard) have negotiated discounts on many drugs that results in a cost that is below that paid by other Federal agencies. Approved VA research studies may use a national dataset of prescription costs in the Management Cost Account (formerly known as DSS) national data extracts. These data includes the cost of the medication itself, as well as the cost of dispensing.
Subscription to databases on the cost of different prescription medications are sold by different vendors, including the following:
These vendors have traditionally reported the Average Wholesale Price (AWP) for each drug in the database. These databases include alternative measures such as Wholesale Acquisition Cost, Average Acquisition Cost, or Suggested Wholesale Price.
AWP may not be an objective statement of the relative costs of pharmaceuticals. In 2009, a lawsuit alleging that Average Wholesale Price had been manipulated to increase reimbursement for certain pharmaceuticals ended in a settlement in which one publisher agreed to stop publishing its AWP estimate (Academy of Managed Care Pharmacy (2009)). Others databases report AWP for only select drugs.
Access to these databases is by subscription; the cost may be prohibitive for academic researchers. Moreover, the annual paperback volume called Pharmacy Redbook is no longer published and access is limited through a subscription with the publisher.
All of these sources need to be adjusted to reflect the actual costs of health care sponsors. FSS and Big 4 costs reflect discounts received by the federal government not available to other sponsors and need to be adjusted upwards. Published AWP exceeds actual cost of sponsors and need to be adjusted downward.
The largest sponsor of prescription drug benefits in the U.S. is Medicare. The prescription drug benefit, enacted in 2003, is also known as Medicare Part D. The amount plans pay pharmaceutical companies is not disclosed. Medicare Part D costs were compared to Medicaid costs by the Department of Health and Human Services Inspector General. Reimbursements for brand name drugs were similar under both programs, but Medicaid paid somewhat higher reimbursement for multi-source drugs than did Medicare Part D (U.S. Government Accountability Office (2009)).
We use information on Medicaid as the next best source of pharmacy costs in the U.S. health care system. We offer the following reasoning. Medicaid and Medicare have been found to have similar costs for brand name prescription medications. Medicare is often the leader in determining U.S. reimbursement amounts. Finally, Medicaid and Medicare account for approximately half of U.S. health expenditures.
Data from historical studies suggest factors than can be used to estimate pharmaceutical costs relative to the FSS. For brand name medications, a study by the Congressional Budget Office found that the FSS averaged about 53% of AWP, and that the costs to Medicaid for brand name drugs was an average of 64% of AWP (U.S. Congression Budget Office (2005)). This implies that Medicaid costs are 121% of the costs in FSS (64%/53%).
VHA (and other Big 4 agencies) pay less for medications than other U.S. health care sponsors (4). For brand name medications, VA costs average about 42% of AWP (U.S. Congression Budget Office (2005)). To convert VA drug costs to a value that is more typical of the U.S. health care system, the VA costs should be multiplied by 152%, the amount by which Medicaid costs for brand name medications exceed VA costs (3).
Some cost-effectiveness studies have used unadjusted AWP as the cost of pharmaceuticals. It is clear that this is an overestimate of the cost, as AWP substantially exceeds the cost paid by health care sponsors (Gencarelli (2002), Gencarelli (2005)).
The U.S. Congressional Budget Office (CBO) has estimated the relationship between Medicaid costs and the AWP amounts for brand name medications. This work is unique, because the analysts from CBO had access to information on the rebates paid to Medicaid, information that is not ordinarily disclosed.
The cost of brand name medications to the U.S. Medicaid program was estimated to be 64% of the Average Wholesale Price (U.S. Congression Budget Office (2005)). This estimate includes the rebates and discounts received by the Medicaid program and the cost the program paid to pharmacies for dispensing medication. CBO found that for brand name drugs, Medicaid payments to pharmacies were an average of 92% of AWP, but Medicaid received rebates from manufacturers that were equivalent to 28% of AWP, resulting in a net final cost to Medicaid of 64%.
There is no similar estimate for generic medications, but available data allow us to estimate that the average cost of generic medications to the U.S. Medicaid program is 26.7% of the Average Wholesale Price. We base this estimate on the U.S Department of Health & Human Services Inspector General finding that the Average Manufacturer Price (AMP) is 30% of the AWP (Levinson (2005)). Centers for Medicare & Medicaid Services (CMS) receives rebates on generic drugs that is 11% of AMP (Bruen (2010)). If state costs are the AMP, then the net cost to the Medicaid program is 89% (100%-11%) of AMP. In terms of AWP, the cost to CMS is 89% of 30% AWP, or 26.7% of AWP.
Although this amount seems low relative to the AWP, it is consistent with the declining cost of generic medications relative to AWP. Medicaid paid pharmacies 58% of AWP in 1994 and 44% of AWP in 1999 (Rehnquist (2002)). We are unaware of any more recent estimates.
A task force was convened by the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) to consider how to assign the cost of pharmaceuticals for cost-effectiveness analysis. Findings from this effort were published in 2010 in a special supplement to the journal Value in Health (Vol 13 No. 1). The task force argued that marginal cost of manufacture and distribution should be used to estimate the cost of pharmaceutical interventions for cost-effectiveness analysis. The papers in the supplement acknowledged the difficulty in estimating marginal cost and recommended analysts use the actual costs paid by particular sponsors instead. Unfortunately, there was no specific guidance provided on how to do this.
The ISPOR group held that marginal cost should be used because development costs are not relevant to the cost-effectiveness analysis since they have already been incurred. They regarded the mark-up of prices over marginal costs as a "monopoly rent," which is not a social cost but a transfer between sectors of society.
Others argue that the marginal cost approach is not consistent with the recommendations that cost-effectiveness analysis use long-run marginal cost, including the cost of product development. Garber (2000) articulates why development cost must be included. According to this view, the mark-up of prices over marginal cost is not a transfer payment; rather, it is a temporary government monopoly granted to developers so that they may recoup their development costs. The mark-up may not exactly represent amortized development cost, but it is the only readily available estimate. The size of the mark-up is disciplined by long-term competition; if it is set too high above the amortized development cost, new competitors will enter the product category. Adopting the societal perspective in a cost-effectiveness analysis requires inclusion of the cost of innovation. Ignoring this cost will bias cost-effectiveness analysis in favor of interventions that can substitute development costs for operating costs.
Analyses of the cost-effectiveness of drugs that are nearing the end of their patent period will want to consider the effect of patent expiration on cost. Hoyle (2011) proposes that cost-effectiveness analyses of medications include the expected effect of patent expiration, which will reduce drug cost in the future. He argues that failure to consider this effect biases cost-effectiveness analysis against pharmaceutical interventions, and in favor of interventions that do not include this effect.
Academy of Managed Care Pharmacy (2009) AMCP Guide to Pharmaceutical Payment Methods, 2009 Update (Version 2.0), J Manag Care Pharm, 15, S3-57, quiz S58-61.
Basu A. (2017) Estimating Costs and Valuations of Non-Health Benefits in Cost-Effectiveness Analysis. In: Neuman PJ, Sanders GD, Russel LB, Siegel JE, Ganiats TG, editors. Cost-Effectiveness in Health and Medicine, Second Edition. Oxford University Press: New York. 2017. pp.201-235.
Bruen, B. K. (2000) Medicaid and Prescription Drugs: an Overview (Washington, D.C., The Kaiser Commission on Medicaid and the Uninsured).
Garber, A. M. (2000) Advances in cost-effectiveness analysis, in: Culyer, A. J. & Newhouse, J. P. (Eds.) Handbook of Health Economics (Amsterdam North-Holland).
Gencarelli, D. M. (2002) Average wholesale price for prescription drugs: is there a more appropriate pricing mechanism?, NHPF Issue Brief, 1-19.
Gencarelli, D. M. (2005) One pill, many prices: variation in prescription drug prices in selected government programs, NHPF Issue Brief, 1-20.
Hoyle, M. (2011) Accounting for the drug life cycle and future drug prices in cost-effectiveness analysis, Pharmacoeconomics, 29, 1-15.
Levinson, D. R. (2005) Medicaid Drug Price Comparisons: Average Manufacturer Price to Published Prices Department of Health and Human Services, Office of Inspector General.
Rehnquist, J. (2002) Medicaid Pharmacy-Actual Acquisition Cost of Generic Prescription Drug Products Department of Health and Human Services, Office of Inspector General.
Render, M. L., Nowak, J., Hammond, E. K. & Roselle, G. (2003) Methods for estimating and comparing VA outpatient drug benefits with the private sector, Med Care, 41, II61-9.
United States Department of Health and Human Services. Office of Inspector General (2009) Comparing pharmacy reimbursement: Medicare part D to Medicaid.
United States Congressional Budget Office (2005) Prices for Brand Name Drugs Under Selected Federal Programs (Washington, D.C., Congress of the United States).