Of the estimated 6.5 million people living in developing countries who currently need treatment with antiretroviral medications, almost 1 000 000 (17%) were estimated to be on antiretroviral therapy as of June 2005 . This achievement, which represents more than a doubling of the numbers of people on treatment over the past 18 months, has been possible through commitment by governments and close international collaboration. However, the financing of drugs remains an area of particular concern, often limiting the quantities procured and consequently the program coverage. This consideration is fundamental for governments and institutions that provide antiretroviral drugs with their own resources and it is equally critical for donors who finance HIV/AIDS programs on a large scale.
The availability of generic forms of antiretroviral drugs has contributed to a significant decrease in their price over the past 5 years. Several multinational drug companies that make originator products have also lowered prices and claim to be selling antiretroviral drugs at ‘no-profit’ [2,3] or ‘at a financial loss’  in the developing world. To help purchasers make better informed procurement decisions, a clearer idea of the actual cost structure of antiretroviral drug production is needed. Studies of the production costs of antiretroviral drug s have also been requested by members of the United States Senate . Previous efforts to analyze the costs of antiretroviral drug production have been based on estimates derived from the published financial reports of research-based pharmaceutical companies . This study fills the gap by presenting real product-specific cost structures of a state-owned generic drug manufacturer based in a developing country that produces finished formulations of antiretroviral drugs, but that does not produce the active pharmaceutical ingredient (API), which it is obligated to purchase on the international market. This will equip those responsible for antiretroviral drug procurement within HIV/AIDS treatment programs with useful data for understanding the price of pharmaceuticals.
Calculation of direct manufacturing costs
Data from 2001 (Table 1)  on the cost structure of antiretroviral drugs produced by the Brazilian state-owned pharmaceutical company Far Manguinhos – a division of the Oswaldo Cruz Foundation (Far/FIOCRUZ) based in Rio de Janeiro – were used to calculate a range of direct manufacturing costs of 19 commonly used antiretroviral drugs, including fixed-dose combinations (FDCs). In these calculations, the cost of API in 2001 was replaced with the lowest and highest API prices quoted in an April 2006 survey by the HIV/AIDS Department of the World Health Organization . Far/FIOCRUZ produced six of these products as of 2001. For those antiretroviral drugs under patent, and thus not produced for sale by Far/FIOCRUZ, estimates of the cost of production were based on data taken from pilot batches made specifically by Far/FIOCRUZ to estimate reference prices and to be used in negotiations between the Brazilian government and drug companies. These pilot batches followed standard guidelines and specifications on the development of new pharmaceutical forms for tablets, capsules, and FDCs, as appropriate [9–18].
Direct manufacturing costs include those expenses that are incurred in the manufacturing process and can be attributed to specific steps in the production chain. These costs were precisely determined using the information from the cost structures of Far/FIOCRUZ, specific to each antiretroviral product. For each production batch, which varied between 250 000 and 500 000 tablets, direct manufacturing costs were categorized and calculated as follows (Table 2).
Active pharmaceutical ingredients
The data from WHO on API prices was used. The cost of API per batch was calculated as follows, and includes the cost of the second or third API, in the case of FDCs.
Equation (Uncited)Image Tools
This line item is comprised of variable costs based on the actual working time of the machines and fixed costs based on downtime. Each of these components in turn includes both ownership costs (depreciation and interest) and operational costs (maintenance, lubricants, fuel, etc.). These costs are also dependent on batch size. A larger batch size requires a larger initial investment in equipment but offers a potential reduction in production costs because of the economies of scale that result from being able to manufacture larger quantities in each production run. To permit the calculation of fixed costs on a per tablet basis, a projected sales volume had to be assumed. This was set to be equal to meet the forecasted needs of the Ministry of Health of Brazil. Batch sizes were equal to those used by Far/FIOCRUZ in 2001.
Equation (Uncited)Image Tools
Labor costs are based on the annual salaries plus benefits (i.e. health insurance, national social security contribution) of the operators directly involved in the different phases of production, including quality control.
Equation (Uncited)Image Tools
These represent the inert ingredients that are included in the formulation of the finished forms which include capsules and tablets.
Equation (Uncited)Image Tools
These costs include those for primary packaging (blister packs, plastic flasks, and strips) and secondary packaging (boxes in which the drugs are shipped). The cost of labeling and informational inserts that are included for the prescriber and patient are also included in this category.
Equation (Uncited)Image Tools
Typical loss rates associated with the manufacturing process of transforming API into the finished formulation ranged between 1–3% at Far/FIOCRUZ. Unit costs for each product were calculated considering the total batch size and including a percentage loss for each product, taken as a historical average over multiple batches of a specific product.
Univariate sensitivity analyses were carried out on one antiretroviral, for two batch sizes (250 000 and 500 000 tablets), to illustrate the effects of changes in the costs of direct labor and API, respectively, on the direct manufacturing costs for that product while controlling for scale of production. Expenses on API and direct labor were selected because it is these production costs that are most likely to vary between drug manufacturers located in different parts of the world. Salaries – and therefore labor costs – vary greatly between developed, developing, and the least developed countries, while API costs are sensitive to overall market demand and supply and will vary between suppliers . Capital costs for the equipment (which vary little between companies as they use the same type of equipment) and the cost of excipients and packaging materials (which show little variation because most are frequently used commodities in pharmaceutical production) were not considered in the sensitivity analysis.
Calculation of factory price
We then calculated the factory price – the price of the product at the factory or warehouse prior to any distribution or importation/exportation, and thus not subject to their associated costs – at which Far Manguinhos would be able to offer these products in the Brazilian market today, under the assumption that its indirect cost structure, for which the data were available, has not changed since 2001. This factory price includes indirect costs – expenses that cannot be directly traced to steps in the production of a specific product (such as general utilities); non-operational expenditures, which are comprised of institutional costs such as administration, human resources, warehousing, and marketing; and an operating margin which is added to ensure that the company is financially sustainable and for use in investment in upgrading manufacturing capacity (Table 2).
Indirect costs were calculated by adding to the direct cost a fixed percentage of the total indirect costs incurred over the entire product portfolio of the firm. In the case of Far/FIOCRUZ, the average portion of total indirect costs that was allocated to the production of antiretroviral drugs was 11%, and included the following categories of expenses.
1. Manufacturing overhead. This includes costs that cannot be traced directly to a product or productive step, such as general utilities, sanitation, and facility cleaning expenses.
2. Non-operational expenses. Institutional spending on administration, finance, human resources and marketing is included, as well as costs associated with research and development (R&D), such as the development of new formulations for pediatric use. Far/FIOCRUZ has over 100 researchers working in different areas, including formulation, synthesis and pharmacology.
Lastly, an operating margin equal to a fixed percentage of the final price for each product was added. This percentage is again determined on a product-by-product basis in order to ensure that the company has sufficient funds remaining after all expenses are paid to sustain operations and re-invest in technology, infrastructure and equipment to maintain or improve production capacity (Table 2). At Far/FIOCRUZ, the average operating margin across all antiretroviral drugs was 14% of the final price.
The price of each product is then equal to the following:
Equation (Uncited)Image Tools
A range of prices was estimated for each antiretroviral drug, based on the lowest and highest costs for the respective APIs taken from the 2006 WHO survey.
Comparison with prices from other producers
For each product we compared the calculated price range to the range of quoted prices from private, for-profit manufacturers from the recent survey of the antiretroviral drug market in developing countries by Medecins Sans Frontieres (MSF) , which includes data from private, for-profit suppliers and one state-owned enterprise (Government Pharmaceutical Organization, Thailand). In this survey, prices are quoted under a variety of commercial conditions that include costs such as freight, insurance, clearance charges and local taxes, also known as Incoterms . Previous analysis of the effects of trade barriers and tariffs on the price of pharmaceuticals documented that the use of different Incoterms would increase factory – also known as ex-works – price up to 15% [21–23], and thus supports the validity of comparison of these prices to the calculated prices.
Calculation of direct manufacturing costs
Direct manufacturing costs for a year's supply of a defined daily adult dose of each of the 19 antiretroviral drugs in the study are shown in Table 3. In general, these costs are lower for first-line drugs than for second-line drugs. For first-line drugs they range from US$ 18 per annum for stavudine (d4T) 30 mg to as high as US$ 312 per annum for efavirenz (EFV), and for second-line drugs from US$ 73 per annum for didanosine (ddI) 400 mg to US$ 1825 for nelfinavir (NFV). Expenditure on API accounted for between 54.5% (for d4T 30 mg) and nearly 99% [for EFV 600 mg and abacavir (ABC) 300 mg] of total direct manufacturing costs. Direct manufacturing costs for the same product vary depending on whether the lowest or highest API price was used in the calculation, ranging from a 33% difference for d4T 30 mg to 292% for lopinavir/ritonavir (LPV/r).
The results of the sensitivity analysis (Fig. 1) show that changes in API price have a much larger impact on direct manufacturing costs than changes in the cost of labor, even when controlling for increased scale of production. For example, a 25% increase in API costs results in a 23% increase in direct manufacturing costs whereas a tripling of labor costs (200% increase) results in only a 6% increase in direct manufacturing costs. A doubling of batch size from 250 000 to 500 000 has minimal impact on direct manufacturing costs, resulting in a 2% decrease at baseline labor costs, and a 4% decrease when three times the labor costs are applied. Increase in scale has no effect on direct costs when API price is the independent variable.
A comparison of the calculated prices with the highest and lowest quoted prices is shown in Fig. 2. For most antiretroviral drugs the estimated prices fall within the lowest quartile of the range of quoted prices, and in some cases are much lower than the lowest quoted price. For example, the estimated prices per patient year for zidovudine (ZDV), the oldest antiretroviral drug on the market, range from US$ 115 to US$ 169 for a year's treatment, while quoted prices from the MSF survey range between US$ 131 and US$ 307. For efavirenz (EFV), the estimated prices range from US$ 238 to US$ 411, while the quoted prices were between US$ 347 and US$ 766.
However for some drugs, particularly those for second-line treatment such as ABC, LPV/r, NFV, saquinavir (SQV), and ritonavir (r), there is greater variability in API prices taken from the WHO survey, and thus the calculated factory prices for the finished product approximate the quoted prices from the MSF survey, and in some cases even exceed the highest quoted price, when the highest cost API is used in the calculation. For NFV, for instance, the calculated prices range from US$ 913 to US$ 2355, while quoted prices vary between US$ 978 and US$ 2211, and for SQV the calculated prices range from US$ 1082 to US$ 2284, while quoted prices range between US$ 989 and US$ 1327. Similarly for ritonavir, estimated prices range from US$ 120 to US$ 440, while quoted prices range between US$ 83 and US$ 438.
This study is the first of its kind to provide information on the cost structure for individual antiretroviral drugs. The precise calculation of direct manufacturing costs of antiretroviral drugs can help purchasers to evaluate the relationship between the antiretroviral drug prices they pay and real production costs. Any difference between the direct cost of the drug and the price of the drug stem from the indirect cost structure and profit margin applied by the drug manufacturer. Indirect costs and profitability vary greatly between manufacturers, but this paper shows what antiretroviral drugs could cost, based on a structure that is both financially sustainable and that produces high-quality antiretroviral drugs to meet a national public health need.
In the case of Far/FIOCRUZ, as of 2001, direct manufacturing costs (including transportation within Brazil), accounted for approximately 75% of the price of the antiretroviral drugs, with API the largest component (approx. 62–70% of final price) (Table 1). Thus 25% remained to account for indirect costs and an operating margin. A comparison of this cost structure with those of private companies necessitates certain assumptions about the generalizability of the former, as the cost structure described above is based on categories used by Far/FIOCRUZ for internal reporting purposes (as it is a state-owned enterprise) while the typical format of published financial statements used by private companies is used for reports to external constituencies.
Given this limitation, it still appears that indirect costs and operating margins (exclusive of taxes, as antiretroviral drug production is not taxed in Brazil) make up a far higher percentage of the prices of drugs produced by private, for-profit companies than is the case with public producers. This percentage ranges from 31 to 53% for Indian generic manufacturers [24–27] to as high as 85% for large, multinational pharmaceutical companies (MNCs) [28,29]. Some of this difference, particularly for the MNCs, is reported as expenditures on R&D (15 to 19%), but details on much of the rest of these costs are not included in published financial reports and have been the subject of much debate [30,31].
The calculation of prices shown here will allow buyers to examine how the prices they currently pay compare with the prices that they could pay from a state-owned company procuring API from the open market, or from a firm that actually sells antiretroviral drugs at marginal cost (the cost of producing one addition unit) to the least developed and developing countries. It also allows comparison with the prices that are currently quoted by other drug manufacturers. However, the fact that some of the estimated prices are greater than the prices quoted by private, for-profit companies suggests that: (1) there are other factors that influence the actual costs for API incurred by antiretroviral drug producers; and (2) there are limitations that must be considered in our selection of manufacturer quotes as comparator benchmarks to the calculated antiretroviral drug prices.
The API prices used in this study are quoted prices given, upon request, to WHO by several manufacturers who were willing to respond to the survey, and therefore may not be representative of the entire API market for antiretrovirals. Actual API costs for drug manufacturers could be lower, either because they are negotiated in a commercial context or because the manufacturer who supplies the finished drug is also an API producer. This would enable a private, for-profit manufacturer to produce antiretroviral drugs at prices that would be competitive with those of a state-owned manufacturer who has bought API on the open market at the prices found in the WHO survey. Furthermore, the APIs for certain antiretroviral drugs, chiefly second-line drugs LPV/r, ABC, NFV, and SQV, are under patent in many countries, thus restricting the market to one or very few suppliers and leading to high prices for these active ingredients. This would subsequently result in high prices for the finished version of these drugs when sourced from generic manufacturers who do not produce the API. Generic drug manufacturers who wish to produce their own API must also undertake the time-consuming, costly and legally risky process of creating a novel route of API synthesis in order to avoid patent infringement, in cases where the patent holder has chosen to patent the production process as well as the product itself. Particularly because the current market for second-line drugs is small relative to that of first line, a result of the relatively early stage of most HIV/AIDS treatment programs in developing countries, there are few economic incentives for API suppliers to take these risks. A matter of additional concern is that new amendments to patent laws and intellectual property restrictions created under bilateral and regional trade agreements may further limit the sources of API [32,33]. Further research is needed to better understand the structure of API prices, including the impact of factors such as patent status, the availability of intermediates, and market size on the API price on the open market.
There are also limitations introduced by our selection of quoted manufacturer prices as comparators to the calculated prices. While quoted prices provide a rough indication of the antiretroviral drug market, they do not necessarily reflect actual practice in transactions between drug companies and purchasers of antiretroviral drugs in the developing world. Yet publicly available purchase data on antiretroviral drugs is relatively sparse and therefore a more stable, validated benchmark was selected in the form of price quotes. Greater transparency on antiretroviral drug transaction prices is needed in order to confirm whether purchasers of antiretroviral drugs actually have access to drugs at quoted prices, as the very limited data available [34–36] document that actual transaction prices are at times much higher than these quotes.
This study demonstrates the importance of the cost of the API in determining the prices of antiretroviral drugs, and shows the relatively minimal impact of other inputs such as labor, scale, and transportation. This finding is consistent with previous analyses of drug production . Significant decreases in antiretroviral drug prices will thus depend largely on a concomitant decrease in the price of the active ingredients.
The direct cost calculations presented in this study indicate that currently there is room for price reduction to reach levels of production cost or marginal cost for antiretroviral drugs purchased by the least-developed and developing countries, despite the absence of major reductions in API price. Even when compared with a production model that includes a sustainable margin of profit and appropriate considerations for quality assurance, current antiretroviral drug prices can be lower. The resolution  by the Group of Eight (G8) countries in support of universal access to antiretroviral therapy by 2010 reinforces the urgent need to promote sustainable access to high quality, low cost antiretroviral drugs. The affordability of antiretroviral drugs for HIV/AIDS treatment programs will be one element of critical importance in determining the ability of the developing world to ensure universal access to and coverage with antiretroviral therapy.
The authors would like to acknowledge Marco Vitoria, Jhoney Barcorolo, and Ivaldo Monteiro for their review and comments on this study.
1. World Health Organization. Progress on global access to HIV antiretroviral therapy: an update on ‘3 by 5’. Geneva: World Health Organization; June 2005.
7. Personal communication with Dr Eloan dos Santos Pinheiro, August 2004. Data adapted from unpublished data, presented publicly in London, UK and Montevideo, Uruguay. Presentations available at URL: http://www.who.int/3by5/amds
8. World Health Organization. Sources and prices of active pharmaceutical ingredients. Geneva: Department of HIV/AIDS. World Health Organization. November 2004. Available at URL: http://www.who.int/3by5/amds/api.pdf
9. Luong ATT. Appareillages et procedes de fabrication. In: Galenica Vol.5 - Les Systèmes dispersés: I'Agents de Surface et Emulsions, Technique et Documentation. Paris, France: Lavoisier;1983.
10. Jarowski CI. The pharmaceutical pilot plant. In: Lieberman, HA, Lachman, L, Schwartz, JB, editors. Pharmaceutical dosage forms: tablets. New York: Marcel Dekker, Inc.; 1990. Vol. 3, Chapter 5.
11. Wadke DA, Serajudin ATM, Jacobson H. Granulation technology and tablet characterization. In: Lieberman, A, Lachaman, L, Schwarz, JB, editors. Pharmaceutical dosage forms: tablets. New York: Marcel Dekker, Inc.; 1990. Vol. 2, Chapter 5.
12. Wadke DA, Serajudin ATM, Jacobson H. Preformulation testing. In: Lieberman, HA, Lachman, L, Schwartz, JB, editors. Pharmaceutical dosage forms: tablets. New York: Marcel Dekker, Inc.; 1989. Vol. 1, Chapter 1.
13. Simatos D. Lyophilisation. In: Galenica Vol. 3: Génie Pharmaceutique, Technique et Documentation. Paris, France: Lavoisier; 1982.
14. Peck GE, Baley GJ, McCurdy VE, Banker GS. Tablet formulation and design. In: Lieberman, H.A., Lachman, L., Schwartz, J.B., editors. Pharmaceutical dosage forms: tablets. New York: Marcel Dekker, Inc.; 1989. Vol. 1, Chapter 2.
15. Multiple Editors. Galenica Vol.6: Formes orales solides: I'poudres et formes unitaires obtenues par division ou agglutination, Technique et Documentation. Paris, France: Lavoisier; 1984.
16. Lachman L, Lieberman HA, Kanig JL. The theory and practice of industrial pharmacy. 3rd edition Philadelphia: Lea & Febiger; 1986.
17. Connoly RJ, Berstler FA, Coffin D. Beach-tablet production. In: Lieberman, HA, Lachman, L, Schwartz, JB, editors. Pharmaceutical dosage forms: tablets. New York: Marcel Dekker, Inc.; 1990. Vol. 3, Chapter 6.
18. Yalkowsky SH. Solubility and solubilization of nonelectrolytes. In: Yalkowsky SH, editor. Techniques of solubilization of drugs. New York: Marcel Dekker, Inc.; 1981.
19. Médecins Sans Frontières. Untangling the web of price reductions, 8th edition. Campaign for Access to Essential Medicines. Geneva, Switzerland: Médecins Sans Frontières; June 2005.
21. Anon. Global HIV/AIDS epidemic: selection of antiretroviral medications provided under US emergency plan is limited. Report to Congressional Requesters. Washington, DC: United States Government Accountability Office; January 2005.
22. Anon. International drug price indicator. Management sciences for health and essential drugs & medicines policy: 2003 Edition. Boston, USA and Geneva, Switzerland: World Health Organization; 2003.
23. Woodward, D. Trade barriers and prices of essential health sector inputs. CMH Working Paper Series No. 9. Geneva: Commission on Macroeconomics and Health (CMH). World Health Organization; June 2001.
30. Angell M. The truth about the drug companies: how they deceive us and what to do about it. New York: Random House; 2004.
31. Goozner M. The $800 million pill: the truth behind the cost of new drugs. Berkeley, California: University of California Press; 2004.
32. United States House of Representatives Committee on Government Reform – Minority Staff Special Investigations Division. Trade Agreements and Access to Medications under the Bush Administration. Washington, DC: US House of Representatives Committee on Government Reform; Available at http://democrats.reform.house.gov/Documents/20050609094902-11945.pdf
Accessed June 2005.
34. World Health Organization. Global price reporting mechanism. Department of HIV/AIDS. Geneva: World Health Organization. Available from URL: http://www.who.int/3by5/amds/price/hdd/
. Accessed 13 June 2005.
36. Vasan A, Hoos D, Mukherjee JS, et al. Pricing and the procurement of antiretroviral drugs: an observational study of data from the Global Fund. Bulletin WHO 2006; 84:393–398.
37. Guimier JM, Lee E, Grupper M. Processes and issues for improving access to medicines: the evidence base for domestic production and greater access to medicines. London: DFID Health Systems Resource Centre; June 2004.
© 2006 Lippincott Williams & Wilkins, Inc.