Bulletin of the World Health Organization

Global strategies to reduce the price of antiretroviral medicines: evidence from transactional databases

Brenda Waning a, Warren Kaplan b, Alexis C King c, Danielle A Lawrence b, Hubert G Leufkens d & Matthew P Fox b

a. Boston University School of Medicine, Boston, MA, 02118, United States of America (USA).
b. Boston University School of Public Health, Boston, MA, USA.
c. Massachusetts College of Pharmacy and Health Sciences, Boston, MA, USA.
d. Utrecht Institute for Pharmaceutical Sciences, Utrecht, the Netherlands.

Correspondence to Brenda Waning (e-mail: bwaning@bu.edu).

(Submitted: 10 September 2008 – Revised version received: 04 December 2008 – Accepted: 22 December 2008 – Published online: 03 June 2009.)

Bulletin of the World Health Organization 2009;87:520-528. doi: 10.2471/BLT.08.058925


New goals on providing universal access to HIV/AIDS services by 2010 were announced in 2007 by WHO, the Joint United Nations Programme on HIV/AIDS (UNAIDS) and the United Nations Children’s Fund (UNICEF).1 The need for life-long HIV/AIDS treatment and the high cost of antiretroviral (ARV) agents present challenges to achieving and sustaining universal access targets. During the past decade, various large-scale strategies have been used to reduce ARV prices in low- and middle-income countries. This paper focuses on three price-reduction strategies: procurement arrangements designed to increase purchase volumes, third-party price negotiation for generic ARVs and differential pricing for branded ARVs.

The first strategy, procurement arrangements to increase purchase volumes, often involves pooled procurement schemes that group multiple purchasers into a single purchasing unit in the hope that economies of scale will lead to lower prices. A pooled procurement mechanism is currently being developed at the Global Fund to Fight AIDS, Tuberculosis and Malaria (Global Fund).2,3

The second large-scale strategy involves third-party consultation and price negotiation with generic ARV suppliers, a practice introduced by the Clinton Foundation HIV/AIDS Initiative (CHAI) in 2003.4 In practice, CHAI attempts to make ARVs more affordable by negotiating price ceilings that reflect suppliers’ costs plus reasonable and sustainable profit margins.4 Moreover, CHAI furthers this strategy by providing direct technical assistance to some suppliers to help lower their production costs.4 The resulting ceiling prices are made available to all members of the CHAI procurement consortium.4 Countries that wish to become part of the consortium sign a memorandum of understanding with CHAI and manufacturers are required to offer ARVs to these countries at prices equal to or less than CHAI-negotiated ceiling prices.4

The third strategy involves differential pricing, sometimes referred to as price discrimination or tiered pricing. In 2000, the Accelerating Access Initiative, a collaborative endeavour of multiple international agencies and pharmaceutical manufacturers, first launched such a strategy for ARVs.5 Whereas CHAI price negotiation deals exclusively with generic ARVs, differential pricing pertains to branded ARVs and was introduced at a time when generic ARVs were not yet available. Under differential-pricing schemes, each manufacturer selects certain branded ARVs to be sold to low- and middle-income countries at prices lower than those charged in high-income countries.5 Each manufacturer determines which countries are eligible to purchase ARVs under their differential-pricing scheme, with eligibility typically being based on the country’s income level and prevalence of HIV infection.

Data on transactions involving the procurement of ARVs with donor funds are made public by the Global Fund and WHO.6,7 The Global Fund and WHO databases can be used to monitor and examine the global ARV marketplace. Although some analyses of these databases have been carried out,811 none has examined the global impact of the various ARV price-reduction strategies mentioned above. We used the Global Fund and WHO databases to test the following hypotheses on three different ARV price-reduction strategies: prices for high-volume ARV purchases are less than for low-volume purchases; prices for generic ARVs purchased within the CHAI consortium are less than for generic ARVs purchased outside the consortium; and prices for branded ARVs purchased under differential-pricing schemes are equal to or less than those for generic ARVs.


Data sources

We used data on ARV procurement transactions from the Global Fund Price Reporting Mechanism and the WHO Global Price Reporting Mechanism (GPRM) for the period between July 2002 and October 2007.6,7 The Global Fund posts details of ARV procurements reported by their international aid recipients on the web-based Price Reporting Mechanism.6 In addition, procurement data from the Global Fund as well as procurement data provided by WHO country offices, international organizations, procurement agencies and others are posted by WHO on the web-based GPRM, which serves as the global repository for data on ARV procurement.7,12 As shown in Fig. 1, data from these two sources were combined in a way that allowed us to remove any overlap in procurement data either within or between data sources. We also made sure that the data concerned valid transactions by removing incomplete records, erroneous reports (e.g. the wrong manufacturer) and suspect data entries with extremely low or high prices. Suspect data entries were identified using standard box-plot equation intervals.

Fig. 1. Flow chart illustrating the removal of duplicate, erroneous and suspect records from combined dataa on the procurement of ARVs in solid form between July 2002 and October 2007a
Fig. 1. Flow chart illustrating the removal of duplicate, erroneous and suspect records from combined data<sup>a</sup> on the procurement of ARVs in solid form between July 2002 and October 2007<sup>a</sup>
ARV, antiretroviral; Global Fund, Global Fund to Fight AIDS, Tuberculosis and Malaria; GPRM, Global Price Reporting Mechanism; US$, United States dollar.
a Data sources were the Global Fund Purchase price report and WHO’s GPRM.
b A price was regarded as a low price outlier if it was < q1="" –="" 3="" ×="" iqr="" and="" as="" a="" high="" price="" outlier="" if="" it="" was=""> Q3 + 3 × IQR, where Q1 was the 25th percentile of price, Q3 was the 75th percentile of price and IQR was the width of the interquartile range.
c The value of procurements before adjustment based on the United States annual consumer price index.

For the current analysis, we restricted our data set to ARV products supplied in a solid form, such as tablets, capsules and caplets. To focus on the more commonly used ARVs and to ensure reasonable sample sizes for regression models, we chose ARVs with procurement sample sizes of 100 or more (i.e. the ARV was purchased at least 100 times between July 2002 and October 2007). As a result, the analysis included 7253 procurement transactions for 24 ARV dosage forms. These 24 dosage forms provide the basis for the regimens commonly used for the prevention of mother-to-child transmission of HIV as well as for first- and second-line treatment of HIV/AIDS. They belong to three major classes of ARV: nucleoside reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitors and protease inhibitors. We adjusted all prices, which were reported by the Global Fund and WHO in United States dollars (US$), to the July 2006–June 2007 time period using the United States annual consumer price index.13

The public data sources provided basic transaction information; however, to examine determinants of price, we created additional independent variables, namely, differential price-eligibility, CHAI-eligibility, volume and quality. Whether or not a branded ARV purchase was eligible for differential pricing was determined using information obtained from the 2001 to 2008 editions of Untangling the web of price reductions, published by Médecins Sans Frontières.14 Whether or not a generic ARV purchase was classified as a CHAI or a non-CHAI purchase was determined from information provided by CHAI on when countries joined the consortium and from CHAI ARV price lists, which indicated the manufacturers and products subject to agreements over the previous 5 years (conversation and material provided by D Ellis, CHAI, December 2007). Relevant ARV purchases were considered eligible for CHAI or differential pricing 6 months after the announcement of new prices offered via CHAI or differential-pricing schemes. This was done to account for the likely scenario that a country may have been locked into a previously negotiated price for an annual procurement cycle and may, therefore, have been unable to access newly announced prices. Volume was categorized as low, medium or high on the basis of thirds of the specific volume distribution for each ARV dosage form. The quality variable indicated whether an ARV was approved or not by a universally accepted, stringent regulatory body. Approved ARVs were those classified as prequalified by WHO or those approved or tentatively approved by the United States Food and Drug Administration (FDA).15,16 A summary of the number and total value of purchases of individual ARVs and their corresponding volume categories is provided in Table 1.

Analytic approach

We used existing and newly created variables to examine determinants of the price of ARVs. We devised separate regression models for each of the 24 ARV dosage forms by using generalized estimating equation linear regression to take account of the correlated nature of the data. Price, our dependent variable, was non-normally distributed; therefore, we adopted the natural log of the price per tablet or capsule as our outcome measure. Data were clustered by country and year of purchase to take into account potential correlations in price within these variables. Candidate predictor variables were: the International Chamber of Commerce standard trade definition (Incoterm),17 the World Bank Country Income Classification,18 eligibility for CHAI or differential pricing14 (D Ellis, CHAI, personal communication, 2007), FDA-approved or WHO-prequalified ARV,15,16 purchase volume third, and analytic year of purchase. For each predictor variable we calculated the percentage change in price for each one-unit increase in the predictor variable by exponentiating the β coefficients from the regression equations and subtracting 1. Multivariate analysis was used to determine the effect of purchase volume, eligibility for CHAI and eligibility for differential pricing on ARV price. The results of the multivariate analysis are presented as percentage price differences between categories, with a negative percentage difference indicating that the ARV price in the comparator group is less than that in the reference category and a positive percentage difference indicating the opposite. Price differences with P-values ≤ 0.05 were considered statistically significant and are highlighted in boldfaced type in Table 2 and Table 3.

To provide a context for interpreting the findings of the regression analysis, Table 2 also lists raw, unadjusted ARV prices for July 2006–June 2007 together with the results of the regression analysis. These raw prices are described in terms of the median annual price per person of individual products (i.e. median price per tablet or capsule × daily dose × 365 days). For ARVs for adults, we used doses for individuals weighing ≥ 60 kg; for paediatric ARVs, we used doses for children weighing 10 kg, as recommended by WHO.1921


Effect of purchase volume

We detected no statistically significant association between purchase volume and price at the country level for 19 of the 24 (79%) dosage forms after adjusting for other variables in the regression model (Table 2). For two of the five dosage forms for which there was a significant association between volume and price, the prices for high-volume purchases were 7% and 21% less, respectively, than for low-volume purchases. For two other dosage forms, the prices for high-volume purchases were less than for both medium- and low-volume purchases, with differences being 4% and 5% less, respectively, for one ARV and 11% and 16% less, respectively, for the other. The other dosage form was 6% less expensive for high-volume purchases than for medium-volume purchases.

Generic prices for CHAI and non‑CHAI purchases

We identified 13 generic ARV products for which CHAI had negotiated price ceilings with manufacturers on behalf of CHAI country consortium members. We compared the actual prices paid for generic ARVs by CHAI and non-CHAI countries and found that the price of 9 of 13 (69%) dosage forms was significantly lower for CHAI purchases than non-CHAI purchases (Table 3, columns A, B and C). Overall, CHAI prices were less than non-CHAI prices by 6–36% (Table 3, column C).

Generic and differentially priced branded prices

Of the 24 (79%) solid ARV dosage forms analysed, 19 were available to some countries through differential-pricing schemes provided by the brand-name manufacturer (Table 3, column D). Of these 19 (95%) differentially priced branded products, 18 could be compared in price with non-CHAI generics (Table 3, column E). For 15 of these 18 (83%) dosage forms, purchases made under differential-pricing schemes were significantly more expensive than non-CHAI generic purchases, with price differences ranging from 23–498% (Table 3, column E). For two of the 18 (11%) ARV dosage forms, prices for differentially priced branded ARVs were 63% and 73% lower, respectively, than prices for non-CHAI generic ARVs. The price difference between the differentially priced branded product and the non-CHAI generic version was significant for all 18 ARV dosage forms, apart from ritonavir 100 mg.


We combined data from medicine procurement databases made public by the Global Fund and WHO with information from other sources to evaluate price-reduction strategies for ARVs for the first time and found some surprising results. The most counterintuitive finding is the absence of an association between purchase volume and price at the country level for 19 of the 24 ARV dosage forms (see Table 2). Although conventional business practice suggests that making a large-volume purchase at the country level will result in a discounted price, this appears not to be the case for these medicines.

The Global Fund has recommended the facilitation of voluntary pooled procurement as a means of increasing procurement efficiency.2,3 Pooled procurement of ARVs will result in much larger purchase volumes than those explored in our study, but it remains difficult to quantify exactly how much money could be saved by pooling purchase orders. Any estimate of potential savings resulting from pooled procurement must be balanced against the programmatic costs of establishing and managing the procurement systems required. While some surveys and desk reviews have described potential pooled procurement mechanisms in developing countries,22,23 insufficient empirical research has been carried out to validate pooled procurement and identify the conditions under which it can operate most efficiently. Pooled procurement may certainly offer other potential supply chain efficiencies beyond increased purchase volumes, but it should be carefully monitored to ensure such efficiencies are achieved.

While interventions for improving procurement efficiency are certainly desirable, they should be designed to develop and increase the technical capability for managing these procurement systems in the countries concerned. New procurement arrangements, whereby donors and international organizations act on behalf of countries for selected diseases, may fail to help strengthen those countries’ health systems. Lastly, pooled purchase arrangements will reduce the number of purchasers and could, therefore, result in a dramatic restructuring of the current global ARV market. Econometric modelling should be used to predict the potential impact of pooled procurement on the global ARV market and the findings should be used to inform the design of these schemes.

Third-party price negotiation by CHAI shows promise. Overall, the price of generic ARVs was less for CHAI purchases than for non-CHAI purchases. However, price differences between CHAI and non-CHAI purchases varied widely across ARVs. While price differences were as high as 27–36% for some ARVs, for others they were less than 10%. The most dramatic price differences between CHAI and non-CHAI generic ARVs were typically observed in the 1 to 2 years immediately following negotiations with suppliers. We recommend, therefore, that additional time-series research be carried out to explore the reasons why these price differences diminish over time and their potential impact on the overall ARV market.

Traditional approaches using differential-pricing schemes have not decreased the prices of branded ARVs to levels that can make these drugs compete with generic ARVs in most scenarios, which suggests that differential pricing alone is insufficient for achieving and sustaining universal access to HIV/AIDS treatment. In this study, nearly all the branded ARVs offered under differential pricing schemes were more expensive than the equivalent generics. There were a few exceptions where generic competition was lacking and differential pricing schemes for branded ARVs offered substantial cost savings over generic ARVs. The most notable exception was for lopinavir, 133.3 mg, plus ritonavir, 33.3 mg, a branded combination purchased under a differential pricing scheme; it was 73% less expensive than its non-CHAI generic equivalent. Likewise, there may be country scenarios in which generics cannot be purchased because of patent protection or other intellectual property barriers, and differential pricing may, therefore, offer substantial cost savings. Clearly, differential pricing of ARVs coexists in an environment that now includes large-scale financing of HIV/AIDS treatment and the maturation of generic ARV markets, two forces that did not exist when the Accelerating Access Initiative began. Additional work is needed to better explain the particular role of differential pricing in providing treatment for HIV/AIDS in today’s global ARV market, with special attention paid to the impact of differential pricing on the price of first-entry generic competitors.

While this research has provided some important findings, our analysis has limitations. Because we were dealing with pre-existing data, we used a standard method to remove outliers that may have come from erroneous reports. However, we repeated our analysis with the outliers included and the results did not differ from those presented in this paper. We also repeated our regression analyses using volume fourths instead of volume thirds and found the same results. Our study examined price–volume relationships at the level of individual purchases and did not consider total volume–price relationships at the level of tender arrangements, as these data were unavailable. Indeed, countries usually tender once or twice a year for larger volumes that are delivered in the multiple smaller purchase volumes reported in the databases. Still, larger tenders are likely to involve larger individual purchase volumes, so an analysis at the tender level would probably reveal similar results. It is notable that an association between volume and price was found for 5 of the 24 ARV dosage forms. This suggests that more work is needed to better understand the exceptional conditions under which the volume purchased at the country level may determine the ARV price.

Our regression models contained many candidate variables thought to be associated with price; however, we lacked access to information on additional factors that may have an influence, such as the timeliness of payment, the lead time between when an ARV is ordered and when it is needed, the presence or absence of drug registration, and intellectual property regulation. While many pricing studies must consider purchase incentives such as bundling, rebates and discounts, we doubt that such purchase arrangements played a major role in our study of donor-funded national ARV procurements. Lastly, our study focused on ARV prices only; other programmatic costs associated with the treatment of HIV/AIDS were not considered.

The quality of medicines on the market varies within and between low-resource countries, which means that previous price research compared medicines of unequal quality. For ARVs purchased with donor funds, however, the policies of the Global Fund and the United States President’s Emergency Plan for AIDS Relief mandate that the purchase of ARVs is approved by the WHO Prequalification Programme, the United States FDA or other stringent regulatory authorities. The WHO Prequalification Programme in particular has increased the availability of lower-priced, high-quality generic ARVs and this has enabled us to compare the prices of ARVs of equal quality.

Alternative strategies for reducing ARV prices should be explored. For instance, financial management systems in donor and country programmes could be improved and generic competition could be promoted by removing barriers to generic entry. Improved forecasting of future demand for ARVs may result in lower prices by preventing or minimizing the need for costly emergency shipments. For ARVs such as protease inhibitors that are expensive to make and are used less often, efforts could be made to consolidate global demand by reaching a consensus on the use of one or two key compounds. Alternatively, interventions aimed at transferring technology to generic producers may result in more timely generic competition for protease inhibitors and subsequent price reductions. Lastly, existing publicly available procurement databases should be expanded and used to guide future policies aimed at increasing access to essential ARV therapies. ■


Emma Back, William Bicknell, Michael Borowitz, Mariah Boyd-Boffa, Boniface Dongmo Nguimfack, Dai Ellis, Susan Fish, Manjusha Gokhale, Bruce Larson, Libby Levison, Maureen Mackintosh, Sapna Mahajan, Audrey McIntyre, Bas Peters, Rose Radin, Michael Reich, Sydney Rosen, Dennis Ross-Degnan, Lyne Soucy, Don Thea, Anita Wagner, Saul Walker and Michael Winter.

Funding: United Kingdom Department for International Development through the Medicines Transparency Alliance (MeTA) Project.

Competing interests: None declared.